inorganic ion exchangers and adsorbents for … · ion-exchangers currently in use. there are...
TRANSCRIPT
IAEA-TECDOC-337
INORGANIC ION EXCHANGERS AND ADSORBENTSFOR CHEMICAL PROCESSING
IN THE NUCLEAR FUEL CYCLEPROCEEDINGS OF A TECHNICAL COMMITTEE MEETINGON INORGANIC ION EXCHANGERS AND ADSORBENTS
FOR CHEMICAL PROCESSING IN THE NUCLEAR FUEL CYCLEORGANIZED BY THE
INTERNATIONAL ATOMIC ENERGY AGENCYAND HELD IN VIENNA, 12-15 JUNE 1984
A TECHNICAL DOCUMENT ISSUED BY THEINTERNATIONAL ATOMIC ENERGY AGENCY, VIENNA, 1985
INORGANIC ION EXCHANGERS AND ADSORBENTSFOR CHEMICAL PROCESSING IN THE NUCLEAR FUEL CYCLE
IAEA, VIENNA, 1985IAEA-TECDOC-337
Printed by the IAEA in AustriaJuly 1985
PLEASE BE AWARE THATALL OF THE MISSING PAGES IN THIS DOCUMENT
WERE ORIGINALLY BLANK
The IAEA does not maintain stocks of reports in this series. However,microfiche copies of these reports can be obtained from
IN IS ClearinghouseInternational Atomic Energy AgencyWagramerstrasse 5P.O. Box 100A-1400 Vienna, Austria
Orders should be accompanied by prepayment of Austrian Schillings 80.00in the form of a cheque or in the form of IAEA microfiche service couponswhich may be ordered separately from the IN IS Clearinghouse.
FOREWORD
Radioactive off-gases and low- to high-activity liquid and solidwastes arise from the operation of power reactors, spent fuelreprocessing plants and other nuclear fuel cycle facilities. The safetreatment and disposal of radioactive wastes is, therefore, an importantengineering task. So is the extraction and recovery of useful fissionproducts and actinides from various sources. Inorganic ion-exchangersand adsorbents have been receiving attention for these purposes becauseof their strong chemical affinity, high retention capacity for certainradionuclides and high resistance to irradiation.
In response to the growing interest in these topics, the IAEAconvened the Technical Committee Meeting on "Inorganic Ion-Exchangers andAdsorbents for Chemical Processing in the Nuclear Fuel Cycle" at itsHeadquarters from June 12 to 15, 1984 with the attendance of29 participants from 15 Member States. This Technical Document containsthe 20 papers presented during the Meeting. A variety of inorganicion-exchangers and adsorbents have been reported: zeolites, ammoniummolybdophosphate, magnetite, sodium titanate, zinc-charcoal mixtures,polyantimonic acid, titanium phosphate, zirconium phosphate, hydratedalumina, silica, bentonites, mordenite, hydrous titanium oxide andothers. The potential usefulness of these inorganic compounds has beenproved in various areas of nuclear fuel cycle technology, especially inthe separation and fixation of fission products and actinides and in thetreatment of effluents from nuclear power plants and of dissolveroff-gases. The use of the inorganic ion-exchangers and adsorbents canentail many advantages over conventional processes in the areas ofradioactive waste treatment and in the recovery of fission products andof actinides.
The Agency wishes to thank all the scientists, engineers andinstitutions who contributed to this Meeting with their papers and theirparticipation. Special thanks are due to the General Chairman,Mr. H.J.Ache (Federal Republic of Germany) and to the Session Chairmen,Messrs. L. Szirtes (Hungary), E.W. Hooper (United Kingdom), C. Madic(France), Weng Haomin (China) and M. Abe (Japan).
The officer of the IAEA responsible for this Meeting wasM. Ugajin of the Division of Nuclear Fuel Cycle. Assistance in preparingthe final document was provided by S. Ajuria.
EDITORIAL NOTE
In preparing this material for the press, staff of the International Atomic Energy Agencyhave mounted and paginated the original manuscripts as submitted by the authors and givensome attention to the presentation.
Tïie views expressed in the papers, the statements made and the general style adopted arethe responsibility of the named authors. The views do not necessarily reflect those of the govern-ments of the Member States or organizations under whose auspices the manuscripts were produced.
The use in this book of particular designations of countries or territories does not imply anyjudgement by the publisher, the IAEA, as to the legal status of such countries or territories, oftheir authorities and institutions or of the delimitation of their boundaries.
The mention of specific companies or of their products or brand names does not imply anyendorsement or recommendation on the part of the IAEA.
Authors are themselves responsible for obtaining the necessary permission to reproducecopyright material from other sources.
CONTENTS
Introduction .................................................................................................................... 7Sodium titanate - A highly selective inorganic ion exchanger for strontium .................... 9
/. Lehto.J.K. MiettinenThe effect of gamma radiation on various inorganic ion exchangers ................................ 19
L. SzirtesTechnology and role of Cs and Sr separation in disposal strategy of high level waste ........ 31
L.H. BaetsleEvaluation of zeolite mixtures for decontaminating high-activity-level water at the
Three Mile Island unit 2 nuclear power station ............................................................ 43E.D. Collins, D.O. Campbell, L.J. King, J.B. Knauer, R.M. Wallace
Application of silver-free zeolites to remove iodine from dissolver off-gases in spent fuelreprocessing plants ........................................................................................................ 53T. Sakurai, Y. Komaki, A. Takahashi, M. Izumo
Radioactive ruthenium removal from liquid wastes of "Mo production process usingzinc and charcoal mixture ............................................................................................ 63R. Motoki, M. Izumo, K. Onoma, S. Motoishi, A. Iguchi, T. Sato, T. Ito
Ion exchange characteristics of hydrous oxides and their use in the separation ofuranium and thorium .................................................................................................... 75N.Z. Misak, H.N. Salama, EM. Mikhail, IM. El-Naggar, N. Petra, H.F. Ghoneimy,S.S. Shafik
Inorganic sorbents in spent resin incineration — The ATOS process ................................ 85C. Airola, A. Hultgren
Precipitated magnetite in alpha effluent treatment ............................................................ 97K. Hording
The application of inorganic ion exchangers to the treatment of alpha-bearing wastestreams ........................................................................................................................ 113E. W. Hooper
Separation and purification of fission products from process streams of irradiatednuclear fuel .................................................................................................................... 133S.A. AU, H.J. Ache
Efficacy of argillaceous minerals used as a migration barrier in a geological wasterepository .................................................................................................................... 143E.R. Merz .
Ten years of experience in extraction Chromatographie processes for the recovery,separation and purification of actinide elements ............................................................ 157C Madic, J. Bourges, G. Koehly
Synthetic inorganic ion exchangers and their potential use in thé nuclear fuel cycle ........ 173A. Clearfield
Fundamental properties of acid salts of tetravalent metals and some considerations onthe perspectives of their application in nuclear technology ........................................ 195G. Alberti
Recovery of cesium-137 from radioactive waste solutions with a new complex inorganicion exchanger ................................................................................................................ 213S. Zhaoxiang, T. Zhigang, H. Zun, L Zhenghao, W. Haomin, L. Taihua, L. Boli
Treatment of liquid wastes containing actinides and fission products using sodiumtitanate as an ion exchanger ........................................................................................ 223Y. Ying, L. Meiqiong, F. Xiannua
Ion-exchange properties of zeolites and their application to processing of high-levelliquid waste ................................................................................................................ 237T. Kanno, H. Mimura
Studies on inorganic ion exchangers — Fundamental properties and application innuclear energy programmes ............................................................................................ 249R.M. lyer, A.R. Gupta, B. Venkataramani
Ion-exchange selectivities on antimonic acids and metal antimonates ................................ 263M. Abe
Panel discussion ................................................................................................................ 277List of participants ............................................................................................................ 279
INTRODUCTION
In spite of a slowing down of construction of new power plants incomparison to previous forecasts, nuclear energy continues to be a provensource of electricity and an essential component of the energy balance ofmany countries. By the end of 1983 there were 317 nuclear power reactorsin operation, which produced about 12% of the world's electricity. Thetotal capacity of these reactors was 191 gigawatts (GWe). The IAEA nowestimates that the total installed nuclear capacity in the world willamount to 275 GWe by 1985, to 380 GWe by 1990 and between 380 and 850 GWeby the turn of the century.
This development of nuclear power will depend on the safe andefficient operation of power reactors and supporting fuel cyclefacilities. Nuclear power production generates actinides, fissionproducts and activation products that have to be safely managed toprotect the public. Radioactive off-gases and low - to high-activityliquid and solid wastes inevitably arise from the operations of powerreactors and plants for uranium ore processing, fuel fabrication, spentfuel reprocessing and other nuclear fuel cycle activities.
Waste treatment is, therefore, one of the great engineering tasksthat the nuclear industry has to face. Other important tasks are toextract and recover useful fission products and fissile and fertilematerials from irradiated nuclear fuels and other sources.
In response to the growing interest in these topics and as a resultof the increasing use of nuclear power, the IAEA decided to hold thisTechnical Committee Meeting within the framework of the Agency'sactivities on nuclear materials and fuel cycle technology. Theapplication of inorganic ion-exchangers and adsorbents to both wastetreatment and the recovery of fission products and actinides were ofprimary concern in this meeting. Because of the importance of thescientific knowledge needed to develop successful technical processes,the meeting covered the two major fields of fundamental studies andindustrial applications.
It has been shown that inorganic ion-exchangers and adsorbents havehigh selectivity and high exchange capacity for certain elements whichare important in the nuclear industry. Moreover, they are virtuallynon-combustible and generally immune to radiation damage and biologicaldegradation, which are significant advantages over the organicion-exchangers currently in use. There are reasonable possibilities touse inorganic exchangers for processes involving extraction, purificationand retention of radioactive materials. Because of their versatility,inorganic ion-exchangers and adsorbents will play an important role innuclear fuel cycle technology in the near future.
The main objectives of the meeting were to provide an opportunityfor experts involved in chemical processing in the nuclear fuel cycle toexchange information on experimental results and experiences and todiscuss the specific problems facing them. This was the first meetingorganized by the Agency to deal with these topics.
SODIUM TITANATE - A HIGHLY SELECTIVEINORGANIC ION EXCHANGER FOR STRONTIUM
J. LEHTO, J.K. MIETTINEN,Department of Radiochemistry,University of Helsinki,Helsinki, Finland
Abstract
A strontium-selective inorganic ion exchanger, sodium titan-ate, has been synthetized from an industrial intermediate of atitanium dioxide pigment process. The distribution coefficientof sodium titanate for strontium is high, nearly 10 at a pHabove 7. The sorption capacity is also high: 1.4 mmol of stron-tium per gram of exchanger. The effect of high concentrationsof sodium chloride, boric acid, sodium citrate and EDTA on thesorption has been studied. The sorption rate of strontium insodium titanate has been found to be rather high. The ex-changer is highly resistant to gamma radiation. A gamma irra-diation dose of 10 Gy has practically no effect on its struc-ture and sorption properties. Decontamination factors higherthan 100 have been obtained for strontium from two actual nu-clear power plant waste solutions.1. Introduction
In the past twenty years a great number of synthetic inor-ganic ion exchangers have been developed for processing nucle-ar waste solutions. Inorganic ion exchangers have many impor-tant advantages in this field.' First, most of them withstandhigh temperatures and radiation doses. Thus, unlike organicion exchange resins, they can be loaded with radionuclides toa high degree and can be utilized in the treatment of hot solu-tions. Second, some of them, like titanates and zeolites, canbe ceramized to a solid final waste product. The best knownceramic waste product is the titanate mineral based SYNROC(1). Third, many inorganic ion exchangers are highly selectiveto certain key radionuclides, like some zeolites are to cesium.
Sodium titanate is known to be effective in the separationof strontium. However, it has often been synthetized from ex-pensive chemicals like titanium tetrachloride and titaniumalcoxides (2,3). In this paper is described a sodium titanateproduct which was produced from an intermediate of an indus-trial titanium dioxide pigment process. The raw material ischeap and readily available. The synthesis, as well as sorp-tion arid radiation resistance properties of sodium titanate,are described.2. Experimental
Sodium titanate was synthetized from "titanium hydrate"which is an intermediate from the titanium dioxide pigmentprocess, the so called sulphate process used at a plant ofKemira Co., Finland. The raw material was slurried in hot buta-nol and sodium hydroxide was added as a 30 w-% water solution.
The mixture was stirred at 93 UC for a few hours. After theproduct had settled, it was filtered, washed with water andfinally dried at 105 °C (4).
The sorption capacity ofpodium titanate for strontium andthe stoichiometry of ttje Sr - Na exchange were studied usingcolumn experiments. Sr solution (0.01-M, pH 5) was pumped at-a flow rate of 0.6 ml/min through 3,0 g sodium titanate col-umns. The grain size of the sodium titanate was 0.315-0.850mm, as in all other experiments. The Sr + and Na+ concentra-tions in the 18 ml fractions collected from the eluate weremeasured by means of an atomic absorption spectrometer (AAS).Using the eluate volume, where the Sr + concentration was 50 %(50 % breakthrough) of the concentration in the feed solution,the sorption capacity was calculated with the following formu-la:
V(50 %} x CCapacity (mmol/g) = —————————,whereV(50 %) = eluate volume at the 50 % breakthroughC = Sr concentration in the feed solution (0.01-M)om = exchanger mass (3.0 g)
The stoichiometry of the Sr +- Na+ exchange was calculatedby comparing+the total amount of Sr + absorbed to the totalamount of Na eluted from the column.
The distribution coefficient (1C) of strontium in sodiumtitanate, as a function of pH, was determined with a batchmethod as follows: 0.2 g samples of exchanger were shaken with20 ml of 0.0005-M Sr + solution containing Sr tracer. ThepH's of the solutions were adjusted to different values usingnitric acid. After two hours shaking, the mixtures wereggentri-fuged and the pH's of the solutions were measured. The Sractivities of the solutions were measured using a Nal crystaland a single channel analyser. The distribution coefficientswere calculated with the following formula:
A -A V«D (ml/g) = "V x m»whereftt-A = 8SSr activity in the solution before shakingA° = Sr activity in the solution after shakingV = solution volume (20 ml)m = exchanger mass (0.2 g)
The effect of boric acid on the sorption of strontium as afunction of pH was studied at two different boric acid concen-trations. The experiments were carried out as described abovefor K_ measurements, but the 0.0005-M Sr solutions were ei-ther 0.5-M or 1.0-M with regard to boric acid.
The effect of different sodium ion concentrations on thesorption of strontium was tested by stirring 1.0 g samplesof sodium titanate in 100 ml of 0.0005-M Sr solutions (pH 5)which where 0.0005, 0.005, 0.05, 0.5, 2.0 or 4.0 molaric withregard to sodium chloride (5). The distribution coefficientswere calculated as described earlier.
10
The effect of organic complex forming agents, EDTA andsodium citrate, on the sorption of strontium was studied asa function of pH at two different EDTA and sodium citrate con-centrations. The experiments were carried out by stirring 1.0g samples of exchanger in 100 ml of 0.0005-M Sr solutionsof different pH values. The Sr solutions contained thesecomplex forming agents in molarities of 0.05 or 0.005. TheSr + concentrations of the solutions before and after twohours stirring were measured using AAS. The sorption percent-ages were calculated with the formula: (A -A/A ) x 100 %,where A is the strontium concentration left in the solutionsafter stirring and A is the initial strontium concentrationin the solutions (4)?
The sorption rate of strontium in sodium titanate wasstudied by stirring 1.0 g samples of exchanger in 100 ml ofSr + solution. Aliquots of 1 ml volume were drawn from thesolution at different time intervals and the Sr activitiesof the aliquots were measured, as described earlier. The sorp-tion half times T ,_ were calculated as the times during whichthe concentration or strontium in the solution decreases by afactor?of two. The effects of the exchanger grain.size and ofthe Sr + concentration were tested (5).
The effects of gamma irradiation on the structure and thesorption properties of sodium titanate wgre studied. Sodiumtitanate samples were irradiated with a Co source. The totalabsorbed gamma dose was 10 Gy and the dose rate 35 kGy/h.After the irradiation, the structure was studied using X-raydiffraction, IR spectrometric and thermoanalytical methods.The specific surface area was determined. The effect on thesorption of strontium was tested by determining the distri-bution coefficient as a function of pH as described earlier(6).
The efficiency of sodium titanate for the separation ofstrontium from two actual power plant (Loviisa, Finland) wastesolutions was tested. The waste solutions contained highamounts of these inactive salts: boric acid, -sodium and potas-sium ions. The total salt contents were 215 g/1 and 2 g/1. ThepH values of the solutions were 12.5 and 10.0, respectively.The solution with pH 12.5, having the higher salt 'content, isa waste concentrate obtained by evaporation from the solutionwith pH 10.0. This solution originates from various sourceslike leakages in the primary circuits, and from decontamina-tions. Sr tracer was added to these waste solutions. Samplesof 450 ml volume from the solutions were eluted through sodiumtitanate columns of masses of 1.0 g and 2.0 g. The élutionrate was 1 ml/min. The Sr activities in the eluates weremeasured using a Ge(Li) crystal and a multichannel analyzer.The decontamination factors (D.F.) were calculated as theratios of the strontium concentration in the feed solution tothe concentration in the eluate.
1 1
3. Results and discussion
Fig. 1 shows a breakthrough curve of strontium from a sodi-um titanate column. The breakthrough curve depicts the concen-trations of Sr + and Na+ in the eluate as a function of theeluate volume.
Sr-2-*- AND No* IN THE EFFLUENT, l/1
20 .
IB .
12 .
B
4 ]
^o—_„o——-200 400 BOO 800
VOLUME OF THE EFFLUENT, ML
Fig. 1. A breakthrough curve of strontium from a 3.0 g sodiumtitanate column. Sr concentration in the feed solution:O.Q106-M, pH: 5. Elution rate: 0.6 ml/min. Solid line (+) =Sr , dashed line (o) = Na+.
The sorption capacity of sodium titanate for strontium, cal-culated using the 50 % breakthrough value of strontium, wasfairly high, 1.4 +_ 0.1 mmol/g.
2+ was ab-In Fig. 1 it can be clearly seen that+when Srsorbed in the sodium titanate column, Na was simultaneouslyeluted out. when Sr began to break through, the Na concen-tration in the eluate decreased respectively. The molar ratioof Na+ eluted from the column to Sr absorbed in the columnwas 2.0 +_ 0.1 averaged from two experiments. This ratio andthe shape's of the curves in Fig, 1 indicate a s^oiciometricchemical ion exchange of two Na ions by one Sr ion.
Fig. 2 shows the distribution coefficient Kp of strontiumin sodium titanate as a function of pH. At the acidic re-gion, 1C increases steeply and reaches a constant high valueof nearly 10 at a pH above 7.
12
5
4 .
3 .
logKD -*-
./
spH
l 2 3 4 5 6 7 8 9 1 0 1 1 1 2
Fig. 2. The distribution coefficient K~ of Sr in sodiumtitanate, as a function of pH. The initial Sr concentrationin the solution 0.0005-M.
Boric acid is used as a neutron absorber in nuclear powerreactors and thus it might be found in high concentrationsin low- and medium-level power plant waste solutions. The ef-fect of boric acid on the sorption of strontium in sodium ti-tanate is shown in Fig. 3. Up to pH 6 there is no effect.Above this pH value, the boric, acid concentration of 1.0-Mreduces the distribution coefficient by a factor of 10.'
1 2 3 4 5 6 7 8 9 1 0 1 1 1 2
.Fig. 3. The distribution coefficient 1C of Sr in sodiumtitanate, as a function of oH, at two different boric acid con-centrations. The initial Sr + concentration in the solution:0.0005-M. * = 0-M H BO o - 0.5-M H B0q (dashed line), + =1.0-M
13
Nuclear waste solutions often contain high amounts ofsodium. A high concentration of 4.0-M of sodium decreases thedistribution coefficient of strontium in sodium titanate onlyby a factor of 10, as can be seen in Fig. 4. Dosch obtaineda similar result with sodium titanate synthetized in theSandia National Laboratories (7).
ID5 KD
104i
1D3J
CNa ]
D. OD1-M 0. Dl-M 0. 1-M 1. D-M 10. D-M
Fig. 4. The distribution coefficient ^ ?titanate, as a function of Na+ concentration. The initial Srof Sr in sodiumion. The i
concentration in the solution: 0.0005-M, pH: 5 (5).In order to fix waste nuclides in inorganic ion exchangers,
the nuclides first have to be eluted from spent organic reac-tor resins. This elution can be carried out by using organiccomplex forming agents. Methods based on this principle havebeen developed in Sweden (8). Figs. 5 and 6 show the effectsof two complex forming agents on the sorption of Sr in sodi-um titanate. EDTA decreases sorption to a great extent, andthus it can be excluded as a possible eluting agent. Sodiumcitrate for its part affects sorption only slightly at pH<6.
100.
80.
60 lI
40.
20.
SORPTION, %-H————!————l————f
W ——
2 3 4 5 B 7 B g2+
10 11 12Fig. 5. The sorption percentage of Sr"' in sodium titanate,as a function of pH, at two different EDTA concentrations.The initial Sr + concentration in the solution: 0.0005-M.+ = 0-M EDTA, o = 0.005-M EDTA (dashed line), * = 0.05-M EDTA (4)
14
100
80
60
40
20
SORPTION, %jj «. O • O~*"'O O"^£**———*———•*»„ ,
1 2 3 4 5 6 7 8 9 10 11 12Fig. 6. The sorption percentage of Sr in sodium titanate,as a function of pH, at two different sodium citrate concen-trations. The initial Sr + concentration in the solution:0.0005-M. o = 0.005-M sodium citrate, * = 0.05-M sodium cit-rate (dashed 1ine)(4).
Sodium titanate is highly resistant to gamma radiation.Practically no change was found in its structure and in itsspecific surface area, when samples of the exchanger were ex-posed to a gamma irradiation dose of 10 Gy. No effect on thesorption of strontium was found either, as can be seen in Fig.7. Kenna found a 50 % decrease in the strontium capacity ofthe sodium titanate synthetized in the Sandia National Labora-tories after having irradiated the exchanger with a gamma irra-diation dose of 2 x 10 Gy (9).
1-9*0
4 .
3 i
2
H2 3 4 5 6 7 8 9 10 11 12
2+Fig. 7. The distribution coefficient K^ of Sr in sodiumtitanate, as a function of pH, when the exchanger was expo:to a gamma irradiation dose of 10 Gy (dashed line(o)) andwhen not exposed (solid line (*)). The initial Sr ------tration in the solution: 0.0005-M (6).
concen-
15
The sorption rate of Sr + in sodium titanate is satisfacto-rily high considering the use of the exchanger in column opera-tions. Sorption half time of 1.6 +_ 0.2 min was obtained forsorption from a 0.001-M Sr + solution in sodium titanate witha grain size of 0.315-0.850 mm. The half time decreases whensmaller grain sizes are used and increases when the Sr con-centration gets higher.
The experiments on the separation of strontium from the twonuclear power plant waste solutions, using sodium titanate,gave satisfactory results. The decontamination factors andthe volume reduction factors obtained are compiled in TableI.Table I. The decontamination factors (D.F," and the volumereduction factors (v.R.F.) obtained from the column experi-ments using sodium titanate for the separation of strontiumfrom two waste solutions taken from the Loviisa nuclear powerplant, Finland. 450 ml of the solutions were eluted throughthe columns. The flow rate was 1 ml/min.
Salt content ofthe waste solution
215 g/1215 g/12 g/12 g/1
Mass of theexchanger
1 P2 g1 g2 g
D. F.
40 +170 +130 +270 +
10402050
V. R. F
280140
280140
4. Conclusions
Sodium titanate is highly selective toward strontium. Thecapacity and the distribution coefficient are high, remainingsufficient even at high concentrations of boric acid and sodiumchloride. Thus this exchanger is suitable for separation ofstrontium from nuclear waste solutions containing high amountsof inactive salts.
The use of the industrial intermediate "titanium hydrate"as a raw material in sodium titanate synthesis reduces prepara-tion costs significantly compared with other synthesis meth-ods. "Titanium hydrate" is available in large amounts and ischeap. Furthermore the synthesis of sodium titanate from "ti-tanium hydrate" is rather straightforward.
16
Separation of waste nuclides using sodium titanate has anadditional advantage in that the loaded exchanger can beturned into a solid final waste form. One method is hot press-ing used, for example, in the SYNROC process. Another methoddeveloped in our laboratory includes mixing the titanate withclay and firing the mixture at 1000 C to a solid form (10).This method also gives a highly resistant waste form.
5. AcknowledgementThis study was supported by the Ministry of Commerce and
Industry, Finland.
6. References
(1). Ringwood, A.E., et al., Nucl. Chem. Waste Manag.2(1981)2.87.
(2). Forberg, S. and Olsson, P.-J., Method of Preparing Titan-ates Suitable as Ion-Exchange Material, U.S.Patent
. 4.161.513, 1979.(3). Lynch, R.W., et al., The Sandia Solidification Process-
A Broad Range Aqueous Waste Solidification Method, inManagement of Radioactive Wastes from the Nuclear FuelCycle, IAEA, Vienna, 1976, p. 361.
(4). Heinonen, O.J., Lehto, J. and Miettinen, J.K., Radiochim.Acta 28(1981)93.
(5). Lehto, J., Heinonen, O.J. and Miettinen, J.K., Radiochem.Radioanal. Letters 46(1981)381.
(6). Lehto, J. and Szirtes, L., Radiochem. Radioanal. Letters50(1982)375.
(7). Dosch, R.G., Final Report on the Application of Titan-ates, Niobates and Tantalates to Neutralized DefenseWaste Decontamination - Materials Properties, PhysicalForms and Regeneration Techniques, Sandia National Labora-tories, Report SAND-80-1212, 1980.
(8). Forberg, S., et al., Fixation of .Medium-Level Wastes inTitanates and Zeolites: Progress Towards a System forTransfer of Nuclear Reactor Activities from Spent Organicto Inorganic Ion Exchangers, in Scientific Basis for Nu-clear Waste Management, Vol. 2., ed. C.J.M. Northrup,Plenum Publishing Co., 1980, p. 867.
(9). Kenna, B.T., Radiation Stability of Sodium Titanate Ex-changer Materials, Sandia National Laboratories, ReportSAND-79-0199, 1979.
(10).Lehto, J., Heinonen, O.J. and Miettinen, J.K., Ceramiza-tion of Inorganic Ion Exchangers Loaded with NuclearWaste into Red Clay Tiles, in Scientific Basis for Nu-clear Waste Management VI, ed. D.G. Brookins, ElsevierScience Publishing Co., 1983, p. 589.
17
THE EFFECT OF GAMMA RADIATIONON VARIOUS INORGANIC ION EXCHANGERS
L. SZIRTESInstitute of Isotopes,Hungarian Academy of Sciences,Budapest, Hungary
Abstract3 7The effect of gamma radiation in the range of 5 x 10 to 10 Gy
has been studied for the following inorganic ion-exchangers: amorphouszirconium, titanium and chromium phosphates, zirconium tellurate andcrystalline zirconium and cerium phosphates. The solubility in acidicsolutions, the ion-exchange capacity and the specific surface area ofthese compounds were examined before and after irradiation. Zirconiumphosphate has a very stable crystalline structure which remains unchangedduring the irradiation process. The crystal lattice of cerium arsenatesis partly destroyed during irradiation but the relatively high stabilityof the bands of AsO groups prove that no chemical change accompaniesthis process. The cerium phosphate crystal is completely destroyed andbecomes amorphous after irradiation.
Using ion exchangers in radiochemical practice in many casesthey are exposed to high doses from charged particles and/orgamma radiation. The organic ion exchangers are submittedto radiolysis which results in changes of different properties,i.e. swelling, capacity etc. For example a Russian anionexchanger type AV-17 lossed 50^ of its original capacityafter irradiation with a dose of about 5x10 Gy (l). Otherobservations in this field were collected later on in thebook of Yegoroff (2). These facts made difficulties in usingion exchange procedures in radiochemical practice, at leastin cases when high activity level exists.It i s a pitiable fact that in contrast to the general viewthat all inorganic ion exchangers have a good resistanceagainst both charged particles and gamma-rays only someworks were found treating this subject (3-5).
19
Therefore, several years ago we made some investigations onamorphous chromium, titanium, zirconium phosphates andcirconium tellurate and on crystalline zirconium, ceriumphosphates and cerium arsenate (6-9).
EXPERIMENTAL
The amorphous zirconium, titanium, chromium phosphates andthe zirconium tellurate were synthetized by the methodsdescribed by us earlier (10-12). The crystalline zirconiumphosphate was prepared by the method of Clearfield (13),•hfeile cerium phosphate and arsenate were synthetizedaccording to Alberti (14-15). The samples were irradiatedwith gamma-rays using the K-80 type gamma irradiationfacility of the Institute of Isotopes (source: Co, doserate: 10 Gy/hr). The samples were irradiated for varioustime in such a way that they were exposed to 5x10 ,4 6 ?10 , 10 and 10 Gy dose, respectively. Doses were
controlled by chlorobenzene dosimetry (17). All experimentswere carried out after irradiation in parallel on thetreated and original samples.The amounts of the components were determined by analyticalchemical methods, i.e. zirconium, titanium and chromiumby usual gravimetric method as oxides; cerium was determinedspectrophotochemically (l?) while phosphate and telluratecolorimetrically (l8) and arsenic iodometrically in theusual way. The X-ray diffractograms were obtained usinga DRON-2 type high-angle powder diffractometer withCuK (Ni-filter) radiation ( X = 0,15418 nm).aThe i.r. investigations were carried out at room temperature(200-4000 cm" ) using a Perkin-Elmer 225 type spectrophoto-meter. The spectra of the solid samples were obtained frommults in Nujol, hexachlorbutadiene and Voltalef.The weight losses were determined by thermal analysisusing MOM type derivatograph in temperature interval25-1000°C by a heating rate of 5°C/min.The specific surface areas of the amorphous materialswere determined by the krypton adsorption method, using avolumetric gas adsorption apparatus at liquid nitrogen
20
temperature and were analysed by the BET method with acomputer program.The ion uptake was examined using tracer technique asfollows: several samples of the ion exchangers (0.1 g.each) were equilibrated with 10 ml of 0.2 M (MCI+ MOH)solution, where M denotes Na*, K+, Rb+ and Cs respectively.After shaking at rooa temperature till attainingequilibrium (usually two days); the amount of alkalinemetals in the samples of the supernatant liquid weredetermined by measuring the radioactivities of the initialand equilibrated solutions, while for assay of the phosphate,tellurate and arsenic content the methods mentioned beforewere used; the process was controlled and the pH measuredby Radiometer type automatic titrimeter. For labelling
fla, T£, Kb and Cs nuclides of high specificactivities were used.
Results and Discussion
The ratios of the ligands to the central metal ions of theexamined materials - both initial and irradiated - and theformulae calculated from the analytical data are collectedin Table 1.The solubility data in various acidic solvents are collectedin Table 2.The values of specific surface area of the amorphousmaterials are given in Table 3.
21
Table 1.
Ratios of ligands/metal for various matériels
Ion exchanger Formulae Ratiosoriginal irradiated
a. ) amorphousZirconium phosphate (ZPa) Zr(lIPO ; i)2 x HgO
Titanium phosphate (TP)- ..... , 1.95
Ti(lIP04)2 x 1.32 1I20 1.90
Chromium phosphate (CrP) Cr gO (lIPO ) g x 3HgO 1.10
Zirconium tellurate (ZTe) ZrO( ] ITeO^) 2 x 21i00 1.22
1.V5
1.85
1.04
1.20
b. ) crys ta l l ine
Zi rcon ium phosphate (Z'P) Zr(lIP04)2 x 1.33»20 2.00
Cerium phosphate (CeP) Ce(llPO ; i) x 0.3311 0 1.99
Cerium arsenate (CeAs) Ce(llAsO ; j)2 x l . U O H g O 2.01
1.98
1.97
"absorbed dose is 1 x 10 Gy.
Table 2.
Solubility data in various acidic solvents (mmole PO.orTeO, or As/g, ion exchanger)
Ion exchangematerials
ZP. originalairradiated
TP originalirradiated
CrP originalirradiated
ZTe originalirradiated
ZP originalirradiated
CeP originalirradiated
CeAs originalirradiated
II SO ÜC1 1IC10
0.05M
0.040.050.050.050.120.110.080.100.020.030.271.570.080.29
0. 5-^
0.050.050.060.050.100.100.070.080.050.050.283.240.610.92
0.1*.
0.020.020.040.060.120.120.08
0.090.020.030.201.600.120.34
1.0"
0.030.020.070.060.110.130.10
0.090.010.020.402.42
0.250.37
0.1*
0.020.030.050.070.12
0.130.09
0.070.020.020.18
1.650.10
0.35
l.O*»
0.020.030.070.060.130.110.070.070.030.030.251.780.22
0.39
Remarks
in allcases thesolubil i tyof tbeMaterialsis strouglyincreasedabovepll > 8
22
Table 3.
Specific surface area values
Ion Exchanger
Zirconium phosphateTitanium phosphateChromium phosphate'Zirconium tellurate
(CrP)(ZT0 )
S(a2/B)before after
irradiation
10.88.726.8
1G.5
10.0
1.32.121.0
The thermoanalytical data of amorphous materials are showuin Fig. 1. (dotted lines for irradiated samples).
meg/g '5
i • r10pH 2
No
10pH 2 10pH 2 10 pH
meg/g
1
Cs* . Rb*
10pH 2
CP
»pH 2
rir****W pH 2
ZT
»pH
meg/g
10pH 2 XJpH 2 10pH 2 10 pH
meg/g R b
\S5 -
Cs Rb Cs'
"iOpH 10pH 2 10pH 10pH
TPFi g . 1 . Ion uptake of alkali metals on amorphous
materials
23
The ion uptake data for alkaline metals are shown inFig. 2.From the data it can be seen that no characteristicchanges occur in solubility, the ion uptake and thermalbehaviour of the zirconium, titanium phosphates andzirconium tellurate.
1000
100 -
Fig. Thermograms of amorphous materials(dotted line for irradiated samples)
In case of chromium phosphate under the irradiation the•specific area value decreased about one order of magnitudeand the exothermic process existing at 250 C on thethermogram of the initial sample was not to be found onthat of the irradiated one.This phenomenon can be explained by supposing that thepolimer chainjVCr -HPO. -Ür/-H2PO^] was broken duringthe irradiation process. In case of crystalline materialsfurther X-ray and i.r. spectrometric investigations werecarried out. The results of these measurements are shownin Fign.3 and 4.
As it can be seen after irradiation zirconium phosphaterevealed exactly the same diffractogram as the originalsample; in case of cerium arsenate some extra very weak
24
V—s
Fig. 3* X-ray diffractograms of crystalline materials1-zirconium phosphate, 2-irradiated zirconiumphosphate, 3-cerium arsenate, 4-irradiatedcerium arsenate, 5-cerium phosphate,6-irradiated cerium phosphate.
peaks appeared and the diffractogram of cerium phosphatecontained no peaks et all.The i.r. spectra of zirconium phosphate (Fig. 4) shovedpractially no difference between the initial and irradiatedsamples. In case of cerium arsenate changes only in thecharacter of the bonds due to As-OH groups (2880, 2360and 1245 cm"1) and H20 (3590, 3485 and 1595 cm"1) were found.
25
l l i 1 1 i . l 1 . L . . . _ J
Fig, Infrated spectra of crystalline materials1-zirconium phosphate (both original andirradiated) 2-cerium phosphate, 3-irradiatedcerium phosphate, 4-cerium arsenate,5-irradiated cerium arsenate
For cerium phosphate the following changes wereobserved: absence of the band at 1420 cm" (P-OH in plane
-1deformation) and the presence of a new band at 980 cmwhich is due to a P-O-P structure.The weight loss data got by thermal analysis are shownin Fig. 5. As it can be ssen both zirconium phosphate andcerium arsenate have practically the same water contentbefore and after irradiation, while the cerium phosphateafter irradiation had about twice higher weight lossthan the initial sample.
26
meg/9 meg/g'
lOpH 2meg/q1
meg/g
No*
10 pHmeg/g
1
10 pH
Cs
10 pH 2Ce P
10pH
Fi g . 3 •
'OpH 2 10pH
Ce As
Ion uptake of alkali metals on crystallinematerials (dotted line shows the solubility)
Comparing the results of these three measurements thefollowing conclusions can be drawn:
1. zirconium phosphate has a very stable crystallinestructure which remains unchanged during the wholeirradiation process.
2. The results concerning cerium arsenate indicatedthat in this case the crystal lattice was partlydestroyed during irradiation but the relatively highstability of the bands of AsOjf~ groups prove that nochemical change accompanies this process.
3. The cerium phosphate crystal lattice is completelydestroyed after irradiation and according to theX-ray diffractogram the material became amorphous.
27
100 200 300 iOO 500 600 700 800 *C
Fig. 6. Weight losses of the crystalline materials•-•-•- zirconium phosphate, o-o-o- irradiatedphosphate, A-i-*- cerium phosphate,A-A-A- irradiated cerium phosphate,m-«-*- cerium arsenate, a-d-D- irradiatedcerium arsenate
Moreover the i.r. spectra indicate an intensive decreaseof HPO; and H2PO~ groups and presence of P-O-Pstructural elements.
The observations suggest that as a result of irradiationthe original structure is split and different phosphorus--oxygen compounds are formed, which are usually rathervolatile (l9)> This assumption seems to be strenghtenedby the ion uptake (Fig. 6) and weight loss data; i.e.the higher amount of soluble materials, the higherextent of vater-loss and the character of this process(the weight loss starts at a relatively low temperatureand it is practically finished at 350°C).Based on the above mentioned it can be stated that theexamined amorphous materials except chromium phosphatedo not change their basic behavious under theirradiation process.The investigated crystalline materials showed individualbehaviours under their irradiation with gamma rays.
28
Taking into consideration these facts, finishing mylecture I want to emphasize that we must not generalizethe good resistance of inorganic ion exchangers againstgamma radiation, but before using them we must investigatetheir behaviour under the irradiation. We must make thisfirst of all in case of newly synhetized materials.
REFERENCES
[I] KISELEVA, E.D., Z. Fizicheskoi Khimii (USSR) 3_§(1962) 2465
[2] YEGQROFF, E.V., NOVIKOV, P.D., Deistvie ionizinyuschihislutsenii ha ionoobmennie matériau Atomizdat Moscov1965 (in Russian)
[3] ANPHLETT C.B., Proc. 2nd Int. Conf. on Peaceful Usesof Atomic Energy Geneva 2J3 (1958) 17.
[4] YASUSHI, I., Bull. Chem. Soc. Japan 36 (1963) 1316[5] GROMOV, V.V., SPICIN, V.l., Atomn. Energ. (USSR) 14
(1963) 491.[6] ZSINKA, L., SZIRTES, L., STENGER, V., Radiochem.
Radioanal Letters 4_ (1970) 257[7] ZSINKA, L., SZIRTES, L., MINK, J., KÂLHAN J.,
Inorg. Nucl. Chem. 36 (1974) 1147[8] SZIRTES, L., ZSINKA, L., J. Radioanal. Chem. £1
(1974) 267[9] ZSINKA, L., SZIRTES, L., MINK, J., J. Radioanal. Chem.
30 (1976) 139[10] SZIRTES, L., ZSINKA, L., ZABORENKO, K.B., IOFA, B.Z.,
Acta Chim. Hung. 54 (1967) 215[II] ZSINKA, L., SZIRTES, L., Acta Chim. Hung. 6_9 (1971)
249[12] SZIRTES, L., ZSINKA, L., Radiochem. Radioanal. Lett.
7 (1971) 61.[13] CLEARFIELD, A., STYNES, J.A., J. Inorg. Nucl. Chem.
26 (1964) 257[14] ALBERTI, G., COSTANTINO, U., DIGREGORIO, F., GALLI, P.,
TORRACCA, E., J. Inorg. Nucl. Chem. 30 (1968) 295
29
[15] ALBERTI, G., COSTANTINO, U., DIGREGORlO, F.,
TORRACCA, E., J. Inorg. Nucl. Chem. 31^
(1969) 3195[ l6j HORVÀTH, Z., BÂNYAI, É., FÖLDIÄK, G., Radiochdn.
Acta 13 (1970-) 150[17] GREENHOUSE, U . C . , FEINBUSH, E.M., GORDON, L.,
Analyt. Chem. 29 (1957) 1531[IS] BERNHARDT, D.N-, V/REATII, A.R. , Analyt. Chem. 27
(1955) 440[19] VAN WASER, J.R., Phoshorus and Its Compounds Vol
I pp. 217 Interscience, New York (1958)
30
TECHNOLOGY AND ROLE OF Cs AND Sr SEPARATIONIN DISPOSAL STRATEGY OF HIGH LEVEL WASTE
L.H. BAETSLECentre d'étude de l'énergie nucléaire,Mol, Belgium
Abstract
The separation of Cs and Sr from the HLW by inorganic ion ex-changers has been extensively studied in the sixties as a treat-ment method to separate "long lived" fission products from shortlived. This approach needs very high decontamination factors and asubsequent partitioning of a-emitters. The separation of Cs and Srin acid medium was thoroughly investigated at S.C.K./C.E.N. Moland the results obtained are briefly discussed in the perspectiveof an over-all waste management strategy. Based on the existingtechnical information different flow sheets are discussed whichidentify the results to be aimed at for each of the separationsteps involved. Though Cs and Sr separation from HLW showed to beof little additional advantage unless it is quantitative i.e.involving very high decontamination factors, a semi-quantitativeseparation approach showed to be very attractive in easing theterminal disposal problem.Any separation of both radionuclides from HLW reduces in a propor-tional way the heat output of the residual waste fraction anddecreases the Cs contamination problems during vitrification. Oneof the limiting factors in disposal of vitrified HLW in conjunc-tion with MLW and a-waste, is the decay heat emitted by the canis-ters. Roughly 1 KW/TEQU after 10 years and 0.6 KW/TEQU after 50years, a 90 o/o removal of both heat producing radionuclides(137Cs and 90Sr) can shorten the standard cooling time in engin-eered structures from 50 to 20 year, or less depending on thethermal load of the formation, the separated Cs and Sr fractionshave to be a-free (i.e. below 100 n Ci/g) and may not use morethan 10 o/o of the initial equivalent glass volume. These re-quirements are fundamental and call for very specific ion ex-changers with high absorption capacity. These matrices constituteafter drying and conditioning a first choice storage form during300 years. Beyond that period the radiation hazard becomes négli-geable.
1. INTRODUCTIONSeparation of Cs and Sr from High Level Waste (HLW) solution is asubject which was intensively studied some 20 years ago. The useof inorganic ion exchangers as separation method seemed obviousbecause none of the classical partition processes e.g. liquidextraction or precipitation could achieve sufficiently high decon-tamination factors.The main aim of the studies was the development of inorganic ionexchangers which possessed sufficiently high selectivity coef-ficients for the two most important fission products 137Cs and90Sr and a reasonably high uptake capacity.
31
In the frame work of waste treatment it was at that time hopedthat by separating Cs and Sr from HLW with Decontamination Factors(DF) of 101* the effluent from this process could be treated asMedium Level Waste (MLW) for which adequate conditioning and dis-posal methods were available.The presence of a-emitters in HLW at concentrations which are notcompatible with MLW compositions was at the beginning not recog-nised as a very important fact, but gradually, with increasingfuel burn-up, it infused the scientific community that even themost perfect Cs-Sr partition method could not transform HLW intoMLW.
2. STATUS OF Cs and Sr SEPARATION FROM HLW IN BELGIUMAROUND THE MIDDLE SIXTIES2.1. Separation of Cs
Several inorganic ion exchangers displaying high selectivitiesfor Cs in concentrated HN03 solution were developed.
• Zirconium phosphate• Titanium phosphate• Ferro-cyanide molybdate
[3)
The first selective inorganic ion exchanger capable of separ-ating Cs from simulated HAW solutions (see fig. 1 for composi-tion and general PUREX flow sheets) was zirconium phosphate.The product was characterized as an amorphous or semi-crystal-line ion exchanger with mono functional phosphate groups. Theacidity of HAW or IWF being about l M HN03 and the chemicalcomposition of the solution rather diluted, zirconium phosphateis technically usable for Cs removal. However the introductionof a Cs partition module between the HAW and IWF or between theIWF and the high activity concentrator is a real technologicalproblem which cannot be realized unless the reprocessing pro-cess has planned the partition module during the design phaseof the plant. Experiments with real HAW solutions have shownthat Cs can be removed from the solution by using a zirconiumphosphate cartridge but that the adsorption is rather irrever-sible. Regeneration and repetitive use of the same zirconiumphosphate cartridge seemed to be problematic.About simultaneously a.second generation ion exchange product"titanium phosphate" was developed and showed to be superior tozirconium phosphate. The functional groups acidity of titaniumphosphate was higher and the reaction kinetics of the ion ex-changer was several times faster than that of zirconium phos-phates. Cs separation experiments were conducted in hot cellconditions with High Activity Waste Concentrate (HAWC) sol-utions produced by the Eurochemic facility in 1966-1967. Thesesolutions typical of PUREX HAWC stream contain 7 M HN03 and awide variety of chemical impurities as shown in fig. 2 (compos-ition 1). The composition 2 was artificially prepared with atenfold lower Al-concentration because this element was notconsidered typical of a PUREX HAWC product but occurred in theEurochemic waste solution. Volume reduction factors of about 50
32
:UN O.B sHN03 0.20Pu.s/1 0.04Flow 353Sp.Gr. C.9&V.Pu »7.9V.U¥ OF.jf Curies 240Pu.g/l
Flow 75T*mp 50«C(i»a>)SpGr. 161
tooV. U MO;DF. ta10*rCuriîs 3x10*
Ns[FP)«6NtiFPJrlD
KjOMNOj 2.4»KÛ.* o.oaFt*** O.OT2Sp.Gr. 109Flow 413
«OH ACTIVTTTCONŒXTRATORU7
I.OKSf.Qr. «3FUv OS
FIG. 1.
with DF's exceeding 100 have been experimentally obtained inreal hot conditions. The definition of the volume reductionfactor is the ratio between the volume of feed solution andthat of the ion exchange cartridge on which selectively all Csions present in the feed solution have been concentrated.Repeated use of one cartridge was shown to be feasible andtherefore the use of titanium phosphate was much more focussedon the separation of 137Cs from fission product solutions as ameans of cheaply producing 137Cs sources for irradiation pur-poses.
33
COMPOSITION OF HAW SOLUTIONS
Element
FeCrNiMnMoCuZrAlUPuRECsSrNaSISO;NH;potH*
9/1
. 1 . 2 50 .540 . 9 40 . 0 21.350 . 0 3 55 . 4 5
2 2 . 63 . 3 5_7 . 21.90 . 8 36 . 80 . 0 3 21.340.1U1.57. M
g/t
1.250 . 5 40 . 9 40 . 0 21.350 . 0 3 50 .05450 . 2 2 6.3.35_
7 . 21.90 . 8 30 . 8 80 . 0 3 21.340.1441.57. M
FIG. 2.
A third type of ion exchange material is an acid coprecitate ofmolybdenum and ferro-cyanides. This type of compound is muchmore crystalline than the M(IV) phosphates and displays thehighest selectivity towards Cs in acid solution. This ion ex-change material called for simplicity "Ferro-cyanide molybdate"is specifically suited for immobilization of 137Cs in view offinal storage. The product is chemically less stable than theM(IV) phosphates and must consequently be used in a continuousmode feed operation. The reaction kinetics is similar to thatof titanium phosphate but the sorption reaction is to be con-sidered as irreversible. Tests with real HAWC solution havebeen carried out till plugging of the cartridge because discon-tinuous mode feed operations had been imposed in hot cell.In conclusion of the Cs sorption paragraph it may be statedthat according to 1965 state-of-the-art two ion exchange ma-terials behaved quite well in real hot cell conditions : tita-nium phosphate and ferro cyanide molybdate ; the first beingpreferred because of chemical stability.
2.2. Separation of 90Sr from HLW [8)A much more difficult partition process to be developed was theseparation of Sr from HAWC solutions. None of those at thattime known sorbents permitted a relatively selective separationof Sr in such a complex mixture of ions and at such high acid-ity. The discovery in 1965 of polyantimonic acid as Sr sorbentin acid medium constituted a real breakthrough in this field.Laboratory tests with simulated HAWC solution showed the sorpt-ion capacity to be high but the reaction kinetics very slow.
34
By working at higher temperature (up to 80 °C) a reasonnablygood adsorption kinetics was obtained. The main ion exchangecompetition came from the rare earths and from sodium ions ifpresent in solution. By continuous mode feed operation at verylow flow rates practical ion exchange saturation can beachieved and volume reduction factors of 15 to 20 according tothe accepted breakthrough percentage have been obtained withHAWC solutions.As a conclusion of the previous paragraphs describing the stateof-the-art on Cs and Sr partitioning from HLW in the I9601s inBelgium it can be stated that selective separation on irrever-sibly loaded inorganic ion exchange materials was feasible withHAWC solutions representative of typical PUREX-HLW solutions.However a better result can be obtained by installing a Cs-Srpartition module between the HAW tank and the 1 WW evaporator.A schematic flow sheet with typical adsorption data for tita-nium phosphate (TiP) and polyantimonic acid (PA) is given inFig. 3. The block diagramme shows that the 137Cs volume reduc-tion factor is adequate but that the 90Sr sorption capacityought to be increased if the combined Cs + Sr cartridge is tobecome smaller than the initial waste vitrification volume (120l/TeqU),[9).
Cs - Sr SEPARATION FROM HAWHAW lEurochemic)
2-4 M HNOjC* S»0 mo/l-
SCi WC$Sr 200 mg/l-
5Ci HSrCorrosion productsAcfmidtsRare earths
LL
48 1 TIP
TIP
TIPt
ÏÏ
1
•180 I PA
PA
PA
-}
1J
••»
1£
3600 1MLLW
4000 I/ton Eq U
TIPmTitinium phosphatePA»Polyantimonic acid
Cartridge Cartridgedisconnection disconnectionDrying compaction Drying compaction-
High activityconcentrator
Dry Csstorage44 1
Dry Srstorage90 1
ENGINEERED STORAGE
Vitrification120 I HLWla" i___DISPOSAL
FIG. 3.
3. ROLE OF Cs-Sr SEPARATION ON LIQUID HLW HANDLING AND STORAGE• The heat output produced by HLW as a function of time is shownon Fig. 4. It clearly illustrates the fact that from 5 till
100 years storage the heat problem is dominated by the presenceof 137Cs and 90Sr. Indeed during this period the over-all heatoutput curve practically coincides with the Cs-Sr curve andstarts deviating consistently from 160 years on, at whichmoment the heat output produced by actinides (21tlAm and 243Amessentially) equals that of the fission products.
35
Watt 5 year10 year25 yearSO year
100 year
(470y)(7600y)
Fission products"Sr - "Y (28 j)137Cs - 137Ba (30
10'
10
25 100 500 103 104
Cooling time (years)
FIG. 4. THERMAL OUTPUT OF A STANDARD HLWGLASS BLOCK CORRESPONDING TO1 TON Eq. URANIUM
In the early storage period, between 5 and 10 years, afterdischarge from the reactor the HLW solution is stored in liquidform in specially equipped HLW storage tanks (see Fig. 5). Thesafety aspects related to this type of storage surpasses by allmeans those associated with the entire PUREX processing schemesA cost benefit analysis of the Cs-Sr separation might thereforehelp to endorse the thesis that even at short notice the sim-plified construction of HLW storage tanks might outweigh theconstruction of the Cs-Sr partition module.
• During the vitrification process liquid HLW solution is heatedup at temperatures of llbO-1200 °C causing a partial evapor-ation of 137Cs (15 to 30 %) which is partly recycled into theme!ter partly plated out on the cool surfaces of the venti-lation system. Suppression of 137Cs deposition on cold surfacesand reduction of liquids recycling is another advantage ofinitial Cs separation from HAW solutions as promoted in thispaper. However it is difficult to evaluate the over-all benefitcompared to the HLW vitrification as a stand alone processsince also MLW treatment costs have to be taken into account.
36
SKETCH OF HIGH-LEVEL LtOUID «ASTESTORAGE TAMK AT TOKAI, JAPAN
Steam (in)
Air purge dn>H A LW concentrate [mi
,- Steam im)S HAU* concern rate
Coding waler<ool) ;•';';--;
'••'.' '.."'v
HA L W concent raie*' ' - "" '
FIG. 5.
4. ROLE OF Cs-Sr PARTITIONING ON LONG TERM STORAGEOF VITRIFIED HLWSolidified HLW resulting from the vitrification process containsall the fission products and as a consequence (explained in theprevious paragraph) produces lots of heat (see Fig. 4) which haveto be dissipated into the geological formation.The actinides in the HLW produce a much lower heat which does notdecrease substantially beyond 500 years.Several scenarios (see Fig. 6) have been considered for geologicaldisposal of HLW and a-waste resulting from a 30 year nuclear elec-tricity production programme. For each scenario discharge fromthe reactor is supposed to be the origin of the time graph at t =0, reprocessing with simultaneous production of a-wastes and li-quid HLW occurs at t = 5 years and vitrification at t = 10 years.
SCENARIO 1Disposal of a-waste 10 years after tt = 50 years. = 0 and disposal of HLW at
37
enario1
10
A 50
enarioIt
301A 30
enario•
501A 50
enarioW
10U 30
.urne in years0 5 10 20 30 40 50 60 70 60
Nucliar programme
'Disposal of a. waste
Storage of HLW
Disposal of HLW
Nuclear programme1 — - • • • • — t — *— . . . ._ storage * waste
Disposal of «wasteStorage of HLW
Disposal of HLW
Nucltar preyMMe .1 • « . , , . - storage «waste
_. « u, , , "Disposal of «wasteStorage of HLW *
Disposal of HLWNuclear programme
, . *••"•• "•"•"•" Storage «waste
Disposal of < waste s|or jgeofHLW|r . .
Disposal of HLW
STORAGE SCENARIO' S OF REPROCESSINGWASTEFIG. 6.
SCENARIO 2Simultaneous disposal of HLW and a-wastes at t = 30 years afterdischarge.
SCENARIO 3Simultaneous disposal of HLW and a-waste at t = 50 years.
SCENARIO 4Disposal of a-waste at t = 10 years and disposal of HLW at t =30 years.
The limiting factor in all scenario's is without any doubt thepermissible heat load of the geological formation which limits thenumber of canisters per unit volume or surface to be disposed offat a given time.
The earliest disposal scenario for HLW is scheduled at t = 30years and stretches out till 50 years after the first fuel elementdischarge from the reactor. The a-wastes can be disposed off in a
38
suitable geologic formation as soon as the repository is avail-able. Economic considerations have shown that simultaneous dis-posal of alpha and HLW wastes is preferable to consecutive dis-posal of a-wastes and HLW wastes. The economic benefits of simul-taneous disposal can be deduced from the operation cost of a repo-sitory beyond the period necessary for a-wastes. In scenario 1 thetime span considered amounts to 40 years whereas in scenario 4 (aformation with high thermal load) the interval between the end ofthe a-waste disposal period and the end of HLW disposal amounts to20 years. Every year a repository has to be kept in operation re-quires the services of about 100 skilled technicians. By integrat-ing these services over 20 respectively 40 years a total expend-iture benefit of 2000 resp. 4000 man year is obtained. By dis-posing a-wastes together with Cs-Sr from HLW another over-all gainin disposal expenditures is expected from the size of the reposi-tory.Presently the repository site is mainly determined by the spacingof heat sources rather than by the total volume of a-waste thoughthese are 10 times more important in volume than HLW. By pursuingthe Cs-Sr from HLW option the small HLW volumes have merely to beadded to the a-waste volume which increases then with 10 % and thedesign differences between HLW and a-waste galleries disappear inthe safety margin of the project.In order to illustrate the influence of the heat load on the up-take capacity of a repository for HLW canisters a typical exampleconnected to the currently investigated Boom-clay layer situatedat 180-280 m below the nuclear site of Mol, Belgium is discussedin some detail, [10].The following constraints were imposed on the design criteria :the temperature at the contact between HLW canister and the clayformation may not exceed 120 °C to avoid drying and local col-lapse, the temperature increase at the interface between clay andsandy layers may not exceed 5 °C in order to avoid convectivecurrents in the sandy aquifer and finally the ground surface tem-perature increase may not exceed 0.5 °C..Taking into account these very conservative hypotheses and relyingon an experimental heat transfer coefficient in clay of 1.69 W/m°C it was calculated that these design conditions lead to a maxi-mum thermal load of 22 kW per hA (hA = 10.000 m2).After 30 years each HLW canister (corresponding to 1.5 TeqU) stillhas a heat output of l kW which corresponds to 22 canisters per hAor the equivalent yearly HLW production of 1 GWel cooled during 30years.By waiting 50 years the heat output per canister decreases to 0.6kW or each hectare has room for 36 canisters.The removal of the heat producing fission products has the sameeffect as a cooling time. By removing 5 years after discharge fromthe reactor 50 % of the Cs-Sr or 100 % of 137Cs or 90Sr corre-sponds to a cooling of about 20 years.
39
Cs - Sr SEPARATION FROM FISSION PRODUCTSREQUIRED DECONTAMINATION FACTORS
HLW 50 YEARS COOLING 390 W.
DF
10095908050
0
EQUIVALENT COOLING TIMEreduced to : years
7 .5
8 . 59 .9
21.50
DECONTAMINATION FACTOR OF «EMITTERSBASIS : 100 n Ci/g Ion exchanger -vVol. red. fact. : 50 / ——
«Activ i ty rCi 674 U3\ decay 250 years 500 years
FIG. 7.
Fig. 7 shows systematically the influence of the removal percen-tage of Cs-Sr on the equivalent cooling time. It clearly showsthat 80 % removal of both heat producing isotopes is sufficient toproduce waste canisters which can be disposed off into a geologi-cal formation immediately after their vitrification, i.e. at t =10 years.The existing separation technology developed earlier is capable ofproviding the necessary base for the removal of Cs-Sr from HAWstreams. Eighty percent removal corresponds in "ion exchange"terms to a DF = 5 which is very easy to obtain in industrial cir-cumstances.The question arises what to do with the separated Cs and Sr frac-tions ? Two aspects have to be addressed : the volume and thepurity of the separated fractions.As already shown in Fig. 3 the separation technique of Cs is suf-ficiently developed to put it into operation since the volume ofthe separated Cs cartridge is only 33 % of that of the vitrifiedwaste (40 1 compared'to 120 1 glass). A decontamination factor of20 (95 % removal) can easily be guaranteed in pilot scale facil-ities and this separation alone reduces the cooling time of theresidual vitrified waste to about 24 years.Further R & D has to be performed in order to bring the Sr separ-ation technique at the same chemical and technological level asthat of Cs.The second aspect which needs .some further comment is the purityof the separated Cs or Sr fraction. If the separation techniquehas to offer new perspectives in the field of geological disposalit must provide pure Cs and Sr fractions free of long-lived o-emitters.
40
As shown in Fig. 7 the decontamination factors of a-emitters mustbe very high in order to get a cartridge whose residual a~activityis equal or lower than 100 nCi/g ion exchanger. This a-activity isnow considered necessary to keep the separated fractions below thelimit of a-waste which must otherwise be disposed off in geol-ogical formation. The second most important R & D task is thedevelopment of ion exchange techniques capable to cope with thesesevere restrictions on the a-purity of the separated Cs-Sr isoto-pic fraction.If the Cs-Sr separation technique can be improved up to the pointwhere the total volume of Cs + Sr fraction does not exceed 20 to30 % of the regular glass volume and if these fractions are freeof a-emitters their storage in bunkers during 300 years is suffi-cient to reduce their health hazard to a low level waste typewhich may be dispersed in the biosphere.In these circumstances the nuclear electricity production hasdisposed off safely all the a-emitters in deep geological forma-tion and the short-lived fission products can be stored safelyduring a period which is short in comparison with the lifetime ofa civilisation.
REFERENCES
[1] BAETSLE, L.H., PELSMAEKERS, J. J. Inorg. Nucl. Chem. (21) pp.124-132 (1961-).
[2) BAETSLE, L.H., HUYS, D., J. Inorg. Nucl. Chem. (21) pp. 131-140 (1961).
[3] BAETSLE, L.H., J. Inorg. Nucl. Chem. (25) pp. 271-282(1963).[4] PIRET, J., HENRY, J., BALOU, G., BEAUDET, C., Bull. Soc.Chim. Fr. , p. 3590, (1965).[5) BAETSLE, L.H., MARKL MUSYL, L., HUYS, D., "Ion Exchange inthe Process Industries", Soc. Chem. Ind. pp. 209-215 (1970).[6) BAETSLE, L.H., VAN DEYCK, D., HUYS, D., J. Inorg. Nucl. Chem.
(27) pp. 683-695 (1965).[7) BAETSLE, L.H., HUYS, D., VAN DEYCK D., J. Inorg. Nucl. Chem.(28) pp. 2385-2394 (1966).[8) BAETSLE, L.H., HUYS, D., J. Inorg. Nucl. Chem. (30) pp. 636-649 (1968).[9) BAETSLE, L.H., HUYS, D., SPEECKAERT, Ph., Belgian Report from
S.C.K./C.E.N., BL6 487 (1973).[10] COMMISSION OF THE EUROPEAN COMMUNITIES, "Nuclear Science andTechnology : Admissible Thermal Loading in Geological Forma-
tions. Consequences on Radio-active Waste Disposal Methods",Vol. 1 Synthesis Report. Report EUR 8179 (1982).
41
EVALUATION OF ZEOLITE MIXTURESFOR DECONTAMINATING HIGH-ACTIVITY-LEVEL WATERAT THE THREE MILE ISLAND UNIT 2 NUCLEAR POWER STATION
E.D. COLLINS, D.O. CAMPBELL, L.J. KING, J.B. KNAUEROak Ridge National Laboratory,*Oak Ridge, Tennessee
R.M.WALLACESavannah River Laboratory,**Aiken, South Carolina
United States of America
Abstract
Mixtures of Linde™ lonsiv™ IE-96 and lonsiv™ A-51zeolites were evaluated for use in the SubmergedDemineralizer System (SDS) that was installed at theThree Mile Island Unit 2 Nuclear Power Station to decon-taminate ~2780 m3 of high-activity-level water (HALW).The original SDS flowsheet was conservatively designedfor removal of cesium and strontium and would haverequired the use of -60 SDS columns. Mixed zeolitetests, which were made on a 10~5 scale, indicated thatthe appropriate ratio of IE-96/A-51 was 3:2. A mathemat-ical model was used to predict the performance of themixed zeolite columns in the SDS configuration and withthe intended method of operation. Actual loadingresults were similar to those predicted for strontiumand better than those predicted for cesium. The numberof SDS columns needed to process the HALW was reduced to-10.
1. INTRODUCTION
Large volumes of high-activity-level water (HALW), containing>100 Ci/m3, were created by the accident at Three Mile Island,Unit 2 (TMI-2). This HALW consisted of two bodies — that in thereactor coolant system (-340 m3) and that which was spilled intothe reactor building floor (-2440 m3). Early selection of aflowsheet for decontamination of the HALW was necessary so thatprocess design and equipment fabrication could be initiated.After careful consideration, the TMI-2 Technical Advisory Groupselected a flowsheet that was based primarily on adsorption of thebulk of the radioactive contaminants, cesium and strontium, ontoan inorganic ion exchanger. This exchanger, Linde™ lonsiv™ IE-95,is a commercially available chabazite type of zeolite, having ahistory of successful, large-scale usage.
* Operated by the Martin Marietta Energy Systems, Inc., under contract DE-AC05-840R2 1400 with the U.S. Department of Energy.
** Operated by E.I. Du Pont de Nemours & Co. under contract DE-AC09-76SR00001with the U.S. Department of Energy.
43
The processing system was designed by Allied Chemical NuclearServices for the Chem Nuclear Company, the prime contractor forequipment fabrication and installation. The system was designedso that the equipment items containing high levels of activitywere housed in one of the spent fuel handling pools in order touse the pool water for shielding. Therefore, the process wascalled the "Submerged Demineralizer System," or SDS, although theprocess was not intended to demineralize the water during itsdecontamination.
The SDS process flowsheet was evaluated in a series of testsmade at Oak RiagÇ National Laboratory during early 1980, using 3 Lof TMI-2 HALW.IÏJ The tests showed that the bulk of the cesiumand strontium was effectively adsorbed on the zeolite, but thatsubsequent treatment with organic-based polishing resins did notprovide additional decontamination from cesium and strontium orremoval of the minor contaminants, 125Sb and 106Ru, which areanionic in nature.
The original SDS flowsheet called for the contaminated waterto be clarified and then passed through a series of ion exchangecolumns, as illustrated in Fig. 1. Four small columns, each con-taining -230 L of sorbent, were to be located within the spentfuel pool. The first three were to contain IE-96 zeolite, thefourth a strong-acid type cation exchange resin.
SPENT FUEL POOL
CLARIFIED
_
mZEOLITE COLUMNSILHUU_ d
CATIONCOLUMN
Ü
*L1- u.~Cm
POLISHINCOLUM!>t?*aj. J.H•lÉa1
FIG. I ORIGINAL SDS FLOWSHEET.
The IE-96 zeolite was found to be an excellent sorbent forcesium and adequate for strontium, if the contact time between thewater and zeolite was sufficiently long (>20 min). Therefore, themanner of operation of the proposed SDS was designed to accom-modate the needed contact time.
The columns were modular and were intended to be used as theradioactive waste containers after being loaded. The flowsheetcalled for the columns to be moved after each 45,000 L of waterhad been processed. This is equivalent to 200 bed volumes, basedon each column. At that point, the column in the first positionwas to be discharged, the other two moved forward one position,countercurrent to the flow of water, and a new column installed inthe third position. In this manner, the cesium would be loaded inthe column in the first position and all three columns would pro-vide a sufficient contact time for strontium sorption.
44
2. EXPERIMENTAL PROCEDURES
2.1. Consideration for increased loading ofcesium and strontium on zeolite
The original flowsheet was conservatively designed withrespect to cesium and strontium removal and would have required atleast 60 columns to process all of the HALW. Thus, a U.S. Depart-ment of Energy Task Force l2J evaluated the technical and financialbenefits of loading the columns with cesium to a higher level thanthat specified by the original*flowsheet. This higher level waslimited by the loading characteristics of strontium.
During evaluation of the original SDS flowsheet, a 1000-bedvolume test had been made. The results indicated that as many as~600 bed volumes could be processed before strontium breakthroughfrom the third column would occur. Since the organic cationresin, which was to be used originally in the fourth column, wasfound to be ineffective for providing additional decontamination,consideration was given to using zeolite in the fourth column inorder to increase the capacity of the system. Then, the through-put could be increased to 600 bed volumes, while maintaining asufficient safety margin. This would reduce the number of SDScolumns needed to ~25.
2.2. Consideration and evaluation of the use of mixed zeolitesA further increase in loading was envisioned after a more
"strontium-specific" zeolite, Linde" lonsiv"1 A-51, was identi-fied. l-*J The use of IE-96 zeolite for cesium sorption and A-51zeolite for strontium sorption in either mixed beds, layered beds,or alternate columns was considered. The use of alternate columnswould involve maintaining two types of columns.that might requiredifferent treatment during subsequent waste solidification opera-tions. If layered columns were used, and if flow distribution atthe bottom of the column was poor, the bottom layer could beadversely affected. Therefore, the use of mixed columns appearedto be most appropriate.
A 1500-bed volume test was made with TMI-2 HALW and a mixedzeolite bed containing equal parts of IE-96 and A-51. The break-through curves obtained for cesium and strontium, using this mix-ture, were compared with those obtained when using only IE-96zeolite (Fig. 2). The cesium breakthrough, when using only IE-96,was <0.01% during the entire 1000-bed volume throughput of thattest.
These results show that, when using the mixed zeolite, thecapacity for strontium sorption is increased by a factor of ~10,even though the kinetics of the strontium sorption, as indicatedby the lesser slope of the breakthrough curve, are somewhatslower. The capacity for cesium is adequate for a throughput ofup to ~2000 bed volumes. That loading would represent a factorof 10 increase over the original flowsheet design.
45
200 400 600 1000 2000 5000COLUMN THROUGHPUT (BED VOLUMES)
FIG. 2. COMPARISON OF PERFORMANCE OFSINGLE ZEOLITE AND MIXED ZEOLITE.
10.000
5000
4000
3000
z
? 3000CTt-
1000
STRONTIUM
CESIUM
O Sr, T R A C E R - L E V E L TEST• Sr, HALW TEST+ Sr, SOSo Cl , T R A C E R - L E V E L TEST• Cs, HALW TEST+ C», SOS
I I __ I I20 40 60 80
% IE-96 IN ZEOLITE MIXTURE
100
FIG. 3. CESIUM AND STRONTIUM DISTRIBUTION COEFFICIENTS
OBTAINED WITH VARIOUS ZEOLITE MIXTURES.
The proportions of the two zeolites necessary to achievebalanced loadings of cesium and strontium were determined bymeans of a series of tracer-level tests. A synthetic solutionformulated to a composition similar to the TMI-2 HALW (2000 ppmof boron, 1600 ppm of sodium, and minor amounts of other elements,with 137Cs and o9-90sr added as radioactive tracers) was used with
46
3/1 RATIOIE96/A51
to4 io2 to3 to4 to2 to3 to"COLUMN THROUGHPUT (BED VOLUMES)
FIG. 4. RESULTS OF TRACER-LEVEL COLUMN TESTS.
the volume of TMI-2 HALW available was sufficient for only onetest. That test was made with a 1:1 zeolite mixture, as describedabove. One of the tracer-level tests was also made with a 1:1zeolite mixture and produced results which were not substantiallydifferent from those obtained with the actual HALW. Other tracer-level tests were made with 2:1 and 3:1 ratios of IE-96/A-51. Acorrelation of distribution coefficients obtained for cesium andstrontium with different zeolite mixtures is shown in Fig. 3,while cesium and strontium breakthrough data are shown on a com-parative basis for the series of tests in Fig. 4. These dataindicate that the proper ratio for balanced loading is between 1:1and 2:1. Thus, a mixture containing 60% of IE-96 and 40% of A-51(a ratio of 3:2) was selected.
2.3. ModelingThe tests provided breakthrough data for only one column.
Therefore, the data were fitted by means of the J-function, usingthe constant separation factor model developed by Thomas,l^J toenable calculation of the mass transfer coefficients and extrapo-lation of the data to obtain the estimated performance of asecond, third, and fourth column in series.
The general Thomas equation for the reaction kinetics of ionexchange in a fixed bed is as follows:
3X•(
3Y X(l - Y) - RY(1 - X) (1)
where X and Y are the dimensionless concentrations of the soluteion in the fluid and solid phases, respectively, and R is the
47
separation factor. The variable X is defined as C/CQ, where C andCQ are the concentrations of the solute ion of interest in theeffluent and feed solutions, respectively. The variable Y isdefined as q/q , where q is the actual concentration in the solidphase, and q is the concentration in the solid phase when it isin equilibrium with fluid at the inlet concentration, CQ. Whenthe concentration of the solute ion is small relative to the con-centration of the replaceable ion in the feed (as it is in thiscase), R approaches unity, and the isotherm is linear.
The variable N represents the length of the exchange columnin transfer units and is defined by the expression
N = K;P K /(f/v) , (2)a b a
in which K^ is the distribution coefficient when X = 1, pg is thebulk density of the ion-exchanger, K& is the mass-transfer coef-ficient characteristic of the system, f is the rate of flow ofsolution through the column, and v denotes the overall volume ofthe sorbent bed, including the void spaces. The throughputparameter, T, is defined approximately by:
T = (V/v)/KjpB , (3)
where V is the volume of solution processed through the column.Note that V/v is the number of "bed volumes" of solution.
Since pg is essentially constant, it is convenient to definea volume-basis distribution coefficient, K^ = qv/C0, where qy isthe concentration of the solute ion per unit volume of the sorbentbed (sorbent plus void space) and C0 is the concentration in thefeed solution. Equations (2) and (3) can then be expressed as
N = K.K /(f/v) , (2a)Q 3
and
T = (V/v)/Kd . (3a)
Equation (1) has been integrated [Eq.(16-128a) in réf. 5] forthe special case of reversible second-order reaction kinetics(appropriate to ion exchange) with the solution being
J(RN.NT)_____________J(RN,NT) + [l-J(N,RNT)j exp [(R-l
where J is a mathematical function* related to the Bessel func-tion, IQ.
48
For large values of RN (a condition approached in SDS opera-tion and in the small-scale tests), C/CQ = "-0.5 when T = 1,independent of the values of RN. This characteristic can beemployed in the data analysis. Experimental data can be used toconstruct logarithmic-probability plots of C/C0 vs V/v. Theseplots are nearly linear and can be used to estimate Kjj, which isapproximately equal to V/v at the point where C/CQ = 0.5. Valuesof R and N can then be obtained from the experimental data throughthe iterative use of Eqs. (3a) and (4).
A numerical solution model of the Thomas equation was devel-oped to accept input data in a form that simulates the cyclicalmode of operation proposed for the SDS. A numerical solution wasrequired to analyze the multibed system in which the partiallyloaded columns are moved forward (countercurrent to the waterflow) periodically, because an analytic solution is not practicalunless the initial loading on each bed is zero.
The four columns of the SDS were represented in the model bytwo 4000-point arrays (one each for X and for Y), using 1000points for each column. Calculations were carried out to simulatethe passage of the desired volume of feed through the four columnsin series, with initial values of zero for X(n) and Y(n) for allpoints.
At the end of the first feed cycle, the values of X(n) andY(n) were replaced by the previously calculated values ofX(n + 1000) and Y(n + 1000) for values of n between 1 and 3000 andwere set equal to zero for values of n between 3001 and 4000.This procedure simulated removing the first colunn, moving thelast three columns forward one position, and putting a new columnin the fourth position. The calculations were then repeated foranother cycle, using this configuration as the initial condition.This modeled rotation procedure was repeated for the number ofcycles necessary to process the total volume of HALW.
2.4. Predicted and actual SDS performanceBy interpolating between the experimentally derived distribu-
tion coefficients and the calculated mass transfer coefficientsand separation factors for 1:1 and 2:1 zeolite mixtures, valueswere derived for the 3:2 mixture. These values are shown in TableI, along with corrected values obtained from early SDS operations.The observed differences were not greatly significant, even thoughthe scaleup factor from the test column size to the SDS columnsize was ~105. Whereas the test column data indicated nearlybalanced loading of cesium and strontium, the actual data showedsimilar performance for strontium but better performance forcesium. As shown in Table II, using the early SDS data, cesiumand strontium breakthroughs were calculated for six loading cyclesin which 2760 m^ of HALW are processed (460 m^ in each cycle).Although the strontium breakthrough continued to increasethroughout the six cycles, the breakthrough from the fourth columndid not exceed the concentration (0.1%) of a nonexchangeable spe-cies of strontium observed in the HALW. The number of zeolitecolumns used to process the bulk of the HALW at TMI-2 was reducedto -10.
49
Table I. Comparison of predicted and actual performanceparameters for cesium and strontium
Test column data SDS column dataParameter Cesium Strontium Cesium Strontium
Distribution coefficient (Kd) 2805 3760 3800 3000
Mass transfer coefficient (Kg) 8.0 x 10"*+ 2.9 x 10~^ 2.2 x 10~3 5.6 x 10~3
Separation factor (R) 1.15 1.65 1.0 1.0
Table II. Calculated cumulative breakthrough of cesium and strontiumfor six cycles of column replacement
Cyclenumber
12
34
56
Column1
0.620.71
0.730.73
0.730.73
CumulativeCesium
Column Column2 3
aa
aa
aa
aa
aa
aa
breakthrough (% of feed)Strontium
Column Column4 1
aa
aa
aa
13.822.9
29.534.5
38.341.4
Column2
0.180.92
2.143.62
5.216.81
Column3aa
a0.17
0.350.60
Column4
aa
aa
aa
Calculated breakthrough less than observed concentrations (0.003% of cesiumand 0.1% of strontium) of nonexchangeable species.
3. SUMMARY
The original SDS flowsheet at TMI-2 was conservativelydesigned for the removal of cesium and strontium and would haverequired the use of -60 SDS columns to decontaminate the HALW. Asa process improvement, the use of mixed zeolites to enable in- .creased, balanced loadings of cesium and strontium was evaluated.Tests made on a 10~5 scale indicated that the appropriate ratio ofLinde™ lonsiv" IE-96 and lonsiv™ A-51 was 3:2. A mathematicalmodel was used to predict the performance of the mixed zeolitecolumns in the SDS configurations and with the intended method ofoperation. Actual loading results were similar to those predictedfor strontium and better than those predicted for cesium. Thenumber of SDS columns needed to process the HALW was then reducedto -10.
50
4. ACKNOWLEDGMENTS
This work was performed at the Oak Ridge National Laboratoryand the Savannah River Laboratory under contracts DE-AC05-840R21400and DE-AC09-76SR00001, respectively, with the U.S. Department ofEnergy.
REFERENCES
[1] CAMPBELL, D. 0., et al., "Evaluation of the SubmergedDemineralizer System (SDS) Flowsheet for Decontamination ofHigh-Activity Level Water at Three Mile Island Unit 2 NuclearPower Station," ORNL/TM-7448 (1980).
[2] Evaluation of Increased Cesium Loading on SubmergedDemineralizer Systems (SDS) Zeolite Beds, DOE/SDS Task Force,DOE/NE-0012 (1981).
[3] COLLINS, E. D., et al., "Water Decontamination ProcessImprovement Tests and Considerations," AIChE Symposium Series213, _78 (1982) 9.
[4] HIESTER, N. K., VERMEULEN, T., KLEIN, G., Chemical Engineers'Handbook, Fourth ed. (PERRY, J. H., et al., Eds) McGraw-Hill,New York (1963) 16-2.
[5] HIESTER, N. K., VERMEULEN, T., Chem. Eng. Prog., 48 (1952) 505.
51
APPLICATION OF SILVER-FREE ZEOLITES TO REMOVE IODINEFROM DISSOLVER OFF-GASES IN SPENT FUEL REPROCESSING PLANTS
T. SAKURAI, Y. KOMAKI, A. TAKAHASHI, M. IZUMOJapan Atomic Energy Research Institute,Tokai-mura, Ibaraki-ken,Japan
Abstract
The present work dealt with the interaction of zeolite13X with such organic iodides as CH3I and C2H5I in thepresence of N02 with the intention of applying silver-freezeolites to the iodine removal process of the dissolver off-gas. On the assumption that the dissolver off-gas would bedried for the purpose of removal of tritium, the experimentswere carried out under a dry condition.
Zeolite 13X removes dilute CH3I from a process gas streamwith high efficiencies at =25 °C. It was found that N02decomposed CH3I on the zeolite. Infrared analysis and massspectrometry confirmed that the following reaction took placeon zeolite 13X and also on a silver-exchanged zeolite at=25 °C.
CH3I + 2 N02———>• CH3N03 + 1/2 I2 + NO.
A similar reaction occurs also for C2tÎ5l on the zeolites:C2H5I + 2 N02———)• C2H5N03 + 1/2 I2 + NO.
When a simulated dissolver off-gas (containing =100 ppm-I2, =10 ppm-CH3I and =1 %-N02) was supplied to a zeolite 13Xbed, the I2 which was produced by the decomposition of CH3Iwas adsorbed at the same place of the bed as another I2 whichwas originally included in the gas.
At =100 °C, the zeolite 13X bed traps only iodine and N02passes through the bed without being adsorbed: the zeolite canseparate iodine chromatographically from N02. On the basis ofthese experimental results, an iodine removal process wasdiscussed.
I. IntroductionCaustic scrubbing was probably the first iodine removal
process used and has been adopted in most of the nuclearreprocessing plants in the world. Because of its relativelypoor net trapping efficiency (especially for organic iodides),caustic scrubbers require a secondary iodine trapping processif high efficiencies are needed.
Recently, several new iodine removal processes with higherefficiencies have been developed in many countries, as analternative to the caustic scrubbing technique [1]. Amongthese alternative processes, a dry process which uses silver-exchanged zeolites or silver-impregnated adsorbents has beenconsidered to be promising by many workers. Both elemental
53
iodine (I2) and organic iodides form stable complexes orcompounds with the silver in these adsorbents [1~3]. Thedisadvantage is in that silver is valuable and expensive; asuccessful recovery process of the adsorbent has not beendeveloped yet [I],
On the other hand, silver-free zeolites, e.g. zeolite13X, are placed in a low position in the field of the off-gastreatment. The reason seems to be that silver-free zeoliteslose their iodine removal efficiencies when co-adsorption ofwater occurs. Wilhelm has reported that there was no usefulremoval efficiency for methyl iodide (CH3I) when zeolite 13Xbed was allowed to reach water adsorption equilibrium with thenormal laboratory air (relative humidity 30^60 %) and at higherrelative humidity (85 %) [3].
In order to remove tritium, however, there is a possibilityin the near future that a filter bed of silica gel is installedin the off-gas treatment system of the dissolver off-gas (DOG).When this filter bed is realized, the water concentration of theprocess gas entering the iodine removal process decreases toconsiderably low levels. This situation makes us reconsider theuse of less expensive silver-free zeolites for the iodine removalprocess of DOG.
On this assumption, we have been studying silver-freezeolites in order to probe their potential abilities of adsorb-ing iodine in the presence of NO . It has been already confirm-ed that some kinds of silver-free zeolites adsorb elementaliodine with high efficiencies [4], Our next concern is inbehavior of organic iodides on the zeolites, which wouldconstitute 1 - 10 % of the total iodine quantity in DOG. Noexperimental data are available for the interaction of organiciodide with silver-free zeolites in the presence of NO . Themain species of the organic iodide has been postulated to bemethyl iodide (CH3I) [5].
The present work was carried out to obtain the informationabout the interactions of organic iodides with silver-freezeolites in the presence of NO ; zeolite 13X was used as thesample. A dynamic method was applied for most part of thepresent work with the aid of a tracer technique using radioio-dine 131I. A static method was adopted as occasion demanded.
II. Experimental1. Materials
The zeolites used were: (i) zeolite 13X in the form of1/16" pellets, and (ii) a silver-exchanged zeolite AgX in theform of 1 mm4> particles for reference.
Commercial grade methyl iodide (CH3I), ethyl iodide (C2H5I),and nitrogen dioxide (N02) were used without further purifica-tion.
Radioiodine, 131I, was purchased from Japan RadioisotopeAssociation in the form of a carrier-free Na131I solution.
54
2. Preparations of radioactive methyl iodide (CH3I ) andelemental iodine (I*)
The methyl iodide was prepared through the liquid-phasereaction between dimethyl sulfate ((0113)2201,) and sodium iodide(Nal) labeled with Na131I [6]. The product was dried on azealite 3A bed at 100 °C.
Heating a mixture of K2Cr207 and Nal labeled with Na131Ito 400 °C yielded radioactive elemental iodine [6], which wasalso dried on zeolite 3A.
Their specific activities were =2 mCi/g-CH3I and =300yCi/g-I2.3. Apparatus and procedure for the dynamic method
The dissolver off-gas entering the iodine removal processof the reprocessing plants, has been postulated to include=1 % of NO and 50^200 ppm of iodine. Organic iodides con-stitute 1 - 10 % of the total iodine quantity.
In the first stage of the experiments, the removal effi-ciency of zeolite 13X for CH3I was studied by supplying an airstream containing CH3I (10 ~ 100 ppm) and N02 (=1 %) to anadsorption column (15 mm i.d. x 200 mm length) packed with ^25 gof zeolite 13X. In the second stage, a simulated DOG was fedinto the adsorption column in order to predict the behavior oforganic iodides in the dissolver off-gas. Figure 1 illustratesthe apparatus for feeding a simulated DOG, together withexperimental conditions. (In the case of the first stageexperiments, the I2-feeding line was closed.) The main airstream was dried with a zeolite 3A bed. Prior to use, thezeolite bed was heated at 400 °C for l h in an air flow.
Air » 193 cc/min —-
rnSHAir36.3 cc/min
I2,30*C(V.P=0.5 lorr)
CH3I*-78*C(VP=Q7torr)
1
I
TEDAfilter
N02,0*C<V.P*280torr)
Zeolite 13X cohm KI-I2 impregnatedcharcoal
Linear gas velocity ~ 2.3 cm/sec[I2] = 100 ppm
CCH3I]=: 10 ppm[NOj] A 1 %
Fig. 1 Apparatus for feeding the simulated dissolver off-gas.(V .P . =Vapor pressure)
55
Nitrogen dioxide represented NO in the present work.This would not affect validity of the results obtained, becauseN02 is a main component of NO and also another component NOis partly oxidized to N02 by oxygen and partly passed throughthe zeolite bed without being adsorbed.
The mass flowmeters 'Model SEC-LU1 from StandardTechnology Co. measured the flow rates of the air flows. Sinceno experimental data had been available for the vapor pressureof CH3I at -78 °C, we checked it with a pressure gauge 'MKSBaratron pressure measurement system' from MKS Instruments INC.and obtained a value (0.7 torr or 93 Pa) close to the oneextrapolated from Antoine's formula [7]„ The triethylenediamine(TEDA) filter and the KI-I2-impregnated charcoal capturediodine released from the adsorption column.
After the gas supply of prescribed time, distribution of131I was examined along the length of the adsorption columnand the filter. In measuring radioactivity, a y~scintillationcounter (Nal) was shielded with Pb blocks of 50 mm thick soas to count it only through a slit of 10 mm wide in the frontPb block.
4. Apparatus and procedure for the static method
The interactions between CH3I and zeolites were alsostudied by means of a static method. The experimental apparatusconsisted of a CH3I container (Monel), a N02 container (Pyrexglass), a Monel reactor to be loaded with zeolite samples (24mm i.d. x 150 mm depth), a Monel cylinder to control theamounts of the gaseous reactants, a pressure gauge (Monel) anda vacuum manifold.
About 500 mg of the zeolite 13X was put in the reactor anddried at 400 °C for l h in a high vacuum. The reactor wasclosed and cooled with liq-N2. On the other hand, CH3I and N02were introduced into the cylinder until the prescribed pressureof each gas was reached therein. The gaseous mixture was thentransferred to the reactor and was condensed at its entranceend. The cold trap was removed from the reactor and the systemwas kept at room temperature (=25 °C) for =1 h, in order toallow the zeolite to adsorb the gas. Subsequently, the reactorwas detached from the apparatus for analysis of the reactionproducts. The infrared analysis and mass spectrometry of theproducts were carried out.
III. Results and discussion
1. Removal efficiency of zeolite 13X for CH3IThe interaction of zeolite 13X with CH3I was studied by
the dynamic method in the presence of N02. An air flow including=100 ppm-CH3I* and =1 %-N02 was fed into the adsorption column(25 CC) at a linear gas velocity of 2.3 cm/sec. In^order torealize a CH3I* concentration of =100 ppm, the CH3I container(Fig. 1) was cooled to -50 °C and an air flow was passed throughit at a rate of 2.6 cc/min. Distribution of 131I in the columnwas measured at every 5 h intervals during the gas supply. Figure2 shows the results. The abscissa of the figure denotes the
56
distance from the inlet of the column and the ordinate the amountof CH3I adsorbed at each position along the length of the column.
Methyl iodide was adsorbed near the inlet of the columnand accumulated there with elapse of time. On the other hand,no radioactivity was detected at the TEDA filter downstream ofthe column. It is therefore evident that zeolite 13X iscapable of retaining successfully CH3I at room temperature;a maximum amount of the adsorbed CH3I exceeded 50 mg/g-zeolite13X. In a different run, a CH3I* concentration of 10 ppm,almost equivalent to that in DOG, was realized by cooling theCH3I container to -78 °C. The same results were obtained asto the feature of 131I distribution in the adsorption column.
60r20 h 20cm
TGos 0 cm
Corner gos , Air[CHjJJ , i 100ppmC N02] , = i l % p p mGos velocity ,
23 cm /secTemperoture , 25 *C
0 2 4 6 8 10 18 20Distonce from the inlet of column,cm
Fig. 2 Distribution of CH3I* in the zeolite 13X columnafter the gas supply of 5, 10, 15 and 20 h.
13
When supply of the gas was started, a yellow adsorptionband of N02 was first noticed at the bottom of the zeolite bed.Behind this band, the zeolite bed turned pink due to theadsorption of iodine. When the air stream containing CH3I*alone (100 ppm) was supplied to another adsorption column,CH3I was also adsorbed on the zeolite, showing a peak ofdistribution at the same position as in Fig. 2. However, inthis case, the zeolite bed was not colored at all. Therefore,appearance of the pink adsorption band described above meansthat the decomposition of CH3I* into I* took place in thepresence of N02. The position of the pink adsorption bandcoincided with that of the peak in the figure. The adsorptionfront of N02 traveled much faster than that of I* in thedirection of flow: after 20 h of the gas supply, the formerreached a position of 14 cm distant from the inlet of thecolumn, whereas the latter was at a dstance of only 3.5 cmfrom the inlet.
57
4000 3200 2400 (500
Wove number cm1000
"1400
Fig. 3 Infrared spectra of © CH3I, (2) N02,(2 NCh^^Mi), (D the free gas remaining
' after the adsorption process, and © CH5NOj.
Heating the adsorption column in an air flow resulted inthe release of N02 at 100 - 200 °C and then desorption of iodinebeyond 300 °C. A large amount of iodine (I2) crystallized inthe Pyrex tubing between the adsorption column and the TEDAfilter. Hence, it became clear that CH3I was decomposed by N02on the zeolite:
Another adsorption column charged with zeolite 13X wastreated with the process gas at 100 °C for 5 h. Methyl iodidewas adsorbed completely, showing a peak of 131I distributioncentered at the position of 3 cm from the inlet of the column;however, the yellow adsorption band of N02 no longer appearedanywhere in the column. Keeping the column at 100 °C in anair flow caused the adsorption band of I* to travel graduallyin the direction of flow with elapse of time. Zeolite 13Xseparates iodine chromatographically from N02 at around 100 °C.
2. Reaction of CH3I and C2H5I with N02
Since N02 was found to decompose CH3I, we tried toidentify the reaction pruduct by means of infrared analysis andmass spectrometry. Figure 3 shows the results of the infraredanalysis.
Fig. 3 - (D and @ are the spectra of CH3I and N02,respectively. Fig. 3 - (D shows the spectrum of the free gaswhich remained in the reactor after the adsorption process.In this case, a gaseous mixture of CH3I and N02 in a volumeratio of 1:2 was introduced into the reactor. The amounts ofCH3I and N02 were 119.4 mg and 17.4 mg, respectively. Thespectrum reveals presence of N0> N20 and other unknown sub-stances. Mass spectrometry indicated that the main componentof the remaining gas was NO and the amount of N20 was verysmall. It further revealed that no free oxygen was present inthe gas. After the analyses, the free gas was evacuated bypumping.
58
Fig. 3-0) suggests the presence of other products adsorbed on the zeolite; so the reactor was warmed slowly in order todesorb the unknown products. Beyond 50 °C, their desorptionbecame observable; the pressure increased gradually withincreasing temperature and reached a value of 30 torr (4 kPa)at 120 °C. Fig. 3 - © shows the spectrum of the gas desorbed.Analyses of the infrared and mass spectra revealed that the gasdesorbed was methyl nitrate (CH3N03) in an almost pure form [8].
When a gaseous mixture of CH3I and N02 was kept in theabsence of the zeolite for one hour, its infrared spectrum wasmerely a superposition of those of the two substances. Thereaction of CH3I with N02 needs the presence of the zeolite;it is therefore written down as follows.
CH3I + 2N02 — =25 C> (1/2)I2 + CH3N03 + NO.
In different runs, the silver-exchanged zeolite (AgX) wastreated with the CH3I - N02 mixture under the same experimentalconditions as described already. The same reactions werenoticed thereof.
When the [N02]/[CH3I] ratio was increased to four in adifferent run, the infrared spectra obtained consisted of theabsorption bands of CH3N03, NO, and excess N02.
Although CH3I has been postulated to be the main componentof the organic iodides in DOG, there is still a possibility thatother organic iodides having molecular weights greater than thatof CH3I may be present therein, e.g. ethyl iodide (C2HsI). Sothe same experiments were performed for C2HsI. The result wasformation of ethyl nitrate (C2H5N03), I2, and NO:
C2H5I + 2N02 ' 1//2^12 + CaHsNOs + NO.
Thus, the present work has confirmed that such organiciodides as CH3I and C2HsI react with N02 in the presence ofzeolites, producing I2, the alkyl nitrates and NO.3. Behavior of CH3I in the simulated DOG on zeolite 13X
In order to obtain the information about the reaction rate,a simulated DOG was supplied to the adsorption column (25 °C)at a linear gas velocity of =2.3 cm/sec. The experimentalconditions were described in Figure 1. Prior to the experiments,either I2 or CH3I was labeled with 131I in order to distinguishtheir behaviors. In the experiments, the simulated DOG wassupplied for 25 h at room temperature, and then the column washeated at 400 °C for 1 h; finally the gas was again supplied tothe column for 6 h at room temperature. At the end of each step,distribution of 131i was examined.
Figure 4 shows the results. The "I2 in the gas stream"curve in the figure shows the distribution of the I2 which wastrapped directly from the gas stream during 25 h of the gassupply. The "I2 from CH3I" curve corresponds to the distribu-
59
Xro
OE
Corner gas , Air,= tOOppm
[CHjI] , = lOppm
Gas velocity ,2 3 cm/sec
Temperature , 25 *C
2 4 6 8 10 18 20Distance from the inlet of column , cm
Fig. 4 Distribution of the adsorbed iodine (12)along the length of the zeolite 13X column,(The gas supply, 25 n;
tion of the I2 which originally came from CH3I in the streamthrough its reaction with N02 during the first gas supply (25 h).The location of the two peaks coincides each other: in bothcases, iodine concentrated in the vicinity of the inlet and noradioactivity was detectable on the TEDA filter. There are twofindings available for assuming the reaction mechanism:(i) as described already, the CH3I - N02 reaction does not take
place in gas-phase, and(ii) zeolite 13X adsorbs CH3I itself and there is little differ-
ence in 131I2 distribution between adsorptions of CH3Ialone and of a CH3I - N02 mixture.
Therefore, the first step of the reaction would be adsorption ofCH3I on the zeolite and then N02 oxidizes in situ the CH3Iadsorbed.
Heating the column in air flow resulted in the desorptionof N02 beyond 100 °C and also in small migration of theadsorbed iodine in the zeolite bed. At 400 °C, more than 90 %of the iodine was desorbed from the bed within 1 h.
Subsequently the simulated DOG was again supplied to thezeolite column for 6 h. The I2 in the gas stream and the I2from CH3I were also located at the same position of the columnas in the first adsorption. The zeolite can be used repeatedlyin removing iodine from DOG.
60
D r y i n g Iodine ennchmen!
H removol)
Iodinefixotion
Cul
Fig. 5 An iodine removal process using zeolite 13X.
IV. Conclusion
Zeolite 13X removes both I2 and CH3I with high efficienciesunder the dry condition; so long as N02 coexists, the presenceof such organic iodides as CH3I and C2HsI does not bring aboutadditional problems. The stability of adsorbed iodine ishowever considerably lower for zeolite 13X than for AgX andother silver-impregnated adsorbents. For the long-term storage,the iodine adsorbed on zeolite 13X should be transferred to anappropriate solid material to form stable complexes or compounds,
We are considering an iodine removal process illustratedin Figure 5. The dissolver off-gas which was dried with afilter bed of silica gel is fed to a zeolite 13X bed at =100 CC,where iodine is trapped and NO passes through the bed withoutbeing adsorbed. The adsorbed iodine is then desorbed from thebed at =400 °C and is allowed to react with copper to form Cul,which would be fed to the long-term storage.
REFERENCES
[1] Burger, L. L. and Scheele, R. D., "The Status of RadioiodineControl for Nuclear Fuel Reprocessing Plants", PNL-4689(1983).
[2] Maeck, W. J., Pence, D. T. and Keller, J. H., "A HighlyEfficient Inorganic Adsorber for Airborne Iodine Species(Silver zeolite development studies)11, CONF-680821, (1968)185.
[3] Wilhelm, J. G., "Trapping of Fission Product Iodine withSilver Impregnated Molecular Sieves", KFK-1065, (1969).
[4] Sakurai, T., Izumo, M., Takahashi, A. and Komaki, T.,J. Nucl. Sei. Technol., 20 (1983) 784.
[5] Atkins, D. H. F. and Eggleton, A. E. J., "Iodine CompoundsFormed on Release of Carrier-Free Iodine-131", AERE-M-1211,(1963).
61
[6] Noguchi, H., Murata, M., Tsuchiya, Y., Matsui, H. andKokubu, M., "A Method of Quantitative Measurement ofRadioiodine Species Using Maypack Sampler", JAERI-M-9408,(1981).
[7] Lange, N. A. (Ed.), "Handbook of Chemistry", HandbookPublishers, Inc., Sandusk, OHIO, (1956) 1424.
[8] Strinhagen, E., Abrahamsson, S. and Mclafferty, F. W.,"Atlas of Mass Spectral Data", Vol. 1, Intersciencepublishers, New York, (1969) 97.
62
RADIOACTIVE RUTHENIUM REMOVAL FROM LIQUID WASTESOF "Mo PRODUCTION PROCESS USING ZINC AND CHARCOAL MIXTURE
R. MOTOKI, M. IZUMO, K. ONOMA, S. MOTOISHI, A. IGUCHI,T.SATO*,T. ITOJapan Atomic Energy Research Institute,Tokai-mura, Ibaraki-ken, Japan
Abstract99 235The production of Mo by the reaction of U(n,f) using
UC>2 as target material was carried out in JAERI. Medium- andhigh-level liquid wastes arising from this production processcontained U, Pu and fission products in which radioactiveruthenium in various valency and complex states was included.
For the removal of radioactive ruthenium, after testingvarious combinations of adsorbents, a novel method using acolumn packed with adsorbents composed of zinc and activatedcharcoal (charcoal) was developed. The effects of compositionof mixtures, flow rates, pH of feed solutions, washing, andrepeat use on ruthenium removal efficiency were investigated.A feature of this column method is that the efficiency ofremoval can be recovered by washing with dilute nitric acid andit is possible to treat a large quantity of liquid wastes. Itwas found that this column method was also effective for theremoval of U, 239pu> 144ce, 155Eu and 125sb.
This method was applied to the treatment of actual medium-and high-level liquid wastes, the quantities of which were 600Land 124L respectively. After the removal of U, Pu and fissionproducts by conventional removal methods such as coprecipi-tation, filtration, electrolysis and adsorption with inorganicadsorbents, residual lO^Ru was removed by using the columnpacked with a mixture of zinc and charcoal. Decontaminationfactors for 106RU ranging from 1()2 to 104 were obtained. Theconcentration of U, 239pUj 144ce, 155Eu, and 125sb after treat-ment were found to be below the limit of detection.1. Introduction »q
Through 1977 into 1979, weekly production of 20Ci Mo fromfission product of 235]j was carried out in JAERI. In the pro-duction, irradiated U02 was dissolved in nitric acid and 99Mowas extracted from the solution with di-2-ethylhexyl phosphoricacid. As the 99Mo production procedure was similar to thePurex process, the liquid wastes resulting from the productionwere also very much like that from the nuclear fuel reprocessing.
It has been well-known that radioactive ruthenium might bepresent in multiple valence and complex states simultaneously innitric acid solution [1-7] and it is one of the most questionablenuclides in the treatment of radioactive wastes [8-15]. Develop-ment of an efficient method of removal of radioactive ruthenium,therefore, has been much expected.
In the course of the treatment of the radioactive wastesarising from the 99Mo production, a novel procedure [16,17] wasdeveloped for the removal of radioactive ruthenium on the basis
* On leave from Mitsui Mining & Smelting Co., Ltd.
63
Trapping ofnoble gas
Trapping ofiodine
r
Irradiated DO.._L
Decanning
DissolutionDistillation
Dilution
ExtractionJ~
Washing (1)
Washing (2)_L
Back extraction~r
Washing
Washing waste of apparatusMedium-levelliquid waste Aqueous phase
Decomposition of
Purification byA1O column
duality control
99Mo products (20Ci)
240g(2.6% 235U)
10M HN03 IL
H.O 0.8L
30% D2EHPA-CC14
5M HNO 0.5L
IM HNO3 IL0.5M HN03 0.5LÛ . 5 M UNO, 0 . 4 8 L
+ 304 }12°2 20m1'
CCI . 0 . 9 L
NaN03 30g in H.
IM NH.OH 0.03L
Fig.l Flow chart of ^Mo production from irradiated UÛ2
of columns packed with a mixture of activated charcoal (charcoal)and metal powder, of which zinc was found to be most effective.The usefulness of this method was demonstrated by the successfulresults obtained in the treatment of total amount of actualwastes. It is natural that the radioactive ruthenium removalprocedure developed here should be applicable to the treatmentof wastes occurring from the various phases of the nuclear fuelcycle.
The present paper describes the results of development of,and of the treatment of the actual wastes with, the novelprocedure of radioactive ruthenium removal.
992. The nature and amount of liquid wastes from Mo productionThe procedure of 99Mo production and the origins of wastes
in the procedure are shown in Fig.l. The amount and character-istics of liquid wastes are summarized in Table I togetherwith the decontamination factor necessary for each nuclide tobe disposed as low-level liquid wastes.
64
Table I Radioactivity concentration of liquid wastes anddecontamination factor (DF) necessary for disposalas low-level liquid wastes (ß-y : 10-3-10-a:<5xlO-5yCi/mL)
Medium-level liquid waste(Assayed in March 1981)
Nuclide
106Ru137CS144Ce
a6-Y
Radioactivityconcentration
(UCi/mL)6 x 10~2
1 x 10"1
2 x 10"2
2 x 10"4
4 x 10"1
DF
60100204
400pHU
NaNO.
-133. Omg/L0.3M
NaNO 2
volume0 . 0 2 7 M600L
High-level liquid waste(Assayed in July 1981)
Nuclide
106RU137Cs144Ce125Sb155Eu90Sr
147Pma6-ï
IRadioactivityconcentration
(IJCi/mL)2.5
11220.2-20.21
11-30.03
48
DF
2.5xl03
11220.2 "0.2 "11-30.6xl03
48 xlO3
IIRadioactivityconcentration
(UCi/mL)5
14550.340.28
14-30.0589
DF
5 XlO3
14
550.3 "0.3 "
14-31 xlO3
89 xlO3
U : 23mg/mLHNO, : I:3.BM, II:4.5M
volume I:73L,
The medium-level liquid waste consisted of the scrubsolution of sodium hydroxide used for the removal of radioactiveiodine and dilute nitric acid used for washing of the apparatus.This waste contained a considerable amount of sodium nitrateand was supposed to be equivalent to the washing effluent inthe nuclear fuel reprocessing plant in terms of the radio-activity level [18]. Since this waste was somewhat alkaline,some part of radioactive ruthenium of chemical forms that wasprone to precipitate was supposed to have formed coprecipitatewith sodium uranate. While the high-level liquid waste consistedof the nitric acid solution of irradiated UC>2 from aqueousphase after extraction with organic solvents (di-2-ethylhexylphosphoric acid and carbon tetrachloride) and washing solutionof the organic solvents and was stored in two tanks, tank-Iand -II. The high-level liquid waste here was equivalent tothe medium-level liquid waste from the nuclear fuel reprocessing.
65
In order to decrease the radioactivity concentration inthese liquid wastes to the acceptable levels of the low-levelliquid wastes defined in our Institute, lO^Ru decontaminationfactors of >60, >2.5xlO^, and >5xlO^ were required respectivelyfor the medium-1evel, high-level liquid wastes-I, and -II (seeTable I).
3. Development of Ru removal techniquesIn preliminary experiments, no effective method was found
for the removal of 106RU of chemical forms whose separationwere difficult by conventional means such as coprecipitation,adsorbent columns, and ion exchangers as previously reported ontreatment of liquid wastes from nuclear fuel reprocessing.In the experiments above, it was observed that several kinds ofmetal powders and charcoal were considerably effective for theremoval of radioactive ruthenium when used separately. Basedon these experiences, a method using a column packed with amixture of metal powders and charcoal was devised and tested.In the experiments a mixture of zinc powder and charcoal wasfound to have high removal efficiency for 106RU.3.1. Radioactivity measurement and decontamination factor
Decontamination factor of each nuclide was calculated fromthe ratio of concentration before and after removal treatmentof sample solutions. The concentration was obtained respec-
tively by measuring y-ray intensity with a y-ray pulse-heightanalyzer using Nal(T£) and Ge(Li) detectors, a-ray intensitywith a 2-rr gas flow counter, and ß-y ray intensities with a gasflow GM counter.3.2. Waste solutions used for experiments
Various chemical species of ruthenium were contained innitric acid solution and in solution of sodium nitrate andspecies which could be removed without difficulty by coprecipi-tation and filtration were also included. In order to examinethe efficiency of removal of ruthenium in the form of anionsor in abnormal chemical state where it is difficult to be sepa-rated from the solution, ruthenium components which could beremoved easily should be separated by pretreatment in advance.Such components were coprecipitated with aluminium hydroxideflocculate generated from electrolysis and filtered in the caseof medium-level liquid waste, and they were coprecipitated withsodium uranate by neutralization of nitric acid with sodiumhydroxide in the case of high-level liquid waste. Decontami-nation factors of 106RU obtained by these coprecipitationmethods were ~4, -20, and ~102 for the medium-level, high-levelliquid wastes-I, and -II respectively.
Each solution obtained by the pretreatment was named thepretreated medium-level, high-level liquid waste-I, and -IIrespectively.
The pH of these waste sample solutions was adjusted withnitric acid or sodium hydroxide prior to experiments. For thepreparation of solution containing all chemical species of 106RU)uranium and 239pu were extracted from the high-level liquidwaste-II with 30% tri-n-butyl phosphate in kerosine. Allchemical species of 106Ru appeared to remain dissolved in theresidual aqueous solution since no l°6Ru radioactivity wasdetected in organic phase. This solution was named the ext-racted solution. The pH and sodium nitrate concentration werecontrolled by adding sodium hydroxide.
66
Table II Decontamination factor of fission products withmetal powder-charcoal column and pH of solution
NO.
1
2345
-x. nu-
Métal \x.Al
SnCuFeZn
a
2> 53
12
>340
155EU
1-
2.9x101
2.8xl03
4.9xl03
-Ce
1161
l.OxlO5
1.9xl05
125Sb
1-
1.4x102
1.3xl03
2.2xl03
106RU
8.3xl03_
8.3x1032
B.3xl03
l.GxlO4
137CS
11311
pH
4.5
6.33.16.08.4
NaNO, : No. IM : 1.6 M, No. 5 : 2.5 M
Flow rate : 0 . 2 4 - 0 . 4 cm/mm
pli of feed : No. 1-4 : 2 . 7 , No. 5 : 1.5
For other detailcs, see text-.
3.3. Removal of 106RU using zinc-charcoal columnIn order to find a combination for an adsorbent of metal
powder and charcoal suitable for 106ßu removal, mixtures of tin,iron, zinc, or copper powder and charcoal were tested. Metalpowders having a purity of 99.9% and a grain size of <100 meshand charcoal (Tsurumicoal Co., Ltd. GVA) of 60 to 300 mesh weremixed in water and packed in columns of a diameter of 8 or 18mm.For the purpose of investigating the removal efficiencies forall chemical species of 106RU iQOmL of the extracted solutionprepared above was passed through each column packed with amixture of 3g each of metal powder and charcoal and thedecontamination factors of each nuclide and the change of pHof solutions were evaluated as shown in Table II. Decontami-nation factor of 104 for 106ßu was obtained by using theseadsorbents except a mixture of copper and charcoal and amongthese a mixture of zinc and charcoal indicated the highestremoval efficiency for all nuclides. No peaks of 144ce, ^Eu,125sb, 10(>Ru, 239pu> ancj u in y-ray and a-ray spectra of thepassed solution were detected except ^"Cs.3.4. Removal condition for Ru with zinc-charcoal column
In order to investigate the effects of zinc to charcoalratio, flow rate of feed solution, and pH on removal efficiencyfor lO^Ru, the following experiments were carried out. Forthese experiments the pretreated medium-level and high-levelliquid waste-I were used as sample solutions, since the objectof removal was chemical species of 106RU whose removal weredifficult.3.4.1. Effect of zinc to charcoal ratio
After adjusting pH to 4.9, 30mL of the pretreated high-level liquid waste-I was passed through columns packed with amixture of 0-lg of zinc and Ig of charcoal. The resultsobtained are shown in Fig.2. Decontamination factor ofincreased with zinc to charcoal ratio, indicated 55 at 40%ratio and increased about twice to 102 at 80%, and at morethan that decontamination factor approached saturation.Passing 30mL of the pretreated medium-level liquid waste throughcolumns packed with Ig of 10-85 wt.% zinc-electrodeposited
67
a.o
10'zo
fl
§ 10D 60 80 100) 20 40
% ZINC TO CHARCOAL
NaN03 : 0 .01MFlow rate : 0.88cm/minBed volume : 3mL
Fig.2 Effect of zinc/charcoal ratio on decontamination factorof 106Ru
10J
KOEH
gZO
£ 10*
EHZOOWa
100 20 40 60 80 100
RATIO OF ELECTRODEPOSITED ZINC («)
NaNO,Flow rateBed volume
0 . 4 M0.9cm/min3mL
Fig.3 Effect of the ratio of electrodeposited zinc to char-coal on decontamination factor of
charcoal, which was developed to improve the mixing conditionof zinc and charcoal, relation between electrodeposition ratioof zinc to charcoal and decontamination factor of lO^Ru wasstudied. The results are shown in Fig.3. Decontaminationfactor of lOORu increased with the electrodeposition ratio ofzinc and approached saturation at more than 30%. From thisresult it is recognized that a charcoal which has the electro-deposition ratio of more than 30% zinc could be utilizedeffectively. Decontamination factor attained saturation atmore than 60% of zinc in the case of a mixture of zinc andcharcoal. It appears that zinc-electrodeposited charcoalreached saturation at low mixture ratio, because of smallparticle size and uniform dispersion of particles.
68
10'
«guS§
ZOuMQ
0 2 4 6 8 10FLOW RATE (cm/min)
Bed volume : 3mL
Fig.4 Effect of flow rate of feed solution on decontami-nation factor of
3.4.2. Effect of flow rateAfter adjusting pH to 5.2, 30mL of the pretreated high-
level liquid waste-I was passed through columns packed with amixture of Ig each of zinc and charcoal at various flow rate.Figure 4 shows experimental results. Decontamination factordecreased with increasing flow rate, being -60 at Icm/min, -30at 2cm/min, and 10-20 at more than 4cm/min.3.4.3. Effect of pH
After adjusting pH to 2.7-7.2, 30mL of the pretreated highlevel liquid waste-I was passed through columns packed with amixture of Ig each of zinc and charcoal. The results are shownin Fig.5. Decontamination factor of lO^Ru was affected withconcentration of nitric acid and became 40 at pH6, ~80 at pH2-3, and it approached saturation when pH was lower than 4.Similar results were obtained in the experiments using thepretreated medium-level liquid waste.
• Zinc-charcoal columnO Charcoal column
NaNO3
Flow rateBed volume
0.01M0.88cm/min3mL
Fig.5 Effect of pH of feed solution of decontaminationfactor of l°6Ru
69
3.4.4. Effect of differences in volume of feed solutionThe optimum condition for 106 U removai obtained from
experimental results was as follows:Ratio of zinc to charcoal = >1:1pH of feed solution = ~3In this condition relation between volume of feed solution
and decontamination factor of 100RU was studied by passing 2.2Lof the pretreated medium-level liquid waste, pH of which wasadjusted to 3.1, through a column packed with a mixture of 30geach of zinc and charcoal. The results are shown in Fig.6.Decontamination factors of 10^RU Were >1C)3 at 0.4L feed and~1()2 at 1.2L feed. The removal efficiency of this column wasexpected to attain to -102 provided that the volume of feedsolution was confined within the range of 40mL per Ig of char-coal. Similar results were obtained by the experiments usingthe pretreated high-level liquid waste-I.3.4.5. Recovery effect by washing columns
The zinc-charcoal column method required a large quantityof adsorbent for treatment of a large amount of waste, sincedecontamination factor decreased as the volume of feed solutionincreased. With the condition above, 15kg (45L of apparentvolume) of adsorbent was necessary for the treatment of totalamount of medium-level liquid waste, 600L. The volume reduc-tion ratio should, therefore, be about 13, since no effectiveprocess was known to concentrate 106RU adsorbed in zinc-charcoalcolumn.
In order to reduce the volume of solid waste arising fromthe treatment, possibility of recovery of efficiency of adsor-bents and of repeat use of the column were examined usingpretreated medium-level liquid waste. Immediately after passinglOOmL of the waste solution through a column packed with a
lü
«g|io3
ouu
0 0.4 0.8 1.2 1.6 2.0 2.4VOLUME OF FEED SOLUTION (L)NaNO, : 0.4MFlow rate : 0.8cm/ininBed volume : lOOmL
Fig.6 Effect of the difference in volume of feed solution ondecontamination factor of
70
mixture of 4.5g each of zinc and charcoal, the column was washedrepeatedly with water or nitric acid solution and change ofthe removal efficiency was observed. It was found that althoughthe column efficiency could be completely recovered by washingimmediately after use with water or dilute nitric acid, washingwith dilute nitric acid of pH3 was preferable because no elutionof 106Ru was observed. The amount of washing solution necessaryfor the recovery of column efficiency was approximately equal tothe volume of the column.
The recovery of column efficiency by washing and repeat useof the column made it possible to increase the amount of liquidwaste to be treated by the column and accordingly to decreasethe volume of the resulting solid waste as demonstrated on thetreatment of actual waste below. '3.5. Removal of 106Ru from actual liquid wastes [19,20]
For chemical species of ruthenium which were difficult tobe removed in the medium-level liquid waste, zinc-charcoalcolumns and zinc-electrodeposited charcoal column were usedafter treatment first by electrolytic flocculation, then withzeolite column, and finally with titanic acid column. For thetreatment of the high-level liquid waste, after dilution,neutralization, coprecipitation with nickel ferrocyanide, andadsorption with a slurry of ortho titanic acid, the solutionwas passed through the zinc-charcoal column.
A column with a diameter of 149mm and 5L of volume packedwith a mixture of 2.5kg of 60-80 mesh zinc powder and 1.5kg of60-300 mesh charcoal was prepared for treatment of the actualmedium- and high-level liquid wastes. The liquid waste dilutedwith equal volume of water to 30L was passed through the columnand then washed repeatedly with 5L of washing solution of pH2~3.Relations between the volume of feed solution and decontami-nation factor of lOoRu in the treatment of the medium-levelliquid wastes were shown in Fig.7.
10" f
S 1022
gOy
solution (UpH of feedsolution
10° 30 60 90 120 ISO IBO 210 240 270 300 339 30} 390 420 450,34,30,3^18^18^8^8^21,19,22.22,22,201
pHof effluent 8 8 9 0 100)0 083 82 7 170 7 0 8 5 8 3 7 2 7 0 71 7 Ol
Flow rate : 1.25cm/min
Fig.7 Removal of 106RU from medium-level liquid waste withzinc-charcoal column
71
The treatment of 450L of the medium-level liquid waste byrepeat use of a single column resulted in decontaminationfactor of 10-20 at pH3, and 102 at pH2. Similar results wereobtained using 31% zinc-electrodeposited charcoal as adsorbent.In the treatment of 280L of the high-level liquid wastedecontamination factor of 102 with a feed solution of pH2 wasobtained by washing with dilute nitric acid.
4. ConclusionThe conclusions based on the experimental results and
interpretation of the data are as follows:(1) Column packed with a mixture of metal powder and charcoal,especially zinc-charcoal mixture, and zinc-electrodepositedcharcoal showed high efficiency of removal for all chemicalspecies of 106RUj in particular for the species that weredifficult to be removed by the conventional methods such ascoprecipitation, aggregation and ion exchange, with highdecontamination factor.(2) Besides for lO^Ru, column packed with zinc-charcoal mixturewas also highly efficient adsorbent for the removal of 239pu>U, 144Cej 15SEu> and 125Sb.(3) Zinc-electrodeposited charcoal was preferably used for thepreparation of packed column because the particles of metalwere dispersed uniformly.(4) The optimum condition of removal of Ru by the presentmethod was to use zinc-charcoal column or zinc-electro-deposited charcoal column of 30-80% of zinc content for thetreatment of liquid waste within the acidic region of pH<3,at the flow rate of ~2cm/min.(5) The efficiency of 106RU removal with zinc-charcoal columncould be recovered by washing with water or dilute nitric acidafter use. The column could, therefore, repeatedly be usedfor the treatment of a large amount of waste.
Although the usefulness of zinc-charcoal column for thewaste management in the field of nuclear fuel cycle wasdemonstrated above, several problems are left to be solved byfurther research as follows:(1) Improvement of the efficiency of the procedure for removalof 106R.li and other radioactive elements.(2) Extension or modification of the present technique to beapplicable for treatment by batch system.(3) Elucidation of the mechanism of radioactive rutheniumremoval.Studies concerning these problems are now in progress and theresults will be soon published elsewhere.
AcknowledgementsThe authors are grateful to Mr. T. Sato, Dr. H. Nakamura,
Mr. Y. Kawakami, Mr. K. Suzuki, and Mr. T. Abe of the JapanAtomic Energy Research Institute for their useful suggestionsand discussions given to us.
References[1] FLETCHER J. M. , et al., J. Inorg. Nucl. Chem. , l_ (1955)
378.[2] ZVYAGINTSEV 0. E., et al., Peaceful Uses of Atomic Energy
(Proc. 2nd Int. Conf. Geneva, 1958) Vol.17, UN, New York(1958) 130.
72
[3] Brown P. G. M., et al., Peaceful Uses of Atomic EnergyCProc. wnd Int. Conf, Geneva, 1958) Vol.17, UN, New York(1958) 118.
[4] FLETCHER J. M. , et al., J. Inorg. Nucl. Chem. , j_2 (1959)154.
[5] KOYAMA M., Nippon Kagaku Zasshi, 82_ (1961) 1182.[6] SUGIMOTO S., Radioisotopes, 28, (1979) 429.[7] AKATSU E., Rep. JAERI-M 9159, JAERI, Tokyo (1980).[8] KRIEGER H. L., et al., Rep. ORNL-1966, Oak Ridge National
Lab., TN (1955).[9] FEBER R. C., Process Chem., 2_ (1958) 247.[10] GARDNER E. R., BROWN P. G. M., UKAEA Rep. AERE-R 3551
(1960).[11] ISHIYAMA T., et al., Radioisotopes _L5 (1966) 181.[12] LYNCH R. W., et al., Rep. SAND-75-5907, Sandia Lab., NM
(1975).[13] KENNA B. T., Rep. SAND-78-2019, Sandia Lab., NM (1978).[14] SUGIMOTO S., SAKAKI T., Radioisotopes 2£ (1979) 361.[15] KANNO T., et al., J. At. Energy Soc. Jpn., 2l_ (1979) 737.[16] NAKAMURA H., et al., JAPAN PATENT SH057-50698 (1982).[17] NAKAMURA H., MOTOKI R., et al., IAEA/WMRA/13-81/11 JP04
(1982).[18] SEGAWA T., J. At. Energy Soc. Jpn., 1_2 (1970) 599.[19] IZUMO M., MOTOKI R., et al., Rep. JAERI-M 83-197, JAERI,
Tokyo (1983).[20] MOTOKI R., IZUMO M., et al., Rep. JAERI-M 84-015, JAERI,
Tokyo (1984).
73
ION EXCHANGE CHARACTERISTICS OF HYDROUS OXIDESAND THEIR USE IN THE SEPARATION OF URANIUM AND THORIUM
N.Z. MISAK, H.N. SALAMA, E.M. MIKHAÏL, I.M. EL-NAGGAR,N. PETRO, H.F. GHONEIMY, S.S. SHAFIKNuclear Chemistry Department,Nuclear Research Centre,Atomic Energy Establishment,Cairo, Egypt
AbstractThe ion exchange characteristics of several hydrous oxides havet.een studied. The kinetic studies done with zireonia and ceria have
s'jown that the increase of the charge of cations has a minor effecton their mobility in the exchangers while the increase of charge orcornplexing ability of the anions decreases their mobility. A. dehydra-tion of the cations is deduced. The mobility of cations and anionsnecreases in the presence of alcohols.
For alkali Cation exchange, it was deduced from thermodynamicOata on hydrous zireonia and ceria that the Eisenman model is quali-tatively, and sometimes quantitatively, applicable, and that it canL>e used for obtaining successful predictions of the trends in thethermodynamie functions, ïhis is not the ease with anion exchange.The thermodynamic studies on hydrous zireonia in methanolic solu-tions show that the cation and anion exchange selectivity constantsc-in be, at least in certain cases, predicted from the values in theaqueous media by considering ion-solvent interactions« However, evenin these cases large changes occur in the solid phase.
The physicochemical properties (acidity, capacity and poroustexture) of hydrous oxides have a profound effect on their ion ex-change behaviour. Besides, water structure effects seem to be ofi:reat importance in the case of alumina. The decrease of porosity andpore size seems to lead to the increase of hydrous oxide selectivity.Sorption studies of uranium and thorium from aqueous solutionson hydrous eerie and from both aqueous and oix*d««olvent media onhydrous tin oxide point to the possibility of the use ef hydrousoxides for the separation of uranium and thorium from each otherand from other elements.
INTRODUCTIONHydrous oxides are an important class of inorganic ion exchangers.Fields of their potential application are: chemical seprations
and purification, in particular for radioactive materials, ion exchangecatalysis at high temperatures, treatment of high- teaperature water,and purification of moderator and coolant water of nuclear reactors(1,2). Of course, a rational use of hydrous oxides in the various fieldsnecessitates a good knowledge of their general and sorption propertiesand in particular their ion exchange selectivity behaviour. For thispurpose, many hydrous oxides were prepared and studied in this labo-ratory. This paper reviews the work done.
EXPERIMENTALAll the samples prepared were initially dried at 50°C and stored
in a nitrogen atmosphere over saturated ammonium chloride. Some of thesamples were heated at different temperatures and the surface charac-teristics were measured by nitrogen adsorption isotherms. Most of thesamples were subjected to X-ray, thermal and IR analyses. When differentsamples of the same hydrous oxide are involved, they are referred toby the symbol of the metal atom of the oxide followed by a Roman num-ber.
15
Hydrous ceria : Ce-I and Ce-II were prepared as previously mentioned(3,4). Ce-I (3) is amorphous and its DTA curve shows a large endother-mic peak (only endothermie peaks indicative of loss of water or hydroylgroups will be mentioned) at about 125°C. Ce-II (4) is slightly crys-talline. Its DTA curve haß a large endothermic peak at a bout 130 C,a very small one at about 100 C and a small one at about 24O C. TheIE spectra show that in both samples, the peaks due to water and dueto both water and. hydroxyl groups decrease continously with increaseof temperature.Hydrous zirconia : Zr-I is a Bio-Rad HZO-1 hydrous zirconium oxide(5), whereas Zr-II was prepared as given before (6). Zr-II ie amor-phous.Hydrated alumina ( or aluminium hydroxide) : Al-I, Al-II, Al-IV andAl-V were prepared according to the methods given before (4). Al-IIIis a ready JHDEX chemical designated as Al(OH),. All the alumina hy-drates prepared are semieryetallins and their "degree of crystalli-nity remains almost the same up to the heating temperature ?50 C. Allthe samples give endothermic peaks at 270-280 C. Besides, very smallendothermic peaks around 100 C are present in the DTA curves of Al-IIIand Al-IV. X-ray, DT and IE analyses (4) show that the aluminas aregenerally composed ef one er more of the trthj-arxidee (gibbeit«», b..y-BT-it», norAct*ttna.ite; with varying amounts of an amorphous oxyhydro-xide. Ho significant changes occur in the samples up to 250 C. At350 C, they are transformed to alumina which sorbs water from theatmosphere, partly as hydroxyl groups.Ferric oxide gel : The methods of preparation of Pe-I. Pe-II andFe-III are given in reference 4-, Fe-I »n Fe-II are virtually amor-phous while Pe-III is Bemicrystalline (4). The erystallinity of allthe samples Increases with the increase of the heating temperature.The DTA curves show small endotheviiic peaks around 100 C in all thesamples, large endothermic peaks in Fe-I and Fe-II at about 250°C, andtwo small endothermic peaks at about 250 and 3OO C in Fe-III.From IEanalysis, it was deduced that the gels begin to lose OH groups at tem-peratures ,150 C. At 400 C, the samples still contain OH groups« Be-sides, molecular water is present, which is probably sorbed from theatmosphere.Hydrous tin oxide : Hydrous tin oxide (B-stannie acid) was prepared aamentioned previously (7). B-etannic acid (4) is slightly crystalline.Its DTA curve shows a large endothermic peak at about 135 C. The IBspectra show that up to the temperature 400 C, the water lost is re-gained by sorption from the atmosphere.
RESULTS AND DISCUSSIONS
1-Review of Previous Kinetic and Equilibria Studies on Zr-I, Zr-IIand Ce-IBritB and Kancollas (5) have shown that Zr-I displays the selecti-vity pattern: Li> Na*> K, i.e., the cation of the smaller bare radius ispreferred. These authors suggested that the alkali eactions are retainedin the exchanger in a substantially dehydrated form« The kinetic studiesdone later by Misak et al. (8,9) on Zr-II showed that the increase ofthe charge of cations has a generally little effect en their mobilityinside zirconia. The mobility of the alkali and earth alkali cations inzirconia decreases inthe order: C6>K>Ca~-Sr~B«>Li>Ia. This ordersupports a dehydration of the cations sorbed,Kinetic studies on Ce-I (3) have shown that the mobility of somealkali and earth alkali cations decreases in the order: Ca~Na> E> Ba.This shows also that the cations are sorbed inside Ce-I in a dehy-drated state. For mono- and bivalent anions, the mobility in Zr-II (9)and Ce-I (?) was found to decrease with the increase of the abilityof the anion for eomplexing with the metal atom of the hydrous oxide»In the presence of alcohols, the internal diffusion eoefficients ofthe Na/H and Cl/OH systems decrease due to a stronger interaction ofthe counter iems with the exchange sites (6)*According to Eisenman (10), the selectivity for alkali ions of so-lids containing fixed charges depends on two factors: ion hydration andelectrostatic ion binding. When ion hydration is the dominant factor,
76
the selectivity pattern IB: Cs> K >ffa>Li,i.e. , a préférence for thealkali ion with a larger bare radius or lesser hydratioa energy is dis-played» The completely reversed sequence results when electrostaticbinding is the dominât facto*. Partial reversals result la the inter«mediate conditions* The Eisenman model considers the sorption of ionsin a completely dehydrated state. Besides, it takes into considerationonly eoulombic electrostatic interactions, and the changes in the entro-py of the solid on replacing one ion in it by another are completelyneglected. Nevertheless, this extremely simple model has been qualita-tively successfully applied to organic resins by Reieheberg (11) andto zeolites by Sherry ( !?.)• Sherry (12) found that the neglected en-tropy changes often rei'.force the selectivity patterns predicted byEisenman.We have tried to test the applicability of this model to hydrousoxides. When applying this model to cation exchange in hydrous oxides,reversed sequences may be expected to be favoured by the increase ofcapacity sjnd pH at which selectivity is measured, and by the decreaseof the acidity of the oxide which increases the field strength of theanionie 0~ sites* All these factors increase the role of electrostaticinteractions. Increase of erystallinity may lead to this by increasingthe lattice energy providing for ion dehydration (5).Equilibria measurements were done for many cations and saions onCe-I (3). '<-'iie results for alkali eations and for many anions on eeriaa in 3 other hydrous oxides were subjected to a preliminary comparison.Regardless of the faet that the comparison often involved equilibriumdistribution coefficients and that these coefficients may pertain onlyto a certain set of experimental conditions, the following observa«tions seemed to be of a general nature. She sequence of preference ofthe ion of a lesser hydration energy is observed only with the moreacidic oxides: silica (13), alumina (14). titaaia (15) amd hydrousmanganese dioxide (16, 1?)» On the other hand. partially eat completelyreversed sequences are observed on less aeidie oxides: «ireonia (5,8),oeria (3) and thoria (15)* For the anions, the résulta with the ma-jority of the hydrous oxides (5,8,18-22) show a preference to the ionof a higher eoaplexing ability with the metal atom of the oxide, whichmeans a dominant role of electrostatic interactions.This comparison has urged us to make detailed «ad thermodynamiestudies on several hydrous oxides with the aim of testing the validi-ty of the Eisenman model and examining the effect of th» various fac-tors such as aridity, capacity (or site density), porous texture anderystallinity ea their exchange behaviour, aad in particular the se-lectivity. The ultimate goal is to obtain a clear general picturefor ion exchange on hydrous oxides*The first studies involved the thermodynamics of alkali eatioaexchange and aaion exchange in Ce-I (23,24-)* These studies involved, &.H , £tS of tthe détermination of the AG , &.H , £tS of the exchanges aad the com-parison of the obtained data with similar studies oa Zr- I (5) aadZr-II (25)* these thermodynamie functions were resolved into two com-ponents: one pertaining to the solution phase and involves the ther-
modyaamic hydration fuactions, Gh, AH*1, £Sh, obtained from Rossein-•ky*a data (26), aad the other pertaining to the solid phase, £5*,£!!*,££*, aad is ebtaiaed from the relatioasi AG°« Gh
The Ei seaman model considers only the free energy change (AG)of the exchange, aad this was the only parameter considered in the app-lication of the model to resins (11) and zeolites (12)* We have triedto apply the ideas in the model also to the enthalpy ehaages ( A.H). Thedifferent data for the replacement of Id in the oxide by other alkaliions for Ce-I (25), Zr-I (5) and Zr-II (25) are givea in Table I.Zr-I has a considerably less water eoateat than Zr-II in the var-ious cationie forms (23)* Heace Zr-I is apparently more dense and hasprobably a higher site density due to a less average spatial separa-tion of the exchange sites* The capacities of the oxides are very close
and the ions seem to be substantially dehydrated in them since the wa-ter contents of the different ionic forms are rather similar. Ceria isexpected to be of lower acidity and hence to have a stronger fieldstrength of its exchange sites.
11
Table I
Thermodynamic datd tor alkali cation exchange at 25°C
on Ce-I*, Zr^I** and
0>inSsi
a
a(D oU) -H§ )n
ü.C COo va ^
•
l-l M1 1 'l-t l-l
o o o o oos K a co^ . ^ . Cl
-t
Vr~*l Si Si ^iO O W CO 0)S <3 < < |'
o>3 |U
0).U]
Li/Ea 5.69 0.00 -19.1 -100.0 -109.2 -30.9 105.7 109.2 11.7Ce-I
Li/Cs 9.08 6.0? -10.1 -227.3 -251.5 -81.2 236.4 257.6 71.1
Zr-I Li/Na 2.39 -0.71 -10.4 -100.0 -109.2 -30.9 102.4 108.5 20.5
Li/Ka 2.05 -9.60 -39.1 -100.0 -109.2 -30.9 102.1 99.6 -8.4
Zr-II Li/Cs _0<02 _2j.46 -78.7 -227.3 -251-5 -81.2 227.3 228.0 2.4
x from Reference 23.
aoe from reference 5-soot from reference 25.
Cosidering the above-mentioned facts and remarks« one may expectthat a preference for the ion of a smaller bare radius would be morepronounced for the less acidic ceria (if porosity is of secondary impe»-tancerelative to acidity), and for Zr-I relative to Zr-II. Besides, thecontribution of £3* is expected to be higher for eeria than for zircon-ia and for Zr-I than for Zr-II. Since SÏÏ6 is positiv* for the given ex-changee (Table I) .AH is expected to be less negative or more positivefor ceria compared to zirconia, and for Zr-I compared to Zr-II, Table Ishows that all these expectations are fulfilled for the given systems,which shows that the Eisenman model is qualitatively applicable tothese systems. Table I shows that a£> is very small in the case of theLi/Cs exchange on Zr-II, which shows the quantitative applicabilityof the model in this ease.
The study of the thermodynamics of NCC/Cl", NOT/Br" and NCC/SClTexchanges on Ce-I and the comparison of the results obtained with thosefor Zr-I (27), Zr-II (25) and another zireonia »ample studied by Ruva-rac and Trtanj (28) have shown the inapplicability of the Eisenmanmodel in the case of anion exchange. This was deduced (24) to be dueto solid entropy changes and aoncoulombic electrstatie interactionsthat hamper the applicability of the model in this «ase.The study of the thermodynamics of Li/Ha and Li/Cs exchanges onZr-II in 30$ and 50$ methanol (29) has shown that th» values of AHand AS0, given in Table I, increase preressively on addition of metha-nol whereas the reverse occurs with AG . Thus, the selectivity constantincreases due to the entropy term. Using the data oa the thermodynamicsof transfer of the alkali ions from agueous to methanolie solutions,the values of the changes in AG , AH and US due to ion-solvent in-teractions on addition of methanol were calculated. From these values,the changes in these functions due to changes in the solid phase werealso calculated. The latter changes were found to be considérable. How-ever, for the Li/Na exchange in J0# nethanel. ÛG calculated theore-tically on the basis of ion-solvent interactions alomc was foumd tobe very close to the experimental value. It seems therefore that thesolid phase changes lead to contributions to &H and UB° that largelycancel each other. The thermodynamics of halide ion/ nitrate ion ex-change on the same exchanger in the same medium (50) revealed similartrends. However, for the nitrate ion/ bromide ion exchange in 60$ me-thanol, AQ changed du* to the enthalpy tern.
78
2-Recent Studies on the Effect of the Pbysieochemieal Properties ofHydrous Oxides on their Ion Exchange Behaviour
a-Kinetics of exchangeThe values of the diffusion coefficients of Cs* (trace ion) in
Al-IV and Al-V (Na-form) from a 0.1M NaOH medium are given in Table IIat different heating temperatures of the exchangers. This table con-tains also the data of the calculated mean poe radii of the samples»
Table II
Dirfusion coefficients at 25°C of Cs in Al-IV and Al-V
heated at different temperatures.
ExchangerHeating teup., ilean pore DxlO
°C radius, A cm Aec
Al-IV
50250350
442722
861676.9
Al-V
50
250
550
776917
398
43.39 A
Table II shows that for one and the same sample, D decreases withthe decrease of the mean pore radius. Since D does not depend only onpore dimensions but also on the capacity of the exchanger, the strengthof interaction with the exchange sites and the degree of dehydration(3l)of the diffusing species, the above-mentioned fact may show that thepore width is the main factor governing the rate of ion diffusion inthe exchangers.
It must be mentioned here that the mean pore radius may not bewell representative of the actual width of the pores in which diffusionoccurs. This ie so because of differences in the geometries of poresan! because the exchange sites may be restricted to pores of certaindimensions. These factors may be a reason of failure of comparison ofdiffusion rates in samples of different preparations on. the basis ofthe mean pore radius, even if the above-mentioned other factors are ta-ken into consideration.Other data on ferric'oxide gels (4) show also the major effect ofpore size on the internal diffusion kinetics.
b-Selectivity behaviourBefore giving and discussing the data ef the selectivity behaviourof alumina and ferric oxide, it is important to show here the'anomalous'behaviour of Al-I heated at 350 C in respect of the variation of sor-ption of Na with pH. Figure 1 shows the variation of Borption of sod-ium from 0.1M solutions (NsjCl+HaOE) with pH on Al-I heated at differenttemperatures. Up to 250 C, the sorption increases with pB, as would beexpected for an ion exchange process. For Al-I heated at 350 C, the sor-ption first decreases sharply and then increases with increase of pH.It has been proved that the sorption of Ha on Al-I heated at 35O Cis an ion exchange process at all pH values. At 350 C, the exchangesites are presumably OE groups resulting from water sorption by thealuminium oxide fromed at this temperature. The first sharp decreaseof sorption with increase of pH tb therefore surprising. A plausible ex-planation of this situation »ay be related with water structure effectsand the porous texture,of the sample. At 350 0, Al-I is mainly compo-
79
sed of narrow pores (4). Water structure is expected to be enhanced atthe oxide/water interface anfl water structure effects aay be more pro-nounced in the narrower pores and may increase with increase of pH,resulting in a more ionisation of the surface protons that are water-structure makers. Therefore, at the relatively low pH values, Na ionswill be excluded from the narrower pores. As the pH increases, exclu-sion from the narrower pores increases and a decrease of sorption withthe increase of pH say occur when this exclusion outweighs the norma-lly -expected inoi-siee of »orptioa with i&ereaae of pH. Inothtl- expla-nation may be the formation of a double layer, whose potential increa-ses with increase of pH , at the mouths of pores* This double layermay constitute a barrier hindering the penetration of ions in the pores.For saving space, selectivity data will be given here only forAl-I and Al-IV. The selectivity coefficients were measured at differentloadings of the exchangers and the values given represent the selecti-vity coefficient at 50$ loading of the Na-form of the exchanger withthe enterinr alktdi ion. This coefficient, designated here as KQ ^ ,
may be equal to or quite close to the selectivity constant (11) and isanticipated to be a good representative of the overall selectivitydisplayed. The values of KQ ,- for the different alkali ion/ alkali ionexchanges in. different media are given in Table III for Al-I heated atdifferent temperatures anfl Al-IV dried at 50 C.
Table III shows that with the exception of the selectivity patterndisplayed in 0.01M hydroxide solutions on Al-IV dried at 50 C, aluminashows a selectivity to Ha over all the other alkali ions. For the otheraluminas, a selectivity to Na over the other alkali ions is also dis-played in all cases (4). The selectivity to Na over K and Cs may beattributed to the large effect of the latter ions in breaking the wa-ter structure and probably also to a stronger electrostatic interactionof Na with the exchange sites. The selectivity to la over Li may bemainly due to the much higher hydration energy and hydration need of Li.
Table III
K« c for alkali ion exchange at 25°C on Al-I and Al-IVunder different conditions.
(l.!=alkali ion, the ion in the superscript is the entering
ion).
i u ,00 -Pg 5 i s g 5 où . v3 S Id g 3Ï3 S S, -Ll -K -CS -K -CS
W(C 0)0 0
50°C 0.0111 MOH 0.024 0.1? 0.16 0.15 0.94 0.94 0.88
50°C 0.10U MOH 0.240 0.18 0.13 0.08 0.72 0.61 0.44
250°C 0.01M UOH 0.025 0.16 0.1? 0.15 1.06 0.86 0.94
250CC 0.10M MOH 0.250 0.19 0.19 0.12 1.00 0.63 0.63Al-I 50°C 0.10U MCI 0.029 0.028 0.032 0.048 1.14 1.50 1.70
(pH=8.3)
250°C 0.10M MCI 0.045 0.014 0.018 0.028 1.29 1.56 2.00(pH=8.3)
350eC 0.10M «Cl 0.150 0.10 0.08 0.18 0.80 2.75 1.80(pH=8.3)
350°C 0.10M MOH 1.01 0.35 0.29 0.32 0.83 1.10 0.91
D.01U UOH 0.085 0.62 1.03 1-26 1.66 1.22 2.03
0.10M MOH 0.342 0.1? 0.19 0.25 1.12 1.32 1.4?Al-IV 50°C
O.Olil MCI 0.029 0.3? 0.50 0.57 1.35 1.14 1.5*0.1011 MCI 0.117 0.09 0.12 0.13 1.28 1.08 1.44
80
In the chloride solutions, the affinity and selectivity to He is muchhigher than in the hydroxide solutions. This seems to be due to a lession exclusion from narrower pores in the former solutioas. In the na-rrower pores» the selectivity to Na over K and Cs increases due to moreprofound water strucure effects, and that over Li increases due to amore role of ion hydration.
With Al-I heated at 50 and 250°C, the increase of site density ge-nerally increases the affinity or selectivity to the ion of a smallererystallographic radius. This may be taken to indicate small differen-ces in the hyiration degrees (large dehydration) of the alkali ionsinside Al-I. This does not seem to be the ease with Al-IV.The increase of the heating temperature of Al-I f som 50 to 250°Cleads to the increase of surface area from 27 to 190 m /g, the increanaof total pore volume (porosity) from 0.08 to 0.54- c.e,/g, amd the dec-rease of the mean pore radius from 57 to 36 &. Xt 350 C, the surfacearea and porosity continue to increase to 437 m /g and 6.87 c.c./while the mean pore radius remains almost at the same value of 4OIn the hydroxide solutions, the site density is at 50and 250 C. The increase of heating temperature from 50 to 250 C leadsto generally lower selectivity in the case of 0.1M hydroxide solutionsand to almost ao change in the case of 0.&1M solutioas. In the formercase, this may be due to the effect of an increasing porosity, outwei-ghing the effect of decreasing pore size. In the latter ease, the twoeffects balance each other probably because narrower pores that becometoo narrow at 250 C are involved in the exchange. For the sample hea-ted at 350 C, the selectivity is in general lowest, which is presumab-ly due to the much higher porosity of this sample, outweighing theeffects of a higher site density (Table III) and acidity (Fig.1). Inthe chloride medium, the selectivity seems to increase with increaseof heating temperature from 50 to 2^0 C. This is probably due to thedecrease of pore size , which may be dominant due to tVv» involvement ofmuch narrower pores in this medium. With Al-II and Al-III. the results
O) indicate that in the conditions favourable for comparison, thedecrease of pore size increases the selectivity.
90
80
70
60
tt
Û 50st-'
«0
30
10
10
7 « 9 10 11 12 UEqui l ibr ium pH
Fig.f 1 »Variation with pH of V. uptake of Na*fromO-1M Na*solutions on Al-l heated at different temperatures
( V/m = I0m!/g )
The capacity-pH curves (not given here for the sake of brevity)show that the acidity of the samples decreases in the order: Al-IV •>A1-I1I>Al-I>Al-II. Table III shows that the selectivity to Na isconsiderably hip.her in Al-I than in Al—IV. Comparisons for other alu-
81
minas (4) show also that this selectivity is higher in the samples oflower acidity. Thus, the selectivity of alumina to Na increases notonly with the decrease of pore size but also with the decrease of siteacidity,
Fe-I and Fe-II have shown the same kinetic and sorption behaviour fin agreement with their similar X-ray, IR, DTA a&d porous texture data(4). The selectivity data on Fe-II sad Fe-III under different condi«.tions are given in Table IV,
Table IV
KQ c for alkali ion exchange at 25°C on Fe-II and Fe-III
under different conditions.
'(M»alkali ion, the ion in ti.e superscript is the enteringion).
v bo u l a *»>bo o o f a o .pa - H . - H O - H -H. . . . „0} 4* Ot r-l iH -P OUÛ v^i TfK V^® V^ v*US „USo S I "30 3 §,> T*a *Na *Na *Li *K *Li
50°C O.ODi MOH 0.110 4.40 6.80 9.30 1.55 1.37 2.11
50°C 0.1011 MOH 0.310 8.60 5.80 3.40 0.67 0.59 0.40
j,e_IZ 200°C 0.0111 UOH 0.091 1.20 1.40 1.90 1.17 1.36 1.58
200°C 0.10U MOH 0.220 1.48 0.63 0.49 0.43 0.78 0.33
400CC 0.0111 MOH 0.058 1.60 2.80 4.10 1.75 1.46 2.56
400°C O.lOli MOH 0.190 1.68 0.48 0.38 0.29 0.79 0.23
50°C 0.01U UOH 0.104 p.80 7.90 10.10 2.10 1.28 2.66
50"C 0.10M MOH 0.410 7.00 6.30 4.10 0.90 0.65 0.59Fe-III 200°C O.OUU 130H 0.090 1.70 2.20 2,?0 1.29 1.23 1.59
200°C 0.10M MOH 0.32 1.40 1.00 0.62 0.71 0.62 0.44
400°C 0.0111 UOH 0.056 3.90 3.80 4.70 0.97 1.24 1.21
400°C 0.1011 MOH 0.170 2.64 1.40 1.19 0.49 0.85 0.42
Table IV shows that at all heating temperatures, the increase ofsite density for both Fe-II and Fe-III leads to a large iacreaee ofthe affinity or selectivity to the ion of a smaller bare radius. Thisprobably shows that the alkali ions are largely dehydrated. Selectivi-ty reversals occur, whereby the ion of a smaller bare radius becomespreferred at the higher site density. This is probably due to the factthat more weakly acidic sites are involved at the higher site density.Before comparing the selectivity sequences, it oust be mentionedthat Fe-II has a porosity of 0.14t 0,36 and 0.24 c,c,/g and mean poreradius of 23, 68 and 69 A at the heating temperatures 50,200 and 400 C,respectively. For Pe-III, these values are 0.29, 0,53 aad 0,19 e.c./gand 40, 65 aad 80 A, respectively. The acidity of Fe-III is higher thanthat of Fe-II at all heating temperatures. At 200 Ct th» acidity ofboth samples is almost the same as that at 50 C up to a capacity ofabout 0,1 meq/g. At higher capacities, the average atlëity is apprec-iably less at 200 C, At 400 C, the acidity decreases considerably forboth samples.In 0.01M solutions, the site density of ïe-H and Pe-III driedat 50 C is almost the same. In these conditions, the selectivity dis-played for the ion of larger bare radius is generally somewhat higherfor Fe-III. Since both the porosity and pore sice are higher for
82
Fe-III, this may b« Aue to the higher acidity of this sample. For both»amples heated at 200 C, The capacity in the 0.01M solutions differslittle from that at 50 C. Besides, the acidity is, as mentioned above,the same» Table IV shows that in these solutions, the selectivity ofboth samples becomes much lower at 200 C, This is clearly due to thelarger porosity and pore size at this heating temperature. In 0.1Msolutions, the site density and acidity at 200 0 are considerablylower than those at 50 C for both samples. In these solutions, theselectivity sequences are largely different at these two heatingtemperatures and with the exception of the Li/Ha exchange they showa generally much higher affinity to the ion of the smaller bare rad-ius at the higher heating temperature, which is presumably due tostronger electrostatic interactions as a result of the lower acidity*However, the lower site density and higher porosity and pore sizeat 200 C reveal theaselves mainly in the much lower selectivity forLi over Na.The selectivity pattern displayed in 0.1M solutions by Fe-IIheated at 2OO C iß: Li>Na>K^Cs. This sequence is reinforced in the
same solution at the heating temperature 4OO°C where, compared to thetemperature 200 C, the acidity and porosity are lower whereas thesite density and pore size are almost the saae. The decrease of aci-dity reinforces this pattern, showing a preference for the ion ofa smaller crystal]ographie radius, ty increasing the role of electro-stativ interactions, whereas the decrease of porosity reinforces itby increasing the selectivity.5-Sorption of Uranium and Thorium by Hydrous Tin Oxide and Ce-II
Uranium and thorium are sorbed on hydrous tin oxide from acidicsolutions (7,52). At a pH of 2 acad 0.01M uranyl sad thorium salt sol-utions, the capacity of this oxide is 0.36 and 0.44 meq/g for uraniumfrom the aqueous nitrate and sulphate media, respectively, and 0.26and 1.1 meq/g, respectively, for thorium. In the aqueous medium, theU/Th separation factors are rather poor. However, these factors ea»become very high in the mixed solvents. Thus, for example, for O.OO1Mnitrate solutions in 80# acetone and 0.01M nitric acid, K- for uran-ium is about 600 and for thorium it is about 25 (V/m-25 ml/g). ForO.O01M sulfate solutions in 20# acetone and 0.1N sulphuric acid, K,for uranium is about 320 while for thorium it is about 1 (V/m « 25°ml/g). Schemes for UAh separation have been suggested (7,32) on thebasis of these and other results.
Relatively large uranium and thorium sorption capacities are alsodisplayed by Ce-II. At pH 3» the thorium capacity from 0.01M nitratesolution is 0,9 meq/g, and at pH 3.2, the uranium capacity from thesame solution is 0.44 meq/g. However, while uranium and thorium seemto be sorbed by hydrous tin oxide by an ion exchange mechanism, thesorption mechanism in ceria seems to be quite complex. Besides the rateof attainment of equilibrium is very slow in ceria, requiring more than-i month. Nevertheless, the majority of exchange takes place in thefirst hour.
Experience shows that many hydrous oxides do not sorb any signi-ficant amounts of alkali, earth alkali or rare earth ions from acidicsolutions. It is therefore anticipated that hydrous oxides can be usedof uranium and thorium separation from many other elements.
ACKNOWLEDGEMENT - The authors wish to thank deeply the InternationalAtomic Energy Agency for financial supprt under the Research ContractNo. 2716/RB, 2716/R1/RB, 2716/R2/RB.
REFERENCES1.A.MPHLETT,C.B.,"Inorganic Ion Exchangers, McGraw-Hill, New York(l964).2.RUVARAC,A.."Inorganic Ion Exchange Materials",Editor: GLEftRl-'IELD.A.,Chap.5, CRC,Florida(l982).3.MISAJC,N.Z.,MIKHaIL,E.M.,J.appl.Chem.Biotechnol.28(1978)499.4.KISAK.N.Z..et al.,Progress reports submitted toTfcEA. .Research Con-
tract No. 2716/RB,2716/R1/RB ,2716/R2/BB(1981-1984).5.BRITZ.D.,NANCOLLAS.G.H.,J.inorg.nucl.Chem.31(1969)3861.6.MISAK,N.Z.,GHONEIMÏ,H.F.,J.Chem.Tech.Bioteehnol.32(1982)709.7.MISAK,N.Z.,SA3AMA,H.N.,EL-NAGGAR,I.M.,Chemica SerTpta 22(1983)74.
83
8.HALLABA,E. ,MISAK,N.Z.,SAIAMà,H.N,,Indian J. Chem.11(1973)580.9.MISAK.N.Z .,SALAMA,H .N.1Inorg.Nucl.Chem.Lettersl1Tl975)559.
10.EISENMAN.G.tBiophys. J.2(1962)259. —11.REICHENBERG,G. "Ion ExeEange"»Editor:MARINBKY.J..Marcel Dekker.
New York(1966)227.12.SHERRYjH.S.,"Ion Exchange"»Editor:MARINSKY,J.,Marcel Dekker,New York
(1969)89.13.TIEN,H .T.,J.Phys.Chem.69_(l963)350.14.CHURMS.S.C..J.S. JLfr.Chem.InBt.19(1966) 98.15.HEITKER-¥IBGUIH.C. .LBtU-YARON,AT. J.inorg.nuel.Chem.28(1966)2379.16.GEREVINI,T.,BMOGLIaBA,R.tEnergia Nucl. Milano 6(19597 39.17.LEONT «EVA.G.V.^VOL'KHIN^.V^Izv.Akad.irauk SSBR",N*org.Mater.^
(1968)728).18.FLOOD,H.,DiBeuBsions Faraday Boe.7(19^9)180.
21 IAHRIAND , B . teAUESOH ,0 ., J. inorg .nuel .Chêm. 33(1971 ) 2229.22.DONAIJ)SON, J J). tPTmLER,M. J. , J.inorB.nucl.C5«m.32(1970)1703.23.MISAK,N.Z. t MIKmiL.E.M. lJ.inorg.nucl.Chem.43TT981)1663.24.MISA.K,N.Z. .MTKmilL.i.M., J.inorg.nucl.Chem.^Tl981)1903.25.GHONEIMY,HiI.f M.Se.. Ain-Shama Univareity, Cairo(19?8).26.ROSSEINSEYtD^.jChem.Rev.65(1965)467.27.NAHCOLIAS,d.E. JaEH),D.S.,77inorg.nucl.Cham.31 (1969)213.28.RUVARAC Jl. ,0?RTANJ.M., J.inors.nucl.Chem.3i(1972)3893) .29.MISAK.N.Z., GHOHEIMY,H.P.,Colloids and BurfaceB 7(1983)89.30.MISAK,N.Z., GHOKEIMY,H.?.,J.Chem.Teeh.Bioteehnol732(1982)893.31.HELFFERICH,F.1"Ion Exchange",MoGraw-Hill,New YorkrT962).32.MISAK,N.Z.,SAIAMA,H.N.,EL-NAGGAE,I.M.,Chemiea Bcripta (in press).
84
INORGANIC SORBENTS IN SPENT RESIN INCINERATION -THE ATOS PROCESS
C. AIROLA, A. HULTGRENStudsvik Energiteknik AB,Nyköping, Sweden
Abstract
The spent resins from the operation of nuclearpower plants form a dominating waste category, byactivity contents and by volume. Volume reduction byincineration is technically feasible but introducesproblems by activity release in the off-gas and inash solidification.
By a chemical pretreatment that transfersradioactive nuclidés to inorganic sorbents a consid-erable improvement is obtained in the incinerationroute. The paper summarizes efforts at STUDSVIK ondevelopment work along this line over the last years,resulting in the ATOS process. Included are resultsfrom wet activity transfer experiments, incinerationof pretreated resins, and ash immobilization. Theoverall conclusion is that the ATOS process offersclear advantages as to clean off-gases, significantvolume reduction and stable products.
1. INTRODUCTION
Spent ion exchange resins from power reactoroperation contain more than 95 % of the total radio-activity of wet reactor wastes. Cementation andbituminization are the two methods applied in Swedenup to now for the immobilization of these resins,both methods leading to final waste volumes by farexceeding the volume of the original resins.
The reactor wastes in Sweden will be disposed ina rock cavern under the sea bottom of the Baltic,1 km from the shore. In order to minimize disposalcosts it is desirable to reduce the volume of thewastes before disposal /!/.
One process developed and discussed in Swedenfor volume reduction of spent resins is the PILOprocess /2/. In the PILO process, the resins areeluted and the eluted activity (99.9 % of Cs-Sr) issorbed in an inorganic material, such as zeolite. Theeluted resin may then be incinerated or, possibly,disposed without further treatment.
In the ATOS process a chemical pretreatmentenables that the resins are incinerated with theactivity retained in the ashes as well as most of thesulphur from the cation resin.
85
2. THE ATOS CONCEPTATOS is an acronym for Activity Transfer from
Organic resins to inorganic Stable solids.Process flowsheets for ATOS and PILO are shown
in Figure 1.In the ATOS process the spent resins, beads or
powder, are mixed simultaneously with an elutingagent and an inorganic sorbent. The slurry is driedand incinerated/calcined under controlled conditions
PILO ATOS
SPENT SPEN1
SORBENT
EV
FIG. 1 Process flowsheets for PILO and ATOS
The incineration residue, a mixture of ash, inorganicsorbent and metal sulphates, may be immobilized indifferent ways. It may be sintered or melted withadditives to a stable end product, or solidified bycementation, depending on local conditions at thesite where ATOS will be applied.
The main objectives of the ATOS process are thefollowing:
transfer of cesium and other radioactive elementsfrom the organic resins to an inorganic sorbentby a pretreatment processsignificant volume reduction of the waste byincineration of the pretreated resinseasy adaptation to already existing processesfor incineration and/or solidificationsimplified cleanup of off-gasesa minimum generation of secondary waste.
86
The treatment stages to reach these objectivesinclude a simple transfer of resin-water slurriesfrom spent resin storage tanks to a mixing vessel,the addition of pretreatment chemicals, removal ofexcess water after treatment, transfer of the dewa-tered resin to a dryer/incinerator, incinerationunder controlled redox conditions thereby ensuringretainment of volatile activity, mainly cesium, andcontainment of most of the sulphur as metal sulphates.
3. PROCESS DEVELOPMENT /3-13/3.1 Pretreatment
One main problem of the combustion of radioactivespent resins is the release of the volatile activity,principally Cs-compounds.
Studies were thus initiated at STUDSVIK assessingdifferent ways by which Cs-compounds could be retainedin the incineration residue during combustion wheredifferent additives such as zeolites and other inor-ganic sorbents were tried. In the drying-calcinationprocess cesium is trapped in the structure of Na-Al-silicate systems, which gives a good retention ofcesium during this stage of the process. Studies ofthe thermodynamic stability of various Cs-compoundsunder typical redox conditions encountered in afurnace were also conducted.
Bench scale tests with different eluting agentssuch as oxalic acid, Na-oxalate, tartaric acid, Na-tartrate, citric acid, Na-citrate, phosphoric acid,various phosphates and various formates showed thatamounts of 2-4 meq's of the eluting agent per ml ofresin is an optimal amount leading to almost exhaus-tive activity elution from the organic resin.
The different sorbents tested include Na-tita-nate, bentonite clay, mordenite, Ca-phosphate, Zr-oxide, Zr-phosphate as well as specially tailoredsorbents like OXTI, PHOZIR A, PHOMIX A, POLYAN M,PERTITANIC and FE-CO-K (Applied Research Sprl,Belgium).
The performance of the above mentioned sorbentsshowed the PERTITANIC to be the best but also themost expensive one. Second choice is Na-titanate andbentonite clay and mixtures/blends of these. As Na-titanate is a good sorbent for Co-60 and bentoniteclay for Cs-137 a standard sorbent consisting of 80 %Na-titanate and 20 % bentonite clay was used as areference. The retention coefficient K, of this mix-ture was for Cs, K,=200 and for Co, K,=600(Kd=Csorbent ' Solution' in ml/^ '
The result of these studies led to the develop-ment of a process for the conditioning of spent resinsprior to combustion. A simplified process flowsheetis shown in Figure 2.
The spent resins are transferred batchwise to amixing vessel. From each batch a sample is withdrawn
87
ELUTINGAGENT
INORGANIC CALCINATIONSORBENT ADDITIVES
TO OFF-GASCLEANING SYSTEM
PRECONDITIONED RESINSTORAGE TANK
Simplified flowsheet for the ATOS pretreatment processFIG. 2
and analyzed. Based upon this analysis eluting agent,inorganic sorbent and calcination additives/sulphurretainers are added. After treatment the conditionedresin is transferred to a storage tank. From thistank the resin is fed to the incinerator.
3.2 IncinerationA market survey made in late 1980 showed that at
that time no proven commercial incineration systemfor resins was available. The development of such asystem has been in progress in the USA, where e.g.the Aerojet company has demonstrated pilot plantscale incineration of ion exchange resins /ll/.
At STUDSVIK incineration tests were performed,at first in laboratory scale to evaluate and optimizethe pretreatment process. Later a prototype resinincinerator was built and tested on the kg/h scale.
Studies of the incineration process by thermalanalysis showed that:
untreated resins decompose in air at tempera-tures up to 300 C . In pure oxygen the gasesignite explosively at about 300 C if driedresins are usedif wet untreated resins are used volatile decom-position products are distilled with the water.Ignition in oxygen now takes place at about400 °Cwet pretreated resins ignite explosively inoxygen at about 260-280 C and are fully com-busted at about 500 C
88
unloaded and untreated resins do not ignite evenin oxygen, but decompose slowly up to about700 °C.Bench scale furnace tests were then made to
verify the above mentioned observations, and resultsfrom these are shown in Table 1. In this table resultsfrom combustion in normal air and in oxygen are com-pared. The reason for using an oxygen atmosphere wasto get a clean and complete incineration. Cesium mayalso be retained as Cs^SO. which is formed in thepresence of sulphur under oxidizing conditions.
From the table it can be seen that during com-bustion of untreated resins in oxygen some 7 % of theactivity are emitted and that this figure can belowered to about 0.1 % by preconditioning of theresin.
TABLE I Activity measurements from incinerationtests
Resintreatment
Untreated
Untreated +inorganicsorbentPretreated
Pretreated +inorganicsorbent
Activitymeasured
TotalCs-137Cs-60TotalCs-137Co- 60TotalCs-137Co-60TotalCs-137Co- 60
Emission% of initial activity
Oxygen atm. Air a tin.6.65.56.81.63.31.30.40.60.40.040.080.2
0.020.040.020.0040.0040.0060.020.050.020.060.020.1
Combustion in air showed a variation of theemission of activity between 0.02-0.06 %. Theselevels are so low that the accuracy of measurement ofthe low amounts of activity used are somewhat uncer-tain.
The tests also showed that good combustioncharacteristics with a high combustion efficiencyindeed was obtained in an oxygen atmosphere whereasthe amount of unburned resin was appreciable aftercombustion in plain air.
A prototype pilot plant resin incinerator with acapacity of about 1 kg/h (dry weight basis) was desig-ned and built on the experiences obtained through thebench scale tests. The pilot plant incinerator isshown in Figure 3.
The resin is fed to the incinerator eitherthrough a screw feeder (powdered resins) or through afunnel feeder (bead resins). Incineration takes placeon a heated turnable grating in an oxygen enrichedatmosphere. The off-gases pass through an after-burner to a condenser and from there through a HEPAfilter and activity monitoring to the laboratorystack.
89
FIG. 3 Pilot plant incinerator
90
The results from a first series of inactivetests are summarized in Table II.
From this table it is seen that a considerablevolume reduction factor may be achieved. The off-gasescontain normally about 3500 ppm S02. This is of thesame order of magnitude as reported by Aerojet. Byadding suitable calcining additives preliminary testsindicate that the S0_ content can be reduced to below100 ppm. Further optimization of the incinerationprocess is required, but the results so far are quitepromising.
TABLE II A summary of the inactive test runs
Resin feedkg/hIncinerationtemp/ CIncinerationtime/minIncinerationresidue/mlresidue/g
Range
0.4 -
700 -
50 -
80 -40 -
0.7
850
170
250170
Average
0.5
800
80
15085
Volume reduction factorby vol 4 - 1 9 11by weight 5 - 2 1 14Offgas comp.oxygen/% 7 - 2 8 17SO, /ppm 1000 -5000 3400NCT /ppm 100 -2500 750
By addition of appropriate calcination additives theSO, content of the off-gases decreased to below 100 ppm
After modification of the pilot plant a seriesof active test runs will be performed.
A preliminary test to incinerate radioactiveresins pretreated according to the ATOS process wasmade in the full scale waste incineration facility atSTUDSVIK. 25 kg of spent resin containing 20 mCiCs-137 and 5 mCi Co-60 were mixed with medical radio-active waste and incinerated. Activity emissionsmeasured showed that only 0.6 ppm Cs and 2 ppm Cowere emitted.
Traces of unburned resin were found in the ashwhich stresses the point that the combustion airshould be enriched in oxygen.
3.3 Ash solidificationThe ashes from resin incineration have to be
solidified for transport and disposal. One. of themethods used in Sweden today for the solidification
91
of various wastes including spent resins is cementa-tion. This method is proven and accepted by theauthorities. A main drawback of this method is thewaste volume increase by a factor of at least two.
In order to maintain the volume reductionadvantage achieved by incineration of the resins, theresidue should be solidified with a minimum volumeincrease in the solidification stage. In Figure 4 thechanges of volume due to different treatments areshown. From this figure it is seen that cementationof the incineration residue still provides an overallreduction of volume by a factor of about 5, comparedto the initial volume.
Screening tests made at STUDSVIK show that theresin ashes may be solidified by sintering at tempera-tures about 1100-1200 C. The adding of suitable sin-tering aids may somewhat reduce this temperature.
DIRECT: CEMENTATIONBITUMINIZATION
CEMENTATIONINCINERATION
MELTING
FIG. 4 Volume changes in different treatments ofspent resins
The tests conducted at 1200 C showed a Cs-emission of about 5 % and a volume reduction factorof about two. Lowering the sintering temperature to1100 °C showed that the Cs-emission decreases toabout 0.5 % but the additives required prohibitfurther volume reduction.
92
Tests have also been made to solidify the ash bycementation. The results so far show that a compara-tively good solid can be obtained with an ash contentof about 35 weight %. The compressive strength ofthis solid was 12 MPa and the volume increase factoronly 1.3 in the cementation stage.
There are two problems encountered when cementingthe ash: the metal sulphates and the carbon content.The former can be solved by using sulphate résistentcements or special cements like the US Gypsum cement.The second problem should be solved by an optimizationof the incineration process, with a sufficiently longresidence time under oxidizing conditions.
Work is now in progress at STUDSVIK to optimizesolidification of the incineration residue by cemen-tation. Results are expected during the fall, 1984.
4. CONCLUSIONSA review of the results of the STUDSVIK studies
supports the conclusion that incineration of resinspretreated as in the ATOS process is a feasible meansto achieve a considerable volume reduction of thespent resins, while at the same time minimizingsecondary waste generation and activity emission.
The use of appropriate inorganic sorbents in thepretreatment process ensures a complete combustionand minimized activity emission in the subsequentincineration step, provided that oxidizing combustionconditions are maintained. This means that the off-gas cleaning system can be relatively simple, thusreducing process and operating costs.
la Eluting agentIb Sorbent2 Mixer-dryer3 Buffer tank4 Metering screw5 Furnace6 Ash conveyor7 Ash storage tankB Afterburner9 Scrubber10 HEPA-filter11 Fan
FIG. 5 Layout of an ATOS pilot plant
93
As to secondary waste generation promisingresults have been obtained by the addition of sulphurretainers to the incinerator/calciner. S02 emissionis below 100 ppm, but more work is needed in order tobring this figure further down. As the overall volumereduction capability decreases with efficient sulphurretention a compromise has to be found between volumereduction and sulphur retention.
Pilot plant studies of a complete ATOS facilityhave also been made and a layout of such a facilityis shown in Figure 5. The cost of such a plant hasbeen estimated to about 10 MSEK at a capacity ofabout 10 m /year of spent resins.
REFERENCES/!/ HULTGREN, A., "Treatment of spent ion exchange
resins - practices and developments in Sweden",Studsvik Technical Report NW-83/428 (1983)Contribution to the IAEA Third Research Coordina-tion Meeting, Cairo, Egypt, 28 Feb-4 March, 1983.
/2/ HULTGREN, A., AIROLA, C. , FORBERG, S.,FÄLTH, L., "Preparation of Titanates andZeolites and Their Uses in Radioactive WasteManagement, particularly in the Treatment ofSpent Resins", Studsvik Technical ReportNW-83/475, KBS TR 83-23, May 1983.
/3/ AIROLA, C., HULTGREN, A., "Final treatment ofspent resins with the ATOS process - a study ofprinciples" (in Swedish), Studsvik TechnicalReport NW-82/237 (1982).
/4/ RINGBERG, H., Unpublished results./5/ RINGBERG, H., "Incineration experiments in air
and oxygen with spent resins under variousconditions" (in Swedish), Studsvik TechnicalReport NW-82/312 (1982).
/6/ RINGBERG, H., "Trial incineration of ATOSpre-treated resins in the STUDSVIK incinerationplant" (in Swedish), Studsvik Technical ReportNW-82/383 (1982).
III OLLINEN, M., AIROLA, C., "Solidification experi-ments with ashes from the ATOS process" (inSwedish), Studsvik Technical Report NW-83/524(1983).
/8/ BRODËN, K., NIRSCHL, N., "Results from trialswith incinerated resin ashes solidified inconcrete" (in Swedish), Studsvik TechnicalReport NW-83/539 (1983).
/9/ AIROLA, C., "ATOS-AMOS plant, technical descrip-tion" (in Swedish), Studsvik Technical ReportNW-83/488 (1983).
/10/ HULTGREN, A., "Development of ATOS technique.Status report September 1983" (in Swedish),Studsvik Technical Report NW-83/545 (1983) .
/ll/ LINDGREN, E., "Incineration of resins in theATOS pilot plant. Results from trials performedduring summer 1983" (in Swedish), StudsvikTechnical Report NW-83/547 (1983) .
94
/12/ MICHLINK, D.L., MARCHALL, R.W., TURNER, V.L.,SMITH, K.R., VANDERWALL, E.M., "The Feasibilityof Spent Resins Incineration at Nuclear PowerPlants", Waste Management '83, Tucson, Arizona,1983.
/13/ PATEK, P.R.M., TSYPLENKOV, V.S., HULTGREN, A.,POTTIER, P.E., "Treatment of Spent Ion ExchangeResins". International Conference on RadioactiveWaste Management, Seattle, WA, USA, May 16-20,1983. IAEA-CN-43/455.
95
PRECIPITATED MAGNETITE IN ALPHA EFFLUENT TREATMENT
K. HARDINGAtomic Energy Establishment Winfrith, UKAEA,Dorchester, Dorset,United Kingdom
Abstract
Iron present as a fortuitous component of most nuclearplant effluents can be precipitated as magnetite after suitablechemical treatment. The spinel crystal lattice of the magnetiteis able to incorporate a wide range of foreign ions whichincludes many important radioactive species.
Provided the operating conditions are well chosen theprecipitated magnetite provides an excellent scavenger for manyradionuclides and gives a denser more granular solid product tobe separated from the effluent than the more visual scavengingagent ferric hydroxide.
The conditions under which magnetite can be formed byprecipitation and its properties are presented and compared withthose of ferric hydroxide. Because magnetite has a highmagnetic susceptibility its separation by high gradient magneticseparation becomes a possibility and data relevant to this ispresented.
1. INTRODUCTION
Nuclear plant effluents frequently contain traces of_j.rontypically at concentrations in the region of 10-200 g m inassociation with radionuclide contaminants which vary from plantto plant. In reactors these are., mainly activated circuitcorrosion products Co, Fe, Mn and Zn with onlyminor amounts of fission products and alpha active nuclides. Infuel reprocessing plant they, „are reversed, fission productsparticularly Sr, Cs, Ru predominate and signifi-cant amounts of alpha active nuclides, ie Pu and Am may bepresent.
On pH adjustment these effluents form a precipitate offerric hydroxide which also carries with it many of the traceradioactive species as co-precipitates or by adsorption.Because these incorporation mechanisms are pH dependent, theefficiency of removal for the various active species varieswidely and also for any given species consistent results are noteasy to achieve. However satisfactory removal of this ferricfloe can produce significant decontamination factors (DF) forsuch effluents. Unfortunately ferric floe is notoriouslydifficult to handle on mechanical filters being gelatinous, thinfilms quickly give rise to high pressure drops on conventionalfilters resulting in low capacity and causing problems with backwashing. Because of this effect settling only is often resortedto, the plant then becoming large and expensive and can oftengive rise to variable DFs because of small particle carry over.
97
In an effort to improve the overall efficiency of effluentdecontamination with particular emphasis on reducing costs theapplication of High Gradient Magnetic Separation (HGMS) toferric floe filtration was investigated [1]. This workconcluded that although HGMS would give process improvements itwas not economically viable because the magnet system neededwould be relatively large. What was required was a precipitatewhich had a higher magnetic susceptibility than ferric hydroxideand so reduce the size of the magnet, provided it was at leastas effective as ferric hydroxide at removing activity fromsolution .
The obvious choice was to precipitate magnetite fromsolution instead of ferric hydroxide. That magnetite can beproduced from cold solutions of ferric ferrous iron has beendemonstrated by other workers [2, 3], The results showed thatthe precipitated magnetite was strongly influenced by thepresence of other ions in solution, some like sulphate beingbeneficial other like silica preventing its formation alltogether except at high temperatures.
However, the potential value for effluent clean up ofprecipitating a dense and strongly magnetic solid from an ironcontaining waste stream was considered to be worth detailedinvestigation. This paper reviews the progress made so far withthe identification of the boundaries within which magnetite canbe formed from cold solutions, its properties and potential foreffluent clean up operations.
2. MAGNETITE - METHODS OF PRODUCTION
Magnetite is ferrous ferric oxide, having the chemicalformula FeO.Fe^O , and is the simplest of a large family ofmixed oxides having the same structure of cubic close packingknown as the spinel structure. The generic term for these mixedoxides is the ferrites. Many examples exist in minerals inwhich the divalent ion is nickel, copper, cobalt or zinc and thetrivalent ion is aluminium or chromium. It is rare for theseminerals to be fully stoichiometric, ie all the ferrous iron tobe replaced by nickel or copper and consequently these mineralspresent a bewildering range of compositions and properties.
Many ferrites have interesting and useful electrical andmagnetic properties and are now prepared commercially on a largescale. These include compounds not found in nature.
In general these synthetic ferrites are prepared byprecipitation of mixed metal hydroxides by the addition ofalkali to carefully formulated mixed salt solutions, followed bycareful calcining and finally sintering to produce regular sizedcrystallites with the required mixed cation array.
In our application we require to produce a similar orderedarray of ferrous and ferric ions with relatively smallproportions of both the ferrous and ferric ions being replacedby the radioactive species also present in the effluent stream.Provided the size of these ions is within the acceptable rangefor the spinel structure they will be locked into the crystal
98
lattice and CED only be removed if the precipitate is fullydissolved, unlike the situation with ferric hydroxide where suchcations are often held by adsorption and are released byrelatively small changes in solution composition.
Reports of the successful removal of very large ions suchas cadmium lead and mercury by incorporation into ferriteprecipitates [4, 5] indicates that a very wide range ofradioactive species could well be held in this form. Morerecently work at Rocky Flats [6] has demonstrated that plutoniumand americium ferrites can be prepared under appropriateconditions.
There are three methods whereby a solution can be made toyield a mixed oxide precipitate which is capable of conversionto a ferrite structure in which the predominant material isiron.
(a) Iron in the ferrous state and ferrous hydroxide isprecipitated followed by partial oxidation usuallywith air.
(b) With iron in either the ferric or ferrous state anappropriate amount of the counter valance state ironis added to give the necessary stoichiometry formagnetite production, followed by pH adjustment.
(c) With iron in the ferric state partial reduction iscarried out followed by pH adjustment to yield a mixedoxide precipitate.
A survey of the effluents to which this process could beapplicable showed that the iron was predominantly present in theferric state and therefore method (b) or (c) would beapplicable. Method (b) involves the addition of further iron tothe system thereby increasing the volume of solid waste to behandled and since this is highly undesirable, this method hasnot been investigated. Our work is being concentrated on method(c), ie the reduction of part of the iron present to yield aferrous to ferric ratio of 0.5 ±0.1.
Reduction of ferric ions to ferrous can be accomplished bya number of reagents and also electrolytically. However, anyprocess designed to treat effluents needs to be both simple andeconomical and therefore only sulphite and hydrazine have beenstudied in detail so far in the programme. However dithionite(sodium hydrosulphite) is an interesting reagent suggested byStreeter et al [7] because in addition to its use for reductionin solution it can under some conditions convert ferrichydroxide to magnetite direct in a process analogous to theoxidation of ferrous hydroxide.3. EXPERIMENTAL
3.1 Effects of Iron Concentration and Dissolved Oxygen
A reducing agent, .suitable for this process should be strongenough to reduce Fe to Fe at an acceptable rate for anefficient process, but not strong enough to react with other
99
oxidising agents particularly nitrate which is always likely tobe present in typical effluents. The presence of dissolvedoxygen is a complicating factor in any likely process sincethere are no known reagents which can reduce Fe to Fein the presence of oxygen and therefore it must be removed by aprior deoxygenation operation, ie nitrogen sparging or reducedalong with part of the iron.
-3At high iron concentrations, ie above 350 g m theamount of sulphite needed to react with the dissolved oxygen isrelatively small, approximately one third of the total needed,and does not seriously interfere with the control of the ironreduction reaction. Jftowever, as the iron concentration isreduced below 100 g m the reduction of oxygen becomes themajor sulphite consumer and control of the ferrous reductionreaction becomes more difficult. Also as the concentration ofiron in solution is reduced the reaction with sulphite becomesslower .and consequently the position where the optimumFe /Fe ratio is reached is more difficult to determine.Clearly some lower concentration of ferric iron will be reachedwhere satisfactory formation of magnetite is not practicable.To examine this aspect the formation of magnetite has beenexamined over a range of concentrations in the presence andabsence of oxygen all at ambient temperature (20-25°C).
-3A stock solution containing 1000 g m iron as ferricnitrate with excess nitric acid to give a pH of 1.5-2.0 was usedto prepare by dilution a series of test solutions containing arange of iron concentrations. The solutions were either purgedwith nitrogen to remove dissolved oxygen followed by theaddition of sufficient sodium sulphite solution to reduce onethird of the iron to the ferrous state, or sufficient sodiumsulphite was added to react with the dissolved oxygen and alsoreduce one third of the iron. After stirring to ensure thereduction reaction was substantially complete the pH wasadjusted to 10.0. On standing the properties of the precipitateformed were assessed visually and with a magnet.
Tables 1 and 2 show typical results from these experiments.
Because of the small scale of the experiments, problemswere encountered with the exclusion of atmospheric oxygen andsome variation in results was obtained. The data presentedrepresents average resuts. From these results it can be seenthat in the absence of oxygen a magnetic, low volume_»roduct isproduced at iron concentrations as low as 30 g m but theprocess is impractical below that. The presence of oxygeninhibits magnetite production; a good quality product only beingformed at iron concentrations of 100 g m or above.3.2 Effect of Ferrous/Ferric Ratio on Magnetite Formation
The essential step in the reaction to .form magnetite is thedevelopment of the optimum ratio of Fe to Fe ions inthe solution prior to precipitation. From the composition ofmagnetite FeO.Fe„0 it is clear that this ratio should beclose to 0.5. However precise control of this ratio on a plantscale will not be simple and therefore if this process is to beconsidered for industrial applications it will be necessary to
100
Table 1Formation of Magnetite from Ferric Nitrate
Solution Free from Dissolved Oxygen
IronConcentration
ppm Fe3"1"150
100
80
70
50
30
15
Mixing Time (mins)before alkaliaddition
5
7
10
10
10
15
15
FinalpH
10.8
10.4
10.5
10.6
10.2
10.2
10.2
Observations
Immediate heavy ppt - denseblack/brown - magnetic
Immediate granular, colour- black - magneticImmediate granular, colour- black - magneticImmediate opt, not as gran-ular as before - magneticInstant fine - black preci-nitate - magneticFine ppt, black/brown -formed after 1 minute -magneticInitial dark green solutionformed a very fine suspensionafter 1 minute, which did notstart to settle until after3 minutes, black/brown -less strongly magnetic
Table 2Formation of Magnetite from Ferric Nitrate Solution
With the Presence of Dissolved Oxygen
IronConcentrationppm Fe3+
150
100
80
70
50
30
2% Na2So2Addedcm3
10
8
6
5
4
3
Mixing*Time +(mins)5
7
10
15
15
15
FinalpH
10.6
10.5
10.7
10.8
11.2
11.2
Observations
Dense brown/black pptformed immediately -magneticBlack granular precipi-tate - magneticDark brown floe quick toform - weakly magnetic
Brown floe, becomingdenser on standing - weak-ly magneticLight brown floe faintlymagneticYellow/brown floe, non-magnetic
»Mixing time considered just sufficient for reaction to becomplete.+Before alkali addition
101
demonstrate that some degree of freedom can be allowed in theadjustment of the Fe /Fe ratio and still produce asatisfactory product.
Table 3 shows the results of a series of experiments todetermine the effect of varying this ratio on the production ofa magnetic produce over a range of iron concentrations. Allsolutions were purged free from oxygen.
Table 3Magnetite Formation from Cold .Aoueous Solutions
Ferrous/ FerricRatio
0.2
0.3
0.4
0.5
0.6
0.7
0.8
1.0
Total Iron Concentration50
BrownKM
Dark BrownWM
DenseBlackSM
DenseBlackS M
ff
BrownWM
GreenKMGreenKM
100BrownKM
Dark BrownWMDenseBlackSM
DenseBlackSM
ii
Dark BrownS M
BrownWMGreenKM
150BrownKM
Dark BrownV;MDenseBlackS M
DenseBlackS Mn
DenseBlackSM
Green/BlackWMGreenNM
300BrownNM
Dark BrownV;MDenseBlackSM
DenseBlackSM
it
DenseBlackSM
Green/BlackV;MGreenNM
600BrownKM
Dark Brov. nWH
DenseBlackSM
DenseBlackSM"
DenseBlackSM
Dark BrownSM
GreenN»'
Kote: KM - Kon-magneticWM - Weakly magneticSM - Strongly magnetic
At low ferrous concentrations the formation of a magneticprecipitate shows no iron concentration dependence. However atFe /Fe ratios beyond 0.6 the formation of a magneticproduct is dependent on_ iron concentration. At high ironconcentrations (600 g m ) even at ratios as high as 0.8 amagnetic product is readily produced.
These results were encouraging for the development of apossible industrial scale process and the scope of these testswere widened to study possible plant control systems.
In a solution containing both oxidised and reduced forms ofan ionic species the potential taken up by an inert electrodehas a constant value under standard conditions of species
102
concentration known as the redox potential of the system. Thiseffect can easily be used to determine when an oxidisation orreduction reaction has been taken to completion because a sharpchange in the solution potential is obtained as the lastfraction of one species is converted to the other. Thesituation is not as clearly defined when one requires to take areduction reaction to an intermediate position.
Experiments were carried out in which the redox potentialof a deoxygenated ferric nitrate solution was followed as it wastitrated with sodium sulphite solution; samples of solutionbeing withdrawn periodically and the Fe /Fe ratio beingdetermined colorimetrically. Figure 1 shows the changes inred_o_x potential against sulphite addition, the region where theFe /Fe ratio was between 0.4 and 0.6 (ie magnetite isproduced) is seen to lie between 400 and 350 mV. This graphclearly shows how the control of conditions needed for magnetiteproduction is easier at higher iron concentrations. Anotherfeature of interest shown in this graph is that the amount ofsulphite required to reduce one third of the iron is greaterthan the theoretical amount.
ooo
5OO
4OO
3OO
2°°
ICO
l l \
I I I .11 Iron I Concenlrotion (ppm)
Ferrous / Ferne Ratio betweenI i O 4 and 0-6
6 8 1C 12 M lu
Sodium SulphiteFig 1 :-Change of Solution Potential as
Ferrous / Ferric Ratio Changes.
18 20 22
It should also be noted that the presence of nitrate ionchanges the position of the solution potential band within whichmagnetite is formed but not the overall width or general shapeof the titration curves.
3.3 Effect of some Interfering Ions on Magnetite Precipitation
The most comprehensive work on the precipitation ofmagnetite from ambient temperature solutions has been carriedout for the treatment of mineral waters [2]. More recentlyreference [8] has also addressed this problem.
103
Table 4
Effect of Magnesium on Magnetite Format i on
StirringTimeMins
10
10
10
10
10
77
7
7
'
7
7
7
7
Fe(III)Coneppm
15
15
15
15
153030
30
30
30
30
30
30
Mg(II)Coneppm
5
10
15
20
25510
20
25
30
35
40
50
Mg/FeRatio
0.33
0.66
1
.33
1.67
0.1670.33
0.66
0.833
1
1.167
1.33
1.667
2% SodiumSulphateAddedcm 30.75
0.75
0.75
0.75
0.75
1.51.5
1.5
1 .5
1 .5
1.5
1.5
1.5
InitialpH
2.1
2.2
2.2
2.2
2. 2
1.91.9
1.9
1.9
1.8
1.9
2.0
2.2
FinalpH
10.7
10.5
10.7
M
1 110.710.7
11.0
10.6
10.6
10.8
1 1
1 1
Observations
Very fine black ppt - 5 minsto settle around magnet -magnetic15 minutes for any settling tooccur around magnet, mostsettled within 30 minutes.
Fine suspension visibleafter 15 mins. 1 hr tosettle ccmpletelv. Formedovernight - weakly magnetic
Fine suspension visibleafter 20 mins. Yellow solu-tion after 30 mins. Lightbrown floe after 1} hrs.Non-macneticNon-magnetic floe after 15 minsImmediate formation ofblack ppt - macnetic
Fine suspension at first, pre-cipitation on magnet after4 minutes - magnetic.Floe-like suspension,settled as magnetite in 20i.iins - magneticVery fine immediate suspension.Fine ppt around magnet after20 mins. Full settling after40 minutes - weakly magnetic.
Greenish-brown ppt after 20min - weaklv magnetic. Darkbrown magnetic ppt after 30min. Floe-like mixture ofmapnetic/non-magnetic pptformed overnightGreen floe-like ppt formedin <. I hr. Formed a floe-like mixture, as above.overnight .
Fine floe suspension -formed overnight, non-magnetic
Fine green floe, formingbrown floe. Non-magnetic
These workers report results from systems where the ironconcentrations and those of the interfering ions are much higherthan is likely to be the case with typical nuclear industryeffluents. However, the main ions which suppress magnetiteformation are identified as magnesium, aluminium and solublesilicate whereas sulphate enhances its production.
104
Table 5
Effect of Aluminium on Magnetite Formation
StirringTimeMin s7
77
7
7
7
5
5
5
5
Fe(III)Coneppm
15
15
15
15
15
15
30
30
30
30
AlCIt)Coneppm0
12
3
5
10
2
3
5
10
Al/FeRatio
0
0.066
0.133
0.2
0.33
0.66
0.066
0.1
0.167
0.33
InitialpH
2.2
2.2
2.2
2.2
2.2
2.3
1.9
1.9
1.9
1.9
FinalpH
10.9
10.510.8
10.8
10.95
10.9
10.9
10.7
10.4
10.5
Observations
Fine ppt of magnetite immed-iately - solution test.Instant ppt of magnetite.Fine suspension showing after3 mins - settled in 10 mins.Magnetic ppt settled in 30mins - brown/black
Fine suspension slo to form,settled as brown floe - notmagnetic
Floe-like brown ppt, notmagneticFine black ppt, settledafter 3 mins - magnetic
Fine suspension visible in5 mins. Settled as faintppt after 30 mins - magneticTine suspension visible in 5t?ins . Settled as faint pptin 20 mins - became denserafter 2 hr - magneticFine suspension slow toform, settled as dark brownfloe, non-magnetic
To assess the effects of these ions on the formation ofmagnetite at relatively low iron concentrations and where theferrous ions are produced by in situ reduction a series of testshave been carried out in the presence of varying concentrationsof magnesium, aluminium and soluble silicate. Again allsolutions were deoxygenated by nitrogen purging prior to thetest.
Tables 4, 5 and 6 present the results obtained from whichthe following conclusions can be drawn.
(a) Magnesium interferes least strongly, with solutionscontaining Mg/Fe ratios up to one able to form amagnetic ppt even at low iron concentrations. As theiron concentration increases the proportion ofmagnesium which can be tolerated also increases.
(b) Aluminium interference is more serious than that ofmagnesium. At low iron concentrations magnetiteformation starts to be suppressed at Al/Fe ratios of0.2 but, as with magnesium, as the iron concentrationincreases a higher Al/Fe ratio can be tolerated.
105
Table 6Effect of Silicate on Magnetite Formation
StirringTimeMins
5
5
5
5
5
7
7
7
_-
Fe(III)Coneppm
30
30
30
30
30
15
15
15
15
15
SilicateConeppm
1
2
3
5
10
0
1
2
5
10
Si/FeRatio
0.033
0.066
0.1
0.167
0.33
0
0.66
0.133
0.33
InitialPH
1.9
1.9
1.9
1.9
1-9
2.1
2.0
2.0
2.0
0.66 2.1
FinalPH
1 1
10.8
10.8
1 1
10.3
10.5
10.1
10.2
10.4
10.7
Observations
Instant black ppt of magne-tite - strongly magneticParamagnetic green floeinstantly formed - darkenedslowly - magnetic
Creen floe - darkenedslowly - became magneticafter 20 minsDark brown floe - notmagnetic even on standingLight brown floe - notmagnetic on standing.
Fine black ppt immediately- solution blank test
Faint black ppt after 10mins - magneticFine suspension coes notsettle after 30 mins.Non-magnetic
. Yellow solution - ferricfloe ppt very slow tosettle. Non-macnetic
Yellow solution, ferric floeslow to settle as before.
(c) Soluble silicate prevents the formation of magnetiteat mole ratios above 0.1. However, unlike aluminium,increasing the iron concentration does not lead to agreater toleration for silicate.
3.4 Activity Uptake by MagnetiteFollowing the identification of a wide region of solution
conditions where magnetite can be precipitated from solution thecapability of this material to remove trace .activity fromsolution was compared with that of ferric floe under similarconditions.
-3A stock solution of 300 g m iron as ferric nitratecontaining traces of typical radionuclide contaminants was used.Its composition is given in Table 7.
Aliquots of this stock solution were taken and eithertreated directly with ammonia to precipitate ferric floe (pH10.5) or dissolved oxygen was removed by purging with nitrogenfollowed by sulphite addition to produce a Fe /Fe ratio
106
Table 7
Analvsis of Active Stock Solution
Solution Component
Iron mg/litre
Magnesium "Plutonium (238, 240, 241) pCi/litreAmericium 241 "
Cobalt 60
Caesium 137 "
Cerium 144
Europium 154 "Ruthenium 106 "
Strontium 90 "
Concentration
300
10151.42.1
2.91 .4
8.04.2
69
Table 8
Comparison of Radio Kuclide Uptake by Ferrie Floe and PrecipitatedMagnetite
RadioIsotope
PlutoniumAmericiumCo 60Cs 137Ce 144Eu 154Ru 106Sr 90
DF
Ferrie Floc10015010NIL5020102
15025015NIL402062
20010010NIL1215101.5
15015015NIL201081.8
Magnetite
800500100NIL501005010
15080080NIL908010020
1000700100NIL1001004015
1000600150SIL801006010
of 0.5 before making alkaline with ammonia to pH 10.5. Afterallowing both precipitates to settle overnight the supernatentliquid was decanted off, filtered through a 0.45 micronmillipore filter and this clear liquor was analysed forradionuclides. The results from A pairs of experiments aregiven in Table 8.
These results confirm the expectation that precipitatedmagnetite should be a more effective getter for trace activespecies than ferric floe. For alpha active species thisimprovement is only marginal since ferric floe is an effectivegetter for the transuratiic elements. The main advantage seen tobe obtained from the use of magnetite is the more consistent DFfor alpha coupled witVu,a signifiant improvement in performancefor the beta emitters Sr and Ru.
107
4. PROPERTIES OF PRECIPITATED MAGNETITE COMPARED WITH FERRICFLOC
4.1 Settling Rates and Densities
Although magnetite has been shown to give improved DFs fora number of radionuclides it is its high magnetic susceptibilitywith the attendant possibility of efficient solids separationusing HGMS which originally prompted this work. In much of thescoping tests directed at finding the boundaries within whichmagnetite could be formed the fact that the precipitate wasblack and attracted by a magnet was taken as sufficient evidencethat magnetite was being produced.
Having established the operational conditions necessary toproduce a "magnetite" precipitate, experimental work wasdirected towards the characterisation of this material and thecomparison of its separation properties with those of ferricfloe precipitated from the same solution comp_osition. Thecomposition of the solution used was 300 g m of iron asferric nitrate with nitric acid to pH 1.5. Treatment used wasas before. The settling rates of the two precipitates weremeasured and are plotted in Figure 2. The magnetite settlessignificantly faster than the ferric floe despite the fact thatthe magnetite particles are a good deal smaller than those ofthe ferric floe (see Figures 3 and 4). Scanning electron photomicrographs (SEM) pictures of the two types of particlesindicate a possible explanation for this behaviour, the ferricfloe particles having a sponge-like appearance whereas themagnetite particles were smooth but "stringy".
T> 1V I
10 20Time
30 40minutes
50 60
:- SETTLING CURVES FOR THE FIRST HOUR
108
18
16
14o^ 12£
10
_ 8o
I 6
2 5Particle S i z e
10 2O 3O 4O SO
Fig 3:-PARTICLE SIZE DISTRIBUTION FOR MAGNETITE.
IB
16
14
12
2 10
I 8
S 6
£ 4û
2 5Particle S i ze -
IO 2O 3O 4O 5O
Fig 4:- PARTICLE SIZE DISTRIBUTION FOR FERRIC FLOC.
The settled densities of the two precipitates varied withtime. At 30 minutes the settled magnetite had five times thedensity of the ferric floe, at 20 hours it was only twice asdense and by 200 hours it was only 20% greater at 8.5% totalsolids. This relatively low final settled density isdisappointing and clearly more work is needed to develop theconditions for the production of a more crystalline form ofmagnetite.
4.2 Magnetic Susceptibility
As indicated in the scoping test results given in Table 4the formation of magnetite or at least the development of strongmagnetic properties is not instantaneous on precipitation.
109
cgs Unitsper mole
0 10 20 30Time a*ler Precipitation ("inutes)
40
Magnetic Susceptibility of Magnetite Directly after PrecipitationFigure 5l
Figure 5 shows the change of magnetic susceptibility of twomagnetite precipitates with time. Both develop strong magneticproperties of the same value as mineral magnetite at 2.6 and 3.0CGS units per mole (ground mineral magnetite had a magneticsusceptibility of 2.7 CGS units per mole). The presence ofsulphate ion retards the development of magnetic propertiesinitially but results in higher values being developedultimately. This material was also more crystalline as shown inelectron micrographs of the particles. These values can becompared with a value of 0.01 CGS units per mole for ferricfloe.5. CONCLUSIONS AND POTENTIAL APPLICATIONS
5.1 Precipitation of MagnetiteIt has been demonstrated that a magnetite-like solid can be
produced from effluents containing iron on suitable treatment ofthe effluent and that this solid:
(a) has high collecting power for radionuclides(b) settles rapidly but its ultimate settled density
requires to be increased(c) has a high magnetic susceptibility.
110
It is not possible to precipitate magnetite directly from allsolution compositions, in that there are a number of common ionswhich suppress its formation direct from ambient temperaturesolution. However, the addition of more iron to that naturallypresent will in the great majority of such cases enable theinterference to be overcome.
The use of the cheap reducing agent sulphite to produce thenecessary conditions for magnetite precipitation without addingto the bulk of waste produced when sufficient iron is naturallypresent in the effluent has distinct advantages for subsequentwaste disposal [9],
5.2 Applications(a) Gravity Settling - the increased density of magnetite gives
rise to faster settling of this material than with ferricfloe enabling smaller settlers to be used for a giventhroughput. However, the smaller particle size gives riseto some problems with carryover.
(b) HGMS - the very high magnetic susceptibility of thismaterial provides a potential application for solidsseparation using a magnetic filter directly in place of asettler. Alternatively, a gravity settler to handle thebulk of the solids followed by an HGMS unit to polish thelast traces of solids appears to be attractive if maximumDFs are needed.
(c) Magnetic Assisted Gravity Settling - again because of thestrong magnetic properties of the precipitate a relativelyweak magnetic gradient which will both flocculate andattract the magnetite particles has been shown to producehigh rates of settling and higher settled solids densitiesof up to 14%.
REFERENCES
[1] HARDING, K., BAXTER, W., "Application of HGMS to FerricHydroxide Filtration" paper 14-4, Intermag, Grenoble(1981).
[2] "US Environmental Protection Agency Water Quality OfficeStudies on Densification of Coal Mine Drainage Sludge",Report 14618 MJT (August 1971).
[3] OKUDA, T., SUGANO, I., SUJI, T. T., "Removal of HeavyMetals from Waste Water by Ferrite Co-precipitationTechnique", Filtration and Separation Journal (January/February 1976).
[4] TAKADA, T., "Removal of Heavy Metal Ions from WasteWater by Ferritisation", Kago to Taisaki 13, 37 (1977),RFP translation 281.
[5] IGUCHI, T., KAMURA, T., INOUE, M., "Ferrite Process forTreatment of Waste Water Containing Hevy Metals", PPM 10,49 (1979), RFP translation 284.
I l l
[6] BOYD, T. E., KOCHEN, R. L., PRICE, M. Y., "FerriteTreatment of Actinide Waste Solutions", RFP 3299 (July1982).
[7] STREETER, R. C., McLEAN, D. C., LOVELL, H. L., "Reductionof Hydrous Ferric Oxide to a Magnetic Form with SodiumDithinite; Implications for Coal Mining DrainageTreatment", Am. Chem. Soc. (September 1971).
[8] VAINSHTEIN, V. I., et al., "Influence of Salt Compositionon the Kinetics of Magnetite Formation, Its MagneticSusceptibility and Mechanical Dehydration", ZhumalPrikladnoi Khinii, f>5_ 1 (1982) 133-138.
[9] HARDING, K., WILLIAMS, R., Provisional UK Patent 2091135 "A '(1981)
112
THE APPLICATION OF INORGANIC ION EXCHANGERSTO THE TREATMENT OF ALPHA-BEARING WASTE STREAMS
E.W. HOOPERChemical Technology Division,Atomic Energy Research Establishment, UKAEA,Harwell, United Kingdom
Abstract
The Department of the Environment IB funding a generic programme of workat AERE, Harwell to examine the application of Inorganic ion exchangers tothe removal of radionuclides from aqueous waste streams.
A literature survey has identified six exchangers of potential use, theyare:- Titanium Phosphate, Zirconium Phosphate, Manganese Dioxide,Polyantimonic Acid, Hydrous Titanium Oxide and Copper Hexacyanoferrate.
A programme of batch contact experiments using spiked solutions andsolutions of Irradiated fuel has examined the extent and rate of absorptionof radlonuclides from solutions 1 Molar in nitrate ion and with hydrogen ionconcentrations between 1 Molar and pH *• 10; 2hi and 4M nitric acid solutionshave also been examined.
This paper presents the data obtained so far on the absorption ofplutonium, americium and neptunium from spiked solutions and of totala-activity removal from solutions of irradiated fuel.
1. Introduction
In 1982 the UK Department of the Environment initiated atAERE Harwell a programme of work with the objective of developingprocesses which would provide efficient and safe concentration ofthe radioactive constituents of waste streams into solid materialswhich can be consolidated into forms suitable for storage ordisposal.
A search of the literature showed that a considerable volumeof information is available on the use of inorganic absorbers forthe decontamination of waste streams but the reported data areoften of limited use in selecting an absorber for a particular
113
waste stream since the candidate material has generally beenexamined only for a few or individual nuclides over a small rangeof operating conditions. Also the information available ona-decontamination is considerably less than that for ß-y removal.
An assessment exercise"- -I , based on the literature
search, identified six ion exchangers with a known affinity foractinides and/or fission products and an experimental programme ofwork is now in progress to obtain a comprehensive knowledge of thecapabilities of these exchangers. The absorbers selected for thisstudy are:- Titanium Phosphate, Zirconium Phosphate, ManganeseDioxide, Polyantimonic Acid, Hydrous Titanium Oxide and SodiumCopper Hexacyanoferrate.
This paper presents the data obtained so far on theabsorption of plutonium, americiuc and neptunium f rom spiked
solutions and of total a-activity removal from solutions ofirradiated fuel.
2. Experimental
2.1 Absorbers
Hydrous Titanium Oxide (HTiO) was prepared by precipitationfrom ti tanium tetrachloride solution by addition of sodium
hydroxide solution. The precipitate was washed and then dried to a
glassy gel.
Manganese Dioxide (MANOX A,Mn02), Titanium Phosphate (PHOTID.TiP), Zirconium Phosphate (PHOZIR A, ZrP), Polyantimonic Acid(POLYAN S, SbA) and Copper Ferrocyanide (FeCu) were obtained fromApplied Research SPRL, Belgium.
Polyantimonic Acid and Titanium Phosphate absorbers have alsobeen prepared at Harwell.
All absorbers were sieved to provide material with theparticle size range 212-425nm.
114
2.2 Waste Simulates
Simulates of salt-free acid wastes were prepared by spiking1-AM nitric acid solutions with radionuclides. Simulates of saltbearing wastes were prepared from IM sodium nitrate solution.
Solutions of Am-241 in 3M HN03, Pu-239 in AM HN03, Np-237 in AMHN03 and irradiated fuel in HN03 ( ~ 3M) were used as sources ofradionuclides.
The pH of the waste simulate was adjusted to the requiredvalue by addition of HN03 or NaOH.
2. 3 Conditioning of Absorbers
Conditioning is necessary in order to maintain a constantsolution pH during the course of a kinetic run. Absorbers werewashed in order to remove adherent fines then contacted with waterfor several days at a pH maintained at the required value by theaddition of either HN03 or NaOH solution from an autotitrator. Theconditioned absorbers were washed with distilled water and, with theexception of HTiO, allowed to air dry for several days. It wasobserved that HTiO undergoes dissolution at a solution pH < 1.0, andthat Mn02 had to be conditioned in the presence of NaN03 since, insolutions of low ionic strength, it tended to break down.
2. A Kinetics of Absorption
The rate of absorption of radio-ions was followed by batchcontact experiments. 0.25 to Ig of conditioned absorber was addedto 50 mL of waste simulate which had been adjusted to the requiredpH. The flasks were then shaken at a fixed rate for 2A hours,samples being removed periodically for analysis.
The percentage removal of an isotope as a function of timewas calculated from the following equation:
A A% removal - (— 2~—— ) 100
n.o
115
where,
Ao « initial activity in (unit volume of) solutionAt * activity at time t in (unit volume of) solution
2. 5 Distribution Coefficient Measurements
A 50 mL aliquot of waste simulate solution was added to anaccurately weighed amount of ion-exchanger. The solution was thenallowed to stand at 25°C, with intermittent shaking, for periods ofup to two weeks.
Distribution coefficients (KD) were calculated from thefollowing:
D _ 2-..»-- V ,_, _-lA wœ
where,
V = total solution volume (mL)w = weight of exchanger (g)
3= Results
3. 1 Plutonium Absorption
The absorption kinetics of Pu(IV) (28 uM) on a variety ofexchangers under different solution conditions are shown in figures1 to 4 inclusive. Initial rates were fairly rapid but equilibriumwas only achieved after several days contact, and possibly particlediffusion may be rate controlling. In going from pH 1.0 to 2.0,the hydrolysis of Pu(IV) produces species with less positiver 2lcharge1- J , which could account for the observed decrease inuptake kinetics (figures 1 and 2). There should also be a decreaseir. competition for exchange sites by H"1" ions, which, in the case ofMnU2, appears to more than compensate for any change in charge of thePu species present in solution.
116
FigurelKinetics of Pu (iv) removal by inorganic ionexchange.
Figure 2
Sb»100
90
BO
% 70
i removal60
50
40
30
20
10
3 4 - 5Time (hrs)
Kinetics of Pu (iv) removal by inorganic ionexchangers.
TOM No NO 3
pH 7-025'C
3 4 5Time (tirs)
Figure 3Kinetics of Pu(iv) uptake by SbA
Figure^
Veremoval
100
90
60
70a!60
.50
40
30
20
10
3 4 5Time (hrs)
24
Kinetics of Pu (iv) removal by TiP
solutions »pH 1C contain VOM NaNDj25'C
pH l C
24
117
Figure 5" pH protile for Pu (ivj removal by SbA25 pM Pu;25°C
tolutuns ï pH 10 contain 1 OH
10
10'
Figure 6 pH profile for Pu dv) removal by inorganic ionexchangers. 25uM Pu;25°C
: solutions ïpH 10 contain 1 DM NaNO;
Kd
10'
10'5 4 3 2 1
!HN03]M
-1
MnOj
pHPH
KD values for Pu(IV) are listed in Table I. Plots of logKD vs pH are shown in figures 5 and 6. Over the range of nitricacid concentrations from 4M to 1M_ the graph is expected to be alinear function of pH with a slope equal to the charge on thespecies being absorbed. The limited data in Fig. 6 suggest thatPu(IV) may be absorbed as Pu*«+, [Pu(N03)]3+ and [Pu(NO)3 )2J2+ fromacid media by different absorbers. With the exception of ZrP, Na~*"ions have little effect on the uptake of Pu. Above pH l colloidalPu interferes with absorption, both ZrP and SbA producing a maximum inthe Kn value at pH ~ 1. This inflection point can be calculatedr 2lfrom the expression for the solubility product"- •> :
Ksp - [PU*+] I
sp 7 x 10-" , [Pui++] - 2.8 x ID"6 M
hence, the pH at which the [OH~] is sufficient to reach Kgp isapproximately 1.35.
118
TABLE Ivalues for Pu(IV)
Acidity/pH
U M
2 hl
1 M^
pH 1
pH 2
pH 7
pH 10
pH 12
ABSORBERS
CSbA
A. 660 x 101
1.220 x 103
1.340 x 101*
5.160 x 10**
4.740 x lO11
5.130 x 102
2.370 x 103
7.890 x 103
ZrP
5.720 x 10°
3.B30 x 101
2.160 x 102
1.150 x 10*
3.420 x 103
n. m.
n. n.
n. m.
Mn02
1.610 x 10°
1.000 x 10°
3.100 x 10°
1.300 x 10 *
1.400 x 102
n. m.
n. m.
n. m.
TIP
2.740 x 102
7.720 x 102
2.890 x 103
8.970 x 102
5.620 x lO1*
n .m.
n. m.
n. m.
HTiO
n .m.
n .m.
n .ID,
1.020 x 103
1.160 x 101*
n. m.
n. m.
n. n.
All Solutions with pH > 1.0 contained l.0 M KaNOn.m. • not measured25°C
The increase in K^ with increasing pH observed for TiP,HTiO and Mn02 can be interpreted as being due to the decrease incompetition by H+ which out-weighs the effects of colloidformation.
3. 2 Amerlcium Absorption
The exchange kinetics for Am(III) are shown in figures 7 to 9inclusive. Exchange is fairly rapid for solution with pH > 1 andshows a dependence on solution pH. These data can be interpretedin terms of the formation of hydrolysis products of Am- •'J.Decrease in competition by IT*" for exchange sites account for rateincreases whereas rate decreases are attributed to the formation ofhydrolysis products of Am with reduced or zero charge.
KD values for Am(III) are listed in Table II. Thevariation in Iog10 KD for Am(III) with solution pH is shown infigures 10 and 11. The general trend is for KD to increase withsolution pH, due to reduced competition by H"*". There is little
119
Figure 7Kinetics of Am removal by POLYAN S
Figure 8
removal60
Kinetics of Am [in] removal by inorganic ionexchangers.
Time (hrs)Time (hrs)
Figure 9Kinetics of Am (in) removal by inorganic ionexchangers
VOM NoHOj
pH 30
25« C
3 4 5Time (hrs)
120
TABLE IIvalues for
Acidi ty /pH
4 M
2 M
1 M
pH l
pH 3
pH 8.8
pH 10
pH 12
ABSORBERS
Polyan S
2.320 x 101
1.370 x 102
2.800 x 103
5.590 x 103
2.720 x 105
9.280 x 10"1
1.430 x 105
1.330 x 1011
ZrP
8.930 x 10°
2.170 x 101
2.710 x 101
2.360 x 101
1.910 x 103
n.m.
n.m.
n.m.
Mn02
1.180 x 101
1.150 x 101
7.900 x 10°
4.400 x 101
6.130 x 101*
n.m.
n.n.
n.m.
TIP
1.010 x 101
1.780 x 10" l
1.060 x 101
7.900 x 101
2.230 x 101*
n.m.
n.m.
n.m.
HT10
n.m.
n.m.
n.m.
2.150 x 101
1.470 x 105
n.m.
n.m.
n.m.
Ail solutions with pH > 1.0 contained 1.0 M NaNO3n.m. » not measured25°C
Kigure 10 pH protiie tor Am im) removal oy inorganic ion6 exchangers.0.76yM Am;25°C
Figure 11 . pH profile for Am (in) removal by POLYAN S0.76 y M Am; 25°C and HTiO
10
Kd
10'
* pH VO tontan 1 0£ SoNO
5 1 3 2 1[HN03]M
iHT.OlH)
C»
I10 12
121
interference from competing Na+ ions for solution pH up to 3-4 butfigure 11 shows a decrease in Am absorption above pH 6 which may bedue to Am hydrolysis or increased Na+ absorption.
The slope of the graph for SbA over the HN03 concentrationrange 4M to IM (figure 10) is about 3.2, which corresponds to theuptake of a tripositive Am species.
3. 3 Neptunium Absorption
Np(V) is the most stable oxidation state of neptunium and itis usually present in solution at NpO?"*~ but disproportionates at
[ •\ *~
—.„.. __ _.._.._-___„„ J.
Np"+ 24-
The exchange kinetics of Np(V) (0.42 mM in IM NaN03) on variousexchangers are shown in figures 12 and 13. Rates increase withincrease in pH, which is interpreted as being due to decreasingcompetition from Ir. In general, uptake is slow, presumably due tothe absorbed species having a relatively large radius and bearing onlya single positive charge; particle diffusion is probably ratedetermining.
Figure 12
100
90
80
70
60rval50
40
30
20
10
% ,removal
Kinetics of Np(v) removal by inorganic ionexchangers.
1 OH No NO,pH 1-0
Figure 13
100
90
80
70
» 60
removal50
30
20
10
Kinetics of Np (v) removal by inorganic ionexchangers.
10M NoNOj
pH 20
48Time
122
Measured KD values for Np(V) in acid and 1.0 Msolutions are listed in Table III. The variation ofNp(V) with solution pH is shown in figure 14. It should be notedthat solutions with pH > 1 are 1.0 M_ in NaN03 . Above pH - 0 thereis a general increase in Kp with increase in solution pHbecause of decreasing competition for exchange sites by H+. BelowpH = 0, the disproportionation of Np02+ to species with highercharge produces an increase in Kp. The net effect is a minimain the K^-pH curve.
TABLE III
values for Np(V) as -.
Acidi ty/pH
4 £
2 M
1 hl
pH 1
pH 2
ABSORBERS
CSbÂ
4.380 x 101
1.980 x 101
2.920 x 101
1.540 x 102
1.800 x 102
ZrP
1.530 x 101
1.030 x 101
1.170 x 101
8.450 x 10°
5.065 x 101
Mn02
1.110 x 101
6.850 x 10°
7.800 x 10" 1
1.130 x 101
1.740 x 101
TiP
7.060 x 10°
4.700 x 10°
~ 0
5.470 x 101
5.310 x 102
HTiO
n. m.
n. n.
n.ni«
5.940 x 10°
1.980 x 103
All solutions with pH > 1.0 contained 1.0 M NaNOn. m. « not measured25'C
Figure 14 -pH profile for Np (v) removal by inorganic ionexchangers
tolutioni »pH 1 0 contain 1 OH NoNOj25'C
123
3» 4 Total g Absorption
Figure 15 illustrates the extent of a-activity removal byd i f f e r e n t absorbers from solutions of irradiated fuel dissolved in
nitric acid and then adjusted to various pH values. Theexperimental data on which these graphs are based is listed in
Table IV.
Figure 1S
Antimomc Acid
Tofaloc removal
Hydrous Titanium Oxidi
Zirconium Phosphate Sodium Copptr Huacyonofcrrotcll l )
Manganese Dioxide
Titanium Phosphate
4 2 0 - 1 Z 4 6 8 1 0 4 2 C H 2 4 6 6 1 0pH (HNOjlM pH
4 2 01 24 6 6 10[HNOjjM
TABLE IV
- stutters Removal
Percentage Recovery
PH /molarity
4 M2 M1 M
0.1 M
pH 2pH 3.8
PH 6pH 8pH 10-11
Time(hrs)
49.3347.5047.7591.0890.0048.8389.5092.421.50
ManganeseDioxide
07.511.836.2318.8299.0099.6199.7461.26
AntimonicAcid
12.3654.4570.1799.1099.2197.7599.8098.3298.37
ZirconumPhosphate
025.9516.0935.0855.9069.5398.1892.5465.80
TitaniumPhosphate
6.4345.4226.5640.8553.6499.6190.4468.3926.54
CopperHexacyanoferrate (II)
////3.5618.86
98.0877.69/
HydrousTitaniumOxide
////
18.91/
98.8498.0698.13
124
It is apparent from figure 15 that StA is a very effectiveabsorber for a-emitters over a wide range of solution H"*"concentration. HTiO and Mn02 also exhibit a strong affinity foractinides at solution pH values betweenA-10 and 4-8 respectively.
ZrP, TiP and Copper Hexacyanoferrate show good affinity fora-emitters at solution pH of 5-6 but the absorption of actinidesfalls off rapidly at higher or lower pH values.
3.5 The Effect of Interfering Ions
Some waste streams contain inactive salts of elements such asfe~ + , Al3"1" and Kg2 + which arise from magnox or stainless steelcladding treatments or are introduced as process reagents such asferrous sulphamate for Pu reduction or Al3"1" for complexing F~ ions.These ions may be present at high concentrations and could seriouslyinterfere with the absorption of radionculides by ion exchangers.
Exchange kinetics for the uptake of Pu(IV) from l.OM NaN03solutions containing 0.28 g IT1 A13+, 40 g L"1 Mg2+ or 7 g IT1 Fe3 +are shown in Figures 16, 17 and 18 respectively. With the
Figure 16Kinetics of Pu dv) removal by inorganic ionexchangers
Figure 17Kinetics of Pudv] removal by inorganic ionexchangers.
pK 10 Sb»
pH 1 0 TiP
PH20 HnO,pH 1 0 Iffp H V O HTiOINoJ
Time Ihrs)
Time (hrs)
125
Figure 16Kinefics of Pu (iv) removal by inorganicexchangers.
ion
3 4 5Time (hrs)
exception of Mn02 (pH 2.0), solution pH had been adjusted to 1.0.It should be noted that degradation of HTiO interfered with kineticmeasurements and therefore these results are not completelyreliable because of change in absorber particle size during theexperiment.
The presence of Fe3* markedly reduces the rate of absorptionand the effective capacity of the absorber for Pu. The effect ofAl3"*" at much lower concentration than in the experiments with Fe3 + ,is to lower the absorption rate and the effective capacity to someextent. When Mg+ is present at 40 g/L (1.66M) the rate ofabsorption is reduced but generally there is little effect onabsorber capacity for Pu.
Pu K~ measurements in the presence of Fe2"1", Fe3"1", A13 +and Mg2"1" are listed in Table V. Because of the reduction of Pu(IV)by Fe2"*", Pu Kp measurements are for the Pu (I II) oxidationstate. Fe2"*" has very little effect on Pu uptake, presumablybecause of its lower charge, which suggests that Pu decontaminationof Fe3"1" bearing waste streams may be improved by pretreatment with areducing agent.
126
TABLE V
K values for Pu(IV) - (interfering ions)
Ion
Fe3+
Fe2+
A13 +
Mg2 +
pH
01
2
2
012
012
ABSORBERS
Polyan S
8.330 x 102
1.580 x 103
1.990 x 103
4.300 x 10k
1.465 x 101*5.230 x 103
9.930 x 101*
6.240 x 102
6.280 x 103
1.940 x 10**
ZrP
3.750 x 101
3.490 x 101
3.100 x 101
1.953 x 103
1.950 x 102
4.970 x 102
4.420 x 102'
6.970 x 101
5.580 x 102
1.230 x 103
Mn02
1.730 x 101
2.460 x 101
3.130 x 101
~ 0
1.690 x 101
8.900 x 10°1.070 x 102
2.760 x 10 l
1.840 x 101
1.380 x 103
TIP
6.410 x 101
1.765 x 102
1.030 x 102
2.352 x 103
5.430 x 102
8.455 x 102
4.130 x 102
1.200 x 103
2.080 x 103
4.570 x 102
HTiO
n.tn.3.290 x 10 *1.440 x 103
1.853 x 101*
n.ra.9.130 x 101
1.500 x 10 3
n «n«2.910 x 101*1.590 x 10 3
IFe] - 7 gL'1[Al] - 0.28 g!'1[Mg] « 40 gf1
All solutions with pH > 1.0 contained 1.0 IM NaNOTi.ro. " not measured25"C* Pu 86 Pu(III)
Am(III) exchange kinetics at pH 3 from solutions containingeither A13+ (0.28 g L"1), or Mg2"1" (40 g L""1) are shown in figures19 and 20 respectively. Experiments with Fe3 + were carried out atpH 2.0, in order to prevent the formation of Fe(OH)3; Fe3 +completely interfered with the uptake of Am and therefore no.results can be shown.
There is a decrease in the uptake kinetics on all exchangematerials, ZrP and TiP being the most affected.
KD values for Am(III) in the presence of the interferingions Fe3 + , Fe2 + , A13+ and Mg2* are listed in Table VI. WhereasFe3+ is absorbed in preference to Am3"*", Fe2+ is not. It should be
127
Figure 19Kinetics of Am (m) removal by inorganicion exchangers
Figure 20Kinetics or Am(ni ) removal by inorganicion exchanger
1-OJJ NoKO
6 24 46
Time (hrs)Time I hrs)
TABLE VIvalues for Am(III) - (interfering ions)
Fe3+
Fe2+
A13+
Mg2+
PH
01
2
2
013
013
ABSORBERS
Polyan S
2.170 x 101
1.460 x 101
~ 0
3.426 x 102
2.100 x 103
1.570 x 103
6.200 x 101*
3.780 x 101
9.720 x 102
5.180 x 103
ZrP
5.580 x 10°~ 0~ 0
2,350 x 101
8.450 x 10°5.050 x 101
5.750 x 101
1.960 x 101
6.140 x 10°3.380 x 10 ï
Mn02
~ 0~ 0~ 0
~ 0
2.000 x 10 *5.480 x 101
3.100 x 103
1.490 x 101
7.520 x 10 *1.380 x 10 5
TiP
~ 0~ 0~ 0
3.970 x 101
6.810 x 10°6.700 x 10 l
1.110 x 103
~ 03'. 780 x 10 !2.400 x 102
HTiO
n«ni.1.400 x 10°
~ 0
4.810 x 101
n.m.5.470 x 10:
7.070 x 10 3
n »in«1.020 x 10 3
2.940 x 10 5
All solutions with pH > 1.0 contained l.OhlNaNO,n.m. • not measured25 CC
[Fe][Al][Mg]
7 gL-10.28 gL'140 gL-1
128
noted, however, that the solution pH at which these Fe measurementswere carried out, was not optimum for Am3 + uptake. In contrast tothis, A13 + and Mg2+ have very little effect on Am removal over thesolution pH range 0-1 but as the pH is increased above 1, the effectof Al and Mg on KD for Am is different for differentabsorbers. Mg2* has little effect on the Am Kp values forMnC>2 and HTiO and may enhance them slightly.
KJJ measurements for Np(V) uptake from solutionscontaining Fe34 (7 glT1 ) , A13 + (9.28 g L'1) or Mg2* (40 g L'1) arelisted in Table VII. The presence of these ions in the wastesimulates reduces the value of KJJ for Np(V) because ofcompetition for exchange sites. The affinity of these exchangersfor NpC>2+ is greater than that for Na+ since, due to its largerionic size, its hydration sphere is held less firmly than that ofNa+ and hence its coordination 1^0 molecules are displaced morereadily.
TABLE VII
K . values for Np(V) as NpO- - (interfering ions)
Ion
Fe3 +
A13 +
Mg2*
pH
012
012
012
ABSORBERS
CSbA
~ 04.850 x 101
4.070 x 101
2.980 x 101
1.120 x 102
1.730 x 102
2.210 x 101
2.350 x 102
1.640 x 102
ZrP
~ 04.320 x 10°
~ 0
9.640 x 10°8.130 x 10°1.250 x 101
1.900 x 101
4.990 x 101
1.160 x 101
Mn02
~ 08.940 x 10°1.680 x 101
1.990 x 101
4.820 x 10°1.340 x 10°
7.820 x 10°3.370 x 10 1
2.860 x 10 1
TIP
~ 02.970 x 10°
~ 0
7.400 x 10°6.060 x 10 l
1.170 x 102
1.460 x 101
5.430 x 102
1.400 x 102
HTiO .
rum.1.02 x 10°2.270 x 10 l
n «m«7.500 x 10°6.180 x 102
n »nu3.330 x 10 3
8.670 x 10 :
[Fe] - 7 glT1[Al] « 0.28 gL-1[Mg] - 40
All solutions with pH > 1.0 containedn.m. « not measured25 °C
129
With exception of Mn02 , the extent of competition at pHl (and in1.0 _M NaNOg ) decreases in the sequence
Fe3 + > A13+ > Mg2*
at the concentrations examined.
These data may be rationalized in terms of the ionic charges andradii of the competing ions. The smaller charge on Mg2"*" compared withthat for both Fe3 + and A13+ is reflected in its smallest interferingeffect. Fe3* and A13 + possess ionic radii of 0.64 and 0. 51Arespectively, and as such, the hydration sphere around A13 + is morefirmly bound than that for Fe3*. It is therefore easier for Fe3+ tolose its water molecules and enter the ion exchange structure.
4.
The experimental data collected so far in this study of
exchangers is showing that there are several inorganic ion exchangersthat can provide high distribution coefficients for actinide ions. Ingeneral, the value of KJJ is greater for Pu(IV) than for Am(III)
and is lowest for Np(V). This behaviour might be expected because of
the different charges on the actinide ions i.e. Pu1*"1", Am3 + and Np02+.
The presence of di- and tri- valent metal ions, such as Fe3+, Al3"1" andKg2"* in the solution, reduces the value of Kp for the actiniae* because ofcoopetition for exchange cites.
The performance of these selected absorbers under flowconditions is now being examined together with a study of recoveryof absorbed radionuclides by elution and absorber recycle.
5. Acknowledgments
The experimental data presented in this paper was obtained byB.A. Phillips, S.P. Dagnall and N.P. Monckton of Separation ProcessesGroup, Chemical Technology Division, AERE, Harwell. Their help in thepreparation of this paper is also gratefully acknowledged.
130
This work was commissioned by the UK Department of the Environmentas part of its radioactive waste management research programme. Theresults will be used in the formulation of Government policy but at thisstage they do not necessarily represent Government policy.
References
[l] Hooper, E.W.; Phillips, B.A; Dagnall, S. P; Monckton, N. P. DoEReport No DoE/RW/83. 171, Harwell Report No. AERE-R11088
[2] Coleman, G.H. The Radiochemistry of Plutonium, Atomic EnergyCommission.
[3] Schulz, W.W., The Chemistry of Americium, TechnicalInformation Center, ERDA(1976).
[A] Burney, G.A; Harbour, R.H., Radiochemistry of Neptunium. NAS-NS-3060, U.S.A. E.G. Washington (1974).
131
SEPARATION AND PURIFICATION OF FISSION PRODUCTS FROMPROCESS STREAMS OF IRRADIATED NUCLEAR FUEL
S.A.ALI.H.J.ACHEInstitut für Radiochemie,Kernforschungszentrum Karlsruhe,Karlsruhe, Federal Republic of Germany
Abstract
New results are presented of studies of the appli-cation of inorganic exchangers in the followingfields:- Concentration and removal of fission product molyb-
denum from solutions of irradiated, short-cooledfuel element targets by attachment to water-containingmetal oxides and subsequent thermal desorption ofMoU3 from the stationary phase. The volatilizationyields were clearly improved by the addition ofwater vapor to the mobile gas phase.
- Chromatographie removal of cesium from the aqueousmedium-active waste streams of nuclear fuel reproc-essing plants on ammonium molybdatophosphate,ammonium hexacyanocobalt ferrate and zirconium phos-phate. In this case, stationary phases of ammoniummolybdatophosphate were found to be by far the mostefficient exchangers among those studied.
Inorganic exchangers, because of their stabilityagainst oxidizing and reducing media and their resist-ance to radiation, have found extensive applicationin radiochemical separation techniques and as station-ary phases in the radionuclide generators used world-.wide, especially in nuclear medicine [1]. Anotherinteresting application has been indicated by morerecent studies of cesium removal from the concentratesof medium-level waste and directly from the cesium-bearing partial streams acidified with nitric acid,respectively, in nuclear fuel reprocessing plants.Cesium removal from these solutions would clearlyreduce the required additional thicknesses of so-called "lost concrete shields" and thus generate con-siderable savings in the costs of final waste con-ditioning.
Metal oxides, because of their extensive use ingenerator applications, are among the best proveninorganic fission product adsorbers.
133
The generator used most frequently by far is thetechnetium-99m generator. Technetium, because of itslow gamma energy of approx. 140 keV and its halflifeof 6 hours, has optimal characteristics for organfunctioning tests. The worldwide success of Tc-99mgenerators began only after suitable exchangers hadbeen found for the parent nuclide, molybdenum-99,which is increasingly being isolated from short-cooled fuel element targets irradiated specificallyfor this purpose, because of the extremely highspecific activities produced by this technique. Themost successful stationary phase used in technetiumgenerators has been acid aluminum oxide. It canimmobilize molybdate anions with high retentioncoefficients from weakly acidified solutions, whilethe technetium passes through the column. Severaldevelopment studies have been conducted in order toutilize the favorable properties of aluminum oxidein fission product molybdenum removal and concen-tration. The results achieved to date have led toinorganic exchanger columns having become establishedparts in most fission product molybdenum removalprocesses. Normally, molybdenum decontaminated to ahigh degree from solutions of low acidity is immobi-lized on Al2C>3 columns and then eluted by ammoniaafter the completion of the washing processes [2].In some production plants all organic impurities areremoved in a time consuming process in which themolybdenum containing solution is evaporated to dry-ness and molybdenum is sublimed as molybdenum tri-oxide at 10OO °C. In these studies [3] it has beenattempted for the first time to streamline the pro-cess by combining the immobilization of molybdenumto hydrated metal oxides and the subsequent directsublimation of the molybdenum from the adsorber matrix.If this modification can be realized, it would clearlyreduce the expenditures in terms of manpower and timeof the respective production campaigns and would de-crease the amount of radioactive waste solutionsaccumulating without entailing any losses of qualityof the final product.
In order for the desired development to be appli-cable to full-scale production processes, volatiliza-tion should be feasible at the lowest possible tempe-ratures. A major reduction in the sublimation tempera-ture of molybdenum in volatilization from the ex-changer matrix can be achieved by making use of theincreased vapor pressure of molybdenum oxohydroxide,Mo02(OH)2; this is a gaseous compound observed in thepure Mo03-water vapor system at temperatures around650 °C [4].
In the first few experiments the steady-state ex-changer behavior of molybdenum was studied in mediaacidified with nitric acid and containing the respec-tive inorganic exchangers, A12Û3, Sn02/ Mn02, andSb205. Fig. 1 is a plot of the molybdenum retention
134
e 100ol 90o>°l BÖ
706050CO3020100
aqMn02 • oqSn02 • aqSb 2 0 5 -oq
5 10 15 20 25 30 35 CO C5Loading (mg Mo/g Exchanger]
Figure 1: Retention of Mo [%] on the oxides vsMo-loading in 0.02 M HNO.,.
of water-bearing metal oxides in an 0.02 M HNÛ3medium as a function of feed loading. The feed loadis the ratio of the amount of molybdenum in thesolution to the amount fed into the exchanger. As thediagram shows, aluminum oxide at this pH has thehighest retention coefficients with 30 mg of Mo/g ofexchanger. The comparatively high retention of A1203in this pH range can be explained by its high poros-ity. Exchanger experiments at higher concentrationsof nitric acid produced mainly decreasing Mo reten-tion capacities in all exchangers.
In view of the intended use under hot conditionsnumerous experiments were conducted in the exchangersystems mentioned above to define their kinetic andwear behavior. Aluminum oxide and tin dioxide exhib-ited the most advantageous preconditions. Antimonypentoxide, because of its slow exchange kinetics,and manganese dioxide, because of its pronouncedwear sensitivity, were excluded from further study.
To optimize the retention of molybdenum, the in-fluence of the pH on sorption to A1203 and Sn02 wasstudied over a wide range at constant molybdenum feedloads (3 mg Mo/g of exchanger) and a molybdenum con-centration of 2 x 10" 3 mole/1. Fig. 2 is a plot ofthe amounts of molybdenum immobilized on the Sn02 andA12O3 exchangers, respectively, as a function of pHafter 20 hours of contact time. A12Û3 acts as ananion exchanger already from a pH of 9 onward, com-pletely immobilizing 3 mg of Mo/g of exchanger froma pH of 8.5 on. When the pH is reduced, complete re-tention is maintained up to pH 1, but afterwards
135
100
•: BOg 70I 60
302010
~- A l i O ,o-o-o Sn Oj
•1 0 1 2 3 I, 5 6 7 B 9 10pH- value
Figure 2: Retention of Mo [%] on the oxides vs. pH
there is a steep decline in capacity. SnC>2 shows asimilar behavior in principle, but the entire curveis shifted more to the acid region. Above pH 8 itacts as an anion exchanger, and at pH 5 it achievescomplete retention of molybdenum, which is maintainedup to a pH of 0. This is followed by a slight declinemuch less pronounced, however, than in the case ofA12Ü3- The reason for the shift in the SnU2 anion ex-change towards lower pH levels is the lower basicityof the -SnO~-group compared with A^O^. The electrondensity of oxygen is deformed strongly to the posi-tive metal ion by the tetravalent oxidation level oftin, which binds more firmly the OH~-group to be ex-changed and allows anion exchange to take place onlyat higher H+-concentrations than in the case ofA1203.
After optimization of the exchange conditions itwas possible to start the volatilization experiments.It was important to elucidate the dependence of theMo03 desorption yield on the loading of the station-ary phase; for this purpose, A12Û3 with differentloads was predried at 100 °C and subsequently sub-jected to a 30-minute thermal treatment at tempera-tures up to 120O °C.
Fig. 3 is a plot of the sublimed Mo03 fractionover the exchanger loads. It indicates that, forhigh volatilization yields to be attained under theseconditions, A1203 must be loaded with the molybdenumcarrier almost to the limit of capacity (dashedline). At lower loads, the volatilization yieldsdecline drastically to levels below 10%. This is onemajor disadvantage of A^OS, for large quantities ofinactive molybdenum carrier would entail low specificactivities of the product.
Another reason for avoiding high loads is themarked decline in the exchange rate. Clear advantages
136
2 100!5"»..
I 9U"o
I «"Br*»
l 7D
60
50
to
30
20
10
0 10 15 20 25 30-Looding |mg Mo/g AI203]
Figure 3: Yield of the sublimation of Mo03 [%] vs.Mo-loading.Temperature = 12OO °C.
in this respect were found in SnC>2. In a number ofseries of experiments the influence on molybdenumsublimation yields of volatilization temperatures,the duration of thermal treatment, and of water vaporin the mobile phase passing through the molybdenum-loaded adsorbers was studied.
Fig. 4 shows the comparable thermal molybdenumdesorption yields from systems containing A12Ü3 andSnC>2 in the presence of water vapor plotted vs. sub-limation time. One characteristic of the developmentof sublimation in tin oxide is the steep initialrise in 11003 sublimation yields which then graduallydeclines with the decrease of MoOß on the exchangersurface. The reason is the drop of the concentrationgradient, which is the prime mover. The much flatterrise in the sublimation yield in A^C^, however, isindicative of volatilization being determined by theslow diffusion step through the pores in the surface.
In another program of specific technologicalorientation conducted within the framework of theReprocessing and Waste Treatment Project, Chromato-graphie techniques are developed mainly for the spe-cific decontamination of individual medium-activewaste streams arising in the reprocessing of nuclearfuels and in fuel element fabrication [53. The under-lying concept is based on the fact that these solu-tions contain largely similar chemical species of the
137
20 id BO 100 120 KO 160 1BD 200Time [mm]
Figure 4: Yield of the sublimation of MoCU [%] vs,time of sublimation at a loading of4 [mg Mo/g oxide].Temperature of sublimation = 1050 °C inthe presence of H_0 vapor.
carriers of activity. Their removal by a directlyconnected specific decontamination step would be muchsimpler and more effective than the procedure devisedso far in which first nearly all aqueous processwastes consisting mainly of nitric acid and sodiumcarbonate solution are co-evaporated. Decontaminat-ing the concentrate of the carriers of activity isto be achieved by cumbersome procedures comprisingessentially a series of precipitation steps.
At the Institute of Radiochemistry (IRCH) modern,simple and effective decontamination procedures byextraction chromatography have been developed withinthe Reprocessing and Waste Treatment Project for suchimportant partial process streams as the basic car-bonate stream [6, 7], the solvent wash in reprocess-ing, and the oxalate bearing mother liquords of Puoxalate precipitation in nuclear fuel fabrication.
The activities in this field are at present beingconcentrated on studies of cesium removal from theaqueous medium active solutions of nuclear fuel re-processing plants. If this project were to be imple-mented, it would offer a possibility for removal ofthe longer-lived Cs-137 source of activity, whichwould entail clear savings in shielding costs of thewaste after final treatment.
The process to be developed should permit ratherwide variations in the composition of the solutionsfrom which cesium is to be removed, thus obviatingthe need for additional expenditures to set optimumprocess conditions, such as the salinity and acidityof the solution. It is known from experience thatthe cesium distribution is limited to the process
138
streams acidified with nitric acid and enters theevaporator concentrate in this way; a powerful methodtherefore would have to be able under all conditionsto allow high cesium decontamination levels to beachieved in nitrate-bearing solutions strongly acidi-fied with nitric acid.
In steady-state and dynamic experiments the distri-bution and retention behavior, respectively, of cesiumfrom aqueous solutions of various acidities andbasicities, respectively, and various sodium nitratecontents on the inorganic exchangers ammonium molyb-datophosphate (AMP-1) made by Bio-Rad and ammoniumhexacyanocobalt ferrate (NCFC), hydrated antimonypentoxide (HAP), and zirconium phosphate (ZPH) made bythe Carlo Erba company was studied [8]. The data deter-mined so far underline the clear superiority of AMP-1under almost all experimental conditions. Fig. 5 is aplot of the cesium retention percentage determineddynamically as a function of the loading of AMP-1 insystems of different pH levels.
100
BO
-.ftD
A 6 H HKDj• 1 H NN03D 0.1 H KK03• pH 7 (HjO)o pH 9.5
10 20 30 tO 50 60 70 BO SO 100Loading [gCs/kg AMP-1]
Figure 5: Retention of cesium on AMP-1 vs. loadingfor various H+-concentrations.
The diagram shows that cesium is retained > 99% byAMP-1 in solutions ranging between higher HNC»3-con-centrations and weakly basic solutions with a capacityof > 60 g/kg of exchanger. The uncommon versatility ofthis exchanger is demonstrated by the comparison belowof data determined under steady-state conditions insystems whose stationary phases in each case were theNCFC and ZPH exchangers. Fig. 6 is a plot of thecesium retention percentage determined under dynamicconditions from aqueous phases of various acidity andbasicity levels, respectively, on the ammo-nium hexacyanocobalt ferrate (NCFC) inorganic ex-changer as a function of the loading of the solidphase.
139
25
pH 12.5
pH 7.0
0.1 M HND3
1M HN03
conc. HN03
10 20 30 40 50 60 70Loading [gCs/kgNCFC]
Figure 6: Retention of cesium on NCFC vs. loadingfor various H+-concentrations.
The diagram shows the very low retention levels inacid media; even in neutral solutions the use of thisexchanger cannot be recommended. Acceptable values canbe achieved only above a pH of 9.
Comparable data take a clearly different turn ifzirconium phosphate is used as the stationary phaseIn neutral systems zirconium phosphate is found to beparticularly powerful. Under these conditions, approx100 g of cesium can be immobilized quantitatively on1 kg of exchanger. However, the high retention de-creases rapidly with increasing shifts both in theacid and in the basic directions; already at a pH of1 only some 8 g of cesium/kg of exchanger can be re-tained. In basic systems the losses of capacity arealso intolerably high. A comparison of the retentionlevels attainable in NCFC and ZPH systems with thoseof AMP-1 indicates the clear discrepancy in the rangeof applicability. It is possible with AMP-1 columnsto retain cesium quantitatively from highly acidi-fied, neutral and basic systems without any pHadjustment.
140
1 100
Figure 7:
75
25
pH 1(0.1 MHNDj l
HN03
pH 12.5
SMHNOj
50 100 150Loading IgCs/kg ZPH]
Retention of cesium on ZPH vs. loadingfor various H+-concentrations.
•1100
90
eo
70
60
50
10
30
20
10
f.*fc\1
CsCs.3.6MHaN03 1MHN03
MAW-simulai«
Figure 8:
3 5 10 30 50 100 300 500 1000Loading [gCs/kg AMP-1]
Retention of cesium on AMP-1 vs. loadingin 1 M HN03 for Cs in (a) pure,(b) NaNO-,-contaminated, (c) simulatedMLW solutions.
141
Another important criterion of the practicalapplicability of an exchanger is its behavior in thepresence of higher nitrate concentrations, such asthose which occur in an MLW concentrate.
Fig. 8 shows the cesium retention levels obtain-able as a function of exchanger loading from1 M HN03, 1 M HN03 + 3.6 M NaN03, and 1 M HN03 + MAWsimulate. While hardly any negative effect could beobserved in the AMP-1 system, the retention capac-ities of all the other exchangers broke down evenunder their most favorable exchange conditions.References[1] "Radioisotope Production and Quality Control",
IAEA, Vienna (1971), Technical Report SeriesNo. 128, pp 659-743
[2] Sameh A. Ali, H.J. Ache,Fourth Intern. Symposium on RadiopharmaceuticalChemistry(Abstracts), Jülich (1982) 168
[3] J. Biirck, Sameh A. Ali, H.J. Ache,KfK-3754 (1984)
[4] 0. Glemser, R.V. Haeseler,Zeitschrift für anorganische und allgemeineChemie,Band 316 ( 1 9 6 2 ) 168-181
[5] W. Faubel, P.-M. Menzler, Sameh A. Ali,KfK-3617 (1984)
[6] J. Haag, Sameh A. Ali,KfK-3015 ( 1 9 8 1 )
[7] J. HaagKfK-3460 (1983)
[8] W. Faubel, P.-M. Menzler, Sameh A. Ali,KfK-3742 ( 1 9 8 4 )
142
EFFICACY OF ARGILLACEOUS MINERALSUSED AS A MIGRATION BARRIERIN A GEOLOGICAL WASTE REPOSITORY
E.R. MERZInstitute of Chemical Technology,Kernforschungsanlage Jiilich,Jülich, Federal Republic of Germany
Abstract
The current strategy for isolating nuclear wastes inan underground geological repository relies on aseries of multiple barriers to restrict the releaseof radionuclides to the biosphere. Waste form andcanister as well as the backfill material used to fillthe space between the waste canister and the hostrock, provide engineered barriers, whereas therepository itself and the overlying geologic mediumrepresent natural barriers.A suitable backfill material, preferentially argil-laceous minerals, exhibits the potential for provi-ding a multifunctional role in the overall performanceof the waste package, significantly retarding radio-nuclide migration via solution flow. The backfillbarrier will serve both as a physical barrier,preventing convective solution flow and allowingradionuclide penetration only by diffusion, and as achemical barrier capable of chemically interactingwith the radionuclides via sorption, ion exchange orprecipitation.The paper presents measurements of the hydraulicproperties and the diffusion and sorption abilitiesof various bentonites and clays. They include in-vestigations dealing with the sorption and retardationbehaviour of the two most important fission productradionuclides strontium and cesium. The Kd-valuesdetermined by batch distribution measurements show astrong dependence from the salt concentration in thesolutions. Brine solutions bring about a displacementaction upon sorbed ions. In highly compacted bentonitesthe transport of radionuclides through the backfillmay be dominated by diffusion through the inter-stitial water. The rate of diffusion will thus dependon normal diffusion and on the sorption properties ofthe clay minerals.
IntroductionThe general strategy in isolating nuclear waste is toimpose a series of barriers between the radioactivematerial and the biosphere. The different barriers
143
considered are in their order of proximity to thebiosphere:
geological repository including overlyingstrata,backfill and sealing material in therepository,canister including overpack,waste form.
It is possible that either one of the first threebarriers, if appropriately choosen, could by them-selves provide the necessary isolation of wastes,irrespective of the quality of the waste form. Thus,the main function of the waste form would be to containand safely isolate the radioactive materials duringhandling and transportation procedures until properdeposition in the geological repository. However, itis common understanding that the waste form should bemade as chemical resistant as reasonable achievable toprovide the maximum possible protection within anacceptable cost range.In this contribution the attention is directed towardsthe usefulness and efficacy of backfill materials onthe basis of argillaceous minerals. Emphasis is putonly on salt evaporites as the geological host mediasince salt domes are the favorite choice for a high-level waste repository in the Federal Republic ofGermany [1].Function of backfill materialsVarious backfill materials are considered to be usedin a geological repository to fill the space betweenthe waste canisters and the host rock. Suitable back-fill materials may also be used to fill the minedrifts and shafts when the repository is sealed.Proper selection of backfill- and displacementmaterials, respectively, and methods of emplacementcreates an additional barrier between the radioactivewaste and the biosphere in case an aqueous solutionis present as the transport medium in the repository.A comprehensive review related to backfill performanceof various materials has been prepared by WHEELWRIGHTet al [2]. Most promising materials are argillaceousminerals, preferably expanding clays like sodium orcalcium bentonites.The backfill has the potential for providing a multi-functional role in the overall performance of thewaste package, significantly retarding or minimizingwaste package degradation and subsequent radionculidesrelease. The primary function is threefold:- to retard or exclude the migration of
ground water or salt brine between thehost geology and the waste canister;
144
- to buffer the pH value and chemicalcomposition of the aqueous solutionand the canister surface in order tosuppress corrosive attack;to retard the migration of selectedchemical species from the waste matrixinto the ground water after breachingof the canister has started.
In addition, supportive functions are the self-sealingof cracks in the backfill or interfacing host geology,the provision of resistance to applied mechanicalforces and a good heat conductibility from the canisterto the host rock system.Experimental InformationThe objective of the investigations was focused on twoitems :
the hydraulic and- the sorptive as well as migration
retardationproperties of selected backfill materials.Only expanding clays with a high content of bentoniteand montmorillonite, respectively, have been consideredas potential candidates, since they combine the twomost important required properties, namely, of solutionconvection suppressing and ion sorption capabilities.They include natural as well as activated bentonitesand different salt clays with high contents ofalcaline or alcaline earth bentonites.The ability of these materials to reduce significantlyor to stop the flow of brine and ground water betweenthe host rock and the waste canisters is of primaryimportance, because it inhibits the single viablemechanism for transporting radionuclides from therepository to the biosphere. Migration tests includestudies of the ability of the clay materials to retardthe outward migration of radionuclides leached fromthe waste form after the structural canister barriersare breached, and secondarly, to slow down the inwardmigration of chemical species that may enhance corro-sion of the canister.Experimental ResultsMaterialsThe materials used are:
1. Natural bentonite ( N B ) , KärlicherTon- und Schamottewerke,
2. Activated bentonite I, II, III ( A B ) ,Kärlicher Ton- und Schamottewerke,
145
3. Grey salt clay (T3), Werk Niedersachsen-Riedel,500 m level,
4. Red salt clay, brown fraction (T4B), WerkNiedersachsen-Riedel,13O5 m level,
5. Red salt clay, red fraction (T4R), WerkNiedersachsen-Riedel,1305 m level,
6. Brick clay (ZT), clay pit Altwarmbuchennear Hannover.
The composition of the various materials determinedby chemical analysis [3] are depicted in table I.The solutions used and their composition are listed intable II.
Table I: Composition of materials used
Component
Si02
TiO2
Al,0,Fef03
MgOCaOSrONa20K2OCSjOH2O(total)
Mass Fraction in %
NB
60.10.88
17.39.51.81.4
< 0.0090.212.10.00087.1
ABI
58,60.51
17,65.14.13.0
< 0.0071.91.80.00067.8
ABU
56.00.3
20.64.73.42.0-
3.01.4-
7.5
AB III
52.50.7
16.98.81.91.3-
8.41.5-
7.6
T3
46.10.68
14.41.5
17.11.00.0430.462.00.0005
14.6
T4B
47.70.296.11.99.3
19.00.110.751.30.00043.1
T4R
45.90.43
12.93.0
11.93.10.0699.92.20.00073.4
ZT
58.10.83
19,45.22.53.0
< 0.0061.43.10.00061.8
Table II: Composition of solutions used
Component
Na+
K+
Mg?+
CI'sol-
Densityf[g/cm3]
Destined
_----
1,00
Quinere Brine(Quit 1)
Ig'U
44,331,843,4
220,44,7
1,233
Quarternere Brine(dual)[g/i]
44,331,842,2
220,4—
1,226
SaturatedNaCI-Brine
tg/i]
14,5--
21,3—
1,197
146
Hydration RateThe rate of water intake was measured for severalbentonite species at room temperature _applying amethod originally proposed by ENSLIN
app[4J.
The samples were compacted to a density of approximately2,0 g/cm3. The water intake velocity as well as thevalue for water saturation show a dependence on thesodium content of the bentonite samples. The highertheir share, the greater their water intake capability[5]. The results obtained are illustrated in figure 1.
NB
0 20 iO 60 80 100 120 140Time [ hours ]
FIG. 1: Temporal progress of water intake of variousbentonite specimen
Swelling PressureIf a compacted bentonite body has access to externalwater and is confined so that its total volume remainsconstant, a swelling pressure is built up. Themechanism of this phenomena is fairly well understood[6] . It is known that the swelling pressure of smectiteclays is a function of their initial compactiondensity.The swelling pressure of compacted high-densitybentonite samples (p = 2.0 -2.1 g/cm3) was measuredin an apparatus shown in figure 2.It allows the determination of the swelling pressureby measuring the force applied on a load cell by awater exposed sample, restrained to a constant colume
The results obtained applying two different large areasexposed to water access are illustrated in figure 3.In the lower curve by far no saturation has beenachieved due to the fact that only about one fifth of
147
:-<!"""* ^ i™* _ ***
Screw RodWoter
Bentonite
Receiver
Force Lood Cell
FIG. 2: Schematic view of the swelling pressuremeasuring device
0 10 20 30 40 50 60Time [days ]
FIG. 3: Swelling pressure of highly compacted sodiumbentoniteCurve A: water exposed surface area 2OO rnm^;
water content after 60 days exposureapprox. 4 % 2Curve B: water exposed surface area 1OOO mm ;water content after 6O days exposureapprox. 16 %
148
the sample surface has been exposed to water access.The upper curve represents a more realistic picture.But even in this case no complete saturation of waterintake has been reached after 60 days of exposure.
PermeabilitySome material transport will take place towards thecanister wall and vice versa from the eventuallybreached canister backwards through the backfillmaterial. The predictive theory which describes theflow of brine or ground water, respectively, in aporous material is DARCY ' s law; the directly derivedparameter is the hydraulic conductivity.
Q = -K • i • AQ = discharge rate [cm- /s]K = hydraulic conductivity [cm/s]i = hydraulic gradient [cm/cm]A = cross sectional area [cm^jIn the formalism of DARCY ' s law, the hydraulicconductivity, K is a constant relating discharge tothe hydraulic gradient. It is a function of both thesolid material and the fluid as indicated by therelationship
K =•y
k = permeability (intrinsic) [cm ]p = fluid density [g/cm3]n = fluid viscosity [g/cm-s^jg = gravitational acceleration [crn/s ]The permeability k is a function of the solid mediaand depends upon variables such as grain size, grainshape, porosity and path tortuosity. The density andviscosity factors depend entirely upon the fluidphase; and the dependence on viscosity clearlyindicates that hydraulic conductivities vary stronglywith temperature.After saturation of the backfill, chemical speciesmay be transported through the backfill either bymovement of liquid, or if the backfill' performs asexpected, by diffusion through the interstitialwater. In either case, the ability of the backfillto retard the migration by sorption, precipitation,or any other mechanism is one of the more importantattributes of the backfill.The hydraulic conductivity of brine through compactedbentonite (density 2.1 g/cm^) has been measured atroom temperature in a constant-volume flow cell. Thebackfill material is confined triaxially betweenporous stone plates, and liquid brine is pumped through
149
the material at a constant hydraulic head pressureuntil saturation is achieved. The time elapsed untilsaturation yields the materials hydration rate; theflow rate after saturation yields the materialsliquid permeability in units of [uarcies], or,equivälently, the hydraulic conductivity in units of[m/s].The hydraulic conductivity dropped rapidly with timebefore stabilizing at a much lower value than at thebeginning (about 1/4 of its original value).Tablelll summarizes the measured data.The high permeability coefficients observed give riseto make the backfill material practically imperviousto brine percolation. However, radionuclide transportis still possible - though at a very low rate - byanother mechanism, namely diffusion. Sufficientquantitative data of radioisotope diffusivity arestill lacking [7,8].
Table III: Permeability measurements of activated Na-bentonite samples.Depicted values after stabilisation (time span elapsed ~ 500 h)
MaterialActivatedNa-bentonite
ABIABUAB III
Density
[g/cm3]
2.112.102.09
Hydraulic HeadPressure
[MPa]
151515
HydraulicGradient
1.3 -105
1.5 -105
1.6 -105
Hydraulic Conductivity[cm/s]
4.8 • 10-13
4.0-10-13
5.4 • 10't3
Diffusion MeasurementsIn a series of experiments the apparent diffusitivity ofcesium, sodium and chlorine ions have been measuredusing a stack of highly compacted bentonite disksforming a diffusion cell and applying radio-tracersof Na-22, Cl-36 and Cs-134 in a saturated NaCl brine.The bentonite was compacted to a density of 1.9 g/cm3.After a period of 2 months the cell was dismounted,and the individual disks measured separately withregard to their radioactivity; the y-radioactivitywith a Ge(Li)detector and the ß-emitters applyingliquid scintillation counting.The diffusitivity (D) for one-dimensional diffusion isgiven by the equation
dcdt -dx dx
which has to be solved by applying appropriate initialand boundary conditions [9].
150
With a plane source containing a limited amount ofsubstance which is diffusing in a cylinder of infinitelength and assuming a concentration independent D,the solution is
cM
12(7T.Dt) 1/2
cMXDt
concentrationtotal amount of diffusing speciesdistance from source [cm|diffusitivity [cm2/s]time [sj
gj
The diffusitivity obtained applying this equation tomeasured concentration profiles yields an apparentdiffusitivity (D~), including the effect of sorptionon the solid [10J. The relation between the apparentdiffusitivity, Da, and the diffusitivity not affectedby sorption, D, is
D = Da(1 + KDp)
p = density of the solid [g/cm ]With the few data available as yet, shown in table IV,it is impossible to draw any meaningful conclusion.More elaborate measurements are needed.
Table IV: Measured diffusivities
Element D, (best estimates)[cm2/s]
CsNaCI
2-10-9
8 • 10-7
5- Id'8
Nuclide MigrationThe physico-chemical mechanismns underlying radio-nuclide .migration through argillaceous clays are rathercomplex. In most cases several effects may be super-imposed in attaining fixation of the released radio-nuclides to the solid backfill. Important are
adsorption and exchange phenomena,complex formation and hydrolysis,precipitation and surface reactions,mixed crystal formation andrecrystallisation,filtration effects if suspendedparticles are present in the liquidmedia.
151
In order to determine the retardation capabilities ofcandidate backfill materials, batch distributionmeasurements of Kd values for strontium and cesiumhave been performed with 6 different argillaceousclays and bentonites using the various solutionslisted in table II [3] . Samples of each (10 and 20 g)were contacted with 50 ml solution for sufficienttime (between 1 and 20 hours). The quinere brine wasapplied only in the Cs system but not in the Sr systemin order to avoid precipitation ofVariables have been the equilibrium concentrations ofSr- and Cs-ions in the respective solutions, tempera-ture (25° and 100°C) , and pressure (1 and 10O bar).
103
102
10°Z T . S r
~H8.Cs
10" 1(T2 10'1 10°Cs, Sr-Equilibr. -Cone, ( g / l l
101
FIG. 4: Batch distribution K<j-values as a function ofequilibrium concentrations for Sr and Cs inaqueous solution at 1 bar and 25°C fordifferent argillaceous clays
The contents of Sr and Cs were determined applyingX-ray fluorescence analysis. The results are shown infigures 4 to 8.
The results show that strontium and cesium are retainedby the bentonites and clays. The substance whichdiffuses through the backfill material will be retardedin transport owing to sorption until equilibrium isestablished. Only then the substance will migratefurther. Retardation increases as the values ofincreases.
152
Kd
102
10°
10,-1
-310
FIG. 5:
HR.Sr
UB.Sf
AB.Cs
I I I I I 1 I
T3.Sr
H8.Cs
10'2 10"1 10°Cs, Sr -Equiliûr. -Cone, l g /1 ]
101
Batch distribution K^-values as a function ofequilibrium concentrations for Sr and Cs insalt brine, quinere and quarternere brines at1 bar and 25°C for different argillaceousclays
101
10°
-£1
'5
10-
10'
AS.NnCI
i i i i i i I__ l I
10" 10° 101
Sr-Pr imary Cone. [ g/l ]102
FIG. 6: Adsorption isothermes of Sr at 1 bar, 25°C
153
10°
co10-
10-z -
10,-2I I
AB. Qui 11
_I I I
10'1 10° 101
Cs -Primary Cone, i g/l]102
FIG. 7: Adsorption isothermes of Cs at 1 bar, 25°C
103
Kd
10
101
10°
10"
(1) Sr, 1 bar, 25°C(2) Sr, 100 bar, 100°C(3) Cs. 1 bar, 25°C(4) Cs, 100 bar, 100°C
J___l I I
10'3 2 10'1 10°Cs. Sr-Equil ibr.-Cone l g / 1 ]
10'
FIG. 8: K^-values as a function of equilibrium concen-trations for Sr and Cs in aqueous solutions atdifferent pressures and temperatures
154
However, the sorptive capacity decreases strongly withincreasing salinity of the solutions. The cations inthe brine give rise to a strong displacement action.An influence of increased temperature and hydrostaticpressure is measurable but comparably small and onlyof minor importance. An interesting behaviour hasbeen found for the strontium retardation, which isabnormally high under certain circumstances. Obviously,it is due to a precipitation of SrS04, originating froma small content of Sr ions in the clay minerals.The rate of diffusion through the backfill is controlledby normal diffusion processes and on the sorptionproperties of the backfill materials. The migration ofvarious ions exhibits a different pattern depending onthe sodium content of the bentonite. It seems highlyprobable that, over the long term, transport ofsolution through the backfill will be so slow thatdiffusion will be the main mechanism of radionuclidetransport and release. In addition to sorption and ionexchange, precipitation as well as remineralisationmay play an important role in the retention of radio-nuclides.
References[l] HERRMANN, A.G., Geowissenschaftliche Probleme bei
der Endlagerung radioaktiver Substanzen in Salz-diapiren Norddeutschlands.Z. dt. geol. Ges. 131 (1980) 433-459
[2] WHEELWRIGHT, E.J., et al., Development of BackfillMaterial as an Engineered Barrier in the WastePackage System.Report PNL-3873 (1981) 1O5 p.
[3] BRODDA, E.G., MERZ, E., Zur Wirksamkeit von Ton-mineralien als Rückhaltebarriere bei der Endla-gerung radioaktiver Abfälle in Salzgesteinen.Z. dt. geol. Ges. 134 (1983) 453-466
[4] ENSLIN, 0., Die chemische Fabrik J_3 (1933)[5] MIRSCHINKA, V., Eigenschaften von Behälterwerk-
stoffen und Verfüllmaterialien für die Endla-gerung hochradioaktiver Spaltprodukte.Report JÜL-Spez-226 (1984) 1OO S.
[6] PUSCH, R., Highly Compacted Sodium Bentonite forIsolating Rock-Deposited Radioactive WasteProducts.Nuclear Technology <45 (1979) 153-157
[7] NOWAK, E.J., Diffusion of Cs(I) and Sr(II) inLiquid-Saturated Beds of Backfill Materials.Report SAND-82-6750 (1982) 52 p.
155
[8] PUSCH, R., ERIKSEN, T., JACOBSSON, A., Ion/WaterMigration Phenomena in Dense Bentonites.Scientific Basis for Radioactive Waste ManagementV, Vol. 11, Elsevier Science Publishing Co,,Amsterdam-New York (1982) 649-658
[9] CRANK, J., The Mathematics of Diffusion.Oxford university Press, London (1956)
[10] TORSTENFELT, H., et al., Transport of Actinidesthrough a Bentonite Backfill.Scientific Basis for Radioactive Waste ManagementV, Vol. 11, Elsevier Science Publishing Co.,Amsterdam-New York (1982) 649-658
156
TEN YEARS OF EXPERIENCE IN EXTRACTIONCHROMATOGRAPHIC PROCESSES FOR THE RECOVERY,SEPARATION AND PURIFICATION OF ACTINIDE ELEMENTS
C. MADIC, J. BOURGES, G. KOEHLYSection des transuraniens,Institut de recherche technologique et de développement industriel,Centre d'études nucléaires,Commissariat à l'énergie atomique,Fontenay-aux-Roses, France
Abstract
An extraction Chromatographie process has been developed for the242 241 2"Î8 2^7recovery of actinide elements ( C m , Am, " Pu, Np, U
and others) from irradiated targets, from alpha liquid wastes and fromaged actinide stocks. An inorganic adsorbent can be used as a stable,inert support for an inorganic extractant, e.g., TBP- impregnatedsilica. Due to the large variety of organic extractants available, thistechnique can be applied successfully to solve difficult actinideseparations in quantities ranging from micrograms to kilograms. Acomparison of extraction chromatography is also made with resinchromatography and liquid-liquid extraction.
I INTRODUCTIONSince about fifteen years the French Atomic Energy Commission has
developed a program for the production of isotopes of actinides elementsfor french and international uses. An overview of the isotopes produced,the accumulated amounts since the beginning of the program and themethod of preparation is presented Table I. Basically there is two types ofprograms.I.I. Major programs(i) Involving the preparation of special targets, their irradiation by
neutrons in nuclear reactors and the chemical treatment of theirradiated targetsproduction of *33u by irradiation and treatment of 232ThC>2 targetsproduction of 238pu by irradiation and treatment of 237(>jp/Ai or237NpO2/MgO targetsproduction of 2^2pu> 243Am, 2^Cm by irradiation and treatment of239pu/Al targets or 2*2pUQ2 targets.
(ii) Involving the milking of large amounts of parent material : produc-tion of 241 Am by chemical treatment of kilogram amounts of PuO2or Pu metal stocks rich in 2*lpu.
(iii) Involving the chemical treatment of large volumes of alpha activeliquid wastes : recovery of 241 Am.
1.2. Minor programsWhich correspond to productions of special isotopes either bymilking of stocks of aged parent (228ih, 229th, 23*u, 237y, 239Np,
, 2<f&Cm) or by chemical treatment of small irradiated targets
157
TABLE IISOTOPES PRODUCED OR RECOVERED
ELEMENT
THORIUM
URANIUM
NEPTUNtUM
PLUTONI
UM
AMERIC
UM
CURIUM
ISOTOPE
22STh
229-rh
233u
23*u
237(j
237Np
239Np
238Pu
240pu
2»2pu
2»lAm
2*3Am
202Cm
2<*«Cm
2<»8cm
METHODOF PRODUCTION
232uS. 228Thmilking of large amounts
of aged 233u rich in 232u
233u^ 229-Thmilking of large amounts
of aged"3u
Irradiation of 232yh by neutronsChemical treatment of
irradiated 232jhO2 targets238pu2, 23*u
milking of large amountsoî aged 2J8puO2
2<Jipu2.237umilking of aged plutonium
rir-h m 2*Ipu isotope
Recovered in special campams mthe reprocessing plant devoted to
treatment of irradiated nuclear fuels
203AmS.239Npmilking of aged 2i*3/almO2
Irradiation of 237Np/TtChemical treatment of irradiated
Np/Al or NpO2/MgO targets
2*4Cm2. 2«>Pumilking of aged 2»<tCmO2
Irradiation of 239pu targets/7)chemical treatment of irradiated
Pu-Al targets
2<*Ip u C2«lAmMilking of aged PuOj or
Pu metal stocks - Chemical treatmentof alpha active liquid wastes
Irradiation of 239pu or 2*2puby neutrons
Chemical treatment of Pu-Alor PuC>2 irradiated targets
Irradiation of 2*1 Am by neutronsChemical treatment of
irradiated AmO2 targets
Irradiation of 239pu Or2'»2puby neutrons - Chemical treatment
of irradiated Pu-Al or PuO2 targets
252CfflL248CmMilking of aged 232d stock
Scale of productionsince the beginning
of the programlO^g
mg
hundredsof grams
grams
*«Ci
10 kg
^g
kg
10 of mg
hundreds
kilogram
100 g
10 mg
too g
10 0*Mg
For the production and purification of all these isotopes chemicalseparations must be achieved ; in few cases these separations are easy tobe performed due to the large difference in chemical properties of the twoelements in presence, this is the case for the production of 239fvjp(239Np(IV))/(243Am(III)) separation or 240Pu (2Wpu(iv))/2Wcm(iiI)) sepa-ration ; for the other isotopes produced difficult separations are needed s
. recovery of traces of 233u(VI) in large amounts of 232Th(IV)recovery of traces of 228, 229rh(IV) in large amounts of 233u(VI)(containing 232y)
. separation 238pu/237Np
. separation 2«Am/2**Cm.
158
At the beginning of the programs these separations were realized usingliquid-liquid extraction (L.L.E.) techniques operated in mixer-settlers butten years ago the extraction Chromatographie technique was developed forpreparative purposes and is now applied for all chemicals separationsneeded for the production of actinides isotopes. That technique appears tobe simple and flexible. It can be used for the production of fig to kilogramamounts of actinide isotopes. This paper focuses on the experience gainedafter ten years and describes some peculiar production of actinide isotopessolved by using extraction Chromatographie techniques.
H EXTRACTION CHROMATOGRAPHY TECHNIQUEMore than ten years ago extraction chromatography technique was
principally used in analytical laboratories for the separation of smallamounts of metallic ions [ 1 ]. Its field of applications in now extended toprocess chemistry especially for nuclear applications £21 to [31. The mainadvantages of extraction chromatography techniques can be compared withthose of more conventional liquid-liquid extraction or ion exchangechromatography methods. Table II presents such a comparison j in spite ofa lower capacity than that possessed by ion exchange chromatography,extraction chromatography seems preferable due to the large versatility ofextracting functions available and the simple method which can be used forthe recovery of actinide trapped in the extracting material. The mainadvantage of extraction chromatography vs liquid-liquid extraction lies inthe simplicity of the device used to perform the separations. Of course thesmall exchange capacities of the columns used in extraction chromato-graphy make that technique only ideal for the production of figrams tokilograms amounts of actinides. For the production of tons amounts ofactinides, extraction chromatography cannot compete with liquid-liquidextraction.
^*\METHOD
CHARACTERISTICS^.
Versatility ofextracting func-tions
Exchangecapacity
Extractingdevice
Number of plates
Sensitivity of per-formance to opera-ting conditions
Recovery of theactinide trapped inthe extractingmaterial
Elimination ofalpha contami-nated material
EXTRACTION
CHROMATOGRAPHY
large
mean(adjustable)
simple(1 column+ 1 pump)
large
small
simpleA diluent washing ofthe column allowsthe recovery of theex tractant and of theactinide
easy(quasi solid
waste)
ION EXCHANGE
CHROMATOGRAPHY
mean
large
simple(! column* 1 pump)
large
small
very difficultA possible method isto calcine the resinand to dissolveashes in acid
easy(solid waste)
LIQUID-LIQUID
EXTRACTION
very large
mean(adjustable)
complex(mixer-s,ettlers)
» pumps)
small
large
simpleTechnique ofsolvent washing
difficult(large volumes of
alpha organicliouid wastes)
TABLE n Comparison of extraction chromatography, resin chromatographyand liquid-liquid extraction techniques for actinides separations
159
n.l. Extracting molecules and their usesThe choice of an extracting molecule suitable for extraction chromato-
graphy application depends on several criteria :. affinity for the actinide to be recovered
simple conditions for loading and elution are neededimpregnation conditions of the extracting molecule on the inertsupportlowest molecular weight possible allowing the highest exchangecapacity of the columnslow solubility in aqueous solutions.
X° FORMULA
C^HgO,.(D C^HcO— P— O
CfcHgO"'
Qcl^P-OC7H15-0-CH2
C6Hl3 Q. ,C2H5© P - CH2 - C - NC6Hl3 Ö"i Ö "C2H5
(o), ,CHbH9© ,P - CH2 - C - N
<£> o 0 *C<»H9
©CgHl 7 -NH*, N03-CgH I 7 '
I(CH3)2-CH-CH2]2-CH-0> /S© P.
[(CH3)2-CH-CH2)2-CH-O OH
@C8H17-0XS
C8H17-0 SH
ABBRE-VIATION
TBP
POX.ll
DHDECMP
<f<?
TOAHNO3
HD(DIBM)P
D2EHDTP
STATE
liquid
liquid
liquid
solid
solid
liquid
liquid
MOLE-CULAR(weight)
266
346
363
371
«16
350
3i4
TYPE OFAPPLICATION
/238pu(iv)/237Np(v)
*2*3Am/2*«Cm
2*lAm(!H)/transitionelements
2»lAm(IH)/transitionelements
2*lAm(in)/transitionelements
237Np«v)/238Pu(ni)
2«lAm<VI)/GdaiI), Ln(III)
23*U(VI>/238pu(iv)
TABLE m Extracting molecules and their uses in actinide separations(D tri-n butylphosphate
0 di-n-hexyloctoxymethylohosphine oxide0 di-n-hexyl diethyl carbamoymethvlene phosphorate
© di phenyl di-n-butyl carbamoyl méthylène phosphme oxide© tri-n-octylammomgm nitrate
© bis(2,6 dimethyl 4 heptyOphosponc acid@ di-2-ethyl hexyl phosphoro dithtoic acid
Table III presents some characteristics of extracting molecules used inour laboratory meeting some or all the criteria described above.
The simplest molecule used is the TBP for which the behavior for theextraction of actinides ions is well known from liquid-liquid extraction(L.LcE.) data. The transposition of the knowledge from L.L.E. to L.LoC«, issimple and easy. The high affinity of TBP for U(VI) and Pu(IV) present inacidic aqueous solutions makes L.L.E. technique using TBP extractingagent ideal for purification of uranium and plutonium. To obtain highdistribution coefficients for trivalent actinides (Am(III) and Cm(III)) withTBP it is necessary to use highly salted aqueous solutions which inducesliquid wastes difficult to manage, that is why a research of extractingmolecules having a high affinity for trivalent actinides in acidic media wasinitiated. In a first step POX.ll (a phosphine oxide) whose properties areclose to those of TOPO was studied. The use of POX.ll instead of TBP forthe extraction of Am(III) allows a decrease of the concentration of thesalting out reagent needed in aqueous solutions : 3.6. M instead of 7 to 8 MLiNO3. The main advantage of POX.ll vs TOPO lies in its liquid nature
160
allowing simple method for loading the extracting molecule on the inertsupport. In a second step search for molecules able to extract Am(III) fromnitric acid solutions without salting out reagent was pursued. For thisapplication DHDECMP, an extracting molecule developped in the U.S.A. [6]C7] andv^fe new molecule) were selected. At first the use of Win L.L.E.was considered difficult because of the solid nature of this compoundinvolving difficulties for the homogeneous loading on the inert support, butit was found that the solvate between <P<f> and nitric acid is liquid and easyto impregnate on the stationary phase.
In the case of TnOAHNC>3, preparation of the loaded material is easy :first liquid TOA is loaded on the support and when the column is filled withthis material, a flow of nitric acid through the column converts the aminein its salt. The high affinity of TOAHNO3 for tetravalent actinides fromnitric acid solutions makes this extracting agent suitable for the purifi-cation of Pu(IV) or Np(IV).
The use of HD(DIBM)P, a highly stericaliy hindered extractant isrestricted to the separation between Am(VI) and trivalent actinides orlanthanides. The selectivity of the separation is very high. The specialproperties of di-2-ethylhexylphosphorodithioic acid able to extract U(VI) ina synergetic combination with TBP from acidic solutions and to reducePu(IV), allow a very special separation U(VI)/Pu(IV) without the need toreduce plutonium to Pu(III) in the feed.
Ü.2. Inert supportThe properties needed for the inert support used in L.L.C. are :
commercial availabilitylow priceshigh specific areahydrophobic propertieshighest specific gravity possible.
Three different inert supports were used : Celite 5*5, Gas Chrom Q(Applied Science Lab) and hydrophobic silica (Merck). Celite 5*5 was madehydrophobic by treatment with dimethyl dichlorosiiane. The particule sizedistribution of Celite 5*5 and Gas Chrom Q is 110 to 1*0 n and theirspecific area ss. 1 m2.g-l. The highest specific gravity of the hydrophobicsilica makes this product more attractive than the two others, this permithighest extracting capacities for the column per unit of volume.
Ü.3. Preparation and characteristics of the stationary phasesImpregnations of the stationary phase with the extracting molecule are
carried out as follows t a certain mass of material is placed in contact witha solution of extractant in hexane or acetone, the solvent is thenevaporated under reduced pressure by means of a Buchi Rotavapor rotaryevaporator. Impregnation levels in mass in the final mixtures are respec-tively : TBP (27 %), POX 11 (30 %), DHDECMP (30 %), *V (30 %), TOA(25 %), HD(DIBM)P (30 %), HD2EHDTP + TBP (30 %). Batches of mixturecan be prepared at the 3 kg scale. Table IV presents the exchange capa-cities of the different stationary phases used, for the extraction of actinideat different oxidation states from nitrate solutions.
Ü.4. Chromatographie equipmentThe chromatograhic equipment includes :
plexiglass columns packed with the stationary phase impregnatedwith the extractantProminent proportioning pumpsstorage tanks for various solutionsgauge for in-line detection at the outlet of the column.
161
STATIONARYPHASE
TBP (27 %)
POX.ll (30%)
DHDECMP (30 %)
«V (30%
TOA (25 %)
HCKDIBM)P (30 %)
' D2EHDTP (20 %)(HA)TBP (10%)
EXTRACTED COMPOUND
.U02(N03)2. (TBP)24Pu<NO3)4. (TBP>2Um(N03)3. (TBP)3
,U02(N03)2 (POX.11)2•|piXNO3^ (POX.U)2lAm(N03)3 (POX.U)i,
Am(N03)j (DHDECMP)3
Am(NO3)3 fcM3
Np(N03)6(TOAH)2(TOAHN03)2
AmO2(D(DiBM)P))2 (HD(DiBMP))2
1 UO2A2-TBP
EXCHANGECAPACITYrn.moJe.g~l
0.5070.5070.338
0.1330.4330.216
0.275
0.269
0.177
0.21*
0.282
TABLE rv Exchange capacities of the different stationary phases used
The column is filled by successive additions and repeated packings ofthe dry stationary phase. Once the packing is completed and the top covercemented, the column is placed in the glove-box or the hot cell, where itundergoes pre-equilibrium designed to produce the chemical conditionsrequired for fixation and to expel the air.
Two kinds of gauges are used for in-line selection at the outlet of thecolumn s
. small Geiger-Muller tube for detection of 241 Am (60 keV)cerium doped glass scintillator for alpha detection.
Depending on the amount of actinide to be extracted the size of columnvaries from 100 g to 9 kg of stationary phase contained. The most usedcolumns contain 2.8 kg of stationary phase and possess a 0 ; 82 mm and aheight h ; 700 mm, their void volume is about 1.5 liters. Degradation of thecolumn can occur by washing or radiolytic damage of the extractant. Thecolumns the most sensitive to these phenomena are those loaded withDHDECMP used for the extraction of americium 2*1. Columns loaded withD2EHDTP are sensitive to oxydation by nitric acid media, thus it isnecessary to add an anti-nitrite agent (NH2SO3H) in the feed. Usedcolumns can be eliminated as alpha solid waste after removal of thesolution contained.
m SEPARATION AND PURIFICATION OF ACTINIDES ISOTOPES BYEXTRACTION CHROMATOGRAPHY
Instead of an overall description of a special process for the productionof an actinide isotope we will describe in the present chapter some peculiarpoints corresponding to different processes :
treatment of irradiated targetsrecovery of actinides from alpha active wastesrecovery of decay products from aged actinide stocks.
OLL Treatment of irradiated targetsIIL L L Neptunium 237 irradiated targets
This program was initiated in France in 1969 in order to prepareplutonium 238 batteries suitable for nuclear pacemakers. That productionof plutonium 238 was based on the irradiation of Np/Al or NpU2/MgOtargets in nuclear reactors, the chemical treatment of the irradiatedtargets after six months of cooling with a double aim : recovery ofplutonium 238 and of neptunium 237 for recycling. The overall process wasbased on liquid-liquid extraction technique operated in mixer-settlers afterthe dissolution of the targets in nitric acid solutions [83-tlOl.
162
In 1975 we started to test extraction chromatography as an alternativefor the L.L.E. process. Here we will focus on two experiments concerningthe difficult separation between plutonium and neptunium. Such a sepa-ration can be achieved by two systems based on :
. Valence adjustment, Pu(IV)/Np(V) then selective extraction of Pu(IV)by TOAHNO3 columns
. Valence adjustment, Np(IV)/Pu(III) then selective extraction ofNp(IV) by TOAHNO3 columns.
(i) Pu(IV)/Np(V) separation on TOAHNO3 columnIn nitric acid media the most stable oxidation states for plutonium and
neptunium are Pu(IV) and Np(V), it is thus possible to perform theseparation by selective extraction of Pu(IV). Such a separation faces with adifficulty : the Np(V) trend to disproportionate according to
2 NpO2+ + 4 H+ ^ Np^+ + NpO22+ + 2 H2O (1)
If in nitric acid of moderate concentration reaction (1) is shifted to theleft, in the presence of an extracting agent having affinities for Np^+,NpO22* or for both, reaction (1) is shifted to the right, thus the neptuniumis coextracted with the plutonium(IV).
To minimize this phenomenon and thus the efficiency of thePu(IV)/Np(V) separation the feed must be adjusted to a low nitric acidity(1 M) compatible with hydrolytic properties of Pu(IV). A dissolution liquorof an NpO2/MgO irradiated target of total volume = 23 1 : Np(V) = 11 g.l'1,Pu(IV) = 0.6 g.l'1 was adjusted to HNC>3 = l M and then injected on a.TOAHNO3/Gas chrom Q column containing 1015 g of stationary phase (0 =60 mm, H = 700 mm) with a flow rate equal to 2.5 l.rr1. After washing thecolumn with HNC>3 = 2 N at 2.5 l.h"1, plutonium 238 was eluted with asolution of following composition : r^SO^ = l M, HNO3 = 0.2 M.
Forteen grams of 238pu were recovered in one cycle, 237(sjp Was free of238pUj decontamination factor F.D. Pu(Np) was only moderate =20 butacceptable for a first cycle of Pu/Np separation, decontamination of 238pufrom fission products was high.
(u) NpOV)/Pu(m) separation on TOAHNO3 columnSeparation between neptunium 237 and plutonium 238 can be achieved
after reduction to Np(IV) and Pu(III) by selective extraction of Np(IV). Sucha separation is difficult to perform with high concentration ofplutonium 238 in solution because of the intense radiolysis leading to thedestruction of the excess of reductor used and thus to the oxidation ofplutonium to Pu(IV). Such a process can be used for final purification ofneptunium 237 contaminated with traces of plutonium 238. Recently thepurification of 300 grams of neptunium 237 contaminated with 17.5 ppm ofplutonium 238 (due to the large difference in the half lives of theseisotopes the alpha contribution of plutonium 238 was 30 %) was realizedusing the following procedure.
Feed The concentration of neptunium in the feed was 100 g.l~* in 6 MHNC>3, the reduction of neptunium to Np(IV) and plutonium to Pu(III)was done by addition of N2H5NO3 to 1.5 M and a warming of thesolution to 80°C during two hours. Thus after cooling the solution wasadjusted to 0.1 M in ferrous sulfamate (Fe(NH2SO3)2) freshly preparedunder nitrogen by dissolution of iron in sulfamic acid. The presence offerrous sulfamate in the feed allows the maintenance of plutonium asPu(III). The presence of small amount of iron(III) in the feed, induces asmall oxidation of plutonium to Pu(IV) and thus a decrease in theperformance of the separation.
163
The feed is then injected on a TOAHNÛ3 column containing fewpercent of octanol 2 in order to decrease a little the distributioncoefficient of tetravalent actinide and thus improve the quality of theseparation.
Washing The column loaded with Np(IV) was washed by HNÜ3 = 3 M toremove plutonium and iron ions present in the void volume.
Elution The elution of Np(IV) was performed with a solution : HNO3 =0.1 M ; H2$O4 = 0.75 M due to the complexing properties of SO^~ ions.In the purified neptunium the concentration of plutonium was 0.24 ppm.The decontamination factor for the entire process was F.D.fsjp(pu) = 73which is a good performance for this low level of 238pu contamination.
JH. 1.2. Plutonium 239 irradiated targets £ 11 j, [ 121This program was initiated for the production of americium 243 and
curium 244 isotopes. The Pu/Al targets containing initially 400 g of fissilematerial were irradiated with an integrated flux of 1L28 n.kb°l. Chemicaltreatment started after three years of cooling time.
The overall process is operated by L.L.C. after dissolution of thetargets in nitric acid which is realized using 88 liters. The highlyradioactive solutions contain 350 Ci.1-1 of /3,7 emitters and 44 g of 242pu .8.35 g of 243Am and 7.44 g of 244cm.
The first step of the process lies in the recovery and purification of theplutonium 242, which is realized by extraction chromatography using aTOAHNO3 column or a TBP column. To prevent clogging of these columnsby fission products precipitates, the feed is filtrated through a smallcolumn containing a bed of silica. After washing of the column with nitricacid solution the elution of the plutonium 242 is realized with a sulfonitricsolution (TOAHNO3 column) or a hydroxylammonium nitrate reducingsolution (TBP column).
The largest difficulty associated with that treatment lies sIn the separation of trivalent actinides (243Am, 244cm) fromtrivalent lanthanides present in the solution in large quantities(=140 g).In the separation Am/Cm. The difficulties have been solved byL.L.C. using a TALSPEAK like system, where TBP extractant wasused instead of HDEHP. Thus in this system the selectivity of theseparation is only due to the DTP A present in aqueous solution.
All the lanthanides possess a higher affinity than Am(III) and Cm(III) forTBP column from LiNOs = 8 M, HNC>3 = 0 + 0.05 M. DTPA = 0.1 M, pH =1.2 solution, the separation factor a= D£^ 3*.D^m 3+ is large for thelight lanthanides and is close to L4 for Tb and heavier Ln(IH). To achieve agood separation (Am, Cm) from Ln, it is necessary to perform severalcycles.
Separation between americium 243^and curium 244 is based for the firststep on the same procedure. Figure 1 presents the separation between1.51 g of curium 244 and 1.75 g of americum 243 using a TBP column and aTalspeak like system.
Final purification of americium 243 is performed by selective extrac-tion of Am(VI) on a HD(DiBM)P column using a process identical with thatdescribed below for the purification of americium 241. That processappears very simple and flexible, easy to operate in the difficult conditionsfound in hot cells. Until now about 100 g of americium 243 and =90 g ofcurium 244 of high purities have been successfully produced.
164
Concentrotion
2.0-
M.
1.2-
•
0.1-
-
(U-
-0
î"Cm or»«Am
(B.I-M
r™"
j —
»*Cm 1.5
11 — •
Si
!~- i
f iri i•j.' iTMllru
Tl
...
1.7'
— ,
•«)"1ii!~!. __ ,i ! i 1 _
' — "*' *" o^s ' ^ ^ ~"vi 1 vfm -1.573 i ¥,^""»3.251 Eluote volumt
Figure 1 Separation 2*3Am/2**Cm by extraction chromatographyon a TBP/Cas Chrom Q column
FEED = DTPA = 0.1 M, Ap* = 0.5 M, LiNO3 7 MSCRUB = LiNO3 g MELUTION = LiNO3 g M DTPA 0.1 M, pH 1.20
Column TBP (25 %) Gas Chrom Q, m = 500 g, L = 64 cm, <t> 4,2 cmFlow rate= 1.2 l.rrl
m.2. Recovery of actinides from alpha aqueous wastesIn 1978 a program was initiated with the goal of producing large
amounts of americium 2*1 used to prepare neutron sources(americium 2*l-beryllium or americium-lithium) for oil industry. The firstsource of americium 2*1 found in the C.E.A. consisted in alpha aqueousliquid wastes :
"Masurca" waste resulting from the reprocessing of certain irra-diated fuels and from criticaility analyses. Cadmium was added tothe solution for safety reasons and then its initial volume of about400 m3 was reduced to * m3 by distillation
. Waste coming from the reprocessing of fabrication scrap fromfabricating (U, PuK>2 fuels intended for fast breeder reactors. Theannual volume of this waste solution amounts to some tens of m^.
The americium 2*1 contents of these wastes were respectively108 mg.1-1 and 28 mg.1-1. As the uranium content of the "Masurca" wastewas high (12g.l~i) it appears that L.L.C. wasn't suitable for its recoveryprior to recovery of higher actinides. Thus conventional L.L.E. using TBPorganic solution in mixer-settlers was used. The neptunium 237 andplutonium 239 were recovered by ion exchange chromatography using ananionic resin (IRA *00) prior to the recovery of americium 2*1 by extrac-tion chromatography. In order to decrease the concentration of the saltingout reagent necessary to obtain high distribution coefficient of Am(III), di-nhexyloctoxymethylphosphine oxide (POX.11) Ü31 was selected as extrac-ting agent. A moderate concentration of LiNC>3 = 3.6 M was found suf-ficient to obtain high KrjAnidlI) (—J.lO^ml.g-*) for a nitric acid concen-tration of 0.1 M. The POX. 11 nevertheless has the disadvantage of extrac-ting Fe(III) with rather slow kinetics, which in the case of the "Masurca"waste gives rise to a sharp drop in Kr}Am(III). The addition of EDTA toaqueous solution in equal concentration to Fe(III) (0.2 M) helps to avoid itsextraction and to restore all the extractive properties of the POX. 11 forAm(III). This is due to the selective formation of the complex FeY~ (withYrfy = EDTA) whose formation constant log(FeY-) = 25.1 tl*3 is muchhigher than that of the americium(III) complex = log (AmY~) = 18.0 115]).
165
Experiments were realized to determine the exchange capacity ofPOX.ll vs Am3+, simulations of Am3+ by Nd^* was also used in order todecrease the irradiation of the chemist working in a glove-box. Figure 2,presents studies of saturation of a POX.ll column by Am(III) or by Nd(III)nitrates. In each case the breakthrough is obtained for a stoichiometricratio POX.11/M(III) close to * indicating that the extracted complexpossesses the following formula: M(NC>3)3 (POX.ll)^. That result wasconfirmed either by the study of POX.ll (30 %)/SiO2 mixture saturation intest tubes or by the saturation of a POX.ll dodecane solution. Such astoichiometry is very unusual because for trioctylphosphine oxide (TOPO)trivalent actinides or lanthanides are extracted as M(NO3)3.(TOPO>3. ThePOX.ll possesses a high affinity for U(VI) and Pu(IV), for these ionsstoichiometries are similar to those obtained with TOPO *(POX.l 1)2 and Pu(NO3)^ (POX.l 1)2.
ov; IS 20 2SVolum» (ml)
Co
_u )f-CH2-O.C7H1sC&H13 / ^0
c =o >,
g-u
Ov, 6 8 10Velum» ( m l )
Figxire 2 Saturation of POXol l/SiC>2 column with 241Am and Nd
Columns = l g POX.ll (29,7 %)/SiO7 0 = 5 mm, H = 80 mm,Void volume = 1.37 ml (Am**), 1,1 ml (Nd3»)
Aqueous phase s
CLiN03) = 3.6 M, IHNO3] i 0.1 M, [Am'»! = 3.81 g.l"1 or [Nd^+l = ! g.l-
Flow rates = 15.1 ml.h-1 (Am^*), 1.25 ml.h-l (Nd^*)
166
During the chemical treatment of the "Masurca" waste rather largePOX.ll column was used containing 9 kilograms of POX.ll (30 %)/SiO2stationary phase. A feed of 685 liters of total volume was treated, thedecontamination factor of the effluent in americium 2*1 was 250 ;americium 2*1 (17 g) was recovered by elution with 20 liters of nitric acid(concentration factor = 3*) with 89 % yield and its decontamination vsCd2+ and Fe3+ was close to 130 1161.
The major drawbacks encountered in the use of POX.ll process for therecovery of americium 2*1 from alpha liquid wastes lies in the rathercomplicated feed adjustment :presence of LiNC>3 3.6 M as salting-outreagent and of EDTA to complex Fe(III) present in the feed. That is whynew molecules able to extract trivaient americium from nitric acid mediawithout need of salt and possessing high affinity and selectivity fortrivaient actinides were considered. The first molecule investigated is di-n-hexyldiethylcarbamoyl-methylenephosphonate (DHDECMP) well studied forL.L.E. in the U.S.A. [5].
Distribution coefficient of Am(III) between nitric acid solutions andDHDECMP (30 %)/SiO? stationary phase were measured at low nitric acidconcentration KoAm^* = 0.9ml.g~l (HNC>3 = 0.1 M) and KpAm3+maximum = 26 ml.g-1 is obtained for HNO3 = 6 M. It is thus possible toextract trivaient americium from moderate volumes of 5 to 6 M HNÛ3 (dueto the relatively low value of K[)Am3-»- maximum) and to perform theelution with water.
Figure 3 presents an experiment related to the separation ofamericium 2*1 from fission products 13*, 137cs and lO^Ru ancj frominactive contaminants contained in a real waste. During the loading step
_t01
oÊEoO
»"Arn
0 0
(DHDECMP)
Figure 3 Purification at americium 241 from an industrial wasteby extraction chromatography
using a DHDECMP/SiOj column
5 V(ml)
Column = l g DHDECMP (30 %)/SiO2, « 5 = 5 mm, H = SO mm
'Ç) alpha active waste free from U and Pu (HNÛ3 - î N)
<2> HNO3 = 5 M
J3> H2O flow rate 2 ml.h" '
Aqueous
Phases
167
fission products and inactive contaminants escaped from the column. Afterthe washing of the column by a small volume of HNC>3 = 5 M, theamericium 241 is eluted with three void volumes of the column.Americium 241 was recovered with a yield of 98 % and after conversion toAmO2» its purity was found 85 % which is a very good performance for onlyone cycle. Adaptation of DHDECMP process to the purification of largeamounts of americium 2*1 (50 grams) was realized but the lives of thecolumns were found to be limited because of the washing and radiolyticdamage of the DHDECMP.
Another molecule able to extract trivalent americium from nitric acidsolution was studied °* the diphenyl di-n-butyl carbamoyl méthylènephosphine oxide (#9. High distribution coefficients of Am3+ were obtainedfrom nitric acid solutions with a maximum for HNO3 = 2.5 M. Theextraction reaction is :
, 3NO3- + 3^^Am(NO3)3$«03 (2)The usefulness of that molecule for the recovery of americium
from nitric acid solutions by L.L.C. was demonstrated in an experimentwhere a nitric acid solution (HNO3 = 2 M) containing 13,4 mg.l-i 241^mwas loaded on a small column containing l g of *PP (30 %)/SiC>2 stationaryphase. Four hundreds voids volume of the column were decontaminatedwith F.D. (Am3+) = 388. Americium 241 can be eluted from the <P<ffSiC>2column, after washing by water, with a K2CO3 solution.
ÏÏL3» Recovery of decay products from aged actiniae stocks
m.3.1. Recovery of americium 241 from aged plutonium metal or PuU2stocks
The recovery of americium 241 from alpha active wastes is a rathercomplicated method due to the large volumes involved and their lowconcentration in americium,, A more attractive way to produce hundreds ofgrams amounts of americium 241 is to treat aged stocks of plutoniummetal or plutonium dioxide initially rich in 241pu> -rne first step of theprocess consists in the dissolution of the starting material on 300 g scale innitric medium (containing F~ ion) under reflux conditions in a tantalumdissolver« After filtration the solution is injected on a TBP column forPu(IV) extraction, americium 241 escapes from the column.
After washing of the column with 5 M nitric acid, the plutonium iseluted as Pu(III) with a hydroxylamine nitrate solution and then convertedin PuO2« The raffinate containing the 241^m js tnen adjusted to thefollowing conditions: H+ = 0 + 0.05, Al(HI) = 1.8 M, DTPA = 0.1 M andloaded on a TBP column in order to decontaminate americium 241 fromimpurities (Pu, inactive metallic ions) and to reduce the volume of theamericiated solution. After a washing of the TBP column with 8 M L1NO3solution, americium is eluted with 2 M nitric acid. Final purification ofamericium 241 is performed with a method described in Figure 4. The firststep consists in the precipitation of Am(OH)3, which is redissolved afterfiltration, in concentrated K,2CC>3 solution.
The addition of K2S2Og allows the precipitation of americium(V),potassium double carbonate. After precipitation, the precipitate is takenup with an oxidizing acidic solution containing K2§2Og and Ag ions actingas catalyst in the oxidation of Am(V) to Am(VI) by S2Og—ions. Afterwarming to 50"C americium is totally converted to Am(VI) (a spectro-metric measurement is performed prior loading the solution into the L.L.C.column). The adjusted Am(VI) feed is then injected on a bis (2,6 dimethyl 4-heptyDphosphoric (HDDiBMP) acid column previously treated by an oxi-dizing solution (S2Og — and Ag(II) ions) until persistance of the dark color
168
N i t r i c eluotf from TBP column
T1 Pr tc i p i fa t ion 1l ^ M l O H J j J
1l Fi t t ra t ion
HNOi ^l ————— * ————— l*| D i Jîo lu t ion
1 M ( O H ) j
, ————— , , .,.,.,. .... ——— . /SpecK) CO) 5,5 M Oxidizing solution Imetri
K, S, 0 , 0 , 2 5 M HNO,1FIH HjPOiJ.IO^M l«>t AirA,l lO-'M ^ [Ag]lOJM KSS20, 0,2SM \^"'
^d) ' 1
P r e c i p i t a t i o n *" DHjO ^ Ka AmOj ICOj], ^
1 U^^Am SOj^^*J
Corhonott «ff lutnts / t?) \ f Htoting io 50«
Ln 111 + Am A/ 50 ü 100 mg,!"' ' ——————————— 'l a s t
Feed— ——————— AmVI ~ 2
HNO) 0,3 M H]pVg1] 10"*M K
»«
Washing (^HNOiO,3M —— T t
H) PO* «''M 1,51 fi-[Ag|lO-JK2SjO,L J 0,1 M
5 g.f1
POt 2.10"2MjSjO, 0,1M
€>(3^ Clution
, L T N,H4NO) 0,5 M~j * l HNO) IN
D
0^\ 1trophoto- B
c control ü.valence P
* \^/
1
HeatingAmV-»Amin
LQ,) i t otion 1cTJ
Fittrat ion Ag Cl
N H« OH ^ '"_ "-
1 A m ( Ci totion>H)| \
0<ata t«
Figure 4 Recovery of americium 241 from aged plutonium stocksFinal purification process for americium 211
of Ag(II) ions in the effluent. Extraction of Am(VI) is apparent on thecolumn by a greenish band. After washing of the column by an oxidizingsolution americium is eluted with a reducing solution (N?H5NO3 = 0.5 M,HN03 = 1 M).
Such a process is applied successfully on a 50 to 60 g. of 2*1 Am scale.The yield of the purification lies on the range 85 to 97 % for routineexperiment. Americium 2*1 purity of 98.5 % was measured by calorimetryafter transformation into AmC>2.
Laboratory experiments were performed in order to determine thestoichiometry of the extracted americium(VI) compound and thus theexchange capacity of the HD(DIBM)P columns. Figure 5 presents the curveobtained on the study of the saturation of a small HD(DIBM)P column withAm(VI), the concentration of 2*1 Am in the effluent is measured in linewith a small Geiger-Muller detector. The break through is obtained for anHD(DIBM)P/Am(VI) ratio equal to *, thus if HA stands for HD(DiBM)P theformula of the extracted americium(VI) compound is = AmO2A2(HA>2. Thesame stoichiometry was found for U(VI) extracted complex.
Using this process, about 500 g. of high purity 2*1 Am have beenproduced.
m.3.2. Recovery of uranium 23* from aged plutonium 23S stocksIn aged plutonium 238 dioxide, uranium 23* accumulates. Due to the
very different half lives of these isotopes (t $ = 87,7 years forplutonium 238 and 2.**.10^ years for uranium 23*) the recovery of highnuclear purity uranium 23* necessitates very efficient separation process.
169
V(mt)
Figure 5 Saturation of HIXDIBMjP/SiOj column with 2*lAm(Vî)
Column = stationary phase = 0.500 g of HD(DIBM)P (29,5« %)SiO2
Aqueous flNaNO3) s 2 M, IHNO3! * 0.1 M, [K2S2Og] = 0.2 M,Phase [_£AgJ = 2.1Q-2M, CAm(VI)J * 3.27 g.l-f
Flow rate = 6 ml.h-1-60 keV 2*'Am detection by in-line Geiger Müller gauge
Ten grams of plutonium 238 on the dioxide form were processed foruranium 234 recovery according to the following scheme s
Dissolution of PuO2 in HNO3 acid with F~ ion acting as catalyst.Precipitation of the bulk of plutonium 238 as oxalate,remains in the mother liquor.Change of medium by Pu(IV), U(VI) coextraction with TBP and backextraction with the addition of alpha bromocapric acid to theorganic phase towards H NO 3 = l N aqueous solution»Extraction chromatography on HD2EHDTP (20 %), TBP (10 %)/GasChrom Q column.
(i)(ii)
(iii)
(iv)
The extraction Chromatographie separations between 238pu ^d 234ywere performed using a column containing 20 g of stationary phase. Duringthe loading step of the feed, extraction of 23*U(VI) is realized by thesynergetic mixture : di 2 ethyl hexyl phosphorodithioic acid (HA)/tributylphosphate. The extracted complex which formula is ; UO2A2.TBP possessesan intense yellow color [17] . The di 2 ethyl hexyi phosphorodithioic acidpossesses reductive properties vs Pu(IV) thus plutonium 238 escapes fromthe column as blue Pu(III). A fraction of the column (at the top) isdestroyed by this reaction.
After washing of the column with a solution of composition HNO3 =1 N, NH2S03H = 0.1 M, 23*u(VI) is eluted with a 0,4 M (NH^C^solution. Forty six milligrams of uranium 23* were recovered and after twocycles of extraction Chromatographie separations the alpha contribution ofplutonium 238 in the purified uranium 234 was only 1.27 %.
Decontamination factors for extraction Chromatographie separationswere respectively : cycle 1 F.D. U(Pu) = 517 ; cycle 2 F.D. U(Pu) = 452.For the entire process F.D. U(Pu) = 4.5.107 was obtained and theuranium 234 recovery yield was 75 %.
170
IV CONCLUSIONSExtraction chromatography technique was developped in the last ten
years for actinides recovery, separations and purifications. Due to thelarge variety of organic extracting ligands available, that technique can beapplied successfully to solve difficult actinides separations. The processesdevelopped for the recovery and purification of microgram to kilogramamounts of actinides isotopes are easy to operate either in glove boxes orin hot cells. Nevertheless there is still some difficulties to solve :especially the extraction of trivalent americium from nitric acid solutionsfor which if some opportunities exist : uses of DHDECMP or W, theirperformances are not really ideal. More research is needed in this area.
REFERENCES
1 BRAUN, T., GHERSINI, G. Extraction Chromatography, Elsevier,Amsterdam (1975)
2 BUIJS, K, MULLER, W., REUL, 3., TOUSSAINT, 3.C., EUR 504e,(1973)
3 ESCHRICH, H., OCHSENFELD, W. Separation Science and Techno-logy 15 *(1980) 697.
4 WENZEL, U., MERZ, E., RITTER, G. Separation Science and Techno-logy 15 it (1980)733.
5 MARTELLA, L.L., NAVRATIL, 3.D., SABA, M.T. "Actinide Recoveryfrom Waste and low grade Sources" Harwood Academic Publishers,Chur, London, New-York (1982) 27.
6 Me ISAAC, L.D., BAKER, 3.D., KRUPA, 3.F., MEIKRANTZ, D.H.,SCHROEDER, N.C. A.C.S. Symposium Series 117, (1980) 395.
7 KNIGHTON, J.B., HAGAN, P.G., NAVRATIL, J.D., THOMPSON, G.H.A.C.S. Symposium Series 161 (1981), 53.
8 KOEHLY, G., MADIC, C„ BERGER, R. Proceedings of the Interna-tional Solvent extraction Conference, The Hague. Society ofChemical Industry, London (1971) 768.
9 for the entire program of production of 238pu see : B.I.S.T. 218,(1976).
10 SONTAG, R., KOEHLY, G. B.I.S.T. 218 (1976) 23.11 BOURGES, a., MADIC, C., KOEHLY, G. A.C.S. Symposium Series
117, (1980)33.12 KOEHLY, G., BOURGES, J., MADIC, C., SONTAG, R., KERTESZ, C.
A.C.S. Symposium Series 161, (1981) 19.13 GUILLAUME, B. Personal communication.14 GUSTAFSON, R.L., MARTELL, A.E. 3. Phys. Chem. 67 (1963) 576.15 MOSKVIN, A.I., KHALTURIN, G.V. GEL'MAN, A.D., Radiokhimiya 1
(1959) 1*1.16 MADIC, C., KERTESZ, C., SONTAG, R., KOEHLY, G., Separation
Science and Technology L5 4 (1980) 745.17 FITOUSSI, R., Ph D Thesis (1980) CEA-R-5152 (1981)
171
SYNTHETIC INORGANIC ION EXCHANGERSAND THEIR POTENTIAL USEIN THE NUCLEAR FUEL CYCLE
A. CLEARFIELDCollege Station,Texas A and M University,Texas, United States of America
Abstract
This paper describes the structure and principal ion exchangecharacteristics of acid salts of polyvalent metals and hydrousoxides. The acid salts can be prepared in crystalline (layered)form, as amorphous gels and in intermediate stages ofcrystallinity. Ion exchange behavior is then related to thecrystallinity of the exchanger. Separations of importance tonuclear technology are also described. The acid salts can beclassified into five types based upon structure. The hydrousoxides are considered to fall into two main types. Those such asZr02, Ti02 and Sn02 consist of small particles of oxide withhydrated surfaces and interparticle water. The nature of thesurface is described and related to the ion exchange properties ofthe hydrous oxide. Other hydrous oxides such as Sb20cj and MnQ2 aredescribed in terms of the pyrochlore and spinel structures,respectively. It is suggested that the properties of theseexchangers can be controlled to greater specificities throughsynthetic procedures.
1. IntroductionThe phenomenon of ion exchange in inorganic compounds is now
known to be very widespread. Among the major classes of compoundswe may cite
1. clays2. zeolites3. hydrous oxides4. acid salts of polyvalent metals5. heteropoly acid salts6. synthetic apatites7. insoluble ferrocyanides8. insoluble sulfides and sulfates9. condensed phosphates and titanatesIn addition a large range of miscellaneous exchangers is known
so that a systematization based on structural principles and
173
exchange behavior is highly desirable. In this paper we shallconfine ourselves to the acid salts of polyvalent cations and to alesser extent to certain hydrous oxides. Even here a bewilderingarray of crystalline and amorphous phases are known. Attempts atsystématisation and new directions for synthesis of exchangers withgreater specificities will be described.
The original discovery and subsequent exploitation of acidsalts of polyvalent cations grew out of the need for ion exchangerswith greater stability to elevated temperatures and the effects ofradiation damage [1,2]. These compounds were initially prepared asamorphous gels and an extensive literature on these gels developed[3, ]. In 196H gelatinous zirconium phosphate was crystallized andits layered nature demonstrated [5]. This compound, Zr(HPOjj)2oH201has now been designated as o-ZrP. Shortly thereafter a secondvariety of zirconium phosphate, Zr(HPOjj)2*2H20, was crystallized[8]. It has an interlayer distance of 12.2A compared to 7.6A foro-ZrP and is very likely a different layer type. It will bereferred to as Y-ZrP.
In the intervening time period, a large number of layeredinorganic ion exchangers have been prepared [7] so that it isuseful to classify them by structure type. In this we follow thescheme proposed by Alberti [8]: (1) Exchangers with a-layeredstructure; (2) Exchangers with Y-type layered structure; (3)Exchangers with fibrous structure; (H) Compounds with a three-dimensional structure; and (5) Acid salts of unknown structure.2. Exchangers of the a-Layered Type
2.1 PreparationA listing of some of the compounds with a-layered structures is
given in Table 1 [8]. They are prepared from their respective gelsby refluxing in strong phosphoric or arsenic acid. Thecrystallization is gradual and the degree of crystallinity of theproducts depends upon the concentration of acid used and the refluxtime [5,9]. This is shown in Fig. 1. To achieve an averagecrystallite size of 1-2ym for o-ZrP requires 1U days of refluxingin 12M H-sPOjj. The process may be greatly speeded up by autoclavingin 1 UM acid at 200-250°C» Large crystals can be prepared bydissolving a titanium or zirconium salt in HF and adding H^PO^,followed by slow evaporation at 60°C [10-12]. For most purposessmall crystallites, such as are obtained in 1-2 hr of refluxing maybe used.
174
Table 1. Some Important add salts of t et ravalentmetals with layered structure of the a-type.
Compound Formula Inter-layer Ion-exchangedistance (A) capacity meq/g
Titanium phosf
Zirconium phos
Hafnium phospt
Germanlum(IV)
T1n(IV) phospt
Lead(IV) phosf
Titanium arser
Zirconium arsi
T1n(IV) arseni
»hate T1(HP04)2-H20 7.56 7.76
iphate Zr(HP04)2-H20 "7.56 6.64
late Hf(HP04)2-H20 7.56 4.1.7
phosphate Ge(HP04)2»H20 7.6 7.08
late Sn(HP04)2-H20 7.76 6.08
»hate Pb(HP04)2-H20 7.8 4.79
täte T1(HAs04)2-H20 7.77 5.78
•nate Zr(HAs04)2-H20 7.78 5.14
ite Sn(HAs04)2*H20 7.8 4.80
Get
i i i t i t i i
0.5=48 A.
i i i i i i i i
0-8=48 A
i i i i i '" "i ' t " ' ' i
2.5:48 A
\ \ f i \ i \ T
J^_JLjL__i__i • -T • r i i i i - - i _.
4.5=48 1 |
. X -II A JLi i i i i i i i
I"'604 I 1 00212:48 1 1...A .AKir A1 1 1 1 I I 1 1
40 30 20 10ANGLE 20
X-ray powder patterns of a-zirconium phosphates of different degrees ofcrystallinity. Numbers above each pattern indicate concentration of H,PO,and time of reflux.(From Ref. 9, with permission.)
Fig. 1.
175
Amorphous gels are prepared by rapid addition of a soluble saltof the cation to the acid. Even if the acid is present in excess,the ratio of phosphate or arsenate to M(IV) is less than two [U].Stirring for several hours or heating is required to obtain astoichiometric product. The non-stoichiometric solids can berepresented as M(IV)(OH)z(HXOJt)2-z/2»YH20 where X - P, As and 0 S. ZS 2. The ion exchange behavior and other properties of these acidsalts depend upon the composition and degree of crystallinity.This will be discussed in more detail following a discussion of thea-structure.2.2 Structure of the a-layered compounds
The crystal structure of both o-zirconiura phosphate [13»1*0 andarsenate [15] have been determined. The layers consist of metalatoms which very nearly lie in a plane (±0.25A from the plane) andare bridged by phosphate (arsenate) groups. These are situatedalternately above and below the metal atom plane. Three oxygenatoms of each phosphate group are bonded to three differentzirconium atoms which form a distorted equilateral triangle. Thecoordination of the metal atoms is therefore six-fold. Theadjacent layers are arranged so as to form six sided cavities asshown in Fig. 2. Each of the oxygen atoms not bonded to a metalatom bears a proton and forms part of the cavity walls. The P-OHgroups hydrogen bond to the water molecules which reside in thecenter of the cavities [16]. There is just one mole of cavitiesper formula weight and the free volume into the cavities is that ofan unhydrated K+.
9Ï*
Fig. 2. Idealized structure of a-ZrP showing one of the zeolite like cavities created by the layer arrangement.(From Albert!, C. in "Study Week on Membranes," Fassino, R., Ed., Fontificiae Académie ScientarumScripta Varia, Rome, 1976, 629, with permission.)
176
pH
I 2 3 4 5 6 7meq OH"/ g a -ZrP
Fig. 3. Potentiometric titration curves for alkali metal ions on a-ZrP(12:336). The Cs curve was obtained on a less crystalline exchanger.(From Ref. 17, with permission.)
2.3 Ion Exchange Behavior of CrystalsThe ion exchange behavior of a-ZrP crystals is conditioned by
the tight binding of the layers and the small entranceways into theinterlayer cavities. This is illustrated by the ion exchangetitration curves of Fig. 3 [7,17]. Li+, Na* and K* are exchangedin acid solution to form half-exchanged phases of the typeZrMH(POn)2'XH20. The ions take up positions in the interlayerspaces which requires diffusion of unhydrated ions into the crystallattice. Thus, the selectivity favors the more weakly hydratedcation viz. K+ > Na+ > Li*. Rb* and Cs* do not exchange in acidsolution since, even in the unhydrated condition, they are toolarge to diffuse in between the layers. However on addition ofbase, it is thought that OH~ removes protons from the latticecausing the layers to open somewhat and permit diffusion of thelarger ions. In alkaline medium a 3/1* phase ZrM^ cHQ c(POjj)'XH20is formed rather than the half-exchanged type [17].
Exchange of ions into the crystalline layered group(IV)phosphates results in discontinuous changes in the interlayerdistances. The interlayer distance depends upon the type of phase,i.e., half, three-quarter or fully exchanged, the size of theentering cation and the amount of water taken up [7,18]. Thus, thehalf-exchanged sodium phase has 5 waters of hydration and aninterlayer distance of 11.8A while the fully exchanged phase,
177
Zr(NaPOJ+)2*3H20 has an interlayer spacing of 9.8A [19]. Thehydrated sodium ion is too large to enter between the a-ZrP layersand therefore must shed some of its waters of hydration to diffuseinto the layers. Rehydration of the Na* with expansion of thelayers then occurs for the pentahydrate. Diffusion begins at theedges and works toward the center so that at intermediate stages ofexchange both the unexchanged or proton form and the exchangedphase are present in the same crystal [18]. This mechanism resultsin irreversible exchange behavior in the forward versus thebackward directions [7,18].
The differehces in ion exchange behavior of crystallinegroup(IV) phosphates which have an a-layered structure arise fromthe different sizes of their unit cell dimensions. The smallerthese dimensions, the smaller the dimensions of the openings intothe cavities, and the greater the steric hindrance to cations.Furthermore, as the product of the cell dimensions within thelayers (a x b) decreases, the density of fixed charges (i.e.phosphate groups) increases .creating a greater barrier to iondiffusion. Therefore, higher pH values are required for exchangeto occur, i.e. to spread the layers apart. In the salt forms ofthe exchanger the forces holding the layers together areessentially coulombic. Thus, it is likely that these forces becomegreater as the density of fixed charges increases. The strongerthe forces holding the layers together, the more rigid they becomeand the more likely are the exchangers to exhibit ion-sieveeffects [8],
From the foregoing it is evident that cation exchange on thea-layered acids will not occur in highly acid solutions. Undermilder conditions where exchange takes place, the exchange behaviorbecomes complicated because of the formation of a second phase.This is usually a half-exchanged cation phase which has a largerinterlayer spacing than the original a-phase. Thus if a thirdcation is present, it may interact with the half-exchanged phase togive complicated phase relationships. A number of studies havebeen carried out to elucidate these phase relationships as anecessary prelude to Chromatographie separations [20,21]. Otherapproaches to utilizing the crystalline phases in Chromatographieprocedures involve expanding the layers prior to exchangingcations. These procedures will be discussed subsequently.2.4 Ion Exchange Behavior of Gels
Amorphous gels of zirconium phosphate behave very differentlyfrom the crystals. In the early literature very little attention
178
was paid to preparative methods and to proper characterization ofthe gels. This led to irreproducible behavior. The reasons forthis are now well understood [M,22]. If during preparation the molratio of phosphate to zirconium is less than two, precipitationstill occurs; but the product is a hydroxyphosphate,Zr(OH) (HPOjj)2_x/2« In sucn SQ^-3 tne lon exchange behavior of thehydroxyl groups must be considered as well as those of themonohydrogen phosphate groups. Even when the ratio of phosphate tozirconium exceeds two in the reactant mix, the precipitate whichfirst forms may have a ratio lower than two [22]. However on agingthe precipitate in the mother liquor, its phosphate contentincreases and in the limit P/Zr - 2. Ratios higher than thisindicate sorbed phosphate [M]. Greater resistance to hydrolysisand more reproducible behavior result if the gel is refluxed inabout 0.5M H^POii for 10 hours or more [9]. When dried thoroughlyover P20c the refluxed gels have the same composition as thecrystals, but they swell in water taking up an additional 4 molesof HpO. This property of the gels allows them to exchange ratherlarge cations as the interlayer distances are greatly increased bythe swelling.
A typical Cs+ titration curve for a gel that has been refluxedin 0.5M H^POjj for 48h is shown in Fig. *». In contrast to thecrystals where the pH remains relatively constant at ca. 7.5-8with ion uptake, each increment of Cs+ exchanged results in anincrease in pH. There are no breaks or endpoints in the curve andthe process is reversible [9]. This is similar to results obtainedwith weak acid organic resin ion exchangers except that the effectis much more marked. It indicates that the cation is uniformlydistributed in the gel to form a solid solution rather thanclustering in adjacent cavities to form a new phase as in thecrystals
The gel behaves as a weak field exchanger in the sense ofEisenman's theory [23] and the selectivity is Cs+ > Rb* > K* > Na+> Li* [24]. As the cation content increases, the selectivitieschange gradually in the way predicted by Eisenman's theory, thatis, to the strong field situation with Li* > Na* > K+ > Rb* > Cs+.The interlayer distance attains a maximum at relatively low cationuptake and then decreases with further loading [9]. Thus thecations are brought closer and closer to the negative sites of thelayers with consequent increase in field strength.2.5 Semi-crystalline Exchangers
By refluxing a gel in phosphoric acid of differentconcentrations and for different lengths of time, it is possible tobuild in any desired amount of crystallinity [5,9]. The particle
179
Fig. 4. Potentiometric titration curve for Cs uptake (Titrant 0.1NCsCl+CsOH) on amorphous zirconium phosphate (0.5:48). Solidline at left is phosphate release to the solution.(From Clearfleld, A., Oskarsson, A. Ion Exchange and Membranes.J_, (1974) 205, with permission.)
size of the refluxed product increases with increase incrystallinity, and the titration curves change -in a continuousfashion from the gel type to that exhibited by the crystals[4,9,25]. This is shown in Fig. 5. As the crystallinity increasesthe solid solution range decreases for each of the phases -formed.For example, with a gel that had been refluxed in 4.5M H^PO^ for 48hours, Na* is initially taken up with no phase change. At aloading of 0.25 meq/g the half-exchanged phase appears andincreases in amount relative to the sodium containing ot-phase untilat 3 meq/g load it is the only phase present. The compositionrange for the half-exchanged phase is approximately 3~3-45 meq/g.This may sound like a contradiction in terms, but by the half-exchanged phase for a semi-crystalline preparation we mean, a solidwhose X-ray pattern resembles that of the true half exchangedphase, Z r M ^ H C P O - X H O , but with broadened reflections.
180
11
10
9
8
7
6
. 5
4
3'
2
0 1 2 3 4 5 6 7 8Meq (OHf/g
Fig. 5. Titration curves for a-ZrP of low and intermediate crystallinity:0.8:48 -O , 2.5:48 -A, 4.5:48 -ft, 12:48 -|.(From Ref. 9, with permission.)
2.6 Cation SeparationsAs a guide to carrying out cation separations several
comprehensive determinations of distribution coefficients foramorphous zirconium phosphate have been published [26,27].Similiar values for amorphous zirconium tungstate and molybdatewere included [27]. The trouble with this data is that theexchangers were poorly characterized so that the results serve onlyas a rough guide. The effect of crystallinity of the exchanger on.measured distribution coefficients of divalent ions was measured byTuhtar [28]. His results are shown in Fig. 6. A dramatic decreasein Kd as a function of increasing crystallinity is observed duelargely to the exclusion of highly hydrated cations by the smallpassageways between the layers. The same trend is exhibited bytri- and tetravalent cations [28].
A rather complete summary of possible cation separations bygroup(IV) phosphate gels has recently been published [29]. In thispaper we shall confine ourselves to those separations related tonuclear processes. ^'Cs has been separated from nuclearreprocessing'solutions [30,31]. Distribution coefficients and some
181
lOOOO
1000
Kd
100
10
Co11
O Zn"O Mn"A Ni"A Co"
0.3:48 45:48
CRYSTALLINITY9:48 12:96
Fig. 6. Variation of distribution coefficient (at almost zero load)with increasing crystallinity of a-ZrP.
Chromatographie data for recovery of Cs+ and Sr2* from reprocessingsolutions has also been given [32,33]. Separation of Ba2"1" fromother divalent cations [34] and Cs* [35] has been reported.
Amorphous zirconium phosphate has received a great deal ofattention for separation of radioactive ions because of itsresistance to ionizing radiation, oxidizing media and strong acidsolutions [36]. The sorption of tracer concentrations of Am(III),Cm(III), Bk(IV), Cf(III), Ce(III), Eu(III) and U(VI) on ZrP withP/Zr - 1.3H was investigated at 75°C [37]. It was found thatseparation factors for ions with the same charge are low, whereasthose for ions with different oxidation states were large. It wassuggested that Bk(IV) could be separated from the trivalent ions.A good separation of americium from curium had also been reported[383.
A number of procedures are based upon the fact that betterseparations are effected on zirconium phosphate if the cationsdiffer in oxidation state. For example, Am(III) is separated fromCm(III) by first oxidizing it to Am(V). The Cm(III) ispreferentially retained on the column [38-40].
This general procedure has also been applied to neptunium-Plutonium separations [4M], Np(V) - Pu(IV) - Pr(III) [42] and
182
Ce(IV) - Pr(III) - Cm(III) separations [43]. It has been foundthat separations which take advantage of different oxidation statesmay be further enhanced by use of non-aqueous solvents [44]. Thesharp decline in Kd for the lower valent state contrasts with theincrease to a maximum for the higher valent state as the alcoholcontent goes to 50$. These differences are attributed to therelative stabilities of the chloro-complexes formed by the two ionsin the changing solvent composition. In fact it has led to amethod of determining the stability constants of these and relateduranium complexes as well as improved separations [45, 46].Distribution coefficients for Th, Pa, U, Np and Pu were alsomeasured and procedures for their separation outlined [4?].Separation of uranyl ion from several di- and trivalent cations hasalso been described2.7. Nuclear Waste Processing and Water Purification.
The original search for new inorganic ion exchangers wasinstigated by the desire for materials stable in intense radiationfields and at high temperatures. Early efforts in this directionhave already been mentioned [1-3,35]. Zirconium phosphate hasindeed been found to be stable to radiation [49,50], but exhibitsincreased hydrolysis at elevated temperatures. Resistance tohydrolysis is increased by use of semicrystalline zirconiumphosphate as shown by Ahrland and Carleson [51]. Even so, a mixedhydrous zirconia-zirconium phosphate column was used to preventphosphate bleeding during purification of dilute solutionscontaining transition metal ions.1 37A method for separating '°'Cs from fission products has beenused by the French Atomic Energy Commission which utilizes a mixedbed of ZrP and ammonium phosphotungstate [52]. We have alreadymentioned the separation of uranium from plutonium on amorphouszirconium phosphate with P/Zr - 1.34. The effect of added fissionproducts was also assessed.
Other interesting applications of zirconium phosphate includethe removal of 10°Ru and 95Zr from purex solvent [53] andseparation of parent -daughter radioisotopes such as ^Cd - 11^mln,^Ce- 1J+V and 210Pb - 210Bi [54].2.8 Ion Exchange with Expanded Layer Zirconium Phosphates
From the earlier discussion on crystalline a-ZrP it is evidentthat its usefulness for separations is severely limited. However,it turns out that it is relatively easy to increase the inter lay erdistance of ot-ZrP by a number of processes. In the expanded layerforms the barrier to diffusion of large cations is drastically
183
reduced. Therefore, exchange processes can take place without theuse of added base. The half-exchanged sodium form,ZrNaH(P04)2'5H20 has an inter-layer distance of 11.8A [19]. Itbehaves as a monofunctional exchanger with a capacity of 2.53 meq/g[55]. Kornyei and Szirtes have reported [56] the separation ofstrontium from radioactive wastewater on ZrNaH(POjj)2«5H20,
If sufficient acid is added to remove the sodium ion fromZrNaH(POj,)2'5H20, a-ZrP is not obtained but rather a higher hydratecontaining 5-6 moles of H20 and named 0-ZrP [19,57]. Thisexchanger has an interlayer spacing of 10.4A so that it acceptslarge cations more readily than ct-ZrP. This is shown in Fig. 7where potentiometric titrations for alkaline earth cations on a-ZrPand 8-ZrP are compared [58]. As yet no cation separations havebeen carried out with 9-ZrP, but this would seem to be a fruitfularea for investigation.
PERCENT OF CONVERSION25 50 75 100
2.0 -
I.O 2.0 3.0 4.0 5.0UPTAKE meq M2*
Fig. 7. Potentiometric titration curves for alkaline earth cations on 6-ZrP(solid lines) and a-ZrP (dashed lines).(From Ref. 58, with permission.)
Another way of expanding the layers of a-ZrP is by amineintercalation [59]. For example, in the butylamine intercalate theinterlayer spacing is 18»6A and this is large enough to put complexions in between the layers [60,61]. In this case the amine acts tokeep the pH high so that no added base is required to load ionsinto the exchanger. Up to U-5 meq/g of divalent ions may beexchanged and multiple ions can be loaded. Thus it would seem thatthis could be the basis for many cation separation techniques.
184
3. Exchangers of the Y-Layered TypeOnly two members of this series are known; Y-ZrP [6] and Y-TiP
[62]. These compounds are not merely more highly hydrated forms ofo-compounds but have different, as yet unknown, types of layers[63,64]. Because the interlayer distances of these two compoundsare large, 12.2A and 11.6A, respectively, they are able to exchangelarge cations and complexes [65,66]. However due to the closenessof the fixed charges, hydrolysis becomes severe at high levels ofexchange. A highly crystalline form has recently been prepared[67] which may be less susceptible to hydrolysis. The selectivitysequence at low to moderate loadings as obtained from titrâtiondata is K* > Rb* > Cs* > Na+ > Li* [65]. I view of this sequence,it is surprising that an extremely high affinity for Cs* by Y-ZrPhas been reported [68], high enough to recommend its use in reactorwater cleanup.4. Acid Salts With Fibrous Structure
Members of this group include cerium(IV) phosphate (CePf) [69]thorium phosphate (ThPf) [70] and titanium phosphate and arsenate[8,71]. These fibrous compounds have been used to prepareinorganic ion-exchange papers, or thin layers, suitable for fastseparations of cations [72]. The fibers are prepared by addingdropwise 0.05M Ce(SOij)2-4H20 in 0.5M HjSO to a well stirredsolution of 6M H^PO^ previously heated to 95°C. Flexible sheetscan be obtained by dispersing the fibrous precipitate in a largevolume of water, then filtering the slurry in a large Buchnerfunnel. The formula of CePf is probably Ce(HPOjj)2*H20, but itsstructure is unknown.5. Acid Salts with Three-Dimensional Structure
This class includes compounds with the general formulaAM2(IV)(X01|)3 [(M(IV) - Ti, Zr, Th, Ge; X - P, As and A - anyunivalent cation]. We have prepared these compounds by thermaldecomposition of exchanged forms of ct-ZrP [7,19] and byhydrothermal means [73,7*0. HZ^CPO^ may be prepared by treatingLiZr2(POjj)o with acid [8] or thermal decomposition of NH^Zr^CPO^),[74]. The structure of these triphosphates is shown in Fig. 8[75]. It is built up of Zr bridged to Zr through 3 phosphategroups. This arrangement produces M(IV) octahedra linked bycorners to the tetrahedra the whole forming a three dimensionalnetwork containing cavities which are of two types.
There are four cavities per formula, one of which is occupiedby Na+. Silicate can replace part or all of the PO " and thisrequires additional Na+ ions to progressively fill up the remainingcavities [76,77]. The silicate ions enlarge the cavitiessufficiently to allow relatively free movement of the univalent
185
Fig. 8. Projection of half the unit cell of NaZr.(PO,), onto the ac plane, showingthe cavity built up by the Zr.CPO,)- groups. Numbers next to the atomsindicate height (in hundredths) above the zero level plane.(From Ref. 76, with permission.)
cations when about 2/3 of the phosphate groups are replaced. TheNa+ ions can now be exchanged by Li4" or H*„ Similiar Li^-H*exchange reactions are possible with HZrfPOjjK. Since the cavityand passageway sizes and exchange capacity can be modified over abroad range by choice of M , the XO^ anion and substitutions forM , it should be possible to prepare exchangers in this systemwhich are highly selective and stable to high temperature,radiation fields and a broad range of pH. This idea is extended toother systems in the next section.6. Hydrous Oxides and other "Amorphous" Exchangers
A number of hydrous oxides including Ti02» Zr02, Mn02 andantimonic acid have been examined for use in nuclear processes.The exchange characteristics of these compounds has been summarizedand reviewed in a recent book [73. Hydrous titania readily sorbsuranium [78] and has been considered for extraction of uranium fromse'a water [79,80]. Hydrous oxides were also used in thepurification and isolation of neptunium, plutonium [81,82]americiura and berkelium [83].
186
A BFig. 9. Schematic representation of hydrous oxide particles. A. ZrO.'nH-O
with hydroxyl ion in the intercrystalline water neutralizing thepositively charged particle surface; B. SnO2"nH20 with H3° in
the intercrystalline water neutralizing the negatively chargedsurface.
A simple structual model of hydrous oxides of the Ti02, Zr02,Sn02 type has recently been given [84]. The solid precipitated byammonia consists of particles ca. 25-30A in diameter. In theinterior of these particles the structure is that of a crystallineanhydrous oxide [85]. These small particles are joined together toform agglomerates containing large cracks and fissures. The fullmetal coordination is preserved at the surface by partialcoordination of HpO. The remaining water molecules form aninterparticle solution, the acidity of which depends upon thedegree of protonation of the surface oxygens. By aging orrefluxing it is possible to obtain more crystalline forms of theseoxides. A detailed description of how amorphous and crystallinehydrous zirconia are formed from solutions of the tetramericzirconyl ion has been given by Clearfield [86]. The formula ofthese hydrous oxides is best represented [84] as[MOa(OH)b(H20)R a+b]1|~(2a"('b) + S. On the assumption that all themetal atoms are fully coordinated, the oxygen-metal ratio, R, willbe greater than 2. For the fluorite structure (Zr02) a 25A cubicparticle will have R - 2.15. The extra oxygen atoms at the surfacemust be supplied by water molecules. The remaining watermolecules, S, form the interparticle solution. Whether S is acidic(Sn02) or basic (Zr02) depends on the degree of protonation of theoxygens at the particle surface. This is illustrated schematicallyin Fig. 9. In the basic limit, the surface oxygens are supplied byterminal H20 and metal bridging OH~; in the acidic limit theoxygens are bare 02~ ions making the particle charge negative withcompensating HoO* in the interparticle solution.
187
I 2 3 4 5 6 7 8 9 10 « \?.pH
Fig. 10. Ion exchange capacity versus pH for hydrous tin and zirconium oxides.
Based on the foregoing picture of hydrous oxides, hydrous tinoxide should readily exchange cations in acid solution because H^O*is present in the interparticle solution. A more basic oxide suchas that of zirconium would require higher pH values to exchangecations because the OH" in the interparticle water must beneutralized first before surface protons will detach from thesurface oxygens. The opposite is then true for anion exchange.This is illustrated in Fig. 10 [35]. The exchange capacity forcations will increase with increasing pH but in the limit dependsupon the total removable protons on the surface. The selectivityfor a particular ion will not only depend upon these factors butalso the complex formation constant of the surface-ion complex.
In contrast to the foregoing picture crystalline antimonic acidexchanges only cations and exhibits the anomalous selectivitysequence [87] Na+ > K+ > Rb* > Cs+ > Li*. This is attributed tothe fact that antimonic acid has a pyrochlore structure with an(SbgOg)2"" framework with almost 3H20 distributed over the ApO*interpenetrating framework [84] . The water molecules would beexpected to be in the I6d and 8b sites with two protons in the A*sites. Thus, the formula SbjO^'S^HgO is structurally representedas (H O SbjOg'O.gHjO in analogy with the A2B2Og pyrochlorerepresentation. There is no interparticle water as it is all inthe crystal lattice. The theoretical exchange capacity, 5.08 meqg , matches the experimental ones 5.1 meq g~ „ Furthermore,antimonic acid behaves as a strong acid exchanger with a relativelyconstant exchange capacity over a large pH range.
Recently a layered form of HSbOg'XHgO was prepared by ionexchange from the ilmenite form of KSbO^ [88]. Similiar reactionswere carried out to prepare HNbO^ and HTaO^. When the startingphases were the pyrochlore thallium salts, heating them in 3N HNO,
188
Fig. 11. Potentiometric titration of hydr.ous MnO_ by the static method. Titrant0.1M (MCI + MOH) with M - Li* -Q . Na* - A and K* - Q .
converted them to the corresponding acids HNbO^'XHgO and HTa03«XH20with retention of pyrochlore structure [88], However if lithiumniobate or tantalate were used as starting phases the acid productshad a perovskite structure [89].
We have recently examined hydrous manganese dioxide preparedelectrolytically [90], The exchange capacity depends upon thetemperature of preparation but for most samples is about U.6 meq/g.The exchanger behaves in some respects as a hydrous oxide, itscapacity changing with pH; but it also exhibits ion sieve behavior.This is shown in the titration curves in Fig. 11. The exchangecapacity for K* is about 1/3 the theoretical capacity and muchlower than that for either Na* or K*. However in acid solution,the distribution coefficients indicate the selectivity sequence K*>Na* > Li*. We interpret this data to indicate that the exchangercontains cavities with different sizes. The largest cavities,perhaps those on the surface prefer the largest unhydrated ion, K*.A second group of intermediate sized cavities can accomodate Na* inacid solution and finally in basic solution Li* becomes the mostpreferred cation. However when the fully exchanged Li* form ofhydrous manganese dioxide was heated at 500°C, it formed a weaklycrystalline A-Mn02 and on washing out the Li* with acid showed aremarkably high selectivity for Li*. This A-Mn02 is related to thespinel LiMn201( [91] and Thackeray, et al. have shown that Li* canbe replaced by H* in this spinel [92]. All of this suggestspossible routes to the synthesis of new inorganic ion exchangers
189
with high selectivities for individual ions and great temperatureand radiation stability. We have mentioned that pyrochlores,spinels, illmenites, perovskites and triphosphates can all beconverted to acid forms and that certain hydrous oxides can take onthese crystalline structures. By synthesizing many of these phaseshydrothermally, or from gels, in the presence of a particular ionit may be possible to alter the structure or size of the cavitiesor tunnels to be highly specific for ions which easily fit into thestructure. A case in point is zirconium phosphosilicate. It isprepared by treating a zirconium silicate gel with H^PO^ [93]°This exchanger has been examined at length for Sorption andseparation of lanthanides [94] and for separation of neptunium fromberkelium [96]. Decontamination factors of 3000 and 200 wereobtained for the separation of Pu from 5Zr + ^5Nb and 10°Ru,respectively [96]. The large change in distribution coefficientswhich this exchanger exhibits with pH change suggests its amorphoushydrous nature. However we have succeeded in converting such gelsinto crystalline triphosphate-like phases under mild conditions.The ion exchange behavior of these exchangers is under activeinvestigation.
7. AcknowledgementThe work carried out in the author's research laboratory was
supported by the National Science Foundation under grants numberedCHE81-14613 (Chemical Division) and DMR80-25184 (Division ofMaterials Research) for which grateful acknowledgement is made.
8. References[1.] RUSSELL, E.Ro, ADAMSON, A.W., SHUBERT, J., BOYD, G.E.,
"Decontamination of product by adsorption on tailor-madeinorganic adsorbents", U.S.A.E.G. Rep. CN-508 (1943).
[2.] KRAUS, K.A., PHILLIPS, H.O., J. Amer. Chem. Soc. 78_ (1956)644.
[3.] AMPHLETT, C.B., Inorganic Ion Exchangers, Elsevier,Amsterdam (1964).
[4.] AHRLAND, S., ALBERTSSON, Jo, JOHANSSON, L., NIHLGARD, Be,NILSSON, L., Acta Chem. Scand. 1^ (1964) 1346, 1357.
[5.] CLEARFIELD, A., STÏNES, J.A., J. Inorg. Nucl. Chem. 26_(1964) 117.
[6.] CLEARFIELD, A., BLESSING, R.H., STYNES, J.A., J. Inorg.Nucl. Chem. 30 (1968) 2249.
190
[7.] CLEARFIELD, A., "Zirconium phosphates", Inorganic IonExchange Materials (Clearfield, A., Ed.) CRC Press, BocaRaton, Florida (1982).
[8.] ALBERTI, G., Ref 7, Ch. 2.[9.] CLEARFIELD, A., OSKARSSOM, A., OSKARSSON, C., Ion Exch.
Membr. (1972) 9.[10.] ALBERTI, G., TORRACCA, E.J., Inorg. Nucl. Chem. 3p_ (1968)
317.[11.] ALBERTI, G., COSTANTINO, U., GIULIETTI, R., J. Inorg. Nucl.
Chem. 42_ (1980) 1062.[12.] INOUE, Y., YAMADA, Y., Bull. Chem. Soc. Jpn. 52. (1979) 3528.[13.] CLEARFIELD, A., SMITH, G.D., Inorg. Chem. £(1969) 431.[14.] TROUP, J.M., CLEARFIELD, A., Inorg. Chem. Jl_6 (1977) 3311.[15.] CLEARFIELD, A., DUAX, W.L., Acta Cryst. B25 (1969) 2658.[16.] ALBERTSSON, J., OSKARSSON, A., TELLGREN, R., THOMAS, J.O.,
J. Phys. Chem., 8l_ (1977) 1574.[17.] KULLBERG, L., CLEARFIELD, A., J. Phys. Chem. 85_ (1981)
1585.[18.] ALBERTI, G., Ace. Chem. Res., 1J_ (1978) 163.[19.] CLEARFIELD, A., DUAX, W.L., MEDINA, A.S., SMITH, G.D.,
Thomas, J.R., J. Phys. Chem. 73_ (1969) 3424.[20.] KULLBERG, L., CLEARFIELD, A., J. Inorg. Nucl. Chem. 43_
(1981) 2543.[21.] KULLBERG, L., CLEARFIELD, A., Solvent Extraction and Ion
Exchange J_ (1983) 77.[22.] AHRLAND, S., ALBERTSSON, J., JOHANSSON, L., NIHLGARD, B.,
NILSSON, L., Acta Chem. Scand. Ij (1964) 707.[23.] EISENMAN, G., Biophys. J. 2_ (1962) 259.[24.] KULLBERG, L., CLEARFIELD, A., J. Phys. Chem. 85 (1981) 1578.[25.] CLEARFIELD, A., KULLBERG, L., OSKARSSON, A., J. Phys. Chem.
78_ (1974) 1150.[26.] AKATSU, E., ONO, R., TSUKUECHI, K., UCHIRYOMA, H., J. Nucl.
Sei. Tech. 2_ (1965) 141.[27.] MAECK, W.J., KUSSY, M.E., REIM, J.E., Anal. Chem. 35_ (1963)
2086.[28.] TUHTAR, D., Ph.D. Thesis Ohio University Athens (1975).[29.] CLEARFIELD, A., Sep. Sei., in press.[30.] PROSPERT, J., "Some Properties of Zirconium Phosphate",
Rappt. CEA-R '2835 Commissariat A L'Energie Atomique Par is( l966) .
[31.] PROSPERT, J., LEFEVRE, J., RAGGENBASS, A., "Recovery ofcesium-137 using a mineral ion exchanger", Communauté Eur.Energ. At. EURATOM EUR-4075 (1968).
191
[32.] BAETSLE, L.H., VAN DEYCK, D., HUYS, D., QUERY, A., "The useof inorganic ion exchangers in acid medium for the recoveryof Cs and Sr from reprocessing solutions", AEC Accession No.7613 Kept. No. BLG-267 (1964).
C33.3 BAETSLE, L. HUYS, D., J. Inorg. Nucl. Chem. Q (1961) 133.[34.] NUNEZ-CORTES, M.P.M., SANCHEZ-BUTANERO, P., Afinidad 37
(1980) 451.[35.] KRAUS, K.A., PHILLIPS, H.O., CARLSON, T.A., JOHNSON, J. S.,
"Ion exchange properties of hydrous oxides", 2nd UN.Conference on Peaceful Uses of Atomic Energy, Geneva (1958)1832.
[36J GAL, I., RUVARAC, A., J. Chromatogr. J_3 (1964) 549.C37-] HOROWITZ, E.P., J. Inorg. Nucl. Chem. 28_ (1966) 1469.[38 J MOORE, F.L., Anal. Chem. 43. (1971) 487.[39.3 SHAFIEV, A.I., EFREMOV, Yu.V., NIKOLAEV, V.M., YAKOVLEV,
G.N., Radiokhimiya _13_ (1971) 129.[40«] SHAFIEV, A.I., EFREMOV, Yu.V., ANDREEV, V „ P o p Radiokhimiya
15 (1973) 265.[41.] CHERNORUKOV, N.G., KORSHUNOV, I.A., PROKOF'EVA, T.V., ZHUK,
M.I., MOSKVICHEV, E.P., S'ezd Obshch. Prikl. Khim j_ (1975)270o
[42.] SHAFIEV, A.I., EFREMOV, Yu.V., NIKOLAEV, V.M., YAKOVLEV,G.N., Radiokhhimiya V3 (1971) 126.
[43°] SHAFIEV, A.I., EFREMOV, Yu.V., Radiokhimiya U_ (1972) 771.[44.] SOUKA, N.f ABDEL-REHIM, F., J. Radioanal. Chem. £3 (1974)
43.[45.] ALY, H.F., ABDEL-RASSOUL, A.A., ZAKAREIA, N., Z. Physik.
Chem. Neue Folge. 9£, (1975) 11.[46.] ALY, H.F., M AR EI, S., ZAKAREIA, N., Mikrochim. Acta J_
(1973).[47.] SOUKA, N., ABDEL-GAWAD, A.S., SHABANA, R., FARAH, K.,
Radiochim. Acta 22_, (1975) 180.[48J GAL, !.. PERIC', N., Microchim. Acta 25J_ (1965) 251.[49.] BORN, E., PAUL, G., Z, Kristallogr. 146 (1977) 19.[50.] ZSINKA, L., SZIRTES, L., MINK, J., KALMAN, A., J. Inorg.
Nucl. Chem. 36, (1974) 447.[51.] AHRLAND, S., CARLESON, G., J. Inorg. Nucl. Chem. 33_ (1971)
2229.[52.] RAGGENBASS, A., COUROUBLE, J.M., LEFEBVRE, J., FRADIN, J.,
PEREBASKINE, C., "ELAN: Installation for the recovery andpackaging cesium-137 in France", AEC Accession No. 39319,Rep. DP-1066, Vol. 2, Paris (1966).
192
[53.] TURCANU, C.N., RADU, D., Radiochera. Radioanal. Letters 3_(1980) 245.
[54.] BHATTACHARYYA, D.K., BASU, S., J. Radioanal. Chem. 4£ (1978)105.
[55.] ALBERTI, G., COSTANTINO, U., GUPTA, J.P., J. Inorg. Nucl.Chem. 36_ (1974) 2103.
[56.] KORNYEI, J., SZIRTES, L., Izotoptechnika, 23_ (1980) 243.[57.] CLEARFIELD, A., LANDIS, A.L., MEDINA, A.S., TROUP, J.M., J.
Inorg. Nucl. Chera. 35_ (1973) 1099.[58.] ALBERTI, G., COSTANTINO, U., GILL, J.S., J. Inorg. Nucl.
Chem. 38. (1976) 1733.[59.] COSTANTINO, U., "Intercalation behavior of group IV layered
phosphates", Inorganic Ion Exchange Materials (Clearfield,A., Ed.) CRC Press, Boca Raton, Florida (1982) chap. 3«
[60.] CLEARFIELD, A., TINDWA, R.M., Inorg. Nucl. Chera. Letter. J_5(1979) 251.
[61.] QUAYLE, L.R., CLEARFIELD, A., Inorg. Chem. 21_ (1982) 4197.[62.] ALLULLI, S., FERRAGINA, C., LAGINESTRA, A., MASSUCCI, J.A.,
TOMMASSINI, N., J. Inorg. Nucl. Chem. 39. 1977 1043.[63.] YAMANAKA, S., TANAKA, M., J. Inorg. Nucl. Chem. 4]_ (1979)
45.[64.] ALBERTI, G., COSTANTINO, U., GIOVAGNOTTI, M. L., J. Inorg.
Nucl. Chem. 41_ (1979) 643.[65.] CLEARFIELD, A., GARCES, J.M., J. Inorg. Nucl. Chem. 4j_
(1979) 879.[66.] CLEARFIELD, A., KALNINS, J.M., J. Inorg. Nucl. Chem 40_
(1976) 1933.[67.] YAMANAKA, S., Inorg. Chem. 1_5 (1976) 2811.[68.] KOMARNENI, S., ROY, R., Nature 299 (1982) 707.[69.] ALBERTI, G., COSTANTINO, U., DI GREGORIO, F., GALLI, P.f
TORRACCA, E., J. Inorg. Nucl. Chera. 3£ (1968) 295.[70.] ALBERTI, G., COSTANTINO, U., J. Chromatog. 50 (1970) 482.[71.] ALBERTI, G.,' COSTANTINO, U., ALLULLI, S. TOMASSINI, N. , J.
Inorg. Nucl. Chem. 40_ (1978) 1113.[72.] A1BERTI, G., MASSUCCI, M.A., TORRACCA, E., J. Chromatog. 30_
(1967) 579.[73.] CLEARFIELD, A., JERUS, P., COTMAN, R.N., PACK, S.P., Mat.
Res. Bull. 1_5 (1980) 1603.[74.] CLEARFIELD, A., ROBERTS, B., SUBRAMANIAN, M.A., Mat. Res.
Bull. ±9_ (1984) 219.[75.] HAGMAN, L.O., KIERKEGAARD, P., Acta. Chem. Scand. ]_2_ (1968)
1822.
193
[76.] HONG, H.Y-P. Mat. Res. Bull. JJ_ (1976) 173«[77.] GOODENOUGH, J., HONG, H.Y-P.. KAFALAS, J.A., Mat. Res. Bull.
1_1_ (1976) 20.[78.] YAMASHITA, H., NAKAJIMA, F., Bull. Chem. soc. Jpn. ISSN
0009-2676, 53. (1980) 1 and 53. (1980) 3050.[79.] KANNO, M. "Extraction of uranium from seawater", Rep. IAEA-
CN 36/161, Vienna (1977).[80.] BETTINALI, C., PANTANETTI, F., "Uranium from seawaters
possibilities of recovery exploiting slow coastal currents",Proc. Advis. Group Meet., 1975 IAEA, Vienna (1976).
[81.] KORSHUNOV, I.A., CHERNORUKOV, N.G., PROKOF'EVA, I.V..Radiochimya (1976) 5.
[82.] KENNEDY, J., PECKETT, J.W., PERKINS, R.t "The removal ofplutonium and certain fission products from alkaline mediaby hydrated titanium dioxide", U.K. At. Ener. Auth. Res.Group At. Ener. Estab. Rep. AERE-R-4516 U.K. At. EnergyEstablishment, Harwell (1964).
[83J MYASOEDOV, B.F.,Usp. Anal. Khira. 148 (1974).[84.] ENGLAND, W.A., CROSS, M.G., HAMNETT, A., WISEMAN, P.J.,
GOODENOUGH, J.B., Solid State Ionics J_ (1980) 231.[85J CLEARFIELD, A., Inorg. Chem. 3. (1964) 146.[86.] CLEARFIELD, A., Rev. Pure Appl. Chem. _1_4 (1964) 91.[87o] ABE, M«, J. Inorg. Hucl. Chem. 4j_ (1979) 85.[88.] CHOWDHRY, U., BARKLEY, J.R., ENGLISH, A.D., SLEIGHT, A.W.,
Mat. Res. Bull. VT_ (1982) 917.[89.] RICE, C.E.-, JACKEL, J.L., J. Solid State Chem. Jn_ (1982)
308.[90.] SHEN, X-M., CLEARFIELD, A», work in progress.[91.] THACKERAY, M . M . , JOHNSON, P.J., DE PICCIOTTA, L . A . , BRUCE,
P.O., GOODENOUGH, J.B., Mat. Res. Bull. J9_ (1984) 179.[92.] DAVID, W.I .F. , THACKERAY, M.M. , BRUCE, P.O., GOODENOUGH,
J.B., Mat. Res. Bull. 1_9_ (1984) 99.[93o] NAUMANN, D., Kernenergie £ (1963) 173.[94.] MAREI, S.A., BOTROS, N., Talanta 2J_ (1980) 599.[95.] OOMS, R., SCHONKEN, P., D'OLIESLAGER, W., BAETSLE, L.,
D'HONT, M., J. Inorg. Nuclc Chem. 36_, (1974) 665*[96.] MYASOEDOV, B « F . 8 BARSUKOVA, K , V , 8 RODÏONAVA, G . N . ,
Radiochem. Radioanal. Lett. J_ (1971) 269.[97.] NAQUI, S.J., Nucleus 8_ (1970 - published 1971) 76.
194
FUNDAMENTAL PROPERTIES OF ACID SALTS OFTETRAVALENT METALS AND SOME CONSIDERATIONS ON THEPERSPECTIVES OF THEIR APPLICATION IN NUCLEAR TECHNOLOGY
G. ALBERT!Department of Chemistry,Perugia University,Perugia, Italy
AbstractBeginning from 1964, acid salts of tetravalent metals
started to be obtained as crystalline materials and researchwas essentially focused for many years in their synthesesand structures. Now that the chemistry of these inorganicion-exchangers is well enough established, interest hasturned to their uses in various technological fields.Among them, their employment in the nuclear fuel cycle aswell as in the recovery, separation and purification of someradionuclides is very attractive.After a brief survey in the syntheses and classificationof the various acid salts and their derivatives and after ashort review of the ion-exchange properties of zirconiumphosphate and its salt forms with a and y lamellar structuressome properties such as hydrolysis and stability towardstemperature and ionizing radiations, that are of fundamentalimportance for their use in nuclear technology^are reportedand discussed.1 . Introduction
Acid salts of tetravalent metals have been known for along time only as amorphous materials Qi] . Because of theirhigh ion-exchange capacity and for their good stability toacid and oxidizing solutions as well as to temperature andionizing radiations, the potential uses of these materialswere subject to intense investigations in the decade 19-54 -1964, especially in nuclear centers. The success was howeververy limited, essentially because of the great tendency of
195
the amorphous acid salts to hydrolyze in alkaline or evenin neutral warm solutions.Beginning from 1964, acid salts of tetravalent metals startedto be obtained as crystalline materials and research wasessentially focused for many years in their syntheses and •structures, as well as in the determination of their funda-mental properties.To-day, after two decady of research carried out on thesematerials in various laboratories, a large number ofcrystalline acid salts with lamellar, fibrous and threedimensional structures has been obtained and their fundamentalproperties determined.A vast literature has accumulated and it is currently wellknown that acid salts of tetravalent metals are members ofa very wide class of inorganic compounds ["2-5J .Now that the chemistry of these materials is well enoughestablished, interest has again turned to their uses invarious technological fields. Among them, their employmentin the nuclear fuel cycle, in the recovery, separation andpurification of some radionuclides as well as in the treatmentof effluents from nuclear powder plants, is still veryattractive. Furthermore the crystalline acid salts, contraryto the amorphous ones, have definite chemical compositionsand structures. The study of their ion-exchange propertiesis therefore considerably facilitated by a knowledge of theircrystalline structure, phase transitions during ion-exchangeprocesses, thermal behaviour and so on.Useful information can be obtained and even some predictionson their potential applications in nuclear technology can bemade,, if the presently known fundamental properties of thesematerials are taken into careful consideration.For this reason, it is useful to divide the crystallineacid salts into various subclasses according to their com-position and crystalline structure and to give a brief account
196
on the known fundamental properties of each subclass withparticular attention to those useful for a better understandingof ion-exchange behaviour.
2. Classification of acid salts of tetravalent metals.
Acid salts of tetravalent metals can be considered asbeing constituted by macroanions carrying fixed negative chargeswhich are neutralized by protons or other counterions.A classification can therefore be made on the basis of thecomposition and structure of their macroanions.To date, two different kinds of macroanions are known:
and.IVwhere M = Ti , Zr, Hf, Ge, Sn , Pb, Ce, Th while X is an element
of the fifth group (usually P and As).Table 1 reports such a type of classification. Note that thestructure of the M ^X04)p macroanion may be eithera fibrous or a lamellar one. In turn, two different lamellarstructures, usually called a and y lamellar structures areknown at present
Table 1
Classification of acid salts of tetravalent metals according to their compositionand crystalline structure of the macroanion.
i2n-Amorphouslinear (or fibrous structure)bidimensional (or lamellar) structure
a-type
V-type
M ( I V ) * Zr, Ti , Ge, Sn; Ce; Th.
amorphous
threedimensional structure
X - P; As; Sb.
at least2 differentmodification
197
To give a complete picture of the classification of the acidsalts of tetravalent metals it must also be remembered that
F IV ~1lamellar compounds having the general formulae M ( RXO ) ;r iv n r iv l L J nM (R-OXO ) and M (0 XRXO ) , where R is an organicL_ o ^l n l «o o _j nradical, have recently been obtained JJ3,7_].,These compounds are known as organic derivatives of acidsalts. If the organic R group contains an acid group such as-COOH and -SO^H, the organic derivative may be considered
o _as an organic-inorganic ion-exchanger I 8,9J .Because of their potential uses in nuclear technology, ourattention will be essentially focused on lamellar zirconiumphosphates and on the threedimensional acid salts with skeletonstructure. It seems that the presently known fibrous acidsalts are not very stable to hydrolysis and to high doses ofr -iionizing radiations hol while the organic derivatives areexpected to exhibit less resistence to drastic working condi-tions than pure inorganic exchangers.
3. Acid salts with g-layered structure.
Some important acid salts with a-layered structure arelisted in Table 2. Among them, the a-zirconium phosphate is
~~ Table 2Some important acid salts of tetravalent metals with layered structure of the a-type.
Compound Formula
Titanium phosphate fTi{P04)_ H2°H 0Zirconium phosphateHafnium phosphateGermanium! IV) phosphateTin(IV) phosphateLead (IV) phosphateTitanium arsenateZirconium arsenateTin(iv) arsenate
_Zr(P04)2_H2.H20
Hf(P04)^H2.H20r6e(P04)2]H2.H20
Sn(P04)2"JH2-H20
Pb(P04)JJH2-H20
Ti( As04>2"jH2'H20
"zr(A304)?]H2-H20
Sn(As04)J H2"H20
Interlayer distance(A)
- 7.56
7.56
7.56
7.6
7.76
7.8
7.8
7.78
7.8
Ion exchange capacitymeq/g
7.76
6.64
4.17
7.08
6.08
4.79
5.78
5.14
4.80
198
7.56 A
7.56 A
• Zr O P O O
Fig. 1. Schematic structure of layered o- JzrlPO ij] H -H O(protons and water are not chown in the figure).
certainly the fullest investigated and the most stable tohydrolysis. Our attention will therefore be focused on thiscompound. Its layered structure was first elucidated byClearfield Mil . A schematic representation of the arrangementof three layers of a- Zr(PO ) planar macroanions is shownin Fig. 1. The distance between adjacent negative charges is
o5.3 A. The layered crystals may be regarded as packed macro-anions with intercalated counterions. Consequently, the inter-layer distance will depend .on the arrangement, volume andsolvation of the inter-lamellar counterions.The packing of the layers creates cavities in the interlayerregion (one for each zirconium atom) where there is roomenough for accommodating one water molecule or large divalent
O r
cations such as Ba .Such cavities are interconnected by narrow windows whosedimensions depend on the interlayer distance.
199
D —, uIr(PO ) JH -H 0 (d = 7.56 A) there is a strong sterichindrance to the diffusion of counterions with ionic diameter
olarger than 2.6 A. Thus, at acid pH values, only a smallnumber of cations (e.g. Li , Na , Ag , Tl , Cu , Ca ) is ableto exchange the protons of a- Zr(PO ) JH "H 0.Potassium and strontium give slow but appreciable ion-exchangerates while, in the case of large monovalent and divalent cationssuch as NH4> Rb , Cs and Ba or highly hydrated divalent or
2+ 3+trivalent cacions such as Mg , Co etc., the protons in theinter-lamellar region cannot be appreciably exchanged evenafter several days of contact with aqueous solutions of thesecations at room temperature. Thus, a- Zr(PO ) H -H 0 seems to
U* .aeJ
be a poor exchanger. However, zirconium phosphate is not sucha rigid exchanger as zeolites; as the interlayer distance isincreased there is a correspondent increase in the dimensionsof the windows connecting the cavities and cations of largesize can be exchanged.Various possibilities are currently known for obtainingconsiderable enlargement of the original interlayer distance.For example, a naked or partially .hydrated Na ion diffusesvery easily in the interlayer region replacing one proton foreach cavity fz, 3~| . Once inside the cavities it tends toréhydrate and water is forced to enter too.Since there is no room in the original cavities for wateraccommodation, the interlayer distance increases until anequilibrium between hydration and expansion of the interlayerdistance is reached. At room temperature this process canbe summarized as follows:
H °H 0 + Na ———»• HNa.SH 0 + HO2 2 a q . 2 3
(7 .56 I ) ( 1 1 - 9 A)
where barred species refer to the exchanger phase whilenumbers in brackets give the corresponding interlayerdistances. Owing to its large interlayer distance, the
200
monosodium form is now able to exchange large cationicspecies such as Cs , Ba as well as highly hydrated divalentor trivalent cations JJ2-14,2,3J.
+ 2 +For example, the ion-exchange processes with Cs , Ba and3+Cr are as follows:
HNa-5H 0 + Cs ————*• HCs + Na2 aq. aq.(11.9 A) (8.0 A)
HNa-5H00 + 0.5 Ba " ———>• HBa^ r.4H 0 + Na2 aq. 0.5 2 aq
(11.9 A) (10.0 A)
HNa-5H O + 0.5 Cr3+ ———>• H_ _ C r _ '4H_0 + 0.5H00+ +Na+
e. O. b 0.5 2 3 aq.(11 .9 A) (11.6 A)
It was found in our laboratory p5,16j that large cations caneasily be exchanged even in the original a- Zr(PO ) H -H 0if small amounts -of Na are present in the external solution.The sodium ion acts as an ion-exchange catalyst in the sensethat there is a considerable increase of the ion-exchange ratewhile, after the ion-exchange process, all the added Na wasfound to be present in the aqueous solution. The mechanism ofthis catalytic behaviour can be understood by taking intoaccount that the diffusion begins at the edges of the crystalsand proceeds towards the inner part. As an example, Fig. 2
2+shows schematically the Ba exchange in the presence of addedNa . Initially only Na ions are able to exchange the protonsof the exchanger by enlarging the interlayer distance of the
Oexternal part of the crystal to 11.8 A.2 +This expansion permits the diffusion of the Ba which exchanges
the Na ions within the enlarged cavities. The exchanged Naions go towards the inner part replacing other protons andenlarging another fraction of the crystals and so on. At tlend of the process, when all the protons are exchanged Na
201
ion must come back to the aqueous solution since the selectivity2+ +of the exchanger for Ba is much higher than for Na .
There exists another possibility for the exchange of largecationic species in crystalline zirconium phosphate
WïT- H,o î.«____________t
!«jw
I !5SJ222HZ&V J>gZBgg>1___ „„,,,,,, s2*f{A j , . . j j j j j . j >. J .JJ| ^.V/AV.-^ ZZZZZZS&' pizazz vzzzzzzzzzz* r~ ^SKSSSSSSSSSS^»L H»-tH,0 Mi _îllC..«» ÏÏ.JH.O llii 8< "'* ie.J»,0 ,..! <Ui.»*t tj À IJ.,„^
Fig. 2-Catalytic a f fec t of Na* en the exchange of large cation»on a-CzrlPOj,] H,-HtO (schematic)
When the monosodium form is regenerated into the hydrogenf— -îform with dilute mineral acids, the original a-jZr(PO ) H -H 0L 4 ^-J ^
is no longer obtained but one has instead a pentahycira-ceform (usually called ^-zirconium phosphate) with an inter-
olayer distance of 10.4 A. The process is the following:
HNa-5H 0 + HO ——————*• H °5H 0 + Na2 3 2 2 a q .
(11.9 A) (10.4 A)
Also, the replacement of the Na ion with some alcohols inacid media (e.g. ethanol acidified with HC1) gives intercalateswith a high interlayer distance. When these intercalatesare put in contact with water, the alcohol is replaced bywater and a- Zr(PO ) H •5H 0 is obtainedThe zirconium phosphate pentahydrate is a very selectiveexchanger for large cations and deserves further attention.When air-dried four molecules of water are lost with an irre-versible conversion intoGHZr(P04)J H2'H 0. Thus a-|Zr(PO ) H .
202
5H 0 must be kept wet. We are presently studying its thermalstability in water and preliminary results seem to show thata-zirconium phosphate possesses a good stability up to 150°C.In conclusion, many possibilities are known to-day for obviatingthe steric hindrance to the diffusion of large cationic speciesin the interlayer region of a-zirconium phosphate. Thisexchanger can therefore be used for. the exchange of the mostpart of monovalent, divalent and trivalent cations.The problem is now the following: what is the chemical andthermal stability of zirconium phosphate for applications athigh temperature? Its high thermal stability in air is wellknown £3,19,20]. At a temperature rate of 2°C/min, the thermalbehaviours is the following:
22o°c r-2 ————— [f < P°4
(7.56 A) ' (7.41 A) (6.80 A)
450°-650°C Q 800°-1000°C> 1
o(6.10 A) (cubic pyrophosphate)
The loss of the hydration water is irreversible and rehydra-tion does not occur even when Zr(PO ) H is dipped in coldwater. The phase transition at 220°C is instead reversible.
t
The condensation of =P—OH groups to lamellar pyrophosphatedepends obviously on the experimental conditions of dehydra-tion [21j (e.g. relative humidity, rate of temperature increaseetc.) and on the degree of crystallinity. For crystallinesamples the condensation does not occur until about 450°Cwhile for its salt-forms condensation cannot occur and thethermal stability is even higher than 500°C.In conclusion, a-zirconium phosphate may be considered a verystable exchanger in anhydrous conditions; but what happensat high temperature in the presence of water?
203
Since no data are available in the literature we have carriedout some preliminary investigations on the stability ofa- |Zr( PO ) l H °H 0 and its monosodium form.L 4 2J z 2It was found that a- Zr(PO ) H -H 0 is stable in water atleast up to 300°C. Only a reversible dehydration at 200°-300°Cwas found.
Hydrolysis is negligible and it seems to be limited to theexternal part of the microcrystals .In the case of the monosodium form the following behaviourwas found :
j— -i 5O°— "ioo°c r~ ~~\a- Zr(PO ) HNa-5H 0 ———————— >• a- ZrCPO ) HNa-H 0L 4 4 J ^ _4H o L «• £J ^2
(11.9 A) (7.9 A)
The decomposition into sodium dizirconium -chreephosphate isirreversible; therefore a-zirconium .phosphate cannot be usedin an aqueous medium at temperature higher than 200°C if Naions are present. This temperature is however high enoughfor many practical uses.Another inconvenience for the use of lamellar acid salts isthat they are usually obtained as small platelets (~ 1 M)Thus very compact Chromatographie columns are usually obtainedThe flows are therefore slow while some particles tend to bereleased into the external solution. The problem can be over-come by using larger particles. Incur laboratory it is nowpossible to prepare crystals of zirconium phosphate havingdimensions of some millimetersNo inconvenience Tor the preparation of the columns isenountered with large crystals; however, it must be taken inconsideration that the rate of exchange decreases when thesize of the crystals is increased.
204
4. Acid salts with y-layered structure.Other than with the tt-structure already discussed, lamellar
acid salts of tetravalent metals have been obtained with adifferent structure usually known as y-structure [3,23-25].Until now single crystals large enough for X-ray structuraldetermination have been not obtained, so present informationon the structure of X-acid salts is very limited. Investigationsperformed by Yamanaka and Tanaka [23j as well as studiescarried out in our laboratory J2a indicate that.the y-structure,similar to a-acid salts, is built up by the packing of planarmacroanions M (*0 ) whose negatives charges are neutra-lized by protons or other counterions. However the structureof the y-macroanion is different from that of the a-one.The planar density of fixed charges is higher in y than ina-n?.c poanions (about 1.37 times). Counterions, water and otherpolar molecules can be arranged in the inter-lamellar region.Owing to its large interlayer distance, steric hindranceto the diffusiono of large cations is much less than ina-acid salts. In particular, y-[zr(PO )JH -2H 0 (d = 12.2 A)|__|_ 4 2J 2 2seems to be very selective for Cs and its use in nucleartechnology has therefore been predicted [26] .
5. Acid salts of tetravalent metals with fibrous structure.
The first fibrous acid salt of a tetravalent metal, ceriumphosphate, was obtained in out laboratory in 1968 [27] .Subsequently, fibrous thorium phosphate was also obtained.[28Jand zirconium and titanium phosphates have also recently beenprepared by hydrothermal methods {~29J .Fibrous inorganic ion exchangers are interesting because theycan be used to prepare ion-exchange papers and membranes. A photo-graph of a sample of Ce ( PO^J H
2 *H 0 showing its fibrousstructure is shown in Fig. 3.Fibrous acid salts have not been widely investigated and nostructures studies are presently available.
205
Fig. 3 - Photograph, taken at optical microscope,of a fragmentof Ce(HPC> ) -H O.42 2
Thermal and chemical stability of fibrous cerium phosphateseems to be lower than that of a-zirconium phosphate. Alsothe resistance to radiation, seems to be lower than that ofa-zirconium phosphate po] . However fibrous cerium phosphateexhibits some particular selectivity (i.e. towards lead) thatcould be utilized for the separation of radionuclides £/,30] •Furthermore, fibrous acid salts, contrary to the lamellar ones,are suitable materials for the preparation of Chromatographiecolumns or for the preparation of very compact ion-exchangefilters.
6 . Acid salts of tetravalent metals with skeleton structures.
Compounds of the general formula M (X04^o M have beenknown for a long time [31,32] . They have usually been preparedby solid state reactions at high temperature ( 1100°-1400°C ) .
206
Their structures may be described in terms of XO tetrahedraand M 0 octahedra which are linked by corners to form a6threedimensional network.Although these compounds may be considered as salt forms ofacid salts M (XO ) |H, the replacement of M counterionswith protons is very difficult. However, it was recently foundthat if the Li-form is prepared by decomposition ofa- [zr(PO ) 1 Lizr(PO ) Li at 600°-800°C
or by solid state synthesis at relatively low temperature j_34Jthe replacement of the Li with protons can easily be carriedout by single contact with acid solutions. The dizirconiumthreephosphate in hydrogen form has been found to behave as a
+ + + +narrow ionic sieve. Only H , Li , Na and Ag can diffuse easilywithin the exchanger while large monovalent and all the poly-valent cations are not exchanged.The acid salts of tetravalent metals with skeleton structure
+ + +can therefore be used for the recovery of Li , Na and Agor for the purification of cationic species, even in veryconcentrated solutions, from traces of Li , Na and Ag .
7 . Some concluding remarks.
Table 3 reports some important characteristics of someacid salts of tetravalent metals for use in nuclear technology.A great deal of experimental work is still needed to seeif some of these materials are indeed useful fro some practicalapplications in nuclear technology.Certainly crystalline acid salts exhibit in many cases acomplete different behaviour from the corresponding amorphousmaterials .
207
Table 3
Important characteristics of some acid salts of tetravalent metals for usein nuclear technology.
Characteristics
Ion-exchange capacityResistance to ionizing radiationsResistance to oxidizing solutionsResistance to reducing solutionsResistance to acid solutionsResistance to alkaline -solutionsThermal stability in air
Thermal stability in water
Ion-exchange rate
Selectivity
Lamellar
very goodM «B
..
„
1« 1«
BadVery good up to45O°CVery good up to30O°C;bad over210»C for its mo-nosodium form.Sufficiently good
High for Cs* Sr2+Ba2* and trivalentcations if Na+ tra-ces are present.
Lamellary-lZr(PO )JH*2H 0L 4 2 J 2 2
Very good?
Very good"
GoodBadGood
T
Good
High for Cs+and NH£ ions
Fibrous
Very goodBadVery goodBadModerateVery badGood
»
SufficientlygoodHigh for Pb2+,Ba +, Cu2+and K+.
Threedimensional
LowVery good (estimated)
" < " >.. ( .. j.. ( .. j
SufficientVery good up to 600°C
Very good
Sufficiently good
Highly selective forLi+, Na+ and AgtLarge monovalentcations and polyva-lent cations are notexchanged .
Thus many negative results previously obtained with amorphousacid salts should be critically examined in order to see ifsome inconveniences may be overcome by usina the newcrystalline acid salts of tetravalent metals. In any case,the investigations on the chemistry of acid salts arecontinuing in several laboratories and new materials andnew properties are continously being found. Referring onlyto new acid salts recently obtained in our laboratory, wecan cite the preparation of pellicular a-zirconium phosphate,[35] the synthesis of a-zirconium phosphate-phosphite {36,37]with composition • (Zr(PO.)_ CC(HFO_)L 4 O.66 3 . _ H „„ and several1 . ooj U .66 0 138J .We are now studying the ion-exchange properties of thesenew materials. In particular it was found that zirconiumphosphate-phosphite, for its structure (Fig. 4),is able toaccommodate in the interlayer region large cationic and polarspecies.
208
(b)
\lx Photphit« Group
O Phosphlt* Group
Fig. 4. a) Idealized structure of a-zirconium phosphate-pho-
sphite of composition a-j~Zr(P04>0 6glHPO >
H0.66b) Phosphate and phosphite groups on one side of
a layer.
Acknowledgements
The work carried out in the author's laboratory wassupported by Finalized Project of F.C.S.C. of C.N.R. andby M.P.I.. The author is indebted to his co-workers U.Costantino and M. Casciola for many discussions and theuse of certain results in advance of full publications.
References
[i .] AMPHLETT, C.B., "Inorganic Ion Exchangers",- Elsevier,Amsterdam (1964).
[2.] ALBERTI , G., Ace. Chem. Res., 1J_ (1978) 163.
209
[j3/] CLEARFIELD, A., Chapter 1 of "Inorganic Ion ExchangeMaterials" (Clearfield, A., Ed.) CRC Press, Boca Raton,Florida (1982).
[4.] ALBERTI, G., Réf. 3, Chap. 2.Qp.] COSTANTINO, U., Réf. 3, Chap. 3.(6.J ALBERTI, G., ALLULLI , S., COSTANTINO, U., TOMASSINI , N.,
J. Inorg. Nucl. Chem., 40 (1978) 1113.[?.] DINES, M.B., DI GIACOMO, P.M., Inorg. Chem.. 20 (1981) 92.[8.] ALBERTI, G., COSTANTINO, U., .LUCIANI GIOVAGNOTTI , M.L.,
J, Chromatog. 180 (1980) 45.fg.] DINES, M.B., DI GIACOMO, P.M., Inorg. Chem. 2!0 (1981) 92.[10.] ZSINKA, L., SZIRTES, L. , MINK, J., KALMAN, K., J. Inorg.
Nucl. Chem. 3_6 (1974) 1147.[11.3 CLEARFIELD, A., STYNES, J.A., J. Inorg. Nucl. Chem. 26
(1974) 117.[12.] ALBERTI, G., COSTANTINO, U., GUPTA, J.P., J. Inorg. Nucl.
Chem. 3J5 (1974) 2103.[13.] ALBERTI, G., BERTRAMI , R., CASCIOLA, M., COSTANTINO, U.,
GUPTA, J.P., J. Inorg. Nucl. Chem 38 (1978) 843.(J4.] ALBERTI, G., BERNASCONI, M.G., COSTANTINO, U., GILL, J.S.,
J. Chromatog. 132 (1977) 477.Qisj ALBERTI, G., COSTANTINO, U., GUPTA, J.P., J. Inorg. Nucl.
Chem. 356 (1974) 2109.[16„] ALBERTI, G., BERTR'AMI . R., COSTANTINO, U., J. Inorg.
Nucl. Chem. _38 (1976) 1729.[17.] ALBERTI, G., COSTANTINO, U., GILL, J.S., J. Inorg.
Nucl. Chem. 38 (1976) 1733.[18.] COSTANTINO, U., J.C.S. Dalton (1979) 402.[19.] LA GINESTRA, A., MASSUCCI, M.A., FERRAGINA, C.,
TOMASSINI, N., Thermal Analysis, Proc. 4th I.C.T.A.Budapest 1974, Vol. I, Heyden, London (1975) 631.
C20.] HORSLEY, S.E., NOWELL, D.V., Thermal Analysis, Proc.3th I.C.T.A., Davos, Vol. 2, Birkhauser Verlag, Basel(1971) 611.
210
.1 COSTANTINO, U., LA GINESTRA, A., Thermochim. Acta 58(1982) 179.
["22.] ALBERTI, G., COSTANTINO, U., GIULIETTI , R., J. Inorg.Nucl. Chem. £2 (1980) 1062.
[23.U YAMANAKA, S., TANAKA, M., J. Inorg. Nucl. Chem. 41_(1979) 45.
[24/] ALLULLI, S., FERRAGINA, C., LA GINESTRA, A., MASSUCCI,M.A., TOMASSINI, N., J. Inorg. Nucl. Chem. _3_9 (1977)1043.
[25.] ALBERTI, G., COSTANTINO, U., LUCIANI GIOVAGNOTTI , M.L.,J. Inorg. Nucl. Chem. 41_ (1979) 643.
[26.] KOMARMENI, S., ROY, R., Nature 299 (1982) 707.[27.] ALBERTI, G., COSTANTINO, U. , DI GREGORIO, F., GALLI , P.,
TORRACCA, E., J. Inorg. Nucl. Chem. 30 (1968) 295.[28.J ALBERTI, G., COSTANTINO, U., J. Chromatog. !50 (1970) 482.[[29.] ALBERTI, G., COSTANTINO, U., LUCIANI GIOVAGNOTTI, M. L.,
to be published.[30J ALBERTI, G., MASSUCCI, M.A., TORRACCA, E., J. Chromatog.
_30 (1967) 579.[31 J HAGMAN, L.O., KIERKEGAARD, P., Acta Chem. Scand . 2_2
(1968) 1822.[32.] SLJUKIC, M., MATKOVIC, B., PRODIC, B., SCAVNICAR,
Croat. Chem. Acta _37 (1965) 115.[33.] ALLULLI, S., MASSUCCI, M.A., TOMASSINI, N., Italian
Patent no. 48852A/79, 26 april (1979).[34.] ALBERTI, G., COSTANTINO, U., MASSUCCI, M. A., TOMASSINI,
N., Italian Patent no. 49546A/83, 20 december (1983).[35.] ALBERTI, G., COSTANTINO, U., Italian Patent no. 484387/A,
17 may (1982).[36.] ALBERTI, G., BARTOLI, F., COSTANTINO, U., DI GREGORIO,
F., Italian Patent no. 22365/A83, 1 august (1983).C.37.] ALBERTI, G., COSTANTINO, U., GIULIETTI, R., Gazz. Chim.
Ital. 113 (1983) 547.[38.] ALBERTI, G., COSTANTINO, U., KORNYEI, J., to be
published.
211
RECOVERY OF CESIUM-137 FROM RADIOACTIVE WASTE SOLUTIONSWITH A NEW COMPLEX INORGANIC ION EXCHANGER
S. ZHAOXIANG, T. ZHIGANG, H. ZUN,L. ZHENGHAO, W. HAOMIN, L. TAIHUA, L. BOLIBeijing Normal University,Beijing, China
Abstract137Recovery of Cs from radioactive waste solution has been
studied by using titanium phosphate - ammonium molybdophosphate. Thecomplex ion-exchanger shows favorable performance for the adsorption and
137elution of Cs. The product Cs has more than 99% radioactive purity103 95 22 90and high decontamination factors for Ru, Zr, Na, Sr,
Am but less for Rb and Ce. A scheme for recovery process137of Cs from radioactive waste solution is suggested from the
experimental results with this new complex ion-exchanger.
Introduction
Utilizion of high-level liquid waste from spent fuel reprocessingcontaining useful isotopes is a matter of great importance for theirpotential nuclear applications. Two isotopes that have receivedconsiderable attention are cesium-137 and strontium-90. Possibletechnique for recovery of cesium and strontium from radioactive wastesolution(IAW) fall into three categories: precipitation, solventextraction amd ion exchange. In general the precipitation and solventextraction processes are rejected because of their adverse effect onwaste management due to the added chemical or because they are noteffective in the acid range easily attainable in IAW. Furthermore,radiation decompositions easily occur in the extractant. Since inorganicexchangers and adsorbents usually have high chemical sélectivités andhigh radiation resistance, they have been considered for recoveringcesium-137 and strontium-90 in
Several exchangers have also been developed for removing cesiumfrom highly acidic solution; these exchangers include the heteropoly acidand acid salts of tungsten and molybdenum, as well as the zirconium,
(3-7titanium, and tin salts of condensed phosphates. ' As is well known, 'these inorganic ion exchangers have some shortcomings, for example theacid salts of tungsten and molybdenum are amorphous. The zirconium and
213
titanium of condensed phosphates easily block up the exchange column, ifpresent in the trivalents state in the feed.
This report presents our preliminary laboratory study on a newcomplex inorganic cation exchanger titanium phosphate-ammoniummolybdophosphate (Tip-AMP), determining its cesium capacities andrelative sélectivités and capabilities. The results are shown below.The new complex exchanger Tip-AMP possesses unique advantages and show noshortcomingsc
Experimental
Preparation of the ion exchanger
The ion-exchanger titaniun phosphate was prepared according tothe method of C. Beaudet ' '. Good granules (50-100 mesh) of Tip wereobtained in our laboratory, then the exchanger was loaded in glass columnor beaker, after being combined with AMP by dynamic or static method; ayellow complex ion-exchanger Tip-AMP was obtained. It was thoroughlywashed with distilled water to remove free acid till the effluent pH wasneutral, and finally was air-dried for three days.
Determination of breakthrough curves and evaluation ofbreakthrough capacity
1 gm exchanger was loaded in a glass column (20 cm x 0.5cm)provided with glass wool padding at the bottom, bed volume being 1 ml/goThe loaded column was washed with 10 column volumes of nitric acid ofappropriate molarity. The synthetic feed solution containing knownamount of carrier and tracer Cs-137 was passed through the exchanger at afixed flow rate (2.0-3.0 ml/hr) at room temperature. 1 ml fraction werecollected with an automatic collector till C/Co reached approx.l. TheC/Co value was plotted against the volume of the feed and breakthroughcapacity calculated where C/Co reached 0.01 (1 % BT). The exchangecapacity is 0.47 meq/g and breakthrough volume is 85 BV« The results areshown in figure 1 and composition of synthetic feed solution in Table I.
Effect of acidity on the cesium capacity of complex exchangerTip-AMPThe exchange capacity was investigated according to the above
static method. Table II shows the results.
214
100 %'
30
60
o 40o
20
Am-241 Cs - 137
123 20 40 60 80 90 100 120 BVFig. l Breakthrough curves of Am241 and Csl37
from Tip-AMP column
Table I. Composition of synthetic feed solution
Ion
C8
Ba
Concentration
2.50.750.0750.50.253.752.02.00.25l H
Table II. The exchange capacity of various acidity
AcidityHNO- 5
Exchangecapacitymeq/g
0.5
0.50
1.0
0.47
2.0
0.46
4.0
0.45
215
Effect of sodium ion concentration on the cesium capacity ofcomplex exchanger Tip-AMP
Both sodium and cesium belong to the alkali metals.It is greatly meaningful to observe the exchange competition
between sodium and cesium, hence the exchange capacity was evaluated withdynamic method at different sodium ion concentration. We found that thesodium ion concentration in feed solution slightly influenced the cesiumcapacity of Tip-AMP as shown in figure II.
10096 •
Vio4»G
8ft
02 1•a a< u
20JÉ
Tg.(Na/Cs)Fig.2 Effect of Na/Cs on the adsorb percent
of Cs-137
Table III. Effect of cycle on the Cesiun capacityof Tip-AMP
Cycle
1234567
Capacity
0„470.490.500.490«470„48Oo45
meq/g
216
Effect of cycle use on the cesium capacity of Tip-AMP
Exchange capacity for cesium was determined under the abovecondition on the complex exchanger Tip-AMP. When the cesium breakthroughreached 1% in the effluent, cesium elution was started using 5.5 NNH.NO., + 0.5 N HNO. at 60°C. Then the exchange column wasregenerated with 10 column volumes of l N HNOß. Seven cycles could beused at least with this complex exchanger, and results are shown inTable III.
Recovery of cesium-137 and determination of decontaminationfactor of Ce-141, Ce-144, Ru-106, Ru-103, Zr-95 and Nb-95
Experiments were also carried out on the decontamination factorof Ce-141, Ce-144, Ru-106, Ru-103, Zr-95 and Nb-95 from feed solution, towhich a fresh fission product was added as an indicator. The feedsolution passed through the exchanger at a fixed flow rate 2.0-3.0 ml/hrat room temperature. Bed volume was 70. The loaded column was washedwith 5 column volumes of 0.5 N HNO, and 15 column volumes of 0.5 NH2C2°4 + °'5 N HN03 f°r removinS Ce-141, Ce-144, Ru-106, Ru-103,Zr-95 and Nb-95 and the purpose of decontamination was achieved.
Cesium elution was carried out by passing 10 column volumes of5.5 N NH.NO. + 0.5 N HNO, at 60°C. The total cesium eluted was4 3 3examined on their activity at 4096 channel analyser. The cesium effluentconcentration indicated a less than 1% loss in loading the column toessentially 100% breakthrough. Washing liquid of 0.5 N HN03 and 0.5 NH2C2°4 + °*5N HN03 were collected and examined for their activityby 4096 channel analyser. The results are shown in Table IV and figures3,4, and 5.
Table IV. Distribution of contaminating nuclldee in the process
Nuclidea '
Ce-141Ce-144Ru-103Zr-95Nb-95Na-22Sr-85+89Am-241Rb-66
feedcpin
23099086110129738364273569940S266705669951164150760856
effluentcpa
22855765600128336292488107765---—
HNO0.5N516216960714102293962112419832680
H2C204 HN030.5H---
313483
-a.-—
0.5N HH4NO,+0.5N HN05--1226403
7106701536
507274
decontamina-tion factor_...1062300
1
11648507581.5 .
217
sa,u
o 36
H
S 24>•H
S00£ 12
o _
iê
-
H ^1* ?
• g» *»0 «
^>
°T0)o
K"V £~O 1e»p-S H
L.'l100 300 500 700 800 kevFig.3« Gama spectrometry of feed
135
90
45
HTHBO
ITMTv
C-3 Z
OH
•H
lüüFig. 4effluent
30ü 500 700 800 kevGama Spectrometry of
144
1-1 108M
«ieso•rt•Ö
72
36
IT»ON
600 800 kevFig. §. Gama Spectrometry ofwashing oxalic acid
Recovery of Cs-137 from IAW using Tip-AMP
137The purity of Cs is higher than 99% by paper chromatographyand loss of Am-241 is less than 1%. The product Cs-137 contained Ce, Sr,Na'and Rb with concentration of 0.6%, 0.1%, 0.28% and 6.7% respectively.Therefore, this product might be separate with regard to cesium andrubidium. The results are shown in figure 6,7,8.
218
au^ 4oHK -i
4»•r«>
"$ 2UAo•r4•Ö 1o)cü
,**
ÇNH
—
- •^fc
K^sH1«
"l
1«ao
•+|C\H
1IQ
JL°.300 600 900 kevFig. 6. Gama apectrometry ofRecovered 137Cg
a so
0 tH »M
£ 4*•!
•Ho p(Ö *o•HT)
« n
m
•
.
-
Sr-90 Ru-106 Ce-144
Cs-137
"I. , r-J , 18 10 cm
Fig. 7. Paper chromatography offeed
° 6<»•or-lM>, 44»
Umo-H•Ocd«
Sr-RuCe
n-90) P.-106-144i__< U__i
Cs-137
4 6 8 10 cmFig. 8.Paper chromatography of
recovered 137Ca
Results and Discussions
1. The results reveal that complex ion exchanger Tip-AMP is muchbetter than titanium phosphate (Tip) or ammonium molybdophosphate (AMP)alone. First, the ion-exchange column block-up when only Tip is used forrecovery of Cs-137 from IAW is overcome and the problem of makinggranulated AMP is solved too. Secondly, the ion-exchange behavior isimproved. Although Tip has a high rate of adsorbing Cs, Cs adsorbed isreadily eluted; however, it almost loses the ability of adsorbing Cs athigh acidity. AMP has a high selectivity capacity for Cs.but Cs adsorbedon it is hard to elute. The new complex exchanger Tip-AMP keeps only theadvantages and thus overcomes the demerit.
219
The complex ion-exchanger Tip-AMP has a higher selectivitycapacity for Cs reaching 0.47 meq/g when C/Co = 0.01, and is morereadily eluted; only 10 column volumes can remove above 99% Cs from theloaded column.
The product Cs contains other radioactive nuclides, totalamount being less than 1% except Rb.
The radioactive purity of recovered Cs-137 by paperchromatography and A096 channel analyser is higher than 99%.
The complex ion-exchanger does not adsorb Am-241s so the lossof Am-241 is less than 1% in the course of ion exchange. This is verysignificant.
2. Tip-AMP exchanger is a complex compound and not astoichiometric compound. This has been confirmed by X ray diffractionanalysis.
3. The results indicate that this new complex inorganic ionexchanger has a good ion-exchange performance for Cs in IAW.
On the basis of these experiments, a scheme for Cs-137recovering process is suggested. It was shown in Figure 9.
IN HN03feed solu-tion
0.5N HNO,
flow rate 2-3BV/hrfluence 5 BVtemperature 15-20°C
eluteCs-137
1
i
0.5N H2C204
flow rate 2-?BV/hrfluence 10 BVtemperature 15-20°<
1
( t
Tip-AMPexchange column
1
Û.5N UNOwashesSr.Ce.Ru
! ———j
Oo5N H2C204wasteZr-95 Mb-95
0.5N NIi4NO +0„5N HNO,flow rate 2-3BV/hrfluence 10 BV
2 tempera-cure 60 C
1
1
fluentsolutionNo Cs-137
Figure 9« Recovery process of Ca-137 from 1AW
220
References
1. V. Vesley et. al. Talanta 19. 219; 19. 1245 (1972)2. A. Clearfield,Ed. Inorganic Ion Exchange Materials (1982) CRC
Press.3. V. Kourin et.al. Atomic Energy Review 12(2) 215(1974)4. L. Baetsle; D. Vam Deyck; D. Hugs and A. Guery
B. L. G. 2 67 (1964)5. L. Baetsle; D. Hugs
B. L. G. 487 (1973)6. Yvonne M. Jones BNWL - 270 (1964)7. D.K. Davis; J.A. Partridge; O.K. Koski
BNWL - 2063 (1977)8. C. Beaudet et. al.
Nuclear Application 5(2) 64 (1968)
221
TREATMENT OF LIQUID WASTES CONTAININGACTINIDES AND FISSION PRODUCTS USING SODIUMTITANATE AS AN ION EXCHANGER
Y. YING, L. MEIQIONG, F. XIANNUAInstitute of Atomic Energy,Beijing, China
Abstract
The ion-exchange selectivity and capacity of sodium titanate havebeen studied as well as the preparation and characterization of thisinorganic compound. Sodium titanate has a high selectivity forpolyvalent ions (Sr, Ce, Fe, Pu, Am, etc.) but not for monovalent ions(Cs) in neutral and basic radioactive waste solutions. Sodium titanatecan be used for the separation and consolidation of these actinides andof fission products into a suitable waste form.
IntroductionNew inorganic ion-exchangers of high adsorption selectivity
have been developed owing to their ideal irradiation, thermaland chemical stability. They might be used in treating and dis-posing of high-level radioactive wastes in nuclear energy field.The techniques of solidification by means of Synroc and monaziteto fix the high-level wastes or high-level o(-wastes all havemade considerable headway. Inorganic ion-exchangers can be usedas adsorbent and additives in solidification of liquid wastes.Natural inorganic materials which have exchange capacity couldbe used as back-fill of multiple barrier system for the dispos-al of nuclear wastes. Among several new ion-exchange materialsof hydrous oxide type, sodium titanate is of most interested.In this work, sodium titanate (ST) was used as an adsorbent fordecontaminating and separating the liquid wastes which containsactinide nuclides and fission product and satisfactory resultwere obtained.
ExperimentPreparation and analysis of sodium titanateLab-Scale ST has been prepared as follow:'1. Tetra-isopropyl-titanate
Titanate tetrachloride was allowed to react with iso-
223
propyl alcohol in anhydrous benzene solution, and the metallicsodium was added later to this reaction mixture with stirring.The reaction mixture was heated and refluxed for 3 hours. Theproduct was centrifuged and isolated by vacuum fractionation.
2. Sodium titanateTetra-isopropyl titanate was added to the 10# NaOH in
methanol solution with stirring, then hydrolysed and filtered.The product was dried at room temperature.
Sodium in ST's product was determined by atomic adsorptionspectrophotometry, and titanium was determined by cupferronreagent.
The components of ST were determined by X-ray diffractiontecnique.
Radioactive measurement:e^c 3 and -activities were measured by silicon surface semi-
conductors plastic scintillation detector, and Ge(Li) detectorrespectively.
Result and discussion- Analysis productThe determined value in ST product agreed with the feed
on Na/Ti ratio(in Mole), but in the case of Na/Ti=l/20 in feedthe result for ST product was lower as shown in Table I.
TABLE IDETERMINATION OP Na/Ti RATIO (IN MOLE) ON ST SAMPLES
NO s
12
3
Na/Ti RATIO(IN FEED)
1/1
1/21/?0
Na/Ti RATIO
(ASSAY)
1/1.041/1.8
1/15.2
The sample of ST (Na/Ti)=1/1 which were sintered underdifferent temperature were determined by X-ray diffraction tech-nique. These data were roughly similar. The sample seemed to bepurer when it had been treated at temperature up to 900°C. Thediffraction data of sodium titanate are shown in Tablel . Thesedata are very similar to that of A2B2Xr such as (Ca, Pe, Na)2(Sb, Ti)2(0,OH)7 type of compound.
224
TABLE 1 DATA OF X-RAY DIFFRACTION ON STRADIATION CONDITION Cu KA( Ni ) Al FOILEXPOSURE CONDITION 55 KV 15 mA 8.5 Hrs
ST SINTERED700° C
Q11.414.216.921.825.127.128.630.932.334.440.0
D3.9003.1032.9592.0761.8171.6921.6421.5011.4431.3651.199
ST SINTERED900" C
Q13.514.816.822.624.628.630.731.732.334.339.9
D3.3023.0182.6672.0061.8521.6101.5101.4671.4431.3681.202
TYPICALA,B,X7D3.2853.1052.9642.0071.8171.7161.5501.4811.4401.3371.178
In addition, both Na and Ti were uniformly distributed ineach micro regions of whole sample when observed under electron-microscope.
- Different way hydrolysisThe ion-exchange performance of ST product depends on di-
fferent condition of hydrolysis. The product hydrolyzed in purewater fomed a hard granules but it's ion-exchange performanceis not good enough. The product hydrolyzed in water-acetonesolution formed loose granules but it's ability for ion-exchangewas 5-6 time greater than in pure water.
The ion-exchange performance of different Na/Ti ratio(inKole) was shown in TableJE. Sodium titanate of Na/Ti=l/l(in Mole)seemed better than Na/Ti=l/2.
- Hydraulics resistanceThe liquid flow experiment was carried out using a column
with ST bed of 1 Cm in diameter(l.D.) by 11 Cm in height. Disti-lled water passed through ST column with a flow rate of 6-7 bedvolume(B.V.) per hour. After 440 hours(3000 B.V. effluent) opper-ation no resistance change was observed. This result indicatedthat ST exchanger has a good performance in column operation.
225
TABLE KION EXCHANGE PERFORMANCE OP ST WITH DIFFERENT Na/TiRATIO (IN MOLE )
NO: Na/Ti RATIO ST . FEED(M) WEIGHT (g) VOLUME
DISTRIBUTION RATIO(Kd)
12
34
1/21/11.5/1
(HTii0,H)
0.150.150.150.15
10101010
4.17X102- 6.6x10*a ILT Qv"\r\J *\ T Y I /™iJ. o Ö« J-U 1, o X A.LU
2 QY T r\* ^ o if 1 n9 7 A J.U J .©ÖÄ X(J
7 15
V, 100N—•
sI—II 8°•
60
40
20-
Sr
10' 10 10" 10*TIMS
PIG. 1 THS EFFECT 0? VARIOUS COÏJTACT TIMEOS ADSORPTION RATIO
» Flow rate of ion-exchangeLike other inorganic ion-exchanger, the flow rate of ST
exchanger is slow. In batch adsorption experiment, the effect ofvarious contact time on adsorption ratio for Sr, Ce and Amare shown in Fig.l. The effects of various flow rate on adsorp-tion capacity and the volume of effluent at 1$ break-through aregiven in TableE. In our experiment, the flow rate of ST columnwas 0.36Cm/min. (4-5B.V./hr.)
- Distribution coefficient, capacity and decontaminationfactorBatch adsorption experiment were performed by contacting
weighted 0.15 gram of ST sample with 10 ml of NaN03 feed solution.
226
TABLE E
FLOW
THE EFFECT OF FLOW
VOLUME
RATE ON CAPACITY
OF EFFLUENT CAPACITY
AT I°/t> BREAKTHROUGH
B.V./hr
1.54.57.51116
Cm/min0.120.360.60.88
1.3
ml
25.52420
1512
B.V.
42.54033.32520
meq/g
2.122.01
1.931.841.5
TABLE IDISTRIBUTION COEFFICIENT (Kd) FOR ST WITH SEVERAL CATION
IONSCs +Srz*Ce3*Fe^Co"Pu''"*Am'*
Cm*
FEED COMPOSITION
0.6MNaNOj-0.02M Cs
0.6KNaN05-0.02M S r
0.6MNaNOj-0.02M Ce
0.6MNaNOj-0.02M Fe
0.6MNaNOj-O.01W Co0 . OMNaNO, -2 . 9x lcf£//l Pu0.6WNaNOs-1.3xlo"c//l Am0.6MNaNOi-8.6xio"o//l Cm
TRACE NUCLIDE'"Cs
*.e?Sr
'*'Ce^Fe"GO
Pu'«'Am*^m
Kd
241.8X103
1.1^10*3.4'(10Z
1.2*103
8.8X103
1.1X10^8.6X103
Each cation to be measured was to feed solution and traced withtheir radionuclidea respectively, and the acidity was adjusted toPH 2. The result shown in Table 1 indicate that ST has a highdistribution coefficient for polyvalent cations than for mono-valent cation.
The equilibrium distribution coefficient, Kd, is difined as:ci W (3)
Where Co: original concentration in liquid phaseCi : remained of solutionW : weight of STV t volume of solution
The column experiment with a flow rate of 4-5 B.V. per hourare given in Table]! . The break-through curves of Sr, Ce, Co andAm on ST column are shown in Fig. 2.
227
oo
80
60
40
20
Co
20 40 60 80 100
VOLUÎu'î 07 EFFLÜSirr (B.V.)
FIG. 2 THE BREAtCfHROUGH CURVES OFSr, Ce. Co„ Am
TABLE 21
CAPACITIES AND DECONTAMINATION FACTOR OP ST FOR SEVERALCATION IN COLUMN OPERATION
NOs
12
345
NUCLIDE
SrCeAmCoPu
FEED
.COMPOSITION
0.6WNaN03-0.02M S r0.6MNaNOj-0.02M Ce0.6MNaN03-0.02M Nd
O.öMNaNO 3-0.01*1 Co
0.6MNaN03 -2 .9*10~e//lC"Pu)
CAPACITY(meq/g)
3.93o8
4.2
1.9
D. F. (BEFORE
\aß> BREAKTHROUGH )
1010
1010
D.F.>-10^in 300 B.V.
EFFLUENT.
- Effect of salt contentThe influence of various quantity of Na, Fes Ca in the feed
solution on distribution coefficients of Cs, Sr, and Am for STare shown in Fig. 3« It indicates that Kd of Cs on ST decreasesag the concentration of Na increases. There is no marked changefor adsorption of Sr, Ce and Am. While the concentrations of Caor Fe in the feed solution are more than 2 grams per litre, the
228
Kd
10'
10* J
10'
Ara
Ce
Ha:i05 (M)
FIG. 3 THE EFFECT OF Ha* ON ADSORPTION RATIO
Kd
io rj
10'
10'
1010 10 10
(K)
FIG. 4 THE EFFEC-r OF Fe" OÏI DISTRIBUTION RATIOFOR Sr
229
JCd
10
10'
10'
1010" 10"
Ça (M)
FIG. 5 THE EFFECT OF Ca Oil DISTRIBUTION RATIOFOR Sr
distribution coefficient of Sr on ST remains greater than 10. andthere is significant change if it is less than 0.5 grams perlitre,(See F.g. 4-5)
- Effect of complexing agent contentWith the organic complexing agent in feed, most polyvalent
cation formed complex with it. In 10ml feed, containing 0.6MNaNOjand 0.02M Sr, 0.15 grams of ST, and various quantity ofNTA, DTPA or EDTA were added. The curves of Fig. 6-8 showed thatthese three complexing agent all effected adsorption ratio ofcations« In simulated high-level liquid wastes(components listedas Table2.), there was little adsorption if 0.15 grams of STwas added« But when the quantity of ST increase to 0.3 grams, STshowed certain effect on adsorption, even if the concentrationof EDTA was much higher than Sr (See TableM).
-= Adsorption performanceThe adsorption of ST for cation are more suitable in neu-
tral and basic solution than in acidic solution. Take Sr for ex-ample, when acidic feed solution (PH 2) passed ST column, a 1^break-through would appear in the effluent at 70-80 B.V. However
230
s
100
80
60
20'
0.001 0.002 0.003 0.004 O.OC5IirA(M)
PIG. 6 THE EFFECT OP NTA ON ABSORPTION RATIO
100
-80
60
40-
20
100
g 80<
60
40-
20-
0.001 0.002 0.003 O.OC4 0.005 0
DTPA (K)
0.01 0.02 0.03SDTA (M)
FIG. 7 THE EFFECT OF DTPA Oil ADSOHPTICH RATIOFIG. 8 THE EFFECT 0? EDTA OH ADSOSPTIOM RATIO
FOR Sr
231
TABLE 31THE EFFECT OF EDTA(M) ON ST ADSORPTION (IN SIMULATED HIGHLEVEL LIQUID WASTES)
Nos
1234567a
EDTA (M)
00.00050.010.050.10.150.20
ST WEIGHT (g)
0.30.30.30.50.30.30.30.3
feed volume(ml)1010101010101010
adsorption(%}
9594.190.489.747.29c77.83
TABLE 3[COMPOSITION OF SIMULATED HIGH LEVEL LIQUID WASTES
COMPOSITION
NaFeCrNiSrZrCsBaCe
COMPOUND
NaN03
Fe(NOi)j9HzOCr(NOj) jNi(N03)3
Sr(NO,)z
ZrO(N03) i2H iO
Cs(NOj)
Ba(N03)z
Ce(N03)2 6H20
WEIGHT
(g/1)
5018
32 .1.2
1.590.141.7
21
CONCKNTHATION
(M)
Oo598
Oo040.013O.COfi5
0.007
Oo059
0.0070.00650.048
then neutral feed (PH 7-8) was passed through, a 0.8 % break-through appeared after 1180 B.V. of effluent.
- Desorption and regenerationThe desorption and regeneration of ST column have been
studied. A feed solution which containing 0.6M NaNO^-O.OlM Ndand Am as tracers passed through column until the effluentreached 80^ break-through. Nitric acid (0.2M) was used fordesorption. After 10 B.V. acid passed through ST column, the
232
TABLE K
THE COMPARISION OP ADSORPTION PERFORMANCE 0V ST BEFORE ANDAFTER IRRADIATION
DOSE
(rad)
05*107
5*lo8
DISTRIBUTION. RATIO( IN BATCH)
ÜCd)Na/Ti=l/l Na/Ti=l/?
4.3X102 6.5*105.6x10* 7.0x108.4x10* 7.1X10.
EXCHANGE CAPACITY(IN COLUMN)(meq/g)
2.78
2.732.75
desorption ratio was higher than 70#. Then the ST column wasregeneration by 3M NaN03(adding NaQH to PH 10). After two cycleof adsorption-desorption, the adsorption performance of ST col-umn did not show any change.
- Irradiation resistanceIrradiation resistance tests for ST have been accompli-
shed, and the result are presented in Table IZ . It shows thatboth the exchange capacity and distribution ratio of ST did'ntchange after up to 5*108rad of irradiation.
Experiment of treatment of actual low-level liquid wastesFeed solutionThe low-level liquid wastes were taken from the waste stor-
age tank in Institute of Atomic Energy. The specific of whichwas l-2*lo" (/l, calcuted as gross ß -activity. Analytical datashowed that the nuclides they contained were mainly Sr, Cs andCo. For the ease of measurement, Sr was added to liquid wastesand the PH value of the liquid wastes was adjected to about 7.
Test facilityThe column was 5-6"mm in internal diameter and 11.5Cm high
containing two grams of ST adsorbent. The liquid wastes passedthrough ST column by peristaltic pump with a flow rate of 3.5-4.5 B.V./hr. The decontamination factor of Sr in 180 B.V. effl-
4 3uent was greater then 10 and remained no less than 10 after1180 B.V. passing through. The f-spectra of low-level liquidwastes before and after decontamination are shown in Fig. 9-10.
233
10
CHO
1C'
IÜ
1C
10
EHERGY(Nev)FIG. 9 THE V-SPSCTRA CF LOW-LEVEL LIQAID WASTES BEFORE
DECOIITAriIiïATICH 0? ST COLUMN
'Co
10 —————————————————————————————————————————————————EJïERCY(Hev)
PIC. 10 THE K-SPECTRA OF LOW-IE/EL LIQUID WASTES AFTERDECOIITAMIiJATIUJ OF 3 T COLUWJ
The adsorption of Cs was interfered by some ions presented inliquid wastes to be treated. After the liquid wastes were treatedby flocculation, the decontamination factor(DoP'e) of Cs raised.The D.P. of Co in low-level liquid wastes was very poor. However,when the liquid wastes passed through an anionic resin column,more than 95^ of cobalt was removed, checking the data of Ü.F.of Co in Table V, it is appears that cobalt exists as anioniccomplex state in liquid wastes«
234
TABLE SDETERMINATION OF LOW-LEVEL LIQUID WASTES AFTER THREE-COLUMN (AVERAGE EFFLUENT OF 1180B.V.) PURIFICATION
LOW-LEVEL LIQUID WASTESDNTRATED PRETREATED WITH
FLOCCULATION
D.F. OF ST COLUMNEFFLUENT
D.F. OF ZEOLITE
COLUMN EFFLUENT
SrC3Co
SrCsCo
1.8X10?2.1X101.1
4.01.53X101
1.1*101.87X10*1.1
3.51.2X101
D.F. OF ANION RESIN CoCOLUMN EFFLUENT C° 9'7 l'°
The further decomtamination of Cs and Co existing in anioniccomplex state was accomplished by using a natural zeolite columnand a macroporous anionic resin column. A three—column purifi-cation system——ST column, zeolite column and anionic resin column-— was assembled for experiment. Table X is the analyticalresults of effluents of low-level liquid wastes passing throughthree column system. It was determined by 4096 multi channel)f-ray spectrometer.
After 3-column purification, gross fl-activity in effluentdropped down to 1X1(5 ci/1, almost reaching the permissible levelof direct discharge, 7xlo"'ci/l.
ConclusionThe cost of ST would be cheaper than other inorganic ion-
exchanger and commercial ST would be much cheaper than ST inLab-scale. After adsorption, ST could be sintered at 1000°Ctranslucent glass-like ceramic form, which reduced volume ofwaste and facilitated waste treatment, In treating actual liquidwastes, the ST-aeolite-anionic resin column system is simplerand easier to operate as compared with existing evaporationmethod. Moreover, it can be installed as a small mobile devicesuitable for liquid wastes treating under local condition.
235
R E F E R E N C E S1 Lynch, R. W. Dosch H. G. et ai., " The SANDIA solidifica-
tion process— A broad range aqueous wastes solidificationmethod" IAEA-SM-207/75 361
2 Schwoebel R. L. et al., " An inorganic ion exchange processfor partitioning high-level commercial wastes "NR-CONF-001 30?
3 Dosch R. G. The use of titanate in decontamination ofdefence wastes solidification SADD-78-0710 (1978)
4 Johnstone J.K. Headly T.J. Hlava P.P. and Stohl F.V." Characterization of a titanate based ceramic for highlevel nuclear waste solidification (Scientific basis fornuclear wastes management ) 1^ (1979) 211-217
5 Schulz W.W. Koenst J.W. and Tallant D.R. "Application ofinorganic sorbents in Actiniae separation process "(Actinide Separation ) (1980) 17.
6 Forberg S. Westermart T. Arnek R. et al., "Fixation ofmedium level wastes in titanate and zeolites:progresstoward a system for transfer of nuclear reactor activitiesfrom spent organic to inorganic ion exchanges" ( ScienceBasis Nuclear Wastes Management ) 2 (1979) 867.
236
ION-EXCHANGE PROPERTIES OF ZEOLITES AND THEIR APPLICATIONTO PROCESSING OF HIGH-LEVEL LIQUID WASTE
T. KANNO, H. MIMURAResearch Institute of Mineral Dressing and Metallurgy,Tohoku University,Sendai, Japan
Abstract
Zeolites which have framework, structuresare excellent inorganic ion exchangers havinghigh stabilities to radioactive irradiation.
Their ion exchange properties and thermaltransformation have been studied for the purposeof their application to the separation of cesiumand strontium from the high-level liquid waste(HLLW). Mordenite, one of the zeolites, selec-tively exchanged cesium ion with high selectivitycoefficient more than 1000, thus cesium ion canbe preferentially exchanged into the mordeniteeven in the acid solution around pH 1, and sepa-rate from other metal ions in HLLW. The selec-tivity for cesium ion decreases in the order,mordenite » zeolite Y > zeolite X > zeolite A.The cesium exchanged mordenite is transformed tocesium aluminosilicate (CsAlSi5O12) by calcinationabove 1200°C.
On the other hand, strontium is selectivelyexchanged into zeolite A and its selectivity isreversed to that of cesium ion,zeolite A > zeoliteX > zeolite Y > mordenite. Strontium forms ofzeolite A, X and Y are recrystallized to triclinicstrontium aluminosilicate (SrAl2Si208) by calcina-tion above 1100°C.
The leaching rates of cesium and strontiumions from these stable minerals are less than10~9 g cm"2 day"1, which are lower by the threeorders of magnitude than those from borosilicateglass products. These excellent properties ofzeolites for the separation and fixation of cesiumand strontium can be applied to the partitioningprocess in the processing of HLLW and to the so-lidification.1. Introduction
Zeolites are crystalline aluminosilicateswith framework structures and have excellentcation exchange properties. Zeolite has somecavities occupied by cations such as Na, K, Caand water molecules. These cations can be ex-changed with other cations, and-the selectivitiesof zeolite for cations seem to be differentwith their structures, Si/Al atomic ratios andthe position of ion exchange sites.
237
Some of zeolites had formed naturally indeep geologic formation by hydrothermal reaction,and also many zeolites are synthesized artifi-cially by various methods. Some of these syntheticzeolites have been used as molecular sieves andcatalyzers in petrolium industry and others. Mostof the naturally occuring zeolites in Japan aremordenite, clinoptilolite or their mixtures.
In the atomic energy industry, the treat-ment and management of high-level liquid waste(HLLW) are the most serious problem. In fissionproducts, the most soluble and long lived radio-active isotopes are 137Cs and 90Sr, thereforethe fixation of these isotopes to insoluble ma-terials is the most valuable to security againstthe management of HLLW.
We have studied on the cation exchange pro-perties of zeolites and their application to thefixation of 137Cs and 90Sr, and some interestingresults were obtained.2c Zeolites
Synthetic zeolite A and X used in this studyare obtained from Toyo Soda Co. Ltd«, and syntheticzeolite Y and synthetic mordenite (Zeolon 900Na,abbreviated as SM) are made by Norton Co. Ltd.Natural mordenite (from Sendai area, Miyagi pre-fecture, Japan, abbreviated as NM) and clinoptilo-lites (from Futatsui, Akita prefecture and fromItaya, Yamagata prefecture, Japan, abbreviatedas CP) were also used.3. Ion exchange properties of zeolites for various
cationsEffects of pH on the distribution of Cs, Sr
and some of transition elements such as Cr, Fe,Co, Ce, Eu and Tb into Na- and H-zeolites fromsimple cation solutions have been investigated.
Distribution curves of these cations for NaAand NaX are shown in FIG. 1, and these distribu-tion curves of cations except for cesium havemaxima around pH 6 - 7. The distribution ratiosof these cations at maxima are about 101* - 10s.
The distribution curves for NaY are alsosimilar to those of NaA and NaX zeolites„ De-creases of distribution ratios at low and highpH ranges seem to be mainly attributable tocompetitive exchange with hydrogen ion andhydrolysis of cations respectively. On the otherhand, distribution ratio of cesium ion decreasesmonotonously with decreasing pH.
In the cases of Na- and H-forms of mordenites,cesium ion is selectively exchanged into thesezeolites even at low pH range as shown in FIG. 2,therefore cesium ion may be separated from othermultivalent cations under this condition.
238
ra
10'
10
- NaA
Tb
10*
103
10'
10
NaX Tb '"V\Eu
Fe
,P .
Cs
Sr
1 2 3 4 5 6 7 8 9 1 0
l 2 3 4 5 6 7 8 9 1 0
pHeq.
FIG. l Cation exchange properties of various cationsinto NaA and NaX
1 2 3 4 5 6 7 8 9 1 0
10'
100 1 2 3 4 5 6
FIG. 2 Cation exchange properties of various cationsinto NaSM and~HSM
239
4. Selectivity of zeolite for cesium ionZeolite has an ability to exchange various
cations, but its selectivity is influenced by thestructure and the position of ion exchange sites.In the case of exchange of cesium with a largeionic radius, sodium ions located in large cagesor large channels can be easily exchanged withcesium, but sodium ions located in narrow channelsor frameworks are hardly exchanged with cesium byion sieve properties of zeolite.
Ion exchange isotherms of NaA for exchangereaction with casium are shown in FIG. 3. Thesigmoid curve seems to show the presence of twokinds of site exchangeable with cesium. Plottsof corrected selectivity coefficients against tothe equivalent fractions of Cs in NaA for Na - Cs ex-change are shown in FIG. 4. In this figure,upper line shows the corrected selectivity.co-efficients at the sites in large a-cages (site1), in which cesium is more selective than sodium.
1.0
0.8
0.6
0.«
0.2
0.2 0.4 0.6 0.8 1.0
FIG- 3 Ion exchange isotherm of NaA for Cs exchange
1.5
1.0
0.5
-1.5
-2.0
-2.5
Site 2
0 02 0.4 0.6 0.8 1.0Zcs
FIG. 4 Ion exchange selectivity of Na-Cs exchange
240
Lower line shows the corrected selectivity co-efficients at the sites in the plane of 6-memberedring (site 2), in which sodium is more selectivethan cesium.
In NaA, 4 sodium atoms in large a-cage ofunit cell are easily exchanged with cesium, whileother 8 sodium atoms in the planes of 6-memberedring are hardly exchanged with cesium. Fromthese results, the cations in a-cage and in theplane of 6-membered ring can be considered ashydrated cations and bared cations, respectively.
Ion exchange isotherm of NaSM for exchangewith cesium are shown in FIG. 5. This simplecurve seems to show that the site exchangeablewith cesium is only one. Ion exchange selec.-tivity for Na - Cs exchange in NaSM are shownin FIG. 6, and a simple straight line was ob-tained.
0.2 0.4 06 08 1.0
FIG. 5 Ion exchange isotherm of NaSM for Cs exchange
3.0
2.5 -
20
1.0
0.5
SM
I________I________I________I
0 0.2 0.4 0.6 08 1.0
FIG. 6 Ion exchange selectivity of Na-Cs exchange
241
From this result, it is known that theselectivity coefficient of NaSM for cesium ex-change is more than 1,000 at initial exchangewith cesium which is very large than that ofother zeolites. The order of the selectivitiesfor cesium exchange is NaSM » NaY > Nax > NaA.5. Thermal transformation of Na-, Cs- and Sr-
zeolitesIt is well known that zeolites are alumino-
silicates having three dimentional structures,and include the various cations in their frame-work structures. After zeolite structure arecontracted by heating, cations contained inzeolites may become to be insoluble to water,since these cations are located in the alumino-silicate frameworks of zeolites, zeolites con-tained cesium and strontium may be also trans-formed to cesium and strontium aluminosilicatesby heating at high temperature.
Thermal transformations of Na-, Cs- and Sr-zeolites are investigated by means of differ-ential thermal analysis (DTA), thermogravimetricanalysis (TGA) and X-ray powder diffraction, inorder to know the possibility of fixation of137Cs and 90Sr.
Phase transformation of Na-, Cs- and Sr-forms of zeolites at high temperature are summa-rized in TABLE I. In the case of Na-A zeolite,the mixture of carnegieite and nepheline wasobtained at 900°C, and then changed to a simplenepheline phase. Cs-A zeolite is transformed tothe mixture of nepheline and pollucite, one ofthe cesium aluminosilicate, above 1200°C.Strontium form of A zeolite is transformed tohexagonal strontium aluminosilicate at 1000°Cand then changed to triclinic strontium alumino-silicate above 1200°C.
Sodium form of X zeolite changed to amorphousphase at 850°C, and then recrystallized to nephelineabove 1000°C. Cesium form of X zeolite is finallytransformed to pollucite above 1000°C, through theamorphous phase and in this case no nephelinephase was detected. The thermal transformationof Sr-X zeolites is similar to that of Sr-Azeolite.
Sodium form of Y zeolite was completelydestroyed at 800°C and does not formed anycrystalline phase. On the other hand, CsYformed pollucite phase above 1100°C, and SrYchanged to strontium aluminosilicate (SrAl2Si208)at 1200°C. Some different results between X-and Y-zeolites may be due to the difference inSi/Al atomic ratios.
Mordenite is most stable zeolite, hencethere is no change in its structure until about900°C. Cesium exchanged synthetic mordenite istransformed to a cesium aluminosilicate
242
TABLE I Transformation of zeolites at high temperature
CationZeolite
A
X
Y
SM
NM
CP
Na
900° >1000°Cam.+ •*• Neph.
Neph.850° >1000Am. -»• Neph.
>600=Am.>900°Am.
>1000°Am.
>1000°Am.
Cs
900° >1000°Neph.
Cam. -»• +Pol.
800-900° >4000°Am. -»• Pol.1000° >1100=Am. -*• Pol. '1100° 1200°Am. -»• R.
>1000°Am.
>1000°Am.
Sr
800° 900-1000° >110Q°
Am. -*• Hex. -»• Trie.
800-900° 1000° >1100°Am. -*• Hex. -*• Trie..
900= 1000s 1100* 1200sAm. - * • ? - » • Am.-»- Trie.1000-1200° >1300°
Tric.+C.+Q. -*- C.>1000°Am.
• >100.0°Am.
Am.: Amorphous, Carn.: Carnegieite(NaAlSiONeph.: Nepheline(NaAlSi04), Pol.: PolluciteCCsAlSi Og),Hex.: Hexagonal(SrAl„Sin00), Trie.: Triclinic(SrAl,Si00Q)Z / o / / oR.: CsAlSi 0 , C.rCristobalite, Q.: a-Qualtz
(CsAlSis012) at 1200°C, and strontium exchangedsynthetic mordenite changes to the mixture oftriclinic strontium aluminosilicate (SrAl2Si208),cristobalite and ot-qualtz at 1000 - 1200°C, andthen changes finally to cristobalite over 1300°C.On the other hand, Na-, Cs- and Sr-forms ofnatural mordenite and clinoptilolite change toamorphous phase over 1000°C, and no crystallinephase is detected.6. reachabilities of cesium and strontium from
calcined zeoliteThe leachability of 137Cs and other radio-
active nuclides from the solidified product ofHLLW is one of the most important parameters inthe safety assessment for the management of HLLW,because the leaching of radioactive nuclide fromsolid waste is the first step of the nuclidemigration with ground water in the repository.In order to realize the ceramic solidificationwith lower leachability, the authors have in-vestigated the solidification procedure by calci-nation of zeolites exchanged with cesium and
243
strontium as mentioned above. Therefore/ theleachability of cesium and strontium from thesecalcined zeolites was investigated.6.1 Effect of calcining temperature on the leach-
ability of cesiumAlkaline ions such as cesium are the mostsoluble elements in the HLLW, therefore, the leach-
ability of 137Cs is often measured as one of theparameters for security assessments. FIGURE 7shows the variation of the leachability of cesiumfrom calcined product into pure water withcalcining temperature. Leachability decreasedwith increasing in calcining temperature, andthat from the calcined CsA and CsX decreasedrapidly between 900° and 1000°C by the crystal-lization of pollucite (CsAlSiOit ) , and nepheline(NaAlSiOit) while the amount of leached cesiumfrom the calcined natural zeolites gradually de-creased with increasing in calcining temperature.The leachability from the calcined natural zeoliteswas a little higher than that from calcied A andX zeolites, since the mordenite and clinoptiloliteare more stable at high temperature than A and Xzeolites. The leachability of strontium is con-siderably lower than that of cesium.
800
N 600
U
0 400
o£
o CsA• CsX• CsNMD CsCP(Itaya)• CsCP(Futatsui)
900 1000 1100Temp. CC)
1200
FIG. 7 Variation of leachability of Cs into distilledwater with calcining temperature(Calcining time; 3 h)
6.2 Leaching rateThe manners of representation for leaching
results were contrived according to the objectand method of leaching test. The leaching ratepresented by ORNL is available and expressed asfollows[1].
244
Leaching rate = (g cm~2 day"1)
where X: Amount of a constituent present in thesample/ g
Y: Quantity leached from the sample inunit time, g/day
A: Surface area of the sample, cm2W: Weight of the sample, g
The evaluation of surface area necessary forthe calculation of leaching rate were made bymeans of BET method using nitrogen gas. Themeasured values of leaching rate are shown inTABLE II. The leaching rates into pure water wereabout 10~8 - 10~9 for cesium and 10~9 - 10~10 forstrontium, respectively. The value of leachingrate of cesium into distilled water was very small,and that from the calcined y zeolite was smallestin the synthetic zeolites by the formation ofpollucite. The amount of cesium leached fromcesium zeolites calcined at 1200°C and that ofstrontium leached from calcined A and X zeoliteswere under the detection limit of atomic absorp-tion analysis.
TABLE II Leaching rate of Cs and Sr for calcined zeolites. -2 , -I,(g cm day )
Zeolite
AXYSMNMCP
(Itaya)CP
(Futatsui)
CsDistilledwater2.5xlO~81.7xlO~81.3xlO~87.4xlO~8<2.3xlO~9<3.8xlO~9
<3.2xlO~9
Seawater1.2xlO"76.0xlO~8i.ixicT78.5xlO~8l.lxlO"81.3xlO~8
1.5xlO~8
SrDistilledwater<2.6xlO~10<2.6xlO-102.8xlO~ 95.0xlO~9
_
_
Seawater1.8xlO~Sl.SxlO"89.7xlO~91.6xlO~8
_
_
Zeolite A, X, Y and SM were calcined at 1100°C for 3 h.Zeolite CP and NM were calcined at 1200°C for 3 h.
7. Volatility of cesium from zeolites incalcination
Volatilization of Cs-zeolites at high tem-perature is investigated for the purpose of thesafety assessment. As shown in FIG. 8, volatili-zation curves of cesium are very d i f f e r e n t wi thzeolites and the amount of volatilized cesium
245
0.24
700 600 900 KOO 1100 1200 1300Temp. ("Ci
FIG. 8 Variation of volatilization rate with calciningtemperature
is increased with increasing temperature. It isshown in FIG. 9 that cesium is volatilized onlyat the first stage of heating at 1200°C, andexcept for CsX zeolite volatilization of césiumis depressed after about 40 min-heating. It seemsthat this depression is owing to the destractionof zeolite structures at high temperature. Vola-tilized species of cesium was determined as Cs20by mass spectrometry. Therefore/volatility ofcesium is decreased in inert gas flow such asargon. The volatilized amounts in calcination ofCs-zeolites except for CsX were very little (Table III).
0.2*
0.20 •
01« -8.
0.1Î -
n 0.08
0.04 <
80 120 160Tim«(min.)
240
FIG. 9 Variation of volatilization rate with calciningtime
246
TABLE irr Volatilization of Cs from zeolites(Calcining condition: 1200°C, 3 h)
ZeoliteCsACsXCsYCsSMCsNMCsCP
Air4.1xlO~212.0x10l.OxlO'17.3xlO~23.4xlO~22.4xlO~2
Ar2.9x10 *28.8x108.4xlO~2l.lxlO~24.4xlO~35.2xlO~3
Volatilization rate '(%) =Volatilized Cs_____Total Cs in zeolite x 100
8. ConclusionZeolites are very excellent materials for
adsorption and fixation of radioactive nuclidesin HLLW as mentioned above.
Mordenite is most selective to cesium ionin the zeolites, and cesium exchanged syntheticmordenite is transformed to cesium aluminosilicate(CsAlSisOia) at high temperature, which is verystable and had difficult leachability. Therefore,mordenite is very suitable for solidification of137Cs.
Zeolite A exchanged with Sr is transformedto strontium aluminosilicate (SrAl2Si208) bycalcination.
Since zeolites have very excellent propertiesfor the treatment of radioactive nuclides inhigh-level liquid waste, it is necessary to promatethe investigation on the application of zeolitesfor the nuclear fuel cycle in atomic energy de-velopment.
REFERENCE
[1] IAEA Technical Report Series, No.82 (1968) 101.
247
STUDIES ON INORGANIC ION EXCHANGERS -FUNDAMENTAL PROPERTIES AND APPLICATION INNUCLEAR ENERGY PROGRAMMES
R.M. IYER, A.R. GUPTA, B. VENKATARAMANIChemistry Division,Bhabha Atomic Research Centre,Bombay, India
Abstract
Investigations on inorganic ion exchangers havebeen carried out with a view to understand thefundamental properties and to assess their suitabilityin specific applications in the Nuclear Fuel Cycle.
Method of preparation had a marked influence onthe structure and properties of this class ofmaterials, as in the case of crystalline thoriumarsenate (TA) and hydrous oxides, especially hydroustitanium oxide (HTiO). The new crystallinemodifications of TA, in contrast to the monohydrate,exchanged larger ions like Na* , K in addition to Li* -,due to the large interlayer spacings. Hysteresis inH*-Li* exchange was observed. The selectivitysequences for alkali, alkaline earth and transitionmetal ions on hydrous thorium oxide (HThO) andmagnetite (MAG) were similar to those observed forother hydrous oxides. While hysteresis in anionexchange involving uni-univalent anions on HThO wasobserved, cation exchange reactions involving Ca andCu * were found to be non-reversible.
Preliminary investigations on the usefulness ofhydrous oxides in the recovery of uranium from seawater, and purification of reactor coolant water arereported in this paper. Different hydrous oxideshaving varying acidities were screened for theiruranium sorption capacities. On HTiO, the mostpromising adsorbent among the hydrous oxides, theinfluence of anions on the sorption of uranium. thepossible mechanism of sorption of uranium fromcarborate solutions and under simulated sea waterconditions were studied. Investigations on thesorption of cationic corrosion products on hydrouszirconium oxide (HZrO), HThO and MAG upto 363 K and ofanions on mixed oxides. like Bi(111)-Th(IV) mixedoxides, from alkaline solutions were carried out with aview to test their utility in reactor coolant waterpurification systems.
Adsorbents when used in Nuclear Energyapplications are liable to see radiation fields, whichcan affect and/or influence their sorption properties.With this view, in-vivo •y-radiation effect on thesorption properties of some representative hydrousoxides has been studied and the results will bediscussed.
249
Results of our recent studies on the propertiesand mechanism of reactions of silver exchangedmolecular sieve for the removal of CH Ï are alsodiscussed.
1.0. INTRODUCTION
Investigations on inorganic ion exchangers likethorium arsenate (TA) [1.23. a variety of hydrousoxides and mixed oxides [3-14], ferrocyanide molybdate,ferrocyanide tungstate CIS] and sorbents likemolecular sieves have been carried out in ChemicalGroup. 8ARC. with a view:
1. to understand the fundamental aspects of ionexchange and/or sorption properties of thesematerials, and
2. to see whether these materials could findpossible application in Nuclear EnergyProgrammes, like purification of reactorcoolant water, recovery of uranium from seawater, treatment of radioactive wastes etc.,
Salient features of the studies will be summarised inthis paper. Results of preliminary experiments on theeffect of irradiation on inorganic ion-exchangers arealso included.2.0. BASIC STUDIES ON INORGANIC ION EXCHANGERS
2.1. Influence of Method of Preparation on theProperties of Inorganic Ion Exchangers
In general. the stability, structure and ionexchange and/or sorption properties of inorganic ionexchangers have been found to depend on the conditionsof preparation CIS]. Our experience has indicated thatit is applicable in the case of thorium arsenate (TA)and hydrous oxides also.
In the case of TA, slight variation in thepreparation conditions resulted in new crystallineforms of TA viz Th(HAs04> 2.5 H20 and ThfHAsO^) .4H 0. [1,2], compared to the one reported in tnelilerature, viz, Th(HAsO^) .H 0. [17]. The interlayerdistance of the new crystalline forms were largernamely, for TA.2.5H 0 it was 8.59 A and TA.4H 0 it was8.23 A - as compared to the material reported inliterature, TA.HgO (7.05A) [173. The new crystallineforms were capable of exchanging larger ions like Na ,K* in addition to Li*, while T A . H O could take up onlyLJL* [17]. 2
Anion and cation sorption properties of hydrousoxides were found to depend on the nature of the alkaliused for precipitation [3.4,7,8,10.11]. Thus, hydrousoxides prepared using NaOH had better cation sorptionand the ones prepared using NH OH had better anion
250
100 -
ÖDoc.o
ce.15I-ZUuttLü(L
0-4 0 -3.0 -20 -10
0 -
Fig. t Effect of method of preparation andcarbonate ions on the sorption of Uranium onhydrous titanium oxide (HTiO) [13]O.HTiOJD); • .HTiO(H}; A .HTiO(D-T) ; A,HTiO<H-T)CD-Direct precipitation; H- Homogeneousprecipitation; T- treated]
sorption properties. In the specific case of hydroustitanium oxide [13], it was established that when NaOHwas used for precipitation, excess Na* was incorporatedin the oxide matrix and when urea was used, thehydrolysis of Ti was incomplete. As a consequence,the two preparations had different sorptioncharacteristics (Fig.1). However, when excess Na* wasremoved by dilute acid treatment and completehydrolysis ensured by alkali treatment. the sorptionchareacteristics of the two materials were very similar.
Our experience. thus. indicates that for betterunderstanding of the properties of inorganic ionexchangers, conditions of preparations should becarefully chosen and impurities present in thematerials should be detected and removed by suitabletreatment.
2.2. Properties of Crystalline Thorium Arsenate [1.2]As mentioned
prepared andTh(HAsO, ), .2.5H.O,materia
earlier. the thorium arsenate (TA)studied had the composition,
and TMHAsO ) .*H 0. Thesestable chemically and thermally than
the Th(HAsO.)..H_0 reported in literature [17].als wereliss
Although both*tneK*. TA.2.5H 0 wasmore stable than
.•* •'•vp*v*«.9Vi ^.n J.At*V.L«llUr
materials exchanged with Li*, Na" andfound to be chemically and thermally
TA.4H.O. The exchangers underwent
251
4.0
10 12
Fig.2 Alkali metal ion/H exchange as afunction of pH on TA.2.5H 0 in aqueous andmethanolic media [13o,metal ion uptake in aqueous medium;•,metalion uptake in methanolic medium ;Q .arsenatereleased in aqueous medium;«.arsenate releasedin methanolic medium;A,Li uptake in aqueousmedium (reverse exchange>;ASarsenate releasedin aqueous medium (reverse exchange)
extensive hydrolysis during exchange. The maximum Li*ion uptake and the corresponding arsenate released,respectively, are: 1.75 and 1.20 mmol/g for TA„4H 0 atpH 9.0 and 3.42 and 0.28 mmol/g for TA.S.SH^O ai pHIt.8. In methanolic medium, it was possibleto incorporate significant amounts of Na and K , ionsin TA 2.5.H 0, without extensive release of arsenateions. Hysteresis in Li -H exchange was observed forTA.2.5H 0 (Fig„2).
The metal ion exchangedcrystal structure as compared toof water molecules associatedexchanged form followed the Samein the various cationic formexchangers [18]: Li*>Na*> K* .
indicated by the absence ofasspectra, andstructures.
bypresumably existed
materials had differentH*-form. The numberwith the metal ionsequence as observed
of synthetic organiclionThe water was not free,
bending modes in IRhydrolysed
H Opartially
252
2.3. Properties of Hydrous oxidesAs hydrous oxides behave as cation exchangers in
alkaline medium and as anion exchangers in acid mediumthe cation and anion exchange behaviour of some hydrousoxides have been investigated.2.3.1. Alkali metal ions
Sorption of alkali metal ions from alkalinesolutions either decreases with increasing ionic radiiof the cations (for example, Li*> Na*2. K* > Cs* > NH^* onhydrous thorium oxide [3]). and with increasinghydrated radii of the cations (for example,Cs - K >Na*> Li*, on magnetite [5])(Table I). This indicatesthat on hydrous thorium oxide, cations are partially orfully dehydrated on sorption, while on magnetitehydration of the exchanging ions is more or lessunaffected.
TABLE I: Sorption of alkali metal ions on hydrous thorium oxide (HThO) andmagnetite(MAG) from 0.1M MeOH solutions(Me: Li*.Na*. K*, Cs*, NH *)[3.5]
Metal ion Crystal ionicradii
Li* 0.68Na* 0.97
K* 1.33Cs* 1.67U U * 1 L n n .L • *r j
Amount sorbed (meq/g)HThO MAG
1110
0
.15 0.54
.03 0.68*
.04 0.78
.96 0.71
.46* From 0.13M solution.
2.3.2. Alkaline earth metal ionsSorption of alkaline earth metal ions is a
complicated process because of the contribution to thesorption from hydrolysis of the cations. As a resultof this, sorption show an abrupt increase over anarrow pH range [12].2.3.3. Transition metal ions
Sorption of transition metal ions from neutralsolutions on hydrous thorium and zirconium oxides,magnetite, La(111)-Fe(111) mixed oxides, followed thesequence: [3,4,5,9,1V],
Cu2* > Zn2* > Ni2* > Co2* > Mn2*The sequence parallels the Irwing-William seriesobserved in the case of stability of metal complexes.
253
H"h»IO~3N1.00.8•4 o.«
Z 0.45 0.2
0.0
1.0
0.8«eU 0.6
0.2
0.0 0.4 0.8 1.0(Cl) C/CC
(0)
0.00.0 0.2 0.4 0.6 0.8 1.0
ee(c<(b)e/ee(Co)
isotherms onFig,3l»} NO --Cl ExchangeHThO[7]_ __ _A. N03"-» Cl"; A. Cl"-» N03"
(b) NH.*--Ca * Exchange isotherms onHThOm — 2 + — 2 + —Ca ;*.Ca -» NH
This indicates that surface hydroxyl groups of thehydrous oxides could be considered »9 ligands andsorption of cations viewed as surface complexformation.2.3.4. Hysteresis in ion exchange isotherms
Hysteresis in anion exchange involving simpleanions like NO ~ . Cl". CNS" and SO ", at constantsolution acidity and ionic strength on hydrous thoriumoxide (HThO) has been observed for the first tim® £73(Fig.3a). Subsequent to our reporting, hysteresis inanion exchange on commercially available hydrous oxideshas been reported C191 , although, in general, it isclaimed that anion exchange reactions on hydrous oxidesare reversible £20.21,22]. The reason for thehysteresis in anion exchange is not yet clear. Onecould, however, tentatively suggest the differences inthe structure of the exchanging sites (for example,5Me-OH2*N03~ , IMe-OH2*Cl", where =Me-OH is a surfacehydroxyl group), in the forward and reverse exchange as
cause of the observed phenomenon. Furtherinvestigations are needed in this area to clearlyunderstand the anion exchange behaviour on hydrousoxides.
Certain cation exchange reactions on HThO havebeen found to be non-reversible C9], for example,between alkali metal ions and Ca .pairs of transitionmetal ions, like Cu -Zn , Cu -Ni . Cu -Co . Thenon-reversibility could be due to selective binding ofions as in the case of Ca * or due to specificsorption. which include electrolyte sorption, as in theease of Cu *. However, NH *-Ca * exchange exhibitedhysteresis (Fig. 3b). A combination of bufferingaction of NH4* and precipitation of Ca2* in solution,has been proposed to explain the observed behaviour.
254
2.4. Effect of t-radiation on the sorption of MetalIons on Hydrous Oxides
The impetus gained in the study of inorganic ionexchangers is due to the radiation stability of thesematerials [231 - It is to be expected that thesematerials, when exposed to radiation in dry conditions,would not appreciably change their physical andchemical characteristics. However, when exposed toradiation in contact with aqueous solutions, as inactual applications, the inorganic ion exchangers couldundergo some changes and hence behave differentiallyas a result of formation and interaction of radiolyticproducts of water such as H, OH, e~ H 0 etc. [2*1. Notmany studies have been reported with regard to theinfluence of in-situ radiation on the sorptionproperties of inorganic ion exchangers.
Some preliminary investigations have been carriedout with HTiO and magnetite (MAG) (representing hydrousoxides of varying acidities) (Fig.4), to find out theinfluence of ^-radiation field on their sorptionproperties. These hydrous oxides also find possibleapplication in radioactive waste disposal (HTiO) and inreactor coolant water systems (MAG). Sorption of Liwas studied, because it is a non-specific ion and itssorption would indicate the changes in the oxides dueto -f-radiation .
9 0.6xo" 0.6o>CC.UJ°- 0.4OUJO)ITo 0.2
0.0
I I I I I 7o-MAG , •- HTiO
5y
Fig. 4hydrous
Sorption oftitanium
Li on magnetite ( MAG ) andoxide(HTiO) from
LiCl*LiOH( -0.0 IM) solutions
The salient results of the investigations arelisted in Table II. Within the limits of theexperimental error, there appears to be no change inthe uptake of Li on the two hydrous oxides even afterin-situ t-dose of >17 Mrad. The significantobservation in the study was that in presence ofradiation, the pH of the supernatant solution incontact with HTiO was higher than the unirradiatedsamples, in contrast to the pH of the supernatant
255
TABLE II: Effect of I -radiation on the sorption of Lland hydrous titanium oxide (HTiO).
on magnetite (MAG)
Time
hNo irradiartion
pHInitial Equil.
Amt. of Li*sorbedmeq/g(±0.02)
Initial
With irradiationpHEquil.
Amt. of Li*sorbedmeq/g(±0.02>
MAGNETITE
4
24
48
67
HTiO
4
24
48
68
11.11.11.11.
11.11.11.11.
27
27
27
28
5757
61
61
10.92
10.95
10.72
10.76
9.578.15
7.76
a. oo
0.0.0.0.
0.0.
0.
0.
0507
05
08
5858
60
60
11.11.11.11.
11.11.11.11.
27
27
27
28
57
57
61
61
10.9010. 77
10. 38
10.4
10.119.00
8.50
8.27
0.0.0.0.
0.0.0.0.
07
06
06
10
5657
57
60
* Time of equilibration or irradiation..Conditions* Oxide 0.4g; Soin. 25ml; for MAS a mixture of LiOH+LiCl ( 0.01M)for HTiO 0.01M LiOH. Strength of f-source 0.26M rad/h.
solution in contact with MAG, which was lower. Therewas no apparent change in the physical nature of MAG.HTiO changed its colour to yellow on exposure toradiation, presumably due to the interaction of H 0 (aproduct of radiolysis) with Ti of HTiO.
Further work is needed in order to understandbetter the pH changes observed in this study.
3.0. APPLIED STUDIES WITH RESPECT TO NUCLEAR ENERGYPROGRAMMES
3.1. Purification of Reactor Coolant WaterOne of the envisaged application of inorganic ion
exchangers was their use as possible purificationmaterials for the reactor coolant water, because of theexpected thermal stability of this class of materials[253. Magnetite (MAG) appeared to be an attractivematerial as it is already present in the reactor andits use could not add a new component to the reactorcoolant system. Sorption studies with MAG would alsolead to a better understanding of the activitytransport phenomenon in reactor coolant systems.
256
TABLE III: Sorption of corrosion product cations on hydrous oxides from amixture of ions at 363 K under flow conditions [4]
Influent Amount of traced ion sorbed, pg/gHydrous _ - _ _ _ _ _ - - _ _ . _ _ - - - - - - - _ - - - - - - - - - _ _ _ _ - - - - - - - _ _ - - - _ _ _ _ _ - . . _ . . _ _ _ . _oxide Total amt. Amt. of
of ions traced ion Cr Fe Mn Copresent in present inmixture the mixture(ug) lug)
HZrOHThOMAG
6400
4000
6400
1 600
1000
1 SOO
816
350
1184
560
340
688
256
70
0
224
50
13Conditions: Hydrous oxide lg; Soln. pH -6; Concn. of each ion 1 ug/ml; flow
rate 50 ml/h.
3.1.1. Sorption of corrosion product cationsSorption of corrosion product cations (Fe *, Cr *.
Co , Mn ) in trace concentrations on hydrouszirconium oxide (HZrO), HThO and MAG under dynamicconditions upto 363 K was investigated [43. Theperformance of the hydrous oxides is given in TableIII. At higher pH, the sorption of Co * and Mn onthese hydrous oxides increased, presumably due topresence of hydrolysed species.3.1.2. Studies on mixed oxides
With a view to develop an adsorbent capable ofsorbing anions from alkaline solutions, differenthydrous mixed oxides were prepared by incorporating ametal ion of higher valency into the bulk oxide matrix.Among the several combination of mixed oxides tested,like Bi(III)-ThdV) , Bi ( 111 )-Zr ( IV ) . Bi( 111 )-Ti ( IV) .TidV)-Sb(V) etc. Bi( III)-Th(IV) , mixed oxide was foundto be most promising, sorbing Cl" ions from alkalinesolutions [63.
Attempts were made to improve the halide ionsorption of Bi(111)-Th(IV) mixed oxide by incorporatingAg* ion in the matrix [8]. By loading Ag ion onBi(III)-Th(IV) mixed oxide, by equilibrating with AgNOsolution, no beneficial effect was observed. However,sorption of halide ion increased with increasing Ag*content in the mixed oxide, when the oxide was preparedby coprecipitating Ag* along with Bi(III) and Th(IV)(Table IV).
These materials may find possible use in theremoval of halide ions from reactor coolant water andfrom spent fuel storage tank water.3.2. Removal of Organic Radioiodine using Ag*
Molecular SieveAnother aspect of our current interest is in
evaluating the basic properties of silver exchanged
257
TABLE IV: Sorption of halj.de ions on hydrous Bi ( 111 ) -Th ( IV ) mixed oxides(Bi/Tn . 4) [8]
TestSolution
0 . 1M NaCl0 . 1M NaBr0 . 1M KIO.OSM KIO.OSM KÏ »10 M NaOH
0 . 05M KI +10 M NaOH
O.OSM KÎ »10 M NaOH
0.05M K I *10* NaOH
Bi-Th
Pureoxide
1 .071 .040.94
1 .0
0.98
0.96
0.92
0.42
Sorptionmixed oxideAg loaded
oxide
0.910.901 .051 .011 .05
0.89
0.90
0.49
capacityBi
Ag*0.
0.
0.
0.
0.
0.
0.
-
-
, meq/g-Th-Ag coprecipita ted oxidepresent in the mixed oxide, meq/g10 0. 29 0.57 1.38
16 0.46 0.70 1.38
12 0.72 0.84 1.76
17 0.50 0.79 1.46
16 0.45 0.77 1.46
15 0.47 0.77 1.51
19 0.47 0.77 1 .54
„
-
(a) Mixed oxide precipitated using NH OH; (b) Ag loading on the mixed oxide0.08 meq Ag*/g of oxide; tc) precipitation effected using NaOH.Conditions : Oxide 0.5 g; solution 25 ml; equil.time 24 h.
TABLE V: Quantity of CH I adsortoed/consumed [26] (x 10" * 0.5 ml)
Temperaturem298
323375
423
A B CHydrated Dehydrated Hydrated
13X 13X Ag* 13X
0.84 23.6 13( 16
0.56 21.1 18
0.42 16.5 23
13.1 21
. 8
.5)*
.2
.3
. 7
0Dehydrated
Ag* 13X
3 1(363028
26
. 5
.2)*
. 1
. 2
. 2Molecular sieve used Ig ; Silver content in Ag 13X,( )* with concommittent t dose o-f 0.4M rad h .
1 .5 mmol
molecular sieves used for the removal of volatileorganic radioactive iodine species from off gas streamsof nuclear installations. The development of anefficient and at the same time less expensive adsorbentthan silver loaded matrices demands a knowledge of thebasic chemical processes involved in thechemisorption-decomposition of CH I and of the role ofsilver.
258
Our studies [26] covered identity and yields ofchemical products formed in the reaction of CH I overplain and silver exchanged (16X) 13X molecular sieve asa function of temperature, humidity and ^-radiationfield. Mechanistic aspects of the reaction wereinferred from supportive evidence . from X-rayphotoelectron spectröscopy. The main results are givenbelow:
(i) Hydrated (182) plain 13X permits only negligiblephysiabsorption of CH I which is uninfluenced byconcommittent f-radiation field (Table V).
, (ii) Dehydrated (at 575 K for 6 hours) plain 13Xshows significant adsorption capacity for CH I whichdecreases with increasing temperature and the capacityis unaffected by ^-irradiation (Table V). No reactionproducts are observera at temperatures below 500 K andthe adsorbed CH I could be desorbed at 600 K. Repeatuse of the same matrix for upto 5 cycles (adsorptionand regeneration) shows no measurable change in itsbehaviour. Ingress of moisture during adsorption cycledrastically reduces the adsorption capacity.(iii) Exposure of CH I to hydrated (18Z) silverexchanged (16Z) 13X produces two products namely. CH OHand CH OCH , th'e yields of which increase withincreasing temperature. The adsorption capacityincreases with temperature and -y-radiation dose (TableV). It appears that more CH I is consumed than isrequired for a stoichiometric reaction between CH I andsilver which is particularly noticeable at highertemperatures (Table V, Column C). The reactionproducts CH OH and CH OCH ^do not react with boundiodine if present in Ag* 13X to regenerate againvolatile organic iodides under the influence oftemperature and/or f-radiation field.(iv) In the case of dehydrated Ag* 13X. although as
expected no oxygenated reaction products are produced,small amounts of the hydrocarbons CH and C H arereleased. the yields of which increase withtemperature. Beyond 575 K free iodine is also releasedfrom the matrix.
A few significant conclusions drawn from thisstudy are
(i) more CH3I is consumed over Ag* 13X than isrequired for a stoichiometric reaction betweenCH3 I and silver indicating participation of Na*of molecular sieve in the fixation of iodine.
(ii) CH3 I interaction with hydrated Ag* 13X isthrough the formation of methyl carbonium ionand
(iii) use of two dehydrated plain molecular sievebeds in series followed by a small bed ofhydrated Ag* molecular sieve could beconsidered as an attractive alternative to theuse of a single Ag* molecular sieve adsorberfor the removal of radioactive CH I.
259
TABLE VI: Sorption of uranium on hydrous oxides from different uranylsalt solutions [11]
Hydrous _ - - _ _ _ _oxide Salt •*
used
HTiO< H )
HTiO( Am)
HTiO(So)
HCeOI Am)HCeO(So)HZrO(Am)HZrO( So)HThOI Am)
HThO(So)
Uptake
Nitrate
0
0
0
0
0
0
0
00
. 86
.66
. 81
.31
.43
.20
.33
.05
. 15
of uraniumn .Chloride
Q00
00Q00Q
. 75
. 52
.68
. 16
.29
. 12
. 1 <>
.02
.09
mmol U/g ofSulphate
0000
Q
0
0
00
. 77
. 56
.60
. 42
.33
. 29
.36
.09
. 08
oxideTricarbonate
0.0.0 .0.0.0.0.0.0.
44
34
26
1?
25
1317
01
05
(H) Oxide homogeneously precipitated. (Am) Oxide precipitated using NH OH. (So)Oxide precipitated using NaOH.Conditions: UO * - 0.01M: pH 3.5; for tricarbonate soin. -9; oxide 0.3g:soin. 25 ml; contact time 24 h.
3.3. Recovery of Uranium from Sea-waterIn recent years there has been a growing interest
in developing sorbents for the recovery of uranium fromsea-water. With this view in mind, different hydrousoxides, HTiO, hydrous cerium oxide (HCeO), HZrO, andHThO were screened in order to find a suitable sorbentfor uranium [103. In general sorption of uraniumdecreased with decreasing acidic nature of the oxide,.Moreover, on acidic oxides like HTiO. HCeO, uranium wassortoed as uranyl ion and on less acidic oxides likeHZrO, HThO. electrolyte sorption, involving uranyl ionand anion, was predominant [103. Uranium sorption alsodepended on the salt used for the sorption experiment[ 1 1 3 (Table VI). Sorption was high when a nitrate wasused and was low when carbonate was used. Among thehydrous oxides investigated, HTiO and HCeO were foundto be more suitable for the sorption of uranium fromcarbonate solutions [113.
More- detailed investigations were carried out withHTiO using uranium in millimolar concentrations [13].HTiO prepared using NaOH did not exhibit good sorptioncharacteristics as Na* was present in the material.The sorption characteristics improved on removing theNa* impurity (Fig.1). HTiO prepared using urea hadlarge surface area and had only partially hydrolysedTi , and exhibited good sorption property as such. Inlarge scale preparation and operation, this particularmaterial could prove useful, unlike theNaOH-precipitated material. because the partially
260
LOG (ANION), M-3.0 -2.0
100 --1.0T- 100
10
Fig.5 Sorption of Uranium on HTiO[13]as a function of : •.pH ;o. NO ~ concentration;A,SO " concentration. Equilibrium pH values foranions are given in brackets.4.Precipitation ofuranium as a function of pH.
hydrolysed Ti * sites could also act as acidic sites,in addition to the surface hydroxyl groups, and bindmore uranyl ions.
pH of the solution had a marked influence on thesorption of uranyl ion (Fig.5) than the added anionslike Cl" . NO " , SO ". Sorption of uranium fromcarbonate solutions was limited by the acidity of thesurface groups, because the sorption involved breakingof the uranium-carbonate complex present in solution[13] (Fig.1). The sorption of uranium from simulatedsea-water was restricted by the competive sorption ofCa present in solution
REFERENCES
[1] SARPAL. S.K., GUPTA, A.R., J. Inorg. Nucl. Chem..il (1981> 1347.
[2] SARPAL. S.K.. GUPTA. A.R.. J.Inorg. Nucl. Chem..43 ( 1981 ) 2043.
[3] VENKATARAMANI. B.. VENKATESWARLU, K.S., SHANKAR,J.. Proc. Indian Acad. Sei.. 87A (1978) 409.
[4] VENKATARAMANI, B., VENKATESWARLU, K.S.. SHANKAR,J.. BEATSLE. L.H.. Proc. Indian Acad. Sei.. 87A(1978) 415.
[5] VENKATARAMANI B., VENKATESWARLU. K.S.. SHANKAR,J., J.Collois. Inrterf. Sei.. 67 (1978) 187.
261
[63 V E N K A T A R A M A N I , 8., V E N K A T E S W A R L U . K.S.. SH A N K A R ,J., 8EATSLE, L.H., 3,Colloid Interf. Sei.. Ifi(1930) 1.
[7] V E N K A T A R A M A N I , B.. V E N K A T E S W A R L U . K.S., J. Inorg.Nucl. Chem., 42 (1980) 909.
.[8] V E N K A T A R A M A N I , 8., V E N K A T E S W A R L U , K.S., Proc.Indian Acad. Sei., (Chem. Sei.), H (1980) 241.
[93 V E N K A T A R A M A N I , 8., V E N K A T E S W A R L U . K.S., J. Inorg.Nucl. Chem., .il (1981) 2549.
[103 MAHAL, H.S., VEN K A T A R A M A N I , 8., V E N K A T E S W A R L U .K o S . , J. Inorg. Nucl. Chem., 43 (1981) 3335.
M 13 MAHAL, H.S., VE N K A T A R A M A N I , 8., VENKATESWARLU,K.S., Proc. Indian Acad. Sei., (Chem. Sei,), 91(1982) 321.
[12] MAHAL, H.S.. VE N K A T A R A M A N I . 8.. Unpublished work.[13] V E N K A T A R A M A N I , 8., 6UPTA, A.R., Paper presented
at the Int. Meeting on the Recovery of Uraniumfrom Sea Water. Tokyo, Japan (Oet*. 17-19, 1983)),Paper No. 32.
[14] MAHAL, H.S., VENK A T E S W A R L U , K.S., Indian J.Chem., 16A (1978) 712.
[153 B H A R G A V A . S.S., VENKATESWARLU, K.S., IndianChem.. 14A (1976Î 800.
[163 CLEARFÏELO. A., NANCOLLAS. S.H., BLESSING. R,,"Ion Exchange and solvent Extraction" (MARINSKY,J.A., M A R C U S , Y. Eds.) Vol.3, Marcel Oekker, NewYork. NY (1973) 1.
[173 A L B E R T I , G., MASSUCCI, M.A.. J.Inorg.Nucl.Chem..32 (1970) 1719.
[183 N ANOAN. D, . GUPTA, A.R., Indian 3. Chem., 12,(1974) 808.
[193 DYER, A.. MALIK, S.A., 3. Inorg. Nuel. Chem., 43(1981) 2975.
[203 NANCOLLAS. G.H., PATERSON. R., J.Inorg.Nucl.Chem.29 (1967) 565.
[213 R U V A R A C . A. Lj., TRSTANJ.. M.I.. J. Inorg. Nuel.Chem., 34 ( 1972) 3893.
[223 MISAK. N.Z., MIKHAÏL, E.M.. J.Inorg. Nucl. Chem.,43 (1981) 1903.
[233 E6ROV, E.V.. NOVIKOV, P.O., "Action of IonizingRadiation on Ion Exchange Materials", IsraelProgr. Sei., Transi., Jerusalem (1967).
[24] SELLERS, R.M., "The Radiation Chemistry ofColloids - A Review". Central Elect. Sen, Board,Berkeley Nucl. Labs., Report, R O / B / N 3707 (1976).
[253 VENKATARAMANI. 8,, V E N K A T E S W A R L U , K.S., "WaterChemistry Studies V!: Role of Inorganic IonExchangers in the Purification of Reactor CoolantWater", Bhabha Atomic Research Centre Rep. BARC/1-123 (1971).
[263 B E L A P U R K A R , A.D., ANNAJI RAO. K., GUPTA. N.M.,IYER, R.M., Surf. Technol., 21 (1984) 263.
262
ION-EXCHANGE SELECTIVITIES ON ANTIMONIC ACIDSAND METAL ANTIMONATES
M.ABEDepartment of Chemistry,
. Tokyo Institute of Technology,Ookayama, Meguro-ku,Tokyo, Japan
Abstract
Antimonic acids and metal antimonates as the inorganicion-exchangers exhibit extremely high selectivity for acertain element or group of elements for comparison withsulfonated polystyrene ion-exchange resin. Variousantimonic acid materials have been obtained with differentcompositions and ion-exchange properties, depending on themethod of their preparations as well as on aging. Thespecies can be divided into three groups -crystalline,amorphous and glassy. The affinity sequence for alkalimetal ions shows Li < Na< K<Rb< Cs on the amorphous andglassy acid, while Li < K<Cs< Rb < Na on the crystallineacid. However, slightly different affinity series havebeen reported by other authors, Cs< Li< Rb< K< Na or Rb=Cs <Na on the crystalline acid. Similar disagreementbetween various authors exists for alkaline earth metalcation series: Mg<Ca<Sr< Ba, Ra- Ba< Ca< Sr, Mg< Ba< Ca<Sr. These selectivities for various metal ions on thecrystalline antimonic acid are discussed in the terms ofrate of adsorption, steric effect, and host-guestco-relations.
The antimonates of quadrivalent metal, such as Ti, Zr,and Sn, act as cation exchangers. The selectivitysequences for alkali metal ions show Li< Na< K < Rb < Cs onZirconium antimonate, Na < K < Rb < Li < Cs on titaniumantimonate and Na < K < Rb < Cs < Li on tin antimonate withmicroamount of metal ions in acid media. However, titaniumand tin antimonates exhibit a regular selectivity sequencefor'alkaline earth metal cations; Mg < Ca < Sr < Ba. Theselectivity on the tin antimonate varies with the loading ofthe alkali metal ions on the exchanger for steric effect.The selectivity sequence for macro amounts shows Li > Na > K >Rb> Cs. These selectivities are discussed in the terms ofsteric effect and entropy changes of the ion-exchangereactions.
1. INTRODUCTIONSynthetic inorganic ion exchangers have both thermal
resistance and radiation stability and have been applied inreactor chemistry such as reprocessing of irradiated nuclearfuel. Much attention has been paid recently to theselective adsorption properties which are not found with thecommercial organic resins.
During the last two decades, new inorganic ionexchangers have been synthesized and studied extensively fortheir ion exchange properties by various authors [1-7].
263
Some of the inorganic ion exchangers exhibit excellentselectivities with respect to certain element or group ofelements. The selectivity on the inorganic ion exchangersincreases usually with decreasing concentration of theelement to be separated. However, the change in theselectivity is much depended on chemical species of theelements, co-ion present, and on the crystallinity of theadsorbent. In the chemical processing of nuclear reactorcycle, the separation of the radionuclides may be neededfrom a large amount of the bulk components. Overall ionexchange reaction from microamount to macroamounts of theelements should be studied for this porpose. Furthermore,high stability in the acidic solution at high concentrationmay be required otherwise in high resistance to hightemperature and radiation field.
Various antimonic acid materials have been obtainedwith different chemical compositions, characteristics, andcation exchange properties. The species can be dividedinto three groups- crystalline( C-SbA ), amorphous( A-SbA ),and glassy( G-SbA ) [ 8-11]. Compounds similar to C-SbAhave been reported as polyantimonic acid ( PA ) and hydratedpentoxide ( HAP ) by different authors [12, 13].
Quadrivalent metal antimonates as ion exchangers havebeen studied first by Abe and Ito [14]. A number of metalantimonate has been reported by various authors.Especially, antimonates of titanium( Ti-SbA ) and tin(Sn-SbA ) exhibit an excellent selectivity for lithium ions[15, 16].
This paper describes comparison of the selectivity ofcations on various antimonic acid and metal antimonates.2. Experimental2. 1. Preparations of antimonic acid cation exchangers.
The antimonic acid cation exchangers of differentspecies were prepared as described previously F 10 ].2. 1. 1. Crystalline antimonic(V) acid( C-SbA ). A 75 mLof antimony pentachloride was preliminarily hydrolyzed by 75mL of cold water, and was then hydrolyzed by 5 L of water.The precipitate obtained was kept in the mother solutionat 40 °C over 20 days, and then washed with colddemineralized water until free from chloride ions using acentrifuge (about 10000 rpm). After drying theprecipitate, the product was ground and sieved in theappropriate fraction of 100-200 mesh size. The collectedsample was rewashed with cold demineralized water in orderto remove any adherent fine particle and to obtain a clearsupernatant solution in the subsequent experiments.2. 1. 2. Amorphous antimonic acid ( A-SbA ).
Similar procedures to those for C-SbA were carried out,except aging overnight in cold mother solution. Thesubsequent procedures are the same as above.2. 1. 3. Glassy antimonic acid ( G—SbA ).
The washed amorphous antimonic acid with colddemineralized water was dissolved in a hot water and cooledrapidly. The water in the solution was as quicklyevaporated as possible by using fan. After drying, thesubsequent procedures are the same as above.
264
2. 2. Preparation of metal antimonate.The SnSbA exchanger was prepared as described
previously[ 15] .2. 2. 1. Preparation of tin antimonate( SnSbA ).
A liquid antimony(V) chloride was prehydrolyzed with anequal amount of demineralized water in order to preventfractional precipitation of antimonic acid by furtherprocedures. The 3.9M antimony(V) chloride solutionobtained was mixed with 4.5M tin(IV) chloride solution at 80°C. The mixed solution was immediately hydrolyzed in50-fold volume of demineralized water at 80°C. The whiteprecipitate was kept in the mother solution overnight, thenwas filtered and water-washed by using a centrifuge(about10000 rpm) until the pH value of the supernatant solutionwas higher than 1.5. After drying at 60°C, the product wasground and sieved to 100-200 mesh size. The collectedsamples were rewashed with water in order to remove adherentfine particle of SnSbA, and finally dried to 60°C.2. 2. 2. Preparation of titanium antimonate( TiSbA ) and
zirconium antimonate( ZrSbA )Similar procedures to those for SnSbA were carried out.
The solutions of titanium(IV) chloride and zirconiumoxide chloride were used instead of tin(lV) chloride.2. 3. Distribution coefficients(Kd).
The equilibrium Kd values of metal ions for microamountsv,ere determined with a batch technique. The exchanger wasintmersed in the metal salt solution containing hydrochloricacid or nitric acid at different concentrations inthermostatted bath. The initial concentrations of metalions were adjusted mainly to 10~4 M. The concentration ofthe metal ions in the exchanger and in the solution wasdeduced from the concentration relative to the initialconcentration in the solution. The concentration of metalions were determined by flame photometry or atomicabsorption spectrometry. The Kd values were calculated asfollows;
_ Amounts of metal ions in exchanger „ mL of solutionAmounts of metal ions in solution g of exchanger
The Kd vakues of metal ions for comparison werenormalized to a 1M nitric acid solution by extrapolating theKd values determined.2. 4. Overall ion exchange reactions
In the forward reactions, the exchanger( 0.10 or 0.20g)in H+ form was immersed in 10.0 or 20.0 mL of a mixedsolution of varying ratio of the metal nitrate/ nitric acidin a sealed glass tube with intermittent shaking atdifferent temperatures. The ionic strength in the mixedsolution was adjusted to 0.1 after attainment of theequilibrium.
In the reverse reaction, the ion exchanger in variousmetal ion forms, corresponding in weight to 0.10 or 0.20 gof the exchanger in H+ form was immersed in 10.0 or 20.0 mLof the mixed solutions with ionic strength of 0.1 afterequilibriation. The concentration of the metal ions in theexchanger and in the solution was deduced from theconcentration relative to the initial concentration in thesolution.
265
2. 5. Preparation of the exchanger in metal ion form.A O.Pî metal nitrate solution was charged continuously
at the top of the column of the exchanger in the hydrogenion form until the change in the concentration of the metalions was negligible between the effluent and feed.2. 6. Reagents. The antimony(V) chloride (Yotsuhata ChemicalCo. Ltd.) and tin(IV) chloride (Wako Chemical Co. Ltd.)were used without further purification. The other reagentsused were all of analytical grade.3. Theoretical Aspects
The ion exchange reactions of metal ions/H+ exchangesystem on the exchanger can be represented by the followingexpression;
n H Mn+ n H ( 1 )
+where bar refers to exchanger and M is metal ions withvalency n.
The thermodynamic equilibrium constant, K, of the abovereaction can be difined as:
_ _ _ , _ « . _ _ _ _ _ _
MV+ ?~n
fH+ H+where K„ is the selectivity coefficient, m + and mj.jn+ arethe molalities of HNOg and M(NC>3 )nin solution, and YH+ andY,,n+are the ionic activity coefficients of H and M"+ insolution. XH+ and X^n+are the equivalent fractions of theexchanging H and M in the exchanger phase and T + and ?,,are the activity coefficients of H+ and Mn+ in the exchangerphase, respectively.
The contribution of the ionic activity coefficiens insolution to the thermodynamic equilibrium constant can beneglected for uni-univalent exchange reactions, while notfor uni-multivalent reactions. The values of the ionicactivity coefficient ratio { y +/y n+)in solution have beenused in the following equation [17]:
nYH+
log ———— =log
2n
S = 1.8252
'±HN03 n S /I1+n
Y±M(N03)n
v in6 (X J_U \e
1 + 1.5/1
0 >1/23T3 >
4 )
where P is the density of water, £ the dielectric constantof water, and T the absolute temperature of the system.
The thermodynamic equilibrium constant can be evaluatedby using the simplified treatment of Gains-Thomas equation[18], assuming that the change of water content in theexchanger and entrance of anion from solution phase arenegligible.
In K = - ( Z„ - Z„ ) +n H n " d XMH M( 5 )
H MWhen the In K„ versus X are linear, the In K,, is given byH M n
= 4-606 In
266
Mwhere the C is the Kielland coefficient! 19], and (In KH)is a value of In K$ obtained by extrapolating XM to zero.The change in the free energy of the ion exchangereaction, A G° , was calculated per cation equivalent,
AG°= -( RT/ZM ZH ) in K 7 )wereThe change in the enthalpy, AH°, and the entropy,A SP
calculated in the usual way [ 20] .The theoretical cation capacity( 5.057 meq/g ) of the
exchangers was evaluated by assuming that one antimony givesone hydrogen ion available for cation exchange reaction.4. Result and Discussion4. 1. Selectivities on different antimonic acids.4. 1. 1. Distribution coefficient Kd ) of microamounts.
The determination of the Kd values gives usefulinformation for the separation of trace and microamounts ofthe elements in the fields of analytical chemistry,radiochemistry, environmental chemistry, and biochemistry.Nitric acid media have been used in order to obtainfundamental ion exchange data in the less complexing media,which are useful for application to the reprocessing ofnuclear fuel.
Equilibrium distribution coefficients of alkalimetal ions on C-SbA, A-SbA, and-G-SbA were plotted in Fig.1, values on Bio-Rad AG 50W-X8 being included forcomparison. The Kd values for alkali metal ions on theirinorganic exchangers, except C-SbA, increase in the order Li<Na < K < Rb < Cs as shown with sulfonate-type organicresins. The separation factors are more favorable than
3 -
-2
0.6 W 1.5• IREffective cryatal ionic radius ( A ) y 574
Nal.02K 1.38Rl>U9CS170
Fig. 1. The log Kd values of alkali metal ion on differentantimonic acidsC-SbA; crystalline, A-SbA; amorphous, G-SbA; glassyalkali metal ions; 10~4M, HNQj 1 M, total vol. 25 mL,exchangrer; 0.25 g.
267
•'chose obtained on Bio-Rad AG 50W-X8. The Kd values ofmetal ions on the Bio-Rad AG 50W-X8 are taken from the O-r.treported by Strelow [21]. The increased selectivity wasobserved as Bio-Rad AG50W-X8 <A-SbA< G-SbA. The adsorptiveproperties on the G-SbA showed essentially the same as thoseon the A-SbA, while the differences in the selectivity maybe due to that the apparent density of the G-SbA is higherthat of A-SbA. However, C-SbA showed an unusualselectivity for alakli metal ions Li< K< Cs< Rb< Na.An extremely low adsorption properties was observed for Lion C-SbA, while an extremely high Kd value was determinedfor Na even in a concentrated nitric acid solution.
Similar unusual selectivity was also found for othermetal ions on the C-SbA: an extremely low selectivity forfig, Ni9 and Al and an extremely high for Ca, Sr, Pb,Cd, and Hg ( Fig, 2).
10
Vs 6c
i i i i i i V i i i i i i i i i l l i i iOS u S U *>
Metal ions
Fig. 2. The In KM of various metal ions on C-SbA in IM HNO,H J
Conditions; same as Fig. 1.
On the organic ion exchange resins, the Kd values ofalkali metal ions are nearly constant for the solutioncontaining alkali metal ion at the concentration lower thanl m mol/L in a relatively high concentration of acid( >0.1 11 ). However, it has been known that the kd valueschange with the concentration of the metal ion loaded onsome inorganic ion exchangers, although the acidconcentration is unity and relatively high, because of theirsteric effects. The log-log plotts of the Kd values vs[H"1], were demonstrated on Fig. 3, with the change in theconcentration of Mg and Ba [ 22] . The ion exchangeideality is maintained up to relatively high concentrationof 10~%, indicating that the slopes, d log Kd/ d log [ H +] ,are about 2. The Kd values increase with decreasing theconcentration of the metal ions even in the sameconcentration of hydronium ions.
A systematic study of the distribution coefficient ofmetal ions on C-SbA has been reported by Abe and co-workers.Girardi and co-workers have reported that about 3,000adsorption and elution cycle were carried out in a
268
I10'
Wjjo
0.01 0.1[HNOj] (mol*/l)
1.0
Fig. 3. The Dependence upon the initial concentration of themetal ions for the Kd of Mg and Ba on C-SbA.
Initial concn.; (---)l(r2M, (——)10~3M, (_._)lCr4M,
standardized way over eleven adsorbents ( including 9inorganic ion exchangers ) from various media forpreliminary screening of possible material to use forradiochemical separation [12], A comparative study of ion'exchange selectivity between C-SbA and HAP reported byGirardi et al. was carried out by Abe. The evaluatedselectivities on HAP are essentially compatible with thoseof C-SbA for 18 but out of 4 metal ions [ 23] .4. 1. 2. maximum uptake of metal ions on C-SbA
The maximum uptakes were determined by the columntechnique mentioned at the section 2. 5. The maximumuptakes on inorganic ion exchangers have been known to varyextensively with the nature of cation species, temperature,concentration of cation in solution, and co-ion present,while the organic ion exchange resins have fairly constantcapacity due to their elastic structure. The maximumuptakes for alkali and alkaline earth metal ions, except Liand Mg on the C-SbA decrease with increasing effectiveionic crystal radii [ 23 ] ( Fig. 4 ). This can beexplained by the effect of steric effect. The maximumuptakes of Mn, Cd, and Pb was about 6 meq/g, which ishigher than the theoretical capacity of 5.057 meq/g. Theadditional capacity may be explained by assuming theadsorption of hydroxo-aquo species or polynuclear species inthe exchanger at high concentration of these metal ions inhigh pH media [23]. The metal ions having small crystalionic radii show strong tendency to be more highly hydratedthan the larger ions. This causes weak electrostatic forceon the surface of the hydrated shell of the small cations.4. 1. 3. Ion exchange 'isotherms on C-SbA.
Ion exchange isotherms have usedgraphically experimental data pertinent tosieve effect. The ordinate of these plots,
to representsteric and ionX , is the
269
6.0
S.O
4.0
2.0
l 1.0D
0.0
TheoreticalCapacity
I I I I_ O ^ « P O w O C o — 3 C - D J O- i Z x i i x u Z u m a X u z u N u a .
Metal ions
Fig. 4. Maximum uptakes of metal ions on C-SbA
equivalent fraction of the ion Mn+in the exchanger phase andthe abbcissa, Xfjj , is in the solution. Cation exchangeresins of strong acid type have essentially elasticstructure and display swelling and shrinking depending uponthe nature of the exchanging cations with hydration. Theydisplay isotherms which do not deviate greatly from theideal ones. Inorganic ion exchangers generally have rigidstructures and display curve in which KJj varies greatlywith the ion exchanging cations. When small cations areexchanged with large cations, the ingoing cation isinitially preferred, and the degree of exchange gives avalue lower than unity. Thus, the ion exchange isothermshows S shape curve for rigid exchanger, indicating reversalselectivity, and the selectivity coefficient decreaseprogressively with increasing concentration of metal ions inexchanger.
Figs. 5-a ) and 6-a ) showed the typical ion exchangeisotherms for alkali metal, alkaline earth metal , and Pb/H+ systems. The selectivity reversal was observed forthe metal ions/ H+ ion exchange reactions listed, except forLi+ and Mg2V H+ reactions [25, 26].
The plots of In KH vs XM , which are referred toKielland plots, are shown in Figs. 5-b and 6-b. Theselectivity coefficients decrease with increasing theconcentration of metal ions in the exchanger. A fairlygood linearlity was observed for the Kielland plots of theall ion exchange system listed, except of Ba^+/H+ and Pb Htsystems. It has been known that the absolute values ofthe Kielland coefficient, C, indicate degree of the stericeffect in the exchangers. The C values for differentexchange systems are plotted as a function of the effectivecrystal ionic radius in Fig. 7, including the results forour earlier results. These plots show that cations with aneffective crystal ionic radius of 1.0-1.2 "A enter theexchanger with minimum steric hindrance. The C-SbA has arigid structure without swelling or shrinking; the change inthe lattice constant is less than 2% for the conversion
270
-6-e
0.1 0.2 0.3 0.4 OS
Fig. 5. Ion exchange isotherms (a ) and plots of In KO vs TM(b) for the system of alkali metal ions/H* on C-SbA.Open mark; 30°C, closed mark; 45°C.
Equivalent fraction of M in solution
-5 -
Equivalent fraction of M in exchanger
M -Fig. 6. Ion exchange isotherms (a ) and plots of In K^ vs X^(b) for the systems of alkaline earth metal ions-and Pb^/H* on C-SbA at 30°C
Open mark; forward reaction,closed mark; reverse reaction
271
ouTJ| 104l
§I 5
Go3' Ça" .pbi-.In3' Ce3'
• .Mn2- ..Cd2'1.0 1.5
E.I.R. (A)
Fig. 7. Plots of absolute value of Kieland coefficient vseffective crystal ionic radius of various metals.
the.+
from the hydrogen ion form to metal ion forms. The smallcations, Lit Mg, and Ni?+ are hydrated strongly in theaqueous solution and may be exchanged in keeping with theirhydration shell or in leaving a few hydrated water fromhydration shell. This brings large steric effects inrigid structure of the C-SbA. The exchange reaction of Hby the large cations with less hydrated becomesprogressively difficult in the rigid structure. Thus, theminimum steric effect of the C-SbA may be interpreted in theboth contributions.4. 2. Ion exchange selectivity of metal antimonates.
Quadrivalent metal antimonates as ion exchangers havebeen studied first by Abe and Ito [14]. Especially,tin(IV) and Titanium antimonates( SnSbA and TiSbA,respectively ) behave as the cation exchangers withrelatively high exchange capacity, and show the excellentselecivities for alkali metal ions in the order: Na<K <Rb<Cs < Li on SnSbA and Na< K<Rb< Li < Cs on TiSbA, whichare not included in 11 Eisenman's selectivity seriespredicted [27].4. 2. I. Selectivity of SnSbA for microamount of metal ions.
Equilibrium distribution coefficients of alkali,alkaline earth, and transition metal ions on SnSbA with amole ratio( Sb/Sn ) of 1.86 were plotted in Fig. 8, valueson Bio-Rad AG 50W-X8 being included for comparison. The Kdvalues for alkali and alkaline earth metal ions on SnSbA,except Li, increase in the order; Na< K <Rb < Cs and Mg<Ca < Sr < Ca, as shown with sulfonate-type organic resins.The separation factors are more favorable than thoseobtained on Bio-Rad AG 50W-X8 [28]. The increasedselectivity, except Na and Mg, was observed to be higherthan that on Bio-Rad AG 50W-X8. Much high selectivity andKd values for transition metal ions were also observed withcomparison of organic ion exchange resins. The selectivesequence of transition metal showed Ni < Mn < Cd < Ccx Cu <Zn on the SnSbA [ 28], and Mn < Zn < Cu < Ni < Co < Cd <on Bio-Rad AG 50W-X8.
272
30
1.0
SnSbA
,0-—°—9--0-""0 Bio Rodor' AG50W-X8
l l -l l l l l l t l l l l l lZ <j <X a 2 u z u i5 3
Ketal ions
Fig. 8. The Kd values of various metal ions on SnSbA inl M HNCL
It has been known that the ion exchange propertiesincluding their selectivities are much dependendt on theirpreparetive conditions and compositions. The Kd value ofLi+ shows a maximum at about 1.7 of the Sb/Sn ratio, whileincrease with increasing Sb/Sn ratio for other alkali metalions [15]. The selectivity sequences of alkali metal ionsmaintain in a range from 0.5 to 2.5 of Sb/Sn ratio. Anextremely high selectivity was observed for Li , which was agood advantage for separation of Li from other alkalimetal ions like seawater [29].4. 2. 2. Selectivity of TiSbA for microamount of metal ions.
On the TiSbA, slightly different selectivities with theSnSbA was observed; Na<K<Rb< Li < Cs for alkali metal ions[16], and Ni < Mn < Co < Cd < Zn < Cu for transitionmetal ions [ 30]. An usual selectivity sequence wasobtained for alkaline earth metal ions. The Kd values onthe TiSbA were smaller than those on the Bio-Rad AG 50W-X8.The Kd value increases with increasing mole ratio ofSb/Ti. The low Kd values for these metal ions, comparedwith SnSbA and Biö-Rad AG 50W-X8, are due to lower value( C.64} on the Sb/Ti ratio (see Fi£> 9).4. 2. 3. Ion exchange isotherms and Selectivitycoefficients for alkali metal ions/ H+ system.
The ion exchange isotherms observed at 30°C are shownin Fig,10. Each plot of the results of the reversereaction can be plotted on tne isotherm curves of theforward reaction. Thus, reversible exchange reactions areestablished and no hysteresis effect is observed. Theisotherms observed showed the S shape curves, indicating theselectivity reversal [31].
The Kielland_plots showed inflection points at about0.04 and 0.14 of XM for Li+/H+ and other metal ions / H+systems. An extremely high selectivity was observed for Liat the concentration lower than X = 0.04. This indicatesthat preference sites for Li are present at lower exchangecomposition. Other alkali metal ions can not approach tothe sites because of large crystal ionic radius. The
273
2.0
1.0
-1.0
--o—o-'"0 Bio RodAG50W-X8
TiSbA
A MS S S lMetal ions
Fig. 9. The Kd values of various metal ions on TiSbA inl H HNO„
0.3
0.2
Li
Fig. 10. Ion exchange isotherms (a) and plots of In K^ vs(b) for the systems of alkali metal ions on SnSbAat 30°C.
maximum uptake was found to be 1.13, 0.98, 0,77, and 0.73meq/g for Li, Na, K, Rb, and Cs, respectively. Thisbehavior can be explained in the terms of the selectivityreversal at these infection points.
REFERENCES[ l] AT1PHLETT, C. B., Inorganic Ion Exchangers, Elsevier,
Amsterdam, 1964.[2] FULLER, M.J., Chromatogr. Rev., 14 (1971) 45.[3] VESELY, V., Pekarek, V., Talanta, 19 (1972) 219,1245.[4] ABE,IL, Bunseki Kagaku, 23 (1974) 1254, 1561.[ 5] ALBERTI, G., Ace. Chem. Res.,11 (1978) 163.[6] De, A. K., Sen, A. K. , Sep. Sei. Technol., 13 .(1978)
517.[7] CLEARFIELD, A., Inorganic Ion-exchange Materials, CRC
Press, Inc., Boca Raton Fl., (1982).
274
[8] ITO, T., ABE, M., Nippon Kagaku Zasshi, 87 (1966)1174.
[9] ASE, II., ITO,T., Bull. Chem. Soc. Jpn., 40 (1967)1013.[10] ABE, II., T. ITO, Bul. Chem. Soc. Jpn., 41 (1968) 333.[il] ABE, M., Bull. Chem. Soc. Jpn., 42 (1969) 2623.Î12] GIRAPJJI, F., SABBIONI, E., J. Radioanal. Chem., 1
(1968) 169.[13] 3AETSLE, S., HUYS, D., J. Inorg. Nucl. Chem., 30
(1968) 242.[14] ABE, M., ITO, T., Kogyo Kagaku Zasshi, 70 (1967) 440.[15] A3E, M., HAYASHI, K., Solv. Extract. Ion Exch.,
1 (1983) 97.[16] ABE, M., TSUJI, II., Chem. Lett. 1983, 1561.Î17Î KRAUS, K..A., RARIDON, R. J., J. Phys. Chem., 63(1959)
1901.[18] GAINS Jr.,G. L., THOIÎAS, H.C., J. Chem. Phys. ,41 (1953)
714.[19] BARRER, R. M., FALCONER, J. D., Proc. R. Soc., A236
(1956) 227.[20] BARRER, R..M.,et al., Proc. R. Soc., A273 (1963) 180.[21] STRELOW, F. W. E., et al., Anal. Chem., 37 (1965) 106.[22] ABE, II., UNO, K., Sepn. Sei. ,Technol. 14, (1979) 355.[23] ABE, M., Sepn. Sei. Tech., 15 (1980) 23.[ 24] SHANNON, D. D., PREWITT, C. T., Acta Crystallogr. B23
(1969) 925.[25] ABE, M., J. Inorg. Nucl. Chem., 41 (1979) 85.[26] ABE, M., SUDOH, K., J. Inorg. Nucl. Chem., 43 (1981)
2537.[27] EISENI1AN, G., Biophys. J., 2, 259 (1962).[28] ABE, M., FURUKI, N., The 44th Annual fleeting of Jpn.
Chem. Soc., paper 2F19, Okayama (1981)[29] AEr, II., HAYASHI, K., Hydrometallgy, 12 (1984) 83.[30] ABE, Tl., CHITRAKAR, R., Unpublished Result.[31] ABE, Tl., FURUKI, N., The 47th Annual Meeting of Jpn.
Chem. Soc., paper 3T15 Kyoto (1983)
275
PANEL DISCUSSION
Chairman: H.J. Ache (Federal Republic of Germany)
SUMMARY AND RECOMMENDATIONS
I. Summary of the discussion
Inorganic ion exchangers and adsorbents have proved their potentialusefulness in various areas of nuclear fuel cycle technology such as:
1) In the separation of fission products and actinides frommedium- and high-activity wastes.
2) In the treatment of effluents from nuclear power plants.
3) As chemical barriers (backfill material) in undergroundrepositories of radioactive wastes.
Improvements can be made in some techniques that are already beingused by the industry, such as:
1) The adsorption of iodine from dissolver off-gases, and
2) The fixation of krypton-85 and 14C02.
Further research and development should be directed towards theoptimization of the chemical and physical properties of these inorganiccompounds for their application in the fields listed above. Considerablefundamental research is still necessary to have a better understanding ofthe mechanisms involved in adsorption and ion-exchange and of thecorrelation between structure and efficiency.
II. Recommendations
Regarding future activities, the participants recommend thatinternational cooperation should be encouraged on the main topic ofresearch on inorganic ion-exchangers and adsorbents for the separation offission products and actinides from medium-and high-activity wastes and
277
for the treatment of effluents front nuclear power plants, including thefollowing sub-topics:
a) Preparation;b) Characterization;c) Applications;d) Process and engineering aspects, ande) Economics.
Special efforts should be made to promote a greater exchange ofinformation and closer cooperation between the specialists who areworking in basic research and those who are primarily concerned with thepractical applications of these compounds in nuclear fuel cycletechnology.
278
LIST OF PARTICIPANTS
BELGIUMBaetsle, L.
Centner, B.
Jacqmin, J.
CHINA. PEOPLE'S REPUBLIC OF
Weng Haorain
Zhu Yong-jun
EGYPTMisak, N.Z.
FINLANDLehto, J.
FRANCEDozol, J.F.
Madie, C.
Chef du Department ChimieCEN/SCKB-2400 Mol, BelgiumIngénieur au bureau d'étuded'Electrobel S.A.Place du trône llOOO Bruxelles, BelgiumBelgonucleaire25, Rue du Champ de MarsB - 1050 Brussels, Belgium
Beijing Normal UniversityDivision of Radiochemistryand Radiation ChemistryBeijing, China, People's Republic ofInstitute of Nuclear Energy TechnologyTsinghua UniversityBeijing, China, People's Republic of
Prof, of Physical ChemistryNuclear Chemistry DepartmentAtomic Energy Post 13759101 Kasr El Einy StreetCairo, Eygpt
University of HelsinkiDepartment of RadiochemistryUnioninkatu 35Helsinki, Finland
CEA/Centre d'Etudes Nucléaires deCadarache
Services d'Effluents et Déchets deFaible et Moyenne ActivitéF-13115 St. Paul Les Durance, FranceCEA, Section des TransuraniensB.P. No. 6F-92260 Fontenay-auz-RosesFrance
279
Regnaud, F.
GERMANY. FEDERAL REPUBLIC OF
Ache, H.J.
Merz, E.
Sameh.A.A.
HUNGARY
Szirtes, L.
INDIAlyer, R.M.
ITALYAlberti, G.
JAPANAbe, M.
Ito, T.
CEA/DCAEA/Service Etudes AnalytiquesB.P. No. 6F-92260 Fontenay-aux-RosesFrance
Nuclear Research Center KarlsruheInstitute of RadiochemistryPostfach 3640D ~ 7500 KarlsruheGermany, Federal Republic ofInstitute of Chemical TechnologyKernforschungsanlage Julich GmbHD-5170 Julich, POB 1913Germany, Federal Republic ofNuclear Research Center KarlsruheInstitute of RadiochemistryPostfach 3640D - 7500 Karlsruhe.Germany, Federal Republic of
Institute of Isotopes of theHungarian Academy of SciencesP.O.Box 77H-1S25 BudapestHungary
Head of Chemistry DivisionBhabha Atomic Research CoentreTrombayBombay 400 085, India
Dipartimento di ChimicaUniversita di PerugiaVia Elce di Sotto 101-06100 Perugia, Italy
Department of ChemistryTokyo Institute of Technology2-12-1 OokayamaHeguro-kuTokyo, JapanDeputy DirectorDepartment of RadioisotopeProductionRadioisotope Center, JAERITokai-mura, Ibaraki-ken, Japan
280
Kanno, T.
Saturai, T.
PARISTAU
Chaudhri, S.A.
SWEDENAirola, C.
UNITED KINGDOM
Harding, K.
Hooper, E.W.
Eccles, H.
Research Institute of Mineral Dressingand Metallurgy
Tohoku University 11 Katahira-ChomeSendai, JapanGeneral Manager, PhysicalChemistry Lab.Department of ChemistryTokai Research Establishment JAERITokai-mura, Ibaraki-kenJapan
Alternate to the Resident RepresentativePermanent Mission of Pakistanto the IAEA
HofzeileA-1190 Vienna
Nuclear Division, Waste SystemsStudsvik Enefgiteknik ABS - 61182 NykoepingSweden
UKAEA ABE WinfrithDorchesterDorset DT2 8DHU. K.
Separation Processes GroupChemical Technology DivisionAERE HarwellOzon 0X11 ORAU. K.
Building 643, BNFLSpringfield WorksSalwickPreston, Lanes., PR4 OXJU. K.
UNITED STATES OF AMERICA
Carl, D. Senior Engineer/Solidification EngineerWest ValleyNuclear Services Co. Inc.P.O. Box 191West Valley, New York14171-0191, U.S.A.
281
Clearfield, A. Department of ChemistryTexas A&M UniversityCollege Station, TX 77843U.S.A.
Collins, E.D. ManagerThree Mile IslandAssistance ProgramsOak Ridge National LaboratoryOak Ridge, TN 37831U.S.A.
YUGOSLAVIA
Ruvarac, A. The Boris Kidric Instituteof Nuclear Sciences11001 BelgradePo 0. Box 522Yugoslavia
ORGANIZATIONS
INTERNATIONAL ATOMIC ENERGY AGENCY (IAEA)
Ajuria, S. Division ofNuclear Fuel Cycle
Rybalchenko, I. Division ofNuclear Fuel Cycle
Ugajia, M. (Scientific Secretary) Division ofNuclear Fuel Cycle
282