resurrection of the iron and phosphorus resource in steel-making slag
TRANSCRIPT
Abstract This research focused on the treatment of steel-making slags to recycle and recover iron and phosphorus.The carbothermal reduction behavior of both synthesizedand factory steel-making slag in microwave irradiation wasinvestigated. The slags were mixed with graphite powderand heated to a temperature higher than 1873K to precip-itate a lump of Fe–C alloy with a diameter of 2–8mm. Thelarger the carbon equivalent (Ceq, defined in the text), thehigher the fractional reduction of iron and phosphorus.An increase in the SiO2 content of slag led to a consider-able improvement in the reduction for both iron and phos-phorus because of the improvement in the fluidity of theslags and an increase in the activity coefficient of P2O5 inthe slags. The extraction behavior of phosphorus fromFe–P–Csatd alloy was also investigated at 1473K by carbon-ate flux treatment. For all the experiments with a process-ing time longer than 10min, the phosphorus in the fluxescould be concentrated to more than 9% (w/w) showing thatit could be used as a phosphorus resource. Compared withK2CO3 flux treatment, that using Na2CO3 was more effec-tive for the extraction of phosphorus, and this was attrib-uted to the lower evaporation of Na2CO3. Finally, a recyclingscheme for steel-making slag is proposed.
Key words Steel-making slag · Recovery · Iron · Phospho-rus · Microwave processing
Introduction
The recycling of used materials and the development ofenvironment-friendly processes are of great significance for
environmental protection. It is a critically urgent matter to develop methods to treat iron- and steel-making slags to realize their intrinsic value. In Japan, 10 million tons ofsteel-making slag is generated every year, and most is castaside without recovering the iron resources it contains.Although some slag is recycled by being used in iron-making processes, the amount is limited because of its highconcentration of phosphorus. However, the concentrationof phophorus in such slags is not high enough for use as aphosphorus resource, and it is considered to be a harmfulby-product of steel-making in spite of the fact that Japanimports a significant amount of phosphorus as fertilizer.Moreover, because of the use of improved slag furnaces andpowder injection of the modified slags, the amount of slaggenerated in the steel-making process will be lower, thusresulting in a higher concentration of phosphorus in theslags. Therefore, it is increasingly important to recover thephosphorus from such slags, and a new system for treatingsteel-making slag is required.
Much research into treatments for steel-making slag hasbeen conducted. Shiomi et al.1 reported that almost all ofthe iron in the slags was recovered, while Takeuchi et al.2
found that 60% of phosphorus was removed from the slagto the gas phase as P2 using Fe–Si alloys as the reducingagent. In addition, phosphorus in steel-making slags can beused as a fertilizer resource, although this has only beendone to a small extent.
On the other hand, since microwave processing is asimple and efficient way of heating materials with highdielectric losses, it has been widely used in many industrialprocesses such as the drying, sintering, and melting ofceramics or composites.3 When substances are exposed toelectromagnetic radiation in the microwave range, theirtemperature increases in a very short time. Various non-combustible waste products can be easily and safely meltedand solidified using microwave heating.4,5 Iron oxide isknown as a highly microwave-susceptible material,3–5 andsteel-making slags were also found to be effectively heatedby microwaves if the Fe3�/(Fe2� � Fe3�) ratio is within anappropriate range, as described in our previous work.6 Inaddition, small particles of conductive materials such as
J Mater Cycles Waste Manag (2002) 4:93–101 © Springer-Verlag 2002
Kazuki Morita · Muxing Guo · Norio Oka · Nobuo Sano
Resurrection of the iron and phosphorus resource in steel-making slag
ORIGINAL ARTICLE
K. Morita (*) · M. Guo · N. OkaDepartment of Materials Engineering, The University of Tokyo, 7-3-1Hongo, Bunkyo-ku, Tokyo 113-8656, JapanTel. �81-3-5841-7106; Fax �81-3-5841-7104e-mail: [email protected]
N. SanoNippon Steel Corporation, Futtsu, Chiba, Japan
Received: March 16, 2001 / Accepted: November 12, 2001
metals and graphite work as very efficient microwave cou-plers, leading to a rapid heating of materials.6–8
In this study, the carbothermic reduction behavior of syn-thesized CaO–SiO2–FetO slags under microwave irradiationwas studied first. Then, that of practical steel-making slagswith a similar composition was investigated. In addition, weexamined the extraction of phosphorus, in the form of phos-phate, by carbonate flux treatment of the Fe–P–Csatd alloysobtained. A new flow system for treating steel-making slagto recover resources was proposed using our experimentalresults.
Experimental
Recovery of metals from slags by microwavecarbothermic reduction
Synthesized and practical steel-making slags were pro-cessed in a commercical microwave oven (RE-6200, 2.45GHz, 1.6kW, Sharp, Osaka). The composition of the synthesized slag was adjusted to 45% (w/w) CaO, 35%(w/w) SiO2, and 20% (w/w) Fe2O3. When phosphorous wasstudied, 5% (w/w) Ca3 (PO4)2 replaced some CaO. Moredetails on the preparation of the synthesized slag were givenin a previous report.6 The compositions of the factory slagsused in these experiments are shown in Table 1. The totaliron content (metallic and iron oxide forms) was about 12%(w/w). The slag contained 4% (w/w) P2O5, and the basicityratio was about 1.5. Fifteen grams of the slag was chargedin each run. Graphite powder was employed as a reductant,and was mechanically mixed with the slag. The amount ofgraphite powder was varied according to the index ofcarbon equivalents, which was defined as the ratio of themixed carbon to the amount theoretically required for thereduction of Fe2O3 in the slag to pure metallic Fe. A quartzcrucible (inner diameter 38mm, height 45mm) was used,and was surrounded by a 50-mm layer of insulating bricks.The crucible and the insulation were considered to be trans-parent to microwaves since they were not heated them-selves in the microwave irradiation. The samples wereirradiated at full power for 2–9min. After the microwavetreatment, the sample was cooled in the furnace, separatedinto slag and metallic phases, and then subjected to chemi-cal analyses.
Recovery of phosphorus from the Fe–P–Csatd alloy bycarbonate flux treatment
Preparation of Fe–P–Csatd alloy
A mixture of high-purity electrolytic iron (99.99%) andreagent-grade graphite powder, in a graphite crucible, wasmelted for 4h at 1823K under a deoxidized argon gasatmosphere (200ml/min). The required amount of analytical-grade Ca3(PO4)2 was then added to adjust thephosphorus content. After the melt had been kept for 18h
at 1823K, it was quenched in water to give Fe–[2.6%–11.3%(w/w)]P–Csatd.
Phosphorus extraction process
Fifteen grams of the Fe–P–Csatd alloy was melted in a MgOcrucible using a SiC resistance furnace in a deoxidizedargon atmosphere (200ml/min) at 1473K. The temperaturewas measured by a thermocouple in the bottom of the cru-cible, and the temperature was controlled to within 3K bya proportional integral derivative (PID) controller. Afterthe sample had melted completely and was at the desiredtemperature, about 8g granular K2CO3 or Na2CO3 wereadded to the molten alloy through a quartz tube. The meltwas held for 2–30min at the experimental temperature, andthen quickly withdrawn from the furnace, and quenched influshing argon gas. To study the extraction ratio of phos-phorus from the alloy, the residual flux was separated fromthe metals. The phosphorus content in the metals and theresidual flux was analyzed, and the weight of the residualflux was also measured.
Chemical analysis
The phosphorus content of the metals and the residual fluxwas determined by ammonium molybdate absorptiometry.The calcium, aluminum, magnesium, and manganese con-tents were analyzed by radio frequency inductively coupledplasma (ICP) spectroscopy, the total iron content of theslags by volumetric titration, and the silicon content by agravimetric method. The carbon content in both metals and slags was determined by oxygen gas fusion–infraredabsorptiometry.
Results and discussion
The experimental results of the microwave treatment ofsteel-making slags and those of phosphorus extraction fromthe Fe–P–Csatd alloy, together with the experimental condi-tions, are summarized in Tables 2 and 3, respectively. Severaleffects were investigated, as shown in the following sections.
Reduction behavior of slags with graphite powder
Reduction behavior of synthesized slags
Before the experiments with factory steel-making slag, syn-thesized slags were treated by microwave carbothermic
94
Table 1. Chemical composition of the steel-making slag employed inthe experiments (% w/w)
T.Fe CaO SiO2 MnO MgO P2O5 Al2O3
11.7 41.9 26.6 5.80 5.20 4.00 3.20
reduction to evaluate the feasibility of this process. Asshown in Fig. 1, a lump of Fe–C alloy with a diameter of 2–8mm was formed during heating, and could easily be sepa-rated from the slag phase after the experiment. Arcing phenomena during the microwave irradiation were oftenobserved when the temperature was higher than 1470K,which was probably due to the metallic phase formation inthe sample. Figure 2 shows the changes in the iron andcarbon content of the slag with time when the carbon equiv-alent (Ceq) is 2.0. The iron content of the slag dropped from15% to 2% (w/w) in 4min, which was the same as the behav-ior of the carbon content. Eventually, the iron content of the
slag was as low as 1% (w/w) and the recovery ratio reachedmore than 90%. The influence of Ceq on the reductionbehavior of the synthesized slag is illustrated in Fig. 3. Therecovery ratio of iron increased and the iron content of theslag decreased with increasing Ceq, and they became almostconstant when Ceq exceeded 1.5. The carbon content of themetals increased to 5.0% (w/w), which is almost carbon saturation, and became constant when Ceq exceeded 1.5,whereas the residual carbon in slags was observed toincrease with Ceq when it was higher than 1.5. The carboncontent of the slag, which is the amount of carbon physi-cally suspended in the slag, increased when Ceq exceeded 1.5
95
Table 2. Composition of slags and metals after microwave processing, recovery ratio of iron and phosphorus, weight of reduced metals, andexperimental conditions
No Heating Carbon Content (% w/w) in slag Content (% w/w) Wmetal Recovery Recoverytime equivalent
T.Fe P2O5 SiO2 CaO MnO MgO Al2O3
in metal (g) ratio of ratio of(min)
C C P Siiron phosphorus
1-1 2.0 1.5 9.91 3.28 25.9 42.6 5.12 5.58 3.12 6.60 3.76 2.65 1.49 0.18 0.14 0.031-2 4.0 1.5 5.33 2.36 26.9 41.5 5.19 5.92 4.02 4.20 4.02 7.58 1.02 0.59 0.44 0.261-3 5.5 1.5 5.61 2.04 24.7 43.0 5.09 6.01 3.96 3.11 3.32 8.06 1.95 0.68 0.51 0.311-4 7.0 1.5 4.25 1.91 28.9 42.2 5.09 5.33 3.78 2.31 3.65 8.34 1.89 0.79 0.58 0.371-5 9.0 1.5 3.94 1.72 28.0 43.9 5.28 4.98 3.97 1.88 3.57 8.56 1.78 0.78 0.58 0.38
2-1 7.0 0.5 7.81 2.48 26.2 41.4 5.07 5.33 3.24 1.16 2.32 8.51 1.57 0.41 0.29 0.192-2 7.0 1.0 6.95 2.28 25.5 42.4 5.35 4.99 3.88 1.34 3.86 6.57 1.87 0.58 0.43 0.212-3 7.0 2.0 3.78 1.55 25.2 46.2 5.28 5.56 3.54 2.31 3.17 9.62 1.57 0.85 0.62 0.462-4 7.0 2.5 3.58 1.16 27.5 46.3 4.97 5.69 4.22 3.97 3.26 9.70 1.98 0.91 0.67 0.51
3-1a 7.0 1.5 3.01 0.38 36.1 44.4 5.33 5.48 4.06 2.89 2.48 13.30 2.83 0.97 0.68 0.74a With extra 10% (w/w) SiO2 addition
Table 3. Experimental conditions, composition of metals and fluxes after flux treatment, weight of residual fluxes, and extraction ratio of phosphorus
No. Processing Species Initial metal Final metal Final flux P in flux Weight of Extractiontime at 1473K of flux content content content
P in metalresidual flux ratio of
(min) (% w/w) (% w/w) (% w/w) (g) phosphorus
Pinit Cinit P C P C
4-1 2.0 K2CO3 6.50 3.87 4.47 3.63 4.09 5.49 0.92 6.69 27.74-2 5.0 6.61 3.50 2.96 3.68 8.94 2.92 3.02 6.11 54.74-3 10.0 6.37 3.31 4.32 3.22 9.33 2.32 2.16 2.98 30.44-4 10.0 6.66 3.82 3.25 3.30 8.92 3.49 2.74 3.98 35.24-5 20.0 6.50 3.87 3.74 3.87 9.67 2.29 2.59 3.60 35.74-6 30.0 6.59 3.85 3.93 3.67 9.56 2.28 2.43 2.56 23.9
5-1 2.0 Na2CO3 6.66 3.82 3.42 3.55 6.78 5.98 1.98 6.14 41.55-2 5.0 6.66 3.82 3.13 3.56 9.25 4.84 2.95 5.91 52.15-3 10.0 6.66 3.82 2.12 3.95 11.7 3.53 5.53 5.61 65.55-4 20.0 6.66 3.82 1.12 4.16 13.6 2.59 12.1 5.42 73.45-5 30.0 6.57 3.81 1.32 4.19 13.6 1.61 10.3 5.44 72.4
6-1 2.0 Na2CO3 11.3 2.25 8.25 2.03 6.56 5.24 0.79 6.28 24.26-2 5.0 11.3 2.25 7.18 2.33 8.91 4.21 1.24 6.01 31.66-3 10.0 11.3 2.25 6.06 2.40 12.8 2.44 2.12 5.78 43.56-4 20.0 11.3 2.25 4.75 2.70 14.9 1.70 3.14 5.32 46.96-5 30.0 11.3 2.25 4.09 2.80 16.2 1.10 3.95 5.46 52.0
7-1 5.0 K2CO3 3.88 3.58 2.26 3.77 4.08 5.41 1.80 5.18 35.87-2 5.0 2.56 4.39 1.62 4.03 2.59 5.95 1.59 5.26 33.47-3 5.0 3.16 3.89 1.50 3.68 4.03 5.56 2.69 5.69 47.67-4 5.0 5.23 3.92 2.90 3.79 5.59 4.52 1.93 5.84 41.27-5 5.0 6.13 3.92 3.36 3.57 7.11 3.66 2.12 5.58 43.17-6 5.0 4.91 3.82 3.06 3.14 5.03 5.89 1.65 5.03 34.37-7 5.0 6.01 3.82 3.88 2.98 6.09 4.98 1.57 5.14 34.4
because the carbon cannot dissolve into the metal phasewhen the carbon content of the metal reaches saturationvalue. The material balance showed that about 30% ofcarbon, which was not accounted for, was probably con-sumed by oxidation during the microwave irradiation in air.The effects of Ceq on the phosphorus content of metals andslags are shown in Fig. 4. The phosphorus content of themetals was about 5% (w/w) when Ceq was lower than 1.25,and then decreased to 3.5% (w/w), while the phosphoruscontent of slags decreased with a Ceq increase. From theseresults, the reduction process with carbon was shown to befeasible for practical use.
Reduction behavior of factory steel-making slags
Based on the results obtained from the experiments withsynthesized slags, the reduction behavior of factory steel-making slag was investigated under similar conditions. Thechanges in the iron, phosphorus, and carbon content of theslag and the metal, and iron reduction ratio for Ceq � 1.5against processing time are plotted in Fig. 5. With theincrease in processing time, both the iron and the phospho-
96
Fig. 1. Particles of Fe–C–P alloy reduced from slag by microwave car-bothermic reduction (processing time, 7min; carbon equivalent, 1.5)
Fig. 2. Change in iron and carbon contents of synthesized slags withprocessing time
Fig. 3. Effect of carbon equivalent on the reduction behavior of synthesized slags
Fig. 4. Effect of carbon equivalent on the phosphorus content ofmetals and slags
rus content of the slag decreased, and the iron reductionratio increased. This was consistent with the decrease in thecarbon content of the slag, which is shown in the topdiagram in Fig. 5. After the maximum processing time (9min), the iron content of the slag was as low as 4% (w/w)and the reduction ratio reached about 60%. The phospho-rus content in the reduced metals reached as high as 9%(w/w), although the phosphorus content of the slagdecreased by only 50%. This is because of the preferentialreduction of iron over that of phosphorus. Figure 6 showsthe time-dependence of the material balances of iron andphosphorus in metals and slags after microwave processing.Both iron and phosphorus in slags were gradually reducedinto the metallic phase with processing time, and the frac-tional reduction of iron was more significant than that ofphosphorus. The unrecovered amount of both elements wasconsidered to be due to uncollectable fumes and dust gen-erated during the reaction, and probably also due to gasifi-cating dephosphorization.
Figure 7 shows the effect of the carbon equivalent on the final composition of the slags and metals after 7minmicrowave processing. As expected, the extent of reductionwas considerably affected by the amount of reductant. Thereduction ratios of iron and phosphorus increased withincreasing Ceq, and reached about 70% and 50%, respec-tively, when Ceq was 2.5. The iron and phosphorus contentof the slag decreased with increasing Ceq, and did not change
97
Fig. 5. Changes in the iron, carbon, and phosphorus contents in slagand metal, and iron recovery ratio with processing time
Fig. 6. Material balances of iron and phosphorus between metals andslags as a function of microwave processing time
Fig. 7. Effect of carbon equivalent on the reduction behavior of ironand phosphorus from slags
when Ceq exceeded 1.5. This is also compatible with thebehavior of carbon shown in the top diagram in Fig. 7. Thecarbon content in the metal increased to 4.0% (w/w), andthen decreased slightly when Ceq was higher than 1.0,whereas the residual carbon in slag increased with Ceq. Asseen in the middle diagram in Fig. 7, the phosphorus contentof the slag decreased with Ceq, and was down to 25% of theinitial content at Ceq � 2.5. The phosphorus content of themetal increased with increasing Ceq, showing that the reduc-tion of phosphorus was affected by the amount of reduc-tant. The distributions of iron and phosphorus between slagand metal after 7min microwave treatment are plotted as afunction of Ceq in Fig. 8. It is clear that the increase in Ceq
results in increases in the distributions of iron and phos-phorus in the metallic phase, which is consistent with theresults shown in Fig. 7. It can also be seen in Fig. 8 that the loss of phosphorus was significantly higher than that ofiron. As mentioned in the results for synthesized slags, thisis thought to result from gasificating dephosphorizationduring microwave processing.
To study the effect of the SiO2 content of the slags on thereduction behavior, an extra 10% (w/w) SiO2 was added tothe factory steel-making slags. The processing time and theCeq were fixed at 7min and 1.5, respectively. Figure 9 showsdifference in the distributions of iron and phosphorus aftermicrowave processing with the addition of SiO2 and withoutit. The material balance of iron and phosphorus indicatedthat the SiO2 addition led to a considerable improvementin the fractional reduction of both iron and phosphorus.Phosphorus in particular was reduced more effectively, and
its reduction ratio reached as high as 74%, with only about10% of phosphorus remaining in slag. The rest was consid-ered to have been removed in the gas phase. These resultscan be interpreted by the change in the slag compositionafter microwave processing, as shown in Fig. 10. The com-position of the original steel-making slag had a smaller ironoxide region, and also showed a saturated two-phase 2CaO·SiO2 region after processing. This leads to a decrease in thefluidity of the slag, thus causing the decrease in the reduc-tion rate. However, with the extra 10% (w/w) SiO2 addition,the slag remained in a single liquid phase throughout theprocessing time, which helps a smooth mass transfer in thereduction behavior. Moreover, the activity coefficient ofphosphorus oxide (P2O5) in the molten slag is significantlydependent on the slag basicity. A decrease in this basicitybrings about increases in the activity coefficient of P2O5.Thus, the fractional reduction of phosphorus oxide wasimproved more than that of iron oxide when an extra 10%(w/w) SiO2 was added to the steel-making slag, as shown in Fig. 9.
Phosphorus recovery from Fe–P–Csatd alloys by carbonate extraction
In order to assess the possibility of phosphorus recoveryfrom steel-making slags as a new resource, the extraction ofphosphorus from Fe–P–Csatd alloy was performed by addingcarbonate flux at 1473K.The weight of the residual flux andthe phosphorus content of the metal and the flux after the
98
Fig. 8. Material balance of iron and phosphorus between metals andslags as a function of carbon equivalent
Fig. 9. Effect of the SiO2 content in slags on the reduction behavior ofiron and phosphorus from slags
99
Fig. 10. Change in slag compositionafter microwave processing
flux treatment are plotted against the processing time in Fig.11.As is clear in the middle and the bottom diagrams in Fig.11, the phosphorus content of the flux increased and that ofmetal decreased with an increase in processing time, andboth of these changes leveled off at around 10min after theaddition. For all experiments with a processing time longerthan 10min, the phosphorus concentration in the flux wasmore than 9% (w/w), which might be high enough to beused as phosphorus fertilizer. In cases with added Na2CO3,the phosphorus content of the flux was higher, while that ofthe metal was lower compared with those with addedK2CO3. As seen in the upper diagram in Fig. 11, less evapo-ration was observed for Na2CO3, which leads to more effec-tive flux utilization than with K2CO3 in the extraction ofphosphorus. As can be seen in Fig. 12, when the initial phos-phorus content of the metal was 11.3% (w/w), the extrac-tion ratio of phosphorus was less than when the initialcontent was 6.6% (w/w). This is probably because the activ-ity coefficient of P2O5 in the slag increased with an increasein the phosphorus content.
The effect of the phosphorus content in metal on theextraction behavior of phosphorus is shown in Fig. 13. Inthese experiments, the phosphorus content of the metalranged from 2.6% to 6.1% (w/w) and K2CO3 flux was used,and the processing time and temperature were fixed at 5minand 1473K, respectively.The phosphorus content of the fluxincreased linearly with the phosphorus content of the metal,and hence the extraction ratio of phosphorus was almostconstant at about 40%.
Figure 14 shows the relationship between the carboncontent and the phosphorus content in the molten metals
after the flux treatment. The carbon content in the metalwas found to decrease almost linearly with increasing thephosphorus content in the metal. Such a decrease in the sol-ubility of carbon in Fe–C–P alloys with phosphorus contentwas also reported by Sanbongi and Otani9 under equilib-rium condition at 1773K.
Future prospects for the factory process
Under our experimental conditions, the energy transfer efficiency of the microwave treatment was as low as 7%owing to heat loss, and further investigations are requiredbefore larger-scale processing is feasible. Since the temper-ature of slag after a hot metal or steel refining process isvery high, the efficiency should be much higher owing tohigher dielectric loss in the factory process. The possibilityof using microwaves as a partial heat source should also beconsidered.
From this study, the combination of microwave carboth-ermic reduction of steel-making slag and the carbonate fluxtreatment would be one possible way toward environmen-tal protection and resource savings. Finally, the flowsheetshown in Fig. 15 is proposed for the recycling of steel-making slags. After iron and phosphorus have been car-bothermically reduced into the metallic phase, the slags canbe recycled or used with blast-furnace slags, and metals sub-jected to phosphorus extraction as a resource for fertilizerproduction. The Fe–C alloy can then be returned to a hotmetal bath.
100
P in
met
al (
%)
P in
flux
(%
)
Fig. 11. Change in the phosphorus content of metals and fluxes andweight of residual fluxes with processing time
Fig. 12. Change in the ratio of the phosphorus content in fluxes to thatin metals and the extraction ratio of phosphorus with processing timeas a function of the change in phosphorus content in alloys
Fig. 13. Effect of the phosphorus content in metals on the extractionbehavior of phosphorus from metals
Fig. 14. Relationship between the content of carbon and that of phos-phorus in molten metals after flux treatment
Conclusions
The reduction of iron and phosphorus from synthesized and practial steel-making slags has been studied usingmicrowave processing. The extraction of phosphorus fromthe reduced Fe–P–Csatd alloys has also been investigated bycarbonate flux treatment at 1473K.The results obtained aresummarized as follows.
1. The experimental results from the synthesized slagsshow that when they were heated to over 1873K, a lumpof Fe–C alloy with a diameter of 2–8mm was formed.The recovery ratio of iron from slags increased withlarger Ceq, and reached 90% at Ceq � 1.5.
2. In the experiments with factory steel-making slags, therecovery ratio of iron reached about 70% at Ceq � 2.5.The phosphorus in the slags decreased with Ceq, and wasabout 25% of the initial content when Ceq � 2.5.
3. An increase in the SiO2 content of the slag led to a con-siderable improvement in the reduction of both iron andphosphorus due to increases in the fluidity of the slag andin the activity coefficients of P2O5.
4. For all experiments with a processing time longer than10min, phosphorus was concentrated in the fluxes atmore than 9% (w/w), and could thus be a possible phos-phorus resource.
5. The treatment of Na2CO3 flux was more effective for theextraction of phosphorus from Fe–P–Csatd alloys thanthat of K2CO3 owing to the lower evaporation ofNa2CO3.
6. A recycling flow of steel-making slag to recover both ironand phosphorus was proposed.
Acknowledgments This work was performed in the framework of the “Research for the Future” Program of the Japan Society for thePromotion of Science (JSPS – RFTF 96R01901). The authors are also grateful to the NKK Corporation for financial support to part ofthis work.
References
1. Shiomi S, Sano N, Matsushita Y (1977) Removal of phosphorus inBOF slags (in Japanese). Tetsu-to-Hagané 63:1520–1528
2. Takeuchi S, Sano N, Matsushita Y (1980) Separate recovery of ironand phosphorus from BOF slags using Fe–Si alloys (in Japanese).Tetsu-to-Hagané 66:2050–2057
3. Sutton WH (1989) Microwave processing of ceramic materials.Ceram Bull 68:376–386
4. Morita K, Ngyuen VQ, Mackenzie JD, Nakaoka R (1992) Immobi-lization of ash by microwave heating. Proceedings of the 11th Incin-eration Conference, University of California Irvine, Irvine, CA, pp507–512
5. Nakaoka R, Morita K, Mackenzie JD (1992) Immobilization processfor transuranic incinerator ash. Proceedings of the 3rd InternationalConference on Advances in the Fusion and Processing of Glass,American Ceramic Society, Westerville, OH, pp 553–560
6. Morita K, Guo M, Miyazaki Y, Sano N (2001) The heating charac-teristics of CaO–SiO2–FetO system slags under microwave irradia-tion. ISIJ Int 41:716–721
7. Kozuka H, Mackenzie JD (1991) Microwave synthesis of metal car-bides. Ceram Trans 21:387–394
8. Bescher E, Mackenzie JD (1991) Microwave heating of cermets.Ceram Trans 21:557–563
9. Sanbongi K, Otani M (1961) Physical chemistry of liquid iron (inJapanese). Tetsu-to-Hagané 47:841–852
101
Fig. 15. Proposed flow system forsteel-making slag recycling BF, blastfurnace; BOF, basic oxygen furnace;De–Si, desiliconization process;De–P, dephasphorization process