evaluation of thermodynamic data on zirconium and … bound... · evaluation of thermodynamic data...

Philips J. Res. 42, 15-40, 1987 R1l49

EVALUATION OF THERMODYNAMIC DATA ONZIRCONIUM AND HAFNIUM HALIDES AND

OXYHALIDES BY MEANS OFTRANSPORT EXPERIMENTS

by G. DITTMER and U. NIEMANNPhi/ips GmbH Forschungslaboratorium Aachen, D-5100 Aachen, Germany

AbstractA consistent set of thermodynamic data for zirconium and hafnium halides,oxides and oxyhalides was achieved. It was found that formation enthal-pies of gaseous compounds could be derived from solubility measurementstogether with theoretical estimations and a revision of literature data. Freeenergy functions were calculated employing statistical mechanics. Data forliquid and solid compounds were obtained via sublimation and vaporiza-tion data. Chemical equilibria of zirconium and hafnium with halogens arediscussed.Keywords: enthalpy, hafnium compounds, halogens, filament lamps, free

energy, zirconium compounds.

1. Introduetion

Zirconium may be used as a convenient getter material for gaseous impuri-ties such as hydrogen and oxygen. The application of this material in halogen-filled incandescent lamps is restricted due to its reactivity towards the gasfilling. How zirconium influences the gas phase composition, however, has notyet been examined. Investigations on the transport behaviour of zirconium inthe presence of halogens were therefore carried out. In order to obtain a moresystematic approach measurements were also carried out on the chemicallyrelated element hafnium. It is the aim of these investigations to provide aconsistent set of thermochemical data of zirconium and hafnium halides, oxy-halides and oxides.

2. Theoretical

2.1. Chemical equilibrium calculations

On heating a metal wire in a closed volume containing an inert gas andsmall additions of reactive gases, gaseous metal compounds are formed until

Philips Journalof Research Vol.42 No. 1 1987 15

(2)

G. Dittmer and U. Niemann

the diffusion fluxes from and to the filament are in equilibrium. The calcula-tions are based on the assumption that the volume is divided into concentric. equilibrium compartments. The local heterogeneous chemical equilibrium atthe wire thus will correspond to a homogeneous equilibrium at the bulb wall.

The validity of the mass balance constitutes the steady-state condition ofthe equilibrium system 1-6). The calculations were done by means of the freeenthalpy minimization method 7,8). The amount of metal dissolved in thegaseous phase m may be calculated as a function of filament temperatures bysummation of the partial pressures Pik within k volumes

m = M~ ViMPik~

R~ t: 'where M is the molecular weight of the metal, R the gas constant, ViM themolar fraction of the metal in component j, Pik the partial pressure of com-ponent j in volume ~ and 1k the temperature in volume ~.For practical purposes it is convenient to deal with the summed pressures of

the metal (SPM) in the gaseous phase within the total volume JIb deriving fromthe weight loss Lt m of the coil

I R·TbSPM = VjMPj""" dm --,M,Vb

(1)

where Tb is the temperature of the bulb wall equivalent to the effectivetemperature of the gasphase.

2.2. Thermodynamic data

2.2.1. Enthalpies of formation

Enthalpy data for gaseous compounds available from literature were takenas a basis for the estimation of unknown formation enthalpies. Literaturevalues and estimations could then be revised by adjusting calculated transportrates to experimental results if transport properties were governed by the par-ticular compound which is of interest in a certain temperature range. The esti-mation method enables enthalpy data to be produced for all halides and oxy-halides of a system, even if experimental values are unknown.If adequate electronegativity values are used for each element, atomic for-

mation enthalpies of compounds are directly proportional to the energy gaindH(M Xn) which occurs on forming a chemical bond between the centralatom and the ligand. The electronegativity values of the ligands compriseionic and covalent terms, whereas the contribution of covalent energy dH(M)of the central atom has to be added

dH8t(MXn) = kn• n(xx - XM) + I:!.H(M)/2, (3)

16 Philip. Journal of Research Vol.42 No. 1 1987

(4)

Evaluation of thermodynamic data

where !1Hat(MXn) is the atomic formation enthalpy of compound MXn,!1H(M) the covalent energy of the central atom, n the valence number, k; theproportionality factor, XM the electronegativity of the central atom and Xx theelectronegativity of the ligand,If the electronegativity XM or the covalence energy !1H(M) are not known, a

more convenient formula is obtained by considering the differences in atomicformation enthalpies of halides which may be related directly to the differencein electronegativities of the substituted halogens Xx and Xy

Atomic formation enthalpies thus may be calculated from differences in elec-tronegativities of different halides or oxyhalides.

Our experiences show that the constant k; slightly depends on the numberof ligands n in respect to the covalence radius r of the central atom approxi-mated by the following expression

(5)

where kl is the energy conversion factor for monovalent bond and ro is thestandardization of r to A units.In figs 1 and 3 atomic formation enthalpies .in the Zr/02/X2 and in the

Hf/02/X2 system are plotted as a function of electronegativities of halogens3

_tJHot

fMJmo(' )

î 2r-~~~~~--~--~

1..35 f,o F

3.3 3 2.66Cl Br I

2 1.38Zr

'-+- xfeV)Fig. 1. Negative atomic formation enthalpies of gaseous zirconium halides and oxyhalides as afunction of electronegativity of ligands n (symbols represent literature values from JANAF).

Philips Journalof Research Vol. 42 No. 1 1987

18 Phlllps Journalof Research Vol.42 No. 1 1987


and divalent oxygen. The intersections of lines correspond to electronegativi-ties of lower (a) or higher (b) valency states of the central atom.

TABLE I

Electronegativities X of halogens, oxygen, zirconium and hafnium

elementX (eY) X (eV)a b literature

F 4.00 3.90-4.00Cl 3.30 3.00-3.30Br 3.00 2.80-2.96I 2.66 2.50-2.660 3.34 4.35 3.30-3.50Zr 0.80 1.38 1.33-1.60Hf 0.80 1.30 1.30-1.40

Electronegativities X(a) and X(b) are listed in table I for comparison withthe range of data published in the literature. Other parameters used in eqs (3),(4) and (5) are given in table 11.

TABLE 11

Constants kt. covalent radii r and covalent energies I!.H(M)/2 ofzirconium and hafnium

kl rI!.H(M)/2

element (kJ. mol:" ey-1) (Á)(kJ· mol-l)a b

Zr 200 1.45 36 277Hf 195 1.44 40 262

Thermodynamic data for solid and liquid compounds have been partlytaken from literature. In other cases data have been derived from those ofgaseous compounds using measured or estimated enthalpies and entropies ofsublimation or vaporization.

Furthermore atomic enthalpies of formation of solid compounds whichcannot be established by experiment may be estimated by means of an inter-polation procedure where enthalpies are plotted as a function of the numberof ligands (figs 2 and 4). .


TABLE. III

Thermodynamic data of gaseous zirconium halides, oxides and oxyhalides

compound I::iH298 ref. 8298 ref. Cp,298 ref.(k.I- mol:") (J .mor" .K-1) (J .mol:" .K-1)

Zr 0 9 181 9 27 9

ZrF - 680 (9) 244 9 33 9ZrF2 -1336 9,11,12 284 9 49 9ZrFs -1962 9,11,12 306 9 68 9ZrF4 -2609 10,11,12 319 10,11 87 10

ZrCI - 536 9,(11) 254 9 35 9z-ei, -1056* 9,11 292 9 58 9z-ei, -1507* 9,11 339 9 76 9ZrCl4 -1974* 10,11 368 9,11,13 98 10,13

ZrBr - 475 (9) 265 9 36 9ZrBr2 - 935* (9) 317 9 61 9ZrBrs -1320* (9) 372 9 79 9ZrBr4 -1713* 10,11 414 10,13 103 10,13

ZrI - 410 (9,14,15) 275 9 37 9Zrh - 800* (9,14,15) 355 9 57 9Zrls -1098* 14,15,(9) 399 9 81 9ZrI4 -1410* 10,11,14,16 446 10,13 105 10,13

ZrO - 811 9,11 228 9 31 9Zr02 -1405 9 274 9 46 9

ZrOF -1450 240 45ZrOF2 -2090 276 64

ZrOCI -1300 285 48ZrOCh -1800 314 68

ZrOBr -1240 298 50ZrOBr2 -1675 335 70

ZrOI -1165 306 52z-ot, -1530 350 76

* Data accessible from transport experiments.

Phlllps Journol of Research Vol. 42 No. 1 1987 19

3 r-----~r_----~r_----_,------~


t-.dHo1

fMJmo('j

cz»

o O~------~------2~----~3~----~4

Fig.2. Negative atomic formation enthalpies of solid zirconium halides and oxyhalides as a func-tion of the number of ligands n (symbols represent literature values from JANAF).

TABLE IV

Thermodynamic data of liquid zirconium halides and oxides

compound !::..H298 ref. 8298 ref. Cp,298 ref.(kJ· mol-I) (J . mol-I. K-I) (J . mol-I. K-I)

Zr - 595 9 53 9 26 9ZrF2 -1693 9 107 9 66 9ZrF4 -2783 10 158 10 104 10z-ei, -1365 (9) 137 9 73 9ZrCl4 -2035 10 252 10 120 10ZrBr2 -1207 9 143 9 91 9ZrBr4 -1787 283 125z-r, -1055 9 181 9 94 9ZrI4 -1495 314 128Zr02 -2126 9 84 9 56 9



TABLE V

Thermodynamic data of solid zirconium halides, oxides and oxyhalides

compound I:::..H298 ref. 8298 ref. Cp,298 ref.(kJ· mol-I) (J .mol-I. K-1) (J .mol-I. K-1)

Zr -a - 620 9 39 9 25 9Zr -p - 616 9 43 9 25 9

ZrF -1150 53 48ZrF2 -1730 9,17 75 9 66 9ZrFs -2290 9 88 9 84 9ZrF4 -2847 10,11,18 105 10,11 104 10

ZrCl - 986 11,23,(19-21) 68 21 55z-ei, -1392 11,21,22,(9,23,24) 90 9,24,(23) 73 9z-ei, -1740 11,21,22,(9,23,24) 146 9,24,(23) 96 9ZrC14 -2085 10,11 181 10,11 120 10

ZrBr - 915 21,22 75 21 60ZrBr2 -1233 9,21,22,25 116 9,21 87 9ZrBrs -1527 21,22,(9,25) 179 9,21 99 9ZrBr4 -1829 10,11 225 10 125 10

ZrI - 840 21,26 93 21 70ZrI2 -1080 9,21,27,28,(26) 150 9,21 94 9z-t, -1300 21,(9,26) 211 9,21 105 9ZrI4 -1537 10,11,26 260 10,16 128 10

ZrO -1600 50 50Zr02 -a -2219 9 51 9 56 9Zr02 -p -2213 9 55 9 56 9

ZrOF -2100 85 68ZrOF2 -2550 110 85

ZrOCl -1880 96 75ZrOCb -2130 (29) 120 90

ZrOBr -1800 114 80ZrOBr2 -2000 135 105

ZrOI -1730 127 90z-ot, -1860 140 105

Phllips Journalof Research Vol. 42 No. 1 1987 21



TABLE VI

Thermodynamic datà of gaseous hafnium halides, oxides and oxyhalides

compound !:!.H29B ref. 829B ref. Cp,29B ref.(kJ· mol-I) (J .mol-I. K-I) (J . mol-I. K-I)

Hf 0 30 187 30 21 30HfF - 663 254 33HfF2 -1340 295 49HfFa -1952 316 69HfF4 -2628 11 330 88HfCl - 527 263 35arci, -1000* 316 54nrci, -1488* 344 76HfCl4 -2000* 11,31 375 13,31 100 13,31HfBr - 468 274 36HfBr2 - 890* 343 57HfBra -1300* 381 80HfBr4 -1735* 423 13 104 13Hfl - 402 283 37nn, - 800* 360 58Hfla -1110* 406 82an, -1455* 455 13 106 13HfO - 790 41 236 40 31 40Hf02 -1430 276 31HfOF -1490 280 50HfOF2 -2120 308 68HfOCI -1340 292 56HfOCh -1830 324 74HfOBr -1275 300 58HfOBr2 -1700 340 78HfOI -1200 305 61HfOh -1565 352 81

• Data accessible from transport experiments.

Evaluation of thermadynamie data

3

4,35 4o F

3.3 3 2.66Cl 8r I

2 1.30

Hf

--- x {eVj

_.dHot

{MJmo(' j

t 2 I------P'~

Fig. 3. Negative atomic formation enthalpies of gaseous hafnium halides and oxyhalides as afunction of electronegativity of Iigands n (symbols represent literature values from JANAF).

TABLE VII

Thermodynamic data of liquid hafnium halides and oxides

compound I::..H298 ref. 8298 ref. Cp,298 ref.(kJ· mol-I) (J . mol-I. K-I) (J . mol-I. K-I)

Hf - 612 32 56 32 26 32

HfF2 -1715 130 66HfF4 -2818 179 106

arci, -1315 156 75HfCl4 -2065 249 122

HfBr2 -1188 183 93HfBr4 -1808 298 126

Hfh -1044 200 97nn, -1540 324 132

Hf02 -2139 31 118 31 60 31

Phlllps Journalof Research Vol. 42 No. I 1981 2.3

24 Phllips Journal of Rese.r~h Vol.42 No. 1 1987


TABLE VIII

Thermodynamic data of solid hafnium halides, oxides and oxyhalides

compound !:J.H29B ref. 829B ref. Cp,29B ref.(k.I- mol-I) (J .mol-I. K-I) (J .mol-I. K-I)

Hf -a - 642 30,32 43 30,32 26 30,32Hf -IJ - 636 30,32 47 30,32 26 30,32HfF -1170 63 48HfF2 -1760 85 66HfFs -2328 98 84HfF4 -2888 32,33 136 32 92 32HfCl -1015 21,(34) 78 21,(34) 55arci, -1400 21,25,31,34,36, 110 21,31,34, 77 31

(35) (35)arci, -1770 21,25,31,34,36 151 21,31,34 100 31HfCl4 -2106 25,31,37,38 191 31 121 31HfBr - 930 21 85 21 60HfBr2 -1260 25,(21) 126 21 87HfBrs -1557 21,25 182 21 99HfBr4 -1866 (25) 215 125Hfl - 860 21 103 21 70an, -1106 21 160 (21) 94Hfls -1344 21 215 (21) 105an, -1569 39 285 39 128HfO -1600 60 50Hf02 -a -2254 31 59 31 60 31Hf02 -IJ -2244 31 65 31 60 31HfOF -2075 95 68HfOF2 -2520 120 85HfOCI -1875 106 75HfOCh -2130 130 90HfOBr -1810 124 80HfOBr2 -2000 145 105HfOl -1732 137 90HfOl2 -1860 150 105


TABLE IX

Sublimation enthalpies (and vaporization enthalpies) of zirconium andhafnium halides and oxides

tl.Hs,298 (kJ· mol-I)

compound 0 F Cl Br I

ZrXI 789 470 450 440 430ZrX2 814 (721) 394 (357) 336 (309) 298 (272) 280 (255)ZrXa 328 233 207 202ZrX4 238 111 (61) 116 (74) 127 (85)

HfXI 810 507 488 462 458HfX2 824 (709) 420 (375) 400 (315) 370 (298) 306 (244)HfXa 376 282 257 234HfX4 260 (190) 106 (65) 131 (80) 114 (85)

TABLE X

Sublimation entropies (and vaporization entropies) of zirconium andhafnium halides and oxidestl.Ss,298 (J . mol-I. K-l)

compound 0 F Cl Br I

ZrXI 178 190 186 190 183ZrX2 , 223 (190) 208 (176) 182 (156) 200 (174) 205 (174)ZrXa 217 193 193 188ZrX4 200 (161) 186 (115) 190 (132) 186 (132)

HfXI 176 190 185 190 180HfX2 217 (158) 190 (165) 206 (160) 217 (160) 200 (160)HfXa 218 193 199 192HfX4 193 (151) 184 (126) 208 (125) 170 (130)

Thermochemical data on zirconium and hafnium halides, oxides and oxy-halides used for the calculations are presented in tables III to X. Where refer-ences are given in brackets, the corresponding data exceed our error limits of± 30 kJ· mol-l for enthalpies and ± 10 J . mol-I. K-l both for entropies and'for specific heats.

Phlllps Journalof Research Vol. 42 No. 1 1987 25


3 ,-------,-------,-------,------,.HfFn

(HfO)

with

2 K;eWe = ---MR

M1·M2MR=----Ml+M2

(6)

(Hf)

o ~-----~----~-----~----~o 2 3 4_n

Fig.4. Negative atomic formation enthalpies of solid hafnium halides and oxyhalides as a func-tion of the number of ligands n (symbols represent literature values from JANAF).

2.2.2. Free energy functions

Free energy functions for zirconium and hafnium halides were calculatedvia the partition function. Constants necessary for zirconium halides weretaken from refs 9 and 10. Due to the lack of experimental values for electroniclevels, vibrational frequencies or interatomie force constants of hafnium hali-des, they have been estimated in analogy to respective zirconium compounds.

M X COMPOUNDS

Vibrational frequencies We (in wave numbers) for diatomic hafnium halideswere calculated from the force constant K; and the reduced mass MR' seeref. 40, eh. III,2

41t2 c2e= -- = 5.8883 . 10-2 cm" S-2 • mol,NA

where c is the velocity of light and NA the Avogrado number. The forceconstants of hafnium and zirconium halides were assumed to be equal. Intable XI constants used to calculate translational, electronic, rotational and

and

vibrational parts of the partition function are given.

26 Phlllps Journalof Research Vol.42 No. 1 .1987

and (9)


TABLE XIConstants for the calculation of the partition functions of HfX compoundsconstant HfF Hf Cl Hffir HfIle (A) 2.16 2.43 2.5 2.77Kc (l06 dyn- cm-I) 3.650 2.866 2.518 1.552We (cm-I) 600 407 278 188ei (cm-I) assumed to be equal to those of ZrX compounds

where le is the HfX bond length and ei the electronic level energy.

MX2 COMPOUNDS

Hafnium dihalides were assumed to be nonlinear symmetric according tothe structure of zirconium dihalides. Vibrational frequencies were calculatedvia the following equations, see ref. 41, ch. II,4,

2 Ko ( 2Mx )CWs = - 1 + --sin2(/J ,Mx MM

(7)

2 2 s; s, ( 2Mx )C(WI+W2) = 2-+ -- 1+ --cos2(/J ,

Mx Mx MM

2 2 s..«; ( 2Mx)CWI W2 = 2· 2 1 + _- cos" (/J,Mx MM

where Mx is the mass of the ligand, MM the mass of the metal atom, K; theforce constant for stretching vibrations, Ko the force constant for bendingvibrations and (/J the angle of ligands. The respective constants are given intable XII.

(8)

TABLE XIIConstants for the calculation of the partition functions of HfX2 compoundsconstant HfF2 HfCb Hffir2 HfI2le (A) 2.16 2.43 2.58 2.77(/J 120° 120° 120° 120°a 2 2 2 2s, (l06 dyn- cm-I) 3.488 2.497 1.202 0.798s; (105 dyn- cm-I) 0.39 0.20 0.15 0.08We (cm-I) 230 156 97.6 60

573 363 181 124601 393 208 149assumed to be equal to those of ZrX2 compounds

where le is the HfX2 bond length and a the symmetry number.

Phllips Journalof Research Vol.42 No. I 1987 27


MXs COMPOUNDS

Hafnium trihalides were assumed to have a pyramidal structure. Vibra-tional frequencies were calculated according to the following equations, seeref. 41, eh. II,4,

2 Ko s, ( Mx )C(W~+ (2) = 3 - + - 1 + 3 -- cos" ffJ ,Mx Mx MM

(10)

2 s,«; ( Mx)C(w~ (2) = 3 --2- 1 + 3 -- cos! (/J,Mx MM

2 2 3Ko x, ( Mx )C(Ws + (4) = -- + - 1 + 3 -- sin" (/J ,2Mx Mx MM

2 2 3 KoKo ( Mx)C(WSW4) = ---2- 2 - sin2(/J + 3--sin2(/J ,4 Mx MM

The respective constants are given in table XIII.

(11)

(12)

and (13)

TABLE XIIIConstants for the calculation of the partition functions of HfXs compoundsconstant HfF s HfCls HfBrs HfIsle (A) 2.16 2.43 2.58 2.77(/J 105° 105° 105° 105°(J 3 3 3 3s, (105 dyn- cm-I) 3.780 2.745 1.675 1.269Ko (105 dyn- cm-I) 0.2403 0.2017 0.1807 0.1360We (cm-I) 187 (1) 135 (1) 97 (1) 72 (1)

234 (1) 161 (2) 105 (2) 74 (2)596 (1) 380 (1) 211 (1) 156 (1)619 (2) 406 (2) 239 (2) 183 (2)

ei (cm-I) assumed to be equal to those of ZrXs compounds

where le is the HfXs bond length.

MX. COMPOUNDS

For hafnium tetrahalides a tetrahedral structure was assumed. Vibrationalfrequencies were calculated via the following equations, see ref. 41, eh, II,4,

2 K;CWI=-

Mx'(14)

2 3KoCW2=--

Mxl; ,(15)

28 PhllIps Journal of Research Vol.42 No. 1 1987

C(W~ +W~) = Ko (1 + 4Mx) + 2KD (1 + 8Mx) (16)Mx 3MM Mx/; 3MM '

2 2 s,KD ( . 4 M X)C(WaW4) = 2 -2-2 1 + -- ,Mx/e MM

(17)


and

The respective constants are given in table XIV.

TABLE XIV

Constants for the calculation of the partition functions of HfX4 compounds

constant HfF4 HfCl4 HfBr4 HfI4

le (Á) 2.16 2.43 2.58 2.77(/J 109.5° 109.5° 109.5° 109.5°(1 12 12 12 12K; (106 dyn- cm-I) 4.159 2.660 2.208 1.730KD (106 dyn- cm-I) 0.1313 0.0813 0.0586 0.0490We (cm-I) 171 (3) 105 (3) 65 (3) 50 (3)

188 (2) 108 (2) 61 (2) 49 (2)610 (1) 357 (1) 216 (1) 152 (1)652 (3) 402 (3) 276 (3) 215 (3)

ei (cm-I) assumed to be equal to those of ZrX4 compounds

where le is the HfX4 bond length.Specific heats and entropies thus calculated by means of statistical mech-

anics were then compared with estimations based on semiempirical equationsdescribed in refs 42 to 46.

3. Experimental

Coiled zirconium and hafnium wires were mounted in glass bulbs of 300 cm avolume. The lamps were baked out at 700 K for 12 hours before filling with0.8 bar Kr premixed with either 10 mbar Hel, 10 mbar HBr or 10 mbar Hl.The gas pressures were measured with a capacitance MKS Baratron pressuregauge. For each experiment related to a specific temperature a new coilwas used. The coils were heated electrically over a period of approximately10 minutes until the current reached a constant value. Filament temperatureswere adjusted pyrometrically. During experiments the glass bulbs were heatedexternally to 700 K to prevent condensation of equilibrium compounds.Finally the coils were reweighed and summed metal pressures was derived viaeq. (2).

Gettering of hydrogen by the colder parts of zirconium or hafnium coilswas investigated by experiments on bulbs filled with considerable amounts of

Philips Journalof Research Vol. 42 No. I 1987 29

5

. -?" ~. .

P70D (Hel) = 24 mîar


H2• From determination of the change in heat conductivity of the gas atmos-phere and the weight change of coils a residual pressure of about 1 mbar H2

could be ascertained. Gettering of oxygen was observed when O2 was added tothe halogen atmosphere. Zirconium or hafnium transport then was not in-fluenced due to the stability of formed oxides and the high solubility of oxygenin the bulk metals. In order to revise sublimation enthalpies of zirconium andhafnium the evaporation rate of coils in evacuated glass bulbs was determinedas a function of filament temperature.Zirconium (99.990/0)and hafnium wires (99.97%) were obtained from Mat-

erials Research Corporation. High purity Kr (99.998%, Messer Griesheim)and the reactive gases HCI (99.995%, L' Air Liquide), HBr (98.8%, Mathe-son) and HI (98.0%, Matheson) were introduced from standard lecturebottles. Prior to use hydrogen halides were purified from excess hydrogen byfractional distillation.

4. Results

For the systems investigated summed pressures of zirconium and hafniumhave been evaluated as a functon of filament temperatures. The curves wereobtained from equilibrium calculations whereas the dots represent experimen-tal values. The corresponding chemical equilibria were calculated under theassumption of constant summed halogen pressures SP (X).The observed H2 partial pressures of 1 mbar in the bulb wall region (see ex-

perimental) correspond to different summed H2 pressures conditioned by thestability of the specific hydrogen halide.Thus in the Zr/HCI system a residual summed H2 pressure SP(H2) of

1.4 mbar results. The corresponding zirconium pressure SP(Zr) (fig. 5) is at a

10

SPZr(mbar)

1500 2000 2500- T(K)

Fig. 5. Summed zirconium pressure in the Zr/HCI system as a function of filament temperature Tr(Tb = 700 K).

30 Phlllps Journal of Research Vol.42 No. 1 1987

2500


constant level between 1000 and 1700K; from 1700 to 2110 K (melting point)a steady increase occurs caused by the high-temperature stability of ZrCl, andZrCl, which is also depicted in the Zr/Cl, partial pressure diagram (fig. 6).With a residual summed H2 pressure of 1.5 mbar the summed zirconium

pressure curve in the Zr/HBr system exhibits a weak maximum of solubility

10->

P zrct,

(bar)

t10-'

2000-- T(K)

Fig. 6. Partial pressures in the Zr/Cl, system as a function of filament temperature Tc(P(CI2) = 5 mbar).

10

5

._~~

__........._..----; .• . • .

P700 (HBr) = 24mbar

I

SPZr(mbar)

t

o1000 2000

-- T(K)Fig. 7. Summed zirconium pressure in the Zr/HBr system as a function of filament temperature Tc(Tb = 700 K).

1500 2500

Philip. Journalof Research Vol. 42 No. 1 1987 31


around 2000 K (fig. 7) due to the decreasing stability of ZrBra with higher tem-peratures (fig. 8).

In contrast to these systems the Zr/HI equilibrium is characterized by anelevated residual summed H2 pressure of 3 mbar. The decrease of the summedzirconium pressure abové 1600 K (fig. 9) results from the fact that it is onlymaintained by Zrh at these temperatures (fig. 10).

///sr////IIIIIII1O.6L_ ~{___..L.- L___J_ __J

1000 1500

10-2

P zrer,

(bar)

t10·'

25002000- T (K)

Fig. 8. Partial pressures in the Zr/Br2 system as a function of filament temperature Ti(P(Br2) = 5 mbar).

10 ~

SPZr(mbar)

t• •

5.-.-.-.-.........

01000 25002000

- T(K)Fig.9. Summed zirconium pressure in the Zr/HI system as a function of filament temperature Te(Tb = 700 K).

32

1500

Phlllps Journal of Research Vol.42 No. 1 1987


10'2

Pzn,

(bar)

t10"

--- T (K)

Fig. 10. Partial pressures in the Zr/I2 system as a function of filament temperature Tr(P(h) = 5 mbar).

2000- T(K)

Fig. 11. Summed hafnium pressure in the Hf/Hel system as a function of filament temperature Tr(Tb = 700 K).

10

SPHf(mbar)

t5

o1000

re- . .. ..-.-. .-. •

P700 (Hel) = 24 mbr1500 2500

The residual summed H2 pressures in the hafnium/halogen systems coincidewith those of the corresponding zirconium/halogen systems.

In the Hf/HCI and Hf/HBr systems the summed hafnium pressures are in-variant from 1000 to 2000 K. Up to 2490 K (melting point) the solubility slowly

PhllIps Journal of Research Vol. 42 No. 1 1987 33


decreases in both cases (figs 11 and 13). From the corresponding partial pres-sure diagrams it is apparent that lower-valency hafnium chlorides andbromides are of minor importance due to the marked high-temperaturestability of HfX4 compounds (figs 12 and 14).

10-2

PHfCl,

(bar)

t10-'

- T(K)

Fig. 12. Partial pressures in the Hf/Ch system as a function of filament temperature Tc(P(Ch) = 5 mbar).

10

5

•• • r--._. _._ • -~~

P700 (HBr) =24mbar

I

SPHf(mbar)

t

o1000 1500 2000

- T(K)

2500

Fig.13. Summed hafnium pressure in the Hf/HBr system as a function of filament temperature Tc(Tb = 700 K).

34 Phlllps Journalof Researeh Vol.42 No. 1 1987


10-2

P(bar)

t10-'

- T(K)

Fig. 14. Partial pressures in the Hf/Br2 system as a function of filament temperature Tc(P(Br2) = 5 mbar).

10

5 ~ -.r..P(HI) =24 mbar

SPHf(mbar)

t

o1000 1500 25002000

- T(K)Fig. IS. Summed hafnium pressure in the Hf/Hl system as a function of filament temperature Tc(Tb = 700 K).

The Hf/Hl system, however, is characterized by a significant drop of so-lubility above 1400 K (fig. 15) which is reflected in the partial pressurediagram by rapid decomposition of HfI4 into HfI2 and I at these tempera-tures (fig. 16).

Phlllps Journalof Research Vol. 42 No. 1 1987 35


10'2

P urt,

(bar)

t10"

-

- T(K)

Fig. 16. Partial pressures in the Hf/h system as a function of filament temperature Tc(P(h) = 5 mbar).

5. Discussion

Measurements of the gas phase solubility of zirconium and hafnium offerthe possibility of determining thermochemical data of molecular species in thetemperature range between 1000 and 2110 or 2490 K, respectively. Forenthalpy data in particular further support is obtained from our estimationmethod which makes use of the dependence of enthalpies on theelectronegativity of ligands.From our evaporation studies it was possible to confirm published sublima-

tion data for zirconium, whereas in the case of hafnium the sublimationenthalpy had to be raised by 23 kJ· mol:" with respect to the revised valuegiven in ref. 30. Accordingly atomic enthalpy data for all hafnium compoundswere adjusted by this amount.

On compiling the sets of enthalpy data for liquid and solid compounds pre-ference was given to maintaining well-established sublimation and vaporiza-tion data, especially if data of gaseous compounds had to be varied in accor-dance with results of transport experiments. Consistent sets of sublimationand vaporization enthalpies and entropies are given in tables IX and X.

5.1. Gaseous compounds

Formation enthalpies of gaseous zirconium tetrahalides have been adoptedfrom the JANAF tables 10). The stability ranges and the decomposition

36 Phlllps Journni of Research Vol.42 No. 1 1987


behaviour calculated with the help of these data coincide with our experimen-tal observations.

As for the zirconium trihalides, data for ZrFs and ZrCla were taken fromJANAF9), the latter being confirmed by our transport measurements. Ourexperimental value for ZrBrs deviates by 66 kJ· mol-I from an estimationgiven in ref. 9. Our transport value for ZrIs ( -1098 kl mol-I) lies in betweenthe estimation in ref. 9 ( - 1162 kI- mol-I) and a mass-spectrometrically de-termined value of -1070 k.I- mol-114,15).

ZrF2 data published in ref. 9 which are based upon mass-spectrometricinvestigations of Murad and Hildenbrand 12)were accepted. Data for ZrCbobtained from a similar experimental set-up 9) are in very close agreement withour transport value. The formation enthalpy of ZrBr2 obtained from ourexperiments (- 935 k.l- mol-I) substantially deviates from the estimated valuein ref. 9 of -1018 kl- mol-I. The transport value for Zrl, (- 800 k.l- mol-I)again falls between the JANAF estimation 9) of - 900 kJ . mol-I and a valueof - 697 kJ· mol-I derived from mass-spectrometric measurements 14,15).

Our tabulated values for zirconium monohalides are merely estimations.Only the JANAF value 9) for ZrCI coincides with our data; estimations for theother monohalides also proposed in ref. 9 seem to be unreliable. An experi-mental value for ZrI 14,15)of - 324 k.l- mol-I deviates considerably from ourenthalpy of - 410 k.l- mol-I. Gaseous zirconium oxides and oxyhalides areunstable in contact with solid zirconium. Data for these compounds given intable III were therefore exclusively obtained from estimations using the elec-tronegativity concept.

Only a limited number of thermodynamic data for gaseous hafnium halidesis available from literature. Our estimations for this system therefore couldonly be based upon enthalpies published for HfF4 11) and HfCl4 lI,SI). In thecase of HfCl4 from our transport investigations a slight variation of theenthalpy from - 2010 to - 2000 kI- mol-I is suggested. In general formationenthalpies of hafnium tetrahalides are about 20 kJ .mol-I more negative thanthose of the corresponding zirconium halides. Our estimations for HfXs andHfX2 compounds could be corroborated by transport studies. For the samereason as discussed in connection with zirconium, data for hafnium oxidesand oxyhalides could not be obtained from transport measurements in thepresence of solid hafnium.

5.2. Solid and liquid compoundsFormation enthalpies of solid and liquid zirconium tetrahalides were

selected in accordance with ref. 10. The values for ZrF4 and ZrCl4 were ob-tained by direct halogenation of highly pure zirconium metal in a bomb calori-

Phlllps Journal of Research Vol. 42 No. I 1987 37

38 Philip, Journal of Research Vol.42 No. 1 1987


meter. Enthalpies of formation of ZrBr4 and ZrI4 were derived from solutioncalorimetry.The estimation for the enthalpy of ZrFs given in ref. 9 deviates by

80 k.I- mol:" from our value. For the calculation ofthe formation enthalpy ofZrCl, disproportionation studies between 670 and 920 K were taken into ac-count 47); the resulting value is in close agreement with other literaturedata 11,21,22).The same procedure was applied to ZrBrs 48) and z-r, 28)yielding enthalpy values which fairly closely approach data given in refs 21and 22.The disproportionation of zirconium dihalides into zirconium metal and the

corresponding tetrahalide was examined to evaluate thermochemical data forthese compounds. In the case of ZrF2 17) an enthalpy close to that given inref. 9 results. The values obtained from disproportionation of ZrCb 2S),ZrBr249) and Zrl, 27,28)at around 670 K agree with literature values.

Our enthalpies for the monohalides estimated via sublimation data are lessnegative by 15 k.l- mol-I in the case of ZrCI11,23) and ZrBr21,22); for ZrI thedifference from values given in refs 21 and 26 amounts to 30 kl- mol-I.The adopted enthalpy of HfF4 was measured by fluorine bomb calori-

metry32,33). Our value for HfCl4 is in close agreement to those obtained fromvapour pressure measurements 26,31,37,38).The value for HfBr 4 was derivedfrom that of the gas using estimated sublimation data. The same treatmentwas applied to HfI4 considering sublimation experiments by Tsirelnikov 39).Our estimations for the hafnium trihalides fit in within our limits with calcula-tions for arci, 21,26,31,34,36),HfBr321,26)and HfI221,37).This situation also applies to HfCl, 21,26,Sl,34,36).As for HfBr2 our enthalpy

of formation lies in between the values given in refs 21 and 25. In the case ofHfI2 our estimation fits in well with data proposed by Lapis 21). Data forhafnium monohalides provided from the same source 21) deviate by about25 k.l- mol:" from our estimations.

6. Conclusion

From transport measurements thermodynamic data for a number of zir-conium and hafnium halides have been derived. These data including thosefor oxides and oxyhalides were checked and completed by means of an esti-mation procedure.With respect to an application of zirconium and hafnium as getter materials

in halogen-filled incandescent lamps the following conclusions may bedrawn:- Zirconium and hafnium are very efficient getters for hydrogen and oxygen

in the temperature range between 400 and 1000 K60).

Evaluation of thermadynamie data

- A regenerative transport cycle of tungsten cannot be maintained in thepresence of zirconium and hafnium up to temperatures of 2000 K becauseof their extreme reactivity towards halogens.

- Zirconium and hafnium only may be applied in halogen atmospheres ifthey are encapsuled by materials permitting rapid permeation of only thosegases 51) wich have to be gettered.

Acknowledgements

Thanks are due to Dr. E. Schnedler for making available free energy func-tions and for helpful discussions. We also appreciate the assistance of Mr.K. H. Christians with computer calculations and of Mr. A. Behrendt and Mr.F. J. Sieben with the preparation of experiments.

REFERENCES1) G. Dittmer, A. Klopfer and J. Schröder, Philips Res. Repts 32,341 (1977).2) G. Dittmer and U. Niemann, Philips J. Res. 36, 87 (1981).8) G. Dittmer and U. Niemann, Philips J. Res. 37, 1 (1982).4) G. Dittmer and U. Niemann, Mat. Res. Bull. 18, 355 (1983).6) G. Dittmer and U. Niemann, Philips J. Res. 40, 55 (1985).6) G. Dittmer and U. Niemann and E. Schnedler, Proc. Sympos. Sci. Tech. High Temp.

Light Sources, Toronto, Canada, 1985, Proc. Vol. 85-2, 185 (1985).7) E. Schnedler, Philips J. Res. 38, 224 (1983).8) E. Schnedler, Calphad 8, 265 (1984).9) JANAF Thermochemical Tables, 2nd ed., Dow Chemicals, Midland, Michigan 1971.10) JANAF Thermochemical Tables, 1978 Supplement, J. Phys. Chem. Ref. Data 7,793 (1978).") D. D. Wagman, W. H. Evans, V. B. Parker, I. Halow, S. M. Bailey, R. H. Schumm

and K. L. Churney, Selected values of chemical thermodynamic properties, NBS TechnicalNote 270-5, Washington D.C. 1971.

12) E. Murad and D. L. Hilden brand, J. Chem. Phys, 45, 4751 (1966).18) V. I. Tsirelnikov, G. K. Selivanov and A. A. Maltsev, Zh. Fiz. Khim. 47, 1046 (1973).14) D. Cubicciotti, D. L. Hildenbrand, K. H. Lau and P. D. Kleinschmidt, Proc. Symp.

High Temp. Met. Hal. Chem., Atlanta, USA, 1977, Proc. Vol. 78-1, 217 (1978).16) P. Kleinschmidt, D. Cubicciotti and D. L. Hildenbrand, J. Electrochem. Soc. 125,

1543 (1978).16) D. Cubicciotti, K. H. Lau and M. J. Ferrante, J. Electrochem. Soc. 125, 972 (1978).17) F. K. McTaggart and A. G. Turnbull, Australian J. Chem. 17,727 (1964).18) L. M. Nij lan d and J. Schröder, Angew. Chem. 76, 890 (1964).19) L. N. Shelest, T. Z. Maiskaya, E. K. Safronov and A. S. Mikhailova, Nauch. Tr.,

Nauch.-Issled. Proekt. Inst. Redkometal. Prom. 1972, No. 44, 5; Ref. Zh., Met. 1973, Abstr.No. 5G322; C.A. 80, 74922 (1974).

20) V. I.Tsirelnikov, Vu. S. Khodeev and S. I.Troyanov, Zh. Fiz. Khim. 47, 1567 (1973).21) Kh. S. Lopis, G. S. Marek, A. S. Izmailovich, S. I.Troyanov and V. I.Tsirelnikov,

Vses. Konf. Kalorim., Rasshir. Tezisy Dokl., 7th, 2, 304 (1977); C.A. 91, 217820 (1979).22) S. I. Troyanov, G. S. Marek and V. I. Tsirelnikov, Izv. Sib. Otd. Akad. Nauk SSSR,

Ser. Khim. Nauk 111, (1974).28) K. Uchimura and K. Funaki, Denki Kagaku 33, 163 (1965); C.A. 63, 1466 (1965).24) A. G. Turnbull and J. A. Watts, Australian J. Chem. 16, 947 (1963).26) S. N. Lungu, Acad. Rep. Populare Romine, Studii Cercetari Fiz. 13, 29 (1962).

Philip. Journal of Research Vol. 42 No. 1 1987 39

40 Phillps Jeumal of Research Vol.42 No. 1 1987


26) R. H. Lamoreuax and D. Cubicciotti, J. Electrochem. Soc. 128,457 (1981).27) F. R. Sale and R. A. J. Shelton, J. Less-Common Met. 9, 60 (1965).28) F. R. Sale and R. A. J. Shelton, J. Less-Common Met. 9, 64 (1965).29) A.!. Morozov and E. V. Kariova, Russ. J. Inorg. Chem. 16, 12 (1971).30) R. Hultgren, R. L. Orr, D. Anderson and K. K. Kelley, Selected Values of Therrno-

dynamic Properties of Metals and Alloys, John Wiley and Sons, New York, 1963.31) I. Barin and O. Knacke, Thermochemical Properties of Inorganic Substances, Springer-

Verlag, Berlin, 1973.32) I. Barin, O. Knacke and O. Kubaschewski, Thermochemical Properties of Inorganic

Substances, Supplement, Springer-Verlag, Berlin, 1977.33) E. Greenberg, J. L. Settle and W. N. Hubbard, J. Phys, Chem. 66, l345 (1962).34) A. S. Izmailovich, V. I.Tsirelnikov and L. N. Komissarova, Russ. J. Inorg. Chem.

24, l309 (1979).35) N. A. Filatova, V. A. Kozhemyakin and B. G. Korshunov, Nauch. Tr., Nauch.-Issled.

Proekt. Inst. Redkometal. Prom. 1972, No. 44, 15; Ref. Zh., Met. 1973, Abstr. No. 12B71;C.A. 80, 31074 (1974).

36) L. P. Ruzinov and S. F. Belov, Tsvetn. Metal. 35, 85 (1962); C.A. 58, 6258 (1963).37) G. L. Galchenko, D. A. Gedakyan and B. I. Timofeev, Russ. J. Inorg. Chem. 13, 159

(1968).38) P. Gross and C. Hayman, Trans. Faraday Soc. 60, 45 (1964).39) V. I. Tsirelnikov and M. I. loffe, Russ. J. Inorg. Chem. 11, 1283 (1966).40) G. Herzberg, Molecular Spectra and Molecular Structure, I. Spectra of Diatomic Mole-

cules, Van Nostrand Reinhold Company, 2nd ed., New York, 1950.41) G. Herzberg, Molecular Spectra and Molecular Structure, 11. Infrared and Raman Spectra

of Polyatomic Molecules, Van Nostrand Company, Princeton, New Yersey, 1945.42) R. R. Wenner, Thermochemical Calculations, McGraw Hill, New York, 1941.43) W. M. Latirn er, J. Am. Chem. Soc. 73, 1480 (1951).44) G. Geiseler, Z. Phys. Chem. 202, 424 (1954).45) E. Sirtl, Z. Naturforsch. 21A, 2001 (1966).46) G. M. Neumann and W. Kn a tz, Z. Naturforsch. 26A, 863 (1971).47) V. S. Emelyanov, A. I. Evstyukhin, I. P. Barinov and A. M. Samonov, Razdelenie

Blizkikh po Svoistram Redkikh Metal 51, (1962).48) H. L. Schlaefer and H. Skoludek, Z. Electrochem. 16, 327 (1962).49) W. C. Schumb and L. K. Morehouse, J. Am. Chem. Soc. 69,2696 (1947).50) G. Kuus and W. Martens, J. Less-Common Met. 75, lil (1980).51) G. Kuus, J. Vac. Sci. Technol. 13, 728 (1976).

Authors

Georg Dittmer; M.Sc. (Physics), University of Vienna, Austria, 1960; Ph. D., University ofVienna,1962; Siemens Entwicklungslabor, Munich, Germany, 1962-1964; Philips Forschungslab-oratorium Aachen, Germany, 1964- . In his thesis work he was concerned with point defectsand the electronic structure of metals. At Siemens he was concerned with vacuum and thin filmtechnology. At Philips he first continued this interest studying thin film photo emission and fieldelectron emission. Since 1972 his work is related to problems in halogen-filled incandescent lamps,particularly to their energy balance and to the thermochemistry of halide systems.

Ulrich Niemann; Diplom-chemiker, Ruhr-University, Bochum, Germany, 1971; Ph.D. (Inor-ganic Chemistry), Ruhr-University, Bochum, 1974; Philips Forschungslaboratorium Aachen,Germany, 1975- . In his thesis he was concerned with perfluoroorganic heterocycles. At Philipshe started with studies on regenerative cycles in halogen-filled incandescent lamps. Subsequentlyhe investigated reactions of lamp-relevant materials via transport experiments. Presently he isworking mainly on problems related to high-pressure discharge lamps.

evaluation of thermodynamic data on zirconium and … bound... · evaluation of thermodynamic data...

Documents