THERMOCHEMICAL STUDY of the GERMANIUM OXIDES USING a MASS SPECTROMETER Thermochemical Study of the Germanium Oxides Using a Mass Spectrometer

Offprinted from the Transactions of The Faraday Society, No. 510, Vol. 61, Part 6, June, 1965

THERMOCHEMICAL STUDY OF THE GERMANIUM OXIDES USING A MASS SPECTROMETER Thermochemical Study of the Germanium Oxides Using a Mass Spectrometer

Dissociation Energy of the Molécule GeO *

By J. DROWART, F. DéGRèVE f, G. VERHAEGEN AND R. COLIN

Laboratoire de Chimie Physique Moléculaire, Université Libre de Bruxelles Brussels, Belgium Received ISth November, 1964

The mass spectrometric study of the vaporization of the compounds GeOzihex) and GeO(am) and of the mixture Ge02(/ie;c)+Ge(c) made it possible to establish their mode of vaporization : Ge02(hex) ->GeO(tf)+iOa (D n GeO(a/7!) -*(GeO)„(ff) (« = 1,2,3), (H) {nl2)Ge02{hex) + (n/2)Ge(c) ^GeO)„(ff) (n = 1,2,3). (UI) The enthalpies of vaporization are A^a,(I) = 124-3 ±2-4 kcal/mole; àHl,i{U, n = 1) = 53-1 ± 1-Okcal/mole; AJÏâgCIII, « = i) = 57-5±1-0kcal/mole. The polymerization énergies are àHl,, (GeO—GeO) = 44-7±3-0 kcal/mole; A//°5s(GeO—GeO—GeO) = 88-5±5-0 kcal/mole. Total pressures given in the literature were re-interpreted taking the présence of the polymers into account. The heat of formation of the metastable compound GeO(ûm) is A///(GeO) = —60-8±l-4 kcal/mole. The dissociation energy of the gaseous molécule GeO is i)S(GeO) = 156-2 ±1-9 kcal/mole.

As part of a study i of the composition of the vapour in equilibrium with the group 4B-group 6B compounds and of the thermochemical détermination 2 of the dissociation energy of the corresponding molécules and of their polymers, the vapour over germanium monoxide, a mixture of germanium dioxide + germanium 3 and germanium dioxide 3 was analyzed mass spectrometrically. Germanium monoxide is amorphous(aw) and metastable and disproportionates to germanium and germanium dioxide.''' 5 Détails concerning this disproportionation were obtained in the présent study. Germanium dioxide can exist in three modifications, the tetragonal, hexagonal(/îex) and glassy(^/).fi Although the low- temperature form is the tetragonal (insoluble) form, the more common one is the hexagonal (soluble) form. The heat of formation ' and thermodynamic properties of the hexagonal form are known.s» To the extent to which similar glassy germanium dioxide samples can be obtained, the thermodynamic properties of the latter are also estabhshed.8* The total pressures of germanium monoxide and of the mixture of hexagonal germanium dioxide and crystalline(c) germanium were determined by Bues and von Wartenberg ^ and by Jolly and Latimer.5 The former authors measured the pressure of both Systems by a manometric method and that of Ge02(/îeA-) + Ge(c) also by the transport method. The latter authors apphed the Knudsen technique to initial GeO samples but concluded that thèse had disproportionated to GeQ+Ge.

* The research reported in this document has been sponsored in part by Air Force Materials Laboratory, Research and Technology Division, AFSC, under contract AF 61(052)—225 through thè European Office of Aerospace Research (OAR), United States Air Force. t Boursier de l'Institut pour l'Encouragement de la Recherche Scientifique dans l'Industrie et l'Agriculture (IRSIA). 1072 J. DROWART, F. DéGRèVE, G. VERHAEGEN AND R. COLIN 1073 From their own measurements and from those of Bues and von Wartenberg, JoUy and Latimer deduced the dissociation energy of the GeO molécule, £)o(GeO) = 157-2+3-0 kcal/mole, compared to the value DÔ(GeO) = 157 + 4 kcal/mole estimated by analogy with SiO from spectroscopic data by Barrow and RowUnson.9 From the data of Bues and von Wartenberg for GeO and for GeOz+Ge, Jolly and Latimer also deduced an approximate heat of dismutation of GeO of about — 7 kcal/mole. The pressure of GeOa, contained in quartz cells, was determined by Davydov,io<» and by Shimazaki, Matsumoto and Niwa lo* using the Knudsen technique. Shchukarev and Semenov n studied the sublimation of the same substance mass spectrometrically and identified the gaseous(5') molécules (GeO)2 and (GeO)3. The systematic mass spectrometric study 1-2 of the subhmation of the group 4B- group 6B compounds (group 4B = C, Si, Ge, Sn, Pb, represented by Me ; group 6B = O, S, Se and Te represented by X) as well as the previous mass spectrometric study of SiOz and SiOa+Si by Porter, Chupka and Inghram 12 indicate that the polymers (MeX)» are typical of the MeX compounds or MeXz+Me mixtures, but not of the MeX2 compounds, which vaporize predominantly by one of the two processes MeX2(c)->MeX(c)+iX2(fir) ; or MeX2(c)^MeX(^)+iX2(5r). The sublimation of GeO(am), of the mixture GeOzihex) + Ge(c) and of GeOzihex) was therefore reinvestigated. The présent paper reports the results of thèse studies.

TECHNIQUE The expérimental technique and procédure have been described previously.i^ The mass spectrometer used i'*-is a single focusing, 20 cm radius of curvature, 60° sector instrument, equipped with a secondary électron multiplier.is The Knudsen ceUs containing the samples were made of quartz and were placed within molybdenum shells heated by radiation from a concentric tungsten loop. The températures were measured with Pt/(Pt—10 % Rh) thermocouples. The dimensions of the aknost circular effusion orifices were measured with a microcomparator. Their area was varied between 5 x 10~3 and 10^2 cm2 and was small compared to the surface of the sample. The thickness of the effusion orifice was also measured to evaluate the Clausingi^ factor (about 0-8). The samples were standard germanium (99-999 % purity) and hexagonal germanium dioxide (checked by X-ray examination). The metastable monoxide was prepared by vacuum sublimation of a stoichiometric Ge02(feA:)+Ge(c) mixture, as described by Bues and von Wartenberg.'' It was brown-black and amorphous as verified by X-ray examination.

RESULTS

COMPOSITION OF THE VAPOUR AND PRESSURES The ions formed by électron impact from the molecular beam issuing from the cell containing either GeO(aw) or Ge02{hex) + Ge(c) are Ge+, GeO+, Ge20+, Ge20J and GesOj. Those formed from cells containing GeOzihex) are OJ, Ge+ and GeO+. The approximate appearance potentials were obtained by the linear extrapolation method, the energy scale being calibrated with the appearance potentials of the H2O+ ion.i8 They are Ge+, 14-0+1; GeO+, 10-l±0-8; OJ, 12-2±0-5; Ge20+, 14-3+1-0; Ge20î, 8-7 +1-0; and Ge30+, 8-6± 1-0eV. The appearance potential of GeO+, 10-1 ±0-8 eV compared to that of a number of isoelectronic ions, N|, 15-6 1»; PJ, 10+0-5 i9 (11-8+0-5 20) ; As+, IO+O-521 (ll-0± 0-5 20); SbJ, 8-4± 0-3 22; CO+, 14-018; SiO+, 10-5 9(10-8 + 0-5 12); and SnO+, 10-0 + 0-7,2 indicates this ion to be formed directly from the GeO molécule. Similarly, the appearance potentials of the ions Ge20J, 8-7 + 1 eV and GeaOJ, 8-6+1 eV indicate thèse to be parent ions from the corresponding molécules. The appearance potential of Ge+, 14 0 eV compared to the spectroscopic value for the

2 'y U û d 3 1074 THERMOCHEMICAL STUDY OF GERMANIUM OXIDES ionization potential, 8-13 eV 23 shows this ion to be formed by fragmentation, mainly of the GeO molécule. The situation is analogous for GezO*, whose appearance potential is much higher than expected for a parent ion. It is considered to be formed by fragmentation, mainly of the GezOz molécule. The prédominant molécules in the vapours of both the GeO(am) and GQ02{hex) + Ge Systems are therefore GeO, Ge202 and GeaOa. In the Ge02 System, they are GeO and O2. The main vaporization processes are thus Ge02{hex)^GeO(g)+i02ig). (I) n GeO(am)^(GeO)„(g), (n = 1,2,3), (II) (n/2)Ge02(/ieA:)+(n/2)Ge(c)^(GeO)„(5), (« = 1,2,3). (III) The partial pressures were determined either by completely subliming samples of a few mg or by determining the weight lost by subUmation during a given time by more important samples. In both cases the différent (GeO)^ intensities were measured and integrated with time. By replacing in the Hertz-Knudsen relation, G = P{MI2nRT)ist (G = weight loss, P = pressure, M = molecular weight of the subUming molécule, R = gas constant, T = température, s = effective area of the effusion orifice, t = interval at température T) the pressure by Pn = InTjanjnk (/» = intensity of species n, c7„ = relative ionization cross-section of species n, 7n = secondary électron multiplier yield for ion n, k = proportionality constant), one obtains, when several species n are simultaneously responsible for the weight loss and when the experiment is carried out at several successive températures during a given time interval At :

with A„ = 2/„riA/. By analogy with a number of diatomic 24 and dimeric 25 molécules, the ratio CGe^Oj/^Geo was takcn equal to 1-6, and the ratio aGe^oJcroeo to 2-1. The relative multiplier yields were read from the cahbration curve of the multiplier.26 Mole• cular effects were corrected for as suggested by Stanton, Chupka and Inghram.27 The relative ay values used are 1,0 61 and 0-52 for GeO, Ge202 and GesOs respectively. The partial pressures of the monomer, dimer and trimer are given separately for GeO(a/w), Ge02(^/) H-Ge(c) and Ge02(/!e;c) + Ge(c) in tables 1-4.

TABLE 1.—HEAT OF SUBUMATION OF THE GeO MOLéCULE FROM GeO(am)

T -logiKGeO) -A[{G°-HM8)/r] expt. AH298 (atm) (kcal mole"') 09-23 754 6-46 (±0-15) 41 1 (±0-6) 53-3 768 6-17 410 53-2 775 5-94 410 530 788 5- 76 40- 9 53-1 09-25 766 6- 22 41- 0 53-2 757 6-43 41-0 53-3 744 6-73 41-1 53-4 769 6-12 41-0 53-1 778 5-90 41-0 52-9 786 5-79 40-9 52-9 795 5-54 40-8 52-7

average 53-1 standard déviation ±0-2 total uncertainty ±1-0 J. DROWART, F. DÉGRÈVE, G. VERHAMEN AND R. COLIN 1075

TABLE 2.—HEAT OF SUBLIMATION OF THE GeO MOLéCULE FROM GeO2(50+Ge(c) T -logp -A[(G°- -HhatIT] AHÎ98 eicpt. (°K) (atm) (cal deg.~i mole"0 (kcal mole-') 09-23 806 5-66 (±0-15) 41-8 (±0-5) 54-6 817 5-57 41-8 (±0-5) 54-9 843 5-13 41-7 54-9 855 4-97 41-7 550 878 4-64 41-7 55-3 09-24 789 5-76 41-9 53-8 768 6-54 41-9 54-2 789 6-10 41-9 551 800 5-87 41-8 550 807 5-77 41-8 551 819 5-60 41-8 551 830 5-42 41-8 55-2 839 5-31 41-8 55-4 845 5-20 41-8 55-3 09-25 844 4-96 41-8 54-3 841 5-05 41*7 54-5 ou 41-8 55-5 769 6-92 41-9 56-6 801 6-04 41-8 54-9 779 6-55 41-9 560 770 6-82 41-9 56-3 782 6-48 419 55-9 793 6-19 41-9 55-7 818 5-46 41-8 54-5 785 6-47 41-9 561 826 5-36 41-8 54-7

average 55-4 standard déviation ±0-8 total uncertainty ±1-0

TABLE 3.—HEAT OF SUBLIMATION OF THB GeO MOLéCULE FROM Ge02(/ieJt:)+Ge(c)

T —log Paeo -&[(.G' -HÎn)IT] A//298 expt. (°K) (atm) (cal deg.-' mole-i) (kcal mole~i) 825 6-20 (±0-15) 42-4 (±0-4) 58-3 784 6-97 42-5 58-3 770 7-17 42-5 580 754 7-72 42-5 58-7 795 6-82 42-5 58-6 840 5-85 42-4 580 806 6-56 42-4 58-4 849 5-63 42-3 56-8 880 4-99 42-3 57-2 863 5-35 42-3 57-6 811 6-44 42-4 58-3 779 7-12 42-5 58-5 827 6-19 42-4 58-4 831 6-14 42-4 58-5 851 5-70 42-3 58-2 857 5-59 42-3 58-1 1076 THERMOCHEMICAL STUDY OF GERMANIUM OXIDES

TABLE 3—contd.

T — log Pqco expt. CK) (atm) (cal deg.-i mole-i) (kcal mole !-') 870 5-29 42-3 57-8 882 501 42-3 57-5 897 4-68 42-2 57-1 915 4-36 42-2 56-8 937 406 42-2 56-9 927 4-20 42-2 56-9 948 3-94 42-1 57-0

average 57-8 standard déviation ±0-7 09-25 14^ points (fig. 1) average 57-0 standard déviation +0-5 total uncertainty ±1-0

TABLE 4.—PARTIAL PRESSURE OF Ge202(ér) AND GeiO^ig) OVER GeO(am), Ge02(^/)+Ge(c) AND Ge02(hex)+Geic)

T -log Kz -log K3 condensed phase expt. ^06303 (°K) (atm) (atm) (atm) (atm) 754 6-92 6-96 6-00 12-42 768 6 56 6-54 5-78 11-97 GeO(a/w) 09-23 775 6-36 6-34 5-52 11-48 788 617 6-21 5-35 11-07 817 6-21 6-36 4-95 10-38 843 5-68 5-83 4-56 9-56 GeO2(^0+Ge(c) 09-23 853 5-52 5-67 4-44 9-27 876 5-24 5-43 408 8-55 768 7-35 5-73 789 6-84 7-01 5-36 11-29 09-24 800 6-73 6-73 5-01 10-88 807 6-43 6-57 511 10-74 819 6-29 6-44 4-81 10-36 830 609 6-26 4-73 10-00 Ge02(hex)+Ge(,c) 09-17 840 7-07 4-65 848 6-78 4-50 851 6-77 4-61 857 6-64 4-55 863 6-42 6-92 4-27 911 871 6-22 6-65 4-36 9-22 879 5-85 6-34 4-14 8-71 882 5-89 6-28 4-12 8-74 897 5-44 5-76 3-90 8-40 914 5-05 5-36 3-69 7-79 927 4-91 5-21 3-53 7-45 937 4-66 4-90 3-49 7-32 950 4-50 4-60 3-32 6-91 09-24 842 5-88 6-04 4-70 9-95 845 5-85 600 4-55 9-57 880 5-76 6-02 4-12 8-80 900 5-62 5-93 3-90 8-35 898 5-70 608 3-94 8-38 914 5-50 5-87 3-60 7-93 933 5-42 5-76 3-56 7-71 J. DROWART, F. DEGREVE, G. VERHAEGEN AND R. COLIN 1077

DISPROPORTIONATION OF GERMANIUM MONOXIDE As expected for a metastable System, the pressures (intensities) of the différent gaseous species in equilibrium with GeO were higher than those for the GeOiihex) + Ge(c) mixture at the same température. (This feature made it possible to study the {GeO)n{g)-*n GeO(g) {n = 2,3) equilibria over a much wider température interval than would have been possible in the Ge02(/ie;c) + Ge(c) System alone.) When increasing the température to about 800°K, the intensity (pressure) of ail three (GeO) n species decreased with time and température and eventually reached a new steady level, indicating disproportionation to have occurred. In four experiments, carried out with samples of comparable size (100 mg) the températures at which the disproportionation took place was the same within some 25°. The ap• parent rate of transformation, which was not studied systematically, was also reproducible.

Il 1-2 1-3 io3/r°K FiG. 1 .•—Disproportionation of GeO(am).

After disproportionation the partial pressures of the three (GeO)„ species were, however, still higher than those in the Ge02(/!ex) + Ge(c) system. It was therefore concluded that the GeOz formed was not the hexagonal but the glassy form. X-ray examination of a sample obtained by interrupting one experiment immediately after the disproportionation took place showed indeed only the présence of crystalline germanium. Another argument for considering the GeOa formed to be glassy form is that the GeO partial pressure as well as its température dependence within the interval 770-830°K was not entirely reproducible from one experiment to the other, indicating sUghtly différent " glasses " to be formed. Furthermore, the slope dlni'/d(l/r) was in one experiment higher, rather than lower, than that for the Ge02(/!ex) + Ge(c) system, which is a thermodynamic inconsistency. When the samples of Ge02(fir/) + Ge(c) obtained by disproportionation of amorphous GeO were heated to about 900°K, a further decrease in partial pressures 1078 THERMOCHEMICAL STUDY OF GERMANIUM OXIDES gradually took place. The relative intensities of the (GeO)» species and the ab- solute pressures became identical with those in the Ge02{hex) + Ge(c) System. X-ray examination of the samples so obtained now showed the présence of hexagonal Ge02 in addition to crystalline germanium. The observations, presented for GeO(g) alone and for one experiment (09-25), are represented in fig. 1. The disproportionation of GeO(am), first to GeOzigl) + Ge(c) and the transformation of glassy into hexagonal GeOz next, are in agreement with Ostwald's rule. In an attempt to observe also the transformation of hexagonal into tetragonal dioxide, a mixture of the former and of crystalline germanium was heated to 1000°K. No transformation took place under the conditions of the experiment. Because of the value of the pressures at the latter température, which are at the limit where Knudsen conditions are still satisfied, the sample was not heated to higher températures.

DISCUSSION Reaction enthalpies A//298 were calculated using the relation, AG° = -RT\nK = RT\n UPl" = AH^gg + TAl{G°-H°2gg)IT'], (AG° = change in Gibbs free energy accompanying the reaction considered ; /"„ = the partial pressure, in atm, of the molécule « ; = the stoichiometric coefficient of molécule n; {G° — H29s)IT = the free energy function). The numerical values for the free functions of Ge(c),29 02,29 GeO(aw),3o GeO(g),2s GeOzigl) « and GeOi (hex) 8 were taken from the hterature.

THERMODYNAMIC PROPERTIES OF THE CONDENSED COMPOUNDS

HEAT OF FORMATION OF GERMANIUM MONOXIDE The ratio of the GeO pressures over the metastable monoxide and over the mixture of GeOzihex) + Ge(c) (fig. 1 and 3) directly gives the free energy of dismuta• tion. Since the ratio of pressures or more precisely, intensities, was obtained each time within one single experiment and was therefore independent of instrumental factors, the accuracy is good. The average value at 800°K is AGgoo = -4-1 ±0-3 kcal/mole. Together with the free energy function estimated by Coughlin 3" it gives AH^gg = — 5-3 ±0-6 kcal/mole. It can be compared with the A/Z&s value derived for the same reaction, viz., GeO(aAM)->iGe02(Âex)-l-iGe(c) determinated by e.m.f. measurements by Jolly and Latimer 3i for the GeO/Ge02 couple : GeO(am)-H H2O(0-* GeO(ftex)-h 2H+ + 2e, E = 0118+0010 V, or AG298 = - 5-442+0-46kcal/mole^' AG298 = 0 (standard électrode) H2(0) + iO2(0)->H2O(Z) AG298 = -56-690+0-001 kcal/mole iGe02(ftex)-H'iGe(c)+^02(0) AG298 = +59-65+ 1-2 kcal/mole'' GeO(am)-^iGe02(hex) + iGe(c) AG^gg = - 2-4+1-3 kcal/mole leadingto A/ffgs = -2-9 +1-3 kcal/mole. Together with the beat of formation of GsOzihex),'^ AiZ&s,/= -132-2+1-2 kcal/mole, the value determined here, AH^gg = -5-3 + 0-6 kcal/mole leads to the beat of formation of amorphous GeO, AH^g^j = — 60-8 +1 -4 kcal/mole. J. DROWART, F. DéGRèVE, G. VERHAEGEN AND R. COLIN 1079

FREE ENERGY OF TRANSFORMATION OF GLASSY INTO HEXAGONAL GER• MANIUM DIOXIDE The free energy of transformation of glassy into hexagonal germanium dioxide was obtained from the pressure ratio in the same way as the heat of dismutation of the monoxide. The average value is AGgoo = -2-3±0-8 kcal/mole, compared to —1-5 kcal listed by Mah and Adami s» for one particular glass sample. If meaningful, the dilference indicates the glasses not to be identical. Since the présent data do not permit the séparation of enthalpy and entropy contributions, the thermodynamic data given for glassy GeOz by Adami and Mah were used in the subséquent calculations.

HEAT OF SUBLIMATION OF GeO The beats of sublimation of GeO from amorphous GeO, from the mixture of glassy Ge02 + crystalline Ge and from the mixture of hexagonal GeOa + crystalline Ge and from the mixture of hexagonal GeOa + crystalline Ge are summarized in table 1-3.

3. REINTERPRETATION OF TOTAL PRESSURES

GERMANIUM MONOXIDE The pressures over samples which were initially metastable GeO, determined by Bues and von Wartenberg, were measured by a manometric method and are therefore the sum of the partial pressure of the monomer, dimer and trimer, provided

FiG. 2.—Equilibrium constants for the dimerization and trimerization of gaseous GeO.

those of higher polymers not observed here can be neglected. The total pressures given by the latter authors were re-interpreted accordingly, using the extrapolated equilibrium constants for the reactions (GeO)„(^)->«GeO(g)(« = 2,3) (see fig. 2). 1080 THERMOCHEMICAL STUDY OF GERMANIUM OXIDES The partial GeO pressures so obtained (table 6a) are represented in fig. 3 which summarizes the data for the différent Systems and investigations. The pressures measured by JoUy and Latimer for initial GeO samples by the Knudsen method can be re-interpreted in a similar way, by writing

.Pc PoeO — PoeO' \ ^ GeO f GeO where Pceo is the apparent GeO pressure. The pressure measurements by the latter authors were carried out in the tem• pérature région where in the présent experiments disproportionation occurred.

lOî/rK FIG. 3.—Review and summary of GeO partial pressures. O, GeO(a/n) ^ x, Jolly and Latimer Ire-interpreted ©, Ge02(i;/)+Ge(c) Vthis work A, Bues and von WartenbergJ • , Ge02(A«)+Ge(c)J

The apparent scatter in the points obtained by them indicates that the same probably occurred during their experiments. Of the eight measurements numbered here (table 5) 1-8 in the séquence in the original publication,^ three (1, 2, 4) were carried out with fresh samples.33 The pressure in run 1 is close to those obtained here before disproportionation occurred and probably corresponded to GeO(aw). The pressures in runs 2 and 4 are close but above those for Ge02(Êf/) + Ge(c), probably indicating that during the measurement disproportionation occurred. The corresponding pressures are therefore considered to represent upper limits for the System Ge02(5^/) + Ge(c). The pressures for runs 3, 5, 6 and 7 are in good agreement with those measured here for Ge02(g'/)-l-Ge(c) and are therefore considered to pertain to that System and to confirm the présent values. The pressure in run 8 J. DROWART, F. DéGRèVE, G. VERHAEGEN AND R. COLIN 1081 finally was obtained at a température at which GeOzigl) has transformed in the présent work into GeOzihex). The same probably occurred during the measurement by Jolly and Latimer which therefore gives an upper limit for the GeOzihex) + Ge(c) System.

TABLE 5.—RE-INTERPRETED PRESSURE DATA FOR GeO(am), Ge02(fl'/)+Ge{c), and GeOzihex)+Ge(c). (JOLLY AND LATIMER, KNUDSEN TECHNIQUE)

T — log p' condensed -A[(G°-H298)/r] no. (°K) (atm) (atm) phase (caldeg.~i mole"i) (kcal mole 1 770 5-89 6-30 GeO(aw) 41-0 53-7 2 788 5-75 608 intermediate 3 816 5-59 5-78 Ge02(Éï/)+Ge(c) 41-3 55-1 4 835 4-96 5-29 intermediate 5 758 7-17 7-23 Ge02(5/)+Ge(c) 41-4 56-5 6 816 5-82 5-95 41-3 55-8 7 790 6-43 6-51 41-3 56-1 8 859 5-11 5-25 Ce02(hex)+Geic) 42-3 56-9

Since in this température range, the dimer and trimer are relatively unimportant, the total pressure is close to the partial pressure of the monomer. Therefore, the Knudsen measurements by Jolly and Latimer are a cross check of the présent measurements, which dépend on the estimate of the ratio of the ionization cross- sections of the dimer and trimer relative to that of the monomer.

GeOzihex)+ Geic) The total pressures by Bues and von Wartenberg, measured by a manometric and by the transport method were re-interpreted in a similar manner as explained above, but taking into account that in the transport method

\ "GeO •'^GeO / The partial GeO pressures obtained and the heat of vaporization of GeO calculated therefrom are summarized in table 6b and fig. 3. Ge02 The pressures measured by Davydov i<"» and by Shimazaki et alA'>'> by the Knudsen method were recalculated to take the stoichiometry of reaction (I) into account. Davydov made the assumption that the gaseous molécule is Ge02 (or a polymer thereof).* The relation between the partial GeO and the apparent pressures is thus

'A^GeoA*/, , Mo, \Ma,o/ \ 2M, Shimazaki et al. used the approximation that the molecular weight is the average of that of GeO and O2 instead of the more accurate relation, G InRT^f^ Me •PGCO — St MceO The recalculated partial GeO and O2 pressures are given in tables la and Ib together with the enthalpy for reaction (I). Davydov's results show some variation with

* That the vaporization process should be GeC)2(hex)-^Ge.O{g)+\02 was already pointed eut by Bergman, Zhurn. Neorg. Khim., 1958, 3, 2422. 1082 THERMOCHEMICAL STUDY OF GERMANIUM OXIDES

TABLE 6C.—RE-INTERPRETED PRESSURE DATA FOR GeO(am) (BUES AND VON WARTENBERG, MANOMETRIC METHOD)

T -logP* -AHG°-Hhs)IT] (•K) (atm) (atm) (cal deg.-' mole-') (kcal mole~0 913 2-63 3-67 40-7 52-6 917 2-36 3-55 40-7 52-2 948 1-89 3-13 40-5 520 978 1-43 2-73 40-4 50-7

average 51-9

TABLE 6è.—RE-INTERPRETED PRESSURE DATA FOR Ge02ihex)+Ge(c) (BUES AND VON WARTENBERG, MANOMETRIC AND TRANSPORT METHODS)

T V j -logf method (atm) (atm) (cal deg.-i mole-i) (kcal mole-i) 1027 manometric 2-63 302 420 57-3 1038 2-36 2-95 42-0 57-6 1042 2-62 2-83 41-9 57-1 1057 216 2-65 41-9 57-1 1084 1-89 2-40 41-9 57-3 1123 1-43 201 41-8 57-2 980 transport 3-05 3-47 42-1 56-8 1081 1-90 2-41 41-9 57-1

average 57-2

TABLE 7a.—RE-INTERPRETED PRESSURE DATA FOR Ge02 (DAVYDOV, KNUDSEN TECHNIQUE)

T -logP* -loiPoeo A//S98 (°K) (atm) (atm) (atm) (cal deg.-' mole"') (kcal mole~0 1159 5-62 5-66 6-18 62-9 119-3 1201 5-43 5-47 5-99 62-8 122-0 1227 5-31 5-17 5-69 62-8 123-5 1248 5-23 5-26 5-78 62-7 124-7 1268 5-13 5-17 5-69 62-6 125-9 1288 4-98 502 5-54 62-6 126-4 1296 4-75 4-79 5'31 62-5 1250 1338 4-43 4-46 4-98 62-4 126-1 1351 4-28 4-32 4-84 62-4 126-0

average 124-3

TABLE 7b.—RECALCULATED PRESSURE DATA FOR Ge02 (SmMAZAKi et al., KNUDSEN TECHNIQUE)

T A//298 (°K) (atm) (atm) (cal deg.-' moIe->) (kcal mole"i) 1313 6-17 6-69 62-5 139-1 1323 606 6-58 62-5 139-1 1333 5-87 6-39 62-5 138-5 1343 5-74 6-26 62-4 138-3 1353 5-60 6-12 62-4 138-0 1363 5-52 604 62-4 138-2 1373 5-34 5-86 62-3 137-5 J. DROWART, F. DéGRèVE, G. VERHAEGEN AND R. COLIN 1083 température, which he attributed to the transformation of hexagonal into tetragonal Ge02. The average third-law reaction enthalpy corresponds, however, to the value calculated from the dissociation energy of GeO to be given below and the beat of formation of hexagonal GeOa. It is therefore accepted that the points correspond to the vaporization of hexagonal GeOi which is the stable form above 1306 + 10°K.6 The high value for reaction (I) calculated from the Shimazaki et al. data is apparently to be attributed to interaction between GeOa and quartz, which was also noticed by Davydov and here when GeOa was vaporized from SiOi crucibles. The observation of the GeaOz and GesOa polymers in the vaporization of GeOi by Shchukarev and Semenov n is completely at variance with the mass spectrometric and thermodynamic results of the présent study. In the investigation referred to, GeOa was vaporized from a platinum filament attached to a nichrome holder. A plausible reason of the discrepancy is therefore that GeOa was reduced by the latter alloy, which would explain the présence of the polymers, characteristic of the GeOa+Ge system.

DISSOCIATION ENERGY OF THE GeO MOLECULE The dissociation energy of the molécule GeO can be calculated from thermochemical cycles based on the beat of sublimation of GeO from amorphous GeO, from the mixture of glassy or hexagonal Ge02 + crystalline germanium and on the beat of formation of hexagonal GeOz- The values used in completing the cycles are : D^sCOa) = 119-2±0-l 34 ; AZ/^gg »(Ge) = 89-5 + 0-5 ; AJ¥^98,/(Ge02,/!ex) = -132-2+1-27; AHlgs^fiGeOzXfl) = -128-9±l-3; AH°29s.f{GeO,am) = -60-8± 1-4 kcal/mole. The values obtained in this work and from the re-interpreted literature data which are now ail in good agreement are summarized in table 8. The average is D^^siGeO) = 157-5 ±1-9, or £>o(GeO) = 156-2± 1-9 kcal/mole (6-7fe± 0-08 eV). '

TABLE 8.—^DISSOCIATION ENERGY OF THE GeO MOLéCULE

D298(GeO) cyde AH298 réf. (kcal mole~i) (kcal mole~0 iGe02{hex)+iGe{c)^GeO(g) 57-5 ±1-0 157-7 ±1-9 this work iGeO2(^0+iGe(c)^GeO(^) 55-4 ±1-0 158-2 ±2-0 »9

GeO(a/«)^GeO(^) 531 ±1-0 156-8 ±2-0 99 iGe02(Aejc)+iGe(c)->GeO(fl') 57-2 ±1-0 158-0±l-94 re-interpreted

GeO(am)->GeO(g) 51-9±l-0 158-0±l-94 91

iGc02(hex)+iGe(c)->G(iO(g) 56-9 ±10 158-3±l-95 99

iGe02(^/)+iGe(c)->GeO(^) 55-9 ±1-0 157-7 ±2-0 5 99

GeO(aff2)^GeO(^) 53-7 ±1-0 156-2 ±2-0 5 99

Ge02(Aejc)^GeO(i?)+i02 124-3 ±2-4 157-0 ±2-710 99 average Dlgg = 157-5 ±1-9 kcal/mole Dl = 156-2±l-9 = 6-77 ±0-08 eV.

ENTROPY AND STABILITY OF GASEOUS Gqpa AND GejOs The relatively large interval accessible and the ratios of intensity (pressures) made it possible to détermine both the entropy and the stabiUty of gaseous Ge202 and GesOa by a second-law treatment (fig. 2). A least-square calculation gave 1084 THERMOCHEMICAL STUDY OF GERMANIUM OXIDES AT/gjo = 43-0±0-75kcal/moIe and A5 350 = 30-2 ±0-9 cal/mole deg. for the re• action Ge202(^)^2 GeO(^?) and AZ/gso = 85-1 ±2-0 kcal/mole and ASgso = 57-0± 2-4 cal/mole deg. for Ge303(5')->3 GeO(g). The error hmits are standard déviations. An estimate of the heat content by analogy with other tetratomic and hexatomic molécules 28 then gave AH29sidim) = 44-7±3-0 kcal/mole and AHlgsitrim) = 88-5 + 5 kcal/mole, the error limits now being estimated over ail uncertainties. The entropies of Ge202 and Ge303 obtained from the above entropy changes and the entropy of gaseous GeO 2» are SgsoCGezOz) = 94-6+2, 585o(Ge303) = 130-2+4, 52°98(Ge202) = 75-1 + 3 and S298(Ge303) = 99-3 + 5 cal/mole deg. As was noted previously for other group 4B-group 6B dimeric molécules, i' 2 the dimerization energy is comparable to the heat of sublimation of the monomer. The same appears to hold here for the trimer, for which the average GeO—GeO bond, 44-3 + 2-5 is the same as the dimerization energy, 44-7 + 3 kcal/mole, the heat of sublimation of GeO from amorphous GeO being 53-4+1 kcal/mole. In the vapour of the other group 4B-group 6B compounds, Si02+Si,i2 SnS, SnSe, SnTe and PbS,i the concentration of the dimeric molécules is, however, small (1 % or less). In the vapour of GeO or Ge02 + Ge it is sufficiently high that one may hope that investigations, e.g., by électron diffraction or by spectroscopic tech• niques could elucidate their structure and show if the suggestion 1 that they have a closed planar or tetrahedral structure is correct. The same applies to the trimer for which a plausible configuration seems to be a hexagonal ring structure.

The authors acknowledge Prof. P. Goldfinger's interest in this study, and thank Prof. W. L. JoUy for communicating détails of his earlier work. They thank Prof. L. Brewer for drawing their attention to an error in the tabulated thermodynamic functions of gaseous GeO. They are grateful to Mrs. S. Smoes for assistance with the measurements and Mr. A. Pattoret for taking and interpreting the X-ray diffrac• tions. The latter were obtained with equipment made available by Dr. F. Bouillon which is acknowledged here. The germanium was kindly made available by the Union Minière du Haut Katanga.

1 Colin and Drowart, /. Chem. Physics, 1962, 37, 1120 ; /. Phyiic. Chem., 1964, 68,428 ; Tram Faraday Soc, 1964, 60, 673 ; Technical Note, no. 10, Contract AF 61(052)—225, 28 Feb., 1963. 2 Drowart and Colin, Technical Note, no. 15, Contract AF 61(052)—225, 15 July, 1963. 3 Dégrève, Ind. Chim. Belg., 1963, 28, 752. * Bues and von Wartenberg, Z. anorg. Chem., 1951, 266, 281. 5 JoUy and Latimer, /. Amer. Chem. Soc, 1952, 74, 5757. 6 Laubengayer and Morton, /. Amer. Chem. Soc, 1932, 54, 2303. 7 Bills and Cotton, /. Physic. Chem., 1964, 68, 802. 8 (a) Kelley and Christensen, U.S. Bur. Mines, R.I. 5710, 1961. (6) Mah and Adami, U.S. Sur. Mines, R.I. 6034, 1962. 9 Barrow and Rowlinson, Proc. Roy. Soc A, 1954, 224, 374. 10 (a) Davydov, Zhur. Neorg. Khim., 1957, 2, 1460. (6) Shimazaki, Matsumoto and Niwa, Bull. Chem. Soc Jap., 1957, 30, 969. 11 Shchukarev and Semenov, Doklady Akad. Nauk. S.S.S.R., 1958, 120, 1059. 12 Porter, Chupka and Inghram, /. Chem. Physics, 1955, 23, 216. 13 Inghram and Drowart, in Proc Symp. High Température Technology (McGraw-Hill Book Co., New York, 1960), p. 219. i"" Drowart and Honig, /. Chem. Physics, 1956, 25, 581 ; /. Physic Chem., 1957, 61, 980. iSDrowart and Goldfinger, /. Chim. Physique, 1958, 55, 721. 16 Ackerman, Stafford and Drowart, /. Chem. Physics, 1960, 33, 1784. 17 Clausing, Z. Physik, 1930, 66, 471. 18 Field and Franklin, in Electron Impact Phenomena (Académie Press Inc., New York, 1957). 19 Carette and Kervin, Can. J. Physics, 1961, 31, 1300. 20 Gutbier, Z. Naturforsch., 1961, 16a, 268. J. DROWART, F. DéGRèVE, G. VERHAEGEN AND R. COLIN 1085

21 Kane and Reynolds, /. Chem. Physics, 1956, 25, 342. 22 De Maria, Drowart and Inghram, /. Chem. Physics, 1959, 31, 1076. 23 Moore, Nat. Bur. Stand., 1949, cire. no. 467. 24 Colin, Ind. Chim. Belg., 1961, 26, 51. Jeunehomme, Thesis (Université Libre de Bruxelles, 1962). Fite and Brackman, Physic. Rev., 1958, 112, 1141. De Maria, Goldflnger, Piacente and Malaspina, Technical Note, no. 1, Contraet AF 61(052)—699. 25Berkowitz, Tasman and Chupka, /. Chem. Physics, 1962, 30, 2170. 26 Aekerman, Thesis (Université Libre de Bruxelles, 1960). 27 Stanton, Chupka and Inghram, Rev. Sci. Jnstr., 1956, 27, 109. 28 Kelley, U.S. Bur. Mines Bull, no. 584 (1960). Kelley and King, U.S. Bur. Mines Bull, no. 592 (1961) (the entropy of GeO(g) tabulated there is 0-4 cal deg.^i mole~i too high; see Brewer and Rosenblatt, Chem. Rev., to be published). 29 StuU and Sinke, Adv. Chem. Ser., no. 18, 1956. 30 Coughlin, U.S. Bur. Mines Bull, no. 542 (1954). 31 JoUy and Latimer, /. Amer. Chem. Soc, 1952, 74, 5751. 32 Rossini, Wagman, Evans, Levine and JalTe, Nat. Bur. Stand., cire. 500 (1952). 33 Jolly, private communication. 34 Brix and Herzberg, Can. J. Physics, 1954, 32, 110. 35 JANAF, Thermochemical Tables (The Dow Chemical Company, Midland, Michigan).

PRINTED IN GREAT BRITAIN AT THE UNIVERSITY PRESS ABEROEEN 36