July, 1960
THEMERCURY-MERCURIC CHLORIDE SYSTEM
909
THE MERCURY-MERCURIC CHLORIDE SYSTEM' BY S. J. YOSIMASD S. W. MAYER Atomics International, A Division of North American Aviation, Inc., Canoga Park, California Received February S, 1960
A phase equilibrium study of the mercury-mercuric chloride system has h e n carried out by thermal analysis and by the visual method. The salt-rich eutectic composition is 4.9 mole yo Hg and occurs at 2 i 3 ' . The syntectic line at 525" extends from 48 to 94 mole % Hg. The solubility of HgC12 in Hg increases to 6.8 mole yoa t 560 . Freezing point depression measurements for HgCI9 suggest that Hg dissolves either as atoms or as Hg2C12 molecules formed by reaction of Hg with HgC12. A thermodynamic analysis of the liquid-solid equilibrium curve between the salt-rich eutectic and the base of the miscibility gap suggests that solution of Hg as atoms is not likely in this region.
Introduction Although there have been many investigations of the interaction of mercury with HgCl2 and of the properties of the reaction product Hg&lz, relatively little research has been carried out to determine the phase diagram of the Hg-HgC1, system. The freezing point depression of HgC1, by calomel, up to 1.2 mole ye,was determined by Beckmann.2 Ruff and Schneider,3 in a series of exploratory esperiments, determined approximate melting temperatures of Hg2C12-HgC12mixtures equivalent in composition to 25-43 mole ye mercury dissolved in HgC12. Smith and M e n z i e ~in . ~their studies of the constitution of caloniel vapor, measured the solubility of Hg,C12 in mercury from 250 to 400". 111 order to obtain a more extensive phaye diagram of this system, a series of phase equilibria mcacjurements therefore was carried out. Experimental
sure adequate surface contact of the two phases. The sample was heated until the second condensed phase completely disappeared. The solution process was found in all cases to be reversible with temperature.
For solithilitv dctcrminations by the visual method, t,he mixture of xtlt ~ n c lmrtal was placed in a thick-w&d i p t r t z tube :tiit1 sc.zled under vacuum. The tube then W:IS fastened to a rii~~kcl rod and suspended in a >farshall fiirnac~, Pquipped with two siqht windows opposite onc anothrr The tube was illitmin,rted from the r e x nncl from a third window at one side. A chroxnel-alumel thermocouplc fastened to the tube was employed for measurement of temperature. The f u r n x e was mounted on a metal base whirh was provided with a rocking mechanism. In this way the entire systrni coiild be inverted frequently to en-
Discussion Thc HgC12 liquidus and the liquid -solid equilibrium curve between the salt-rich eutectic and the base of the miscibility gap were examined in order to see whether information on the mechanism of solution of Hg in HgCl, could be obtained. The mechanisms which can be considered for the solution of Hg in HgCl, fall into two classes. The
Results Figure 1 shows the phase diagram of the Hg-HgCl2 system under its own pressure. Mercuric chloride was found to freeze a t 279.5" compared with the literature value of 277°.5 The salt-rich eutectic temperature was 273" and the eutectic composition was 4.9 mole yo Hg. The curve between the saltrich eutectic and the base of the miscibility gap is in sharp disagreement with that of Ruff and Schneider.3 However, in their admittedly exploratory work, the liquidus temperatures were determined by heating and were defined as temperatures a t 11-hich mixtures formed droplets. These temperatures are therefore likely to be considerably lower than those reported here. The syntectic temperature, or that temperature a t which Hg2C12forms A . Materials.-Reagent grade HgClz and Hg:PC12, dried two liquid phases, was found to be 525", in good under vacuum a t 110' for 24 hours, and triple-di~tillerl agreement with Ruff and Schneider's value of mercury were used. B. Apparatus and Procedure.-Two techniques were used 532'. The fact that the syntectic line extends to in this work, thermil analysis and the visual method. For compositions less than 50 mole % Hg suggests that freezing point determinations by thermal analysis, the Pome disproportionation of HgZC12 takes place on salts were contained in a s e a l d 18 mm. Pyrex or T'ycor liquefying a t the syntectic temperature. tube which had a thin-walled thermoronple well sealed into Visual observations of the salt-rich region the bottom of the titbe. Temperatures were measured with a ralibrated rhromel-alumel thcrmorouple, using a showed an interesting behavior. As mercury or Rubicon €3 potentiomvter The rold junction of the Hg,Cl, is added in increasing concentrations to thermocouple was inaintnincd in an oil-filled tube, placed in I1gClz which is almost colorless, the solution bea distilled water-ice-bath. The sample n-as heated in a comes yellow, red and eventually black. As a furnace which was automatically controlled by a regnlatorgiven solution containing Hg or HgZC12 is heated pyrometer operating through a variable transformer. Because thP high vapor pressures of HgClr and Hg above the same darkening in color, \\hich is reversible, 390' could lead to explosions of the tubes containing thc takes place.6 samples, a stninlws stccl pressure vessel was used to holtl The solubility of the salt in the metal rises from the quartz titbrs for thermal arrest measitrements on Ramples in the 15.9-85 0 mole % H g range. In the pres- 0.2 mole % HgCl, (or Hg?C12)at 280" to 6.8 mole % sure vessel, e\ternal pressures as high as 50 atmosphercs of HgC12 (7.13 mole % HgzClz) a t 555". The solunitrogen werr applied to the sample tubes. Thermal bilities measured by Smith and Rlcnzies4 are in arrest mcn6iirc.ments were not made above GOO" sinre th2t was considerd to he the safety limit of the pressure vessel. good agreement with these values.
(1) This work was siippoited by the Rtzsearch
Division of the
Atomic Energy Commission. ( 2 ) E. Beckmann, Z. nnorg. Ckem.. 55, 175 (1007). (3) 0. Ruff and R. Schneider, Z. anorg. allgem. Chem., 170, 42 (1928). (4) .4. Smith and A. Jlenaies, J . A m . Ciiem. Soc., 32, 1561 (1010).
( 5 ) L. Brewer, ' The Chemistry and Metallurgy of Miscellaneous Materials-Thermodynamics," L. L. Quill, Ed., N N E S IV-19B, hloGran-Hi11 Book Co., Neu York, N. Y..1950. (6) For example, a n HgCI? solution containing 10% Hg was yellow a t 415', orange at 46j0,red a t 610° and dark transparent red a t BAS0 wliile a more concentrated solution, 25 mole % Hn, was very dark red and Laiely transparent at 415' and black a t 4G7".
S.
910
J. YOSIMAND S. W. MAYER
600 L , + L*
,
550 500 -
0' E 3
450
-
c Q
LL
4w-
5c 350
-
m-
-
-
~
ng2CI2+HgCIZ
250 IO
20
40
30
50
60
70
80
90
f
100
MOLE PERCENT Hg.
Fig. 1.-The mercury-mercuric chloride system. Circles denote the results obtained by thermal analyais. Data obtained by the visual technique are represented by the triangles. Smith and bfenries' results' are denoted by the squares, and the cross-hatched line summarizes the work of Ruff and Schneider.3
first is solution as Hg atoms, dimers or higher polymers and the second is solution by reaction of Hg with HgC12 to form the lower-valent compound, mercurous chloride. (The fact that solid HgzCI2is stable with respect to its disproportionation products in the pure state does not prove its existence in the liquid solution). In the case of the HgClz liquidus the cryoscopic number, the apparent number of particles formed in molten HgClz per molecule of solute, was calculated from the Raoult-van't Hoff equation.' The calorimetric value of the heat of fusion of HgCI,, 4640 + 50 cal./mole,8 was used for this calculation. The cryoscopic number was found to he 1.17 0.03 (Table I). The discrepancy between this value and unity may be due to deviations of the solvent from idealit,yor to the fact that two or more processes of soluti(in are taking place. The cryoscopic number of about unity indicates that Hg
Val. 64
mechanisms of solution which yield a cryoscopic number of unity. These are solution of mercury as atoms or solution by reaction to form HgzC4. If the latter mechanism takes place, the HgzCla is presumably un-ionized since the addition of small amounts of Hg2C1, to HgCI, does not increase the electrical conductivity of the latter.9 Molten HgClz has the relatively low specific of 8 X lo-' ohms-' cm.-l which suggests that its chloride ion concentration is correspondingly low. It can be seen from Table I that mercurous chloride solute also has a cryoscopic number of about unity. As in the case of Hg, solution of mercurous chloride as the un-ionized HgzCI2or by dlsproportiouation to form Hg plus HgCI, are mechanisms which are consistent with a cryoscopic number of one. The above evidence plus the visual observations of similar color changes suggest that the resulting solutions formed by dissolving Hg or HgzC12are identical, but the data of the HgCL liquidus curve do not distinguish whether these solutions contain Hg atoms or Hg2CI, molecules. The liquid-olid equilibrium curve between the salt-rich eutectic and the base of the miscibility gap (547 mole % Hg) cannot be examined on the basis of freezing point depressions since the heat of fusion of Hg2C12is not known. However, the possibility that this curve can be interpreted in terms of a molten solution containing only Hg and HgClz can be examined. If one assumes that solid HgzCl! is in equilibrium only with its disproportlonatlon products, i.e.
+
-;f Hg(1) HgC1d1) (1) then the variation of the Hg concentration with temperature is given by HgClAs)
*
TmLE FREEZING
PolNT I)lOTRESsl0NS
I IN
MoIIrEN
b~UllCUllll:
CHI.URIDE sol,llc. Solu1e
Hg Hg Hg Hg HP Hi HgsCIa Hg& HK&L H&lz Hg,C!lr
mole
%
0.69 1.09 2.15 3.16 4.27 4.85 0.87 1.37 2.50 3.41 4.76
CrY"SC",,iC
A T ! . "C.
1.05 1.73 3.31 5.04 6.34 7.04 1.32 2.14 3.90 4.99 6.78
""_
1.16 1.21 1.17
1.22 1.13 1.11 1.16 1.19 1.19 1.12 1 .09
does not dissolve in molten HgClz predorninant,lgas a metal polymer or as HgCl. There are two (7) S. W.brayer, S. . I . Yorim and L. E. Topol, THIB lounn~b.64, 238 (1960). (8) L. E. Topol and 1.D. Ransom, ibid., in prear.
where NTlcis thc mole fraction of Hg, AHo is the st,andard heat of reaction 1, AII is thc relative partial molal heat. of solution and y is the activity, coefficient,the standard statc being the pure liquid.1L (9) 11. Foate and N. Marlin, Am. Chcm. J . . 11. 451 (1908). (10) W. Klemm and W. Riltr. E . anorb. ollgcm. Chamie. 161, 226 ( I ma). (11) The derivation of oqualiou 2 is omslant IC for reaction 1 i. given b y
BS
K = awmot.
follows: the equilibrium
(4
where "a" is the activity. the sthndard alate8 bcing tho uure liquida. Then
where the aubacript "sat" rcfers l o the a~liiralioncurve. Sinae. at the d u r a t i o n CIITVO. holh 1om"wnliir~and eomposilioo m e varying. it is necessary l o evaluate the effect of each on the aalivilv oaefficiant. apply in^ R treatment aimilar l o that of Williamson (Trona. Faradav Sor.. 10, 421 (1944)). i t can he shown that
July, 1960
THERXODYNAMIC PROP~XTIES OF THORIUM SULFATE
To illustrate the magnitude of the terms of equution 2, two temperature regions will be considered, one near the syntectic temperature at a composition of 38 mole yoHg where d log NH,/d(l/T) = - 1600 and one near the eutectic region at a composition of 7 mole yo Hg where the slope = -1500. Substituting these values and the value 13,0005for A H o of reaction 2 into equation 2, one obtains for the high temperature arid low temperature, respectively
I n the usual biliary y s t e m in which a miscibility gap exists, the A€$ terms and therefore their ~ sum are positive. Also the term ( b In y l / b N , ) is negative, ie., at small values of K,, 7 , is greater than unity due to positive deviations from ideality while, a t values of S,close to unity, y1 approaches one. Between these two limits, y1 decreases continuously with increasing concentration in the single phase region and is of course constant in the two
(d) Applying the Gibbs-Duhem relation, one then obtains equations 2.
911
liquid phase region. Thus in the uwal binary system in which a miscibility gap exists the difference between the term ( b In Y i / b N i ) T and the term consisting of the partial molal heats of solution would be negative. However the initial assumption that HgeClz disproportionates on dissolving in HgCI, leads to positive values for this difference-a fact which suggests that the initial assumption is incorrect. While the activity coefficients aiid the relatiye partial molal heats of solution of the HgCl,-Hg*Cl, system are not known, it is not anticipated that such apparent inconsistencies arise if this model is assumed for the solutions discussed above. The fact that the melt obtained when Hg2C12is heated to 580" is diamagnetic12 is consistent with this mechanism. As stated earlier the heat of fusioii of Hg2Cl2 is not known. However, if one assumethat Hg2Cl2rather than Hg exists in the salt melt, and calculates the heat of fusion from tht. slope of :i log ~ V H us. ~ ~1/T C plot ~ ~ near the miscibility gap, a value of 11 kcal./niole ( A S fus = 14 e.11.) is obtained. Such a value is not unreasonable. Tht.rrfore, the above arguments suggest that Hg di+ solves in HgClz as HgzC12 rather than as Hg atom. Acknowledgment.-The author. are grateful to Dr. D. E. McKenzie and to Professor H . Flood for many helpful discussions. 112) J. Farquharson and E Heyman, T r a m . Faradaq Soe., 31, ( 2 ) 1004 (1935).
HIGH-TEMPERATVRE FREE ENERGY, ENTROPY, EKTHA1,PY ,iNI) HEI'l' CAPACITY OF THORIUJil SULFATE1 BY S. IT. h f A 4 Y E R , B. B. OWENS,T. H. RUTHERFORD AND R. B. SERRISS Research Department, A-ltoniicsInternational, -4 Diozsion o j .Vorth .4me~ican duiatioii, Canoga P m k , ('aliiornia Receited I'ebruary 18, 1960
Decomposition pressures of Th(s04h have been measured from 908 to 1057°K. The results show that the tl~~c.oiiipo~itioii Oz(g). The heat raparity of Th(SO,)? has been deternmiiwi from 623 reaction is: Th(S04)2(~)= ThOz(s) -t 2SO2(g) to 897°K. by a drop calorimeter technique and the results are given bv: C, = 25.0 55.2 x l o - 3 ~cal. molr-1 deg.-' Least squares treatments of the decomposition pressure and heat capacity data have been used to obtain equations for thc thermodynamic functions of the decomposition reaction Equations have also been calculated for thc xhwlritr en t rcqv of Th(SOa)nand for the standard free energy and enthalpy of formation of Th(SO& from its elements.
+
Introduction A series of measurements of the decomposition pressures of Th(S032 has been made as part of a program for determining the high-temperature thermodynamic functions for thorium and uranium compounds. Heat capacities a t elevated temperatures have also been determined in thiq study to obtain free energy, enthalpy and entropy equations for Th(SO& applicable from 298 to 1057'K. Th(SO& is of particular interest because of its possible use as the fertile material in reactor fuels based on a molten alkali metaphosphate-sulfate system.2 The only previous reported thermochemical values for this compound are a heat of (1) This research was performed under t h e auspices of the t- S. 4tomic Energy Commission ( 2 ) S. W. Rfayer, R. S. Ginell and D. IC AlLIienzio, Tta?is. A m . .\'7cclaar Soc , 2, 196 (1059).
+
forrnat'ion3-j a t 298'K. and a heat c.apncityfi.7at 298'K. S o thermochemical values haye hitherto been reported for Th(S04)z at elewt et1 tempmitures. Experimental Materials.-Thorium sulfate, prepared I)?. I hr. 5. 11'. Shattuck Chemical Company, mas dried in :t vwiiuiii for nine days; the temperature of the s:mplr> w'its in by fifty degrees each day until it had been dried at 475' for 24 hours. Chemical analysis of the product showed that thc wat,er content was lese than 0 content corresponded to 99.9 powder X-ray diffraction patt lines, with no evidence of a hydr (3) F. D. Rossini, et at., Circular 500, National Bureau of Standards, 1952. (4) J. J. Katz and G. T. Seaborp, "The Chemistry of the Actinide Elements," John Wiley and SOLIS.Inc., Xew York. N. \-., 1057. ( 5 ) G. Beck, 2. anorg. Chem., 174, 31 (1928). ( 6 ) I,..'k Nilson and 0. Petterson, Ber., 13, 1459 (1880). (7) I