DISSOCIATION PRESSURE AND STABILITY OF BERYLLIUM CARBIDE
April, 1959
1a.xation time at 20” of the metal-free porphyrazine molecule in benzene solution, which agrees with the observed value 302 X 1O-I2 as well as or better than any of the calculated values in Table VI. Since the volume of the iron porphyra-
587
z h e complex is almost the same as that of the metalfree molecule, the relaxation time calculated for this molecule by the Debye equation is considerably too small, but not as much too emall as three of the four different calculated values in Table VI.
DISSOCIATION PRESSURE AND STABILITY OF BERYLLIUM CARBIDE’ BY B. D. POLLOCK Contributionfrom Atomics International, A Division of North American Aviation, Inc., Canoga Park, California Receiued September 1 1 , 1968
+
Equilibrium pressures for the reaction l/zBezC(s) = Be(g) l/&(s) were measured in the temperature range 14301669 K. by the Knudsen technique. The dissociation pressure in this tem erature range is given by the equation log P(atm.) = 7.026 i 0.347 - (19,720 i 637)/T. The heat and free energy orfornation of beryllium carbide were derived from the above equation in combination with the literature vapor pressure data for solid beryllium.
Introduction Beryllium carbide is of potential interest in atomic power applications because of its nuclear properties; however, experimental thermodynamic data concerning it are meager. Krikorian2 estimates heat, free energy and entropy of formation of Be& at 298°K. to be - 13.0 f 5 kcal., - 12.4 I 5 kcal., and -2.0 =t1.0 entropy units, respectively, and Quirk3estimates -7.83 kcal. for the freeenergy of formation at 2400°K. However, the latter quotes, for various temperatures, estimated vapor or dissociation pressures of beryllium carbide4 which are equal to or greater than those of pure liquid beryllium. These pressures are inconsistent with a negative free energy of formation for a compound having no stable gaseous carbide species.’ The absence of such species and the observation that beryllium carbide loses beryllium preferentially at elevated temperatures with deposition of graphite3 indicate that the carbide dissociates according to the reaction 1,616
l/~BetC(s)= Be(g)
+ l/zC(s)
(1)
Thus a determination of the dissociation pressure as a function of temperature for this compound would yield results which could be combined with data on the vapor pressure of beryllium metal from the literature to obtain the free energy and heat of formation of the solid compound. Experimental Dissociation pressures were measured by the Knudsen method and the experiments were performed in the vacuum induction furnace illustrated in Fig. 1. The power source was a, 6 kilowatt arc-type high frequency converter. The vacuum system was capable of maintaining residual gas pressures in the range 10-’-10-6 mm. during the major part of each run following the initial warm-up. Knudsen cells and orifice plates were turned from National Carbon Company AT2 grade graphite. Temperatures were read with a
disap earing-filament type optical pyrometer by sighting into Elack-body holes in the bottom of the Ihudsen cell, through a prism and optical window. The instrument, made by the Pyrometer Instrument Company of Bergenfield, New Jersey, was calibrated against a secondary standard which consisted of a tungsten ribbon lamp for which current versus temperature data had been obtained a t the National Bureau of Standards. The window and prism correction was checked periodically. These corrections did not change significantly during the investigation. The over-all uncertainty in temperature measurements was estimated to be about 10”. A layer of beryllium carbide was formed on the interior of the effusion cell prior to a group of experiments by heating a few pellets of metallic beryllium in the apparatus t o about 1400” for a few minutes and then removing the excess metal. The carbide formed in this manner was hard, dense and adherent to the graphite. Spectroscopic examination showed only trace amounts of iron and silicon and X-ray diffraction showed Be& and no detectable amount of BeO. The interior of the cell was inspected visually after each run to ensure the presence of a large evaporating surface of beryllium carbide throughout each experiment. The surface area of the carbide layer was estimated to be at least 100 times the orifice area in all cases. Vapor effusing from the cell was condensed on the interior of the quartz vacuum cell and no detectable deposit was found on the glass below the level of the top of the graphite cell. This observation showed that no error occurred by virtue of diffusion of beryllium through the graphite. It was found that the deposit adhered to the glass if it were less than about 3 mg., otherwise there was a tendency to flake off. Accordingly, variables were chosen so as to obtain beryllium losses of about 1 mg. After the run, the deposit was dissolved easily in a €ew cc. of dilute HC1 and the amount of beryllium determined by a spectrophotometric methods with an estimated accuracy of 2-3%.
Results and Discussion The vaporization data are summarized in Table I. The pressure of beryllium was calculated by use of the equation P(atm) = 0.02254(m/at)(K)-1(T/M)’/z
(2)
where (rnlat) is the rate of effusion in g./cm.2-second, K is the Clausing orifice correction, and T and M are the temperature in degrees Kelvin and the atomic weight of beryllium, respectively. Results of these calculations are listed in the last column of Table I and also shown in Fig. 2. A least squares treatment of the data yielded the equation
( 1 ) This work was carried out as part of Contract AT-11-1-GEN-8 with the United States Atomic Energy Commission. (2) 0. H. Krikorian, “High Temperature Studies; Part 11, Thermodynamic Properties of the Carbides” UCRL 2888, Apr. 1955, p. 82. (3) J. F. Quirk, “Beryllium Carbide,” A E C - T I S Reactor Handbook, S, Section 1 , Ch. 1.5 (1955). (4) Jacob Kielland and Leif Tronstad, “Det Kongelige Norske Videnskabers Selskab,” Vol. VIII, N o . 42. 1935, pp. 147, 150. (5) R. B. Holden, R.Speiser and H. L.Johnston, J . A m . Chem. SOC., log P(atm) = (7.026 i 0.347) (19,720 f 537)/T (3) TO, 3897 (1948). (6) E . A. Gulbransen and K. F. Andrew, J . Electrochem. Sac., 97, the slope of which corresponds to a mean enthalpy 383 (1950). (7) W. A. Chupka, J. Berowitz, C. F. Giese and M. G. Inghram, (8) M. E. Smith, “A Spectrophotometric Method for Determining T ~ r JOURNAL, s 6a, 611 (1958). Small Amounts of Beryllium in Uranium” LA-1585, Auguet, 1953.
-
B. D. POLLOCK
n
Vol. 63 TABLB I Wt. loss
Expt. no.
Temp.,
OK.
Time, aec.
2 1 4 3 13 11 16 12 7 6 8 14 10 9
1669 1659 1047 1643 1628 1590 1580 1578 1536 1515 1446 1442 1424 1430
4680 4260 4500 6120 4980 6720 9660 4200 10,800 12,600 27,600 26,900 25,000 23,450
Be, mg.
Orifice area. om.*
C!ausIng
factor
P (atm.)
1.60 0.0085 0.69 1.79 X 10-6 1.50 .0085 .69 1 . 8 3 X lo-& 0.94 .0085 .69 1.08 X 10-6 1.00 .0085 .69 8.5 x 10-6 0.83 ,0085 .69 8 . 6 X 0.50 .0085 .69 3.80 x 10-0 2.80 ,0324 .81 3 . 3 X 10-6 0.45 .0270 .65 2.18 X .45 ,0085 .69 2.08 X 10-0 .95 ,0270 .65 1.25 X 10-6 .29 ,0270 -65 1.72 X 10-7 .33 .0270 .65 2.00X 10.14 .0085 .69 2.72 X .09 .0085 .69 1.86 X lo-'
and condensed phases must be known and the existI ence of equilibrium between the solid and vapor during the experiment must be shown. Chupka, Berkowitz, Giese and Inghram showed in a mas8 1 spectrometric study that Be(g) is the only imporLHAGNET OPERATED SHUTTER tant gaseous species above Be2C a t approximately 1900°K.' This was confirmed by an experiment performed a t 1580°K. in this Laboratory in which the gross weight loss of a Knudsen cell was found to be equal to the amount of beryllium in the sublimate. Solid solubility in beryllium carbide is assumed to be negligible. The existence of equilibFig. 1. rium is shown by the non-dependence of measured pressures on orifice area. Thus, equation 3 may be used for thermodynamic calculations. The standard free energy of formation of beryllium carbide may be obtained by combining the data for reaction 1 with data for the vapor pressure of beryllium by the usual thermodynamic treatment J TO VACUUM PUMP
+ +
2Br(g) C(s) = B%C(s) AFO = 2(4.576T) log P (4) 2Be(s) = 2Be(g) AFO -- 2(4.576T) log Po (5) C(s) = BepC(s) AFO = 2Be(s) 9.152T(log P log PO) (6)
-
The vapor pressure of solid beryllium is6 log Po(atm.) = 6.186
+ 1.454 X 10-4 T - 16 734T(*SO) (7)
If equations 3 and 7 are substituted in (6) one obtains for the free energy of formation of Be2C AFo(f) = (7.688 i 3.176)T 1.331 X lo-* T* - 27,328 ( f 5 0 0 )
(8)
The heat of formation may be obtained from the expression
-
b(AFo/T) = dT
--- AHa
T2
which gives I
6.0
6.2
I
+
6.4
Fig. 2.-Dissociation
I
I
I
6.6
6.8
7.0
x 104.
pressure of Be&.
AHo(f) = 1.331 X 10-'Ta
- 27,328 (h5000)
The entropy of formation may be obtained from the expression AH0 = AFO
+ TAP
change of 90.2 f 2.5 kcal. for reaction 1 in the temAt 1500"K., the approximate mid-point of the experature range 1430-1669°K. For the data to be uaed for calculation of ther- perimental temperature range, these equations give modynamic quantities, the identity of the gaseous -18.8 kcal. for AFO(f), -24.3 kcal. for AHo(f) and
1
April, 1959
HEATS OF MIXINGIN
THE
SYSTEMCCL~-CYCLOHEXANE-BENZENE
-3.7 e.u. for ASo(f). For reactions in which react,ants and products are solids entropy changes are usually small, Le., 1-2 e.u. per gram atom of constituents. Thus the value found above indicates that the magnitude and temperature dependence of the pressures measured in this investigation are consistent. To obtain values of A F O ~ Q and ~ AH02~8 in the ab-
589
sence of high temperature heat capacity data it is assumed that ACP for reaction 6 is zero and the approximate expression AF(T)
AH298
- T(A8298)
will be used together with Krikorian's estimate of ASo2Q8(f). These assumptions lead to a value of -21.8 f 5.0 kcal. for AHOZQ8, and -21.2 f 5.0 for AF'zQ~.
HEATS OF MIXING IN THE SYSTEM CARBON TETRACHLORIDE-CY CLOHEXANE-BENZENE B Y J. REXGOATES, RALPHJ. SULLIVAN AND J. BEVANOTT contribulion from the Department of Chemistry, Brigham Young University, PYOUO, Utah Received September l i , 1968
Heats of mixing were measured calorimetrically for the three binary solutions that can be formed from carbon tetrachloride, cyclohexane and benzene at several temperatures in the range of 10-40". The heats of mixing in the two systems containing benzene varied appreciably with temperature, (dAH?/bT), a t the equal mole fraction composition being 0.33 cal. mole-' deg.-l for CC14-C6H6,and -2.00 cal. mole-' deg.-l for C d k C & . The heat of mixing of CCl4-Ct& was independent of temperature over the 25" range investigated. Analytical expressions for the heats of mixing as a function of temperature and composition were derived. These equations were combined with free energy data from the literature to obtain free energy expressions from which all the thermodynamic roperties at constant pressure can be calculated. Heats of mixing at 25" are also reported for the two ternary systems 8Cl~-C&-CeH, and CClrCHCls-CH?Cl2. I n the first system a ternary term that accounts for 3 4 % of the total heat of rmxing is necessary to adequate1 describe the data. The heats of mixing in the second system were found to be within 1% of the values redicted from d e properties of the binaries. Freezing point measurements of the binar solutions of CCl,, Ca, and &He were made, from which partial phase diagrams could be constructed. The CCl&6$ diagram showed the existence of a 1 :1 addition compound, with a heat of dissociation of approximately 3.5 kcal./mole. It is suggested that r-bondmg between the aromatic ring and the empty 3-d level of chlorine ma be responsible for this complex. Replacing benzene with p-xylene to increase the electron density of the ring increased d e heat of dissociation of the resulting complex to a value of approximately 6.3 kcal./mole. On the other hand, either decreasing the electron density of the ring by using nitrobenzene, or decreasing the electronegativity of the chlorine atoms by substituting CHCll for CCla prevented complex formation.
Some years ago Scatchmd, Wood and Mochell made a careful study of the vapor-liquid equilibria of the three binary systems that can be formed from carbon tetrachloride, cyclohexane and benzene. From these data they calculated the thermodynamic properties at constant pressure and expressed this information analytically as fuiictions of both composition and temperature. Later Wood, e l al.,2 measured thc volumcs of mixing of those systems over a wide temperature rmge, making possible the calculation of the therinodyriamic properties at constant volume. Very extensive use of the information obtained in these two studies has been made in the evaluation of theories of solutions. We felt that additional data on the heats of mixing, especially the effect of temperature on the heat of mixing, were needed; and have measured calorimetrically the heats of mixing for these systems in the temperature range of 10-40'. The heats of mixing in the systems containing benzene were found to vary appreciably with temperature, while CC14-CeH12showed no detectable change over the 25' range that was investigated. Solid-liquid equilibria also were investigated. Our original intention was to calculate the free energies of mixing from these data. Formation of solid solutions, however, precluded the use of the (1) (a) G . Scatchard, 8. E. Wood and J. M. Mochel, THISJOURNAL, 43, 119 11939); (b) J . A m . Chem. Soc., 61, 3206 (1939); (a) 62, 712
freezing point measurements for this purpose. This part of the study did, however, prove to be rewarding, for the phase diagrams show the formation of a compound in the system CClrCaHa. The phase diagram evidence of compound formation, together with the effect of temperature on the heat of mixing, shows that this system is considerably less ideal than it was formerly thought to be.
Experimental Reagent grade chemicals were purified by both fractional crystallization and distillation and then stored under an atmosphere of dry nitrogen gas. Indices of refraction of the final products agreed to the fourth decimal place with those given in the Dow Chemical tabless; and the mole per cent. impurities, calculated from freezing points, were less than
0.02%.
The calorimeter consisted of two 15-ml. stainless steel compartments separated by an aluminum or tin foil disc. Mixing was accomplished by causing a steel plunger inside one of the compartments to pierce the metal foil. The calorimeter is m constructed that no vagor phase needs to be present; however, by urposely leavlng a small approximately 2% vapor spacey changes in pressure broug t about by volume changes in mixing are minimized, and the effect of pressure on the heat of mixing becomes insignificant. The heat effects produced by change in composition of the vapor during mixing are calculated to be less than 0.1 cal./ mole. Further details have been published previously. The over-all uncertainty of these calorimetric measurements is calculated to be less than either one calorie, or one per cent., whichever is larger. The freezing point data were obtained from cooling curves. A calibrated 10 junction thermocouple connected to a 10
6
(1940).
(2) (a) S. E. Wood and J. P. Bruaie, ibid., 65, I891 (1943); (b) S. E. Wood and A. E. Austin, ibid., 67, 480 (1945); ( 0 ) S. E. Wood and J. A. Gray, 111,ibid., 74,3729 (1952); (d) 74,3733 (1952).
(3) R. R. Dreisbaoh, "Physical Properties of Chemical Substanccs,"
Dow Chemical CO., Midland, Michigan, 1952. (1) J. R. Gostes and R. J. Sullivan, THISJOURNAL, 62, 188 (1958)
