ENTHALPY OF FORMATION OF BORON NITRIDE
Fluorine Bomb Calorimetry. XVI.
7
The Enthalpy of
Formation of Boron Nitride’]’
by Stephen S. Wise, John L. Margrave, Harold M. Feder, and Ward N. Hubbard3 Chemicd Engineering Division, Argonne National Laboratory, Argonne, Illinois,and Department of Chemistry, University of Wisconsin, Madison, Wisconsin (Received September 7, 1966)
The energy of combustion of boron nitride in fluorine was measured in a combustion bomb calorimeter. This value was combined with the enthalpy of formation of BF3, determined by a similar method, to give the standard enthalpy of formation, A H f O 2 8 8 . 1 ~ , of hexagonal BN. The value obtained, -59.97 f 0.37 kcal. mole-l, is in excellent agreement with several recent determinations by other methods.
Introduction When this work commenced, the enthalpy of forms tion of boron nitride was of considerable interest because of an apparent discrepancy between the cdorimetric values and the values derived from early vaporization studies. This laboratory undertook to determine the enthalpy of formation of boron nitride by a calorimetric method which differed from those previously used, When boron nitride is burned in excess fluorine, the products of combustion are gaseous BF3and N2,4--B according to the reaction BN(c)
+ 3/&’2(g)
-+BFdg)
+ l/zNz(g)
(1)
This simple reaction is particularly amenable to fluorine bomb calorimetry and its enthalpy may be combined with the enthalpy of formation of BF3 to obtain the enthalpy of formation of BN. The experimental portion of this work was completed in 1961. We delayed publication of the results, however, because of our uncertainty about the correct value for AHf”(BF3). This value has now been redetermined’ to our satisfaction, and the corresponding value for AH? (BN) is herewith reported.
Experimental Section Materials. Boron nitride powder w a ~furnished by the National Carbon Co. Microscopic examination indicated particle sizes of 1 to 10 p . X-Ray powder diffraction patterns of the material corresponded to that reported by Pease8 for hexagonal boron nitride. Chemical analysis showed the following impurities (in
%): 0, 0.40 f 0.08; C, 0.06 i 0.01; Si, 0.033 f 0.003; H, 0.010 A 0.003. Spectrographic analysis indicated approximately 0.008% titanium as the only metallic impurity present. Fluorine was puritied by distillation in a low-temperature ~ t i l l . ~ Mercury 9~~ titration analysis showed the distilled fluorine to contain less than 0.1% impurities; the impurities found by mass spectrometric analysis were oxygen (0.04%), nitrogen (0.01%), and traces of helium and argon. Calorimetric System. The calorimetric system and operational procedures of this investigation were similar to those previously des~ribed.~The reaction vessel (1) Work done under the auspices of the U. S. Atomic Energy Commission. (2) Abstracted from a thesis submitted by 8. 8. Wise to the faculty of the University of Wisconsin in partial fulfillment of the requirements for the Ph.D. degree. (3) To whom inquiries should be directed (Argonne National Laboratory). (4) G. E. Coates, J. Harris, and T. Sutcliffe, J. Chem. SOC.,2762 (1951). (5) 0.Glemser and H. Haeseler, 2. anorg. allgem. Chem., 279, 141 (1956). (6) W. C. Schumb and R. F. O’Malley, Inorg. Cham., 3, 922 (1964). (7) G. K. Johnson, H. M. Feder, and W. N. Hubbard, J . Phy8. C h . ,70, 1 (1966). ( 8 ) R.8. Peaae, Acta Cryst., 5 , 356 (1952). (9) E. Greenberg, J. L. Settle, H. M. Feder, and W. N. Hubbard’ J. Phy8. cham., 65, 1168 (1961). (10) L. Stein,E.Rudzitis, and J. L. Settle, “Purification of Fluorine by Distillation,” Argonne National Laboratory, ANL-6364, 1961. (Available from Office of Technical Services, U. 8. Department of Commerce, Washington, D. C.)
Volume 70, Numbe~1 January lG66
S. S. WISE,J. L. MARGRAVE, H. M. FEDER,AND W. N. HUBBARD
8
was the two-chambered device’l for use with materials that burn spontaneously in fluorine. The volume of the fluorine tank was 0.24 1.; the total volume of the tank and bomb combined was 0.57 1. The calorimetric system was calibrated in the standard manner by combustion in oxygen of benzoic acid (NBS Standard Sample 39g) whose certsed 0.0026 abs. energy of combustion was 26.4338 kjoules g.-l under prescribed conditions. During the calibration, the fluorine tank was evacuated, the valve was kept closed, and oxygen at 30 atm. pressure was contained only within the bomb proper. No nitric acid or carbon monoxide was observed in the combustion products. The mean value for E(ca1or.) the energy equivalent of the calorimetric system, determined by seven combustions was 3348.08 i 0.5 0 joules.) (std. dev.) cal. deg.-l. (1 cal. ~ 4 . 1 8 4 abs. Combustion Technique. Boron nitride burns spontaneously in fluorine. In a glass bomb,12 combustion of a pellet was observed to cause a deep blue glow in the surrounding gas when the fluorine pressure was about 100 torr. The blue glow may be evidence of the formation of excited nitrogen atoms during combustion. At higher pressures, pellets burned white hot without spattering or melting. The following experimental procedure was found satisfactory. Approximately 0.4-g. samples of boron nitride were pelleted in a 0.95-cm. diameter stainless steel die. A pellet was weighed on the thin nickel dish which supported it during combustion. The dish was attached to the bomb head with a small nickel clamp. The bomb was sealed and the tank, containing fluorine at 8 atm. pressure, was attached. The bomb was pumped down to 10-3 torr, 150 torr of BF3 was introduced, and the assembled bomb was placed in the calorimeter. After the usual foreperiod measurements, the tank valve was opened and fluorine expanded from the tank into the bomb to a final pressure of 3.4 atm. After the calorimetric measurements were the Product gases were removed for analysis andIthe vessel was filled with helium and then opened in 8 helium-atmosphere glove box for examination. Between runs, the bomb was washed and dried. was usuPostc”ustion Ezamination* ally 99.9% Complete. The residue, a thin white layer adhering to the nickel dish, was identified as boron nitride by X-r,ay examination. However, the residue was slightly hygroscopic, POSSibb’ because of adsorbed BFa. The amount of residue in each run was determined as folwows: the nickel dish was weighed in the glovebox’ the residue was brushed Off’ and the dish was then reweighed. The weight of the nickel dish
*
The Journal of P h l & d Chmiatry
remained unchanged from run to run; apparently, no appreciable attack by fluorine occurred during the combustions. I n selected experiments the bomb gases were analyzed. One portion was treated with mercury to remove fluorine and the residue was examined mass spectrometrically. Nz, BFa, and small amounts (presumably, from the sample impurities) of 0 2 , SD4,and CF4 were the only species detected. Another portion was condensed at liquid nitrogen temperature, fluorine was pumped off, and the residue was transferred to a cell for infrared spectra1 analysis. No nitrogenfluorine compounds were detected in this sensitive test. Schumb and O’Malleye found only 0.01% NF3 in the product gases from the passage of fluorine over boron nitride.
Results Listed in Table I are the data for seven acceptable runs. The corrections to standard states were applied in the usual manner for fluorine bomb calorimetry.la Item 6 in Table I is the net thermal correction for the energy of expansion of Fzfrom the tank into the bomb and for its subsequent reaction with the bomb surfaces. This correction, assumed to be constant, was determined by blank experiments in which F2 was allowed to expand into the bomb under the same conditions &g in the combustion experiments except that no sample was present. The other entries in the table have been previously explained. For the calculations of item 5, the second virial coefficient of the gaseous mixture and its temperature derivative were estimatedl6 from intermolecular force constant data for Fz,16Nz, l6 and BF,. l7 Auxiliary data used to calculate various numerical quantities are given in Table 11.
Derived ~~h The energy of combustion of the sample in fluorine was -8525.4 cal. g.-I, with an uncertainty (twice the combined standard deviation) of 8.5 tal. g.-l resulting (11) R. L.Nuttall, S. Wise, and W. N. Hubbard, Reo. Sci. Instr., 32, 1402 (1961). (12) R. L. Nuttall, M. A. Frisch, and W. N. Hubbard, ibid., 31, 461 (1900). (13) W. N, Hubbard in “Experimental Thermoche-try,T, vel, 11, H. A. Skinner, Ed., Interscience Publishera, London, 1962. Chapter 6. (14) E. Greenberg, C. A. Natke, and W. N. Hubbard, J. Phya. 698 2089 (1965)* (15) J. 0. Hirschfelder, C. F. Curtiss, and R. B. Bird, “Molecular Theory of Gases and Liquids,” John Wiley and Sons, Inc., New York, N. Y . , 1954. (16) D. m t e , J. H. Hu, m d H. L. Johnston, J. Chem. Phys., 21, 1149 (1953). (17) G. L. Brooks and C. J. G. Raw, Trans. B’“& sot., 54, 972 (1958); w.H. Mears, Allied Chemical COW., unpublished data.
ENTHALPY OF FORMATION OF BORON NITRIDE
9
Table I: Results of Boron Nitride Combustions 1. 2. 3. 4. 5. 6. 7.
m', g. Ato deg. &(calor.)(- Ato), cal. AEcontents, cal. iiEgBI, cal. A.Ebisnk, Cal. AEcO/M(sctmple), cal. g.-1
0.39273 0.99797 -3341.28 -9.06 0.27 4.0
0,39439 1.00263 -3356.89 -9.11 0.27 4.0
0.67533 1 71661 -5747.35 -15.67 0.42 4.0
0.49098 1.25051 -4187.17' -3.23 0.32 4.0
0.48438 1.23344 -4130.01" -3.18 0.32 4.0
0.49723 I.26635 -4240.21" -3.28 0.33 4.0
0.49959 1.27324 -4263.28" -3.29 0.33 4.0
-8520.0
-8523.9
-8527.1
-8526.0
-8524.0
-8525.6
-8531.5
I
Mean AEc'/M(sample) = -8525.4 cal. g.-l Std. dev. of mean 1 . 3 cal. g.-l
" For these runs &(calor.)was changed to 3348.37 cal. deg.-l owing to bomb modifications.
Table III: Enthalpy of Formation of BN at 25"
Table II: Auxiliary Data (25') cpo, cal. deg.-l g.-l C,", cal. deg.-l mole-' p, g. cc.-1 AH!', kcal. mole -l
Ni (0.1061)," BN (0.190)' Fz (5.50),' Nz (4.973),dBFs (10.07)' BN (2.20), Ni (8.907) Be03(amorph) ( -299.74),4 B4C ( -12.7),b HzO ( -68.32),d Si3N4 ( -179.3),d TiN (-80.47),' BFI (-271.65); CF4 ( -221),A S i F d (-385.98),' HF ( -64.8),A TiF4 (-394.19),' HsBOs(c) (-261.47)6
a R. Hultgren, R. L. Orr, P. D. Anderson, and K. K. Kelley, "Selected Values of Thermodynamic Properties of Metals and Alloys," John Wiley and Sons, Inc., New York, N. Y., 1963, p. 198. "JANAF Thermochemical Tables," The Dow Chemical Co., Midland, Mich. W. H. Evans, T. R. Munson, and D. D. Wagman, J. Res. Natl. Bur. Std., 55, 147 (1955). F. D. Rossini, D. D. Wagman, W. H. Evans, S. Levine, and I. Jaffe, "Selected Values of Chemical Thermodynamic Properties," National Bureau of Standards Circular 500, U. S. Government Printing Office, Washington, D. C., 1952. e W. D. Good, M. Miinsson, and J. P. McCullough, Symposium on Thermodynamics and Thermochemistry, Lund, Sweden, 1963. G. L. Humphrey, J. Am. Chem. SOC., 73, 2261 (1951). See ref. 7. W. H. Evans, National Bureau of Standards, Washington, D. C., private communication. S. S. Wise, J. L. Margrave, H. M. Feder, and W. N. Hubbard, J. Phys. Chem., 67, 815 (1963). j E. Greenberg, J. L. Settle, and W. N. Hubbard, ibid., 66, 1345 (1962).
Method
Equilibrium Decomposition pressure 1928 Mass spectrometric, Knudsen effusion 1959 Langmuir free evaporation 1962 1963 Torsion effusion Calorimetric Combustion in 0 2 1954 Direct combination of elements 1959 1961 Combustion in Fz Combustion in NF3 1961 Combustion in Oz-HzO 1962 This work, combustion in Fz 1965
'
'
from the errors of calibration, reproducibility, and blank correction. To convert this value to that for eq. 1 required a correction for impurities in the sample. The H, 0, Si, and Ti impurities were assumed to be present as HzO, B2O3, SiaNd, and TiN, respectively. The corresponding corrections were found to be fairly insensitive to the assumptions made about the states of combination. Consideration of the relevant phase diagrams indicated that the carbon impurity could be present as free carbon, as B4C, or &g a mixture of both; moreover, the correction for carbon was quite
-28.1" -60' -60 -59.8 -59.7 -60.7 -59.Sg -60.1 -59.9 -59.97
f 2' f0.6d
*O.T f 0.7'
*2' f1.5' f 0.37
" R. Lorenz and J. Woolcock, 2. anorg. allgem. Chem., 176, 289 (1928). Values ranging up t o -50 kcal. mole-' have been P. Schissel and W. Williams, calculated from their data. Bull. Am. Phys. SOC.,[2] 4,139 (1959). L. H. Dreger, V. V. Dadape, and J. L. Margrave, J. Phys. Chem., 66, 1556 (1962). D. L. Hildenbrand and W. F. Hall, ibid., 67, 888 (1963). A. S. Dworkin, D. J. Sasmor, and E. R. Van Artsdalen, J. Chem. Phys., 22, 837 (1954). The reported value, -60.7, has been corrected by use of a more recent determination of Mf" (BaOa,amorph) by Good, et al.: see footnote e in Table 11. G. L. Gal'chenko, A. N. K o d o v , B. I. Timofeev, and S. M. Amd. Sci. USER, Chem. Sect., 127, 635 (1959). Skuratov, PTOC. P. Gross, Fulmer Research Institute, Stoke Poges, England, private communication. A C. J. Thompson and G. C. Sinke, reported in "JANAF Thermochemical Tables," The Dow Chemical Co., Midland, Mich., 1964. J. J. Keavney, Stamford Research Laboratories, American Cyanamid Co., Stamford, Conn., private communication. The reported value, -60.6, has been corrected by use of a more recent determination of AHf"(H3BOs,c) by Good, et al.: see footnote e in Table XI.
'
(I
'
sensitive to its distribution between the two states. The B4C and free carbon contents of the specimen were therefore differentiated analytically by a method Volume 70,Number 1 January 1966
S. S. WISE, J. L. MARGRAVE, H. M. FEDER,AND W. N. HUBBARD
10
based on a suggestion18that at 900" (in the absence of a catalyst) free carbon should burn in oxygen, whereas B,C should be essentially noncombustible. A check experiment showed insignificant oxidation of B4C at 900" in oxygen. Free carbon was therefore determined by combustion in oxygen at 900" ; total carbon was determined by combustion in oxygen at 1100" in the presence of a copper catalyst. The results of these analyses were: free carbon, 0.027 f 0.003%; total carbon, 0.06 f 0.01%; carbon as BIC (by difference) 0.033 f 0.01%. The correction for impurities in the sample waa calculated using the enthalpies of formation given in Table 11. The total correction was -3.9 cal. g.-l with an uncertainty, equal to twice the standard deviation, of 8.3 cal. g.-l. Combination of the impurity correction with the energy of combustion of the sample gave a value of -8529.3 f 11.9 cal. g.-l for the energy (and enthalpy) of combustion of hexagonal BN. With the molecular weight of BN taken as 24.818,19 a value of -211.68 f 0.30 kcal. mole-' was obtained for AEc"298.15 = mC"298.15. Combination of the enthalpy of combustion with the enthalpy of formation of BF3 (-271.65 f 0.22 kcal. mole-')' yielded a value of -59.97 f 0.37 kcal. mole-' for the standard enthalpy of formation, mfo298.16, of hexagonal BN from crystalline boron and nitrogen gas.
The Journal of Ph&d
Chemistry
Discussion The value derived for the enthalpy of formation of BN from this work is listed together with the results of nine other determinations in Table 111. Wherever possible, the originally reported values have been corrected, as noted, for subsequent changes in auxiliary data. The stimulus for much of the work done after 1954 was the large discrepancy between the results based on decomposition pressure and the results based on oxygen bomb calorimetry. The range of the nine modern determinations, only 1.0 kcal. mole-', is very gratifying considering that seven distinct methods, three based on 'high-temperature equilibria and four on calorimetry, were used. The massive evidence in Table I11 leaves little doubt that the value reported in this work is very close to the true value.
Acknowledgments. Grateful acknowledgment is extended to B. D. Holt and I. Fox for the performance of special analyses, to Miss R. Terry for checking the data, and to G. K. Johnson for his painstaking preparation of the manuscript. (18) We are indebted to 0. H. Kriege, Westinghouse Electric Corp., for this suggestion.
(19) A. E. Cameron and E. Wichers, J . Am. C h . SOC.,84, 4176 (1962).