3512
G. D. Foss AND D. A. PITT
Calorimetric Studies of Bis (fluoroxy) perfluoromethane’
by G. D. Foss and D. A. Pitt Central Research Laboratories, Minnesota Mining & Manufacturing Company, St. Paul, Minnesota 66101 (Rcceivcd April $6,1968)
Calorimetric measurements of the explosive decomposition of gaseous bis(fluoroxy)perfluoromethane indicate a value of -134.9 3 kcal mol-l for the value of the heat of formation at 298.15”K. The bonding energy of the fluoroxy group in CF,(OF)2 is 133.5 kcal, which is 6 kcal weaker than in CFsOF. Introduction The thermochemistry of the single fluoroxy substituent on carbon was studied by Porter and Cady.2 A value was determined for the enthalpy of the thermal decomposition of CF30F according to eq 1 by the measurement of the equilibrium constants. These
ZCOFzk) + Fz(g)
>3250
(:FaOF(g)
(1)
data have been reanalyzed by Stull and coworkers,a who utilized significant trends in the equilibrium constants to estimate the importance of a simultaneous reaction of CFBOFand COF2 to give perfluorodimethyl peroxide.2 Although this uncertainty could not be brought to account, their third-law analysis provides a value of 32.3 kcal mol-1 for the heat of reaction 1, which minimizes the errors in the equilibrium constants and which results in a value of - 184.0 f 2.5 kcal mol-’ for the heat of formation of CF,OF(g) at 25” by utilizing the selected value4 for the heat of formation of COF2(g), -151.7 A 2 kcal mol-I. Methods for the synthesis of bis(fluoroxy)perfluoromethane have recently been reported from this laboratory5 and elsewhere.6--s The availability of the bissubstituted compound has provided the opportunity for determining the extent to which the fluoroxy group energy is altered by multiple substitution. Experimental Section
Materials. The samples of bis(fluoroxy)perfluoromethane were prepared, according to the techniques discussed by Thompson,6 by the direct fluorination of sodium oxalate. The crude sample was trapped and fractionated at -78”. Purification was effected with a preparative gas-liquid chromatograph having an acidwashed Celite column carrying 3hI inert fluid FC-43. The purity of the sample was monitored with an analytical glpc unit having a 0.5 in. X 8.5 ft column packed in the same manner, which showed 4.8 mol yo CF, for the sample used in run 123 and less than 1 mol u/o contamination in that for runs 135-139. The infrared spectrum of the latter sample was in all respects identical with that reported by T h o m p ~ o n . ~ Storage prior to charging the bomb was in glass The Journal of Physical Chemistry
traps that had been passivated with fluorine, and the sample was maintained at liquid nitrogen temperature. Within a few minutes of the time any sample was charged to the bomb, the purity was rechecked both by analytical gas chromatography and by infrared spectroscopy. This precaution may have been unnecessary; it has been established that a pure sample of CFZ(0F)z is stable in a glass bulb at room temperature for several years.5 Safety Precautions. This compound is potentially hazardous because of its high energy content and oxidizing reactivity. It is recommended that suitable protective equipment be used when working with the material, as explosions of considerable force have occurred.6 Although the toxicology of CF2(OF), has not yet been reported, it should be treated as highly toxic in view of the known toxicity9 of OFz. CalorimetricEquipment. The isothermal combustion calorimeter with a rotating-bomb mechanism is constructed according to the design of the Argonne Na,tional Laboratory. A description of this calorimeter design is given by Good, et aZ.1° Since only gaseous reaction products were involved in this work, the moving-bomb feature of the calorimeter was not used. The bomb was made from grade A nickel with a Teflon main gasket; it had a volume of 0.313 1. The reaction was initiated without auxiliary substances by a (1) This work was carried out with the support of the Advanced Research Projects Agency, administered by the Bureau of Naval Weapons under Contract NOrd-18688. (2) R. S. Porter and G. H. Cady, J . Amer. Chem. Soc., 7 9 , 5628 (1957). ( 3 ) D. R. Stull, et a;., “JANAF Thermochemical Tables,” Clearinghouse for Federal Scientific and Technical Information, Document PB 168370, U. S.Department of Commerce, SprinFfield, Va., 1965. (4) H. von Wartenburg, 2. Anorg. Chem., 258, 356 (1949); W. H. Evans, National Bureau of Standards Report 8504, U. S. Government Printing Office, Washington, D. C., July, 1964, p 164 ff. ( 5 ) P. G. Thompson, J . Amer. Chem. Soc., 89, 1811 (1967). (6) F. A. Hohorst and J. M.Shreeve, ibid.,89, 1809 (1967). (7) R.L. Cauble and G. H. Cady, ibid., 89, 1962 (1967). (8) M. Lustig, A. R. Pitochelli, and J. K. Ruff, ibid., 89, 2841 (1967). (9) D. Lester and W. R. Adams, Amer. Ind. Hyg. Assoc. J., 26, 562 (1965). (10) W. D. Good, D. W. Scott, and G. Waddington, J . Phys. Chem., 60, 1080 (1956).
CALORIMETRIC STUDIES OF BIS(FLUOROXY)PERFLUOROYETHANE platinum wire having a diameter of 0.002 in., which dissipated 0.2 cal when fired with a 12.5-V transformer. The bomb valves were specially designed for use under vacuum, with Rlonel needles and Teflon packing. Prior to charging, the bomb was thoroughly dried in vacuo while heating to 150" and was then exposed to fluorine at room temperature for 12 hr. The calorimeter water was added to the calorimeter can after all electrical connections were made and the can cover and its rubber gasket were attached. From a globular weighing bulb of 3-1. capacity, the water was added through the thermometer port. The weight of water was determined by difference on a balance having a 1-mg sensitivity. The calorimeter temperature was read from a knifetype platinum resistance thermometer by a thermostated and shielded G-2 RIueller bridge. The bridge output voltage, amplified electronically, was fed into a strip chart recorder. By periodically recording the bridge zero, the time at which an exact thermometer resistance obtains was determined.'"12 The corrected resistance rise and temperature rise were determined by a computer program from time-temperature data pairs. A least-squares fit to a straight line was carried out in the fore and after drift periods. The midtime of the main period was calculated by the equal-area method. Calibration. The energy equivalent of the calorimeter was determined by electrical calibration. The e q ~ i p r n e n t 'makes ~ use of paired electronic analog-tofrequency converters and pulse counters for readout. These units indicate heater voltage and the total coulombs passed through the calorimeter heater as the time integral of voltage across a precision 1-ohm series resistor during 3, cycle of arbitrary duration. A source of constant voltage for the heater is assured by electronic regulation. The mean energy equivalent, determined from five electrical calibration experiments, was found to be 3545.4 i 3.3 cal deg-l, for temperature rises of 0.2". The precision and accuracy of the digital electronic technique h.as been studied by comparison with standard benzoic acid, showing the standard deviation of the mean to be ea. 0.01% for both methods, with an agreement of the means within 0.01%.13 Units of Measurement. The results of the calorimetric experiments are expressed in terms of the defined thermochemical calorie, equal to 4.154 J (exactly) and refer to the isothermal process at 25" and to the sample mass in vacuo. The results were computed on the basis of the atomic weights14adopted in 1961. Procedures. The samples were charged at a pressure of ea. 100 torr by expansion from a glass weighing bulb into the passivated and evacuated bomb. After closing the bomb valves, the sample remaining in the lines of the manifold was retrapped into the weighing bulb for gravimetric determination. The bomb and covered calorimeter were then placed into the iso-
3513
thermal jacket, and tempered calorimeter water was introduced from a large weighing flask to minimize evaporation losses. Analytical determinations of the explosion products were carried out within 30 min after the removal of the bomb from the calorimeter. The pressure in the bomb after thermal equilibration was measured to 1 0 . 5 torr. A vacuum manifold of stainless steel tubing with Monel needle valves was used for charging the bomb and for product analysis. The determination of pressures on this manifold was carried out with a Pace Model P3D stainless steel differential transducer. Fluorine in the product gases was determined with a Cary Rfodel 11 ultraviolet spectrophotometer, using a 10.3-cm nickel cell equipped with CaFz windows. Measurements were made at 3100 8 to minimize interference from OF compounds, which are weak ultraviolet absorbers at the Fz maximum15 of 2845 8. This determination was found to be accurate to iO.01 of Fz pressure at a total pressure of 150 torr. The analysis of the remaining product gases was carried out chromatographically with the analytical column described above. The temperature of the column was maintained at -30". The conversion of peak area per cent to mole per cent was accomplished by the relation nl/nz = (A~/Az)~Mz/M where ~ , n is the number of moles, A is the peak area, and M is the molecular weight. This relation16 has been found to be descriptive within an uncertainty of 3% of a variety of fluorinated compounds of one carbon atom, e.g., the compounds reported by Rebertus, et a1.,17and was assumed valid for CF300CFB and CFBOOF also. Identification of these minor constituents was from the published ir spectra2~'*~'@ and fluorine nmr spectra.20 The reaction of the product gases with mercury, detected both manometrically and gravimetrically, provided corroborative information on the bomb product distribution. The increase in weight of the mercury provides a measure of fluorine plus CFSOF, which reacts, according to Porter and Cady,21a22to give COFz and
(11) G. T. Armstrong, P. K. Wong, and L. A. Krieger, Rea. Sci. Instrum., 30, 339 (1959). (12) G. S. Ross and H. D. Dixon, J . Res. Nat. Bur. Stand., C , 64, 271 (1960). (13) G. D. Foss and D. A. Pitt, Rev. Sci. Instrum., in press. (14) A. E. Cameron and E. W'ichers, J . Amer. Chem. Soc., 84, 4175 (1962). (15) R. K. Steunenberg and R. C. Vogel, ibid., 78, 901 (1956). (16) L. F. Herk, private communication. (17) (a) R. L. Rebertus, J. J. McBrady, and J. G. Gagnon, J . Org. Chem., 32, 1944 (1967); (b) R. L. Rebertus and B. W. Nippoldt, ibid., 32, 4044 (1967). (18) A. J. Arvia and P. J. Aymonino, Spectrochim. Acta, 18, 1299 (1962). (19) P. G. Thompson, Fluorine Symposium, Inorganic Division of the American Chemical Society, Ann Arbor, Mich., June 1966. (20) P. G. Thompson, J . Amer. Chem. Soc., 89, 4316 (1967). (21) R. S. Porter and G. H. Cady, ibid., 79, 5625 (1957).
Volume 72, Number 10 October 1968
3514
G. D. Foss AND D. A. PITT
Hg2Fz. Corrections for CF300CF3and any unreacted CF2(0F)2 are necessary. The latter was shown to react c ~ m p l e t e l with y ~ ~ mercury ~~ to form C02. The reaction of CF300CF3with mercury is not known. The mercury method of fluorine analysis generally showed greater scatter than the uv method owing to side reactions and was considered to be less reliable.
Results The sample of CF2(0F)zwas sufficient for six acceptable measurements of the heat of explosion, each associated with a particular distribution of the explosion products. The explosion products of bis(fluoroxy)perfluoromethane are carbonyl fluoride, fluoroxyperfluoromethane, fluorine, and oxygen. Thermal decomposition6 goes predominantly to carbonyl fluoride, fluorine, and oxygen, as described by eq 2. However, an equilibrium, given by eq 1, is obtained between carbonyl fluoride and fluorine and fluoroxytrifluoromethane, and there is evidence of further reaction to form anorganic peroxide2CFBOOCF3andaminor amount of theperoxyfluoride, lo CF300F. Thestoichiometries of these reactions are given in eq 2 and 3. By the comCF2(0F)2 --+ COF2 COFz
+ CFaOF
+ Fz + 0.502
(2)
250-300'
CFaOOCFa
(3)
bination of the above equations, one may write stoichiometric eq 4 and 5 to augment eq 2 in describing the explosion process in the calorimeter bomb.
+ 0.502 0.5CF300CF3 + 0.5Fz + 0.502
CFZ(0F)z +CFaOF CF2(0F)z ----f
(4)
(5)
Measurement of the final pressure in the bomb was used to calculate the partial pressure of COF2. It is apparent from the consideration of reactions 2, 4, and 5 that the pressure of the explosion products will exceed 1.50 times the pressure of the CF2(0F)2 by an amount equal to the partial pressure of carbonyl fluoride, provided that no CF2(0F)2remains, and that the partial pressure of carbonyl fluoride is equal to the partial pressure of fluorine minus the partial pressure of CF300CFa. The product distribution shown in Table I was determined from the final bomb pressure, uv spectrophotometry, and the analysis of the product gases by glpc. The unassigned material had a chromatograph retention time equal to that for the CF2(OF)%. It could not be established whether this represents unreacted starting material or an inert impurity. This could have been sulfur hexafluoride, which is known to have nearly the same retention time and is often a contaminant of commercial fluorine. This small amount of material was considered inert during the bomb process. The thermal measurements on this series of determinations are given in Table I. The correctionsz3 to standard thermodynamic states were simplified by the The Journal of Physical Chemistry
~
Table I : Summary of Thermal Data" rn' Run no.
(compound), g
123 135 136 137 138 139
0,19910 0.49892 0.53788 0.50542 0.43122 0.40401
AEr
EBPP &.to,
(- A t d ,
deg
cal
kcal mol-1
0.01459 0.04617 0.04993 0.04644 0.03909 0.03634
51.40 163.48 176.82 164.47 138.38 128.64
-30.98 -39.32 - 39.45 -39.05 -38.51 -38.21
a Symbols used are those of ref 23, the subscript r referring to the explosion reaction to give the product distribution of Table 111.
Table 11: Summary of Calculated Heats of Formation of Bis(fluoroxy)perfluoromethane a t 298.15"K (in kcal mol-') Run
AH~%B.~J, kcal mol-1
no.
123 135 136 137 138 139 Mean
-134.3 -132.8 -136.0 -135.9 -135.4 -135.2 -134.9
Std dev
0.5
absence of auxiliary materials, water, air, and 30 atm of oxygen, which are usual in combustion experiments. Further, the low charging and final pressures in the bomb reduce the corrections for both the pressure dependence of the internal energy and for the specific heat of the bomb contents to negligible proportions. The enthalpy of formation of bis(fluoroxy)perfluoromethane is shown in Table 11, as computed from the information of Tables I and 111. The estimates for the peroxy compounds are based on the work of Kirshenbaum, Grosse, and A ~ t o in n ~combination ~ with the recent results of King and A r m s t r ~ n g . ~ It~should be noted that an uncertainty of f10 kcal in the assumed energy of the 0-0 bond in these compounds leads to an uncertainty of less than kO.2 kcal mol-' in the heat of formation of CF2(0F)2. Earlier analyses of these calorimetric experiments, based on Porter and Cady's result of 26.9 kcal mol-' for the heat of reaction of CF30F and different estimates for the peroxy compounds, had led to the value of -130.5 kcal mol-l for the heat of (22) P. G. Thompson, private communication (we independently determined quantitatively the stoichiometry of these reactions). (23) W. D. Good and D. W. Scott in "Experimental Thermochemistry,'' Vol. 2,H. A. Skinner, Ed., Interscience Publishers, New York, N. Y.,1962, Chapter 2. (24) A. D. Kirshenbaum, A. V. Grosse, and J. G. Aston, J . A n e r . Chem. Soc., 81, 6398 (1959). (25) R. C. King and G. T. Armstrong, National Bureau of Standards Report 9500,U. S. Government Printing Office, Washington, D. C., Jan 1967,pp 140-226.
CALORIMETRIC STUDIESOF
BIS(FLUOROXY)PERFLUOROMETHANE
3515
Table 111: Distribution (mol/mol of CF2(0F)?) and Heats of Formation of the Reaction Products Total
Run no.
COF2
CFsOF
123 135 136 137 138 139
0,514 0.310 0.275 0 289 0.320 0.340
AHf'zss, kcal mol-'
--151.7 =k 2"
0,0098 0,0205 0,0174 0,0206 0.0146 0,0102
0.416 0.612 0.683 0.665 0.645 0.634
I
' Reference 3.
CFsOOCFs
-184.0 If 2 . 5 "
-358 =k l o b
Discussion The reported enthalpies of formation of the gaseous fluoroxy compounds, in combination with the heats of formation of the relevant atoms as listed in Table IV, Table IV : Heats of Formation and of Atomization of Relevant Gaseous Species (in kcal mol-')
c 0 F OF2
CF4 C'FsOF CE'n(0F)z
.
1
.
0,0148 .
I
.
... ...
... -177 If lob
CF4
Unassigned
gas
0.048
...
2,041
... ...
0.0140 0.0084
1.851
... ... ...
0.0050 0.0058 0.0058
1.794 1.802 1.835 1.855
,220.5 A 2.55
Estimated value.
formation of CF2(0F)2Jas reported by Thompson.6 The uncertainties of the experimental work and of the heats of formation of the reaction products, as given in Table 111, may be combinedJZ6giving for the heat of formation of CF2(0F)2(g) a value of -134.9 A= 3 kcal mol-'.
Compd
CFaOOF
AHf
170.89 59.559 18.86 5.86 -220.5 -184.0 -134.9
Qa
Ref
91.42 466.83 489,89 500.38
3 3 3 25 3 3 This work
were used in the computations of the heats of atomization, Q., of the molecular species. Neglecting bondbond interactions, in view of the few available data, the application of the C-F bond energy in CFI, 116.7 kcal, to fluoroxyperfluoromethane gives a value of 139.8 kcal for the group energy. Similarly, it may be calculated that each fluoroxy group in bis(fluoroxy)perfluoromethane has an energy of 133.5 kcal. The tabulated heat of atomization of OFzprovides an 0-F bond energy of 45.7 kcal. Application of this value to the fluoroxy group energy indicates C-0 bond energy values for the mono- and bis-substituted compounds of 94.1 and 87.8 kcal, respectively.
Acknowledgment. The authors are indebted to Dr. R. L. Talbott for the preparation of the samples of bis(fluoroxy)perfluoromethane used in these studies, to Dr. R. L. Rebertus for consultation on analytical aspects of this work, and to Dr. P. G. Thompson for many helpful discussions. (26) F. D. Rossini in "Experimental Thermochemistry," Vol. 1, F. D. Rossini, Ed., Interscience Publishers, New York, N. Y., 1962, pp 313-320.
Volume 72,Number 10 October 1968