FeaC (described else2 . Hagg carbide and cementite have narrow- reaction Fe2C-IIagg+ Ft. ranges of composition and apparently do not where) starts below 300“ 4. The “unstable” or “labile” carbide of Kolform intermediate metastable solid solutions. 3. The reaction 3FezC-Hagg -+ 2Fe8C+C be- bel and co-workers is simply Hagg carbide. comes appreciable only a t about 450°, while the RRUCETOS, PA RECEIVED MARCH23, 1950 t
1 CONTRIBUTION No 19
PROM THE
THERMODYNAMICS LABUKA1 0 ~ 1 , I’FTKULEUM MINES!
EXPERIMENT STATION,BUREAT:CF
Spiropentane: Heat Capacity, Heats of Fusion and Vaporization, Vapor Pressure, Entropy and Thermodynamic Functions’ BY 1). CY. S c o n , H. L. FINKE,W. N. HUBBARD, J. P. MCCULLOUGH, M. E. GROSS,K. U. WILLIAMSON, GUYWADDINGTON AND H. M. HUFF MAN^ The hydrocarbon spiropentane, first prepared in a relatively pure state by Murray and Stevenson,3is of interest because of its highly condensed structure, and it seemed desirable to have accurate values of certain of its thermodynamic properties. Therefore, an investigation of spiropentane was incIuded in the program of the Bureau of Mines for studying the thermodynamic properties of hydrocarbons and related substances. The present paper reports experimental values of the heat capacity in the solid, liquid, and vapor states, heats of fusion and vaporization, vapor pressure, and entropy, and in addition gives values of the thermodynamic functions calculated from spectroscopic and molecular structure data for selected temperatures up to 1500’ K.
Experimental The Material.-The
spiropentane used for the calorimetric and vapor-pressure studies was purified from crude material, moot of which was prepared in this Laboratory but which also included two smaller quantities obtained from outside sources. The preparation was carried out by reducing pentaerythrityl tetrabromide with zinc in an ethanol-water mixture containing sodium carbonate and sodium iodide, as described by Slabey.* Batchwise reduction of 44.2 moles (17.2 kg.) of pentaerythrityl tetrabromide in this manner yielded 2.47 kg. (3.36 1.) of hydrocarbon product, which was then distilled in an efficient fractionatkg column t o separate the spiropentane from the methylenecyclobutane, 2-methyl-l-butene, and other byproducts of the reduction. Distillation fractions having refractive indices, *BOD,between 1.399 and 1.418 were combined t o give 870 ml. of crude spiropcntaae. To this of spiropentane obtahed in part was added about 76 d e from Dr. Ralph Spitzer, of the State,College of Oregon, and in part from the Automotive Section of the National Bureau of Standards through the courtesy d Frank L. Howard. The combined material wiu treated with bromine to convert the aleiinic impurities into higher-boiling brominated compounds. (Bromine was added untU the yellow color would persist for a number of seconds after the addition of a single drop.) The spiropentane was then separated from the brominated materials by a second fractional distillation. Center cuts from this distillation, all having refractive indices, n%, between 1.41198 and 1.41201, were combined into two samples. Sample A, ‘1) Not subject to copyright ( 2 ) Deceased (8) Murray und Stevcasoa. Tms
P. 812. (4) Slrbcy, ibid., U,11186 ‘IQ4.W
JOTJRVAL,
66, 814 (1944): rbsd,
having a volume of about 160 ml., was used for the heatof-vaporization measurements and the initial series of vapor-heat-capacity measurements. Its purity, as determined by the time-temperature freezing-point method, was 9S.6770. Sample B, having a volume of about 65 ml., was used for the first series of low-temperature studies and for the vapor-pressure measurements. From observations of the melting point ds a function of the fraction melted, its purity w a s found to be 99.6970. In the ebulliometric vapor-pressure studies, the difference between the boiiing and condensation temperatures of this sample a t the normal boiling point was observed t o be 0.009”. After completion of the measurements with these two samples, they were combined and further purified by repeated fractional crystallization. The crystallization was done in a closed, all-glass apparatus, which will be described in detail in a forthcoming publication from the Chemistry and Refining Section of this Station. The product from the fractional crystallization, Sample C (about 70 ml.), was used for the second series of low-temperature studies. I t s purity was 99.87%, as determined by the change of melting point with fraction melted. A summary of the melting-point data for Sample C is given in Table I. The larger volume of less pure material rejected in the fractional crystallization, Sample D, was used for the final series of vapor-heat-capacitv measurements. 1‘.4BLf* I SPIXOPENTANE MELTING POINTSUMMARY 0°C. = 273.16’K., A72/F = 0.02ROAT”
T , “R
Melted, %
Obsd
Calcd.
10.3
166.6969 166.0450 0722 0860
165.6848 166.0450 .0716 .OS60 ,0907 1373
50.4b
x
70 91 I ’
100
Pure Triple point = 166.14 i. 0.05’K. Impurity = 0.13 mole yo. -V* = mole fraction of impurity; F = fraction of sample in liquid form. These points used to obtain calculated values.
Low-Temperature Heat Capacity.-The low -temperature measurements were made in the apparatus described by Ruehrwein and Huffman.6 Very briefly, the method is as follows: About 0.61 mole of the material under investigation was contained in a sealed copper calorimeter, which was mounted in the adiabatic calorimetric system. A measured amount of electrical energy was supplied t o the calorimeter, and at all times the temperature nf tb* (61 Ruelarwoiu aad Huffmrn, Wd., $6, 1620 (19481,
Oct., 1950
THERMODYNAMIC PROPERTIES
environment was maintained at that of the calorimeter to prevent heat interchange. The initial and iinal temperatures of the calorimeter were measured b means of a platinum resistance thermometer. The ecctrical measurements required for determining the resistance of the thermometer and the electrical energy were made on a "White:' double potentiometer in conjunction with a high-sensitivity galvanometer and accurately calibrated resistances. The potential was in terms of a bank of six saturated cadmium cells, which had been certified by the National Bureau of Standards. Time measurements were made with an electric stop clock, which was operated by an a.c. power supply, the frequency of which was accurate fo 0.001%. The precision of the measurements was, in general, better than O.l%, and above 30'K. it is believed that the accuracy uncertainty should not be greater than 0.2%. The energy measurements were made in terms of the international joule and converted t o calories by dividing by 4.1833. Two complete sets of low-temperature measurements were made, one with Sample B and one with Sample C. The two sets of data did not differ significantly, except in the premelting region, and they yielded virtually identical values for the entropy of the liquid a t 298.16'K. Only the data obtained with the purer material, Sample C, will
TABLE I1 THE MOLALHEATCAPACITY OF SPIROPENTANE 0°C. = 273.16aK., mol. wt. = 68.114. Csst.,
1', OK.
AT
12.56 14.12 15.88 16.99 18.21 19.65 21.12 22.54 24.46 26.82 28.10 29.41 31.49 33.26 37.52 41.73 45.88 50.43 54.48 55.08 59.83 60.90 65.22 66.72 70.70 72.55 76.26 78.44 81.90 84.60 87.32 89.91 91.04 93.07 95.62 97.17 99.08
1.511 1.597 1.929 2.595 2.730 2.733 3.088 3.052 3.601 3.497 3.671 3.691 3.107 4.016 4.498 3.920 4.374 4.729 5.593 5.569 5.109 6.063 5.668 5.585 5.298 6.080 5.808 5.695 5.475 6.615 5.359 5.319 6.270 6.151 6.099
cal. deg. -1
0.649 0.919 1.262 1.504 1.776 2.120 2.480 2.835 3.309 3.643 4.205 4.515 4.984 5.393 6.262 7.020 7.697 8.344 8.870 8.941 9.494 9.614 10.067 10.222 10.594 10.766 11.115 11.336 11.658 11.904 12.131 12.332 12,423 12.578 12.767 6.888 12.889 6.897 13.082
Cast.,
T, OK.
AT
cal. deg.-l
102.07 103.53 105.34 109.19 110.10 116.42 116.46 122.52 123.86 128.43 131.61 134.56 139.29 140.91 146.68 153.78 159.96
6.801 6.706 6.604 7.430 6.438 6.201 7.119 6.003 7.692 6.820 7.819 6.442 7.529 6.250 7.256 6.945 5.396 Liquid 5.698 5.701 5.658 9.420 11.154 10.979 10.795 10.610 10.413 10.220 10.020 9.814 9.043 9.460 4.665 7.701 10.687 0.719
13.256 13.365 13.507 13.800 13.864 14.356 14.350 14.800 14.900 15.246 15.498 15.753 16.104 16.249 16.745 17.605 19.993
176.44 176.84 182.12 184.40 194.69 205.76 216.64 227.35 237.86 248,17 258.29 268.21 277.94 287.49 294.55 297.20 301.91 312.11
25.944 25.956 26.076 26.126 26.402 26.781 27.224 27.700 28.240 28.804 29.426 30.068 30.708 31.366 31.890 32.089 32.428 38,194
4865
OF SPIROPENT-4NE
be presented in detail. The results of the heat-capacity measurements are given in Table 11, and the values of CWt. at integral temperatures, as selected from a smooth curve through all the data, are given in Table 111. Two measurements of the heat of fusion gave the same value, 1537.6 cal./mole
.
TABLE I11 MOLALHEATCAPACITYOF SPIROPENTANR AT INTEGRAL TEMPBRATURRS Cast., T , OK. cal. deg.-l
12 13 14 15 20 25 30 35 40 45 50 55 60 65 70 75
0.555 .720 .895 1.080 2.210 3.435 4.645 5.750 6.720 7.560 8.285 8.930 9.510 10.050 10.530 11.000
?', "K.
80 85 90 95
Cmt., cal. deg.-l
11.480 11.935 12.340 12.725 100 13.100 110 13.860 120 14.615 130 15.370 140 16.125 150 16.885 160 17.640 166.14 18.110 Liquid 166.14 25.740 170 25.810
T,O K .
Cast., cal. deg.-1
180 190 200 210 220 230 240 250 260 270 280 290 298.16 300 310
26.020 26.265 26.570 26.945 27.365 27.830 28.350 28.910 29.530 30.180 30.850 31.550 32.150 32.280 33.030
Vapor Heat Capacities and Heats of Vaporization.Measurements of vapor heat capacities and heats of vaporization were carried out with a flow calorimeter and cycling vaporizer as described in previous publications from this laboratory.6 The glass cycling vaporizer formerly used has been replaced by one made of metal. This will be doscribed in a forthcoming publication. Sample A was used in the determination of the heat of vaporization. As the 0.33% of volatile impurity present in this sample probably was similar in physical properties to spiropentane, it is believed that it had a negligible effect upon the heat of vaporization. From three experiments at each of the three temperatures 10.00,25.00 and 38.98" the mean values 6763 * 5, 6572 f 4 and 8393 * 1 cal./ mole were obtained. The uncertainties listed are rnaximum deviations from the mean. The accuracy of these values is believed t o be about f0.1%. For interpolation within the temperature range covered by the experimental measurements, the following equation may be used AH,,,. = 8023 2.727T - 2.547 X 10-'T2 (1) The heat capacity of the vapor was determined a t two or more pressures a t each of five temperatures ranging from 318 to 488" K. Two complete series of measurements were made. The first of these, made on sample A, was rejected because of inconsistencies and a lack of precision in the data indicative of experimental difficulties of unknown nature. The second series was made on Sample D, which at the beginning of the measurements contained approximately 0.4 mole per cent. of volatile impurity but during the measurements suffered an additional contamination of 0.6 mole per cent. of non-volatile impurity. The non-volatile impurity affected the heat of vaporization by an amount mainly due t o a heat of mixing term. The cycling process involved the vaporization of the volatile material and its return in liquid form t o the vaporizer. An analysis of the heat effects attending the changes of state entering into the cycling process indicated that the presence of the small percentage of non-volatile material had little effect upon the flow rates determined from the power input and the heats of vaporization of the uncontaminated sample. The quantity of sample available mas insufficient
+
(6) (a) Waddingtoo, Todd and Huffmn, Tma Jomnu, 0, 11 (1047); (b) Wddlngtom .ad Dousfin, ibid., W, 287s (1W7).
to repeat the measurements 011a repurified sample; hence, the results of the second series are reported here in Table IV. An accuracy uncertainty of *0.4% is aqsigned to these results instead of the usual value of *0.2%.
1AIIL': It" ~ A P u RH E A r 2
-
-f >
1 J I
, 1
1150
1
87r i 77b 581 305 '71'
1
I
2 i
Thermod amic Functions. -Valurs uf the functions H ; - F$T, (H+ -- H!),50, aiid COP were calculated for a range of temperatures. The moments of iurrtia were computed from the electron diffraction teaults of Donohue, Humphrey and Schomaker17to be fA = 77.28 >( 10-40 g. cm and JB = Ti = IPB.R:? Y lO-$" g C I I I . ~ IAHILLi
THERMUDYVAhIIC
I?(
I ' V C P J O X S I )I. SPlKUPEX I A b t ;
Vol. 7 2
The symmetry number is 4. Values of the physical conNo corrections for anharmonicity, rotational stretching or interaction between rotation and vibration were included, as the last two egects were estimated to be negligible, and insufficient data are available for any adeqwate treatment of anharmonicity . The d u e s calculated for the entropy at 383.16, 298.16, and 312.14'' K., 66.40, 67.46 and 68.45 cal./deg./mole agree well with the calorimetric values of 68.37,67.45 and 68.44 cal./deg./mole from Table VI. The values calculated for the heat capacity aTe compared with those determined experimentally in Table 1V. The differences at the higher temperatures between the observed and calculated values of c; are no greatcarthan is to be expected from the neglect of anharmonicity in computing the latter. Table IXlists the values of the thermodynamic fanctions lor selected temperatures up t o 1500' X . Some entries in the table are given to more decimal places than is justified by their absolute accuracy in order to retain internal I-onsistency among the several functions. stants given by Wagman, id &,'*were used.
Acknowledgment.-The authors wish to thank J. W. Moore and K. L. Hopkins of the Chemistry and Refining Section of this station for assistance in the synthesis of the compound. The fractional distillations were done by H. J Coleman of that section. Summary 'The heat capacity of spiropentane in the solid and liquid states was measured over the temperature range 12-298' K. The melting point (166.14 * 0.05" I(.)and heat of fusion (1538 cal./mole) were determined. The heat of vaporization was ineasured at three temperatures, and the values iound were 6753, 6572 and 6393 cal./mole, at 10.00, 25.00 and 38.98', respectively. The heat capacity of the vapor was measured at five different temperatures in the range 318 to 487' K. 'I3e experimental values of C, the heat capacity iii the ideal gas state, may be represented by the equation. C$ = -7.078 0.1085OT k.799 X !0-5Tz. The vapor pressure was ineasured over the temperature range from 3 to Ti', and the following equation was found to lit the va or-pressure data: loglop = 6.91794 1090.589/t 231.165). Xn equation for the second virial coefficient, B , in the equation of state PV = RT BP was obtained from thermal data. This equation is B(cc.) = -57 - 136.7 exp(65O/T). The entropy of liquid spiropentane is 46.29 * 0.10 cal./deg./mole, and the entropy of the vapor in the ideal gas state at one atmosphere pressure is 67.45 * 0.15 cal./deg./mole, both at
+
+
+
298.16
300 400 .500 fioo
700 800 ROO
I
loo0 1100 1200 1300 1400 1500
M 92 55 00 59 03 ti2 88 66.60 70.22 7a 72 77 11 80.40 83.56 86 62 89.68 92.45 95 22
3 739 3 775 6 256 9 451 13 21 17 44 22.06 27 01 32.23 37.70 43 36 49.22 55 22 61 35
67 46 67.59 74 70 81.76 88.62 95 13 101.30 107.12 112.63 117 83 122.76 127.44 131 89 136 12
21.06 21 19 28 5.5 34 91 40.10 44.3ti 47.91 50.93 53 51 35.73 57.64 59.29 60.73 61 9s
-*--
1: _I
IJn,,nh,i. J0i
n,inIl,iuc~
IIU
k t L q J t i i ~ , k ~i i m
t ')i'n1v?i. 67
.'93.16° K. .'ivibrational assignment was made for spiro-
pentane with the aid of detailed nonnal coordiiiate calculations. Using .this vibrational assignment and other molecular structure data, values of the functions (Hi Pr)/T,~FT- H!, So and CpJ were computed for selected temperatures up to 1500' K.
-
BARTLESVILLB, OKLA.
RECEIVED APRIL24, 1950
