/COKTRIBUTION FROM THE
THERMODYNAMICS SECTION,
KATIOSAL
BUREAU OF
STAXDARDS]
Thermodynamic Properties of Gaseous Difluorodichloromethane (Freon-12) BY JOSEPH F. ~ I A S I RECEIVED APRIL 12. 3052 The heat capacity of gaseous difluorodichloromethane (Freon-12) has been measured at -30, 0, 45 and 90” at pressures up to 1.5 atmospheres. The results, accurate within *to lye, have been used t o obtain more accurate values of Cg and AC,/AP, the change of heat capacity with pressure, than hitherto available. Based on the experiments and on a recently puhlished assignment of frequencies, new tables of the heat capacity, heat content, entropy and free energy function have 1)erii calculated A new equation of state has beeu developed to conform to the values obtained for lC,,/AP.
Introduction BuEngton and Gilkeys published data of state Measurements of heat capacities of gases, with with an accuracy of about 1%, and Buffington and an accuracy of one-tenth per cent., over a range of Fleischer“ obtained three values of the gaseous temperatures and low pressures, are of fundamental heat capacity a t one atmosphere with an estimated importance in a.t least two ways. First, the experi- uncertainty of about 1%. In the present research, heat capacities were mental values may be extrapolated to zero pressure a t each temperature to yield values of the heat measured with the flow calorimeter from the boiling capacity of the hypothetical ideal gas. For most point of CFAX up to go”, and a t pressures up to gases whose molecules have more than three atoms, 1.5 atmospheres. The ranges covered were apthe heat capacities thus obtained are more reliable proximately those over which the apparatus could than those calculated from molecular data and can conveniently be operated. New tables of thermodynamic properties of frequently be used to decide questions of structure or frequency assignment. Second, the values ob- Freon-12 gas, both ideal and real, as a function of tained for the change in heat capacity with pres- temperature and pressure, are being prepared for sure are an order of magnitude more accurate than the NBS-NACX series. the same quantity calculated from data of state of Experimental good accuracy. Thus there is provided an opporMaterial.-Ten pounds of CFiClz, specially purified by tunity to check or modify existing equations of state. fractional distillation, was obtained from the Jackson LaboThe flow calorimeter used in this Laboratory was ratory of E. I . du Pont de Semours Company. The suprecently shown? to have an accuracy of better than plier furnished an infrared analysis which showed the CF2t o contain no detectable impurities. The most likely one-tenth per cent. The apparatus is a modifica- Cls impurities were stated to be CClF3, CFCla and CHClFi, tion of a calorimeter described by n’acker, Cheney and the minimum amounts detectable by infrared analysis were 0.02, 0.01 and 0.05%, respectively. and S ~ o t t . ~ About one pound of the material was vented from the The investigation of difluorodichloromethane cylinder, then about two pounds was taken from the vapor (“Freon-12”) was begun because of the scarcity of phase for use. This sample was further purified by twice data in the literature from which to compile accu- evaporating it a t the boiling point and condensing i t in a rate tables of thermodynamic properties of both trap surrounded by liquid nitrogen, while pumping with a ideal and real gas for the series of “NBS-?;SCAI high-vacuum apparatus. Initial and final portions were Finally, the sample was repeatedly frozen and Tables of Thermal Properties of Gases.”.’ The discarded. evacuated until a pressure of about 0.2 micron was observed recent work of Plyler and Benedictj on the infrared at the temperature of liquid nitrogen. A mass spectrograph analysis of the sample as used indispectrum and the assignment of fundamental vibration frequencies, gives a calculated heat capacity of cated presence of small amounts of one or more hydrogenderivatives such as CHClF?. The analysis was the ideal gas about 3:; higher than the previous containing essentially qualitative, but seemed t o indicate that the values of Thompson and Most of the sample was 99.870 CFzCll or better. The maximum poscharts and tables’ of thermodynamic properties sible error in heat capacity caused by impurities is estimated which have been made available have resulted from to be 0.05%. and Method.-The flow calorimeter and the experimental work of I3uffington, et o / . ~ - - ’ its ~ Apparatus operation have been described in detail elsewhere.? -4 i i j This work 15.183 supported in parr h y the Sational Ad\-isory Committee for Aeronautics. ‘ 2 ) j F. Xlasi ami €3 P e t k * l f . J . R-si~arch .\-til/ R U P . ~ i u u * / ~ i r IJ S - ,,
P. I? Wdckcr. R u t h K Cheney and K. H S r o l t . i h i i l , 38. 051 , 1 5 1 4 7 ) ,R P 1804. ( 4 ) Information about these tables may be obtained b y writing Rlr. Joseph Hilsenrath, Heat and Power Division, National Rrireau o f SLandards, Washington 2 5 , D. C . 13) E. K . Plyler and W. S. Benedict, .I K e s e i i v r h Y i i f l Hiir,. , S ’ L [ I ~ ~ I , I T d S , 47, 202 (1951). K P 2245. ( 6 ) H. W. Thompson and R. B. Temple, J . Chrm. S o c , 1422 f19481. 17) For example, R. h l . Buffington and \V. K. Gilkry, Amer. Refrig. Engr., Circular Nc,. 12 ;1931j. ( 8 ) K. h f . Buffington and 11’ K . Oilkey, 117~1 E n g . i’liev!., 23, 2.74 I 1 g.3 1i , ! I ) IV, K. C,ilke?-, F IV, ( k r ; + r d and SI I i Bixler, i h i f i , 23. :iiil fl!lRl~, ! I O 1 I;. R . Bichowsky and \T. K. Gilkey, ibid , 23, 366 f ! I I ) I-near the temperature of the heated gas. ‘The gas entering the calorimeter at I was brought to bath temDerature in the helix. its temuerature was measured at SI, electrical energy was added by means of the heater H, and the temperature was again measured a t S ? . S i and Si : I r p nickP1 rrsistance thrrmomrters:. The power and rate of
Oct. 5, 1932
THERMODYNAMIC PROPERTIES
OF GASEOUS
flow were always chosen so that the rise in temperature was close to 10’. The sample of Freon was contained in a steel cylinder i l l an ice-water bath (not shown in Fig. 1). Reduction in pressure was accomplished, and the rate of flow was accurately controlled, by means of a sensitive diaphragm valve. The mean pressure and pressure drop in the calorimeter were observed by means of a three-column manometer, two columns of which were connected to the calorimeter as shown in Fig. 1. The gas was led from the calorimeter outlet a t 0 to a receiver a t liquid nitrogen temperature, through a snapthrow valve which was used to divert the flow, within 0.1 second, from one receiver to another. A heat capacity experiment consisted of the data taken during a steady state period when the gas was removing heat a t the same rate it was being added. Nearly continuous measurements were made of the exit temperature a t Nz, and slight adjustments of the flow were made t o keep that temperature as constant as possible. The experiment was begun and ended by directing the flow of gas to and away from a weighed receiver by means of the snap-throw valve. The readings of Nz were identical at these two moments so that there was no gain or loss of energy by the calorimeter. Readings of the current and potential in the heater, the resistance of NI and a platinum resistance thermometer in the bath, height of the manometer columns, barometer, time interval to 0.01 second and weight of sample to 1 mg. were the other data obtained. The electrical circuits used were similar to those previously described.13 A G-2 Mueller Bridge was used for the measurement of resistances, and a sensitive five-dial potentiometer, with calibrated standard resistor and volt-box, was used for the energy measurements. I n order to eliminate, as far as possible, the effect of residual heat leaks, the heat capacity was determined a t a number of different flow rates a t each temperature and pressure of measurement, and extrapolated to zero reciprocal rate of flow. Also, “blank” determinations were made a t several rates a t each temperature and pressure, to measure the amount of cooling, 6T,experienced by the gas in passing through the calorimeter when no heat was applied. The value of 6Tvaried from 0.008 to 0.35’.
Results Heat Capacity of the Ideal Gas.-If W is the power supplied to the heater, F is the rate of flow of the gas, AT is the observed rise in temperature, and 6T is the fall in temperature in a corresponding blank experiment, the apparent heat capacity is calculated by WF-1 e=---AT 6T
+
DIFLUORODICHLOROMETHANE4780
VA C
1I
MAN t
t
S
.
b--N*
..
T
-.
,.
L-_TC2
-
0 0
I
0 0 - .
0 0
0 0
0
6i
p;’ /I
The values of C were corrected for the deviation of the observed mean pressure and mean temperature from the nominal values. These corrections were small, never larger than 0.1%. The corrected
Fig. l.-&ale drawing of flow calorimeter: B, constant temperature bath level; FI and F2,flanges; H,, calorimeter heater; Ha, shield heater; Ht, tube heater; I, inlet; J, TABLE I bellows seal; M, mica spacer; M A S , manometer tubes; HEATCAPACITY OF GASEOUS DIFLUORODICHLOROMETHANE ICl and N2, nickel thermometers; 0, outlet; S, radiation (SUMMARY O F RESULTS) shield; T?, helical tube; TCI, tube thermel; TC2, shield -3O.OOO O.OOo 45.00’ 90.00 thermel; Th, metal thimble; T‘, throttle valve; VAC, Cp, calories mole-1 degree-’ vacuum line. At press., atm. *
1.50 1.oo 0.67 .50
.33 Ci, obsd. ACJAP, cal. mole-’ deg.-I atm.-’
17.274 18.276 17.048 18.161
19.251 19.180
16.114 16.844 15.893 15.672 16.650 0.664
0.410
18.058 19.114 17.946
19.045
0.218
0.137
(13) R . B . Scott, C. H.Meyers, R . D. Rands, Jr., F. G . Brickwedde and N. Bekkedahl, J. Research N a t l . Bur. S t a n d a d s , 56, 39 (1945).
RP 1661.
heat capacity, called C,(observed), for all of the experiments with CF2C12, are plotted in Fig. 2 against the reciprocal of the rate of flow. The straight lines drawn through the points in Fig. 2 were determined by the method of least squares. The average standard deviations from these lines was =t0.03%. The intercepts are taken as the true heat capacities and are reproduced in Table I. Thermodynamic Functions of the Ideal Gas.The final values of heat capacity a t finite pres-
4740
Vol. 74
JOSEPH F. MASI
182€
1818
-
18 I C
__
-
1/2 A T M
-
&-Po--
I
w
wl802
a
0 W
n
l70€
cc
W LQ
m
0 4 7 OC
i)
-
203-
I 6 9;
I 6 8'
16 1 4
160C Temperature " C
Fig. 3.-Heat capacity of ideal gas CFzClz: the base line (zero ordinate) is the calculated heat capacity as given in Table 11; 0.this research; @, Buffington and Fleischer"; 8 , Eucken and Bertram,l6 ----,calculated from assignment of Plyler and Benedict' (without anharmonicity correction); . , , same, changing YS from 473 to 450 cm.-l.
I 5 3@
I 5 9C
RECIPROCAL OF RATE, SECONDS GRAM
Fig. 2.-Heat
-'.
capacity data for CF2C1,.
sures in Table I were extrapolated linearly to zero pressure at each temperature, and the results are
I n the usual manner, the heat capacity, heat content, entropy and free energy function have (14) W. H. Stockmayer, G. >I. Kavanaugh and H. S. Mickley, J . Chem, Phrs, la, 408 (1944) (1:) A. Eucken and A. Bertram. z.physlk. Chem., B31, 361 (1936)
THERMODYNAMIC PROPERTIES OF GASEOUS DIFLUORODICHLOROMETHANE 4741
Oct. 5 , 1932
been calculated, including the anharmonicity were calculated and plotted against P. It was corrections noted above, and are given t o 1500°K. found that a better fit of the P-V-T data could be in Table 11. The product of the moments of inertia obtained if b was a linear function of temperature was obtained by the method of Hirschfelder,lBusing rather than a constant. C was assumed to be inthe bond distances and angles given by Brockway. l 7 versely proportional to the temperature. The final expressions for the virial coefficients to be TABLEI1 used in eq. 2 were FUNCTIONS OF THE IDEALGAS, CFtC11 THERMODYNAMIC c; T,
OK.
200 273.16 298.16 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500
s9
9
B = 216
+ 0.391T - 152e445/Tcc. mole-'
C' = 14.6
-
and
3
cal. mole-) deg. -1
cal. mole-' deg. - 1
14.043 16.654 17.407 17.456 19.831 21.469 22.605 23.411 24.003 24,448 24,791 25.060 25.277 25.454 25.602 25.726
65.648 70.430 71.920 72.027 77.394 82.004 86.024 89,571 92.742 95.599 98,194 100.57 102.76 104.79 106.68 108.45
2.0100 3,1372 3.5628 3.5948 5.4664 7.5354 9.7419 12.044 14.418 16.845 19.308 21.800 24.317 26.853 29.406 31.973
55.598 58.945 59.971 60.044 63.728 66.933 69.788 72.366 74.719 76.883 78.886 80.751 82,495 84.132 85.676 87.136
=
RT
+ BP + CP2
(3)
and B was assigned the form proposed by Hirschfelder, McClure and WeekslB B =b
- &IT
(4)
The values of c and a were then chosen to fit the experimental data. Then, for all the experimental P-V-T data obtained by Buffington and Gilkey,8 the values of the left-hand side of the equation V
RT -+ ceaIT P
=
b
+ CP
, 0 TEMPERATURE,
(2)
could be neglected. The values of AC,/AP listed in the last line of Table I1 were set equal to
(5)
(16) J. 0. Hirschfelder, J . Chcm. Phys., 8 , 431 (1940). (17) L. 0. Brockway, J . Phys. Cnem., 41, 747 (1937). (18) J. 0. Hirschfelder, F. T. RlcClure and I. F. Weeks, J . Chem. Phys., 10, 201 (1942).
cc. mole-' atm.-'
(6)
From eqs. 6 the derivative of heat capacity with respect to pressure was calculated a t one atmosphere and plotted as the solid curve in Fig. 4.
Equation of State.--At the low pressures of the heat capacity measurements, it was assumed that the term involving P 2in the virial equation PV
' 9
I
I
50
100
'C.
Fig. 4.-Pressure coefficient of heat capacity of CFZCI~: 0,this research; -, equations 6; -----, equation of Buffington and Gilkey!
The same quantity calculated from the equation of state of Buffington and Gilkey is shown by the dashed curve. The virial coefficients of eqs. 6 reproduce all of the data of state of Buffington and Gilkey with a root-mean-square deviation of *0.5% compared with =kO0.40j,deviation from the equation derived by those authors. Acknowledgments.-The author wishes to thank Mr. Benjamin Petkof, Mr. Richard Sherman, Mr. Joseph Ratti and Mr. Howard Flieger for performing calculations and assisting with the measurements a t various times. Thanks are due Mr. R. C. McHarness for information concerning the purity of the sample. WASHINGTON, D. C.
