29!14 [COSTRIBC r I O S FROM T H E
J. 1'. (;. HICKS,JK.,
AKD
J. (;II,H&KT H O O L I ~ Y
Vol.
(io
RESEARCH LABORATORY O F PHYSICAL CHEMISTRY, hlASSACIIUSETTS I N S r I T U TL O F TFCHNOLOGY, KO. 4061
The Heat Capacity of Potassium Sodium Tartrate Tetrahydrate from 15 to 340OK.l B Y 1.F.
c.HICKS, J R . , ~.4ND J. GILBERTH0OLEY3
The heat capacity of Rochelle salt (KNaC4H4064HzO) in the regions about the Curie points (+24 and - I So) is of iiiterest in correlating the remarkable dielectric properties of this compouad.4 The present investigation was carried out a t the suggestion of Professor Hans Mueller and was designed to cover not only the temperature region -30 to +30", but the entire range a t our disposal. M ~ e l l e r ,by ~ means of his phenomenological theory of Rochelle salt behavior, was able to calculate approximately the change in heat capacity a t the upper Curie point. His figure,fi Ac-1.5 X lo5 ergs/cc./degree-O.; cal./mole,/ degree is rather too small to be positively shown by our measurements. The heat capacity of the salt a t the upper Curie point where our experimental accuracy is 0 3 % , is 92.5 cal./mole/degree. On the other hand, Kobeko and Xelidow7 and RusterholzYreport a sharp discontinuity in the heat capacity-temperature curve at 26'. In both of these iiivestigations the observed rise iii heat capacity above the normal is about 5yo. Material.-The salt was of c. P. reagent quality and contained less than O . O l ~ oimpurities. I t was allowed to stand, being mised frequently, for tWo weeks at room temperature in a desiccator over 40 wt. per cent. sulfuric acid (vapor pressure 13.5 mm.) to ensure the presence of the single solid phase, K N ~ C I H ~ O ~ ~ H Subsequently, ?O.~ it was found that the material a t our disposal did not require this treatment. Analysis for water by loss of weight in vacuum ( mm. and 180 to 200") showed 25.52 and 25.49%; theoretical, 2.5.54. For purposes of analysis, we were unable to drive off more t h a n 3.8 of the 4 moles of water by heating in a drying oven a t 130 for one week, b u t found that conditions under which the tartrates theniselves exhibited incipient decomposition were necessary to remove the last bit of water. -.___
4nalysis for sodium plus potassiuni by ignition and conversion to the chlorides showed thc sample to be 99.98 and lOO.Oii% Rochelle salt, A solution containing 30.0 g. of Rochelle salt per 100.0 1111. of solution gave [ a ] * %+ 2 2 . 2 " ml.,/gram decimeter. .ill measurenients were made on a single calorimeter ~ , 0.31459 mole, the molecular loading of 88.760 g. i n L ~ U C Zor weight of KNaC4H40e,4H20 taken as 282.14. Method.-The apparatus, method of measurement, and calorimeter all have been described in a n earlier paper of this series.10 The temperatures were measured by means of the platinum-rhodium resistance thermometer-heater having the laboratory designation R197. It had been calibrated previously against a helium gas thermometer," and the ice-point resistance showed no appreciable departure from its calibration value. In loading the calorimeter, special precautions were taken to avoid any decomposition arising from heating the salt above 40°.12 The two ring seals of soft solder joining the monel lid to the tnonel collars of the calorimeter barrello were made while the calorimeter and weighed contents were immersed to within 15 rnni. of the top in ice and water. The solder was then flowed on as rapidly as possible. I n this manner, a tight seal could be made without the salt ever attaining the temperature of 15'. However, the contents could not be recovered without decomposition. Prior to the above operation, several centimeters of thinwalled nickel-silver tubing had been sealed t o the bottom of the empty calorimeter and the end of the tube closed. After soldering the lid to the barrel, the free end of the tube was opened, cemented to the vacuum line, the calorimeter and its contents cooled to 80°K., and exhausted. .Wer several flushings with pure helium, sufficient helium was finally admitted to exert a pressure of one atmosphere at room temperature. The calorimeter was allowed to warm, the tube clipped and soldered, and the unit suspended in the apparatus. The e. m . f.'s were read on an Eppley microvolt potentiometer having a range of one-ninth volt t o one microvolt on the dials. The calorie used in this work is defined equal to 4.1833 itit. joules. The absolutc tctripcrature of the ice-point is assumed to bc 273.19O.
i 1) Low. Temperature Studies, S o . 4. ( 2 ) Present address: Corning Glass LVorks, Corning, New York. ( 3 ) Submitted in partial fulfilment of t h e requirements for t h e
Heat Capacity Measurements.-The data are presented in chronological order in Table I. Sedegree of Doctor of Philosophy. (4) For a reyiew of recent x o r k , see S t a u h , ~VaIwr7r~issrrischuft~11,ries I has been omitted because of the large error 23, 728 (1935). Other substances behaving similarly are listed by in reproducibility of these points. The inaccura-
Busch, H d u . P h y s . A c l o , 10, 981 (1937). (>) hlueller, Phys. Kez'., 47, 173 (1935). (6) Mueller, ibid.. 46, 736 (1934). This value is, of course, dependent upon the Lorentz factor, If this factor he three times as great as t h a t assumed in t h e calculation, t h e change in heat capacity ahould have been detected. ( 7 ) Kobeko and Selido\v, Phys. Z.So7r~.ielz~?iioiz, 1, 382 (103%). (8) Kusterholz, H e l s . Phys. A d a , 8, 39 (1933). (0) 1,owry and Morgan, THISJOUKXAI., 46, 2102 (1924).
( I O ) Hicks, ibid., 60, 1000 (1938). (11) Blue and Hicks, i b i d . , 59, l9ti2 (1937). (12) 1,owry and Morgan9 and van Leeuwen, Z.g h y s i k . Chtw7., 23, 33 (18(17), ol,served t h a t the \.apor pressures of Rochelle salt a t lower temperatures were not reproducible, if a t a n y previous time 1,owry and Morgan do not s t a t e t h e salt had been held ahove 40'. t h e composition of t h e other solid phase in equilibrium with Rochelle salt and water vapor when reproducible vapor pressures are obtained
Dec., 10BS
cies were traced to a poor contact in one of the potentiometer dials and the defect was remedied before observations labelled Series 11. TABLE I Cp in cal. mole-'
T,O K . degree-' Series I1 59.93 24.29 63.85 26.08 69.06 28.24 73.00 30.39 77.07 31.51 81.08 33.92 85.00 35.37 88.74 36.68 92.42 38.13 96.09 39.07 99.98 41.03 103.94 41.76 107.53 43.72 111.51 44.98 115.97 46.70 120.47 48.06 125.05 48.91 130.04 50.22 134.19 51.73 139.22 53.34 144.22 54.54 149.29 56.25 155,67 57.81 161.33 59.49 167.20 61.42 173.04 62.91 178.84 64.73 184.89 66.02 187.11 66.75 192.96 68.08 201.10 69.96 210.54 72.45 217.36 73.79 223.66 75.55 229.89 77.13 226.08 76.13 232.36 77.41 238.82 78.83 243.42 79.77 244.29 76.10
2!)!K
HIZ.\T C.4PACITY OF POTASSIUM S O D I U M TARTRATE TETRAIIYDKATI:
Cp in cal. T , OK.
mole-' degree-1
246.64 248.86 250,98 253,12 255.27 252.56 254.70 256.78 259.08 265.16 270.38 275.60 280.10 286.80 293.01 296.48 298.72 300.94
80. 59 81.31 82.01 81.96 82.38 81.83 82.21 83.26 83.80 85.07 85.80 87.06 88.63 89.82 91.60 92.23 92.98 92.40
Series I11 246.13 245.76 248.32 250.91 253.45 255.89 258.32 263.60 267.35 287.90 292.74 295.70 298.15 299.90 302.15
Series 16.72 18.95 21.14 23.62
78.73 80.37 80.57 81.50 82.01 82.56 83.92 84.56 84.96 89.96 91.63 91.72 92.59 93.43 91.79
3.31 2.98 4.04 4.74
OK.
C p in cal mole-' degree-'
26.28 29.04 31.72 34.04 36.75 40.28 43.97 47.81 52.15 56.68 61.06 65.24 69.74 74.61
6.03 7.23 8.61 9.88 11.16 13.20 15.25 17.39 19.94 22.08 24.67 26.89 28.93 31,28
T,
Series V 244.19 79.58 246.32 80.17 248.54 80.83 251.08 8 1 . 4 4 252.06 81.73 254.23 82.33 256.34 82.73 258.39 83.07 258.62 83.56 261.30 83.60 263.71 84.90 Series VI1 244.73 80.36 300.11 93.07 303.72 94.20 307.41 95.16 311.00 95.59 314.49 96.80 317.94 98.97 321.35 100.76 324.67 104.07 323.33 99.93 326.57 107.07 332.87 191.35 335.25 188.24
Before Series 11, the calorimeter was held for two weeks a t 80'K. and between Series I1 and 111, the calorimeter was cooled directly to hydrogen temperatures and was below 80'K. for about two days during the course of Series IV measurements. It was then allowed to warm up slowly over a period of four days to 245'K. where it was held for six days. Series V data were then taken. Following these runs the calorimeter was immediately cooled to 240'K. and held just below 250'K. for one week after which Series VI observations
were made. Thereafter, Series VI1 and the heat of transition were measured ; these required three days. It is evident from the data that the variety of thermal treatments produced no effect on the heat capacity below 310'K. The data of Table I are shown graphically in Fig. 1. The data of Series I11 and V are displaced, but all three curves are identical. The three dots are from Wilson's13 work; his straight line is indistinguishable from that portion of our curve when the data are plotted 0x1 this scale. Figure 2 is a deviation plot with ordinate the observed heat capacity minus that calculated from the straight line, C, = 0.300T 7.00. The lengths of the vertical lines represent 0.5% of the total heat capacity a t that temperature. The upper curve in Fig. 2 is the portion of the deviation curve from 240 to 300'K. replotted on a larger scale abscissa so that the experimental points can be distinguished. The measurements of Series V I are listed in Table I1 and plotted in Fig. 3 . They represent relative total heat capacities of the full calorimeter in the neighborhood of the lower Curie temperature. Each heating period was made through approximately a 0.3' interval with the shield sur-
+
TABLE I1 T , OK.
Cal. degree -1
Series VI 249.18 55.16 249.63 55.01 250.07 54.83 250.50 55.09 250.97 54.89 251.42 55.10 251.85 54.75 252 29 55.36 252.78 54.48 253.17 55.39 253.60 55.34 254.05 55.18 254.50 55.17 254.98 55.41 255.40 54.98 256.07 55.59 256.82 55.54 257.28 55.61 257.71 55 67 258.15 65.15 258.59 55.58 259.04 55.77 259.48 55.35
TABLE TI1 Cp in cal.
T ,O K .
15 20
25 30 35 40 4
