June 20, 1953
HEATCAPACITIES AND THERMODYNAMIC PROPERTIES OF (NH4)2O AND NHiOH
2811
ICONTRIBUTION FROM THE CHBbfICU LABORATORY OF THE UNIVERSITY OF h I F O R h ' I A ]
Ammonium Oxide and Ammonium Hydroxide. Heat Capacities and Thermodynamic Properties from 15 t o 300'K.l BY D. L. HILDENBRAND AND W. F. GIAUQUE RECEIVEDFEBRUARY 13, 1953
It has been shown that ammonium oxide crystallizes as a pure compound to within 0.02 mole % even a t the two eutectic regions where it is present. Ammonium hydroxide was shown to be pure to within 0.05% at the two eutectic regime where it is a n equilibrium phase. These are upper limits for solid solutions under the above conditions, and these two compounds must crystallize in very pure form at their normal melting points. The heats of fusion of (NH&O and NH4OH were found to be 2352 and 1568 cal. mole-' a t their respective melting points, 194.32 and 194.15"K. The heat capacity of (NH&O was measured from 15 to 273'K., and that of NHdOH from 15 to 293°K. Sharp double maxima occur in the heat capacity curve of ammonium oxide in the region near 52'K. Tliese show the characteristics of cooperative phenomena and are believed t o be due t o angular motions of NH, groups. The gradual transition was studied in detail and total heats through the
from isothermal data available a t 298.16'K. The entropies of (NX&O and NH'OH are 63.94 and 39.67 cal. deg.-' mole-', respectively, at 298.16'K. It is concluded that (NH4)*0and NHIOH approach zero entropy at low temperatures and thus all hydrogen bonds dnd other structural details must be essentially ordered in the crystalline states of these substances a t low temperatures.
This paper reports a low temperature thermodynamic investigation of ammonium oxide and ammonium hydroxide. The work is a portion of the series of investigations in this Laboratory on possible non-equilibrium disorder in the solid state a t low temperatures due to random hydrogen bonding. Preparation.-Pure dry ammonia was prepared from C.P. ammonium chloride and potassium hydroxide, as described by Overstreet and Giauque? and then mixed with the necessary amounts of pure. water to make ammonidm oxide or ammonium hydroxide. The determination of the exact composition will be discussed below. Freezing point composition data for the ammonia-water system have been obtained by Piekering,' Rupert,' Smits and Postma,' Postma' and Elliott.? Apparatus.-It waa planned to use a calorimeter which has been given the laboratory designation Gold Calorimeter I F b ; however, it was found that a crack had developed in the heavy gold bottom. The cause of this is worth noting. During the first fifteen years, of the some 26-year period in which this calorimeter has been used, a thermocouple was soldered to the bottom and diffusion of lead into the gold was responsible for the cracking. This possibility was reduced in later calorimeters of this type by welding one or two platinum thermocouple wells to the gold. These wells were filled with either Rose's or Wood's alloy. It was expected that the platinum tube would afford some protection to the gold. A tube of this type was used on Gold Calorimeter I1 during approximately the last ten years it was in use but there was evidence that the Rose's alloy diffused through the platinum tube in both this and in another similar calorimeter. Gold Calorimeter IVg was removed from its surrounding lead-copper protective cylinder and installed in place of Gold Calorimeter I1 which was scrapped. Otherwise the apparatus and standard thermocouple were substantially those which have been described previously as Gold Calorimeter 11. Wood's alloy had been used in the platinum thermocouple well of Gold Calorimeter IV. The standard thermocouple, which has the laboratory (1) This work was supported in part by The Offlce of Naval Research, United States Navy (2) R. Overstreet and W. F. Giauque, TEIS JOURNAL, 69, 254
(1937). (3) S. U. Pickering, J. Chem. Soc.. 68, 141 (1893). (4) F.Rupert, THISJOURNAL, 81, 800 (1909); 84, 748 (1910). (8) A. Smits and S. Postma, 2. anorg. Chcm., 71,250 (1911). (6) S. Postma, Rec. Iror. chim., 89,516 (1920). (7) L. D. Elliott, J. Phys. Chcm., 48,887 (1024). (8) (a) W.F. Giauque and R. Wiebe, THISJOWRNAL, 60,101 (1928); (b) R. W. Blue and W. F. Giauque, ibid., 6'7, 991 (1935). (9) J. D.Kemp and W. F. Giauque, ibid., 69, 70 (1937).
desimation W. and which was oridnallv calibrated in 1926. was-compared with the vapor $ r e s h e of hydrogen af 14.70, 16.02, 18.21 and 20.41'K. (b.p. 20.37"). The results averaged 0.06 & 0.02' low. It was also comDared with the vaior pressure of nitrogen at 63.15' (T.P.), 64.97, 70.76 and 77.32OK. (b.p. 77.34'). The results averaged 0.07 zk 0.02" low. Appropriate corrections were made a t these and other temperatures. One defined calorie was taken as 4.1840 absolute joules. 0°C. = 273.16"K. The molecular weights were taken as 35.048 and 52.080 for NHrOH and (NH4)20, respectively. Filling the Calorimeter.-In previous work it has been a simple matter t o distil volatile pure substances into the calorimeter from a closed system. In this case the ammonia distilled in first and it was impractical to make the water rich residue follow it. An attempt to distil ammonia into water which had been placed in the calorimeter proved impractical because the lower density of the solution, formed near the surface, made mixing slow. The dilute solution could not be cooled much below the ice point during this procedure because separation of ice could have caused damage to the calorimeter due to expansion. It is quite practicable to produce convection stirring in such calorimeters, by heating the bottom section, when density differences are small but they were too great in the present case. It was finally necessary t o cool the ammonia solution to a low temperature t o reduce the pressure and introduce the liquid through a fine stainless steel capillary. It was not possible to maintain composition accurately during this procedure so that excess ammonia was used and then removed by pumping out small increments until the composition of pure compound was approximated. This could be ascertained by measuring the amount of heat required for eutectic melting after each increment of ammonia was removed. The total amounts in the calorimeter were determined by distilling them out and weighing a t the end of the measurements. All weights were corrected to vacuum. Before solidifying the liquids, helium gas was always admitted when the temperature was near the melting point, so that essentially all of the ammonia gas was swept into the calorimeter. The helium gas, a small amount of which also served to improve thermal conduction, was pumped out a t a low temperature before measurements were made on the liquid state. The protective copper-lead cylinder, which enclosed the calorimeter, was then kept at a higher temperature than the calorimeter at all times to avoid possible distillation effects. The Densities of Ammonium Oxide and Ammonium Hydroxide.-In order to know the amounts of (NH4)zO and NHdOH which could be held by the calorimeter, and also for calculating the gas space in connection with vaporization corrections, it was necessary t o know the densities approximately. Densities of ammonia water solutions near ordinary temperatures are tabulated in the I.C.T.'" These were (10) "International Critical Tables," Vol. 3, McGraw-Hill Book N. Y., 1928,p. 59.
Co., Inc., New York,
2812
n.L. IJILDENBRAND AND I f r . F. GIAUQUE
Yol. 7 ;
supplemented by two measurements, accurate to about 2%, by means of a rough calibration on a measuring bulb. NHIOH = 0.89 at 206°K.; (NHJ20 = 0.82 a t 203°K. Linear interpolation was sufficiently accurate for the several intermediate compositions.
the eutectic on the NHs side was identical to within an accuracy of about 0.02 mole % with the (NH4)20 phase in the eutectic on the NH40H side. We believe that no case is known where such a Proof that Ammonium Oxide and Ammonium situation corresponds to other than a pure comHydroxide Crystallize as Pure Compounds and pound and it is very difficult to imagine that a solid solution of fixed composition could be in equilibrium Not as Solid Solutions in the Eutectic Regions.with two liquid phases of widely different composiWhen the calorimeter contained an excess of ammonia, over the amount corresponding to (NH4)20, tions a t the same temperature. Accordingly we a heat of fusion of the NHs-(NH&O eutectic was conclude that pure (NH4)20 is the equilibrium observed a t about 181'K. Postmae gives 180.6'K. phase in each of the two eutectics where it is for the (NH&O-NHs eutectic temperature and present. The eutectic melting data as a function of ammonia removed are given in Table I and the Rupert4found 179.1'K. The eutectic heat of fusion was reduced linearly intersection of the two straight lines a t zero heat of when increments of ammonia gas were removed fusion is shown in Fig. 1. 1.9421 moles of (NH4)20 a t higher temperatures followed by resolidification was in the calorimeter during this experiment. and fusion. This linearity would be expected TABLEI whether or not the eutectic consisted of solid HEATSOF FUSIONOF EUTECTIC PHASES AS A FUNCTION OF solutions, provided that all of the solid solutions AMMONIAREMOVED had the composition corresponding to the eutectic 1.9421 moles (NH4)20 in calorimeter equilibrium. In a case where non-equilibrium Total moles Eutectic heat Euctectic solid solutions persisted a t the eutectic temperaNHs removed of fusion, cal. phases ture small deviations from linearity would be 0 190.2 (NHa)*O-NHa possible. When stlfficient ammonia had been 188.9 removed the NH40H-(NH.&O eutectic appeared .I3429 97.2 ( SH4)20-NHz and the heat of fusion increased linearly as addi95.6 tional increments of ammonia were removed. .OS02 19.0 (NHM-NHs The (NH&O--NH40H eutectic temperature was 17.7 about 187.2'K. Postmaa found 187.4'K. and .0920 9.8 NH4OH-(NHd)zO Rupert4 reported 186.1'K. 10.2 When each of the linear curves was drawn to .1398 150. I ~ I O H - NH4)zO ( zero heat of fusion they were coincident within a The purity of the NH40H phase was not infew tenths of a calorie per mole of (NH4)20. This shows that the (NH4)20 phase in equilibrium with vestigated in such detail but it was shown in the following manner: an eutectic heat of fusion of 10 200 cal. was obtained at about 173'K. This indicated an excess of water in the sample which contained 3.1498 moles of NH40H. Various heat content data on the ammonia water system enable a rough estimate of the heat of fusion of the H20-NH40H eutectic as about 2000 cal. for the amount of eutectic associated with one mole of I50 ice. The observed 10 cal. heat of fusion of the amount of eutectic in the calorimeter corresponds to only 0.005 mole of ice and a requirement of 0.005 CAL, mole of NHa to make pure NH4OH. Thus an accurate heat of fusion is not of great importance. An amount of 0.0133 mole of ammonia was added, I oc followed by convection stirring of the solution, solidification and finally eutectic melting. The eutectic melting near 173'K had disappeared completely and a heat of fusion of 35 cal. appeared a t the (NH4)20-NH40H eutectic temperature. This corresponds to 0.0105 mole of excess ammonia, 50 compared to 0.0083 mole calculated. The agreement is probably within the limit of error and in any case indicates that the NH4OH phase is a pure compound to within about 0.0570. 0 M O L E S OF N H s REMOVED
Fig. 1.--Heats of fusion of NHa-(NH&O eutectic and (NH4)z0-XH40H eutectic: 1.9421 moles of (NH4)20 in the calorimeter.
The Melting Points of Ammonium Oxide and Hydroxide. -The melting points were determined by adding increments of energy and waiting a t least two hours for equilibrium. Although observations were made as a function of the amount melted the freezing point composition curves are so flat that the NHdOH in the (NHa)20 and wice versa produced very small effects. The results are given in Table I1 Heats of Fusion of (NH4)20and NH40H.-The heats of fusion of ammonium oxide and hydroxide were measured by
June 20, 1953
HEAT CAPACITIES AND
THERMODYNAMIC PROPERTIES OF (NH4)20 AND NHaOH
TABLEI1 MELTING POINTS OF ( NH4)*0AND NH40H Melted,
Resistance thermometer, T , OX.
7,
Thermocouple
NH4OH 194.123 194 152 194.156 194.162 194.153 194.160
8 25 43 60 79 91 Accepted value Other observers
194.15 194.18 194.15 194.15 194.15 194.14 194.15 193.8 194.1
~
Rupert' (1910) Postma8 (1920) (NHJzO 194.319 194.323 194.317 194.318 194.3 19 194.325
14 30 45 61 76 92 Accepted value Other observers
TABLEV HEATCAPACITY OF AMMONIUM OXIDE 1.9391 moles (NH4)zO $0.0030 mole NHlOH in calorimeter. Results have been corrected to pure ( NH4)20by subtracting CP of NHlOH, cal. deg. --I mole-' (NH4)20 Series I CP
T , OK.
CP
T , OK.
CP
14.53 16.13 17.83 19.63 21.65 24.11 26.74 29.39 32.29 35.73 39.80 43.68 47.42
1.000 1.314 1.661 2.059 2.533 3.129 3.784 4.492 5.274 6.303 7.557 8.883 10.42
58.05 61.98 66.37 71.47 76.80 82.32 88.31 94.08 99276 105.40 111.14 116.96
11.39 11.55 12.02 12.70 13.43 14.19 15.02 15.80 16.56 17.29 18.04 18.74
122.86 128.84 134.89 141.10 147.34 153.66 159.85 166.13 172.42 178.59 183.70 186.94
19.46 20.13 20.82' 21.48" 22.20" 22.90" 23.58" 24.21" 24.87" 25.48" 26.07" 28.49*
T , OK.
AT
CP
T , OK.
AT
3.192 3.205 3.070 3.220 3.559 4.522 3.538 2.371 2.263 0.521 .450 .289 ,279 .260 .0864 ,0255
2.661 3.436 4.243 5.111 6.110 7.353 8.670 9.837 11.02 12.00 12.51 12.97 13.50 14.59 21.93 76.75
51.65 51.66 51.67 51.69 51.77 51.90 52.20 52.65 53.11 53.88 55.02 56.28 57.51 58.91 60.98
0.0077 ,0052 ,0074 .0310 .127 .136 .435 .463 .476 1.030 1.253 1.239 1.211 1.567 2.572
T,
194.32 194.34 194.31 194.32 194.32 194.32 194.32 194.1 194.3
OK.
Series I1
22.15 25.37 starting below the (NH4)zO-NHrOH eutectic temperature 28.52 and heating somewhat beyond the melting points which 31.70 were only about seven degrees above the eutectic tempera- 35.10 ture. Corrections were made by means of the heat capaci- 39.14 ties and for the fusion of the small amounts of the other eu43.20 tectic phases present. The data are given in Table 111. 46.20 TABLEI11 48.55 HEATSOF FUSION OF (NH4)20 AND NHdOH AT THEIR MELT- 49.97 ING POINT 50.46 NHiOH (NH4)20 50.84 Initial AH, Initial AH, 51.13 T, O K . cal. mole-' T ,OK. cal. mole-' 51.40 2352 1565 186.20 186.07 51.58 1568 185.75 2352 186.25 51.63 2351 1569 185.84 185.63 Accepted values 1568 i 3 2352 =I= 3 Rupert' (1910) Postma6 (1920)
TABLEI V HEATCAPACITY OF AMMONIUM HYDROXIDE 3.1288 moles NH4OH 0.0105 mole (NH4)zO in calorimeter. Results have been corrected to pure NH4OH by subtracting C, of (NH4)20, cal. deg.-' mole-' NHaOH
+
T , OK.
CP
T,OK.
CP
T,
OK.
2813
CP
15.39 0.443 87.22 9.28 194.15' = m.p. 17.22 .620 93.36 9.94 197.12 28.36 10.56 201.38 28.74 18.94 .805 99.24 20.81 1.022 105.00 11.15 206.38 29.15 22.83 1.262 110.70 11.74 211.61 29.62 25.33 1.572 116.31 12.29 216.93 30.10 28.10 1.924 122.01 12.86 222.37 30.60 31.03 2.309 127.82 13.40 227.89 31.09 34.27 2.761 133.68 13.94" 233.47 31.60 37.79 3.250 139.60 14.46" 239.13 32.12 41.97 3.794 145.62 15.03" 244.84 32.63 46.66 4.399 151.71 15.58" 250.62 33.17 51.45 5.033 157.84 16.19" 256.44 33.65 56.17 5.627 164.04 16.71" 262.35 34.19 61.17 6.259 170.26 17.29" 268.28 34.69 66.28 6.892 176.34 17.81" 274.12 35.15 71.19 7.477 182.47 18.38" 279.81 35.62 76.33 8.073 186.76 23.70b 285.33 36.06 81.71 8.666 290.21 36.39 * Eua Temperature rise obtained from thermocouple. tectic heat absorption, A T = 2.248'
CP
281. 429. 301. 71.33 13.73 12.50 12.52 12.95 14.01 12.53 11.77 11.48 11.36 11.35 11.45
Series I11
T,OK.
AT
CP
T , OK.
AT
CP
50.46 50.87 51.46 51.97 52.10 52.22 52.35 52.48 52.60 52.74 52.87
0.509 .292 .877 .1245 .1253 .1236 .1221 .1210 .1307 .1297 .1270
12.48 13.01 25.49 12.40 12.34 12.55 12.72 12.83 13.09 13.17 13.48
53.00 53.13 53.26 53.38 53.58 53.87 54.20 54.62 55.19 55.77
0.1262 .1235 .1203 .1240 .2778 ,2812 .3779 .4712 .6580 .5132
13.52 13.84 14.26 13.66 12.84 12.43 12.10 11.93 11.71 11.56
Series I11 above m.p. 194.32"K. T , OK.
CP
T , OK.
CP
T, OK.
CP
196.82 47.21 223.36 49.97 251.63 52.76 201.08 47.66 228.55 50.56 257.78 53.31 205.86 48.22 234.14 51.12 264.00 53.83 211.08 48.70 239.90 51.66 270.16 54.32 217.74 49.43 245.65 52.21 a Temperature rise from thermocouple. * Includes eu2.373. tectic heat A T =D
The Heat Capacities of (NH4)zO and NHaOH.-The heat of capacities of ammonium oxide containing 0.155 mole NHdOH, and ammonium hydroxide containing 0.33 mole % ' of (NH4)zO were measured and corrected for the foreign phases present. The correction was assumed additive in the case of the liquids and while this is only approximately
VOl. 75
D. L. HILDENBRAND AND W. F. GIAUQUE
2814
true the correction is so small that any error is negligible. The results are summarized in Tables II' and V. The entire range above 15'K. was investigated since each run began where the previous run terminated. The temperature rise in each run may be estimated from the spacing. All measurements made in temperature regions where there was an appreciable vapor pressure were corrected for vaporization heat by means of pressure observations on a 2-1" capillary manometer. The temperatures and volumes of the small gas space in the calorimeter and its connecting tube were known at all times. The densities of the liquids have been given above. The heat capacities recorded as C, in Tables I V and V are C(aat.vapor) except for the effect of a very small pressure of helium gas on the solid a t low temperatures. However, the correction to obtain Cp by means of the following equation is within the limit of error below a vapor pressure of one atmomhere.
1 atm.
Over the 60' regions below the melting points of both xH40H and (NH&O, the resistance thermometer showed evidence of a small amount of strain which was relieved when small fractions of the substances were melted. I t is for this reason that the measurements so designated in Tables IV and V were obtained in terms of the standard thermocouple. One possibility was that crystallization was accompanied by some slow internal rearrangement which caused expansion after the initial solidification. To check this point about 5% of the solid K&OH sample was melted and allowed to cool slowly by radiation. It required about 26 hours to resolidify the melted portion. Subsequent measurements showed that the strain persisted. The Anomaly h the Heat Capacity of ( NH4)20.-Figure 2 shows two maxima which occurred in the heat capacity of ammonium oxide. They are typical of the maxima which have been reported for other compounds containing hydrogen in S H I and other groups, such as NH4Cl,11a,h SH4-
S03,12(NH4)2S0412J3CH4.14 The energy absorption is believed to be due to an increase in angular motion and the narrow high maxima are typical of codperative molecular action. Other examples with more than one maximum are HBrl6 with t h e e , H P and NH4I1lbwith two. We have considered that angular motions in more than one crystal plane were a plausible explanation for multiple maxima. In the present case an additional possibility would appear to be present since the two KH4 groups may play different role? within the structure. Details of the heat measurements through the anomalous region are given in Table V. The short runs in Table V were designed to give the shape of the curve rather than high accuracy. The few long runs measure the total beat absorption through the region. They are given in Table VI.
TABLE VI HEATABSORBEDBY (XH&O IN THE ANOMALOUS REGION XEAR52 OK. Run
Temperature interval, OK. (made in series 1, Table V)
AH,cal. mole-'
1 2
49.385-52.607 52.612-56.248
51.5 44.8
(made after series 3, Table V)
3
49.946-56.354 Calcd. from Runs 1
+2
90.8
50-55
74.8
Calcd. from Run 3 50-55
74.5
The resistance thermometer showed a small amount of strain near the lower maximum but this disappeared when the transition was complete. There is evidently a volume change associated with the gradual transition. Runs 1 and 2 make it possible to calculate separately the amounts of energy in the two peaks above the minimum of heat capacity between them. This was desirable in connection with the entropy calculation. Runs 1 and 2 have been combined in Table V I for comparison with run 3 on the basis of the arbitrary interval 50-55'K. The substance had cooled to only 49' before run 3. The value calculated from runs 1 and 2 was accepted for calculating the thermodynamic properties because it was obtained in a continuous series from liquid hydrogen temperatures. However, the 0.3 cal. mole-' difference is probably within the experimental error.
The Heat of Formation of Ammonium Oxide from Ammonia and Ammonium Hydroxide.-The heat of the reaction (NH4),0(S) = NHIOH(s)
I
Fig. 2.-Anomaly mole-' degree-':
-
I
,
in heat capacity of (NH4)ZO; Cp in cal. 0 , series I ; A, series 11; m, series 111.
(11) (a) F. Simon, Ann. Physik, 68, 241 (1922); (b) F. Simon, 0. Simson and M. Ruhemann, Z . physik. Chcm. l S B , 339 (1927).
+ SHa(g)
(1)
was measured in the following manner. After the removal of the small increments of NH3 to prove that (T\;H&O was a pure solid compound, the remaining substance was solidified. It was eventually possible to ascertain its composition accurately after its complete removal from the calorimeter. Several heat capacity measurements were made on the solid at temperatures below the eutectic to eliminate conceivable supercooling or premelting as sources of error. The total heat required to convert all of the solid to liquid a t a temperature somewhat above the melting point of ( N H & 0 was then measured. This was followed by heat capacity measurements on the liquid at increasing temperatures until the vapor pressure of NHB reached a value slightly above one atmmphere. A measured amount of heat was then added to (12) J. L.Crensham and I. Ritter, ibid., B16, 143 (1932). (13) C.H. Shomate, THIS] O u R S A L , 67, 1096 (1845). (14) K. Clusius and A. Perlick, Z. physik. Chem., Ba4, 513 (1934). (15) W.F. Giauque and R. Wiebe, THISJ O U R N A L , 60, 2195 (192R). (16) TI', F. Giauque a n d R . Wiebe. ibid , 51, 1441 (1929)
June 20, 1953
2815
HEATCAPACITIES AND THERMODYNAMIC PROPERTIESOF (NH4)20 AND NHiOH
evaporate a considerable amount of ammonia a t constant pressure. The heat was added by means of the portion of the heater on the lower half of the calorimeter to ensure that ammonia gas leaving the calorimeter would not be superheated. This was absorbed in a glass weighing bulb containing 6 N sulfuric acid. The solution in the weighing bulb was protected by guard tubes against the transfer of water to or from the atmosphere. The ammonia reached the acid by means of a glass tip immersed in mercury to prevent the solution from being sucked back. A few measurements of vapor pressure were made as a function of composition and temperature. These were supplemented by temperature coefficients which could be cakuldted, with sufficient accuracy by means of the measured heats of vaporization. These data were used to determine the average temperature a t which the ammonia gas left the calorimeter. Following the heat of vaporization measurement the substance was resolidified and the heat of fusion and the heat capacity were again measured from below the eutectic temperature to the condition existing a t the end of the vaporization measurement. The change of heat content of ammonia gas with pressure a t constant temperature was calculated by means of the tables of the U. S. Bureau of Standards1’ which were extrapolated below their lower limit of about one third atmosphere. The heat capacity of the ideal gas was calculated from the spectroscopic observations on ammonia gas summarized by Herzberg.lB A small correction was made for the effect of rotational stretching by means of the equations given by Wi1s0n.l~ These data gave the heat of Reaction 1 a t the eutectic temperature. Essentially the method uses the pure solid phases in the eutectic as references for heat content. Since heat capacity data were available on all substances concerned the heat of the reaction (NH4)20(1) = NHdOH(1)
+ NHl(idea1 gas)
(2)
could be calculated a t higher temperatures. In calculating the value at 25’ it was necessary to extrapolate the heat capacity data on (NH4)23 above 273’K. but this should not cause appreciable error. The data are summarized in Table VII. TABLE VI1 HEATOF THE REACTION (NHa)zO(pure solid) = NH40H(pure solid) -t NH,(ideal gas) Moles at start
Run (NH4)zO N&OH
Moles T NHI vaporiz. removed av.
AH, mole-, NHdOH formed
1.7296 0.2124 0.4255 255.5 AHis7.40K. = 7532 ,3085 258.2 AHisT.iox. = 7523 2 1.3041 ,6379 Run 1 was given no weight AH of Reaction 2 Values from this research A H O W . = 5863 AHQF,OC. = 5598 Scatchard, et aZ.,20plus Bureau of Standards Tables16 AHooc. = 5879 A H Z ~ O=C5609 1
(17) U. S. Bureau of Standards Circular No. 142, April 19 (1923). (18) G.Herzberg, “Infrared and Raman Spectra,” D. Van Nostrand Co.. Inc.. New York. N. Y.. 1945,PD. 295,437. (19) E. B. Wilson, Jr., J . C h m P h y s . , 4, 636 (lQa6).
During run no. 1 the heat leak into the calorimeter from the surrounding copper block seemed to be larger than the normal rate both at the beginning and end of the run. This could have been due to a small delayed heat effect from incomplete mixing within the calorimeter. On account of this small uncertainty Run 1 is given no weight, although it agrees rather well with Run 2. A very small amount of water left the calorimeter with the ammonia. The tables of Scatchard, Epstein, Warburton and Codym were used to make a correction of -0.7 cal. mole-’ in Run 2 for the weight and heat of vaporization of the water. The correction in Run 1 was negligible. The values of A H in Table VI1 attributed to Scatchard, et a1.,20 are based primarily on the calorimetric observations of Zinner.21 They were calculated by combining the heat of mixing of the pure saturated liquids with data from the tables of the properties of pure ammonia given in the U. S. Bureau of Standards Tables,” and the properties of water from the tables of Keenan and KeyesZ2 Heat Capacity of Some Aqueous Ammonia Solutions.Incidental to the measurements of the heat of reaction discussed above, the heat capacity of aqueous ammonia at three concentrations was measured. As mentioned above, there was some evidence of incomplete mixing in these measurements, particularly a t the concentration 0.6411 weight fraction of ammonia, however, any such effect should have had little influence on the heat capacity. The observed results are given in Table VIII. They should be accurate to 0.1%.
X
TABLE VI11 HEATCAPACITY OF AQUEOUSAMMONIA = weight fraction NH3 in liquid, Cp in cal. deg.-’ g.-l X = 0.6411 X = 0.6123 X = 0.5883
T,OK. 201.56 207.30 213.40 219.35 225.14
0.910 .923 .935 .947 ,958
....
...
237.26 243.37 249.63 256.22
.981 .992 1.005 1.016
Cp
T , OK.
201.30 207.13 213.35 219.76 225.87 232.04 238.18 244.18 250.31 256.64
Cp
0.896 ,908 ,921 .934 .947 ,960 .972 .983 ,996 1.008
T , OK.
Cp
201.33 207.30 213.72 220.14 226.42 232.63 238.82 245.09 251.46 257.91
0.883 .897 .909 ,924 .937 .951 ,964 ,977 ,991 1.005
Table I X contains values taken from smooth curves through the data. These values, together with those corresponding to the concentrations X = 0.48596 ((NH4OH). and
TABLE IX HEATCAPACITY OF AQUEOUS AMMONIA FROM 200 TO 260 OK. X = weight fraction NHa in liquid, Cp in cal. deg.-l g.-I X = 0.6411 X = 0.6123 X = 0.5883 T,OK. 200 210 220 230 240 250 260
CP
CP
0.907 ,928 .948 .967 .987 1.005 1,023
0.893 .914 ,935 .955 ,976 .995 1.014
CP
0.880 .902 ,923 ,945 ,967 ,988 1,009
(20) G. Scatchard, L. F. Epstein, J. Warburton and P. J. Cody, Refvig. Eng., 58, 418 (1947). (21) K. Zinner, 2. gssamte KUte-Ind., 41, 21 (1934). (22) J. K. Keenan and F. G. Ryes, “Thermodynamic Properties of Steam, Including Data for the Liquid and Solid Phases,‘’ John Wiley and Sons, Inc., New York, N. Y., 1936
X
= 0.65407 ((NH&O) at the same temperatures, serve to define the heat capacity as a function of composition between those compounds.
Smooth curves through the data for ammonium oxide and ammonium hydroxide are believed to represent the heat capacities to 0.1-0.2'% a t all temperatures above 35'K. except for the short runs in the region of the anomaly in (NH&O. The error may be 1% a t 20'K. and several per cent. a t l5OK. principally because of decreasing sensitivity of the resistance thermometer. Values taken from the curves a t even values of the temperature have been used to obtain the thermodynamic properties of ammonium oxide and ammonium hydroxide which are given in Tables X and XI. TABLEX THE THERMODYNAMIC PROPERTIES OF AMMONIUM OXIDE, CAL.MOLE-1D E G . - ~ (Ho -(Fo CP S H8)/ T H8/T T,O K . 0.291 0.103 1.085 0.394 15 .618 .229 2.151 .847 20 ,411 1.453 1.042 25 3.344 1.534 .643 4.649 2.177 30 2.079 .920 6 075 2.999 35 2.674 1.236 7.666 3.910 40 1.588 4.905 3.317 45 9.683 4.039 1.973 6.012 50 12.032
-
55 60 70 80 90 100 110 120 130 140 150 160 170 180 190
Vol. 75
AND W. F. GIAUQUE D. L. HILDENBRAND
2816
11.776 11.388 12.491 13.873 15.246 16.592 17.889 19.108 20.268 21.392 22.485 23.562 24.616 25.654 26.667
Transition region 7.444 5.032 8.439 5.565 10.269 6.468 12.027 7.307 13.741 8.113 15.417 8.895 17.060 9.654 18.669 10.391 20.24.5 11.107 21.788 11.802 23.301 12.477 24.787 13.137 26.247 13.781 27.683 14.411 29.098 15.031
194.32 Melting point 200 47,552 43.168 210 48.603 45.513 220 49.659 47.798 230 50.695 50.029 240 51.685 52.208 250 52.617 54.337 260 53.491 56.418 270 54.310 58.452 280 55.075" 60.442 290 55,795" 62,387 298.16 56.355" 63,942 300 56.477" 64,290 a Extrapolated above 273°K.
27.962 28.920 29.838 30.723 31.576 32.400 33.194 33.961 34.702 35.417 35.983 36.108
2.412 2.874 3.801 4.720 5.628 6.522 7.406 8.278 9.138 9.986 10.824 11.650 12.466 13.272 14.067 15.206 16.593 17.960 19.306 20.632 21.937 23.224 24.491 25.740 26.970 27.959 28.182
The Entropies of Ammonium Oxide and Ammonium Hydroxide.-The entropy calculations are summarized in Tables XI1 and XIII. In obtaining the entropy in the anomalous region of heat capacity of (NH&O the calculation was made graphically except for the amount repre-
TABLE XI THE THERMODYNAMIC PROPERTIES OF AMMONIUM HYDROXIDE, CAL.D E G . - MOLE-' ~ T,OK. 15 20 25 30 35 40 45 50 55 60 70 80 90 100 110 120 130 140 150 160 170
180 190 194.15 200 210 220 230 240 250 260 270 280 290 298.16 300
CP 0.419 0.924
5
0.140 .325 1.533 .596 2.185 .933 2.860 1.320 3.536 1.747 4.195 2.201 4.&15 2.679 5.485 3.170 6.114 3.673 7.333 4.708 8.486 5.763 9.583 6.826 10.643 7.891 11.646 8.953 10.011 12.650 13.603 11.062 14.531 12.104 15.441 13.137 16.344 14.221 17.245 15.181 18.142 16.192 19.030 17.197 Melting point 28.566 26.530 29.474 27.946 30.380 29.338 31.285 30.708 32.187 32.059 33.081 33.391 33.957 34.706 34.807 36.004 35.624 37.285 36.400 38.549 37.022 39.567 37.134 39.795
(HaH$/T
0.104 .242 .439 .675 .938 1.222 1.515 1.818 2.121 2.427 3.042 3.651 4.249 4.836 5.410 5.973 6.524 7.063 7.591 8.168 8.621 9.125 9.623 18.210 18.725 19.234 19.738 20.238 20.734 21.226 21.714 22.197 22.673 23.057 23.143
-(Fo
- Hi)/ T
0.036 ,083 .157 .258 .382 .525 .686 .861 1.049 1.246 1.666 2.112 2.577 3.055 3.543 4.038 4.538 5.041 5.546 6.053 6.560 7.067 7.574 8.320 9.221 10.104 10.970 11.821 12.657 13.480 14.290 15.088 15.876 16.510 16.652
sented by the peaks above C, = 12.32 cal. deg.-' mole-'. In this case the center of gravity of 1/T was estimated and used along with the total energy TABLEXI1 ENTROPY OF AMMONIUM OXIDE,CAL.DEG.-' MOLE-1 0-15°K. Debye extrapolation 0.394 15-194.32°K., graphical 29.308 Fusion 2352/194.32 12.104 17.279 194.32-273.16 OK., graphical Entropy of liquid at 273.16"K. 273.16-298.16 OK., graphical extrap. Entropy of liquid at 298.16"K.
59.085 4.857 63.94
TABLEXI11 ENTROPY OF AMMONIUM HYDROXIDE, CAL.DEG.-I MOLE-' 0-15 OK., Debye extrapolation 0.140 17.472 15-194.15 OK., graphical Fusion 1568/194.15 8.076 10.722 194.15-273.16 OK., graphical Entropy of liquid at 273.16OK. 273.16-298.16 OK., graphical
36.410 3.157
Entropy of liquid at 298.16 OK.
39.57
June 20, 1953
HEATCAPACITIES AND THERMODYNAMIC PROPERTIES OF (NH4)20AND NH40H
above the arbitrary cut off point. This energy could be obtained accurately from the long runs through the anomalous region. Other entropy values which will be needed for the comparisons to be made below are
For Reaction 4
H20(1), S~08.16= 16.72, Sz7a.is = 15.14 cal. deg.-l mole-' NHa(g$ = I), Szos.te = 46.03, Szra.ia = 45.29 cal. deg.-l mole-'
NHdOH(1)
The entropy and heat content of ammonia gas were calculated from the spectroscopic data summarized by Herzberg. The natural constants used are those given by Du Mond and Cohen and listed by Rossini, Gucker, Johnston, Pauling and The moments of inertia of the ground state of the ammonia molecule, which were calculated in terms of the above constants, are the average of those obtained from the inversion doublets of the rotational levels. 11 = 12 = 2.815 X lo-" and 18 = 4.435 X 10-@ g.cm.9 The vibrational constants used were vl, 3335.9
Data from Bureau of Standard's Tables"
2817
AFP5o = 1298 AH250 -5032 ASSO = -21.230
Combining Reactions 3 and 4
+ NHa(g,f = 1)
(NHdzO(1)
(6)
AFz5 +864 AH25 = -5609 AS25 = -21.705
If we accept the above value of AF for Equation 6 and combine it with the heat of reaction measured in the present research, A H 2 5 = - 5598, the entropy change, AS26 = -21.67 cal. deg.-' mole-'. Applying the Third Law 63.94
- 39.57 - 46.03 = ASzs = -21.66
cal. deg.-l mole-'
The almost exact all around agreement is somewhat fortuitous. The free energy and entropy of the reaction HzO(1)
+ NHa(1, sat.) = NH40H(l)
(7)
may also be obtained by graphical interpolation in the tables of Scatchard, et aLZ0 For Reaction 7
(1/2), 3337.5 (1/2); ~ 2 931.58 , (1/2), 968.08 (1/2); v3, 3414 (2); and ~ 4 1627.5 , (2) cm.-'. Harmonic = - 1527 cal. mole-' oscillation was assumed since there is essentially AH26' = -2109 cal. mole-' no excitation above the states Y = 1, and that only AS250 = -1.953 cal. deg.-l mole-' in v2 and v4. Since the inversion doublets of the rotational states have a separation of only about combining the data of Equations 4 and 7 0.8 cm.-', equal distribution was assumed and HzO(1) NHa(g,f 1) NHdOH(1) (8) the symmetry number taken as three in calculating AFzbO = -229 AHzbo = -7141 the entropy of ammonia. The entropy and heat ASzjo -23.18 content calculation were corrected for rotational stretching of the molecule by means of the equa- Applying the third law tions given by Wilson. l9 39.57 - 16.72 - 46.03 = A& = -23.18 cal.deg.-l mole-' Comparison of the Third Law Data with EnAgain the exact agreement is somewhat fortropy Changes from the Free Energy Equation.tuitous since the tables of Scatchard, et a1.,20while The free energy entropy, and heat content changes carefully prepared, have utilized data which will in the reaction not ensure this degree of accuracy. That this is so NHdOH(1) NH3(l, sat.) = (NH&O(l) (3) may be shown by using the tables a t 0' to make the were obtained by graphical interpolation from the comparisons a t that temperature. tables of Scatchard, et aLZ0 These tables have For Reaction 7 been based on available data on the liquid-vapor Table of Scatchard, et ~ 1 A. F p~ = -1559 composition and vapor pressure of the ammoniaAHoQ = -2122 water system, together with calorimetric observaASoo -2.06 tions of Zinner2' on the heat of mixing liquid For Reaction 4 ammonia and water. Data from Bureau of Standards' Tables'? AF"0 = -751 Data relating to the reaction A H O " = -5300
+
+
NHa(g,f = 1) = NHa(1, a t . )
(4)
ASoa = -22.15
Combining Equations 4 and 7
were obtained from Tables of the U. S. Bureau of HzO(1) NHa(g,f = 1) = NHdOH(1) (8) Standards.17 The tables for ammonia require a AFoo = -808 cal. mole-' small correction below a pressure of 5 p.s.i.a. in AH00 = -7422 cal. mole-' obtaining data for the ideal gas state cf = 1). AS00 = -24.21 cal. deg.-l mole-' Since the heat of vaporization of ammonia is a Applying the Third Law large item in the heat of reaction we compared this quantity from the table with the more recent S N H < O H ( I ) - SH20(1) - &'H;(s,f-l) = 36.41 - 15.14 - 45.29 ASOO= -24.02 cal. deg.-l mole-' determination of Overstreet and Giauque.2 The discrepancy is 0.19 cal. deg.-' mole-'. NHa(1) = NHa(g) (5) Bur. Std' Tables" AH,.,. = 5580 mole-' The present heat capacity measurements should Overstreet and Giauque Azfb.,. = 5581 cal. mole-' lead to an entropy value of NH40H which is accurate to about 0.05 cal. deg.-l mole-'. For Reaction 3 The entropies of water and ammonia are accuTables of Scatchard, et alePo rately known, and the sign of the discrepancy is AF2s0 = -434 cal. mole-' such that either A S from isothermal considerations AH?s* = -576 cal. mole-I AS200 = -0.475 cal. deg.-l mole-' is too small, or C, d In Tfor NHdOH is too large. "fT (23) F D. Kossini, F. T Gucker. Jr., H. L. Johnston, L. Pauling Considering the limits of error this latter situation a n d G. W. Vinal, THISJOURNAL, 74, 2699 (1952).
+
D. L. HILDENBRAND AND W. F. GIAUQUE
2818
is thermodynamically inadmissible and the error must come from the tables which give AS00 = -24.21 cal. deg.-' mole-'. A similar test for Reaction 6 a t 0" is as follows For Reaction 3 Tables of Scatchard, et al. AFoo = -436 AH00 = -579 AS00 = -0.52
Combining the data of Reactions 3 and 4 for 0" N&OH(I) AFpo
+ 315cal. NKa(g,f = 1 ) = (NHd)$O(l) mole-' =
AH00 = -5879 cal. mo1e-I AS00 = -22 6'7 cal. deg -I mole-'
TABLE XIV COMPARISON OF CALORIMETRIC
HZSO - HOQ,cal. mole-'
(NHM(1) (NHIOH(1)
NHa(idea1gas) (6)
VOl. 75
AH250
- AH00
DATA
This research
Table of Scatchard, el at.
1387
1425 945
902 Spectroscopic data
Bureau of Standards tables
219.7 265
210.2 270
Reaction 6
Table XIV. Although the disagreement of the several terms is considerable the errors nearly cancel. However, a similar calculation for ReS(NHdrO(1) - SNH*OB(I) - S N H & . f = i ) = 59.09 - 36.41 action 8 involves the subtraction of the third from 45.29 the second entry and this involves a discrepancy A s p = -22.61 cal. deg.-l mole-' of 47 cal. mole-', which is equivalent to'an error This is excellent agreement again and yet the 0" of 0.17 cal. deg.-' mole-' in entropy a t 273'K. comparison for reaction 8 shows an internal inI t is evident that the Bureau of Standard's consistency in the combined tables of Scatchard, Ammonia Tables17 could be improved by making et and the Bureau of Standards.l7 them consistent with the thermodynamic properThe free energy of formation of ammonia-water ties of ideal ammonia gas, which are very accusolutions, from the components, in the range 0-25' rately known from spectroscopic data. should be accurate. *4 principal weakness in the thermodynamic data of the ammonia-water system is due to the fact that A F = R 7 f In f N H 3 dA'hs3 and a t ordinary temperatures either the pressures of Zinner*O determined the temperature c&cient ammonia and water are low or when the pressure of the heat of mixing by measuring the heats of of ammonia is high there is essentially no water mixing over a range of temperature. It is very vapor present. Thus the fugacity, f, may be cal- difficult to obtain accurate temperature coefficients culated from the properties of pure ammonia gas. by such a method. The preferred procedure is Most of the data are in the form of total pressures to measure the heat of reaction under the most and Scatchard, et al.,?@have made a small correction favorable conditions and then obtain the temperafor water vapor in terms of the Gibbs-Duhem equa- ture coefficients by means of heat capacity measurements. From the above comparisons it would tion. This should be an accurate procedure. In order to find the source of the discrepancy the appear that Zinner's measurements are fairly heat content tables of Scatchard, et al., were care- accurate near 2.5 O . Despite the minor discrepancies in available fully interpolated by means of graphical plots and tables relating to ammonia and the ammoniathe heat content values in the Bureau of Standards Ab.monia Tables were extrapolated graphically water system the over-all entropy agreement is from 5 p.s.i.a. to zero pressure. The changes in good and we believe that it is safe to conclude that heat content between 0 and 25" for NH40H and the entropies of ammonium oxide and ammonium (NH&O were compared with the present heat hydroxide both approach zero a t the absolute zero capacity results. The similar value for NHs(g,f = of temperature within 0.1 cal. deg.-' mole-'. l), was obtained from the limiting values of the The hydrogen bonding and all other structural Bureau of Standards Tables and compared with the details must be essential& ordered in the crystalline results of calculation based on the spectroscopic states of these substances at low temperatures. We thank W. P. Cox for assistance with the data given above. calorimetric measurements, and R. H. Sherman The comparison is shown in Table XIV. The sum of the second and third entries in Table and particularly E. W. Hornung for assistance XIV subtracted from the first entry gives AH250 - with the calculations. IIIoo for Reaction G. The result has been given in BERKELEY, CALIFORNIA Applying the third law
