Vor,. XX'XIV.
No. 3.
MARCH, 1912.
THE JOURNAL OF THE
American Chemical Society NINETEENTH ANNUAL REPORT OF THE COMMITTEE ON ATOMIC WEIGHTS. DETERMINATIONS PUBLISHED IN 191I . BY F. W. CLARKE. Received January 22, 1912.
During the year 1911a fair number of investigations relative to atomic weights appeared, some of them being of a fundamental character. The actual data obtained are, briefly, as follows : Chlorine.-Burt and Gray' continued their research upon the density of hydrochloric acid, confirming their earlier conclusions. For the weight of a normal liter of the gas a t o o , 760 mm., and the latitude of London they give the following figures : 1.63977 I .63999 I . 64007
I. 63999
I . 64016
I .64049
1.64037 I .64030
I I
.63982 ,64009
Reduced to latitude 45' this mean becomes
-I
Mean, I .6401 I ,63915, whence C1 = 35.460.
Chlorine and Potassium.-Since the time of Berzelius the analysis of potassium chlorate has been of fundamental importance in the deterrnination of atomic weights. The data, however, were discordant, and not in harmony with recent investigations. Now, with all modern precautions, Staehler and Meyer2 have reanalyzed the compound, with the subjoined results. Absolute weights are given. Chem. News, 103, 161, 170. Their former determinations appear in the report of this committee for 1909. Z.anorg. Chent., 71, 368.
F. W. C L A W .
226
PRELIMINARY. Weight KClOz.
Weight KCl.
Ratio.
12.38248 I I.28213 12.22480 11.52268 12 e44913
7.53218 6.86340 8.65366 7.00963 7.57331
1 .643943
I .643811 1.643791 1 ' 643835 I ,643816
Mean of the last four,
I .643813
Hence KCI
-
74.5558. The &st experiment was rejected.
FINAL SERE%. Weight Kclog.
Weight KCl.
Ratio.
6.24370 6.13362 6.10269 6.47073 7.33096
10.26355 IO ,08261 IO.OSI77 IO.6365I 12.05095
I.643824 I.643826 I.643828 I.643786 I.643842
Mean,
-
I
.643819
Hence, KCI 74.5551. From this the authors deduce K = 39.097,and C1= 35.458,'valuesin accord with the results obtained by other methods.
Sodiwm.-Some interesting analyses of sodium chloride and bromide are due t o Goldbaum.' The salts were electrolyzed with a mercury cathode and a weighed silver anode, and on the latter the halogen was fixed in weighable form. Omitting a preliminary series of analyses of the chloride, the following data, with vacuum corrections, are given :
SODIUMCHLORIDE. Welght NaCl. I
.oza34
Weight Cl.
At. wt. Ne.
0.62014
22.997 22.997 22.996
I .02221
0.62006
43474 1.46370 0.56934 1 .OO793 I .06501 2.16720 2.75219 0.92900 1'83527
I .47692 0.88789 0.34534 0.61141 0.64600 I .31460 1 * 66939 0 * 56349 I. 11324
2*
Computed with C1
THISJOURNAL,
3
35.458. If Na
33, 35.
-
22.995
22.999 22.995 22.999 22 * 997 22.999 23 .ooo 22.998
Mean, 22.997 23.00,C1 = 35.462.
REPORT OF THE COMMITTEE ON' ATOMIC WEXGRTS.
227
SODIUM BROMIDE. Weight NaBr.
Weight Br.
At. wt. Na.
1.05343 1 * 33360 1.95652 5.02976 2.09332 6 46697 5.54733 7.03901
0.81803 I .03561 1 ,51936 3.90586 1.62554 5.02178 4.30768 5.46606
22.998 22.997 22.995 22.997 22.998
Computed with Br = 79.92. If Na =
23.00,
23.000
22.999 22.998 Mean, 22.998 Br = 79.927.
Sulfur.-The ratio between nitrogen and sulfur has been determined by Burt and Usher,l through analyses of nitrogen sulfide, N,S,. The substance was decomposed by passing its vapor a t a red heat over quartz wool in a quartz tube, and from the volume of nitrogen so liberated its weight was computed. The following abbreviated table gives the essential data : Weight N&.
0.469455 0.442787 0.456326 0.470072 0.4669I 8 0.491307 0 ' 484307
Weight Ne.
Ratio, N/S.
0.142726 0.134627 0.138736 0.142919 0.141969 0.149380 0.147257
0.43685 0.43688 0,43684 0.43686 0.43690 o .43688 0.43690
Mean, o ,43687 Hence, if N = 14.009,S = 32.067. In short, the new ratio confirms the accepted values for both nitrogen and sulfur.
Calcium.-Richards and Honigschmid2 have analyzed calcium chloride, and confirmed their former determinations of the atomic weight of calcium. The ratio 2Ag : CaC1, was measured, by the usual Harvard methods, with the results given in the next table. Vacuum weights are stated, and the reductions are based upon Ag = 107.88 and C1 = 35.457. Weight CaC12.
Weight Ag.
At. wt. Ca.
4.60350 4.82401 4.81846
8 ' 94908
40.075 40.074 40.076 40.076 40.073 40.073 40.072
9.37780 9.36688
5'29799
10.29911
5,40550 5.24539
10.50832 IO. 19715
5.34110
10.38328
Mean, 40.074 40.074. The earlier bromide analyses gave Ca = 40.070. Proc. Roy. Sot., 85A, 82.
Hence, Ca
=
THISJOURNAL, 33, 28.
E'. W. CLARKE$.
228
A very rough determination of the atomic weight of calcium is due to Oechsner de C0ninck.l He ignited calcium formate, and in four experiments found values for CaO from 55.94-56.11. I n one other experiment he dissoll-ed calcium carbonate in hydrochloric acid, precipitated the lime as oxalate, and finally ignited the latter. The mean of his five discordant values is Ca = 40.02. Cadmium.-In order t o determine the atomic weight of cadmium, Perdue and Hulett2 have analyzed the hydrated sulfate electrolytically. The salt was electrolyzed over mercury, in which the liberated cadmium dissolved. The water of the sulfate was also determined a t temperatures between 670' and 700'. The data, with vacuum weights, are as follows: PERCENTAGE OF W A T E R .
Weight H 2 0 .
Weight CdS0,.8/3H20.
Per cent. 1 1 2 0
1856 I . 25986 I . 2886
6.32863 6.72493 6 ' 87537 5.65027 6.81125 7.34977 7.74837 7 ' 8843 6.6100
18.734 18.734 18.742 18.7?9 18.727 18.736 18.727 18.734 18.730
I.
I .05822 I
.27557
I . 37703
1.3572 1,477'3 I . 2480
Mean, 18.733 PER a w r . OF CADMIUM ~ i CdS0,.8/,H,0. i Weight sulfate.
Weight Cd.
Per cent. Cd.
7.90902 9.07468 7.32787 6.48847 5.11684 8.02954 5 .os743
3.46335 3 ' 97434 3.20936 2.84186 2.24157 3.51755 2.22827
43 ' 790 43.796 43.796 43.799 43.808 43.807 43.599
Mean, 43.799 Weight CdSO,.
P B R CENT. Cd I N CdSO,. Weight Cd.
5,14303 5 ' 46507 5.58677 5'53568 6.29717 6.40718 5.37196
2.77196 2.94566 3 .or076 2.98276 3.39295 3.45255 2.89457
Per cent. Cd
53'897 53.898 53.891 53.883 53.880 35.887 53.883
Mean, 53.888 Compt. reiid., 153, 1579. j . Physic. Ckem., 15, 147.
REPORT OF THE COMMITTEE ON ATOMIC #IGHTS.
229
From these data, when S = 32.07 and H = 1.008, the authors deduce Cd = 112.30, a Galue lower than that generally accepted. These new determinations have been criticized by Richards,‘ who suggests t h a t the cadmium sulfate possibly contained an excess of water in “solid solution.” Hulett is continuing his research with other cadmium compounds, and therefore judgment may well be suspended until the new evidence is published. Iron.-Atomic weight redetermined by Baxter, Thorvaldson and Cobb2 from analyses of ferrous bromide. The figures obtained are as forlows, with all corrections applied : PRELIMINARY SERIES. Weight FeBrZ.
3.45339 3.04933 2.9007 3.0873 3.50278 4.05239 4.085 IG
Weight Ag.
At. wt. Fe.
3.45481 3.05055 2 .go19 3.0885 3.50426 4.05404 4.08683
55.840 55.840 55.839 55.844 55.837 55.840 ‘55.840
Mean, 55.840
Weight AgBr.
At. wt. Fe.
6.01358 5.31029
55.853 55 * 844
6.10033 7.05752
55.831 55.831
....... .......
......
...... ......
.......
Mean, 55.840
FINALSERIES. Weight FeBrl
Weight Ag.
At. wt. Fe.
Weight AgBr.
At. wt. Fe.
5.03555 6.06309 5.59258 5.89767 4.48546 5.41562 6.50002 3.56564 5.32434 6’38845 6.37952 8.51818
5.03744 6.06557 5.59482 5.90014 4.48742 5.41799 6.50277 3.56719 5.32642 6,39094 6.38213
55.834 55.840 55.482 55.838 55.834 55.834 55.837 55.834 55.844 55.844 55’840
8.76950 10.55889 9.73974 10.27507
55.837 55.839 55.834 55.844
9,43171 11.31958 6.20987 9.27237 11.12536 I .I097 I 14.83468
55.830 55 * 843 55.829 55.839 55.842 55.844 55.836
.......
......
Mean, 55.838
........
......
Mean, 55.838
In a second paper Baxter and Thorvaldson3 give another series of determinations, like the foregoing, but starting with meteoric rather than terrestrial iron. The results obtained are esseptially the same, as is shown by the subjoined figures. THIS JOURNAL, 33, 888. Ibid., 33, 319. Ibid., 333 337.
c 30
F. W. CLARKE. Weight FeBrl.
Weight Ag.
3 ' 95460 4.66954 4.75335 6.95582 3.20762
3.9563 I 4.67177 4.75550 6.95854 3 * 20904
At. wt. Fe.
55.835 55.825 55.831 55'844 55.833
Mean, 55.834
Weight AgBr.
At. wt. Ft.
6.88720 8.13282 8 * 27855 12.11329 5.58632
55,831 55.818 55.824
55.844 55.830
Mean, 55.829
The authors reject the second and third of these pairs of determinations, leaving to be accepted the means 55.837 and 55.835. The calculations are based upon Ag = 107.88 and Rr = 79.916. Tantalum.-The determinations of this atomic weight by Chapin and Smith1 were made by t h e hydrolysis of the pentabromide. The weights given below are corrected to a vacuum. Br = 79.92, Weight T&rs.
0.86837 1.50903 I ,56554 1.23239 I .31815
Weight TanOs.
At. wt. Ta.
0'. 331 I 7
181.68 181.80 181.75 181.91 181.85 181.So 181.91 181.74
I .AI702
I . 20090 1 .0405O
0.39688
Mean, 181.80
This \value is higher than that previously found by Balkc, i 8 ~ . 5 ? , from similar analyses of tantalum pentachloride. Selenium-Kuzma and Krehlik2 have redetermined the atomic weight of selenium by reduction of SeO, with SO,. Special precautions were taken io secure a perfect reduction, and to avoid losses or impurities. Thr essential figures are as follows : Weight SeOl.
Weight Se.
A t wt.Se.
0.44245 0.61918 I .39106 0.66740 0,65154
0.31523 0.44122 0.99109 0.47544 0.46474 0.68417 0.86256 0 53760 0.27486 I .07594
79.290 79.338 79.292 79.257 79.255 79.253 79.243 79.249 79.302 79'249
0.96042 I . 2 1088
0.75468 0.38577 I . 5 I040
Mean, 79.273 Reduced to a vacuum standard, Se = 79.26. THISJOURNAL, 33, 1497. ?'ram. Bohemialz Acad. of Emperor Francis Joseph, 19, No. 13 (1910). In Bohemian. I am indebted for the details to the kindness of Professor Brauner, at whose suggestion the work was done.
REPORT OF THE COMMITTEE O N ATOMIC WEIGHTS.
23 1
Te1luriuna.--Harcourt and Baker' have criticized the work of Flint? upon the supposed complexity of tellurium, and conclude that the portions of low atomic weight which Flint obtained were contaminated by some impurity. They suggest that the basic nitrate employed in Flint's determination probably contained tellurium trioxide. Repeating his process of fractionation, they used the fourth fraction for atomic weight determinations, with the results shown below. The bromide method of Baker was employed. Weight Te.
Weight TeBq
0.87822 0.59706 0.69189 0.62732 0.58307
2.20103 I '49640 I . 73442 1.57254 I ,46162
At. wt. Te.
127.55 127.55 127.53 127.53 127.53 Mean, 127.54
I am informed that Flint is continuing his investigation, so that the question a t issue still remains open. The value now given by Harcourt and Baker agrees with that found by Baker and Bennett in 1907. Uranium.--In three very short notices Oechsner de Coninck3 gives approximate determinations of the molecular weight of UO,. First, UO,Cl, was reduced to UO, by heating in hydrogen. In mean, U 0 2 = 270.07. Similar reductions of UO,.H,O gave UO, = 270.66. Another series with ITO,.zH,O gave UO, = 270.46. The last value corresponds to U = 238.46. Iridiuiit.-Hoyermann' has determined the atomic weight of iridium by reducing (NHJ21rClb in a stream of hydrogen. His figures are subjoined, with deductions based upon H = 1.008, N = 14.01,and C1 = 35.46. Weight chloride.
Weight Ir.
At. wt. Ir.
.72348 1.77984 1.78837 I . 15161 1,73794
0.77205 0.77654 0.7801I 0.50249 0.75838
192.645 192.598 192.533 192.635 192.654
I
-_
Mean, 192.613
Holmium.-The atomic weight of holmiLtm has been determined by H ~ l m b e r g ,by ~ the well known sulfate method. His syntheses are as follows : Jouv. Chem. Soc., 99, 1311. A m . J . Sci., [4] 30, 209. Cited in this report for 1910.
Comfit. rend., 152, 7 1 1 , 1.179; 153, 63. Sit%.phj,r. m e d . Soz. Erlaltgen, 42, 278 2.anorg. Chem 71,2 2 6 .
RICHARD C. TOLMAN AND ALFRED L. FERGUSON.
232
Weight H o 2 0 ~ .
Weight HoZ(S04)3.
At. wt. Ho.
0.3467 0.3400 0.3960 0.7631 0.6877 0.5378
0.5687 0.5579 0.6496 I . 2524 I . 1286 0.8822
163.57 163.40 163.55 163.31 163.33 163.55
Mean, 163.45
Argon.-Fischer and Froboesel have made numerous fractional distillations of liquid argon, and found its density as gas to be practically constant. The final result is d . 19.94-19.95, and A = 33.9. Niton.-For the atomic weight of niton, the gaseous emanation of radium, Gray and Ramsay’ give determinations ranging from 2 I 8-22 7. The mean is 2 2 3 ; but the value N t = 222.6 is preferred. Miscellaneous Notes.-Hinrichs3 has reconsidered all the evidence relative to the atomic weight of hydrogen, and concludes that H = 1.00781. I n another paper4 he discusses the atomic weight of vanadium, which he places a t 51 precisely. A brief note by Ter Gazarian5 defends his work on the density of PH,. C. Henrys has considered the proper mode of calculating atomic weights. Relations between the atomic weights are studied by Loring,’ by Emersons and by Nicholson.@ Emerson’s “helix chimica” is an arrangement of the elements on a spiral, while Nicholson develops a structural theory of their formation.
THE FREE ENERGY OF DILUTION OF HYDROCHLORIC ACID. BY RICHARD C. TOLMAN AND ALFREDL. FERGIJSON. Received December 19, 1911.
I.
Introduction.
The free energy of dilution of an electrolyte is usually obtained from measurements of the electromotive force of concentration cells. I n the case of hydrochloric acid, apparently accurate measurements have been made by Jahn‘O on concentration cells of the type, Ag : AgCl HCl : HC1 AgCl : Ag.
c,
c2
Ber., 44, 92. a Proy. Roy. Soc., 84A, 536. *Rev. gln. chim., 13, 351, 377 (1910). Proc. Am. Phil. Soc., 50, 191. J . ckim. phys., 9, loo. 6 C. R . Assoc. F r a m . Avance Sci., 269 (1909). Phys. Z., 12, 1 0 7 . Am. C h m . J., 45, 160 (1911). Phil. Mag., [6] 22, 864. lo Jahn, 2. physik. Ch~nz., 33, 545 (1900);35, I (1900).
’
