I945
ATOMIC REFRACTION.
7. The following data are submitted. &at of inversion of sucrose by hydrochloric acid at 20' = 10.4 * 0.06 gram calories per gram. Heat of solution of sucrose in water a t 20' to ca. 4% sucrose concentration = 3.43 f 0.02 gram calories per gram. Heat of solution of sucrose in I .64molar hydrochloric acid, a t 20' to ca. 4% sucrose = 4.23 * 0.05gram calories per gram. Heat of solution, anhydrous a-glucose in water and in I .64molar hydrochloric acid, a t 20' to ca. 4% glucose = 13.9* 0.1gram calories per gram. NBWYORK CITY.
[CONTRIBUTION FROM
THE
POLYTECHNIC INSTITUTEOF WARSAW.]
ATOMIC REFRACTION. BY W. SWIENTOSLAWSKI. Received March 17, 1920.
In 191I, F. Eisenlohrl published new calculations of the refractivities of carbon, hydrogen, and other atoms in organic compounds. These calculations do not differ from those used by J. W. Briihl, because both authors assumed that the molecular refractions of the organic compounds are equal to the sums of the refractions of the separate atoms. MD = ZAD. According to this supposition, the computation of atomic refraction depends on the average value of the refraction of the CH2 group and those of certain atoms. If, however, the individual atomic refractions show even small deviations, this method cannot be used. In order to determine the limits of variations in the refractions caused by carbon and hydrogen atoms in the compounds which contain only the linkings C-C and C-H, I have used the following method of calculation. I . If we denote by rc, TH, and rCHl the refractions for the D lines of carbon, hydrogen, and the CH2 group, the molecular refraction, MD, of the hydrocarbon C,H, can be expressed by the following equation, MD = nrCHa (m - 2n)rU Z Arc C&H, (1) or M D = W C H z (m - 2 n ) r ~ 2 Ar, where 2 Ar = I:Arc zATH. Equation I cannot be solved when the increments I:Ar are not known; therefore, we can compute the quantities rc and TH only by reference to several chosen compounds, thus obtaining the average values of rc and r H on the supposition that I :Ar = 0.
+
+
+
+
+
Z. physik. Chem., 75, 605 (1911).
+
I946
W. SWIENTOSLAWSKI.
and In some cases it is more advantageous to use the quantities corresponding to refractions of the atomic linkings C-C and C-H, especially when it is desirable to compare these quantities with the thermochemical data. It is clear that the quantities rc-c and rC-H can be computed when rc and TH are known, for we have rc-c = o.5rc 7C-H = f 0.25rc. I have selected as the basis of calcuiation the following compounds, n-pentane, a-methyl-butane, n-hexane, n-octane, di-isobutyl, di-isoamyl, cyclopentane, and cyclohexane. Table I contains the values of the molecular refractions M D and the equations connecting rCH2, rII, and Z Ar. These equations were solved by the method of least squares. 2.
IC-H,
TABLEI. Name. I
z 3
4 5
6 7 8
Formula.
n-Pentane.. . . . . . . . . . . . . &-Pentane.. . . . . . . . . . . . n-Hexane.. . . . . . . . . . . . . . n-Octane.. . . . . . . . . . . . . . Di-isobutyl.. . . . . . . . . . . . Di-isoamyl. . . . . . . . . . . . . Cyclopentane. .......... Cyclohexane.. . . . . . . . . . .
M,.
Equation.
+ + + +
+ + + + +
CjH1, 2 j ,z3 jrCH2 zrH ZA r = z j . 23 CsHll z j . 2 5 jrCHz zrH Z A r = 25 . z j '&Hi4 29.84 6YC& 2TH Z A Y = 29.84 CBHIR 39. 16 81'CHz 2 r H ZA r = 39.16 39. I I 8f'CHz f 2rH Z A Y = 39. I 1 C8H18 Z A Y = 48.33 CloHzz 48.33 IOrCHy 2rH' Cj1310 23. IZ y C H 2 2 A Y = 23. 1 2 27.72 GrCHz Z A r = 27.72 C6H12
+ +
Upon the assumption that Z A r = 0, we obtain the average values of rC = 2.490, TH = I ,066, and rCHI = 4.622, for our 8 selected hydrocarbons. Table I1 contains a comparison of the observed and calculated values of M D . TABLE11. Name.
%-Pentane.............. iso-Pentane . . . . . . . . . . . . . %-Hexane............... %-Octane. . . . . . . . . . . . . . . Di-isobutyl.. ............ Di-isoamyl.. ........... Cyclopentane ............ Cyclohexane ............
M, (obs.). CJln 25.23 CBHn 25.25 29.84 C6H14 39.16 CBHIS CSH18 39.11 C ~ O H P B 48.33 GHio 23.12 27.72 C6H13
Formula.
(calc.).
25.24 25.24 29.86 39.11 39. 11 48.35 23.11 27.73
Ar. -0.01
4-0.01 -0.02 +o.oj 0.00 -0.02
A%.
-0.04 $0.04 -0.07 +0.13 0.00
-0.04
$0.04 -0.04 A% = *0.07
+O.OI -0.01
The small differences between the observed and calculated values show that Ar is very small, and that, therefore, it is very probable that the atomic refractions remain constant in the 8 compounds, at any rate within = t o .0 7 7 ~ . Comparing the values of rc, r H , and rCHz thus obtained with those of I?. Ejisenlok
I947
ATOMIC REFRACTION. 1C.
m.
F. Eisenlohr .... . . . . . . . . . . . . . . . . . 2.418 W. Swientoslawski.. , , . . , . . . . . . . . . 2.490
I . IO0
+3.0%
= -3.1%
ICHv
4.618 I ,066 4.622 Deviation % Deviation % Deviation % E
=
+O.Og%
we see that his value for rCHz is in sufficient agreement with mine; those for rc and rH differ widely ( = t 3.o%>. In order to explain why Eisedohr’s calculations of molecular refractions show a sufficient agreement with th; direct measurements, it should be observed that the value of MD for numerous organic substances arises from two terms MD = nrCHz 4- Z r x . The first term nrcEIz,is very large, while the second one, corresponding to the refraction of other atoms or groups in the molecule is always small. Therefore, small errors in the determination of the values of rc and YH, etc., do not exert any marked influence on the calculated values of MD. Yet the determination of the real values of rc and TH is of great importance for our science. As regards our computation, we can only maintain that the values found for rc and r H correspond to the average values of atomic refractions in the cases of the 8 compounds cited. By using a larger number of equations, for example, MD = %yCHz 2rH I:Ar for hydrocarbons C,H2,+2. containing only the linkM‘D = nrCHl I: Ar for hydrocarbons C,H2,. M”D = nrCHz- 2 r H Z A r for hydrocarbons CnH21t-2. ings C-C and D = nrCHz - 4?‘H 4- ZAr for hydrocarbons C,H2,+. C-H; and by introducing new hydrocarbons CnH2n-2, CnN2,--4, etc., it is probable that we might obtain slightly different values for rc, YH, and YCH~. Therefore, this problem cannot be considered definitely solved. In order to show that the refraction of carbon and hydrogen depends on the constitution of the hydrocarbons, in Table I11 are given the observed values of MD for several compounds and those calculated from the formula, MD = Zrc I::PH, where rc and YH are the average values of atomic refraction obtained from the 8 hydrocarbons cited. For this purpose the methyl derivatives of polymethylene hydrocarbons have been chosen. It is interesting to observe that all the deviations vary between 0 . 0 0 and +0.94%, excluding one case (1;I-dimethyl-cyclohexane, where A = -0. 05Yc according to Lange’s value). In some cases the deviations are very remarkable ; for example, in the case of methyl-cyclohexane, which has been investigated so carefully by Kishner and Eisenlohr, the considerable deviation of between + o s 34y0 and +o. 37Yc is observed.
+
+
+ +
I
+
1948
W. SWIENTOSLAWSKI.
Name.
Methyl-cyclohexane . . . . . . . . . . . . . . . Methyl-cyclohexane . . . . . . . . . . . . . . . I, I-Dimethyl-cyclohexane . . . . . . . . . . I, I-Dimethyl-cycloh6xane. . . . . . . . . . I ,2-Dimethyl-cyclohexane. . . . . . . . . . I ~-Dimethyl-cyclohexane. ......... I ,4-Dimethyl-cyclohexane. . . . . . . . . . Dihydro-laurolene . . . . . . . . . . . . . . . . . Dihydro-iso-laurolene . . . . . . . . . . . . . . I,z ,4-Trimethyl-cyclohexane. . . . . . . . 1,~,4-Trimeth~l-cyclohexane.. ...... 1,3,5-Trimethyl-cyclohexane.. ...... Pulegane.. ....................... Dihydro-campholene.. . . . . . . . . . . . . .
M,.
Author.
32.47 Kishner 32.46 Eisenlohr 37. 12 Perkin 36.96 Lange 37.02 Eykmann 37,26 Eykmann 37.28 Eykmann 37. 12 Perkin 36.99 Perkin 41.66 Eykmann C9H18 41.75 Eykmann CgHls 41.99 Eykmann C9H1.5 41.74 Eykmann C ~ H I S4 1 ~ 6 6Eykmann
nrCH2.
32.35 32.35 36.98 36.98 36.98 36.98 36.98 36.98 36.98 41.60 41.60 41.60 41 .Go 41.60
Ar.
A%.
$0.37 f 0 . 1 1 $0.34 fo.14 +0.38 4-0.12
-0.02
---o.oJ
+0.04
+O.II
f0.76 +0.30 f o . 8 1 f o . 1 4 +0.38 +O.OI +0.03 +o.o6 +0.14 $0.15 +0.36 + 0 , 3 9 +o. 94 +o. 14 $0.34 $0.06 $0.14 $0.28
The only possible conclusion to be drawn from the data of Table I11 is, that the refraction of carbon and hydrogen in the 8 selected compounds is not identical with that in the methylenic polymethylene hydrocarbons: the introduction of the CH2 group produces a small, but appreciable increase, Ar, in the molecular refraction. In another paper it will be shown that the methyl group exerts a similar eJect u p o n the heat of formation of the atomic linkings C-C and C-If. In the case of compounds which contain other atoms in the molecule (for example, oxygen) the results obtained from F. Eisenlohr's values differ from those obtained by me, and the differences are found to be greater, the larger the value of Z r x in the equation where nrCHzcorresponds to the refraction of the CH2 group and 2 r x to that of the rest of the molecule. For example, in the case oi ketones and aldehydes, we have Zrx = r o f 2 A r ,
where ro corresponds to the refraction of oxygen in the carbonyl group. In this case my calculations and those of Eisenlohr are in sufficient agreement because Z r x is very small. The calculations for ketones and aldehydes are given in Table IV. Eisenlohr assumes ro = 2 . 2 1 I , which value differs from mine (To = 2 . 1 9 6 ) by about $0.67%. On comparing the various values of r o in Table IV, considerable deviations from the average value, r o = 2 . 2 0 , are evident. These can be accounted for by supposing that the refraction of the carbonyl group, or of the oxygen, is variable and depends on the constitution of the molecule.
I949
ATOMIC Rl3FRACTION.
TABLEIV. Formula.
Name.
..... . . . . . .. . .. . . . . . .. . . . . . . . . . . .
Acetone , , . . , , ., Methylethyl-ketone. . . . . .. . .. . . .. Diethyl-ketone. ... Methylpropyl-ketone.. , . . ... Methyl-iso-propylketone.. . . . . . . . . . Ethylpropyl-ketone ....... . . . . . . . . . Methyl-iso-butylketone.. , . . , . ,.. . . Oenanthol ...... , . . , , . . . . . . . . . . . Methylhexyl-ketone. . . . . . . . . . . . . Methylnonyl-ketone . . . . . . . . . . . . . . n-Butyl-aldehyde ........ . . . . . . . . . . iso-Butyl-aldehyde., , . , . . . . . . . . . . . I
.
.
# . .
...
. . .
.
.
nrCEv
MD.
16.15 CHsCOCHs 20.67 CH3COC& 25.18 CZH&OC& 25.20 CH3COCaH7 CHaCOCaHi 25.24 C Z H ~ C O C ~ H ~29.71 30.01 CH3COC4Hg C1Hi40 34.79 CH3COCeHis 39.28 CHsCOCsHie 53,OO 20.64 C4HsCOH 20.68 CdHgCOH Average, y o
13.87 18.49 23.11 23.11 23.11 27.73 27.73 32.35 36.98 50 84 18.49 18.49 I
i o 4-
Z Ar.
2.28
2.18 2.07 2.09
2.13 I .98 2.28 2.44 2.30 2.16 2.15
2.19
+ 2 A r = 2,196
This can be proved, for if we assume that Ar = 0, we may calculate the values of rcI12 and yo by the method of least squares from the series of equations M D = nrCH, ro 2Ar. Thus, we obtain r C H z = 4.631 and r o = 2,137. The differences between M D and the calculated values (1zrCI12 y o ) are given in Table V.
+ +
+
TABLE V. Name.
Acetone... ....................... Methylethyl-ketone.. . . . . . . . . . , , . , . Diethyl-ketone . . . . , . . . . . . , . , . . , , . , Methylpropyl-ketone. . . . , . . . . , , . . . Methyl-iso-propylketone. , , , . , , . . . . Methyl-iso-butylketone., , . , , , . , . , Ethylpropyl-ketone.. , , , , . . . . . Oenanthol ......... . . . . , . . . , , . . ... Methylhexyl-ketone. . . . . , , , . , , , , . , Methylnonyl-ketone.. . . . . . . . . . , , n-Butyl-aldehyde.. . . . . . . . . , . . . . , iso-Butyl-aldehyde. . . . . , . . . . . , , , , ,
.
.. . . . .
..
.
M,.
n7cH2 + y o .
A%..
A7.
16.15 20.65 25.18
16.03 20.66
+o.rz
+0.74
-0.01
-0.05
25.29
+.II
-a
25.20
2 j.29
-0.09
-0.35
25.24 30.01 29.71 34.79 39.28 53.00 20.64 20.68
2 j,229 -0.05 29.92 +d.09 29.92 -0.21 34.55 +0.24 39.19 +o.og 53.08 -0.08 20.66 -0.02 20.66 So.02 Average, A% =
44
--a.20
+o.30 -0.70
+0.70 +0.23 -0 15 I
-0.
IO
$ 0 . IO
+0,34
A comparison of the percentage deviations in Tables I1 and V shows that the refraction of the carbonyl group or of carbonyl oxygen varies within wide limits, and that the deviations cannot be explained as being due to experimental errors. If we examine the following data
I950
W. SWIENTOSLAWSKI.
we observe that the replacement of a methyl group by an ethyl or by a n-propyl group is accompanied by a considerable diminution in the value of y o . These diminutions are almost identical in all cases and vary within narrow limits (-0. og and -0. I I ) . Likewise, we must note that the methyl group, when it is substituted for hydrogen in aldehydes, effects a very noticeable change equal to +o.og, when the rest of the molecule is iso-C4H&O, for example. The data of F. Eisenlohr which concern the atomic refractions of ether compounds, are not in agreement with those computed above. These variations pertain to the alcohols, ethers, and esters. This paper is not intended to give complete discussion of the question, but only to demonstrate by some example that the exact determination of the refractions rc and YH is indispensable in all calculations which deal with this problem. If we calculate values for M D for the alcohols and ethers by the formula M D = nrCHz 2 r H ro 4- Z Ar it is clear that an error in the value assumed for YH causes an error in the value of yo. Table VI contains the calculated values of yo assuming, as above, that rCH2 = 4.622, rH = I .066, and that Z Ar = 0.
+
+
TABLEVI. Name, alcohol.
Methyl., .................... Ethyl ........................ n-Propyl.. . . . . . . . . . . . . . . . . . . . iso-Propyl. . . . . . . . . . . . . . . . . . . . %-Butyl...................... iso-Butyl ..................... Trimethyl-carbbol.. , , . , . , . , . . &-Amyl ..................... iso-Amyl (ferm.) . . . . . . . . . . . . . . n-Heptyl.. . . . . . . . . . . . . . . . . . . . Methylhexyl-carbinol . . . . . . . . . .
The average value, ro =
I
Formula.
M,.
CHaOH CzHsOH CaHjOH C3H7OH C4H90H C4HsOH C4HsOH C5HllOH C~HIIOH C~HI~OH CsH1jOH
8.22
niCHl
12.74 17.52 17.54 22.13
+ 2rH. + Ar.
22.16
20.62 20.62
22.22
20.62
26.74 26.77 36.05 40,56
io
1.47 1.36 1.52 1.54 1.51 1.54 1.60
6.'75 11.38 16.00 16.00
25.24 I .50 25.24 I ,53 34.49 1.56 39.11 1.35 Average, = I ,494
,494, differs from that of Eisenlohr
( i o=
1.525) by about +2.03%.
In Table VII, the data for the refractions of oxygen in the ethers are given. TABLEVII. Name.
Methylal. . . . . . . . . . . . . . Acetal. . . . . . . . . . . . . . . . Ethylpropyl ether. . . . . . Ethyl ether,. . . . . . . . . . .
M,.
Formula.
19.19 33.13 26.95 22.43
CHs.O.CHz.0.CHs (CHs.CHzO)g.CH.CHa CzH6.0.CsHr CzH5.O.CzHs
The value obtained by I?. Eisenlohr is by about -1.2%.
I
+
w C H Z27,.
ro
+ Z: Ar. .60
16.00
I
29.86 25.24
I .64 I .71
20.62
I .SI
Average, =
I . 663
,643, which differs from
I . 663
THE MEASUREMENT OF DISSOCIATION PRESSURES.
1951
It is to be noted that quite aside from this disagreement, the values of rc, rH, T o , etc., are not constant. The calculation of average values of atomic refraction cannot be accepted as a satisfactory solution of the question before us. In another paper I shall compare these results with the thermochemical data. In comparison I shall attempt to justify the selection of the 8 hydrocarbons as a basis €or the calculation of the average values YCH2, rc and r H . WARSAW,POLAND.
[CONTRIBUTION FROM THE DEPARTMENT O F CHEMISTRY OF PRINCETON UNIVERSITY. ]
THE EXPLANATION OF AN APPARENT ANOMALY OUTSTANDING IN T H E RESULTS OF MEASUREMENT O F DISSOCIATION PRESSURES. BY ALAN W. C. MENZIES. Received Mrrch 29, 1920.
After brief reference to the nature of the anomaly in question and to the various explanations offered to account for it, it is proposed to adduce experimental results, supported by results of a parallel nature drawn from the work of others, to show that this apparent anomaly is disposed of by a knowledge of the real facts of the case. The Nature of the Anomaly.-In 1888, Tammann,' applying a form of the gas-current saturation method to the measurement of the dissociation pressures of salt hydrates near 35 O, obtained results which, while somewhat erratic,2 were uniformly higher by from 2 to 5% than the results obtained by Frowein3 with the tensimeter. This anomaly was confirmed by S ~ h o t t k y working ,~ under Nernst's guidance, who found that the initial dissociation pressures developed in tensimetric measurements were higher than the equilibrium values. In 1911,Partingtons added further confirmation, again using the gas-current saturation method, although not in a form identical with Tammann's. Explanation of the Anomaly.-Thoughtful elucidations and critical discussions have been offered by Tammann,6 Nernst,' Partingtoq6 Brereton Baker,* and C a m ~ b e l lthose , ~ of Nernst and of Campbell being especially instructive. Lack of space forbids their outlining or consideration here. 1
Tammann, Ann. Physik., 33, 3 2 2 (1888). Cf.Menzies, THISJOURNAL, 42, 978 (1920). Frowein, 2. physik. Chem., I , 5 (1887). Schottky, ibid., 64, 415 (1908). Partington, J . Chem. SOC.,gg, 466 (1911).
6 LOC. 8
cit.
Nernst, 2. physik. Chem., 64, 425 (1908). Baker, Ann. Rep, Progress Chem., 8, 34 (1912). Campbell, Trans. Faraday SOC.,IO, 195 (1914).
