WALTERF. EDGELLAND CURTISH. WARD
6486
the order of the shift of the 0-H and N-H frequencies is C1 > N > 0 > F. As is shown in Table V, this is in good accord with the order of the van der Waals radii of these atoms, The above result is not to be expected from a point charge type of model, and suggests that if Y is large and polarizable the important effect a t large distances is an overlap of the outer electrons of Y with the proton. Both the very low frequencies found in strong hydrogen bonds and the recent neutron diffraction data (Fig. 8 and Table VI) indicate that in strong hydrogen bonds the X-H distance increases exponentially as the X-H . . . Y distance decreases. This seems to be in serious disagreement with semitheoretical treatments of hydrogen bonding, such
1.10
Vol. 77
as that of Coulson and Danielsson.l4 In several recent treatmentsl5 i t is either assumed or found that the amount of covalent character in the long H . . .Pbond is negligible, and it seems possible that these treatments fail to account for the rapid increase of the X-H distance in strong hydrogen bonds because they grossly underestimate the covalent character in the H . . .Ybond as Y approaches the hydrogen. I n any case, the bad disagreement with hydrogen positions in strong hydrogen bonds does not justify faith in recent theoretical treatments. TABLE VI THE RELATIONBETWEEN 0-H.. .O DISTANCEAND 0-H DISTANCE 0-H., .O Compound distance (A,) ( C O O H ) ~Z H ~ O 2 . 8 5 5 CaSOr.ZH20 2.70b a-HIOa 2.686 o 0094 RDzAsOp 2.52“ 2.525 KH2AsOi (COOH)22H*O 2.518 i O.OOSa KHzPOa 2.487 0.005~
0-H hfethodf Distance (A,) 0.96a N.D. P.M.R. 0.98C N.D. 0.990 0.0175 N.D. 1.03‘ N.D. 1.06‘ N.D. 1.057 f 0.016’ 1.085 -i 0 . 0 1 3 ~ N.D.
*
B. S. Garrett, Oak Ridge National Laboratory Report, No. 1745. * W . A . Wooster, 2. Krist., 94, 375 (1936). G. E. G. E. Pake, J. Chem. Phys., 16, 327 (1948). Bacon and R . S. Pease, Proc. Roy. SOC.(London), A220, 397 (1953). S. W. Peterson, H. A. Levy and S. H. Simmonsen, J . Chem. Phys., 21,2084 (1953). f N. D., neutron diffraction; P.M.R.. proton magnetic resonance.
?
5. Q)
2 1.05 G
Figure 8 suggests that for hydrogen bonds of -2.45 8.the hydrogen will be centered, tending to confirm that the hydrogen bond in nickel dimethylglyoxime, where 0-H-0 = 2.44 A., is symmetric. Acknowledgment.-The authors are indebted to Mr. Richard Hedges and Mr. Robert Kross for aid in obtaining spectra and to the Ames Laboratory of the Atomic Energy Commission for use of equipment. The Corn Industries Research Foundation has kindly provided a fellowship for one of us (K. N.).
M
2
1 .oo
0.95 2.5 2.6 2.7 2.8 2.9 3.0 0-H. . .O distance (A,). Fig. 8.-O-H . , . 0 distance oerszis 0-H distance 2.4
(14) C. A. Coulson and U. Danielsson, Arkiv F y s i k , 8 , 239, 245 (1955). (15) For example, N. D. Coggeshall, J . Chem. P h y s . , 1 8 , 9 7 8 (1950). AMES, I O W A
[CONTRIBUTIOSFROM THE DEPARrXEST O F CHEMISTRY, PURDUE UNIVERSITY]
The Raman and Infrared Spectra of Tin Tetramethyll BY WALTERF. EDGELL AND CURTISH . WARD^ RECEIVED FEBRUARY 8, 1955
The Raman and infrared spectra of Sn(CHd4have been studied and compared Tvith the results of earlier investigations. The fundamental frequencies of vibration are determined and the large number of combination and overtone bands in both spectra are explained in terms of them. The coupling between the skeletal vibrations and those modes characteristic of the methyl groups is discussed and the results compared to those for the other tetramethyl compounds of Group IVA elements.
Introduction The Raman and infrared spectra of several organotin compounds have been studied recently in this Laboratory. Such a study is of theoretical interest in connection with the nature of “group frequencies” and in particular with regard to the extent of coupling between the skeletal modes and (1) Based upon a portion of the Ph.D. thesis of C. H . Ward, Purdue University, September, 1953. (2) Purdue Research Foundation Fellow, 1950-1952; Allied Chemical and Dye Fellow, 1952-1953.
those of the methyl groups. This paper deals with the study of tin tetramethyl. Although several previous workers have studied this c o m p ~ u n d , ~ - ~ considerable disagreement still appears to exist as (3) C . F . Kettering and W.W. Sleator, Physics, 4, 39 (1933). (4) G. Pai, Pvoc. R o y . SOC.( L o n d o n ) , 1498, 29 (1935). (5) C. W. Young, J. S. Koehler and D . S. McKinney, THISJOURK A L , 69, 1410 (1947). ( 6 ) H . Siebert, 2. anoyg. Chem., 2 6 3 , 82 (1950). (7) H. Siebert, i b i d . , 268, 177 (1952). (8) E. R . Lippincott and M. C. Tobin, THISJ O U K N A L , 76, 4141 (1953).
Dec. 20, 1955
RAMAN AND INFRARED SPECTRA OF TIN TETRAMETHYL
to the actual observed spectrum and to its interpretation. The recent paper of Lippincott and Tobiqs which appeared as the present paper was in preparation, reports the most complete study to date. However, our results are not in accordance with those reported above in several important aspects. Experimental Preparation.-The tin tetramethyl was prepared by the method of Waring and Hortong and by an improved method employing n-butyl ether.10 The first method requires the separation from large amounts of diethyl ether, the last portions of which are difficult to remove. Some previous spectroscopic work has been done on samples so contaminated. The second method reduces these difficulties. Both samples were purified by fractionation through a Todd column (40-50 plates) and the fraction distilling a t 76.6” (748 mm.) (reported 76.8’, 76’) was used in the spectroscopic work. The progress of the purification was followed by infrared spectroscopy. Both samples gave identical results. Raman Spectrum.-The Raman spectrum was recorded photoelectrically and photographically using an Applied Research Laboratories Raman spectrometer. This three prism instrument has a Schmidt t y e camera of focal aperature f/3.5 and a dispersion of 15 l . / m m . a t 4358 A. For photoelectric recording an exit slit periscope arrangement traverses the focal plane and directs the radiation to an external 1P21 photomultiplier tube. Low pressure, watercooled mercury arcs served as the source. They were enclosed in a cylindrical, “light furnace” type of reflector system, which was coated with a silicone resin impigmented with TiOt. No filters were used except in those special runs to distinguish lines excited by 4046 and 4077 from those excited by 4358 A . A Wratten 2B filter was used for this purpose. The Raman spectrum was observed a t 40 f 5 ” . Exposure times varied from 15 seconds to 18 hours and slit widths from 50 to 500 p for the photographic work. Frequencies were measured by graphical comparison to an Argon reference spectrum printed on the same film. Polarization meas-
urements were made by the method described by Rank and Kagarise.I1 The spectrum of liquid Sn(CH3)r is shown in Fig. 1 and in Table 1.
,
I
500
0
loo0
I
I
I
I
1500
zoo0
2500
3030 CM;)
Fig. 1.-The
Raman spectrum of Sn(CH8)d.
Infrared Spectrum.-The infrared spectra were obtained with a double beam infrared spectrometer (Perkin-Elmer Model 21). The spectral region from 2 to 25 p was studied using CaFt, NaCl and KBr optics. Sealed liquid ~ e l l s ~ 2 ~ ~ 3 of thickness from 0.025 to 0.196 mm. were used for the study of the liquid. The extremely strong band a t 13 p was also studied using a capillary liquid cel112813and in dilute CS2 solution. The spectrum of gaseous S ~ I ( C Hwas ~ ) ~obtained with 10 cm. cells a t pressures from 5 to 90 mm. The infrared spectrum of the gas is shown in Fig. 2 and of the liquid in Fig. 3. The frequencies of all the bands are given in Tables I1 and 111.
2
3
6
5
4
0
7
9
10
I2
II
ern.-]
(obsd.)
4130 4108 3235 2987 2915 2840 2368 2086 1974 1880 1719 1548 1205 1194 1072 996 832 768 664 530 508 370 298 157
Relative intensity
Polarization factor
V.W.
..
V.W.
..
V.W.
3.5 7.9 V.W. W. V.V.W. V.W. V.W. V.W. V.W. 4.5 3.8 V.W.
V.W. V.W. m. V.W.
3.9 10.0 V.W.
.. 0.72 0.02
..
.. .. .. .. .. .. 0.10 0.65
.. .. .. .. .. 0.83 0.04
V.W.
.. ..
6.1
0.80
+ + +
14
I5 MICRONS
GAS
27’C.
IOCM.
Interpretation
2915 1205 = 4120 1194 = 4109 2915 2(157) = 3229 2915 d E ) , via(Fz) pi(Ai). vic(Fz) 2987 - 157 = 2830 2(1194) = 2388 508 157 = 2110 1445 768 = 1973 1205 508 157 = 1870 1205 508 = 1713 1205 2(768) = 1536 vz(Ai) Y16( FZ) 2(530) = 1060 2(508) = 1016 2(157) = 822 508 4 E ) , vidFz) 508 157 = 665 vis(Fz)
+ + + +
13
(CH314Sn
TABLE I THERAMAN SPECTRUM OF Sn( CH3)4 Frequency,
6487
+ +
+ +
AI)
530 - 157 = 373 2(157) = 314 4 E ) . YIO(FZ)
(0) C. E.Waring and W.S. Hortoo, THISJOURNAL, 67, 640 (1046). (10) Walter F. Edge11 and C. H.Ward, ibid., 76, I160 (1964).
17
10
20
19
Fig. 2.-The
,
f
21
21
MICRONS
infrared spectrum of Sn(CH3)4 (gas).
0.05MY
cs,
2
4
Fig. 3.-The
6
8
IO
I2
SOL*
14
infrared spectrum of Sn( CH3)4 (liquid).
Results and Discussion Previous Raman spectrum studies have been reported by Pai,4 by Siebert,8J and by Lippincott (11) D. H. Rank and R . E. Kagarise, J. Opt. SOC.A m . , 40, 89 (1950). (12) V. Z. Williams, Rev. Sci. Insfr., 19, 136 (1948). (13) R. C. Lord, R. S. McDonald and Foil A. Miller, 1.O p f . SOC. Am.,41, 140 (19%2).
WALTERF. EDGELL AND CURTISH. WARD
6488
TABLE I1 THEINFRARED SPECTRUMS I ~ ( C H(GAS) ~)~ Frequency cm.-l (obsd )
4439 4378 4 189 4112 3685 3595 3457 3118 3047 2991 2919 2812
Description
\v .
S.
\v .
S.
V.W.
S.
V.7V.
2991 1447 = 4438 1447 = 4366 2919 1200 = 4191 2991 1200 = 4119 2919 2919 4- 771 = 3690 2( 1205) 1200 = 3610 529 = 3448 2919 157 = 3148 2991
4422 4378 4 168 4103 3676 3605 3544 3445 2982 2914 2363 2216 2131
sh., br.
yc
S.
1%'.
S.
S.
rv . \v , m.
sh., br. sh., br. sh., br.
V.W. V.W.
2982 1445 = 4427 1445 = 4359 2914 1190 = 4172 2982 1190 = 4104 2914 764 = 3678 2914 1190 = 3600 2(1205) 524 = 3506 2982 2914 4-524 = 3438
S.
S.
dF2)
S.
S.
S.
m.
1962
br .
w.
1861
br .
V.W.
1717 1697 1445 1412 1348 1190 1050 1029 764 524
s.
m.
S.
m.
v.br. sh., br. sh., br.
v14(F2) 2(1190) = 2380 1445 764 = 2209 764 157 = 1205 2126 or 1445 524 157 = 2126 1445 524 = 1969 or 1205 764 = 1969 508 157 = 1190 1855 1190 524 = 1714 508 = 1698 1190 dFz)
m. m.
S.
S.
sh., br. br. v.v.br. br .
V.W.
+ + + +
rv .
V.V.\v .
S.
S.
S.
s.
via(F2)
s. sh., br.
s. V.V.W.
~14(F2) 1447 1205 157 = 2809 2(1200) = 2400 771 = 2218 1447 771 157 = 1205 2133 1447 529 = 1976 or 771 = 1976 1205 508 157 = 1200 1865 1205 4- 529 = 1734 or 529 = 1739 1200
br . v.br.
1975
br .
V.R.
1858
br .
V.W.
1740 1730 1722 1717 1708 1698 1447 1354 1207 1200 1192 1056 1034 800 771 536 529 523
Description
V.W.
+ +
+
+
m.
S.
+
+ + + + + + +
V.W. V.W.
+
m.
S.
1200
+ 508 = 1708
br. br.
m.
vis(Fz)
W.
1200
PQR
S.
vis(Fa)
br. br . Sh. v.v.br.
V.W.
w. V.V.W.
2(529) = 1058 508 = 1037 529 Impurity YidFz)
PQR
V.S.
+ 157 = 1357
+
V.V.S.
Sn MI,
PHOTOELECTRI$
'1
RAMAN RECORDING
MATTEFILTER 40 DB
Relative intensity
(obsd.)
S.
2384 2232 21 19
cn-1
Interpretation
sh., br. sh., br. sh., br.
V.W.
TABLE I11 IKFRARED SPECTRIJM O F SIl(CHa)r (LIQUID)
Frequency,
Relative intenqity
sh., hr.
THE
VOl. 7'7
I N 0 FILTER
30 DB
b a Fig. 4.-Raman spectrum of SnMe4 between 1100-1500 crn.-l: a, without Wratten filter; b, with Wratten filter.
S.
br . br .
.
W.
V.W. V.W.
S.
W.
V.V.S. V.S.
Interpretation
+ + + + + + +
+ +
+ +
+
+ + + + + +
...... ..
+
1190 157 = 1347 vi6 (F2) 2(524) = 1048 508 524 = 1032 vi7 (F2) via (F2)
+
and Tobin.8 Our results are most nearly in agreement with those of Siebert. We resolve the line a t 1200 ern.-' into a doublet a t 1194 and 1205 cm.-' in agreement with the findings of Lippincott and Tobin. Apparently, however, we did not achieve as high a resolution as these workers since we did not obtain partial resolution of the line a t 157 ern.-'. We do not observe the lines reported only by Pai a t 262, 952 and 1262 cm.-'; they are certainly due to impurities. The same may be true for the line found by Siebert a t 1434 and by Lippincott and Tobin a t 1451 cm.-' and the line reported a t 1046 by Pai and a t 1053 by Lippincott and Tobin. We do not observe either of them. Raman lines have been reported recentlys a t 1346 and 1400 ern.-', and the latter of these was assigned to a fundamental frequency of vibration. Figure 4a shows the detailed photoelectric recording of this region of the spectrum. The lines in question (which we find a t 1335 and 1406 cm.-l) are seen t o the right of the 1200 doublet. However, these positions also correspond to the strong Raman lines 2915 and 2987 ern.-' if excited by the mercury line 4077 k . instead of by the 4358 A. line. If this latter interpretation were correct, these same lines also should appear, with greater intepsity, excited by the stronger mercury line 4046 A. The 2915 cm.-' line does indeed appear displaced 1144 cm.-' from the 4358 k . mercury line; the presence of the 2987cm.-'line (at about 1220 an.-') is not obvious. And perhaps this fact led to the choice of 4358 k . as the exciting mercury line and hence t o a belief
RAMAN AND INFRARED SPECTRA OF TINTETRAMETHYL
Dec. 20, 1955
in the reality of displacements a t 1346 and 1400 cm.-’. A Wratteq 2B gelatin filter strongly absorbs light in the 4100 A. region while passing 4358 A. with a slight diminution in intensity. Figure 4b shows the photoelectric recording of the same region of the spectrum taken with the Raman tube wrapped with several layers of this filter. The gain has been increased more than threefold to substantially overcompensate for the lessened 4358 A. intensity reaching the sample. I n this trace the Raman lines excited by 4358 A. should appear stronger than in Fig. 4a. The “1144”line has been greatly reduced in intensity while the “ 1346” and 1400” lines have disappeared into the systemic noise. Moreover, while the apparent intensity of the 1200 cm.-’ doublet has increased, it is much more narrow than shown in Fig. 4a. Thus a relatively strong line adjacent to and slightly overlapping 1205 cm.-l was removed by the filter. I t can be concluded that the lines “1144” and ca. “1220 cm. -l” are the Raman displacements 2715 and 2987 cm.-l excited by the mercury line 4046 A. and the lines a t “1346“ and “1400,” these same displacements excited by the mercury line 4077 A. Infrared spectra of tin tetramethyl have been previously reported by Kettering and Sleator3 and by Lippincott and T o b h 8 This work differs rather substantially from the above a t wave lengths shorter than 5 p . Presumably this results from the higher pressure, greater resolution and purer samples used in this study. At wave lengths longer than 5 p our results are in reasonably good agreement with the earlier work except for the absence of an occasional impurity band. The very strong infrared band found a t 771 cm.-’ in the gas phase and a t 764 cm.-’ in the liquid phase has been reporteds as a multiplet band in both phases. We have observed the peaks a t 760 and 774 cm.-‘, reported8 as strong for the gas phase, only with samples containing traces of ethyl ether and never when using the highly purified material. No shoulders were observed in this study for this band in the liquid state. The infrared band previously founds a t 1118 ern.-' is also due to ethyl ether. About 1% ether is sufficient to give rise to this bitnd. ‘ I
Assignment of Frequencies The Sn(CH3)4 molecule, like the other tetramethyl compounds of Group IV elements, appears to have the symmetry of the tetrahedral point group Td. The number of fundamental vibrations and the selection rules are listed in Table 1V. TABLE IV SELECTION RULES FOR FUNDAMENTAL VIBRATIOXS Sn(CHd4 Fundamental No. of fundamentals type
Ai
3
Raman activity
Pol.
A2
1
....
E FI F2
4 4 7
Depol. .... Depol.
OF
Infrared activity
.... .... .... .. .
Active
In describing the vibrations it will be convenient to use the ide:i of specific ( U T Jhypothetical) ~ cliurnr-
6489
teristic motions.14 Each such definite set of atomic displacements is chosen to be a possible fundamental vibration and hence each is orthogonal t o all the others. It is useful to deal with the frequency which is characteristic of this definite motion. The actual vibrations are combinations of these simplified motions and the observed vibrational frequencies may be regarded as combinations of the corresponding characteristic frequencies. Because the tin atom is much heavier than the other atoms of this molecule these characteristic motions can be chosen t o be good approximations to the actual vibrations in this case. Consider first of all an “isolated” CH3 group. Its modes may be described as the symmetrical stretching and deformation vibrations and the doubly degenerate, asymmetrical stretching and deformation vibrations. These motions in the “combined” molecule take place against a stationary tin atom. Orthogonality now requires that the methyl group be rigid in the remaining modes of motion. These may be described as skeletal stretching and deformation motions plus the torsional and rocking motions of the methyl group. The set of these simplified motions which form a possible group of fundamental modes for Sn(CH3)4 are found in Table V. When a methyl group vibrates, i t deforms the Sn-C bond, and the extent t o which the Sn atom moves depends upon the mass of the Sn atom and the stiffness of the bond. Since the mechanical strength of this bond is not great and the Sn mass is large, the displacement of the Sn atom is small. Thus there is little mass coupling between the CH3 groups and little coupling between the skeletal and methyl vibrations and the descriptions of Table V are good approximations t o the actual motions. Consider the vibrations v1 and v14. They consist of sym: metrical stretching motions of the CH, groups and differ by the phase relationships between the motions of the several groups. It is a consequence of the above that these two frequencies should be nearly the same and might not be resolvable experimentally. The same may be expected for all motions of the methyl groups which differ only in this phase relationship. The confirmation of this expectation is the primary feature of the experimental spectra. TABLE V DESCRIPTION O F FUNDAMENTAL VIBRATIO~ s OF Sn( CH.,)4 Description
CH, asymmetrical stretch CH3 symmetrical stretch CH3 asymmetrical deformation CH3 symmetrical deformation CHs rocking Skeletal stretching Skeletal deformation CHa torsion
Ai
Az
E
FI Pz
V6
V9
VI3
V8
VI0
V15
vi
VI1
v17
vi
“14
vz
v16
v3
VI8 VS VI
VI9 v12
The fundamental frequencies of Sn(CH3)4 are readily assigned in terms of these ideas and the observed polarization data of the Raman spectrum. The strong, polarized Raman line a t 2915 crn.-l is assigned to the symmetrical CHBstretching (14) Welter F. Bdgrll, t o be piiblisbril
6490
IijALTER
cmr'
0I
F.EDGELL AND
CURTIS
VOl. 77
H. \YARD
1500 3000 I 1
1000
500 I
1
(CH,), Pb ~
~~
Fig. 5.-Fundamental
_ _ _ ~ ~
frequencies of M( CH3)(molecules.
fundamentals V I and ~ 1 4 . The latter fundamental is observed in the infrared a t 2914 cm.-l for the liquid and 2919 for the gas. The strong, depolarized Raman line a t 2987 cm.-' is readily assigned to the asymmetrical CHI stretching modes u5 and v13. An infrared band corresponding to ~ 1 3was observed a t 2982 cni.-l in the liquid state and a t 2991 ern.-' in the gaseous state. The strong infrared band a t 1445 cm.-' in the liquid and 1447 in the gas arises from the asymmetrical deformation vibration VIS. The corresponding Raman line arising from either or ~ 1 was 5 not found. Similar vibrations are either missing or give weak Raman lines in the various methyl tin halides.15 The strong, polarized Raman line a t 1205 ern.-' arises from the AI symmetrical CHBdeformation vibration v 2 . Another strong but depolarized Raman line is found a t 1194 ern.-' and must be assigned to the F2 symmetrical CH3 deformation vibration v16. The resolution of these two lines was achieved by both photographic and photoelectric techniques. The v16 fundamental appears in the infrared spectrum of the liquid as a strong band a t 1190 c m - ] and for the gas and a strong band a t 1200 cm.-' with PQR branches. Here two non-skeletal fundamentals have been resolved which differ only in the phase relationship of the methyl group motions. The coupling is greatest here because the C atoms move by far the largest amount along the CH3axis, and hence the Sn-C bond, in the symmetrical deformation vibration. The symmetrical skeletal stretching vibration va gives rise to the polarized Raman line a t 505 cm.-l while the F2 vibration gives rise to the Raman line a t 530 cm:-' and the infrared bands at 524 and 529 cm.-' for the liquid and gaseous states, respectively. :Z very strong infrared band is found a t S i 1 for the gas and 764 cm.--l for the liquid which arises from t h e methyl rocking vilir;ii i(.jri
I..;)
W ~ l l i r r1'. I:
