Cyclic Imides. II. Silylsuccinimides

Cyclic Imides. II. Silylsuccinimides A. F. JANZEN AND E. A. KRAMER Department of Chemistry, University of Manitoba, Winnipeg, Manitoba

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 180.242.48.105 on 06/03/13 For personal use only.

Received June 1 , 1971 A number of silylsuccinimides containing silicon substituents -H, - C H 3 , - C 2 H S , -CH=CH2, -CH2CH=CH2, -C6Hs, as well as disuccinimides, have been prepared and characterized by i.r., n.m.r., and mass spectrometry. The results are compared with previous studies of silylphthalimides. On a prkpark et caractkrisk par i.r., r.m.n., et spectre de masse un certain nombre de silylsuccinimides -CH2CH=CH2, contenant des substituants du silicium, tels -H, -CH,, -C2H,, -CH=CH2, -C,Hs, ainsi que des disuccinimides. On compare les rksultats avec des Btudes antkrieures sur les silylphthalimides. Canadian Journal of Chemistry, 49, 3456 (1971j

We have previously reported the preparation and characterization by i.r. and mass spectrometry of a number of silylphthalimides (I). As a continuation of this study we now describe a variety of silylsuccinimides 1-9 listed in Table 1. Silicon substituents include -H, -CH,, -C2H5, -CH=CH,, -CH2CH=CH,, -C,H5, and compounds 8 and 9 are disuccinimides. Although a number of silylsuccinimides have been prepared previously' few details of their spectral properties or synthetic applications have been reported. Silylsuccinimides 1-9 were conveniently prepared from the corresponding chlorosilane and potassium succinimide and purified by vacuum distillation or recrystallization. In general, silylsuccinimides appeared more susceptible to hydrolysis than silylphthalimides, in particular disuccinimides, and they were handled in a vacuum system or nitrogen atmosphere. All compounds were liquids or low melting solids and, on the basis of i.r. and n.m.r. studies, compounds 1-9 are N-silylated and no evidence was found for 0-silylation. The i.r. spectra were examined in the region 4000-400 cm-l. Two bands were found in the carbonyl region. The more intense band was found in the region 1696 4 cm-' and may be assigned to the vibrationally coupled asymmetric mode of vibration (3) analogous to the situation in silylphthalimides where C=O (asymmetric) bands were found in the region 1700 7 cm-' (1). The symmetric mode was of weaker intensity and appeared at higher frequency, in the region 1764 f 4 cm-l. Values of C==O (asymmetric)

+

+

'For the preparation of N-(trimethylsilyl)succinimide see e.g. ref. 2.

and C==O (symmetric) are recorded in Table 1. Typical i.r. spectra recorded as liquid films, clearly showing two bands in the carbonyl region, are reproduced in Fig. 1. Splitting of the C==O bands due to vibrational interactions is consistent with the presence of two equivalent carbonyl groups (4) and therefore Nsilylation, rather than O-silylation, must have occurred. Compounds 3 and 4 showed Si-H stretching vibrations at 2201 and 2199 cm-l, respectively, in the region expected for silicon-hydrogen bonds (5). A band of strong intensity was found for all silylsuccinimides in the region 1165-1 178 cm-', illustrated in Fig. 1, and this band may tentatively be assigned to a Si-N vibration. Wallwitz et al. have previously assigned a strong band in the region 1170-1 180 cm- to a Si-N vibration in N-silylheterocycles (6). The n.m.r. spectra at 40" of silylsuccinimides 19 showed a single sharp peak in the region - 2.66 to -2.54 p.p.m. assigned to the succinimido hydrogens. The succinimido hydrogens must therefore be equivalent, consistent with N-silylation, and the possibility of isomeric structures resulting from the addition of silicon-hydrogen (7) or silicon-nitrogen (8) bonds across the carbonyl bond is eliminated. The equivalence of the succinimido hydrogens is of significance with regard to the stereochemistry at the nitrogen atom. If the succinimido ring defines a plane of symmetry then the hydrogens will be equivalent if the silicon atom is in the plane of the succinimido ring or if rapid nitrogen inversion is occurring. On the other hand, pyramidal geometry at the nitrogen atom, or slow

'

JANZEN AND KRAMER: CYCLIC

IMIDES. I1

TABLE1. The i.r. data of silylsuccinimides

"'

-

N-Si-R2

0 R3


No.

R1

Rz

CH3 CzHs CH3 CH3 CH3 CH3 CH3 SUE? SUCC

1 2 3 4 5 6 7 8 9

R3

CH3 CZHS CH3 C6HS CH3 C6H5 CH3 CH3 CH3

Infrared (cmW1)*

c==o

(asymmetric)

(symmetric)

1696 1697 1700 1699 1692 1694 1699 1698 1698

1761 1763 1764 1766 1761 1764 1768 1764 1768

CH3 CzHs H H C6Hs C6HS CHzCH=CH2 CH3 CH=CH2

Si-H

2201 2199

'Band positions are accurate to i.2 cm-I. Compounds 5 and 6 were dissolved in Nujol; all others were

neat liquid films.

tSucc = succinimido. MICRONS 2 5

30

35

a0

MICRONS 50

e,]

6,O

7 5

80

00

90

PO0

50

00

I

I

smo

rim

aom

2sm

im

ram

,300

i~m c i.3

FREQUENCY (cM'!

~o

MICRONS 30

35

40

900

8M

im

600

sm

I

FREQUENCY ICM'I

MICRONS 50

60

80

I

I

i FREQUENCY ICM.'!

FREQUENCY ICM?

(b)

FIG.1. Typical i.r. spectra: (a) N-(dimethylsily1)succinimide (3) and (b) N-(N'-succinimidodimethy1silyl)succinirnide (8) recorded as neat liquid films.

pyramidal inversion, would destroy the equivalence and a complex spectrum would be expected. Since all compounds 1-9 showed single sharp peaks at + 40°, therefore changing substituents on silicon has no effect on the equivalence of the succinimido hydrogens nor does the introduction of a second succinimido ring. The n.m.r. spectrum of N-(triethylsily1)succinimide (2) in CHCl, was recorded down to - 80" with the aim of establishing temperature dependent inversion. Although the succinimido

peak is somewhat broadened relative to the sharp singlet triethylsilyl peak, the effect is small, and at - 80" silylsuccinimides must still be undergoing rapid inversion as has previously been found for N-substituted imines (9). Attempts to slow down the rate of inversion at room temperature by forming complexes with SnCl, (10) and thereby introducing steric hindrance via carbonyl-metal complexing were unsuccessful. In a similar experiment, AlCl, was added to 2 and the n.m.r. spectrum recorded down

CANADIAN JOURNAL OF

CHEMISTRY. VOL. 49,

1971


[31

0

FIG.2. The 'H n.m.r. spectrum of N-(triethylsily1)succinimide (2) in benzene showing triethylsilyl region only.

to - 80°, but the line broadening was not greater than without added AlCl,. It was noted above that the spectrum of N-(triethylsily1)succinimide(2) in CHC1, solution gave a sharp singlet assigned to the triethylsilyl group. To confirm that the singlet was due to an A3B2 type of spectrum expected for an ethyl group, the sample was re-run in benzene solution and the complex pattern shown in Fig. 2 was obtained, consistent with an A3B, spectrum (1 1).

Mass Spectra Silylsuccinimides generally showed weak molecular ions a and the most prominent peaks resulted from loss of alkyl, aryl, or hydrogen,

-R a ----t b (eq. 1). If methyl and phenyl substituents, or if methyl and hydrogen substituents were present, then in both cases loss of methyl gave the more abundant peak. There were no peaks in the mass spectra of silylsuccinimides which could be assigned to CO, elimination. This was rather interesting in view of the prominent loss of CO, observed in silylphthalimides (1) (eq. 2). The loss of CO, must therefore be characteristic of the phthalimido group, since N-alkyl- and N-silylphthalimides readily lose CO, (12, 13) whereas N-alkyl- and N-silylsuccinimides do not (13, 14). Loss of CO has been observed in N-alkylsuccinimides and is also found in N-silylsuccinimides. For example, the ion d, of 23% relative abundance, in the mass spectrum of N-(dimethylsily1)-

0

c, m/e 157 (8%)

d, m/e 129 (23%)

succinimide 3 (eq. 3) may result from loss of CO, by analogy with the loss of CO from N-methylsuccinimide (eq. 4). The latter process was confirmed by high mass spectrometry and the presence of a metastable ion (14).

[41

4-.H3 0

d

-':O

fCH3 0

Silylsuccinimides showed peaks of varying in-

+

+

tensity assigned to R3SiOSiR,, R,SiOH, and R3Sif. These peaks may originate in the hydrolysis and condensation of silyl derivatives (eq. 5),

as has been noted before in mass spectral studies (15). As a further possibility siloxanes may result from skeletal rearrangements as has been found for trimethylsilyl ethers (16). Additional peaks found at mle 56, 55, 28 have also been found in N-alkylsuccinimides (13, 14) and must be characteristic of the succinimido ring.

Experimental General All compounds were manipulated in a conventional vacuum manifold or under dry nitrogen atmosphere. Organochlorosilanes were purchased from Peninsular Chemresearch and succinimide from Aldrich Chemical Co. Solvents were dried and distilled prior to use. Elemental analyses were carried out by A. Bernhardt, West Germany. The i.r. spectra were determined on a Perkin-Elmer 337 spectrophotometer using KBr windows and calibrated against polystyrene film. Mass spectra were obtained on a Finnigan 1015 quadrupole or Hitachi Perkin-Elmer RMU-6D mass spectrometer at 50 eV. Proton n.m.r. spectra were obtained on a Varian A-56/60A spectrometer at 60 MHz in 10% (v/v) CDC13 solutions using a small amount of CHC13 as internal standard. Chemical shifts are reported relative to external tetramethylsilane

JANZEN AND KRAMER: CYCLIC IMIDES.

3459

I1

TABLE2. Elemental analyses, boiling points, and n.m.r. data


Elemental analyses (%) Calculated

Found

C H

C H

Chemical shift (.-v .- ~ . m. . ) *

Boiling voint (meltini point) ("C)

Succ

CH3

-0.42, d J=3.5Hz

9

--i

100 at

mm

-2.62, s

H

-4.68, sep J=3.5Hz

-0.78, s

*s = singlet; d = doublet; q = quartet; sep = septet; succ = succinimido. ?Reactivity of 9 prevented satisfactoryelemental analysis.

(TMS) using the formula 6 p.p.m. (CHCI, internal) 7.23 p.p.m. = F p.p.m. (TMS external). All chemical shifts were found to be downfield from TMS.

Preparation of N-(Triethylsilyl)succinimide (2) Triethylchlorosilane (4.50 g, 29.9 mmol) was slowly added to potassium succinimide (4.10 g, 29.9 mmol), prepared from potassium metal and succinimide in tetrahydrofuran, at 0" under nitrogen atmosphere. Potassium chloride was filtered off and solvent and unreacted triethylchlorosilane pumped off leaving a pale yellow liquid. Vacuum distillation gave a clear colorless liquid identified as N-(triethylsily1)succinimide. The typical procedure outlined for N-(triethylsily1)succinimide was repeated for each silylsuccinimide.Products were purified by vacuum distillation, with the exception of 5 and 6 which were recrystallized from benzene, and the yield of purified product was about 25-35% for compounds 1-7 and 10-15% for 8 and 9. The i.r. data for 1-9 are reported in Table 1 and elemental analyses, boiling points or melting points, and n.m.r. data are reported in Table 2. The financial assistance of the National Research Council of Canada is gratefully acknowledged. We thank the University of Manitoba for a Manitoba Graduate Fellowship (to E. A. K.). 1. A. F. JANZEN and E. A. KRAMER.Can. J. Chem. 49, l o l l (1971). 2. L. BIRKOFER and H. DICKOPP. Chem. Ber. 101,2585 (1968); A. E. PIERCE. Silylation of organic compounds. Pierce Chemical Co., Rockford, Illinois, 1968.

3. L. J. BELLAMY.Advances in infrared group frequencies. Methuen, London, 1968. p. 129. 4. H. K. HALL,JR. and R. ZBINDEN.J. Am. Chem. SOC.80, 6428 (1958). 5. B. J. AYLETT. Advan. Inorg. Chem. Radiochem. 11, 249 (1968). 6. U. WALLWITZ,H. SCHMIDT, and W. GOSDA. J. prakt. Chem. 32, 274 (1966). 7. A. F. JANZEN and C. J. WILLIS. Can. J. Chem. 43, 3063 (1965). 8. M. F. LAPPERT and B. PROKAI. Advan. Organomet. Chem. 5, 225 (1967). 9. A. T. BOTTINIand J. D. ROBERTS.J. Am. Chem. SOC.80, 5203 (1958). 10. S. C. JAINand R. RIVEST.J. Inorg. Nucl. Chem. 31, 399 (1969). 11. J. W. EMSLEY,J. FEENEY,and L. H. SUTCLIFFE. High resolution nuclear magnetic resonance spectroscopy. Pergamon Press, 1965. p. 355, 675. 12. R. A. W. JOHNSTONE, B. J. MILLARD,and D. S. MILLINGTON.Chem. Commun. 600 (1966); T. W. BENTLEY and R. A. W. JOHNSTONE. J. Chem. Soc. C, 2354 (1968); C. M. ANDERSON, R. N. WARRENER, and C. S. BARNES.Chem. Comrnun. 166 (1968). 13. J. H. BOWIE,M. T. W. HEARN,and A. D. WARD. Aust. J. Chem. 22, 175 (1969). H. BUDZIKIEWICZ, and C. DJERASSI. 14. A. M. DUFFIELD, J. Am. Chem. Soc. 87, 2913 (1965). 15. H. BUDZIKIEWICZ, C . DJERASSI, and D. H. WILLIAMS. Mass spectrometry of organic compounds. HoldenDay, San Francisco, California, 1967. p. 476. 16. J. DIEKMAN,J. B. THOMSON, and C. DJERASSI.J. Org. Chem. 33, 2271 (1968).