Molecular structure of N-trimethylsilylaziridine in the gas phase

Article abstract of DOI:10.1039/a910248p

N-Trimethylsilylaziridine, Me₃SiNC₂H₄1, prepared by a literature procedure has been studied by gas-phase IR and NMR spectroscopy of the nuclei ¹H, ¹³C, ¹⁵N, ²⁹Si. The molecular structure of 1 has been determined in the gas phase by analysis of electron diffraction (GED) data augmented by geometric and vibrational amplitude restraints derived from ab initia calculations employing the second order Moller-Plesset (MP2) level of theory and the 6-31 lG(d, p) basis set. The molecule was found to adopt C₄ symmetry and has a pyramidal arrangement around the nitrogen atom with a slightly elongated Si-N bond [1.770(4) A], relative to those in other Si-N species having a planar nitrogen co-ordination. The results for the principal distances (r_a) and angles from the combined GED/ab initia study of Me₃SiNC₂H₄ are (uncertainties: one sigma): r(Si-N) = 1.770(4), r(Si-C_in-plane) = 1.868(4), r(Si-C_out-of-plane) = 1.881(2), r(C-N) = 1.459(3) A, C-N-C 61.1(5), N-Si-C_in-plane 109.7(7), N-Si-C_out-of-plane 107.8(10), Si-N-C 121.9(4)°. The structure data are complemented by natural bond orbital analyses, and calculation of the barrier to inversion at the nitrogen centre, which is predicted to be 21.4 kJ mol^-1. Calculations on the simpler model compound H₃SiNC₂H₄ [up to QCISD/6-31 lG(d, p)] gave geometrical and energetic results similar to those of 1. The steeply pyramidal nitrogen co-ordination is attributed to the strain of the aziridine ring, as was found by comparison with the results of other calculations on acyclic systems and those with larger rings. The Royal Society of Chemistry 2000.

Full text of DOI:10.1039/a910248p

View Article Online / Journal Homepage / Table of Contents for this issue
Molecular structure of N-trimethylsilylaziridine in the gas phase
Norbert W. Mitzel,*a Elizabeth M. Pageb and David A. Riceb
a Anorganisch-chemisches Institut, Technische Universität München, Lichtenbergstr. 4,
85747 Garching, Germany
b Department of Chemistry, University of Reading, PO Box 224, Whiteknights, Reading,
UK RG6 6AD
Received 22nd December 1999, Accepted 22nd March 2000
N-Trimethylsilylaziridine, Me3SiNC2H4 1, prepared by a literature procedure has been studied by gas-phase IR and
NMR spectroscopy of the nuclei 1H, 13C, 15N, 29Si. The molecular structure of 1 has been determined in the gas phase
by analysis of electron diffraction (GED) data augmented by geometric and vibrational amplitude restraints derived
from ab initio calculations employing the second order Møller–Plesset (MP2) level of theory and the 6-311G(d,p)
basis set. The molecule was found to adopt Cs symmetry and has a pyramidal arrangement around the nitrogen atom
with a slightly elongated Si–N bond [1.770(4) Å], relative to those in other Si–N species having a planar nitrogen
co-ordination. The results for the principal distances (ra) and angles from the combined GED/ab initio study of
Me3SiNC2H4 are (uncertainties: one sigma): r(Si–N) = 1.770(4), r(Si–Cin-plane) = 1.868(4), r(Si–Cout-of-plane) = 1.881(2),
r(C–N) = 1.459(3) Å, C–N–C 61.1(5), N–Si–Cin-plane 109.7(7), N–Si–Cout-of-plane 107.8(10), Si–N–C 121.9(4)Њ. The
structure data are complemented by natural bond orbital analyses, and calculation of the barrier to inversion at the
nitrogen centre, which is predicted to be 21.4 kJ molϪ1. Calculations on the simpler model compound H3SiNC2H4
[up to QCISD/6-311G(d,p)] gave geometrical and energetic results similar to those of 1. The steeply pyramidal
nitrogen co-ordination is attributed to the strain of the aziridine ring, as was found by comparison with the results
of other calculations on acyclic systems and those with larger rings.
electrostatic repulsion between the substituents at the nitrogen
Introduction
atom.7 Only for compounds with very electronegative substitu-
The structural chemistry of silicon–nitrogen compounds has
been the focus of interest of many studies in the past as the
Si–N bond was found to be unexpectedly strong and is distinct
in its chemistry and structure from the related C–N bond.
ents as in (H3Si)MeNOMe, a significant amount of electrostatic
attraction occurs thus leading to steeply pyramidal silylamines.
The influence of ring strain on the geometry around the
nitrogen atom in Si–N compounds is of current interest. Very
Nitrogen atoms in typical alkylamines form bonds in a pyr-
recently a crystal structure of the silyl aziridine Ph3SiNC2H4
amidal array and have considerable basicity. Exceptions do
was published, showing this compound to have a steeply pyr-
occur where there is massive steric overcrowding, an example
amidal nitrogen atom in the crystal lattice.8 This poses the
being (iPr)3N which was reported to have a planar nitrogen NC3
question as to whether the observation of a pyramidal nitrogen
core in the gas phase,1 but is pyramidal in the crystal.2 In con-
core is a consequence of crystal packing as was seen in the case
trast, silylamines generally possess planar nitrogen atoms and
of (iPr)3N, see above. The molecule Ph3SiNC2H4 is thought
the compounds show very little basicity. In most cases just one
unlikely to exhibit sufficient stability in the gas phase to allow
silyl substituent bound to the nitrogen is sufficient to achieve a
investigation of the influence of crystal packing. However, the
complete flattening of the nitrogen co-ordination sphere. The
trimethyl analogue, Me3SiNC2H4 1, which was prepared by
whole process of flattening of the nitrogen containing core
Scherer and Schmidt as early as 1965,9 was thought likely to
is associated with a decrease in the height of the barrier to
be suitable for a gas-phase investigation. Therefore in this paper
inversion in alkylamines caused by the presence of silicon sub-
stituents. While in most cases the barrier vanishes, in some cases
there remains either a very shallow potential or a small barrier.
Such small barriers probably exist in H3SiNMe2, which exhibits
are reported the results of ab initio calculations and gas-phase
electron diffraction studies on 1.
Results and discussion
a slightly pyramidal NSiC2 core,3 and in compounds with very
bulky silyl substituents carrying a NH2 group.4 Recently in-
herently pyramidal SiN containing molecules were found when
the structures of N-silylhydroxylamines were investigated. The
observation was ascribed to the influence of the high electro-
negativity of the alkoxy substituent bound to the nitrogen atom
which causes a high barrier to inversion, thus overcoming the
counter influence of the silicon substituent. Examples are
provided by (H3Si)2NOMe5 and (H3Si)MeNOMe6 which have
slightly and steeply pyramidal cores respectively.
The background to our study is as follows. Initially attempts
were made to synthesize H3SiNC2H4. Two methods were used.
The first required the condensation of bromosilane and
aziridine in the presence of 2,6-dimethylpyridine as an auxiliary
base, while in the second lithioaziridine, LiNC2H4, was allowed
to react with bromosilane. The first method was completely
unsuccessful, probably because of the sensitivity of aziridine
towards ring opening by acids. The second method produced a
volatile compound, which condensed onto the walls of a cold
trap at Ϫ60 ЊC after the volatile constituents of the reaction
mixture (when the mixture was warmed to maximum Ϫ78 ЊC)
were pumped through to other cold traps. However, upon
removing the coolant, the liquid condensate that had been
Electronic reasons for planar nitrogen containing cores in
silylamines are the subject of many publications, the latest, in
which N(SiH3)3 and H2NSiH3 are discussed, contains argu-
ments in favour of a polar σ effect, which is said to give rise to
DOI: 10.1039/a910248p
J. Chem. Soc., Dalton Trans., 2000, 1491–1497
This journal is © The Royal Society of Chemistry 2000
1491

View Article Online
trapped at Ϫ60 ЊC turned immediately and irreversibly to a
solid as the temperature was raised slightly. The solid was not
soluble in common aprotic organic solvents. If silylaziridine
was formed by the second reaction, which could not be proved
unequivocally, it probably decomposed by reaction of the
Si–H functions with the aziridine ring, a reaction which has
previously been described as alkali metal catalysed, e.g. see the
reaction of Et2SiH2 and HNC4H4.10 It should also be noted that
this was not the first attempt to prepare this compound.11
Having failed to make the parent compound of the homo-
logous series it was decided to study 1, so eliminating the
problem of the incompatibility of Si–H and aziridine functions.
A new preparative method was adopted in which lithioaziridine
was allowed to react with Me3SiCl, eqn. (1). The desired
Fig. 1 Experimental scattering intensity (s4I mol(s)) and final differ-
ence curves (vs. model) for gaseous Me3SiNC2H4.
Me3SiCl ϩ LiNC2H4 → Me3SiNC2H4 ϩ LiCl (1)
1
material was easily purified by fractional condensation in a high
vacuum system to give a sample suitable for structure deter-
mination by gas-phase electron diffraction. Characterisation
of 1 in solution was undertaken by multi-nuclear NMR spec-
troscopy. The 1H and 13C NMR data deserve no comment and
are listed in the Experimental section. In the 29Si NMR spec-
trum a multiplet appears at δ Ϫ19.7, which is about 20 ppm
to lower frequency than chemical shifts reported for other
typical silylamines [(Me3Si)2NH (δ 2.2), (Me3Si)3N (2.4)12 and
Me2NSiMe3 (6.5)].13 The unique value obtained here for a
silylamine suggests the presence of a different bonding situation
at silicon and a different type of Si–N bond in 1 to those
observed in other silylamines. This interpretation is confirmed
by data from 15N NMR spectroscopy which gave a signal at
δ Ϫ348.5, which bears two pairs of satellites identified by their
relative intensities as coupling to the ring carbon atoms with
1JNC = 2.9 Hz and a coupling to the silicon centre with
1JSiN = 4.8 Hz. The latter value is exceptionally low for a mono-
silylamine (coupling constants of about 20 Hz are seen
normally14) suggesting the presence of a weak and relatively
Fig. 2 Radial distribution and difference curves for the electron
diffraction refinement of Me3SiNC2H4. Before Fourier inversion the
data were multiplied by sؒexp[(Ϫ0.002s2)/(ZSi Ϫ fSi)/(ZN Ϫ fN)]. Vertical
lines indicate atom pairs with their height being proportional to their
scattering contribution.
1
long Si–N bond. Such small JSiN coupling constants have,
however, been observed for N-silylhydroxylamines, which have
been shown to have pyramidal nitrogen containing cores and
weak Si–N bonds.15
Gas-phase electron diffraction data for compound 1 were
obtained using the Reading apparatus and an all-glass nozzle
as described below. The data were interpreted using the
SARACEN method,16 to overcome the difficulties often
experienced in refining a complete molecular geometry from
electron diffraction data alone. Geometry restraints derived
from ab initio calculations using the second-order Møller–
Plesset (MP2) level of theory up to the 6-311G(d,p) basis set
[MP2/6-311G(d,p)] were used to augment the experimental
data, and a force field calculated at MP2/6-31G(d) was
employed to calculate the vibrational amplitudes, perpen-
dicular amplitude corrections and centrifugal distortions by the
normal co-ordinate analysis program ASYM40.17
A model with Cs symmetry was proposed for compound 1
with C3, Si1, N2 and the mid-point of the C6–C7 bond [see
Fig. 3 for the atomic numbering scheme] defining the plane of
symmetry. The model was defined by the seventeen parameters
and restraints listed in Table 1. Shown in Fig. 1 are the molec-
ular scattering intensity curves while in Fig. 2 is depicted the
radial distribution curve obtained from Fourier transformation
of the molecular intensity data. The small residuals exhibited in
the difference curves [see Figs. 1 and 2] indicate that a good fit
was achieved between the experimental and calculated curves,
the latter being obtained using the proposed molecular model
and the values for the parameters given in Table 1. The validity
of the proposed model was also substantiated by the final RG
value of 0.0647.
Fig. 3 Molecular structure of Me3SiNC2H4 in the gas phase with atom
numbering scheme.
Si–N bonds for comparable gas-phase geometries cover the
18
range of 1.710(5) Å in Me3SiNMe2 to 1.755(3) Å in
(Me3Si)3N,19 with the values for Me3SiNCS [1.743(6) Å],20 for
Me3SiNCO [1.740(4) Å],20 for H3SiNMe2 [1.713(5) Å]3,22 and
for Me3SiNHMe [1.738(5) Å]23 lying in between. The major
reason for this lengthening can be attributed to the very differ-
ent co-ordination geometry at nitrogen in compound 1, where
it is steeply pyramidal, as is indicated by the sum of angles
formed at nitrogen by the three bonds [305.0(11)Њ] and the
declination of the Si–N vector from the C2N plane of the
aziridine ring of 52.1(5)Њ. In two of the compounds mentioned
18
above, Me3SiNMe2 and (Me3Si)3N,19 the nitrogen atom and
the three atoms bonded to it are coplanar. Only in H3SiNMe2
a slight deviation from planarity is observed as revealed by
the sum of the angles subtended at the nitrogen atom
[353.9(8)Њ].3,22
The geometry of the ring is similar to that of the parent
The Si–N bond, at 1.770(4) Å, is comparatively long. Typical
aziridine24 and deserves no further comment. The distortion of
1492
J. Chem. Soc., Dalton Trans., 2000, 1491–1497

View Article Online
Table 1 Parameter definition, parameter values and restraints for the gas-phase electron diffraction (GED) least squares refinement of the geometry
of Me3SiNC2H4. Results from the ab initio calculations are listed for comparison
Parametera
Experimental result
Restraintb
Theoretical results MP2/6-311G(d,p)
Independent
p1
r(Si–N)
1.770(4)
1.881(2)
1.868(4)
1.459(3)
1.082(4)
1.091(2)
107.8(10)
109.7(7)
121.9(4)
61.1(5)
116.9(16)
111.4(6)
110.9(8)
117.5(11)
Ϫ66.4(75)
104.4(28)
Ϫ109.7(27)
1.753
1.884
1.872
p2
r(Si–C(4))
r(Si–C(3))
r(N–C)
p2 Ϫ p1 = 0.131(10)
p3 Ϫ p2 = Ϫ0.012(5)
p3
p4
1.462
p5
r(C–Hring
)
1.086/1.090
1.094–1.096
106.8
112.6
123.6
61.3
111.4
110.6
109.9
p6
r(C–Hmethyl
N–Si–C(4)
N–Si–C(3)
Si–N–C
)
p6 Ϫ p5 = 0.007(5)
p7
p8
p9
p10
p11
p12
p13
p14
p15
p16
p17
C–N–C
C4–Si–C5
Si–C(4)–H
Si–C(3)–H
N–C–H
τCSiCH
τCNCH(8)
τCNCH(9)
p13 Ϫ p12 = Ϫ0.7(10)
p8 = 118.8(15)
118.8
Ϫ57.6
106.3
p16 = 106.3(30)
p17 = Ϫ110.4(30)
Ϫ109.7
Dependent
p18
p19
p20
r(C–C)
C–C–N
Σ angles (at N)
1.484(7)
59.4(2)
305.0(11)
p18 Ϫ p4 = Ϫ0.028(10)
1.490
59.4
308.5
a Distances (ra) are in a˚ngströms and angles are in degrees. Parenthesised values are σ. For atom numbering see Fig. 3. b Derived from ab initio
calculations at MP2/6-311G(d,p), see next column. For derivation of uncertainties on restraints see text.
the SiC3 moiety in compound 1 has been observed in a range of
other compounds that contain the Me3SiN fragment with one
N–Si–C angle being different from the other two. The difference
between the two angles is small in 1 [N–Si–C4 107.8(10), N–Si–
C3 109.7(7)Њ], as in the case of (Me3Si)2NH [109.5 and 111.9Њ],25
while those in Me3SiNSO26 show a greater difference [101.5(18)
and 107.9(13)Њ].
The gas-phase geometry around the nitrogen atom of
C2SiNC2H4 in compound 1 is similar to that found in the solid
state for Ph3SiNC2H4,8 although the Si–N distances are differ-
ent [1.731(1) Å in Ph3SiNC2H4, 1.770(4) Å in 1]. The sum of
angles subtended at the nitrogen atom is 312.3Њ in Ph3SiNC2H4
[305.0(11)Њ in 1], and the declination of the Si–N vector from
the C2N plane is 47.2Њ(52.1Њ in 1). The solid state structure
of a related species having a four-membered NCCC ring,
Ph3SiNC3H6, was described as being much closer to the geom-
etry of a planar nitrogen centre.8
Fig. 4 Molecular structure of H3SiNC2H4 as predicted by ab initio
calculations at the QCISD/6-311G(d,p) level of theory.
Table 2 Ab initio molecular geometry of H3SiNC2H4 as calculated at
the MP2/6-311G(d,p) and the QCISD/6-311G(d,p) level of theory (core
electrons not included in the correlation calculations)
Parametera
MP2
QCISD
Taking the typical deviations between ra (electron diffraction)
and re parameters (calculations) into account, the gas-phase
geometry of Me3SiNC2H4 is in reasonable agreement with the
theoretical predictions obtained from Møller–Plesset calcu-
lations at the 6-311G(d,p) level of theory. The ra/re differences
are most pronounced in the Si–N distances [GED 1.770(4),
MP2 1.753 Å]. There is also a substantial deviation between
theory and experiment in the value of the C4–Si–C5 angle,
which lies between the two symmetry-equivalent methyl groups.
Refinement of experimental data gave 116.9(16)Њ, whereas a
value of 111.4Њ was calculated ab initio. To investigate the
discrepancy further refinements were executed. Initially the
angle under consideration was set at the ab initio value, and
secondly it was allowed to refine with a restraint of 1.0Њ in the
uncertainty. In both cases a substantially poorer fit to the
experimental scattering data was observed to that obtained
previously when the angle was allowed to refine freely. The
experimentally observed values for N–Si–C angles associated
with the geometry of the trimethylsilyl group also differ
somewhat from ab initio predictions. A value of 107.8(10)Њ was
obtained for N–Si–C(4) and N–Si–C(5) whereas it is calculated
at 106.8Њ. However, the in-plane angle N–Si–C(3) [109.7(7)Њ] is
closer to the ab initio value of 112.6Њ.
r(Si–N)
1.743
1.487
1.475
1.465
1.489
108.5
111.8
121.2
61.1
1.741
1.488
1.476
1.464
1.491
108.5
112.1
122.2
61.2
r(Si–H(3))
r(Si–H(4))
r(N–C)
r(C–C)
N–Si–H(4)
N–Si–H(3)
Si–N–C
C–N–C
Σ angles (at N)
303.5
305.6
a Distances (re) are in a˚ngströms and angles are in degrees.
not possible to compare experimental data for this compound
with those obtained for 1. Theoretical calculations for
H3SiNC2H4 were carried out using the same basis set and level
of theory as for 1 and, additionally, calculations with an
improved treatment of electron correlation at the QCISD/
6-311G(d,p) level. The results are shown in Table 2 and Fig. 4.
A comparison of the theoretical structure of H3SiNC2H4 with
that of 1 (Table 1) shows the geometries of the Si–N–C2 cores to
be similar. An improvement of the theoretical level does not
lead to significant changes in geometry. Thus the description of
the structure and the suggested bonding model (see below) for
1 is transferable to H3SiNC2H4.
As attempts to prepare H3SiNC2H4 were unsuccessful it was
J. Chem. Soc., Dalton Trans., 2000, 1491–1497
1493

View Article Online
with the simplistic expectation that a sp2 hybridised nitrogen
atom makes stronger bonds to its neighbours than one which is
sp3 hybridised. There is also a larger Si/N charge separation in
the transition state (the atomic charges are 1.378 e for Si and
Ϫ0.730 e for N, the “natural charges” being 1.92 and Ϫ1.00 e)
than the ground state (Si atom bears ϩ1.343 e and the N atom
Ϫ0.668 e, the “natural charges” being 1.90 and Ϫ0.95 e) accord-
ing to Mulliken population analyses.
We have performed further calculations for energy calibra-
tion purposes and to provide information suitable to rationalise
further the steeply pyramidal co-ordination geometry at nitro-
gen in compound 1. These results are compiled in Fig. 6. The
barrier to nitrogen inversion of H3SiNC2H4 was calculated at
the MP2/6-311G(d,p) to be 21.9 kJ molϪ1 and to be 22.4 kJ
molϪ1 at the more sophisticated QCISD(T)/6-311G(d,p) level
of theory [based on a QCISD/6-311G(d,p) geometry]. This
shows that the MP2/6-311G(d,p) level is suitable to provide a
realistic estimate of the height of this barrier. The barrier for
H3SiNC2H4 is close to that in 1, which leads to the conclusion
that the substituents (Me and H) at the silicon atom do not
change this barrier significantly.
Fig. 5 Molecular structure of the transition state to inversion/rotation
of Me3SiNC2H4 as predicted by ab initio calculations at the MP2/6-
311G(d,p) level of theory.
Table 3 Ab initio geometry of the transition state of Me3SiNC2H4 as
calculated at the MP2/6-311G(d,p) level of theory (core electrons not
included in the correlation calculations)
Parametera
Value
Calculations on N-silylazetidine H3SiN(CH2)3, which con-
tains a larger ring system than compound 1, and the open
chain systems H3SiNMe2 and H3SiNH2 have been performed
to demonstrate the effect of the R–N–R angle on the co-
ordination geometry at nitrogen in R2NSi systems. The
compound H3SiN(CH2)3 has a less steeply pyramidal nitrogen
co-ordination, i.e. the declination of the Si–N vector from the
C2N plane is 33.8Њ [52.1(5)Њ in 1]. However the ring strain
exerted by the four- and three-membered rings in H3SiN(CH2)3
and H3SiN(CH2)2 (and 1) leads to Si–N–C angles in these com-
r(Si–N)
1.718
1.873
1.878
1.431
1.433
1.505
106.0
110.9
148.4
148.3
63.4
r(Si–C(3))
r(Si–C(4))
r(N–C(7))
r(N–C(4))
r(C(3)–C(4))
N–Si–C(5)
N–Si–C(3)
Si–N–C(7)
Si–N–C(6)
C(6)–N–C(7)
8
a Distances (re) are in a˚ngströms and angles are in degrees.
pounds [125.9 and 121.2Њ, compare that in Ph3SiN(CH2)3
127.6(1)Њ], which are slightly larger than those in the open-chain
compounds H3SiNMe2 and H3SiNH2, the latter adopting
Si–N–H angles of 119.4 and 119.6Њ and almost flat nitrogen
co-ordination spheres. The barrier to nitrogen inversion of
these open-chain systems is consequently much lower: 4.9 kJ
molϪ1 for H3SiNMe2 and 1.9 kJ molϪ1 H3SiNH2 at MP2/
6-311G(d,p). Improvement of the treatment of electron corre-
lation in the calculations for H3SiNH2 does not change
the height of this barrier substantially, as it is predicted to be
2.3 kJ molϪ1 at QCISD(T)/6-311G(d,p) on the basis of a
QCISD/6-311G(d,p) geometry (compare 2.1 kJ molϪ1 in an
earlier calculation by Gordon27).
Calculations of the transition state to nitrogen inversion
The main question that remains to be addressed is the reason
for the non-planar co-ordination geometry of the nitrogen
atom. A transition state having a planar C2NSi skeleton was
located by ab initio calculations. Such a geometry becomes a
transition state to nitrogen inversion only if the Me3Si group is
concurrently rotated such that one of the methyl carbon atoms
becomes coplanar with the C2NSi core. The transition state is
characterised by the occurrence of one imaginary frequency in
the calculations such that it is a combined rotation/inversion
process, i.e. rotation about the Si–N bond and inversion of
the nitrogen centre. This transition state is predicted to be
21.4 kJ molϪ1 higher in energy than the ground state [MP2/
6-311G(d,p)], which is a substantial amount relative to the very
low or even non-existent barriers in non-cyclic silylamines.
The geometry of the transition state is shown in Fig. 5 and
the important geometric parameters are listed in Table 3. On
adopting planarity the angles of the C2NSi core open (Si–N–
C(6) 148.3, cf. 121.9(4)Њ in ground-state experimental structure;
C(6)–N–C(7) 63.4, cf. 61.1(5)Њ in ground-state experimental
structure). This places the NC(6)C(7) ring under some strain
compared to that in the parent 1-methylaziridine structure.24
The biggest difference to the ground-state geometry is in the
Si–N bond length, which is much shorter in the planar trans-
ition state (1.718 Å) than that calculated in the ground state
(1.753 Å). The natural bond orbital (NBO) analysis of the Si–N
bond in the transition state geometry indicates it consists of Si
sp2.25d0.05 and N sp1.76 hybrid orbitals, which represents a much
smaller p contribution than in the pyramidal ground state,
where the Si–N bond is described to be formed by Si sp3.35d0.08
and N sp1.78 hybrid orbitals. It is worthy of note that it is the
hybridisation of the silicon atom which changes between the
transition state and ground state. In contrast the nitrogen atom,
around which the co-ordination geometry undergoes a major
change, retains the same hybridisation. This is not consistent
Conclusion
The determination of the gas-phase structure of Me3SiNC2H4
shows the steeply pyramidal nitrogen co-ordination to be an
inherent phenomenon of the class of N-silylaziridines. This
geometry is not imposed on the molecules by packing or dipole
forces when studied in their crystalline forms. The pyramidal
nitrogen co-ordination restricts the Si–N–C angles to 121.9(4)Њ
(cf. 148.4Њ calculated for the planar transition state) and thus
circumvents angle strain in addition to the ring strain of the
aziridine ring.
Experimental
N-Trimethylsilylaziridine was prepared from aziridine (2.4 cm3,
46 mmol), which was deprotonated in diethyl ether (20 cm3) by
46 mmol of n-butyllithium (1.6 M in hexane) at Ϫ78 ЊC. Chloro-
trimethylsilane (50 mmol) was added, the resulting suspension
stirred for 1 h, and warmed to ambient temperature. The
volatile products were distilled off in vacuo and separated by a
trap to trap condensation, with the product being held in a
Ϫ45 ЊC trap and obtained in 84% yield. N-Trimethylsilylazirine
has been prepared before by another procedure9 and our
sample has the same properties as described therein. The
following list contains new spectroscopic data. The NMR
1494
J. Chem. Soc., Dalton Trans., 2000, 1491–1497

View Article Online
Fig. 6 Calculated molecular structures and selected geometrical parameters of the ground states (GS) of Me3SiNC2H4 1, H3SiNC2H4,
H3SiN(CH2)3, H3SiNMe2 and H3SiNH2 at the MP2/6-311G(d,p) level of theory. The geometries and relative energies for the transition states (TS) of
nitrogen inversion are given for 1, H3SiNC2H4, H3SiNMe2 and H3SiNH2. For the model systems H3SiNC2H4 and H3SiNH2 energy differences are
also provided at the QCISD(T)/6-311G(d,p) level of theory, which are based on geometries optimised at QCISD/6-311G(d,p).
Table 4 Camera distances, data ranges, trapezoidal weighting function points, correlation parameters, scale factors, wavelengths and R values for
the GED structure determination of Me3SiNC2H4
smin
/
smax
/
s1/
s2/
Data
set
Camera
distance/mm
Sample
T/ЊC
Nozzle
T/ЊC
∆s/
Correlation
parameter
Scale
factor
Wavelength/
Å
ÅϪ1
ÅϪ1
ÅϪ1
ÅϪ1
ÅϪ1
Rg
1
2
496.12
245.54
33–36
38–40
42–47
48–50
0.2
0.4
2.4
6.8
12.8
27.6
4.4
8.8
11.8
23.6
0.1524
Ϫ0.2077
0.454(4)
0.413(10)
0.058561
0.058561
0.0686
0.0581
spectra were recorded at 21 ЊC on a JEOL JNM-LA400 spec-
trometer from a sample in a sealed tube with C6D6 as a solvent
directly condensed onto the sample from K/Na alloy: 1H NMR
δ 0.0 (s, 9 H, H3C) and 1.52 (s, 4 H, H2C); 13C NMR δ Ϫ2.87 (qp,
NC2H4 were recorded on Kodak Electron Image film using the
University of Reading diffraction apparatus operating at 42 kV
accelerating voltage.28 Two sets of data were recorded at each
of the two camera distances using nozzle temperatures between
42 and 50 ЊC. The wavelength (0.058561 Å) was calibrated
against diffraction patterns of benzene. Optical densities were
measured using a commercial AGFA II scanner.29,30 Further
experimental conditions and general parameters concerning
the refinements are listed in Table 4. The least squares refine-
ments were carried out using the program ED9631 and applying
the scattering factors established by Fink and co-workers.32
The refined molecular parameters, their definition and the
applied restraints, with selected interatomic distances including
vibrational amplitudes and applied restraints, are given in Table
5; elements of the correlation matrix are given in Table 6.
4
1
1JCH = 118.3, JCH = 1.8, (CH3)3Si) and 20.07 (qt, JCH = 168.8,
3JCNCH = 2.7 Hz, (CH3)2N); 15N-{1H} NMR δ Ϫ348.5 (s,
1JNC = 2.9, 1JSiN = 4.8 Hz); 29Si NMR δ Ϫ19.7 (m). A vibrational
spectrum was recorded in a gas cell in an FTIR spectrometer
(Midac Prospect FTIR): 3042m, 2965s, 2913m, 2880w, 1288s,
1276s, 1258s, 1143m, 1073w, 955s, 949vs, 943s, 845vs, 744m,
682w, 649w, 556w cmϪ1
.
Electron diffraction experiments
GED data. Electron scattering intensity data for Me3Si-
J. Chem. Soc., Dalton Trans., 2000, 1491–1497
1495

View Article Online
Table 5 Selected interatomic distances (ra/Å), amplitudes of vibration (u/Å) and amplitude restraints as obtained in the least squares refinement of
Me3SiNC2H4
Atom paira
Distance
Amplitude
Restraintb
d1
Si(1)–N(2)
Si(1)–C(3)
Si(1)–C(4)
N(2)–C(6)
C(6)–C(7)
C(6)–H(8)
C(3)–H(12)
Si(1) ؒ ؒ ؒ C(6)
N(2) ؒ ؒ ؒ C(3)
N(2) ؒ ؒ ؒ C(4)
C(3) ؒ ؒ ؒ C(4)
C(4) ؒ ؒ ؒ C(5)
C(3) ؒ ؒ ؒ C(6)
C(4) ؒ ؒ ؒ C(6)
C(4) ؒ ؒ ؒ C(7)
N(2) ؒ ؒ ؒ H(8)
N(2) ؒ ؒ ؒ H(9)
C(6) ؒ ؒ ؒ H(10)
C(6) ؒ ؒ ؒ H(11)
Si(1) ؒ ؒ ؒ H(12)
Si(1) ؒ ؒ ؒ H(13)
Si(1) ؒ ؒ ؒ H(15)
Si(1) ؒ ؒ ؒ H(16)
Si(1) ؒ ؒ ؒ H(17)
1.770(4)
1.868(4)
1.881(2)
1.459(3)
1.484(7)
1.082(4)
1.091(2)
2.827(4)
2.939(18)
2.985(13)
2.999(27)
3.206(28)
3.295(25)
4.316(9)
3.724(11)
2.182(13)
2.145(13)
2.177(26)
2.213(86)
2.478(11)
2.507(11)
2.495(7)
2.503(7)
2.502(7)
0.051(3)
0.058(3)
0.058(3)
0.046(3)
0.044(3)
0.064(2)
0.064(2)
0.044(10)
0.041(14)
0.045(15)
0.096(11)
0.113(11)
0.184(16)
0.121(9)
0.266(22)
0.106(10)
0.106
0.106
0.106
0.099(5)
0.097
0.097
0.097
0.097
u1 = 0.049(5)
d2
u2/u1 = 1.096(55)
u3/u2 = 0.983(49)
d3
d4
d5
u5/u4 = 0.961(48)
Tied to u6
d6
d7
d8
d9
d10
d11
d12
d13
d14
d15
d16
d17
d18
d19
d20
d21
d22
d23
d24
u10/u9 = 1.094(55)
u11 = 0.106(11)
u12 = 0.108(11)
u13 = 0.186(19)
u14 = 0.115(12)
u15 = 0.257(26)
Tied to u16
Tied to u16
Tied to u16
Tied to u20
Tied to u20
Tied to u20
Tied to u20
a Uncertainties represent σ. For atom numbering see Fig. 3. b Restraints derived from ab initio at MP2/6-31G(d).
Table 6 Correlation matrix elements (×100) for Me3SiNC2H4 with absolute values >50%
p1
p4
p8
p11
p12
u1
u2
u4
u5
u6
u7
u9
u10
u16
p9
Ϫ61
55
51
p10
p11
p15
u2
Ϫ90
100
59
59
64
76
55
100
62
u3
u5
55
61
100
55
u7
100
65
u8
70
65
Ϫ51
79
98
80
u10
u20
k2
55
65
GED model. The geometrical model for Me3SiNC2H4 was
defined in Cs symmetry. The atom numbering scheme is
provided in Fig. 3. While fixing the differences between the
parameters defining hydrogen atom positions (C–H distances,
angles and torsion angles) to calculated values, a total of 17
parameters was used to refine the structures. The parameter
definitions are listed in Table 1. Eight geometrical restraints
based on the ab initio calculated values (MP2/6-311G(d,p))
were used to reduce correlation between parameters of similar
nature (differences) or those describing hydrogen positions
(absolute restraints). Their definition and values with assigned
uncertainties based on an educated guess (experience with
related structures) are also given in Table 1.
All vibrational amplitudes belonging to a pair of scatterers
contributing more than 5% of the most important pair (Si–N)
were refined. Restraints for these amplitudes were calculated
from harmonic force fields obtained at the MP2/6-31G(d) level
of theory, which were transformed into amplitudes by means
of the program ASYM40.17 Ratios between amplitudes were
assigned an uncertainty of 10%, absolute restraints 20% of the
calculated value. These amplitudes and restraints are given in
Table 5.
optimisations and vibrational frequency calculations were per-
formed from analytic first and second derivatives at the SCF
and MP2 levels of theory. Calculations were undertaken at the
SCF level using the standard 3-21G(d)34–36 and 6-31G(d)37–39
and 6-311G(d,p) basis sets, while the larger basis set was used
for calculations at the MP2 level. For two model compounds
higher level geometry optimisations were also undertaken with
quadratic configuration interaction calculations including
single and double substitutions (QCISD) and the energies were
calculated single point further including the triple contribution
substitutions [QCISD(T)].40
References
1 H. Bock, J. Göbel, Z. Havlas, S. Liedle and H. Oberhammer, Angew.
Chem., Int. Ed. Engl., 1991, 30, 187.
2 R. Boese, D. Bläser, M. Y. Antipin, V. Chaplinski and A. de Meijere,
Chem. Commun., 1998, 781.
3 G. Gundersen, R. A. Mayo and D. W. H. Rankin, Acta Chem.
Scand., Ser. A, 1984, 38, 579.
4 K. Ruhlandt-Senge, R. A. Bartlett, M. M. Olmstead and P. P. Power,
Angew. Chem., Int. Ed. Engl., 1993, 33, 425.
5 N. W. Mitzel, E. Breuning, A. J. Blake, H. E. Robertson, B. A. Smart
and D. W. H. Rankin, J. Am. Chem. Soc., 1996, 118, 2664.
6 N. W. Mitzel and H. Oberhammer, Inorg. Chem., 1998, 37, 3539.
7 Y. Mo, Y. Zhang and J. Gao, J. Am. Chem. Soc., 1999, 121,
5737.
Ab initio calculations. Ab initio molecular orbital calculations
were carried out using the Gaussian 94 program.33 Geometry
1496
J. Chem. Soc., Dalton Trans., 2000, 1491–1497

View Article Online
8 G. Huber, A. Jockisch and H. Schmidbaur, Eur. J. Inorg. Chem.,
1998, 1, 107.
9 O. J. Scherer and M. Schmidt, Chem. Ber., 1965, 98, 2243.
10 N. S. Nametkin, V. N. Perchenko and L. G. Batalova, Dokl. Chem.
(Engl. Transl.), 1964, 158, 935.
11 C. Glidewell and D. W. H. Rankin, J. Chem. Soc. A, 1970, 279.
12 H. C. Marsmann and H. G. Horn, Z. Naturforsch., Teil B, 1972, 27,
1448.
13 B. Heinz, H. Marsmann and U. Niemann, Z. Naturforsch., Teil B.,
1977, 32, 163.
14 N. W. Mitzel, J. Riede, A. Schier, M. Paul and H. Schmidbaur,
Chem. Ber., 1993, 126, 2027.
15 N. W. Mitzel, M. Hofmann, E. Waterstradt, P. v. R. Schleyer and
H. Schmidbaur, J. Chem. Soc., Dalton Trans., 1994, 2503.
16 A. J. Blake, P. T. Brain, H. McNab, J. Miller, C. A. Morrison,
S. Parsons, D. W. H. Rankin, H. E. Robertson and B. A. Smart,
J. Phys. Chem., 1996, 100, 12280.
17 L. Hedberg and I. M. Mills, ASYM20, ASYM40, Programs for
force constants and normal coordinate analysis, Version 3.0, June
1994. See also: L. Hedberg and I. M. Mills, J. Mol. Spectrosc., 1993,
160, 117.
18 A. J. Blake, E. A. V. Ebsworth, D. W. H. Rankin, H. E. Robertson,
D. E. Smith and A. J. Welch, J. Chem. Soc., Dalton Trans., 1986, 91.
19 D. G. Anderson, D. W. H. Rankin, H. E. Robertson, G. Gundersen
and R. Seip, J. Chem. Soc., Dalton Trans., 1990, 161.
20 C. M. Huntley, D. W. H. Rankin and H. E. Robertson, J. Mol.
Struct., 1987, 156, 103.
26 K. I. Gobbato, C. O. Della Vedova and H. Oberhammer, J. Mol.
Struct., 1995, 350, 227.
27 M. S. Gordon, Chem. Phys. Lett., 1986, 126, 451.
28 C. J. Holwill, PhD Thesis, University of Reading, 1987.
29 S. Gundersen and T. G. Strand, J. Appl. Crystallogr., 1996, 29, 638.
30 K. Aarset, K. Hagen, E. M. Page and D. A. Rice, J. Mol. Struct.,
1999, 478, 9.
31 N. W. Mitzel, P. T. Brain and D. W. H. Rankin, ED96, Version 2.0,
1998. A program developed on the basis of formerly described ED
programs: A. S. F. Boyd, G. S. Laurenson and D. W. H. Rankin,
J. Mol. Struct., 1981, 71, 217.
32 A. W. Ross, M. Fink and R. Hilderbrandt, International Tables
for X-Ray Crystallography, ed. A. J. C. Wilson, Kluwer Academic
Publishers, Dodrecht, Boston, 1992, vol. C, p. 245.
33 Gaussian 98, Revision A.6, M. J. Frisch, G. W. Trucks, H. B.
Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, V. G.
Zakrzewski, J. A. Montgomery, Jr., R. E. Stratmann, J. C. Burant,
S. Dapprich, J. M. Millam, A. D. Daniels, K. N. Kudin, M. C.
Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi, R. Cammi,
B. Mennucci, C. Pomelli, C. Adamo, S. Clifford, J. Ochterski, G. A.
Petersson, P. Y. Ayala, Q. Cui, K. Morokuma, D. K. Malick, A. D.
Rabuck, K. Raghavachari, J. B. Foresman, J. Cioslowski, J. V. Ortiz,
B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi,
R. Gomperts, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham,
C. Y. Peng, A. Nanayakkara, C. Gonzalez, M. Challacombe,
P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong, J. L. Andres,
C. Gonzalez, M. Head-Gordon, E. S. Replogle and J. A. Pople,
Gaussian, Inc., Pittsburgh, PA, 1998.
34 J. S. Binkley, J. A. Pople and W. J. Hehre, J. Am. Chem. Soc., 1980,
102, 939.
35 M. S. Gordon, J. S. Binkley, J. A. Pople, W. J. Pietro and W. J. Hehre,
J. Am. Chem. Soc., 1982, 104, 2797.
36 W. J. Pietro, M. M. Francl, W. J. Hehre, D. J. Defrees, J. A. Pople and
J. S. Binkley, J. Am. Chem. Soc., 1982, 104, 5039.
37 W. J. Hehre, R. Ditchfield and J. A. Pople, J. Chem. Phys., 1972, 56,
2257.
38 P. C. Hariharan and J. A. Pople, Theor. Chim. Acta, 1973, 28, 213.
39 M. S. Gordon, Chem. Phys. Lett., 1980, 76, 163.
40 J. A. Pople, M. Head-Gordon and K. Raghavachari, J. Chem. Phys.,
1987, 87, 5968.
21 S. Cradock, C. M. Huntley and J. R. Durig, J. Mol. Struct., 1985,
127, 319; A. J. Careless, M. C. Green and H. W. Kroto, Chem. Phys.
Lett., 1972, 16, 414; K. Kimura, K. Katada and S. H. Bauer, J. Am.
Chem. Soc., 1966, 88, 416.
22 C. Glidewell, D. W. H. Rankin, A. G. Robiette and G. M. Sheldrick,
J. Mol. Struct., 1970, 6, 231.
23 W. R. Roper and C. J. Wilkins, Trans. Faraday Soc., 1962, 58,
1686.
24 N. W. Mitzel, J. Riede and C. Kiener, Angew. Chem., Int. Ed. Engl.,
1997, 36, 2215.
25 T. Fieldberg, J. Mol. Struct., 1984, 112, 159; A. G. Robiette, G. M.
Sheldrick, W. S. Sheldrick, B. Beagley, D. W. J. Cruickshank, J. J.
Monaghan, B. J. Aylett and I. A. Ellis, Chem. Commun., 1968, 909.
J. Chem. Soc., Dalton Trans., 2000, 1491–1497
1497