On the presence or absence of geminal Si...N interactions (α-effect) in pentafluorophenylsilyl compounds with SiCN, SiNN and SiON backbones

Article abstract of DOI:10.1039/b808388f

The silanes C₆F₅SiF₂CH₂NMe ₂ (1), C₆F₅SiF₂N(SiMe ₃)NMe₂ (2) and C₆F₅SiF ₂ONMe₂ (3) with pentafluoroph

Full text of DOI:10.1039/b808388f

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
On the presence or absence of geminal Si ◊ ◊ ◊ N interactions (a-effect) in
pentafluorophenylsilyl compounds with SiCN, SiNN and SiON backbones†
Markus Woski, Raphael J. F. Berger and Norbert W. Mitzel*
Received 19th May 2008, Accepted 11th July 2008
First published as an Advance Article on the web 8th September 2008
DOI: 10.1039/b808388f
The silanes C6F5SiF2CH2NMe2 (1), C6F5SiF2N(SiMe3)NMe2 (2) and C6F5SiF2ONMe2 (3) with
pentafluorophenyl substituents and geminal N atoms have been prepared by the reaction of C6F5SiF3
with LiCH2NMe2, LiN(SiMe3)NMe2 and LiONMe2, respectively. The compounds have been
characterised by spectroscopic methods and crystal structure determination. Comparison of measured
and calculated IR spectra has provided insight into the conformational composition of the vapour of 3.
Whereas 2 and 3 show interactions between the geminal Si and N atoms, 1 does not. Further analysis of
the bonding situation has been undertaken by quantum chemical calculations of the rotation and
bending potentials of the C6F5SiF2–X–NMe2 units.
14
and F3SiN(SiMe3)NMe2 for the hydrazines. Despite the many
Introduction
structural parameters determined in the solid and the gas phase,
The last 40 years have seen many debates about the so called a-
effect in silicon chemistry. This was initially deduced from reduced
basicities of a-silylated (germylated and stannylated) amines as
compared to alkylated ones.1 The reason for this effect was
discussed in terms of the postulate of a dative bond between the
nitrogen and the silicon atom, i.e. the acceptor quality of silicon
reducing in this way the N basicity. It should be noted, that the term
“a-effect” is also used in a different context in silicon chemistry to
describe the stabilisation of a-carbanions by silyl groups;2 these
concepts should not be confused.
the Si ◊ ◊ ◊ N interaction remained difficult to describe, as only
very little electron density is found between Si and N atoms.
In electron density topologies analysed in terms of the quantum
theory of atom in molecules (QTAIM) no bond critical points for
the Si ◊ ◊ ◊ N interactions are found.13 Moreover, sophisticated levels
of theory are necessary for a correct reproduction of exper◦imental
findings.15 In the gas phase F3SiCH2NMe2 has an 110.3(7) Si–C–
N angle (gas electron diffraction), not at all indicating an attractive
Si ◊ ◊ ◊ N intera◦ction16 [as also found in H3SiCH2NMe2, angle Si–
C–N 114.7(3) ].17 By contrast to this and the monomeric nature
of all SiON and SiNN compounds studied in the solid state so far,
F3SiCH2NMe2 is a dimer with two intermolecular N→Si dative
bonds in a six-membered Si2N2C2 ring in its crystals. It adopts even
widened Si–C–N angles of 120.3(1) and 120.4(1)◦ in the crystalline
phase.16
In contrast to SiON and SiNN backbones, no example for the
SiCN compounds has been found so far, which shows the described
features of small Si–C–N angles or short Si ◊ ◊ ◊ N distances. This
was surprising as these are exactly those compounds for which an
“a-effect” is claimed and found to be active, but obviously has a
different origin than that proposed in the original postulate.
Instead of intriguing structural features, F3SiCH2NMe2 has an
unexpectedly flat bending potential for the SiCN unit and a charge
distribution inconsistent with the picture of a classical Si ◊ ◊ ◊ N
interaction.16
Many additional properties of a-silylated amines and other
geminal systems such as magnetic susceptibility, dipole moments,3
various spectroscopic parameters4 and structural parameters5
have since then been attributed to the a-effect. The higher
reactivity of the silicon atom in such compounds is manifest from
the uncommon ease of cleavage of the Si–C(N) bond6 and from the
observed increased reaction speed towards substitution reactions
at Si. The latter effect has found widespread application in the
chemistry of organofunctional silanes and led to the development
of the so called a-silanes, which are nowadays employed as surface
mediators, adhesion promoters and generally as covalent linkers
between the inorganic and organic parts of composite materials.7
Despite this, the a-effect is not really well understood in terms
of its electronic details. We have shown that there are classes of
compounds with Si–X–N backbones (X = spacer group), clearly
featuring an attractive force between the Si and N atoms. This
holds true for X = O and N, i.e. the O-silylhydroxylamines8 and -
oximes9 and the silylhydrazines.10 The examples with the most pro-
nounced of such interactions are compounds displaying the feature
of a three-membered SiXN ring with short Si ◊ ◊ ◊ N distances and
small Si–X–N angles. Such examples include ClH2SiONMe2,11
In this contribution we present a set of pentafluorophenylsilanes
with SiON, SiNN and SiCN backbones, which allows for the first
time a consistent comparative study of the ground state structures
of a series of compounds monomeric in the solid state and with
the same substitution pattern at the silicon atom.
12
13
F3SiONMe2 and (F3C)F2SiONMe2 for the hydroxylamines
Results and discussion
Fakulta¨t fu¨r Chemie, Universita¨t Bielefeld, Universita¨tsstrasse 25, 33615,
Bielefeld, Germany. E-mail: mitzel@uni-bielefeld.de; Fax: +49 521 106
6029; Tel: +49 521 106 6163
Synthesis
The three compounds C6F5SiF2CH2NMe2 (1), C6F5SiF2N-
(SiMe3)NMe2 (2) and C6F5SiF2ONMe2 (3) were prepared by
† CCDC reference numbers 688419–688421. For crystallographic data in
CIF or other electronic format see DOI: 10.1039/b808388f
5652 | Dalton Trans., 2008, 5652–5658
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
similar procedures from C6F5SiF3 with the respective lithiated
reagents LiONMe2, LiN(SiMe3)NMe2 and LiCH2NMe2 to yield
the products by salt elimination.
C6F5SiF3 + LiXNMe2 → C6F5SiF2–X–NMe2 + LiF
[X = CH2 (1), N(SiMe3) (2), O (3)]
(1)
Compound 1 is extremely hydrophilic and has to be strictly
handled under inert gas. It can be crystallised at low temperatures
from solution, but is a liquid at ambient temperature (mp -5 ◦C).
Compound 2 can be recrystallised at low temperatures to give thin
needles melting at 35 ◦C. All three compounds were identified and
checked for purity by 1H, 13C, 19F and 29Si NMR (for 3) and further
characterised by IR and MS spectra.
Fig. 2 Measured gas phase IR spectrum of compound 3.
NMR spectroscopy
1
The signal for the NMe2 group of 1 is found in the H NMR
spectrum at 2.32 ppm in C6D6 solution (compare: F3SiCH2NMe2
1.96 ppm).16 In contrast, the corresponding resonances of com-
pounds 2 (2.54 ppm) and 3 (2.68 ppm) are found at higher
frequencies. Very similar in their chemical shifts are the 13C NMR
data (1: 49.4 ppm, 2: 48.0 ppm, 3: 49.5 ppm, F3SiCH2NMe2:
48.5 ppm). This indicates that the electronic situation at the
nitrogen donor centre is not very different for the three attached
spacer groups in solution. The fluorine atoms at the acceptor atom
Si also lead to surprisingly similar chemical shifts in the 19F NMR
spectra. Unfortunately we could not obtain meaningful 29Si NMR
spectra for 1 and 2, probably owing to the weakness of the signals
due to their multiple splitting by coupling to F nuclei. The 29Si
NMR shift of -48.3 ppm for 3 is between those of Cl3SiONMe2
and H3SiONMe2 (-42.4 ppm and 49.5 ppm).4,5
Compound 3 can be crystallised from the melt by application
of low-temperature in situ techniques (see below). It can be
evaporated and transferred in a high vacuum line and this allowed
measurements of a gas phase IR spectrum. Calculated vibrational
spectra were derived from a force field obtained at the B3LYP/6-
31G* level of theory. A comparison between calculated and
measured spectra can be used to derive information about the
conformation of the free molecules of 3. In particular the region
between 1200 and 700 cm-1 is very sensitive to the conformation of
the molecules in this case as is shown in Fig. 1. A comparison with
the calculations indicates strongly the predominant presence of the
gauche conformer of 3 in the gas phase. The obtained simulation
of the IR spectrum of the gauche conformer fits very well with
the experimental spectrum as shown in Fig. 2. Comparison with
the calculated spectra also suggests that some smaller bands, in
particular those at 1538, 1078 and 1029 cm-1, are likely to stem
from the second, less abundant, anti conformer in the vapour
phase.
Molecular and crystal structures
Single crystals of compounds 1, 2 and 3 were grown by three
different methods: 1 by the OHCD-method18 (optical heating and
crystallisation device, a laser-assisted microscale zone-melting), 2
by crystallisation from solution and 3 by in situ methods from a
solid–liquid equilibrium (Fig. 3, 4 and 5). The structural data are
collected in Table 1 along with the results of ab initio calculations
at the MP2/TZVPP level of theory.
Fig. 3 Molecular structure of compound 1 in the crystal as obtained
by low-temperature X-ray crystallography. The displacement factors are
represented by ellipsoids at the 50% probability level.
Fig. 1 Calculated (B3LYP/6-31G*) gas phase IR spectra in the region
between 1800 and 500 cm-1 for the anti and gauche conformer of
compound 3.
All three compounds are monomeric in their crystals, which has
an analogy in the related compounds F3SiXNMe2 for X = NR
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 5652–5658 | 5653
©

View Article Online
◦
˚
Table 1 Selected bond lengths [A], atom distances, angles [ ] and torsion angles of C6F5SiF2–X–NMe2 [X = CH2 (1), N(SiMe3) (2), O (3)] as determined
by low temperature X-ray crystallography (XRD) and by ab initio calculations (MP2/TZVPP)
1
2
3
XRD
MP2
1.597
XRD
MP2
1.606
XRD
MP2
1.597
d(Si–Fanti
)
1.569(3)
1.583(2)
1.580(3)
d(Si–Fgauche
d(Si–X)
)
1.540(4)
1.831(5)
1.488(6)
2.676(4)
1.421(8)
1.445(8)
1.868(4)
1.596
1.867
1.460
2.750
1.454
1.454
1.870
1.574(3)
1.660(3)
1.494(5)
2.241(4)
1.461(5)
1.449(5)
1.854(4)
90.4(2)
1.598
1.698
1.461
2.230
1.455
1.456
1.872
1.576(2)
1.633(3)
1.495(4)
2.117(3)
1.455(5)
1.466(5)
1.857(4)
85.1(2)
103.4(2)
117.2(2)
66.3
1.593
1.662
1.481
2.140
1.455
1.453
1.868
d(X–N)
d(Si ◊ ◊ ◊ N)
d(N–C1)
d(N–C2)
d(Si–C11)
∠(Si–X–N) (a)
∠(F–Si–X) (b)
∠(C11–Si–X) (g)
t(C11SiXN)
107.0(3)
110.8
89.5
85.6
108.6(2)
110.4(2)
55.6(4)
26.5
109.9
109.8
29.4
109.3(2)
115.8(2)
65.5(2)
109.4
112.8
63.1
2.4
104.5
115.6
63.9
2.1
t(lpNNXSi)
47.3
3.1
1.7
Fig. 5 Molecular structure of compound 3 in the crystal as obtained
by low-temperature X-ray crystallography. The displacement factors are
represented by ellipsoids at the 50% probability level.
Fig. 4 Molecular structure of compound 2 in the crystal as obtained
by low-temperature X-ray crystallography. The displacement factors are
represented by ellipsoids at the 50% probability level.
and O, but is in variance to them for X = CH2 (F3SiCH2NMe2 is
a dimer in the solid). Thus it is here for the first time that these
situations can be compared directly with reliable data from the
solid state. Moreover, all three compounds are found in the same
conformation with the pentafluorophenyl group in gauche position
[t(C11SiXN) ª 60◦], as was already indicated to be the most stable
conformer in the gas phase. It is therefore the most electronegative
substituent, fluorine, which is found in anti position relative to the
Si–X–N skeleton—a situation that was not unexpected, as in a
number of related cases different preferences were found.11,12
Whereas in the cases of 2 and 3, the conformation of the NMe2
group relative to the N–X–Si skeleton is such that these parts
of the structures are almost of mirror symmetry, as is necessary
for maximising the Si ◊ ◊ ◊ N interactions, in compound 1 there
is a marked deviation from such a conformation. As shown in
Fig. 6 the position of a hypothetical lone pair (lp) of electrons
at nitrogen (calculated as a vector enclosing three equal angles
lp–N–C) deviates by 26.5◦ from an idealised position of the
Fig. 6 View of compound 1 along the N–Si axis showing the orientation
of the NMe2 group relative to the SiCN plane. lpN denotes the idealised
position of the lone pair of electrons at the nitrogen atom and its deviation
from a hypothetical orientation in the SiCN plane shown in faded print
(including the corresponding hypothetical positions of the atoms of the
NMe2 group).
NMe2 group, lying in the SiCN plane (by contrast, the lone
pairs are approximately aligned in the SiON and SiNN planes
5654 | Dalton Trans., 2008, 5652–5658
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
in compounds 2 and 3). This leads to a substantial difference
between the torsional angles Si–C–N–C to both Me groups being
94.1 and -144.3◦. In contrast the corresponding torsional angles
for compound 2 are 113.2 and -119.5◦ and for 3 they are 118.1
and -122.2◦.
the compounds with SiNRNMe2 and SiONMe2 units have smaller
Si–N–N and Si–O–N angles for an F3 rather than an (F5C6)F2
substitution pattern at silicon.
The Si–N–N angle in 2 is 90.4(2)◦ and the Si ◊ ◊ ◊ N distance is
˚
2.241(4) A. These values are much smaller than the corresponding
Comparing the backbone structures of compounds 1–3, only
the spacer group X (X = CH2, N(SiMe3), O) varies and affects
the Si ◊ ◊ ◊ N interaction, which is structurally characterised by the
Si–X–N angle and the Si ◊ ◊ ◊ N distance. In the case of 1 the
ones in the aminomethylsilane 1. Ab initio calculations at the
MP2(RI)/TZVPP level of theory predicted this angle to be
89.5◦, which is close to the crystal structure value. This angle is
about seven degrees wider than in the trifluorosilyl compound
F3SiN(SiMe3)NMe2, but is equal to the one found in the N-
trimethylstannyl substituted compound F3SiN(SnMe3)NMe2 with
89.6◦.14
˚
Si ◊ ◊ ◊ N distance of 2.676(4) A and the corresponding Si–C–N
angle of 107.0(3)◦ are not indicative of an attractive Si ◊ ◊ ◊ N
interaction. A comparison of this angle determined in the solid
state, i.e. in a polar surrounding of many molecules with relatively
high dipole moments, with the situation of the free molecule
described by the calculations shows a deviation of almost 4◦,
the largest deviation for the series 1–3. These 107.0(3)◦ for the
Si–C–N angle of 1 are, however, much smaller t◦han the Si–
C–N angle in the dimer of F3SiCH2NMe2 (120.4 ).16 We can
also compare the structure of compound 1 with the zwitterionic
ammoniomethyl(trifluorosilicate) F4SiCH2NHMe2, which is not
capable of forming a direct Si ◊ ◊ ◊ N interaction due to the already
hypercoordinate Si and tetracoordinate N atom. In this compound
the Si–C–N angles are 118.5 and 119.2◦ for two independent
molecules in the crystal.19
In compound 2 the Si–N–Si angle is very wide at 140.7(2)◦ and
marks the substantial deformation of the planar coordinate spacer
nitrogen atom (sum of angles 359.5◦). The C5-methyl group of the
SiMe3 unit is slightly bent towards the fluorine atom connected
to the silicon in anti position relative to N, which results in a
˚
distance H51 ◊ ◊ ◊ F of about 2.60 A. This interpretation of an
attractive F ◊ ◊ ◊ H interaction is supported by an N1–Si2–C5 angle
of 104.6◦, which is much smaller than the angles N1–Si2–C3
(111◦) and <N1Si2C4 (112◦), respectively. In contrast to 1 the
orientation of the lone pair at the NMe2 nitrogen (calculated
idealised orientation) is oriented towards the silicon atom, as is
described by a torsional angle lpN–N1–N2–Si1 of 3.1◦.
In contrast to 1 the other two compounds 2 and 3 have
small Si–X–N angles [2: 90.4(2)◦, 3: 85.1(2)◦], clearly indicating
attractive Si ◊ ◊ ◊ N interactions to be operative. This leads to
Likewise as in compound 1 and 2 the pentafluorophenyl group
in 3 is placed in gauche position with a torsion angle t(C11–Si–O–
N) = 66.3(2)◦. This was also predicted by ab initio calculations
at a MP2(RI)/TZVPP level of theory. As expected, 3 shows
the smallest Si–X–N angle [<SiON = 85.1(2)◦] of all three
pentafluorophenyl compounds in this paper. This Si–O–N angle is
thus more than 5◦ smaller than the Si–N–N angle in 2 [<Si–N–N =
90.4(2)◦] and 12◦ smaller than the Si–C–N angle in 3 [<Si–C–N =
107.0(3)◦)] (see Fig. 4). Further structural parameter values are
listed in Table 1.
˚
˚
relatively short Si ◊ ◊ ◊ N distances of 2.241(4) A and 2.117(2) A,
respectively. C6F5Si(F)2CH2NMe2 (1) is the first SiCN compound
for which no widened Si–C–N angle (bigger than the ideal 109.4◦)
16
is observed. This was the case for F3SiCH2NMe2 with ∠(Si–
C–N) = 110.3(7)◦ and for H3SiCH2NMe2 with ∠(Si–C–N) =
17
114.7(3)◦ for which structures from the gas phase are available
(gas electron diffraction).
In order to evaluate the possible influence of the electronic
effect of substituents at the C atom of the SiCN linkage on
the structures of such a-silanes, we calculated the structures
of compounds of the type R3Si–C(R¢¢)–NMe2. The results are
compiled in Table 2 which shows that a contraction of the
Si–C–N angle is possible when electropositive substituents are
introduced. One or two trimethylsilyl groups at the spacer carbon
atom in combination with electronegative substituents at silicon,
F or F5C6, lead to such structures with small Si–C–N angles.
The minimum among the investigated systems is predicted to
be 83.5◦ for (F5C6)F2SiC(SiMe3)2NMe2. Interestingly this angle
is more than two degrees smaller than in the trifluorosilyl
compound F3SiC(SiMe3)2NMe2 (86.0◦), which was unexpected, as
We have already discussed the preference of the gauche con-
formers for compound 3 in the context of its IR spectrum (see
above). In order to confirm this and to learn more about the
energetics of conformations and structural changes between them,
we have calculated a part of the potential surface by successive
variation of the C11SiON and C11SiCN dihedral angles in 1
and 3 [PBE(RI)/TZVPP geometry optimization followed by single
point energy calculations at the MP2(RI)/TZVPP level of theory].
Corresponding calculations for 2 have not been undertaken for
reasons of computational feasibility, and because 1 and 3 mark
both extremes for the presence and absence of Si ◊ ◊ ◊ N attractive
interactions. Fig. 7 shows the MP2 energies of 1 and 3 as a function
of both the C11–Si–C–N (C11–Si–O–N) dihedral angle and the
Si–C–N (Si–O–N) bending angle for 1 (3) in the forms of three-
dimensional perspective plots. The highest barrier to the transition
between gauche to anti position of the C6F5 group in 1 is about
14 kJ mol-1, whereas it is almost twice as high for 3 with 26 kJ
mol-1. The transition between both gauche positions is lower for
both compounds and is 6 kJ mol-1 for 1 and more than twice as
high for 3 at 16 kJ mol-1. Whereas the barriers are lower for 1, the
absolute energy differences between gauche and anti conformers is
higher for 1 with 8 kJ mol-1, while 3 has a corresponding energy
difference of about 6 kJ mol-1. Nevertheless, the gauche conformer
is the preferred for both compounds. Thus the reason why only
gauche conformers are observed in the crystal structures and for 3
Table 2 Dependence of calculated SiCN angles on the substituents at the
spacer carbon atom of the molecules R3Si–C(R¢2)–NMe2. Calculations are
at the PBE/TZVPP(RI) level of theory, angles are given in ◦
R¢2 =
R3 =
H2
(CH3)2
H(SiMe3)
(SiMe3)2
(CH3)3
H3
F3
116.8
114.6
113.1
113.9
109.1
108.7
107.8
107.0
105.3
104.9
102.7
88.9
105.4
110.7
86.0
(F5C6)F2
83.5
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 5652–5658 | 5655
©

View Article Online
Fig. 7 Dependence of the energy and angle SiXN (X = C, O) on the conformation/torsional angle C11SiXN in compounds 1 (left) and 3 (right). For
better comparison, the axes are drawn on the same scale. The plots are shown as a three-dimensional graph representing the individual calculation points
with projections into two dimensions showing a dependence angle on torsion (bottom) and energy/torsion (back). Calculations were performed as single
point calculations (MP2/TZVPP) based on relaxed geometries with restricted torsional angles (PBE/TZVPP).
in the gas phase is clearly an inherent property and not forced by
intermolecular forces in the solid state.
Syntheses
F6C5SiF2CH2NMe2 (1). A suspension of 0.15 g (2.31 mmol)
N,N-dimethylaminomethyllithium in 20 mL dry pentane was
added dropwise from a funnel equipped with a magnetic stirrer
into a solution of C6F5SiF3 (0.64 g, 2.54 mmol) in pentane (10 mL)
It is also interesting to observe the angles Si–C–N and Si–
O–N during the rotation of the silyl group about the Si–C and
Si–O bonds, respectively. The Si–C–N angle in the case of 1
varies over a range of only a few degrees during such a rotation,
whereas the corresponding Si–O–N angle in 3 changes by about
20◦ during a full rotation. This indicates also, that in 3 certain
orientations of the silyl group only allow for an interaction
with the geminal N atom, which is an argument against purely
electrostatic interaction and in favour of orientationally dependent
orbital interactions like a negative hyperconjugation of the type
lp(N)→s*(Si–X). However, it has to be mentioned that we have
already investigated the latter possibility in terms of the electron
density topologies and could not find evidence for this for other
SiON systems, but found indications in favour of lp(N)→s*(Si–
F) interactions in F3SiONMe2 or lp(N)→s*(Si–Cl) interactions
in ClH2SiONMe2 by means of NBO analyses.11,12 ClH2SiONMe2
is also a nice example for which the orientation dependence of the
electronegative substituents at silicon becomes manifest, as the
anti conformer (with the Cl atom in the SiON plane) has a much
stronger Si ◊ ◊ ◊ N attractive interaction than the gauche conformer,
in which a less electron withdrawing H atom is in anti position
relative to the nitrogen atom.
at -50 ◦C. After allowing warming to -30 ◦C during h, the
1
2
reaction was completed by stirring for 2 h at this temperature. Then
the suspension was filtered through a glass wool filter fitted to a
PTFE cannula and the residue was washed with 2 ¥ 5 mL pentane.
The volume was reduced by one third. Storing at a temperature
of -78 ◦C led to precipitation of the product. Yield: 0.15 g,
◦
1
0.52 mmol, 22%, mp -5 C. H NMR (CDCl3; 200 MHz): d
2.32 ppm (s, 6 H, (CH3)2), 2.40 (s, 2 H, CH2), 13C NMR (CDCl3;
50.3 MHz): d 46.4 ppm (s, CH2), 49.4 (s, N(CH3)2). 125.0 (m, CF3,
1J(C,F) = 317 Hz). 19F NMR (CDCl3, 188.3 MHz): d -158.9 ppm
(m-F), -145.3 (p-F, 3JF,F = 20 Hz), -133.8 (t, SiF, 1JSi,F = 126 Hz),
-126.5 (o-F). IR(gas) 518 (m) n(SiC), 730 (w) n(SiCH2), 755 (vw),
812 (w) r(CH2), 890 (s) n(SiF), 941 (s) nas(SiF), 977 (vs) nas(CF),
1042 (w) r(NCH2), 1103 (vs) n(pCF), 1157 (vw) t(SiH2), 1180
(vw) w(SiH2), 1251 (w) d(CH2), 1305 (s) n(CF), 1390 (s) n(CC),
1483 (vs) das(CC), 1521 (vs), 1650 (s), 2780–2989 (w) n(CH). MS
(EI) m/z (%) 291 (M+, 92), 289 (M+ - 2H, 100), 123 (M+ - C6F5,
3.8), 57 (NMe3 - 2H, 6).
C5F6SiF2N(SiMe3)NMe2 (2).
A
solution of 0.48
g
(3.5 mmol) N-lithio-N-trimethylsilyl-N¢,N¢-dimethylhydrazine
(generated from N-trimethylsilyl-N¢,N¢-dimethylhydrazine20 and
n-butyllithium) in 10 mL dry pentane was added dropwise to a
solution of 0.97 g (3.85 mmol) trifluoro-pentafluorophenylsilane
in 15 mL dry pentane at -40 ◦C. After allowing warming to
ambient temperature over 2 h the mixture was filtered through
a glass wool fitted PTFE cannula and washed two times with
5 mL pentane. After reducing the volume by one half the product
precipitated at -78 ◦C over night in the form of fine clear
crystalline needles, which ◦are sensitive to moisture. Yield: 1.10 g,
Experimental
General
All reactions were carried out under an atmosphere of dry nitrogen
in anhydrous and degassed solvents with standard Schlenk and
high vacuum techniques. NMR spectra were recorded on AC200,
Avance 400 or Varian Inova 500 spectrometers in deuteriated
solvents and were referenced relative to (residual) solvent reso-
nances (1H, 13C, 19F) or relative to an external standard (29Si; neat
SiMe4). Infrared spectra were measured on a Midac Prospect-IR
spectrometer. Mass spectra were recorded on a Varian MAT 212
or on a Pfeiffer QMG422/SEM.
1
3.02 mmol, 86%, mp 35 C. H NMR (C6D6; 200 MHz): d =
0.29 ppm (s, 9H, Si(CH3)3), 2.54 ppm (s, 6 H, (CH3)2), 13C NMR
(C6D6; 50.3 MHz): d = 2.0 ppm (s, Si(CH3)3), 48.0 (s, N(CH3)2).
5656 | Dalton Trans., 2008, 5652–5658
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
125.0 (m, C6F5), 19F NMR (CDCl3, 188.3 MHz): d = -160.7 ppm
Table 3 Crystal and refinement data for the crystal structures of 1–3
3
1
(m-F), -149.2 (p-F, JF,F = 21 Hz), -133.9 (t, SiF, JSi,F = 175
Hz), -127.0 (o-F). IR(gas) = 508 (s), 700 (m), 760 (m), 809 (s)
r(CH3), 841 (s), 904 (s) n(SiF), 974 (s) nas(SiF), 1021 (m), 1080
(bs) n(SiN), 1173 (m), 1230 (m), 1259 (s), 1291 (m), 1388, 1473,
1516, 2789–2964 (w) n(CH). MS (EI) [m/z(%)]: 364 (M+, 100%),
349 (M+ - CH3, 30%), 257 (M+ - F - CH3 - SiMe3, 26%), 233
(M+ - NSiMe3)NMe2, 7%), 168 (C6F5H, 12%), 73 (SiMe3, 83%).
Compound
1
2
3
Formula
C9H8F7NSi
291.25
C11H15F7N2Si2
364.41
Orthorhombic
C8H6F7NOSi
293.23
Formula weight
Crystal system
Space group
Diffractometer
Triclinic
Orthorhombic
P212121
Nonius CAD4
6.325(1)
10.809(1)
15.633(2)
90
90
90
1068.8(2)
1.822
4
1.033
143(2)
-0 ≤ h ≤ 8
-17 ≤ k ≤ 17
-20 ≤ l ≤ 5
54.30
3503
2320
1793
0.072
¯
P1
Pna21
STOE IPDS I
5.897(1)
10.362(2)
10.892(2)
68.54(3)
80.39(3)
74.53(3)
593.1(2)
1.631
Bruker APEX
28.767(11)
8.542(3)
6.264(2)
90
˚
a/A
˚
b/A
F6C5SiF2ONMe2 (3). 1.26
g (5.00 mmol) of trifluoro-
˚
c/A
pentafluorophenylsilane were condensed onto a suspension of
an excess of diethyl ether and 0.34 g (5.0 mmol) O-lithio-N,N-
dimethylhydroxylamine. After warming to RT and stirring for 3 h
the mixture was filtered through a glass wool filter fitted to a PTFE
cannula. The solvent and remaining C6F5SiF3 were removed under
reduced pressure until the residue became oily. Some remaining
LiF precipitated by addition of 10 mL dry hexane. After filtration
a/◦
b/◦
g /◦
90
90
3
˚
V/A
1539.2(10)
1.573
4
0.913
153(2)
-40 ≤ h ≤ 40
-12 ≤ k ≤ 12
-8 ≤ l ≤ 8
60.08
17049
4448
r
calc/g cm-3
Z
2
m/mm-1
T/K
Index range
1.113
213(2)
-6 ≤ h ≤ 6
-12 ≤ k ≤ 12
-13 ≤ l ≤ 13
51.65
7718
5631
2020
0.123
◦
the vessel was cooled to -10 C and the solvent was completely
removed under reduced pressure to give 3 as a colourless liquid.
Yield: 0.76 g, 2.9 mmol, 48%, mp 10 ◦C. 1H NMR (CDCl3;
400 MHz): d = 2.68 ppm (s, 6 H, (CH3)2). 13C NMR (CDCl3;
100.6 MHz): d = 49.5 ppm (s, NCH2), 100.0 (s, Cipso). 149.3–137.3
(m, C6F5), 19F NMR (CDCl3, 376.5 Hz): d = -159.2 ppm (m-F),
◦
2Jmax
/
Measured refl.
Unique refl.
Observed refl.
Rint
2893
0.252
224
1
-145.9 (p-F), -133.8 (t, SiF, JSi,F = 126 Hz), -126.9 (o-F). 29Si
Parameters
165
188
Absolute strucure
parameter
0.2(2)
0.0(3)
NMR: (CDCl3; 79.5 MHz). d = -48.3 ppm. IR(gas) = 518 (m)
n(SiC), 735 (w), 816 (w), 959 (s) n(SiF), 982 (s) nas(SiF), 1109 (vs)
(rCF), 1308 (s) n(CF), 1399 (s) n(CC), 1483 (vs), 1522 (vs), 1649
(s), 2876–2977 (w) n (CH). MS (EI) [m/z(%)]: 293 (M+, 0.2%), 168
(C6F5H, 81%), 148 (Si(ONMe2)2, 100%), 126 (M+ - C6F5, 20%).
R[I > 2s(I)]/wR2
0.081/0.179
0.096/0.190
0.74/-0.31
0.085/0.151
0.106/0.156
1.05/-0.33
0.050/0.118
0.080/0.131
0.44/-0.36
R[all refl]/wR2
-3
˚
r
fin(max/min)/e A
Crystal structures
Hydrogen atoms were included in calculated positions and refined
with a riding model with fixed angles and distances for 1, with
group-wise refining distances and fixed angles for 2 and located in
difference Fourier maps and refined with isotropic displacement
parameters for 3. See Table 3 for crystal and refinement data.
The experimental setup for low temperature crystal growth of 1
consists of an X-ray diffractometer (Stoe¨ IPDS 1) with an attached
nitrogen cryostream. An IR-laser source is set up in such a way
that the sample can be heated with the focused laser-beam. The
intensity and position of the laser focus is computer controlled.
The liquid samples were added to thin-walled glass capillaries
(approx. 0.25 mm inner diameter) and were then sealed. After
placing a capillary on the goniometer head, the sample was cooled
to ca. 20 K below the melting point (polycrystalline solidification).
The IR laser (hitting a tiny portion of ca. 0.2 mm of the capillary)
was then applied at an intensity sufficient to partially melt the
polycrystalline compound. It was then scanned 4 times along
the length of the capillary very slowly (approx. 2.5 mm h-1) at
-50 ◦C. During this procedure the temperature was kept constant.
After starting at the bottom (0%) of the capillary the next turn
will be started at a position of 10% of the scanning range. The
last run was started at 40% and ended at 100%. After finishing
the crystallisation procedure a suitable single crystal can be
identified by means of a microscope equipped with a polarisation
filter.
Conclusions
With the series of compounds C6F5SiF2CH2NMe2 (1),
C6F5SiF2N(SiMe3)NMe2 (2) and C6F5SiF2ONMe2 (3) we have
for the first time a set of a-donor-functionalised silanes with
identical donor and acceptor functions separated by different
spacer units CH2, N(SiMe3) and O, which are all monomeric in
the solid state (in the series F3SiCH2NMe2, F3SiNRNMe2 and
F3SiONMe2 such a comparison was impossible due to the fact that
F3SiCH2NMe2 dimerises, whereas the other two compounds are
monomeric in the crystalline phase). This allowed a comparison
of their structures in this phase. Only for 2 and 3 are there clearly
attractive structure-determining interactions between the geminal
Si and N atoms, as indicated by their Si–X–N angles [2: 90.4(2)
and 3: 85.1(2)◦]. C6F5SiF2CH2NMe2 (1), in contrast, has an Si–C–
N angle of 107.0(3)◦, which shows that—if at all—only extremely
weak attractive forces are present. This is also supported by the
potential curves for the rotation about the Si–X axes for 1 and
3, which show a strong dependence of the Si ◊ ◊ ◊ N distance and
Si–O–N angle on the orientation of the F5C6F2Si group, whereas
these parameters are not very much affected by rotating this group
about the Si–C axis in 1.
A crystal of 3 was grown on a CAD4 diffractometer with
cryostream from a seed crystal, which was selected in a carefully
established solid–liquid equilibrium, by melting other crystals
followed by slowly lowering the temperature until the whole
capillary was filled with a single crystal.
The structures were solved by direct methods and refined using
standard crystallographic software.21 Non-hydrogen atoms of all
structures were refined with anisotropic displacement parameters.
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 5652–5658 | 5657
©

View Article Online
1997, 109, 2897, (Angew. Chem., Int. Ed. Engl., 1997, 36, 2807); (c) U.
Losehand and N. W. Mitzel, Inorg. Chem., 1998, 37, 3175; (d) N. W.
Mitzel and U. Losehand, Eur. J. Inorg. Chem., 1998, 2023; (e) N. W.
Mitzel, U. Losehand and A. D. Richardson, Inorg. Chem., 1999, 38,
5323; (f) N. W. Mitzel, U. Losehand and D. W. H. Rankin, J. Chem.
Soc., Dalton Trans., 1999, 4291; (g) N. W. Mitzel, U. Losehand and B.
Bauer, Inorg. Chem., 2000, 39, 1998; (h) U. Losehand and N. W. Mitzel,
J. Chem. Soc., Dalton Trans., 2000, 1049; (i) N. W. Mitzel, U. Losehand,
S. L. Hinchley and D. W. H. Rankin, Inorg. Chem., 2001, 40, 661; (j) N.
W. Mitzel and U. Losehand, Z. Naturforsch., B: Chem. Sci., 2001, 56,
630; (k) N. W. Mitzel and K. Vojinovic, J. Chem. Soc., Dalton Trans.,
2002, 2341; (l) K. Vojinovic´, N. W. Mitzel, T. Foerster and D. W. H.
Rankin, Z. Naturforsch., B: Chem. Sci., 2004, 59, 1505.
9 U. Losehand and N. W. Mitzel, J. Chem. Soc., Dalton Trans., 1998,
2537.
10 (a) N. W. Mitzel, H. Schmidbaur, D. W. H. Rankin, B. A. Smart, M.
Hofmann and P. v. R. Schleyer, Inorg. Chem., 1997, 36, 4360; (b) N. W.
Mitzel, Chem.–Eur. J., 1998, 4, 692.
11 N. W. Mitzel and U. Losehand, J. Am. Chem. Soc., 1998, 120,
7320.
12 N. W. Mitzel, U. Losehand, A. Wu, D. Cremer and D. W. H. Rankin,
J. Am. Chem. Soc., 2000, 112, 4471.
13 N. W. Mitzel, K. Vojinovic´, R. Fro¨hlich, T. Foerster and D. W. H.
Rankin, J. Am. Chem. Soc., 2005, 127, 13705.
Acknowledgements
This work was supported by the Deutsche Forschungsgemein-
schaft. We are grateful to Tania Pape for collecting the crystal-
lographic data for compound 2 and to Dr Alexander Hepp for
NMR measurements.
Notes and references
1 (a) R. G. Kostyanovskii and A. K. Prokov’ev, Dokl. Akad. Nauk SSSR,
1965, 164, 1054; (b) V. V. Kharpov, V. I. Gol’danskii, A. K. Prokov’ev
and R. G. Kostyanovskii, Zh. Obshch. Khim., 1967, 37, 3; (c) E. W.
Randall, C. H. Yoder and J. J. Zuckerman, Inorg. Chem., 1967, 6,
744; (d) Z. Pacl, M. Jakoubkova, Z. Papouskova and V. Chvalovsky,
Collect. Czech. Chem. Commun., 1971, 36, 1588; (e) A. Marchand, J.
Mendelsohn, M. Lebedeff and J. Valade, J. Organomet. Chem., 1969,
17, 379.
2 M. A. Brook, Silicon in Organic, Organometallic and Polymer Chem-
istry, John Wiley & Sons, New York, 2000, p. 500.
3 (a) G. Mital and R. R. Gupta, J. Am. Chem. Soc., 1969, 91, 4664;
(b) J. M. Bellama and A. G. MacDiarmid, J. Organomet. Chem., 1970,
24, 91; (c) N. Egorochkin, S. E. Skobeleva, E. I. Sevast’yanova, I. G.
Kosolapova, V. D. Sheludyakov, E. S. Rodionov and A. D. Kirilin, Zh.
Obshch. Khim., 1976, 46, 1795.
14 K. Vojinovic´, L. McLachlan, D. W. H. Rankin and N. W. Mitzel,
Chem.–Eur. J., 2004, 10, 3033.
4 (a) W. K. Musker and G. K. Larson, J. Organomet. Chem., 1966, 6,
627; (b) M. E. Freeburger and L. Spialter, J. Am. Chem. Soc., 1971, 93,
1894.
5 L. G. L. Ward and A. G. MacDiarmid, J. Inorg. Nucl. Chem., 1961, 20,
345.
6 (a) N. F. Lazareva, E. I. Brodskaya and G. V. Ratovsky, J. Chem. Soc.,
Perkin Trans. 2, 2002, 2083; (b) N. F. Lazareva, V. P. Baryshok and M.
G. Voronkov, Russ. Chem. Bull., 1995, 104, 374; (c) D. Labreque, K. T.
New and T. H. Chan, Organometallics, 1994, 13, 332.
7 (a) A. Bauer, T. Kammel, B. Pachaly, O. Scha¨fer, W. Schindler, V.
Stanjek and J. Weis, in Organosilicon, Chemistry V, Wiley-VCH,
Weinheim, 2003, p. 527; (b) One step ahead – organofunctional
silanes from Wacker, Brochure of the Wacker company:
http://www.wacker.com/internet/webcache/de_DE/BrochureOrder/
GENIOSIL_Brosch_e.pdf; (c) W. Schindler, Adhaes.–Kleben Dichten,
2004, 48, 29.
15 K. Vojinovic´, N. W. Mitzel, M. Korth, R. Fro¨hlich and S. Grimme,
in Organosilicon Chemistry VI, ed. N. Auner and J. Weis,Wiley VCH,
Weinheim, 2005, vol. 1, pp. 156–159.
16 N. W. Mitzel, K. Vojinovic´, T. Foerster, H. E. Robertson, K. B.
Borisenko and D. W. H. Rankin, Chem.–Eur. J., 2005, 11, 5114.
17 N. W. Mitzel, C. Kiener and D. W. H. Rankin, Organometallics, 1999,
18, 3437.
18 R. Boese, M. Nussbaumer, In situ Crystallization Techniques, in
Correlations, Transformations, and Interactions in Organic Crystal
Chemistry, IUCr Crystallographic Symposia, ed. D. W. Jones and A.
Katrusiak, Oxford University Press, Oxford, UK, 1994, vol. 7, p. 20–
37.
19 R. Tacke, J. Becht, O. Dannappel, R. Ahlrichs, U. Schneider, W. S.
Sheldrick, J. Hahn and F. Kiesgen, Organometallics, 1996, 15, 2060.
20 U. Wannagat and F. Hoefler, Monatsh. Chem., 1966, 97, 976.
21 (a) G. M. Sheldrick, Acta Crystallogr., Sect. A: Found. Crystallogr.,
1990, 46, 467; (b) G. M. Sheldrick, Universita¨t Go¨ttingen, Germany,
1997.
8 (a) N. W. Mitzel, A. J. Blake and D. W. H. Rankin, J. Am. Chem. Soc.,
1997, 119, 4143; (b) N. W. Mitzel and U. Losehand, Angew. Chem.,
5658 | Dalton Trans., 2008, 5652–5658
This journal is
The Royal Society of Chemistry 2008
©