Silylation products of cyclic tri-aminal carbanions and their lithiation

Article abstract of DOI:10.1039/c1dt10943j

The structure of 1,3,5-trimethyl-1,3,5-triaza-cyclohexane (TMTAC) was determined by single crystal X-ray diffraction and compared with earlier gas-phase data. It shows a preference for an aee-conformation in all phases. Lithiated TMTAC, [(RLi)₂

Full text of DOI:10.1039/c1dt10943j

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Dalton
Transactions
Cite this: Dalton Trans., 2012, 41, 104
www.rsc.org/dalton
PAPER
Silylation products of cyclic tri-aminal carbanions and their lithiation†
Constantin Sicking, Andreas Mix, Beate Neumann, Hans-Georg Stammler and Norbert W. Mitzel*
Received 19th May 2011, Accepted 15th September 2011
DOI: 10.1039/c1dt10943j
The structure of 1,3,5-trimethyl-1,3,5-triaza-cyclohexane (TMTAC) was determined by single crystal
X-ray diffraction and compared with earlier gas-phase data. It shows a preference for an
aee-conformation in all phases. Lithiated TMTAC, [(RLi)2·(RH)] (1) (R = 2,4,6-trimethyl-2,4,6-triaza-
cyclohex-1-yl), was reacted with Et3SiCl, Ph3SiCl and PhMe2SiCl to afford the substituted silanes
Et3SiR (1), Ph3SiR (2) and PhMe2SiR (3) in moderate yields. They were characterised by NMR
spectroscopy (1H, 13C, 29Si). 1 reacts with Me2SiCl2 and Ph2SiCl2 to give Me2SiR2 (5) and Ph2SiR2 (6)
which were characterised by NMR spectroscopy. 5 was also identified by crystal structure
determination. Analogous triple substitution could not be observed by employing trichlorosilanes.
Quantumchemical calculations explain this by sterical overcrowding of the silicon atom. The reaction
of 1 with SiCl4 did not yield fourfold substitution but a formal insertion product of SiCl2 into a C–N
bond of the TMTAC ring (2,4,6-trimethyl-2,4,6-triaza-1,1-dichloro-1-sila-cycloheptane, 7) in very small
quantities. It was identified by X-ray crystallography and shows an intramolecular Si ◊ ◊ ◊ N dative bond.
The reactions of (3) and (5) with n-butyl lithium afforded lithiation of the silicon bound methyl groups
in both cases. The products, 8 and 9, were characterised by NMR spectroscopy (1H, 13C, 29Si), 8 was
also characterised by X-ray crystallography.
used in this sense throughout this report),8 and the product
consists of an endless chain aggregate of a dimer of the metal-
lated substrate molecules coordinated by the non-deprotonated
Introduction
Known examples of the direct lithiation of amines are still scarce.
This is due to the fact that they are formally non-stabilised or
even destabilised lithium carbanions and thus one of the most
favourable procedures to prepare such a-lithiated amines is still
the transmetalation route developed by Seyferth et al.1
The few examples of direct deprotonation of amines in-
clude the lithiation of Me2N(CH2)2NMe2 (TMEDA)2 and
MeN[(CH2)2NMe2]2 (PMDTA).3 These findings are important as
these amines are often employed to increase the carbanion activity
of the butyl lithiums. Other examples are (1,4,7-trimethyl-1,4,7-
triazacyclononane,4 N-methylpiperidine,5 and (R,R)-tetramethyl-
1,2-diaminocyclohexane6), which are always singly deprotonated
at one of their terminal methyl positions. An exception in
this context is the high-yield double deprotonation of aminals
RMeNCH2NMeR to give LiCH2(Me)NCH2N(Me)CH2Li, again
lithiated in terminal positions.7
substrate: [(RLi)2·(RH)]• (1). This reactivity pattern appears to
be general for cyclic aminals,9,10 and provides mechanistic insights
into precoordination of the butyl lithium bases, which determine
regioselectivity of metallation.9–11 These compounds can serve
as acylation agents for carbonyl compounds analogous to the
Corey/Seebach reagents, but in contrast to the latter with the
advantage of a heavy-metal-free workup procedure.8
Transmetallation of this lithiated TMTAC (1) is possible and
a series of aluminium and gallium derivatives with MR¢2 groups
replacing lithium (M = Al, Ga; R¢ = Me, Et) has been reported.12
A trimethylsilyl derivative RSiMe3 has also been reported and
demonstrated to be capable again to serve as metallation substrate,
however, then not at the aminal position but rather at the SiMe3
group.12 This carbanion was used again to generate aluminium
and gallium derivatives.
We discovered recently that nBuLi metallates 1,3,5-trimethyl-
1,3,5-triaza-cyclohexane (TMTAC) selectively at the methylene
unit between two nitrogen atoms, leading to the first doubly
amino-substituted carbanion, RLi (R = MeN(CH2NMe)CH–,
Interestingly, the triaminal group R does not always stay
intact when 1 is reacted with element halides. Bulkily substi-
tuted substrates lead to decomposition of 1 into MeN CH2
and MeN CHLi fragments (TMTAC itself is a trimer of
MeNCH2) and these fragments are then found included in the
products.13 Cp2YCl for instance reacts with RLi to give the
4
Fakulta¨t fu¨r Chemie, Lehrstuhl fu¨r Anorganische Chemie und Struk-
turchemie, Universita¨t Bielefeld, Universita¨tsstraße 25, 33615, Bielefeld,
Germany. E-mail: mitzel@uni-bielefeld.de; Fax: +49 (0) 521 106 6026
† CCDC reference numbers 821375–821378. For crystallographic data in
CIF or other electronic format see DOI: 10.1039/c1dt10943j
complex Cp2Y[h (RCH2NMe)], with a new tetrapodal ligand
and this monoanionic ligand is formed in one step (showing
MeN CH2 reactivity). t-Bu2GaCl reacts with 1 to give cyclo-
[t-Bu2GaCH NMe]2 (showing MeN CHLi reactivity).
104 | Dalton Trans., 2012, 41, 104–111
This journal is
The Royal Society of Chemistry 2012
©

View Article Online
Table 1 Selected bond lengths and angles of TMTAC a) for one of the
four independent molecules, b) average values for the four molecules with
parameters grouped by approximate molecular symmetry, c) gas-phase
value (GED16) with the assumption of all N–C distances, C–N–C and
N–C–N angles to be equal
It remained of interest whether more than one triaminal group
R could be transferred to other elements and thus we explored
this chemistry further with a range of silicon reagents and report
here about the results. Prior to this we report the crystal structure
of TMTAC for further comparison.
XRD
XRD
mol 1
average all
mols
GED16
Results and discussion
Crystal structure of TMTAC
N(1)–C(1)
N(1)–C(3)
N(2)–C(1)
N(3)–C(3)
1.454(2)
1.452(2)
1.475(2)
⎫
⎫
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎬
⎪
⎭⎪
1.453
1.473
Due to the growing importance of the chemistry of TMTAC, it
is desirable to determine the molecular structure in the crystal.
TMTAC was reported to have a preference for the axial orientation
of the methyl substituents, as was also found in other hexahydro-
s-triazines.14 TMTAC exists in the liquid as a 1 : 1 mixture of
the diaxial and monoaxial conformers, with no molecules having
the triequatorial conformation.14b In contrast, the gas-phase
conformer composition and structure has been reported earlier
by Vilkov et al.15 TMTAC vapour contains a sole conformer with
one axially oriented and two equatorially methyl substituents (aee-
configuration). Its main structural parameters (only mean values
⎫
⎪
⎪
⎬
⎪
1.475(2)
1.457(2)
1.455(2)
1.465(2)
⎪
⎭
⎫
⎪
⎪
N(2)–C(2)
N(3)–C(2)
N(1)–C(4)
⎬ 1.463(3)
1.458
1.468
⎬
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎭
⎫
⎪
⎪
⎬
⎪
⎪
⎭
⎫
⎪
⎪
1.456(2)
1.457(2)
N(2)–C(5)
N(3)–C(6)
1.456
108.7
⎬
⎪
⎪
⎭
⎪
⎭
⎫
⎪
⎪
C(1)–N(1)–C(3)
109.0(1)
⎬
⎪
⎫
⎪
˚
⎪
⎭
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
determined) were reported to be rg(C–N) = 1.463(3) A, rg (C–H) =
◦
◦
15
˚
1.117(5) A, ∠(C–N–C) = 110.91(1) and ∠(N–C–N) = 111.1(1) .
⎫
⎪
⎪
C(1)–N(1)–C(4)
C(3)–N(1)–C(4)
112.4(1)
112.4
⎬
⎪
The reason for this conformational behaviour is the electrostatic
repulsion between the lone pairs of electrons at the nitrogen atoms
and also anomeric contributions.14
112 5 1
. ( )
⎪
⎭
⎪
⎪
110.91(1)
111.1(1)
⎬
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎫
⎪
⎪
C(1)–N(2)–C(2)
C(2)–N(3)–C(3)
109.0(1)
109.0(1)
109.1
110.0
⎬
⎪
We have now succeeded in growing a single crystal of TMTAC
for X-ray diffraction in a sealed capillary on an X-ray diffractome-
ter by generating manually a crystal seed at 247 K and cooling
down slowly to 100 K. It belongs to the monoclinic system, space
group P21/n. The asymmetric unit contains four molecules, three
of which are oriented with their N3 planes parallel and one with
an angle of 65◦ between it and the average N3 planes of the other
molecules (Fig. 1).
⎪
⎭
⎫
⎪
⎪
C(2)–N(2)–C(5)
C(2)–N(3)–C(6)
N(1)–C(1)–N(2)
110.1(1)
110.1(1)
⎬
⎪
⎪
⎭
⎪
⎭
⎫
⎪
⎪
111.8(1)
⎫
⎪
111.7
110.4
⎪
⎪
⎪
⎬
⎪
⎪
⎪
⎬
⎪
N(1)–C(3)–N(3)
N(2)–C(2)–N(3)
111 7 1
. ( )
⎪
⎭
⎫
⎪
⎪
110.4(1)
⎬
⎪
⎪
⎭
⎪
⎭
mean parameters assuming an idealised molecular CS symmetry.
This allows also a comparison with the values determined by
gas electron diffraction (GED).15 In this GED refinement, N–
C distances and all angles of the same type (N–C–N and C–N–C)
were assumed to be equal. Consequently, this gas-phase structure
has been derived with only three parameters defining the heavy
atom skeleton. Our new crystal structure parameters suggest that
this assumption is only of limited validity. For instance, the largest
C–N–C angle is that describing the position of the axial methyl
group (C(1)–N(1)–C(4) etc.) with an average of 112.4◦; the smallest
C–N–C angle is the endocyclic one involving the nitrogen atom
with the equatorial methyl group (C(1)–N(2)–C(2) etc.) with an
average of 109.1◦. The reported very high precision of the gas
electron diffraction parameter for C–N–C at 110.91(1)◦ reflects the
value for an average, but does not reflect the spread in chemically
distinguishable parameters; as with more parameters refining and
closely related distances leading to correlation the provided error
is probably misleading in terms of absolute accuracy.
Fig. 1 Asymmetric unit of the crystal structure of TMTAC with four
The chemical meaningful information to be extracted from the
new crystal structure determination is that there is a systematic and
significant difference between C–N bond lengths, N–C–N and C–
N–C bond angles of the same type relating to the distinguishable
chemical surrounding of the contributing atoms. The differences
molecules. Hydrogen atoms omitted for clarity.
All four independent molecules adopt the aee-configuration –
as in the gas-phase – and only marginally different structures.
Table 1 lists the parameters of one molecule together with the
This journal is
The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 104–111 | 105
©

View Article Online
in angles have already been mentioned, for the three existing types
of N–C bonds, the group including C(1)–N(2) has much longer
Reaction of lithiated TMTAC (1) with the dichlorosilanes
Me2SiCl2 and Ph2SiCl2
˚
values (1.473 A on average) than the two other groups (1.453 and
Reaction with Me2SiCl2. The reaction of lithiated TMTAC (1)
with dichlorodimethylsilane, Me2SiCl2 (Scheme 2) afforded the
formation of dimethyl-bis(2,4,6-trimethyl-2,4,6-triaza-cyclohex-
1-yl)silane (5) in relatively low yields of 35% after recrystallisation.
This is due to the presence of free TMTAC in 1, which complicates
purification by recrystallisation. Compound 5 is a colourless
(slightly yellowish) solid and is soluble in hydrocarbons. It was
characterised by NMR spectroscopy and its crystal structure was
determined.
˚
1.458 A on average).
In order to establish a parameter describing the conformation
of TMTAC concerning the positions of the methyl groups, we
introduce the angle c defined as depicted in Fig. 2 as the angle
between the N–C vector of the corresponding bond and the plane
defined by the three nitrogen atoms. c[C(4)ax] is 98.1◦ for the
axial methyl group in the molecule shown in Fig. 2 and 97.4◦
on average for the four independent molecules in the unit cell. The
exocyclic methyl group about C(5) adopts a corresponding angle
of c[C(5)eq] = -19.4◦.
Scheme 2 Reaction of 1 with Me2SiCl2.
The 1H NMR spectrum of a C6D6 solution shows broad signals
for the hydrogen atoms attached to the rings at 3.37 and 3.02
ppm. The methyne proton results in a resonance at 3.26 ppm. The
exocyclic methyl groups are detected as two singlets at 2.69 and
1.80 ppm, the methyl groups at silicon at 0.43 ppm. The 13C NMR
spectrum shows peaks 79.4 and 75.2 ppm for the ring carbon
atoms, at 43.9 and 39.9 ppm for the N-methyl groups and at -1.1
ppm for the methyl groups at the silicon atom. The { H} Si NMR
spectrum contains a single resonance at 0.0 ppm. The molecular
structure of 5 in the crystal is shown in Fig. 3.
Fig. 2 Molecular structure of one molecule of TMTAC with definition
of the angle X describing the orientation of the axial methyl group relative
to the N3 plane. This angle is 97.4◦ on average for the four independent
molecules. Hydrogen atoms omitted for clarity.
1
29
Reaction of lithiated TMTAC (1) with monochlorosilanes. We
reacted lithiated TMTAC, 1, with various trialkylsilanes, namely
Et3SiCl, Ph3SiCl and Me2PhSiCl (Scheme 1). In all cases the single
substitution could be achieved without greater difficulties, yielding
R–SiEt3 (2), R–SiPh3 (3) and R–SiMe2Ph (4).
Fig. 3 Molecular structure of 5 in the crystal. Hydrogen atoms
˚
have been omitted for clarity. Selected distances and angles [A, deg]:
Si(1)–C(1) 1.923(2), Si(1)–C(7) 1.929(2), Si(1)–C(13) 1.882(2), Si(1)–C(14)
1.878(2), N(1)–C(1) 1.462(2), N(1)–C(3) 1.451(2), N(1)–C(4) 1.472(2),
C(1)–Si(1)–C(7) 106.6(1), C(13)–Si(1)–C(14) 111.2(1), Si(1)–C(1)–N(2)
114.6(2), N(1)–C(1)–N(2) 110.2(2); tilt angles: c[C(4)ax] 97.0◦, c[C(5)eq]
-19.4◦, c[C(11)ax] 98.2◦.
Scheme 1 Reaction of 1 with trialkylchlorosilanes.
Purification of 2 and 3 was possible by distillation and crystalli-
sation, respectively, but these techniques were not applicable to 4
due to glassy solidification and decomposition before reaching the
boiling temperature. The preparative yields were always moderate
(2 69%, 3 78%, 4 69%). Compounds 2 and 3 were characterised by
The Si–C bonds to the methyl groups (1.882(1) and 1.878(1)
˚
A) in compound 5 are shorter than those to the carbon atoms of
˚
the triazacyclohexyl rings (1.923(2) and 1.929(2) A). The latter are
elemental analyses and NMR spectroscopy of the nuclei H, 13C
and 29Si. As most of these data deserve no detailed comments, we
refer to the experimental part.
larger than in the trimethylsilylated TMTAC, R–SiMe3, at 1.918(2)
1
12
˚
A. The reason is probably due to steric factors, as the H ◊ ◊ ◊ H
˚
distances between the two R substituents are as close as 2.48 A.
106 | Dalton Trans., 2012, 41, 104–111
This journal is The Royal Society of Chemistry 2012
©

View Article Online
A distorted tetrahedral coordination geometry characterises the
bonding situation of the silicon atom; there are compressed angles
like ∠(C(1)–Si(1)–C(7)) at 106.6◦, while others are larger than
the tetrahedral angle, such as ∠(C(13)–Si(1)–C(14)) at 111.2◦.
Both N,N,N-trimethyl-triazacyclohexyl units in 5 adopt the aee-
configuration, alike to free TMTAC. The angle c characterising
the tilt of the axial methyl group measures 97.6◦ on average, which
is almost identical to TMTAC (97.4◦).
Reaction with Ph2SiCl2. Analogous to the reaction described
above, that of lithiated TMTAC (1) with dichlorodiphenylsilane,
Ph2SiCl2 afforded Ph2SiR2 (6), but only in very low isolated
yield of 4%. Characterisation of the compound was effected by
spectroscopic means and is in general analogous to that of 5. The
data are listed in the experimental section.
Fig. 4 Molecular structure of product 7 isolated in minor quantities
from the reaction of 1 with SiCl4. Hydrogen atoms are omitted for
clarity. Selected bond lengths and angles [A,deg] Si(1)–N(1) 1.705(2)
˚
Si(1)–Cl(1) 2.102(1) Si(1)–C(3) 1.876(2) N(1)–C(1) 1.444(3) N(1)–C(4)
1.449(2) N(2)–C(2) 1.480(2) N(3)–C(3) 1.463(3), Si(1) ◊ ◊ ◊ N(2) 2.115(2);
C(3)–Si(1)–N(1) 125.2(2), N(2) ◊ ◊ ◊ Si(1)–Cl(2) 166.7(1), C(3)–Si(1) ◊ ◊ ◊ N(2)
85.0(1), N(1)–Si(1) ◊ ◊ ◊ N(2) 70.0(1), Si(1)–N(1)–C(1) 104.7(2).
Reaction of lithiated TMTAC (1) with trichlorosilanes and
tetrachlorosilane, SiCl4
We have undertaken various attempts to isolate triply R-
substituted silanes by reactions of 1 with MeSiCl3 and HSiCl3,
but failed to isolate new products. Attempts to achieve partially
substituted products like R-SiCl3, R2SiCl2 etc. from the reactions
of 1 with excess of SiCl4 were also unsuccessful. A possible
explanation is that reactions of TMTAC (included in 1) or the
substitution products with SiCl4 occur leading to ionic products
remained unsuccessful. However, note that the loosely related
tri(cyclohexyl)silanes are known compounds.19
Quantumchemical estimation of possible limitations for the
substitution grade with 2,4,6-trimethyl-2,4,6-triaza-cyclohex-1-yl
groups at silicon
3
of the type [(h -TMTAC)SiCl3]+.16 We could only isolate some
Due to the experimental observation, that a higher substitution
grade than two with 2,4,6-trimethyl-2,4,6-triaza-cyclohex-1-yl
groups could not be achieved, we tried to estimate the feasibility
of such attempts on the basis of quantumchemical calculations on
the M05-2X/6-311G(dp) level (and 6-311(2df) for silicon).
Two isodesmic reactions were investigated. The first is the
reaction of a silane with two 2,4,6-trimethyl-2,4,6-triaza-cyclohex-
1-yl groups R, R2SiMe2, which exchanges formally an R group
with R–H = TMTAC.
crystalline material in one experiment; however, the amount was so
small that identification of this product (7) was solely possible on
the basis of X-ray diffraction and we report this structure displayed
in Fig. 4 to show that quite unexpected chemical processes occur
within such mixtures. Compound 7 is the product of a formal
insertion of a SiCl2 unit into the TMTAC ring, widening it from six
to seven ring members. The Lewis acidity of the silicon atom with
a substituent pattern SiCNCl2 is sufficient to induce the formation
of a dative bond to the nitrogen atom N(2) within the ring. This
leads to a bicyclic system comprising a four- and a five-membered
ring. Ring strain leads to deformation of the coordination spheres
of the involved atoms. The coordination geometry at silicon is
that of a strongly distorted trigonal pyramid with the dative
bonds and one Cl substituent adopting the axial positions (angle
Cl(2)–Si(1) ◊ ◊ ◊ N(2) 166.7(1)◦). The equatorial substituents are
Cl(1), N(1) and C(3); their angles involving the silicon atom
are C(3)–Si(1)–N(1) 125.2(2), Si(1)–N(1)–C(1) 104.7(2) and C(3)–
Si(1)–Cl(1) 110.7(1)◦, clearly showing the strong deviation from a
trigonal planar arrangement.
R2MeSi–Me + R–H → R2MeSi–R + Me–H
While breaking and forming the same number of Si–C and C–
H bonds, this reaction is calculated to be endothermic by 46 kJ
mol-1. As electronic effects can be assumed to be less important,
this number mainly reflects the energy increase by steric crowding.
The second reaction investigated by this method is one involving
an overcrowded silane with three R groups. The steric stress is
somewhat released by exchanging the fourth substituent Me by H
by a formal exchange reaction with trimethylsilane.
Nitrogen atom N(1) is planar coordinated and placed at a
R3Si–Me + Me3Si–H → R3SiH + Me3Si–Me
˚
distance of 1.705(2) A to the silicon atom; this is longer than
17
18
in Me2N-SiHCl2 [1.664(2) A] and Me2N-SiCl3 [1.665(3) A], as
is expected from the involvement of the bond within a strained
four-membered ring. The two Si–Cl distances are different (Si(1)–
The calculations predict this reaction to release 15 kJ mol-1,
again demonstrating the importance of steric effects. It is impor-
tant to note that these values were calculated for components
with the lowest energies conformers, i.e. the aee-conformers. The
calculated minimum structure of R3SiMe is shown in Fig. 5. It also
shows the short distance between hydrogen atoms exemplified by
the distance between the Si bound methyl group and the axially
positioned methyl group of one R substituent.
˚
˚
˚
Cl(1) 2.102(1), Si(1)–Cl(2) 2.144(1) A) due to their different roles
in the coordination sphere of silicon.
This result demonstrates that it is obviously not possible to
achieve a fourfold substitution at silicon with 2,4,6-trimethyl-
2,4,6-triaza-cyclohex-1-yl groups R – this is most likely to be
due to steric reasons. Without going into details we observed
similar problems with attempts to prepare triply substituted
compounds, e.g. using MeSiCl3 as a substrate. These attempts
Table 2 lists energies calculated for the change between two
of the conformational possibilities of some compounds, TMTAC
itself, R2SiMe2 (2), R3SiMe and R3SiH, in each case for the ground
This journal is
The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 104–111 | 107
©

View Article Online
Table 2 Energy differences for different conformers of TMTAC, R2SiMe2 (2), R3SiMe and R3SiH. The conformation assignment is for all R groups
Conformational change
DE [kJ mol-1]
TMTAC (aee) → TMTAC (eee)
20
47
89
58
[MeN(CH2NMe)2CH]2-SiMe2 (aee) → [MeN(CH2NMe)2CH]2-SiMe2 (eee)
[MeN(CH2NMe)2CH]3-SiMe (aee) → [MeN(CH2NMe)2CH]3-SiMe (eee)
[MeN(CH2NMe)2CH]3-SiH (aee) → [MeN(CH2NMe)2CH]3-SiH (eee)
Scheme 3 Lithiation of R2SiMe2 (5).
at 40.5 and 39.9 ppm correspond to the methyl groups at nitrogen
in 2- and 4-positions, respectively. Compared to the non-lithiated
5, these chemical shifts differ only marginally. A resonance in the
29Si NMR spectrum at 0.0 ppm completes the spectroscopic data.
The result of a crystal structure determination of 8 is shown
in Fig. 6. The compound crystallises in the triclinic space group
Fig.
5
Molecular structure of R3SiMe calculated on the
M05-2X/6-311G(d,p) level (and 6-311(2df) for silicon) of theory.
Only selected hydrogen atoms are shown for clarity. Selected bond lengths
and angles [A,deg] Si(1)–C(1) 1.945, Si(1)–C(7) 1.937, N(1)–C(1) 1.455,
N(1)–C(2) 1.444, N(1)–C(4) 1.465, H(42) ◊ ◊ ◊ H(44) 2.102, C(1)–Si(1)–C(7)
124.4, C(1)–Si(1)–C(13) 105.8, C(1)–Si(1)–C(19) 105.4, C(1)–N(1)–C(4)
113.0, c[C(4)ax] 98.0, c[C(5)eq] -19.7.
¯
P1. Despite several attempts of crystallisation and examination of
˚
various crystals, the crystal quality of the best measurement was
still limited. There is an inversion centre lying in the middle of the
C2Li2 ring, formed by dimerisation of two units of monomeric 8.
The coordination spheres of the two lithium atoms are completed
state aee-conformations and the eee-conformation (for all R sub-
stituents). The values demonstrate that the conformation-based
energies are coarsely additive. R2SiMe2 (2) with two R groups
needs about two times the energy needed for the conformational
change in TMTAC, while R3SiMe and R3SiH with three groups
require about three times that value.
Lithiation of R2SiMe2 (5) and RSiMe2Ph (4)
Lithiated R2SiMe2 (5). It was earlier shown that the
trimethylsilyl derivative of TMTAC, R-SiMe3, can be metallated
with butyl lithium. This metallation takes place at the silicon-
bound methyl groups rather than at the nitrogen heterocycle.
The question arose, whether it would be possible to metallate
R2SiMe2 (5) and if so, where the metallation would take place.
We reacted 5 with n-butyl lithium at -78 ◦C in hexane and
obtained a product, metallated at the silicon bound methyl
group: [{MeN(CH2NMe)2CH}SiMe(CH2Li)]2 (8) (Scheme 3).
1
7
Compound 8 was characterised by H, Li, 13C and 29Si NMR
spectroscopy in C6D6 solution.
A broad resonance in the 1H NMR spectrum at 0.14 ppm
represents the lithiated SiCH2 unit, while a signal at 0.46 ppm
corresponds to the unchanged SiCH3 group. The 7Li NMR
spectrum contains a resonance at 1.2 ppm. A 13C NM resonance at
-2.8 ppm corresponds to the carbanionic function SiCH2Li, while
a signal at -1.1 ppm represents the Si-bound methyl group and one
at 77.5 ppm, the ring carbon atom attached to silicon (SiCHN2).
The methylene NCH2N units give a signal at 79.3 ppm. Resonances
Fig. 6 Molecular structure of 8 in the solid state. Selected bond
lengths and angles [A, deg]: Li(1)–C(13) 2.210(7), Li(1)–N(1)
˚
2.168(7), Li(1¢)–N(4) 2.181(8), Si(1)–C(1) 1.956(3), Si(1)–C(13) 1.803(4),
Si(1)–C(14) 1.887(3), N(1)–C(1) 1.492(4), N(1)–C(4) 1.477(4), Li(1)–Li(1¢)
2.542(13), Li(1)–C(13)–Li(1¢) 70.4(3), Li(1)–C(13)–Si(1) 98.4(2),
C(13)–Li(1)–C(13¢) 109.6(3), C(13)–Si(1)–C(1) 109.7(2), N(1)–C(1)–Si(1)
111.6(2), N(1)–C(1)–N(2) 115.4(3), c[C(4)ax]C 95.0, c[C(5)ax] 108.0,
c[C(6)eq] -20.4.
108 | Dalton Trans., 2012, 41, 104–111
This journal is
The Royal Society of Chemistry 2012
©

View Article Online
by two contacts each to nitrogen atoms of the 2,4,6-trimethyl-
2,4,6-triaza-cyclohex-1-yl rings.
Most of the bond lengths in 8 are relatively similar to the
corresponding ones in 5. The bonds between the silicon atom and
˚
the ring carbon atoms at 1.956(3) to C(1) and 1.930(4) A to C(7)
˚
are slightly longer than in 5 (1.923(2) A). Close are the distances to
˚
˚
the methyl group in 8 (1.887(3) A) and in 5 (1.882(2) A). Shorter
˚
though is the bond Si(1)–C(13) to the carbanion at 1.803(4) A.
This is the structural proof of the concept of anion stabilisation
by adjacent silicon atoms through hyperconjugation.
˚
The distance Li(1)–C(13) is 2.210(7) A and that to the lithium
˚
symmetry equivalent, Li(1)¢–C(13), 2.200(8) A. The coordinative
bonds between lithium and nitrogen atoms are shorter than these:
˚
N(1)–Li(1) 2.168(7) N(4¢)–Li(1) 2.181(8) A. Despite the relative
Fig. 7 1H NMR spectrum of a solution of compound 9 in d8-THF
(* denotes a resonance due to silicone grease).
˚
close proximity of the two lithium atoms at 2.542(13) A, they
are pushed apart from one another as becomes obvious from the
deformation of the Li2C2 ring; the C–Li–C angles are widened to
109.6(3)◦, whereas the Li–C–Li angles are compressed to 70.4(3)◦;
this leads to an Li2C2 rhomb.
Conclusion
The above experimental results show, that up to two 2,4,6-
trimethyl-2,4,6-triaza-cyclohex-1-yl groups can easily be intro-
duced into the coordination sphere of silicon by reacting chlorosi-
lanes with lithiated TMTAC. However, threefold substitution pro-
vokes a problem due to steric overcrowding. This was confirmed by
quantumchemical calculations. Upon attempts to achieve fourfold
substitution of silicon in SiCl4 with 2,4,6-trimethyl-2,4,6-triaza-
cyclohex-1-yl groups by lithiated TMTAC, the system swerved
and in small quantities a product of a formal insertion of SiCl2
into the C–N bonds of the TMTAC ring was observed. In most
of the compounds the 2,4,6-trimethyl-2,4,6-triaza-cyclohex-1-yl
units adopt the same conformation as in free TMTAC, the aee-
conformation.
The conformation of the 2,4,6-triaza-cyclohex-1-yl groups is
unusual. As shown above, the preferred conformation is aee. In
8 one of the rings and its substituents adopts the aee- (bound to
C(7)) while the other has the aae-conformation (involving C(1)).
This is presumably due to the spatial preorgansisation by Li ◊ ◊ ◊ N
interactions which leads to steric repulsion between the methyl
groups at the nitrogen and silicon atoms C(14) and C(5).
Lithiated RSiMe2Ph (4). A similar reactivity pattern towards
metallation with butyl lithium is observed when RSiMe2Ph is
employed as a substrate (Scheme 4).
Systems containing methyl groups at silicon, like R2SiMe2 and
RSiMe2Ph (R = 2,4,6-trimethyl-2,4,6-triaza-cyclohex-1-yl) can be
deprotonated by butyl lithium. The lithiation does not take place
at the heterocyclic unit, but at the Si–Me groups.
Experimental
General methods
Scheme 4 Lithiation of RSiMe2Ph (4).
All manipulations were carried out under an atmosphere of dry
nitrogen in anhydrous and degassed solvents (distilled before
use from standard drying agents) using Schlenk vacuum-line
techniques or an MBRAUN UNILAB glovebox. NMR spectra were
recorded in C6D6, d8-toluene and d8-THF on a BRUKER AVANCE
DRX 500 and a BRUKER AVANCE 600 spectrometer. All NMR
chemical shifts were referenced to the residual peaks of the
protons of the used solvents. Elemental analyses were performed
using an ELEMENTAR VARIO EL III CHNS and a LECO CHNS 932 (the
carbon contents of silicon compounds are systematically low due
to carbide formation).9 Bis[(2,4,6-trimethyl-2,4,6-triazacyclo-hex-
1-yl)-lithium]-(1,3,5-trimethyl-1,3,5-triaza-cyclohexane)-adduct
([(RLi)2·(RH)]) (1) was prepared as previously described by us.1
Product 9 is obtained in 33% yield directly, as it precipitates after
addition of butyl lithium to a cold solution of 4. Despite various
attempts, obtaining crystals suitable for structural analysis was not
possible. Compound 4 was characterised by NMR spectroscopy.
1
Fig. 7 shows the H NMR spectrum recorded from a d8-thf
solution of 4. Despite there being no information about the degree
of aggregation in this donor solvent, the data suggest an aee-
conformation, because there are three signals for the N-bound
methyl groups. This is unusual as in many other cases we observed
only two signals for these methyl groups in TMTAC derivatives
indicating rapid exchange. The non-metallated methyl group at
silicon and the Li–CH2–Si unit can easily be distinguished by their
integrals and their chemical shifts of 0.29 and -1.91 ppm.
Consistently, the 13C NMR spectrum also contains three
resonances for the methyl groups at nitrogen at 41.2, 40.5 and
39.3 ppm. The Si-bound methyl group causes a resonance at 2.0,
the carbanionic group at -12.1 ppm. The 29Si NMR shows a single
peak at -2.3 ppm.
Preparations
[MeN(CH2NMe)2CH]-SiEt3 (2). 2.0 g (5.0 mmol) lithiated
TMTAC (1) are placed in a vessel and 20 mL diethyl ether
were added at -78 ◦C. Chlorotriethylsilane (1.7 mL, 1.5 g,
This journal is
The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 104–111 | 109
©

View Article Online
2
10 mmol) diluted with 10 mL diethyl ether added dropwise. The
resulting suspension was allowed to warm to ambient temperature
overnight. The precipitated lithium chloride was filtered off and
washed three times with 5 mL of diethyl ether. After removal of
the volatile solvent, the product was yielded as yellowish liquid.
The pure product was yielded by distillation at 80–84 ◦C and 0.1
mbar. Yield 1.79 g (7.36 mmol, 73%). 1H NMR (500 MHz, C6D6)
d 3.41 (d, 2H, NCH2N, 2JHH = 10 Hz), 3.02 (s, 1H, SiCH), 2.92 (d
br, 2H, NCH2N, 2JHH = 10 Hz), 2.59 (s, 6H, 2-NCH3), 1.80 (s, 3H,
NMR (500 MHz, C6D6) d 3.37 (d, br, 4H, NCH2N, JHH = 8.7
Hz), 3.26 (s, 2H, SiCH), 3.02 (s, br, 4H, NCH2N), 2.69 (s, 12H,
2-NCH3), 1.80 (s, 6H, 4-NCH3), 0.43 (s, 6H, SiCH3); 13C-NMR
(125 MHz, C6D6) d 79.4 (br, NCH2N), 75.2 (br, SiCH), 43.9 (br, 2-
NCH3), 39.9 (4-NCH3), -1.1 (SiCH3); 29Si-NMR (99 MHz, C6D6)
d 0.0 (s); MS (EI, 70 eV) m/z 128 (TMTAC, 100%), 101 (TMTAC-
(Me)2, 19%), 85 (TMTAC-(Me)3, 51%), 59 (SiMe2, 26%), 44 (SiMe,
90%).
[MeN(CH2NMe)2CH]2-SiPh2 (6). The procedure was similar
to that described for 5. Employed reagent: dichlorodiphenylsilane
(1.03 mL, 1.26 g, 5.00 mmol, 20 mL hexane), 1 (2.0 g, 5.0 mmol,
20 mL hexane). After filtration and concentration of the solution
under reduced pressure, the product precipitated as fine powder
upon cooling to 4 ◦C and was yielded by filtration. Yield 0.1 g
(0.2 mmol, 4%). 1H-NMR (500 MHz, C6D6) d 7.83 (d, 2H, SiPh,
3JHH = 6.3 Hz), 7.21 (m, 10H, SiPh), 3.77 (s, 2H, SiCH), 3.44 (d,
4H, NCH2N, 2JHH = 10.3 Hz), 2.91 (d, 4H, NCH2N, 2JHH = 10.3
Hz), 2.63 (s, 12H, 2-NCH3), 1.72 (s, 6H, 4-NCH3); 13C-NMR (125
MHz, C6D6) d 137.7 (SiPh), 137.6 (SiPh), 135.0 ((SiPh), 129.5
(SiPh), 78.9 (br, NCH2N), 74.0 (SiCH), 44.7 (br, 2-NCH3), 39.8
(4-NCH3); 29Si-NMR (99 MHz, C6D6) d -23.3 (s).
3
4-NCH3), 0.97 (t, 12H, SiCH2CH3, JHH = 7.9 Hz), 0.65 (q, 6H,
3
SiCH2CH3, JHH = 7.9 Hz); 13C NMR (125 MHz, C6D6) d 79.7
(br, NCH2N), 75.5 (SiCH), 43.9 (br, 2-NCH3), 39.8 (4-NCH3), 7.5
(CH2CH3), 2.9 (CH2CH3); 29Si-NMR (99 MHz, C6D6) d 3.5 (s).
Found: C, 59.20; H, 12.01; N, 17.26. Calc. for C12H29N3Si (243.21):
C, 57.62; H, 11.75; N, 17.09%.
[MeN(CH2NMe)2CH]-SiPh3 (3). The procedure was the same
as described above for 2. Employed reagents: 1 (3.99 g, 10.0 mmol,
40 mL Et2O), chlorotriphenylsilane (5.90 g 20.0 mmol, 20 mL
Et2O). The compound was isolated by reducing the volume of
the raw solution after filtration and crystallisation. Yield 3.02 g
1
(7.80 mmol, 78%). H NMR (500 MHz, C6D6) d 7.88 (m, 6H,
SiPh), 7.19 (m, 9H, SiPh), 3.98 (s, 1H, SiCH), 3.42 (d, 2H, NCH2N,
2JHH = 10.1 Hz), 2.90 (d, br, 2H, NCH2N, 2JHH = 10.0 Hz), 2.64 (s,
6H, o-NCH3), 1.68 (s, 3H, 4-NCH3); 13C NMR (125 MHz, C6D6)
d 136.9 (SiPh), 136.0 (SiPh), 129.6 (SiPh), 128.0 (SiPh), 79.0 (br,
NCH2N), 76.9 (SiCH), 44.9 (br, 2-NCH3), 39.7 (4-NCH3); 29Si
NMR (99 MHz, C6D6) d -20.7 (s) Found: C, 74.37; H, 7.54; N,
10.84. Calc. for C24H29N3Si (387.21): C, 72.39; H, 7.79; N, 9.83%.
2,4,6-Trimethyl-2,4,6-triaza-1,1-dichloro-1-sila-cycloheptane
(7). A few tiny crystals of this compound were obtained in the
following way. Freshly distilled tetrachlorosilane (0.14 mL, 0.20
g, 1.2 mmol) were condensed onto freshly prepared 1 (1.0 g,
2.5 mmol) under vacuum at liquid nitrogen temperature. The
mixture was warmed to -78 ◦C and allowed to slowly warm to
ambient temperature overnight. The lithium chloride was filtered
off and the obtained clear solution cooled to 30 ◦C, whereby a few
crystals of 7 precipitated within one week. Characterisation was
undertaken solely by X-ray diffraction.
[MeN(CH2NMe)2CH]-SiMe2Ph (4). The procedure was simi-
lar to that described for 2. Employed reagents: 1 (2.0 g, 5.0 mmol,
20 mL hexane), chlorodimethylphenylsilane (1.77 mL, 1.71 g,
10.0 mmol, 20 mL hexane). The compound was isolated by
reducing the volume of the raw solution after filtration and
crystallisation. Yield 3.02 g (7.80 mmol, 78%). The volatile solvent
was removed by application of reduced pressure. The product
remains as an orange oil, which contained some impurities
and was not purified, as distillation led to decomposition and
crystallisation was not possible. It was employed as such in further
[{MeN(CH2NMe)2CH}2-SiMe(CH2Li)]2 (8). n-Butyl lithium
(1.6 M in hexane, 1.0 mL, 1.6 mmol) was dropped into a solution
of [MeN(CH2NMe)2CH]2SiMe2 (5) (0.43 g, 1.4 mmol) in 10 mL
hexane at -78 ◦C. Warming the reaction mixture to ambient
temperature resulted in precipitation of a colourless solid, which
was filtered and washed three times with 5 mL of hexane and dried
in vacuum. Yield 0.21 g (0.33 mmol, 46%). 1H NMR (500 MHz,
C6D6) d 3.52 (d, 8H, NCH2N, 2JHH = 9.9 Hz), 3.04 (br, 4H, SiCH),
2.96 (d, 8H, NCH2N, 2JHH = 10.0 Hz), 2.71 (s, 24H, 2-NCH3), 2.21
1
reactions. Yield 1.84 g (6.98 mmol, 69%). H NMR (500 MHz,
3
C6D6) d 7.67 (d, 2H, SiPh, JHH = 7.8 Hz), 7.20 (m, 3H, SiPh),
2
7
3.46 (d, 2H, NCH2N, JHH = 10.0 Hz), 3.14 (s, 1H, SiCH), 2.80
(12H, 4-NCH3), 0.46 (s, 6H, SiCH3), 0.14 (s, 4H, SiCH2Li); Li
2
NMR (194 MHz, C6D6) d 1.2 (s); 13C NMR (125 MHz, C6D6) d
79.3 (NCH2N), 77.5 (SiCH), 40.5 (2-NCH3), 39.9 (4-NCH3), -1.1
(SiCH3), -2.8 (SiCH2Li); 29Si NMR (99 MHz, C6D6) d 0.0 (s).
(d, 2H, NCH2N, JHH = 10.0 Hz), 2.48 (s, 6H, 2-NCH3), 1.79 (s,
3H, 4-NCH3), 0.43 (s, 6H, SiCH3); 13C NMR d (125 MHz, C6D6)
139.7 (SiPh), 134.5 (SiPh), 129.1 (SiPh), 80.2 (br, NCH2N), 77.8
(SiCH), 42.9 (2-NCH3), 39.9 (4-NCH3), -1.8 (SiCH3); 29Si NMR
(99 MHz, C6D6) d -6.8 (s); MS (EI, 70 eV, liquid) 176 (M+-Ph-Me,
66%), 162 (M+-PhMe2, 30%), 135 (SiMe2Ph, 100%), 128 (TMTAC,
100%), 121 (SiMePh, 12%), 105 (SiPh, 13%), 85 (TMTAC-Me3,
19%), 42 (SiMe, 49%).
[MeN(CH2NMe)2CH]-SiMePh(CH2Li) (9). The procedure
was similar to that described for 8. Employed reagents:
n-butyl lithium (1.6
M in hexane, 2.2 mL, 3.5 mmol),
[MeN(CH2NMe)2CH]2SiMePh (6) (0.94 g, 3.3 mmol). Yield 0.30
g (1.1 mmol, 33%). 1H NMR (500 MHz, d8-THF) d 7.77 (d, 2H,
3
3
[MeN(CH2NMe)2CH]2SiMe2 (5). A stirred suspension of
lithiated TMTAC (1 3.0 g, 7.5 mmol) in 20 mL pentane was
dropped into a solution of dichlordimethylsilane (0.90 mL, 1.0 g,
7.5 mmol) in 30 mL pentane at -78 ◦C. The resulting suspension
was allowed to warm to ambient temperature overnight. The pre-
cipitated lithium chloride was filtered off. The solvent was removed
under reduced pressure, which yielded crystalline material in the
SiPh, JHH = 7.0 Hz), 7.16 (m, 3H, SiPh, JHH = 7.0 Hz), 3.65
2 2
(d, 1H, NCH2N, JHH = 9.2 Hz), 3.60 (d, 1H, NCH2N, JHH
9.3 Hz), 2.85 (s, 1H, SiCH), 2.72 (d, 1H, NCH2N, JHH = 8.9
Hz), 2.68 (d, 1H, NCH2N, JHH = 8.3 Hz), 2.51 (s, 3H, NCH3),
=
2
2
2.29 (s, 3H, NCH3), 2.02 (s, 3H, NCH3), 0.30 (s, 3H, SiCH3),
-1.91 (s, 2H, CH2Li); 13C NMR (125 MHz, d8-THF) d 148.2
(SiPh), 133.9 (SiPh), 126.5 (SiPh), 126.3 (SiPh), 80.7 (NCH2N),
80.5 (NCH2N), 80.3 (SiCH), 41.2 (NCH3), 40.5 (NCH3), 39.3
1
last stages of concentration. Yield 0.84 g (2.7 mmol, 35%). H
110 | Dalton Trans., 2012, 41, 104–111
This journal is
The Royal Society of Chemistry 2012
©

View Article Online
Table 3 Crystal and refinement data for the structure determinations of compounds 1, 5, 7 and 8
Compound
TMTAC
5
7
8
Formula
Mr
C6H15N3
129.21
0.25 ¥ 0.25 ¥ 0.25
(Capillary)
Monoclinic
P21/n
11.5902(2)
18.9804(3)
14.0480(2)
90
95.5696(9)
90
3075.78(8)
16
1.116
0.071
3.0–30.0
78902
8927
0.041
6917
565
0.041/0.103
0.057/0.116
0.251/-0.202
821375
C14H34N6Si
314.56
0.06 ¥ 0.16 ¥ 0.28
C6H15Cl2N2Si
228.20
0.24 ¥ 0.28 ¥ 0.30
C28H66Li2N12Si2
640.99
0.16 ¥ 0.28 ¥ 0.30
cryst. size [mm]
cryst. syst.
Monoclinic
P21/c
8.2646(5)
13.3023(8)
16.6558(7)
90
96.950(4)
90
1817.65(17)
4
1.149
0.134
3.0–27.5
17917
4048
0.039
3211
198
0.0428/0.1038
0.0600/0.1150
0.362/-0.280
821376
Orthorhombic
Pbca
6.7426(2)
15.1934(3)
21.4697(5)
90
Triclinic
P¯ı
Space group
˚
a/A
7.0976(11)
12.225(2)
12.2926(17)
61.357(6)
77.351(7)
86.350(6)
912.3(2)
1
1.167
0.134
3.0–27.5
10304
3755
0.146
1688
214
0.0658/0.1196
0.1894/0.1512
0.032/-0.339
821378
˚
b/A
˚
c/A
a (◦)
b (◦)
g (◦)
90
90
3
˚
V/A
2199.42(9)
8
1.378
0.656
3.3–27.5
20157
2505
0.031
2428
169
0.0327/0.0709
0.0339/0.0714
0.314/-0.252
821377
Z
Dc/Mg m-3
m/mm-1
q-range [◦]
reflns collected
unique reflns
Rint
Reflections I > 2s(I)
refined param.
R1, wR2 [I > 2s(I)]
R1, wR2 [all data]
-3
˚
Drmax/min/e A
CCDC no.
(NCH3), 2.0 (SiCH3), -12.1 (CH2Li); 29Si NMR (99 MHz, d8-
THF) d -2.3 (s).
4 J. Arnold, V. Knapp, J. A. R. Schmidt and A. Shafir, J. Chem. Soc.,
Dalton Trans., 2002, 3273.
5 (a) M. Schlosser and J. Hartmann, Angew. Chem., 1973, 85, 544; M.
Schlosser and J. Hartmann, Angew. Chem., Int. Ed. Engl., 1973, 12,
508; (b) W. Bauer and L. Lochmann, J. Am. Chem. Soc., 1992, 114,
7482.
6 C. Strohmann and V. H. Gessner, Angew. Chem., 2007, 119, 8429; C.
Strohmann and V. H. Gessner, Angew. Chem. Int. Ed. Engl., 2007, 46,
828.
Crystallographic structure determinations. The crystallo-
graphic data sets for TMTAC, 5, 6 and 8 were collected on a
Bruker Nonius Kappa CCD, operated with monochromatic Mo-
˚
Ka radiation (l = 0.71073 A). Structure solutions and refinements
were undertaken with the programs SHELXS-9720 and SHELXL-
97.20 The crystal quality of 8 was limited. The low scattering
power of the crystal is the reason for the relative small number
of reflections with I > 2s(I). We used data up to qmax = 27.5◦
and thereby could include some 50 reflections (between q = 25
and 27◦) to improve the validity of structural information. This
leads to a lower completeness and worse R and Rint factors, but
increases the amount of information and thus the validity of
the structure. Experimental details are listed in Table 3. CCDC
reference numbers 821375 (TMTAC), 821376 (5), 821377 (6),
821378 (8). For crystallographic data see ESI.†
7 H. H. Karsch, Chem. Ber., 1996, 129, 483.
8 (a) D. Bojer, I. Kamps, X. Tian, A. Hepp, T. Pape, R. Fro¨hlich and N.
W. M i t z e l , Angew. Chem., 2007, 119, 4254; (b) R. D. Ko¨hn, G. Seifert
and G. Kociok-Ko¨hn, Chem. Ber., 1996, 129, 1327.
9 I. Kamps, A. Mix, R. J. F. Berger, B. Neumann, H.-G. Stammler and
N. W. Mitzel, Chem. Commun., 2009, 5558.
10 I. Kamps, D. Bojer, S. A. Hayes, R. J. F. Berger, B. Neumann and
Norbert W. Mitzel, Chem.–Eur. J., 2009, 15, 11123.
11 C. Strohmann and V. H. Gessner, Chem.–Asian J., 2008, 3,
1929.
12 I. Kamps, I. Langlitz, A. Mix, B. Neumann, H.-G. Stammler and N.
W. M i t z e l , Dalton Trans., 2009, 8363.
13 B. J. Hellmann, I. Kamps, B. Neumann, H.-G. Stammler and N. W.
Mitzel, Chem. Commun., 2010, 46, 6536.
14 (a) R. A. Y. Jones, A. R. Katritzkya and M. Snarey, J. Chem. Soc. (B),
1970, 135; (b) C. H. Bushweller, M. Z. Lourandos and J. A. Brunelle,
J. Am. Chem. Soc., 1974, 96, 1591; (c) R. O. Hutchins, L. D. Kopp and
E. L. Eliel, J. Am. Chem. Soc., 1968, 90, 7174; (d) N. J. Harris and K.
Lammertsma,, J. Am. Chem. Soc., 1997, 119, 6583; (e) J. G. Jewett, J. J.
Breeyear, J. H. Brown and C. H. Bushweller, J. Am. Chem. Soc., 2000,
122, 308.
15 V. G. Zverev, V. V. Kuznetsov, A. N. Rykov, N. M. Karasev, N.
N. Makhova and L. V. Vilkov, Russ. J. Phys. Chem. A, 2007, 81,
1639.
Acknowledgements
This work was supported by Deutsche Forschungsgemeinschaft
and by Fonds der Chemischen Industrie. We are grateful to Peter
Mester for NMR measurements and Brigitte Michel for elemental
analyses.
16 C. Sicking, Dissertation, University of Bielefeld, 2010.
17 D. G. Anderson, A. J. Blake, S. Cradock, E. A. V. Ebsworth, D. W. H.
Rankin, H. E. Robertson and A. J. Welch, J. Chem. Soc., Dalton Trans.,
1987, 3035.
18 N. W. Mitzel, Z. Naturforsch., 2003, 58b, 369.
19 B. E. Lindfors, J. L. Male and T. J. R. Weakly, Acta Crystallogr., Sect.
C: Cryst. Struct. Commun., 1998, 54, 694.
Notes and references
1 D. Seyferth and M. A. Weiner, J. Org. Chem., 1959, 24, 1395.
2 (a) F. H. Ko¨hler, N. Hertkorn and J. Blu¨mel, Chem. Ber., 1987, 120,
2081; (b) V. H. Gessner and C. Strohmann, J. Am. Chem. Soc., 2008,
130, 14412.
3 V. H. Gessner and C. Strohmann, Angew. Chem., 2007, 119, 4650; V.
H. Gessner and C. Strohmann, Angew. Chem. Int. Ed. Engl., 2007, 46,
4546.
20 G. M Sheldrick, Acta Crystallogr., Sect. A: Found. Crystallogr., 2008,
A64, 112.
This journal is
The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 104–111 | 111
©