Modelling nucleophilic substitution at silicon in solution, using hypervalent silicon compounds based on 2-pyridones

Article abstract of DOI:10.1039/a904402g

A novel method for performing structure correlations in solution is described. Examination of how the ¹³C chemical shifts of the ring carbons of substituted 2-pyridones change on complexation of the oxygen with silicon has enabled the % Si-O bond formation to be determined in solution for a number of pentacoordinate silicon species with 2-pyridones as ligands. The % pentacoordination in these complexes has been determined from the ²⁹Si chemical shift using model compounds for the tetracoordinate and pentacoordinate limiting cases. Correlation of the % Si-O bond formation with % pentacoordination enables the pathway for substitution at silicon to be mapped in solution. The generality of these techniques is examined using a series of related aromatic ligands.

Full text of DOI:10.1039/a904402g

View Online / Journal Homepage / Table of Contents for this issue
Modelling nucleophilic substitution at silicon in solution, using
hypervalent silicon compounds based on 2-pyridones
Alan R. Bassindale,* Moheswar Borbaruah, Simon J. Glynn, David J. Parker and
Peter G. Taylor*
The Chemistry Department, The Open University, Walton Hall, Milton Keynes, UK MK7 6AA
Received (in Cambridge, UK) 2nd June 1999, Accepted 27th July 1999
A novel method for performing structure correlations in solution is described. Examination of how the 13
C
chemical shifts of the ring carbons of substituted 2-pyridones change on complexation of the oxygen with silicon
has enabled the % Si–O bond formation to be determined in solution for a number of pentacoordinate silicon
species with 2-pyridones as ligands. The % pentacoordination in these complexes has been determined from the
29Si chemical shift using model compounds for the tetracoordinate and pentacoordinate limiting cases. Correlation
of the % Si–O bond formation with % pentacoordination enables the pathway for substitution at silicon to be
mapped in solution. The generality of these techniques is examined using a series of related aromatic ligands.
We sought to develop a method that would enable similar
Introduction
structure correlations to be obtained in the solution phase,
using 13C and 29Si NMR spectroscopy as the monitoring tools.
Initial results have been published7,8 and this paper provides a
detailed description of the technique.
X-Ray crystallographic studies have been used by Dunitz,
Bürgi1 and others2–6 to determine the molecular pathways of a
variety of reactions. Such crystal structure correlations involve
collecting together the crystal structures of a range of structur-
ally similar compounds and then sequencing them on the basis
of key crystal data (usually inter-atom distances). Thus, a
picture of the gradual molecular deformation on going from
reactants to products is built up. Each individual structure rep-
resents a snapshot of the reaction at a particular point on the
modelled reaction profile. This methodology has been applied
to nucleophilic substitution at silicon by a number of groups.3–6
The general picture that has emerged involves formation of the
nucleophile–silicon bond accompanied by lengthening of the
silicon–leaving group bond. The tetrahedral reactants are con-
verted into a trigonal bipyramidal structure with the non-
participating groups equatorial followed by reversion to a
tetrahedral structure. For example, Baukov4 and Pestunovich5
separately examined the crystal structures of a range of related
amidomethylhalosilanes (1). The resulting trajectory for nucleo-
Unlike X-ray crystallography, which gives detailed inform-
ation on the bond lengths and spacial arrangement of atoms,
NMR only provides an insight into the connectivity and local-
ised environment around particular nuclei.9 Furthermore quan-
titative molecular coordinates can usually only be obtained
through the use of model compounds. Thus, structural corre-
lations using NMR require careful design of appropriate
systems. The minimum requirement for studying the progress
of a reaction is two parameters that measure the extent of bond
formation or breaking and/or the change in geometry around a
key site. NMR chemical shifts are sensitive to a number of
parameters such as bond order, coordination number and local-
ised electronic environment. Thus deconvolution of NMR data
to provide the required information to construct a reaction
profile is potentially very difficult. Clearly the task is made
much easier if the system is designed so that particular chemical
shifts essentially depend upon only one of the variables.
The 29Si chemical shift is particularly susceptible to the
coordination number around the silicon, having a value of
between ϩ24 ppm and Ϫ76 ppm for four coordinate silicon,
Ϫ47 ppm and Ϫ110 ppm for five coordinate silicon10 and Ϫ142
ppm and Ϫ220 ppm for six coordinate silicon.11 Although the
regions overlap for four and five coordination, the addition of
an extra ligand to a four coordinate silicon leads to a substan-
tial upfield shift. Thus, within a series of structurally similar
compounds, the 29Si chemical shift can be used as an indicator
of the gradual change in coordination number.12
The extent of bond formation or breaking is a little more
difficult to measure. For example, the attacking or leaving atom
may not be NMR active or its chemical shift may depend upon
a number of factors. One alternative is to monitor the chemical
shifts of atoms attached to the attacking atoms. However, this
may lead to a smaller variation in the chemical shifts and there
is still no guarantee that the chemical shift will only depend
upon the extent of bond formation or breaking. A better strat-
egy is to use conjugated systems where bonding changes at the
nucleophilic atom are relayed to more distant atoms that are
not affected by other changes at the reaction centre. Further-
more the chemical shift change in such conjugated systems is
often substantial. We chose to base our initial studies on the
philic substitution at silicon exhibited a hyperbolic relation-
ship between the Si–O and Si–X bond lengths and a regular
variation of the position of the silicon atom relative to the
plane of the three equatorial carbon atoms.
J. Chem. Soc., Perkin Trans. 2, 1999, 2099–2109
This journal is © The Royal Society of Chemistry 1999
2099

View Online
substituted 2-pyridones (2). Judicious choice of substituent Y
and leaving group X leads to a series of compounds with differ-
ing extents of O–Si bond formation and Si–X bond cleavage
enabling us to model the pathway for substitution at silicon.
Results and discussion
The model compounds 3a–d, 4a–d, 5a–d, 6a–d, 7a–d and 8a–d
were prepared from the commercially available 2-pyridones, as
shown in Scheme 1. Treatment of the substituted 2-pyridone
Fig. 1 The change in chemical shift as a function of Si–O bond
formation for compounds 5a–d.
similar to those expected of compound 2 with 0% O–Si bond
formation. There is the possibility of some Si–O bond form-
ation with compound 9, but this was shown to be negligible
by the 29Si chemical shifts, which are consistent with a four
coordinate disiloxane.15 Similarly, the 13C chemical shifts of the
ring carbons are equivalent to those of non-silicon containing
N-alkyl-2-pyridones. X-Ray crystal structures of amidomethyl-
halosilanes 1, where X is phenoxide, have a PhO–Si bond length
very close to that of a covalent O–Si bond and little distortion
of the tetrahedral arrangement around silicon.6 Compound 10
provides a good model for 100% O–Si bond formation. The 13
C
Scheme 1 Synthesis of model compounds.
chemical shifts of the ring carbons are similar to those of
2-methoxypyridinium salts.16 The only drawback with using 10
as the limiting case is that compound 2 with 100% Si–O bond
formation involves formation of a five membered ring. This
may lead to geometry changes in the aromatic ring that are not
present in compound 10.
with diethylaminotrimethylsilane gave the corresponding 2-
trimethylsiloxypyridine.13 Reaction of this with chloromethyl-
dimethylchlorosilane gave the chloro derivatives 3b–8b,14
whereas treatment with chloromethyldimethylsilyl triflate gave
the triflato derivatives 3d–8d. The bromo derivatives were pre-
pared by treating the substituted 2-trimethylsiloxypyridine with
bromomethyldimethylchlorosilane. Reaction using chloro-
methyldimethylfluorosilane gave the chloro derivative, thus the
fluoro derivatives were prepared by treating the chloro deriv-
atives with antimony trifluoride.
We have previously shown that when a 10% w/w solution of
N-methyl-2-pyridone in CDCl3 is titrated with successive
amounts of trimethylsilyl trifluoromethanesulfonate, silylation
is complete and exchange between the 2-pyridone to the pyr-
idinium forms is fast on the NMR time scale.7 There is a good
linear correlation between the 13C chemical shifts of each ring
carbon and the mole fraction of the pyridinium form. However,
the direction and extent of the change of the chemical shift
varies greatly from one carbon to another. In calculating the
extent of Si–O bond formation in compounds of the type 2, we
have assumed that the 13C chemical shifts of the ring carbons
vary in a linear fashion with the extent of Si–O bond formation
between the two extremes of 0% bond formation, as repre-
sented by compound 9, and 100% bond formation, as repre-
sented by compound 10. The justification for this assumption is
best demonstrated in the 6-chloro-2-pyridone system. Fig. 1
shows a plot of the extent of bond formation against the 13C
chemical shifts of the ring carbons. The data show that on
going from the 2-pyridone to the pyridinium forms the chem-
ical shift of each position changes in a coherent fashion and can
increase or decrease. In general for all the simple 2-pyridones
studied, the chemical shifts of carbons 4, 5 and 6 increase with
Si–O bond formation whereas that of carbon 3 often decreases.
The increase arises from the shift in π electron density away
from the ring towards the exocyclic oxygen on Si–O bond
formation. The decrease at carbon 3 is not easy to explain.
However, compared to the 1-alkyl-2-pyridone, a similar pattern
is observed when boron trifluoride complexes with 2-methoxy-
pyridine.16
The 13C spectra of these pentacoordinate compounds showed
that as predicted the chemical shift of the ring carbons changed
systematically with leaving group, a similar pattern being
observed for each substituted 2-pyridone. However, in order to
carry out a structure correlation, the extent of bond formation
or breaking needs to be calculated from this chemical shift data.
This is achieved using the corresponding model compounds
9 and 10 that represent 0 and 100% O–Si bond formation
Small deviations from the line are observed which may arise
from experimental error, specific interactions within the model
compound or a non-linear response of the chemical shift to the
extent of bond formation. In each case the % Si–O bond form-
ation has been estimated to give the best fit to the complete set
of data. This can either be achieved visually or by calculating
the % Si–O bond formation for each 13C resonance based on
those observed with the limiting cases 9 and 10. The weighted
respectively. Since these two compounds are the limiting
cases that provide the anchor for the calculation of the % Si–O
bond formation, it is important that they are chosen with
care. Compound 9 was employed, rather than the simpler
N-methyl-2-pyridone, to ensure that the electronic effects of the
N-substituent on the chemical shifts of the ring carbons were
2100
J. Chem. Soc., Perkin Trans. 2, 1999, 2099–2109

View Online
Table 1 % Si–O bond formation for each of the model compounds
3–8 as calculated from the 13C chemical shifts of the ring carbons,
together with the % pentacoordination calculated from the 29Si chemical
shifts
% Si–O
bond
% Penta-
Compound
Y
X
formation
coordination
3a
3b
3c
3d
4a
4b
4c
4d
5a
5b
5c
5d
6a
6b
6c
6d
7a
7b
7c
7d
8a
8b
8c
8d
H
H
H
H
5-Cl
5-Cl
5-Cl
5-Cl
6-Cl
6-Cl
6-Cl
6-Cl
6-Me
6-Me
6-Me
6-Me
3-OMe
3-OMe
3-OMe
3-OMe
3-NO2
3-NO2
3-NO2
3-NO2
F
Cl
Br
OTf
F
30
50
70
90
30
65
70
90
40
50
70
90
50
70
70
90
20
45
72
95
12
40
58
80
74
102
73
10
63
71
66
19
79
101
84
37
93
81
77
21
61
88
68
23
23
78
85
44
Cl
Fig. 2 % Pentacoordination versus the % Si–O bond formation.
Br
OTf
F
Cl
In order to carry out a structural correlation it is necessary to
have some measure of the coordination state of the silicon. This
is measured using the 29Si chemical shift. Again the conver-
sion from chemical shift information to % pentacoordination
requires the use of model compounds. In this instance three
anchor points are required. One to model 0% pentacoordin-
ation involving 0% Si–O bond formation, compound 11, one to
model 0% pentacoordination involving 100% Si–O bond form-
ation, compound 13 and one to model 100% pentacordination,
compound 12. The 29Si NMR chemical shifts of Me3SiCl,
Me3SiF and Me3SiBr all appear within 2 ppm of ϩ28 ppm
and so this value was used to anchor the scale at 0% penta-
coordination with 0% O–Si bond formation.15 The small vari-
ation of the chemical shift with leaving group confirms that the
major arbiter of the 29Si chemical shift is the coordination state
at silicon. The X-ray crystal structure of 14 has been reported
and shows one silicon to be tetracoordinated and the other
completely pentacoordinated.4,19 We have measured the 29Si
chemical shifts in CDCl3 and found that the tetracoordinate
silicon gives rise to a resonance at 26.8 ppm whereas the penta-
coordinate silicon gives rise to a resonance at Ϫ39.9 ppm. The
value for the tetracoordinate silicon is within experimental error
of the limiting value of 28 ppm. We have used the chemical shift
of the pentacoordinate silicon to define 100% pentacoordin-
ation as a chemical shift of Ϫ40 ppm.20 Formation of a fully
pentacoordinate silicon in compound 12 should involve 50%
Si–O bond formation. Compounds 3b and 5b both exhibit 50%
Si–O bond formation and their 29Si chemical shifts are both
within 2 ppm of this proposed limiting value for a pentacoordi-
nate silicon. The second tetrahedral limiting value, correspond-
ing to compound 13, was set to ϩ36 ppm based on the 29Si
chemical shift of O-trimethylsilylated N-methyl-2-hydroxy-
pyridine. For compounds that exhibited a % Si–O bond form-
ation between 0 and 50% the extent of pentacoordination was
calculated using the limiting cases 11 and 12, that is ϩ28 and
Ϫ40 ppm respectively. For compounds that exhibited a % Si–O
bond formation between 50 and 100%, the extent of penta-
coordination was calculated using the limiting cases 12 and 13,
that is Ϫ40 and ϩ36 ppm respectively. Table 1 shows the %
pentacoordination of compounds 3–8.
Br
OTf
F
Cl
Br
OTf
F
Cl
Br
OTf
F
Cl
Br
OTf
average % Si–O bond formation is then calculated with weight-
ings proportional to the size of the variation in the chemical
shift. Using such techniques the uncertainty in the % Si–O bond
formation is estimated to be approximately 5%.
The % Si–O bond formation for each of the model com-
pounds 3–8 are given in Table 1. The extent of bond formation
for each substituted 2-pyridone follows the expected order with
the leaving group ability TfO > Br > Cl > F.17,18 However, the
% Si–O bond formation for each leaving group is modified by
the substituent. Structures 11, 12 and 13 represent the various
structural extremes. Structure 11 involves no O–Si bond form-
ation and 13 complete O–Si bond formation with loss of the
leaving group. In this latter structure there will be some build up
of charge at the oxygen, but most of the charge will be located
around the nitrogen. Structure 12 involves a pentacoordinate
silicon with partial O–Si bond formation and Si–X bond cleav-
age. In this latter structure there is partial build up of positive
charge on the nitrogen. Thus, substituents that stabilise the
build up of positive charge at the nitrogen will favour O–Si
bond formation. The 3-nitro group is strongly electron with-
drawing, thus will resist the build up of positive charge at the
nitrogen and thus, of all the substituents it has the lowest extent
of bond formation for each leaving group. The 6-methyl group
will stabilise the adjacent build up of positive charge and thus
has the greater extent of bond formation. The orders of the
extents of bond formation for each leaving group do vary a
little, but the trend is 6-Me > 5-Cl > 6-Cl > H > 3-OMe >
3NO2.
Interestingly the size of the variation in the % Si–O bond
formation with leaving group decreases as the substituent
changes from nitro through hydrogen to methyl. A similar pic-
ture emerges if one focuses on each leaving group and calculates
the variation of the % Si–O bond formation with substituent.
The triflate leaving group has the largest extent of bond form-
ation and thus a relatively small change in the % Si–O bond
formation is observed with substituent. On the other hand the
fluoro leaving group has the smallest extent of bond formation
and this leads to a larger range of % Si–O bond formation with
substituent.
For each substituent, as the leaving group ability improves
in the order fluoro, chloro, bromo and triflate, the % penta-
coordination first increases then decreases. This change occurs
earlier for substituents that stabilise the positive charge on
nitrogen, that is those that favour Si–O bond formation. Fig. 2
shows how the % pentacoordination varies as a function of %
Si–O bond formation. At first, as the extent of bond formation
increases, the % pentacoordination gets larger as the compound
moves from resembling 11 to resembling 12. At 50% Si–O bond
formation the maximum pentacoordination is observed. This is
associated with the most negative 29Si chemical shifts. As the
extent of bond formation increases from this point, the leaving
group becomes more detached and thus the % pentacoordin-
ation decreases. Eventually, the Si–O bond is completely
J. Chem. Soc., Perkin Trans. 2, 1999, 2099–2109
2101

View Online
formed and the leaving group lost with the silicon reverting to
a completely tetrahedral structure, 13. As expected the com-
pounds with the poorest leaving group, the fluoro derivatives, lie
on the left of the graph and show only an increase in % penta-
coordination. The chloro derivatives are clustered around the
point of full pentacoordination and the bromides exhibit about
70% Si–O bond formation with decreasing pentacoordination.
Finally the triflates, which contain the best leaving group, show
almost complete Si–O bond formation with little pentacoordin-
ation.
An alternative explanation is that rather than a continuum of
structures from 11 through 12 to 13, the reaction mixture con-
tains only 11, 12 and 13 in equilibrium. As the substituent and
leaving group changes so the equilibrium position first moves
from favouring 11 to favouring 12 and then 13. Such a series of
equilibria would have to be rapid on the NMR timescale in
order to give single sets of resonances in the NMR spectra of
each nucleus. X-Ray crystallography demonstrates that a con-
tinuum of structures is possible in the solid state and thus we
would assume that a similar variety is possible in the solution
phase.4–6 This is supported by conductivity studies on solutions
of the 2-pyridones in acetonitrile. If the equilibria hypothesis
were correct, compounds with about 75% Si–O bond formation
should conduct electricity since they will contain equal
amounts of covalent 12 and ionic 13. However, none of the
bromides studied exhibited any conductivity.
tions only becomes a problem if the data is correlated with
other inherent parameters such as Hammett sigma values and
this may explain the variation in the magnitude and/or order
with respect to the various substituents.
Voronkov has measured the IR stretching frequencies of
(aroyloxymethyl)trifluorosilanes and shown that whilst the
medium has a large effect on the frequencies, suggesting a
change in bond orders, non-coordinated species could only be
observed in the gas phase.22,23 Non-coordinated species could
be observed in organic solvents with the mono- and di-fluoro
analogues suggesting that the equilibrium between coordinated
and non-coordinated species becomes important for com-
pounds with low % pentacoordination.24 IR studies on the
corresponding lactams revealed the absence of uncoordinated
species in solution as would be expected from the stronger
coordination, however, the frequency was less susceptible to
changes in the solvent. Nevertheless the warning suggested by
Kummer is worth heeding and all further comparisons have
been carried out at constant temperature, concentration and
solvent.
Assuming 3–8 are discrete compounds, Fig. 2 represents a
solution phase structure correlation for substitution at silicon.
As the extent of reaction increases the nucleophile–silicon bond
is formed at the expense of the leaving group forming a genuine
pentacoordinate species followed by loss of the nucleophile. A
similar picture has emerged in the solid state. However, this
is the first time that such a process has been demonstrated in
solution.
Kummer has recently shown that the 29Si, 13C and 1H NMR
chemical shifts of the 2-pyridone derivatives 3b and 3c are
dependent upon the temperature, solvent and concentration.21
For example, with 3c in CDCl3, the 29Si chemical shift changes
from Ϫ27.3 ppm at 80 ЊC to ϩ3.3 ppm at Ϫ55 ЊC. Similarly, at
25 ЊC, the 29Si chemical shift of 3c is about Ϫ10 ppm in CDCl3
and CD3CN but Ϫ28 ppm in C6D5CN and Ϫ39.2 in C6D6. In
CDCl3 a ten-fold change in the concentration can lead to an 8
ppm shift in the 29Si resonance. This variation in chemical shift
was explained in terms of a change in solvent parameters which
disturbs the balance of 11, 12 and 13 between the various
equilibria.
We were interested in extending the series to include related
2-pyridone structures. We thus prepared the 2-quinolone
series 15a–d using a similar methodology to that employed for
Throughout this study we have carried out all NMR meas-
urements at 25 ЊC and at constant concentration. As a result of
solubility problems we were forced to use a range of solvents
and solvent mixtures. However, the main solvents employed,
CDCl3 and CD3CN, had similar solvent parameters such that
the variation was kept to a minimum. For example, at 25 ЊC,
Kummer has shown that 3b had a 29Si chemical shift of Ϫ39.0
in CDCl3 and Ϫ42.1 in CD3CN, whereas 3b had a 29Si resonance
in CDCl3 of Ϫ11.1 and in CD3CN of Ϫ9.5. Thus any variation
in the chemical shifts measured, and thus the % pentacoordin-
ation or Si–O bond formation, is relatively small and within the
experimental error of the method ( 5%). As evidence for this,
the fluorides 3a–8a were measured in CDCl3, yet the plot of %
pentacoordination against % Si–O bond formation shows a
similar scatter for the fluorides as for the other halogens, even
though these latter compounds were measured in a range of
solvents or solvent mixtures.
Just as specific interactions in the crystal lead to changes in
bond lengths and bond angles so variations of the solvent
parameters will be accompanied by changes in the bond lengths
and bond angles of the pentacoordinate species in solution. For
example, Kummer argues that variation in the acceptor proper-
ties of the solvent will disturb the equilibria through changes in
the coordination of the solvent with the ‘leaving group’. Such
varying coordination will also affect the Si–X bond length and
other geometric parameters. Thus, if variation of the solvent,
concentration and temperature affect the solvent parameters
such that the equilibria are disturbed, these must be accom-
panied by wholesale changes in the structure. If these changes
in structure result in a coherent variation of the 13C and 29Si
chemical shifts, they remain valid models for use in our solution
phase mapping technique. Using a range of measuring condi-
the 2-pyridones. However, the triflate derivative was prepared
by reaction of the chloro derivative 15b with trimethylsily tri-
fluoromethanesulfonate. We have found that pentacoordinate
silicon species Si1–X1 undergo exchange with tetracoordinate
silicon species Si2–X2 to give pentacoordinate Si1–X2 when X1 is
below X2 in the series Cl < Br < TfO. Thus this provides an
easier route to triflate derivatives than the use of chloromethyl-
dimethylsilyl triflate. 8-Hydroxyquinoline has been used as a
bidentate ligand for silicon,25 tin26 and antimony.27 However,
the 2 isomer has only been used as a ligand for rhenium.28
Compounds 16 and 17 were used as model compounds to
anchor the calculation of the % Si–O bond formation. Titration
of 16 with aliquots of trimethylsilyl triflate gave a linear
relationship between the ring 13C chemical shifts and the pro-
portion of trimethylsilyl triflate added. After two equivalents of
the triflate had been added, no further chemical shift change
2102
J. Chem. Soc., Perkin Trans. 2, 1999, 2099–2109

View Online
Table 2 % Si–O bond formation and % pentacoordination for the
2-quinolones 15a–d
In conclusion we have successfully mapped substitution at
silicon in solution using model compounds. We have examined
the strengths and weaknesses of this NMR technique for a
range of 2-pyridone ligands. In future publications we will
explore the applicability of this technique to other ligands
and nucleophilic systems.
Leaving
group
% Si–O bond
formation
% Penta-
coordination
Compound
15a
15b
15c
15d
F
Cl
Br
OTf
35
63
78
93
79
97
79
38
Experimental
Melting points were determined on a Buchi 510 melting point
apparatus and are uncorrected. Infrared spectra were obtained
as Nujol mulls or thin films using sodium chloride plates or as
KBr discs on a Pye Unicam SP1050 or a Nicolet 205 FT-IR
spectrometer. NMR spectra were recorded as solutions in
deuteriochloroform with tetramethylsilane as internal standard
on a JEOL FX 90Q or a JEOL EX 400 NMR spectrometer
(J values are given in Hz and the precision of the chemical shifts
is appropriate for the subsequent calculations). All NMR Mass
spectra were obtained using a Cresta MS 30 instrument or a
VG20-250 quadrupole instrument.
occurred and the NMR data were consistent with 17. For
compounds 15a–d all of the aromatic 13C chemical shifts were
used to calculate the % Si–O bond formation, however, in this
compound the variation of the chemical shift was generally
smaller than that observed with the 2-pyridones. The extent of
Si–O bond formation is given in Table 2. To determine the
extent of pentacoordination, values of 28 and Ϫ40 ppm were
used to anchor the chemical shift scale to a tetrahedral silicon
with no Si–O interaction and a fully pentacoordinate silicon
respectively. O-Trimethylsilylated N-methyl-2-hydroxyquinol-
ine has a 29Si chemical shift of 12.2 ppm and this was used to
anchor the scale to a tetrahedral silicon with 100% Si–O bond
formation. Values for the % pentacoordination of 15a–d are
also given in Table 2. Again, the same ordering of leaving groups
is observed and as the % Si–O bond formation increases,
the extent of pentacoordination first increases to a maximum
and then decreases. The % Si–O bond formation for the 2-
quinolones is consistently higher than those of the correspond-
ing unsubstituted 2-pyridones. This is the result of further
delocalisation of the positive charge on formation of the quino-
linium species. This data overlaps well with that shown in Fig. 1
for the simple 2-pyridones. However, above 50% Si–O bond
formation, the extent of pentacoordination is consistently a
little higher than that obtained with the 2-pyridone system.
This may be a result of differences between the two series in the
relationship between the chemical shift and the extent of Si–O
bond formation. Alternatively, a ‘tighter’ pentacoordinate
species may be formed in the 2-quinolone series where the loss
of the leaving group is less advanced. To extend the series
further we examined the diaza derivatives 18, 19 and 20. The
electron withdrawing effect of an aza group is similar to that of
a nitro group in the same position.29 We might therefore expect
18b and c to have a similar structure to 8b and c respectively.
Compound 18b has a 29Si resonance at Ϫ33.7 ppm similar to
that of 8b at Ϫ25 ppm. Compound 18c has a 29Si resonance at
Ϫ27.6 ppm very close to that of 8c, Ϫ27.9 ppm, confirming the
similarity between the two series. The 29Si chemical shift of 19b,
the corresponding 5 isomer, is Ϫ16.7 ppm suggesting that in
this position the nitrogen has a larger electron withdrawing
influence. The 29Si chemical shift of 20a (Ϫ6.1 ppm) and 20b
(Ϫ27.1 ppm) are both less negative than the corresponding
2-quinolones (15a, Ϫ25.4 ppm and 15b, Ϫ38.3 ppm). This
again suggests that electron withdrawal by the aza group at the
4 position decreases the extent of O–Si bond formation and
pentacoordination. Comparison of the 29Si chemical shifts sug-
gests that electron withdrawal is not as great in 20a and b as in
19a and b. In this instance, the order of O–Si bond formation
reflects that of O protonation.30
Conductivity measurements
Conductivity measurements were carried out using a PTI-10
digital conductivity meter (quoted 0.5%, repeatability
1
digit). All experiments were perfomed under nitrogen. Cali-
bration of the meter was checked by using a standard solution
of potassium chloride. The compound whose conductivity was
to be determined was dissolved in acetonitrile and introduced
into the cell by a syringe. The conductivity of the following
solutions were measured: nBu4NBr (0.23 g, 7 ml, CH3CN, 0.1 M
solution) 182 µS cmϪ1; 3c (X = Br, Y = H) (0.17 g, 7 ml,
CH3CN, 0.1 M solution) 8 µS cmϪ1; 3d (X = OTf, Y = H)
(0.22 g, 7 ml, CH3CN, 0.1 M solution) 29 µS cmϪ1
.
General procedure for the silylation of 2-pyridones with
diethylaminotrimethylsilane13
2-Pyridone (25 mmol) was dissolved in 10 ml of benzene.
Diethylaminotrimethylsilane (25 mmol) was added to the reac-
tion mixture and refluxed for 5 h under nitrogen. The volatile
materials were removed under reduced pressure and the prod-
uct isolated by distillation. The siloxypyridines were used for
the synthesis of silyl-2-pyridones without any characterisation
other than NMR.
The following siloxypyridines were obtained:
2-Trimethylsiloxypyridine. 3.0 g, 72%, bp 37 ЊC/1 mmHg;
δH (90 MHz, CDCl3, Me4Si) 0.35 (9H, s, SiMe3) and 6.6–8.1
(4H, m, arom); δC (22.5 MHz, CDCl3, Me4Si) 0.52 (SiMe3),
112.3 (C3), 116.2 (C 5), 138.3 (C4), 146.7 (C6) and 162.2 (C2);
δSi (17.8 MHz, CDCl3, Me4Si) 19.9.
6-Methyl-2-trimethylsiloxypyridine. 2.9 g, 65%, bp 45 ЊC /0.2
mmHg; δH (90 MHz, CDCl3, Me4Si) 0.34 (9H, s, SiMe3), 2.33
(3H, s, Me) and 6.4–7.4 (3H, m, arom); δC (22.5 MHz, CDCl3,
Me4Si) 0.57 (SiMe3), 24.0 (Me), 109.2 (C3), 115.6 (C5), 139.0
(C4), 156.2 (C6) and 162.2 (C2); δSi (17.8 MHz, CDCl3, Me4Si)
19.3.
We prepared the corresponding disiloxanes to use as model
compounds for 0% Si–O bond formation. However, we were
unable to titrate these compounds with trimethylsilyl triflate to
form the O-trimethylsilylated derivatives, which were to act as
models for 100% Si–O bond formation. In all cases the presence
of the second aza group led to exchange of the trimethylsilyl
group between the oxygen and the nitrogen leading to line
broadening of the ring carbon signals or multiple peaks. With-
out an appropriate anchor for the 13C resonances of the ring
carbons it was not possible to calculate the precise % Si–O bond
formation.
3-Methoxy-2-trimethylsiloxypyridine. 3.9 g, 80%, bp 66 ЊC/
0.5 mmHg; δH (90 MHz, CDCl3, Me4Si) 0.36 (9H, s, SiMe3),
3.75 (3H, s, OMe) and 6.8–7.7 (3H, m, arom); δC (22.5 MHz,
CDCl3, Me4Si) Ϫ0.57 (SiMe3), 55.0 (OMe), 116.6 (C5), 117.8
(C4), 137.0 (C6), 144.4 (C3) and 152.5 (C2); δSi (17.8 MHz,
CDCl3, Me4Si) 20.7.
5-Chloro-2-trimethylsiloxypyridine. 4.2 g, 84%, bp 60 ЊC/2
mmHg; δH (90 MHz, CD3CN, Me4Si) 0.34 (9H, s, SiMe3) and
6.5–8.1 (3H, m, arom); δC (22.5 MHz, CD3CN, Me4Si) Ϫ0.23
J. Chem. Soc., Perkin Trans. 2, 1999, 2099–2109
2103

View Online
(SiMe3), 110.6 (C3), 116.2 (C4), 140.7 (C6), 147.9 (C5) and
161.8 (C2); δSi (17.8 MHz, CD3CN, Me4Si) 21.4.
231/233 (3:1, Mϩ Ϫ Me), 211, 185/187 (3:1, Mϩ Ϫ NO2), 165
(Mϩ Ϫ ClNO2) (Found: C, 39.21; H, 4.52; N, 11.39. C8H11N2-
O3ClSi requires 38.95; H, 4.49; N, 11.35%).
3-Nitro-2-trimethylsiloxypyridine. 4.2 g, 79%, bp 92 ЊC/0.5
mmHg; δH (90 MHz, CD3CN, Me4Si) 0.40 (9H, s, SiMe3) and
7.0–8.3 (3H, m, arom); δC (22.5 MHz, CD3CN, Me4Si) Ϫ0.34
(SiMe3), 116.9 (C5), 134.8 (C4), 142.5 (C3), 151.6 (C6) and
154.7 (C2); δSi (17.8 MHz, CD3CN, Me4Si) 25.5.
1-(Chlorodimethylsilylmethyl)-6-chloro-2-pyridone 5b. 0.47 g,
85%, mp 99–101 ЊC; δH (90 MHz, CDCl3, Me4Si) 0.67 (6H, s,
SiMe2), 3.7 (2H, s, NCH2) and 6.8–7.8 (3H, m, arom); δC (22.5
MHz, CDCl3, Me4Si) 7.3 (SiMe2), 42.3 (SiCH2), 112.5 (C5),
113.1 (C3), 140.5 (C4), 143.0 (C6) and 164.2 (C2); δSi (17.8
MHz, CDCl3, Me4Si) Ϫ40.9; m/z (EI) 235/237/239 (10:6:1,
Mϩ), 220/222/224 (10:6:1, Mϩ Ϫ Me), 202/204 (3:1, Mϩ Ϫ
Cl), 165 (Mϩ Ϫ 2Cl) (Found: C, 40.80; H, 4.63; N, 6.27. C8H1l-
NOCl2Si requires C, 40.69; H, 4.69; N, 5.93%).
6-Chloro-2-trimethylsiloxypyridine. 4.3 g, 85%, bp 67 ЊC/3
mmHg; δH (90 MHz, CDCl3, Me4Si) 0.36 (9H, s, SiMe3) and
6.5–7.6 (3H, m, arom); δC (22.5 MHz, CDCl3, Me4Si) Ϫ0.29,
113.5 (C3), 123.8 (C4), 138.4 (C5), 145.3 (C6) and 160.8 (C2);
δSi (17.8 MHz, CDCl3, Me4Si) 22.4.
General procedure for the synthesis of bromodimethylsilylmethyl-
2-pyridones
General procedure for the synthesis of chlorodimethylsilylmethyl-
2-pyridones
The synthesis was identical to that for the synthesis of
chlorodimethylsilylmethyl-2-pyridones except that chloro-
methyldimethylchlorosilane was replaced by bromomethyl-
dimethylchlorosilane (0.32 g, 2.5 mmol).
The following bromodimethylsilylmethyl-2-pyridones were
obtained:
Chloromethyldimethylchlorosilane, (0.3 g, 2.5 mmol) in dry
diethyl ether was added slowly to a stirred solution of
2-trimethylsiloxypyridine (2.5 mmol) in dry diethyl ether under
nitrogen. The reaction mixture was stirred for 1 h. The solid was
filtered under nitrogen and dried under vacuum.
The following chlorodimethylsilylmethyl-2-pyridones were
obtained:
1-(Bromodimethylsilyl)methyl-2-pyridone 3c. 0.44 g, 86%,
mp 102–104 ЊC; δH (90 MHz, CDCl3, Me4Si) 0.72 (6H, s,
SiMe2), 3.9 (2H, s, NCH2) and 6.9–8.1 (4H, m, arom); δC (22.5
MHz, CDCl3, Me4Si) 1.9 (SiMe2), 44.8 (SiCH2), 115.0 (C3),
115.7 (C5), 141.0 (C4), 146.1 (C6) and 163.1 (C2); δSi (17.8
MHz, CDCl3, Me4Si) Ϫ18.4; m/z (EI) 166 (Mϩ Ϫ Br), 136
(Mϩ Ϫ BrMe2), 106, 78 (Found: C, 38.95; H, 4.76; N, 5.51.
C8H12NOBrSi requires C, 39.01; H, 4.92; N, 5.69%).
1-(Chlorodimethylsilylmethyl)-2-pyridone 3b. 0.43 g, 87%, mp
91–94 ЊC; δH (90 MHz, CDCl3, Me4Si) 0.64 (6H, s, SiMe2), 3.7
(2H, s, NCH2) and 6.8–7.9 (4H, m, arom); δC (22.5 MHz,
CDCl3, Me4Si) 7.5 (SiMe2), 42.2 (SiCH2), 113.6 (C5), 115.8
(C3), 140.0 (C4), 143.3 (C6) and 160.1 (C2); δSi (17.8 MHz,
CDCl3, Me4Si) Ϫ41.1 (Found: C, 47.79; H, 6.06; N, 6.78; Cl,
17.14. C8H12NOClSi requires C, 47.63; H, 6.01; N, 6.94; Cl,
17.57%).
1-(Bromodimethylsilylmethyl)-6-methyl-2-pyridone 6c. 0.54 g,
87%, bp 108–111 ЊC; δH (90 MHz, CDCl3, Me4Si) 0.45 (6H, s,
SiMe2), 2.7 (3H, s, Me), 3.7 (2H, s, NCH2) and 6.8–8.0 (3H, m,
arom); δC (22.5 MHz, CDCl3, Me4Si) 1.7 (SiMe2), 20.7 (Me),
37.9 (SiCH2), 111.5 (C5), 115.2 (C3), 144.8 (C6), 151.6 (C4) and
163.7 (C2); δSi (17.8 MHz, CDCl3, Me4Si) Ϫ21.3 (Found: C,
41.67; H, 5.37; N, 5.40. C9H14NOBrSi requires C, 41.54; H,
5.42; N, 5.38%).
1-(Chlorodimethylsilylmethyl)-6-methyl-2-pyridone 6b. 0.40 g,
84%, mp 99–104 ЊC; δH (90 MHz, CDCl3, Me4Si) 0.53 (6H, s,
SiMe2), 2.7 (3H, s, Me) 3.7 (2H, s, NCH2) and 6.8–8.0 (3H, m,
arom); δC (22.5 MHz, CDCl3, Me4Si) 1.21 (SiMe2), 20.6 (Me),
37.8 (SiCH2), 111.3 (C5), 115.6 C3), 144.8 (C6), 151.0 (C4) and
163.4 (C2); δSi (17.8 MHz, CDCl3, Me4Si) Ϫ24.7 (Found: C,
49.90; H, 6.46; N, 6.42; Cl, 16.50. C9H14NOClSi requires C,
50.10; H, 6.54; N, 6.49; Cl, 16.43%).
1-(Bromodimethylsilylmethyl)-3-methoxy-2-pyridone 7c. 0.45
g, 90%, mp 105–108 ЊC; δH (90 MHz, CDCl3, Me4Si) 0.73 (6H,
s, SiMe2), 4.0 (3H, s, OMe), 4.2 (2H, s, NCH2) and 7.1–7.9 (3H,
m, arom); δC (22.5 MHz, CDCl3, Me4Si) Ϫ0.63 (SiMe2), 44.9
(Me), 58.0 (SiCH2), 116.1 (C5), 122.1 (C3), 132.1 (C6), 147.3
(C4) and 154.9 (C2); δSi (17.8 MHz, CDCl3, Me4Si) Ϫ14.3.
1-(Chlorodimethylsilylmethyl)-3-methoxy-2-pyridone 7b. 0.31
g, 88%, mp 96–99 ЊC; δH (90 MHz, CD3CN, Me4Si) 0.62 (6H,
s, SiMe2), 3.7 (2H, s, NCH2), 3.9 (3H, s, OMe) and 6.7–7.6 (3H,
m, arom); δC (22.5 MHz, CD3CN, Me4Si) 6.4 (SiMe2), 43.6
(SiCH2), 56.6 (OMe), 110.9 (C5), 118.0 (C3), 129.5 (C6), 147.2
(C4) and 157.6 (C2); δSi (17.8 MHz, CDCl3, Me4Si) Ϫ32.0; m/z
(EI) 231/233 (3:1, Mϩ), 216/218 (3:1, Mϩ Ϫ Me), 200/202 (3:1,
Mϩ Ϫ OMe), 196 (Mϩ Ϫ Cl), 181 (Mϩ Ϫ MeCl).
1-(Bromodimethylsilylmethyl)-5-chloro-2-pyridone 4c. 0.65 g,
88%, mp 107–112 ЊC; δH (90 MHz, CDCl3, Me4Si) 0.44 (6H, s,
SiMe2), 4.4 (2H, s, NCH2) and 7.4–8.6 (3H, m, arom); δC (22.5
MHz, CDCl3, Me4Si) Ϫ0.34 (SiMe2), 42.9 (SiCH2), 115.8 (C5),
122.0 (C3), 138.9 (C6), 145.4 (C4) and 160.3 (C2); δSi (17.8
MHz, CDCl3, Me4Si) Ϫ12.7; m/z (EI) 279/281/283 (3:4:1, Mϩ),
264/266/268 (3:4:1, Mϩ Ϫ Me), 200/202 (3:1, Mϩ Ϫ Br), 186/
188 (3:1, Mϩ Ϫ MeBr) (Found: C, 34.00; H, 4.09; N, 4.99.
C8H11NOClBrSi requires C, 34.24; H, 3.95; N, 4.99%).
1-(Chlorodimethylsilylmethyl)-5-chloro-2-pyridone 4b. 0.56 g,
88%, mp 99–106 ЊC; δH (90 MHz, CD3OD, Me4Si) 0.46 (6H, s,
SiMe2), 4.0 (2H, s, NCH2) and 7.2–8.4 (3H, m, arom); δC (22.5
MHz, CD3OD, Me4Si) 1.2 (SiMe2), 43.0 (SiCH2), 116.2 (C5),
120.8 (C3), 138.2 (C6), 144.7 (C4) and 160.9 (C2); δSi (17.8
MHz, CD3OD, Me4Si) Ϫ17.1; m/z (EI) 235/237/239 (10:6:1,
Mϩ), 220/222/224 (10:6:1, Mϩ Ϫ Me), 200/202 (3:1, Mϩ Ϫ Cl),
165 (Mϩ Ϫ 2Cl) (Found: C, 40.54; H, 4.72; N, 5.92 C8H11-
NOCl2Si requires C, 40.69; H, 4.69; N, 5.93%).
1-(Bromodimethylsilylmethyl)-3-nitro-2-pyridone 8c. 0.69 g,
90%, mp 110–112 ЊC; δH (90 MHz, CD3CN, Me4Si) 0.76 (6H, s,
SiMe2), 3.9 (2H, s, NCH2) and 7.1–8.7 (3H, m, arom); δC (22.5
MHz, CD3CN, Me4Si) 1.5 (SiMe2), 45.0 (SiCH2), 111.2 (C5),
135.7 (C3), 141.4 (C6), 146.5 (C4) and 157.(C2); δSi (17.8 MHz,
CD3CN, Me4Si) Ϫ27.9; m/z (EI) 292/290 (1:1, Mϩ), 211 (Mϩ Ϫ
Br), 181 (Mϩ Ϫ BrMe2).
1-(Chlorodimethylsilylmethyl)-3-nitro-2-pyridone 8b. 0.51 g,
77%, mp 96–100 ЊC; δH (90 MHz, CD3CN, Me4Si) 0.50 (6H, s,
SiMe2), 3.7 (2H, s, NCH2) and 6.7–8.4 (3H, m, arom); δC (22.5
MHz, CD3CN, Me4Si) 0.57 (SiMe2), 43.8 (SiCH2), 105.9 (C5),
136.7 (C3), 140.1 (C6), 146.4 (C4) and 156.6 (C2); δSi (17.8
MHz, CD3CN, Me4Si) Ϫ25.0; m/z (EI) 246/248 (3:1, Mϩ),
1-(Bromodimethylsilylmethyl)-6-chloro-2-pyridone 5c. 0.60 g,
88%, mp 105–109 ЊC; δH (90 MHz, CDCl3, Me4Si) 0.55 (6H, s,
SiMe2), 3.9 (2H, s, NCH2) and 6.8–8.1 (3H, m, arom); δC (22.5
MHz, CDCl3, Me4Si) 1.7 (SiMe), 40.l (SiCH2), 113.0 (C5),
2104
J. Chem. Soc., Perkin Trans. 2, 1999, 2099–2109

View Online
114.7 (C3), 141.3 (C4), 144.7 (C6) and 163.8 (C2); δSi (17.8
MHz, CDCl3, Me4Si) Ϫ27.4; m/z (EI) 279/281/283 (3:4:1, Mϩ),
264/266/268 (3:4:1), 200/202 (3:1), 186/188 (3:1) (Found: C,
34.58; H, 4.22; N, 4.99. C8H11NOClBrSi requires C, 34.24; H,
3.95; N, 4.99%).
SiMe2), 3.2 (2H, s, NCH2) and 6.6–7.6 (3H, m, arom); δC (22.5
2 2
MHz, CDCl3, Me4Si) 1.9 (d, JCSiF 19.4, SiMe), 38.0 (d, JCSiF
46.6 Hz, SiCH2), 109.8 (C5), 114.3 (C3), 140.4 (C6), 141.0 (C4)
1
and 164.1 (C2); δSi (17.8 MHz, CDCl3, Me4Si) Ϫ26.0 (d, JSiF
257.8); m/z (EI) 219/221 (3:1, Mϩ), 218/220 (3:1, Mϩ Ϫ H),
204/206 (3:1, Mϩ Ϫ Me), 184 (Mϩ Ϫ FMe) (Found: C, 43.10;
H, 5.00; N, 6.21. C8H11NOClFSi requires C, 43.73; H, 5.05; N,
6.37%).
General procedure for the synthesis of fluorodimethylsilylmethyl-
2-pyridones
Chloromodimethylsilylmethyl-2-pyridone (5 mmol) was dis-
solved (or suspended) in 5 ml dry benzene under nitrogen.
Antimony trifluoride (0.30 g, 1.7 mmol) was added and the
reaction mixture was stirred for 0.5 h. The reaction mixture was
then diluted with excess water and extracted with chloroform
(3 × 75 ml). The washed extract was dried over anhydrous
magnesium sulfate and the solvent removed using a rotary
evaporator to obtain a colourless crystalline solid, which was
dried under vacuum.
General procedure for the synthesis of trifluoromethylsulfonyl-
oxydimethylsilylmethyl-2-pyridone
The synthesis was identical to that for the synthesis of chloro-
dimethylsilylmethyl-2-pyridones except that chloromethyl-
dimethylchlorosilane was replaced by chloromethyldimethyl-
silyl triflate (0.62 g, 2.5 mmol).
Chloromethyldimethylsilyl triflate. Trifluoromethanesulfonic
acid (5 ml, 37.8 mmol) was added to 3.34 ml (37.8 mmol)
chloromethyldimethylchlorosilane, with stirring under nitro-
gen. The reaction mixture was heated at 60 ЊC for 5 h. The
product was isolated by distillation. (8.6 g, 89%), bp 62 ЊC/8
mmHg; δH (90 MHz, CDCl3, Me4Si) 0.60 (6H, s, SiMe2) and 3.0
(2H, s, CH2); δC (22.5 MHz, CDCl3, Me4Si) Ϫ3.0 (SiMe2), 27.6
The following fluorodimethylsilylmethyl-2-pyridones were
obtained:
1-(Fluorodimethylsilylmethyl)-2-pyridone 3a. 0.57 g, 92%, mp
3
82–86 ЊC; δH (90 MHz, CDCl3, Me4Si) 0.32 (6H, d, JHSiF 6.8,
SiMe2), 3.2 (2H, s, NCH2) and 6.5–7.7 (4H, m, arom); δC (22.5
2
2
1
MHz, CDCl3, Me4Si) 1.8 (d, JCSiF 25, SiMe), 39.5 (d, JCSiF
44.0, SiCH2), 109.2 (C5), 116.8 (C3), 139.1 (C4), 141.4 (C6) and
163.2 (C2); δSi (17.8 MHz, CDCl3, Me4Si) Ϫ22.3 (d, 1JSiF 256.8);
m/z (EI) 185 (Mϩ), 184, 170, 166 (Found: C, 51.60; H, 6.51; N,
7.45. C8H12NOFSi requires C, 51.86; H, 6.53; N, 7.56%).
(SiCH2) and 118.9 (q, JCF 317.1, CF3); δSi (17.8 MHz, CDCl3,
Me4Si) 31.7; m/z (EI) 256/258 (3:1, Mϩ), 207 (Mϩ Ϫ MeCl),
191 (Mϩ Ϫ ClMe2) (Found: C, 18.95; H, 3.13. C4H8O3C1F3SSi
requires C, 18.71; H, 3.15%).
The following dimethylsilylmethyl-2-pyridone triflates were
obtained:
1-(Fluorodimethylsilylmethyl)-6-methyl-2-pyridone 6a. 0.61 g,
91%, mp 87–89 ЊC; δH (90 MHz, CDCl3, Me4Si) 0.28 (6H, d,
3JHSiF 5.9, SiMe2), 2.5 (3H, s, Me), 3.0 (2H, s, NCH2) and 6.5–
1-(Trifluoromethylsulfonyloxydimethylsilylmethyl)-2-pyridone
3d. 0.63 g, 88%, mp 120–123 ЊC; δH (90 MHz, CD3CN, Me4Si)
0.66 (6H, s, SiMe2), 4.0 (2H, s, NCH2) and 7.2–8.3 (4H, m,
arom); δC (22.5 MHz, CD3CN, Me4Si) 0.17 (SiMe2), 42.8
(SiCH2), 115.2 (C5), 118.3 (C3), 142.3 (C4), 147.5 (C6) and 162.9
(C2); δSi (17.8 MHz, CD3CN, Me4Si) 32.3.
2
7.5 (3H, m, arom); δC (22.5 MHz, CDCl3, Me4Si) 2.4 (d, JCSiF
27.2, SiMe), 20.9, 35.8 (d, 2JCSiF 50.5, SiCH2), 110.2 (C5), 112.5
(C3), 141.2 (C6), 148.6 (C4) and 163.8 (C2); δSi (17.8 MHz,
CDCl3, Me4Si) Ϫ35.5 (d, 1JSiF 252.9) (Found: C, 54.26; H, 7.11;
N, 6.96. C9H14NOFSi requires C, 54.24; H, 7.08; N, 7.03%).
1-(Trifluoromethylsulfonyloxydimethylsilylmethyl)-6-methyl-2-
pyridone 6d. 0.69 g, 90%, mp 128–132 ЊC; δH (90 MHz, CDCl3,
Me4Si) 0.64 (6H, s, SiMe2), 2.6 (3H, s, Me), 3.8 (2H, s, NCH2)
and 7.0–8.1 (3H, m, arom); δC (22.5 MHz, CDCl3, Me4Si) 3.2
(SiMe2), 20.5 (Me), 39.8 (SiCH2), 111.5 (C5), 117.9 (C3), 121.0
1-(Fluorodimethylsilylmethyl)-3-methoxy-2-pyridone 7a. 0.66
g, 92%, mp 88–92 ЊC; δH (90 MHz, CDCl3, Me4Si) 0.35 (6H, d,
3JHSiF 5.1, SiMe2), 3.3 (2H, s, NCH2), 3.9 (3H, s, OMe) and 6.2–
2
6.4 (3H, m, arom); δC (22.5 MHz, CDCl3, Me4Si) 1.4 (d, JCSiF
2
1
23.3, SiMe), 40.2 (d, JCSiF 40.2, SiCH2), 56.1 (OMe), 107.7
(q, JCF 321.0, CF3), 146.1 (C6), 152.7 (C4) and 163.0 (C2);
(C5), 114.6 (C3), 129.7 (C6), 148.3 (C4) and 158.3 (C2); δSi (17.8
MHz, CDCl3, Me4Si) Ϫ13.5 (d, JSiF 258.8) (Found: C, 50.02;
δSi (17.8 MHz, CDCl3, Me4Si) 23.0 (Found: C, 36.52; H, 4.48;
N, 4.80. C10H14NSF3O4Si requires C, 36.46; H, 4.28; N, 4.25%).
1
H, 6.64; N, 6.39. C9H14NO2FSi requires C, 50.21; H, 6.55; N,
6.51%).
1-(Trifluoromethylsulfonyloxydimethylsilylmethyl)-3-
methoxy-2-pyridone 7d. 1.1 g, 85%, mp 127–130 ЊC; δH (90
MHz, CDCl3, Me4Si) 0.69 (6H, s, SiMe2), 3.9 (3H, s, OMe), 4.0
(2H, s, NCH2) and 7.1–7.9 (3H, m, arom); δC (22.5 MHz,
CDCl3, Me4Si) 3.1 (SiMe2), 42.9 (SiCH2), 57.2 (OMe), 116.9
1-(Fluorodimethylsilylmethyl)-5-chloro-2-pyridone 4a. 0.66 g,
90%, mp 91–95 ЊC; δH (90 MHz, CDCl3, Me4Si) 0.31 (6H, d,
3JCSiF 7.8, SiMe2), 3.2 (2H, s, NCH2) and 6.7–7.7 (3H, m, arom);
2
1
δC (22.5 MHz, CDCl3, Me4Si) 1.2 (d, JCSiF 19.4, SiMe), 40.4
(C5), 120.3 (q, JCF 319.2, CF3), 123.5 (C3), 131.3 (C6), 146.5
(d, 2JCSiF 41.4, SiCH2), 115.0 (C5), 118.0 (C3), 137.0 (C6), 142.1
(C4) and 161.9 (C2); δSi (17.8 MHz, CDCl3, Me4Si) Ϫ14.6 (d,
1JSiF 259.8); m/z (EI) 219/221 (3:1, Mϩ), 218/220 (3:1, Mϩ Ϫ
H), 204/206 (3:1, Mϩ Ϫ Me), 184 (Mϩ Ϫ MeF) (Found: C,
43.26; H, 5.00; N, 6.07. C8H11NOFClSi requires C, 43.73; H,
5.05; N, 6.37%).
(C4) and 155.8 (C2); δSi (17.8 MHz, CDCl3, Me4Si) 21.3
(Found: C, 34.92; H, 3.92; N, 4.10. C10H14NSF3O5Si requires C,
34.78; H, 4.09; N, 4.06%).
1-(Trifluoromethylsulfonyloxydimethylsilylmethyl)-5-chloro-
2-pyridone 4d. 0.74 g, 88%, mp 136–139 ЊC; δH (90 MHz,
CDCl3, Me4Si) 0.67 (6H, s, SiMe2), 4.0 (2H, s, NCH2) and 7.1–
8.3 (3H, m, arom); δC (22.5 MHz, CDCl3, Me4Si) 1.7 (SiMe2),
41.5 (SiCH2), 115.8 (C5), 118.8 (q, 1JCF 319.7, CF3), 123.9 (C3),
139.6 (C6), 147.0 (C4) and 61.8 (C2); δSi (17.8 MHz, CDCl3,
Me4Si) 25.2; m/z (EI) 349/351 (3:1, Mϩ), 334/336 (3:1,
Mϩ Ϫ Me), 200/202 (Mϩ Ϫ CF3SO3) (Found: C, 31.02; H, 3.32;
N, 4.27. C9H11O4F3CINSSi requires C, 30.90; H, 3.17; N,
4.00%)
1-(Fluorodimethylsilylmethyl)-3-nitro-2-pyridone 8a. 0.69 g,
90%, mp 87–90 ЊC; δH (90 MHz, CDCl3, Me4Si) 0.38 (6H, d,
3JCSiF 7.8, SiMe2), 3.6 (2H, s, NCH2) and 6.4–8.4 (3H, m, arom);
δC (22.5 MHz, CDCl3, Me4Si) Ϫ0.52 (d, 2JCSiF 18.1, SiMe), 42.7
2
(d, JCSiF 29.8, SiCH2), 105.1 (C5), 137.4 (C3), 145.9 (C4) and
1
155.6 (C2); δSi (17.8 MHz, CDCl3, Me4Si) 12.5 (d, JSiF 272.5)
(Found: C, 42.01; H, 4.95: N, 11.83. C8H11N2O3FSi requires
C, 41.72;H, 4.82; N, 12.17%).
1-(Trifluoromethylsulfonyloxydimethylsilylmethyl)-3-nitro-2-
pyridone 8d. 0.80 g, 89%, mp 129–131 ЊC; δH (90 MHz, CDCl3,
Me4Si) 0.66 (6H, s, SiMe2), 4.0 (2H, s, NCH2) and 7.2–8.8 (3H,
1-(Fluorodimethylsilylmethyl)-6-chloro-2-pyridone 5a. 0.68 g,
92%, mp 90–95 ЊC; δH (90 MHz, CDCl3, Me4Si) 0.31 (6H, br s,
J. Chem. Soc., Perkin Trans. 2, 1999, 2099–2109
2105

View Online
m, arom); δC (22.5 MHz, CDCl3, Me4Si) 3.2 (SiMe2), 42.1
(SiCH2), 114.8 (C5), 120.3 (q, JCF 319.4, CF3), 135.5 (C3),
General procedure for the synthesis of 1,1,3,3-tetramethyl-1,3-
bis(2-trimethylsiloxypyridiniomethyl)disiloxane ditriflate
1
142.4 (C6), 147.4 (C4) and 158.0 (C2); δSi (17.8 MHz, CDCl3,
Me4Si) 4.5 (Found: C, 29.96; H, 3.64; N, 8.86. C9H11OF3N2SSi
requires C, 30.0; H, 3.08; N, 7.78%).
The corresponding 1,1,3,3-tetramethyl-1,3-bis(2-oxo-1,2-di-
hydro-1-pyridylmethyl)disiloxane (2.0 mmol) was dissolved in
2.0 ml of deuterochloroform in a 5 mm NMR tube. Aliquots
of trimethylsilyl triflate were added and the NMR spectra
obtained. The NMR spectrum did not significantly change
after addition of 1 equiv. of trimethylsilyl triflate.
1-(Trifluoromethylsulfonyloxydimethylsilylmethyl)-6-chloro-
2-pyridone 5d. 1.3 g, 92%, mp 135–140 ЊC; δH (90 MHz, CDCl3,
Me4Si) 0.65 (6H, s, SiMe2), 3.9 (2H, s, NCH2) and 7.2–8.0
(3H, m, arom); δC (22.5 MHz, CDCl3, Me4Si) 1.7 (SiMe2),
The following disiloxanes were obtained:
1
41.8 (CH2), 112.8 (C5), 116.8 (C3), 120.3 (q, JCF 319.7, CF3),
1,1,3,3-Tetramethyl-1,3-bis(2-trimethylsiloxypyridinio-
methyl)disiloxane ditriflate. δH (90 MHz, CDCl3, Me4Si) 0.21
(6H, s, SiMe2) 4.0 (2H, s, NCH2) and 7.2–8.3 (4H, m, arom); δC
(22.5, CDCl3, Me4Si) 0.40 (SiMe2), 42.5, (SiCH2), 115.1 (C5),
117.8 (C3), 119.0 (1JCF 317.0, CF3), 141.5 (C4), 146.9 (C6) and
162.5 (C2); δSi (17.8 MHz, CDCl3, Me4Si) 2.5.
142.8 (C6) 146.4, (C4) and 165.0 (C2); δSi (17.8 MHz, CDCl3,
Me4Si) 10.7; m/z (EI) 349/351 (3:1, Mϩ), 334/336 (3:1,
Mϩ Ϫ Me) and 200/202 (3:1, Mϩ Ϫ Tf) (Found: C, 31.56; H,
3.21; N, 4.49. C9H11O4F3ClNSSi requires C, 30.90; H, 3.17; N,
4.00%).
General procedure for the synthesis of 1,1,3,3-tetramethyl-1,3-
bis(2-oxo-1,2-dihydro-1-pyridylmethyl)disiloxane
1,1,3,3-Tetramethyl-1,3-bis(6-methyl-2-trimethylsiloxy-
pyridiniomethyl)disiloxane ditriflate. δH (90 MHz, CDCl3, Me4Si)
0.20 (6H, s, SiMe2), 2.6 (3H, s, Me), 3.8 (2H, s, NCH2) and 7.0–
7.9 (3H, m, arom); δC (22.5 MHz, CDCl3, Me4Si) 2.00 (SiMe2),
20.8 (Me), 39.9, (SiCH2), 111.6 (C5), 116.9 (C3), 119 (1JCF
318.1, CF3), 145.1 (C6), 151.3 (C4) and 162.7 (C2); δSi (17.8
MHz, CDCl3, Me4Si) 7.4.
The corresponding chlorodimethylsilylmethyl-2-pyridone (0.5
g) was added to 1 ml of distilled water. The reaction mixture
was stirred for 1 h and then extracted with chloroform (3 × 75
ml). The washed extract was dried over anhydrous magnesium
sulfate and distilled under reduced pressure and dried under
vacuum.
The following disiloxanes were obtained:
1,1,3,3-Tetramethyl-1,3-bis(3-methoxy-2-trimethylsiloxy-
pyridiniomethyl)disiloxane ditriflate. δH (90 MHz, CDCl3,
Me4Si) 0.20 (6H, s, SiMe2), 4.0 (2H, s, NCH2), 4.0 (3H, s, OMe)
and 7.2–7.7 (3H, m, arom); δC (22.5 MHz, CDCl3, Me4Si) 0.98
(SiMe2), 42.8, (SiCH2), 57.2 (OMe), 116.6 (C5), 119.5 (1JCF
318.4, CF3), 122.7 (C3), 130.8 (C6), 146.6 (C4) and 156.0 (C2);
δSi (17.8 MHz, CDCl3, Me4Si) 9.5.
1,1,3,3-Tetramethyl-1,3-bis(2-oxo-1,2-dihydro-1-pyridyl-
methyl)disiloxane. δH (90 MHz, CDCl3, Me4Si) 0.16 (6H, s,
SiMe2) 3.4 (2H, s, NCH2) and 6.0–7.2 (4H, m, arom); δC (22.5
MHz, CDCl3, Me4Si) 0.63 (SiMe2), 42.4 (SiCH2), 105.6 (C5),
119.7 (C3), 138.3 (C4), 138.6 (C6) and 162.2 (C2); δSi (17.8
MHz, CDCl3, Me4Si) 4.2.
1,1,3,3-Tetramethyl-1,3-bis(5-chloro-2-trimethylsiloxy-
pyridiniomethyl)disiloxane ditriflate. δH (90 MHz, CDCl3,
Me4Si) 0.10 (6H, s, SiMe2), 4.1 (2H, s, NCH2) and 7.1–8.2 (3H,
m, arom); δC (22.5 MHz, CDCl3, Me4Si) 0.80 (SiMe2), 42.9,
(SiCH2), 115.5 (C5), 119.0 (1JCF 318.0, CF3), 123.5 (C3), 139.0
(C6), 146.4 (C4) and 161.6 (C2); δSi (17.8 MHz, CDCl3, Me4Si)
7.2.
1,1,3,3-Tetramethyl-1,3-bis(6-methyl-2-oxo-1,2-dihydro-1-
pyridylmethyl)disiloxane. δH (90 MHz, CDCl3, Me4Si) 0.29
(6H, s, SiMe2), 2.5 (3H, s, Me), 3.5 (2H, s, NCH2) and 6.3
Ϫ7.4 (3H, m, arom); δC (22.5 MHz, CDCl3, Me4Si) 2.20
(SiMe2), 21.4 (Me), 38.2, (SiCH2), 109.6 (C5), 114.2 (C3),
139.9 (C6), 147.5 (C4) and 163.2 (C2); δSi (17.8 MHz, CDCl3,
Me4Si) 4.2.
1,1,3,3-Tetramethyl-1,3-bis(3-nitro-2-trimethylsiloxy-
pyridiniomethyl)disiloxane ditriflate. δH (90 MHz, CDCl3,
Me4Si) 0.10 (6H, s, SiMe2), 3.9 (2H, s, NCH2) and 7.2–8.8 (3H,
m, arom); δC (22.5 MHz, CDCl3, Me4Si) 42.4, (SiCH2), 114.3
(C5), 120.0 (1JCF 318.4, CF3), 135.8 (C3), 142.4 (C6), 146.9 (C4)
and 158.1 (C2); δSi (17.8 MHz, CDCl3, Me4Si) 2.6.
1,1,3,3-Tetramethyl-1,3-bis(3-methoxy-2-oxo-1,2-dihydro-1-
pyridylmethyl)disiloxane. δH (90 MHz, CDCl3, Me4Si) 0.20
(6H, s, SiMe2), 3.6 (2H, s, NCH2) 3.8 (3H, s, OMe) and 6.1–
6.9 (3H, m, arom); δC (22.5 MHz, CDCl3, Me4Si) 0.40
(SiMe2), 55.8 (OMe), 42.7 (SiCH2), 105.1 (C5), 112.2 (C3),
129.3 (C6), 149.7 (C4) and 157.8 (C2); δSi (17.8 MHz, CDCl3,
Me4Si) 3.4.
1,1,3,3-Tetramethyl-1,3-bis(6-chloro-2-trimethylsiloxy-
pyridiniomethyl)disiloxane ditriflate. δH (90 MHz, CDCl3,
Me4Si) 0.10 (6H, s, SiMe2), 3.9 (2H, s, NCH2) and 7.1–8.1 (3H,
m, arom); δC (22.5 MHz, CDCl3, Me4Si) 1.61 (SiMe2), 41.5,
(SiCH2), 113.3 (C5), 116.3 (C3), 119.0 (1JCF 318.0, CF3), 141.8
(C6), 145.8 (C4) and 161.0 (C2); δSi (17.8 MHz, CDCl3, Me4Si)
7.2.
1,1,3,3-Tetramethyl-1,3-bis(5-chloro-2-oxo-1,2-dihydro-1-
pyridylmethyl)disiloxane. δH (90 MHz, CDCl3, Me4Si) 0.20 (6H,
s, SiMe2), 3.8 (2H, s, NCH2) and 6.2–7.2 (3H, m, arom); δC (22.5
MHz, CDCl3, Me4Si) 0.51 (SiMe2), 43.1, (SiCH2), 112.4 (C5),
120.5 (C3), 136.0 (C6), 139.9 (C4) and 160.9 (C2); δSi (17.8
MHz, CDCl3, Me4Si) 3.5.
1-(Fluorodimethylsilylmethyl)-2-quinolone 15a
1,1,3,3-Tetramethyl-1,3-bis(3-nitro-2-oxo-1,2-dihydro-1-
pyridylmethyl)disiloxane. δH (90 MHz, CDCl3, Me4Si) 0.15 (6H,
s, SiMe2), 3.7 (2H, s, NCH2) and 6.3–8.3 (3H, m, arom); δC (22.5
MHz, CDCl3, Me4Si) 0.57 (SiMe2), 44.5, (SiCH2), 103.7 (C5),
138.2 (C3), 138.2 (C6), 146.1 (C4) and 154.9 (C2); δSi (17.8
MHz, CDCl3, Me4Si) 5.0.
1-(Chlorodimethylsilylmethyl)-2-quinolone (15b) (0.75 g, 2.98
mmol) was dissolved in benzene (4 cm3) under a nitrogen
atmosphere. Granular antimony trifluoride (0.18 g, 1.01 mmol)
was introduced and the reagents stirred together for 1 hour at
room temperature. The organic phase was decanted from the
oily antimony-containing by-products and hydrolysed by
shaking with distilled water (50 cm3). The organic phase was
extracted into dichloromethane (3 × 50 cm3). The extracts were
combined and dried over magnesium sulfate and the solvents
removed by rotary evaporation. The remaining solid was
recrystallised from chloroform to give the pure product (0.40 g,
57%); δH (400 MHz, CDCl3, Me4Si) 0.38 [6H, d, J 6.4,
1,1,3,3-Tetramethyl-1,3-bis(6-chloro-2-oxo-1,2-dihydro-1-
pyridylmethyl)disiloxane. δH (90 MHz, CDCl3, Me4Si) 0.20 (6H,
s, SiMe2), 3.8 (2H, s, NCH2) and 6.4–7.2 (3H, m, arom); δC (22.5
MHz, CDCl3, Me4Si) 0.92 (SiMe2), 39.8, (SiCH2), 106.9 (C5),
117.0 (C3), 137.9 (C6), 137.9 (C4) and 162.8 (C2); δSi (17.8
MHz, CDCl3, Me4Si) 3.5.
2106
J. Chem. Soc., Perkin Trans. 2, 1999, 2099–2109

View Online
Si(CH3)2], 3.18 (2H, s, N-CH2), 6.82 (1H, d, J 9.4, H3), 7.38
(1H, m, H6), 7.73–7.64 (3H, m, H5/H7/H8) and 7.95 (1H, d,
J 9.4, H4); δC (100 MHz, CDCl3, Me4Si) 2.0 [d, J 23.8,
Si(CH3)2], 33.9 (d, J 47.5, N-CH2), 115.9 (C8), 117.1 (C3), 121.4
(C4a), 123.6 (C6), 129.0 (C5), 131.8 (C7), 139.3 (C8a), 141.4
(C4) and 162.9 (C2); δF (90 MHz, CDCl3) Ϫ114.2; δSi (79 MHz,
CDCl3, Me4Si) 25.4 (d, J 254.3); m/z 235 (Mϩ, 35%), 234
(Mϩ Ϫ H, 100), 220 (Mϩ Ϫ CH3, 99) (Found: C, 60.83; H, 5.98;
N, 5.86. C12H14FNOSi requires C, 61.25; H, 6.00; N, 5.95%).
42.24; H, 3.99; N, 3.77. C13H14F3NO4Si requires C, 42.73; H,
3.86; N, 3.83%).
1,1,3,3-Tetramethyl-1,3-bis(2-oxo-1,2-dihydro-1-quinolyl-
methyl)disiloxane 16
1-(Chlorodimethylsilylmethyl)-2-quinolone 15b (4.54 g, 15.33
mmol) was dissolved in acetone (20 cm3) with stirring. Distilled
water (20 cm3) was added and stirring continued for a further 16
hours at room temperature. The solution was extracted with
chloroform (3 × 50 cm3) and the aqueous phase discarded. The
extracts were combined and dried over magnesium sulfate.
Removal of the solvent initially by rotary evaporation and
latterly by high vacuum pumping afforded 16 as a colourless,
highly viscous oil (5.2 g, 76%); δH(400 MHz, CDCl3, Me4Si)
0.13 [12H, s, Si(CH3)2], 3.66 (4H, s, N-CR2), 6.72 (2H, d,
J 9.2, H3), 7.20 (2H, m, H6), 7.39 (2H, m, H8), 7.57 (4H, m,
H5/H7) and 7.66 (2H, d, J 9.2, H4); δC (100 MHz, CDCl3,
Me4Si) 1.1 [Si(CH3)2], 35.4 (N-CR2), 115.3 (C8), 120.8 (C4a),
120.9 (C3), 121.8 (C6), 128.7 (C5), 130.3 (C7), 138.4 (C4), 139.9
(C8a) and 161.8 (C2); δSi (79 MHz, CDCl3, Me4Si) 1.5; m/z (EI)
433 (Mϩ Ϫ CH3, 3%), 290 (Mϩ Ϫ C9H6NO, 100), 216 (Mϩ Ϫ
C11H12NOSi, 97) [Found: Mϩ Ϫ CH3, 433.1404. C23H25N2O3Si2
requires 433.1404; Found Mϩ Ϫ C9H6NO, 290.1033 (EI).
C14H20NO2Si2 requires 290.1033].
1-(Chlorodimethylsilylmethyl)-2-quinolone 15b
A solution of 2-(trimethylsiloxy)quinoline31 (0.65 g, 2.99
mmol) was prepared in benzene (4 cm3) under a nitrogen
atmosphere. Chloro(chloromethyl)dimethylsilane (0.44 g, 2.66
mmol) was added drop-wise with stirring and the solution
allowed to stand for 30 minutes at room temperature. After
removal of the solvent the product was recrystallised from
acetonitrile to give colourless needle-like crystals (0.67 g, 89%);
δH (400 MHz, CDCl3, Me4Si) 0.75 [6H, s, Si(CH3)2], 3.65 (2H, s,
N-CH2), 6.93 (1H, d, J 9.2, H3), 7.49 (1H, m, H6), 7.83–7.77
(3H, m, H5/H7/H8) and 8.09 (1H, d, J 9.2, H4); δC (100 MHz,
CDCl3, Me4Si) 7.4 [Si(CH3)2], 38.6 (N-CH2), 115.4 (C3), 116.5
(C8), 121.9 (C4a), 124.8 (C6), 129.3 (C5), 132.7 (C7) , 138.1
(C8a), 143.1 (C4) and 163.4 (C2); δSi (79 MHz, CDCl3, Me4Si)
Ϫ38.4 ; m/z 253 (Mϩ, 13), 252 (Mϩ Ϫ H, 40), 251 (Mϩ, 37), 250
(Mϩ Ϫ H, 100), 238 (Mϩ Ϫ CH3, 31), 236 (Mϩ Ϫ CH3, 76)
(Found: C, 57.15; H, 5.86; N, 5.35. C12H14ClNOSi requires C,
57.24; H, 5.60; N, 5.56%).
1,1,3,3-Tetramethyl-1,3-bis(2-trimethylsiloxyquinoliniomethyl)-
disiloxane ditriflate 17
In a 10 mm NMR tube capped by a rubber septum was
prepared a solution of 16 (0.45 g, 1.00 mmol) in deuterio-
choroform (2.5 cm3). To the solution was added, by syringe,
trimethylsilyl triflate (0.45 g, 2.02 mmol) and the reagents
shaken briefly together. Compound 17 could only be observed
in solution and is characterised by its NMR spectra; δH (400
MHz, CDCl3, Me4Si) 0.08 [18H, s, Si(CH3)3], 0.73 [12H, s,
Si(CH3)2], 3.96 (4H, s, N-CH2), 7.22 (2H, d, J 9.4, H3), 7.67
(2H, m, H6), 7.88 (2H, m, H8), 7.95 (2H, m, H7), 8.03 (2H, m,
H5) and 8.56 (2H, d, J 9.4, H4); δC (100 MHz, CDCl3, Me4Si)
1.6 [Si(CH3)2], 2.0 [Si(CH3)3], 37.1 (N-CR2), 113.7 (C3), 117.1
1-(Bromodimethylsilylmethyl)-2-quinolone 15c
A solution of 2-(trimethylsiloxy)quinoline31 (0.65 g, 2.99
mmol) was prepared in benzene (5 cm3) under a nitrogen
atmosphere. (Bromomethyl)chlorodimethylsilane (0.56 g, 2.99
mmol) was added drop-wise with stirring and the solution
allowed to stand for 30 minutes at room temperature. After
removal of the solvent the product was recrystallised from
acetonitrile to give colourless cubic crystals (0.74 g, 84%);
δH (400 MHz, CDCl3, Me4Si) 0.81 [6H, s, Si(CH3)2], 3.94 (2H, s,
N-CH2), 7.04 (1H, d, J 8.5, H3), 7.55 (1H, m, H6), 7.91–7.84
(3H, m, H5/H7/H8) and 8.22 (1H, d, J 8.5, H4); δC (100 MHz,
CDCl3, Me4Si) 6.1 [Si(CH3)2], 39.0 (N-CH2), 114.6 (C3), 117.1
(C8), 122.3 (C4a), 125.5 (C6), 129.4 (C5), 133.3 (C7), 137.8
(C8a), 144.1 (C4) and 163.4 (C2); δSi (79 MHz, CDCl3, Me4Si)
Ϫ28.9. m/z 297 (Mϩ, 4%), 296 (Mϩ Ϫ H, 9), 295 (Mϩ, 4), 294
(Mϩ Ϫ H, 9), 282 (Mϩ Ϫ CH3, 1), 280 (Mϩ Ϫ CH3, 6), 216
(Mϩ Ϫ Br, 100), 128 [Mϩ Ϫ CH2Si(CH3)2Br, 31] (Found: C,
49.37; H, 4.88; N, 4.79. C12H14BrNOSi requires C, 48.65; H,
4.76; N, 4.73%).
1
(C8), 120.9 (q, JCF 320.0, SO3CF3), 123.2 (C4a), 126.5 (C6),
130.1 (C5), 133.9 (C7), 137.6 (C8a), 146.3 (C4) and 163.3 (C2);
δF (90 MHz, CDCl3) Ϫ80.0; δSi (79 MHz, CDCl3, Me4Si) Ϫ0.8
[Si(CH3)2] and 11.7 [Si(CH3)3].
2-(Trimethylsiloxy)pyrimidine
Dried 2-hydroxypyrimidine hydrochloride (3.0 g, 23 mmol) was
suspended in dry benzene (20 ml) under nitrogen. N,N-Diethyl-
trimethylsilylamine (4.6 g, 32 mmol) was then added and the
mixture stirred under reflux for three hours, during which time
the 2-hydroxypyrimidine hydrochloride dissolved and a solid
(diethylamine hydrochloride) sublimed into the condenser.
After cooling, the solvent was removed under vacuum, and the
product purified by vacuum distillation to give 2-(trimethyl-
silyloxy)pyrimidine as a yellow liquid in a yield of 3.3 g (85%);
νmax(neat)/cmϪ1 850, 939, 1253, 1350, 1560, 1429 and 2960;
δH (400 MHz, CDCl3, Me4Si) 0.39 [9H, s, Si(CH3)3], 6.90 (1H, t,
J 4.8, H4) and 8.47 (2H, d, J 4.8, H3, H5); δC (100 MHz,
CDCl3, Me4Si) 0.0 (SiMe3), 114.8 (C5), 159.2 (C4 and 6) and
163.7 (C2); δSi (79 MHz, CDCl3, Me4Si) 23.2.
1-(Trifluoromethylsulfonyloxydimethylsilylmethyl)-2-
quinolone 15d
To a solution of 15b (0.37 g, 1.47 mmol) under a nitrogen
atmosphere, was added trimethylsilyl triflate (0.34 g, 1.53
mmol) by syringe, drop-wise, with stirring. After 30 minutes the
solvent was removed by distillation at atmospheric pressure
under nitrogen followed by pumping at high vacuum. The crude
product was washed with diethyl ether (2 × 5 cm3) and dried
under high vacuum for 2 hours. Recrystallisation from benzene
gave colourless cubic crystals (0.48 g, 89%); δH (400 MHz,
CDCl3, Me4Si) 0.69 [6H, s, Si(CH3)2], 3.88 (2H, s, N-CH2), 7.11
(lH, d, J 9.2, H3), 7.64 (1H, m, H6), 7.95–7.88 (3H, m, H5/H7/
H8) and 8.36 (lH, d, J 9.2, H4); δC (100 MHz, CDCl3, Me4Si)
2.7 [Si(CH3)2], 37.2 (N-CH2), 113.8 (C3), 117.2 (C8), 119.8 (q,
1JCF 316.4, SO3CF3), 122.8 (C4a), 126.4 (C6), 129.7 (C5), 133.9
(C7), 137.7 (C8a), 145.5 (C4) and 163.4 (C2); δF (90 MHz,
CDCl3) Ϫ79.9; δSi (79 MHz, CDCl3, Me4Si) Ϫ8.1; m/z 365 (Mϩ,
20%), 364 (Mϩ Ϫ H, 27), 350 (Mϩ Ϫ CH3, 18), 216 (Mϩ Ϫ
SO3CF3, 100), 128 [Mϩ Ϫ CH2Si(CH3)2SO3CF3, 49] (Found: C,
N-(Chlorodimethylsilylmethyl)-2-pyrimidone 18b
2-(Trimethylsiloxy)pyrimidine was dissolved in dry diethyl ether
under nitrogen. Chloro(chloromethyl)dimethylsilane was then
added, the flask agitated briefly to mix the reactants, and then
left for three hours. After this time a yellow precipitate had
formed. The solvent was removed and the precipitate washed
with further dry diethyl ether before being dried under vacuum.
νmax(neat)/cmϪ1 836, 1564, 1634 and 1743; δH (400 MHz, CDCl3,
Me4Si) 0.62 [6H, s, Si(CH3)2], 3.65 (2H, s, NCH2), 6.73 (1H, dd,
J. Chem. Soc., Perkin Trans. 2, 1999, 2099–2109
2107

View Online
J 6.5, 4.0, H5), 8.14 (1H, dd, J 6.5, 2.4, H4) and 8.60 (1H, dd,
J 4.0, 2.4, H6); δC (100 MHz, CDCl3, Me4Si) 7.4 (SiMe2), 44.9
(SiCH2), 109.8 (C5), 149.6 (C4), 160.9 (C6) and 167.3 (C2); δSi
(79 MHz, CDCl3, Me4Si) Ϫ33.7; m/z (EI) 204, 202 (Mϩ), 189,
187 (Mϩ Ϫ Me), 164 (Mϩ Ϫ Cl) (Found: Mϩ, 202.033. C7H11-
N2OSiCl requires 202.033).
(1H, m, H7), 7.96 (1H, d, 3JH5H6 8.0, H5) and 8.40 (1H, s, H3);
2
δC (100 MHz, CDCl3, Me4Si) 1.3 [d, JCF 22.1, Si(CH3)2], 33.4
(d, 2JCF 40.5, N-CH2), 115.1 (C8), 124.8 (C6), 130.6 (C5), 131.9
(C7), 132.9 (C4a), 134.4 (C8a), 146.9 (C3) and 155.5 (C2);
δF (90 MHz, CDCl3, Me4Si) Ϫ128.7; δSi (79 MHz, CDCl3,
Me4Si) Ϫ6.1 (d, 1JSiF 260.3); m/z (EI) 236 (Mϩ), 235 (Mϩ Ϫ H),
221, 158, 77 (Found C, 56.40; H, 5.49; N, 12.30. C11H13FN2OSi
requires C, 55.91; H, 5.54; N, 11.85%).
N-(Bromodimethylsilylmethyl)-2-pyrimidone 18c
2-(Trimethylsiloxy)pyrimidine was dissolved in dry diethyl
ether. (Bromomethyl)chlorodimethylsilane was then added and
the flask briefly agitated to mix the reactants. After standing
for two hours, a white precipitate had formed. The solvent was
removed and the precipitate dried under vacuum; νmax(neat)/
cmϪ1 839, 1101, 1215, 1583, 1734 and 2960; δH (400 MHz,
CDCl3, Me4Si) 0.84 [6H, s, Si(CH3)2], 4.02 (2H, s, NCH2), 7.20
(1H, dd, 1H, J 6.3, 4.7, H5), 8.61 (1H, dd, J 6.3, 2.4, H4) and
8.89 (1H, dd, J 4.7, 2.4, H6); δC (100 MHz, CDCl3, Me4Si) 7.5
(SiMe2), 45.4 (SiCH2), 111.9 (C5), 151.3 (C4), 161.2 (C6) and
167.5 (C2); δSi (79 MHz, CDCl3, Me4Si) Ϫ27.6 [Found:
Mϩ Ϫ Me, 167.0641 (EI). C7H11N2OSi requires 167.0641].
1-(Chlorodimethylsilylmethyl)quinoxalin-2-one 20b
Chloro(chloromethyl)dimethylsilane (0.56 g, 3.90 mmol) was
added drop-wise with stirring to a solution of 2-(trimethyl-
siloxy)quinoxaline (0.85 g, 3.89 mmol) in 10 ml of benzene
under a nitrogen atmosphere. The solution was allowed to
stand for 30 minutes at room temperature. The solvent was
removed by distillation under nitrogen and latterly by high vac-
uum pumping. By washing the resulting brown solid residue
with diethyl ether (2 × 2 cm3) and drying under high vacuum
for 1 hour, a creamy white powder was obtained (0.80 g, 81%);
δH (400 MHz, CDCl3, Me4Si) 0.76 [6H, s, Si(CH3)2], 3.54 (2H, s,
N-CH2), 7.58 (1H, ddd, 3JH6-H5 8.4, 3JH6-H7 7.6, 4JH6-H8 1.1, H6),
7.67 (1H, dd, 3JH8-H7 8.8, 4JH8-H6 1.1, H8), 7.80 (1H, ddd, 3JH7-H8
8.8, 3JH7-H6 7.6, 4JH7-H5 1.6, H7), 8.05 (1H, dd, 3JH5-H6 8.4, 4JH5-H7
1.6, H5) and 8.53 (1H, s, H3); δC (100 MHz, CDCl3, Me4Si) 6.9
[Si(CH3)2], 37.3 (N-CH2), 115.7 (C8), 126.0 (C6), 130.9 (C5),
131.5 (C4a), 132.7 (C7), 135.1 (C8a), 144.3 (C3) and 155.8
(C2); δSi (79 MHz, CDCl3, Me4Si) Ϫ27.1; m/z (EI) 252/254 (3:1)
(Mϩ) , 251/253 (3:1) (Mϩ Ϫ H) 237/239 (3:1) (Mϩ Ϫ Me), 217,
129 [Found: Mϩ, 252.0486 (EI). C11H1335ClN2OSi requires
252.0486].
3-(Fluorodimethylsilylmethyl)-4-pyrimidone 19a
As a result of competing desilylation, the proportion of
antimony trifluoride was reduced to approximately 0.25 equiv-
alents. Thus, using a similar procedure to that above, 3-(chloro-
dimethylsilylmethyl)-4-pyrimidone 19b (4.95 g, 24.43 mmol)
was suspended in 20 cm3 benzene and reacted with antimony
trifluoride (1.09 g, 6.10 mmol). This gave 3-(fluorodimethylsilyl-
methyl)-4-pyrimidone as a pale yellow oil (1.9 g, 42%); δH (400
MHz, CDCl3, Me4Si) 0.37 [6H, d, 3JHF 7.6, Si(CH3)2], 3.32 (2H,
3
3
s, N-CH2), 6.51 (1H, d, JH5H6 7.2, H5), 7.98 (1H, d, JH6H5
7.2, H6), 8.27 (1H, s, H2); δC (100 MHz, CDCl3, Me4Si)
Ϫ0.2 [d, 2JCF 18.3, Si(CH3)2], 37.3 (d, 2JCF 32.9, N-CH2), 114.0
(C5), 151.8 (C2), 154.6 (C6) and 162.3 (C4); δF (90 MHz,
CDCl3, Me4Si) Ϫ141.6; δSi (79 MHz, CDCl3, Me4Si) 8.8 (d, 1JSiF
269.0) [Found: MHϩ, 187.0703 (FAB). C7H12FN2OSi requires
187.0703].
References
1 (a) D. Britton and J. D. Dunitz, J. Am. Chem. Soc., 1981, 103, 2971;
(b) H. B. Bürgi and J. D. Dunitz, Acc. Chem. Res., 1983, 16, 153.
2 A. J. Kirby, Adv. Phys. Org. Chem., 1994, 29, 87.
3 M. J. Barrow, E. A. V. Ebsworth and M. M. Harding, J. Chem. Soc.,
Dalton Trans., 1980, 1838.
4 A. A. Macharashvili, V. E. Shklover, Yu. T. Struchkov, G. I.
Oleneva, E. P. Kramarova, A. G. Shipov and Yu. I Baukov, J. Chem.
Soc., Chem. Commun., 1988, 683.
5 V. F. Sidorkin, V. V. Vladimirov, M. G. Voronkov and V. A.
Pestunovich, THEOCHEM, 1991, 228, 1.
6 Yu. E. Ovchinnikov, A. A. Macharashvili, Yu. T. Struchkov,
A. G. Shipov and Yu. I. Baukov, J. Struct. Chem., 1994, 35, 91.
7 A. R. Bassindale and M. Borbaruah, J. Chem. Soc., Chem.
Commun., 1991, 1499.
3-(Chlorodimethylsilylmethyl)-4-pyrimidone 19b
Using a similar procedure to that above, chloro(chloromethyl)-
dimethylsilane (1.48 g, 10.35 mmol) was reacted with 4-(tri-
methylsiloxy)pyrimidine (1.74 g, 10.34 mmol) in 5 ml of
benzene. This gave 3-(chlorodimethylsilylmethyl)-4-pyrimidone
as a white crystalline solid (1.8 g, 86%) δH (400 MHz, CDCl3,
Me4Si) 0.68 [6H, s, Si(CH3)2], 3.53 (2H, s, N-CH2), 6.63 (1H,
8 A. R. Bassindale and M. Borbaruah, J. Chem. Soc., Chem.
Commun., 1991, 1501.
3
3
d, JH5H6 6.4, H5), 8.13 (1H, d, JH6H5 6.4, H6) and 8.38 (1H,
s, H2); δC (100 MHz, CDCl3, Me4Si) 5.7 [Si(CH3)2], 39.4
(N-CH2), 112.7 (C5) 150.9 (C2), 156.0 (C6) and 163.3
(C4); δSi (79 MHz, CDCl3, Me4Si) Ϫ16.7; m/z (EI) 202/204
(3:1) (Mϩ) (Found: Mϩ, 202.0329. C7H1135ClN2OSi requires
202.0329), 167 (Mϩ Ϫ Cl) (Found: Mϩ Ϫ Cl, 167.0641. C7H11-
N2OSi requires 167.0641).
9 C. J. Jameson and J. Mason, in The Chemical Shift, ed. J. Mason,
Plenum Press, New York, 1987.
10 Y. Takeuchi and T. Takayama, in The Chemistry of Organosilicon
Compounds Volume 2, ed. Z. Rappoport and Y. Apeloig, Wiley, New
York, 1998, ch. 6, pp. 267–354.
11 A. R. Grimmer, F. Von-Lampe and M. Magi, Chem. Phys. Lett.,
1986, 132, 549.
12 A. I. Albanov, L. I. Gubanova, M. F. Larin, V. A. Pestunovich and
M. G. Voronkov, J. Organomet. Chem., 1983, 244, 5.
13 M. J. Buchanan, R. H. Cragg and A Steltner, J. Organomet. Chem.,
1976, 120, 189.
1-(Fluorodimethylsilylmethyl)quinoxalin-2-one 20a
1-(Chlorodimethylsilylmethyl)quinoxalin-2-one 20b (11.02 g,
43.60 mmol) was suspended in benzene (40 ml) under a nitrogen
atmosphere. Granular antimony trifluoride (2.60 g, 14.55
mmol) was introduced and the reagents stirred together for 1
hour at room temperature. The organic phase was decanted
from the oily antimony-containing by-products and hydrolysed
by shaking with distilled water (50 cm3). The organic phase was
extracted into dichloromethane (3 × 50 cm3). The extracts were
combined and dried over magnesium sulfate and the solvents
removed by rotary evaporation. The residue was recrystallised
from acetonitrile to give 1-(fluorodimethylsilylmethyl)quin-
oxalin-2-one as a white crystalline solid (5.4 g, 53%); δH (400
MHz, CDCl3, Me4Si) 0.43 [6H, d, 3JHF 7.8, Si(CH3)2], 3.23 (2H,
s, N-CH2), 7.47 (1H, m, H6), 7.56 (1H, d, 3JH8H7 8.4, H8), 7.71
14 R. W. Hillyard, C. M. Ryan and C. H. Yoder, J. Organomet. Chem.,
1978, 153, 369.
15 H. Marsmann, in NMR 17; Oxygen-17 and Silicon-29, ed. P. Diehl,
E. Fluck and R. Kosfeld, Springer-Verlag, Berlin, 1981, ch. 2,
pp. 65–235.
16 A. Fratiello, G. A. Vidulich, V. K. Anderson, M. Kazazian,
C. S. Stover and H. Sabounjian, J. Chem. Soc., Perkin Trans. 2, 1983,
475.
17 A. R. Bassindale, S. J. Glynn and P. G. Taylor, in The Chemistry
of Organosilicon Compounds Volume 2, ed. Z. Rappoport and
Y. Apeloig, Wiley, New York, 1998, ch. 9, pp. 495–511.
18 D. Kost and I. Kalikhman, in The Chemistry of Organosilicon
Compounds Volume 2, ed. Z. Rappoport and Y. Apeloig, Wiley,
New York, 1998, ch. 23, pp. 1339–1446.
19 K. D. Onan, A. T. McPhail, C. H. Yoder and R. W. Hillyard Jr.,
J. Chem. Soc., Chem. Commun., 1978, 209.
2108
J. Chem. Soc., Perkin Trans. 2, 1999, 2099–2109

View Online
27 V. K. Jain, J. Mason and R. C. Mehrotra, J. Organomet. Chem.,
20 A. R. Bassindale and T. Stout, J. Chem. Soc., Chem. Commun.,
1984, 1388.
1986, 309, 45.
21 D. Kummer and S. H. Abdel Halim, Z. Anorg. Allg. Chem., 1996,
622, 57.
28 M. Leeaphon, K. Rohl, R. J. Thomas, P. E. Fanwick and R. A.
Walton, Inorg. Chem., 1993, 32, 5562.
22 M. G. Voronkov, Yu. L. Frolov, V. M. D’Yakov, N. N. Chipanina,
I. I. Gubanova, G. A. Gavrilova, L. V. Klyba and T. N.
Aksamenova, J. Organomet. Chem., 1980, 201, 165.
23 A. I. Albanov, L. I. Gubanova, M. F. Larin, V. A. Pestunovich and
M. G. Voronkov, J. Organomet. Chem., 1983, 244, 5.
24 Yu. L. Frolov and M. G. Voronkov, J. Mol. Struct., 1990, 217, 265.
25 G. Klebe, K. Hensen and J. Vonjouanne, J. Organomet. Chem., 1983,
258, 137; A. I. Albanov, V. A. Petunovich, O. M. Trofimova,
N. F. Chernov and M. G. Voronkov, Z. Obshch. Khim., 1996, 66,
1949.
29 O. Exner, in Correlation Analysis in Chemistry, ed. N. B. Chapman
and J. Shorter, Plenum Press, New York, 1978, ch. 10, pp. 439–540.
30 D. D. Perrin, Dissociation Constants of Organic Bases in Aqueous
Solution, Butterworths, London, 1965; V. A. Palm, Tables of rate
and equilibrium constants of heterolytic organic reactions, Moscow,
1976, vol. 2.
31 E. Lukevics, I. Segal, I. Birgele and A. Zablotskaya, Khim.
Geterolsikl. Soedin, 1994, 9, 1255.
26 V. K. Jain, J. Mason., B. S. Saraswat and R. C. Mehrotra, Poly-
hedron, 1985, 4, 2089.
Paper 9/04402G
J. Chem. Soc., Perkin Trans. 2, 1999, 2099–2109
2109