Aggregative activation in heterocyclic chemistry. Part 4. Metallation of 2-methoxypyridine: Unusual behaviour of the new unimetal superbase BuLi-Me2N(CH2)2OLi (BuLi-LiDMAE)

Article abstract of DOI:10.1039/a701914i

A series of potential unimetal superbases BuLi-ROLi has been studied in order to increase the basicity/ nucleophilicity ratio ([B/N]R) of BuLi. The best [B/N]R ratio is found with BuLi-LiDMAE. This complex base apparently metallates 2-methoxypyridine at t

Full text of DOI:10.1039/a701914i

View Article Online / Journal Homepage / Table of Contents for this issue
Aggregative activation in heterocyclic chemistry. Part 4. Metallation
of 2-methoxypyridine: unusual behaviour of the new unimetal
superbase BuLi–Me2N(CH2)2OLi (BuLi–LiDMAE)
Philippe Gros, Yves Fort and Paul Caubère*
Laboratoire de Chimie Organique I, URA CNRS 457, Faculté des Sciences,
Université Henri-Poincaré, Nancy I, BP 239, F-54506 Vandoeuvre-les-Nancy, France
A series of potential unimetal superbases BuLi–ROLi has been studied in order to increase the basicity/
nucleophilicity ratio ([B/N]R) of BuLi. The best [B/N]R ratio is found with BuLi–LiDMAE. This
complex base apparently metallates 2-methoxypyridine at the unexpected C-6 position. It is shown that no
actual metallated species are formed in the reaction medium, the reaction occurring as the result of a
common radical precursor stabilized by an aggregate cluster. Finally, as an application, C-6 substituted
2-methoxypyridines have been obtained in good to excellent yields.
Introduction
The structure of lithium reagents has attracted, and still
N
OMe
attracts, considerable attention.1–4 Numerous very elegant
works 4,5 have definitively established that the intricate nature of
such derivatives results from equilibria between aggregates and
sometimes the corresponding monomer. The structure of the
aggregates and the equilibrium positions depend on the lithium
reagent, its concentration, the solvating property of the solvent
and temperature. All this shows that it is not easy to establish
the identity of the actual reactive species during a reaction
between a lithium reagent and a substrate. Further, little is gen-
erally known about the influence of the substrate undergoing
reaction on the reactivity of the starting lithium reagent.5,7
Finally the very intimate mechanism of such reactions is not yet
well elucidated. It thus appears that the study of the chemical
behaviour of aggregated lithium reagents might provide some
helpful information in the understanding of the properties of
lithium derivatives. As part of our studies on aggregative acti-
vation6,7 we undertook the investigations dealing with the gen-
eration of unimetal superbases (USB)7 from BuLi and lithium
alkoxides. Indeed our work on sodium amide-containing com-
plex bases showed that sodium alkoxides were very efficient in
the generation of a series of NaNH2–RONa reagents with
specific properties.6–8 We thought that analogous activation
might take place with lithium reagents although differences
between NaNH2–RONa and RLi–R1OLi had to be expected in
the structure and in the optimal amounts of the alkoxides lead-
ing to the most efficient bases. From a synthetic point of view a
large number of alcohols are commercially available and their
lithium salts may be generated in situ very easily. In a short
communication we reported our first observations about
unexpected reactions between a number of super bases BuLi–
ROLi (complex bases) and 2-methoxypyridine.9 It appeared
that under our conditions, unusual functionalization of the
substrate took place at the C-6 position instead of the C-3 as
might have been expected from the directed ortho metallation
(DOM) principles.3
1
BunLi–ROLi
Hexane, 0 °C
Li
Li
H
+
+
Bun
N
OMe
Li
N
OMe
N
OMe
H
3a
2a
Me3SiCl
4a
Me3SiCl
Me3SiCl
SiMe3
H
SiMe3
+
+
Bun
N
OMe
N
OMe
Me3Si
N
OMe
H
4b
3b
2'a
H2O
H
H
Bun
N
OMe
H
2b
Scheme 1
Results
Influence of ROLi on the properties of BuLi-ROLi USBs
Use of Me3SiCl as a trapping agent showed that the most
general reaction taking place during these metallations was that
illustrated in Scheme 1. Addition product 2b is the only product
usually obtained10 from the reaction of BuLi with 1 while 3b is
currently prepared after metallation of 1 with lithium dialkyl-
amides.11 To the best of our knowledge 4b has never been
obtained from 1 by direct proton abstraction so we focused our
attention on the formation of the corresponding lithium deriva-
tive 4a. First we carried out an exploratory study in order to
understand the behaviour of the new complex bases BuLi–
Here we report our investigations aimed at producing the
most efficient activating agent ROLi and use of the correspond-
ing new complex bases in the synthesis of 6-substituted 2-
methoxypyridines. We also discuss some new observations
on the reactivity of aggregates supporting our hypothesis con-
cerning their propensity to promote single electron transfer
(SET).6–7
E-Mail: caubere@lco1.u-nancy.fr Fax: (33) 03 83 40 45 58
J. Chem. Soc., Perkin Trans. 1, 1997
3071

View Article Online
Table 1 Metallation of 1 with BuLi–ROLia
Complex base
Yields (%)b
Runs
BuLi (equiv)
Activating agent (equiv)
Rec. 1
2b
3b
4b
1
2
3
4
5
6
7
8
9
10
11
12
13
14
1
1
1
1
1
1
1
1
2
4
4
4
4
4
none
0
0
100
100
0
28
16
26
1
3
13
5
100
100
83
40
0
0
0
0
31
0
17
23
0
0
0
0
0
0
0
0
42
0
ButOLi (1)
Et(OCH2CH2)2OLi (1)
Me2N(CH2)2OLi (1)
TMEDA (1)
100
30
53
66
30
39
27
1
0
0
5
3
EtOCH2CH2OLi (1)
Me2N(CH2)2OLi (2)
Me2N(CH2)2OLi (3)
Me2N(CH2)2OLi (2)
Me2N(CH2)2OLi (4)
Pri2N(CH2)2OLi (4)
Me2N(CH2)3OLi (4)
Me2NCH2CHMeOLi (4)
Me2NCHMeCHPhOLi (4)
7
52
35
60
94
0
0
10
56
0
15
16
17
4
4
4
LiOCH2CH(CH2)3CH2NMe (4)
CH2(CH2)2CH2N(CH2)2OLi (4)
CH2(CH2)3CH2N(CH2)2OLi (4)
3
2
2
32
10
27
0
0
0
64
86
67
a Metallations were performed in a constant volume of hexane (40 ml) at 0 ЊC for 1 h; condensations were performed with 1.5 equiv. of Me3SiCl
(based on BuLi) in 5 ml of hexane at 0 ЊC for 1 h. b Yields determined by GC capillary analysis.
ROLi. The main results are reported in Table 1. Important
differences were soon noted between heterogeneous12 NaNH2–
RONa and homogeneous BuLi–ROLi complex bases with
respect to the structure of the activating agents. While ButONa
and Et(OCH2CH2)2ONa were among the best activating agents
of NaNH2,8 their corresponding lithium salts were found to
be completely inefficient. In fact with the former salt, 2b
was obtained as the only product (run 2) and with the latter
the starting material was quantitatively recovered (run 3). In
contrast, while the sodium salt of dimethylaminoethanol
(NaDMAE) only moderately activated NaNH2,8c lithium 2-
dimethylaminoethanolate (LiDMAE) efficiently increased the
[B/N]R ratio of BuLi (runs 4 and 7–10) and, under appropriate
conditions, strongly favoured the formation of 4b (runs 9, 10).
Thus, we first hypothesised that the properties of BuLi were
influenced by the simultaneous presence of dimethylamino and
lithium alkoxide groups in the activating agent. This hypothesis
was supported (run 5) by the reaction performed with
N,N,NЈ,NЈ-tetramethylethylenediamine (TMEDA) in which a
N,N-dimethylamino group replaced the alkoxide one. This clas-
sical complexing agent3b,4d,11b,13 led only to small amounts of 3b
and 2b but to no 4b. Interestingly, besides 2b EtOCH2CH2OLi
(run 6) led to a small amount of 4b. This result and the fact that
dilithium glycolate (unreported results) was inefficient as an
activating agent supported the conclusion that the structure of
any activating agent must simultaneously contain an alkoxide
group and a lone-paired neutral atom. With regard to the ratio
BuLi/activating agent, differences with NaNH2/RONa were
also observed. Indeed, with sodium amide-containing complex
bases, a ratio of NaNH2/RONa = 2 induced the best activation
whatever the reactions performed.6–8 With BuLi–ROLi, and as
far as the selective formation of 4b was concerned, a ratio of
BuLi/ROLi = 1 led to the best results (runs 4, 9, 10). This ratio
was adopted in the following investigations particularly because
complex bases with a ratio of BuLi/ROLi < 1 led to the com-
petitive formation of 3b (runs 7, 8), an observation which will
deserve further investigation. An excess of lithiating reagent
considerably increased the yield of 4b (runs 9, 10). This result
parallels the literature data according to which, for some
unknown reason, an excess of lithiating reagents is usually
necessary to metallate nitrogen-containing heterocycles.14 It is
interesting to note that under our conditions, the presence of a
large excess of BuLi–ROLi led only to very low yields of add-
ition product 2b (runs 9, 10), the best yields of 4b being
obtained when the ratio of 1/BuLi/ROLi = 1/4/4 was used (run
10). According to this observation, we retained this ratio in
investigations on the effect of the structure of aminoalkoxides
(see Table 1; runs 11–17).
Comparison of runs 10 and 11 shows that bulky substituents
on the nitrogen of the ethanolamine alkoxide completely
suppressed the formation of aggregates appropriate to the C-6
proton abstraction from 1. This behaviour may be due to a
strong steric hindrance which impeded intermolecular inter-
actions together with the classical B strain15 which flattens
nitrogen and decreases its cation complexation ability. Our
results completely agree with the well established fact4,5,7 that
stereoelectronic factors play an important part in the form-
ation and properties of aggregates. This was also illustrated by
the inefficiency of lithium 3-dimethylaminopropanolate to
increase the [B/N]R ratio of BuLi (run 12). From the other
data it appeared that with one exception (run 13) lithium
salts of aminoethanols with available nitrogen lone pairs consti-
tuted good activating agents within the framework of the
reaction performed. Interestingly, lithium 2-(pyrrolidinyl)-
ethanolate (run 16) was only slightly less efficient than LiD-
MAE. Without other physicochemical information we cannot
further discuss the data reported and particularly the part
played by the substituents of alkoxides such as those used in
runs 13 or 14. The above reported study encouraged us to sub-
sequently continue our investigations with the most efficient
BuLi–LiDMAE base.
Influence of LiBr on the properties of BuLi-LiDMAE
From a synthetic point of view we were concerned with the
serious drawbacks that could be brought about by the excess of
base necessary to obtain 4a in excellent yields. On the other
hand, it is well known that addition of some additives to
lithium reagents changes the structure of their aggregates and
thus their properties.4–5 After short exploratory experiments, we
selected LiBr, an additive well known for its ability to interact
with lithium reagent aggregates.5a,16 From an unreported sys-
tematic study, a practical general trend emerged as far as the
condensation of Me3SiCl was concerned. (i) In the presence of
3 or 4 equiv. of complex base, addition of LiBr was deleterious.
(ii) In contrast, on avoidance of a large excess of base, it was
found that in the presence of 1 or 2 equiv. of complex base,
addition of LiBr substantially improved the formation of 4b.
However, the best results (see Table 2, runs 1 and 2) were
obtained with 2 equiv. of base. Since we have no information
about the structure of the aggregates we are unable to discuss
these results in detail. Moreover, the empirically determined
optimal amount of added LiBr, cannot be of help since this salt
3072
J. Chem. Soc., Perkin Trans. 1, 1997

View Article Online
Table 2 Metallation of 1 with BuLi–LiDMAE and trapping with electrophiles: influence of LiBr and electrophilesa
R
R′
3) H2O
H
H
Bun
N
OMe
Bun
N
OMe
(1) n [BunLi–LiDMAE]
H
H
Additive; hexane; 0 °C; 1 h
2
(2) E; 0 °C; 1 h
b, c R′ = H
d R′ = D
N
OMe
1
R
N
OMe
4
b R = SiMe3
c R = SMe
d R = D
Additive
LiBr
(equiv.)
Yields (%)b
E
Runs
n
(equiv)
Rec. 1
2b–d
4b–d
1
2
3
4
5
6
2
2
2
2
2
2
—
0.25
—
0.25
—
0.25
Me3SiCl (2)
Me3SiCl (2)
MeSSMe (2.5)
MeSSMe (2.5)
DCl/D2O (10)
DCl/D2O (10)
27
5
20
10
5
13
15
15
15
61
53
60
75
63
72
34c
41c
6
a Metallations were performed in a constant volume of hexane (40 ml) at 0 ЊC for 1 h; Me3SiCl or MeSSMe were added in hexane (5 ml). DCl–D2O
were added without solvent. All electrophiles were allowed to react at 0 ЊC for 1 h. b Yields determined by GC capillary analysis. c Deuteriation yields
were determined by 1H NMR spectroscopy.
is sparingly soluble in the reaction medium. Before exploring
the synthetic potential of the above results we wanted to check
that the yield of 4b reflected the yield of 4a in the reaction
medium. Indeed it might be thought that aggregates containing
4a or a precursor of 4a were formed in low concentration
during an equilibrated reaction and that addition of Me3SiCl
displaced the equilibrium towards 4b. This possibility was
investigated in runs 3 to 6 performed with 2 equiv. of base. The
results obtained with MeSSMe (runs 3, 4) completely agreed
with those obtained with Me3SiCl. The results obtained with
DCl–D2O (runs 5, 6) were puzzling. Indeed a much larger
amount of 2a was trapped with DCl–D2O than with Me3SiCl or
MeSSMe. Since the metallations were performed under the
same conditions, it must be concluded that 2a was formed in
50–60 % yields before the trapping with Me3SiCl or MeSSMe
whereas at the end of the condensations, the yields of 2b and
2c never exceeded 15% (runs 1–4). A possible explanation
could be in the reversible addition of BuLi to 1 associated with
a displacement towards 2a or 4a during the addition of the
electrophile. We discarded this hypothesis for the following
reasons. Since the formation of 4a should be irreversible due
to the formation of butane, longer reaction time and/or higher
temperature should have led to a considerable increase in the
formation of 4a, which was never observed. Moreover, an
increase in the complex base concentration corresponded to a
substantial increase in the ratio 4a:2a (see Table 1, runs 9 and
10) when decreased or at least unchanged ratios were expected.
Last but not least, no butane evolution was observed during the
metallation step. We shall see later that there may be another
explanation for this surprising result.
could be obtained and that the simultaneous addition of THF
with the carbonyl substrate could improve the reaction yields.
Moreover, from unreported results, a number of useful observ-
ations emerge. (i) Only low to fair yields of 5 were obtained
when metallations were performed with a stoichiometric
amount of base. However these reactions evidenced that LiBr
and/or THF improved the yield of 5 to the detriment of 2b. (ii)
While the use of 4 equiv. of base had led to excellent conden-
sation yields with Me3SiCl (see Table 1) low to fair yields were
obtained with carbonyl electrophiles which had to be used in
large excess (5 equiv.). Curiously, under these conditions, it was
also observed that LiBr had no effect. It even negated the effect
of THF which alone strongly favoured the formation of 5. (iii)
Finally, the reactions performed in the presence of 2 equiv. of
base appeared to be the most interesting (see Table 3). Under
such conditions, as illustrated with acetone, addition of LiBr or
THF considerably favoured the formation of 5 (runs 2, 3) and
their effects were additive (run 4).
The same behaviour was generally found with other represen-
tative carbonyl derivatives (runs 5–12) and under appropriate
conditions we were able to obtain the alcohols 5 in good to
very good yields. In addition to its influence on the aggregates,
it is likely that, thanks to a classical electrophilic assistance,16
LiBr promotes the condensations of 4a with the carbonyl
derivatives. On the other hand, it appeared that under the
same metallation conditions the yield of 2a deduced from the
yields of 2b varied with the nature of the condensed carbonyl
derivative! This observation must be related to that made
when we compared the behaviour of Me3SiCl or MeSSMe with
DCl–D2O (vide supra). Since the reversibility of the formation
of 2a is unlikely, the only rational hypothesis is that neither 2a
nor 4a was actually formed. Both should potentially exist in the
form of a common precursor related to the complex base
aggregates which would evolve either towards 2a or 4a in the
presence of the electrophile, depending on the reactivity of the
latter. The nature of such a precursor is further discussed
hereunder.
Metallation of 2-methoxypyridine with BuLi–LiDMAE and
reaction with carbonyl derivatives
As part of the present study, carbonyl compounds were con-
sidered as interesting electrophiles. Indeed with appropriate
structures, such substrates may undergo nucleophilic addition
and competitive enolization. Given the complex nucleophilic as
well as basic properties of our reaction medium, the behaviour
of such substrates was expected to be very informative. More-
over, these reactions could provide good access to 6-hydroxy-
methyl-2-methoxypyridine derivatives. In fact, exploratory
experiments showed that the expected pyridine derivatives
At first, it seemed that the observed THF effect disagreed
with the hypothesis of an aggregated common precursor.
Indeed, the strong solvating power of this solvent relative to
hexane, should have led to substantial destruction of the inter-
mediate aggregates with consequent back formation of BuLi
J. Chem. Soc., Perkin Trans. 1, 1997
3073

View Article Online
Table 3 Metallation of 1 with BuLi–LiDMAE and condensation with carbonyl derivativesa
(1) n [BunLi–LiDMAE]
H
Additive; hexane; 0 °C; 1 h
H
R1
R2
+
R1
N
OMe
Bun
N
OMe
N
OMe
H
O; solvent; 0 °C; 1 h
(2)
OH
R2
1
2b
5
5a R1 = R2 = Me; 5b R1, R2 = (CH2)4
5c R1 = Me, R2 = Et; 5d R1 = R2 = c-C3H5
5e R1, R2 = (CH2)7; 5f R1, R2 = CH CH(CH2)3
5g R1 = R2 = Ph; 5h R1 = H, R2 = But; 5i R1 = H, R2 = Hex
Carbonyl compound
R1, R2
Additive
LiBr
(equiv.)
Solventb
THF
(ml)
Yields (%)c
Runs
n
equiv
Rec. 1
2b
5a–i
1
2
3
4
5
6
7
8
9
2
2
2
2
2
2
2
2
2
2
2
2
Me, Me
Me, Me
Me, Me
Me, Me
(CH2)4
Me, Et
c-C3H5, c-C3H5
(CH2)7
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
—
0.25
—
—
—
40
40
40
40
40
40
40
40
40
40
10
10
8
9
7
4
2
9
5
80
40
49
28
27
9
4
45
43
54
63
80
83
45
49
43
88
53
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
9
45
44
44
5
CH᎐CH-(CH2)3
᎐
10
11
12
Ph, Ph
H, But
H, Hex
10
5
10
32
a Metallations were performed in a constant volume of hexane (40 ml) at 0 ЊC for 1 h. b Carbonyl compounds were added in hexane (5 ml) or in THF
(see Table) and allowed to react at 0 ЊC for 1 h. c Yields determined by GC capillary analysis.
and an increase in the yield of 2b. In order to clarify this point,
we studied the influence of THF on the trapping by DCl/D2O.
Reinforcing our hypothesis, we found that 2d was formed quan-
Discussion
Comparison of the results obtained during the present work
led to the apparently confusing conclusion that the reaction rate
titatively when THF was added 5 min before the trapping.
of the electrophile added to our reaction medium was the main
In contrast, addition of a solution of DCl–D2O in THF
factor determining the nature of the species formed during the
metallation step. Low rates favoured 2a and high rates did 4a. It
gave formation of 4d (82%). In other words, THF favours the
electrophilic condensation of 4a when this species is formed
should be noted that the results obtained with the very reactive
but it is detrimental to its formation from the precursor. The
electrophile DCl–D2O also agreed with this observation. Indeed
in Table 2, the reactive species were trapped (runs 5 and 6)
under heterogeneous conditions since the metallations were
performed in hexane. Thus, the trapping rate was lower than in
the presence of hydrophilic THF, which increased the inter-
action between the metallated species and the quenching
nature of the products formed resulted from these reverse reac-
tions. However, the evolution of the precursor towards either 2a
or 4a must be essentially determined by the nature of the
electrophile.
Metallation of 2-methoxypyridine and reaction with alkylating
agents
reagent and favoured the formation of 4d. We confirmed the
part played by the reactivity of the electrophiles by trapping
with Bu3SnCl, a very reactive electrophile. Under the metal-
lationconditionswhichusuallyledtolargeamountsof 2b(seefor
example Tables 2, 3; runs 5 and 1, respectively) we observed the
formation of 2-methoxy-6-(tributylstannyl)pyridine 10 (70%).
This result confirms that the reagent obtained by metallation of
1 with BuLi–LiDMAE may quantitatively behave like 4a. In
other words had our study had been performed only with
strong electrophiles we should have concluded that metallation
at the C-6 position was nearly quantitative. Finally, we verified
(see Experimental section) that the products of condensation
on the C-6 position were not due to a second metallation of 2a
followed by elimination of BuLi. All these experimental results
supported our hypothesis that 2a and 4a do not actually exist in
our reaction medium but are potentially present in the form of
a common reactive precursor. This hypothesis also explains our
failure to observe butane evolution during the metallation step
throughout the present work. In fact, this took place during the
condensation of the electrophiles (see Experimental section). A
similar observation was made many years ago during the reac-
tion of the complex bases NaNH2–RONa with Ph3CH.
Although a deep red colour had appeared during the metal-
lation, strong evolution of NH3 was observed only during the
condensation of electrophiles such as PhCH2X.18 In accordance
with our aggregative activation principles we proposed that the
From exploratory experiments performed with these electro-
philes, the following data emerged. (i) In hexane without addi-
tive, yields were generally low whatever the amounts of complex
base used. (ii) No valuable effect of LiBr was observed. (iii)
According to our observations during the condensation of car-
bonyl derivatives, an increase in solvent polarity by addition of
THF also increased the reaction yields. However, 4 equiv. of
base and 5 equiv. of electrophile were needed to obtain accept-
able yields. With only a limited decrease in the yields, this stoi-
chiometry may be reduced to 2 and 2.5 equiv., respectively, by
addition of a catalytic amount of cuprous iodide.17 (iv) The
product formed from 2a aromatised during the work-up to give
6. Taking these observations into account we performed the
reactions reported in Table 4.
From the reported results, it appears that alkyl iodides or
sulfates must be preferred to less reactive bromides. Practical
and satisfactory yields may be obtained under appropriate con-
ditions with only a limited excess of base and electrophiles
(runs 2, 4, 6, 8, 10 and 14). Examination of comparable runs in
Tables 4 and 5 which differ only in the nature of the added
electrophile confirms that the formation of the putative 2a and
4a depends on the nature of the electrophiles and reinforces our
hypothesis about the involvement of a common reactive
precursor.
3074
J. Chem. Soc., Perkin Trans. 1, 1997

View Article Online
Table 4 Metallation of 1 with BuLi–LiDMAE and condensation with alkylating agentsa
(1) n [BunLi–LiDMAE]
R
Hexane; 0 °C; 1 h
+
N
OMe
Bun
N
OMe
R
N
OMe
(2) E; additive; solvent; 0 °C
1
6
7
a: R = Me; b: R = Et
c: R = Hex; d: R = c-C6H13(CH2)3
Additive
CuI
(equiv.)
Yields (%)b
THF
(ml)
Runs
n
E (equiv.)
R
Rec. 1
6a–d
7a–d
1
2
3
4
5
6
7
8
9
10
11
12
13
14
4
2
4
2
4
2
4
2
4
2
4
2
4
2
MeI (5)
MeI (2.5)
Me2SO4 (5)
Me2SO4 (2.5)
EtI (5)
EtI (2.5)
Et2SO4 (5)
Et2SO4 (2.5)
Hex-I (5)
Hex-I (2.5)
Hex-Br (5)
Me
Me
Me
Me
Et
Et
Et
Et
Hex
Hex
Hex
Hex
c-C6H11(CH2)3
c-C6H11(CH2)3
—
0.2
—
0.2
—
0.2
—
0.2
—
0.2
—
0.2
—
0.2
80
40
80
40
80
40
80
40
80
40
80
40
80
40
0
5
7
18
0
4
5
20
0
6
0
25
35
14
17
29
42
16
19
34
54
64
68
44
50
70
59
70
62
64
54
71
60
60
40
30
25
52
40
Hex-Br (2.5)
c-C6H11(CH2)3I (5)
c-C6H11(CH2)3I (2.5)
5
2
8
a Metallations were performed in a constant volume of hexane (40 ml) at 0 ЊC for 1 h; electrophiles were added in THF and allowed to react at 0 ЊC
for 1 h. b Yields determined by GC capillary analysis.
2) THF
+
(1)
2b: 60%
MeO
N
N
O
OMe
0 °C; 12 h
11: 30%
HO
2)EDCH; THF
– 78 °C
(2)
(3)
+
HO
2b: 55%
6a: 65%
room temp.
12: 30%
7a : 25%
1) Hexane
0 °C; 1 h
2) MeI; THF
0 °C; 1 h
1 + 2 [BunLi–LiDMAE]
+
2) THF–HMPA (1/1)
0 °C; 1 h
(4)
(5)
(6)
+
+
+
11: 90%
12: 65%
7a : 42%
2b: 5%
2) EDCH; THF –HMPA (1/1)
2b: 25%
6a: 54%
– 78 °C
room temp.
2) MeI; THF–HMPA (1/1)
0 °C; 1 h
EDCH = 4,4-ethylenedioxy-2,6-dimethylcyclohexane-2,5-dione
Fig. 1
red colour was due to aggregates containing the anion radical
NaH–RONa stabilise electron-rich low-oxidation state metal
species6 which can normally survive only in the presence of
ligands such as phosphines and carbon monoxide etc. We
attribute this effect, at least partially, to the cluster effect of the
aggregates. Taking these published results together with those
of the present work, we conclude that the precursor of 2a and
[Ph3CH]Ϫ stabilised by a cluster effect.19 The presence of such a
ؒ
radical has since been confirmed by EPR.20 This hypothesis was
also supported by the finding that the SET propensity of NaH
was considerably increased when this hydride was included in
appropriate aggregates NaH–RONa.8 Such a propensity was
exacerbated when low-oxidation state metal species were intro-
duced inside the aggregates to give complex reducing agents
(CRA).6,20 It is worth noting that matrices of aggregates of
4a contains [2-MeOC5H4N]Ϫ in a stabilised state. This being so,
ؒ
it became possible to direct the evolution of the precursor
towards the customary free-radical reactions, simply by placing
J. Chem. Soc., Perkin Trans. 1, 1997
3075

View Article Online
1 in contact with BuLi–LiDMAE in an appropriate solvent.
The reactions reported in Fig. 1 completely agree with our
hypothesis.
Me
N
Li
(BunLi)m
Me
Bun
+
([Me2N(CH2)2OLi])n
In hexane–THF without an electrophile [Eqn. (1)] we
observed the formation in significant yield of a radical-coupling
product 11. Newkome et al.21 showed that under analogous
conditions, (i.e. the addition of HMPA, a well known radical
stabilising solvent22), the amount of coupling product was con-
siderably increased. We observed a similar increase [Eqn. (4)] as
well as a very large increase in the reaction rate. The radical
nature of the reactive intermediate is also supported by the
reactions reported in Eqns. (2) and (5) using 4,4-ethylenedioxy-
2,6-dimethylcyclohexane-2,5-dione (EDCH) as radical trap-
ping agent.23 It should be noted that the yield of 12 due to the
trapping of the radical intermediate corresponds to the yield of
the coupling product 11, respectively, obtained in Eqns. (1) and
(4) under the same solvent conditions. In addition, an EPR
study of a sample containing 1 and 2 equiv. of the complex base
in hexane showed a large signal (∆H = 11 G) with a g-factor
of 2.0023. We were unable to obtain the fine structure but this
observation reinforces our hypothesis. Finally, Eqns. (3) and (6)
show that, under the corresponding conditions and in the
presence of an electrophile, the products resulting from radical
reactions were replaced by condensation products. This obser-
vation supports the hypothesis of a common precursor with
radical-like properties. To sum up, we had to reconcile the
following data. (i) Mixing of BuLi and LiDMAE gives a
complex base in which the properties of each constituent are
much modified and reflect a tight association to give mixed
aggregates. (ii) Metallation of 2-methoxypyridine 1 with the
complex base does not obey the known DOM effect3 and
takes place at the C-6 position instead of C-3. (iii) Conden-
sation of electrophiles leads to the formation of products com-
ing from the apparent irreversible addition of BuLi to 1 as
well as its apparent metallation at the C-6 position. Under
given conditions, the ratio of the product formed, unusually,
depends on the nature of the trapping electrophile. As a corol-
lary, neither 2a nor 4a actually exists in the reaction medium.
(iv) A common precursor must potentially contain 2a and 4a.
(v) The common precursor must possess some radical char-
acter stabilised by the aggregate cluster effect. The reactions
reported in Scheme 2 tentatively account for the above
requirements.
Li
p
O
1
Me
N
H
O
Me
Me
Bun
Li
Li
N
1
O
q
–
•
N
a
Me
Li
Products
Products
[ 2a ]
[ 4a ]
a
H
O
b
b
a
+
Me
Me
Bu
b
•
Li
N
O
q
HMPA
Li+
Li+
–
H
O
H
–
–
2 LiH + 11
Li+
•
Me
N
N
OMe
N
Me
O
N
H
H
+
n
•
Bu
Li
Solvent
Bun
Bun – Bun + BunH
Li
r
O
Scheme 2
1) 2 [BunLi / LiDMAE]
Hexane; 0 °C; 1 h
In such a mechanism, the aggregates of the complex bases
and, consequently, the alkoxides must have an appropriate
structure to fit with the stereoelectronic requirements of the
complexation of 1. This hypothesis explains how a change in
the nature of the lithium aminoalkoxides led to an accompany-
ing change in the behaviour of the complex base. Such require-
ments and their consequences were previously found during our
study of syn eliminations performed with the CB NaNH2–
RONa.24 At the present time we are unable to provide either a
detailed mechanism for the condensation of the electrophile or
an explanation for the effect of THF. However, the mechanism
proposed opens new areas of investigation and also explains the
reactions observed during the metallation of dithioketal with
NaNH2–Et(OCH2CH2)2ONa.25
N
X
R
N
X
2) ClSiMe3 or MeI; CuI
or
MeCOEt; LiBr
THF; 0°C
8 X = SMe
9 X = NMe2
8a X = SMe; R = SiMe3; 80%
8b X = SMe; R = Me; 62%
8c X = SMe; R = MeEtCOH; 75%
9a X = NMe2; R = SiMe3; 52%
9b X = NMe2; R = Me; 45%
9c X = NMe2; R = MeEtCOH; 44%
Scheme 3
of 2-methoxypyridine reflects a surface effect6 due to an associ-
ation of the substrates with the aggregates. A similar role of the
ions distribution at the aggregates surfaces explained the
unique ability of NaNH2–RONa to perform syn-eliminations.7
On the other hand, experimental evidence strongly suggests
that the actual metallated species come from a precursor during
the electrophile addition. It seems likely that this precursor is of
a radical nature and stabilised by a cluster effect, according to
aggregative activation.19 In other words, the role of proton
abstraction is once more questioned. It is interesting to note
the analogy as illustrated by Eqns. (1) and (2) between the
Finally, we have briefly checked that the reactions presently
reported were not limited to 2-alkoxypyridines and that
replacement of the oxygen atom by sulfur and nitrogen atoms
led to similar results (see Scheme 3).
Conclusion
Complex bases BuLi–ROLi have been obtained. According to
aggregative activation principles,7,8 the nucleophilicity/basicity
ratio of the BuLi included in these new unimetal superbases7
may be changed by simply modifying the nature of the activat-
ing agent ROLi. A thorough study of BuLi–LiDMAE showed,
once again, how the intrinsic properties of bases included in
mixed aggregates can be modified. The unusual C-6 metallation
(R .. Mgϩ) XϪ
RMgX (1)
Ϫ
ϩ
ؒ
ؒ
RX ϩ Mg
(RX) Mg
Ϫ
ϩ
ؒ
ؒ
RH ϩ CB
(RH) CB
RϪ Mϩ
(2)
3076
J. Chem. Soc., Perkin Trans. 1, 1997

View Article Online
mechanism of generation of organometallic reagents from
organic halides and a metal and the presently proposed
proton abstraction mechanism. Such an analogy also appeared
during our study with complex reducing agents.6 Finally, we
have provided a simple preparation of C-6 substituted 2-
methoxypyridines. We are actively pursuing further work in this
area.
10% aqueous HCl (40 ml). The mixture was then extracted
twice with ether (20 ml), and the combined extracts were dried
(MgSO4) and evaporated under reduced pressure. The crude
product was then purified on a Chromatotron (AcOEt–hexane,
5:95, as eluent) to give 2-methoxy-6-methylthiopyridine 4c (830
mg, 67%).
2-Butyl-2,5-dihydro-6-methoxypyridine 2b.10 δH 0.75–1.05 (m,
3H, CH ), 1.15–1.80 (m, 6H, CH ), 2.70 (d, 2H, J 6.2, ᎐CCH ),
3.70 (s, 3H, CH3O), 3.90–4.25 (m, 1H, N-CH) and 5.65 (s, 2H,
Experimental
᎐
3
2
2
General methods
CH᎐CH).
᎐
1H and 13C NMR spectra were recorded on a Bruker spec-
trometer at 400 and 100 MHz, respectively, with TMS as
internal standard and CDCl3 as solvent. J Values are given in
Hz. GC analysis were performed with an internal standard on a
Shimadzu GC-14A apparatus using a HP1 25m column and
temperature programming. GC–MS measurements (EI) were
performed on a HP5971A spectrometer. HRMS were per-
formed by the Centre de Spectrochimie Organique de l’Univer-
sité Pierre et Marie Curie (Paris). Elemental analyses were
performed by the Service Central d’Analyse du CNRS (Vernai-
son, France).
2-Methoxy-6-trimethylsilylpyridine 4b. δH 0.25 (s, 9H,
CH3Si), 3.90 (s, 3H, CH3O), 6.65 (d, 1H, J 8.2, H-3), 7.10 (d,
1H, J 8.2, H-5), 7.45 (t, 1H, J 8.2 H-4); δC Ϫ1.25 (CH3Si), 54.9
(CH3O), 111.8 (C-5), 124.0 (C-3), 138.5 (C-4), 165.2 (C-2) and
167.2 (C-6); m/z 181 (Mϩ); m/z (EI) 181.0923 (Mϩ C9H15NOSi
requires 181.0923).
2-Methoxy-6-methylthiopyridine 4c. δH 2.55 (s, 3H, CH3S),
3.90 (s, 3H, CH3O), 6.40 (d, 1H, J 7.8, H-3), 6.75 (d, 1H, J
7.7, H-5) and 7.40 (t, 1H, J 7.8, H-4); δC 12.9 (CH3S), 52.95
(CH3O), 104.9 (C-5), 113.2 (C-3), 138.1 (C-4), 156.8 (C-6)
and 163.3 (C-2); m/z 155 (Mϩ) (Found: C, 54.05; H, 5.47; N,
9.21; S, 20.37 C7H9NOS requires C, 54.19; H, 5.81; N, 9.03,
S. 20.65%).
Materials
2-Methoxy[6-2H]pyridine 4d. This compound was obtained
as a mixture with 2-methoxypyridine; δH 3.90 (s, 3H, CH3O),
6.70 (d, 1H, J 8.7, H-3), 7.15 (d, 1H, J 7.7, H-5) and 7.40 (t, 1H,
J 7.8, H-4).
BuLi (1.6  solution in hexane), 2-methoxypyridine and 2-
(N,N-dimethylamino)pyridine were purchased from Aldrich. 2-
Methylthiopyridine was prepared by treating 2-bromopyridine
with BuLi at Ϫ78 ЊC followed by quenching with dimethyl
disulfide. Hexane and THF were distilled and stored over
sodium wire before use. Ether refers to diethyl ether. Alcohols
and amino alcohols were commercially available and purified
by the usual methods when necessary. LiBr and CuI were dried
at 100 ЊC under reduced pressure for 24 h and used immediately.
Chlorotrimethylsilane, dimethyl disulfide, ketones, aldehydes,
alkyl halides and acyl chlorides were distilled or recrystallized
before use. 4,4-Ethylenedioxy-2,6-dimethylcyclohexane-2,5-
dione (EDCH) was prepared from 2,6-dimethylbenzoquinone23
according to a published procedure.26
Typical procedure for metallation of 2-methoxypyridine with
2(BuLi–LiDMAE) in the presence of LiBr and condensation with
butan-2-one in THF (run 6, Table 3)
To the above prepared base was added solid, anhydrous LiBr
(2 mmol, 175 mg). The mixture was stirred at 0 ЊC for 0.5 h
after which 2-methoxypyridine (8 mmol, 870 mg) in hexane
(10 ml) was added dropwise to it. The reaction mixture was
stirred at 0 ЊC for 1 h after which a solution of butan-2-one
(20 mmol, 1.44 g) in THF (40 ml) was added dropwise to it.
After being allowed to react for 1 h the product yield (GC)
was 80%. The mixture was finally hydrolysed at 0 ЊC with
10% aqueous HCl (40 ml). The aqueous layer was extracted
with ether (2 × 20 ml) after which it was made alkaline with
aqueous NaOH and extracted with ether. The combined
extracts were dried (MgSO4) and evaporated and the crude
product was purified on a Chromatotron [AcOEt–hexane
(1:9) as eluent] to give 2-(2-methoxy-6-pyridyl)butan-2-ol 5c
(1.16 g, 70%).
1-(2-Methoxy-6-pyridyl)-1-methylethanol 5a. δH 1.50 (s, 6H,
CH3), 3.95 (s, 3H, CH3O), 4.65 (s, 1H, OH), 6.60 (d, 1H, J 7.5,
H-3), 6.90 (d, 1H, J 7.5, H-5), 7.60 (t, 1H, J 7.6, H-4); δC 30.4
(CH3), 53.2 (CH3O), 71.6 (C-OH), 108.5 (C-5), 110.7 (C-3),
139.6 (C-4), 162.6 (C-6) and 163.9 (C-2); m/z 167 (Mϩ) (Found:
C, 64.97; H, 8.07; N, 8.62. C9H13NO2 requires C, 64.67; H, 7.78;
N, 8.38%).
1-(2-Methoxy-6-pyridylcyclopentanol 5b. δH 1.50–2.00 (m,
8H, CH2), 3.95 (s, 3H, CH3O), 4.55 (s, 1H, OH), 6.60 (d, 1H, J
7.9, H-3), 6.95 (d, 1H, J 7.9, H-5) and 7.60 (t, 1H, J 7.8 H-4);
δC 24 (CH2), 42.2 (CH2), 52.9 (CH3O), 71.2 (C-OH), 107.9
(C-5), 110.8 (C-3), 139.1 (C-4), 161.8 (C-6) and 162.7 (C-2); m/z
193 (Mϩ) (Found: C, 68.48; H, 8.07; N, 7.43. C11H15NO2
requires C, 68.39; H, 7.77; N, 7.25%).
2-(2-Methoxy-6-pyridyl)butan-2-ol 5c. δH 0.80 (t, 3H, J 8,
CH3CH2), 1.50 (s, 3H, CH3C), 1.80 (q, 2H, J 8, CH3CH2), 3.95
(s, 3H, CH3O), 4.75 (s, 1H, OH), 6.60 (d, 1H, J 8.2 H-3), 6.85
(dd, 1H, J 8.2, H-5) and 7.60 (dt, 1H, J 8.1, H-4); δC 8.1
(CH3CH2), 28.6 (CH3C), 36.2 (CH2C), 53.3 (CH3O), 74.0
(C-OH), 108.4 (C-5), 111.5 (C-3), 139.5 (C-4), 162.7 (C-6) and
163.0 (C-2); m/z 181 (Mϩ) m/z (CI) 182.1160 [(M ϩ H)ϩ
C10H15NO2 requires 182.1161].
Typical procedure for metallation of 2-methoxypyridine with
4(BuLi–LiDMAE) and condensation with chlorotrimethylsilane
(run 10, Table 1)
A three-necked flask (100 ml) cooled at 0 ЊC under a nitrogen
atmosphere was charged with BuLi (32 mmol, 20 ml). N,N-
Dimethylaminoethanol (16 mmol, 1.42 g) in hexane (10 ml) was
then added dropwise to the flask followed after 0.5 h at 0 ЊC, by
2-methoxypyridine (4 mmol, 440 mg) in hexane (10 ml), also
added dropwise. The orange-coloured solution was stirred at
0 ЊC for 1 h after which chlorotrimethylsilane (16 mmol, 1.76 g)
in hexane (5 ml) was introduced. The reaction was allowed to
continue for 1 h at 0 ЊC after which the product yield (GC) was
94%. The solution was finally hydrolysed at 0 ЊC with 10%
aqueous HCl (40 ml). The aqueous layer was extracted twice
with ether (20 ml) and the combined extracts were dried
(MgSO4) and evaporated under reduced pressure. The crude
product was then purified on a Chromatotron (AcOEt–hexane,
5:95, as eluent) to give 2-methoxy-6-trimethylsilylpyridine 4b
(440 mg, 74%).
Typical procedure for metallation of 2-methoxypyridine with
2(BuLi–LiDMAE) in the presence of LiBr and condensation with
dimethyl disulfide (run 4, Table 2)
To the above prepared base was added anhydrous LiBr (2
mmol, 175 mg) as a solid. The mixture was stirred at 0 ЊC for 0.5
h after which 2-methoxypyridine (8 mmol, 870 mg) in hexane
(10 ml) was added dropwise to it. The reaction mixture was
stirred at 0 ЊC for 1 h after which a solution of dimethyl
disulfide (20 mmol, 1.9 g) in THF (5 ml) was then added drop-
wise to it. After being allowed to react for 1 h the product yield
(GC) was 72%. The solution was finally hydrolysed at 0 ЊC with
2-Methoxy-6-pyridyl(dicyclopropyl)methanol 5d. δH 0.15–0.55
(m, 8H, CH2C), 1.20 (m, 2H, CHCH2), 3.90 (s, 3H, CH3O), 4.60
J. Chem. Soc., Perkin Trans. 1, 1997
3077

View Article Online
(s, 1H, OH), 6.65 (d, 1H, J 7.8, H-3), 7.10 (d, 1H, J 7.7, H-5)
and 7.60 (t, 1H, J 7.8, H-4); δC 0.79 (CH), 19.7 (CH2), 52.7
(CH3O), 71.2 (C-OH), 107.9 (C-5), 111.8 (C-3), 139.1 (C-4),
161.77 (C-6) and 162.9 (C-2); m/z 219 (Mϩ) (Found: C, 70.93;
H, 8.05; N, 6.15 C13H17NO2 requires C, 71.23; H, 7.76; N,
6.39%).
1-(2-Methoxy-6-pyridyl)cyclooctanol 5e. δH 1.50–2.00 (m,
14H, CH2), 3.90 (s, 3H, CH3O), 4.50 (s, 1H, OH), 6.60 (d, 1H,
J 7.4, H-3), 6.90 (d, 1H, J 7.4, H-5) and 7.55 (t, 1H, J 7.5,
H-4); δC 21.9 (CH2), 24.3 (CH2), 28.3 (CH2), 37.3 (CH2), 53.2
(CH3O), 75.7 (C-OH), 108.5 (C-5), 111.5 (C-3), 139.3 (C-4),
162.5 (C-6) and 164.1 (C-2); m/z 235 (Mϩ) (Found: C, 71.60;
H, 9.02; N, 6.15. C14H21NO2 requires C, 71.46; H, 8.99; N,
5.95%).
1.6 (m, 4H, CH2), 2.7 (t, 4H, J 8, CH2), 3.95 (s, 3H, CH3O), 6.65
(d, 1H, J 7, H-5) and 7.80 (d, 1H, J 7, H-4).
2-Butyl-3-hexyl-6-methoxypyridine 6c. δH 1.00 (m, 6H, CH3),
1.4–1.7 (m, 12H, CH2), 2.8 (t, 4H, J 8, CH2), 3.95 (s, 3H,
CH3O), 6.70 (d, 1H, J 7, H-5) and 7.90 (d, 1H, J 7, H-4).
2-Butyl-3-cyclohexylpropyl-6-methoxypyridine 6d. δH 0.8–0.9
(m, 3H, CH3), 1.15–1.20 (m, 8H, 4 CH2), 1.5–1.75 (m, 11H,
CH2 ϩ CH), 2.7 (t, 4H, J 8, CH2), 3.95 (s, 3H, CH3O), 6.75 (d,
1H, J 7, H-5) and 7.75 (d, 1H, J 7, H-4).
6-Methyl-2-methoxypyridine 7a. δH 2.4 (s, 3H, CH3), 3.90 (s,
3H, CH3O), 6.50 (d, 1H, J 7.8, H-3), 6.80 (d, 1H, J 7.8, H-5)
and 7.5 (t, 1H, J 7.7, H-4); δC 24.1 (CH3), 53.2 (CH3O), 107.1
(C-5), 115.6 (C-3), 138.6 (C-4), 156.2 (C-6) and 163.1 (C-2); m/z
123 (Mϩ) (Found: C, 68.13; H, 7.46; N, 11.45. C7H9NO requires
C, 68.27; H, 7.37; N, 11.37%).
1-(2-Methoxy-6-pyridyl)cyclohex-2-enol 5f. δH 1.10–2.20 (m,
6H, CH2), 3.90 (s, 3H, CH3O), 4.85 (s, 1H, OH), 5.60–6.00 (m,
6-Ethyl-2-methoxypyridine 7b. δH 1.25 (t, 3H, J 8, CH3), 2.6
(m, 2H, CH2), 3.95 (s, 3H, CH3O), 6.50 (d, 1H, J 7.6, H-3), 6.70
(d, 1H, J 7.5, H-5) and 7.45 (t, 1H, J 7.7, H-4); δC 13.2 (CH3),
30.6 (CH2), 52.8 (CH3O), 106.8 (C-5), 113.9 (C-3), 138.4 (C-4),
160.2 (C-6) and 163.3 (C-2); m/z 137 (Mϩ) (Found: C, 70.13; H,
8.24; N, 10.05. C8H11NO requires C, 70.04; H, 8.08; N, 10.21%).
6-Hexyl-2-methoxypyridine 7c. δH 0.95 (t, 3H, J 8, CH3),
2H, CH᎐CH), 6.60 (d, 1H, J 8.3, H-3), 6.80 (d, 1H, J 8.4, H-5)
᎐
and 7.50 (t, 1H, J 8.3, H-4); δC 19.1 (CH2), 24.9 (CH2), 38.9
(CH2), 53.3 (CH3O), 73.2 (C-OH), 108.9 (C-5), 112.6 (C-3),
139.3 (C-4), 162.5 (C-2) and 181.1 (C-6); m/z 205 (Mϩ) (Found:
C, 70.35; H, 7.29; N, 6.87. C12H15NO2 requires C, 70.22; H,
7.37; N, 6.82%).
1-(2-Methoxy-6-pyridyl)-1,1-diphenylmethanol 5g. δH 3.90 (s,
3H, CH3O), 5.90 (s, 1H, OH), 6.60 (d, 1H, J 6.9, H-3), 6.70 (d,
1H, J 6.8, H-5), 7.25 (m, 10H, Ph) and 7.5 (t, 1H, J 6.9, H-4); δC
53.2 (CH3O), 76.7 (C-OH), 109.2 (C-5), 115.3 (C-3), 127.9
(CAr), 138.7 (C-4), 145.8.5 (C-6), 164.1 (C-2); m/z 291 (Mϩ)
(Found: C, 78.48; H, 6.07; N, 4.75. C19H17NO2 requires C,
78.33; H, 5.88; N, 4.81%).
1.35–1.70 (m, 8H, CH ), 2.85 (t, 2H, J 7.8, CH C᎐N), 3.90 (s, 3H,
᎐
2 2
CH3), 6.50 (d, 1H, J 7.7, H-3), 6.80 (d, 1H, J 7.7, H-5) and 7.45
(t, 1H, J 7.8, H-4); δC 13.8 (CH3), 22.3 (CH2), 22.3–37.6 (CH2),
52.8 (CH3O), 106.7 (C-5), 114.6 (C-3), 138.2 (C-4), 160.1 (C-6)
and 163.3 (C-2); m/z 193 (Mϩ) (Found: C, 74.65; H, 10.02; N,
7.33. C12H19NO requires C, 74.57; H, 9.91; N, 7.25%).
6-Cyclohexylpropyl-2-methoxypyridine 7d. δH 1.15–1.25 (m,
1-(2-Methoxy-6-pyridyl)-2,2-dimethylpropanol 5h. δH 0.95 (s,
9H, CH3C), 3.90 (s, 3H, CH3O), 4.05 (s, 1H, CHO), 4.30 (s, 1H,
OH), 6.65 (d, 1H, J 7.3, H-3), 6.80 (d, 1H, J 7.3, H-5) and 7.50
(t, 1H, J 7.4, H-4); δC 25.8 (CH3C), 36.2 (C), 53.1 (CH3O), 80.0
(CH-OH), 109.1 (C-5), 115.4 (C-3), 138.2 (C-4), 157.5 (C-6) and
162.9 (C-2); m/z 194 (Mϩ) (Found: C, 67.93; H, 8.53; N, 7.07.
C11 H17NO2 requires C, 67.69; H, 8.71; N, 7.18%).
1-(2-Methoxy-6-pyridyl)heptan-1-ol 5i. δH 0.95 (m, 3H, CH3),
1.10–1.50 (m, 8H, 4 CH2), 1.80 (m, 2H, CH2), 3.90 (s, 3H,
CH3O), 4.10 (d, 1H, CHO), 4.25 (s, 1H, OH), 6.70 (d, 1H, J 7.7,
H-3), 6.85 (d, 1H, J 7.7, H-5) and 7.50 (t, 1H, J 7.7, H-4); δC
14.0 (CH3), 22.5 (CH2), 22.6–38.2 (CH2), 53.2 (CH3O), 79.5
(CHOH), 106.5 (C-5), 113.3 (C-3), 138.5 (C-4), 160.5 (C-6) and
163.1 (C-2); m/z 223 (Mϩ) (Found: C, 69.72; H, 9.68; N, 6.12.
C13H21NO2 requires C, 69.95; H, 9.41; N, 6.28%).
8H, 4 CH2), 1.50–1.75 (m, 7H, CH ϩ CH2), 2.60 (t, 2H, J 8,
CH C᎐N), 3.90 (s, 3H, CH ), 6.50 (d, 1H, J 7.1, H-3), 6.70 (d,
᎐
2
3
1H, J 7.1, H-5) and 7.45 (t, 1H, J 7.3, H-4); δC 7.3 (CH), 26.1–
37.9 (CH2), 52.8 (CH3O), 106.7 (C-5), 114.6 (C-3), 138.2 (C-4),
160.1 (C-6) and 163.2 (C-2); m/z 233 (Mϩ) (Found: C, 77.35; H,
10.12; N, 6.28. C15H23NO requires C, 77.21; H, 9.93; N, 6.00%).
2-Methylthio-6-trimethylsilylpyridine 8a. δH 0.30 (s, 9H,
CH3Si), 2.50 (s, 3H, CH3S), 7.10 (d, 1H, J 7.9, H-3), 7.20 (d, 1H,
J 8, H-5) and 7.35 (t, 1H, J 7.8, H-4); δC Ϫ1.5 (CH3Si), 12.5
(CH3S), 120.4 (C-5), 123.8 (C-3), 133.2 (C-4), 158.4 (C-2) and
167.6 (C-6); m/z 197 (Mϩ) (Found: C, 55.08; H, 7.93; N, 7.08.
C9H15NSSi requires C, 54.82; H, 7.61; N, 7.10%).
6-Methyl-2-methylthiopyridine 8b. δH 2.45 (s, 3H, CH3), 2.55
(s, 3H, CH3S), 6.80 (d, 1H, J 7.4, H-3), 6.95 (d, 1H, J 7.4, H-5)
and 7.40 (t, 1H, J 7.3, H-4); δC 13.8 (CH3S), 24.8 (CH3), 118.2
(C-5), 119.5 (C-3), 136.5 (C-4), 158.7 (C-2) and 159.5 (C-6); m/z
139 (Mϩ) (Found: C, 60.65; H, 6.22; N, 10.42. C7H9NS requires
C, 60.43; H, 6.47; N, 10.07%).
2-Methylthio-6-pyridylbutan-2-ol 9c. δH 0.75 (t, 3H, J 8, CH3),
1.50 (s, 3H, CH3), 1.80 (q, 2H, J 8, CH2), 2.55 (s, 3H, CH3S),
5.00 (s, 1H, OH), 6.95 (d, 1H, J 8.1, H-3), 7.05 (d, 1H, J 8.2,
H-5) and 7.50 (t, 1H, J 8.1, H-4); δC 7.9 (CH3), 13.1 (CH3S),
28.5 (CH3), 35.9 (CH2), 74.1 (C-OH), 114.3 (C-5), 1193 (C-3),
136.6 (C-4), 157.6 (C-2) and 164.7 (C-2), m/z 197 (Mϩ) (Found:
C, 60.80; H, 7.76; N, 6.92. C10H15NOS requires C, 60.91; H,
7.61; N, 7.11%).
2-Dimethylamino-6-trimethylsilylpyridine 9a. δH 0.25 (s, 9H,
CH3Si), 3.10 (s, 6H, CH3N), 6.45 (d, 1H, J 7.8, H-3), 6.75 (d,
1H, J 7.8, H-5) and 7.45 (t, 1H, J 7.9, H-4); δC Ϫ1.8 (CH3Si),
37.7 (CH3N), 104.9 (C-5), 117.0 (C-3), 135.1 (C-4), 159.1 (C-2)
and 165.8 (C-6); m/z 194 (Mϩ) (Found: C, 61.56; H, 9.12; N,
14.62. C10H18N2Si requires C, 61.85; H, 9.28; N, 14.43%).
2-Dimethylamino-6-methylpyridine 9b. δH 2.40 (s, 3H, CH3),
3.15 (s, 6H, CH3N), 6.30 (d, 1H, J 8, H-3), 6.40 (d, 1H, J 8, H-5)
and 7.30 (t, 1H, J 7.9, H-4); δC 24.8 (CH3), 38.2 (CH3N), 102.8
(C-5), 111.1 (C-3), 137.6 (C-4), 156.9 (C-6) and 158.9 (C-2); m/z
136 (Mϩ) (Found: C, 70.23; H, 9.10; N, 20.71. C7H9NS requires
C, 70.59; H, 8.82; N, 20.59%).
Typical procedure for metallation of 2-methoxypyridine with
2(BuLi–LiDMAE) and condensation of iodomethane in the
presence of CuI and THF (run 2, Table 4)
A three-necked flask (100 ml) cooled at 0 ЊC under a nitrogen
atmosphere was charged with BuLi (32 mmol, 20 ml) to
which N,N-dimethylaminoethanol (16 mmol, 1.42 g), in hexane
was then added dropwise. After 0.5 h at 0 ЊC, the mixture was
treated with 2-methoxypyridine (8 mmol, 880 mg) in hexane (10
ml), added dropwise. The mixture was stirred at 0 ЊC for 1 h
after which a solution of iodomethane (20 mmol, 2.84 g) in
THF (40 ml) and anhydrous CuI (1.6 mmol, 300 mg) were
added simultaneously to it. After 1 h at 0 ЊC (GC yield was then
59%), the mixture was hydrolysed at 0 ЊC with water (40 ml).
The mixture was extracted with ether (2 × 20 ml) and the
extract dried (MgSO4) and evaporated. The crude product
was purified on a Chromatotron (hexane as eluent) to give 2-
methoxy-6-methylpyridine 7a (510 mg, 52%).
Compounds 6a, 6b, 6c and 6d were found to be identical
with authentic samples prepared according to a published
method.11b
2-Butyl-3-methyl-6-methoxypyridine 6a. δH 0.95 (m, 3H,
CH3), 1.5 (m, 4H, CH2), 2.3 (s, 3H, CH3), 2.8 (t, 2H, J 8, CH2),
4.00 (s, 3H, CH3O), 6.70 (d, 1H, J 7, H-5) and 7.90 (d, 1H, J 7,
H-4).
2-Dimethylamino-6-pyridylbutan-2-ol 9c. δH 0.75 (t, 3H, J 8,
CH3), 1.45 (s, 3H, CH3), 1.75 (q, 2H, J 8, CH2), 3.10 (s, 6H,
CH3N), 6.45 (d, 1H, J 7, H-3), 6.50 (d, 1H, J 7, H-5) and 7.50 (t,
2-Butyl-3-ethyl-6-methoxypyridine 6b. δH 0.9–1 (m, 6H, CH3),
3078
J. Chem. Soc., Perkin Trans. 1, 1997

View Article Online
1H, J 6.9, H-4); δC 8.1 (CH3), 28.5 (CH3), 35.9 (CH2), 37.9
(CH3N), 73.4 (C-OH), 103.2 (C-5), 106.1 (C-3), 138.1 (C-4),
157.2 (C-2) and 162.8 (C-2); m/z 194 (Mϩ) (Found: C, 68.28; H,
9.12; N, 14.27. C11H18N2O requires C, 68.04; H, 9.28; N,
14.47%).
Procedure for measurement of butane evolution
The reaction was performed in a Schlenck apparatus connected
to a vacuum pump (250 Torr) via a trap immersed in liquid
nitrogen. After each reaction step, the trap was disconnected
from the Schlenck apparatus, stopped at one extremity whilst
the other was connected to a volume measurement device. The
trap was then allowed to heat to room temperature whilst the
volume of butane was measured. The volumes measured for a
typical run [see Table 2; run 1): BuLi (32 mmol), DMAE 16
mmol), 2-methoxypyridine (8 mmol), ClSiMe3 (20 mmol)] are
reported in the following table.
Preparation of 2-methoxy-6-tributylstannylpyridine 10
A three-necked flask cooled at 0 ЊC under a nitrogen atmos-
phere was charged with BuLi (32 mmol, 20 ml) to which N,N-
dimethylaminoethanol (16 mmol, 1.42 g) in hexane (10 ml) was
then added dropwise. After 0.5 h at 0 ЊC the mixture was treated
with 2-methoxypyridine (8 mmol, 880 mg) in hexane (10 ml),
added dropwise. The solution was stirred at 0 ЊC for 1 h after
which a solution of tributyltin chloride (24 mmol, 7.82 g) in
THF (40 ml) was added dropwise to it. After 0.5 h at 0 ЊC, the
yellow solution was hydrolysed at 0 ЊC with water (40 ml) and
then extracted with ether (2 × 20 ml). The combined extracts
were dried (MgSO4) and evaporated and the crude product
was diluted with hexane (40 ml) and passed down a column of
neutral alumina. After evaporation of the eluate, the crude
stannylpyridine (GC yield: 94%) was distilled under reduced
pressure (140 ЊC/10 Torr) to yield 2-methoxy-6-tributylstannyl-
pyridine 10 (2.23 g, 70%).
Base
preparation
2-methoxypyridine 1
(1 h; 0 ЊC)
TMSCl
(1 h; 0 ЊC)
H2O
V/mla
300 (384)
V/mla
0 (0)
V/mla
85 (115)
V/mla
160 (243)
a Corrected by deducing dilation volume (50 ml). Butane was character-
ised by GC–MS: m/z 58 (Mϩ). The number in parentheses are the theor-
etical volumes taking into account the GC analysis of the reaction
medium after hydrolysis (Recovered 1, 25%; 2b, 13%; 4b, 60%)
2-Methoxy-6-tributylstannylpyridine 10. δH 0.90 (t, 9H, J 7,
CH3CH2), 1.10 (m, 6H, CH2), 1.35 (m, 6H, CH2), 1.5 (t, 6H, J 8,
CH2Sn), 3.90 (s, 3H, CH3O), 6.55 (d, 1H, J 7.8, H-3), 7.00 (d,
1H, J 7.8, H-5) and 7.35 (t, 1H, J 7.9, H-4); δC 9.6 (CH2Sn), 13.4
(CH2CH3), 27 (CH2), 28.7 (CH2), 52.8 (CH3O), 108.6 (C-5),
125.6 (C-3), 135.5 (C-4), 162.8 (C-2) and 170.3 (C-6) (Found: C,
54.32; H, 3.44; N, 3.23. C18H33NOSn requires C, 54.27; H, 3.27;
N, 3.52%).
Acknowledgements
We thank Professor G. Queguiner for helpful recommendations
and Professor A. Merlin for EPR measurements. We are also
grateful to Dr Ph. Amard for correction of our English and to
the editor for useful corrections.
References
Homocoupling of 1 by addition of THF–HMPA: synthesis of
6,6Ј-dimethoxy-2,2Јbipyridine 11
1 (a) L. Lochmann, J. Pospisil, J. Vodnansky, J. Trekoval and D. Lim,
Collect, Czech. Chem. Commun., 1965, 30, 2187; (b) L. Lochmann
and D. Lim, J. Organometal. Chem., 1973, 50, 9.
2 J. F. McGarrity and C. A. Ogle, J. Am. Chem. Soc., 1984, 107,
1805.
3 For reviews on Directed Ortho Metallation (DOM) see for example:
(a) P. Beak and A. I. Meyers, Acc. Chem. Res., 1986, 19, 356;
(b) V. Snieckus, Chem. Rev., 1990, 90, 879.
4 (a) M. Marsch, K. Harms, L. Lochmann and G. Boche, Angew.
Chem., Int. Ed. Engl., 1990, 29, 308; (b) P. Hall, J. Gilchrist,
A. Harrison, D. Fuller and D. Collum, J. Am. Chem. Soc.,
1991, 113, 9575; (c) R. Mulvey, Chem. Soc. Rev., 1991, 20, 167;
(d) D. Collum, Acc. Chem. Res., 1992, 25, 448 and references cited
therein; (e) G. Delong, D. Pannell, M. Clarke and R. Thomas,
J. Am. Chem. Soc., 1993, 115, 7013.
5 (a) F. Romersberg and D. Collum, J. Am. Chem. Soc., 1994, 116,
9187; (b) J. Saà, G. Martorell and A. Frontera, J. Org. Chem., 1996,
61, 5194; (c) M. Davidson, R. Davies, P. Raithby and R. Snaith,
J. Chem. Soc., Chem. Commun., 1996, 14, 1695; (d) B. Lucht and
D. Collum, J. Am. Chem. Soc., 1996, 118, 2217.
6 P. Caubère, Rev. Heteroatom Chem., 1991, 4, 78 and references cited
therein.
7 P. Caubère, Chem. Rev., 1993, 93, 2317 and references cited therein.
8 (a) P. Caubère, Acc. Chem. Res., 1974, 7, 301; (b) P. Caubère,
Top. Curr. Chem., 1978, 73, 50 and references cited therein;
(c) G. Ndebeka, P. Caubère, S. Raynal and S. Lecolier, Polymer,
1981, 22, 347.
2-Methoxypyridine (8 mmol, 870 mg) was allowed to react
for 1 h with the 2(BuLi/LiDMAE) base in hexane (50 ml) at
0 ЊC; a mixture of THF and HMPA (1:1; 100 ml) was then
added dropwise to it. The mixture, turning from dark green
to dark brown and then becoming colourless, was stirred for
1 h at 0 ЊC. After work-up, the crude product was treated
with ether–hexane mixture (1:1; 100 ml). The precipitated
solid was filtered off and recrystallized from hexane to yield
10 (780 mg, 90%) as a pale yellow solid, mp 118 ЊC (lit.,27
119 ЊC).
Radical-anion trapping using EDCH
2-Methoxypyridine (8 mmol, 870 mg) was allowed to react for 1
h with the base 2 (BuLi–LiDMAE) in hexane (50 ml) at 0 ЊC,
the mixture being cooled to Ϫ78 ЊC. A solution of EDCH24 (32
mmol, 2.83 g) in THF (50 ml) was added dropwise to the mix-
ture which was then kept for 1 h at Ϫ78 ЊC. After this the mix-
ture was allowed to warm to room temperature. Hydrolysis of
the mixture at 0 ЊC was effected with a THF–H2O mixture (1:1;
50 ml). Work-up and base extraction (10% aqueous NaOH)
gave 1229 (400 mg, 30%).
4-(2-Hydroxyethoxy)-2,6-dimethylphenol 12.29 δH 2.25 (s, 6H,
CH3), 3.95, 4.10 (m, 4H, CH2OAr ϩ CH2OH) and 6.40 (s, 2H,
Ar).
9 Ph. Gros, Y. Fort, G. Queguiner and P. Caubère, Tetrahedron Lett.,
1995, 36, 4791.
10 E. W. Thomas, J. Org. Chem., 1986, 51, 2184.
11 (a) D. L. Comins and D. La Muyon, Tetrahedron Lett., 1988, 29,
773; (b) F. Trécourt, M. Mallet, F. Marsais and G. Quéguiner, J. Org.
Chem., 1988, 53, 1367.
12 (a) P. Caubère and B. Loubinoux, Bull. Soc. Chim. Fr., 1969,
2483; (b) P. Caubère and G. Coudert, Bull. Soc. Chim. Fr., 1971,
2234.
13 P. Beak, Chem. Rev., 1984, 84, 471 and references cited therein.
14 (a) G. Queguiner, F. Marsais, V. Snieckus and J. Epsztajn, Adv.
Heterocycl. Chem., 1991, 52, 187 and references cited therein;
(b) N. Plé, A. Turck, K. Couture and G. Queguiner, J. Org. Chem.,
1995, 60, 3781.
15 (a) H. C. Brown, J. Am. Chem. Soc., 1945, 2, 1452; (b) H. C. Brown,
H. Pearsall, J. Am. Chem. Soc., 1945, 2, 1765.
Attempted second metallation of 2a
2-Methoxypyridine (8 mmol, 870 mg) in hexane (5 ml) was
added dropwise to a solution of BuLi (8 mmol, 5 ml) at 0 ЊC;
after 0.5 h, 2a was quantitatively formed (checked by deuteri-
ation). The solution of 2a was transferred via a nitrogen-
flushed line into freshly prepared BuLi–LiDMAE base (2
equiv.) at 0 ЊC in hexane. After 1 h at this temperature, chloro-
trimethylsilane (20 mmol, 2.2 g) was added to the mixture
which was then kept for 1 h at 0 ЊC. Work-up gave 2b (95%)
with no trace of 4b.
J. Chem. Soc., Perkin Trans. 1, 1997
3079

View Article Online
16 (a) M. Majewski and D. M. Gleave, J. Org. Chem., 1992, 57, 3599;
(b) M. Imai, A. Hagihara, H. Kawasaki, K. Manabe and K. Koga,
J. Am. Chem. Soc., 1994, 116, 8829; (c) B. H. Lipshutz, M. Wood
and C. Lindsley, Tetrahedron Lett., 1995, 36, 4385.
17 B. H. Lipshutz and S. Sengupta, Org. React., 1992, 41, 135.
18 P. Caubère and B. Loubinoux, Bull. Soc. Chim. Fr., 1968, 9, 3857.
19 P. Caubère, Pure Appl. Chem., 1985, 57, 1875.
24 G. Guillaumet, V. Lemmel, G. Coudert and P. Caubère, Tetrahedron,
1974, 30, 1289.
25 (a) M. C. Carré, G. Ndebeka, A. Riondel, P. Bourgasser and
P. Caubère, Tetrahedron Lett., 1984, 25, 1551; (b) Ph. Gros,
Ph. Hansen and P. Caubère, Tetrahedron, 1996, 52, 15147.
26 M. Franck-Neumann, M. Miesch and F. Barth, Tetrahedron, 1993,
49, 1409.
20 M. Perdicakis and J. Bessière, C. R. Acad. Sci. Paris, Série II, 1982,
879.
27 R. Held, F. Dietz and P. Thomas, Z. Chem., 1972, 12, 346.
28 S. Fukuzawa, S. S. Sakai, Bull. Chem. Soc. Jpn., 1992, 65, 3308.
21 G. Newkome and D. Hager, J. Org. Chem., 1982, 47, 600.
22 J. Chaudhuri, S. Kume, J. Jagur-Grodzinski and M. Szwarc, J. Am.
Chem. Soc., 1968, 90, 6421.
23 (a) D. Liotta, M. Saindane and L. Waykole, J. Am. Chem. Soc.,
1983, 105, 2923; (b) D. Liotta, J. Arbiser, J. W. Short and
M. Saindane, J. Org. Chem., 1983, 48, 2932.
Paper 7/01914I
Received 18th March 1997
Accepted 29th May 1997
3080
J. Chem. Soc., Perkin Trans. 1, 1997