The mechanism of lithiation and nitrile insertion reactions of β-methylazines: Evidence from the structure of 3-C5H4NCH=C(Ph)N(H)C(Ph)=NLi · PMDETA

Article abstract of DOI:10.1016/S0022-328X(97)00553-6

A 1:1:1 reaction of 3-methylpyridine, 1, with LDA and PhCN in the presence of PMDETA gives the title complex 6, shown by X-ray crystallography to be the first monomeric iminolithium. The isolation of 6, and of a cyclised product 8 when a 3:10:3 reaction o

Full text of DOI:10.1016/S0022-328X(97)00553-6

Ž
.
Journal of Organometallic Chemistry 550 1998 457–461
Short communication
The mechanism of lithiation and nitrile insertion reactions of
b-methylazines: evidence from the structure of
ž / ž / ž /
3-C5H4 NCH5C Ph N H C Ph 5NLiPPMDETA
1
)
Sarah C. Ball, Robert P. Davies, Paul R. Raithby, Gregory P. Shields, Ronald Snaith
Department of Chemistry, UniÕersity of Cambridge, Lensfield Road, Cambridge CB2 1EW, UK
Received 7 April 1997
Abstract
A 1:1:1 reaction of 3-methylpyridine, 1, with LDA and PhCN in the presence of PMDETA gives the title complex 6, shown by X-ray
crystallography to be the first monomeric iminolithium. The isolation of 6, and of a cyclised product 8 when a 3:10:3 reaction of
1:LDA:PhCN is carried out, suggest a mechanism for previously published cyclisation reactions of b-methylazines. q 1998 Elsevier
Science S.A.
Keywords: Lithium; Reaction mechanisms; Nitriles; X-ray structure
1. Introduction
only moderate even under forcing conditions: in this
specific case, 63% after heating at 1008C for 36 h.
Ž
Better yields under milder conditions 90% after heating
Several of our recent papers have concerned the
isolation and structural characterisation of lithium-con-
taining species present during ‘one-pot’ organic synthe-
.
Ž
at 408C for 4 h are gained when an excess two-fold or
more of lithiating reagent is used Scheme 1, bottom
equation . Such improvement was attributed 6 to for-
mation of a dilithiated intermediate 2a.
.
Ž
.
w x
w
x
ses involving lithiations 1–5 . In all cases, the origi-
nally proposed reaction mechanismrsequence involved
the generation of both mono- and di-lithiated organic
molecules. However, so far the latter species have
proved very elusive. For example, our most recent study
w x
5 examined briefly a versatile protocol whereby b-
Ž
.
methylazines pyridines, quinolines, pyrimidines etc.
undergo lithiation and nitrile insertion, then ring closure
w x
6 . Scheme 1 shows the reaction sequence proposed in
the original paper, taking 3-methylpyridine 1 as the
precursor. Monolithiation top equation and benzoni-
trile insertion, followed by thermolysis and work-up,
give the final organic product 4. However, yields are
Ž
.
)
Corresponding author.
1 Dedicated to Professor Ken Wade on the occasion of his 65th
birthday and in recognition of his outstanding contributions to Chem-
istry. R.S. in particular thanks Ken, his erstwhile PhD supervisor, for
his strong support and valued friendship over many years.
Scheme 1.
0022-328Xr98r$19.00 q 1998 Elsevier Science S.A. All rights reserved.

(
)
458
S.C. Ball et al.rJournal of Organometallic Chemistry 550 1998 457–461
w x
In our preliminary study 5 we could not isolate such
lent of PhCN is not 2, but rather it is 5. This is
significant in that the C5C bond of the pyrrole ring in
the final product 4 is formed already in 5.
Ž
a species although dilithiation of the related but much
.
more acidic 2-methylpyridine followed by PhCN inser-
tion did give in high yield a dilithiated product: a
complicated Li12 aggregate containing two types of
dianion, viz, C5H4 NPCHC Ph N2yP2Li 2 P C5H4 N
w
Ž
.
qx
w
yŽ
.
Ž
.
qx
y
PCH LiPTHF PC Ph N Li 4. Here we describe the
results of a more detailed study of the reactions shown
in Scheme 1. The major conclusions are: i that the
Ž .
monolithium species 2 is not that formulated in Scheme
Ž
.
1
and below but rather is probably C5H4 N P
Ž
.
CH5C Ph NHLi, 5, i.e., the C5C bond apparent in the
five-membered ring of the final product 4 is formed at
this stage; ii that dilithium species 2a and 3a are not
Ž .
The identity of 6 was confirmed by X-ray crystallog-
Ž
.
isolable as solid materials even if they might be present
in the reaction solution; and iii , from i and ii , that
product 4 results from cyclisation then work-up of the
raphy. The molecular structure Fig. 1 is monomeric
and shows the features apparent from solution NMR
spectroscopy. The single Li centre is approximately
Ž
.
Ž .
Ž .
Ž
w
Ž .
Ž .
monolithium species 5 whose efficient production and,
tetrahedrally coordinated by an imino N N 1 –Li 1
˚
x
.
Ž .
disordered PMDETA ligand mean of six N–Li dis-
possibly, whose cyclisation depends upon the presence
distance, 1.963 8 A and by the three N atoms of a
Ž
.
w
of excess two- or more-fold LDA.
˚
x
Ž
.
tances, 2.124 10 A . The central backbone of the
molecule consists of a zig-zag-NC Ph N H C Ph C H –
Ž
. Ž . Ž . Ž .
Ž . Ž .
unit involving as the key atoms N 1 C 1 -
Ž . Ž . Ž
.
N 3 C 14 C 15 . Although the bonding within this unit
Ž . Ž . Ž .
can be represented formally by N 1 5C 1 –N 3 –
Ž
.
Ž
.
w
x
C 14 5C 15 see structural formula 6 it is apparent
from the bond lengths that there is considerable delocal-
Ž . Ž .
isation along the backbone; these lengths are N 1 –C 1
Ž . Ž . Ž . Ž . Ž . Ž .
Ž
.
1.320 5 , C 1 –N 3 1.331 5 , N 3 –C 14 1.381 5 and
˚
Ž .
Ž
.
Ž
.
C 14 –C 15 1.352 5 A. The C–N distances can be
compared with those found within the pyridyl ring
mean 1.326 6 A and the C–C one with those dis-
2. Results and discussion
˚
x
w
Ž .
w
Ž .
Initially we attempted to synthesise a monolithium
complex of type 2 by reacting 3-methylpyridine 1 with
one equivalent each of LDA and PhCN in THF as
tances found in the two phenyl groups mean 1.383 7
2
˚
x
A . The essentially sp and planar natures of the imino
Ž .
Ž
.
carbon atom C 1 and the alkenyl carbon atom C 14
Ž
.
solvent Scheme 1, top equation . Cooling of the reac-
tion solution over several days failed to afford crystals
even after removal of much of the THF. Accordingly,
the reaction was repeated in the presence of one equiva-
lent of an alternative Lewis base, PMDETA
are further confirmed by the summations of the angles
around them, 360.08 and 359.38 respectively.
It was noted above that 6 could be considered as an
Ž .
iminolithium complex of type Ph R C5NLiPPMDETA
where
R is the alkenylamino-group 2-C5 H4 N P
w
Ž
. x
Ž
. Ž .
MeN CH2CH2 NMe2 . Refrigeration of this reaction
solution over a week then gave crystals of 6 in reason-
CH5C Ph N H –. As such it is the first example—al-
beit a somewhat elaborate one—of a monomeric imino-
lithium species. In this context, it is pleasing and appro-
priate to note that the first structurally-characterised
iminolithium, the uncomplexed hexamer Bu2C5NLi ,
was reported in 1979 by Shearer et al. 7 . Since then,
further hexameric structures have been revealed 8,9 as
have pseudo-cubane tetrameric complexes such as
2
Ž
.
able yield 51% . Complex 6 was characterised initially
by elemental analyses and by 1H, 13C NMR spec-
troscopy. The species has several key and interesting
features. Firstly, two equivalents of PhCN have been
inserted into lithiated 1, 2-C5H4 NCH2 Li; the implica-
tions are that lithiation is slow andror incomplete and
that the nitrile insertion reactions are much faster. Sec-
ondly, as shown by NMR spectroscopy, 6 contains a
t
Ž
.
6
w x
w
x
Ž
Ž
.
Ph2C5NLiPPyr and dimeric complexes such as
4
t
.
w
x
Bu2C5NLiPHMPA 8–11 .
2
Ž
.
Ž
.
–CH5C Ph NH– unit rather than a –CH2–C Ph 5N–
one, as in 7. Put another way, while both 6 and 7 can be
considered to be iminolithium species of type
In the light of the above results on the monolithiation
of 1 and subsequent double PhCN insertion, we have
examined more closely the course of the supposed
dilithiation of 1 Scheme 1, bottom equation . Treat-
ment of 1 in THF with LDA and then PhCN 3:5:3
equivalents , followed after 1.5 h by five further equiva-
Ž
.
Ž .
Ž
.
Ph R C5NLi, in 6 R is an alkenylamino-group whereas
Ž
in 7 it is an imino-group. This suggests that the product
of lithiation of 1 followed by insertion of one equiva-
.

(
)
S.C. Ball et al.rJournal of Organometallic Chemistry 550 1998 457–461
459
w
Ž
.
Ž .
Fig. 1. Molecular structure of 6 hydrogen atoms other than H 3a and H 15 , and the disorder within the PMDETA molecule, are omitted for
x
clarity .
lents of LDA gives a clear solution. No crystalline
material could be isolated from this solution even after
prolonged refrigeration. Detailed NMR experiments re-
main to be done but certainly so far we have no
evidence for the formation of a dilithiated species such
as 2a. However, heating of this reaction solution at
408C for 4 h did, after cooling, give a monolithium
product, 8, in 60% yield. Its crystal structure is awaited
depend on the rate and extent of production of lithiated
1 in step i . If this species is produced only slowly or
Ž .
incompletely, then in effect PhCN is present in excess
and 6 will be the dominant product. Herein, we believe,
lies the explanation for the efficacy of using two or
more equivalents of LDA in the first stage of these
syntheses. The excess of LDA is not there to produce
1
but its empirical identity has been proved by H, 13C
NMR spectroscopy. Most importantly, 8 is an obvious
Ž
.
immediate forerunner cf., 3, 3a in Scheme 1 of the
organic product 4. Indeed, hydrolysis of 8 followed by
work-up gives 4 in 90% yield. In reverse, lithiation of 4
by Bun Li reproduces 8.
Tying together all these results, our preliminary con-
wŽ .x
i
clusions are expressed in Scheme 2. The first step
is clearly the monolithiation of 1. The product can insert
Ž .
w
x
one equivalent of PhCN to give 5 step ii or, as we
w
have shown, it can insert two PhCN units to give 6 step
Ž
.x
iii . Which of these processes dominates will clearly
Scheme 2.

(
)
460
S.C. Ball et al.rJournal of Organometallic Chemistry 550 1998 457–461
1H relative to C5H4 N and 2=Ph resonances . 13C
Ž
.
.
dilithiated species such as 2a Scheme 1 , for which we
can find no evidence. Rather, it ensures the more rapid
andror more extensive production of lithiated 1. Evi-
Ž
.
NMR spectrum 400 MHz, DMSO alkenyl CH carbon
Ž
.
resonance at d 103.7 ppm H odd by APT scans .
1
dencing this, the H NMR spectrum of a 1:1 mixture of
1 and LDA in d8-THF at 258C indicates a mixture of 1
and lithiated 1 ca. 50% of lithiated 1 , whilst a similar
3.2. Crystallographic data for 6
Ž
.
spectrum of a 3:5 mixture shows more complete mono-
lithiation of 1 ca. 75% . Efficient production of lithi-
ated 1 will lead to 5, not 6, as the dominant insertion
C29 H39 LiN6; Ms478.60, colourless crystals, di-
mensions 0.41=0.34=0.22 mm, monoclinic, P21rc
Ž
.
˚
Ž .
Ž
.
Ž .
Ž .
No. 14 , as10.21 2 , bs15.81 3 cs17.76 4 A,
product. The key feature of our precise formulation of 5
3
bs96.8 2 ; Vs2849 10 A , Dcalc s1.116 Mgrm3,
˚
.
Ž .
Ž
<
Ž
.
Zs4, F 000 s1032, graphite-monochromated Mo-Ka
is its –HC5C –NH– unit which sets it up perfectly for
ring closure to give, eventually, the final organic prod-
uct 4. Such formulation, and the presence of such a unit,
is in part inferred from the structure of 6. However,
further to this inference, we can cite the fact that
treatment of 2-methylpyridine with LDA in the presence
y1
˚
Ž
.
radiation, ls0.71073 A, m Mo-Ka s0.067 mm
,
Ž .
Ts153 2 K. Data were collected on a Stoe four-circle
diffractometer equipped with an Oxford Cryostream
crystal cooling apparatus. A total of 6113 reflections
Ž
.
3692 independent, Rint s0.065 were collected in the
of TMEDA Me2 N CH2 2 NMe2 and then ButCN af-
range 7.08FuF22.508. Structure solved by direct
w
Ž
.
x
t
Ž
.
methods SHELXTL-PLUS and refined with all non-
hydrogen atoms anisotropic by full-matrix least squares
Ž
.
fords not 2-C5H4 NPCH2C Bu 5NLiPTMEDA but
t
2
Ž
.
rather 2-C5H4 NPCH5C Bu NHLiPTMEDA.
The
based on F2 SHELXL-93 ; H atoms were included in
Ž
.
final uncertain and largely unexplored stage is precisely
w
Ž
.
Ž
.
Ž
.
idealised positions whilst H 15 on C 15 and H 3a on
N 3 could be located in the difference electron density
map their coordinates were not refined freely . The
w
Ž .x
how 5 cyclises on heating to give 8 step iv . In
Scheme 2, we imply that this step formally involves
elimination of dihydrogen. However, it is more conceiv-
able that excess LDA still in the system lithiates 5 at the
N–H position and that the resulting geminal –NLi2
Ž .
x
PMDETA ligand is positionally disordered over two
different sites and was refined with partial occupancies
and with appropriate restraints to positional and dis-
placement parameters. The final cycle of refinement
included 310 parameters with unweighted R1 s0.0714
Ž
)
species or even a Õicinal one such as 3a is the
immediate precursor to complex 8, formed by LiH
elimination.
Ž .
and weighed wR2 s
on 2694 data with I)2s
I
0.1939 on all data. G.o.f. 1.062, weighing scheme wy1
2
2
2
w
2Ž
x
.
Ž
.
x
w
s s
F
q 0.0589P q3.04P where Ps Fo q
3. Experimental
2
2F r3 . Highest peak in final difference map 0.575 e
c
y3. Atomic coordinates, bond lengths and angles, and
˚
A
3.1. Synthesis and characterisation of 6
vibrational parameters have been deposited at the Cam-
bridge Crystallographic Data Centre. This information
may be obtained, on request, from the Director, Cam-
bridge Crystallographic Data Centre, 12 Union Road,
Cambridge, CB2 1EZ, UK.
Ž
.
3-methylpyridine 0.48 ml, 5 mmol was added to a
Ži
.
chilled solution of LDA Pr2 NLi 0.535 g, 5 mmol in
THF 8 ml . The resulting yellow suspension was chilled
to y788C and PhCN 0.51 ml, 5 mmol was added. The
mixture was stirred for 1.5 h at 08C, resulting in a clear
red-brown solution. The THF was removed to produce
an orange powder which was recrystallised from
PMDETA 0.7 ml, 5 mmol and toluene 8 ml . Refrig-
eration for 1 week afforded orange crystals of 6
C5H4 N CHC Ph NHC Ph NLiPPMDETA. First batch
yield 0.60 g 51% . Analysis for 6 C29 H39 N6 Li. Found
Ž
.
Ž
.
3.3. Synthesis and characterisation of 8
Ž
.
Ž
.
Ž
.
3-methylpyridine 3 mmol, 0.3 ml was added to a
Ž
.
solution of LDA 5 mmol . On stirring a yellow suspen-
Ž
.
Ž
.
.
Ž
.
Ž
sion resulted. This was chilled to y788C and PhCN 3
Ž
.
mmol, 0.3 ml added. The mixture was stirred at 08C for
C72.44, H8.22, N17.72, Li 1.39%; Calc. C72.75, H
8.22, N17.56, Li 1.47%. H NMR spectrum 250 MHz,
DMSO alkenyl CH proton d 5.72 ppm s, integration
1.5 h resulting in a clear red-brown solution. A further
1
Ž
Ž
.
portion of LDA solution 5 mmol was then added and
the solution heated for 4 h at 408C. The resulting red
solution was filtered to remove a small amount of fine
precipitate. Refrigeration for two days afforded orange
crystals of 8, C21H25LiN2O2. First batch yield 0.72 g
.
Ž
Ž
.
41% , mp 127–1298C. Analysis found C72.70, H7.30,
N8.26, Li1.84%. Calc. C73.21, H7.32, N8.10, Li2.04%.
2 The crystal structure of this complex has been solved; it is a
polymer with NLi rings linked together by TMEDA molecules.
Ž
.
2
1
Ž
.
Ž
.
H NMR 250 MHz, DMSO d 7.96 dd, 2H, Ph , 7.81
dd, 1H, Pyr , 7.51 dd, 1H, Pyr , 7.27 dd, 2H, Ph ,
S.C. Ball, R.P. Davies, P.R. Raithby, R. Snaith, unpublished observa-
Ž
.
Ž
.
Ž
.
tions.

(
)
S.C. Ball et al.rJournal of Organometallic Chemistry 550 1998 457–461
461
Ž
.
Ž
.
Ž
.
7.07 t, 1H, Ph , 6.54 s, 1H , 6.48 dd, 1H, Pyr , 3.60
References
13
Ž
.
Ž
.
Ž
m, 8H, THF , 1.77 m, 8H, THF . C NMR 100.6
MHz, DMSO 159.3 H even , 149.1 H even , 140.2
w x
1
S.C. Ball, I. Cragg-Hine, M.G. Davidson, R.P. Davies, A.J.
Edwards, P.R. Raithby, R. Snaith, Angew. Chem. Int. Ed. Engl.
.
Ž
.
Ž
.
.
Ž
Ž
.
Ž
.
Ž
.
H even , 137.5 H odd , 127.7 H odd , 125.2 H odd ,
Ž
.
34 1995 921.
Ž
.
Ž
.
Ž
.
Ž
124.6 H odd , 124.4 H even , 123.3 H odd , 110.0 H
w x
2
S.C. Ball, I. Cragg-Hine, M.G. Davidson, R.P. Davies, P.R.
Raithby, R. Snaith, J. Chem. Soc., Chem. Commun. 1996
.
Ž
.
Ž
.
Ž
.
odd , 91.7 H odd , 66.9 H even , 25.1 H even .
Ž
.
4 was prepared via hydrolysis of a solution of 8,
1581.
w x
3
M.G. Davidson, R.P. Davies, P.R. Raithby, R. Snaith, J. Chem.
Soc., Chem. Commun. 1996 1695.
R.P. Davies, P.R. Raithby, R. Snaith, Organometallics 15 1996
4355.
S.C. Ball, J. Cobb, R.P. Davies, P.R. Raithby, G.P. Shields, R.
Snaith, J. Organomet. Chem. 534 1997 241.
M.L. Davis, B.J. Wakefield, J.A. Wardell, Tetrahedron 48
followed by extraction and purification as described in
1
Ž
.
w x
Ž
.
Ref. 6 . mp 209–2108C, H NMR 250 MHz DMSO ,
w x
4
Ž
.
Ž
.
Ž
.
Ž
.
d 12.16 s, 1H, NH , 8.20 dd, 1H , 7.95 m, 3H , 7.47
dd, 2H , 7.39 t, 1H , 7.09 dd, 1H , 6.94 s, 1H .
Reaction of 4 with one equivalent of BuLi in a 1:1
THF:hexane solvent afforded crystals of 8 in moderate
yield.
Ž
.
Ž
.
Ž
.
Ž
.
w x
5
n
Ž
.
w x
6
Ž
.
1992 939.
w x
7
H.M.M. Shearer, K. Wade, G. Whitehead, J. Chem. Soc.,
Ž
.
Chem. Commun. 1979 943.
D. Barr, R. Snaith, W. Clegg, R.E. Mulvey, K. Wade, J. Chem.
Soc., Chem. Commun. 1986 , 295.
D. Barr, R. Snaith, W. Clegg, R.E. Mulvey, K. Wade, J. Chem.
w x
8
Ž
.
Acknowledgements
w x
9
Ž
.
Soc., Dalton Trans. 1987 1071.
Ž
.
We thank the EPSRC S.C.B., R.P.D., G.P.S. , the
w
w
x
x
Ž
.
10 R.E. Mulvey, Chem. Soc. Rev. 20 1991 167.
11 K. Gregory, P.v.R. Schleyer, R. Snaith, Adv. Inorg. Chem. 37
Ž
.
Ž
.
CCDC G.P.S. , the Associated Octel Co. Ltd. S.C.B.
and the Royal Society low-temperature X-ray diffrac-
tion apparatus, P.R.R. for financial support.
Ž
Ž
.
1991 47.
.