Metal 4-alkylidene-4H-pyridin-1-ides and 2H-imidazol-4-ones from novel highly N-(pyridin-4-yl)methyl-substituted azomethines

Article abstract of DOI:10.1002/ejoc.200400460

Novel azomethines 3 synthesized from 4-(aminomethyl)pyridine (1) and various ketones 2 exhibit a higher acidity of the Ca-protons compared with other 4-alkylpyridines like 4-benzylpyridine or 4-(tosylmethyl)pyridine. Therefore, they can easily be metalated to give a variety of specific anions 3(-), which deserve interest for further synthetic applications. Important properties of compounds 3 and their lithiated derivatives 4 (lithium 4H-pyridin-1-ides, synthesized by deprotonation of the α-CH₂ group in 3) can be controlled by the carbonyl reactant 2. These carbanions 3(-) are remarkable intermediates regarding their NMR properties. Comparing the NMR spectra of the lithium 4H-pyridin-1-ides 4 with their parent azomethines 3 reveal very distinct changes in the chemical shifts from which important structural and electronic properties of the carbanions can be deduced. The observation of E/Z isomers in the case of 3c and their mutual conversion at higher temperature inspire DFT calculations on the B3LYP/6-311++G(d,p) level of theory to get insight into the energetic demand of such isomerization processes. Moreover, calculations on the same level of theory were carried out on the model azomethine anion 5 and different monomeric lithium coordination isomers 6a-d. Experimentally, two selected examples will exemplify the reactivity of 4H-pyridin-1-ide anions towards heterocumulenes. The 2-(pyridin-4-yl)acetate 7 obtained from the reaction of 4e with carbon dioxide appears to be a CO₂ storage molecule as it easily releases CO₂ under protolysis conditions comparable with the vitamin B₆ reaction mode. Interesting five-membered heterocycles were isolated after reaction of 4 with isocyanates followed by hydrolysis and oxidation. Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2004.

Full text of DOI:10.1002/ejoc.200400460

FULL PAPER
Metal 4-Alkylidene-4H-pyridin-1-ides and 2H-Imidazol-4-ones from Novel
Highly N-(Pyridin-4-yl)methyl-Substituted Azomethines
Diana Hampe,[a] Wolfgang Günther,[a] Helmar Görls,[b] and Ernst Anders*[a]
Keywords: Metalated azomethines / Heterocumulene fixation / Carboxylation / DFT calculations
Novel azomethines 3 synthesized from 4-(aminomethyl)pyri-
dine (1) and various ketones 2 exhibit a higher acidity of the
Cα-protons compared with other 4-alkylpyridines like 4-
benzylpyridine or 4-(tosylmethyl)pyridine. Therefore, they
can easily be metalated to give a variety of specific anions
3(−), which deserve interest for further synthetic applica-
tions. Important properties of compounds 3 and their lithiated
conversion at higher temperature inspire DFT calculations on
the B3LYP/6-311++G(d,p) level of theory to get insight into
the energetic demand of such isomerization processes. More-
over, calculations on the same level of theory were carried
out on the model azomethine anion 5 and different mono-
meric lithium coordination isomers 6a−d. Experimentally, two
selected examples will exemplify the reactivity of 4H-pyri-
derivatives 4 (lithium 4H-pyridin-1-ides, synthesized by de- din-1-ide anions towards heterocumulenes. The 2-(pyridin-
protonation of the α-CH2 group in 3) can be controlled by the
carbonyl reactant 2. These carbanions 3(−) are remarkable
4-yl)acetate 7 obtained from the reaction of 4e with carbon
dioxide appears to be a CO2 storage molecule as it easily
intermediates regarding their NMR properties. Comparing releases CO2 under protolysis conditions comparable with
the NMR spectra of the lithium 4H-pyridin-1-ides 4 with their
parent azomethines 3 reveal very distinct changes in the
chemical shifts from which important structural and elec-
the vitamin B6 reaction mode. Interesting five-membered
heterocycles were isolated after reaction of 4 with isocyan-
ates followed by hydrolysis and oxidation.
tronic properties of the carbanions can be deduced. The ob- ( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
servation of E/Z isomers in the case of 3c and their mutual
Germany, 2004)
Introduction
Further, in biological processes of the amino acid metab-
olism azomethines also play an important role as key inter-
mediates (Scheme 1): In reactions mediated by the deriva-
tives of the vitamin B6 complex, the formation of imines
from amino acids and pyridoxal or α-keto acids and pyri-
doxamine, respectively, is the precondition for build-up and
decomposition of amino acids, α-keto acids and biogenic
amines by transamination or decarboxylation.[12]
Imines, azomethines or Schiff bases Ϫ synonyms for one
and the same species Ϫ are condensation products of pri-
mary amines with aldehydes or ketones.[1,2] Since azome-
thines are known they were introduced in a wide range of
chemistry.
Their great synthetic importance reflects in a broad appli-
cation: as protecting group for the CϭO double bond or
the amine function, as chiral auxiliaries in asymmetric sub-
stitution reactions of amino acids,[3] as reagent for the
quantitative transfer of aldimines into aza-enolates or the
synthesis of primary and secondary amines by reduction of
the CϭN double bond,[4] as ligands for various metals and
the use of the resulting complexes as catalysts (e.g. polymer-
ization, epoxidation of alkenes).[5Ϫ10] Azomethine ylides are
widely used as 1,3-dipoles in cycloaddition reactions with
several multi-bond systems like alkenes, acetylenes or car-
bonyls to give a great variety of different 5-ring hetero-
cycles.[11]
The 2-[(pyridin-4(1H)-ylidenemethyl)imino] moiety (bold
in II, Scheme 1) Ϫ a 4-alkylidene-1,4-dihydropyridine sys-
tem extended by an imino function at the Cα center Ϫ is
proposed as conceivable intermediate which facilitates the
rearrangement of the CϭN bond of the aldimine (structure
I, Scheme 1) to the ketimine (structure III, Scheme 1) and
the other way round by abstraction of a proton in α or αЈ
position as well as the loss of carbon dioxide (not shown in
Scheme 1; instead of a proton, CO2 is abstracted then).
The 4-alkylidene-1,4-dihydropyridine system is an object
of thorough investigations by our work group since a longer
time. Simple 4-alkylpyridines like 4-CH3-, 4-iPr- or 4-Bn-
pyridine as well as more functionalized derivatives like 4-
halomethyl-, 4-Me3SiCH2- or 4-(RSO2CH2)-pyridines were
subjected to N-alkylations, N-acylations, deprotonations
and subsequent reactions with various electrophiles.[13aϪ13h]
In connection with investigations of several group transfer
reagents, N-acyl-4-alkylidene-1,4-dihydropyridines and N-
[a]
Institut für Organische Chemie und Makromolekulare Chemie
der Friedrich-Schiller-Universität,
Humboldtstrasse 10, 07743 Jena, Germany
[b]
Institut für Anorganische und Analytische Chemie der
Friedrich-Schiller-Universität,
Lessingstrasse 8, 07743 Jena, Germany
Fax: (internat.) ϩ49-(0)3641-948212
E-mail: Ernst.Anders@uni-jena.de
Eur. J. Org. Chem. 2004, 4357Ϫ4372
DOI: 10.1002/ejoc.200400460
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4357

D. Hampe, W. Günther, H. Görls, E. Anders
FULL PAPER
Scheme 1. 1,4-Dihydropyridine intermediate (bold, structure II) in the vitamin-B6-catalyzed transamination reactions
acyl-4-alkylpyridinium salts emerge as useful tools in syn-
thesis.[13i]
Structural and electronic backgrounds and explanations
are given on the fundament of some high level DFT calcu-
The interesting properties of the 1,4-dihydropyridine in- lations which include the starting imines 3 Ϫ especially,
termediates of amino acid aldimines in pyridoxal catalyzed their Z/E stereoisomerism Ϫ as their anions 3(؊) and lithi-
reactions (Scheme 1) inspired us to investigate the chemistry ated counterparts 4 as well. They are summarized in the
of the related azomethines 3 from 4-(aminomethyl)pyridine chapter DFT calculations.
(1) and the carbonyl compounds 2 as shown in Scheme 2
A brief outlook should convey an insight into interesting
and 3. Moreover, the properties of the carbanions 3(؊) de- reactivity properties of 4 towards heterocumulenes.
rived from 3, especially, their potential ability to fixate and
activate heterocumulenes, and their suitability to act as car-
bon dioxide transfer reagents in accordance to natural
Results and Discussion
models like biotin or the above mentioned vitamin B6 will
be part of this paper (Scheme 2 gives an impression of
alternative reaction paths).
Azomethines
Furthermore, the introduction of the CϭN double bond
The synthesis of the azomethines 3 is a typical conden-
dilates the reaction diversity and opens the possibility to sation reaction between 4-(aminomethyl)pyridine (1) and
interesting cyclization reactions similar to 1,3-dipolar cyclo- ketones 2 (Scheme 3) which tolerates a wide variety of sub-
additions.
stituents R1 and R2 of the carbonyl compounds 2. Thus,
Scheme 2. Possible carbon dioxide adducts and their reaction modes as observed in natural processes. a) CO2 transfer comparable with
the biotin mode via a carbamate intermediate, b) Reaction mode resembling vitamin B6 (to give the same intermediate as in c), c)
Vitamin-B6-like way of reaction
4358
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 4357Ϫ4372

Metal 4-Alkylidene-4H-pyridin-1-ides and 2H-Imidazol-4-ones
FULL PAPER
the reactivity and electronic properties of 3 and their carb- quadruplet for the methine proton nor a doublet of the cor-
anions 3(؊) can easily be adjusted to novel synthetic tasks responding methyl group (aldimine).
(Scheme 3).
An in-depth structure determination was carried out by
two dimensional NMR experiments (HMQC, HMBC,
COSY, NOESY) in CDCl3 at 0 °C. The exact configuration
1
assignment could be realized by a two dimensional H,1H
NOE spectroscopy (0 °C, CDCl3). The interaction or non-
interaction among the protons of the α-methylene and the
methyl group is a reliable indicator for this differentiation:
a cross signal will only appear for the E isomer. This NMR
study showed, that the E isomer is present in much lower
concentration (Z: 83%, E: 17%). This finding goes parallel
with the result of our DFT calculations which predicted the
Z isomer to be 2.1 kcal/mol more stable than the E isomer.
Raising the temperature to 75 °C during the 1H NMR
measurement causes a change of the relative ratio from
83:17 to 60:40 as a thermally induced isomerization of the
CϭN bond takes place, presumably by inversion,[15,16]
which additionally proves the coexistence of E/Z isomers.
Further information concerning mechanism and activation
energy was gained by DFT calculations (vide infra).
Comparing the chemical shifts of the α-methylene pro-
tons, protons H3/5 and H2/6 of both isomers of 3c with
those of the other imines 3a,bϪe (Table 1) indicates that the
corresponding signals of Z-3c are shifted upfield while the
signals of E-3c are in a similar range as the other imines.
Furthermore, a clear tendency is recognizable for the chemi-
cal shift differences between the protons Hα, H3/5 and H2/
6 of Z-3c and E-3c: the difference becomes smaller with
greater spatial distance from the naphthyl substituent Ϫ
∆δ ϭ 0.6 ppm, 0.28 ppm and 0.1 ppm, respectively, i.e., the
naphthyl residue causes a slight shielding of these protons
in Z-3c. Thus, the 4-pyridylmethyl moiety must be orien-
tated above or beneath the naphthyl ring plane so that the
protons Hα, H3/5, H2/6 are affected by the ring current in
decreasing strength with increasing distance. A further hint
on the three-dimensional arrangement gives the NOESY
spectrum where crosspeaks exist between the methyl group
and the protons H10 and H16, respectively. Hence, the aro-
matic π system of the naphthyl substituent cannot be in
the same plane with the π bond of the CϭN double bond
(otherwise only one crosspeak would be observable). Both
these presumptions are confirmed by an X-ray crystal struc-
ture analysis of the Z isomer of 3c (Figure 2).
Scheme 3. Synthesis of azomethines 3 from 4-(aminomethyl)pyrid-
ine (1) and ketones 2
From the standard procedure applied,[14] 3aϪe are ob-
tained in yields between 50 and 65% as colorless or pale
yellow crystal solids or powders.
With one exception (2e), the ketone reagents 2 carry at
least one aromatic substituent as such moieties stabilize the
CϭN double bond due to conjugation with the aromatic π
system.[2,14] Nevertheless, compounds 3 are in general sensi-
tive towards moisture and show a distinct tendency for de-
composition depending on the nature of R1 and R2 (cf. the
Exp. Sect.). Therefore, it is recommended to store them un-
der an inert gas atmosphere. Analogous aldimines 3 were
not sufficiently stable. They polymerized to brown, highly
viscous products.
E/Z Stereoisomerism
Due to the unsymmetrically substituted CϭN double
bond the imines 3bϪe can exist as E/Z stereoisomers. Inter-
estingly, the appearance of these two isomers is only detect-
able for 3c (Scheme 4) by NMR: Two signal sets (Table 1,
Exp. Sect.) can be observed in 1H and 13C spectra in a ratio
of about 83:17 (Figure 1). An equilibrium between the ket-
imine and its tautomeric aldimine in solution which differ
from each by the position of the CϭN double bond can be
The dihedral angle for imine-N (Nim), C7, C9 and C10
amounts to Ϫ96.1°, i.e., the CϭN double bond does not
profit from the stabilizing effect of the conjugation with the
π system of the naphthyl residue. The planes spanned by
Cα, Nim, C7 and Nim, C7, C8, respectively, enclose an di-
hedral angle of 178.2°, so these atoms almost lie in one
plane. The dihedral angle between the pyridine ring and
the C4ϪCαϪNim plane is 61.7°. The naphthyl and pyridine
rings are situated on opposite sides of the CϭN plane and
point away from each other so that they have the largest
possible distance. Thus, the α-H atoms have different chemi-
cal surroundings. Therefore, the 1H NMR signal of the dia-
stereotopic α-methylene protons splits into four lines of an
1
excluded because the H NMR (Figure 1) shows neither a
Scheme 4. E/Z isomers of 3c (the numbering does not follow IU-
PAC rules, but should facilitate the description of the NMR results) AB system (Figure 1). Another interesting observation is a
Eur. J. Org. Chem. 2004, 4357Ϫ4372
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4359

D. Hampe, W. Günther, H. Görls, E. Anders
FULL PAPER
Figure 1. 1H NMR of E/Z-3c measured at 400 MHz in CDCl3 at 0 °C; the CH3 groups are split into triplets because of a homoallyl
coupling to the α-CH2 group
ments reveal only crosspeaks between CH2 and CH3 group
for all three imines 3b,d,e. Thus, they exist in the E configu-
ration.
Metalation of the Azomethines
Treatment of the azomethines 3 with various strong bases
like metal organyls (nBuLi, MgEt2), metal amides [LDA,
KN(SiMe3)2] or alkoxides (KOtBu) results in the metalated
structures (M ϭ Li: 4aϪe) after deprotonation of the α-
methylene group (Scheme 5) in yields up to 95% depending
on the metalating reagent.
The negative charge is strongly delocalized but of the
many possible mesomeric structures Scheme 5 shows only
those which mainly contribute to the resonance hybrid: The
4-alkylidene-1,4-dihydropyridine A from which the major
participation is assumed, the 2-aza-allyl anion B in which
the former positively polarized imine carbon is partially
Figure 2. Molecular structure of Z-3c with selected bond lengths
homoallyl coupling between the α-CH2 and CH3 group of negatively charged, and a 4-alkylidene-3,4-dihydropyridine
both isomers E/Z-3c. The CH3 signals split into triplets C which can be deduced from the calculated charge distri-
5
with a small coupling constant of J ϭ 1.28 Hz for Z-3c butions of E-3c(؊) and 5 (vide infra, Table 5, 9) as C3 and
5
and J ϭ 0.7Ϫ0.8 Hz for E-3c, respectively (Figure 1, en- C5 bear a considerable negative partial charge of Ϫ0.28 (cf.
largement of both triplets). This homoallyl coupling has the chapter DFT calculations). Of course, the effect of the
also been observed for the CH3 signal of 3b (5J ϭ 0.78 Hz). metal coordination on the electron distribution should not
Interestingly, in the case of the other unsymmetrically be neglected (vide infra).
substituted imines 3b,d,e the NMR spectra show only one
The substantially higher acidity of the α-methylene hy-
signal set what can be explained by formation of only one drogen atoms in comparison with that of the methyl hydro-
isomer in the synthesis, or a fast isomerization in regard to gen atoms in the azomethines 3bϪd (i.e. R1 ϭ CH3) follows
the NMR relaxation time takes place. But NOESY experi- our expectation as in the anions resulting from the Cα de-
4360
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 4357Ϫ4372

Metal 4-Alkylidene-4H-pyridin-1-ides and 2H-Imidazol-4-ones
FULL PAPER
Scheme 5. Deprotonation of the 4-[(dialkylmethylenamino)methyl]pyridines 3 in Cα position (coordination modes of the metal cations
are not considered)
protonation the charge is extremely well stabilized. If R1 to be the best model to describe the reaction behavior of
and/or R2 are aryl residues they additionally can extend the such lithiated ambident structures in solution.[18]
conjugated system in the anions 3(؊).
The common reaction pattern of the azomethine anions
While KOtBu, KN(SiMe3)2 or LDA are sterically de- 3(؊) towards electrophiles (heterocumulenes, aldehydes, es-
manding and therefore non-nucleophilic bases, the metal ters) is the attack on the Cα atom with or without sub-
organyls nBuLi and MgEt2 could act as nucleophiles and sequent reactions (cf. the chapter Reaction with heterocu-
transfer an alkyl residue onto the CϭN double bond or mulenes), which is in some contradictions with the calcu-
alkylate the pyridine ring in the 2-position. Fortunately, lated charge distribution of the anions E-3c(؊) and 5, and
these reactions have not been observed under the applied the lithium species 6aϪd (cf. chapter DFT calculations),
reaction conditions. No deprotonation of the Cα atom was where the pyridine nitrogen bears the main part of the
observed with ZnEt2 in contrary to nBuLi and MgEt2 what charge (qN: Ϫ0.49 to Ϫ0.88) in comparison to the Cα atom
can be correlated to the more ionic metalϪcarbon bond in (qC: Ϫ0.11 to Ϫ0.26). Nevertheless, the preceding forma-
lithium and magnesium organyls and thus the more basic tion of a weak bond between the pyridine nitrogen and the
alkyl residues.[17] Moreover, no alkylation of the CϭN electrophilic center cannot be excluded as this path would
double bond takes place with ZnEt2.
lead to a high reactive group transfer reagent which then
Depending on the residues R1 and R2, the compounds can perform the attack on the Cα atom.
4aϪe are intensively colored: deep red (4e), deep magenta
(4a,d) and deep blue violet (4b,c). The influence of the metal
NMR Investigations
cation on the color of the carbanions 3(؊) is only small;
the color of the corresponding potassium and magnesium
compounds is similar to that of the lithium complexes 4.
The potassium compounds K3(؊) were mainly synthesized
by deprotonation with KN(SiMe3)2 in diethyl ether, but the
use of KH or KOtBu is also possible. The magnesium com-
pounds Mg[3(؊)]2 are accessible with MgEt2 (THF, Ϫ20
°C). As selected examples K3c(؊) and Mg[3c(؊)]2 are given
in the Exp. Sect.
By comparison of the NMR spectra of imines 3aϪe
(Table 1) and the lithium compounds 4aϪe (Table 2) infor-
mation were gained regarding the charge distribution in the
1
Table 1. H and 13C NMR spectroscopic data of the azomethines
3 (CDCl3; 3a,b,d,e: 250/62.5 MHz, room temp.; 3c: 400/100 MHz,
0 °C)
As expected, all described metalated compounds
M[3(؊)]n are very sensitive to moisture and protic solvents,
and must be handled exclusively under inert conditions.
The lithium complexes 4 are in fact interesting intermedi-
ates, but difficult to handle so that they can only satisfac-
torily be characterized by NMR spectroscopy. Their ex-
treme moisture sensitivity complicates other analyzing tech-
niques (elemental analysis, MS, IR).
Some conceivable structures are calculated at the B3LYP/
6-311ϩϩG(d,p) level of theory, indicating, that the lithium
cation can migrate without significant energy demand to
different positions resulting in the structures 6aϪd (see
DFT calculations, Figure 5). The most stable monomeric
structure appears to be 6a while in general the reactivity
pattern in solution will be determined by dimers such as
[4e(THF)2]2 (Figure 4) so that the structural and electronic
[a]
3
a
b
Z-c[b]
E-c[b]
d
e
δH
H2/6
H3/5
Hα
CH3
δC
C2/6
C3/5
C4
8.52
7.28
4.56
8.57
8.47
8.57
8.59
7.43
4.69
2.32
8.53
7.33
4.43
1.86
[c]
7.13
4.19
2.49
7.41
4.79
2.43
4.70
2.35
Ϫ
149.7
122.7
149.7
56.1
170.2
Ϫ
149.8
122.8
149.7
54.4
166.6
16.0
149.7
149.9
149.9
122.8
150.0
54.2
166.2
15.7
149.6
122.5
150.3
53.2
177.6
13.8
122.9
149.1
55.2
171.6
29.8
123.0
149.2
54.9
170.8
21.1
Cα
C7
CH3
For numbering of compounds 3 cf. Scheme 4.[b] Two signal sets
emerge for the E/Z mixture.
the signals of the meta and para protons of the 4-phenyl substituent
[a]
[c]
The signal of H3/5 overlaps with
properties of the anion of the least stable isomer 6d appears to a multiplet at δ ϭ 7.34Ϫ7.47 ppm.
www.eurjoc.org
Eur. J. Org. Chem. 2004, 4357Ϫ4372
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4361

D. Hampe, W. Günther, H. Görls, E. Anders
FULL PAPER
carbanions. The chemical shift changes for the residues R1,
Because of the pyridine ring rotation around the C4ϪCα
R2 of 4aϪe are not so significant as for the (methyleneami- bond in the azomethines 3 the proton pairs H2/6 and H3/
1
no)pyridine moiety, hence, only the corresponding H and 5 and the corresponding carbon atoms C2/6 and C3/5 are
13C signals are listed in Table 1 and Table 2, respectively.
chemically equivalent and thus have the same chemical
shift. The signals of H2/6 in the compounds 3 are very simi-
lar; the H3/5 signals of Z-3c are slightly upfield shifted ver-
sus the H3/5 signals of the other imines 3. The above men-
tioned shielding of the α-H in Z-3c brings about an upfield
shift. The 1H methyl group shifts of 3bϪd are similar except
1
Table 2. H and 13C NMR spectroscopic data of the lithium 4H-
pyridinides 4 ([D8]THF; 4a,b,d,e: 250/62.5 MHz, room temp.)
4[a]
a
b
c[b][c]
d
e
1
for 3e. In the H NMR spectrum of 3e an upfield shift is
observed due to the second aliphatic substituent (tBu). The
13C signals of the pyridine ring and the Cα position are very
akin within the compounds 3. The imine carbon C7 of 3e
shows the greatest downfield shift followed by Z-3e because
the stabilization effect from the conjugation with an aro-
matic π system is lacking, thus, the imine carbon is more
positively polarized. The 13C CH3 signal of E/Z-3c are
clearly downfield shifted. Presumably, they are deshielded
by the ring current effect of the naphthyl residue.
δH
[d]
[d]
[e]
[e]
H2
H3
H5
H6
Hα
CH3
δC
C2
C3
C4
C5
C6
Cα
C7
CH3
7.07
6.96
6.81
5.79
6.89
5.96
1.93
6.73
6.42
5.52
6.67
5.56
1.66
6.95
5.99
7.05
6.21
2.22
5.73
7.01
6.04
6.02
[e]
6.19
2.02
Ϫ
145.7
111.7
148.1
113.5
145.3
106.8
130.4
Ϫ
144.9
110.6
146.6
112.9
144.5
105.9
123.6
11.1
144.8
144.2
109.4
145.1
111.6
144.0
103.7
125.2
11.2
143.3
107.4
142.7
109.1
143.2
99.4
109.8
146.6
112.4
144.7
106.0
126.7
15.7
The strong upfield shift of the protons H2, H3, H5 and
H6 (∆δ of H2/6: 1.41Ϫ1.86 ppm, H3: 0.18Ϫ0.91 ppm, H5:
1.14Ϫ1.81 ppm) and the carbon atoms C2, C3, C5 and C6
(∆δ of C2/6: 3Ϫ7 ppm, C3: 9Ϫ15 ppm, C5: 7Ϫ13 ppm) in
4 compared with 3 demonstrates a great amount of negative
charge delocalization into the pyridine ring which is in ac-
cordance with the major participation of mesomeric struc-
ture A in the resonance hybrid (Scheme 5). In contrast to
138.4
10.8
[a]
For numbering of compounds 4 cf. Scheme 4.[b] Only the E-
isomer is observable. [c] Measurement at 400/100 MHz (1H/13C) and
[d]
0 °C.
The signals of H2/3 overlap with proton signals of the
[e]
phenyl substituent to a multiplet at 7.11Ϫ7.17 ppm. The signals
of H2, H3 and H6 are overlapped to a multiplet at 7.09Ϫ7.18 ppm. the imines 3, separated signals appears for the protons and
1
Figure 3. H NMR of the lithium 4H-pyridin-1-ide 4c ([D8]THF, 0 °C, 400 MHz)
4362
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 4357Ϫ4372

Metal 4-Alkylidene-4H-pyridin-1-ides and 2H-Imidazol-4-ones
FULL PAPER
carbon atoms 2, 3, 5 and 6 as the rotation of the pyridine 7.65 ppm in E/Z-3c to 9.65 ppm in 4c (Figure 3). The
ring around the bond C4ϪCα is no longer possible owing NOESY spectrum displays crosspeaks of H16 to H3 and
to its great double bond character. Thus, they have different H2, but not to the methyl protons, thus, providing an evi-
chemical surroundings. While H3/5 and C3/5 are clearly se- dence for the E configuration of 4c (see Scheme 3). Possibly,
pararted to ca. 1 ppm (1H) and 2 ppm (13C), H2/6 and C2/ a weak hydrogen bond exists between H16 and the imine-
6 do not differ so much from each other: the doublets (1H) N (six-membered ring) or H16 reaches into the negative
are located closely abreast or appear as a triplet. The largest area of the anisotropy cone of the 2-aza-allyl system or the
proton shift differences to the starting materials emerge for CϭN double bond, respectively. On the B3LYP/6-
4d (∆δ: 1.62/1.69 ppm for H2/6, 0.6/1.81 ppm for H3/5) and 311ϩϩG(d,p) theory level, the H16ϪNim distance was de-
˚
4e (∆δ: 1.8/1.86 ppm for H2/6, 0.91/1.81 ppm for H3/5). In termined to 2.202 A (vide infra) which is very short for a
4d the 4-methoxy substituent enhances the electron density hydrogen bond.[19] Therefore, the down field shift is sup-
in the phenyl ring and, therefore, restricts the charge de- posed to be primarily attributed to the electronic influence
localization into the aromatic ring. In 4e no aromatic sub- of the 2-aza-allyl anion.
stituent is available for charge distribution. Thus, the elec-
Finally, a structure motif of a dimer of 4e is presented
tron density in the 4-[(methyleneamino)methyl]pyridine in Figure 4 to back up the results of the above described
moiety is higher in 4d,e than in 4aϪc.
NMR investigations.
The C4 signals of the 4H-pyridinide in 4 are shifted up-
field in the range of 1Ϫ7 ppm, whereat 4e shows the largest
difference in comparison with 3e.
The Hα and Cα signals of 4 are shifted downfield by
∆δ ϭ 1.13Ϫ1.48 ppm (1H) and 46Ϫ51 ppm (13C) because
of the change in carbon hybridization from sp3 to sp2 after
deprotonation of 3. In relation to other olefinic shift values
(110Ϫ130 ppm), the Cα atoms experience a great amount
of shielding due to the negative partial charge. In 4e, the Cα
atoms has the lowest value because no aromatic π system
stabilizes the negative charge. With respect to the azome-
thines 3, the imine carbon atoms C7 in 4 undergo the larg-
est upfield shift of 39Ϫ44 ppm which is in good agreement
with the mesomeric formula B of an 2-aza-allyl anion
(Scheme 5). The above mentioned reason for the lower Cα
shift value in 4e compared to 4aϪd can be given for the
higher C7 shifts in 4e versus 4aϪd, too. But 4a also shows
a somewhat lower upfield shift compared to 4bϪd, because
the phenyl rings cannot be localized in the same plane as
the CϭN double bond due to the repelling forces between
the ortho-positioned hydrogen atoms. Hence, the delocali-
zation into the aromatic substituents is reduced (i.e. less
electron density is stabilized on C7). The H and C signals
of the methyl groups in 4bϪe are slightly shielded compared
to 3bϪe to about ∆δ ϭ 0.20Ϫ0.38 ppm (1H) and 3Ϫ14 ppm
(13C) by the negative partial charge of the 2-aza-allyl anion.
Compound 4c shows the largest shift difference of
14.1 ppm.
[33Ϫ37]
Figure 4. Structure motif of [4e(THF)2]2
Ϫ a dimer of 4e
The 4H-pyridin-1-ide anion in 4e is E configured regard-
ing the CϭN double bond. The lithium ions are coordi-
nated by the N atoms of the 4H-pyridinide rings in an
asymmetrically way and each by two solvent molecules
(THF) in a tetrahedral mode.
Because of these results we conjecture that compounds
4b,d also adopt E configuration, but in order to unambigu-
ously prove it further NOESY measurements must be car-
ried out.
DFT Calculations
A further interesting observation can be made for the
E/Z Isomerization of 3c: As mentioned above orientating
lithium 4H-pyridin-1-ide 4c which reveals a remarkable DFT calculations on the B3LYP/6-311ϩϩG(d,p) level of
downfield shift of the naphthyl proton H16 from ca. 7.5/ theory using the GAUSSIAN 98 program package[20] were
Table 3. Absolute and relative energies for the calculated structures Z-3c, E-3c and the E/Z isomerization TS
Structure
abs. E[a]
[a.u.]
ZPE
[a.u.]
ZPE-corr. E
[a.u.]
rel. E
NIMAG[b]
∆E
[kcal·molϪ1
]
Z-3e
E-3c
TS
Ϫ805.250789
Ϫ805.247460
Ϫ805.208817
0.292666
0.292688
0.290602
Ϫ804.958123
Ϫ804.954772
Ϫ804.918215
0.0
2.1
25.0
[0]
[0]
[1]
[a]
[b]
B3LYP/6-311ϩϩG(d,p) optimizations. Number of imaginary frequencies
www.eurjoc.org
Eur. J. Org. Chem. 2004, 4357Ϫ4372
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4363

D. Hampe, W. Günther, H. Görls, E. Anders
FULL PAPER
Scheme 6. Calculated structures of Z-3c, E-3c, and the TS for the E/Z isomerization: an inversion mechanism (B3LYP/6-311ϩϩG(d,p)
results). For completeness, the structure of the E-configured anion of 3c(؊) is depicted, too
carried out to get insight into the isomerization mechanism
of Z/E-3c. The calculated structures of Z-3c, E-3c and the
transition state TS are depicted in Scheme 6. The crystal
structure data were the starting point for the calculations
Table 5. Charge distribution in the anion E-3c(؊) (Scheme 6) as
result from an NBO analysis (NBO version 3.0[21]) on the B3LYP/
6-311ϩϩG(d,p) level of theory. Charges of the protons are omitted
Atom
Charge
Atom
Charge
˚
(using the RES file as input). Bond lengths (in A) of the
molecular structure of Z-3c (Figure 2) and the calculated
structures are listed in Table 4. Moreover, the anion of 3c
with E configuration (Scheme 6) was optimized on the same
N1
Ϫ0.55
0.03
Ϫ0.28
Ϫ0.03
Cα
Nim
Ϫ0.12
Ϫ0.47
C2/6
C3/5
C4
C7
0.12
˚
Table 4. Selected bond lengths (BL, in A) of the crystal structure
of Z-3c and the calculated structures of Z-3c, E-3c, the E-anion
3c(؊), and the TS for the E/Z isomerization; the naphthyl bonds level of theory to perform an NBO[21] analysis. The charge
are omitted for clearness
distribution obtained is summarized in Table 5.
In general, two mechanisms can be discussed: a rotation
Bond
Exp. BL
calcd. BL
about the CϭN double bond or an inversion of the (pyri-
din-4-yl)methyl residue on the imine nitrogen, respec-
tively.[15,16] A coupled mechanism with both inversional and
torsional components must also be taken into account. The
transition state TS, as simplified depicted in Scheme 6, is
characterized by a linear arrangement of the Cα, Nim and
C7 atoms (bond angle 180°). Both bonds CαϪNim (1.394
Z-3c[a]
Z-3c[b] E-3c TS (inversion) E-3c(؊)
N1ϪC2
N1ϪC6
C2ϪC3
C3ϪC4
C4ϪC5
C5ϪC6
C4ϪCα
1.342 (3) 1.336
1.332 (3) 1.338
1.377 (4) 1.394
1.384 (3) 1.395
1.381 (3) 1.398
1.380 (3) 1.392
1.519 (3) 1.517
1.336 1.336
1.338 1.338
1.395 1.395
1.393 1.391
1.398 1.397
1.391 1.391
1.516 1.531
1.451 1.397
1.278 1.245
1.520 1.527
1.497 1.521
1.349
1.345
1.383
1.422
1.421
1.384
1.428
1.331
1.325
1.517
1.456
˚
˚
A) and NimϪC7 (1.245 A) are shortened compared with
Z-3c and E-3c (Table 4). Hence, the hybridization of the
imine nitrogen ranges between sp2 and sp. Thus, the iso-
merization mechanism can be denoted as an inversion. The
bonds C7ϪC8 and C7ϪC9 are somewhat lengthened to
avoid steric hindrance between the naphthyl and the methyl
CαϪNim 1.463 (3) 1.454
Nim-C7
C7ϪC8
C7ϪC9
1.275 (3) 1.273
1.500 (4) 1.509
1.505 (3) 1.508
[a]
[b]
Data taken from the crystal structure. Calculated values.
4364
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 4357Ϫ4372

Metal 4-Alkylidene-4H-pyridin-1-ides and 2H-Imidazol-4-ones
FULL PAPER
group on C7. The pyridine ring almost lies in the C7 (ϩ0.12). The high negative charge on N1 (Ϫ0.55) and
CαϪNimϪC7ϪC8 plane, the naphthyl ring almost stands the carbon atoms C3/5 as well as the extended planar sys-
perpendicular on that plane (Scheme 6).
tem (pyridine, Cα, Nim, C7) indicate the main contribution
Surprisingly, the Z-isomer is about 2.1 kcal/mol more of the 4H-pyridin-1-ide (form A, Scheme 5) on the reso-
stable than its E counterpart. The energy barrier for their nance hybrid.
transformation (TS) is with 25.0 kcal/mol somewhat higher
One possible explanation for the observed attack of the
than for some experimentally investigated CϭN double electrophiles at the Cα position (vide infra) is the regener-
bond systems. Kessler et al. determined the free activation ation of the aromatic system (regain of resonance energy)
enthalpies for E/Z isomerizations of different 2,4,6-triiso- in the pyridine ring.
1
propyl-substituted anils by inversion with H NMR coalesc-
The comparison of the experimentally obtained and the
ence measurements of diastereotopic groups bound on the calculated 13C NMR signals of Z/E-3c and the E-anion
imine carbon. Their highest experimentally obtained value 3c(؊) exhibits a little overestimation of the chemical shifts
was 19.8 kcal/mol (coalescence temperature 102 °C).[15a] about 4Ϫ10 ppm by the theoretical values, but the direction
Marullo and Wagener determined the free activation ener- of the changes caused by the negative charge are well repro-
gies of the azomethines derived from aceton and aniline duced (vide supra).
and benzylamine, respectively, to 21 kcal/mol and Ͼ 23
Table 6. Comparison of the experimental and calculated 13C NMR
kcal/mol by measurement of the coalescence tempera-
shifts for selected carbon atoms of Z-3c, E-3c and the E-anion
ture.[16c] Hofmann et al. calculated on the STO-3G level of
3c(؊), B3LYP/6-311ϩϩG(d,p) level of theory (GIAO-DFT calcu-
theory values of 23Ϫ30 kcal/mol for the linear inversion
lations[22]
)
transition state of three different anil types.[16d]
Comparing the crystal structure of Z-3c (Figure 2) with
its calculated analogue (Scheme 6) the most striking differ-
ence is a conformational one. The imino moiety (torsion
angle 61.7° in the crystal) is turned for 55.2° almost into
the plane of the pyridine ring by rotation along the
CαϪNim bond (torsion angle 16.5°). That is certainly the
result of packing effects in the solid state. Since the crystal
structure data was chosen as input for the optimization the
geometry of Z-3c, such a structure does not represent a
minimum in the gas phase. The CϪC bonds of the pyridine
ring are somewhat too long calculated, but apart from that
the selected DFT method (B3LYP/6-311ϩϩG(d,p)) is
proved to be suitable for those investigations.
C-atom
Exp. δ (ppm)
E-3c E-3c(؊)
Calcd. δ (ppm)
Z-3c E-3c E-3c(؊)
Z-3c
C2
C3
C4
C5
C6
Cα
C7
C8
149.7 149.9 144.8
122.9 123.0 109.8
149.1 149.2 146.6
122.9 123.0 112.4
149.7 149.9 144.7
158.0 158.5 155.0
127.7 128.6 118.6
157.7 157.3 156.4
126.3 126.8 121.0
156.6 156.5 154.9
55.2
171.6 170.8 126.7
29.8 21.1 15.7
54.9 106.0
60.8
179.6 178.7 123.6
33.8 22.2 15.8
58.9 115.7
Different Lithium Coordination to the 4-(Iminometh-
yl)pyridine Anion System ؊ A Model Study: Additionally to
In the E-configured anion (Scheme 6) the (pyridin-4-yl)- the experimentally gained results we investigate the mono-
methyl moiety and the CαϪNim bond are located in the mer coordination modes of the lithium cation to the 4-
same plane from which the naphthyl residue is turned out [(methyleneamino)methyl]pyridine anion with DFT meth-
˚
about 26.2°. The C4ϪCα (1.428 A) and the CαϪNim (1.331 ods. The selected model system derives from 4-(aminome-
˚
˚
A) bond are shortened whereas the NimϪC7 (1.325 A) thyl)pyridine (1) and formaldehyde (Figure 5, 5: free anion;
bond is elongated in comparison to the neutral species. The 6aϪd: lithium species). Those theoretical investigations can
pyridine ring is also affected regarding the bond lengths: be considered as extension to the 4-(picolyl)lithium system
C2ϪC3/C5ϪC6 are somewhat shortened and the remaining formerly examined in our work group.[23]
bonds are slightly longer (Table 4). The similar bond
As depicted in Figure 5, four different coordination
lengths of CαϪNim and NimϪC7 remember a 2-aza-allyl modes 6aϪd of the lithium cation to the formaldimine
anion (mesomeric form B, Scheme 5), but the results of the anion 5 are conceivable following experimental data[23,24]
NBO analysis (Table 5) reveal a slight positive charge on and theoretical studies,[23,25] whereat π coordination (6aϪc)
Figure 5. The formaldimine anion 5 and its different coordination modes 6aϪd to the lithium counterion
Eur. J. Org. Chem. 2004, 4357Ϫ4372
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4365

D. Hampe, W. Günther, H. Görls, E. Anders
FULL PAPER
is favored to the Nσ lithiation (6d).[26,27] The most stable solid state (X-ray crystallography),[23] the lithium cation is
monomer is structure 6a, in which the lithium ion is η4- exclusively coordinated to the N1.
coordinated to the atoms C3, C4, Cα, and Nim in a sickle-
As found in previous investigations of lithium 4-alkylid-
shaped fashion with distinct distortion towards the Nim ene-4H-pyridin-1-ides in solution (HOESY NMR) stabili-
˚
atom (cf. Table 8 for LiϪC/N distances, in A). This coordi- zation of that mode is achieved by aggregation to dimers
nation mode was also found in MNDO calculations of (cf. [4e(THF)2]2, Figure 4) and additional coordination of
Würthwein et al. for lithium 1-phenyl-2-aza-1,3-pentadienyl solvent molecules (e.g. THF in [4e(THF)2]2) or co-ligands
compounds, in which the lithium is bound to ortho-C (Ph (e.g. TMEDA) to obtain a saturation of the metal coordi-
ring), ipso-C (Ph ring), C1, and N2.[27a] In 6b, the lithium nation sphere, which obviously overcompensates the ener-
ion is symmetrically coordinated in an η3 mode to the getically favored π coordination in the monomers.[26b] Col-
atoms Cα, Nim, and C7 in a 2-aza-allyl analogous manner. lum et al. had shown by NMR investigations that lithio
The LiϪC distances are slightly shorter, and the LiϪN dis- imines and lithio amides form dimers in solution.[28]
tance is slightly longer than in 6a.
Comparing the binding relations of 5 and 6aϪd with the
The lithium monomer 6c represents a ‘‘borderline’’ case experimental bond lengths of the methyleneamino-pyridine
between η5 and η6 coordination because the LiϪC4 dis- moiety in Z-3c (Table 4) an elongation of the C3ϪC4/
tance can or cannot be regarded as a binding contact. The C4ϪC5 bonds in the anionic systems is discernible, which
LiϪC2/6 distances are shorter than the LiϪC distances in is greatest for 6c followed by 6d. The C4ϪCα and CαϪNim
6a and 6b, but the LiϪC3/5 distances in 6c are clearly bonds are clearly shortened and therefore have a distinct
longer. The LiϪN1 distance is slightly longer than in 6a double bond character, which is greatest for the C4ϪCα
and 6b. In contrary to the lithium coordination observed bond in 6c,d and for the CαϪNim bond in 6b and the
in the structure motif of [4e(THF)2]2 (Figure 4), the related anions E-3c(؊) (Table 4) and 5.
monomer model 6d is the least stable species in this series.
The NimϪC7 bond is only slightly lengthened except for
This can be attributed to the coordinative unsaturation of 6b where it is stretched to a similar length of the CαϪNim
the η1 (Nσ) bound lithium cation. The LiϪN1 distance is bond (an almost symmetrical 2-aza-allyl anion). In 6a, the
the shortest one in this series of monomers. The LiϪC2/6 C2ϪC3 bond is strongly stretched in comparison with the
distances are too long for binding interactions, but the lith- naked anion 5 due to the lithium coordination while the
ium ion shows a slight shift towards C2 (Table 8). In the C3ϪC4 and C4ϪCα bonds are only slightly affected. The
Table 7. Absolute and relative energies for the calculated structures of the anion 5 and the four lithium isomers 6aϪd
Structure
abs. E[a]
[a.u.]
ZPE
[a.u.]
ZPE-corr. E
[a.u.]
rel. E
NIMAG[b]
∆E [kcal·molϪ1
]
5
Ϫ380.566295
Ϫ388.080721
Ϫ388.067755
Ϫ388.067050
Ϫ388.062981
0.122602
0.126578
0.125956
0.126564
0.126026
Ϫ380.443693
Ϫ387.954143
Ϫ387.941799
Ϫ387.940486
Ϫ387.936955
Ϫ
[0]
[0]
[0]
[0]
[0]
6a
6b
6c
6d
0.0
7.7
8.6
10.8
[a]
[b]
B3LYP/6-311ϩϩG(d,p) optimizations. Number of imaginary frequencies.
˚
Table 8. Calculated bond lengths and distances to lithium (in A) of the anion 5 and the lithium species 6aϪd
Bond
5
6a
6b
6c
6d
N1ϪC2
N1ϪC6
C2ϪC3
C3ϪC4
C4ϪC5
C5ϪC6
C4ϪCα
CαϪNim
Nim-C7
1.353
1.349
1.380
1.432
1.431
1.381
1.408
1.348
1.299
1.330
1.356
1.397
1.438
1.430
1.375
1.418
1.373
1.342
1.339
1.387
1.410
1.409
1.389
1.453
1.336
1.371
1.368
1.378
1.457
1.454
1.380
1.377
1.373
1.381
1.377
1.359
1.447
1.445
1.361
1.379
1.370
1.296
1.329
1.283
1.284
LiϪN1 1.805
LiϪNim 1.910
LiϪCα 2.161
LiϪC4 2.195
LiϪC3 2.171
LiϪCα 2.147
LiϪNim 1.919
LiϪC7 2.151
LiϪN1 2.055
LiϪC2 2.121
LiϪC3 2.276
LiϪC4 2.423
LiϪC5 2.288
LiϪC6 2.131
LiϪC2 2.786
LiϪC6 2.822
4366
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 4357Ϫ4372

Metal 4-Alkylidene-4H-pyridin-1-ides and 2H-Imidazol-4-ones
FULL PAPER
lithium η5/η6 coordination in 6c causes a strong elongation
of the N1ϪC2/N1ϪC6 bonds in comparison with the neu-
tral Z-3c and the naked anion 5, while the C2ϪC3/C5ϪC6
bonds are barely influenced. The isomer 6d show an length-
ening of N1ϪC2/N1ϪC6 bonds as well, but is accompanied
with a shortening of the C2ϪC3/C5ϪC6 and C4ϪCα
bonds.
The NBO analysis of the anion 5 reveals five centers with
increased negative charge density: N1, C3/C5, Cα, Nim,
and C7. The interaction with the lithium cation changes the
charge distribution as expected in direction to the coordi-
nating centers. In 6a, C3, C4, Cα, and Nim bear a higher
negative charge, C7 a smaller one than in 5. The lithium 2-
aza-allyl isomer 6b shows a high negative charge on C7 and
Cα; the charge on the imine nitrogen is a little bit smaller
than in 6a. Coordination of the metal ion above the pyri-
dine ring plane in 6c causes the accumulation of negative
charge into the ring, which results in the negativation of
C2/C6. The Nσ interaction of the lithium ion in 6d entails
a high concentration of negative charge onto the pyridine
nitrogen N1.
In contrast to 5, the highly substituted anion E-3c(؊)
(Table 5) contains only four centers of higher negative
charge density: N1, C3/5, Cα, and Nim. The carbon C7 is
positively charged (about the same amount like Cα), which
should be traced back to the bulky naphthalene residue.
This result is more consistent with the experimental obser-
vations of favoring the Cα attack.
Scheme 7. Formation and decay of the carboxylate 7 under mild
protic conditions
As 2D NMR measurements exhibit the attack of CO2
occurs on the Cα atom, which results in formation of the
lithium 2-[(dialkylmethylene)amino]-2-(pyridin-4-yl)acetate
7. In the HMBC spectrum (1H,13C correlation) the cross-
peaks between Hα and C7 (imine carbon) and C13 (car-
boxyl carbon), respectively, unambiguously pinpoint the
carboxyl group to the α position. If CO2 would be bound
on C7 a crosspeak between the methyl protons and the car-
boxyl carbon have to appear, but such a signal was not ob-
served. The carboxylate experienced a fast loss of carbon
dioxide in protic solvents (like water or alcohols) under re-
covery of the starting material imine 3e. Presumably, the
proton is primarily attached to the pyridine nitrogen as in
an amino acid zwitterion, which facilitates the escape of
CO2 via formation of a neutral 4-alkylidene-1,4-dihydro-
pyridine. Spontaneous decarboxylation is also described for
other free pyridin-4/2-yl acetic acids.[30] Thus, the 4H-pyri-
Table 9. Calculated charge distribution of the anion 5 and the lith-
ium species 6aϪd as result of an NBO analysis (B3LYP/6-
311ϩϩG(d,p) level of theory, NBO version 3.0[21]); charges of the
hydrogen atoms are omitted
Atom
5
6a
6b
6c
6d
N1
C2
C3
C4
C5
C6
Cα
Nim
C7
Li
Ϫ0.59
0.02
Ϫ0.29
Ϫ0.05
Ϫ0.30
0.02
Ϫ0.16
Ϫ0.45
Ϫ0.28
Ϫ
Ϫ0.52
0.09
Ϫ0.46
Ϫ0.11
Ϫ0.27
0.05
Ϫ0.23
Ϫ0.60
Ϫ0.16
0.93
Ϫ0.49
0.05
Ϫ0.26
Ϫ0.05
Ϫ0.26
0.05
Ϫ0.26
Ϫ0.51
Ϫ0.45
0.92
Ϫ0.66
Ϫ0.05
Ϫ0.34
Ϫ0.11
Ϫ0.34
Ϫ0.05
Ϫ0.12
Ϫ0.46
Ϫ0.11
0.94
Ϫ0.88
0.01
Ϫ0.26
Ϫ0.06 din-1-ide anion in 4e is able to act as carbon dioxide reser-
Ϫ0.26
0.02
Ϫ0.11
Ϫ0.45
Ϫ0.13
0.95
voir by fixation and release of CO2 following the vitamin
B6 similar way (Scheme 2, b) under mild conditions. At-
tempts to isolate the corresponding hydrochloride of the
acid failed due to the large pace of the decarboxylation and
the instability of the imine functionality against an acidic
Reactions with Heterocumulenes ؊ an Outlook
Two trendsetting examples are described here to give a
first expression on the chemical behavior of the lithium 4H-
pyridin-1-ides 4 towards heterocumulenes XϭCϭY: the re-
action of 4e with carbon dioxide[29] and phenyl isocyanate
(8), respectively (Scheme 7 and 8).
Bubbling dry carbon dioxide gas through a clear, deep
red THF solution of 4e at room temp. causes an instan-
taneous change of color to bright yellow, a slight warming
of the solution (5Ϫ10 °C) and the formation of a volumi-
nous, fluffy precipitate. The reaction is completed after less
than 5 min.
Scheme 8. Reaction sequence leading to the (pyridin-4-yl)-5(2H)-
imidazolone 9. i) THF, Ϫ30 °C to room temp.; ii) H2O/CHCl3; iii)
CH3OH/O2 (air), ∆T
Eur. J. Org. Chem. 2004, 4357Ϫ4372
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4367

D. Hampe, W. Günther, H. Görls, E. Anders
FULL PAPER
medium. Derivatization reactions for carboxyl anions to get nation mode of lithium in 6a is the most stable found
a more stable carboxyl system like an ester or an amide are monomer species in the gas phase. The E configuration of
rare. But it turned out that the ethyl ester prepared from 4e two lithium compounds (4c,e) could be proved by a
and diethyl carbonate is unstable against protic solvents, NOESY experiment and a structure motif, respectively. On
too (formation of CO2 and ethanol).
the contrary to the calculated charge distribution the attack
The addition of phenyl isocyanate (8) to a solution of 4e of different heterocumulenes preferentially occurs on the α-
followed by hydrolysis and treating the crude product with carbon. The carboxylate 7 obtained as product of treating
hot methanol resulted in the 4-(pyridin-4-yl)-5(2H)-imid- 4e with CO2 undergoes a fast decarboxylation under mild
azolone 9 (Scheme 8). Based on NMR spectroscopic data protic conditions and, thus, can serve as a CO2 storage. The
the crude foamy product obtained after hydrolysis consists primarily formed carboxamide 10 by reaction of 4e with
mainly of the carboxamide 10 resulting from the attack on phenyl isocyanate and following hydrolysis cyclizes to the
the Cα atom of 4e. The percentage of 10 in the foam imidazolidinone 11 which is easily oxidized on the
amounts to 70Ϫ95% in dependence of the working-up rate. CαHϪNH bond to the imidazol-4-one 9.
The attempt to obtain 10 in pure substance by crystalliza-
tion or column chromatography failed due to the fast cycli-
Experimental Section
zation-oxidation or the decomposition of the imino func-
tion. Heating 10 in methanol for several minutes under vig-
orous stirring causes ring closure and oxidation of the
CαHϪNH bond of the built imidazolidone derivative 11 to
9. Compound 11 was only observable in NMR spectra (cf.
Exp. Sect.) during the oxidation process. The carboxamide
10 and the imidazolidone 11 differ clearly from each other
in the chemical shift of the Hα and the NH signal in the
1H NMR spectrum (10: Hα δ ϭ 5.00, NH 9.86 ppm; 11:
Hα δ ϭ 4.64, NH 9.46 ppm). The structural determination
of 9 was carried out with two-dimensional NMR spec-
troscopy. The result of the reaction sequence can be grasped
in the widest sense as 1,3-dipolar cycloaddition (11) fol-
lowed by an oxidation (9). The formation of an extended
conjugated π-system between aromatic pyridine ring and
CϭO double bond is considered as driving force for the oxi-
dation.
General Remarks: Spectra were measured with following technical
devices: NMR Ϫ Bruker AC 250 and AC 400, respectively, IR-
ATR Ϫ BIORAD FTS-25, MS Ϫ Finnigan MAT SSQ 710 and
Finnigan MAT 900 XL TRAP, respectively. Elemental analyses (C,
H, N, S) Ϫ Leco CHNS-932 and Vario EL III. NMR spectra were
recorded at 250 or 400 MHz and 62.5 or 100 MHz, for proton and
carbon, respectively. For 1H and 13C, [D8]THF (δΗ ϭ 1.73, 3.58
ppm, δC ϭ 25.2, 67.4 ppm) and CDCl3 (δΗ ϭ 7.24, δC ϭ 77.0 ppm)
were used as solvents with TMS as internal standard. Melting
points (uncorrected) were obtained with a Lindstrom copper block
apparatus. The ketones 2aϪe and 4-(aminomethyl)pyridine (1) are
commercially available, but ketones 2b,c,d were prepared in
FriedelϪCrafts acylation according to standard procedure. The
amine 1 was purified by distillation and used immediately. n-Butyl-
lithium (1.6  solution in hexane fraction) were purchased from
Aldrich. The diethylmagnesium solution was prepared after litera-
ture procedure.[32] The concentration of magnesium was deter-
mined by titration with 0.025  EDTA solution against Eriochrom-
scharz T. The nomenclature of all new compounds was obtained
using the programme ACD/IUPAC Name Free v8.05 with the
ACD/I-Lab service. For more information about ACD/IUPAC
Name Free v8.05 for Windows, please contact ACD (www.acdlab-
s.com).
The 1,3-dipolar cycloaddition of azomethine ylides, de-
rived from the thermal ring opening of aziridines, or other
nitrogen ylides with iso(thio)cyanates resembles on the
above presented reaction, but a following oxidation is often
not possible due to the substituent on the aziridine N atom
(not a proton).[31]
General Procedure for the Synthesis of 3aϪe: A solution of ketone
2
(0.1 mol), amine 1 (0.1 mol) and p-toluenesulfonic acid
Conclusion
(75Ϫ100 mg) in toluene (125 mL) was refluxed in a DeanϪStark
apparatus until almost all reaction water (0.1 mol) has separated
(5Ϫ14 h). After cooling to room temp. the reaction solution is
washed with diluted Na2CO3 solution (1 ϫ 75 mL), brine (3 ϫ 75
mL) and dried with Na2SO4. Further purification was carried out
as described at the corresponding imine 3.
A new class of azomethines is easily synthesized by con-
densation of 4-(aminomethyl)pyridine with various ketones.
The unsymmetrically substituted imines can appear in E/Z
isomers but only in one case (3c) both isomers could be
observed at NMR spectroscopy. DFT calculations give in-
N-(Diphenylmethylene)-(1-pyridin-4-yl)methanamine (3a): Reflux:
sights on mechanistical details of the E/Z isomerization of 14 h. After complete removal of toluene in vacuo the resulting yel-
low oil was crystallized with diethyl ether (30 mL), maybe under
cooling. The imine precipitated as pale yellow powder. From the
filtrate pale yellow crystal were obtained after cooling to ϩ 4 °C.
the imines; according to them it takes place via an inver-
sion. The α-methylene group shows a remarkable CH acid-
ity and can be deprotonated almost completely by n-butyl-
lithium. Comparison of the NMR spectra of the azome-
thine starting materials 3 and the corresponding anions 4
allows a conclusion on the delocalization of the negative
charge into the pyridine ring and the CϭN double bond.
This is supported by DFT calculations on a model system
derived from 4-(aminomethyl)pyridine and formaldehyde.
1
Yield: 16.1 g (59%), m.p. 236Ϫ238 °C (decomposition). H NMR
(250 MHz, CDCl3, room temp.): δ ϭ 4.56 (s, 2 H), 7.14Ϫ7.18 (m,
2 H), 7.28 (dd, 3J ϭ 4.43, 4J ϭ 1.75 Hz, 2 H), 7.33Ϫ7.38 (m, 3 H),
7.43Ϫ7.47 (m, 3 H), 7.66Ϫ7.70 (m, 2 H), 8.52 (dd, 3J ϭ 4.43, 4J ϭ
1.75 Hz, 2 H) ppm. 13C NMR (62.5 MHz, CDCl3, room temp.):
δ ϭ 56.1, 122.7, 127.5, 128.2, 128.5, 128.7, 128.8, 130.4, 136.3,
139.3, 149.7, 170.2 ppm. IR (ATR): ν˜ ϭ 3025 (ϭCϪH), 1622 (Cϭ
According the DFT results, the sickle-shaped η4 coordi- N), 1596 (aryl), 1444 (CH2) cmϪ1. MS (ESI): m/z (%) ϭ 273 (100)
4368  2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.eurjoc.org
Eur. J. Org. Chem. 2004, 4357Ϫ4372

Metal 4-Alkylidene-4H-pyridin-1-ides and 2H-Imidazol-4-ones
FULL PAPER
[M ϩ H]. C19H16N2 (272.34): calcd. C 83.79, H 5.92, N 10.29;
found C 83.73, H 5.91, N 10.34.
δ ϭ 13.8, 27.7, 40.9, 53.2, 122.5, 149.6, 150.3, 177.6 ppm. IR
(ATR): ν˜ ϭ 3028 (ϭCϪH), 2967 (CH2, CH3), 1652 (CϭN), 1600,
1557, 1472 (aryl), 1410 (CH2, CH3) cmϪ1. MS (EI): m/z (%) ϭ 191
(100) [M ϩ H]ϩ. C12H18N2 (190.28): calcd. C 75.73, H 9.55, N
14.72; found C 75.67, H 9.51, N 15.06.
N-[1-(4-Phenylphenyl)ethylidene]-(1-pyridin-4-yl)methanamine (3b):
Reflux: 10 h. The solution was concentrated to one third of the
original volume and cooled to ϩ4 °C. The yellow precipitate was
filtered off and washed with cold diethyl ether. From the filtrate
another fraction could be collected. The imine is sufficiently pure
for further reactions, but additional purification is possible by ex-
traction with n-hexane. Yield: 15.2 g (53%), m.p. 100 °C (decompo-
sition). 1H NMR (250 MHz, CDCl3, room temp.): δ ϭ 2.35 (t, 5J ϭ
0.78 Hz, 3 H), 4.70 (s, 2 H), 7.34Ϫ7.47 (m, 5 H), 7.61 (dd, 3J ϭ
General Procedure for the Synthesis of 4aϪe: The compounds 4aϪe
were directly prepared for NMR in deuterated [D8]THF without
prior isolation. The synthesis is carried out under argon. A yield
would not given since a very small inaccuracy can already lead
to an incomplete deprotonation. In a small Schlenk tube 0.3 mL
(0.48 mmol) n-butyllithium (1.6  solution in hexane) is filled in
and the solvent is removed completely by distillation (cold). The
oily residue is dried in vacuo for 1 1/2 h and then dissolved in
0.3Ϫ0.5 mL [D8]THF. To a Ϫ70 °C cold solution of 0.48 mmol
imine 3 in 0.8Ϫ1.0 mL [D8]THF the freshly prepared n-butyllith-
ium solution is added dropwise about a period of 10 minutes. Sub-
sequently the cooling is removed. The reaction mixture is stirred
overnight and then 0.5 mL are filled into a NMR tube.
4
3
4
8.35, J ϭ 1.43 Hz, 2 H), 7.63 (dd, J ϭ 8.70, J ϭ 2.08 Hz, 2 H),
3
4
3
4
7.95 (dd, J ϭ 8.53, J ϭ 1.90 Hz, 2 H), 8.57 (dd, J ϭ 4.43, J ϭ
1.58 Hz, 2 H) ppm. 13C NMR (62.5 MHz, CDCl3, room temp.):
δ ϭ 16.0, 54.4, 122.8, 127.0, 127.1, 127.2, 127.7, 139.4, 140.4, 142.7,
149.7, 149.8, 166.6 ppm. IR (ATR): ν˜ ϭ 3029 (ϭCϪH), 1679 (Cϭ
N), 1600 (aryl), 1412 (CH2) cmϪ1. MS (Micro-ESI): m/z (%) ϭ 287
(100) [M ϩ H]. C20H18N2 (286.37): calcd. C 83.88, H 6.34, N 9.78;
found C 83.50, H 6.24, N 9.74.
Lithium 4-{[(Diphenylmethylene)amino]methylene}-4H-pyridin-1-ide
(4a): 1H NMR (250 MHz, [D8]THF, room temp.): δ ϭ 5.73 (dd,
N-[1-(1-Naphthyl)ethylidene]-(1-pyridin-4-yl)methanamine (3c): Re-
flux: 4Ϫ5 h. Working up like 3b: beige needles, which can be recrys-
tallized from n-hexane. Yield: 15.6 g (60%), m.p. 87Ϫ90 °C. 1H
NMR (400 MHz, CDCl3, room temp.), Z-3c: δ ϭ 2.49 (t, 5J ϭ
3
3J ϭ 6.13, 4J ϭ 1.23 Hz, 2 H), 6.04, (s, 1 H), 6.65 (t, J ϭ 7.16 Hz,
1 H), 6.93 (t, 3J ϭ 7.36 Hz, 2 H), 7.01 (d, 3J ϭ 6.12 Hz, 1 H),
7.11Ϫ7.17 (m, 5 H), 7.31 (t, 3J ϭ 8.14 Hz, 2 H), 7.42 (dd, 3J ϭ
8.53, 4J ϭ 1.15 Hz, 2 H) ppm. 13C NMR (62.5 MHz, [D8]THF,
room temp.): δ ϭ 106.8, 111.7, 113.5, 121.5, 124.2, 125.9, 127.7,
128.8, 130.4, 131.6, 141.5, 145.0, 145.3, 145.7, 148.1 ppm.
3
4
1.28 Hz, 3 H), 4.19 (mAB, 2 H), 7.13 (dd, J ϭ 4.40, J ϭ 1.44 Hz,
2 H), 7.18 (dd, 3J ϭ 6.96, 4J ϭ 1.00 Hz, 1 H), 7.46Ϫ7.54 (m, 3 H),
3
4
3
7.61 (dd, J ϭ 7.40, J ϭ 1.08 Hz, 1 H), 7.85 (d, J ϭ 8.24 Hz, 1
3
4
3
H), 7.90 (dd, J ϭ 7.16, J ϭ 1.76 Hz, 1 H), 8.47 (dd, J ϭ 4.52,
4J ϭ 1.52 Hz, 2 H) ppm. E-3c: 2.44 (t, 5J ϭ 0.7Ϫ0.8 Hz, 3 H), 4.79
Lithium 4-({[(1E)-1-(4-Phenylphenyl)ethylidene]amino}methylene)-
(s, 2 H), 7.41 (dd, 3J ϭ 4.48, 4J ϭ 1.52 Hz, 2 H), 8.04 (d, 3J ϭ 4H-pyridin-1-ide (4b): 1H NMR (250 MHz, [D8]THF, room temp.):
3
4
3
4
9.76 Hz, 1 H), 8.56 (dd, J ϭ 4.40, J ϭ 1.52 Hz, 2 H) ppm. 13C δ ϭ 2.02 (s, 3 H), 6.02 (dd, J ϭ 6.20, J ϭ 2.11 Hz, 1 H), 6.19 (s,
NMR (100 MHz, CDCl3, room temp.), Z-3c: δ ϭ 29.8, 55.2,
1 H), 7.09Ϫ7.18 (m, 4 H), 7.33 (t, 3J ϭ 7.88 Hz, 2 H), 7.38 (d, 3J ϭ
122.89, 122.94, 124.4, 125.5, 126.6, 127.1, 128.5, 128.6, 128.8, 8.87 Hz, 2 H), 7.59 (dd, 3J ϭ 8.50, 4J ϭ 1.32 Hz, 2 H), 7.74 (d,
133.4, 137.3, 149.1, 149.7, 171.6 ppm. E-3c: 21.1, 54.9, 123.0, 124.4,
125.2, 126.0, 125.2, 125.0, 129.1, 130.0, 133.8, 140.7, 149.2, 149.9,
3J ϭ 8.67 Hz, 2 H) ppm. 13C NMR (62.5 MHz, [D8]THF, room
temp.): δ ϭ 11.1, 105.9, 110.6, 112.9, 122.7, 123.6, 125.2, 125.3,
170.8 ppm. IR (ATR): ν˜ ϭ 3053 (ϭCϪH), 1643 (CϭN), 1600 125.4, 128.2, 132.2, 141.8, 143.7, 144.5, 144.9, 146.6 ppm.
(aryl), 1414 (CH2) cmϪ1. MS (EI): m/z (%) ϭ 259 (100) [M Ϫ Hϩ].
C18H16N2 (260.33): calcd. C 83.03, H 6.21, N 10.76; found C 83.22,
H 6.31, N 10.73.
Lithium 4-({[(1E)-1-(1-Naphthyl)ethylidene]amino}methylene)-4H-
pyridin-1-ide (4c): H NMR (400 MHz, [D8]THF, 0 °C): δ ϭ 2.22
1
(s, 3 H), 5.99 (dd, 3J ϭ 6.21, 4J ϭ 2.04 Hz, 1 H), 6.21 (s, 1 H), 6.95
(dd, 3J ϭ 6.31, 4J ϭ 1.91 Hz, 1 H), 7.05 (d, 3J ϭ 6.20 Hz, 1 H),
7.07 (d, 3J ϭ 6.33 Hz, 1 H), 7.18Ϫ7.25 (m, 5 H), 7.58 (d, 3J ϭ
9.55 Hz, 1 H), 9.65 (d, 3J ϭ 9.87 Hz, 1 H) ppm. 13C NMR
(100 MHz, [D8]THF, 0 °C): δ ϭ 15.7, 106.0, 109.8, 112.4, 121.3,
121.5, 123.2, 124.0, 124.8, 126.9, 130.5, 131.7, 135.2, 142.2, 144.7,
144.8, 146.6 ppm.
N-[1-(4-Methoxyphenyl)ethylidene]-(1-pyridin-4-yl)methanamine
(3d): Reflux: 3Ϫ4 h. Toluene was completely removed and the dark
yellow solid was extracted with n-hexane. After cooling to ϩ4 °C
a white precipitate could be isolated. The imine must be stored
under argon as it decomposes quickly due to reaction with air
1
moisture. Yield: 14.4 g (60%), m.p. 78Ϫ80 °C (decomposition). H
NMR (250 MHz, CDCl3, room temp.): δ ϭ 2.32 (s, 3 H), 3.86 (s,
3
4
Lithium
4-({[(1E)-1-(4-Methoxyphenyl)ethylidene]amino}meth-
3 H), 4.69 (s, 2 H), 6.94 (dd, J ϭ 6.79, J ϭ 2.15 Hz, 2 H), 7.42
(dd, 3J ϭ 4.45, 4J ϭ 1.54 Hz, 2 H), 7.88 (d, 3J ϭ 6.82, 4J ϭ 2.11 Hz,
2 H), 8.59 (dd, 3J ϭ 4.44, 4J ϭ 1.59 Hz, 2 H) ppm. 13C NMR
(62.5 MHz, CDCl3, room temp.): δ ϭ 15.7, 54.2, 55.3, 113.7, 122.8,
128.2, 130.6, 133.3, 149.9, 150.0, 161.1, 166.2 ppm. IR (ATR): ν˜ ϭ
3004 (ϭCϪH), 1675, 1629, 1597 (aryl), 1411 (CH2) ppm. MS
(ESI): m/z (%) ϭ 241 (100) [M ϩ H]ϩ. C15H16N2O (240.30): calcd.
C 74.97, H 6.71, N 11.66, O 6.67; found C 74.71, H 6.70, N 11.48.
1
ylene)-4H-pyridin-1-ide (4d): H NMR (250 MHz, [D8]THF, room
temp.): δ ϭ 1.93 (s, 3 H), 3.68 (s, 3 H), 5.79 (dd, J ϭ 6.33, J ϭ
3
4
3
4
2.23 Hz, 1 H), 5.96 (s, 1 H), 6.64 (dd, J ϭ 6.95, J ϭ 2.23 Hz, 2
H), 6.81 (dd, 3J ϭ 6.33, 4J ϭ 2.05 Hz, 1 H), 6.89 (d, 3J ϭ 6.15 Hz,
1 H), 6.96 (d, 3J ϭ 6.33 Hz, 1 H), 7.59 (dd, 3J ϭ 6.95, 4J ϭ 2.23 Hz,
2 H) ppm. 13C NMR (62.5 MHz, [D8]THF, room temp.): δ ϭ 11.2,
54.2, 103.7, 109.4, 111.6, 112.6, 123.6, 125.2, 138.1, 144.0, 144.2,
145.1, 155.7 ppm.
1-(Pyridin-4-yl)-N-(1,2,2-trimethylpropylidene)methanamine
(3e):
Reflux: 5 h. After completely evaporating the solvent the yellow oil
was purified by distillation. The white liquid crystallized at rubbing
with a glass stick. The imine must be also stored under argon.
Yield: 14.1 g (74%), b.p. 108Ϫ110 °C (0.2 mbar), m.p. 38Ϫ42 °C.
1H NMR (250 MHz, CDCl3, room temp.): δ ϭ 1.17 (s, 9 H), 1.86
Lithium 4-({[(1E)-1,2,2-Trimethylpropylidene]amino}methylene)-4H-
pyridin-1-ide (4e): 1H NMR (250 MHz, [D8]THF, room temp.): δ ϭ
3
4
1.09 (s, 9 H), 1.66 (s, 3 H), 5.56 (s, 1 H), 5.51 (dd, J ϭ 6.42, J ϭ
2.19 Hz, 1 H), 6.41 (dd, 3J ϭ 6.54, 4J ϭ 2.28 Hz, 1 H), 6.67 (d,
3
3J ϭ 6.41 Hz, 1 H), 6.73 (d, J ϭ 6.51 Hz, 1 H) ppm. 13C NMR
3
3
(s, 3 H), 4.43 (s, 2 H), 7.31 (d, J ϭ 5.00 Hz, 2 H), 8.53 (d, J ϭ (62.5 MHz, [D8]THF, room temp.): δ ϭ 10.8, 28.3, 38.8, 99.4, 107.4,
5.00 Hz, 2 H) ppm. 13C NMR (62.5 MHz, CDCl3, room temp.): 109.1, 138.4, 142.7, 143.2, 143.3 ppm.
Eur. J. Org. Chem. 2004, 4357Ϫ4372
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4369

D. Hampe, W. Günther, H. Görls, E. Anders
FULL PAPER
Potassium
4-({[(1E)-1-(1-Naphthyl)ethylidene]amino}methylene)-
extracted with CHCl3 (3 ϫ 30 mL), the organic layer was washed
4H-pyridin-1-ide [K3c(؊)]: The preparation is carried out with with brine (1 ϫ 50 mL), and dried with Na2SO4. The chloroform
standard Schlenk technique. To a solution of KN(SiMe3)2 (0.751 g,
3.76 mmol) in diethyl ether (30 mL) 3c (0.979 g, 3.76 mmol) is ad-
ded in small portions at 0 °C, whereat the color changes immedi-
was completely removed resulting in a yellow foam, which consists
mainly of the carboxamide 10. The attempt to obtain 10 in purified
form by crystallization or column chromatography failed due to
ately to dark blue violet. After 3 h stirring at room temp. an half the fast cyclization-oxidation or the decomposition of the imino
oily, half crystalline precipitate is formed, which becomes solid at
standing overnight. The solution is decanted and the dark blue
violet residue is washed twice with 10 mL diethyl ether and then
dried in vacuo. Yield: 1.10 g (83%) (0.75 molecules of diethyl ether
function.
N-Phenyl-2-(pyridin-4-yl)-2-[(1,2,2-trimethylpropylidene)amino]-
acetamide (10): (As crude product, ca 80%): 1H NMR (CDCl3,
250 MHz): δ ϭ 1.29 (s, 9 H), 1.80 (s, 3 H), 5.00 (s, 1 H), 7.12 (t,
1
included). H NMR ([D8]THF, 250 MHz): δ ϭ 2.22 (s, 3 H), 6.12
3J ϭ 7.39 Hz, 1 H), 7.29Ϫ7.39 (m, 4 H), 7.58 (d, J ϭ 7.47 Hz, 2
3
(d, 3J ϭ 5.63 Hz, 1 H), 6.18 (s, 1 H), 6.88 (dd, 3J ϭ 5.71, 4J ϭ
153 Hz, 1 H), 7.17Ϫ7.39 (m, 7 H), 7.60 (d, 3J ϭ 6.83 Hz, 1 H),
9.46 (d, 3J ϭ 9.98 Hz, 1 H) ppm. 13C NMR ([D8]THF, 62.5 MHz):
δ ϭ 14.3, 102.7, 106.9, 111.4, 118.9, 119.0, 121.5, 122.3, 123.2,
125.3, 128.1, 129.4, 133.7, 140.0, 145.1, 145.4, 145.6, 147.5 ppm.
H), 8.57 (dd, 3J ϭ 4.50, 4J ϭ 1.62 Hz, 2 H), 9.86 (br. s, 1 H, ) ppm.
13C NMR (CDCl3, 62.5 MHz): δ ϭ 15.0, 26.4, 41.3, 66.3, 119.4,
122.0, 129.0, 129.6, 137.5, 147.6, 149.8, 168.8, 181.0 ppm. IR
(ATR): ν˜ ϭ 3278 (NH), 3060, 3034, 2969 (alkyl), 1697 (CϭO),
1597, 1546, 1496, 1443, 1369, 1307 cmϪ1. MS (Micro-ESI): m/z
(%) ϭ 332 (100) [M ϩ Na]ϩ. Exact mass for C19H23N3ONa calcd.
332.174; found 332.174. The yellow foam was dissolved in 30 mL
methanol and the solution heated for 10 min while bubbling air
through it. Evaporation of the solvent gives the 4H-imidazol-4-one
9, which is contaminated with 13Ϫ15% of the imidazolidin-4-one
11, in 49% yield.
Magnesium
Bis[4-({[(1E)-1-(1-naphthyl)ethylidene]amino}meth-
ylene)-4H-pyridin-1-ide] {Mg[3c(؊)]2}: The preparation was carried
out with standard Schlenk technique. To a solution of 3c (1.102 g,
4.23 mmol) in THF (20 mL) a MgEt2 solution (4.0 mL, c ϭ 0.53
mol/L in THF)[32] is added at Ϫ20 °C over a period of 40 min,
whereat the color changes from beige to orange-brown with a violet
touch. The solution is warmed up to room temp. and stirred at
room temp. for 3 h. Over night the solution takes on an intensive
blue-violet color and precipitate of small red crystals is formed,
which is isolated and dried in vacuo. Yield: 1.53 g (52%) (2.1 mol-
ecules of THF included). 1H NMR ([D6]DMSO, 250 MHz): δ ϭ
2.15 (s, 3 H), 6.03 (d, 3J ϭ 4.35 Hz, 1 H), 6.11 (s, 1 H), 6.79 (d,
2-tert-Butyl-2-methyl-3-phenyl-5-(pyridin-4-yl)imidazolidin-4-one
(11): (Not isolated). 1H NMR (CDCl3, 250 MHz): δ ϭ 1.15 (s, 9
H), 2.16 (s, 3 H), 4.64 (s, 1 H), 7.29Ϫ7.41 (m, 7 H), 8.60 (dd, J ϭ
3
4.49, 4J ϭ 1.67 Hz, 2 H), 9.46 (br. s, 1 H) ppm. 13C NMR (CDCl3,
62.5 MHz): δ ϭ 24.7, 44.3, 59.4, 77.3, 119.6, 121.9, 124.6, 129.0,
129.6, 149.1, 149.9, 169.6 ppm. The crude product 9 can be purified
by column chromatography: Silica gel 60 (0.063Ϫ0.2 mm, Fluka),
eluent ethyl acetate. 2-tert-Butyl-2-methyl-3-phenyl-5-(pyridin-4-yl)-
2,3-dihydro-4H-imidazol-4-one (9): Isolated yield: 0.190 g (12.8%),
3
4
3J ϭ 4.51 Hz, 1 H), 7.19Ϫ7.35 (m, 7 H), 7.70 (dd, J ϭ 7.45, J ϭ
3.46 Hz, 1 H), 9.50 (dd, J ϭ 5.85, J ϭ 3.85 Hz, 1 H) ppm. 13C
NMR ([D6]DMSO, 62.5 MHz): δ ϭ 16.5, 105.5, 108.9, 112.2,
121.1, 121.2, 123.8, 124.6, 125.3, 127.4, 129.4, 130.6, 134.7, 141.1,
146.0, 149.3, 149.4 ppm.
3
4
1
m.p. 106Ϫ108 °C. H NMR (CDCl3, 250 MHz): δ ϭ 0.94 (s, 9 H),
1.87 (s, 3 H), 7.26Ϫ7.51 (m, 5 H), 8.31 (dd, 3J ϭ 4.54, 4J ϭ 1.50 Hz,
2 H), 8.77 (d, 3J ϭ 5.93 Hz, 2 H) ppm. 13C NMR (CDCl3,
62.5 MHz): δ ϭ 21.4, 25.9, 39.9, 93.6, 122.0, 127.5, 128.2, 129.6,
137.4, 137.7, 150.4, 160.0, 162.7 ppm. IR (ATR): ν˜ ϭ 3067, 3051,
Lithium Pyridin-4-yl[(1,2,2-trimethylpropylidene)amino]acetate (7):
To a Ϫ78 °C cold solution of 3e (0.913 g, 4.8 mmol) in THF (20
mL), nBuLi (3.0 mL, 4.8 mmol, 1.6  solution in hexane) are added
dropwise over a period of 30Ϫ40 min. The resulting deep red solu-
tion is allowed to come to room temp. and stirred overnight. Dry
carbon dioxide gas is bubbled through the solution at room temp.
whereat the color changes instantaneously to bright yellow, the
temperature raises slightly (5Ϫ10 °C) and a voluminous precipitate
is formed. The filtered product is carefully dried in vacuo (2 d).
Because of varying amounts of THF in the product the exact yield
cannot be given. Yield: 1.12 g (73%) (1.1 molecules of THF in-
cluded). 1H NMR ([D6]DMSO, room temp.): δ ϭ 1.09 (s, 9 H),
1.73 (s, 3 H), 4.82 (s, 1 H), 7.47 (d, 3J ϭ 6.15 Hz, 2 H), 8.37 (d,
2956, 2875, 1697, 1596, 1499, 1412, 1387, 1368, 1306, 1125 cmϪ1
.
MS (Micro-ESI): m/z (%) ϭ 330 (100) [M ϩ Na]ϩ. C19H21N3O
(307.39): calcd. C 74.24, H 6.89, N 13.67, O 5.20; found C 74.59,
H 6.95, N 13.94.
Crystal Structure Determination: The intensity data for the com-
pounds were collected on a Nonius KappaCCD diffractometer,
using graphite-monochromated Mo-Kα radiation. Data were cor-
rected for Lorentz and polarization effects, but not for absorp-
tion.[33,34] The structures were solved by direct methods
(SHELXS[35]) and refined by full-matrix least-squares techniques
against F2o (SHELXL-97[36]). The hydrogen atoms for the com-
pound Z-3c were located by difference Fourier synthesis and re-
fined isotropically. All other hydrogen atoms were included at cal-
culated positions with fixed thermal parameters. The absolute con-
figuration of compound Z-3c could not be determined. All non-
hydrogen atoms were refined anisotropically.[36] XP (SIEMENS
Analytical X-ray Instruments, Inc.) was used for structure represen-
tations.
4
3J ϭ 5.83, J ϭ 1.40 Hz, 2 H) ppm. 13C NMR ([D6]DMSO, room
temp.): δ ϭ 14.2, 27.5, 39.7, 69.9, 122.9, 148.5, 152.1, 172.3,
176.5 ppm. IR (ATR): ν˜ ϭ 2963, 2871, 1643, 1604, 1364, 1304
cmϪ1. MS (ESI negative): m/z (%) ϭ 473 (39) [Li(C13H17N2O2)2]Ϫ,
713 (63) [Li2(C13H17N2O2)3]Ϫ.
2-tert-Butyl-2-methyl-3-phenyl-5-(pyridin-4-yl)-2,3-dihydro-4H-
imidazol-4-one (9): To a Ϫ78 °C cold solution of 3e (0.913 g,
4.8 mmol) in 20 mL THF, nBuLi (3.0 mL, 4.8 mmol, 1.6  solution
in hexane) were added dropwise over a period of 30Ϫ40 min. The
resulting deep red solution was allowed to come to room temp. and
stirred overnight. After addition of 0.525 mL (4.8 mmol) phenyl
isocyanate (8) to the solution at Ϫ30 °C the cooling was removed.
A very voluminous orange precipitate formed after 15 min, which
stands for 6 h at room temp. The following workup was carried out
under not-inert conditions. The solvent was almost evaporated to
dryness, the residue hydrolyzed with 30 mL water, the water phase
Crystal Data for [4e(THF)2]2:[37] C40H66Li2N4O4, Mr ϭ 680.85
g·molϪ1, colourless prism, size 0.05 ϫ 0.04 ϫ 0.04 mm3, ortho-
rhombic, space group Cmca, a ϭ 13.8551(3), b ϭ 12.4606(2), c ϭ
3
˚
˚
24.7271(6) A, V ϭ 4268.96(15) A , T ϭ Ϫ90 °C, Z ϭ 4, ρcalcd.
ϭ
1.059 g·cmϪ3, µ(Mo-Kα) ϭ 0.67 cmϪ1, F(000) ϭ 1488, 13819 reflec-
tions in h(Ϫ16/17), k(Ϫ16/14), l(Ϫ29/32), measured in the range
4.40° Յ Θ Յ 27.46°, completeness Θmax. ϭ 99.2%, 2523 indepen-
4370
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 4357Ϫ4372

Metal 4-Alkylidene-4H-pyridin-1-ides and 2H-Imidazol-4-ones
FULL PAPER
[11] [11a] E. Lopez-Calle, M. Keller, W. Eberbach, Eur. J. Org. Chem.
dent reflections, Rint ϭ 0.055, 1604 reflections with Fo Ͼ 4σ(Fo),
132 parameters, 0 restraints, R1obsd. ϭ 0.106, wR2obsd. ϭ 0.298,
R1all ϭ 0.149, wR2all ϭ 0.342, GOOF ϭ 1.029, largest difference
[11b]
2003, 8, 1438Ϫ1453.
A. S. Gothelf, K. V. Gothelf, R. G.
Hazell, K. A. Jørgensen, Angew. Chem. 2002, 114, 4410Ϫ4412.
[11c]
K. V. Gothelf, K. A. Jørgensen, Chem. Rev. 1998, 98,
Ϫ3
˚
peak and hole: 0.662/Ϫ0.418 e·A
.
863Ϫ909.
[12] [12a]
Crystal Data for Z-3c:[37] C18H16N2, Mr ϭ 260.33 g·molϪ1, colour-
less prism, size 0.03 ϫ 0.03 ϫ 0.02 mm3, orthorhombic, space
A. E. Martell, Acc. Chem. Res. 1989, 22, 115Ϫ124 and
[12b]
references cited therein.
M. D. Toney, Biochemistry 2001,
40, 1378Ϫ1384. [12c] B. Witkop, T. W. Beiler, J. Am. Chem. Soc.
˚
group Pca2(1), a ϭ 12.2813(5), b ϭ 15.9945(8), c ϭ 7.1037(3) A,
1954, 76, 5589Ϫ5597. [12d] J. E. Dixon, T. C. Bruice, Biochemis-
3
V ϭ 1395.4(1) A , T ϭ Ϫ90 °C, Z ϭ 4, ρcalcd. ϭ 1.239 g·cmϪ3
,
˚
[12e]
try 1973, 12, 4762Ϫ4766.
J. R. Maley, T. C. Bruice, J. Am.
µ(Mo-Kα) ϭ 0.73 cmϪ1, F(000) ϭ 552, 3032 reflections in h(Ϫ15/
15), k(Ϫ20/20), l(Ϫ9/9), measured in the range 3.04° Յ Θ Յ 27.49°,
completeness Θmax. ϭ 99.5%, 3032 independent reflections, 2000
reflections with Fo Ͼ 4σ(Fo), 245 parameters, 1 restraints, R1obsd. ϭ
0.0487, wR2obsd. ϭ 0.0894, R1all ϭ 0.0981, wR2all ϭ 0.1045, GOOF ϭ
0.985, Flack-parameter 2(4), largest difference peak and hole:
[12f]
Chem. Soc. 1968, 90, 2843Ϫ2847.
E. H. Abbott, M. E.
Bobrick, Biochemistry 1973, 12, 846Ϫ851.
[13] [13a]
[13b]
E. Anders, Synthesis 1978, 8, 586Ϫ588.
E. Anders,
W. Will, Disser-
[13c]
Tetrahedron Lett. 1978, 38, 3547Ϫ3548.
[13d]
tation, University of Erlangen-Nürnberg, 1982.
A. Stan-
kowiak, Dissertation, University of Erlangen-Nürnberg, 1986.
[13e]
Ϫ3
˚
A. Opitz, Dissertation, University of Erlangen-Nürnberg,
0.146/Ϫ0.206 e·A
.
1992. [13f]E. Anders, W. Will, A. Stankowiak, Chem. Ber. 1983,
[13g]
116, 3192Ϫ3204.
1984, 1039Ϫ1041.
E. Anders, A. Stankowiak, Synthesis
[13h]
E. Anders, U. Korn, A. Stankowiak,
[13i]
Chem. Ber. 1989, 122, 105Ϫ111.
R. Wagner, B. Wiedel, W.
Günther, H. Görls, E. Anders, Eur. J. Org. Chem. 1999,
2383Ϫ2390.
S. Pawlenko, Houben-Weyl: Methods of Organic Chemistry,
1990, vol. E 14b, part I, 222Ϫ281.
Acknowledgments
Financial support by the Deutsche Forschungsgemeinschaft (Sond-
erforschungsbereich 436) and by the Thüringer Ministerium für
Wissenschaft, Forschung und Kultur (Erfurt, Germany) is grate-
fully acknowledged. Moreover, we thank Dr. W. Günther (Jena) for
his excellent advices in the course of our NMR investigations.
[14]
[15] [15a]
H. Kessler, D. Leibfritz, Chem. Ber. 1971, 104, 2143Ϫ2157
and references cited therein. [15b] H. Kessler, D. Leibfritz, Tetra-
[15c]
hedron 1970, 26, 1805Ϫ1820.
H. Kessler, Angew. Chem.
1970, 82, 237Ϫ253.
[1] [1a]
H. Schiff, Justus Liebigs Ann. 1864, 131, 118.
[16] [16a]
M. Raban, J. Chem. Soc., Chem. Commun. 1970,
[2] [2a]
W. R. Layer, Chem. Rev. 1963, 63, 5, 489Ϫ510.[2b] P. Y.
[16b]
1415Ϫ1416.
E. Carlson, F. B. Jones, Jr., M. Raban, J.
Sollenberger, R. B. Martin, in Chem. Amino Group (Ed.: S. Pa-
tai), Interscience Publ., London, 1968, 349Ϫ406.[2c] P. G.
Taylor, in Chemistry of the Double-Bonded Functional Groups
(Ed.: S. Patai), Wiley 1977Ϫ1997, Chichester, UK, 1997, 3(2),
1103Ϫ1133.[2d] M. Bartok, G. Schneider, in Chemistry of the
Double-Bonded Functional Groups (Ed.: S. Patai), Wiley
1977Ϫ1997, Chichester, UK, 1997, 3(2), 1223Ϫ1251.[2e] D. J.
Berger, J. M. Tanko in Chemistry of the Double-Bonded Func-
tional Groups (Ed.: S. Patai), Wiley 1977Ϫ1997, Chichester,
UK, 1997, 3(2), 1281Ϫ1354.[2f] P. Hlavica, M. Lehnerer, in
Chemistry of the Double-Bonded Functional Groups (Ed.: S. Pa-
tai), Wiley 1977Ϫ1997, Chichester, UK, 1997, 3(2),
1625Ϫ1663.
[16c]
Chem. Soc., Chem. Commun. 1969, 1235Ϫ1237.
N. P. Ma-
rullo, E. H. Wagener, Tetrahedron Lett. 1969, 30, 2555Ϫ2558.
[16d]
H.-J. Hofmann, T. Asano, R. Cimiraglia, R. Bonaccorsi,
Bull. Chem. Soc., Jpn. 1993, 66, 130Ϫ134.
[17] [17a]
C. Lambert, P. v. R. Schleyer, Houben-Weyl: Methods of
Organic Chemistry, 1993, vol. E 19d, 1Ϫ113. [17b] M. Schlosser,
Organometallics in Synthesis.A Manual, 2nd ed., John Wiley &
[17c]
Sons, New York, 2002, 1Ϫ352.
L. Pauling, The Nature of
the Chemical Bond, 3rd ed., Cornell University Press, Ithaca,
New York, 1960. [17d]T. Chan, I. Fleming, Synthesis 1979,
761Ϫ786.
S. O. Nilssonj Lill, U. Köhn, E. Anders, Eur. J. Org. Chem.
2004, 13, 2868Ϫ2880.
[18]
[19]
[20]
[3]
T. Abellan, R. Chinchilla, N. Galindo, G. Guillena, C. Najera,
J.-A. van den Berg, K. R. Seddon, Chryst. Growth Design 2003,
J. M. Sansano, Eur. J. Org. Chem. 2000, 2689Ϫ2697 and refer-
3, 643Ϫ661.
ences cited therein.
[4]
Gaussian 98, Revision A-11; M. J. Frisch, G. W. Trucks, H. B.
Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, V. G.
Zakrzewski, J. A. Montgomery, R. E. Stratmann, J. C. Burant,
S. Dapprich, J. M. Millam, A. D. Daniels, K. N. Kudin, M. C.
Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi, R. Cammi,
B. Mennucci, C. Pomelli, C. Adamo, S. Clifford, J. Ochterski,
G. A. Petersson, P. Y. Ayala, Q. Cui, K. Morokuma, D. K.
Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J.
Cioslowski, J. V. Ortiz, B. B. Stefanov, G. Liu, L. A., P. P., I.
Komaromi, R. Gomperts, R. L. Martin, D. J. Fox, T. Keith,
M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, C. Gonzalez,
M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W.
Wong, J. L. Andres, C. Gonzalez, M. Head-Gordon, E. S. Re-
plogle, J. A. Pople, Gaussian, Inc., Pittsburgh PA, 1998.
G. Alvaro, D. Savoia, Synlett 2002, 5, 651Ϫ673.
[5] [5a]
J. F. Larrow, E. N. Jacobsen, J. Org. Chem. 1994, 59,
[5b]
1939Ϫ1942.
S. E. Schaus, B. D. Brandes, J. F. Larrow, M.
Tokunaga, K. B. Hansen, A. E. Gould, M. E. Furrow, E. N.
Jacobsen, J. Am. Chem. Soc. 2002, 124, 1307Ϫ1315 and
ref.[17] therein.
[6]
M. E. Bluhm, M. Ciesielski, H. Görls, M. Döring, Angew.
Chem. 2002, 114, 3104Ϫ3107; Angew. Chem. Int. Ed. 2002, 41,
0000Ϫ0000 !!!!!!!.
[7] [7a]
D. J. Darensbourg, P. Rainey, J. Yarbrough, Inorg. Chem.
[7b]
2001, 40, 986Ϫ993.
D. J. Darensbourg, J. C. Yarbrough, J.
Am. Chem. Soc. 2002, 124, 6335Ϫ6342.
[8a] D. Walther, S. Geßler, U. Ritter, A. Schmidt, K. Hamza, W.
[8]
[8b]
Imhof, H. Görls, J. Sieler, Chem. Ber. 1995, 128, 281Ϫ287.
[21] [21a]
A. E. Reed, L. A. Curtiss, F. Weinhold, Chemical Reviews
D. Walther, U. Ritter, R. Kempe, J. Sieler, B. Undeutsch, Chem.
Ber. 1992, 125, 1529Ϫ1536.
[21b]
(Washington, DC, United States) 1988, 88, 899Ϫ926.
E.
[9] [9a]
D. Gledening, A. E. Reed, J. E. Carpenter, F. Weinhold, Natu-
ral Bond Orbital, NBO 3.0 Program Manual, Theoretical
Chemistry Institute and Department of Chemistry, University
of Wisconsin, Madison, WI, 1988.
M. Cheng, D. R. Moore, J. J. Reczek, B. M. Chamberlain,
E. B. Lobkovsky, G. W. Coates, J. Am. Chem. Soc. 2001, 123,
[9b]
8738Ϫ8749.
B. M. Chamberlain, M. Cheng, D. R. Moore,
T. M. Ovitt, E. B. Lobkovsky, G. W. Coates, J. Am. Chem. Soc.
2001, 123, 3229Ϫ3238.
H. Hao, S. Bhandari, Y. Ding, H. W. Roesky, J. Magull, H.-G.
Schmidt, M. Noltemeyer, C. Cui, Eur. J. Inorg. Chem. 2002,
1060Ϫ1065.
[22] [22a]
K. Wolinski, J. F. Hinton, P. Pulay, J. Am. Chem. Soc.
[10]
1990, 112, 8251Ϫ8260. [22b] R. Ditchfield, Mol. Phys. 1997, 27,
[22c]
789Ϫ807.
D. B. Chesnut, C. K. Foley, Chem. Phys. Lett.
1985, 118, 316Ϫ321.
Eur. J. Org. Chem. 2004, 4357Ϫ4372
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4371

D. Hampe, W. Günther, H. Görls, E. Anders
FULL PAPER
[23]
[29b]
E. Anders, A. Opitz, N. J. R. van Eikema Hommes, F. Frankel,
2851Ϫ2859.
M. Bräuer, J. Weston, E. Anders, J. Org.
[29c]
J. Org. Chem. 1993, 58, 4424Ϫ4430.
Chem. 2000, 65, 1193Ϫ1199.
A. Opitz, R. Koch, A. R.
[24] [24a]
Katritzky, W. Fan, E. Anders, J. Org. Chem. 1995, 60,
A. Sygula, P. W. Rabideau, J. Am. Chem. Soc. 1992, 114,
[24b]
3743Ϫ3749.
821Ϫ824.
M. Bühl, N. J. R. v. Eikema Hommes, P. v. R.
[30] [30a]
P. J. Taylor, J. Chem. Soc., Perkin Trans. 2 1972,
1077Ϫ1086. [30b] R. G. Button, P. J. Taylor, J. Chem. Soc., Per-
kin Trans. 2 1973, 557Ϫ567.
H. Benhaoua, F. Texier, R. Carrie, Tetrahedron 1986, 42,
2283Ϫ2291. [31b] R. Huisgen, Angew. Chem. 1963, 75, 604Ϫ637.
S. J. Storfer, E. I. Becker, J. Org. Chem. 1962, 27, 1868Ϫ1876.
COLLECT, Data Collection Software; Nonius B.V., Nether-
lands, 1998.
Z. Otwinowski, W. Minor, ‘‘Processing of X-ray Diffraction
Data Collected in Oscillation Mode’’, in Methods in Enzy-
mology, vol. 276, Macromolecular Crystallography, Part A
(Eds.:C. W. Carter, R. M. Sweet), pp. 307Ϫ326, Academic
Press, 1997.
G. M. Sheldrick, Acta Crystallogr., Sect. A 1990, 46, 467Ϫ473.
G. M. Sheldrick, SHELXL-97, University of Göttingen, Ger-
many, 1997.
CCDC-238419 (for [4e(THF)2]2) and -238420 (for Z-3c) con-
tains the supplementary crystallographic data for this paper.
These data can be obtained free of charge via www.ccdc.cam.a-
c.uk/conts/retrieving.html (or from the Cambridge Crystallo-
graphic Data Centre, 12 Union Road, Cambridge CB2 1EZ,
UK; Fax: ϩ 44-1223-336-033; or deposit@ccdc.cam.ac.uk).
Received July 1, 2004
Schleyer, U. Fleischer, W. Kutzelnigg, J. Am. Chem. Soc. 1991,
113, 2459Ϫ2465.
[25] [25a]
P. S. Patterman, I. L. Karle, G. D. Stucky, J. Am. Chem.
[31] [31a]
[25b]
Soc. 1970, 92, 1150Ϫ1157.
M. A. Beno, H. Hope, M. M.
Olmstead, P. P. Power, Organometallics 1986, 4, 2117Ϫ2124.
[32]
[33]
[25c]
W. Zarges, M. Marsch, K. Harms, G. Boche, Chem. Ber.
[25d]
1989, 122, 2303Ϫ2310.
G. W. Frenking, G. Boche, Chem. Ber. 1991, 124, 543Ϫ549.
W. Zarges, M. Marsch, K. Harms,
[34]
[26] [26a] J. Kaneti, P. v. R. Schleyer, T. Clark, A. Kos, G. W. Spitzna-
gel, J. G. Andrade, J. B. Moffat, J. Am. Chem. Soc. 1986, 104,
1481Ϫ1492. [26b] R. Glaser, A. Streitwieser, J. Org. Chem. 1991,
56, 6612Ϫ6624.
[27] [27a]
G. Wolf, E.-U. Würthwein, Chem. Ber. 1991, 124,
[35]
[36]
[27b]
655Ϫ663.
G. Wolf, E.-U. Würthwein, Chem. Ber. 1991,
124, 889Ϫ896.
[28] [28a]
[37]
J. H. Gilchrist, A. T. Harrison, D. J. Fuller, D. B. Collum,
[28b]
J. Am. Chem. Soc. 1990, 112, 4069Ϫ4070.
Roth, E. M. Michaelides, D. B. Collum, J. Am. Chem. Soc.
1988, 110, 2658Ϫ2660.
A. S. Galiano-
[28c]
J. S. DePue, D. B. Collum, J. Am.
Chem. Soc. 1988, 110, 5518Ϫ5524, 5524Ϫ5533.
For related experimental and theoretical studies see:
[29]
[29a]
M.
Piffl, J. Weston, E. Anders, Eur. J. Org. Chem. 2000,
4372
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 4357Ϫ4372