Diastereoselective dimerisation of alkenylthiazolines: A combined synthetic and computational study

Article abstract of DOI:10.1002/ejoc.200500360

The acylative dimerisation of 2-alkenyl-1,3-thiazolines 1 gives compounds 3 and 8 upon treatment with trichloroacetyl chloride and trifluoroacetic anhydride, respectively. This reaction is completely diastereoselective, in particular giving only a single

Full text of DOI:10.1002/ejoc.200500360

FULL PAPER
Diastereoselective Dimerisation of Alkenylthiazolines: A Combined Synthetic
and Computational Study
Ariane Beutler,[a] Christopher D. Davies,[b] Mark C. Elliott,*[a] Natasha M. Galea,[a]
Matthew S. Long,[a] David J. Willock,[a] and John L. Wood[a]
Keywords: Heterocycles / Thiazolines / Dimerisation / Diastereoselectivity / Semiempirical calculations
The acylative dimerisation of 2-alkenyl-1,3-thiazolines 1 gives
compounds 3 and 8 upon treatment with trichloroacetyl chlo-
ride and trifluoroacetic anhydride, respectively. This reaction
is completely diastereoselective, in particular giving only a
single double-bond isomer. The scope of the reaction has
been evaluated synthetically, while a computational study
has elucidated the mechanism of the reaction and the origin
of stereocontrol.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2005)
acylated dimer was formed once again (Scheme 2). From
this observation it was clear that the trichloroacetyl chloride
was reacting with the thiazoline, and furthermore this acyl-
ation was so rapid that in the presence of zinc-copper cou-
ple, the ketene did not have time to form. The relatively low
yield is a result of losses on chromatography, and we were
able to confirm from NMR spectra of the crude reaction
products that only a single stereoisomer had been formed.
The cis relationship between the phenyl rings was initially
suggested due to the presence of a 3.5-Hz W-coupling be-
tween the two benzylic protons, and was subsequently con-
firmed by nOe experiments. (Scheme 2).
Introduction
Over the last few years we have investigated the annu-
lation reactions for unsaturated azolines with heterocumul-
enes.[1] These reactions provide access to a range of novel
heterocycles with excellent stereocontrol.[2]
As a continuation of this study we anticipated that reac-
tion of a simple alkenylthiazoline– such as 1a– with dichlo-
roketene would be followed by loss of HCl to give com-
pound 2a (Scheme 1). In the event, this reaction turned out
to be far from straightforward, and the results of this study
are described in full herein.[3]
Scheme 1.
Results and Discussion
Scheme 2. a) Cl3CCOCl, DME, Et2O, 25 °C, 3 h, 29%.
When trichloroacetyl chloride was added to a solution of
the alkenylthiazoline 1a containing zinc-copper couple, the
expected product 2a was not obtained. Detailed examina-
tion of the spectroscopic data allowed us to propose struc-
ture 3 for the product, although we were unable at this stage
to elucidate the double-bond geometry. When the reaction
was carried out without the zinc-copper couple present, the
There appears to be no direct precedent for this transfor-
mation, although a related dimerisation of the 2-(benzyl-
ideneamino)benzothiazoline
4
to give
5
is known
(Scheme 3).[4]
Upon consideration of the possible mechanisms, the for-
mation of a single double-bond isomer is particularly sur-
prising. Reaction of an acylated alkenylthiazoline 6 with
compound 1 will ultimately give compound 7, either by a
concerted or stepwise pathway (vide infra). Loss of a proton
from this compound will then give the observed product 3
[a] Cardiff School of Chemistry, Cardiff University,
Park Place, Cardiff, CF10 3AT, UK
[b] AstraZeneca,
Alderley Park, Macclesfield, Cheshire, SK10 4TG, UK
Eur. J. Org. Chem. 2005, 3791–3800
DOI: 10.1002/ejoc.200500360
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3791

M. C. Elliott et al.
FULL PAPER
Scheme 5. a) TFAA, DME, Et2O, 25 °C, 3 h, 41 %.
The alkenylthiazolines 1a–1f were prepared by the iod-
ine-catalysed condensation of 2-methylthiazoline with 4-
substituted benzaldehydes (Scheme 6). The yields for this
reaction, although comparable to those reported in the lit-
erature, are modest.[5,6] As an alternative, compounds 1a
and the tert-butyl analogue 10 were prepared by a two-step
method (Scheme 7).
Scheme 3. a) Xylene, reflux, 60 h, 65%.
(Scheme 4). The stereoselective formation of tetrasubsti-
tuted double bonds is particularly challenging, and since
the environment around the tetrasubstituted double-bond
in 3 is almost symmetrical it would be optimistic to expect
much stereocontrol in the deprotonation of compound 7.
Indeed, semi-empirical calculations showed that the (E) and
(Z) isomers of compound 8 have negligible differences in
energy. We therefore sought to confirm the double-bond ge-
ometry and probe the mechanism of this new reaction.
Scheme 6. a) RC6H4CHO, I2, toluene, reflux, 16 h.
Scheme 4. The basic reaction mechanism.
Despite extensive efforts, compound 3 did not produce
crystals suitable for X-ray diffraction. We therefore decided
to examine the scope and limitations of the reaction in the
hope of producing a crystalline analogue. A range of other
acylating agents was screened for their ability to promote
the dimerisation of compound 1a. Triflic anhydride, 4-tolu-
enesulfonyl chloride, acetyl chloride, acetic anhydride and
methanesulfonyl chloride returned only starting materials
Scheme 7. a) nBuLi, THF,–78 °C then PhCHO, 90%; b) TFA, tolu-
ene, 4-Å molecular sieves, 1 h, 75%; c) i. nBuLi, THF,–78 °C then
tBuCHO; ii. TFA, toluene, reflux, 3 h, 50% over two steps.
The reaction conditions employed to form the parent
under a variety of conditions. The dimerisation was compound 8a were suitable for the 4-methoxyphenyl ana-
smoothly induced by trifluoroacetic anhydride, acylated di- logue 8d, while the 4-bromophenyl, 4-nitrophenyl and 4-
mer 8a being formed in 41% yield (Scheme 5). Compared phenylphenyl analogues 8b, 8c and 8e called for a change
to compound 3, the 5-H in compound 8a was shifted ap- in solvent and elevated temperature (75 °C). The 2-(4-meth-
proximately 0.5 ppm upfield in the proton NMR spectrum. ylphenyl)ethenyl-1,3-thiazoline (1f) did not react under any
We tentatively ascribe this as a difference in the through- conditions; the tert-butyl substrate 10 was also unreactive
space effect between the COCCl3 and COCF3 groups; this (Table 1). Within this series, it is impossible to see any trend
supports the double-bond geometry shown, although it cer- in either reactivity or yield. We tentatively attribute the lack
tainly does not constitute proof.
of reactivity of compound 10 to steric effects, although the
3792
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2005, 3791–3800

Diastereoselective Dimerisation of Alkenylthiazolines
FULL PAPER
apparent lack of reactivity of compound 1f is bewildering.
Having failed to prove the double-bond geometry crys-
None of these compounds gave crystals suitable for X-ray tallographically, we sought indications of the likely geome-
diffraction. The 13C NMR spectroscopic data for these try from the NMR spectroscopic data. Apart from the dif-
compounds warrant brief comment. It is not surprising, ference in the 5-H between compounds 3 and 8a mentioned
given the anticipated low intensity of quaternary carbons, above, the only hint was the upfield shift of two aromatic
that the expected quartets for the trifluoroacetamide car- protons away from the body of the aryl signals, presumably
bonyl peaks could not be assigned with any degree of confi- due to the proximity of the trifluoroacetyl group. These
dence. The CF3 carbon itself in these compounds was also were shown by nOe studies to be the ortho protons on the
of low intensity, and in the case of compound 8b was diffi- phenyl ring at the 5-position, tentatively supporting the
cult to observe. More surprisingly in all of compounds 8b, double-bond geometry as drawn throughout. It seemed rea-
8c and 8d only two distinct aromatic CH peaks were ob- sonable that upon replacement of the trifluoroacetyl group
served in the 13C NMR spectra, whereas the 1H NMR spec- with an acetyl group a nOe from the acetyl CH3 to one of
tra clearly showed the presence of two different, albeit sim- the benzylic ring protons might be seen, and the double
ilar, aromatic rings. Compound 3 also shows only three dis- bond geometry therefore unambiguously proven. The tri-
tinct aromatic CH peaks, and compound 8a shows four. fluoroacetyl group was cleaved reductively using sodium
This coincidence of peaks is unexpected, but all other data borohydride, presenting the dimer as the free enamine 13.
(including the aromatic quaternary carbon atoms in the This reaction was far from clean and the free enamine was
same 13C NMR spectra) are fully consistent with the pro- intolerant to chromatography. Re-acylation of this com-
posed structures.
pound with trifluoroacetic anhydride returned an impure
sample of 8a as the same double-bond isomer. Reaction
with acetic anhydride gave a very impure sample of the de-
sired compound 14 such that no meaningful nOe data could
be obtained (Scheme 9). However, the relative stability and
facile acylation of compound 13 appears to preclude an al-
ternative mechanism for the reaction whereby compound
1a undergoes dimerisation, this being followed by acylation.
Table 1. Acylative dimerisation of 2-alkenyl-1,3-thiazolines.
Com-
R
Conditions
Yield%
pound
8a
8b
8c
8d
8e
Ph
Et2O, DME, 25 °C, 3 h 41
4-BrC6H4
4-O2NC6H4
4-MeOC6H4
4-PhC6H4
CHCl3, reflux, 6 h
CHCl3, reflux, 3 h
43
50
Et2O, DME, 25 °C, 3 h 31
CHCl3, reflux, 6 h
25[a]
[a] Compound 8e was obtained with difficulty from a complex mix-
ture and was not obtained analytically pure.
We also briefly investigated whether the reaction could
be extended to alkenyloxazolines. The 2-styryl-1,3-oxazol-
ine (12) was prepared in two steps from cinnamoyl chloride
and ethanolamine via the amide 11 (Scheme 8). This com-
pound failed to react with trifluoroacetic anhydride under
any of the conditions used for the corresponding thiaz-
olines.
Scheme 9. a) NaBH4, EtOH, 25 °C, 48 h, 80% crude; b) TFAA,
Et3N, DMAP, CH2Cl2; c) Ac2O, DMAP, Et3N, DMAP, 25 °C,
24 h, see discussion.
Returning to our original synthetic goal, the formation
of compounds of general structure 2, it was clear that while
trichloroacetyl chloride is a sufficiently strong electrophile
to induce dimerisation of compounds 1, dichloroacetyl
chloride is not. Since dichloroacetyl chloride can be used as
a dichloroketene precursor,[7] we added triethylamine to a
mixture of compound 1d and dichloroacetyl chloride in
THF. This gave the desired compound 2d in 89% yield
(Scheme 10).[8]
Scheme 8. a) NaOH, CH2Cl2, H2O, 25 °C, 24 h, 78%; b) MsCl,
Et3N, CH2Cl2, 25 °C, 16 h, 90%.
Eur. J. Org. Chem. 2005, 3791–3800
www.eurjoc.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3793

M. C. Elliott et al.
FULL PAPER
grees of freedom were optimised at each point of a search.
The approximate transition states were optimised using the
eigenvector follower approach[11] without constraints and
both minima and transition states were confirmed by a fre-
quency analysis at the same level of theory, minima being
recognised by a complete set of positive normal modes and
transition states by a single negative (imaginary force con-
stant) mode.
The formation of intermediate 7 could proceed via a con-
certed aza-Diels–Alder pathway, or a stepwise mechanism
involving sequential conjugate addition reactions.[12] Proton
loss from this compound then leads to the formation of
compound 8a. We initially had reservations about the abil-
ity of PM3 calculations to predict the stereochemical out-
come of this proton loss. In order to allay these concerns,
we elected to calculate the reaction pathway leading to com-
pound 7, since if we were able to distinguish between con-
certed and stepwise pathways and rationalise the stereo-
chemistry of this reaction, we could be confident that this
level of theory was indeed adequate.
Scheme 10. a) Cl2CHCOCl, Et3N, THF, 25 °C, 2.5 h, 89%.
This reaction may proceed by attack of 1d on in situ-
generated dichloroketene. However, there is a distinct col-
our change on addition of dichloroacetyl chloride to 1d,
which leads us to favour reaction between these two species
followed by deprotonation and cyclisation as the more
likely course of reaction (Scheme 11). Alkenylthiazoline 1a
did not react cleanly under these conditions, possibly re-
flecting the lower nucleophilicity of this compound.
The Concerted Reaction Pathway
This pathway is more straightforward to investigate com-
putationally, as the transition states would be relatively ri-
gid. We investigated both the exo and endo Diels–Alder
pathways for this reaction, but were unable to locate a tran-
sition state in either case. These transition states were also
unsuccessfully sought using DFT calculations (B3LYP 6-
31G** basis set). We were therefore satisfied that the reac-
tion was unlikely to proceed via an initial concerted Diels–
Alder pathway, and so turned our attention to the eventu-
ally more productive stepwise mechanism. This gave en-
tirely reasonable results which explain all of the stereochem-
ical features of the pathway, and so we are confident that
the PM3 level of theory is perfectly adequate in this case.
Scheme 11. Mechanism of formation of compound 2d.
The Stepwise Reaction Pathway
Since the double-bond geometry is defined in the final
abstraction of a proton from an intermediate such as 15,
Computational Studies
Since we had been unable to unequivocally determine the we examined the barrier to rotation of the acylated thiaz-
double-bond geometry by chemical and spectroscopic oline ring in this compound. The rotational profile is essen-
methods, and were unable to rationalise the completely ster- tially sinusoidal with a substantial barrier to rotation
eoselective formation of a tetrasubstituted double bond in (33 kJ·mol–1). Both minima 15a and 15b have similar ener-
an almost symmetrical local environment, we undertook a gies, with the acylated thiazoline ring essentially perpendic-
computational study of the reaction of compound 1a with ular to the bicyclic fragment of the molecule (Figure 1).
trifluoroacetic anhydride in order to attempt to answer Therefore, no preference for either clockwise or anticlock-
these questions. This study was carried out using the PM3[9] wise rotation to give a single product isomer is indicated.
Hamiltonian in the MOPAC program.[10] All calculations
It seemed reasonable that if intermediate 15 was formed
were performed with an SCF tolerance of 10–6 eV and ge- in a given conformation, it would be deprotonated to form
ometry optimisations and transition state searches used a the final product 8a rapidly. This would preclude rotation
gradient cut-off of 4 kJ·mol–1 Å–1. The potential energy sur- about the C6–C2Ј bond and so give rise to a single
face for the reaction was searched using a series of con- double-bond isomer. In order to investigate this possibility,
strained optimisations to approximately locate transition we considered all steps in the proposed reaction mechanism
states between minima. In the following discussion the de- in order to understand the conformational bias at each
gree of freedom constrained will be identified, all other de- stage. Since much of this study did not eventually shed light
3794
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2005, 3791–3800

Diastereoselective Dimerisation of Alkenylthiazolines
FULL PAPER
the intermediate 15 (Scheme 13) and its rotamer. The analo-
gous transition states to form the 5,7-anti diastereomeric
transition states were found to be 19 kJ·mol–1 and
33 kJ·mol–1 higher in energy than 20a and 20b, respectively.
Therefore, irrespective of the double-bond geometry the
5,7-syn arrangement of the phenyl groups is overwhelm-
ingly favoured. It is encouraging that the calculations are
able to reproduce this aspect of the experimental observa-
tions.
Figure 3.
Figure 1. Minimum conformations 15a and 15b of intermediate 15.
on the reason behind the formation of a single double-bond
isomer, it will be presented in a condensed form. The s-trans
conformation 16 of the free thiazoline is slightly favoured
(3.5 kJ·mol–1), while after acylation the s-cis conformation
17 is more favoured (13 kJ·mol–1) (Figure 2). While these
conformations do in fact give rise to the lowest energy tran-
sition states and intermediates along the reaction coordi-
nate, the entire reaction pathway was investigated beginning
with the s-cis and s-trans conformers of each.
Scheme 13.
Although these calculations had established the likely
conformation of the various intermediates and transition
states along the reaction pathway, conformations 15a and
15b (of essentially identical energy) were shown to be fav-
oured from 20a and 20b, respectively. We therefore still had
no basis upon which to predict which double-bond isomer
would be formed, or even that any selectivity would be ob-
served. Up to this point the anion for the cationic interme-
diates had been omitted for simplicity, not least because its
location is difficult to determine. However, for the final step
of hydrogen abstraction the inclusion of a trifluoroacetate
anion to receive the proton was found to be crucial. At this
point the positioning of the anion is also less ambiguous
since it must be placed with the carboxylate group in close
proximity to C6-H. The bulk of the phenyl substituents also
place additional limitations on its position. With the anion
in place the calculated energy difference between the two
Figure 2.
Location of the transition state 18 for nucleophilic attack
of 16 onto 17 proved straightforward. In intermediate 19
the double-bond geometry for the ring closure is fixed
(Scheme 12), with the transition state forming the opposite
double-bond isomer being 13 kJ·mol–1 higher in energy.
–
–
intermediate structures 15a·CF3CO2 and 15b·CF3CO2 is
now significant as shown in Figure 4.
The introduction of the anion causes significant confor-
mational change of the phenyl substituents to allow its in-
teraction with the C6 proton. This change is more notable
at the 5-phenyl than the 7-phenyl which is held more rigidly
in place by the C=C bond of the six membered heterocycle.
However the C6-H and C–N bonds are still found to be
near co-planar and so the geometry of the product double
Scheme 12.
Of the four possible transition states for the actual ring- bond is still ambiguous at this stage. Transition states for
closure step, structures 20a and 20b are shown schemati- the proton removal were located using two methods. Firstly
cally in Figure 3. These were located by systematically in- a series of constrained optimisations with the C6-H bond
creasing the corresponding carbon–carbon bond length in length systematically increased was explored and secondly
Eur. J. Org. Chem. 2005, 3791–3800
www.eurjoc.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3795

M. C. Elliott et al.
FULL PAPER
tra were recorded with a Perkin–Elmer 1600 FTIR spectrophotom-
eter. Mass spectra were recorded with a Fisons VG Platform II
spectrometer. High-resolution mass spectra were performed at the
EPSRC centre for Mass Spectroscopy in Swansea. NMR spectra
were recorded at 25 °C with a Bruker DPX 400 spectrometer op-
erating at 400 MHz for 1H and at 100 MHz for 13C or with a
Bruker Avance 500 spectrometer operating at 500 MHz for 1H and
125 MHz for 13C. All chemical shifts are reported in ppm down-
field from TMS. Coupling constants (J) are reported in Hz. Multi-
plicity in 1H NMR is reported as singlet (s), doublet (d), double
doublet (dd), triplet (t), and multiplet (m). Multiplicity in 13C
NMR was obtained using the DEPT pulse sequence. Flash
chromatography was performed using matrex silica 60 35–70 mi-
cron.
4,5-Dihydro-2-[(E)-2-phenylethenyl]-1,3-thiazole (1a): 2-Methylthi-
azoline (5.06 g, 50 mmol), and benzaldehyde (5.36 g, 50.0 mol)
were heated under reflux in toluene (50 mL) under Dean–Stark
conditions. After 1 h iodine (100 mg) was added and reflux allowed
to continue for 16 h. The reaction mixture was washed with satu-
rated sodium thiosulfate solution (100 mL), dried with MgSO4, and
the solvent removed in vacuo. The resulting thick oil was dissolved
in diethyl ether and filtered, the filtrate was concentrated in vacuo
and distilled under reduced pressure (kugelrohr, oven temperature
180 °C at 0.5 Torr), to give 1a (2.8 g, 30%) as a white crystalline
Figure 4.
solid, m.p. 103–105 °C (ref.[6] m.p. 101–102 °C). IR (CH Cl ): ν =
˜
2
2
by rotation of the acylated thiazoline ring from each of the
intermediates 15a·CF3CO2– and 15b·CF3CO2– without con-
straining the C6-H distance. Both approaches led to the
same conclusion that proton loss occurs as the five-mem-
bered ring rotates to become co–planar with the six–mem-
bered heterocycle, so that the formation of the new double
bond occurs simultaneously with proton loss to form the
neutral acylated dimer. From 15a·CF3CO2– we obtained the
1583, 1462, 1377 cm–1. 1H NMR (400 MHz, CDCl3, 25 °C): δ =
7.43 [dd, 3J(H,H) = 7.9, 4JH,H = 1.3, 2 H, aromatic CH], 7.40–7.25
3
(m, 3 H, aromatic CH), 7.05 (d, JH,H = 16.2, 1 H, one of alkene
3
3
CH), 6.95 (d, JH,H = 16.2, 1 H, one of alkene CH), 4.30 (t, JH,H
= 8.2, 2 H, CH2N), 3.25 (t, JH,H = 8.2, 2 H, CH2S) ppm. 13C
3
NMR (100 MHz, CDCl3, 25 °C): δ = 168.3 (thiazoline C), 141.6
(alkene CH), 135.7 (aromatic C), 129.8 (aromatic CH) 129.3 (aro-
matic CH), 127.9 (aromatic CH), 123.1 (alkene CH), 65.1 (CH2N),
33.4 (CH2S) ppm. MS (APCI): m/z (%) = 190 (100) [M + H]+.
–
preferred transition state 21a and from 15b·CF3CO2 the
transition state 21b was located. Both structures therefore
lead to the same (E) double bond isomer. Rotational se-
arches to find the corresponding (Z) isomer failed since the
steric hindrance between the trifluoroacetyl group and the
7-phenyl prevented the five-membered ring rotating to be-
come co-planar with the piperidine ring. It appears, then,
that selectivity is controlled by the ease of rotation of the
acylated thiazoline ring to form the new C=C bond. This
is impeded by steric interactions between the trifluoroacetyl
group and the phenyl substituents. In the case of the (Z)
isomer, the 7-phenyl group has less conformational mobility
so that rotation to give the (E) isomer is the preferred pro-
cess.
2-[(E)-2-(4-Bromophenyl)ethenyl]-4,5-dihydro-1,3-thiazole (1b):
2-Methylthiazoline (2.2 g, 20 mmol) and 4-bromobenzaldehyde
(3.6 g, 20 mmol), were heated under reflux in toluene (100 mL) un-
der Dean–Stark conditions for 1 h. Iodine (100 mg) was added and
the reflux continued for 16 h. The resulting reaction mixture was
washed with saturated sodium thiosulfate solution (100 mL), dried
with MgSO4, and the solvent removed in vacuo. The dark oil re-
sulting was dissolved in diethyl ether (40 mL) and filtered. The di-
ethyl ether was removed in vacuo to give a brown solid which was
purified by column chromatography (eluting with diethyl ether/
CH2Cl2, 1:9), late fractions yielding 1b (1.6 g, 30%) as a white so-
lid; m.p. 160–163 °C. IR (CDCl ): ν = 1634, 1574, 1489, 1007 cm–1.
˜
3
1H NMR (400 MHz, CDCl3, 25 °C): δ = 7.39 (d, JH,H = 7.4, 2 H,
3
3
aromatic CH), 7.29 (d, JH,H = 7.4, 2 H, aromatic CH), 6.95 (ap-
3
parent s, 2 H, 2×alkene CH), 4.30 (t, JH,H = 8.0, 2 H, CH2N),
3.25 (t, 3JH,H = 8.0, 2 H, CH2S) ppm. 13C NMR (100 MHz, CDCl3,
25 °C): δ = 168.0 (thiazoline C), 140.1 (alkene CH), 134.7 (aromatic
C) 132.5 (aromatic CH), 129.3 (aromatic CH), 123.9 (aromatic C),
123.7 (alkene CH), 65.2 (CH2N), 33.5 (CH2S) ppm. MS (APCI):
m/z (%) = 270 (100) [M + H]+, 268 (100) [M + H]+, 71 (58). HRMS
(CI) C11H1179BrNS [M + H]+: 267.9795; found 267.9797.
Conclusions
A novel stereoselective acylative dimerisation reaction of
alkenylthiazolines has been investigated experimentally and
computationally. The calculations show the importance of
including counterions in the structure in order to fully ex-
plain the stereoselectivity in such reactions.
4,5-Dihydro-2-[(E)-2-(4-nitrophenyl)ethenyl]-1,3-thiazole (1c):
2-Methylthiazoline (2.20 g, 20 mmol) and 4-nitrobenzaldehyde
(3.0 g, 22 mmol) were heated under reflux in toluene (100 mL) un-
der Dean–Stark conditions for 1 h. Iodine (100 mg) was added and
the reflux continued for 16 h. The reaction mixture was washed
with sodium thiosulfate solution (100 mL), dried with MgSO4 and
the toluene removed in vacuo. Diethyl ether (40 mL) was added to
Experimental Section
General Remarks: Melting points were determined with a Gallen-
kamp melting point apparatus and are uncorrected. Infrared spec-
3796
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2005, 3791–3800

Diastereoselective Dimerisation of Alkenylthiazolines
FULL PAPER
the resulting dark oil, and 1c (1.45 g, 31%) isolated by filtration as
a light brown solid; m.p. 213–216 °C, (ref.[6] m.p. 214–216 °C)
Dean–Stark conditions for 1 h. Iodine (100 mg) was added and the
reflux continued for 16 h. The reaction mixture was washed with
saturated sodium sulfite solution (40 mL). The solvent was re-
moved in vacuo to give a brown solid which was purified by column
chromatography (eluting with CH2Cl2), late fractions affording 1e
which was used without further purification. IR (CH Cl ): ν= 1629,
˜
2
2
1604, 1508, 1167 cm–1. 1H NMR (400 MHz, CDCl3, 25 °C): δ =
3
3
8.25 (d, JH,H = 8.8, 2 H, aromatic CH), 7.64 (d, JH,H = 8.8, 2 H,
3
aromatic CH), 7.17 (d, JH,H = 16.4, 1 H, one of alkene CH), 7.10
(1.8 g, 35%) as a white solid, m.p. 148–150 °C (ref.[6] m.p. 148–
(d, 3JH,H = 16.4, 1 H, one of alkene CH), 4.40 (t, JH,H = 8.2, 2 H,
150 °C). IR (CH Cl ): ν = 1634, 1579, 1318, 1177 cm–1. H NMR
3
1
˜
2
2
CH2N), 3.35 (t, JH,H = 8.2, 2 H, CH2S). 13C NMR (100 MHz, (400 MHz, CDCl3, 25 °C): δ = 7.31 (d, JH,H = 8.0, 2 H, aromatic
3
3
3
3
CDCl3, 25 °C): δ = 167.6 (thiazoline C), 148.3 (aromatic C), 141.9
(aromatic C), 139.5 (alkene CH), 128.3 (aromatic CH), 127.1 (al-
kene CH), 124.6 (aromatic CH), 65.4 (CH2N), 33.7 (CH2S) ppm. alkene CH), 4.30 (t, JH,H = 8.1, 2 H, CH2N), 3.26 (t, JH,H = 8.1,
MS (ammonia CI): m/z (%) = 235 (26) [M + H]+, 220 (33), 205 2 H, CH2S), 2.30 (s, 3 H, CH3) ppm. 13C NMR (100 MHz. CDCl3,
(100). HRMS (CI) C11H11N2O2S [M + H]+: 235.0541; found 25 °C): δ = 168.3 (thiazoline C), 141.6 (alkene CH), 140.1 (aromatic
CH), 7.11 (d, JH,H = 8.0, 2 H, aromatic CH), 6.99 (d, JH,H =
3
15.8, 1 H, one of alkene CH), 6.93 (d, JH,H = 15.8, 1 H, one of
3
3
235.0545.
C), 133.0 (aromatic C), 130.0 and 127.8 (aromatic CH), 122.1 (al-
kene CH), 65.1 (CH2N), 33.4 (CH2S), 21.8 (CH3) ppm. MS
(APCI): m/z (%) = 204 (100) [M + H]+.
4,5-Dihydro-2-[(E)-2-(4-methoxyphenyl)ethenyl]-1,3-thiazole (1d):
2-Methylthiazoline (2.5 g, 25 mmol) and 4-methoxybenzaldehyde
(2.8 g, 24 mmol) were heated under reflux in toluene (100 mL) un-
der Dean–Stark conditions for 1 h. Iodine (50 mg) was added and
the reflux continued for 18 h. The resulting brown solution was
washed with saturated sodium thiosulfate solution (50 mL), dried
with MgSO4 and the solvent removed in vacuo. The brown oil re-
maining was dissolved in diethyl ether (30 mL), filtered and the
diethyl ether was removed in vacuo, to yield a brown solid which
was purified by column chromatography (eluting with diethyl ether)
late fractions yielding 1d (1.8 g, 33%,) as colourless plates, m.p.
6-Chloro-2,3-dihydro-7-(4-methoxyphenyl)-5H-[1,3]thiazolo[3,2-a]-
pyridin-5-one (2d): 4,5-Dihydro-2-[(E)-2-(4-methoxyphenyl)eth-
enyl]-1,3-thiazole (1d) (520 mg, 2.4 mmol) was dissolved in dry
THF (8 mL). Dichloroacetyl chloride (230 µL, 2.24 mmol) was
added and the resulting bright red mixture stirred at 25 °C under
N2 for 20 min. Triethylamine (340 µL, 2.4 mmol) was added and
the resulting cloudy brown mixture stirred for a further 2 h. The
resulting suspension was filtered and the filtrate dried with MgSO4,
and concentrated in vacuo. Purification by column chromatog-
raphy (eluting with diethyl ether/CH2Cl2, 1:5) yielded 2d (625 mg,
100–102 °C (ref.[6] m.p. 99–101 °C). IR (CH Cl ): ν = 1463,
˜
2
2
1377 cm–1. H NMR (400 MHz, CDCl3, 25 °C): δ = 7.45 (d, JH,H
1
3
89%) as a light brown solid m.p. 128–130 °C. IR (CHCl ): ν =
˜
3
= 8.7, 2 H, aromatic CH), 7.07 (d, 3JH,H = 16.1, 1 H, one of alkene
1636, 1513 cm–1. H NMR (400 MHz, CDCl3, 25 °C): δ = 7.33 (d,
1
3
3
CH), 6.93 (d, JH,H = 16.1, 1 H, one of alkene CH), 6.89 (d, JH,H
3JH,H = 8.6, 2 H, aromatic CH), 6.87 (d, 3JH,H = 8.6, 2 H, aromatic
3
= 8.7, 2 H, aromatic CH), 4.35 (t, JH,H = 8.5, 2 H, CH2N), 3.85
3
CH), 6.15 (s, 1 H, 8-H), 4.52 (d, JH,H = 7.4, 2 H, CH2N), 3.79 (s,
(s, 3 H, OCH3), 3.30 (t, JH,H = 8.5, 2 H, CH2S) ppm. 13C NMR
3
3 H, OCH3), 3.41 (t, JH,H = 7.4, 2 H, CH2S) ppm. 13C NMR
3
(125 MHz. CDCl3, 25 °C): δ = 168.5 (thiazoline C), 160.8 (aromatic
C), 141.2 (alkene CH), 129.1 (aromatic CH), 128.0 (aromatic C),
120.2 (alkene CH), 114.3 (aromatic CH), 64.3 (CH2N) 55.4
(OCH3), 32.9 (CH2S) ppm. MS (APCI): m/z (%) = 220 (100) [M +
H]+.
(100 MHz. CDCl3, 25 °C): δ = 159.0 (C=O), 157.6 (SCN), 149.6
(CCl), 144.4 (aromatic C), 128.9 (aromatic CH), 128.3 (aromatic
C), 116.7 (aromatic C), 112.7 (aromatic CH), 101.4 (8-C), 52.3
(OCH3), 50.8 (CH2N), 27.9 (CH2S) ppm. MS (APCI): m/z (%) =
2 9 6 ( 3 5 ) [ M + H ] + , 2 9 4 ( 1 0 0 ) [ M + H ] + . H R M S ( C I )
C12H13NO4S35Cl [M + H]+: 294.0355; found 294.0359.
2-[(E)-2-(4-Phenylphenyl)ethenyl]-4,5-dihydro-1,3-thiazole (1e):
2-Methylthiazoline (1.75 g, 17 mmol) and 4-phenylbenzaldehyde
(3.15 g, 17 mmol) were heated under reflux in toluene (100 mL)
under Dean–Stark conditions for 1 h. Iodine (100 mg) was added
and the reflux continued for 16 h. The reaction mixture was washed
with saturated sodium thiosulfate solution (100 mL), dried with
MgSO4, and the toluene removed in vacuo. The resulting dark oil
was dissolved in diethyl ether (40 mL) and filtered. The diethyl
ether was removed in vacuo to give a brown solid which was puri-
fied by column chromatography (eluting with CH2Cl2), late frac-
tions yielding 1e (960 mg, 21%) as an off-white crystalline solid,
(5RS,7RS,E)-6-(4,5-Dihydro-3-trichloroacetyl-1,3-thiazol-2-ylid-
ene)-2,3,5,7-tetrahydro-5,7-diphenyl[1,3]thiazolo[3,2-a]pyridine (3):
4,5-Dihydro-2-[(E)-2-phenylethenyl]-1,3-thiazole (1a) (250 mg,
1.3 mmol) was dissolved in dry diethyl ether (1 mL) and dry DME
(4 mL). Trichloroacetyl chloride (15 µL, 0.8 mmol) was added
dropwise with stirring under N2 at 25 °C. The reaction mixture was
stirred for 3 h, washed with aqueous NaHCO3 solution (10 mL)
then with brine (10 mL), and dried with MgSO4. The solvent was
removed in vacuo to yield a glassy orange solid, which was purified
by chromatography on neutral alumina (eluting with CH2Cl2) late
m.p. 141–144 °C. IR (CH Cl ): ν = 1633, 1582, 1488, 1177 cm–1.
˜
2
2
fractions yielding compound 3 as a colourless solid (98 mg, 29%),
1H NMR (400 MHz, CDCl3, 25 °C): δ = 7.74–7.56 (m, 6 H, aro-
1
m.p. 186–188 °C. IR (CHCl ): ν = 2920, 1691, 1625 cm –1. H
˜
3
3
matic CH), 7.46 (apparent t, JH,H = 7.5, 2 H, aromatic CH), 7.36
NMR (400 MHz, CDCl3, 25 °C): δ = 7.20–7.10 (m, 8 H, aromatic
3
3
(t, JH,H = 7.4, 1 H, aromatic CH), 7.26 (d, JH,H = 16.2, 1 H,
4
CH), 6.90 (m, 2 H, aromatic CH), 5.05 (d, JH,H = 3.5, 1 H, 5-H),
3
3
alkene CH), 7.08 (d, JH,H = 16.2, 1 H, alkene CH), 4.40 (t, JH,H
3
4.55 (d, JH,H = 11.8, 1 H, alkene CH), 3.95–3.85 (m, 2 H,
= 8.2, 2 H, CH2N), 3.36 (t, JH,H = 8.2, 2 H, CH2S) ppm. 13C
3
CH2NCO), 3.50–3.35 (m, 3 H, 7-H and CH2N), 3.10–2.95 (m, 2
H, CH2S), 2.90–2.80 (m, 2 H, CH2S) ppm. 13C NMR (100 MHz.
CDCl3, 25 °C): δ = 175.6 (C=O), 168.7, 167.2 (both thiazoline C),
139.6 and 138.7 (aromatic C), 128.5, 127.8, 127.2 (all aromatic
CH), 97.5 (CCl3), 96.7 (alkene C), 64.5 (CH2NCO), 60.5 (alkene
CH), 52.9 (CH2N), 50.0 (5-C), 43.6 (7-C), 33.3 (CH2S), 27.0
(CH2S) ppm. MS (APCI): m/z (%) = 525 (9) [M + H]+, 523 (10) [M
NMR (100 MHz. CDCl3, 25 °C): δ = 168.3 (thiazoline C), 142.2
(aromatic C), 141.1 (alkene CH), 140.7 (aromatic C), 134.7 (aro-
matic C), 129.3, 128.3, 128.1, 127.9 and 127.4 (all aromatic CH),
123.0 (alkene CH), 65.2 (CH2N), 33.5 (CH2S) ppm. MS (APCI):
m/z (%) = 266 (100) [M + H]+. HRMS (CI) C17H16NS [M + H]+:
266.1003; found 266.0997.
2-[(E)-2-(4-Methylphenyl)ethenyl]-4,5-dihydro-1,3-thiazole (1f): + H]+, 190 (100) [M + H]+ (molecular ion shows expected isotopic
2-Methylthiazoline (2.5 g, 25 mmol) and 4-tolualdehyde (2.84 g,
25 mmol) were heated under reflux in toluene (100 mL) under
distribution pattern). HRMS (CI) C24H2235Cl3N2O2S [M + H]+:
523.0239; found 523.0240.
Eur. J. Org. Chem. 2005, 3791–3800
www.eurjoc.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3797

M. C. Elliott et al.
FULL PAPER
(5RS,7RS,E)-6-(4,5-Dihydro-3-trifluoroacetyl-1,3-thiazol-2-ylid-
ene)-2,3,5,7-tetrahydro-5,7-diphenyl[1,3]thiazolo[3,2-a]pyridine (8a):
2-[(E)-2-Phenylethenyl]-4,5-dihydro-1,3-thiazole (1a) (150 mg,
0.8 mmol) was dissolved in dry diethyl ether (1 mL) and dry DME
(4 mL). Trifluoroacetic anhydride (70 µL, 0.5 mmol) was added
dropwise with stirring under N2 at 25 °C. The reaction mixture was
stirred for 3 h, washed with aqueous NaHCO3 solution (10 mL)
then with brine (10 mL), and dried with MgSO4. The solvent was
removed in vacuo to yield a glassy orange solid, which was purified
by chromatography on neutral alumina (eluting with CH2Cl2), late
fractions yielding 8a (78 mg, 41 %) as a white solid, m.p. 196–
= 8.7, 2 H, aromatic CH), 7.40 (d, 3JH,H = 8.6, 2 H, aromatic CH),
3
3
7.06 (d, JH,H = 8.7, 2 H, aromatic CH), 4.66 (d, JH,H = 11.1, 1
H, alkene CH), 4.61 (d, 4JH,H = 4.1, 1 H, 5-H), 3.71–3.65 (m, 2 H,
3
4
CH2NCO), 3.49 (dd, JH,H = 11.1, JH,H = 4.1, 1 H, 7-H), 3.45–
2
3
3.40 (m, 1 H, one of CH2N) 3.31 (ddd, JH,H = 10.6, JH,H = 7.7,
3JH,H = 2.9, 1 H, one of CH2N), 3.17–2.98 (m, 4 H, 2×CH2S)
ppm. 13C NMR (100 MHz. CDCl3, 25 °C): δ = 168.4, 166.7 (both
thiazoline C), 148.5, 148.1, 147.7, 145.4 (all aromatic C), 129.2,
1
123.8 (both aromatic CH), 118.1 (q, JC,F = 290 Hz, CF3), 98.2
(alkene C), 64.7 (CH2NCO), 59.7 (alkene CH), 53.3 (CH2N), 49.6
(5-C), 41.7 (7-C), 33.9 (CH2S), 27.7 (CH2S) ppm. MS (APCI): m/z
(%) = 565 (100) [M + H]+, 65 (57). HRMS (CI) C24H20N4O5F3S2
[M + H]+: 565.0828; found 565.0857.
198 °C. IR (CHCl ): ν = 1691, 1625, 1499 cm–1
.
1H NMR
˜
3
(400 MHz, CDCl3, 25 °C): δ = 7.32–7.12 (m, 8 H, aromatic CH),
3
4
6.85 (dd, JH,H = 7.5, JH,H = 1.7, 2 H, aromatic CH), 4.56 (d,
(5RS,7RS,E)-2,3,5,7-Tetrahydro-5,7-bis(4-methoxyphenyl)-6-(4,5-di-
hydro-3-trifluoroacetyl-1,3-thiazol-2-ylidene)[1,3]thiazolo[3,2-a]pyri-
dine (8d): 2-[(E)-2-(4-Methoxyphenyl)ethenyl]-4,5-dihydro-1,3-thi-
azole (1d) (200 mg, 0.93 mmol) was dissolved in dry diethyl ether
(1 mL) and dry DME (4 mL). Trifluoroacetic anhydride (90 µL,
0.64 mmol) was added dropwise and the solution stirred under N2
at 25 °C for 3 h. The orange solution resulting was washed with
aqueous NaHCO3 solution (10 mL), then brine (10 mL), and dried
with MgSO4. The solvent was removed in vacuo to give an orange
solid which was purified by flash chromatography on alumina (elut-
3JH,H = 11.1, 1 H, alkene CH), 4.51 (d, JH,H = 4.2, 1 H, 5-H),
4
3.79 (m, 2 H, CH2NCO), 3.41 (dd, 3JH,H = 11.1, JH,H = 4.2, 1 H,
4
7-H), 3.35–3.30 (m, 2 H, CH2N), 3.00–2.80 (m, 4 H, 2×SCH2)
ppm. 13C NMR (100 MHz. CDCl3, 25 °C): δ = 166.5 and 164.6
(both thiazoline C), 139.5 and 136.9 (both aromatic C), 127.6,
1
127.1, 126.9, 126.3 (all aromatic CH), 117.1 (q, JC,F = 290 Hz,
CF3), 97.9 (alkene C), 70.7 (NCOCH2), 59.1 (alkene CH), 51.8
(NCH2), 48.8 (5-C), 40.7 (7-C), 32.3 (CH2S), 29.2 (CH2S) ppm.
MS (APCI): m/z (%) = 475 (100) [M + H]+, 102.5 (29). HRMS
(EI) C24H22F3N2OS2 [M + H]+: 475.1127; found 475.1125.
ing with CH2Cl2) yielding 8d (77 mg, 31%) as a pale yellow solid;
1
m.p. 183–185 °C. IR (CH Cl ): ν = 1683, 1624, 1463 cm–1. H
(5RS,7RS,E)-2,3,5,7-Tetrahydro-5,7-bis(4-bromophenyl)-6-(4,5-di-
hydro-3-trifluoroacetyl-1,3-thiazol-2-ylidene)[1,3]thiazolo[3,2-a]pyri-
dine (8b): 2-[(E)-2-(4-Bromophenyl)ethenyl]-4,5-dihydro-1,3-thi-
azole (1b) (200 mg, 0.74 mmol), was dissolved in dry chloroform
(5 mL). Trifluoroacetic anhydride (75 µL, 0.53 mmol) was added
dropwise and the reaction mixture heated under reflux for 6 h. The
resulting orange solution was washed with aqueous NaHCO3 solu-
tion (10 mL) then brine (10 mL), and dried with MgSO4. The sol-
vent was removed in vacuo to give a light brown solid which was
purified by column chromatography on alumina (eluting with
CH2Cl2) early fractions yielding 8b (100 mg, 43%) as a white solid,
˜
2
2
3
NMR (400 MHz, CDCl3, 25 °C): δ = 7.10 (d, JH,H = 8.8, 2 H,
aromatic CH), 6.80–6.69 (m, 6 H, aromatic CH), 4.52 (d, JH,H
3
=
4
11.1, 1 H, alkene CH), 4.45 (d, JH,H = 3.9, 1 H, 5-H), 3.83 (m, 2
H, CH2NCO), 3.72 (s, 3 H, OCH3), 3.71 (s, 3 H, OCH3), 3.40 (m,
3 H, CH2N and 7-H), 3.05 (m, 2 H, CH2S), 2.93 (m, 2 H, CH2S)
ppm. 13C NMR (100 MHz. CDCl3, 25 °C): δ = 167.9, 167.7 (both
thiazoline C), 160.0, 158.9, 133.0, 130.2 (all aromatic C), 129.5
1
(aromatic CH), 118.0 (q, JC,F = 290 Hz, CF3), 113.6 (aromatic
CH), 99.7 (alkene C), 64.6 (CH2NCO), 60.1 (alkene CH), 55.6
(OCH3), 55.4 (OCH3), 53.2 (CH2N), 50.4 (5-C), 41.1 (7-C), 33.7
(CH2S), 27.6 (CH2S) ppm. MS (APCI): m/z (%) = 535 (100) [M +
H]+. HRMS (EI) C24H20N4O5F3S2 [M]+: 534.1259; found
534.1260.
m.p. 200–203 °C. IR (CH Cl ): ν = 1692, 1625, 1500, 1464 cm–1.
˜
2
2
1H NMR (400 MHz, CDCl3, 25 °C): δ = 7.38 (d, JH,H = 8.4, 2 H,
3
3
aromatic CH), 7.30 (d, JH,H = 8.4, 2 H, aromatic CH), 7.08 (d,
3JH,H = 8.5, 2 H, aromatic CH), 6.73 (d, 3JH,H = 8.5, 2 H, aromatic
3
4
(5RS,7RS,E)-2,3,5,7-Tetrahydro-5,7-bis(4-phenylphenyl)-6-(4,5-di-
hydro-3-trifluoroacetyl-1,3-thiazol-2-ylidene)[1,3]thiazolo[3,2-a]pyri-
dine (8e): 2-[(E)-2-(4-Phenylphenyl)ethenyl]-4,5-dihydro-1,3-thi-
azole (1e) (150 mg, 0.57 mmol) was dissolved in dry chloroform
(5 mL). Trifluoroacetic anhydride (70 µL, 0.5 mmol) was added
dropwise and the solution heated under reflux for 6 h. The resulting
orange solution was washed with aqueous NaHCO3 solution
(10 mL), then brine (10 mL), and dried with MgSO4. The solvent
was removed in vacuo to give a yellow solid which was purified
twice by flash chromatography on alumina (eluting with CH2Cl2),
yielding 8e (45 mg, 25%) as a pale yellow solid. This compound
exhibited spectroscopic data in line with compounds 8a–8d, but
could not be purified to allow full characterisation.
CH), 4.50 (d, JH,H = 11.1, 1 H, alkene CH), 4.47 (d, JH,H = 4.2,
1 H, 5-H), 3.80 (apparent t, JH,H = 8.3, 2 H, CH2NCO), 3.40–
3
3.30 (m, 3 H, 7-H and CH2N), 3.11–3.03 (m, 2 H, CH2S), 2.95–
2.88 (m, 2 H, CH2S) ppm. 13C NMR (125 MHz. CDCl3, 25 °C): δ
= 166.5 and 165.8 (both thiazoline C), 138.4 and 136.0 (both aro-
matic C), 130.1, 128.6 (both aromatic CH), 121.7, 120.3 (both C–
Br), 97.5 (alkene C), 63.2 (CH2NCO), 58.5 (alkene CH), 51.8
(NCH2), 48.4 (5-C), 40.0 (7-C), 32.4 (SCH2), 26.2 (SCH2) ppm.
MS (APCI): m/z (%) = 635 (58) [M + H]+, 633 (100) [M + H]+,
631 (48) [M + H]+. HRMS (CI) C24H1979Br2F3N2OS2 [M + H]+:
630.9336; found 630.9335.
(5RS,7RS,E)-6-(4,5-Dihydro-3-trifluoroacetyl-1,3-thiazol-2-ylid-
ene)-2,3,5,7-tetrahydro-5,7-bis(4-nitrophenyl)[1,3]thiazolo[3,2-a]pyri-
dine (8c): 4,5-Dihydro-2-[(E)-2-(4-nitrophenyl)ethenyl]-1,3-thiazole
(1c) (200 mg, 0.85 mmol) was dissolved in dry chloroform (5 mL).
Trifluoroacetic anhydride (83 µL, 0.6 mmol) was added dropwise
and the solution heated under reflux for 3 h. The resulting orange
solution was washed with aqueous NaHCO3 solution (10 mL), then
brine (10 mL), and dried with MgSO4. The solvent was removed
in vacuo to give a yellow solid which was purified by flash
chromatography on alumina (eluting with CH2Cl2), yielding 8c
2-(4,5-Dihydro-1,3-thiazol-2-yl)-1-phenyl-1-ethanol (9):[13] 2-Methyl-
thiazoline (2.0 g, 20 mmol) was dissolved in dry THF and cooled
to–78 °C. n-Butyllithium (8.7 mL of a 2.5  solution in hexanes,
22 mmol) was added and the resulting orange solution stirred at–
78 °C under N2 for 1 h. Benzaldehyde (2.1 g, 20 mmol) was added
and the reaction mixture stirred for a further 2 h at–78 °C then
allowed to warm to 25 °C over 1 h. Saturated ammonium chloride
solution (20 mL) was added, and the products extracted with di-
(120 mg, 50 %) as a pale yellow solid, m.p. 210–212 °C. IR ethyl ether (3 × 30 mL), the organic portions were washed with
(CH Cl ): ν = 1683, 1625, 1462 cm–1. 1H NMR (400 MHz, CDCl ,
25 °C): δ = 8.13 (d, JH,H = 8.3, 2 H, aromatic CH), 8.09 (d, JH,H
brine (20 mL), dried with MgSO4, and concentrated in vacuo to
give compound 9 (3.6 g, 90%) as an orange solid, m.p. 76–79 °C
˜
2
2
3
3
3
3798
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2005, 3791–3800

Diastereoselective Dimerisation of Alkenylthiazolines
(dec.). IR (CH Cl ): ν = 3267, 1624, 1434, 1124 cm–1. 1H NMR
FULL PAPER
120.6 (alkene CH), 62.7 (CH2O), 43.1 (CH2N) ppm. MS (APCI):
m/z (%) = 192 (100) [M + H]+.
˜
2
2
3
(400 MHz, CDCl3, 25 °C): δ = 7.30 (apparent t, JH,H = 7.5, 2 H,
3
aromatic CH), 7.26 (d, JH,H = 7.6, 2 H, aromatic CH), 7.19 (t,
4,5-Dihydro-2-[(E)-2-phenylethenyl]-1,3-oxazole (12): (E)-N-(2-
Hydroxyethyl)-3-phenyl-2-propenamide (11) (1.1 g 5.75 mmol), was
dissolved in CH2Cl2 (50 mL), triethylamine (580 mg, 5.75 mmol)
was added and the mixture cooled to 0 °C. Methanesulfonyl chlo-
ride (665 mg, 5.75 mmol) was added and the resulting solution
stirred at 25 °C for 24 h. The reaction mixture was washed with
NaHCO3 solution (10 mL), dried with MgSO4, and concentrated
in vacuo to afford a green oil which was purified by column
chromatography (eluting with ethyl acetate/diethyl ether, 1:10),
early fractions yielding compound 12 (895 mg, 90%) as a white
3
3JH,H = 6.9, 1 H, aromatic CH), 5.04 (apparent t, JH,H = 6.3, 1
H, CHOH), 4.70 (broad s, 1 H, OH), 4.18 (m, 2 H, CH2N), 3.21
3
(apparent t, JH,H = 8.2, 2 H, CH2S), 2.72 (m, 2 H, CH2CHOH)
ppm. 13C NMR (100 MHz. CDCl3, 25 °C): δ = 171.0 (thiazoline
C), 142.3 (aromatic C), 128.8, 128.0 and 126.2 (aromatic CH), 71.8
(CHOH), 64.6 (CH2N), 43.1 (CH2S), 33.8 (CH2CHOH) ppm. MS
(APCI): m/z (%) = 208 (40) [M + H]+, 190 (100).
Alternative Preparation of 2-[(E)-2-Phenylethenyl]-4,5-dihydro-1,3-
thiazole (1a): 2-(4,5-Dihydro-1,3-thiazol-2-yl)-1-phenyl-1-ethanol
(9) (1.5 g, 7.2 mmol) was dissolved in toluene (15 mL). Trifluoro-
acetic acid (2 drops) was added followed by 4-Å molecular sieves,
and the mixture heated under reflux for 1 h. After cooling to room
temperature, the reaction mixture was diluted with dichlorometh-
ane (50 mL), washed with water (50 mL) and brine (50 mL). The
resulting solution was dried with MgSO4 and the solvent removed
in vacuo to give an orange solid which was purified by kugelrohr
distillation (180 °C, 0.5 Torr) to give the title compound (1.03 g,
75%) as a colourless solid. Data were as previously reported.
solid, m.p. 51–54 °C (ref.[15] m.p. 53–55 °C). IR (CHCl ): ν = 2924,
˜
3
1651, 1450 cm–1. H NMR (400 MHz, CDCl3, 25 °C): δ = 7.28 (d,
1
3JH,H = 7.6, 2 H, aromatic CH), 7.21–7.08 (m, 4 H, aromatic CH
3
and one of alkene CH), 6.43 (d, JH,H = 15.1, 1 H, alkene CH),
3
3
4.13 (t, JH,H = 9.3, 2 H, CH2N), 3.79 (t, JH,H = 9.3, 2 H, CH2O)
ppm. 13C NMR (100 MHz. CDCl3, 25 °C): δ = 166.2 (oxazoline
C), 140.2 (alkene CH), 135.6 (aromatic C), 129.8, 129.2 and 127.9
(all aromatic CH), 115.5 (alkene CH), 67.6 (CH2O), 55.3 (CH2N)
ppm. MS (APCI): m/z (%) = 174 (100) [M + H]+.
(5RS,7RS,E)-5,7-Diphenyl-6-(4,5-dihydro-1,3-thiazol-2-ylidene)-
2,3,6,7-tetrahydro[1,3]thiazolo[3,2-a]pyridine (13): (5RS,7RS,E)-6-
(4,5-Dihydro-3-trifluoroacetyl-1,3-thiazol-2-ylidene)-2,3,5,7-tetra-
4,5-Dihydro-2-[(E)-3,3-dimethyl-1-butenyl]-1,3-thiazole (10): 2-
Methylthiazoline (2.5 g, 25 mmol) was dissolved in THF (15 mL)
and cooled to–78 °C. n-Butyllithium (7.8 mL of a 2.5  solution in
hexanes, 26 mmol) was added dropwise and the resulting orange
anion was stirred for 30 min. 2,2-(Dimethyl)propionaldehyde
(1.86 g, 21 mmol) was then added dropwise and the clear orange
solution resulting was stirred for 1 h at–78 °C then warmed to
25 °C over 2 h. Water (20 mL) was then added, and the organic
layer separated, the aqueous layers was extracted with diethyl ether
(3×20 mL). The combined organic layers were concentrated in
vacuo. The resulting red solid was dissolved in toluene (40 mL),
trifluoroacetic acid (30 mg) was added and the mixture heated un-
der reflux, under Dean–Stark conditions for 1 h. The brown solu-
tion remaining was washed with saturated sodium hydroxide solu-
tion (30 mL), dried with MgSO4, and the solvent removed in vacuo
to yield the essentially pure 10 (2.1 g, 50%) as a yellow oil. IR
hydro-5,7-diphenyl[1,3]thiazolo[3,2-a]pyridine
(8a)
(200 mg,
0.4 mmol) was dissolved in ethanol (10 mL). Sodium borohydride
(30 mg, 0.8 mmol) was added and the mixture stirred at 25 °C for
24 h. Water (10 mL) and DCM (10 mL) were added and the phases
separated. The organic phase was dried with MgSO4, and concen-
trated in vacuo to give compound 13 (127 mg, 80%) as a white
solid. 1H NMR (400 MHz, CDCl3, 25 °C): δ = 7.31–7.15 (m, 10
3
H, aromatic CH), 4.29 (d, JH,H = 11.3, 1 H, alkene CH), 4.11 (d,
3
4JH,H = 5.1, 1 H, 5-H), 3.60 (m, 2 H, CH2N), 3.49 (dd, JH,H
=
11.3, 4JH,H = 5.1, 1 H, 7-H), 3.21–2.99 (m, 4 H, CH2N and CH2S),
3
2.80 (t, JH,H = 8.5, 2 H, CH2S) ppm.
Acknowledgments
(CH Cl ): ν = 1647, 1460 cm–1 1H NMR (400 MHz, CDCl3,
.
˜
2
2
3
25 °C): δ = 6.26 (d, JH,H = 15.3, 1 H, one of alkene CH), 6.23 (d,
We are indebted to the EPSRC for a studentship (GR/L53687 to
MSL) and for support of our NMR facilities (GR/S25456) and
AstraZeneca (JLW) and Cardiff University for additional financial
support. We would also like to thank the EPSRC Mass Spectrome-
try Centre, Swansea, for provision of high resolution mass spectro-
metric data.
3JH,H = 15.3, 1 H, one of alkene CH), 4.24 (t, JH,H = 8.1, 2 H,
3
3
CH2N), 3.20 (t, JH,H = 8.1, 2 H, CH2S), 1.03 (s, 9 H 3×CH3)
ppm. 13C NMR (100 MHz. CDCl3, 25 °C): δ = 171.7 (thiazoline
C), 155.2 (alkene CH), 120.6 (alkene CH), 64.5 (CH2N), 34.0
[C(CH3)3], 32.8 (CH2S), 29.0 (3×CH3) ppm. MS (APCI): m/z (%)
= 170 (100) [M + H]+. HRMS (CI) C8H16NS [M + H]+: 170.1003;
found 170.1000.
[1] For a review of this area see M. C. Elliott, M. S. Long, E. Krui-
swijk, Tetrahedron 2001, 57, 6651–6677.
(E)-N-(2-Hydroxyethyl)-3-phenyl-2-propenamide (11): (E)-Cinna-
moyl chloride (1.0 g, 6 mmol) and ethanolamine (336 mg, 6 mmol)
were dissolved in CH2Cl2 (20 mL). A saturated solution of sodium
carbonate (54 mL) was added and the mixture stirred at 25 °C for
16 h. Brine was added and the products extracted into CH2Cl2
(3×30 mL), dried with MgSO4, and concentrated in vacuo to af-
ford a white solid. Recrystallisation from ethyl acetate gave com-
pound 11 (0.91 g, 78%) as a white crystalline solid, m.p. 67–69 °C
(ref.[14] m.p. 67–68 °C). IR: ν (melt) = 3298, 1650, 1597 1556,
[2] a) M. C. Elliott, E. Kruiswijk, Chem. Commun. 1997, 2311–
2312; b) M. C. Elliott, E. Kruiswijk, D. J. Willock, Tetrahedron
Lett. 1998, 39, 8911–8914; c) M. C. Elliott, D. E. Hibbs, D. S.
Hughes, M. B. Hursthouse, E. Kruiswijk, K. M. A. Malik, J.
Chem. Crystallogr. 1998, 28, 663–671; d) M. C. Elliott, A. E.
Monk, E. Kruiswijk, D. E. Hibbs, R. L. Jenkins, D. V. Jones,
Synlett 1999, 1379–1382; e) M. C. Elliott, E. Kruiswijk, J.
Chem. Soc., Perkin Trans. 1 1999, 3157–3166; f) M. C. Elliott,
E. Kruiswijk, D. J. Willock, Tetrahedron 2001, 57, 10139–
10146; g) M. C. Elliott, M. S. Long, Tetrahedron Lett. 2002,
43, 9191–9194; h) M. C. Elliott, M. S. Long, Org. Biomol.
Chem. 2004, 2, 2003–2011.
[3] For a preliminary account of this work see: M. C. Elliott,
N. M. Galea, M. S. Long, D. J. Willock, Tetrahedron Lett.
2001, 42, 4937–4939.
[4] a) M. A. Abdel-Rahman, Synth. Commun. 1993, 23, 1427–
1436; For a related transformation see: b) M. E. A. Zaki, A. A.
1
3
1456 cm–1. H NMR (400 MHz, CDCl3, 25 °C): δ = 7.60 (d, JH,H
= 15.6, 1 H, alkene CH), 7.55 (m, 2 H, aromatic CH), 7.25 (m, 3
3
H, aromatic CH), 6.37 (d, JH,H = 15.6, 1 H, alkene CH), 6.30
(broad s, 1 H, NH), 3.75 (m, 2 H, CH2O), 3.52 (m, 2 H, CH2N),
3.00 (broad s, 1 H, OH) ppm. 13C NMR (100 MHz. CDCl3, 25 °C):
δ = 167.6 (amide C=O), 142.0 (alkene CH), 135.0 (aromatic C),
130.3 (aromatic CH), 129.3 (aromatic CH), 128.3 (aromatic CH),
Eur. J. Org. Chem. 2005, 3791–3800
www.eurjoc.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3799

M. C. Elliott et al.
FULL PAPER
Fadda, K. Samir, F. A. Amer, Chem. Heterocycl. Cmpds. 2003,
39, 1242–1248.
[5] H. L. Wehrmeister, J. Org. Chem. 1962, 27, 4418–4420.
[6] R. Caujolle, G. Baziard-Mouyssett, J. D. Favrot, M. Payard,
R. R. Loiseau, Eur. J. Med. Chem. 1993, 28, 29–35.
[7] G. H. Posner, E. Asirvathan, S. F. Ali, J. Chem. Soc., Chem.
Commun. 1985, 542–543.
[8] For a closely related transformation of a benzothiazole see: R.
Niwa, N. Katagiri, T. Kato, Chem. Pharm. Bull. 1984, 32,
4149–4153.
[9] J. P. P. Stewart, J. Comp. Chem. 1989, 10, 209–220.
[10] J. P. P. Stewart, Journal of Computer-Aided Molecular Design
1990, 4, 1–105.
[13] a) A. I. Meyers, J. L. Durandetta, R. Munavu, J. Org. Chem.
1975, 40, 2025–2029; b) A. I. Meyers, R. Muvanu, J. Durand-
etta, Tetrahedron Lett. 1972, 13, 3929–3932; c) J. P. Guthrie,
K. J. Cooper, J. Cossar, B. A. Dawson, K. F. Taylor, Can. J.
Chem. 1948, 62, 1441–1446.
[14] O. K. Behrens, J. Corse, D. E. Huff, R. G. Jones, Q. F. Soper,
C. W. Whitehead, J. Biol. Chem. 1948, 175, 771–792.
[15] a) P. Elfeldt, Ber. Dtsch. Chem. Ges. 1891, 24, 3218–3228; b)
W. Seeliger, E. Aufderhaar, W. Diepers, R. Feinauer, R. Nehr-
ing, W. Thier, H. Hellman, Angew. Chem. Int. Ed. 1966, 5, 875–
888.
Received: May 19, 2005
Published Online: July 7, 2005
[11] J. Baker, J. Comp. Chem. 1986, 7, 385–395.
[12] For a related reaction see a) K. Takasu, N. Nishida, M. Ihara,
Synthesis 2004, 2222–2225; b) K. Takasu, N. Nishida, M.
Ihara, Tetrahedron Lett. 2003, 44, 7429–7432.
3800
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2005, 3791–3800