Synthesis of chiral iminoalkyl functionalised N-heterocyclic carbenes and their use in asymmetric catalysis

Article abstract of DOI:10.1016/j.tetlet.2007.10.049

A library of new chiral iminoalkyl imidazolium salts has been synthesised from amino acids using a modular design approach. Deprotonation with silver oxide yields silver carbene transfer reagents, which can be used as ligand sources in asymmetric catalysis. Preliminary testing has shown that the ligands induce enantioselectivity in the palladium-catalysed allylic alkylation of 1,3-diphenylprop-3-enyl acetate with dimethyl malonate.

Full text of DOI:10.1016/j.tetlet.2007.10.049

Available online at www.sciencedirect.com
Tetrahedron Letters 48 (2007) 8914–8917
Synthesis of chiral iminoalkyl functionalised N-heterocyclic
carbenes and their use in asymmetric catalysis
Mahboub Merzouk, Theo Moore and Neil A. Williams*
School of Pharmacy and Chemistry, Kingston University, Penrhyn Road, Kingston upon Thames, Surrey KT1 2EE, UK
Received 15 February 2007; revised 28 September 2007; accepted 10 October 2007
Available online 14 October 2007
Abstract—A library of new chiral iminoalkyl imidazolium salts has been synthesised from amino acids using a modular design
approach. Deprotonation with silver oxide yields silver carbene transfer reagents, which can be used as ligand sources in asymmetric
catalysis. Preliminary testing has shown that the ligands induce enantioselectivity in the palladium-catalysed allylic alkylation of 1,3-
diphenylprop-3-enyl acetate with dimethyl malonate.
Ó 2007 Elsevier Ltd. All rights reserved.
The design of new classes of chiral ligands is of great
importance to extend the substrate scope, activity and
enantioselectivity of asymmetric catalysis. To date, most
research has focused on phosphine and amine ligands
due to the large variation of electronic and steric prop-
erties that is possible and their relative ease of synthesis.1
Stable N-heterocyclic carbenes (NHCs), first isolated by
Arduengo,2 are recognised as being valuable alternatives
to phosphines displaying stronger r-donor and poorer
p-acceptor properties.3 N-Heterocyclic carbenes have
been shown to be superior to phosphine ligands in a
number of metal-catalysed reactions, including palla-
dium-catalysed Suzuki–Miyaura cross coupling of steri-
cally hindered aryl chlorides4 and ruthenium-catalysed
ring-opening and closing metathesis.5 Progress in the
synthesis and application of chiral N-heterocyclic car-
bene catalysts was initially slow; most of the early exam-
ples were monodentate with chiral N-substituents,
which were presumably too flexible to yield a catalyst
with a fixed conformation conducive to enantioselectiv-
ity.6 More recently, excellent enantiomeric excesses have
been reported for hydrogenation7 and alkylation reac-
tions8 using chiral bidentate ligands containing NHCs.
In each case the chirality is located in the backbone of
the ligand and the constrained geometry of the ligand
is critical to the excellent enantioselectivity. The work
described here highlights a novel route to chiral NHC
ligands using amino acids and a modular design
approach.
Chiral mixed donor imine phosphine ligands display
high ees for a wide range of asymmetric catalysed reac-
tions, such as alkene hydrogenation,9 allylic alkylation10
and conjugate addition.11 Our aim was to replace the
phosphine donor group with a NHC donor group in
such ligand structures. We were particularly interested
in allylic alkylation where the much stronger r-donor
ability of the carbene compared to the imine would
allow significant electronic differentiation of the terminal
carbons of the meso allyl intermediate. We anticipated
that the chiral backbone of our bidentate ligands,
derived from a chiral amino acid, would create a suitable
fixed-geometry chiral pocket to induce enantioselection.
Without detailed mechanistic information for our
ligands it would be difficult to predict which combination
of alkyl and aryl substituents would give the optimum
performance. Therefore, we adopted
a
modular
approach to the design of the ligands that would allow
the synthesis of a large library of ligands from small
libraries of amino acids, carbonyl compounds and
N-substituted imidazoles Figure 1. The use of cheap
amino acids, available from the chiral pool, was an
attractive feature of the design.
One of the attractions of NHCs is that they can be read-
ily prepared by the deprotonation of imidazolium salts,
which are air stable, unlike trialkylphosphines. The imi-
dazolium salts were synthesised by alkylation of N-
substituted imidazoles with chiral iminoalkyl halides.
*
Corresponding author. Tel.: +44 (0)20 8547 2000; fax: +44 (0)20
8547 7562; e-mail: n.a.williams@kingston.ac.uk
0040-4039/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2007.10.049

M. Merzouk et al. / Tetrahedron Letters 48 (2007) 8914–8917
Table 1. Iminoalkyl bromides synthesised
8915
R1
1
Yield of 3
Yield of 4
Yield of 5
CO2H
R3
R4
1a R1 = Bn
1b R1 = i-Pr
1c R1 = i-Bu
55
90
85
90
88
87
H2N
O
R2
N
N
70
R1
N
N
R3
N
R2
Me
Me
Me
(Boc)2O
95-85%
R4
i) MeSO2Cl
H2N
OH
BocHN
Br
BocHN
OH
ii) LiBr
Figure 1. Modular design of ligands.
6
7
2d
Me
Me
H3N
Ph2C=NH
90%
HCl / dioxane
95%
The chiral iminoalkyl halides were prepared from com-
mercially available amino acids or amino alcohols.
Cl
N
Br
Br
Ph
Ph
Amino acids 1 were reduced with sodium borohydride in
the presence of iodine to give amino alcohols 2. The
aminoalkyl bromide hydrobromide salts 3 were pre-
pared in high yields by stirring with thionyl bromide
and DMF in cyclohexane for 16 h, according to the
method of Jung (Scheme 1, Table 1).12 The amino group
enhances the rate of bromination and the reaction pro-
ceeds with retention of configuration. Diphenylmethy-
lideneamine derivatives 4 were conveniently prepared
by reaction of 3 with benzophenone imine. Benzylidene-
amine alkyl bromides 5 were prepared by condensation
of 3 with benzaldehyde in the presence of molecular
sieves.13
3d
4d
Scheme 2. Preparation of an iminoalkyl halide derived from alaninol.
8 to give the imidazolium salt derivatives 9 that are con-
venient precursors to imidazolylidene ligands (Table 2).
The salts were most conveniently obtained by heating
the reactants at 85 °C in the absence of a solvent. After
12 h the reaction mixture was allowed to cool and the
crude product formed was triturated and washed with
dry diethyl ether, then dried under vacuum. The salts
were purified by recrystallisation from dichloromethane
and diethyl ether. Lower yields were obtained with N-
aryl imidazoles (25–46%) than with N-alkyl imidazoles
(44–79%). It is thought that sterically hindered alkyl ha-
lides can undergo elimination in competition with nucle-
ophilic substitution when reacted with the bulkier N-aryl
imidazoles. In addition, the aryl imidazoles produced
more viscous melts during the reaction which impeded
The aminoalkyl bromide derived from alaninol could
not be made by bromination using thionyl bromide.
An alternative method involving protection was used.
Alaninol was treated with di-tert-butyl dicarbonate in
CH2Cl2 at 0 °C followed by warming to room tempera-
ture to give the protected product in near quantitative
yield.14 The protected alaninol was mesylated in 80–
95% yield by methanesulfonyl chloride and converted
into bromide 7 by stirring with anhydrous lithium bro-
mide in dry acetone.15 Derivative 7 was deprotected by
treatment with HCl in dioxane to give product 3d
(Scheme 2).
1
stirring. The H NMR spectra display a signal between
d 10 and 11 characteristic of the C1 proton of the
imidazolium cation. The diphenylmethylideneamine
imidazolium derivatives are air stable white solids and
were not hygroscopic. The benzylideneamine imidazol-
ium derivatives were however moisture sensitive and
prone to hydrolysis. The yields of imidazolium salts
not used extensively in catalyst testing, such as 9d and
9g, were not optimised. Douthwaite and co-workers
reported that dimethylmethylideneamino functionalised
NHCs gave the best results for asymmetric allylic
Imidazolium salts have proved very convenient precur-
sors to imidazol-2-ylidene ligands. The iminoalkyl
halides 4 or 5 were coupled to N-substituted imidazoles
R1
R1
R1
SOBr2, DMF
NaBH4, I2
70-90%
CO2H
H3N
Br
Br
H2N
H2N
OH
1
2
3
Ph2C=NH
88-92%
PhCHO
70%
R1
R1
N
Br
N
Br
Ph
Ph
Ph
5
4
Scheme 1. Preparation of iminoalkyl halides.

8916
M. Merzouk et al. / Tetrahedron Letters 48 (2007) 8914–8917
Table 2. Synthesis of iminoalkyl imidazolium salts
use of silver oxide is a convenient method for the gener-
ation of monodentate imino functionalised imidazol-
2-ylidene complexes. Therefore, Ag2O was used to
R1
R1
R3
N
˚
Δ
deprotonate imidazolium salts 9 in the presence of 4 A
N
Br
N
R2
Ph
N
N
+
N
R2
molecular sieves. After filtering the solution through
Celite and removing the solvent, the silver carbene com-
plex was added to a solution of [Pd(g3-C3H5)Cl]2 in
THF and stirred for 30 min before the addition of 1,3-
diphenylpropenyl acetate (1 equiv), dimethyl malonate
(3 equiv) and NaH (2.9 equiv). Conversions were mea-
sured by GC and ees were determined by 1H NMR anal-
ysis using Eu(hfc) as a chiral shift reagent.
Ph
R3
Br
9
4 or 5
8
R1
Bn
i-Pr
i-Bu
Me
i-Bu
i-Bu
i-Bu
i-Bu
R2
R3
Yield %
Bn
Bn
Bn
Bn
Me
Mesityl
Ph
Bn
Ph
Ph
Ph
Ph
Ph
Ph
Ph
H
9a 44
9b 57
9c 79
9d 66
9e 63
9f 46
9g 25
9h 70
Table 3 summarises conversions and enantiomeric
excesses for a screen of some representative ligands from
our library in the palladium catalysed asymmetric allylic
alkylation of 1,3-diphenylprop-3-enyl acetate with
dimethyl malonate. In the work of Douthwaite and
co-workers16 the catalyst mixture was filtered to remove
any precipitated silver bromide. In our work, we found
that filtration had little effect on the yield or ee of the
reaction. Entries 1–5 indicate that the alkyl substituent
on the ligand backbone has only a minor effect on the
ee of the reaction. The data reveal that changing the
imidazol-2-ylidene substituent has a more dramatic
effect. The N-methyl imidazol-2-ylidene derivative (entry
6) has a similar ee to entries 1–5 but the activity is
severely diminished compared to the N-benzyl imidazol-
ylidene derivatives. The N-phenyl derivative (entry 8)
gave modest results. In contrast, entry 7, involving the
bulkier N-mesityl derivative displayed a greater activity
and also gave the best ee (53%). Greater steric bulk on
the imidazol-2-ylidene N-substituent is effective in
increasing activity and selectivity. In the work of Dou-
thwaite, reducing the steric bulk around the imine nitro-
gen, by employing imines derived from acetone,
increased the enantioselectivity.16,18 However, dimethyl-
methylideneamine derivatives were not accessible using
our synthetic method. Here the less sterically demanding
benzylideneamine derivative was tested (entry 9), how-
ever, a lower ee was observed. This may well be due to
alkylation;16 however, attempts to make dimethylme-
thylideneamine functionalised imidazolium salts were
unsuccessful.
Mori and co-workers reported the first example of palla-
dium-catalysed allylic alkylation using N-heterocyclic
carbene ligands.17 Around the same time, Douthwaite
reported the first asymmetric allylic alkylation using
NHC–imine ligands derived from trans-1,2-diamino-
cyclohexane.16 These ligands gave ees of up to 92% and
represent one of the most successful examples of asym-
metric catalysis using chiral imidazol-2-ylidenes to date.
Related chiral NHC–phosphine ligands gave poorer
results.18 Hoveyda has reported copper catalysed enantio-
selective allylic alkylation with bidentate imidazolidin-2-
ylidene ligands.19 Wang and Lin first demonstrated that
silver carbene complexes can be obtained easily by treat-
ing azolium salts with Ag2O and that they are efficient
ligand transfer agents for generating palladium carbene
species.20 This mild approach to the formation of imida-
zol-2-ylidenes obviates the need for strong bases.
Danopoulos21 and Douthwaite22 have shown that the
Table 3. Asymmetric allylic alkylations
i) 9, Ag2O
OAc
Ph
MeO2C
Ph
CO2Me
Ph
+
H2C(CO2Me)2
Ph
ii) [Pd(η3—C3H5)Cl]2 ,NaH
Entry
Imidazolium salt
R2
Time (h)
Conversiona (%)
eeb (%)
R1
R3
1 9a
2 9b
3 9c
4 9c
5 9d
6 9e
7 9f
8 9g
9 9h
Bn
Bn
Bn
Bn
Bn
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
H
18
18
18
87
18
18
18
18
18
52
83
60
99
80
26
99
99
99
35
31
30
40
33
38
53
7
i-Pr
i-Bu
i-Bu
Me
i-Bu
i-Bu
i-Bu
i-Bu
Bn
Me
Mesityl
Ph
Bn
9
[Pd(g3-C3H5)Cl]2 (2.5 mol %), silver carbene complex (6 mol %) generated from 9 and Ag2O were stirred for 30 min in THF and then 1,3-diphe-
nylpropenyl acetate (1 equiv), dimethyl malonate (3 equiv) and NaH (2.9 equiv) were added.
a Measured by GC.
b Measured by 1H NMR using Eu(hfc)3 as a chiral shift reagent.

M. Merzouk et al. / Tetrahedron Letters 48 (2007) 8914–8917
8917
2. Arduengo, A. J., III; Dias, H. V. R.; Harlow, R. L.; Kline,
M. J. Am. Chem. Soc. 1992, 114, 5530–5534.
3. Herrmann, W. A. Angew. Chem., Int. Ed. 2002, 41, 1290–
1309.
4. (a) Navarro, O.; Kelly, R. A.; Nolan, S. P. J. Am. Chem.
Soc. 2003, 125, 16194–16195; (b) Altenhoff, G.; Goddard,
R.; Lehmann, C. W.; Glorius, F. Angew. Chem., Int. Ed.
2003, 42, 3690–3693.
5. Trnka, T. M.; Grubbs, R. H. Acc. Chem. Res. 2001, 34,
18–29.
6. Perry, M. C.; Burgess, K. Tetrahedron: Asymmetry 2003,
14, 951–961.
7. Perry, M. C.; Cui, X.; Powell, M. T.; Hou, D.-R.;
Reibenspies, J. H.; Burgess, K. J. Am. Chem. Soc. 2003,
125, 113–123.
8. (a) Clavier, H.; Coutable, L.; Guillemin, G.; Mauduit, M.
Tetrahedron: Asymmetry 2005, 16, 921–924; (b) Bonnet, L.
G.; Douthwaite, R. E.; Kariuki, B. M. Organometallics
2003, 22, 4187–4188; (c) Douthwaite, R. E. Coord. Chem.
Rev. 2007, 251, 702–717.
9. Pfaltz, A.; Blankenstein, J.; Hilgraf, R.; Hormann, E.;
Mcintyre, S.; Menges, F.; Schonleber, M.; Smidt, S. P.;
Wustenberg, B.; Zimmermann, N. Adv. Synth. Catal.
2003, 345, 33–43.
10. Dawson, G. J.; Frost, C. G.; Williams, J. M. J. Tetra-
hedron Lett. 1993, 34, 3149–3153.
hydrolysis of the benzylideneamine function, the result-
ing aminoalkyl imidazol-2-ylidene is unlikely to chelate
resulting in a loss of enantioselectivity in the reaction.
The ees of our reactions are modest: it is possible that
the imine nitrogen is only weakly co-ordinating and
may in fact be hemi-labile. Evidence to support this
comes from an attempt to synthesise palladium
complexes containing a chelating ligand. Mass spectral
analysis of the reaction product indicates the presence
of a bis(imidazol-2-ylidene)palladium species. Related
bis(iminocarbene) palladium complexes containing two
bulky, C-bonded iminocarbene moieties have been
isolated and characterised by Tilset and co-workers23
This group has also demonstrated the hemi-labile nature
of bulky iminocarbene ligands in DMSO by 1H NMR.24
In conclusion, we have synthesised a small library of
chiral imidazol-2-ylidene ligands by adopting a modular
approach with readily available amino acids as the chiral
building blocks. The ligands have shown enantioselec-
tivity for allylic alkylation, but further modification of
the ligand structure to give smaller imino-substituents
and bulkier N-imidazol-2-ylidene substituents is likely
to be required to enhance the enantioselectivity. Given
the highly successful application of related imino-phos-
phine ligands to a wide range of catalytic asymmetric
syntheses, we aim to test ligands from our library on
other reactions, such as copper-catalysed conjugate
addition and iridium-catalysed hydrogenation of
arylalkenes.
11. Helmchen, G.; Pfaltz, A. Acc. Chem. Res. 2000, 33, 336–
345.
12. Nagle, A. S.; Salvatore, R. N.; Chong, B.-D.; Woon Jung,
K. Tetrahedron Lett. 2000, 41, 3011.
13. De Kimpe, N.; De Smaele, D. Tetrahedron 1995, 51, 5465–
5478.
14. Myllymaki, V. T.; Lindvall, M. K.; Koskinen, A. M. P.
Tetrahedron Lett. 2001, 57, 4629–4635.
15. Higashiura, K.; Morino, H.; Matsuura, H.; Toyomaki, Y.;
Ienaga, K. J. Chem. Soc., Perkin Trans. 1 1989, 1479–
1481.
Acknowledgements
16. Bonnet, L. G.; Douthwaite, R. E.; Hodgson, R.; Hough-
ton, J.; Kariuki, B. M.; Simonovic, S. Dalton Trans. 2004,
3528–3535.
17. Sato, Y.; Yoshino, T.; Mori, M. Org. Lett. 2003, 5, 31–32.
18. Hodgson, R.; Douthwaite, R. E. J. Organomet. Chem.
2005, 690, 5822–5831.
Financial support from EPSRC is gratefully acknowl-
edged (M.M.). The service provided by the EPRSC mass
spectrometry service centre (Swansea) is also gratefully
acknowledged.
19. Larsen, A. O.; Leu, W.; Oberhuber, C.; Campbell, J. E.;
Hoveyda, A. H. J. Am. Chem. Soc. 2004, 126, 11130–
11131.
Supplementary data
20. Wang, H. M. J.; Lin, I. J. B. Organometallics 1998, 17,
972–975.
21. Tulloch, A. A. D.; Danopoulos, A. A.; Winston, S.;
Kleihenz, S.; Eastham, G. J. Chem. Soc., Dalton Trans.
2000, 4499–4506.
Supplementary data including synthetic procedures and
spectroscopic characterisation are available. Supplemen-
tary data associated with this article can be found, in the
online version, at doi:10.1016/j.tetlet.2007.10.049.
22. Bonnet, L. G.; Douthwaite, R. E.; Kariuki, B. M.
Organometallics 2003, 22, 4187–4188.
23. Froseth, M.; Netland, K. J.; Tornroos, K. W.; Dhindsa,
A.; Tilset, M. Dalton Trans. 2005, 1664–1674.
24. Froseth, M.; Dhindsa, A.; Roise, H.; Tilset, M. Dalton
Trans. 2003, 4516–4524.
References and notes
1. Pfaltz, A.; Yamamoto, H.; Jacobsen, E. N. Comprehensive
Asymmetric Catalysis; Springer, 2004.