Cyclophane-type imidazolium salts with planar chirality as a new class of N-heterocyclic carbene precursors

Article abstract of DOI:10.1002/chem.200800942

Cyclophane-type imidazolium salts with planar chirality were synthesized from enantiopure 2-amino alcohols, of which the N(1) and N(3) positions were connected with a bridge. The structural profiles of the imidazolium salts and their derivative N-heterocy

Full text of DOI:10.1002/chem.200800942

FULL PAPER
DOI: 10.1002/chem.200800942
Cyclophane-Type Imidazolium Saltswith Planar Chirality asa New Clasof
N-Heterocyclic Carbene Precursors
Yuki Matsuoka, Yasuhiro Ishida,* Daisuke Sasaki, and Kazuhiko Saigo*[a]
Abstract: Cyclophane-type imidazoli-
um salts with planar chirality were syn-
thesized from enantiopure 2-amino al-
cohols, of which the N(1) and N(3) po-
sitions were connected with a bridge.
The structural profiles of the imidazoli-
um salts and their derivative N-hetero-
cyclic carbenes (NHCs) were investi-
gated by means of several analyses.
The chiral NHCs derived from these
imidazolium salts were found to cata-
lyze the asymmetric cross-annulation
of an a,b-unsaturated aldehyde with a
ketone by means of the “conjugated”
umpolung of the enal to give the target
g-lactone with good to excellent enan-
tioselectivity (up to 94% ee). Based on
the expected structure of the NHCs
and their intermediates, together with
the absolute configuration of the prod-
ucts, a plausible mechanism for the ste-
reocontrol was proposed.
Keywords: carbenes · chirality · cy-
clophanes · heterocycles · umpolung
Introduction
these reactions feature the creation of a stereogenic carbon
atom(s) in products, the stereocontrol of these reactions by
chiral NHCs has been a challenging target over the last
decade.[4] At the earlier stage of the development of NHC
precursors, simple azolium/azolinium salts with chiral N sub-
stituent(s) were studied extensively,[5] and more recently
NHC precursors with topological chirality have appeared,
including polycyclic (A–C),[6] axially chiral (D–F),[7] planarly
chiral (G–I),[8] oligopeptidyl (J),[9a] rotaxane-containing
(K),[9b] and chirality-relayed (L)[9c] azolium/azolinium salts.
In fact, some of these examples were found to provide NHC
organocatalysts for the benzoin condensation and the Stetter
reaction, of which the chirality induction ability was superior
to that of traditional NHCs with simple point chirality.
Through our ongoing study on the development of novel
NHC precursors with topological chirality, we focused on
imidazolium salts (1a–c) with cyclophane-type planar chiral-
ity, of which the N(1) and N(3) positions are bridged with a
chiral linker. When the two nitrogen atoms of an imidazoli-
um ring are linked by a dissymmetric bridge with a proper
length, the steric demand forces the bridge to cover either
of the two faces of the prochiral imidazolium ring, and as a
result, a planar-chiral cyclophane-type molecule will be
readily generated without resorting to complicated/bulky
substituent(s).[10,11] Because planar-chiral cyclophanes with
such a structure are anticipated to easily isomerize by means
of the rope-skipping motion of the bridge going over the
imidazolium ring, a sufficient chiral auxiliary might be re-
quired to distinguish two atrop isomers in stability and/or re-
activity; either of the two isomers would work as a net reac-
Topological chirality, as represented by helical, axial, and
planar chiralities, has been regarded as a key principle in
overcoming the limits of traditional tetrahedral chirality and
expanding the scope of chiral molecular devices.[1] In every
field of chiral technology, topologically chiral molecules
often provide useful chiral materials, such as ligands,[2a,g]
auxiliaries,[2b] selectors,[2d,e] and sensors.[2c,f] Their well-de-
fined, three-dimensionally dissymmetric shape seems to be
advantageous not only for chirality induction/recognition
but also for further development of materials with an opti-
mized structure and elucidation of mechanisms. As repre-
sentative examples, numerous metal-based catalysts with
topologically chiral ligands have been developed to date,[2a]
and recently the utility of topological chirality has been also
proved in the realm of organocatalysis.[3]
Among various classes of organocatalysts, N-heterocyclic
carbenes (NHCs) have attracted exceptional attention.
These offer a unique carbon–carbon bond-forming ability by
means of the umpolung of aldehydes.[4] Because most of
[a] Dr. Y. Matsuoka, Prof. Y. Ishida, D. Sasaki, Prof. K. Saigo
Department of Chemistry and Biotechnology
Graduate School of Engineering
The University of Tokyo
Hongo, Bunkyo-ku, Tokyo 113-8656 (Japan)
Fax : (+81)3-5802-3348
E-mail: ishida@chiral.t.u-tokyo.ac.jp
saigo@chiral.t.u-tokyo.ac.jp
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200800942.
Chem. Eur. J. 2008, 14, 9215 – 9222
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
9215

Results and Discussion
Synthesis of cyclophane-type imidazolium salts: The N(1)–
N(3)-bridged imidazolium salts 1a–c were synthesized as
shown in Scheme 1. According to the procedure we have re-
cently reported,[12] the imidazoles 7a and 7b with a chiral N
substituent were directly synthesized by the cyclocondensa-
tion of the enantiopure 2-amino alcohols 6a and 6b with
glyoxal, formaldehyde, and ammonium acetate. On the
other hand, the analogous benzimidazole 7c was prepared
by the following series of reactions: the nucleophilic substi-
tution of 1-fluoro-2-nitrobenzene with 6a, the hydrogenoly-
sis of the nitro group into an amino group, and the conden-
sation of the resultant 1,2-diaminobenzene derivative with
triethyl orthofomate.[13] The imidazoles 7a–c thus obtained
were then treated successively with sodium hydride and 1,8-
dichlorooctane to afford the 8-chlorooctylated imidazoles
8a–c. Finally, the intramolecular quaternization of the resul-
tant imidazoles 8a–c gave the cyclophane-type imidazolium
salts 1a–c.[10a,b,d] For the last step, a highly elevated tempera-
tive species. In the context of chirality type, 1a–c might be
classified in the same category as the types G–I (planar chir-
ality). However, the carbene moiety of 1a–c is incorporated
in a macrocyclic structure, which seems to be advantageous
in directly controlling spatial environment around the car-
bene reaction center. Despite such a simple and promising
structure, NHCs derived from imidazolium salts like 1a–c
have not yet been explored so far. Here we report the syn-
thesis, characterization, and application of a new type of
NHC/NHC precursor with cyclophane-type planar chirality.
Scheme 1. Synthesis of cyclophane-type imidazolium salts 1a–c. Top:
i) glyoxal, HCHO, NH4OAc, MeOH/H2O, 80 8C, 5 h; ii) NaH, ClACHTREUNG(CH2)8Cl,
DMF, 0 to 558C, 20 h; iii) DMAc, 1508C, 48 h. Bottom: i) 1-fluoro-2-ni-
trobenzene, DMSO, RT, 60 h; ii) Pd/C, H2 (1 atm), MeOH, RT, 40 h;
iii) CH
ACHRTUNEG(OEt)3, p-TsOH·H2O, 1008C, 44 h; iv) NaH, ClCAHTRE(UGN CH2)8Cl, DMF, 0
to 558C, 17 h; v) DMAc, 1508C, 140 h; vi) KPF6, CH3CN, RT, 10 h.
9216
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 9215 – 9222

Chiral Cyclophane Imidazolium Salts
FULL PAPER
ture was required to ensure a reasonable reaction rate, be-
cause the length of the alkyl chain had little margin to con-
nect the two imidazolium nitrogen atoms. Although the in-
termolecular quaternization was a possible undesired side
reaction, only a small amount of a mixture of oligomers was
generated, as long as the initial concentration was at an
order of 10–100 mm. In addition, the extensive degas of
oxygen from the system before heating efficiently prevented
side reactions, most likely caused by the heat-promoted gen-
eration of NHCs and a subsequent reaction with internal
oxygen.
suggests that the activation energy for the rope-skipping
process is rather low. The above-mentioned molecular mod-
eling also indicates that the rope-skipping process easily
takes place at the C(2) side of the imidazolium ring, even
though the bridge barely goes through the C(4)/C(5) side
because of steric hindrance.
Considering the structural similarity between imidazolyli-
denes and their parent imidazoliums, it might be reasonable
to anticipate that the imidazolylidene 1aNHC, derived from
1a, also exists as a mixture of two atrop isomers ((Sp)-and
(Rp)-1aNHC in Scheme 2) that are isomerizable into each
Structural profilesof the cyclophane-type imidazolium salts :
The structural profiles of the cyclophane-type imidazolium
salts 1 were investigated by 1H NMR spectroscopy, X-ray
crystallography, and molecular modeling. The 1H NMR
spectroscopy signals for the methylene protons at the
middle of the N(1)–N(3) bridge in 1a were observed at an
upfield-shifted region (Ddꢀ0.5 ppm, overlapped with the
À
signal of the isopropyl group (CH3)2CH ), which indicates
that the bridge overlays the imidazolium ring to suffer a
shielding effect.[10,14] Given such a “hover-overed” structure,
two atrop isomers (diastereomers) arising from the planar
chirality of the imidazolium ring are possible for 1a (Sp and
Rp). In good agreement with this expectation, a theoretical
calculation at the HF/6-31G level gave two constitutions
with a “hover-overed” structure minimal in potential energy,
both of which correspond to two possible conformers (Fig-
ure 1a). Moreover, an X-ray crystallographic study revealed
that 1a takes one of the two conformations ((Rp)-1) in the
crystalline state (Figure 1b).[14] However, variable-tempera-
ture (VT) 1H NMR spectroscopy measurements clearly
proved that the isomerization by means of the rope-skipping
motion of the N(1)–N(3) bridge easily takes place in a solu-
tion; 1a showed one set of sharp signals at a temperature
range from 25 to À808C, and even at À958C only partial
broadening of signals was observed.[10a,b,14] The phenomenon
Scheme 2. Relative reactivity of two conformers of 1NHC
.
other. However, there seems to be large difference in nucle-
ophilicity between (Sp)-and ( Rp)-1aNHC, because of the dif-
ference in steric congestion around the carbene moieties. In
the case of (Rp)-1aNHC, the isopropyl group on the stereo-
genic carbon in the bridge would overlay the C(2) to reduce
the nucleophilicity of this NHC species. Contrary to this,
(Sp)-1aNHC seems to have enough space around the C(2) to
facilitate nucleophilic attack toward electrophiles such as
AHCTREUNG
ed imidazolium salt 1aMe revealed that the isomerization of
(Sp)-and ( Rp)-1aMe hardly occurs, even when standing at
room temperature for several months; the methyl group is
bulky enough to suppress the rope-skipping of the N(1)–
N(3) bridge across the imidazolium plane. To confirm the
deduction described above, the following model reaction
was conducted. The imidazolium salt 1a was treated with
butyllithium to generate the imidazolylidene 1aNHC, which
was then electrophilically trapped with methyl iodide to give
1aMe [15]
The conformation of the Sp and Rp isomers should
.
be locked after the C(2)-methylation (Scheme 2). As a
result, we could roughly but directly estimate the relative re-
activity of the two imidazolylidene conformers (Sp)-and
Figure 1. Structure of cyclophane-type imidazolium salt 1a. a) Optimized
structure (HF/6-31G level) of the two conformers of 1a with cyclophane-
type planar chirality of Rp (left) and Sp (right). b) X-ray crystal structure
of 1a (bromide salt). The counteranion is omitted for clarity.
(Rp)-1aNHC from the product ratio of (Sp)-and ( Rp)-1aMe
.
We should emphasize here that, because the isomerization is
Chem. Eur. J. 2008, 14, 9215 – 9222
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9217

Y. Ishida, K. Saigo et al.
such a fast process as estimated by variable-temperature
NMR spectroscopy, the product ratio will not be governed
by the population of the two imidazolylidene conformers,
but will reflect their relative reactivity.
Through the deprotonation and the subsequent methyla-
tion at the C(2) position, 1a gave (Sp)-1aMe as the main
product (85% conversion, Sp/Rp =82:18). The absolute con-
figuration of the major isomer (determined by an X-ray
crystallographic study) is in good agreement with our ex-
pectation that (Sp)-1aNHC would be more reactive than (Rp)-
1aNHC [10d]
In the cases of the other imidazolium salts 1b and
.
1
1c, essentially the same results (by means of H NMR spec-
troscopy and molecular modeling) were obtained as those of
1a, which strongly implies a general tendency in the reactiv-
ity order (Sp isomer>Rp isomer), regardless of the side-
chain residue on the stereogenic carbon and of the substitu-
ents at the C(4) and C(5) positions. Thus, (Sp)-1 was found
to preferentially react even with methyl iodide relatively
little in comparison with carbonyl compounds widely used
as substrates in NHC-catalyzed reactions.
Scheme 3. Cross-annulation of the enal 11 and the ketone 12 by means
of the conjugated umpolung catalyzed by NHC, and a possible side
AHCTRErGNU eaction.
Cross-annulation of the enal 11 and the ketone 12 catalyzed
by the NHC derived from imidazolium salt 1b: With these
cyclophane-type imidazolium salts 1a–c in hand, we next ap-
plied them as the precursors of chiral NHC organocatalysts.
For the first attempt the NHC-catalyzed annulation of cin-
namaldehyde (11) and 2,2,2-trifluoroacetophenone (12) was
selected, because only those from imidazoliums are known
to catalyze this reaction efficiently among four types of
NHCs based on triazolium, thiazolium, imidazolium, and
imidazoliunium salts.[15b] The annulation involves multiple
steps, such as the generation of an acyl anion equivalent
through the NHC-catalyzed umpolung of the enal unit, the
conversion of the acyl anion equivalent into a homoenolate
equivalent through the conjugation, the electrophilic trap-
ping of this homoenolate by the carbonyl compound, and
the intramolecular nucleophilic attack of the resultant alk-
oxide on an activated ester moiety (Scheme 3).[16] Despite
the synthetic importance of this annulation, attempts to
extend the reaction to an enantioselective version are at a
premature stage.[6f,g,17,18] Thus, this reaction seemed to be a
suitable entry to demonstrate the utility of our imidazolium
salts 1 as precursors of chiral NHC organocatalysts.
11 and 12 were isolated from the reaction mixture
(Scheme 3), whereas other possible byproducts generated
from an acyl anion equivalent such as acyloins and the Stet-
ter products were not detected at all. The homoenolate gen-
erated showed undesired chemoselectivity (i.e., the cross-an-
nulation/self-annulation ratio) to lower the yield of the
target lactone 13; this problematic side reaction has been
commonly observed in the cases of other imidazolidene cat-
alysts.[15a] However, upon raising the reaction temperature,
the cross-/self-annulation ratio was improved to some extent
(from 26:74 to 46:54, Table 1, entries 2–4).
The stereochemical outcomes of the cross-annulation ex-
hibited an unexpected dependence on the reaction condi-
tions. The diastereomeric ratio (trans-13/cis-13) showed little
dependence on the reaction temperature, solvent, and base.
On the other hand, the enantiomeric excesses of trans-and
cis-13 significantly increased as the reaction temperature
became higher, which is a contrast to the commonly ob-
served tendency of asymmetric reactions.
Overall, the conditions given in entry 4 in Table 1 were
found to be optimal from the viewpoints of yield and selec-
The effect of reaction condi-
tions on the efficiency and se-
lectivity of this reaction was
Table 1. Cross-annulation of the enal 11 and the ketone 12 catalyzed by 1bNHC
.
Entry[a]
Solvent
Base
T [8C]
Conv.[b] [%]
Ratio[b] 13/14
Yield[c] (ee[d]) [%]
thoroughly
investigated
by
trans-13
cis-13
using 1b (Table 1). The NHC
1bNHC was generated in situ by
treatment of 1b (0.2 equiv)
with a base (0.2 equiv), which
was directly used for the annu-
lation of 11 (1.0 equiv) and 12
(4.0 equiv). In all entries, the
target cross-annulated lactone
13, the self-annulated lactone
1
2
3
4
5
6
THF
THF
THF
THF
CH2Cl2
THF
KN
KN
KN
KN
KN
N
À78
À20
0
25
25
<1
59
62
>99
69
60
–
–
–
26:74
31:69
46:54
41:59
43:57
6 (19)
8 (32)
21 (43)
14 (25)
13 (20)
3 (29)
3 (51)
9 (69)
4 (44)
3 (45)
LiN
G
25
[a] Conditions: NHC generation: 1b (0.2 equiv), a base (0.2 equiv), a solvent, À788C for 1 h. Cross-annulation:
NHC, 11 (1.0 equiv), 12 (4.0 equiv), a solvent, from À788C to a temperature in a period of 30 min and then at
that temperature for 16 h. [b] Conversion determined by 1H NMR spectroscopy. [c] Isolated yield. [d] Deter-
mined by chiral HPLC. In all entries, (4R,5R)-and (4 R,5S)-13 were obtained as the major enantiomers of
14,[19] and the starting materials trans-and cis-13, respectively.
9218
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 9215 – 9222

Chiral Cyclophane Imidazolium Salts
FULL PAPER
Figure 2. Possible reactive intermediates in the present reaction.
tivity. By taking account of the unexpected changes in enan-
tioselectivity depending on the reaction temperature and on
the metal cation, the reactive intermediates might include a
metal homoenolate complex (Figure 2, III and/or IV), in ad-
dition to the Breslow intermediate and its tautomer (I and
II).
ed lactone 13 because of unsatisfactory chemoselectivity re-
Cross-annulation of the enal 11 and the ketone 12 catalyzed
quired to generate a considerable amount of the self-annula-
by the NHCsderived from variousimidazolium aslts1–5
:
tion product 14 (Scheme 3).[19] Among 1a–cNHC 1cNHC
,
To address the characteristic profile of the cyclophane-type
planarly chiral NHCs 1NHC, the same reaction was conducted
by using conventional chiral NHCs based on the imidazoli-
um salts 2–5; C2-symmetric 2 is one of the most widely used
chiral imidazolium precatalysts,[5c,d] whereas the tricyclic and
bicyclic imidazolium salts (3 and 4) are limited examples of
precedent chiral imidazolium precatalysts with a fused ring
structure. In addition to them, an open-chained analogue of
1b (5) was also employed as a precatalyst, which would clar-
ify the effect of the cyclophane structure in 1 (Table 2).
showed better cross-annulation selectivity (entries 1 and 2
versus entry 3), which might be elucidated as follows. Be-
cause of the electron-withdrawing and/or charge-delocaliz-
ing effect of the benzo moiety, 1cNHC generates a homoeno-
late equivalent with lower nucleophilicity than do 1aNHC and
1bNHC, and the resultant homoenolate species prefers the
electronically favorable electrophile 12 rather than the steri-
cally favorable electrophile 11 (Scheme 3).
For the diastereoselectivity of this reaction catalyzed by
NHCs, it is generally known that the isomer possessing two
aromatic groups at the same side of the g-lactone ring
(trans-13) is more predominantly formed than is the other
isomer (cis-13). In good agreement with this propensity,
1bNHC afforded trans-13 as the major product in a trans/cis
ratio of around 70:30 (entry 2, Table 1). Worth noting is the
peculiar diastereoselectivity of 1aNHC and 1cNHC, the cyclo-
phane-type NHCs bearing an isopropyl group on the stereo-
genic center; the ratio of the cis isomer increased, and the
diastereomer ratio reached around 50:50 (entries 1 and 3,
Table 2). Because such a peculiar diastereoselectivity was
not observed for 1bNHC, the bulky isopropyl group on the
stereogenic carbon of 1aNHC and 1cNHC was likely to contrib-
ute to suppress the formation of trans-13.
Table 2. Cross-annulation of the enal 11 and the ketone 12 catalyzed by
NHCs from the imidazolium salts 1–5.
Entry[a]
Precat.
Conv.[b] [%]
Ratio[b] 13/14
Yield[c] (ee[d]) [%]
trans-13
cis-13
1
2
3
4
5
6
7
1a
1b
1c
2
3
4
93
>99
79
>99
94
45:55
46:54
59:41
45:55
>99:1
44:56
36:64
10 (57)
21 (43)
16 (74)
22 (23)
48 (28)
23 (66)
16 (18)
10 (89)
9 (69)
17 (94)
7 (45)
20 (14)
5 (58)
6 (31)
>99
>99
5
[a] Conditions: NHC generation: imidazolium precatalyst (Precat.)
(0.2 equiv), KN
(SiMe3)2 (0.2 equiv), THF, À788C for 1 h. Cross-annula-
ACHTREUNG
tion: NHC, 11 (1.0 equiv), 12 (4.0 equiv), THF, from À788C to 258C in a
period of 30 min and then at 258C for 16 h. [b] Determined by 1H NMR
spectroscopy. [c] Isolated yield. [d] Determined by chiral HPLC. In all
entries except for entry 6, (4R,5R)-and (4 R,5S)-13 were obtained as the
major enantiomer of trans-and cis-13, respectively. In entry 6, the abso-
lute configuration of the major enantiomers were inverted for both of
trans-and cis-13.
Quite interestingly, the NHCs derived from the N(1)–
N(3)-bridged imidazoliums 1a–c showed exceptionally high
enantioselectivity (entries 1–3, Table 2) compared with con-
ventional chiral NHCs. Especially, the enantiomeric excesses
of trans-and cis-13 obtained by the reaction catalyzed by
1cNHC were no less than 74 and 94%, respectively (entry 3).
To the best of our knowledge, these are the highest enantio-
selectivity (ee) values for this reaction up to now.[6f,g,17,18]
A
The conversion of the enal 11 revealed that the NHCs
1a–cNHC exhibited a sufficient catalytic activity comparable
to that of the conventional NHCs 2–4NHC, although the reac-
tivity of the NHC 1cNHC was lower than that of the others to
some extent (Table 2, entries 1–3 versus entries 4–6). De-
spite the adequate amount of catalytic activity thus ob-
served, the NHCs 1a–cNHC as well as the NHCs 2NHC and
4NHC could not achieve sufficient yield of the cross-annulat-
control experiment using the analogous ring-opened 5NHC
strongly suggests that such high enantioselectivities are as-
cribable to the conformationally constrained cyclophane
structure (entry 2 versus entry 7).
Elucidation of the mechanism of the present stereocon-
trolled lactone formation: Direct g-lactone formation by
means of the conjugated umpolung of enals is expected to
Chem. Eur. J. 2008, 14, 9215 – 9222
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9219

Y. Ishida, K. Saigo et al.
become an important benchmark reaction to evaluate the
potential of chiral NHCs, especially those derived from
chiral imidazolium salts. For the detailed elucidation of the
mechanism for chirality induction, the determination of the
absolute configuration of the annulation products would be
highly necessary. Up to now, however, studies on the abso-
lute configuration of 4,5-diaryl-g-lactones have been limited
to a few reports dealing with the simplest type of 4,5-di-
phenyl-g-lactone.[20] Moreover, all of these conventional as-
signments were deduced from the pattern of circular dichro-
ism spectra. Therefore, we tried to determine the absolute
configuration of the enantiomers of trans-and cis-13 by an
unequivocal method based on the chromatographic separa-
tion of the enantiomers, their derivation by the aminolysis
with (R)-1-phenylethylamine, and the X-ray crystallography
of the resultant g-hydroxypentanamides. As a result, the
major enantiomers of trans-and cis-13 obtained by the
NHC catalysts 1a–cNHC were found to be 4R,5R and 4R,5S,
respectively (Figure 3).[14]
Based on the configurations as determined above, a plau-
sible reaction path elucidating the enantioselection by 1NHC
can be proposed; it involves the Si-selective attack of the E
homoenolate (Scheme 4, I and II). Although 1NHC is likely
to exist as a mixture of the Rp and Sp isomers easily isomer-
izable into each other, the Sp isomer was proved to be the
net reactive species of 1NHC, as described above (Scheme 2).
Apparently, the Z homoenolate might seem to be more fa-
vorable than the E isomer as an intermediate of the first
step, considering steric congestion around the N(1) of the
NHC moiety. However, we believe that the E homoenolate
is more stable owing to the following “strain-relayed” mech-
anism; our hypothesis is strongly supported by the X-ray
crystal structure of a homoenolate-intermediate analogue
(the PF6 salt of (Sp)-1aMe), of which the C(2) substituent is
À
replaced with a methyl group (Figure 4).[10d] Taking account
of steric repulsion, the bulky side-chain residue on the ste-
reogenic carbon (iPr for 1a,cNHC and Ph for 1bNHC) should
orient perpendicularly to the imidazolium ring (Figure 4,
arrow i). In addition, the ether unit in the N(1)–N(3) bridge
might restrict the conformation around the N(1) end of the
bridge, owing to the partial sp2 character of the oxygen
atom. These conformational demands would relay to the
neighboring atoms in the bridge (arrows ii and iii). To cancel
the strain thus accumulated, the other end of the bridge ad-
jacent to the N(3) is pushed out to the C(2) side, which in-
duces the N(3) end of the bridge to occupy the space
around the C(2) (arrow iv). Such a dissymmetrically distort-
ed structure of the N(1)–N(3) bridge seems to be suitable
Figure 3. Circular dichroism spectra of the enantiomers of a) trans-and
b) cis-13 in acetonitrile at RT and X-ray crystal structures of the g-hy-
droxypentanamides derived from c) trans-(4R,5R)-and d) cis-(4S,5R)-13
by treatment with (R)-1-phenylethylamine.
Scheme 4. A plausible reaction path for the stereocontrolled cross-annulation of the enal 11 and the ketone 12 catalyzed by NHCs from 1a,c.
9220
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 9215 – 9222

Chiral Cyclophane Imidazolium Salts
FULL PAPER
high enantioselectivity in the cross-annulation reaction of an
enal and a ketone (up to 94% ee), owing to their character-
istic cyclophane structure. As far as we know, this is the first
attempt to apply planar chirality to the development of
NHC-based chiral organocatalysts. Although the yield of the
target lactone was unsatisfactory at present, the ee values
observed here were at the highest level for this reaction.
Considering the easy access, simple and well-defined struc-
ture, and the high chirality induction ability, the cyclophane-
type planarly chiral imidazolium salts developed here would
provide us a new structural motif in the field of NHC
chemistry.
À
Figure 4. X-ray crystal structure of (Sp)-1aMe (PF6 salt): a) top view,
b) side view, and c) front view. Hydrogen atoms, counteranions, and the
other symmetrically independent imidazolium cations in a unit cell are
omitted for clarity.[10d]
Experimental Section
Chemicals: Chemicals were purchased and used as delivered unless oth-
erwise indicated. Tetrahydrofuran (THF) was distilled from sodium wire
and benzophenone before use. Dichloromethane (CH2Cl2), acetonitrile
(CH3CN), and dimethylsulfoxide (DMSO) were distilled from CaH2 and
stored over activated molecular sieves. N,N’-Dimethylformamide (DMF)
was dried over P2O10, distilled from CaH2, and stored over activated mo-
lecular sieves. N,N’-Dimethylacetamide (DMAc) was distilled just before
use. Cinnamaldehyde (11) and 2,2,2-trifluoroacetophenone (12) were dis-
tilled just before use. A solution of potassium hexamethyldisilazane in
toluene was purchased and used as received. Imidazolium salts 2–4 were
prepared according to the literature cited in the Supporting Information.
For the synthesis of the cyclophane-type imidazolium salts 1a–c and the
open-chained analogue 5, see the Supporting Information.
for the preferential formation of the E homoenolate. Ac-
cording to the molecular modeling of the E homoenolate,
the Re face is significantly shielded by the N(1)–N(3) bridge
of the imidazolylidene moiety, whereas there is essentially
no steric hindrance on the Si face. Consequently, the elec-
trophilic attack of 12 occurs mainly on the Si face of the ho-
moenolate to generate the g-lactone with R configuration at
the C(4) position. The expected stereochemical outcome is
in good agreement with that observed for both of the trans
and cis isomers (Scheme 4).
On the other hand, the relative stereochemistry at the
C(4) and C(5) positions (i.e., the diastereocontrol) of 13
might be mainly determined by the inherent nature of sub-
strates 11 and 12. Considering the favorable arrangement of
the dipoles of the homoenolate equivalent and the ketone
12, as well as p–p interaction between the electron-rich and
electron-deficient phenyl groups, the transition states of I in
Scheme 4 are expected to be the most favorable. The subse-
quent electrophilic trapping of the homoenolate and lactone
formation would give the trans isomer, which is in good
agreement with the general tendency of this reaction includ-
ing our observations. The exceptional bias in the diastereo-
selectivity observed for the reactions catalyzed by 1aNHC and
by 1cNHC might be explained in terms of steric repulsion be-
tween the bulky isopropyl group in the imidazolylidene and
the phenyl group in 12.
C(2)-methylation of the imidazolium salt 1a: A solution of butyllithium
in hexane (1.56m, 0.24 mL, 0.36 mmol) was added dropwise to a solution
(3 mL) of 1a (90 mg, 0.30 mmol) in CH2Cl2 at À788C, and the mixture
was stirred at À788C for 1 h. After adding a solution (1.5 mL) of methyl
iodide (70.2 mg, 0.49 mmol) in CH2Cl2, the mixture was stirred at À788C
for 30 min and then at RT for 3 h. The insoluble materials were filtered
off (ADVANTEC filter paper 5A), and the filtrate was concentrated
under reduced pressure to give a crude mixture of (Sp)-and ( Rp)-1aMe
,
and unreacted 1a (85% conversion, (S)-1aMe/(Rp)-1aMe =82:18, deter-
1
mined by H NMR spectroscopy).
Typical procedure for the cross-annulation of the aldehyde 11 and the
ketone 12: A solution of potassium hexamethyldisilazane in toluene
(0.50m, 200 mL, 0.10 mmol) was added dropwise to a suspension of 1a
(30.1 mg, 0.10 mmol) in THF (3 mL) at À788C under an argon atmos-
phere, and the mixture was stirred at that temperature for 1 h. Then,
2,2,2-trifluoroacetophenone (11, 280 mL, 2.0 mmol) and cinnamaldehyde
(12, 65 mL, 0.50 mmol) were successively added to the mixture at À788C.
The mixture was stirred at that temperature for 30 min, and then allowed
to warm up to RT. After being stirred at RT for 16 h, the mixture was
treated with methanol (0.5 mL) and concentrated under reduced pres-
sure. The resultant residue was subjected to silica gel column chromatog-
raphy (eluent: hexane/CH2Cl2 =2:1, v/v) to give a mixture of trans-and
cis-13 as a pale yellow oil (29.0 mg, 0.09 mmol, 19%). The ratio of trans-
and cis-13 was 50:50 (estimated by a 1H NMR spectroscopy measure-
ment), and the enantiomeric excesses of trans-and cis-13 were 57 and
89%, respectively (estimated by a chiral HPLC measurement). Chiral
HPLC conditions: column, Daicel CHIRALCEL AS-H (4.6”153 mm);
eluent, hexane/2-propanol (90:10, v/v); flow rate, 1.0 mLminÀ1; detection,
UV absorption at l=210 nm; retention time, 8.41 and 13.16 min for
trans-13 and 14.06 and 23.44 min for the cis-13.
Conclusion
Imidazolium salts with cyclophane-type planar chirality (1a–
c) were developed as precursors of a novel class of chiral
NHCs. The structural profiles of the imidazolium salts and
their derivative NHCs were investigated by several methods,
which revealed that one of the two possible conformers for
the imidazolylidenes acted as a main active species. The
NHCs derived from these imidazoliums showed remarkably
[1] For selected reviews of topologically chiral molecules, see: a) A.
Togni, Angew. Chem. 1996, 108, 1581–1583; Angew. Chem. Int. Ed.
Engl. 1996, 35, 1475–1477; b) C. Piguet, G. Bernardinelli, G. Hopf-
gartner, Chem. Rev. 1997, 97, 2005–2062; c) S. Grimme, J. Harren,
Chem. Eur. J. 2008, 14, 9215 – 9222
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9221

Y. Ishida, K. Saigo et al.
A. Sobanski, F. Vçgtle, Eur. J. Org. Chem. 1998, 1491–1509;
d) M. M. Green, K.-S. Cheon, S.-Y. Yang, J.-W. Park, S. Swansburg,
W. Liu, Acc. Chem. Res. 2001, 34, 672–680; e) J. M. Brunel, Chem.
Rev. 2005, 105, 857–898.
[10] a) Y. Ishida, H. Miyauchi, K. Saigo, Chem. Commun. 2002, 2240–
2241; b) Y. Ishida, D. Sasaki, H. Miyauchi, K. Saigo, Tetrahedron
Lett. 2004, 45, 9455–9459; c) Y. Ishida, H. Miyauchi, K. Saigo, Het-
erocycles 2005, 66, 263–273; d) Y. Ishida, D. Sasaki, H. Miyauchi, K.
Saigo, Tetrahedron Lett. 2006, 47, 7973–7976.
[11] For examples of cyclophanes containing two imidazoliums, see:
a) E. Alcalde, M. Alemany, L. Pꢅrez-Garcꢁa, M. L. Rodriguez, J.
Chem. Soc. Chem. Commun. 1995, 1239–1240; b) P. Cabildo, D.
Sanz, R. M. Claramunt, S. A. Bourne, I. Alkorta, J. Elguero, Tetrahe-
dron 1999, 55, 2327–2340; c) M.-M. Luo, S.-J. Guo, C.-H. Zhou, R.-
G. Xie, Heterocycles 1995, 41, 1421–1424; d) C.-H. Zhou, R.-G. Xie,
H.-M. Zhao, Org. Prep. Proced. Int. 1996, 28, 345–347.
[12] Y. Matsuoka, Y. Ishida, D. Sasaki, K. Saigo, Tetrahedron 2006, 62,
8199–8206.
[2] a) R. Noyori, H. Takaya, Acc. Chem. Res. 1990, 23, 345–350; b) O.
de Lucchi, Pure Appl. Chem. 1996, 68, 945–949; c) Y. Kubo, Synlett
1999, 161–174; d) V. A. Davankov, Enantiomer 2000, 5, 209–223;
e) T. Nakano, Y. Okamoto, Chem. Rev. 2001, 101, 4013–4038; f) J.
Lacour, V. Hebbe-Viton, Chem. Soc. Rev. 2003, 32, 373–382; g) M.
Shibasaki, S. Matsunaga, Chem. Soc. Rev. 2006, 35, 269–279;
h) R. P. Lemieux, Chem. Soc. Rev. 2007, 36, 2033–2045.
[3] For examples of organocatalysts with topological chirality, see the
following and references therein: P. I. Dalko, L. Moisan, Angew.
Chem. 2004, 116, 5248–5286; Angew. Chem. Int. Ed. 2004, 43, 5138–
5175.
[4] a) D. Enders, T. Balensiefer, Acc. Chem. Res. 2004, 37, 534–541;
b) J. S. Johnson, Angew. Chem. 2004, 116, 1348–1350; Angew.
Chem. Int. Ed. 2004, 43, 1326–1328; c) M. Christmann, Angew.
Chem. 2005, 117, 2688–2690; Angew. Chem. Int. Ed. 2005, 44, 2632–
2634; d) K. Zeitler, Angew. Chem. 2005, 117, 7674–7678; Angew.
Chem. Int. Ed. 2005, 44, 7506–7510; e) N. Marion, S. Dꢁez-Gon-
zꢂlez, S. P. Nolan, Angew. Chem. 2007, 119, 3046–3058; Angew.
Chem. Int. Ed. 2007, 46, 2988–3000.
[5] a) J. C. Sheehan, T. Hara, J. Org. Chem. 1974, 39, 1196–1199; b) D.
Enders, K. Breuer, J. Runsink, J. H. Teles, Helv. Chim. Acta 1996,
79, 1899–1902; c) Y. Suzuki, K. Yamauchi, K. Muramatsu, M. Sato,
Chem. Commun. 2004, 2770–2771; d) A. Chan, K. A. Scheidt, Org.
Lett. 2005, 7, 905–908; e) T. Kano, K. Sasaki, K. Maruoka, Org.
Lett. 2005, 7, 1347–1349.
[13] a) K. L. Tan, R. G. Bergman, J. A. Ellman, J. Am. Chem. Soc. 2001,
123, 2685–2686; b) M. Shi, H.-X. Qian, Appl. Organomet. Chem.
2005, 19, 1083–1089.
[14] See the Supporting Information.
[15] a) U. Zoller, Tetrahedron 1988, 44, 7413–7426; b) F. M. Rivas, U.
Riaz, A. Giessert, J. A. Smulik, S. T. Diver, Org. Lett. 2001, 3, 2673–
2676.
[16] For examples of the NHC-catalyzed annulation of enals and alde-
hydes/ketones/aldimines, see: a) S. S. Sohn, E. L. Rosen, J. W. Bode,
J. Am. Chem. Soc. 2004, 126, 14370–14371; b) C. Burstein, F. Glo-
rius, Angew. Chem. 2004, 116, 6331–6334; Angew. Chem. Int. Ed.
2004, 43, 6205–6208; c) M. He, J. W. Bode, Org. Lett. 2005, 7, 3131–
3134; d) S. S. Sohn, J. W. Bode, Org. Lett. 2005, 7, 3873–3876; e) V.
Nair, S. Vellalath, M. Poonoth, R. Mohan, E. Suresh, Org. Lett.
2006, 8, 507–509; f) V. Nair, S. Vellalath, M. Poonoth, E. Suresh, J.
Am. Chem. Soc. 2006, 128, 8736–8737.
[6] a) R. L. Knight, F. J. Leeper, Tetrahedron Lett. 1997, 38, 3611–3614;
b) A. U. Gerhard, F. J. Leeper, Tetrahedron Lett. 1997, 38, 3615–
3618; c) R. L. Knight, F. J. Leeper, J. Chem. Soc. Perkin Trans. 1
1998, 1891–1894; d) M. S. Kerr, T. Rovis, Synlett 2003, 1934–1936;
e) F. Glorius, G. Altenhoff, R. Goddard, C. Lehmann, Chem.
Commun. 2002, 2704–2705; f) J. R. Struble, J. Kaeobamrung, J. W.
Bode, Org. Lett. 2008, 10, 957–960; g) Y. Matsuoka, Y. Ishida, K.
Saigo, Tetrahedron Lett. 2008, 49, 2985–2989.
[7] a) J. Pesch, K. Harms, T. Bach, Eur. J. Org. Chem. 2004, 2025–2035;
b) T. J. Seiders, D. W. Ward, R. H. Grubbs, Org. Lett. 2001, 3, 3225–
3228; c) O. Winkelmann, D. Linder, J. Lacour, C. Nꢃther, U. Lꢄning,
Eur. J. Org. Chem. 2007, 3687–3697.
[17] For examples of the NHC-catalyzed annulation of enals and electro-
philes other than simple carbonyl compounds, see: a) E. M. Phillips,
T. E. Reynolds, K. A. Scheidt, J. Am. Chem. Soc. 2008, 130, 2416–
2417; b) J. Seayad, P. K. Patra, Y. Zhang, J. Y. Ying, Org. Lett. 2008,
10, 953–956; c) A. Chan, K. A. Scheidt, J. Am. Chem. Soc. 2008,
130, 2740–2741.
[18] By using special electrophiles such as nitrobenzene/nitrone/1-acyl-2-
aryldiazene in this reaction (ref. [17]), good to excellent enantiose-
lectivity was often achieved (up to 93% ee, ref. [17a]). However, it
is still difficult to realize satisfactory selectivity by using a simple al-
dehyde/ketone/aldimine as an electrophile (up to 66% ee, ref. [6g]).
[19] By using the NHC catalyst generated from 1a, the self-annulation of
12 to form 14 proceeded with appreciable diastereo-and enantiose-
lectivities. The optimization of the reaction conditions for the self-
annulation is under progress. Details will be reported elsewhere in
the near future.
[8] a) Y. Ma, C. Song, C. Ma, Z. Sun, Q. Chai, M. B. Andrus, Angew.
Chem. 2003, 115, 6051–6054; Angew. Chem. Int. Ed. 2003, 42, 5871–
5874; b) A. Fꢄrstner, M. Alcarazo, H. Krause, C. W. Lehmann, J.
Am. Chem. Soc. 2007, 129, 12676–12677.
[9] a) S. M. Mennen, J. D. Gipson, Y. R. Kim, S. J. Miller, J. Am. Chem.
Soc. 2005, 127, 1654–1655; b) Y. Tachibana, N. Kihara, T. Takata, J.
Am. Chem. Soc. 2004, 126, 3438–3439; c) Y. Matsuomoto, K. To-
mioka, Tetrahedron Lett. 2006, 47, 5843–5846.
[20] C.-J. Chang, J.-M. Fang, L.-F. Liao, J. Org. Chem. 1993, 58, 1754–
1761.
Received: May 17, 2008
Published online: August 27, 2008
9222
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 9215 – 9222