Highly enantioselective biphasic iminium-catalyzed epoxidation of alkenes. on the importance of the counterion and of N(sp2)-C(sp3) Rotamers

Article abstract of DOI:10.1002/adsc.200800778

Diastereomeric biaryliminium cations made of an (Ra)-5,5′,6,6′, 7,7′,8,8′-octahydrobinaphthyl core and exocyclic appendages derived from (S)or (R)-3,3-dimethylbutan-2-amine are effective asymmetric epoxidation catalysts for unfunctionalized alkenes. Herei

Full text of DOI:10.1002/adsc.200800778

FULL PAPERS
DOI: 10.1002/adsc.200800778
Highly Enantioselective Biphasic Iminium-Catalyzed Epoxidation
2
À
of Alkenes. On the Importance of the Counterion and of N
(sp )
C
ACHTUNGTRENNUNG
Roman Novikov,a Gꢀrald Bernardinelli,b and Jꢀrꢁme Lacoura,*
a
Dꢀpartement de Chimie Organique, University of Geneva, Quai Ernest-Ansermet 30, CH-1211 Geneva 4, Switzerland
Fax : (+41)-(0)22-379-3215; e-mail: Jerome.Lacour@unige.ch
Laboratoire de Cristallographie, University of Geneva, Quai Ernest Ansermet 24, CH-1211 Geneva 4, Switzerland
b
Received: December 16, 2008; Published online: March 4, 2009
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/adsc.200800778.
Abstract: Diastereomeric biaryliminium cations “lack” of stereochemical control from the chiral exo-
made of an (Ra)-5,5’,6,6’,7,7’,8,8’-octahydrobinaphth- cyclic appendage in this type of catalysts is due to
2
À
yl core and exocyclic appendages derived from (S)- the existence of atropisomers around the N
or (R)-3,3-dimethylbutan-2-amine are effective asym-
(sp )
E
(sp3) bond that links the azepinium core to the exo-
metric epoxidation catalysts for unfunctionalized al- cyclic stereocenter. Finally, we develop a general
kenes. Herein, we report that the negative counter- model to predict with certainty the high selectivity in
ion of the iminium salts has to be chosen wisely. the formation of non-racemic epoxides of defined
À
While the hexafluoroantimonate anion [SbF6 ] is op- absolute configuration.
timal for reliable results, one has to be careful about
À
other anions and tetraphenylborate [BPh4 ] in partic- Keywords: counterion; epoxidation; iminium salts;
ular. We also detail that the so far unexplained organocatalysis; rotamers
Introduction
fins in particular. Moreover, the propensity of
iminium ions to react with
Oxoneꢂ
Non-racemic epoxides are useful precursors and (2KHSO5·KHSO4·K2SO4) to generate the oxaziridini-
building blocks in synthetic organic chemistry. Quite a um species renders the development of catalytic pro-
few effective systems have been developed for their cesses possible [Eq. (1)].[12,13]
preparation,[1] and the catalytic asymmetric epoxida-
tion of olefins has proven to be one of the most pow-
erful approaches. Most of them are based on transi-
tion metals such as the Katsuki–Sharpless epoxidation
of allylic alcohols with chiral titanium catalysts[2], va-
nadium-catalyzed epoxidation of allylic[3] and homoal-
lylic[4] alcohols and the Katsuki–Jacobsen protocol for
unfunctionalized olefins[5]. During the last decade,
much effort has been devoted to the development of
organocatalytic systems that afford metal-free proce-
In the recent years, quite a few successful enantio-
dures, such as asymmetric epoxidation catalyzed by selective variants of the reaction have been report-
chiral ketones and iminium salts.[6] In this field, three- ed,[14–16] many of them based on configurationally
membered ring heterocycles such as dioxiranes,[7,8] stable biarylazepinium skeletons. Some of these deriv-
oxaziridines,[9] ammonium and oxaziridinium salts[7] atives are doubly bridged biphenyl 1,[17] binaphthyl
have been shown to be effective oxidation agents.
2[18,19] and 5,5’,6,6’,7,7’,8,8’-octahydrobinaphthyl 3 aze-
Oxaziridinium ions are attractive alternatives to the pinium cations[20] that are displayed in Figure 1 and
commonly used dioxiranes.[10] These organic salts are Figure 2. Of importance for this study, these catalysts
effective oxygen transfer reagents towards nucleophil- contain an exocyclic chiral appendage R* in addition
ic substrates[11] and electron-rich unfunctionalized ole- to the twisted biaryl axis.
596
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 596 – 606

Highly Enantioselective Biphasic Iminium-Catalyzed Epoxidation of Alkenes
FULL PAPERS
links the azepinium core to the chiral appendage. Sig-
nificantly, these rotameric conformations are the key
to an understanding of the lack of the stereochemical
influence of the exocyclic stereocenter. Finally, and
also importantly, we report on the importance of the
counterion associated with the cationic catalytic moi-
eties to afford reliable reactivity in these enantioselec-
tive epoxidation reactions.
Figure 1. Doubly bridged biphenyl-, 1, and binaphthyl-, 2,
azepinium cations containing a chiral exocyclic appendage
R* (e.g., l-acetonamine or 3,3-dimethylbutan-2-amine).
Results and Discussion
On the Importance of the Counterion
In fact, in our previous study on 5,5’,6,6’,7,7’,8,8’-octa-
hydrobinaphthylazepinium catalysts with appendages
derived from (S)- and (R)-3,3-dimethylbutan-2-amine,
catalysts 3a and 3b (Figure 3), or from (S)- and (R)-1-
Figure 2. 5,5’,6,6’,7,7’,8,8’-Octahydrobinaphthyliminium
3
catalysts and q and F dihedral angles (measured inside and
outside the N-containing seven-membered ring, respective-
ly).
Recently, it was proposed that the larger the dihe-
dral angles q and F around the central bond joining
the aromatic rings (Figure 2) are, the stronger is the
stereocontrol of the reaction by the biaryl axis over
the exocyclic chiral appendage.[20] In fact, salts (Ra,S)
and (Ra,R) of catalysts 2 (unlike 1) provide non-race-
mic epoxides with the same absolute configuration
Figure 3. Diastereomeric 5,5’,6,6’,7,7’,8,8’-octahydrobinaph-
thyliminium catalysts 3a and 3b derived from (S)- and (R)-
3,3-dimethylbutan-2-amine, respectively.
and with virtually the same enantiomeric purity de- phenylpropylamine, a somewhat puzzling observation
spite the diastereomeric relationship.[19] Based on this was made.[20] Rather large differences in enantioselec-
dihedral angle hypothesis, partially hydrogenated aze- tivity were observed for the reactions performed with
pinium cations of type 3 were prepared and they dis- the iminium catalysts combined with bromide or with
played excellent levels of enantioselectivity for certain tetraphenylborate counterions respectively. This dif-
olefins such as 1-phenyl-3,4-dihydronaphthalene 5a ference (up to 5% ee with 3b and 14% with the imini-
(ee up to 94%, Oxone/NaHCO3/18-C-6/CH2Cl2/H2O). ums derived from 1-phenylpropylamine) was depen-
However, for other simple alkenes, such as trans-a- dent upon the nature of the olefin and catalysts at
methylstilbene, only a moderate level of induction play; the highest ee values being observed for the
was obtained (ee up to 77%). While such a difference BPh4 salts.
in selectivity for these two substrates has been ob-
We thus decided to characterize this effect better
served in all previously reported examples of epoxida- and looked more closely at what was happening with
tion with chiral iminium catalysts, there has been, to the tetraphenylborate salts. First, a study of the enan-
our knowledge, no attempt to account for this differ- tioselectivity of the epoxidation of 1-phenylcyclohex-
ence rationally. No general model has been provided ene as a function of catalyst loading was performed
so far to predict with certainty which type of olefins using salt [3a]ACTHUNRGTNEUNG[BPh4]. The results are reported in
reacts more selectively than others with this class of Table 1. The stereoinduction is greatly affected by the
organocatalysts and which absolute configuration is amount of active salt in the medium and, to a bit of
afforded for the epoxides.
our surprise, the larger the amount of catalyst the
Herein, based on an extensive experimental study, lower the selectivity.
we report a rationalization to foresee with confidence
This unusual observation was difficult to rationalize
which trisubstituted olefins lead to high levels of at first glance. One possible explanation was, at low
enantioselectivity. We also provide details on the solu- amount of catalyst, the occurrence of a degradation
tion (and solid-state) behaviour of the iminium biaryl generating a more selective catalyst in situ. Care was
catalysts of type 3 and demonstrate the existence of then taken to fully characterize the result of these re-
2
(sp3) bond that actions and those performed with the largest amount
À
atropisomers around the N
(sp ) C
Adv. Synth. Catal. 2009, 351, 596 – 606
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
597

Roman Novikov et al.
FULL PAPERS
Table 1. Epoxidation of 1-phenylcyclohexene using catalyst
Catalyst Preparation
[3a]ACHTUNGTRENNUNG
[BPh4].[a]
Hexafluorophosphate was first evaluated as an alter-
native. The PF6 iminium salts of cations 3a and 3b
were easily prepared by ion exchange metathesis
from the corresponding bromide salts (~70% yield).
However, these salts were stable only in the solid
state. 19F and 31P NMR spectroscopic analysis of the
salts revealed a hydrolysis of the counterion already
Catalyst [mol%]
ee[b] [%]
1.0
2.5
5
90
90
80
66
10
[a]
Conditions: 1-phenylcyclohexene (0.2 mmol), catalyst (x
mol%), 2.5 mol% of 18-C-6, 1.1 equiv. Oxoneꢂ, 4.0 equiv.
in CD2Cl2 only; the signals of the cations remaining,
1
NaHCO3, CH2Cl2/H2O (3:2) , 2 h, 08C. Average of at however, unchanged in the H and 13C NMR spectra.
least two runs.
Determined by CSP-GC (Chiraldex Hydrodex b-3P).
Hexafluoroantimonate, which was studied next, was
found to be suitable for the catalysis work (see the
[b]
next section). The synthesis of catalysts [3a]
and [3b][SbF6] was performed in two steps following
a general and highly-reproducible procedure [Eq. (2)
ACHTUNGTRENNUNG[SbF6]
AHCTUNGTRENNUNG
of catalysts in particular. Relatively high amounts of and Experimental Section].
biphenyl could be identified in the GC-MS analyses
of the reaction mixtures, a compound that was fur-
thermore isolated in pure form from the crudes. Dif-
ferent hypotheses could then account for this observa-
tion as biphenyl might (i) be an impurity brought in
solution along with the starting alkene, (ii) result from
degradation and aromatization under the reaction
conditions of in situ formed 1-phenylcyclohexene
oxide or (iii) be derived from tetraphenylborate anion
by an oxidative decomposition pathway.[21]
To distinguish these hypotheses, a few tests were
Tertiary
5,5’,6,6’,7,7’,8,8’-octahydrobinaphthylaze-
performed. First of all, the purity of 1-phenylcyclo- pines (Ra,S)-4a and (Ra,R)-4b, of which a synthesis in
hexene (commercial, 98%) was checked by GC-MS 6 steps has been previously reported, were used as
and no trace of biphenyl could be found. Then, 1-phe- starting materials.[20] Their treatment with N-bromo-
nylcyclohexene was submitted to the epoxidation pro- succinimide (NBS) in dichloromethane provided a
cedure with bromide salts as catalysts (e.g., [3a][Br]); fast and clean formation of the desired iminium bro-
no biphenyl was observed rendering unlikely the exis- mide salts in 5 min only. Subsequent anion metathesis
tence of a pathway from 1-phenylcyclohexene oxide in acetone with sodium hexafluoroantimonate (5 min)
to biphenyl (for details see Supporting Information). and trituration in EtOH provided the pure catalysts
Finally, [3a]ACHTUNGTRENNUNG[BPh4] (5 mol%) was treated under the (Ra,S)-[3a]ACHTNURTGENN[GNU SbF6] and (Ra,R)-[3b]ACHTNGUTRENUN[GN SbF6] in good yield
reaction conditions in the absence of alkene and, for two steps (81–87%).
after two hours of vigorous stirring at 08C and filtra-
1
tion of the biphasic mixture, a H NMR analysis was
performed which revealed a complete disappearance Enantioselective Epoxidation Reactions
of the tetraphenylborate species and only signals of
biphenyl.
Initial Reactivity/Selectivity Screen
Clearly, under the oxidative reaction conditions,[21]
the lipophilic counterion undergoes a decomposition The first task at hand was the verification that differ-
to form biphenyl as by-product. This process is of par- ent loading of salts [3a][SbF6] and [3b][SbF6] would
A
ACHTUNGTRENNUNG
ticular importance as it modifies profoundly the solu- afford non-racemic epoxides with identical levels of
bility and/or the partition ability of the [iminium/ox- enantiomeric excesses. To test this hypothesis, 1-phe-
ACHTUNGTRENNUNGaziridinium] species in the two-layer situation of the nylcyclohexene was used as substrate along with the
reaction conditions. The epoxidation, which is most following reaction conditions that are suitable for
probably occurring in the organic layer exclusively even very sensitive epoxides (Oxoneꢂ/NaHCO3/18-
when the lipophilic anion is intact,[15] might now occur crown-6/CH2Cl2/H2O/08C). The results are summar-
at the interface of the heterogeneous solvent mixture ized in Table 2.
or even in the aqueous phase upon degradation.
In view of these results, a search for a stable alter- mol% of [3a]
native to BPh4 was started and that of a “light” coun- (1S,2S)-1-phenylcyclohexene oxide with, as desired,
terion rather than a “heavy” TRISPHAT.[22]
invariable 89 and 92% ee, respectively. Clearly, cata-
The reactions were performed with 2.5, 5.0 and 10
A
ACHTUNGTRENNUNG
598
asc.wiley-vch.de
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 596 – 606

Highly Enantioselective Biphasic Iminium-Catalyzed Epoxidation of Alkenes
FULL PAPERS
Table 2. Asymmetric epoxidation of 1-phenylcyclohexene
with SbF6 salts. Catalyst loading influence.[a]
[3a]
[SbF6]
[3b]ACTHNGUETRNNU[G SbF6]
Catalyst
[mol%]
ee[b]
[%]
Conv.[c]
[%]
ee[b]
[%]
Conv.[c]
[%]
2.5
5.0
10
89
89
89
97
>99
>99
92
92
92
60
66
68
[a]
Conditions: 1-phenylcyclohexene (0.2 mmol), catalyst (x
mol%), 2.5 mol% of 18-C-6, 1.1 equiv. Oxoneꢂ, 4.0 equiv.
NaHCO3, CH2Cl2/H2O (3:2), 2 h, 08C. Average of at
least two runs.
[b]
[c]
Determined by CSP-GC (Chiraldex Hydrodex b-3P). In
all cases, the absolute configuration of 1-phenylcyclohex-
ene oxide is (À)-(1S,2S).
Figure 4. Tested commercial alkenes.
Determined by GC analysis of crude reaction mixture
using naphthalene as an internal standard.
ents being actually needed for high selectivity (in
comparison with 5c). Looking at the larger enantio-
meric excess of 5a over 5d, we reasoned that the pres-
ence of a third a-substituent next to the aromatic
lyst [3a]ACHTUNGTRENNUNG[SbF6] is more reactive than [3b]ACHTUNGTNER[NUGN SbF6] and it group was also beneficial. This was tested with trans-
can be used in the epoxidation reaction in amounts as a-methylstilbene 5e which afforded its epoxide with a
low as 2.5 mol% without significant loss of activity 65% ee value, quite a bit better than that of 5b (6%
(97% conversion in 2 h). Under the same conditions, ee); the moderate ee value being the result of antago-
only a 60% conversion to the desired epoxide is nistic effects of the favorable a-Me and unfavorable
reached with [3b]
meric excess of 92% that remains constant with larger
amounts of 3b. While [3a][SbF6] is the most reactive, to the general formula displayed in Figure 5 should
[3b][SbF6] is the most selective. It was also noted that lead to high enantiomeric excesses in the enantiose-
the epoxidation reaction was very clean and no traces lective epoxidation catalyzed by salt [3b][SbF6]; the
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
of ring opening by-products were detected by GC-MS size of the a-substituent compatible with the process
analysis of reaction mixtures. The pure epoxide was remaining to be determined in the course of the
obtained by simple filtration through a silica gel plug study.
necessarily buffered though with triethylamine.
In short, both catalysts performed as expected. For
further studies, it was decided to use the most selec-
tive salt [3b]ACHTUNGTRENNUNG[SbF6] as its reactivity is addressable by
increasing either reaction time or catalyst loading.
Figure 5. Proposed model type of alkenes giving epoxides
with high enantiomeric excesses.
Initial Substrate Screen
The epoxidation was then performed with a few
simple commercially available substrates (Figure 4)
looking for leads that would indicate which olefins Extended Substrate Screen
are best for this asymmetric catalyzed process.
Whereas the epoxidation was highly disappointing To test this hypothesis, a range of substrates was pre-
with trans-stilbene 5b and a-methylstyrene 5c (6% pared and their asymmetric epoxidation with catalyst
and 12% ee, respectively), much higher selectivity was [3b]CAHTUNGTERN[NUGN SbF6] performed. All results are presented in
obtained with trans-b-methylstyrene (5d, 64% ee) Table 3 and Table 4.
and, as already mentioned, 1-phenylcyclohexene (5a,
It should be mentioned that for all the alkenes, ex-
92% ee). In view of these results and considering that periments with different catalyst loading (mol%) led,
one aromatic group is beneficial (as in 5a), we rea- as for 1-phenylcyclohexene, to constant levels of
soned that the second aromatic group on 5b was too asymmetric induction. For 5a, full conversion to the
large for the chiral pocket of the catalyst and that a desired epoxide was obtained after 24 h at 08C. 1-
methyl or a methylene group at the same trans-b-posi- Phenyl-3,4-dihydronaphthalene 5f, another classical
tion was, however, favorable; these smaller substitu- substrate, was very reactive as well. The resulting ep-
Adv. Synth. Catal. 2009, 351, 596 – 606
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
599

Roman Novikov et al.
FULL PAPERS
Table 3. Asymmetric epoxidation of various alkenes mediated by catalyst [3b]
ACHTUNGTRENNUNG
Substrate
Configuration[i]
5a
5f
2.5
2.5
24
24
85 (>99[b])
83 (>97[c])
92[e]
(À)-(1S,2S)
91[f], (98[j]) (+)-(1R,2S)
5g
5h
5
48
48
95 (>99[b])
80 (>97[c])
90[g]
88[f]
(À)-(S)
(À)-(S)
20
5i
10
43
87 (>99[b])
91[f]
(À)
5j
10
5
47
84 (>99[b])
67 (>99[b])
94 (>99[b])
67 (75[b])
91[f]
94[f]
90[f]
98[f]
93[h]
(À)
5k
5l
24
(À)
10
20
10[h]
24
(À)-(S)
(+)
5m
5m
91
66[h]
85[h] (>99[h])
(+)
(E)-5n
2.5
5
48
48
91 (>99[b])
90[e]
76[e]
(À)-(2S,3S)
(Z)-5n[k]
93[k] (>99[b])
–
[k]
[a]
Conditions: substrate (0.2 mmol), catalyst (x mol%), 2.5 mol% of 18-C-6, 1.1 equiv. Oxoneꢂ, 4.0 equiv. NaHCO3, CH2Cl2/
H2O (3:2), 08C.
[b]
[c]
[d]
[e]
[f]
Determined by GC-MS (HP-5MS) analysis of the crude reaction mixture.
1
Determined by H NMR analysis of the crude reaction mixture (400 MHz).
Isolated yields of pure epoxides.
Determined by CSP-GC (Chiraldex Hydrodex b-3P).
Determined by CSP-HPLC (see Supporting Information).
[g]
[h]
[i]
1
Determined by H NMR spectroscopy in the presence of (+)-Eu
N
Results obtained with catalyst [3a]ACTHNUGTRENUNG[SbF6].
The absolute configuration of major enantiomers was determined by comparison of optical rotation with that reported in
the literature.
After single recrystallization from n-hexane.
[j]
[k]
A 95:5 mixture of (Z)-5n and (E)-5n was used. Due to the stereospecificity of the reaction and the occurrence of a minor
diastereomeric epoxide, the optical rotation was not measured.
600
asc.wiley-vch.de
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 596 – 606

Highly Enantioselective Biphasic Iminium-Catalyzed Epoxidation of Alkenes
FULL PAPERS
Table 4. Ring size effect on the epoxidation of few cycloalkenes with catalyst [3b]ACHTUNGTRNEUNG
[SbF6].[a]
Substrate
Yield[c] (Conversion [%][b])
ee [%]
Configuration[f]
5a
5o
5p
85 (>99)
92[d]
(À)-(1S,2S)
85 (>99)
92 (>99)
72[e]
83[d]
(À)-(1S,2S)
(À)
[a]
Conditions: substrate (0.2 mmol), catalyst (2.5 mol%), 2.5 mol% of 18-C-6, 1.1 equiv. Oxoneꢂ, 4.0 equiv. NaHCO3,
CH2Cl2/H2O (3:2), 24 h, 08C.
Determined by GC-MS (HP-5MS) analysis of the crude reaction mixture.
Isolated yields.
Determined by CSP-GC (Chiraldex Hydrodex b-3P).
Determined by CSP-HPLC (OD-H column).
The absolute configuration of major enantiomers was determined by comparison of optical rotation with that reported in
the literature.
[b]
[c]
[d]
[e]
[f]
oxide was isolated in 83% yield and 91% ee. Similar bent conformation intermediate between boat and
results were obtained with 1,1-diphenylprop-1-ene 5g chair forms and the aromatic groups adjacent to the
(90% ee); the system being however slightly less reac- double bond are twisted by 45 and 658 out of the
tive. Interestingly, moving from this substrate to the plane of the alkene.[24] Despite this very large differ-
analogous allylic alcohol (5h, 88% ee), only a very ence, both 5l and 5m afforded the corresponding ep-
small decrease in enantioselectivity was measured. oxides in high to very high enantiomeric purity (90%
The reactivity was however much lower as 5 mol% and 98% ee respectively). The reaction with 5m was
and 20 mol% of catalyst [3b]ACTHNUTRGENUNG[SbF6] were required to however much slower than others as, with 20 mol%
achieve full conversion with 5g and 5h, respectively of catalyst and 4 days, the isolated epoxide was ob-
under the same reaction time. This difference in reac- tained in only 67% yield (75% conversion).[25] In view
tivity may be explained by the electron-withdrawing of this result, 5m was treated with the more active
effect of the hydroxy moiety that renders the double catalyst [3a]ACTHNUGTRENUNG[SbF6]. As expected, the reaction was
bound of allylic alcohol less electrophilic and also faster and only required 10 mol% of iminium salt to
possibly by a less favorable partition of the olefin in yield the epoxide in 85% yield and still a decent 93%
the biphasic CH2Cl2/water layers. Then, a range of ee after ca. 3 days.
substrates structurally analogous to 5g was prepared
The impact of the olefin geometry was also tested
bearing either electron-withdrawing (F, Cl: 5i and 5j, using the E and Z geometrical isomers of 1-methyl-1-
respectively) or electron-donating (Me: 5k) groups at phenylpropene 5n. If the reaction could be performed
the para-positions. In the cases of 5i and 5j, essentially with an essentially pure E isomer (>99%), that of the
the same level of asymmetric induction was observed Z isomer was conducted with a sample only 95%
as for 5i (ee 91% vs. 90% ee), whereas electronically- pure in the cis form. In the first case, (E)-5n was ep-
rich alkene 5k led to an increase in the enantiomeric oxidized using just 2.5 mol% of catalyst [3b]ACHTNUTRGNE[NUG SbF6] in
excess value (94% ee). If only a moderate impact was 48 h and 90% ee, whereas (Z)-5n afforded its epoxide
seen on the selectivity, it was not the case on the reac- in only 76% ee. Moreover, the Z isomer was less reac-
tivity. While only 5 mol% of catalyst was required to tive; 5 mol% of catalyst being required to achieve full
achieve full conversion after 48 h at 08C with 5g, its conversion after 48 h.
electron-rich analogue 5k reacted in half the time for
Finally, after these largely successful examples, two
the same catalyst loading. Electron-poor derivatives other 1-phenylcycloalkenes, namely the 5- and 7-
5i and 5j required however quite a lot more of the membered ring 5o and 5p, were tested in the epoxida-
catalyst as full conversion was observed only after 2 tion reaction (Table 4). Only 2.5 mol% of [3b]
ACHTUNGTRENNUNG[SbF6]
days with 10 mol% of [3b][SbF6]. was required to achieve full and clean conversion of
AHCTUNGTRENNUNG
We then turned our attention to 5l and 5m, cyclic olefins to the corresponding epoxides that were isolat-
analogues of 5g derived from 9H-fluoren-9-one and ed in good yields (85–92%), albeit with lower enan-
dibenzosuberone, respectively. Whereas all carbons tiomeric excesses (72 and 83% ee, respectively). The
are virtually coplanar in 5l,[23] this is not the case for results are compared to that of alkene 5a.
5m. The central cycloheptadiene ring of 5m adopts a
Adv. Synth. Catal. 2009, 351, 596 – 606
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
601

Roman Novikov et al.
FULL PAPERS
Validity of the Alkene Model Hypothesis and
Absolute Sense of Stereoinduction
chiral face of the alkene; the top face in the geometri-
cal description provided on Figure 7. For substrates
5a, 5f, 5g, 5l, (E)-5n and 5o, it is the si face of the b-
Clearly, at this stage, the alkene model type proposed carbon of the alkene. For 5h, due to the presence of
earlier for double bonds that would provide high the oxygen at the allylic position which changes the
levels of enantioselectivity in reactions catalyzed by priority sequence in the CIP rule, it is the re face. It is
[3b]ACHTUNGTRENNUNG[SbF6] is a working model. From 5f to (E)-5n, the then highly probable that compounds 5i, 5j, 5k, 5m
lowest ee value is 88% – and this corresponds to 5h and 5p have reacted in a similar fashion.
that might react differently in the biphasic catalytic
system due to the presence of hydroxy group. Having
at the trans-b-position from the aromatic group a Slow Rotation around the N
small alkyl moiety R’ of type CH3, CH2OH or CH2 Importance of the Resulting rotameric Situation
2
(sp3) Bond and
À
(sp ) C
(cyclic) is something favorable. The group at the a-
position R can be possibly quite large and even twist- Previously, it has been shown that a slow rotation
2
(sp3) bonds can occur when bulky
À
ed out of the plane of the double bond as in 5m.
around N
(sp ) C
If (E)-5n also fits the model, it is however not the substituents are attached to a stereogenic carbon
case for the (Z) isomer. For that substrate, the b- linked to an sp2-hybridized nitrogen atom.[26] It results
methyl group is positioned in a quadrant usually occu- in atropisomeric conformations that can have an
pied by an H-atom only and this most probably leads impact on the stereochemical outcome of reactions.
to an unsatisfactory interaction with the iminium cata- For some acridinium cations, a rotation barrier of
lyst and hence the lower enantioselectivity (Figure 6). 20.1 kcal·molÀ1 was recently measured for the 3,3-di-
The exact reasons why 1-phenylcyclopentene 5o and methylbutan-2-amine derivative.[27] In view of this
1-phenylcycloheptene 5p lead to lower ee values result, it was likely that such a hindered rotation
2
(sp3) bond that links the azepini-
À
remain undetermined for the moment, the different around the N
G
conformations of the three rings systems possibly play um core to the chiral appendage would also occur for
a role.
[SbF6].[16] Care was thus taken
salts [3a]ACHTGNUTRENNU[G SbF6] and [3b]ACTHUNGTRENNNUG
to characterize the diarylazepinium cations by varia-
ble temperature (VT) NMR in a search for atropiso-
meric anti- and synperiplanar conformations.
1
While the H NMR analysis of [3b]
N
displayed sharp signals at 298 K, a gradual decrease
of the temperature led to a broadening of all signals
(with a coalescence around 243 K) before they
became sharp again at 193 K (Figure 8). Then, two
sets of signals appeared indicative of the existence
Figure 6. Different recognition of E and Z isomers.
Finally, the stereochemical outcome was analyzed indeed of two atropisomeric conformers and of their
for the reactions providing epoxides of known abso- slow interconversion on the NMR time scale at
lute configuration and the results are summarized in 193 K. The signal of the iminium proton at d=
Figure 7. All the major enantiomers produced come 8.6 ppm was particularly useful for monitoring. If one
from the addition of the O-atom onto the same pro- considers that the NMR coalescence is reached at
243 K, then an approximate energy barrier in the
range of 11–12 kcal·molÀ1 can be estimated. {The rela-
tionship DG° =RT
cACHTUNTRGENN[GU 22.96+lnACHTUNGTREN(NUGN Tc/Dn)] was used to de-
termine the activation energy, DG°, from the coales-
cence temperature, Tc (243 K), and the frequency sep-
aration of the peaks of the iminium protons, Dn
(115.1 Hz at 193 K)}. An atropisomeric enrichment
was also observed between the anti- and synperiplanar
conformers and a 85:15 ratio was measured by inte-
gration of the respective signals.
In 13C NMR spectroscopy, an analogous behavior
was observed although with a broadening of some of
the signals appearing already at 298 K; these enlarge-
ments disappearing at 193 K along with the splitting
of all carbon signals (see Supporting Information).
The determination of the conformations of the
major and minor atropisomer in solution was realized
Figure 7. Facial selectivity.
602
asc.wiley-vch.de
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 596 – 606

Highly Enantioselective Biphasic Iminium-Catalyzed Epoxidation of Alkenes
FULL PAPERS
Figure 9. NOESY experiment (CD2Cl2, 500 MHz) of [3b]-
[SbF6] at 193 K. A color variant of this drawing is included
AHCTUNGTRENNUNG
in the Supporting Information.
Figure 8. 1H NMR (d=9.5–3.5 ppm, CD2Cl2, 500 MHz) of
[3b][SbF6] at (a) 298 K and (b) 193 K. A color variant of
ACHTUNGTRENNUNG
this drawing is included in the Supporting Information.
by a NOESY experiment at 193 K. In the case of the
minor conformer, a strong cross-peak was observed
between the signal of the iminium proton (H1’, d=
8.68 ppm) and that of the hydrogen a to t-Bu group
(d=4.09 ppm, Figure 9) whereas, for the major atrop-
isomer, a weak cross-peak was only observed for the
analogous signals. For that major isomer, a strong
cross-peak was nevertheless seen between the H2
signal and that of the ortho-aromatic proton next to
the CH2N group. It strongly points towards syn- and
antiperiplanar conformations for the minor and major
conformers, respectively (Figure 8 and Figure 9).
Figure 10. ORTEP view of [3b]ACHTNUGRTENUNG[SbF6]. Ellipsoids are pre-
sented at the 50% probability level.
Monocrystals of compound [3b]ACTHNUTRGENUNG[SbF6] were ob-
tained and an X-ray structural analysis was per-
formed. The ORTEP view is reported in Figure 10.
Clearly, in the solid state, the antiperiplanar confor- drogen a to t-Bu group (d=4.05 ppm). Then, only a
mer is the preferred atropisomer as well.[28]
synperiplanar geometry can account for the spectrum
Interestingly, for catalyst [3a][SbF6], a totally differ- of the unique atropisomer of 3a. The reason for this
ent situation was observed (Figure 11). change in the preferred atropisomeric population be-
At 193 K, 1H and 13C NMR analyses revealed a tween 3a and 3b is unclear at the moment.
lone set of signals indicating the presence of a single However, this dichotomy in the atropisomeric situa-
AHCTUNGTRENNUNG
conformer (>98%). A NOESY experiment was per- tion is important as it is a possible reason for (i) the
formed at low temperature (193 K) and a strong “lack” of observed stereochemical influence from the
cross-peak was observed between the signal of the exocyclic auxiliary and (ii) the higher reactivity of dia-
iminium proton (H1’, d=8.62 ppm) and that of the hy- stereomer 3a over 3b during the course of the study.
Adv. Synth. Catal. 2009, 351, 596 – 606
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
603

Roman Novikov et al.
FULL PAPERS
Figure 12. Importance of the rotameric situation for the
“lack” of stereochemical influence of the R* appendage. A
color variant of this drawing is included in the Supporting
Information.
Conclusions
Herein, we have looked for an optimal chiral iminium
cation/achiral anion combination and reported on (i)
the origin of the lack of stereochemical control from
the chiral exocyclic appendage, (ii) the unexpected
dangers of using BPh4 as counterion and (iii) on a
general model to predict high selectivity in the forma-
tion of non-racemic epoxides of defined absolute con-
figuration.
Figure 11. 1H NMR (d=9.5–3.5 ppm, CD2Cl2, 500 MHz) of
[3a]
ACHTUNGTRENNUNG[SbF6] at (a) 298 K and (b) 193 K. A color variant of this
drawing is included in the Supporting Information.
Indeed, the fact that 3a and 3b are exclusively and Experimental Section
predominantly synperiplanar and antiperiplanar, re-
For analytical data of all new compounds, HPLC, GC and
spectively, creates a situation in which the bulky tert-
butyl side chain of the chiral auxiliary is always posi-
tioned on the bottom (si face, C1 atom) of the imini-
um ion irrespective of the diastereomeric situation
(Figure 12). As such, iminium ions 3a and 3b react
with the same sense of stereoinduction and little influ-
ence from the chiral auxiliary.
This geometrical arrangement of the side chain fur-
thermore reinforces the natural preference for a top
face attack of the peroxymonosulfate anion on the
1H NMR determination of the enantiomeric purity of the
chiral epoxides, see Supporting Information.
General Procedure for the Synthesis of the Iminium
SbF6 Salts
To a solution of substrate (4a or 4b)[20] in CH2Cl2 (2 mL for
100 mg of starting tertiary amine), NBS (1.0 equiv.) was
added in small portions (exothermic reaction). The resulting
deep yellow solution was stirred for 5 min at ambient tem-
(Ra)-iminium species to generate oxaziridinium rings perature.
A solution of sodium hexafluoroantimonate
(1.0 equiv. in 1.0 mL of acetone) was added in one portion,
and the resulting mixture was stirred for 5 min. The suspen-
sion was diluted with CH2Cl2 (15 mL), water (15 mL) and
the resulting two-phase mixture separated. The aqueous
phase was extracted with dichloromethane (2ꢄ10 mL). The
combined organic layers were dried (Na2SO4), filtered and
concentrated under vacuum. The crude product was purified
by trituration from ethanol. The resulting salts were filtered,
washed with a small amount of ethanol, diethyl ether, n-pen-
tane and dried under high vacuum at 708C for at least one
day.
of RC,RN configuration in accordance with the calcula-
tions from Washington and Houk[29] and the experi-
mental results from Zavada and collaborators.[30]
Finally, in the case of 3a, the fact that the atropiso-
meric situation is shifted towards a single rotamer
may also entice a faster reaction of this diastereomer
with the peroxymonosulfate anion than with 3b for
which the mixed antiperiplanar/synperiplanar popula-
tion retards possibly the formation of the RC,RN ox-
ACHTUNGTRENNUNGaziridinium species.
604
asc.wiley-vch.de
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 596 – 606

Highly Enantioselective Biphasic Iminium-Catalyzed Epoxidation of Alkenes
FULL PAPERS
Zhang, A. Basak, Y. Kosugi, Y. Hoshino, H. Yamamo-
to, Angew. Chem. 2005, 117, 4463–4465; Angew. Chem.
Int. Ed. 2005, 44, 4389–4391.
Typical Biphasic Enantioselective Epoxidation
Procedure
All reactions were performed in
a standard test-tube
[4] W. Zhang, H. Yamamoto, J. Am. Chem. Soc. 2007, 129,
286–287; N. Makita, Y. Hoshino, H. Yamamoto,
Angew. Chem. 2003, 115, 971–973; Angew. Chem. Int.
Ed. 2003, 42, 941–943.
[5] E. N. Jacobsen, in: Catalytic Asymmetric Synthesis,
(Ed.: I. Ojima), VCH, New York, 1993; T. Katsuki, in:
Catalytic Asymmetric Synthesis, (Ed.: I. Ojima), VCH,
New York, 2000; Q. H. Xia, H. Q. Ge, C. P. Ye, Z. M.
Liu, K. X. Su, Chem. Rev. 2005, 1603–1662.
[6] O. A. Wong, Y. Shi, Chem. Rev. 2008, 108, 3958–3987.
[7] W. Adam, C. R. Saha-Mçller, P. A. Ganeshpure, Chem.
Rev. 2001, 101, 3499–3548.
[8] R. W. Murray, Chem. Rev. 1989, 89, 1187–1201.
[9] F. A. Davis, A. C. Sheppard, Tetrahedron 1989, 45,
5703–5742.
[10] W. Adam, C. R. Saha-Moeller, C.-G. Zhao, Org. React.
2002, 61, 219–516; X.-Y. Wu, X. She, Y. Shi, J. Am.
Chem. Soc. 2002, 124, 8792–8793; R. Curci, L. D’Ac-
colti, C. Fusco, Acc. Chem. Res. 2006, 39, 1–9; D.
Goeddel, L. Shu, Y. Yuan, O. A. Wong, B. Wang, Y.
Shi, J. Org. Chem. 2006, 71, 1715–1717; H. Q. Tian,
X. G. She, J. X. Xu, Y. Shi, Org. Lett. 2001, 3, 1929–
1931; H. Q. Tian, X. G. She, L. H. Shu, H. W. Yu, Y.
Shi, J. Am. Chem. Soc. 2000, 122, 11551–11552; Z. X.
Wang, Y. Tu, M. Frohn, J. R. Zhang, Y. Shi, J. Am.
Chem. Soc. 1997, 119, 11224–11235; P. Barbaro, C.
Bianchini, Chemtracts 2001, 14, 274–277.
[11] G. Hanquet, X. Lusinchi, Tetrahedron Lett. 1993, 34,
5299–5302; L. Bohe, M. Lusinchi, X. Lusinchi, Tetrahe-
dron 1999, 55, 155–166; A. Gluszynska, I. Mackowska,
M. D. Rozwadowska, W. Sienniak, Tetrahedron: Asym-
metry 2004, 15, 2499–2505; R. E. delRio, B. Wang, S.
Achab, L. Bohe, Org. Lett. 2007, 9, 2265–2268.
equipped with a magnetic stirring bar. NaHCO3 (67.0 mg,
0.80 mmol, 4.0 equiv.) was dissolved in 800 mL of water.
Oxoneꢂ (132.0 mg, 0.21 mmol, 1.1 equiv.) was then added as
a solid in one portion and the solution stirred for few mi-
nutes until effervescence subsided. 500 mL of a 0.4 mol/L so-
lution of an alkene (0.20 mmol, 1.0 equiv.) in CH2Cl2 were
added and the resulting biphasic mixture was cooled to 08C
with a cryostat bath. A catalyst was added in CH2Cl2
(500 mL) in one portion followed by a solution of 18-C-6
(1.0 mg, 5.0 mmol, 2.5 mol%) in CH2Cl2 (200 mL) and the re-
sulting mixture was then vigorously stirred (very important!)
at 08C. After the indicated amount of time, the reaction
mixture was diluted with dichloromethane (10 mL), water
(10 mL) and the layers were separated. The aqueous phase
was extracted with CH2Cl2 (2ꢄ10 mL). The combined or-
ganic layers were dried (Na2SO4), filtered, concentrated
under vacuum and purified.
Crystal Structure Determination of [3b]ACTHUNRGTNEUNG[SbF6]
Cell dimensions and intensities were measured at 150 K on
a Stoe IPDS diffractometer with graphite-monochromated
MoKa radiation (l=0.71073 ꢅ). The structure was solved
by direct methods (SIR-97),[31] and all other calculations
were performed with the XTAL system[32] and ORTEP[33]
programs. The Flack parameter (x) was refined to x=
À0.03(3) and corresponds to the absolute configurations ob-
served from the syntheses (see Supporting Information for
the CD spectra of the measured crystals).
CCDC 713527 contains the supplementary crystallograph-
ic data for this paper. These data can be obtained free of
charge from The Cambridge Crystallographic Data Centre
via www.ccdc.cam.ac.uk/data_request/cif or on application to
CCDC, 12 Union Road, Cambridge CB2 1EZ, UK [fax:
(internat.) + 44 1223/336–033; e-mail: deposit@ccdc.cam.
ac.uk].
[12] G. Hanquet, X. Lusinchi, P. Milliet, C. R. Acad. Sci.,
Ser. II: Mec., Phys., Chim., Sci. Terre Univers 1991,
313, 625–628; X. Lusinchi, G. Hanquet, Tetrahedron
1997, 53, 13727–13738.
[13] Recently, it has been shown that hydrogen peroxide
can be used in some instances as an alternative to
Oxone, see P. C. B. Page, P. Parker, G. A. Rassias, B. R.
Buckley, D. Bethell, Adv. Synth. Catal. 2008, 350, 1867.
[14] L. Bohe, G. Hanquet, M. Lusinchi, X. Lusinchi, Tetra-
hedron Lett. 1993, 34, 7271–7274; P. C. B. Page, G. A.
Rassias, D. Bethell, M. B. Schilling, J. Org. Chem. 1998,
63, 2774–2777; A. Armstrong, G. Ahmed, I. Garnett,
K. Goacolou, J. S. Wailes, Tetrahedron 1999, 55, 2341–
2352; S. Minakata, A. Takemiya, K. Nakamura, I. Ryu,
M. Komatsu, Synlett 2000, 1810–1812; P. C. B. Page,
G. A. Rassias, D. Barros, D. Bethell, M. B. Schilling, J.
Chem. Soc. Perkin Trans. 1 2000, 3325–3334; P. C. B.
Page, G. A. Rassias, D. Barros, A. Ardakani, B. Buck-
ley, D. Bethell, T. A. D. Smith, A. M. Z. Slawin, J. Org.
Chem. 2001, 66, 6926–6931; M.-K. Wong, L.-M. Ho, Y.-
S. Zheng, C.-Y. Ho, D. Yang, Org. Lett. 2001, 3, 2587–
2590; P. C. B. Page, G. A. Rassias, D. Barros, A. Arda-
kani, D. Bethell, E. Merifield, Synlett 2002, 580–582;
P. C. B. Page, D. Barros, B. R. Buckley, A. Ardakani,
B. A. Marples, J. Org. Chem. 2004, 69, 3595–3597;
P. C. B. Page, D. Barros, B. R. Buckley, B. A. Marples,
Acknowledgements
We thank the Swiss National Foundation, and the Federal
Office for Education and Science for the financial support
this work. We also thank Dr. Damien Jeannerat, Andrꢀ Pinto
and Bruno Vitorge for helpful discussions and NMR experi-
ments.
References
[1] P. Besse, H. Veschambre, Tetrahedron 1994, 50, 8885–
8927; A. H. Li, L. X. Dai, V. K. Aggarwal, Chem. Rev.
1997, 97, 2341–2372; C. Bonini, G. Righi, Tetrahedron
2002, 58, 4981–5021.
[2] R. A. Johnson, K. B. Sharpless, in: Catalytic Asymmet-
ric Synthesis, (Ed.: I. Ojima), VCH, New York, 2000.
[3] N. Murase, Y. Hoshino, M. Oishi, H. Yamamoto, J.
Org. Chem. 1999, 64, 338–339; Y. Hoshino, H. Yama-
moto, J. Am. Chem. Soc. 2000, 122, 10452–10453; W.
Adv. Synth. Catal. 2009, 351, 596 – 606
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
605

Roman Novikov et al.
FULL PAPERS
Tetrahedron: Asymmetry 2005, 16, 3488–3491; P. C. B.
Page, B. R. Buckley, H. Heaney, A. J. Blacker, Org.
Lett. 2005, 7, 375–377; P. C. B. Page, B. R. Buckley, D.
Barros, A. J. Blacker, H. Heaney, B. A. Marples, Tetra-
hedron 2006, 62, 6607–6613.
[25] The slower kinetics are possibly the result of the in-
creased hindrance around the double bond generated
by the twisted phenyl groups.
[26] A. T. Balaban, C. Uncuta, A. Dinculescu, M. Elian, F.
Chiraleu, Tetrahedron Lett. 1980, 21, 1553–1556; A. R.
Katritzky, S. N. Vassilatos, M. Alajarin Ceron, Org.
Magn. Reson. 1983, 21, 587–595; A. T. Balaban, A.
Dinculescu, Tetrahedron 1984, 40, 2547–2553; A. T. Ba-
laban, M. D. Gheorghiu, T. S. Balaban, J. Mol. Struct.
1984, 114, 363–366; T. S. Balaban, M. D. Gheorghiu, C.
Roussel, A. T. Balaban, Z. Naturforsch. B 1985, 40,
1555–1557; C. Uncuta, I. Paun, C. Deleanu, M. Plaveti,
A. T. Balaban, C. Roussel, New J. Chem. 1997, 21,
1055–1065; J. I. Seeman, R. Galzerano, K. Curtis, J. C.
Schug, J. W. Viers, J. Am. Chem. Soc. 1981, 103, 5982–
5984.
[15] J. Lacour, G. Bernardinelli, V. Russel, I. Dance, Crys-
tEngComm 2002, 4, 165–170.
[16] J. Vachon, C. Pꢀrollier, D. Monchaud, C. Marsol, K. Di-
trich, J. Lacour, J. Org. Chem. 2005, 70, 5903–5911.
[17] J. Vachon, S. Rentsch, A. Martinez, C. Marsol, J.
Lacour, Org. Biomol. Chem. 2007, 5, 501–506; R. Nov-
ikov, J. Vachon, J. Lacour, Chimia 2007, 61, 236–239.
[18] V. K. Aggarwal, M. F. Wang, Chem. Commun. 1996,
191–192; P. C. B. Page, B. R. Buckley, D. Barros, A. J.
Blacker, B. A. Marples, M. R. J. Elsegood, Tetrahedron
2007, 63, 5386–5393; P. C. B. Page, M. M. Farah, B. R.
Buckley, A. J. Blacker, J. Org. Chem. 2007, 72, 4424–
4430.
[19] P. C. B. Page, R. Buckley Benjamin, A. J. Blacker, Org.
Lett. 2004, 6, 1543–1546; J. Vachon, C. Lauper, K. Di-
trich, J. Lacour, Tetrahedron: Asymmetry 2006, 17,
2334–2338.
[20] R. Novikov, G. Bernardinelli, J. Lacour, Adv. Synth.
Catal. 2008, 350, 1113–1124.
[27] B. Laleu, C. Herse, B. W. Laursen, G. Bernardinelli, J.
Lacour, J. Org. Chem. 2003, 68, 6304–6308.
[28] For this compound, the q and F angle values are
À58.88 and À61.28, respectively. These q and F angles
are quite a bit larger and lower, respectively, than
those reported for saturated analogues of type 4 (mean
average, À55.58 and À65.78, respectively). However, it
is difficult to draw a conclusion as the azepines previ-
ously studied by X-ray crystallography were derived
from the 1-phenylpropylamine and not from 3,3-dime-
thylbutan-2-amine.
[21] J. D. Wilkey, G. B. Schuster, J. Org. Chem. 1987, 52,
2117–2122.
[22] The TRISPHAT anion [tris(tetrachlorobenzenediola-
to)phosphate(V)] is the ate complex of phosphorus(V)
and made up of electron-poor catecholate units. As
such, this lipophilic anion, is amenable to a use as
counterion in asymmetric iminium-catalyzed epoxida-
tion reactions.
[29] I. Washington, K. N. Houk, J. Am. Chem. Soc. 2000,
122, 2948–2949.
[30] M. Tichy, M. Budesinsky, J. Gunterova, J. Zavada, J.
Podlaha, I. Cisarova, Tetrahedron 1999, 55, 7893–7906.
[31] A. Altomare, M. C. Burla, M. Camalli, G. L. Cascara-
no, C. Giacovazzo, A. Guagliardi, A. G. G. Moliterni,
G. Polidori, R. Spagna, J. Appl. Crystallogr. 1999, 32,
115–119.
[32] S. R. Hall, H. D. Flack, J. M. Stewart (Eds.), XTAL3.2
User:s Manual, Universities of Western Australia and
Maryland, 1992.
[23] T. W. Bell, V. J. Catalano, M. G. B. Drew, D. J. Phillips,
Chem. Eur. J. 2002, 8, 5001–5006.
[24] S. Bindya, W. T. Wong, M. A. Ashok, H. S. Yathirajan,
R. S. Rathore, Acta Crystallogr., Sect. C: Cryst. Struct.
Commun. 2007, 63, O546-O548; C. L. Klein, J. Lear, S.
Orourke, S. Williams, L. Liang, J. Pharm. Sci. 1994, 83,
1253–1256; C. L. Klein, T. A. Banks, D. Rouselle, Acta
Crystallogr., Sect. C: Cryst. Struct. Commun. 1991, 47,
1478–1480.
[33] C. K. Johnson, ORTEP II, Report ORNL-5138, Oak
Ridge National Laboratory, Oak Ridge, TN, 1976.
606
asc.wiley-vch.de
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 596 – 606