Asymmetric induction during electron transfer mediated photoreduction of carbonyl compounds: Role of zeolites

Article abstract of DOI:10.1039/b517069a

Photochemistry of 17 aryl alkyl ketones included within cation exchanged zeolites has been examined. In solution five of the 17 ketones undergo intramolecular hydrogen abstraction reaction even in the presence of a chiral amine and the rest are photoreduced to the corresponding alcohol. Within zeolites all 17 ketones yielded in presence of a chiral amine, the corresponding alcohol as the major product. When a chiral amine was used as the coadsorbent within alkali ion exchanged zeolites, enantiomerically enriched alcohol was formed in all cases. The best chiral induction was obtained with phenyl cyclohexyl ketone (enantiomeric excess: 68%). ¹H-¹³C Cross Polarization Magic Angle Spinning (CP-MAS) experiments, with a model ketone (perdeuterated acetophenone) and chiral amine (pseudoephedrine) included within MY zeolites, suggested that the cation brings the reactant and the chiral amine closer. The role of the cation in such a process is also revealed by the computation results. The results presented here highlight the importance of a supramolecular structure in forcing a closer interaction between a reactant and a chiral inductor that could be used to achieve asymmetric induction in photoproducts. The Royal Society of Chemistry 2006.

Full text of DOI:10.1039/b517069a

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Asymmetric induction during electron transfer mediated photoreduction of
carbonyl compounds: role of zeolites
b
a
a
a
J. Shailaja, Lakshmi S. Kaanumalle, Karthikeyan Sivasubramanian, Arunkumar Natarajan,
b
b
a
Keith J. Ponchot, Ajit Pradhan and V. Ramamurthy*
Received 1st December 2005, Accepted 7th February 2006
First published as an Advance Article on the web 14th March 2006
DOI: 10.1039/b517069a
Photochemistry of 17 aryl alkyl ketones included within cation exchanged zeolites has been examined.
In solution five of the 17 ketones undergo intramolecular hydrogen abstraction reaction even in the
presence of a chiral amine and the rest are photoreduced to the corresponding alcohol. Within zeolites
all 17 ketones yielded in presence of a chiral amine, the corresponding alcohol as the major product.
When a chiral amine was used as the coadsorbent within alkali ion exchanged zeolites, enantiomerically
enriched alcohol was formed in all cases. The best chiral induction was obtained with phenyl cyclohexyl
1
13
ketone (enantiomeric excess: 68%). H– C Cross Polarization Magic Angle Spinning (CP-MAS)
experiments, with a model ketone (perdeuterated acetophenone) and chiral amine (pseudoephedrine)
included within MY zeolites, suggested that the cation brings the reactant and the chiral amine closer.
The role of the cation in such a process is also revealed by the computation results. The results presented
here highlight the importance of a supramolecular structure in forcing a closer interaction between a
reactant and a chiral inductor that could be used to achieve asymmetric induction in photoproducts.
Introduction
During the last decade interest has been growing in the topic
of asymmetric induction in photochemical reactions and a few
1
remarkable methodologies have been developed. In the early days
of development, chiral solvents, circularly polarized light and
chiral sensitizers were utilized to conduct enantioselective pho-
toreactions and unfortunately they all yielded less than satisfactory
2
,3
results. Recent efforts with chiral auxiliaries, chiral sensitizers
Fig. 1 Six possible modes of distribution of two different molecules, C
and R, within zeolite supercages.
4–6
and chiral templates have led to remarkable success. Organized
media have also been used to achieve asymmetric induction in
photoreactions and of these, the crystalline state and solid host–
7
of the inductions that occur in cages of types III, IV (racemic
products) and V (enantiomerically enriched product).
8
guest assemblies have provided the most encouraging results. Our
contribution to this topic has been to establish the utility of zeolites
as media to achieve asymmetric induction. In the absence of
chiral zeolites, our approach has been to render the zeolite cages
For the above approach to be successful (to avoid products from
cages of types III and IV) every reactant molecule (R) has to be
placed next to a chiral inductor molecule (C). Recognizing the
current lack of knowledge concerning the distribution of guest
molecules within a zeolite and the highly unlikely probability
of placing every chiral inductor molecule next to a reactant
molecule in the absence of any specific interaction between them,
we have explored a strategy that would limit the photoreaction
of interest to reactant molecules next to a chiral inductor. Such
a condition eliminates the possibility of formation of the product
of interest in the cages of types III and IV. The photoreaction
we have investigated in this context is the well-known electron
transfer initiated intermolecular hydrogen abstraction reaction
of carbonyl compounds. Under the conditions employed, one
of the guest molecules (e.g., ephedrine) assumes the dual role
of a chiral inductor and an electron donor. We term such a
strategy the ‘chiral inductor as a reagent’. The goals of the current
investigation are: (a) to explore the generality of the chiral inductor
as a reagent strategy and (b) to establish that both the reactant
ketone and chiral inductor molecules are present in a single
supercage of a zeolite. For the former purpose we have investigated
9,10
‘chiral’ by absorption of an optically active molecule.
The strategy of employing chirally modified zeolite as a reaction
medium requires inclusion of two different molecules: C (a chiral
inductor) and R (a reactant), within the interior spaces of a
zeolite. This strategy by its very nature does not allow quantitative
chiral induction. When two different molecules C and R are
included within a zeolite, the distribution is expected to follow
the pattern shown in Fig. 1. The six possible modes of distribution
of guest molecules are: type I (cages containing a single C), type
II (two C’s), type III (single R), type IV (two R’s), type V (one
C and one R molecule), and type VI (empty cages). The chiral
induction obtained from the photoreaction of R is an average
11
a
Department of Chemistry, University of Miami, Coral Gables, Florida,
3124, USA. E-mail: murthy1@miami.edu; Fax: 305 284 4571; Tel: 305
84 1534
3
2
b
Department of Chemistry, Tulane University, New Orleans, LA, 70118,
USA
This journal is © The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 1561–1571 | 1561

View Article Online
the photoreduction of 17 aryl alkyl and diaryl ketones (1a–1q;
Scheme 1) within chirally modified Y zeolites. To reach the second
goal we have probed the model systems through solid state NMR
and computational methods.
filtered and the organic contents were extracted using acetonitrile.
The chiral inductor was removed from the solution by column
chromatography (silica gel/hexane–ethyl acetate) and the solution
was concentrated and analysed using chiral GC/HPLC column.
Both the reactant ketones and the product alcohols are known
compounds and were synthesized by the literature methods (see
experimental). Structures of the product alcohols were confirmed
by comparing with authentic samples.
Results
To establish the versatility of zeolites to obtain high enantiose-
lectivity, the chiral amine mediated photoreduction of 17 ketones
All 17 substrates, when irradiated within NaY zeolite in the
presence of a suitable chiral amine, gave significant amounts of
optically active alcohol (Scheme 2). The enantiomeric excesses
(ee) obtained with ketones 1a–1q included in NaY are listed in
Table 1. For the sake of brevity we have provided results only in
NaY. We have carried out experiments with several chiral amines
in each case and the results reported correspond to the chiral
amine that gave the best number. This clearly illustrates that there
is no single chiral amine that works best for all ketones. The results
(ee) obtained with various chiral inductors in the case of 1a are
summarized in Table 2. Data in Table 2 show that the extent of
chiral induction depends on the structure of the chiral amine.
Since detailed study was carried out only with ketone 1a, the
absolute configuration of the product alcohol in this case alone
(
(
1a–1q) was investigated within alkali ion exchanged Y zeolite
Scheme 2). In solution, five of these ketones gave only products
of intramolecular c-hydrogen abstraction even in the presence of
chiral amines. The remaining twelve ketones underwent photore-
duction to the corresponding alcohol in the presence of chiral
amines. No asymmetric induction was observed in solution. On
the other hand, within zeolites all ketones gave the corresponding
alcohol via an electron transfer process. Photoreduction was
attempted with numerous chiral inductors whose structures are
provided in Scheme 3.
A typical experimental procedure involved the addition of
2
1
50 mg of activated NaY zeolite to a solution of 7 mg (3.87 ×
−
5
0
moles) of 1a–1q, 26–30 mg of chiral inductor and 20 mL of
12
hexanes. The mixture was allowed to stir under nitrogen for 2–3 h
and then irradiated for 3 h. The conversions in most experiments
were in the range of 20 to 35% as monitored by GC. The zeolite was
was identified by following the literature method. In all other
cases the first peak to elute from the chiral column is arbitrarily
assigned as A and the second peak as B (Table 1). Once again
Scheme 1 List of aryl alkyl ketones investigated in this study.
Scheme 2 Photoreduction of ketones within chirally modified NaY zeolite.
562 | Org. Biomol. Chem., 2006, 4, 1561–1571
1
This journal is © The Royal Society of Chemistry 2006

View Article Online
Scheme 3 Structures of chiral inductors explored in this study.
we wish to emphasize that no detailed studies on ketones 1b–1q
were conducted and a brief study on them helped us establish the
generality of the zeolite based chiral induction strategy.
the chiral induction process and also that NaY is the best zeolite
for the present purpose.
To establish the role of a cation in bringing the reactant
ketone and the chiral amine closer we have carried out NMR
polarization experiments of solid zeolite samples. Solid state NMR
To probe the importance of alkali ions present in a zeolite in
the chiral induction process several experiments were carried out
with 1a. The following observations are noteworthy: while the
dry NaY containing optically pure norephedrine gave 68% ee,
the same zeolite complex when it was intentionally made ‘wet’ by
adsorption of water and irradiated gave <2% ee. Although the ee
was influenced by the cation, the ratio of 2a to 3a remained the
same under dry and wet conditions (for structures 2a and 3a see
1
13
experiments such as intermolecular H– C Cross Polarization
Magic Angle Spinning (CP-MAS) can provide insight into the
existence of interactions between adsorbed molecules within a
zeolite and can also provide evidence regarding the relative
1
3–16
proximity of the two molecules.
In a CP-MAS experiment,
1
the CP part of the experiment involves selective H magnetization
followed by dipolar transfer to the less abundant C nuclei under
+
13
Scheme 5). We believe that hydrated Na ions are unable to interact
with the reacting ketone and the electron donor amine. If the cation
present in a zeolite is important in aiding the necessary interaction
between the ketone and the chiral inductor, the photoreduction
and ee should depend on the nature and number of the cation. Y-
Sil, a zeolite with morphology similar to Y zeolite but with very few
cations gave negligible ee (<5%) with norephedrine as the chiral
inductor. Similarly norephedrine when adsorbed onto MCM-41,
a mesoporous zeolite, and silica gel containing no cation failed
to induce enantioselectivity in the product alcohol. Examination
of the photoreduction, with norephedrine as the chiral inductor
within various alkali ion exchanged zeolites (LiY, NaY, KY, RbY
and CsY), revealed that the nature of the cation is also important
in the process of chiral induction. For the identical loading level
of 1a (one molecule per ten supercages) and norephedrine (one
molecule per supercage), the ee within LiY, NaY, KY, RbY and
CsY were 45%, 68%, 30%, 10% and zero respectively. From the
above studies it is clear that the cation plays a crucial role during
Hartman–Hahn matching conditions. Our study was prompted
by the work of Garcia-Garibay and co-workers on benzophenone
16
13
and cyclohexane in NaY. By comparison of the C signal
1
13
intensities of benzophenone, due to H– C Cross Polarization
Magic Angle Spinning (CP-MAS) measured with deuterated and
non-deuterated benzophenone samples, Garcia-Garibay and co-
workers established that the intermolecular C–D–H–R distances
between carbon atoms of deuterated benzophenone and hydrogens
˚
of cyclohexanes have an average value of ca. 2.2 A within
NaY. This suggested that CP-MAS experiments could be useful in
establishing the presence of an interaction between the cation, the
ketone and the chiral inductor.
The CP-MAS experiments were carried out with perdeuterated
acetophenone as the ketone and pseudoephedrine as the chiral
amine. The choice of acetophenone as the model compound was
dictated by the ready availability of deuterated acetophenones such
as C
6
H
5
COCD
3
, C
6
D
5
COCH
3
and C
6
D
5
COCD . The observation
3
This journal is © The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 1561–1571 | 1563

View Article Online
a,b
Table 1 Enantiomeric excess in the alcohol products of ketones 1a–1q included in NaY zeolite in presence of a chiral inductor
c,d
Substrate
Zeolite
Chiral inductor
%Enantiomeric excess
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
a
b
b
c
d
e
e
f
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
(+)-Norephedrine
(−)-Norephedrine
(+)-Norephedrine
(+)-Pseudoephedrine
(+)-Pseudoephedrine
(+)-Ephedrine
68R
67S
50
30
25
47A
53B
30
(−)-Ephedrine
(+)-Pseudoephedrine
(+)-Ephedrine
g
h
h
i
35
(1R,2R)-(1,2-Diaminocyclohexane)
(1S,2S)-(1,2-Diaminocyclohexane)
(−)-Ephedrine
42B
44A
40B
42A
43A
43B
36B
24A
23B
28A
22A
22B
24A
34B
24A
23B
30A
27B
64B
65A
57B
58A
i
(+)-Ephedrine
j
(+)-Pseudoephedrine
(−)-Pseudoephedrine
(−)-Pseudoephedrine
(+)-Ephedrine
j
k
k
k
k
l
(−)-Ephedrine
(+)-Pseudoephedrine
(−)-Pseudoephedrine
(+)-Pseudoephedrine
(+)-Pseudoephedrine
(−)-Ephedrine
l
m
m
n
n
o
o
p
p
q
q
(+)-Pseudoephedrine
(−)-Pseudoephedrine
(+)-Pseudoephedrine
(−)-Pseudoephedrine
(−)-Ephedrine
(+)-Ephedrine
(−)-Ephedrine
(+)-Ephedrine
a
b
The numbers reported are the average of at least three independent runs. The loading level was maintained at one ketone per 10 supercages and one
c
chiral amine per cage. In the case of ketone 1a, the absolute configuration of the product alcohol has been identified. R and S refer to the isomer being
d
enhanced. In ketones 1b–1q the absolute configuration of the product alcohol was not identified. The first peak to elute from the GC or HPLC column
was noted as A and the second as B.
Table 2 Enantiomeric excess of alcohol product of ketone 1a with various
13
of C signals from perdeuterated acetophenone relies on
a,b
chirally modified NaY zeolites
1
the polarized H source and cannot be observed unless we have a
hydrogen source that is relatively rigid and is in closer proximity
to the perdeuterated acetophenone molecule. The absence of
C signals in the H– C CP-MAS spectra of perdeuterated
acetophenone adsorbed within NaY indicates that there is no
source of hydrogen present inside the zeolite from which the
polarization transfer could take place. But when H– C CP-MAS
spectra of perdeuterated acetophenone and (−)-pseudoephedrine
adsorbed within NaY were recorded, along with the signals of the
c
Medium Chiral inductor
%Enantiomeric excess
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
NaY
(+)(S,R)-Norephedrine
68R
67S
26R
54S
9S
10R
16R
11S
37R
35S
11R
26S
27R
23S
13
1
13
(−)(R,S)-Norephedrine
(+)(S)-Aminophenylethanol
(S,S)-Aminophenylpropanediol
(S)-Valinol
1
13
(R)-Valinol
(+)(S,R)-Ephedrine
(−)(R,S)-Ephedrine
1
3
(+)(S,S)-Pseudoephedrine
(−)(R,R)-Pseudoephedrine
(S)-Phenylglycinol
chiral inductor, a new C signal at d 25 ppm was observed. Based
13
on the fact that the C signal of CH
3
in acetophenone comes at
d 25 ppm, we assign the new C signal at d 25 ppm to be that of
CD , perdeuterated acetophenone. The maximum distance up to
which intermolecular polarization transfer can take place is 10 A˚
1
3
(S,S)-Diaminocyclohexane
(+)(S,R)-N-methylephedrine
(−)(R,S)-N-methylephedrine
3
(S,S)-Aminomethoxyphenylpropanol 30S
˚
and since the diameter of the supercage is 13 A, it is quite possible
that both the chiral inductor and perdeuterated acetophenone are
present in the same or adjacent supercages of NaY.
Support for the above interaction between the ketone and the
amine also comes from the contact time variation study. In this
study we assume that the polarization transfer occurs through
(S)-Phenylalaninol
(S)-Methylbenzylamine
(S)-Alaninol
13R
8S
2S
25R
(S)-Diphenylpyrrolidine methanol
a
The numbers reported are the average of at least three independent runs.
b
The loading level was maintained at one ketone per 10 supercages and one
c
chiral amine per cage. In the case of ketone 1a, the absolute configuration
of the product alcohol has been identified. R and S refer to the isomer
being enhanced.
1
13
interaction between the H of (−)-pseudoephedrine and the
C
of perdeuterated acetophenone. Cations are assumed not to play
1
13
13
a role in this process. In a H– C CP-MAS experiment, the
C
1
564 | Org. Biomol. Chem., 2006, 4, 1561–1571
This journal is © The Royal Society of Chemistry 2006

View Article Online
1
13
signals grow during H– C contact times and then decay by virtue
of their spin lattice relaxation. If the polarization transfer is taking
place between two molecules (intermolecular), the intensity of
25 ppm) changed with the relative intensity being large in LiY and
smaller in KY. Because of the poor baseline, the intensities of the
signals are not reliable and therefore no quantitative comparison
between the cations could be made. Alternative experiments of
varying the contact time to achieve the same intensity were
not conducted. However the importance of zeolite framework
1
3
the new C signal observed should increase with the increase in
1
13
contact time. As shown in Fig. 2, when H– C CP-MAS spectra of
−)-pseudoephedrine and perdeuterated acetophenone adsorbed
within NaY were recorded with varying contact times (2500 ls to
(
1
13
and cations is revealed when H– C CP-MAS spectra of (−)-
pseudoephedrine and perdeuterated acetophenone adsorbed on
silica gel were recorded. From the spectra in Fig. 3 it is clear that
the intensity of the new signal was least when silica gel was used
as a medium, suggesting that silica gel that lacks cations is not
able to interact strongly with the adsorbed organic molecules and
thus bring them closer for intermolecular polarization transfer to
occur.
1
3
5
0 ls), the intensity of the C signal due to CD at d 25 was at
3
a maximum when the contact time was 2500 ls and it decreased
with the decrease in contact time. In order to provide further
support for the above proposition of intermolecular polarization
transfer, a contact time variation study of (−)-pseudoephedrine
and acetophenone (C
6
H
5
COCH ) adsorbed within NaY was
3
performed. Unlike the perdeuterated acetophenone case, here the
proton source required for polarization transfer is present within
the acetophenone molecule itself and hence one would expect no
significant change in the relative intensity of the signal with varying
contact times used in this study. The contact time variation did
1
13
Results of H– C CP-MAS experiments confirm that there are
cages of type V (Fig. 1) within NaY zeolites. We believe that
photoreduction occurs in such cages. To provide further support to
our assumption that the cation plays an important role in the chiral
induction process we decided to probe, through computations, the
possibility of a cation forcing an interaction between the reactant
ketone 1a and norephedrine. Computations were carried out with
the Gaussian 98 suite of programs at the RB3LYP level with the
1
3
not affect the relative intensity of the C signal due to the CH of
acetophenone at d 25 in the contact time range of 50–2500 ls.
3
17
6
–31G(*) basis set. The goal of the computations is simple: can
an alkali ion force an interaction between a reactive ketone and
a chiral amine? We recognize that the computationally generated
structures that refer to the gas phase may be of limited value
in the development of a mechanistic model where the cations are
embedded within the walls of a zeolite. Although the computations
relate to free cations, we felt that the trends observed in the gas
phase would be useful in gaining an insight into the origin of chiral
induction within a zeolite.
1
13
Fig.
−)-pseudoephedrine and perdeuterated acetophenone adsorbed within
NaY zeolite. Only a narrow region corresponding to the C signal due
2 Contact time dependent H– C CP-MAS spectra for
(
13
Computations with (+)-norephedrine, phenyl cyclohexyl ketone
+
13
(1a) and Li were performed in three stages. In stage 1, the most
to CD
3 3
is shown. The arrow points to the C signal due to CD . Note
the intensity change with contact time. The other signals correspond to
−)-pseudoephedrine. (a) 2500 ls, (b) 1500 ls, (c) 500 ls, (d) 300 ls and
e) 50 ls.
computationally favored conformations of the chiral amine, (+)-
+
(
(
norephedrine, were determined. In stage 2, the Li ion was allowed
to interact with the conformers of (+)-norephedrine to determine
+
the computationally favored structures for the Li –chiral amine
The importance of the charge density of the cation in bring-
ing the ketone and chiral amine closer is evident when H–
complex. In the final stage, the ketone was allowed to interact
1
+
with the Li –chiral amine complex.
1
3
C CP-MAS spectra of perdeuterated acetophenone and (−)-
Chiral inductor molecule (+)-norephedrine on geometry opti-
+
+
+
pseudoephedrine adsorbed within Li , Na and K exchanged
Y zeolites for the same loading level of ketone and chiral inductor
were recorded. It can be inferred from Fig. 3 that, on going
mization at RB3LYP/6-31G* level, resulted in three conformers
−1
close in energy (difference in energies <1 kcal mol ). Energies
and optimized structures of the conformers are shown in Fig. 4.
The three conformers represent the three Newman projections that
would be obtained for norephedrine. In two of these, as expected,
there is hydrogen bonding between the amino and hydroxyl groups.
The structures and energies reported here are similar to the ones
recently computed at the MP2/6-311 + G** level by Simons and
+
+
13
from Li to K , the intensity of the C signal due to CD
3
(d =
18
co-workers.
+
Each conformer was independently allowed to interact with Li
+
ion, and re-optimized. During the calculation, Li ion was free
to move to find the most stable position. Seven structures were
obtained, inclusive of bidentate and monodentate interactions
1
13
Fig. 3 H– C CP-MAS spectra for (−)-pseudoephedrine and perdeuter-
+
of Li with (+)-norephedrine (Fig. 5). In the bidentate modes
ated acetophenone adsorbed within alkali ion exchanged Y zeolites and
+
13
shown in Figs. 5a–c, Li interacted with amino nitrogen and
onto silica gel. Only a narrow region corresponding to the C signal due
to CD is shown. The arrow points to the C signal due to CD . Note the
3 3
intensity change with the nature of the alkali ion present in a zeolite. Other
signals correspond to (−)-pseudoephedrine. (a) LiY, (b) NaY, (c) KY and
13
hydroxyl oxygen in a, amino nitrogen and phenyl in b, hydroxyl
oxygen and phenyl in c and in the monodentate modes shown in
Figs. 5d–g it interacted with amino nitrogen in d, hydroxyl oxygen
in e, phenyl from face-1 in f and with phenyl from face-2 in g.
(
d) silica gel.
This journal is © The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 1561–1571 | 1565

View Article Online
stable position. Calculations were repeated by placing the alkali
ion at various locations. Independent of the starting location of
the cation, structure h shown in Fig. 6 was obtained for the
+
supramolecular complex of Li , cyclohexyl phenyl ketone and
norephedrine. It is important to note in structure h that a newer
+
interaction between Li cation and the carbonyl of the cyclohexyl
Fig. 4 The three conformers of (+)(1S,2R)-norephedrine chiral inductor
optimized at the RB3LYP level with the 6-31G* basis set in Gaussian 98.
Assuming the least stable structure as zero, the relative energies
of various structures are listed in Fig. 5. Perusal of Fig. 5 reveals
that structures in which cation is bound to two functional groups
(
a–c in Fig. 5) are more stable than those possessing a single
interaction (d–g in Fig. 5). According to computation, the cation
interacting with amino nitrogen and alcohol oxygen is the most
stable (structure a in Fig. 5).
Fig.
6
Interaction between the ketone, lithium cation and
Structure a in Fig. 5 was re-optimized at the RB3LYP/6-31G*
level in the presence of cyclohexyl phenyl ketone. During the
calculation, the alkali ion was free to move to find the most
(
+)-norephedrine chiral inductor in the gaseous state as obtained
−1
by computations. The relative energy differences in kcal mol between
the given complex k and the conformers h–j are shown in parentheses.
Fig. 5 Various modes of lithium cation binding to (+)-norephedrine chiral inductor (Gaussian 98/RB3LYP/6-31G*). The relative energy differences in
kcal mol between the given conformer g and the most stable conformers a–f are shown in parentheses.
−
1
1
566 | Org. Biomol. Chem., 2006, 4, 1561–1571
This journal is © The Royal Society of Chemistry 2006

View Article Online
phenyl ketone has developed. To identify other possible structures
for the ternary complex, structures b–g in Fig. 5 were indepen-
dently optimized in the presence of cyclohexyl phenyl ketone at
the RB3LYP/6-31G* level. Three supramolecular structures (i–k
in Fig. 6) were obtained. Vibrational frequencies with no Raman
intensities were computed for all four structures and were found
to be positive. Positive frequencies suggest that all three cation
bound structures are true minima. We do not know which one of
the four structures is present within a zeolite. Independent of this
limitation the computational results unequivocally suggest that
an alkali ion can bring a reactive ketone and an electron transfer
agent, chiral amine, closer. Existence of closer interaction between
alkali ion, reactive ketone and chiral amine is consistent with
the results of CP-MAS experiments and photochemical product
studies.
about 23.5% ee in the DL-pinacol formed from acetophenone. It is
important to note that in our studies no pinacol was obtained and
in the majority of the cases the only product is the corresponding
alcohol and the ee obtained within zeolites is in the range between
22–68%. When the irradiation was carried out in solution in all
cases only racemic products were obtained.
The general mechanism of amine-mediated photoreduction of
11
carbonyl compounds is shown in Scheme 4. This process involves
three steps: first, an electron transfer from the amine to the excited
carbonyl chromophore, second, proton transfer from the radical
cation to the radical anion and third, the reaction of the radical
intermediates, yielding either a pinacol or the corresponding
alcohol. One of the important aspects of this reduction is that
it is independent of the nature of the lowest excited state (np* or
pp*) of the reacting carbonyl compound. The mechanism shown
in Scheme 4 works best for primary and secondary amines, but
tertiary amines also initiate photoreduction through an electron
transfer pathway. We believe that photoreduction of all ketones
examined in this study proceeds by the electron transfer pathway
shown in Scheme 4. Although we carried out preliminary studies
on 17 ketones, in depth investigation was pursued only on ketone
1a. As emphasized earlier, the study on the other sixteen ketones
helped us establish the generality of the zeolite based ‘chiral
inductor as a reagent’ strategy. However, we are unable to establish
any structure–asymmetric induction relationship with the ketones
investigated. The excited state behavior of 1a within zeolites
discussed in detail below helps to establish the various factors that
need to be controlled during the chiral photoreduction process.
Discussion
From the results presented above it is clear that the phenomenon of
asymmetric induction during electron transfer mediated photore-
duction within alkali ion exchanged zeolites is general. Cations
present within a zeolite play an important role in this process.
The results of solid-state NMR experiments and computational
studies suggest that alkali ions help to bring the reactant ketone
and the chiral amine closer.
Perusal of Table 1 reveals that ketones 1a, 1b, 1e, 1i, 1p
and 1q give moderate enantioselectivity (e.e) within NaY: 1a,
(
+)-norephedrine (68%); 1b, (+)-norephedrine (50%); 1e, (−)-
20,21
ephedrine (53%); 1i, (+)-ephedrine (42%); 1p, (−)-ephedrine (64%)
and 1q, (−)-ephedrine (57%). We recognize that the ee obtained is
certainly not in the range to be synthetically useful. However,
we wish to emphasize that our studies are directed towards
understanding supramolecular interactions and in this context the
results reported here are significant. One of the earliest reports of
the use of amines as chiral reagents is by Seebach and coworkers in
Phenyl cyclohexyl ketone (1a)
in isotropic solution gives
a c-hydrogen abstraction Norrish–Yang cleavage product (2a;
Scheme 5). As expected based on solution behavior, irradiation
as a hexane slurry of 1a included within NaY gave 2a as the
only product. On the other hand, irradiation of 1a included
within optically pure ephedrine, pseudoephedrine or norephedrine
modified NaY gave intermolecular hydrogen abstraction product
3a, in addition to 2a. The alcohol 3a was the product of electron
transfer from the amino group of the above chiral inductors to
19
the solution phase photopinacolization of aromatic ketones.
mixture of 1 : 5 chiral amine to pentane as solvent at −72 C gave
A
◦
Scheme 4 Mechanism of electron transfer mediated photoreduction.
Scheme 5 Inter and intramolecular photoreduction of cyclohexyl phenyl ketone.
Org. Biomol. Chem., 2006, 4, 1561–1571 | 1567
This journal is © The Royal Society of Chemistry 2006

View Article Online
the excited ketone. Absence of formation of 3a when ephedrine
hydrochloride and (−)-diethyltartarate (no amino group) were
used as chiral inductors confirms the need for a good electron
donor to initiate the hydrogen abstraction process.
If the role of the amine was to serve as the electron donor, the
ratio of inter vs. intramolecular hydrogen abstraction products
same cage. In addition one should keep in mind that the ability
to dry zeolites with various alkali ions might differ. It is especially
important to note that LiY is harder to dry than other zeolites. In
+
general we have found that, to achieve chiral induction, Na and
+
+
+
K ions are better than Rb and Cs ions.
One disappointing observation is that despite the entire reaction
occurring within chirally modified cages, the ee is not quantitative.
This, we believe, is due to the multi-step nature of the reaction
involving at least three distinct intermediates, the triplet ketone, the
radical ions and the radicals (Scheme 4). All three intermediates
possess pro-chiral faces and the factors that control the addition to
the pro-chiral faces of these intermediates are likely to be different.
Both computational and solid state NMR results relate only to
the ground state ketone and at this stage we do not have any
experimental results that provide an understanding of the nature
of interaction between the three reactive intermediates and the
chiral amines. In spite of this deficiency we have been able to make
some generalizations that are presented below.
(
2a to 3a) should depend on the electron donating ability of
the chiral inductor. The ratio of 2a : 3a is expected to be lower
when secondary amine chiral inductors such as ephedrine and
pseudoephedrine are the donors than when the chiral inductor is
a primary amine such as norephedrine. As predicted, for the same
loading level of the chiral amine (one molecule per supercage)
the ratio of 2a : 3a with norephedrine was 5 while that with
pseudoephedrine was 0.20.
As indicated above, ketone 1a included within a chirally
modified (bearing amino group) NaY zeolite undergoes two
primary photoreactions, one dependent and the other independent
of the chiral inductor. For example, 1a present in cages of types
III and IV (Scheme 1) would give only 2a whereas those present
in cages of type V are expected to yield both 2a and 3a. This
generalization predicts that the 2a : 3a ratio would depend on the
ratio of the cages that contain the chiral inductor and those that
do not, which in turn depends on the loading level of the chiral
inductor. This expectation was also realized. Upon irradiation of
We believed that to make generalizations it is important to
assign the absolute configuration of the two alcohols formed
during the photoreduction of cyclohexyl phenyl ketone and
identify the peaks A and B in the GC/HPLC traces. This was
achieved by synthesizing the optically enhanced a-cyclohexyl
12
benzyl alcohol by a known procedure. When ketone 1a was
reduced thermally by a known literature procedure using chiral
inductor (S)-diphenyl pyrrolidine methanol, enantiomerically
enriched a-cyclohexyl benzyl alcohol was obtained. About 65%
enrichment of the R enantiomer was obtained in the product
mixture which corresponded to the A isomer on the chiral GC
trace. The authenticity of the configuration of the enantiomer was
confirmed with the literature report. In Table 2 the enantiomer
that is enhanced by each chiral amine is indicated.
1a included within NaY containing pseudoephedrine the ratio
of 2a : 3a varied; at a loading level of one chiral amine per five
supercages the ratio was 1.27 whereas at a loading level of one
chiral amine per supercage the ratio was 0.20. Clearly, the amount
of intermolecular reduction product increased with the increased
loading level of pseudoephedrine. An additional important point
to note is that although the ratio of 2a to 3a was dependent on the
loading level of the chiral inductor, the ee was independent of the
loading level of the chiral inductor. This observation supported
the view that reduction occurred only in cages containing the
chiral inductor. Had there been intermolecular reduction (to yield
racemic products) in cages that do not contain the chiral inductor
the ee would have increased with increased loading levels of the
chiral inductor
On examination of the results of chiral induction with various
chiral amines a few generalizations result. The ephedrine family of
inductors comprising norephedrine, ephedrine, pseudoephedrine,
N-methylephedrine, aminophenylethanol, aminophenylpropane-
diol and aminomethoxyphenylpropanol provided an interesting
comparison. It was observed that chiral inductors with the S
configuration at the benzylic alcohol part always favored the R
enantiomer of a-cyclohexyl benzyl alcohol. This point becomes
clear when one notices that the inductor (S)-aminophenylethanol
that does not have any chiral carbon connected to the amino group
gave an ee of 26% R. We tentatively conclude that the configuration
at the benzylic alcohol portion of the chiral inductor determines
the absolute configuration of the alcohol product.
Perusal of Table 2 reveals that the probability of getting chiral
induction from chiral inductors possessing two or more functional
groups like amino and hydroxyl was greater than for those with
monofunctional groups. For example, norephedrine (68% ee)
and pseudoephedrine (37% ee) gave higher ee’s compared to a-
methylbenzylamine (8% ee). It was also seen that the phenyl group
was important in the chiral induction process. Norephedrine and
other ephedrine family inductors that possess hydroxyl, phenyl
and amino groups were very successful compared to the ones
lacking a phenyl group such as valinol and alaninol. The nature
of the amine also seems to matter: Comparing norephedrine,
ephedrine and N-methyl ephedrine (Scheme 3), it can be seen
that all three inductors have similar structural features and chiral
configurations (1S,2R), except that norephedrine is a primary
While the ee obtained in this study is the highest amongst the
19
thus far reported examples of photoreduction of achiral ketones,
unfortunately we are unable to develop a model that could be used
to predict the extent of chiral induction within a zeolite. Several
observations suggest that one must include the alkali ion in any
mechanistic understanding of the asymmetric induction process
within zeolites. The following experimental observations relating
to the measured ee in the case of 1a should be noted in this context:
(
a) the ee depends on the water content within NaY. Dry zeolites
give higher ee than wet ones, (b) the ee depends on the number
of alkali ions, (c) most importantly, the % ee depends on the
nature of the alkali ion. Thus it is clear that to obtain maximum
ee one should use a dry zeolite with moderate number of alkali
ions (preferably Y zeolite). At this stage we are unable to predict
which alkali ion would work best. Alkali ions play two roles in the
chiral induction process. They forge a closer interaction between
the ketone and the chiral amine. In addition they modulate the
+
electron donating ability of the amine. Smaller ions such as Li
force a closer interaction but they drastically reduce the ability of
+
the amine to donate electrons. Further larger ions such as Rb and
+
Cs may not allow both the ketone and the amine to occupy the
1
568 | Org. Biomol. Chem., 2006, 4, 1561–1571
This journal is © The Royal Society of Chemistry 2006

View Article Online
amine, ephedrine is a secondary amine and N-methyl ephedrine is
a tertiary amine. The ee’s obtained from these vary from 68%
to 16% to 27%, all enhancing the R(+) enantiomer in the a-
cyclohexyl benzyl alcohol. Thus there is a complex relationship
between the chiral inductor and the extent of chiral induction.
With the results on hand it is too difficult to present a model for
electron transfer mediated photoreduction within zeolites. We have
established that with the help of zeolites one can obtain significant
chiral induction in the electron transfer mediated photoreduction
of carbonyl compounds. The mechanistic details are yet to be
established.
tophenone was obtained from Cambridge Isotope Laboratories,
Inc.
Zeolite NaY (Si/Al 2.4) was obtained from PQ Corporation.
Zeolites LiY, KY, and RbY and CsY were prepared from the
NaY zeolite sample by cation exchange using appropriate nitrate
◦
solution and refluxing at 90 C for at least 3 h. A 10 mL aliquot of
a 10% nitrate solution was used for each gram of zeolite, and the
cation exchange was repeated three times. The exchanged samples
were then thoroughly washed with water to remove sulfate and
chloride ions. Exchange loading was typically between 60 and
84%.
Characterization of reactant ketones. Reactants 1a, 1b, 1c, 1e,
h, 1k, 1l, 1m, 1n and 1o are commercially available from Sigma-
Conclusions
1
Aldrich and Lancaster synthesis.
We have shown with several examples that electron transfer medi-
ated photoreduction of carbonyl compounds could be achieved
within zeolites. More importantly, the product obtained when
a chiral amine is used is enantiomerically enriched. The role of
zeolite is obvious when one recognizes that the product alcohol is
formed as a racemic mixture in solution. The best enantiomeric
excess obtained within a zeolite is 68%. This chiral induction
is far higher than that reported in solution during the electron
transfer mediated photopinacolization of acetophenone at room
temperature. The role of cations in forcing a closer interaction
between a reactant and a chiral amine is revealed by solid state
NMR and computational results. At this stage we are unable to
present a model that would predict the absolute configuration of
the optical isomer that would be enhanced. Further experiments
are needed to achieve this goal.
1
Reactant 1d. H NMR (400 MHz, CDCl
3
) d: 7.54–7.51 (m, 2H),
7
.42–7.36 (m, 3H), 2.06–1.99 (m, 9H), 1.73–1.72 (m, 6H).
GC-MS (EI): 240 (M 88), 135 (41), 105 (100), 77 (67).
+
1
Reactant 1f. H NMR (400 MHz, CDCl ) d: 1.12–2.22 (10H,
3
m), 2.92–3.24 (1H, m), 3.82 (3H, s), 6.82 (2H, d, J = 9 Hz), 7.84
(2H, d, J = 9 Hz).
+
GC-MS (EI): 218 (M 5), 135 (100), 107 (8), 92 (11), 77 (14).
+
Reactant 1g. GC-MS (EI): 256 (M 16), 201 (18), 187 (40), 173
(100), 145 (48), 83 (38).
1
Reactant 1i. H NMR (400 MHz, CDCl
3
) d: 7.8–7.6 (m, 4H),
7
3
.5–7.3 (m, 3H), 6.9–6.8 (2H, m), 3.8 (3H, s).
I
Reactant 1j. H NMR (400 MHz, CDCI
3
) d: 1.01 (t, J = 8 Hz,
H), 3.93 (q, J = 8 Hz, 2H), 6.94 (br d, J = 8.4 Hz, 1H), 7.02 (m,
J = 7.1, 0.9 Hz, 1H), 7.36–7.46 (m, 4H), 7.52 (m, J = 7.4, 1.2 Hz,
1
H), 7.78 (dd, J = 8.4, 1.1 Hz, 2H).
Experimental
1
Reactant 1p. H NMR (400 MHz, CDCl
3
) d: 1.14–1.19 (m, 2H),
2
3
7
.0–2.06 (m, 2H), 3.45 (t, 1H, J = 1.3 Hz), 3.64–3.65 (dd, 2H, J =
.6 Hz), 7.12 (m, 2H), 7.19 (m, 2H), 7.45 (m, 2H), 7.55 (m, 1H),
.95 (m, 2H)
GC-MS (EI): 248 (M , 30), 247 (3), 229 (3), 143 (21), 133 (17),
28 (92), 115 (21), 105 (100), 89(2), 77 (35).
(
a) Materials
Sigma-Aldrich samples of phenyl cyclohexyl ketone (1a),
phenyl isopropyl ketone (1b), phenyl cyclopentyl ketone
(
+
ꢀ
1c), 4 -methoxyacetophenone (1e), 2-methylbenzophenone (1k),
1
4
-methylbenzophenone (1l), 2-chlorobenzophenone (1m), 4-
chlorobenzophenone (1n) and 2-methoxyacetophenone (1o), and
-methoxybenzophenone (1h) from Lancaster synthesis were used
as received. Ketones 1d, 1f and 1g were synthesized following the
1
Reactant 1q. H NMR (400 MHz, CDCl
3
) d: 1.14–1.19 (m, 2H),
2
3
.0–2.06 (m, 2H), 3.45 (t, 1H, J = 1.3 Hz), 3.64–3.65 (dd, 2H, J =
2
.6 Hz), 7.1–7.22 (m, 6H), 7.96–8.0 (m, 2H).
+
GC-MS (EI): 266 (M , 7), 247 (1), 151 (19), 143 (20), 129 (12),
22
procedure reported in the literature. Ketone 1i was synthesized
by refluxing sodium methoxide (1 eq.) and the corresponding
hydroxybenzophenone (1 eq.) and dimethyl sulfate (excess) in
methanol for 12 h. Methanol was removed under reduced pressure,
and the reaction mixture was dissolved in ether and washed
several times with water. Products were isolated from the reaction
mixtures by silica gel column chromatography using 10% ethyl
acetate and hexane mixture as eluent. Ketone 1j was prepared by
refluxing 2-hydroxybenzophenone (1 eq.), potassium carbonate
1
28 (88), 123 (100), 115 (22), 95 (35), 89(3), 75 (5).
(b) Activation of zeolites
About 300 mg of the zeolite was placed in a silica crucible and
heated at 500 C for 12 h. The freshly activated zeolite samples
were rapidly cooled in air to ca. 50 C and then transferred into
glass sample tubes, connected to a vacuum manifold, and degassed
at 80 C under low pressure (10 Torr). We observed that such
a procedure facilitates the complete removal of water and other
adsorbents.
◦
◦
◦
−3
(
1 eq.), dibenzo-18-crown-6 (0.5 eq.) and ethyl iodide (2 eq.) in
acetonitrile for 12 h. Product was isolated by silica gel column
chromatography using 10% ethyl acetate and hexane mixture as
eluent. The procedure for the synthesis of ketones 1p and 1q was
kindly provided by J. R. Scheffer and C. Scott. The chiral inductors
used in this study were obtained from Aldrich and Fluka and used
as received.
(c) Characterization of photoproducts
Photoproducts were identified to be the corresponding alcohols
from 1a–1q. Their authenticity was confirmed by comparing with
Commercial samples (Aldrich) of acetophenones for solid state
NMR studies were distilled twice prior to use. Perdeuterated ace-
synthetic samples prepared from 1a–1q through NaBH reduction.
Spectral data are provided below.
4
This journal is © The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 1561–1571 | 1569

View Article Online
1
1
Alcohol from 1a. H NMR (400 MHz, CDCl
3
) d: 0.85–1.42
Alcohol from 1q. H NMR (400 MHz, CDCl ) d: 1.14–1.18
3
(
(
m, 6 H), 1.54–1.81 (m, 4 H), 1.89 (s, 1H), 1.92–2.03 (m, 1 H), 4.35
(m, 2H), 1.98–2.03 (m, 2H), 2.17–2.2 (m, 1H), 2.7 (m, 1H), 3.6
(m, 1H), 4.5 (d, 1H), 6.8–7.1 (m, 4H), 7.19 (m, 2H), 7.33–7.38 (m,
2H).
d, 1 H, J = 7.2 Hz), 7.22–7.36 (m, 5 H).
+
GC-MS(EI): 190 (M , 7), 107(100), 79(34).
+
GC-MS (EI): 268 (M , 23) 250 (2), 229 (5), 219 (5), 151 (13),
1
Alcohol from 1b. H NMR (400 MHz, CDCl
3
) d: 0.79 (d, 3H),
1
9
44 (16), 143 (100), 130 (62), 129 (24), 128 (48), 116 (22), 115 (50),
7 (19), 77 (16).
1
.02 (d, 3H), 1.92–1.99 (m, 2H), 4.35 (d, 1H), 7.25–7.37 (m, 5H).
+
GC-MS (EI): 150 (M , 13), 107 (100), 79 (50).
1
Alcohol from 1c. H NMR (400 MHz, CDCl
3
) d: 7.35–7.22
(d) Analysis of the enantiomeric excess of optically active alcohol
products
(
m, 5H), 4.39 (d, 1H), 2.24 (m, 1H), 1.92–1.15 (m, 8H).
GC-MS (EI): 176 (M , 8), 158 (52), 129 (32), 115 (36), 107 (100),
+
Enantiomers were resolved either with a HP 5890 series II
gas chromatograph with Supelco b-dex 350/OV-1701, b-dex 120
capillary chiral columns or in Rainin HPLC with chiralcel OD,
chiralpak AD and chiralpak AD-RH chiral columns depending
upon the benzylic alcohol products. Enantiomeric excesses (%ee)
were calculated using the formula [(area of A − area of B)/(area
of A + area of B)] × 100.
9
1 (40), 79 (56), 67 (64).
1
Alcohol from 1d. H NMR (400 MHz, CDCl
3
) d: 7.25–7.33
(
m, 5H), 4.2 (d, 1H), 1.96 (m, 3H), 1.84 (d, 1H), 1.47–1.72 (m,
1
2H).
+
GC-MS (EI): 242 (M , 4), 135 (100), 107 (13), 93 (18), 79 (30).
1
Alcohol from 1e. H NMR (400 MHz, CDCl
3
) d: 1.44 (d, J =
6
6
.3 Hz, 3H), 2.76 (br, 1H), 3.77 (s, 3H), 4.78 (q, J = 6.3 Hz, 1H),
(e) Asymmetric reduction of phenyl cyclohexyl ketone (1a)
.83–6.91 (d, J = 8.3 Hz, 2H), 7.22–7.31 (d, J = 8.3 Hz, 2H).
+
GC-MS (EI): 152 (M , 31), 137 (100), 109 (65), 94 (38), 77 (43).
Freshly distilled trimethylsilyl chloride (130 mg, 1.2 mmol) was
added to a suspension of NaBH
4
(45 mg, 1.2 mmol) in dry
1
Alcohol from 1f. H NMR (400 MHz, CDCl
3
) 7.18 (d, J =
◦
THF (5 mL). After the mixture was heated at 70 C for 1 h
and allowed to cool to room temperature, a solution of (S)-a,a-
diphenylpyrrolidinemethanol (25 mg, 0.1 mmol) in THF (2 mL)
was added. When there was no gas emitted, a solution of phenyl
cyclohexyl ketone (188 mg, 1 mmol) in THF (2 mL) was added very
slowly with a gas-tight syringe. After the addition was complete,
the mixture was hydrolyzed with 2 M HCl (5 mL) and extracted
with ether (3 × 10 mL). The combined organic layers were washed
with brine, and dried with sodium sulfate. Filtration followed by
evaporation of solvent yielded the alcohols 65% enriched in the
R isomer as determined by analysis on GC with Supelco b-dex
8
3
3
.4 Hz, 2H), 6.84 (d, J = 8.4 Hz, 2H), 4.25 (d, J = 7.6 Hz, 1H),
.77 (s, 3H), 1.99–1.97 (m, 2H), 1.75–1.72 (m, 1H), 1.63-1.54 (m,
H), 1.34–1.31 (m, 1H), 1.29–0.83 (m, 5H).
+
GC-MS (EI): 220 (M , 3), 202 (8), 137 (100), 121 (10), 109 (20),
9
1
1
4 (13), 77 (10).
+
Alcohol from 1g. GC-MS (EI): 258 (M , 4), 175 (100), 147 (8),
27 (30), 83 (92).
1
Alcohol from 1h. H NMR (400 MHz, CDCl
3
) d: 3.12 (s, br,
H), 3.80 (s, 3H), 6.05 (s, 1H), 6.89 (d, 1H, J = 8.3 Hz), 6.92–6.96
(
m, 1H), 7.20–7.33 (m, 5H), 7.37–7.41 (m, 2H).
350/OV-1701 chiral column. The configurations of the alcohols
1
were confirmed by comparing and verifying them with various
Alcohol from 1i. H NMR (400 MHz, CDCl
3
) d: 2.19 (s, br,
12
literature reports on a-cyclohexyl benzyl alcohol.
f) Sample preparation for solid-state NMR
1
7
H), 3.78 (s, 3H), 5.80 (s, 1H), 6.84–6.89 (m, 2H), 7.24–7.39 (m,
H).
(
1
Alcohol from 1k. H NMR (400 MHz, CDCl
CH
), 2.38 (s, 1H), 6.00 (s, 1H), 7.15–7.36 (m, 8H), 7.53 (d, J =
.0 Hz, 1H).
3
) d: 2.26 (s, 3H,
The preparation of samples for solid state NMR experiments
consisted of two steps. The first one involved the loading of chiral
inductors within the zeolite. About 300 mg of the activated zeolite
was added to a solution (9.0 mL hexane and 1 mL chloroform) of
about 20–25 mg of chiral inductor (ephedrine, pseudoephedrine
or norephedrine) and was stirred for 12 hours. The loading of
chiral inductor within zeolite was one molecule per supercage.
The solution was then filtered using a sintered crucible. The
zeolite/chiral inductor complex was then degassed on a vacuum
3
8
1
6
1
1
Alcohol from 1l. H NMR (400 MHz, CDCl
3
) d: 2.00 (s, br,
H), 2.32 (s, 3H), 5.76 (s, 1H), 7.05–7.38 (m, 9H).
1
Alcohol from 1m. H NMR (400 MHz, CDCl
3
) d: 2.46 (s, 1H),
.11 (s, 1H), 7.11–7.30 (m, 8H), 7.51 (d, J = 7.5 Hz, 1H).
1
Alcohol from 1n. H NMR (400 MHz, CDCl
3
) d: 2.21 (s, br,
◦
−3
H), 5.76 (s, 1H), 7.20–7.45 (m, 9H).
line at 60 C, typically for 12 hours under low pressure (10
Torr). To the degassed zeolite/chiral inductor complex, about
mg of acetophenone sample (acetophenone or perdeuterated
1
Alcohol from 1o. H NMR (400 MHz, CDCl
3
) d: 1.50 (d, 3H,
4
J = 6.6 Hz), 2.72 (s, 1H), 3.86 (s, 3H), 5.08 (m, 1H), 6.86–7.36 (m,
acetophenone) (equivalent to one molecule per four supercages of
MY zeolite) was added in a glovebag under a nitrogen atmosphere.
The sample was then transferred to a vacuum manifold and
4
H).
1
Alcohol from 1p. H NMR (400 MHz, CDCl
m, 2H), 2.0–2.06 (m, 2H), 2.17–2.22 (m, 1H), 2.7 (m, 1H), 3.6 (m,
H), 4.5 (d, 1H), 6.8–7.1 (m, 4H), 7.19 (m, 2H), 7.3–7.42 (m, 3H).
GC-MS (EI): 250 (M , 30), 230 (4), 219 (10), 204 (16), 190 (13),
76 (6), 148 (18), 143 (100), 133 (47), 130 (87), 129 (24), 128 (77),
16 (35), 115 (60), 107 (34), 105 (27), 91(10), 77 (37), 58(27).
3
) d: 1.14–1.19
◦
−3
(
1
degassed at 60 C for typically 12 h under low pressure (10 Torr).
Such a heating of the sample ensured the uniform adsorption
of acetophenone or perdeuterated acetophenone. The degassed
sample was then packed into a zirconia rotor with an O ring in a
+
1
1
1
glove bag under nitrogen atmosphere. H NMR studies of these
1
570 | Org. Biomol. Chem., 2006, 4, 1561–1571
This journal is © The Royal Society of Chemistry 2006

View Article Online
solid samples revealed that there was no adsorption of atmospheric
moisture during the course of sample preparation.
4 B. Grosch and T. Bach, Chiral Photochemistry, ed. Y. Inoue and
V. Ramamurthy, Marcell Dekker, New York, 2004.
5
Y. Inoue, H. Ikeda, M. Kaneda, T. Sumimura, S. R. L. Everitt and T.
Wada, J. Am. Chem. Soc., 2000, 122, 406.
(
g) Solid-state NMR spectra
6 J.-P. Pete and N. Hoffmann, Chiral Photochemistry, ed. Y. Inoue and
V. Ramamurthy, Marcell Dekker, New York, 2004.
7 J. R. Scheffer, Chiral Photochemistry, ed. Y. Inoue and V. Ramamurthy,
Marcell Dekker, New York, 2004.
K. Tanaka and F. Toda, in Organic Photoreactions in the Solid State,
ed. F. Toda, Kluwer, New York, 2002.
9 J. Sivaguru, A. Natarajan, L. S. Kaanumalle, J. Shailaja, S. Uppili, A.
Joy and V. Ramamurthy, Acc. Chem. Res., 2003, 36, 509.
10 V. Ramamurthy, A. Natarajan, L. S. Kaanumalle, S. Karthikeyan, J.
Sivaguru, J. Shailaja and A. Joy, Chiral Photochemistry, ed. Y. Inoue
and V. Ramamurthy, Marcell Dekker, New York, 2004.
1
3
Solid-state C NMR spectra were recorded at room temperature
with a Bruker AVANCE DSX 300 spectrometer at a C resonance
frequency of 75.47 MHz. Each acquisition typically consisted of
1
3
8
1
13
1
2000 to 16000 free induction decay scans. The H– C cross-
1
3
polarization spectra ( C CP-MAS) were obtained with a recycle
delay of 2 s using a 90 pulse width of 6.5 ls and a contact
time of 1.5 ms. All CP-MAS experiments were processed with
line broadening at 100 Hz. The spectra were referenced at room
temperature relative to TMS. All NMR figures were plotted using
Bruker XWINPlot software.
◦
11 S. G. Cohen, A. Parola and G. H. Parsons, Chem. Rev., 1973, 73, 141.
1
1
1
2 B. Jiang, Y. Feng and J. Zheng, Tetrahedron Lett., 2000, 41, 10281.
3 J. Schaefer and E. O. Stejskal, J. Am. Chem. Soc., 1976, 98, 1031.
4 J. Schaefer and E. O. Stejskal, Top. Carbon–13 NMR Spectrosc., 1979,
3
, 283.
1
1
1
5 E. O. Stejskal and J. D. Memory, High-Resolution NMR in the Solid
State, Oxford University Press, New York, 1994.
6 D. Cizmeciyan, L. B. Sonnichsen and M. A. Garcia-Garibay, J. Am.
Chem. Soc., 1997, 119, 184.
(
h) Computational details
All the computations were done using the Gaussian 98 suite of
programs with the RB3LYP method and 6-31G(d) basis set. For
all the computed structures, frequencies were calculated and were
found to be positive.
7 M. J. Frisch, G. W. Trucks, H. B. Schlegel, P. M. W. Gill, B. G. Johnson,
M. A. Robb, J. R. Cheeseman, T. Keith, G. A. Petersson, J. A. Mont-
gomery, K. Raghavachari, M. A. Al-Laham, V. G. Zakrzewski, J. V.
Ortiz, J. B. Foresman, J. Cioslowski, B. B. Stefanov, A. Nanayakkara,
M. Challacombe, C. Y. Peng, P. Y. Ayala, W. Chen, M. W. Wong, J. L.
Andres, E. S. Replogle, R. Gomperts, R. L. Martin, D. J. Fox, J. S.
Binkley, D. J. Defrees, J. Baker, J. P. Stewart, M. Head-Gordon, C.
Gonzalez and J. A. Pople, in Gaussian 98 Rev. A11, Pittsburgh, PA,
Acknowledgements
V. R. thanks the NSF for financial support (CHE-0212042).
1
998.
1
8 P. Butz, R. T. Kroemer, N. A. Macleod, E. G. Robertson and J. P.
Simons, J. Phys. Chem. A, 2001, 105, 1050–1056.
References
19 D. Seebach, H. Dorr, B. Bastani and V. Ehrig, Angew. Chem., Int. Ed.
Engl., 1969, 8, 982.
20 F. D. Lewis, R. W. Johnson and D. E. Johnnson, J. Am. Chem. Soc.,
1974, 96, 6090.
21 J. H. Stocker and D. H. Kern, J. Chem. Soc. D, 1969, 204.
22 D. E. Bergbreiter and J. M. Killough, J. Org. Chem., 1976, 41, 2750.
1
Y. Inoue, Chiral Photochemistry, ed. Y. Inoue and V. Ramamurthy,
Marcell Dekker, New York, 2004.
2
3
Y. Inoue, Chem. Rev., 1992, 92, 741.
H. Rau, Chem. Rev., 1983, 83, 535.
This journal is © The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 1561–1571 | 1571