Cyanide-free enantioselective synthesis of nitriles: Synthetic proof of a biocatalytic concept and mechanistic insights

Article abstract of DOI:10.1002/cctc.201402612

The first ( by definition ) cyanide-free enantioselective synthetic approach towards chiral α - and β -branched nitriles is reported. This process is based on a biocatalytic dehydration of racemic aldoximes by using an aldoxime dehydratase and proceeds with high conversion and excellent enantioselectivity (up to 98% ee) with water as the only side-product when starting from a racemic substrate with a high E/Z ratio. Thus, in combination with the facile generation of aldoximes through condensation of readily accessible aldehydes with hydroxylamine, this methodology offers an attractive and efficient path to chiral nitriles with excellent atom economy in aqueous solution. Furthermore, this study shows a surprising enzymatic dependency of the enantiopreference on the E or Z configuration of the aldoxime moiety. Notably, the whole stereochemical course of this enzymatic reaction has been rationalized by means of a computational study.

Full text of DOI:10.1002/cctc.201402612

CHEMCATCHEM
COMMUNICATIONS
DOI: 10.1002/cctc.201402612
Cyanide-free Enantioselective Synthesis of Nitriles:
Synthetic Proof of a Biocatalytic Concept and Mechanistic
Insights
Richard Metzner,[a, b] Seiji Okazaki,[a, c] Yasuhisa Asano,*[a, c] and Harald Grçger*[b]
The first (“by definition”) cyanide-free enantioselective synthet-
ic approach towards chiral a- and b-branched nitriles is report-
ed. This process is based on a biocatalytic dehydration of race-
mic aldoximes by using an aldoxime dehydratase and pro-
ceeds with high conversion and excellent enantioselectivity
(up to 98% ee) with water as the only side-product when start-
ing from a racemic substrate with a high E/Z ratio. Thus, in
combination with the facile generation of aldoximes through
condensation of readily accessible aldehydes with hydroxyla-
mine, this methodology offers an attractive and efficient path
to chiral nitriles with excellent atom economy in aqueous solu-
tion. Furthermore, this study shows a surprising enzymatic de-
pendency of the enantiopreference on the E or Z configuration
of the aldoxime moiety. Notably, the whole stereochemical
course of this enzymatic reaction has been rationalized by
means of a computational study.
anide in the synthesis of racemic or prochiral substrates for
subsequent resolutions[5–7] or asymmetric reactions[8–10] with
the formed nitriles, a range of asymmetric catalytic cyanations
have been developed.[11–16] Cyanide, however, is highly toxic
and there is a strong demand for developing methods that
avoid the use of cyanide in the asymmetric or preceding syn-
thetic steps for the nitrile substrates. Ideally, such alternative
methods should also be attractive for applications on the in-
dustrial scale. Notably, until now there has not been a single
asymmetric method for synthesizing a-branched chiral nitriles
that “by definition” does not rely on the use of cyanide: for ex-
isting routes cyanide is either required or conceivably used in
the substrate synthesis (if chiral, racemic, or prochiral nitriles
serve as substrates) or in the enantioselective step (through
enantioselective cyanation reactions).
Looking at Nature’s way to prepare nitriles, which notably
are found in several natural products,[17] we became inspired
to use enzymes for the enantioselective synthesis of chiral a-
branched nitriles without need for any cyanide in the whole
synthetic sequence. In the following we report the first exam-
ple of such a (bio)catalytic enantioselective process for a-
branched nitriles as a further, alternative, and complementary
method that does not require cyanide. The use of aldehydes
as starting materials that are commercially available or easily
accessible (e.g., by hydroformylation) with a broadly variable
substitution pattern represents a further advantageous key fea-
ture of this new approach. The synthetic concept is based on
the spontaneous condensation of a racemic aldehyde rac-
1 with the bulk chemical hydroxylamine[18] towards the corre-
sponding E/Z aldoxime rac-E/Z-2, which is subsequently enan-
tioselectively transformed into the desired nitrile (R) or (S)-3 in
a resolution process with an aldoxime dehydratase (Scheme 1).
Aldoxime dehydratases catalyze the elimination of water
from aldoximes under the formation of the corresponding ni-
trile. As water is not considered as waste and with a nearly
perfect atom economy, this three-step process concept is at-
tractive from an economic and environmental perspective. Al-
doxime dehydratases have been discovered, characterized, and
applied by Asano and Kato in the synthesis of a broad range
From the early days of organic chemistry, (chiral) nitriles have
been widely used as versatile intermediates in the synthesis of
carboxylic acids (by hydrolysis) and amines (by hydrogena-
tion).[1] A selected recent example underlining the importance
of chiral nitrile intermediates is the synthesis of boceprevir
(Merck) as a drug for treatment of hepatitis C.[2] Furthermore,
recently chiral a-branched nitrile motifs have increasingly been
part of the key structural framework in novel final drug mole-
cules that have already been clinically approved, such as vilda-
gliptin (Novartis) and saxagliptin (Bristol Myers Squibb).[3,4]
Thus, the search for efficient enantioselective synthetic strat-
egies towards chiral nitriles has gained tremendous signifi-
cance for organic chemists and the pharmaceutical industry.
Hitherto, the preferred introduction of a chiral nitrile moiety
into a molecule has proceeded through reactions with cyanide
(either by substitution or addition reactions). Besides using cy-
[a] R. Metzner, Dr. S. Okazaki, Prof. Dr. Y. Asano
Biotechnology Research Center and Department of Biotechnology
Toyama Prefectural University
5180 Kurokawa, Kosugi, Imizu, Toyama (Japan)
[b] R. Metzner, Prof. Dr. H. Grçger
Faculty of Chemistry
Bielefeld University
Universitꢀtsstr. 25, 33615 Bielefeld (Germany)
[c] Dr. S. Okazaki, Prof. Dr. Y. Asano
Asano Active Enzyme Molecule Project
ERATO, JST
5180 Kurokawa, Kosugi, Imizu, Toyama (Japan)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cctc.201402612.
Scheme 1. Concept for an enantioselective synthesis of a-branched chiral
nitriles by enzymatic resolution with an aldoxime dehydratase.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 3105 – 3109 3105

CHEMCATCHEM
COMMUNICATIONS
www.chemcatchem.org
of achiral and structurally less-demanding nitriles.[19] So far,
a few chiral substrates have also been studied in earlier pre-
liminary screenings, for example, (E/Z)-2-phenylpropionaldox-
ime (E/Z-2a) and (E/Z)-mandelic aldoxime. However, either no
conversion or formation of racemates was observed in these
experiments, thus indicating a non-enantioselective reaction
course.[20–22]
even at increased reaction time or catalyst amount. Interesting-
ly, taking into account the E/Z ratio of 4:1 for the racemic al-
doxime E/Z-rac-2a as a starting material, this observed conver-
sion of 60% exactly matches a complete conversion of both Z
enantiomers and one of the two E enantiomers: in other
words, this result implies an excellent enantioselectivity of al-
doxime dehydratase from Bacillus sp. OxB-1 for the resolution
of the racemic E aldoxime E-rac-2a, and a highly enantioselec-
tive nitrile formation can be expected when starting from (Z-
free) pure E racemate E-rac-2a. To verify this hypothesis, we
prepared racemic E/Z aldoximes E/Z-rac-2a with an increased
E/Z ratio and studied them as substrates in biocatalytic dehy-
dration reactions (entries 3 and 4). We were pleased to find
that, as expected, an increased ratio of 9:1 led to the nitrile (S)-
3a with a high enantioselectivity of 88% ee (entry 3). Further-
more, when starting from the nearly pure E racemate E-rac-2a
(E/Z ratio=99:1), an excellent enantiomeric excess of 98% ee
was found for the resulting nitrile (S)-3a (entry 4), underlining
the proposed high enantioselectivity of the aldoxime dehydra-
tase in the resolution of the E aldoxime racemate. To find fur-
ther support for this explanation, racemic Z-enriched aldoxime
Z-rac-2a (E/Z ratio 1:11.5) was used and, as expected, rapid
conversion of both enantiomers, corresponding to a low enan-
tioselectivity for the differentiation of Z enantiomers, was
found (entries 5 and 6). These findings are not only synthetical-
ly interesting but also remarkable from an enzymatic mecha-
nistic perspective because the enantiopreference of the
enzyme for one specific absolute configuration (here the R
configuration) depends on the isomeric structure of the aldox-
ime moiety (E or Z).
For our initial study towards an enantioselective aldoxime
dehydratase-catalyzed nitrile synthesis by dehydration we
chose rac-E/Z-2a (E/Z ratio 4:1) as an easily accessible model
substrate in combination with the aldoxime dehydratase from
Bacillus sp. OxB-1[19] as biocatalyst. This enzyme was chosen for
detailed (re)investigation owing to its availability in recombi-
nant form overexpressed in Escherichia coli and high activities
in the conversion of (Z)-phenylacetaldoxime to phenylacetoni-
trile (with up to 14.000 ULÀ1 of culture) and other achiral aro-
matic, heteroaromatic, and aliphatic substrates.[20] We were
pleased to find that this aldoxime dehydratase also showed
a high activity of 3.190 ULÀ1 (at 308C) rac-E/Z-2a, correspond-
ing to 32% relative activity compared to the reaction with E/Z-
phenylacetaldoxime (9.820 ULÀ1 at 308C) as an excellent achi-
ral substrate. In the biotransformation, a quantitative conver-
sion of rac-E/Z-2a was reached within less than 2 h (Table 1,
entry 1). Unfortunately, the resulting nitrile 3a was again a race-
mate.
However, on changing the experimental protocol and oper-
ating at a lower temperature of 8 instead of 308C, the resulting
nitrile product 3a showed for the first time an enantiomeric
excess (65% ee) with a preference for the S enantiomer[23]
(Table 1, entry 2). Notably, the conversion did not exceed 60%
As a next step we became interested if other types of sub-
strates can also be enantioselectively dehydrated by the
enzyme. For this initial study of the substrate scope we chose
completely different types of substrates, ranging from pure ali-
phatic cyclic aldoximes to heterocyclic aldoximes and aldox-
imes with a stereogenic center in the b position (Scheme 2).
After obtaining an excellent enantioselectivity for the non-
cyclic aryl and alkyl-substituted aldoxime E-rac-2a with 50%
conversion and 98% ee [Table 1, entry 4 and Scheme 2, Eq. (1)],
we chose the cyclic aldoxime E/Z-rac-2b as a challenging sub-
strate as its stereochemical differentiation owes only to the
presence of a C=C double bond in the cyclic framework. Nota-
bly, the aldoxime dehydratase also proved suitable for this al-
doxime leading to a high enantioselectivity of 97% ee (at 3%
conversion) and 83% ee (at 14% conversion) for nitrile 3b
even when starting from a low E/Z ratio of the substrate of
2.2:1 [Scheme 2, Eq. (2)]. In this case, conversion had to be
stopped at an early stage (14%) to obtain a high ee value
owing to this low E/Z ratio. Over a prolonged reaction time,
the reaction stopped at 76% (data not shown), thus indicating
that, in analogy to rac-2a, one isomer aldoxime substrate re-
mained untouched and that this biotransformation with the E
racemate also proceeded with excellent enantioselectivity. In-
terestingly, in contrast to the high enantioselective courses for
E-rac-2a (98% ee at 50% conversion) and E/Z-rac-2b (83% ee
at 14% conversion), the tetrahydrofuran-substituted aldoxime
E/Z-rac-2c was not a suitable substrate: here a negligible ee
Table 1. Whole-cell biotransformations of aldoxime dehydratase from
Bacillus sp. OxB-1 with different mixtures of E/Z isomers of rac-2.
Entry[a]
E/Z ratio of 2
T [8C]
Conversion[b] [%]
ee [%]
1
2
3
4
5
6
4:1
4:1
9:1
99:1
1:11.5
1:11.5
30
8
8
8
8
>99
60 (52)
53
0
65 (S)[c]
88 (S)[c]
98 (S)[c]
67 (R)[c]
4 (S)[c]
50
15 (13)[d]
96
8
[a] For a general protocol, see the Supporting Information; t=24 h unless
otherwise stated. [b] Isolated yields for purified products obtained after
column chromatography on silica (15% ethyl acetate in n-hexane) given
in parentheses. [c] Absolute configuration determined by comparison of
the direction for optical rotation with a literature value given in Ref. [23].
[d] Reaction stopped deliberately for verification of the absolute configu-
ration.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 3105 – 3109 3106

CHEMCATCHEM
COMMUNICATIONS
www.chemcatchem.org
Further insight into the interesting preference of the aldox-
ime dehydratase from Bacillus sp. OxB-1 for three stereoiso-
mers (Z-(R)-2a, Z-(S)-2a, and E-(S)-2a) while leaving the fourth
isomer E-(R)-2a untouched was found when studying the ki-
netic course of the dehydration reaction starting from a 9:1 E/
Z mixture of rac-2a (Figure 1). As shown in the chiral HPLC
Figure 1. Biotransformation of E/Z-rac-2a with a recombinant aldoxime de-
hydratase from Bacillus sp. OxB-1 as an E. coli-whole cell-catalyst: reference
mixture of E/Z-rac-2a (t=0) with rac-3a and reaction mixture after 0.5, 2,
and 24 h.
data, Z-(R)-2a, Z-(S)-2a, and E-(S)-2a are consumed very rapidly
whereas E-(R)-2a remains unchanged even after 24 h, thus un-
derlining an excellent enantioselectivity of the enzyme for the
dehydration process. The resulting nitrile is then formed in
a whole cell-catalyzed biotransformation with 53% conversion
and over 88% ee (Figure 1 and Table 1, entry 3). Thus, this en-
zymatic stereoselective dehydration reaction towards nitrile
(S)-3a with enantiomeric excesses of up to 98% ee represents
a combination of two types of enzymatic differentiation pro-
cesses: a diastereoselective differentiation (E vs. Z) and an
enantioselective differentiation (R vs. S).
Scheme 2. Enzymatic enantioselective dehydration of further aldoxime sub-
strates.
value was obtained independent of the conversion [Eq. (3)]. A
promising enantioselective course, however, was again found
when using the aldoxime E/Z-rac-2d bearing the stereogenic
center not in the a but in the b position. In spite of the low E/
Z ratio of the substrate (1:1), the desired nitrile E/Z-rac-3d was
obtained with 61% ee at 29% conversion [Eq. (4)]. This result
also indicated a high enantioselective course when starting
from E-enriched aldoximes. Considering that E-enrichment of
aldoximes is well known (as demonstrated in our case for 2a),
this novel biocatalytic route not only represents the first cya-
nide-free enantioselective synthesis of chiral nitriles of practical
use but also appears to have a significant substrate scope
(comprising substrates with stereogenic centers in the a or
b position, substrates with a pure aliphatic and aryl substitu-
tion pattern, and cyclic and acyclic substrates).
These results gave an insight into the kinetic picture of this
enantioselective nitrile synthesis by enzymatic dehydration,
however, an open question remained regarding the lack of
enantioselectivity in the reaction at 308C (Table 1, entry 1). In-
terestingly, we found that this was not due to an enzymatic
but a chemical effect: on testing the stability of both substrate
and product in buffered solution at this elevated temperature
(data not shown), the enzymatically non-converted E-(R) aldox-
ime isomerized rapidly to its active Z isomer, whereas the S ni-
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 3105 – 3109 3107

CHEMCATCHEM
COMMUNICATIONS
www.chemcatchem.org
trile product (S)-3a remained unchanged. As both Z isomers
were converted by the enzyme, this led to the conversion of
the enzymatically untouched E-(R)-2a by isomerization to the
Z-(R)-2a isomer and subsequent dehydration, thus explaining
the full conversion and racemic nitrile (rac-3a) formation
under these reaction conditions.
A remaining challenge was to rationalize, by means of a com-
putational study, the unusual highly enantioselective differen-
tiation for the E enantiomers combined with a very low enan-
tioselectivity for the Z enantiomers. Homology modelling of
the aldoxime dehydratase from Bacillus sp. OxB-1[19] and subse-
quent docking studies of all four isomers in the active center
of this enzyme were performed. As recent work determined
binding of the nitrogen atom in the aldoxime to the ferrous
heme essential for the conversion of an aldoxime to its ni-
trile,[24] we analyzed these FeÀN bond distances obtained from
the docking studies as a key indicator of a dehydration reac-
tion (Table 2 and Figure 2). Notably, the three isomers of 2a
Figure 2. Molecular model of the aldoxime dehydratase from Bacillus sp.
OxB-1 showing the positions of a) E-(S)-2a (c) and b) E-(R)-2a (c) in the
active site. Images were created by using CCP4mg.[25] Calculated bond dis-
tances for FeÀN and FeÀO are given in Table 2, entries 3 and 4, respectively.
In summary, we have reported the first enantioselective de-
hydration of a racemic aldoxime, which proceeds in the pres-
ence of an aldoxime dehydratase. Furthermore, this novel bio-
catalytic resolution represents the first “by definition” cyanide-
free approach towards enantiomerically enriched a-branched
nitriles, leading to the desired nitriles with high conversion
and excellent enantioselectivities of up to 98% ee with water
as the only side-product when starting from racemic aldoximes
with a high E/Z ratio. The use of commercially available or
easily accessible aldehydes as starting materials for aldoximes
represents a further advantageous key feature of this new ap-
proach. As a biocatalyst, a recombinant E. coli whole-cell cata-
lyst, overexpressing the aldoxime dehydratase from Bacillus sp.
OxB-1, is used directly. Thus, in combination with the facile
generation of aldoximes by condensation of hydroxylamine
with the corresponding racemic aldehydes, this proof-of-con-
cept gives a perspective towards an efficient, sustainable, and
broadly applicable novel synthetic pathway to chiral nitriles
with excellent atom economy in aqueous solution. Further-
more, we identified a highly surprising enzymatic dependency
of the enantiopreference on the E or Z configuration of the al-
doxime moiety. In addition, the whole stereochemical course
of this enzymatic reaction is rationalized by means of a compu-
tational study. Future work will focus on the selectivity towards
chiral Z aldoximes by molecular biological alteration of the
enzyme and process development of the biotransformations.
Combination with in situ racemization of the remaining E-(R)
aldoxime for a dynamic kinetic resolution process will also be
investigated.
Table 2. Calculated bond distances for FeÀN and FeÀO coordination of
aldoxime isomers to aldoxime dehydratase from Bacillus sp. OxB-1.
Entry
Substrate
FeÀN [ꢁ]
FeÀO [ꢁ]
1
2
3
4
5
6
(E)-phenylacetaldoxime
(Z)-phenylacetaldoxime
E-(S)-2a
E-(R)-2a
Z-(R)-2a
2.87
2.76
2.88
3.21
2.73
2.69
3.26
3.50
3.31
2.80
3.66
3.22
Z-(S)-2a
converted by the enzyme showed a FeÀN bond distance
(2.69–2.88 ꢁ) in the same range as those of the well-tolerated
achiral substrates (E) and (Z)-phenylacetaldoxime (entries 1–3,
5, and 6). However, the E-(R) isomer of 2a as the only substrate
not accepted by the enzyme had a relatively long FeÀN bond
distance of 3.21 ꢁ. Additionally, the calculated bond distance
of 2.80 ꢁ for FeÀO was significantly shorter than for all other
isomers, suggesting binding of this substrate via the oxygen of
the aldoxime. This could have owed to a steric repulsion be-
tween the methyl group of the E-(R) isomer and the porphyrin
ring of the heme. Thus, these simulations were in excellent ac-
cordance with the experimental results and the experimentally
observed stereopreference of the aldoxime dehydratase from
Bacillus sp. OxB-1 matched exactly with the calculated FeÀN
and FeÀO bond distances from the docking studies.
Acknowledgements
R.M. thanks the German Academic Exchange Service (DAAD) and
the Japan Society for the Promotion of Science (JSPS) for a stu-
dent exchange scholarship. This work was supported in part by
a Grant-in-Aid for Scientific Research (A) (23248015) from JSPS
(Y.A.) and Exploratory Research for Advanced Technology (ERATO)
of the Japan Science and Technology Agency (JST) (Y.A.).
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 3105 – 3109 3108

CHEMCATCHEM
COMMUNICATIONS
www.chemcatchem.org
[13] C. Mazet, E. N. Jacobsen, Angew. Chem. Int. Ed. 2008, 47, 1762–1765;
Angew. Chem. 2008, 120, 1786–1789.
Keywords: biocatalysis
·
diastereoselectivity
·
[14] Y. Tanaka, M. Kanai, M. Shibasaki, J. Am. Chem. Soc. 2010, 132, 8862–
8863.
enantioselectivity · nitriles · reaction mechanisms
[15] Asymmetric Strecker reaction (review): H. Grçger, Chem. Rev. 2003, 103,
2795–2827.
[1] Science of Synthesis Vol. 19, Three Carbon–Heteroatom Bonds: Nitriles,
Isocyanides, and Derivatives (Ed.: S.-I. Murahashi), Georg Thieme Verlag,
Stuttgart, 2004.
[2] T. Li, J. Liang, A. Ambrogelly, T. Brennan, G. Gloor, G. Huisman, J. La-
londe, A. Lekhal, B. Mijts, S. Muley, L. Newman, M. Tobin, G. Wong, A.
Zaks, X. Zhang, J. Am. Chem. Soc. 2012, 134, 6467–6472.
[3] F. F. Fleming, L. Yao, P. C. Ravikumar, L. Funk, B. C. Shook, J. Med. Chem.
2010, 53, 7902–7917.
[16] J. Guin, G. Varseev, B. List, J. Am. Chem. Soc. 2013, 135, 2100–2103.
[17] F. F. Fleming, Nat. Prod. Rep. 1999, 16, 597–606.
[18] Hydroxylamine is a readily available industrial bulk chemical used for
the large-scale production of e-caprolactam. Although it represents
a hazardous substance, hydroxylamine represents an attractive and
less-toxic alternative than highly toxic cyanide.
[19] Y. Asano, Y. Kato, FEMS Microbiol. Lett. 1998, 158, 185–190.
[20] Y. Kato, K. Nakamura, H. Sakiyama, S. G. Mayhew, Y. Asano, Biochemistry
2000, 39, 800–809.
[21] S.-X. Xie, Y. Kato, H. Komeda, S. Yoshida, Y. Asano, Biochemistry 2003, 42,
12056–12066.
[22] Y. Kato, S. Yoshida, S.-X. Xie, Y. Asano, J. Biosci. Bioeng. 2004, 97, 250–
259.
[23] The absolute configurations of the remaining aldoxime and product
were determined by comparing the direction of the optical rotation
with a literature value for (S)-2-phenylpropionitrile: see V. V. Thakur, A.
Sudalai, Indian J. Chem. 2005, 44B, 557–562.
[4] S. A. Savage, G. S. Jones, S. Kolotuchin, S. Ann Ramrattan, T. Vu, R. E.
Waltermire, Org. Process Res. Dev. 2009, 13, 1169–1176.
[5] M. Wieser, T. Nagasawa in: Stereoselective Biocatalysis (Ed.: R. N. Patel),
Marcel Dekker, New York, 2000, p. 461.
[6] C. Martinez, D. Yazbeck, J. Tao, Tetrahedron 2004, 60, 759–764.
[7] G. DeSantis, Z. Zhu, W. A. Greenberg, K. Wong, J. Chaplin, S. R. Hanson,
B. Farwell, L. W. Nicholson, C. L. Rand, D. P. Weiner, D. E. Robertson, M. J.
Burk, J. Am. Chem. Soc. 2002, 124, 9024–9025.
[8] E. Brenna, F. G. Gatti, A. Manfredi, D. Monti, F. Parmeggiani, Catal. Sci.
Technol. 2013, 3, 1136–1146.
[9] B. Kosjek, F. J. Fleitz, P. G. Dormer, J. T. Kuethe, P. N. Devine, Tetrahedron:
Asymmetry 2008, 19, 1403–1406.
[10] J. Choi, G. C. Fu, J. Am. Chem. Soc. 2012, 134, 9102–9105.
[11] Asymmetric olefin hydrocyanation (review): T. V. RajanBabu, A. L. Casal-
nuovo in: Comprehensive Asymmetric Catalysis, Vol. 1 (Eds.: E. Jacobsen,
A. Pfaltz, H. Yamamoto), Springer, Heidelberg, 1999, p. 267–378.
[12] T. Mita, K. Sasaki, M. Kanai, M. Shibasaki, J. Am. Chem. Soc. 2005, 127,
514–515.
[24] R.-Z. Liao, W. Thiel, J. Phys. Chem. B 2012, 116, 9396–9408.
[25] S. McNicholas, E. Potterton, K. S. Wilson, M. E. M. Noble, Acta Crystallogr.
Sect. D 2011, 67, 386–394.
Received: August 4, 2014
Published online on October 8, 2014
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 3105 – 3109 3109