Design of a well-defined, silica-supported chiral Zn scaffold for enantioselective catalysis

Article abstract of DOI:10.1039/c003552c

New compound Zn[(S,S)-iPr-pybox](Et)₂ (1) was fully characterized, including by X-ray diffraction structural studies. Its grafting onto partially dehydroxylated silica affords material 2, which bears well-defined chiral (SiO)Zn[(S,S)-iPr-pybox]

Full text of DOI:10.1039/c003552c

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/dalton | Dalton Transactions
Design of a well-defined, silica-supported chiral Zn scaffold for
enantioselective catalysis†
Je´re´my Ternel,a,b Laurent Delevoye,a,b Francine Agbossou-Niedercorn,a,b Thierry Roisnel,c Re´gis M. Gauvin*a,b
and Christophe M. Thomas*d
Received 24th December 2009, Accepted 1st March 2010
First published as an Advance Article on the web 16th March 2010
DOI: 10.1039/c003552c
New compound Zn[(S,S)-iPr-pybox](Et)2 (1) was fully char-
acterized, including by X-ray diffraction structural stud-
ies. Its grafting onto partially dehydroxylated silica affords
Along these lines, we have designed a hybrid material bearing
well-defined, chiral zinc centers, by grafting of a new pybox
zinc alkyl complex onto a conveniently treated silica surface.
Thorough characterization of this supported catalyst is presented,
along with preliminary investigations of the reactivity of these
new homogeneous and heterogeneous chiral catalysts in aldehyde
silylcyanation.
≡
material 2, which bears well-defined chiral ( SiO)Zn[(S,S)-
iPr-pybox](Et) sites. Hybrid material 2 displays significantly
better catalytic performances than 1 in enantioselective silyl-
cyanation, thus demonstrating a beneficial grafting effect.
The 1 : 1 reaction of the tridentate pro-ligand (S,S)-iPr-pybox
with ZnEt2 in toluene at room temperature cleanly affords the
diethyl–zinc complex 1 in 90% yield (Scheme 1). Single crystals
suitable for X-ray diffraction were obtained by slow evaporation of
a saturated hexane solution at room temperature. The molecular
structure as well as selected bond lengths and angles are given
in Fig. 1. The solid-state structure of 1 features a monomeric
molecule with a five-coordinated zinc center, with the pybox ligand
Since their early discovery, organozinc compounds and their
derivatives have been explored as reagents and catalysts in numer-
ous organometallic and organic transformations.1 In particular,
the chemistry of alkyl–zinc derivatives bearing nitrogen-based
bi- or tridentate ligands has recently attracted much attention
due to their rich structural chemistry,2 and the broad scope of
their catalytic applications in fine chemicals synthesis and in
polymerization.3
Owing to their accessibility, modular nature, and applicability
in a wide range of metal-catalyzed transformations, compounds
containing a chiral oxazoline ring have become one of the most
versatile and commonly used classes of ligands for (asymmetric)
catalysis.4 An example is the chiral tridentate 2,6-bis(oxazolinyl)-
pyridine ligand (pybox), which has proved to be a valuable ligand
for a variety of asymmetric reactions.5,6
As heterogenization of catalysts is considered beneficial as it
eases products/catalyst separation and recycling,7 stereoselective
chiral supported catalysts have attracted considerable interest.8
Dramatic changes in the catalyst’s performances in terms of
activity and/or selectivity are often observed. If significant efforts
have led to some degree of understanding of these phenomena,8c
the impact of grafting on the performances of a given supported
catalyst is still difficult to rationalize. Progress in this matter
is expected from a better understanding of the structure of
the surface active sites, which can be achieved by an approach
combining controlled synthesis and thorough characterization.
Scheme 1 Synthesis of 1 and 2.
aUniversite´ Lille Nord de France, France
bUCCS, UMR CNRS 8181, Cite´ Scientifique, BP 90108, 59652,
Villeneuve d’Ascq Cedex, France. E-mail: regis-gauvin@ensc-lille.fr;
Fax: +33(0)320436585; Tel: +33(0)320436754
cCentre de Diffractome´trie X, Universite´ de Rennes 1, UMR 6226 Sciences
Chimiques de Rennes, 35042, Rennes Cedex, France
dChimie ParisTech, UMR CNRS 7223, 75005, Paris, France.
E-mail: christophe-thomas@chimie-paristech.fr; Fax: +33(0)143260061;
Tel: +33(0)144276721
† Electronic supplementary information (ESI) available: Full experimental
details, additional spectroscopic data on 1 and 2, catalytic results table and
X-ray crystallographic data of 1. CCDC reference number 765759. For
ESI and crystallographic data in CIF or other electronic format see DOI:
10.1039/c003552c
˚
Fig. 1 Solid-state structure of 1. Selected bond lengths (A) and angles
(◦): Zn(1)–C(1), 2.000(3); Zn(1)–C(3), 2.011(3); Zn(1)–N(1), 2.666(2);
Zn(1)–N(2) 2.304(2), Zn(1)–N(3), 2.517(2); C(1)–Zn(1)–C(3), 148.31(12);
N(1)–Zn(1)–N(3), 137.69(12).
3802 | Dalton Trans., 2010, 39, 3802–3804
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
coordinating in a tridentate fashion. Overall, the structure of 1
is closely related to that of other zinc pybox complexes.9 The
geometry at Zn(1) could be described as a trigonal bipyramid with
N(1) and N(3) as the axial donors, or as a square pyramid with an
apical pyridinic N(2). The latter suggestion is favored by the C(1)–
Zn(1)–C(3) angle of 148.31◦, which is significantly larger than
120◦. Furthermore, the geometric parameter t defined by Addison
and Reedijk is 0.18, which implies that the coordination geometry
is best described as a square pyramid.10 The main distortion of the
square pyramid is due to the N(1)–Zn–(1)–N(3) angle of 137.69◦,
which results from the constraints imposed by the ligand.
(4.2 ppm) and aliphatic protons (1.0 ppm). Furthermore, the
intense high-field peak comprises shoulders centered at 1.8 and
0.2 ppm. The former may originate from the iPr groups methynic
protons, while the latter is characteristic of ZnCH2 protons. The
corresponding groups give rise to signals at 1.70 and 0.67 ppm in
1 in C6D6 solution, respectively.
The 13C CPMAS NMR spectrum of 2 also confirms the
≡
formation of the chiral surface zinc complex ( SiO)Zn[(S,S)-
iPr-pybox](Et).† Noteworthy, the a- and b-carbons of the zinc–
ethyl moiety resonate at -1.8 and 9.8 ppm, respectively, in good
agreement with the corresponding values for 1, of 3.29 and
14.72 ppm. The observed shielding is rather surprising considering
that a donating alkyl ligand was substituted by an electron-
Comparison of the Zn–Noxazolyne bond lengths reveals a sig-
nificant discrepancy, as Zn(1)–N(1) is significantly longer than
˚
Zn(1)–N(3) (2.666(2) and 2.517(2) A, respectively). The existence
attracting siloxide. The pybox ligand gives rise to the expected
2
=
of diverse oxazolinyl bonding modes is reflected in the IR
spectrum of 1, which features two n(C N) bands at 1670 and
signals: most particularly, the O–C N oxazoline sp carbon
=
resonates at 160.6 ppm (161.4 ppm in 1), the oxazoline ring
CH2CH fragments at 72.5 ppm (73.14 and 71.68 ppm in 1), and
the ortho-pyridyl carbons at 143.2 ppm (145.85 ppm in 1).
1640 cm-1, the latter value being close to that of the free pro-
ligand (1639 cm-1). These observations may be indicative of some
degree of fluxionality in the pybox framework. Indeed, as 1H and
13C NMR spectra of complex 1 in C6D6 at room temperature
show one set of resonances for equivalent (on the NMR timescale)
oxazolinyl and Zn–ethyl groups, lowering the temperature caused
decoalescence of the pybox’s pyridinyl and methyl signals into two
species featuring symmetrical pybox frameworks in a 1 : 8 ratio.11†
As zinc alkyls are highly reactive towards protic reagents,
complex 1 is well-suited for immobilization by reaction with a silica
surface bearing silanol groups. In order to obtain well-defined
1H-13C CP HETCOR spectrum was recorded using the PMLG
sequence during proton evolution to partially cancel homonuclear
dipolar couplings (Fig. 2).13 1H-13C correlations are detected at
0.2/-1.8 ppm and 1.0/9.9 ppm, accounting for the zinc–ethyl
fragment’s CH2 and CH3 groups, respectively. Furthermore, a
long-distance 1H-13C correlation appears at 1.0/-1.8 (ZnCH2–
CH3 coupled pair) due to the non-selective property of the CP
transfer. The spectrum also features the expected correlations
between the ligand framework’s H/C pairs: 4.2/72.5 ppm for the
(unresolved) CH2–CH fragments of the oxazolinyl groups, and
1.3/31.7 ppm and 0.9/17.0 ppm for the isopropyl groups’ CH and
CH3, respectively. Aromatic C/H pairs are clearly observed on the
correlation spectrum obtained under different CP and decoupling
conditions: the pairs 8.4/143.2 ppm and 7.7/123.8 ppm account
for the pyridyl meta and para CH, respectively.†
◦
grafted species, we chose to use silica dehydroxylated at 800 C,
which only bears isolated silanols and which has successfully been
used for this purpose.12 This would lead to the selective formation
≡
of ( SiO)Zn[(S,S)-iPr-pybox](Et), that bears a reactive zinc–ethyl
moiety available for catalysis or further reactions, thus opening
the way to some degree of active site tuning through substitution
with electron-attracting ligands, for instance. Grafting of 1 onto
non-porous Degussa Aerosil 380 silica heated under vacuum at
800 ◦C (SiO2-800, specific area of 350 m2 g-1) in toluene proceeds
efficiently to afford after work-up a pale orange powder (2).†
Fast ethane evolution was detected in NMR scale experiments
involving 1 and suspended SiO2-800 in C6D6. Elemental analysis
shows a zinc weight content of 2.18%, which, when combined with
the C,H,N analysis results, indicates the formation of a species
featuring the expected stoichiometry (N/Zn = 2.94; C/Zn =
19.45, respective theoretical values, 3 and 19). Infrared spectrum
of the material shows full consumption of the silanol groups,
as evidenced by the disappearance of the sharp SiO–H peak at
3747 cm-1. Accordingly, new sets of signals indicating the presence
of grafted species on the surface are observed, more specifically
3
=
at 2964–2810 (n(C(sp )–H)), 1662 (n(C N) from oxazolinyl ring),
1591 and 1469–1446 cm-1 (aromatic ring vibrations). Interestingly,
=
observation of a single n(C N) band may indicate that the
dynamic process observed for 1 in solution is not operative after
immobilization. The slight shift in this band’s position compared
to those of 1 is indicative of electronic and structural modifications
of the chiral Zn–pybox fragment after grafting.
Solid-state NMR studies of 2 have been performed on a high-
field NMR spectrometer (18.8 T). 1H NMR data are in agreement
with the expected structure, and closely match those of the
molecular precursor 1.† Indeed, the pybox ligand gives rise to
signals accounting for the aromatic (8.1 ppm), oxazolinic ring
Fig. 2 Aliphatic region of the 1H-13C CP HETCOR NMR spectrum of 2.
(201.21 MHz, 10 kHz spinning speed, relaxation delay of 5 s; contact time
was set to 1.5 msec, at a radio-frequency field of 50 kHz. 1H decoupling at
85 kHz was applied during the acquisition.) The PMLG decoupling was
performed at a RF field strength of 50 kHz.
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 3802–3804 | 3803
©

View Article Online
The enantioselective addition of trimethylsilylcyanide to alde-
hydes catalyzed by chiral metal complexes is a powerful method
for the synthesis of enantiomerically pure cyanohydrins, which are
versatile synthons for the preparation of various useful synthetic
intermediates.14 If pybox aluminium complexes have been used
for this reaction,15 cheap and environmentally-benign zinc has
attracted little attention and has not met much success in this
reaction, leaving room for improvement.16
The catalytic potential of 1 and 2 in benzaldehyde silylcyanation
has been probed. Thus, using optimized conditions,† a 5% loading
of 1 converts benzaldehyde into silylated mandelonitrile with
18% enantiomeric excess, with a conversion of 20% in 20 h. In
comparison, under the same conditions, heterogeneous catalyst 2
gives rise to a significantly improved enantiomeric excess of 66%,
while conversion rises to 40% in 20 h.
Notes and references
1 D. Seyferth, Organometallics, 2001, 20, 2940.
2 B. Luo, B. E. Kucera and W. L. Gladfelter, Polyhedron, 2006, 25, 279
and references cited therein.
3 For recent examples, see: (a) Z. Zheng, G. Zhao, R. Fablet, M.
Bouyahyi, C. M. Thomas, T. Roisnel, O. Casagrande, Jr. and J.-F.
Carpentier, New J. Chem., 2008, 32, 2279; (b) B. M. Trost and C.
Mu¨ller, J. Am. Chem. Soc., 2008, 130, 2438.
4 (a) J. S. Johnson and D. A. Evans, Acc. Chem. Res., 2000, 33, 325;
(b) G. Helmchen and A. Pfaltz, Acc. Chem. Res., 2000, 33, 336;
(c) O. B. Sutcliffe and M. R. Bryce, Tetrahedron: Asymmetry, 2003,
14, 2297; (d) A. K. Ghosh, G. Bilcer and S. Fidanze, in Chemistry of
Heterocyclic Compounds, Hoboken, NJ, United States, 2004, vol. 60,
529; (e) A. Alaaeddine, A. Amgoune, C. M. Thomas, S. Dagorne, S.
Bellemin-Laponnaz and J.-F. Carpentier, Eur. J. Inorg. Chem., 2006,
3652.
5 H. Nishiyama, H. Sakaguchi, T. Nakamura, M. Horihata, M. Kondo
and K. Itoh, Organometallics, 1989, 8, 846.
6 (a) G. Desimoni, G. Faita and P. Quadrelli, Chem. Rev., 2003, 103, 3119
and references cited therein; (b) J. Jankowska and J. Mlynarski, J. Org.
Chem., 2006, 71, 1317.
7 (a) U. Kragl and T. Dwars, Trends Biotechnol., 2001, 19, 442; (b) B.
Clapham, T. S. Reger and K. D. Janda, Tetrahedron, 2001, 57, 4637;
(c) R. A. Sheldon, Chem. Ind. (London), 1992, 903.
8 (a) C. E. Song and S. Lee, Chem. Rev., 2002, 102, 3495; (b) D. Rechavi
and M. Lemaire, Chem. Rev., 2002, 102, 3467; (c) J. M. Fraile, J. I.
Garcia, C. I. Herrerias, J. A. Mayoral and E. Pires, Chem. Soc. Rev.,
2009, 38, 695.
9 (a) M. Jiang, S. Dalgarno, C. A. Kilner, M. A. Halcrow and T. P. Kee,
Polyhedron, 2001, 20, 2151; (b) F. B. Panosyan, A. J. Lough and J. Chin,
Acta Crystallogr., Sect. E: Struct. Rep. Online, 2003, 59, m864; (c) J. Li
and J. Zhao, Struct. Chem., 2006, 17, 617.
Indeed, direct anchoring of the poorly active and selective
catalyst 1 proved to be most beneficial, as the catalytic perfor-
mances of the derived material 2 outpace those of its molecular
parent compound. The rise in activity may find its origin in the
introduction of an electron-withdrawing siloxide group in the Zn
coordination sphere. As described by Fraile and Mayoral,8c the
selectivity increase may partly originate from the breakdown in
symmetry (from C2 to C1) upon immobilization. It is also worth
mentioning that immobilization prevents the exchange mechanism
observed in solutions of complex 1, and thus contributes to reduce
the number of different, potentially catalytically active species in
the reaction medium.
Further efforts will be targeted at the tuning of the metal
coordination sphere through exchange of the Zn–Et group by
anionic, electron-poor ligands, and in the extension of this
methodology to other chiral auxiliaries, in order to develop further
examples of silica-derived, zinc-containing hybrid materials for
enantioselective catalysis.
10 The t value ranges from 0 (square pyramidal) to 1 (trigonal bipyra-
midal). a and b are the angles that are opposite each other in the xy
plane (with Zn(1)–N(2) oriented along the z-axis).A. W. Addison, T. N.
Rao, J. Reedijk, J. van Rijn and G. C. Verschoor, J. Chem. Soc., Dalton
Trans., 1984, 1349.
3
2
11 The operating equilibrium is not a h -h equilibrium consisting in
reversible coordination of an oxazolinyl ring, that would afford an
unsymmetrical species. Isoelectronic Cu(I)-pybox complexes have been
shown to display complex coordination behavior. See (a) J. Diez, M. P.
Gamasa and M. Panera, Inorg. Chem., 2006, 45, 10043; (b) M. Panera,
J. Diez, I. Merino, E. Rubio and M. P. Gamasa, Inorg. Chem., 2009, 48,
11147.
Acknowledgements
12 C. Cope´ret, M. Chabanas, R. Petroff Saint-Arroman and J.-M. Basset,
Angew. Chem., Int. Ed., 2003, 42, 156.
13 E. Vinogradov, P. K. Madhu and S. Vega, Chem. Phys. Lett., 1999, 314,
443.
14 (a) M. North, D. L. Usanov and C. Young, Chem. Rev., 2008, 108,
5146; (b) M. North, Tetrahedron: Asymmetry, 2003, 14, 147.
15 I. Iovel, Y. Popelis, M. Fleisher and E. Lukevics, Tetrahedron: Asym-
metry, 1997, 8, 1279.
16 See for instance: (a) S. Gou, X. Chen, Y. Xiong and X. Feng, J. Org.
Chem., 2006, 71, 5732;(b) B. M. Trost and S. Martinez-Sanchez, Synlett,
2005, 627.
We thank the CNRS and MNESR (Ph.D. grant for J. T.) for
financial support, Bertrand Revel and Marc Bria (CCM RMN,
USTL) and Julien Tre´bosc for assistance with NMR spectroscopy,
and Bernard Hallipret for technical assistance. Nord/Pas de
Calais Region, Europe (FEDER), CNRS, French Ministry of
Science and USTL are granted for funding the Bruker 18.8 T
spectrometer. Financial support from the TGE RMN THC
Fr3050 for conducting the research is gratefully acknowledged.
3804 | Dalton Trans., 2010, 39, 3802–3804
This journal is
The Royal Society of Chemistry 2010
©