The asymmetric addition of trimethylsilyl cyanide to aldehydes catalyzed by chiral (salen)titanium complexes

Article abstract of DOI:10.1021/ja984197v

The use of chiral (salen)TiCl₂ complexes to induce the asymmetric addition of trimethylsilyl cyanide to aldehydes has been investigated. The complexes are catalytically active at substrate-to-catalyst ratios as high as 1000:1, and the optimal c

Full text of DOI:10.1021/ja984197v

3968
J. Am. Chem. Soc. 1999, 121, 3968-3973
The Asymmetric Addition of Trimethylsilyl Cyanide to Aldehydes
Catalyzed by Chiral (Salen)Titanium Complexes
Yuri N. Belokon’ ,*,‡ Susana Caveda-Cepas,† Brendan Green,† Nicolai S. Ikonnikov,‡
Viktor N. Khrustalev,‡ Vladimir S. Larichev,‡ Margarita A. Moscalenko,‡ Michael North,*,†
Charles Orizu,† Vitali I. Tararov,‡ Michela Tasinazzo,† Galina I. Timofeeva,‡ and
Lidia V. Yashkina‡
Contribution from the Department of Chemistry, UniVersity of Wales, Bangor, Gwynedd,
LL57 2UW, UK, and A. N. NesmeyanoV Institute of Organoelement Compounds,
Russian Academy of Sciences, 117813 VaViloV 28, Moscow, Russia
ReceiVed December 7, 1998
Abstract: The use of chiral (salen)TiCl2 complexes to induce the asymmetric addition of trimethylsilyl cyanide
to aldehydes has been investigated. The complexes are catalytically active at substrate-to-catalyst ratios as
high as 1000:1, and the optimal catalyst (2e) which is derived from (R,R)-1,2-diaminocyclohexane and 3,5-
di-tert-butyl-2-hydroxybenzaldehyde produces trimethylsilyl ethers of cyanohydrins with up to 90% enantiomeric
excess at ambient temperature. Water plays a key role in these reactions since under strictly anhydrous conditions
much lower enantiomeric excesses are produced. The role of water has been shown to be to generate dimeric
complexes of the form [(salen)Ti(µ-O)]2 (4) which are the real catalyst precursors. A structure for one of these
complexes (4a derived from (R,R)-1,2-diaminocyclohexane and 2-hydroxybenzaldehyde) has been determined
by X-ray crystallography. The dimeric complexes are more active than the dichloride precursors, and at substrate-
to-catalyst ratios between 100 and 1000:1 give cyanohydrin trimethylsilyl ethers with up to 92% enantiomeric
excess in less than 1 h at ambient temperature.
Introduction
and difficult undertaking, and attempts to modify the structures
of the synthetic peptides to improve their substrate tolerance
have been unsuccessful.5 In contrast, the structural modification
of chiral transition metal complexes is a straightforward
undertaking, thus offering the potential for this class of
asymmetric catalysts to generate any desired cyanohydrin with
high enantiomeric excess.
Previous work on transition metal induced asymmetric
cyanohydrin synthesis has largely concentrated on titanium
complexes, though scattered reports of the use of boron,6
aluminum,7 tin,8 magnesium,9 rhenium,10 bismuth,11 and lan-
thanide salts12 have appeared. The chiral ligands which have
previously been utilized in conjunction with titanium include
the following: taddol,13 binol,7,14 tartrate esters,15 sulfoximines,16
and salicylimines.17 In each case, complexation of the ligand
to a suitable titanium salt generated a chiral complex that
Cyanohydrins are highly versatile synthetic intermediates,
which can easily be converted into a wide variety of important
synthetic intermediates including R-hydroxy acids, R-amino
acids, and â-amino alcohols.1 They are also components of
commercially important compounds such as the pyrethroid
insecticides cypermethrin and fluvalinate.2 In view of their
importance, there is currently considerable interest in the
asymmetric synthesis of cyanohydrins, especially by methods
that utilize a chiral catalyst.3 A number of catalysts for the
asymmetric addition of cyanide to aldehydes have been reported,
the most useful of which are enzymes,1,3 synthetic peptides,3,4
and chiral transition metal complexes.3 While impressive
enantiomeric excesses have been obtained in some cases for
the enzyme- and peptide-catalyzed addition of HCN to alde-
hydes, in other cases much lower enantiomeric excesses and/or
poor chemical yields were obtained.3 The structural modification
of enzymes to increase their substrate compatibility is a long
(5) Noe, C. R.; Weigand, A.; Pirker, S. Monatsh. Chem. 1996, 127, 1081.
Noe, C. R.; Weigand, A.; Pirker, S.; Liepert, P. Monatsh. Chem. 1997,
128, 301. Thoen, J. C.; Lipton, M. A. Tetrahedron: Asymmetry 1997, 8,
3947. Hulst, R.; Broxterman, Q. B.; Kamphuis, J.; Formaggio, F.; Crisma,
M.; Toniolo, C.; Kellogg, R. M. Tetrahedron: Asymmetry 1997, 8, 1987.
(6) Reetz, M. T.; Kunisch, F.; Heitmann, P. Tetrahedron Lett. 1986, 27,
4721.
(7) Reetz, M. T.; Kyung, S.-H.; Bolm, C.; Zierke, T. Chem. Ind. 1986,
824. Ohno, H.; Nitta, H.; Tanaka, K.; Mori, A.; Inoue, S. J. Org. Chem.
1992, 57, 6778. Mori, A.; Ohno, H.; Nitta, H.; Tanaka, K.; Inoue, S. Synlett
1991, 563. Kaptein, B.; Monaco, V.; Broxterman, Q. B.; Schoemaker, H.
E.; Kamphuis, J. Recl. TraV. Chim. Pays-Bas 1995, 114, 231. Iovel, I.;
Popelis, Y.; Fleisher, M.; Lukevics, E. Tetrahedron: Asymmetry 1997, 8,
1279.
† University of Wales. E-mail: m.north@bangor.ac.uk.
‡ Russian Academy of Sciences. E-mail: yubel@ineos.ac.ru.
(1) Griengl, H.; Hickel, A.; Johnson, D. V.; Kratky, C.; Schmidt, M.;
Schwab, H. Chem. Commun. 1997, 1933 and references therein. Kruse, C.
G. In Chirality in Industry; Collins A. N., Sheldrake, G. N., Crosby, J.,
Eds.; Wiley: Chichester, 1992; Chapter 14.
(2) Shan, G. M.; Hammer, R. P.; Ottea, J. A. J. Agric. Food Chem. 1997,
45, 4466. Galin, F. Z.; Rakhimov, R. G.; Tolstikov, G. A. Russ. Chem.
Bull. 1998, 47, 530 and references therein.
(3) For reviews see: North, M. Synlett 1993, 807. Effenberger F. Angew.
Chem., Int. Ed. Engl. 1994, 33, 1555. North, M. ComprehensiVe Organic
Functional Group Transformations; Katritzky, A. R., Meth-Cohn, O., Rees,
C. W., Pattenden, G., Eds.; Pergamon Press: Oxford, 1995; Vol. 3, Chapter
18.
(8) Kobayashi, S.; Tsuchiya, Y.; Mukaiyama, T. Chem. Lett. 1991, 541.
(9) Corey, E. J.; Wang, Z. Tetrahedron Lett. 1993, 34, 4001.
(10) Garner, C. M.; Fernandez, J. M.; Gladysz, J. A. Tetrahedron Lett.
1989, 30, 3931.
(4) Oku, J.-I.; Ito, N.; Inoue, S. Makromol. Chem. 1979, 180, 1089. Mori,
A.; Ikeda, Y.; Kinoshita, K.; Inoue, S. Chem. Lett. 1989, 2119.
(11) Wada, M.; Takahashi, T.; Domae, T.; Fukuma, T.; Miyoshi, N.;
Smith, K. Tetrahedron: Asymmetry 1997, 8, 3939.
10.1021/ja984197v CCC: $18.00 © 1999 American Chemical Society
Published on Web 04/08/1999

Asymmetric Addition of Trimethylsilyl Cyanide to Aldehydes
J. Am. Chem. Soc., Vol. 121, No. 16, 1999 3969
induced the asymmetric addition of hydrogen and/or trimeth-
ylsilyl cyanide to aldehydes. However, all of these complexes
suffer from one or more of the following limitations: need to
use large amounts (10-100 mol %) of the chiral complex; need
to carry out reactions at low temperatures (-80 °C); poor
enantiomeric excesses and/or chemical yields with many alde-
hydes. These characteristics have so far prevented the industrial
exploitation of transition metal based catalysts for asymmetric
cyanohydrin synthesis.
use of complexes 2b,d-g are reported, along with further studies
culminating in the synthesis of an even more active catalyst
for this reaction. The use of similar titanium salen complexes
to catalyze the asymmetric addition of trimethylsilyl cyanide
to aldehydes has also been reported,23 and very recently Jacobsen
has reported the use of chiral aluminum salen complexes to
induce the asymmetric addition of hydrogen cyanide to imines.24
Inspired by the pioneering work of Jacobsen18 and Katsuki19
on asymmetric epoxidation using chiral manganese salen com-
plexes,20 we initiated a research program aimed at the develop-
ment of chiral titanium salen complexes as catalysts for the
asymmetric addition of cyanide to aldehydes. Early results21
using the readily available cyclohexane-1,2-diamine derived
ligands 1a-e and titanium tetraisopropoxide demonstrated the
feasibility of our approach. However, the resulting catalysts were
subject to many of the limitations that restricted the applicability
of earlier catalysts, namely the need to use 20 mol % of the
catalyst and the need to carry out reactions at -80 °C to obtain
high enantiomeric excesses. In a recent preliminary communica-
tion,22 we have reported the use of complex 2e derived from
ligand 1e and titanium tetrachloride as a highly active (only
0.1 mol % of catalyst required) catalyst for the asymmetric
addition of trimethylsilyl cyanide to benzaldehyde at ambient
temperature. In this paper, full details of the preparation and
(12) Abiko, A.; Wang, G.-q. J. Org. Chem. 1996, 61, 2264. Yang, W.
B.; Fang, J. M. J. Org. Chem. 1998, 63, 1356. Qian, C.; Zhu, C.; Huang,
T. J. Chem. Soc., Perkin Trans. 1 1998, 2131. Abiko, A.; Wang, G.-q.
Tetrahedron 1998, 54, 11405.
(13) Minamikawa, H.; Hayakawa, S.; Yamada, T.; Iwasawa, N.; Narasa-
ka, K. Bull. Chem. Soc. Jpn. 1988, 61, 4379. Narasaka, K.; Yamada, T.;
Minamikawa, H. Chem. Lett. 1987, 2073.
(14) Mori, M.; Imma, H.; Nakai, T. Tetrahedron Lett. 1997, 38, 6229.
(15) Hayashi, M.; Matsuda, T.; Oguni, N. J. Chem. Soc., Chem. Commun.
1990, 1364. Hayashi, M.; Matsuda, T.; Oguni, N. J. Chem. Soc., Perkin
Trans. 1 1992, 3135.
Results
(16) Bolm, C.; Mu¨ller, P. Tetrahedron Lett. 1995, 36, 1625. Bolm, C.;
Mu¨ller, P.; Harms, K. Acta Chem. Scand. 1996, 50, 305.
To optimize the catalytic activity of complex 2, the effect of
varying each aspect of its structure was investigated, starting
with the substituents attached to the salen ligands. Catalysts
2b,d-g were readily prepared by treating equimolar quantities
of the appropriate ligand 1b,d-g with titanium tetrachloride in
dichloromethane at room temperature. In each case, the titanium
complex 2 was isolated as a crystalline solid that was found by
NMR studies to exist as a single species in chloroform solution.
Catalyst 2e was also analyzed by X-ray crystallography,22 and
the complex was determined to have a C2 symmetric structure
with the two chlorine atoms trans to one another.25 The NMR
studies indicated that this C2 symmetric structure was also
present in solution. These results are in marked contrast to our
earlier work with titanium tetraisopropoxide derivatives, where
the catalyst had to be prepared in situ,21 and attempts to analyze
the complex showed that a mixture of at least three species was
present. The complex derived from ligand 1a was also prepared;
however, this compound was found to be highly insoluble in
organic solvents which has prevented its purification and
characterization.
(17) Hayashi, M.; Miyamoto, Y.; Inoue, T.; Oguni, N. J. Org. Chem.
1993, 58, 1515. Hayashi, M.; Inoue, T.; Miyamoto, Y.; Oguni, N.
Tetrahedron 1994, 50, 4385. Hayashi, M.; Miyamoto, Y.; Inoue, T.; Oguni,
N. J. Chem. Soc., Chem. Commun. 1991, 1752. Jiang, Y.; Zhou, X.; Hu,
W.; Wu, L.; Mi, A. Tetrahedron: Asymmetry 1995, 6, 405. Jiang, Y.; Zhou,
X.; Hu, W.; Li, Z.; Mi, A. Tetrahedron: Asymmetry 1995, 6, 2915. Mori,
A.; Nitta, H.; Kudo, M.; Inoue, S. Tetrahedron Lett. 1991, 32, 4333. Nitta,
H.; Yu, D.; Kudo, M.; Mori, A.; Inoue, S. J. Am. Chem. Soc. 1992, 114,
7969.
(18) Zhang, W.; Leobach, J. L.; Wilson, S. R.; Jacobsen, E. N. J. Am.
Chem. Soc. 1990, 112, 2801. Zhang, W.; Jacobsen, E. N. J. Org. Chem.
1991, 56, 2296. Jacobsen, E. N.; Zhang, W.; Muci, A. R.; Ecker, J. R.;
Deng, L. J. Am. Chem. Soc. 1991, 113, 7063. Jacobsen, E. N.; Zhang, W.;
Guler, M. L. J. Am. Chem. Soc. 1991, 113, 6703. Deng, L.; Jacobsen, E.
N. J. Org. Chem. 1992, 57, 4320. Chang, S.; Lee, N. H.; Jacobsen, E. N.
J. Org. Chem. 1993, 58, 6939. Brandes, B.; Jacobsen, E. N. J. Org. Chem.
1994, 59, 4378. Jacobsen, E. N.; Deng, L.; Furukawa, Y.; Martinez, L. E.
Tetrahedron 1994, 50, 4323. Chang, S.; Galvin, J. M.; Jacobsen, E. N. J.
Am. Chem. Soc. 1994, 116, 6937. Jacobsen, E. N.; Larrow, J. F. J. Am.
Chem. Soc. 1994, 116, 12129. Pospisil, P. J.; Carsten, D. H.; Jacobsen, E.
N. Chem. Eur. J. 1996, 2, 974.
(19) Irie, R.; Noda, K.; Ito, Y.; Matsumoto, N.; Katsuki, T. Tetrahedron
Lett. 1990, 31, 7345. Irie, R.; Noda, K.; Ito, Y.; Katsuki, T. Tetrahedron
Lett. 1991, 32, 1055. Irie, R.; Noda, K.; Ito, Y.; Matsumoto, N.; Katsuki,
T. Tetrahedron: Asymmetry 1991, 2, 481. Irie, R.; Ito, Y.; Katsuki, T. Synlett
1991, 265. Hashihayata, T.; Ito, Y.; Katsuki, T. Synlett 1996, 1079.
(20) For reviews of this work, see: Katsuki, T. Coord. Chem. ReV. 1995,
140, 189. Katsuki, T. J. Mol. Catal. A: Chem. 1996, 113, 87. Dalton, C.
T.; Ryan, K. M.; Wall, V. M.; Bousquet, C.; Gilheany, D. G. Top. Catal.
1998, 5, 75.
(21) Belokon’, Y.; Ikonnikov, N.; Moscalenko, M.; North, M.; Orlova,
S.; Tararov, V.; Yashkina, L. Tetrahedron: Asymmetry 1996, 7, 851.
Belokon’, Y.; Flego, M.; Ikonnikov, N.; Moscalenko, M.; North, M.; Orizu,
C.; Tararov, V.; Tasinazzo, M. J. Chem. Soc., Perkin Trans. 1 1997, 1293.
(22) Tararov, V. I.; Hibbs, D. E.; Hursthouse, M. B.; Ikonnikov, N. S.;
Malik, K. M. A.; North, M.; Orizu, C.; Belokon’, Y. N. Chem. Commun.
1998, 387.
(23) Pan, W.; Feng, X.; Gong, L.; Hu, W.; Li, Z.; Mi, A.; Jiang, Y.
Synlett 1996, 337. Jiang, Y.; Gong, L.; Feng, X.; Hu, W.; Pan, W.; Li, Z.;
Mi, A. Tetrahedron 1997, 53, 14327. Belokon’, Y. N.; Yashkina, L. V.;
Moscalenko, M. A.; Chesnokov, A. A.; Kublitsky, V. S.; Ikonnikov, N. S.;
Orlova, S. A.; Tararov, V. I.; North, M. Russ. Chem. Bull. 1997, 46, 1936.
Belokon’, Y.; Moscalenko, M.; Ikonnikov, N.; Yashkina, L.; Antonov, D.;
Vorontsov, E.; Rozenberg, V. Tetrahedron: Asymmetry 1997, 8, 3245.
Hwang, C.-D.; Hwang, D.-R.; Uang, B.-J. J. Org. Chem. 1998, 63, 6762.
(24) Sigman, M. S.; Jacobsen, E. N. J. Am. Chem. Soc. 1998, 120, 5315.
(25) After this manuscript was submitted, a report of the synthesis and
structure of a range of achiral titanium salen complexes was published:
Chen, H.; White, P. S.; Gagne´, M. R. Organometallics 1998, 17, 5358.

3970 J. Am. Chem. Soc., Vol. 121, No. 16, 1999
Belokon’ et al.
Table 3. Asymmetric Addition of Trimethylsilyl Cyanide to
Scheme 1
Aldehydes Catalyzed by Catalyst 2ea
RCHO, R )
ee
RCHO, R )
ee
Ph
86
62
74
72
72
78
84
2,4-(MeO)2C6H3
3,4-(MeO)2C6H3
3,5-(MeO)2C6H3
4-F3CC6H4
4-O2NC6H4
Me3C
86
80
84
50
30
46
44
2-MeC6H4
3-MeC6H4
4-MeC6H4
2-MeOC6H4
3-MeOC6H4
4-MeOC6H4
Table 1. Effect of Varying the Ligand Structure on the Activity of
Catalysts 2a
enantiomeric excess
(absolute configuration)
ligandb
MeCH2
1b
1d
1e
1f
63 (S)
67 (S)
86 (S)
58 (S)
21 (S)
a All reactions were carried out using 0.1 mol % of the catalyst for
24 h at ambient temperature.
carried out at -80 °C using catalyst prepared in situ from
titanium tetraisoproxide.21 However, no nonlinear effect26 was
observed, a result that is consistent with either a monomeric or
oligomeric active species. Attempts to modify the metal ion have
so far proved unsuccessful as the corresponding Sn(II), Sn(IV),
Mn(III), and Zr(IV) complexes gave compound 3 with less than
25% ee. The use of hydrogen cyanide as an alternative to
trimethylsilyl cyanide was also investigated, but no reaction
occurred.
Having optimized the catalyst structure and reaction condi-
tions, the asymmetric addition of trimethylsilyl cyanide to a
range of aromatic and aliphatic aldehydes catalyzed by 0.1 mol
% of catalyst 2e was investigated. The results shown in Table
3 indicate that good enantiomeric excesses are obtained with
electron-rich aromatic aldehydes, while electron-deficient aro-
matic aldehydes give lower enantiomeric excesses. There
appears to be a relationship between the steric properties of the
aromatic aldehyde and the ee of the product in these reactions
as ortho-substituted aromatic aldehydes generally gave lower
ee values than the meta and para isomers. The two aliphatic
aldehydes that were investigated gave cyanohydrin silyl ethers
with moderate ee values and did not appear to be sensitive to
steric effects.
While carrying out these experiments, occasionally incon-
sistent results were obtained. In particular, occasionally much
lower conversions and ee values were observed. The cause of
this problem was eventually traced to the effect of water on the
catalysts. Experiments showed that under rigorously anhydrous
conditions, neither the dichloride complexes 2 nor the in situ
prepared isopropoxide complexes reported earlier21 are effective
catalysts. In the case of complexes 2, it was also found
advantageous to add triethylamine to the reaction mixture to
scavenge any HCl that was generated. Representative results
for both catalyst systems are shown in Table 4. The reaction
between complexes 2 and water also explains the increase in
ee that was observed as the concentration of catalyst 2e was
decreased as discussed earlier.
1g
a All reactions were carried out using 0.1 mol % of the catalyst for
24 h at ambient temperature. b The unpurified complex derived from
ligand 1a gave 18% ee (S).
Table 2. Effect of Varying the Amount of Catalyst 2e on the
Enantiomeric Excess of 3a
mol % of catalyst
enantiomeric excess of 3
conversion
12
3
1
66
68
78
82
86
86
100
100
100
100
100
80
0.5
0.1
0.01
a All reactions were carried out for 24 h at ambient temperature.
Each catalyst was tested for its ability to catalyze the
asymmetric addition of trimethylsilyl cyanide to benzaldehyde
at ambient temperature as shown in Scheme 1. As the results
shown in Table 1 illustrate, the optimal catalyst was found to
be compound 2e, which is derived from the ligand-bearing tBu
groups in the 3- and 5-positions of both aromatic rings. The
enantiomeric excess of silyl ether 3 was determined by chiral
GC analysis. The effect of varying the halogen ligands attached
to the titanium atom was also investigated by the preparation
of the titanium dibromide and titanium difluoride complexes
of ligand 1e. The difluoride complex gave similar results to
complex 2e; however, the dibromide complex proved to be
rather unstable and gave inferior results to complex 2e.
Having determined that structure 2e was the optimal catalyst,
the effect of varying the catalyst-to-substrate ratio was inves-
tigated. As Table 2 shows, the enantiomeric excess of silyl ether
3 increases as the amount of catalyst used decreases. The optimal
amount of catalyst was found to be 0.1 mol % with lower
amounts giving the same enantiomeric excess (86%) but lower
conversions of benzaldehyde to silyl ether 3. Since the amount
of solvent used for these reactions was kept constant, the most
likely explanation for this unusual relationship is that catalyst
2e reacts with some species present in the dichloromethane
solvent to produce a more reactive catalyst. Reducing the amount
of catalyst 2e will increase the relative concentration of this
more active catalyst present in the reaction mixture. The cause
of this effect is discussed later in this paper.
This beneficial effect of water also provides an explanation
for another effect that had been observed when using the
titanium tetraisopropoxide catalyst system at -80 °C. Under
these conditions, the asymmetric addition of trimethylsilyl
(26) Oguni, N.; Matsuda, T.; Kaneko, T. J. Am. Chem. Soc. 1988, 110,
7877. Kitamura, M.; Okada, S.; Suga, S.; Noyori, R. J. Am. Chem. Soc.
1989, 111, 4028. Puchot, C.; Samuel, O.; Dunach, E.; Zhao, S.; Agami,
C.; Kagan, H. B. J. Am. Chem. Soc. 1986, 108, 2353. Mikami, K.; Terada,
M. Tetrahedron 1992, 48, 5671. Guillaneux, D.; Zhao, S.-H.; Samuel, O.;
Rainford, D.; Kagan, H. B. J. Am. Chem. Soc. 1994, 116, 9430. Mikami,
K.; Motoyama, Y.; Terada, M. J. Am. Chem. Soc. 1994, 116, 2812.
Kitamura, M.; Suga, S.; Niwa, M.; Noyori, R. J. Am. Chem. Soc. 1995,
117, 4832. Noyori, R.; Suga, S.; Kawai, K.; Okada, S.; Kitamura, M. Pure
Appl. Chem. 1988, 60, 1597. Kagan, H. B.; Girard, C.; Guillaneux, D.;
Rainford, D.; Samuel, O.; Zhang, S. Y.; Zhao, S. H. Acta Chem. Scand.
1996, 50, 345. Girard, C.; Kagan, H. B. Angew. Chem., Int. Ed. Engl. 1998,
37, 2923.
Only a small temperature coefficient was observed for
reactions using catalyst 2e. Thus, while reactions carried out at
room temperature gave compound 3 with 86% ee, reactions
carried out at -80 °C gave a 90% ee but took much longer,
typically 209 h for 68% conversion. To investigate whether the
catalytically active species obtained from titanium(IV) and
ligands 1a-e was monomeric or oligomeric, the relationship
between the enantiomeric excess of ligand 1e and the enantio-
meric excess of silyl ether 3 was investigated for reactions

Asymmetric Addition of Trimethylsilyl Cyanide to Aldehydes
J. Am. Chem. Soc., Vol. 121, No. 16, 1999 3971
Figure 1. Variation of ee of 3 with time using ligands 1 and Ti(OiPr)4 as the catalyst system.
Table 4. Effect of Water and Triethylamine on the Catalytic
Activity of the Titanium Salen Complexes
catalyst
system
reaction conversion
water/Et3N treatment
ee time (h)
(%)
a
1b/Ti(OiPr)4 0.007% water present
56
70
74
62
223
24
24
24
24
100
100
100
80
1b/Ti(OiPr)4 0.12% water present
a
1b/Ti(OiPr)4 1 equiv of water added
a
2bb
2bb
anhydrous
1 equiv of water/2 equiv 80
100
of Et3N
anhydrous
1 equiv of water/2 equiv 86
2eb
2eb
40
24
24
42
100
of Et3N
a,bAll reactions were carried out at ambient temperature: (a) using
1.5 mol % of the catalyst and (b) using 0.1 mol % of the catalyst.
cyanide to benzaldehyde using 20 mol % of the catalyst is very
slow, with reactions requiring 100-200 h to go to completion.
Monitoring of the progress of these reactions showed that the
enantiomeric excess of product 3 increased as the reaction
progressed as shown in Figure 1. This effect is consistent with
the initially formed anhydrous catalyst that has only low activity
and gives relatively low ee values, reacting with adventitious
water during the course of the reaction to generate a new, more
active species which gives a higher ee.
Figure 2. X-ray structure of complex 4a‚4CHCl3. Titanium atoms are
shown in black, oxygen atoms are striped, chlorine atoms are hashed,
and nitrogen atoms are completely white.
In view of these results, it seemed likely that the same
catalytically active species was being generated from both the
in situ titanium tetraisopropoxide/ligand 1 catalyst system and
the titanium dichloride complexes 2. Both systems gave similar
ee values with a wide range of aldehydes and required water
for activation. The main difference between the two systems
was the need to use 20 mol % of the in situ catalyst system and
to work at -80 °C. This could be explained by the fact that the
in situ catalyst system gives a mixture of species, and the low
temperature is necessary to avoid unwanted catalysis by achiral
complexes such as unreacted titanium tetraisopropoxide. To
prove this hypothesis, a preformed mixture of ligand 1e and
titanium tetraisopropoxide was treated with 1 equiv of water,
and complex 2e was treated with 1 equiv of water and 2 equiv
of triethylamine. In both cases, the same species 4e was obtained
as a crystalline solid. Analogous complexes 4a,g were obtained
when dichloride complex 2g was treated with water and
triethylamine, and when ligand 1a was treated with titanium
tetraisopropoxide followed by water. The structure of com-
pounds 4a,e,g is based upon their spectral and analytical
properties and the X-ray structure of complex 4a as shown in
Figure 2.
The X-ray structure of compound 4a suggested that com-
plexes 4e and 4g are probably also dimeric with a central four-
membered ring comprising two titanium atoms and two oxygen
atoms. The oxygen atoms of the four-membered ring have
appreciable basicity as each is hydrogen bonded to a chloroform
molecule (O-H distances 2.369 and 2.238 Å; C-H-O angles
151.1 and 151.4°). The four bonds within the four-membered
ring are not of equal length, rather there are two short Ti-O
bonds (1.806 and 1.821 Å) and two long Ti-O bonds (1.857
and 1.878 Å). In the dimeric structure, the requirement for the
two bridging oxygen atoms to adopt positions cis to one another
means that the salen ligands cannot adopt a planar coordination
around the titanium atoms. Rather, the salen ligands have to
adopt a conformation in which one of the coordinating atoms
(a phenolic oxygen) is nonplanar with the other three coordinat-

3972 J. Am. Chem. Soc., Vol. 121, No. 16, 1999
Belokon’ et al.
Table 5. Asymmetric Addition of Trimethylsilyl Cyanide to
ing groups. The complex retains overall C2 symmetry, however,
since the ligands adopt a ∆-configuration around both titanium
atoms.
Aldehydes Catalyzed by Catalyst 4ea
RCHO, R )
Ph
2-MeC6H4
3-MeC6H4
4-MeC6H4
2-MeOC6H4
3-MeOC6H4
4-MeOC6H4
2,4-(MeO)2C6H3
ee (abs config)
RCHO, R )
ee (abs config)
A survey of the Cambridge crystal structure database revealed
that while there are many crystal structures of titanium
complexes which contain a Ti-O-Ti-O ring system, in the
vast majority of these the oxygen atoms are also bound or
coordinated to additional atoms. Only a few crystal structures
in which the µ-oxygens are both divalent have been reported,27
and only one of these, a titanium/palladium mixed complex,
contains salen ligands.28 The O-Ti-O and Ti-O-Ti bond
angles in these complexes do not vary significantly (82 and 98°,
respectively), but the Ti-O bond varies between 1.810 and
1.923 Å depending upon the nature of the trans ligands.
The catalytic activity of the dimeric complexes 4a,e,g was
investigated using benzaldehyde as the substrate. In each case,
the dimeric complexes gave the same or better enantiomeric
excess to that obtained using the titanium dichloride complexes
under non-anhydrous conditions (56, 86, and 72%, respectively).
However, the reactions with the dimeric catalysts were much
faster, being complete in 1-2 h at ambient temperature as
opposed to the 24 h required for catalysts 2. If the reaction
temperature was reduced then higher ee values could be
obtained. Thus, catalyst 4e gave 86% ee at ambient temperature
(reaction complete in <5 min), 92% ee at 0 °C (reaction
complete in <5 min), and 96% ee at -50 °C, though the latter
reaction was complete after only 3 days. Similarly, catalyst 4g
gave 72% ee at ambient temperature (reaction required 2 h),
84% ee at 0 °C (reaction was complete after 18 h), and 92% ee
at -50 °C, though this reaction was only 75% complete after 3
days. Significantly, however, the ee of the product observed in
these reactions stayed constant throughout the course of the
reaction, as would be expected if the true catalyst was present
throughout the reaction. The use of catalyst 4e to induce the
asymmetric addition of trimethylsilyl cyanide to a range of
aldehydes was subsequently investigated as shown in Table 5.
In each case, the aldehyde was completely consumed after a 1
h reaction at ambient temperature, and in every case the enan-
tiomeric excess obtained using catalyst 4e was the same as or
higher than the enantiomeric excess obtained using catalyst 2e.
Finally, the dimeric complex analogous to 4e but derived from
racemic ligand 1e was prepared. In principle, this could have
given three dimeric complexes: the two enantiomeric ho-
modimers containing two ligands both with the same absolute
configuration and the heterodimer containing one ligand with
the (R,R)-configuration and one ligand with the (S,S)-configu-
ration. The latter complex would be an achiral meso compound
and would be diastereomeric with the homodimers. There are
further diastereomeric possibilities if the relative orientation of
the two salen ligands or the configuration around the titanium
atoms are allowed to vary from those shown in Figure 2.
86 (S)
76 (S)
90 (S)
87 (S)
88 (S)
92 (S)
84 (S)
88 (S)
3,4-(MeO)2C6H3
3,5-(MeO)2C6H3
4-F3CC6H4
4-O2NC6H4
Me3C
Me2CH
MeCH2
85 (S)
90 (S)
86 (S)
50 (S)
66 (S)
64 (S)
52 (S)
a All reactions were carried out using 0.1 mol % of the catalyst for
1 h at ambient temperature.
explains the absence of a nonlinear relationship26 between the
ee of the ligand and the ee of the product in trimethylsilyl
cyanations carried out using ligand 1e and titanium tetraiso-
propoxide. An alternative explanation for these resultssthat the
dimeric complexes dissociate in solutionscan be discounted
since the molecular weight of complex 4a determined by
ultracentrifugation showed it to be dimeric in solution. Further
studies on the structure and mode of action of the catalytically
active species in these reactions will be reported soon, and the
synthetic utility of these and related catalysts is currently being
investigated and will be reported in due course.
Conclusions
This work has resulted in the discovery of a new class of
catalysts for the asymmetric addition of trimethylsilyl cyanide
to aldehydes. The catalysts are dimeric (salen)titanium com-
plexes in which the titanium atoms are bridged by two oxygen
atoms, and the optimum catalyst 4e possesses tert-butyl sub-
stituents in the 3- and 5-positions of the aromatic rings. High
levels of asymmetric induction are obtained using catalyst 4e
and aromatic aldehydes, the product cyanohydrin trimethylsilyl
ethers being obtained with between 76 and 92% enantiomeric
excess at ambient temperature. Aliphatic aldehydes give cy-
anohydrin trimethylsilyl ethers with lower enantiomeric excesses
(52 to 66%). Notably, the reactions are complete within 1 h at
ambient temperature.
Experimental Section
1H NMR spectra were recorded at 250 MHz on a Bruker AM250
1
spectrometer fitted with a H-13C dual probe, and were recorded at
293 K in CDCl3. Spectra were internally referenced either to TMS or
to the residual solvent peak, and peaks are reported in ppm downfield
of TMS. Multiplicities are reported as singlet (s), doublet (d), triplet
(t), quartet (q), some combination of these, broad (br), or multiplet
(m). 13C NMR spectra were recorded at 62.5 MHz on the same
spectrometer as 1H NMR spectra, at 293 K and in CDCl3. Spectra were
referenced to the solvent peak, and are reported in ppm downfield of
TMS. Infrared spectra were recorded on a Perkin-Elmer 1600 series
FTIR spectrometer; only characteristic absorptions are reported. Mass
spectra were recorded using the FAB technique (Cs+ ion bombardment
at 25 kV) on a VG Autospec spectrometer. Only significant fragment
ions are reported, and only molecular ions are assigned. Optical rotations
were recorded on an Optical Activity Ltd. Polar 2001 or a Perkin-
Elmer 241 polarimeter, and are reported along with the solvent and
concentration in g/100 mL. Melting points are uncorrected. Elemental
analyses were performed within the Chemistry department on a Carlo
Erba Model 1106 or Model 1108 analyzer. Analytical ultracentrifugation
was conducted using a MOM 3180 (Hungary) ultracentrifuge with
differential Philpot-Svensson’s optics. Chiral GC was carried out on a
DP-TFA-γ-CD, fused silica capillary column (32m × 0.2 mm) using
helium as the carrier gas.
1
However, the H and 13C NMR spectra of the racemic dimer
were identical with those of complex 4e derived from enantio-
merically pure ligand 1e. This indicates that for this system,
the homodimer is more stable than the heterodimer, and this
(27) Borgias, B. A.; Cooper, S. R.; Koh, Y. B.; Raymond, K. N. Inorg.
Chem. 1984, 23, 1009. Clark, T. J.; Nile, T. A.; McPhail, D.; McPhail, A.
T. Polyhedron 1989, 8, 1804. Qian, Y.; Huang, J.; Chen, X.; Li, G.; Chen,
W.; Li, B.; Jin, X.; Yang, Q. Polyhedron 1994, 13, 1105. Gomez-Sal, P.;
Irigoyen, A. M.; Martin, A.; Mena, M.; Monge, M.; Yelamos, C. J.
Organomet. Chem. 1995, 494, C19. The synthesis and catalytic activity in
asymmetric carbonyl-ene reactions of a di-µ-oxo-titanium-binaphthol
complex has also been reported: Terada, M.; Mikami, K. J. Chem. Soc.,
Chem. Commun. 1994, 833. Kitamoto, D.; Imma, H.; Nakai, T. Tetrahedron
Lett. 1995, 36, 1861.
X-ray crystallographic measurements on complex 4a were made at
193 K using a Siemens P3/PC diffractometer, Mo KR radiation, and a
(28) Kless, A.; Lefeber, C.; Spannenberg, A.; Kempe, R.; Baumann, W.;
Holz, J.; Bo¨rner, A. Tetrahedron 1996, 52, 14599.

Asymmetric Addition of Trimethylsilyl Cyanide to Aldehydes
J. Am. Chem. Soc., Vol. 121, No. 16, 1999 3973
graphite monochromator. Crystal data: C43H43N4O6Cl2Ti2‚4CHCl3 (fw
1126.66); tetragonal; space group P43212; a ) 18.411(3) Å, c )
30.568(6) Å; V ) 10361(3) Å3; Z ) 8; Dc ) 1.445 g cm-3. The structure
was solved by direct methods and refined by full-matrix least-squares
techniques with anisotropic approximations for non-hydrogen atoms.
The crystal contains four solvate chloroform molecules, two of which
are disordered. One of the two disordered molecules occupies a special
position on the C2 axis. The chlorine atoms of the disordered molecules
were refined within the isotropic approximation. All hydrogen atoms
in calculated positions were included in the refining with fixed positional
and thermal parameters (riding model). The absorption correction was
not necessary (µ ) 0.820 mm-1) and therefore was not applied. The
absolute configuration was determined by the refined Flack parameter
which became equal to 0.01(1), thus confirming that the absolute
structure had been determined correctly. The final divergence factors
were R1 ) 0.091 for 2450 independent reflections with I > 2σ(I) and
wR2 ) 0.290 for 4093 independent reflections. All calculations were
carried out using SHELXTL PLUS (PC Version 5.0) programs.29
Preparation of Catalysts 2b,d-g. A mixture of ligand 1b,d-g (1
equiv) and titanium tetrachloride (1 M solution in dichloromethane; 1
equiv) was stirred at room temperature under argon in dry dichlo-
romethane for 2 h. The reaction was concentrated in vacuo to give a
brown solid that was washed with ether and with 50% ether/50%
petroleum ether. The resultant brown solid was recrystallized from
chloroform to give catalysts 2b,d-g.
then filtered through a pad of silica eluting with a 5 /1 mixture of
hexane/ethyl acetate. Concentration of the resulting solution in vacuo
gave the cyanohydrin silyl ethers, the spectroscopic properties of which
were consistent with those reported in the literature.3,6-17 The chemical
yields and enantiomeric excesses of the products are given in Tables
1-3.
Synthesis of the Dimeric Complexes 4e,g from the Dichlorides
2e,g. A solution of the dichloride complex 2e,g (1 equiv), water (1
equiv), and triethylamine (2 equiv) in dichloromethane was stirred at
room temperature for 3 h. The resulting yellow solution was washed
with water, dried over magnesium sulfate, and concentrated in vacuo
to yield the dimeric complex 4e,g as a yellow solid.
4e: Yield 95%; mp 315 °C dec; [R]22 -267 (c 0.0125, CHCl3);
D
(Found: C, 70.9; H, 8.7; N, 4.8. C72H104N4O6Ti2 requires: C, 71.0; H,
8.6; N 4.6); νmax(CH2Cl2) 2962 s, 2865 m, 1625 s, and 1555 cm-1 m;
δH 1.04 (9H, m), 1.22 (9H, m), 1.31 (9H, m), 1.40 (9H, m), 1.8-1.9
(4H, m), 2.5-2.6 (4H, m), 4.0-4.1 (2H, m), 6.95 (1H, s), 7.05 (1H,
s), 7.23 (1H, s), 7.42 (1H, s), 7.75 (1H, s), 8.15 (1H, s); δC 24.6, 24.9,
28.1, 29.7, 30.1, 31.5, 31.7, 34.1, 35.0, 35.7, 65.7, 69.8, 121.0, 121.9,
125.8, 127.5, 128.0, 128.5, 137.6, 138.0, 139.1, 139.4, 157.1, 161.3;
m/z (FAB) 1217 (MH+, 6), 1201 (14), 744 (100), 609 (20).
4g: Yield 93%; mp >320 °C dec); [R]22D -297 (c 0.0125, CHCl3);
(Found: C, 56.7; H, 5.7; N, 9.4. C56H68N8O14Ti2‚0.5H2O requires: C,
56.9; H, 5.8; N 9.5); νmax(CH2Cl2) 2949 s, 2862 m, 1727 w, 1633 s,
1596 s, 1572 m, 1552 w, and 1510 cm-1 s; δH 1.15 (9H, s), 1.50 (9H,
s), 2.5-2.6 (4H, m), 2.8-2.9 (4H, m), 3.85 (2H, m), 8.11 (1H, s),
8.26 (2H, m), 8.37 (3H, m); δC 23.9, 24.4, 25.328.1, 29.2, 29.5, 35.4,
35.9, 67.9, 68.9, 121.7, 122.0, 126.3, 126.9, 128.7, 137.5, 138.0, 138.9,
140.0, 141.1, 144.3, 158.0, 161.2.
Synthesis of Dimers 4a,e from in Situ Catalyst. A solution of
ligand 1a,e (1 equiv) and titanium tetraisopropoxide (1 equiv) in dry
dichloromethane was stirred at room temperature under argon for 2 h.
Water (1 equiv) was then added and the reaction mixture was stirred
at room temperature for 3 h. The resulting yellow solution was
concentrated in vacuo, and the residue was washed with dichlo-
romethane to leave complex 4a,e as a yellow solid.
2b: Yield 19%; mp 250-300 °C; [R]21 +390 (c 0.033, CHCl3);
D
(Found: C, 59.0; H, 6.8; N, 5.2. C28H36N2O2Cl2Ti‚1/3CH2Cl2 requires:
C, 59.3; H, 6.4, N, 4.9); νmax(Nujol) 1618 s, 1583 m, and 1560 cm-1
m; δH 1.4-1.6 (2H, m), 1.53 (9H, s), 2.0-2.1 (1H, m), 2.5-2.6 (1H,
m), 4.0-4.1 (1H, m), 7.01 (1H, t, J ) 7.7), 7.41 (1H, d, J ) 7.7), 7.58
(1H, d, J ) 7.7), 8.31 (1H, s); δC 24.1, 28.4, 29.8, 35.4, 67.7, 122.2,
126.2, 133.5, 133.7, 137.6, 160.2, 161.7.
2d: Yield 76%; mp 280-300 °C dec; [R]24D +153 (c 0.01, CHCl3);
(Found: C, 53.4; H, 6.3; N, 4.4. C30H40N2O4Cl2Ti‚CH2Cl2 requires:
C, 53.5; H, 6.1, N, 4.0); νmax(Nujol) 1615 s, 1594 s, and 1560 cm-1 s;
δH 1.2-1.4 (2H, m), 1.51 (9H, s), 2.0-2.1 (1H, m), 2.4-2.6 (1H, m),
4a: Yield 50%; [R]25D -410 (c 0.04, CHCl3); (Found: C, 61.7; H,
5.4; N, 7.3; Ti, 12.3. C40H40 N4O6Ti2 requires: C, 61.8; H, 5.4; N, 7.3;
Ti, 12.3); δH 1.3-1.4 (7H, m), 1.8-1.9 (7H, m), 2.3-2.4 (2H, m),
2.7-2.8 (2H, m), 3.8-3.9 (2H, m), 6.95 (1H, s), 6.75-7.45 (16H, m),
8.07 (2H, s), 8.24 (2H, m). Molecular weight determined by analytical
ultracentrifugation30 773 (c 3-6 g/L, THF, 25 °C), calculated for
C40H40N4O6Ti2 768 (2% deviation).
Asymmetric Addition of Trimethylsilyl Cyanide to Aldehydes
Induced by Catalyst 4. To a stirred solution of the aldehyde (1.25
mmol) and catalyst 4a,e (0.00125 mmol) in dry dichloromethane (1.5
ml) under an argon atmosphere was added dropwise trimethylsilyl
cyanide (1.38 mmol). The reaction mixture was stirred for 24 h and
was then filtered through a pad of silica eluting with a 5/1 mixture of
hexane/ethyl acetate. Concentration of the resulting solution in Vacuo
gave the cyanohydrin silyl ethers, the spectroscopic properties of which
were consistent with those reported in the literature.3,6-17 The enan-
tiomeric excesses of the products are given in Table 5.
3.80 (3H, s), 4.0-4.1 (1H, m), 6.85 (1H, d, J ) 3.0), 7.16 (1H, d, J )
3.0), 8.25 (1H, s); δC (DMSO-d6) 23.9, 29.0, 29.7, 35.1, 55.8, 67.8,
115.7, 124.8, 128.1, 138.3, 141.0, 152.2, 161.1; m/z (FAB) 692 ((M -
Cl2 + o-nitrobenzyl alcohol)+, 100), 676 (10), 555 (12), 541 (15).
2e: Yield 95%; mp 330 °C dec; [R]22 +736 (c 0.0125, CHCl3);
D
(Found: C, 63.8; H, 8.1; N, 4.4. C36H52N2O2Cl2Ti‚0.7H2O requires:
C, 63.9; H, 8.0; N, 4.1); νmax(KBr) 3449 br m, 2955 s, 1617 s, and
1561 cm-1 m; δH 1.35 (9H, s), 1.4-1.65 (2H, m), 1.54 (9H, s), 2.05-
2.1 (1H, m), 2.5-2.65 (1H, m), 4.0-4.1 (1H, m), 7.35 (1H, d, J )
2.3), 7.61 (1H, d, J ) 2.3) 8.32 (1H, s); δC 24.1, 28.5, 29.9, 31.4, 34.5,
35.5, 67.7, 125.7, 130.1, 131.1, 136.7, 144.6, 159.7, 160.5; m/z (FAB)
663 (MH+, 20), 629 (100).
2f: Yield 99%; mp 280-300 °C; [R]21 +112 (c 0.0125, CHCl3);
D
(Found: C, 73.5; H, 6.3; N, 2.6. C66H64N2O2Cl2Ti requires: C, 73.2;
H, 6.0, N, 2.6); νmax(Nujol) 1611 s, and 1552 cm-1 m; δH 1.35 (9H, s),
1.5-1.7 (2H, m), 1.8-2.0 (1H, m), 2.3-2.5 (1H, m), 3.9-4.1 (1H,
m), 6.9-7.4 (17H, m), 8.11 (1H, s); δC 24.0, 29.7, 35.4, 64.6, 67.8,
125.3, 126.2, 127.9, 131.0, 135.3, 136.3, 137.3, 140.5, 146.4, 160.0,
160.5.
Acknowledgment. The authors thanks the EPSRC, INTAS,
the EU (INCO-COPERNICUS and SOCRATES), and the
Russian Fund for Fundamental Research (Grant No. 98-03-
32862) for financial support. Mass spectra were recorded by
the EPSRC national mass spectrometry service.
2g: Yield 100%; mp 280-300 °C dec; [R]21 +268 (c 0.0125,
D
CHCl3); (Found: C, 52.0; H, 5.7; N, 8.3. C28H36N4O6Cl2Ti requires:
C, 52.3; H, 5.6, N, 8.7); νmax(Nujol) 1620 s, 1592 s, and 1513 cm-1 m;
δH 1.2-1.5 (2H, m), 1.56 (9H, s), 2.1-2.2 (1H, m), 2.6-2.7 (1H, m),
4.0-4.1 (1H, m), 8.4-8.5 (3H, m); δC 23.9, 28.6, 29.4, 35.9, 68.4,
125.6, 128.8, 129.1, 139.7, 157.0, 159.5, 160.2.
Supporting Information Available: Tables of crystal data,
structure solution and refinements, atomic coordinates, bond
lengths and angles, and anisotropic thermal parameters for 4a;
table of GC conditions for ee determinations; 1H NMR spectra
for complexes 2b, 2e, and 4g (PDF). This material is available
free of charge via the Internet at http://pubs.acs.org.
Asymmetric Addition of Trimethylsilyl Cyanide to Aldehydes
Induced by Catalysts 2. To a stirred solution of the aldehyde (1.25
mmol) and catalyst 2b,d-g (0.00125 mmol) in dry dichloromethane
(1.5 mL) under an argon atmosphere was added dropwise, trimethylsilyl
cyanide (1.38 mmol). The reaction mixture was stirred for 24 h and
(29) Sheldrick, G. M. SHELXTL PC Version 5.0. An integrated System
for Solving, Refining, and Displaying Crystal Structures from Diffraction
Data; Siemens Analytical X-ray Instruments Inc.: Madison, WI, 1994.
JA984197V
(30) Van Holde, K. E.; Baldwin, R. L. J. Phys. Chem. 1958, 62, 734.