Insight into the mechanism of the asymmetric ring-opening aminolysis of 4,4-dimethyl-3,5,8-trioxabicyclo[5.1.0]octane catalyzed by titanium/BINOLate/ water system: Evidence for the Ti(BINOLate)2-bearing active catalyst entities and the role of water

Article abstract of DOI:10.1021/ja801847r

The mechanism of the enantioselective ring-opening aminolysis of 4,4-dimethyl-3,5,8-trioxabicyclo[5.1.0]octane with benzylamine, catalyzed by the titanium-BINOLate species generated in situ from a mixture of enantiopure BINOL (1,1′-bi-2-naphthol), Ti(OiPr)₄, and H₂O in the presence of benzylamine in toluene, was investigated in detail using a combination of reaction profile measurements, nonlinear effect (NLE) studies, solution ¹H NMR analysis, electrospray ionization mass spectrometry (ESI-MS), as well as the results obtained from screening of dynamic catalyst library of complexes L_a/Ti/L_b (L_a or L _b = chiral diol ligands). The BINOL-to-titanium ratio and the presence or absence of water in the catalytic system were found to exert profound influences on both reactivity and enantioselectivity of the reaction. The NLE studies revealed that the titanium species involved in the catalysis should contain more than one BINOL unit, either within or at the periphery of the catalytic cycle. ESI-MS analysis of the catalytic systems provided strong support in favor of the mechanistic proposal that titanium complexes bearing the Ti(BINOLate)₂ moiety should be the active species responsible for the catalysis, which was further confirmed by the observation of synergistic effect of the heteroligand combinations during screening of the dynamic catalyst library. ESI-MS analysis of the reaction system indicated that water does not take part in the catalyst generation, which is an unprecedented finding in contrast to the previous mechanistic understandings in the titanium catalytic chemistry involving the participation of water. Most probably, water functions as a proton shuttle in the catalysis, facilitating the proton transfer between the reactants. Furthermore, the origin of (+)-NLE observed in the present catalytic system is rationalized on the basis of the ESI-MS analysis of the catalyst system prepared from a 1:1 pseudoracemic mixture of (S)-BINOL and (R)-3,3′,6,6′-D₄-BINOL. Finally, the reactivity differences between several couples of epoxide/amine combinations were tentatively rationalized on the basis of the arguments on their relative coordination preferences, and several other aliphatic amines were also found to efficiently ring-open the titled epoxide in excellent enantioselectivities. The results from this study are expected to shed some light on the often elusive chemistry of Ti(IV)-based catalytic systems where water or molecular sieves (or alcohols, etc.) are found to play an important yet inexplicable role and may help the search for effective asymmetric Ti(IV) catalysts for other types of reactions.

Full text of DOI:10.1021/ja801847r

Published on Web 07/11/2008
Insight into the Mechanism of the Asymmetric Ring-Opening
Aminolysis of 4,4-Dimethyl-3,5,8-trioxabicyclo[5.1.0]octane
Catalyzed by Titanium/BINOLate/Water System: Evidence for
the Ti(BINOLate)2-Bearing Active Catalyst Entities and the
Role of Water
Hongli Bao,†,‡ Jing Zhou,† Zheng Wang,*,† Yinlong Guo,† Tianpa You,‡ and
Kuiling Ding*,†
State Key Laboratory of Organometallic Chemistry and Shanghai Mass Spectrometry Center,
Shanghai Institute of Organic Chemistry, Chinese Academy of Sciences, 354 Fenglin Road,
Shanghai 200032, P. R. China, and Department of Chemistry, UniVersity of Science and
Technology of China, Hefei, Anhui 230026, China
Received March 12, 2008; E-mail: wzsioc@mail.sioc.ac.cn; kding@mail.sioc.ac.cn
Abstract: The mechanism of the enantioselective ring-opening aminolysis of 4,4-dimethyl-3,5,8-
trioxabicyclo[5.1.0]octane with benzylamine, catalyzed by the titanium-BINOLate species generated in situ
from a mixture of enantiopure BINOL (1,1′-bi-2-naphthol), Ti(OiPr)4, and H2O in the presence of benzylamine
in toluene, was investigated in detail using a combination of reaction profile measurements, nonlinear effect
(NLE) studies, solution 1H NMR analysis, electrospray ionization mass spectrometry (ESI-MS), as well as
the results obtained from screening of dynamic catalyst library of complexes La/Ti/Lb (La or Lb ) chiral diol
ligands). The BINOL-to-titanium ratio and the presence or absence of water in the catalytic system were
found to exert profound influences on both reactivity and enantioselectivity of the reaction. The NLE studies
revealed that the titanium species involved in the catalysis should contain more than one BINOL unit,
either within or at the periphery of the catalytic cycle. ESI-MS analysis of the catalytic systems provided
strong support in favor of the mechanistic proposal that titanium complexes bearing the Ti(BINOLate)2
moiety should be the active species responsible for the catalysis, which was further confirmed by the
observation of synergistic effect of the heteroligand combinations during screening of the dynamic catalyst
library. ESI-MS analysis of the reaction system indicated that water does not take part in the catalyst
generation, which is an unprecedented finding in contrast to the previous mechanistic understandings in
the titanium catalytic chemistry involving the participation of water. Most probably, water functions as a
proton shuttle in the catalysis, facilitating the proton transfer between the reactants. Furthermore, the origin
of (+)-NLE observed in the present catalytic system is rationalized on the basis of the ESI-MS analysis of
the catalyst system prepared from a 1:1 pseudoracemic mixture of (S)-BINOL and (R)-3,3′,6,6′-D4-BINOL.
Finally, the reactivity differences between several couples of epoxide/amine combinations were tentatively
rationalized on the basis of the arguments on their relative coordination preferences, and several other
aliphatic amines were also found to efficiently ring-open the titled epoxide in excellent enantioselectivities.
The results from this study are expected to shed some light on the often elusive chemistry of Ti(IV)-based
catalytic systems where water or molecular sieves (or alcohols, etc.) are found to play an important yet
inexplicable role and may help the search for effective asymmetric Ti(IV) catalysts for other types of reactions.
Introduction
classes of chiral catalysts and have been used in a wide variety
of highly enantioselective reactions.1–3 Despite this fact, how-
In the context of asymmetric catalysis, titanium complexes
of BINOL and its derivatives are one of the most versatile
(2) Balsells, J.; Davis, T. J.; Carroll, P.; Walsh, P. J. J. Am. Chem. Soc.
2002, 124, 10336–10348, and references therein.
† Shanghai Institute of Organic Chemistry.
‡ University of Science and Technology of China.
(3) For recent examples, see: (a) Dong, H. M.; Lu, H. H.; Lu, L. Q.;
Chen, C. B.; Xiao, W. J. AdV. Synth. Catal. 2007, 349, 1597–1603.
(b) Kim, J. G.; Camp, E. H.; Walsh, P. J. Org. Lett. 2006, 8, 4413–
4416. (c) Mahrwald, R.; Schetter, B. Org. Lett. 2006, 8, 281–284. (d)
Kano, T.; Hashimoto, T.; Maruoka, K. J. Am. Chem. Soc. 2006, 128,
2174–2175. (e) Gao, G.; Wang, Q.; Yu, X. Q.; Xie, R. G.; Pu, L.
Angew. Chem., Int. Ed. 2006, 45, 122–125. (f) Qin, Y. C.; Liu, L.;
Sabat, M.; Pu, L. Tetrahedron 2006, 62, 9335–9348. (g) Zhang, Y. L.;
Zhang, F.; Tang, W. J.; Wu, Q. L.; Fan, Q. H. Synlett 2006, 1250–
1254. (h) Yang, F.; Xi, P.; Yang, L.; Lan, J.; Xie, R.; You, J. J. Org.
Chem. 2007, 72, 5457–5460. (i) Chen, C. A.; Wu, K. H.; Gau, H. M.
(1) For an early review, see: (a) Mikami, K.; Shimizu, M. Chem. ReV
1992, 92, 1021–1050. For recent reviews, see: (b) Chen, Y.; Yekta,
S.; Yudin, A. K. Chem. ReV. 2003, 103, 3155–3211. (c) Brunel, J. M.
Chem. ReV. 2005, 105, 857–897. (d) Ramon, D. J.; Yus, M. Chem.
ReV. 2006, 106, 2126–2208. (e) Mikami, K.; Terada, M. In Lewis Acids
in Organic Synthesis; Yamamoto, H., Ed.; Wiley-VCH: Weinheim,
Germany, 2000, Vol. 2, pp 799-847. (f) Yuan, Y.; Ding, K.; Chen,
G. In Acid Catalysis in Modern Organic Synthesis; Yamamoto, H.,
Ishihara, K., Eds.; Wiley-VCH: Weinheim, Germany, 2008, Vol. 2,
pp 721-823.
9
10116 J. AM. CHEM. SOC. 2008, 130, 10116–10127
10.1021/ja801847r CCC: $40.75
2008 American Chemical Society

BINOL/Ti(IV)-Catalyzed Ring-Opening Aminolysis
A R T I C L E S
ever, the mechanistic elucidation of the multifarious aspects of
the BINOL/Ti catalysis still remains an extremely challenging
task for most reactions, and only some progress has been made
in recent years.2,4 The catalytic species generated in situ from
various titanium precursors and BINOL derivatives often display
a rich structural chemistry and complicated behaviors in
solution,2,5 as a result of the well-known flexibility of 1,1′-
binaphthylic core,6 the variable coordination geometries of
titanium, as well as its strong tendency to oligomerize into
complicated supramolecular assemblies.5,7 Furthermore, varia-
tions in reaction parameters such as the stoichiometric ratio of
chiral ligand-to-metal,8,9 the presence or absence of water5f,7f,10
and molecular sieves,11 solvent, temperature, and so forth11e,12
have also been found to exert a profound influence on the
catalytic behavior of the BINOL/Ti systems as a result of distinct
catalyst structures involved in the reactions. Indeed, these can
be regarded as common features for asymmetric Ti(IV) catalysis
mediated with various chiral diols, whereby a slight modification
on the catalyst composition can result in dramatically different
catalyst structures and catalytic behaviors.13 A classical example
of the latter can be found in the pioneering studies by Kagan et
al.14 and Modena et al.15 in the development of titanium
alkoxide/tartrate catalysts for enantioselective thioether oxida-
tion. Although using the standard Sharpless reagent for epoxi-
dation of allylic alcohols (1:1 titanium alkoxide/tartrate deriva-
tive) in the sulfide oxidation reaction only gave racemic
sulfoxides, variations made on the ligand-to-Ti ratios and/or
the addition of water resulted in dramatic improvements in the
enantioselectivity. In general, the titanium complexes of the
chiral diols used in various asymmetric catalyses are generated
in situ, which often results in the formation of a dynamic mixture
of metal-containing species equilibrating with each other in
solution. While this situation may be beneficial for facile fine-
tuning of the catalysis through modifications on a diverse range
of structural and reaction parameters, it often makes the
unequivocal identification of the nature of active catalyst an
extremely difficult and somewhat elusive goal.
Optically active anti- or syn-2-aminobutane-1,3,4-triol (ABT)
equivalents, a class of highly functionalized compounds in
various protected or unprotected forms, are extremely versatile
chiral C4 building blocks for the efficient synthesis of a diversity
of biologically active compounds, such as phytosphingosine,16
statine,17 and nelfinavir.18 Accordingly, several useful strategies
for the synthesis of enantiopure ABT equivalents have been
developed in recent years, either through the functional group
manipulation of naturally occurring chiral compounds19 or via
asymmetric syntheses where new stereogenic centers are ste-
reoselectively formed in the products.18a For practical utilization
of ABT derivatives as versatile chiral synthons, the compounds
should be available on large scales, with desired relative and
absolute stereochemical structures, as well as in suitably
protected forms for the selective transformation of their three
hydroxyl groups and amino group. Unfortunately, many ABT
synthetic routes20 developed thus far cannot satisfy these
requirements. Among the few exceptions, the Ti(IV)/BINOL-
catalyzed asymmetric aminolysis of the meso epoxide 3,5,8-
Angew. Chem., Int. Ed. 2007, 46, 5373–5376. (j) Gou, S.; Chen, X.;
Xiong, Y.; Feng, X. J. Org. Chem. 2006, 71, 5732–5736. (k) Lin, L.;
Chen, Z.; Yang, X.; Liu, X.; Feng, X. Org. Lett. 2008, 10, 1311–
1314.
(4) For examples, see: (a) Walsh, P. J. Acc. Chem. Res. 2003, 36, 739–
749. (b) Terada, M. J. Synth. Org. Chem. Jpn. 2007, 65, 748–760.
(5) (a) Mikami, K.; Matsukawa, S. J. Am. Chem. Soc. 1993, 115, 7039–
7040. (b) Boyle, T. J.; Eilerts, N. W.; Heppert, J. A.; Takusagawa, F.
Organometallics 1994, 13, 2218–2229. (c) Waltz, K. M.; Carroll, P. J.;
Walsh, P. J. Organometallics 2004, 23, 127–134. (d) Pescitelli, G.;
Di Bari, L.; Salvadori, P. Organometallics 2004, 23, 4223–4229. (e)
Mikami, K.; Terada, M.; Matsumoto, Y.; Tanaka, M.; Nakamura, Y.
Microporous Mesoporous Mater. 1998, 21, 461–466. (f) Terada, M.;
Matsumoto, Y.; Nakamura, Y.; Mikami, K. Inorg. Chim. Acta 1999,
296, 267–272. (g) Hanawa, H.; Uraguchi, D.; Konishi, S.; Hashimoto,
T.; Maruoka, K. Chem.-Eur. J. 2003, 9, 4405–4413. (h) Salvadori
observed a tetrameric titanium species with a formula of (BINOLate)6-
Ti4(µ3-OH)4 in Uemura et al.’s catalyst system (see ref 10e,f) for
asymmetric sulfoxidation prepared by using a 2:1 molar ratio of
BINOL and Ti(OiPr)4 in the presence of 10 equiv of H2O. See:
Pescitelli, G.; Di Bari, L.; Salvadori, P. J. Organomet. Chem. 2006,
691, 2311–2318.
(6) (a) Whitesell, J. K. Chem. ReV. 1989, 89, 1581–1590. (b) Telfer, S. G.;
Kuroda, R. Coord. Chem. ReV. 2003, 242, 33–46.
(7) (a) Boyle, T. J.; Barnes, D. L.; Heppert, J. A.; Morales, L.;
Takusagawa, F.; Connolly, J. C. Organometallics 1992, 11, 1112–
1126. (b) Wu, Y. T.; Ho, Y. C.; Lin, C. C.; Gau, H. M. Inorg. Chem.
1996, 35, 5948–5952. (c) Davis, T. J.; Balsells, J.; Carroll, P. J.; Walsh,
P. J. Org. Lett. 2001, 3, 699–702. (d) Pandiaraju, S.; Chen, G.; Lough,
A.; Yudin, A. K. J. Am. Chem. Soc. 2001, 123, 3850–3851. (e) Liu,
B.; Jiang, F. Y.; Song, H. B.; Li, J. S. Tetrahedron: Asymmetry 2006,
17, 2149–2153. (f) Mikami, K.; Ueki, M.; Matsumoto, Y.; Terada,
M. Chirality 2001, 13, 541–544. (g) Corden, J. P.; Errington, W.;
Moore, P.; Partridge, M. G.; Wallbridge, M. G. H. Dalton Trans. 2004,
1846–1851. (h) Guo, Q. S.; Song, H. B.; Li, J. S. Z. Anorg. Allg.
Chem. 2006, 632, 721–722.
(11) (a) Aoki, S.; Mikami, K.; Terada, M.; Nakai, T. Tetrahedron 1993,
49, 1783–1792. (b) Terada, M.; Mikami, K. J. Chem. Soc., Chem.
Commun. 1994, 833–834. (c) Terada, M.; Matsumoto, Y.; Nakamura,
Y.; Mikami, K. J. Mol. Catal. A: Chem. 1998, 132, 165–169. (d)
Mikami, K.; Motoyama, Y.; Terada, M. J. Am. Chem. Soc. 1994, 116,
2812–2820. (e) Posner, G. H.; Dai, H.; Bull, D. S.; Lee, J. K.; Eydoux,
F.; Ishihara, Y.; Welsh, W.; Pryor, N.; Petr, S., Jr J. Org. Chem. 1996,
61, 671–676.
(12) Keck, G. E.; Krishnamurthy, D. J. Am. Chem. Soc. 1995, 117, 2363–
2364.
(13) (a) Finn, M. G.; Sharpless, K. B. J. Am. Chem. Soc. 1991, 113, 113–
126. (b) Seebach, D.; Plattner, D. A.; Beck, A. K.; Wang, Y. M.;
Hunziker, D. HelV. Chim. Acta 1992, 75, 2171–2209. (c) Potvin, P. G.;
Fieldhouse, B. G. Tetrahedron: Asymmetry 1999, 10, 1661–1672. (d)
Hintermann, L.; Broggini, D.; Togni, A. HelV. Chim. Acta 2002, 85,
1597–1612.
(8) Keck, G. E.; Li, X. Y.; Krishnamurthy, D. J. Org. Chem. 1995, 60,
5998–5999.
(9) For examples, see: (a) Weigand, S.; Bruckner, R. Chem.-Eur. J. 1996,
2, 1077–1084. (b) Yamazaki, S.; Tanaka, M.; Yamabe, S. J. Org.
Chem. 1996, 61, 4046–4050. (c) Feng, X.; Shu, L.; Shi, Y. J. Am.
Chem. Soc. 1999, 121, 11002–11003. (d) Feng, X.; Shu, L.; Shi, Y.
J. Org. Chem. 2002, 67, 2831–2836. (e) Wang, B.; Feng, X.; Cui, X.;
Liu, H.; Jiang, Y. Chem. Commun. 2000, 1605–1606. (f) Wang, B.;
Feng, X.; Huang, Y.; Liu, H.; Cui, X.; Jiang, Y. J. Org. Chem. 2002,
67, 2175–2182. (g) Wooten, A. J.; Kim, J. G.; Walsh, P. J. Org. Lett.
2007, 9, 381–384.
(14) (a) Pitchen, P.; Kagan, H. B. Tetrahedron Lett. 1984, 25, 1049–1052.
(b) Brunel, J.-M.; Diter, P.; Duetsch, M.; Kagan, H. B. J. Org. Chem.
1995, 60, 8086–8088.
(15) (a) Furia, F. D.; Modena, G.; Seraglia, R. Synthesis 1984, 4, 325–
326. (b) Bortolini, O.; Furia, F. D.; Licini, G.; Modena, G.; Rossi, M.
Tetrahedron Lett. 1986, 6257–6260.
(16) Hertweck, C.; Sebek, P.; Svatos, A. Synlett 2001, 1965–1967.
(17) Kwon, S. J.; Ko, S. Y. Tetrahedron Lett. 2002, 43, 639–641.
(18) (a) Inaba, T.; Birchler, A. G.; Yamada, Y.; Sagawa, S.; Yokota, K.;
Ando, K.; Uchida, I. J. Org. Chem. 1998, 63, 7582–7583. (b) Inaba,
T.; Yamada, Y.; Abe, H.; Sagawa, S.; Cho, H. J. Org. Chem. 2000,
65, 1623–1628.
(10) For examples, see: (a) Massa, A.; Lattanzi, A.; Siniscalchi, F. R.;
Scettri, A Tetrahedron: Asymmetry 2001, 12, 2775–2777. (b) Massa,
A.; Siniscalchi, F. R.; Bugatti, V.; Lattanzi, A.; Scettri, A. Tetrahedron:
Asymmetry 2002, 13, 1277–1283. (c) Kurosu, M.; Lorca, M. Synlett
2005, 1109–1112. (d) Kurosu, M.; Lorca, M. T. Synlett 2005, 1109–
1112. (e) Komatsu, N.; Hashizume, M.; Sugita, T.; Uemura, S. J. Org.
Chem. 1993, 58, 4529–4533. (f) Komatsu, N.; Hashizume, M.; Sugita,
T.; Uemura, S. J. Org. Chem. 1993, 58, 7624–7626. (g) Kitamoto,
D.; Imma, H.; Nakai, T. Tetrahedron Lett. 1995, 36, 1861–1864.
(19) (a) Badorrey, R.; Cativiela, C.; Diaz-De-Villegas, M. D.; Diez, R.;
Galvez, J. A. Eur. J. Org. Chem. 2003, 2268–2275. (b) Kwon, S. J.;
Ko, S. Y. J. Org. Chem. 2001, 66, 6833–6835.
(20) Ikunaka, M. Chem.-Eur. J. 2003, 9, 378–388.
9
J. AM. CHEM. SOC. VOL. 130, NO. 31, 2008 10117

A R T I C L E S
Bao et al.
Scheme 1. BINOL/Ti(OiPr)4-Catalyzed Asymmetric Aminolysis of
the Epoxide 1 with Benzylamine
trioxabicyclo[5.1.0]octane (1) using benzylamine and its de-
rivatives, developed by Inaba et al., can be regarded as one of
the most efficient routes to optically pure ABT derivatives
(Scheme 1).18,21 In this procedure, the reactions were performed
in toluene (or 9:1 heptane/toluene) at 40 °C in the presence of
the catalyst generated from (S)-BINOL and Ti(OiPr)4 and water
in a ratio of 1:1:10-20, affording the products in excellent
stereoselectivity (>97% ee with 1-phenylethanamine as nu-
cleophile 93% ee with benzylamine, respectively) with high
conversions. The protected enantiopure 2-aminobutane-1,3,4-
triols 3 and their enantiomers are accessible in bulk quantities
using this protocol and have been successfully used as chiral
intermediates in the practical syntheses of sphinganines16 and
Nelfinavir (a potent HIV-protease inhibitor).18,20,21 Although a
simplified molecular model of the catalytic species chelated by
the epoxide 1 has been proposed by Inaba et al. on the basis of
force field calculation of the model complex 1-Ti(OMe)4, no
detailed mechanistic information was reported in the original
procedure. The remarkable practical importance of the reaction
coupled with the intriguing role of water in this and some other
Ti-diol-catalyzed reactions prompted us to undertake a further
study on the reaction mechanism as a part of our ongoing
interests in developing titanium catalytic chemistry.22 Herein,
we report the mechanistic investigation of the titled reaction
using a combination of reaction profile measurements, nonlinear
Figure 1. Reaction profiles of the (R)-BINOL/Ti(OiPr)4/H2O-catalyzed
aminolysis of 1 with benzylamine using different (R)-BINOL/Ti molar ratios
at 1 mol % loading of Ti(OiPr)4 or racemic BINOL/Ti(OiPr)4/H2O-catalyzed
aminolysis of 1.
of the derivatized product using a chiral AD column (Figure
1). Control experiments showed that in the absence of BINOL
the reaction was negligible after 24 h, and with a BINOL/Ti
molar ratio of 0.5 only 11% conversion of 1 was observed within
the same period. As shown in Figure 1, using a BINOL/Ti molar
ratio of 1:1 (corresponding to Inaba et al.’s standard aminolysis
conditions),21 the reaction took 24 h to reach a conversion of
96%. Further increase of the BINOL/Ti molar ratios up to 2:1
resulted in a considerable enhancement in the initial rate for
the reaction, which took only 6 h to reach full conversion of 1
and did not show any incubation period in contrast to the 1:1
case. Moreover, the enantioselectivity of the reaction was also
improved when the BINOL/Ti molar ratio was switched from
1:1 to 2:1 (92 vs 96% ee). Notably, further increment of the
BINOL/Ti molar ratio to 3:1 only led to an almost identical
reaction profile as the 2:1 system with a slight deterioration in
enantioselectivity (92% ee). A plausible explanation of this
behavior is the existence of some dynamic equilibria involving
the active species, whose concentration can be affected by the
subtle variation on the ligand-to-metal ratios in the present
system.2,5a–d,7g Therefore, under the present conditions, the
BINOL/Ti molar ratio of 2:1 turns out to be optimal for the
reaction in terms of both catalytic activity and enantioselectivity.
Impact of Water on Catalysis. We next proceeded to
investigate the catalytic behavior of the aminolysis in the
presence of various amounts of water. Water has often been
shown to play an important yet somewhat intriguing role in
chiral diol/Ti catalysis, usually by facilitating the formation of
oxo- and/or hydroxo-bridged aggregates/oligomers through
partial hydrolysis of the titanium alkoxide complexes.5,7,10g,23
For the aminolysis of 1 with benzylamine, the reactions were
carried out by variation of the molar percentage of water in the
system under the otherwise optimized conditions as above
(BINOL/Ti ratio ) 2:1), and the results were depicted in Figure
2. From the figure, it is clear that, although the reaction can
still proceed without the addition of water, the presence of a
certain amount of water (5-25 mol % relative to that of 1) in
1
effect (NLE) studies, solution H NMR analysis, electrospray
ionization mass spectrometry (ESI-MS), as well as the screening
of dynamic catalyst library of complexes La/Ti/Lb (La or Lb )
chiral diol ligands).
Results and Discussion
Reaction Profiles: Influence of the Ratios of BINOL/Ti on
Reactivity and Enantioselectivity. The work started with the
examination on the potential impact of the BINOL/Ti molar
ratios on the reactivity and enantioselectivity of the aminolysis
of 1 with benzylamine. The reactions were performed in toluene
(2.0 M of the epoxide and an equimolar amount of benzylamine)
at 40 °C in the presence of 15 mol % water, with 1 mol % of
Ti(OiPr)4 loading under several different BINOL/Ti molar ratios.
The reaction time courses were monitored by 1H NMR analysis
of aliquots taken from reaction mixtures at specified time
periods, and ee values were assessed by HPLC determination
(21) Sagawa, S.; Abe, H.; Hase, Y.; Inaba, T. J. Org. Chem. 1999, 64,
4962–4965.
(22) (a) Long, J.; Hu, J.; Shen, X.; Ji, B.; Ding, K. J. Am. Chem. Soc.
2002, 124, 10–11. (b) Yuan, Y.; Zhang, X.; Ding, K. Angew. Chem.,
Int. Ed. 2003, 42, 5478–5480. (c) Yuan, Y.; Wang, X.; Li, X.; Ding,
K. J. Org. Chem. 2004, 69, 146–149. (d) Wang, X.; Wang, X.; Guo,
H.; Wang, Z.; Ding, K. Chem.-Eur. J. 2005, 11, 4078–4088. (e) Guo,
H.; Wang, X.; Ding, K. Tetrahedron Lett. 2004, 45, 2009–2012. (f)
Ji, B.; Yuan, Y.; Ding, K.; Meng, J. Chem.-Eur. J. 2003, 9, 5989–
5996. (g) Yuan, Y.; Li, X.; Sun, J.; Ding, K. J. Am. Chem. Soc. 2002,
124, 14866–14867. (h) Yuan, Y.; Long, J.; Sun, J.; Ding, K. Chem.-
Eur. J. 2002, 8, 5033–5042.
(23) Mikami, K.; Matsumoto, Y.; Xu, L. Inorg. Chim. Acta 2006, 359,
4159–4167.
9
10118 J. AM. CHEM. SOC. VOL. 130, NO. 31, 2008

BINOL/Ti(IV)-Catalyzed Ring-Opening Aminolysis
A R T I C L E S
Figure 2. Influence of water content (X: mol % relative to 1) on the BINOL/
Ti(OiPr)4/H2O (molar ratio 1:2:X) catalyzed aminolysis of 1 with benzyl-
amine at 1 mol % loading of Ti(OiPr)4.
Figure 3. Positive NLE observed in the BINOL/Ti/H2O-catalyzed enan-
tioselective ring-opening aminolysis of epoxide 1 by benzylamine.
the system demonstrated a beneficial effect on both reactivity
and enantioselectivity of the reaction (X ) 5, 15, or 25 vs X )
0). Further increasing the water loading to 50 mol %, however,
led to a significant deterioration in both reactivity and enanti-
oselectivity, suggesting a partial decomposition of the catalyst
might have occurred in the presence of too much water. Herein,
15 mol % of water (relative to the epoxide) is again found to
be the optimal value. It is noteworthy that the catalyst composi-
tion of the present system is quite similar to that reported
previously by Uemura et al. for asymmetric sulfoxidation,10e,f
both containing the (R)-BINOL/Ti(OiPr)4/H2O in a ratio of 2:1:
15 (or 20). Since the solution structure of the catalyst (or
precatalyst) of Uemura et al.’s catalyst system has been
unequivocally elucidated by Salvadori et al. to be a tetrameric
titanium species (BINOLate)6Ti4(µ3-OH)4,5h we wondered
whether it would be the same species that is working in the
present reaction. For this purpose, we prepared the titanium
catalyst following an analogous Uemura et al.’s procedure5h,10e,f
(i.e., adding 15 equiv of H2O (relative to Ti) to the mixture of
BINOL/Ti(OiPr)4 (2:1) in toluene followed by BnNH2), and the
result is also depicted in Figure 2. It should be noted that here
water is added before the addition of benzylamine nucleophile.
Obviously, compared with the present reaction protocol, sig-
nificant degradation in both reactivity and ee was observed by
using Uemura et al.’s catalyst. Moreover, in this case only 2%
conversion of 1 was obtained after 1 h reaction, indicating an
incubation period is required for active catalyst formation. This
behavior indicated that different catalytic species should be
operating in the two reaction systems, and the tetrameric
titanium species observed in Uemura et al.’s system should not
be directly involved in the catalytic cycle of the current
aminolysis reaction.
configurations.24 Therefore, the study of NLE with nonenan-
tiomerically pure catalysts has often been used as a powerful
diagnostic tool for probing the nature and especially the
aggregation behavior of the species involved in asymmetric
catalysis. The presence of NLE has been observed in many
BINOL/Ti-catalyzed enantioselective reactions (e.g., the gly-
oxylate-ene reactions,25 the allylation26 and aldol condensation
of aldehydes,27 sulfoxidations,10e,f and Diels-Alder reac-
tions11d). Although the exact reason for observation of NLE in
these BINOL/Ti catalytic systems varies from case to case, the
simultaneous presence of various oxo- or hydroxo-bridged
diastereomeric dimeric or oligomeric BINOL/Ti species with
different thermodynamic stabilities and kinetic reactivities,
formed by partial hydrolysis or under the influence of molecular
sieves, has been proposed to be responsible for the origin of
NLE.11,28
Following the general procedure outlined in the previous
section, the NLE of the present reaction was examined with
BINOL of varying ee’s under the otherwise identical conditions
as above. As can be seen by the convex line plot of the data in
Figure 3, positive NLE was observed for this catalytic system,
which obviously indicates that diastereomeric complexes are
formed either in the catalytic cycle or at its periphery. In other
(24) (a) Girard, C.; Kagan, H. B. Angew. Chem., Int. Ed. 1998, 37, 2923–
2959. (b) Blackmond, D. G. Acc. Chem. Res. 2000, 33, 402–411. (c)
Kagan, H. B. AdV. Synth. Catal. 2001, 343, 227–233. (d) Kagan, H. B.;
Luukas, T. O. In ComprehensiVe Asymmetric Catalysis; Jacobsen,
E. N., Dfaltz, A., Yamamoto, H., Eds.; Springer: Berlin, 1999; Vol.
1, pp 101-118. (e) Heller, D.; Drexler, H. J.; Fischer, C.; Buschmann,
H.; Baumann, W.; Heller, B. Angew. Chem., Int. Ed. 2000, 39, 495–
499.
(25) (a) Terada, M.; Mikami, K.; Nakai, T. J. Chem. Soc., Chem. Commun.
1990, 1623–1624. (b) Mikami, K.; Terada, M. Tetrahedron 1992, 48,
5671–5680.
NLE Study. Given the pronounced effects of BINOL/Ti ratios
and water on the activity/enantioselectivity observed in the
aminolysis reaction, we proceeded to investigate the potential
aggregation behavior of the titanium species through an NLE
study. NLE has been established as an ubiquitous phenomenon
in enantioselective catalysis and can usually be ascribed to the
presence of diastereomeric homochiral or heterochiral catalytic
species which are generated from the self-assembly of the metal
ions with more than one chiral ligand of the same or opposite
(26) (a) Keck, G. E.; Krishnamurthy, D.; Grier, M. C. J. Org. Chem. 1993,
58, 6543–6544. (b) Faller, J. W.; Sams, D. W. I.; Liu, X. J. Am. Chem.
Soc. 1996, 118, 1217–1218. (c) Costa, A. L.; Piazza, M. G.; Tagliavini,
E.; Trombini, C.; Umani-Ronchi, A. J. Am. Chem. Soc. 1993, 115,
7001–7002. (d) Gauthier, D. R.; Carreira, E. M. Angew. Chem., Int.
Ed. 1996, 35, 2363–2365.
(27) Soriente, A.; De Rosa, M.; Villano, R.; Scettri, A. Curr. Org. Chem.
2004, 8, 993–1007.
(28) See ref 11b and the following refs: (a) Terada, M.; Matsumoto, Y.;
Nakamura, Y.; Mikami, K. Chem. Commun. 1997, 281–282. (b)
Mikami, K.; Matsumoto, Y. Tetrahedron 2004, 60, 7715–7719.
9
J. AM. CHEM. SOC. VOL. 130, NO. 31, 2008 10119

A R T I C L E S
Bao et al.
equilibrating aggregates or oligomeric species. The isopropoxide
methine proton signal at δ 4.53 of Ti(OiPr)4 (spectrum a) was
shifted upfield to 3.65 (spectrum b), suggesting the complete
protonation of isopropoxide by BINOL and its release from the
Ti binding to become the free isopropyl alcohol.30 With the
addition of 10 equiv of benzylamine [relative to Ti(OiPr)4] and
aging the system for 30 min, however, the complex proton
signals at the aromatic region became much more well-resolved
with significant simplification of the spectrum (spectrum c),
indicating that addition of benzylamine promotes a convergence
of a manifold of species into a limited number of species (this
is in accord with the ESI-MS data that only a limited number
of species evolved from a complex spectrum upon addition of
BnNH2, vide infra). Visual inspection of the solution in the
NMR tube showed that the dark red BINOL/Ti solution turned
yellow immediately upon addition of benzylamine, suggesting
the coordination of the benzylamine to titanium occurred in the
system. Further addition of 15 equiv of water [relative to
Ti(OiPr)4] to the system did not result in much alteration in the
appearance of the spectrum (spectrum d), suggesting that there
were no appreciable changes in the solution structures of the
Ti/BINOL species, implying water is not necessary for the
formation of the active catalyst at least up to this stage, as it
did not cause any obvious structural reorganization of the
catalyst by actions such as hydrolysis.
ESI-MS Analysis of the Catalytic System before Addition
of the Epoxide. To further probe the nature of the various
titanium species equilibrating in solution under the reaction
conditions and to get information on their evolution to the active
species, ESI-MS was employed as an analytical tool to assess
the reaction system before and after the addition of the epoxide.
ESI-MS has been applied as a powerful tool to characterize
inorganic and organometallic species in solution, and plenty of
precedents can be found in the literature where ESI-MS has
been applied to the identification of different polynuclear metal
complexes in rapid exchanging systems.31 Furthermore, ESI-
MS has also been successfully used to understand the unique
nature of the catalytic species in a variety of reactions.32 For
ESI-MS characterization of the present catalytic system before
the addition of the epoxide, the spectra were collected in
Figure 4. 1H NMR spectra taken in C6D6. (a) Ti(OiPr)4, (b) [(R)-BINOL/
Ti(OiPr)4 (2:1)], (c) [(R)-BINOL/Ti(OiPr)4/BnNH2 (2:1:10)], (d) [(R)-
BINOL/Ti(OiPr)4/BnNH2/H2O (2:1:10:15)], and (e) (R)-BINOL.
words, the pronounced (+)-NLE usually suggests the involve-
ment of two or more BINOL ligands in the formation of active
or inactive titanium species, and, if interpreted using Kagan’s
ML2 model (vide infra),24a,29 the heterochiral BINOL/Ti com-
plexes should be less reactive than any of the homochiral ones.
Indeed, when the rac-BINOL/Ti(OiPr)4 (2:1) system was used,
the reaction rate was lower than that of the enantiopure (R)-
BINOL/Ti(OiPr)4 (2:1) system under otherwise identical condi-
tions (Figure 1).
To probe the possibility of the involvement of enantioenriched
amino alcohol product in the active species, the variation of
the enantioselectivity with time was determined by using
isopropylamine instead of benzylamine as the nucleophile in
the aminolytic reaction of 1, since in the former case the product
ee could be easily determined by chiral GC without the need
of further derivatization. The experiment was performed under
identical conditions as above using enantiopure BINOL, and
the ee’s of the product were monitored during the reaction course
by removing aliquots at varying time durations. The reaction
was completed at 8 h, and the ee’s were found to be constant
over the whole reaction course. The absence of ee variations
during the reaction makes it unlikely that the amino alcoholic
product is involved in the catalytically active species. Instead,
the enantiopure BINOL ligand must be the sole cause respon-
sible for the efficient enantiocontrol of the reaction.
(30) Gottlieb, H. E.; Kotlyar, V.; Nudelman, A. J. Org. Chem. 1997, 62,
7512–7515.
(31) For reviews, see: (a) Colton, R.; D’Agostino, A.; Traeger, J. C. Mass
Spectrom. ReV. 1995, 14, 79–106. (b) Henderson, W.; Nicholson,
B. K.; McCaffrey, L. J. Polyhedron 1998, 17, 4291–4313. (c) Traeger,
J. C. Int. J. Mass Spectrom. 2000, 200, 387–401. (d) Plattner, D. A.
Int. J. Mass Spectrom. 2001, 207, 125–144. (e) Plattner, D. A. Top.
Curr. Chem. 2003, 225, 153–203. (f) Di Marco, V. B.; Bombi, G. G.
Mass Spectrom. ReV. 2006, 25, 347–379.
(32) For examples, see: (a) Adam, W.; Mock-Knoblauch, C.; Saha-Moller,
C. R.; Herderich, M. J. Am. Chem. Soc. 2000, 122, 9685–9691. (b)
Bonchio, M.; Licini, G.; Modena, G.; Bortolini, O.; Moro, S.; Nugent,
W. A. J. Am. Chem. Soc. 1999, 121, 6258–6268. (c) Enquist, P. A.;
Nilsson, P.; Sjoberg, P.; Larhed, M. J. Org. Chem. 2006, 71, 8779–
8786. (d) Gagliardo, M.; Chase, P. A.; Brouwer, S.; van Klink,
G. P. M.; van Koten, G. Organometallics 2007, 26, 2219–2227. (e)
Maheswari, P. U.; de Hoog, P.; Hage, R.; Gamez, P.; Reedijk, J. AdV.
Synth. Catal. 2005, 347, 1759–1764. (f) Nagataki, T.; Tachi, Y.; Itoh,
S. J. Mol. Catal. A: Chem. 2005, 225, 103–109. (g) Santos, L. S.;
Rosso, G. B.; Pilli, R. A.; Eberlin, M. N. J. Org. Chem. 2007, 72,
5809–5812. (h) Taccardi, N.; Paolillo, R.; Gallo, V.; Mastrorilli, P.;
Nobile, C. F.; Raisanen, M.; Repo, T. Eur. J. Inorg. Chem. 2007,
4645–4652. (i) Vicent, C.; Viciano, M.; Mas-Marza, E.; Sanau, M.;
Peris, E. Organometallics 2006, 25, 3713–3720. (j) Guo, H.; Qian,
R.; Liao, Y.; Ma, S. M.; Guo, Y. L. J. Am. Chem. Soc. 2005, 127,
13060–13064. (k) Qian, R.; Guo, H.; Liao, Y.; Guo, Y.; Ma, S. Angew.
Chem., Int. Ed. 2005, 44, 4771–4774. (l) Shibasaki, M.; Kanai, M.
Org. Biomol. Chem. 2007, 5, 2027–2039. (m) Suzuki, M.; Kato, N.;
Kanai, M.; Shibasaki, M. Org. Lett. 2005, 7, 2527–2530.
1H NMR Characterization of Catalyst System. Until now little
was known about the nature of the catalytic entity responsible
for the enantioselective reaction. The room-temperature 1H
NMR spectra of the catalytic system (Figure 4) before the
addition of the epoxide was taken to provide a preliminary
assessment of the nature of the various titanium species
equilibrating in solution (under the reaction conditions) before
the reaction. As shown in Figure 4, upon mixing 2:1 (R)-BINOL/
1
Ti(OiPr)4 in C6D6, we saw that the H NMR spectrum of the
mixture shows broad bulgelike resonances at the aromatic region
(spectrum b), indicating the presence of a mixture of slowly
(29) Guillaneux, D.; Zhao, S.-H.; Samuel, O.; Rainford, D.; Kagan, H. B.
J. Am. Chem. Soc. 1994, 116, 9430–9439.
9
10120 J. AM. CHEM. SOC. VOL. 130, NO. 31, 2008

BINOL/Ti(IV)-Catalyzed Ring-Opening Aminolysis
A R T I C L E S
Figure 5. ESI-MS spectra of the catalyst systems before the addition of the epoxide: (a) (R)-BINOL/Ti(OiPr)4/BnNH2 (2:1:10), (b) experimental isotopic
distribution and theoretical simulation of the peak at m/z ) 724, (c) (R)-BINOL/Ti(OiPr)4/Bn15NH2 (2:1:10), and (d) (R)-3,3′,6,6′-D4-BINOL/Ti(OiPr)4/
BnNH2 (2:1:10).
positive-ion mode by stepwise diluting the toluene solutions of
(R)-BINOL/Ti(OiPr)4 (2:1), (R)-BINOL/Ti(OiPr)4/BnNH2 (2:
1:10), or (R)-BINOL/Ti(OiPr)4/BnNH2/H2O (2:1:10:15), re-
spectively, in acetonitrile. Tandem mass spectrometry (MS/MS)
analysis provided useful information on the relative metal-ligand
bond strengths. The simulation of the observed isotopic distribu-
tions of the characteristic well-resolved peaks and the use of
isotope-labeled benzylamine or BINOL gave an unambiguous
identification of all the Ti-containing ions in the system. A
combined application of these ESI-MS techniques provided clear
insight into the structural elucidation of the Ti(IV) complexes
under study. First, the positive ion spectrum collected for the
2:1 mixture of (R)-BINOL/Ti(OiPr)4 showed a complex pattern
with a multitude of poorly resolved peaks (Figure S1 in
Supporting Information (SI)), indicating the presence of many
titanium species which is in keeping with the NMR result and
the well-known complexity of BINOL/Ti(OiPr)4 mixtures.4–12
Then, adding 10 equiv [relative to that of Ti(OiPr)4] of
benzylamine to the toluene solution of (R)-BINOL/Ti(OiPr)4
(2:1) and aging the system for 30 min resulted in a dramatic
evolution of the ESI-MS spectrum: as shown in Figure 5a, only
four intense peaks at m/z ) 724 (C47H34NO4Ti+), 1010 (C67H48-
NO6Ti+), 1117 (C74H57N2O6Ti+), and 1224 (C81H66N3O6Ti+),
compositionally corresponding to [(BINOLate)2Ti + BnNH2 +
H]+ (I), [(BINOLate)2Ti + BnNH2 + BINOL + H]+ (II),
[(BINOLate)2Ti + 2BnNH2 + BINOL + H]+ (III), and
[(BINOLate)2Ti + 3BnNH2 + BINOL + H]+ (IV), respectively,
predominate in the clean spectrum. This is in agreement with the
corresponding NMR result discussed above. Isotopic pattern
simulation of the base peak at m/z ) 724 agreed well with the
observed isotopic distribution (Figure 5b), thus unambiguously
establishing the elemental composition of the observed ions. Further
evidence of the number of benzylamine and BINOL molecules
present in each of these ions came from the ESI-MS of a similar
system using 15N-labeled benzylamine or 3,3′,6,6′-teradeuterium-
labeled (R)-1,1′-bi-2-naphthol ((R)-3,3′,6,6′-D4-BINOL) in catalyst
preparation, respectively (Figure 5c,d). It is interesting to note that
all four detected species contain only one Ti atom in their formula,
despite the well-known tendency of aggregation for titanium
complexes. Remarkably, further addition of 15 equiv of water to
the above system [(R)-BINOL/Ti(OiPr)4/BnNH2 ) 2:1:10] did not
result in any significant changes in the ESI-MS spectrum (Figure
S2a in SI), suggesting that water does not play any role in new
species (or precatalyst) formation despite the significant amount
of water relative to Ti. This is somewhat surprising in that most of
the Ti/alkoxide systems reported thus far demonstrate a facile
tendency to hydrolysis, forming oxo-, hydroxo-, or water-containing
(or bridged) multinuclear (or dinuclear) Ti species (aggregates) in
the presence of water.
Tandem MS spectroscopic analysis of ions I-IV was
performed to ascertain the relative metal-ligand bond strengths
and to confirm the assignment of their identities outlined above
(Figure 6). MS/MS analysis shows that the base peak ion I
fragments by losing one benzylamine molecule (724 - 617 )
107 Da); such a fragmentation pattern is strong evidence for
the coordination of a benzylamine to the titanium center
surrounded by two BINOLates (Figure 6a). Similarly, the MS/
MS spectrum of ion II (m/z ) 1010) (Figure 6b) collected under
identical experimental conditions only fragments to ion I,
resulting from the loss of one molecule of BINOL (286 Da),
indicating a distinct difference in bonding strengths exists among
the three BINOL(ate) moieties and titanium in the ion. One
BINOL molecule only interacts relatively weakly with the rest
part of the complex. Remarkably, the MS/MS spectra of ions
III (m/z ) 1117) (Figure 6c) and IV (m/z ) 1224) (Figure 6d)
collected under the same experimental conditions produce only
9
J. AM. CHEM. SOC. VOL. 130, NO. 31, 2008 10121

A R T I C L E S
Bao et al.
Figure 6. MS/MS spectra of the major ions I-IV observed in Figure 5.
fragments corresponding to ions with the loss of one or two
molecule(s) of benzylamine. The persistent occurrence of the
common structural feature [(BINOLate)2Ti] strongly indicates
that the presumably double-chelated core structure remained
intact through the electrospray process, whereas the third
relatively labile BINOL in weak peaks II, III, and IV should
be linked to the core through weak interactions such as
H-bonding.33 It is worth noting here that although a Ti complex
with the composition [(BINOL)Ti(BINOLate)(OiPr)2] somewhat
similar to that of [(BINOLate)2Ti] was proposed by Mikami
and Matsukawa in 1997 on the basis of 13C NMR determina-
tion,34 the MS data discussed above is the first unequivocal
evidence of the species.
Furthermore, ESI-MS spectrum of Inaba et al.’s catalyst
system for the same reaction [(R)-BINOL/Ti(OiPr)4/BnNH2/H2O
(1:1:10:15)] (Figure S2b in SI) is very similar to that in Figure
5a without the appearance of new peaks, suggesting that the
same mechanism is operating in both systems [(R)-BINOL/
Ti(OiPr)4 ) 2:1 or 1:1], and the rate difference observed in
Figure 1 for the two systems can be attributed to the difference
in the concentration rather than the identity of the involved active
species. Remarkably, this might also provide an alternative
interpretation for the observed distinction in catalytic perfor-
mance in similar 2:1 and 1:1 diol/Ti systems.
ESI-MS Analysis of the Catalytic System after the Addition
of the Epoxide. Subsequently, ESI-MS determinations were
performed on the reaction system after addition of the epoxide
[(R)-BINOL/Ti(OiPr)4/BnNH2/H2O/epoxide ) 2:1:10:15:10].
The spectra were collected under identical ESI-MS experimental
conditions as above, and the sample was taken immediately after
addition of the epoxide (1 min) to get the information of the
system before completion of the reaction. As shown in Figure
7, a quite clean and simple-looking spectrum demonstrates the
appearance of a number of new species. Data analysis and peak
Figure 7. ESI-MS spectrum of the [BINOL/Ti(OiPr)4/BnNH2/H2O/epoxide
) 2:1:10:15:10] system.
assignment using procedures similar to the last section (for
details, see SI) revealed that all the significant peaks contain a
mononuclear Ti species with different number of BINOL(ate)
units, and all the species except the one at m/z ) 833 can be
assigned to bear a Ti(BINOLate)2 core together with the weak
association of varying numbers of BINOL (0 or 1), benzylamine
(0 to 3), epoxide (0 to 5), and/or the amino alcohol product (0
to 3) molecules. Again, no water- or hydroxo- or oxo-containing
species was observed, confirming the innocent (or spectator)
nature of the water on the catalyst structure formation. The Ti
species coordinated with epoxide alone is not detected in the
reaction mixture, which seems to imply its high reactivity. The
base peak at m/z ) 833 (C48H53N2O8Ti+) compositionally
corresponds to the [(BINOLate)Ti]2+ unit coordinated with one
or two deprotonated amino alkoxide products (3-H+), that is,
[(BINOLate)Ti(BnNH2)(epoxide)(3-H+)]+ or [(BINOLate)Ti +
2(3-H+) + H+]+. MS/MS analysis of this ion (Figure S4 in SI)
demonstrates no fragments corresponding to the loss of either
benzylamine or the epoxide, suggesting that the latter should
(33) Nishimura, A.; Nagahori, N. Angew. Chem., Int. Ed. 2005, 44, 571–
575.
9
10122 J. AM. CHEM. SOC. VOL. 130, NO. 31, 2008

BINOL/Ti(IV)-Catalyzed Ring-Opening Aminolysis
A R T I C L E S
Chart 1. Collection of Chiral Diol Ligands
proposed by Mikami et al.34,35 and has been successfully applied
in a variety of catalytic asymmetric reactions.35–37 The asym-
metric activation can in principle take effect whenever the
enantioselectivity-determining transition state for a catalytic
asymmetric reaction contains the metal M simultaneously
coordinated with two ligands La and Lb (same as or different
from each other). Upon mixing of two such ligands La and Lb
with metal M, two homocombinations MLaLa and MLbLb as
well as the heterocombination MLaLb can be obtained which
are in equilibration with each other in the catalyst solution. When
MLaLb is more active and more selective than either of the
complexes MLaLa or MLbLb, the mixture of all three catalysts
may well lead to an enhanced enantioselectivity than either of
the catalyst MLaLa or MLbLb used alone.36 On the one hand,
asymmetric activation can be used as a powerful strategy for
catalyst optimization; on the other hand, it should also provide
unequivocal evidence that two or more ligands are directly
involved in the enantioselectivity-determining transition state
and thus may well serve our purpose. From this point of view,
we further screened a library of Ti/diol complexes with various
La/Ti(OiPr)4/Lb (molar ratio 1:1:1) combinations in the present
reaction under otherwise identical conditions as above, through
the use of a collection of enantiopure chiral diols depicted in
Chart 1. As can be seen from the results in Table 1, the
heterocombinations of L4/L7 indeed afforded a dramatically
enhanced ee value (92%, entry 29) relative to either of the two
homocombinations (both are 7%, entries 4 and 7). Remarkably,
this improvement in ee can only be achieved in a catalytic cycle
where the enantioselectivity-determining transition state simul-
taneously bears more than one chiral ligand in the active site.
Furthermore, slight improvement in ee values can also be
observed for L1/Ti/L2, L1/Ti/L3, L3/Ti/L6 combinations relative
to their homocounterparts L1/Ti/L1, L2/Ti/L2, L3/Ti/L3, and L6/
Ti/L6 (entries 9, 10, and 24 vs entries 1-3, 6). Similar
Figure 8. Common structural motif 4 in the Ti species observed in the
ESI-MS spectra of the reaction system before or after the addition of
epoxide 1.
be the plausible species corresponding to the peak at m/z )
833. This is unlikely a catalytically active species for the
following reasons. First, if Ti is chelated with two amino alcohol
molecules and one BINOLate, then without ligand dissociation
the complex [BINOLate/Ti/(3)2] would be coordinatively satu-
rated and no longer accessible for epoxide activation. Second,
dissociation of an amino alcohol 3 from this species would
produce a Ti complex (BINOLate/Ti/3) bearing one molecule
of BINOLate and one chiral amino alcohol 3, respectively. Since
the latter is also highly enantioenriched, it is thus expected to
exhibit some perturbation on ee values during the reaction
course, which is in contradiction with the observation mentioned
above (vide supra). The considerable decrease in the relative
peak intensity at m/z ) 724 after the addition of epoxide 1
(compared with the base peak I in Figure 5) suggests that the
complex [(BINOLate)2Ti(BnNH2)] might be the catalyst precur-
sor whose concentration level should decrease significantly
during the reaction course by consumption with the stronger
coordinating epoxide 1 and/or the amino alcohol product 3.
To summarize this section, the potential presence of any
polynuclear titanium species is not detected in the positive ion
ESI-MS of the systems either before or after the addition of
the epoxide, which provides the strong support that water does
not take part in the catalyst formation. Moreover, a mononuclear
Ti coordinated by two BINOLates, Ti(BINOLate)2 (4), is found
to represent the dominant structural feature which is present in
almost all the detected species (Figure 8) and thus might be
persistent throughout the whole catalysis.
Screening of Dynamic Catalyst Library of Complexes La/
Ti/Lb. Although the ESI-MS analysis and the NLE study is
strongly indicative of the involvement of Ti(BINOLate)2-
containing species in the catalysis, it is still not clear whether
it is acting as the protagonist within or simply as a catalyst
precursor at the periphery of the catalytic cycle. Here, the
observation of synergistic effect in the present Ti(IV)-catalyzed
reaction through the combinatorial use of binary mixtures of
chiral diol ligands provided the ultimate proof in favor of the
direct involvement of Ti(BINOLate)2-containing species in the
catalytic cycle. The concept of asymmetric activation was
(35) For a review on asymmetric activation, see: Mikami, K.; Tereda, M.;
Korenaga, T.; Matsumoto, Y.; Ueki, M.; Angeluad, R. Angew. Chem.,
Int. Ed. 2000, 39, 3532–3556.
(36) For other related reviews, see: (a) Mikami, K.; Terada, M.; Korenaga,
T.; Matsumoto, Y.; Matsukawa, S. Acc. Chem. Res. 2000, 33, 391–
401. (b) Mikami, K.; Yamanaka, M. Chem. ReV. 2003, 103, 3369–
3400. (c) Faller, J. W.; Lavoie, A. R.; Parr, J. Chem. ReV. 2003, 103,
3345–3367. (d) Gennari, C.; Piarulli, U. Chem. ReV. 2003, 103, 3071–
3100. (e) Walsh, P. J.; Lurain, A. E.; Balsells, J. Chem. ReV. 2003,
103, 3297–3344. (f) Ding, K.; Du, H.; Yuan, Y.; Long, J. Chem.-
Eur. J. 2004, 10, 2872–2884. (g) de Vries, J. G.; Lefort, L. Chem.-
Eur. J. 2006, 12, 4722–4734. (h) Ding, K. Chem. Commun. 2008,
909–921.
(34) (a) Mikami, K.; Matsukawa, S. Nature (London) 1997, 385, 613–
617. (b) Matsukawa, S.; Mikami, K. Enantiomer 1996, 1, 69–73. (c)
Matsukawa, S.; Mikami, K. Tetrahedron: Asymmetry 1997, 8, 815–
816. (d) Mikami, K.; Matsukawa, S.; Volk, T.; Terada, M. Angew.
Chem., Int. Ed. 1998, 36, 2768–2771.
(37) For a more recent review, see: (a) Mikami, K.; Aikawa, K. In New
Frontiers in Asymmetric Catalysis; Mikami, K., Lautens, M., Eds.;
Wiley-Interscience: Hoboken, NJ, 2007, pp 221-257.
9
J. AM. CHEM. SOC. VOL. 130, NO. 31, 2008 10123

A R T I C L E S
Bao et al.
Table 1. Screening of Various La/Ti/Lb Catalyst Combinations (La
and Lb ) Chiral Diols in Chart 1) in the Ring-Opening Reaction of
1 with Benzylamine
be the underlying structural reason for the observation of the
synergistic effect of two-component ligands.
Proposed Mechanism for the Aminolysis of Epoxide 1
Catalyzed by the BINOL/Ti(OiPr)4/H2O System. On the basis
1
of the afore-discussed reaction profile studies, H NMR and
ESI-MS analyses, and the beneficial synergistic effect of the
binary ligands exhibited in dynamic catalyst library of com-
plexes La/Ti/Lb, we propose a plausible mechanism for the
enantioselective aminolysis of epoxide 1 by benzylamine using
the BINOL/Ti(OiPr)4/H2O catalyst system. As outlined in
Scheme 2, mixing of BINOL and Ti(OiPr)4 in toluene results
in the formation of a multitude of BINOL/Ti species complexes
equilibrating with each other, which reorganized into a limited
number of mononuclear Ti species (such as 5) bearing the
Ti(BINOLate)2 moiety upon the addition of an excess amount
of benzylamine. The mononulear Ti species 5, containing at
least one coordinated benzylamine for which the exact structure
is still not clear at the present stage, is proposed to be composed
of the titanium coordinated with two chelating BINOLate in
cis configuration. Further addition of a certain amount of water
does not cause any significant structural variation of the observed
Ti species. This apparently exotic behavior can be attributed to
the coordination power of benzylamine, which can not only
resolve the Ti/BINOL mixtures into well-defined structures, but
also effectively suppress the competitive coordination of water
to Ti and therefore prevent the otherwise hydrolysis. Further-
more, this ligating ability of benzylamine might also have some
influences on the coordinating activation of the epoxide 1. Since
Inaba et al. demonstrated that the bichelating coordination of
an epoxide seems to be necessary for the reaction to occur,21 it
is likely that the coordinated benzylamine has to be replaced
by the chelating epoxide 1 via ligand exchange to form complex
6 before sufficient activation of the epoxide can be achieved. It
is notable here that, in the ESI-MS spectrum of the reacting
system (Figure 7), the relative abundance of the peak at m/z )
724 (species 5 in Scheme 2) decreases sharply as compared
with that in Figure 5a, consistent with 5 being further trans-
formed into other species such as 6 during the reaction process.
Intriguingly, the expected peak corresponding to the proposed
active species 6 did not appear in Figure 7, presumably as a
result of its high reactivity (thus low concentration). Now that
Ti-activated epoxide species 6 is ready for the nucleophilic
attack from the amine, then what is the function of water in the
catalysis? From the ESI-MS analyses of the system before and
after the reaction, it is unlikely that water acts as a Lewis base
donor (OH-, O2-, or H2O itself) to take part in the active catalyst
formation even though coordination vacancy (vacant site) seems
to be available on the (BINOLate)2Ti(BnNH2) species. However,
from the reaction profile study there is significant difference in
both reactivity and enantioselectivity in the presence or absence
of water; therefore, water must have taken a part in the rate-
and enantioselectivity-determining transition state. Most prob-
ably, water can affect the reaction by acting as a proton shuttle
to facilitate the catalytic turnover (i.e., as both proton acceptor
and donor) to transfer protons between the amine group and
forming (or formed) amino alkoxide group through hydrogen
bonding with both. The backward attack of the amine on the
epoxide carbon requires some species to assist the accompanying
proton transfer. Although the amine in solution can also
accomplish the proton transfer, it is less efficient than the smaller
amphoteric water which can enter into the active site of the
catalytic assembly. Although this type of behavior has also been
proposed in some water-containing organometallic catalytic
entry
La/Lb
yield (%)
ee (%)a
1
2
3
4
5
6
7
8
L1/L1
L2/L2
L3/L3
L4/L4
L5/L5
L6/L6
L7/L7
L8/L8
L1/L2
L1/L3
L1/L4
L1/L5
L1/L6
L1/L7
L1/L8
L2/L3
L2/L4
L2/L5
L2/L6
L2/L7
L2/L8
L3/L4
L3/L5
L3/L6
L3/L7
L3/L8
L4/L5
L4/L6
L4/L7
L4/L8
L5/L6
L5/L7
L6/L7
L6/L8
L7/L8
96
94
94
26
0
96
95
96
7
8
14
7
12
16
99
95
81
92
90
85
91
96
84
90
94
98
90
95
93
91
95
86
15
32
27
31
12
6
26
98
98
85
94
90
92
80
94
78
96
96
96
92
94
96
98
97
96
28
19
92b
14b
0
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
34
35
36
10
0
47
0
21
a The absolute configuration of all the amino alcohol products was
determined to be (5S, 6R) unless otherwise specified. b The absolute
configuration of all the amino alcohol products was determined to be
(5R, 6S).
enhancement in reactivity can also be found for the L6/Ti/L8
heterocombination as compared with their respective homo-
combinations (entry 35 vs 6 and 8). This is the ultimate proof
that the Ti species bearing the Ti(BINOLate)2 moiety is(are)
the active catalyst(s) within the catalytic cycle, directly partici-
pating in the aminolytic ring-opening of the epoxide. This
interpretation is also supported by the poor reactivity and
enantioselectivity of the L4/Ti(OiPr)4/L4 and L5/Ti(OiPr)4/L5
systems (entries 4 and 5), since the use of 3,3′-disubstituted
BINOLs is expected to be unfavorable for the formation of
Ti(BINOLate)2 moiety which is sterically demanding at the Ti
center. Last but not least, ESI-MS of the catalyst system with
the composition (R)-BINOL/Ti(OiPr)4/(R)-3,3′,6,6′-D4-BINOL
provided unequivocal evidence for the presence of heteroligand
complex along with the corresponding homocombinations in
the mixture (vide infra, Figure 9a). It is therefore conceivable
that, in related La/Ti(OiPr)4/Lb systems, the presence of similar
heterocombination complexes (such as L4/Ti/L7) with an activity
higher than those of their homocombination counterparts should
9
10124 J. AM. CHEM. SOC. VOL. 130, NO. 31, 2008

BINOL/Ti(IV)-Catalyzed Ring-Opening Aminolysis
A R T I C L E S
Figure 9. ESI spectra of (a) (R)-BINOL/Ti(OiPr)4/(R)-3,3′,6,6′-D4-BINOL/BnNH2 (1:1:1:10), and (b) (S)-BINOL/Ti(OiPr)4/(R)-3,3′,6,6′-D4-BINOL/BnNH2
(1:1:1:10).
Scheme 2. Proposed Mechanism for the Asymmetric Ring-Opening Aminolysis of Epoxide 1 Using BINOL/Ti/Water Catalyst System
systems,38 it has never been demonstrated or reported for chiral
diol/Ti systems presumably as a result of the well-known
aggregation behavior of Ti complexes. This finding may also
shed some light on the often elusive chemistry of Ti(IV)-based
catalytic systems where water or molecular sieves (often
containing a certain amount of water) play an important yet
inexplicable role8,11,39 and may help to further enrich the already
extremely versatile titanium/diol chemistry. While the introduc-
9
J. AM. CHEM. SOC. VOL. 130, NO. 31, 2008 10125

A R T I C L E S
Bao et al.
Scheme 3. Plausible Model of NLE in the Title Reaction System
tion of a certain amount of water (5-15 mol%) is beneficial to
the title reaction, the presence of excessive water will more or
less destroy the delicate coordination equilibria and reduce the
concentration of active species by the formation of inactive oxo-
bridged Ti oligomers, leading to a lowering in reactivity. An
optimal balance is reached somewhere in Figure 2, which is in
agreement with the mechanistic interpretation.
A transition-state model 7 was tentatively proposed for the
aminolysis step by taking into account these considerations. The
Ti center in Ti(BINOLate)2 is chelated with the epoxide 1,
which, via the assistance of water as the proton shuttle,
undergoes the backward nucleophilic attack from benzylamine
to form the amino alcohol product bound on the Ti(BINOLate)2
moiety 8. Subsequent dynamic ligand exchange of 8 with either
benzylamine or the epoxide will release the amino alcohol
product 3 with the regeneration of catalytic species 5 or 6, which
can resume the further catalytic cycle.
et al.’s original ML2 model (ca. 0.4). These results in combina-
tion with the absence of the heterochiral species MLRLS as
shown in Figure 9b suggest that Kagan et al.’s original ML2
model might not be applicable to the present system. Although
the origin of NLE can be rationalized in many ways, the
formation of some types of Ti species containing both (R)- and
(S)-BINOL ligands should be essential. Thus, we propose a
tentative explanation similar to Girard and Kagan’s reservoir
model,24a where the active monomeric homochiral complexes
MLRLR (or MLSLS) (active) and their higher homochiral and
heterochiral aggregates (inactive) are taken into account (Scheme
3). The formation of a reservoir of inactive homochiral (Km)
and heterochiral (Kn) dimeric or oligomeric assemblies with Km
< Kn will reduce the relative amount of the less abundant
enantiomer in the catalytically active species and accordingly
result in a positive NLE in the catalysis. In fact, this proposed
model can be considered a more sophisticated version of
Blackmond’s reversible dissociation model,24b with the inactive
higher aggregates, (MLRLR)m, playing the same role as the
MLRLR in the latter, and MLRLR and MLSLS are active in the
present system like the MLR or MLS in Blackmond’s model.
Rationalize the Difference in Reactivity of the Epoxide/
Amine Combinations Catalyzed by the Present System. Sub-
sequently, we attempted to extend the scope of the current
catalysis by examining the ring-opening reactions between
several other couples of amines and epoxides. Although catalytic
enantioselective ring-opening of epoxides by the use of amines
as the nucleophiles represents a very useful approach to the
synthesis of various chiral amino alcohols,40 it often suffers from
an inherent problem of compatibility (i.e., the possible deactiva-
tion of the Lewis acidic catalyst by its irreversible complex
formation with the Lewis basic amine and/or with the generated
ꢀ-amino alcohol).21,40a Perhaps this is the reason that aliphatic
amines are rarely successful in the desymmetrization of various
meso epoxides, which typically involves the use of aromatic
amines (such as aniline or its derivatives) as the nucleophiles.
For comparison purposes, the aminolysis of epoxide 1 and
cyclohexene oxide (CHO) with benzylamine and aniline,41
respectively, was examined using BINOL/Ti(OiPr)4 (2:1) cata-
lyst, and the results are summarized in Table 2. No reactions
were observed for aminolysis of CHO using benzylamine as
the nucleophile, which is consistent with Inaba et al.’s observa-
tion that several tested monodentate meso epoxides did not
undergo the reaction. However, the aminolysis of CHO using
aniline proceeded smoothly under similar conditions. These
According to this proposed mechanism, it seems plausible
that the NLE observed in the reaction can be rationalized using
Kagan et al.’s ML2 model.24,29 By this model, three different
BINOL/Ti complexes, that is, two homochiral combinations (R)-
BINOLate/Ti/(R)-BINOLate and (S)-BINOLate/Ti/(S)-BINO-
Late, and one heterochiral combination (R)-BINOLate/Ti/(S)-
BINOLate (here we temporarily omit the presence of BnNH2
in the discussion for the sake of clarity), can be generated in
situ by mixing racemic or enantioenriched BINOL with Ti(O-
iPr)4 in toluene, and it is tempting to assume that the pronounced
(+)-NLE might be caused by the more stable and less reactive
heterochiral complex. To probe the structural origin of the NLE,
ESI-MS spectrum (Figure 9b) was collected on the catalyst
system prepared from a 1:1 pseudoracemic mixture of (S)-
BINOL and (R)-3,3′,6,6′-D4-BINOL under the otherwise identi-
cal conditions as above so as to minimize the stereoelectronic
effect of ligand composition on the formation of the heterochiral
complex. Surprisingly, while both of the homochiral combina-
tions are clearly distinguishable in Figure 9b [e.g., the peaks at
m/z ) 724 for (S)-BINOLate/Ti/(S)-BINOLate and 732 for (R)-
3,3′,6,6′-D4-BINOLate/Ti/(R)-3,3′,6,6′-D4- BINOLate], the ex-
pected peak at m/z ) 728 for the heterochiral complex (S)-
BINOLate/Ti/(R)-3,3′,6,6′-D4-BINOLate did not appear at all.
The same is true for other groups of peaks in Figure 9b, which
is in contradiction to the prediction by the ML2 model. For
comparison purpose, ESI-MS spectrum (Figure 9a) was also
collected on a catalyst system with compositions identical to
those in Figure 9b except that (R)-BINOL is used instead of
(S)-BINOL. Remarkably, all three expected assemblies (i.e., La/
Ti/La, Lb/Ti/Lb, and La/Ti/Lb) are detected in this case as shown
in Figure 9a, indicating that the BINOL ligands with same
absolute configurations should have a stronger tendency to form
ML2-type complexes in the present reaction system.
According to Blackmond’s approach,24b a kinetic analysis of
the NLE can provide an independent confirmation of the models
of NLE. Kinetic analysis of the NLE for the present system
using the data of reaction rate vs ligand ee (Figure S12) revealed
the racemic system has ca. 67% activity of the enantiopure
system, considerably higher than the value deduced using Kagan
(38) For examples, see: (a) Shi, F.-Q.; Li, X.; Xia, Y.; Zhang, L.; Yu, Z.-
X. J. Am. Chem. Soc. 2007, 129, 15503–15512. (b) Irebo, T.; Reece,
S. Y.; Sjo¨din, M.; Nocera, D. G.; Hammarstro¨m, L. J. Am. Chem.
Soc. 2007, 129, 15462–15464.
(40) (a) Pastor, I. M.; Yus, M. Curr. Org. Chem. 2005, 9, 1–29. (b)
Schneider, C. Synthesis 2006, 3919–3944. (c) Jacobsen, E. N. Acc.
Chem. Res. 2000, 33, 421–431.
(41) (a) Kureshy, R. I.; Singh, S.; Khan, N. U. H.; Abdi, S. H. R.; Suresh,
E.; Jasra, R. V. Eur. J. Org. Chem. 2006, 1303–1309. (b) Kureshy,
R. I.; Singh, S.; Khan, N. U. H.; Abdi, S. H. R.; Agrawal, S.; Mayani,
V. J.; Jasra, R. V. Tetrahedron Lett. 2006, 47, 5277–5279.
(39) (a) Carree, F.; Gil, R.; Collin, J. Org. Lett. 2005, 7, 1023–1026. (b)
Arai, K.; Lucarini, S.; Kobayashi, S. J. Am. Chem. Soc. 2007, 129,
8103–8111.
9
10126 J. AM. CHEM. SOC. VOL. 130, NO. 31, 2008

BINOL/Ti(IV)-Catalyzed Ring-Opening Aminolysis
A R T I C L E S
Table 2. Aminolytic Ring-Opening of Epoxides Catalyzed by
BINOL/Ti(OiPr)4/H2O Systema
system was progressively uncovered on the basis of reaction
profile studies, NLE investigation, 1H NMR and ESI-MS
analysis, and the observation of synergistic effect of component
ligands in catalysts La/Ti/Lb. The titanium species bearing
Ti(BINOLate)2 moiety is unequivocally established as the active
species responsible for the catalysis. Apart from the function
as a nucleophile, benzylamine also helps to reorganize the
catalyst structures and prevent the hydrolytic oligomerization
of titanium species. Water is acting as a cocatalyst in the reaction
to shuttle the proton between reactants, which represents an
unprecedented finding in Ti/diol-mediated catalysis. ESI-MS
detection of the catalyst system prepared from a 1:1 pseudo-
racemic mixture of (S)-BINOL and (R)-3,3′,6,6′-D4-BINOL
indicates the exclusive formation of homochiral Ti((S)-BINO-
Late)2 and Ti((R)-D4-BINOLate)2 complexes without appearance
of any heterochiral [((R)-D4-BINOLate)Ti((S)-BINOLate)] spe-
cies. Accordingly, the origin of (+)-NLE observed in the present
catalytic system is proposed to be caused by the preferential
formation of heterochiral aggregates between Ti((S)-BINOLate)2
and Ti((R)-BINOLate)2 species to that of their corresponding
homochiral counterparts. Furthermore, the reactivity differences
between several couples of epoxide/amine combinations was
rationalized on the basis of the arguments on their relative
coordination strengths. The results from this study are expected
to provide new insights into the often elusive chemistry of
Ti(IV)-based catalytic systems, where water or molecular sieves
(or alcohols, etc.) are found to play an important yet inexplicable
role. Finally, the clarification of the reactivity behavior may
help the search for effective asymmetric Ti(IV) catalysts for
other types of reactions.
entry
amine
epoxide
yield (%)
ee (%)
1
2
3
4
BnNH2
BnNH2
PhNH2
PhNH2
1
96
96
CHO
1
CHO
no reaction
no reaction
90
25(35)b
a The reactions were run in the presence of 1 or 5 mol % of BINOL/
Ti(OiPr)4 (2:1) in the presence of water (15 equiv relative to Ti). b The
value within the parentheses corresponds to the reaction without addition
of water under the otherwise identical conditions.
Scheme 4. Enantioselective Ring-Opening Aminolysis of 1 with
Aliphatic Amines Using the BINOL/Ti/Water Catalytic System
results, coupled with the good reactivity of the epoxide 1 with
benzylamine, suggest that a preferential sequence of coordination
power toward the titanium center should exist for the reactants
(i.e., epoxide 1 > amino alcohol 3 > BnNH2 > CHO > PhNH2).
To achieve an adequate activation, the epoxide (1 or CHO) has
to compete with either the aliphatic amine (BnNH2) or the
aromatic amine (PhNH2) for coordination to the metal site. Here
the competitive coordination of the reaction partners (the amine
and the epoxide) to the Lewis acidic catalyst seems to be the
key factor dictating the reactivity. However, it is not clear at
the moment why no reaction was observed between epoxide 1
and the aniline, whereas the similar reaction with benzylamine
proceeds efficiently under mild conditions.
Finally, the ring-opening of epoxide 1 using aliphatic amines
nBuNH2, iPrNH2, or tBuNH2 proceeded smoothly under the
optimized reaction conditions, affording the corresponding
amino alcohol products in moderate to high yields with excellent
enantioselectivity, irrespective of the steric bulkiness of the alkyl
group near the reactive center (NH2) (Scheme 4). Remarkably,
by using the binary ligand catalyst L1/Ti(OiPr)4/L3 (1:1:1) (Chart
1) at a Ti(OiPr)4 loading of 0.5 mol % under the otherwise
identical conditions as above, the product 3 could be obtained
in 97% yield with 99% ee in a scale of 50 mmol (12.3 g).
Acknowledgment. Financial support from the National Natu-
ral Science Foundation of China (No.20532050, 20423001), the
Chinese Academy of Sciences, the Major Basic Research
Development Program of China (Grant No. 2006CB806106), the
Science and Technology Commission of Shanghai Municipality,
and Merck Research Laboratories is gratefully acknowledged.
This article is dedicated to Prof. Ryoji Noyori on the occasion
of his 70th birthday.
Supporting Information Available: Detailed experimental
procedures for (1) the ring-opening aminolysis of the epoxides
with amines, (2) reaction profile measurements, (3) NLE study,
1
(4) measurements of H NMR and ESI-MS spectra of catalyst
systems, and (5) chiral GC or HPLC analyses of the products.
This material is available free of charge via the Internet at http://
pubs.acs.org.
Conclusions
In summary, the mechanism of the asymmetric ring-opening
aminolysis of the epoxide 1 catalyzed by the BINOL/Ti/H2O
JA801847R
9
J. AM. CHEM. SOC. VOL. 130, NO. 31, 2008 10127