Asymmetric hydrogenation of trisubstituted olefins with iridium-phosphine thiazole complexes: A further investigation of the ligand structure

Article abstract of DOI:10.1021/ja057178b

New chiral bidentate phoshine thiazoles have been prepared and successfully applied as ligands in the homogeneous iridium-catalyzed asymmetric hydrogenation of aryl alkenes and aryl alkene esters. The ligands are designed to be highly modular and have one common chiral intermediate, from which diversity can be introduced at a late stage in the synthetic pathway. It was found that a six-member-ring backbone of the rigid ligand structure was preferred over seven- or five-member rings. In this study it is shown that the substituent pattern of the ligands has a major influence on the stereochemical outcome of the products. By applying the selectivity model proposed in this study, it is possible to match different substrates against different catalysts. In this way, good to excellent enantioselectivity can be obtained for typically difficult substrates. Geometrically different derivatives of α- and β-methyl cinnamic acid ethyl esters were hydrogenated, to demonstrate the validity of the selectivity model and to verify the importance of steric and electronic matching of the catalyst and the substrate.

Full text of DOI:10.1021/ja057178b

Published on Web 02/11/2006
Asymmetric Hydrogenation of Trisubstituted Olefins with
Iridium-Phosphine Thiazole Complexes: A Further
Investigation of the Ligand Structure
Christian Hedberg,† Klas Ka¨llstro¨m,† Peter Brandt,‡ Lars Kristian Hansen,§ and
Pher G. Andersson*,†
Contribution from the Organic Chemistry, Department of Chemistry, Uppsala UniVersity,
Box 599, 751 24 Uppsala, Sweden, Department of Chemistry, BioVitrum AB, 112 76 Stockholm,
Sweden, and Institute of Chemistry, Faculty of Mathematics-Natural Sciences,
UiT, 9037 Tromsø, Norway
Received October 21, 2005; E-mail: Pher.Andersson@kemi.uu.se.
Abstract: New chiral bidentate phoshine thiazoles have been prepared and successfully applied as ligands
in the homogeneous iridium-catalyzed asymmetric hydrogenation of aryl alkenes and aryl alkene esters.
The ligands are designed to be highly modular and have one common chiral intermediate, from which
diversity can be introduced at a late stage in the synthetic pathway. It was found that a six-member-ring
backbone of the rigid ligand structure was preferred over seven- or five-member rings. In this study it is
shown that the substituent pattern of the ligands has a major influence on the stereochemical outcome of
the products. By applying the selectivity model proposed in this study, it is possible to match different
substrates against different catalysts. In this way, good to excellent enantioselectivity can be obtained for
typically difficult substrates. Geometrically different derivatives of R- and â-methyl cinnamic acid ethyl esters
were hydrogenated, to demonstrate the validity of the selectivity model and to verify the importance of
steric and electronic matching of the catalyst and the substrate.
Introduction
iridium-catalyzed asymmetric hydrogenation is still highly sub-
strate dependent and the development of efficient chiral ligands
that tolerate a broader range of substrates remains a challenge.
Pfaltz has developed his original ligand structure to obtain better
enantioselectiveties with a broader range of substrates.9a-i
In this publication, we report a new class of iridium phosphine
thiazole complexes, which are highly enantioselective for a wide
range of substrates. Recently, a number of papers have appeared
in the literature dealing with the mechanism of the Ir-catalyzed
asymmetric hydrogenation of aryl alkenes.10-12 Since little is
known about the mechanism or the enantioselectivity determin-
Enantioselective hydrogenation is one of the most powerful
methods in asymmetric catalysis.1 While ruthenium- and
rhodium-catalyzed asymmetric hydrogenations of chelating
olefins have a long history,2 unfunctionalized olefins still
represent a challenging class of substrates. Early work in this
field has mainly focused on chiral metallocene complexes.3a-e
During the past few years, Pfaltz and co-workers have success-
fully developed and used phoshine and phosphinite oxazoline
Ir complexes for the asymmetric hydrogenation of weakly
coordinating olefins4 and imines,5 subsequently followed by
others.6a-j Recently there have been more structural variations
in ligand design such as the oxazoline carbene ligand structure
of Burgess and co-workers and the use of a non-carbon chiral
sulfoxide phosphine ligand reported by Ellman.7,8 However,
(6) (a) Hou, D,-R.; Reibenspies, J. H.; Colacot, T. J.; Burgess, K. Chem.s
Eur. J. 2000, 7, 5391. (b) Powell, M. T.; Hou, D.-R.; Perry, M. C.; Cui,
X.; Burgess, K. J. Am. Chem. Soc. 2001, 123, 8878. (c) Bernardinelli, G.
H.; Kundig, E. P.; Pfaltz, A.; Radkowski, K.; Zimmermann, N.; Neuburger-
Zehnder, M. HelV. Chim. Acta 2001, 84. 3233. (d) Tang, W.; Wang, W.;
Zhang, X. Angew. Chem., Int. Ed. 2003, 42, 943. (e) Bunlaksananusorn,
T.; Polborn, K.; Knochel, P. Angew. Chem., Int. Ed. 2003, 42, 3941. (f)
Bolm, C.; Focken, T.; Raabe, G. Tetrahedron: Asymmetry 2003, 14, 1733.
(g) Xu, G.; Gilbertson, S. R. Tetrahedron Lett. 2003, 953. (h) Cozzi, P.
G.; Menges, F.; Kaiser, S. Synlett 2003, 833. (i) Liu, D.; Tang, W.; Zhang,
X. Org. Lett. 2004, 4, 513. (j) For a recent review see: Cui, X.; Burgess,
K. Chem. ReV. 2005, 105, 3272.
† Uppsala University.
‡ Biovitrum AB.
§ UiT.
(1) Blaser, H.-U.; Schmidt, E. Asymmetric Catalysis on Industrial Scale, 1st
ed.; Wiley & Sons: New York, 2004.
(2) Brown, J. M. In ComprehensiVe Asymmetric Catalysis; Jacobsen, E. N.,
Pfaltz, A., Yamamoto, H., Eds.; Springer-Verlag: Berlin, 1999; Vol. I, pp
121-182.
(7) Perry, M. C.; Cui X.; Powell, M. T.; Hou, D.-R.; Reibenspies, J. H. Burgess,
K. J. Am. Chem. Soc. 2003, 125. 113.
(3) (a) Cesarotti, E.; Ugo, R.; Kagan, H. B. Angew. Chem., Int. Ed. Engl. 1979,
18, 10. (b) Halterman, R. K.; Vollhardt, K. P. C.; Welker, M. E.; Blaeser,
D.; Boese, R. J. Am. Chem. Soc. 1987, 109, 8105. (c) Paquette, L. A.;
McKinney, J. A.; McLaughlin, M. L.; Rheingold, A. L. Tetrahedron Lett.
1986, 27, 5599. (d) Waymouth, R.; Pino, P. J. Am. Chem. Soc. 1990, 112,
4911. (e) Broene, R. D.; Buchwald, S. L. J. Am. Chem. Soc. 1993, 115,
12569.
(8) Schenkel, L. B.; Ellman, J. A. J. Org. Chem. 2004, 69, 1800.
(9) (a) Hilgraf, R.; Pfaltz, A. Synlett 1999, 1814. (b) Cozzi, P. G.; Zimmermann,
N.; Hilgraf, R.; Schaffner, S.; Pfaltz, A. AdV. Synth. Catal. 2001. 343, 450.
(c) Blankenstein, J.; Pfaltz, A. Angew. Chem., Int. Ed. 2001, 40, 4445. (d)
Menges, F.; Neuburger, M.; Pfaltz, A. Org. Lett. 2002, 4, 4713. (e) Menges,
F.; Pfaltz, A. AdV. Synth. Catal. 2002, 344, 40. (f) Durry, W. J., III;
Zimmerman, N.; Keenan, M.; Hayashi, M.; Kaiser, S.; Goddard, R.; Pfaltz,
A. Angew. Chem., Int. Ed. 2004, 43, 70. (g) Smidt, S. P.; Menges, F.;
Pfaltz, A. Org. Lett. 2004, 6, 2023. (h) Smidt, S. P.; Menges, F.; Pfaltz, A.
Org. Lett. 2004, 6, 3653. (i) Hilgraf, R.; Pfaltz, A. AdV. Synth. Catal. 2005,
347, 61.
(4) Lightfoot, A.; Schnider, P.; Pfaltz, A. Angew. Chem., Int. Ed. 1998, 37,
2897
(5) Schnider, P.; Koch, G.; Pretot, R.; Wang, G.; Bohnen, F. M.; Kruger, C.;
Pfaltz, A. Chem.sEur. J. 1997, 3, 887.
9
10.1021/ja057178b CCC: $33.50 © 2006 American Chemical Society
J. AM. CHEM. SOC. 2006, 128, 2995-3001
2995

A R T I C L E S
Hedberg et al.
Figure 1. Calculated complex structures, from left (a) oxazole complex and then thiazole complexes with (b) five-, (c) six-, and (d) seven-membered cyclic
backbones.15
Scheme 1. Going from Oxazole to Thiazole Increasing both
Ligand Stability and Late Stage Diversification of the Structure
ing factors of asymmetric iridium-catalyzed hydrogenations, our
group recently undertook a computational and kinetic study of
the hydrogenation of (E)-1,2-diphenyl-1-propene with Pfaltz
iridium-PHOX complex.12 Encouraged by the results from our
previous studies, we became interested in further investigating
the structure selectivity relationship of these systems. We
therefore decided to develop a new set of ligands that relies on
the same ligand principle that we previously have reported.13
These new ligands are highly modular and allow for the late
stage diversity of the ligand. The variations in the ligands can
be used to study how structural changes influences the enan-
tioselectivity of the hydrogenation. Moreover the selectivity
also desirable to change the labile P-O bond into the more
stable P-CH2 bond and thereby gain a more stable ligand
structure compared to the previously reported oxazole ligand
structure (Scheme 1).13
model derived in this study can be used to rationalize the results.
The preparation of thiazoles is much more expedient than
oxazoles, and introduction of a halogen at the 2-position allows
for more structural variation in the ligand structure.14 It was
Results and Discussion
To systematically vary the critical selectivity governing
factors of the ligand structure, we started to evaluate the
importance of the cyclic backbone by calculating the distance
between the ipso-carbon of the substituent in the 2-position on
the heterocycle and the chelated Ir atom. This was done for
five-, six-, and seven-membered cyclic backbones of the ligand
structure and the previously reported oxazole structure (Figure
1).13 The calculated structures showed a large difference between
(10) Fan, Y.; Cui, X.; Burgess, K.; Hall, M. B. J. Am. Chem. Soc. 2004, 126,
16688.
(11) Dietiker, R.; Chen, P. Angew. Chem., Int. Ed. 2004, 43, 5513.
(12) Brandt, P.; Hedberg, C.; Andersson, P. G. Chem.sEur. J. 2003, 9, 339.
(13) Ka¨llstro¨m, K.; Hedberg, C.; Brandt, P.; Bayer, A.; Andersson, P. G. J.
Am. Chem. Soc. 2004, 126, 14308.
(14) Kikelj, D.; Urleb, U. Sci. Synth. 2002, 11, 627.
9
2996 J. AM. CHEM. SOC. VOL. 128, NO. 9, 2006

Hydrogenation of Olefins with Iridium−Phosphines
A R T I C L E S
Table 1. Comparison between Catalysts 1-3 and the Corresponding Oxazole Complex 413
a Conditions: P ) 10 bar (entry 1) and 50 bar (entries 2-6); all reactions performed at room temperature overnight; catalyst loading 0.5 mol % in all
entries. All reactions were performed in freshly distilled (CaH2, ethanol free) CH2Cl2 as 0.25 M solutions. b Reaction time 2 h.
Scheme 2. Synthesis of Complexes 1-3a
the five- and six-membered ring size and a smaller difference
between the six- and seven-membered rings suggesting that
altering the ring size of the backbone may have a major impact
on the stereochemical outcome. It is also favorable to change
the oxygen atom to a sulfur atom in the heterocyclic backbone
since this results in a shorter distance between the ipso-carbon
and the Ir atom, resulting in a more congested complex.
To investigate any differences among the five-, six-, and
seven-membered cyclic backbones, the ligands were synthesized
and the corresponding Ir complexes evaluated in the hydrogena-
tion of olefins.
Reterosynthetic analysis suggested that bromination in the
least substituted position of cyclic â-keto esters, followed by
Hantzsch condensation with thiobenzamide, should result in the
desired thiazoles.16 Reduction of the ester functionality followed
by preparative chiral chromatography should give chiral alco-
hols, suitable for conversion into phosphinothiazoles (Scheme
2).
a Reagents and conditions: (i) Br2, Et2O, 0 °C, 3 h; (ii) PhC(S)NH2,
MeOH (5c) or EtOH (5a,b), 1 h rt, then reflux over night, and then aqueous
K2CO3; (iii) LiAlH4, THF, 0 °C to rt, yielding alcohols 9-11; (iv)
preparative HPLC, Chiracel OD; (v) TsCl, pyridine, CH2Cl2, 0 °C to rt;
(vi) Ph2P(BH3)H, BuLi, THF, DMF, -78 °C to rt over 16 h, yielding 15-
17; (vii) Et2NH (large excess), 18 h, rt; (viii) [IrCl(COD)]2, CH2Cl2, reflux
30 min and then H2O, NaBArF‚3H2O, rt, 1 h.
Synthesis of the five- (1), six- (2), and seven- (3) membered
cyclic backbone complexes (Scheme 2) starts with treatment
of 5a-c with Br2, under previously established conditions,17
followed by a Hantzsch condensation with PhC(S)NH2 to give
the corresponding thiazole esters 6-8. Reduction of 6-8 with
LiAlH4 gave the corresponding racemic alcohols 9-11 in good
yields, followed by resolution of racemates 9-11 into their
enantiomers by preparative chiral HPLC (Chiracel OD). The
resolved alcohols were then converted to their corresponding
tosylates 12-14, followed by substitution with Ph2P(BH3)Li,
and deprotection with Et2NH gave the free phosphines 18-20,
which were subsequently transformed into their corresponding
Ir complexes 1-3 according to the standard procedure.4
Initial hydrogenation results revealed a difference in selectivi-
ties among complexes 1-3, with complex 2 as the overall most
selective catalyst (Table 1). For substrate 21 (entry 1, Table 1)
a slight effect on enantioselectivity could be detected at 10 bar,
in favor of catalyst 2 over 1 and 3. However, at 50 bar (entry
2, Table 1) the structural differences between the catalysts do
not affect the enantioselectivity for substrate 21; this was also
the case for substrate 22 (entry 3, Table 1). For substrate 23
(entry 4, Table 1), catalyst 2 is superior to 1 and slightly better
than 3. Substrate 24 gave, as expected, close to racemic material
and low conversions (entry 5, Table 1), in agreement with the
proposed selectivity model (Figure 5).
The results in Table 1 show that a six-membered cyclic
backbone is preferable and also the thiazole moiety proved to
be much better than the oxazole; therefore, an improved
synthetic protocol for the enantiomerically pure six-membered
cyclic backbone was developed (Scheme 3).
(15) Geometrical minimization was preformed at the B3LYP/LACVP level of
theory in the Jaguar software from Schro¨dinger LLC.
(16) Hantzsch, A. Ber. Chem. Ges. 1890, 23, 1474.
(17) Wilkening, D.; Mundy, B. P. Synth. Commun. 1984, 14, 227.
9
J. AM. CHEM. SOC. VOL. 128, NO. 9, 2006 2997

A R T I C L E S
Hedberg et al.
Scheme 3. Synthesis of Ir Complexes 2 and 36-38a
25·DBTA with aqueous potassium carbonate gave the enan-
tiopure free base (S)-25. Converting (S)-25 in situ to its hydrogen
bromide salt, followed by diazotation under anhydrous condi-
tions employing t-BuONO in acetonitrile with CuBr2, as halogen
source, gave 26 in good yield.19
Intermediate 26 could then be converted either into bro-
mothiazole alcohol 27 by treatment with 2 equiv of DIBAL at
-40 °C or into the debrominated 2-H-thiazole derivative 28 by
the action of an excess of DIBAL at 0 °C. Compound 27
smoothly underwent Suzuki coupling with PhB(OH)2 according
to a standard protocol, which gave 10 in high yield.20 No
racemization of (S)-10, 26, 27, or 28 could be detected by chiral
HPLC during the synthetic sequence, by comparison of retention
times of previously synthesized racemic material. Alcohols 10
and 28 were then converted into their corresponding tosylates
13 and 29 in good yields. Treatment of tosylates 13 and 29
with Ar2P(BH3)Li at 0 °C in THF, followed by stirring at room
temperature overnight, in DMF, yielded the P-borane-protected
phosphines 16 and 30 to 32 in high yields. At this point the
borane adducts are stable to hydrolysis and oxidation, compared
to the unstable first generation phosphinite ligands. Removal
of the borane-protecting group by stirring in neat Et2NH gave
the deprotected phosphines 19 and 33-35. Complex formation
was performed according to the previously employed protocol,
and complexes 2 and 36-38 were isolated in high yields
(Scheme 3).
All attempts to obtain crystals suitable for single-crystal X-ray
analysis from the aminothiazole-(L)-DBTA salt ((S)-25·DBTA)
failed; a crystalline bis-tosylated derivative of 25 was synthe-
sized instead.21 The absolute configuration of this material was
established by X-ray crystallography and was found to be (S).22
Comparison of the HPLC elution order and optical rotation of
the five-, six-, and seven-membered structures established the
(S) configuration of the five- and seven-membered structures.23
The complexes 2 and 36-38 proved to be highly efficient in
the asymmetric hydrogenation of a wide variety of trisubstituted
olefins (Table 2). As expected, the steric bulk of the heteroaro-
matic substituent is important for achieving high enantioselec-
tivity, clearly seen in substrate 21 where a dramatic drop in
enantioselectivity is observed for complexes 37 and 38, lacking
the phenyl group. Also the size of the phosphine substituents,
occupying the semihindered quadrant, is of importance for
substrates such as 22 and 39 (entries 2, 3, Table 2), where
changing from 2 (phenyl) to 36 (o-tolyl) leads to a dramatic
loss of enantioselectivity in both cases. All complexes showed
a dramatic difference in enantioselectivity between substrates
40 (entry 4, Table 2) and 41 (entry 5, Table 2), caused by the
presence of an electron-donating methoxy group in 40. Allylic
alcohol 43 (entry 7, Table 2) worked exceptionally well for all
complexes 2 and 36-38. Ethyl R-methyl trans-cinnamate 44
(entry 8, Table 2) gave good enantioselectivities only with the
a Reagents and conditions: (i) Br2, Et2O, 0 °C, 3 h, 95%; (ii) thiourea,
absolute EtOH, rt (room temperature), 18 h, 94%; (iii) 10% aqueous K2CO3,
warm (40 °C) CHCl3, yielding rac-25, quantitative; (iv) (L)-DBTA‚H2O,
80:20 MeCN/H2O, 37% over two recrystallizations, +99.5% ee of (S)-
25‚DBTA; (v) 10% aqueous K2CO3, warm (40 °C) CHCl3, quantitative;
(vi) HBr, CuBr2, t-BuONO, MeCN, 0 °C, 2 h, 78%, +99.5% ee; (vii) 2.2
equiv of DIBAL, THF, -40 °C, 1.5 h, 85%, +99.5% ee; (viii) 4 equiv of
DIBAL, THF, -40 to 0 °C, 4 h, 85% of 28, +99.5% ee; (ix) PhB(OH)2,
DPPF‚PdCl2 (5 mol %), K2CO3, toluene/H2O, 80 °C, 3 h, 87%, +99.5%
ee; (x) TsCl, pyridine, 0 °C to rt overnight; (xi) Ar2P(BH3)H, BuLi, THF,
DMF, -78 °C to rt, 16 h, yielding 16 and 30-32; (xii) Et2NH (large excess),
18 h, rt; (xiii) [IrCl(COD)]2, CH2Cl2, reflux 30 min and then H2O,
NaBArF‚3H2O, rt, 1 h.
With the encouraging results from this initial study, the
synthesis of second generation ligands is outlined in the
following way: bromination of ketoester 5a with Br2 in Et2O
at 0 °C gave the corresponding 3-bromo derivative in almost
quantitative yield. Treatment with thiourea in absolute ethanol
at room temperature gave the aminothiazole hydrobromide (rac-
25·HBr) in excellent yield.18 The free base (rac-25) was then
resolved as its (L)-dibenzoyltartrate salt, formed by addition of
(L)-DBTA‚H2O to rac-25 in a carefully controlled volume of
hot 80% aqueous MeCN. The resolution step was optimized in
terms of mass yield and enantioselectivity. The reaction condi-
tions yielding 37% of (S)-25·DBTA proved to be the best
compromise, producing a crystal crop of 95-98% enantiomeric
excess, directly from the racemic mixture. One single recrys-
tallization yielded material of more than 99% enantiomeric
excess. The unwanted enantiomer (R)-25, isolated from the
mother liquids, can be recycled by racemization employing a
catalytic amount of sodium ethoxide in refluxing absolute
ethanol. This is according to our knowledge the first example
of an optical resolution of a chemical compound by salt
formation on a remote aminothiazole moiety. Treatment of (S)-
(19) Okonya, J. F.; Al-Obeidi, F. Tetrahedron Lett. 2002, 43, 7051.
(20) Hodgetts, K. J.; Kershaw, M. T. Org. Lett. 2002, 4, 1363.
(21) See the Supporting Information for experimental details.
(22) Crystallographic data for the bis-tosylate have been deposited with the
Cambridge Crystallographic Data Centre as supplementary publication no.
CCDC 286224 by Professor Lars Kristian Hansen (University of Tromsø).
Copies of the data can be obtained free of charge, on application to CCDC,
12 Union Road, Cambridge CB2 1EZ, U.K. (fax +44-(0)1223-336033 or
e-mail deposit@ccdc.cam.ac.uk).
(23) On the basis of comparison of HPLC (Chiracel OD-H) elution order among
9-11 with elution order of 10, synthesized from the route in Scheme 3.
Also the products from hydrogenations had the same absolute configuration
as products from 2.
(18) Marinko, P.; Kastelic, J.; Krbavcic, A.; Kikelj, D. Tetrahedron Lett. 2001,
42, 8911.
9
2998 J. AM. CHEM. SOC. VOL. 128, NO. 9, 2006

Hydrogenation of Olefins with Iridium−Phosphines
A R T I C L E S
Table 2. Results of Hydrogenation with Complexes 2 and 36-38a
a Absolute configurations of products given in parentheses after the ee-value. Conditions: P ) 50 bar in all entries; all reactions performed at room
temperature for 2 h, except entries 7 and 10 were reactions run overnight instead; catalyst loading 0.5 mol % in all entries, except 7, where 1 mol % was
used. All reactions were performed in freshly distilled (CaH2, ethanol free) CH2Cl2 as 0.25 M solutions.
Figure 2. Schematic representation of electronically neutral trisubstituted olefin coordination to the catalyst.
o-tolyl-complexes 36 and 38. For substrate 24, (entry 10, Table
2), the 2-H-thiazole complexes 37 and 38 prove to be superior,
giving high conversions and excellent enantioselectivities for
the previous very unreactive substrate. Conversely, substrate
23 (entry 9, Table 2) gave excellent conversions and enantio-
selectivities with complexes 2 and 36.
Computational Details. Geometries of all substrates were
calculated with the Jaguar program24 using the B3LYP hybrid
density functional,25 together with the LACVP** ECP and basis
set. Normal-mode analysis revealed one imaginary frequency
for each transition state. LACVP in Jaguar defines a combination
of the LANL2DZ basis set26 for iridium and the 6-31G basis
set for other atoms. LACVP implies the use of an effective core
potential for 60 core electrons of iridium and a (5s,6p,3d)
primitive basis contracted to [3s,3p,2d]. Final energies were
retrieved from single point calculations at B3LYP/LACV3p+**.
LACV3p+** differs from LACVP** by using the 6-311+G**
basis set in place of 6-31G**.
Selectivity Model. To establish a model for the enantiode-
termining step, calculations were performed on the full-sized
catalyst 2, with 39 as the substrate. By addition of a quadrant
scheme over the catalyst with the olefin coordination site
centered over the origin in front of the iridium atom, with the
phosphine ligand to the right and the thiazole amine to the left,
the steric bulk will be oriented into the lower left quadrant. This
schematic description of the catalytic system is depicted in
Figure 2. The bulky substituent in the 2 position of the
heterocycle on the ligand creates an almost perfect pocket;
especially well adapted to host trisubstituted olefins. In case of
trisubstituted olefins the least sterically demanding substituent,
(24) Jaguar 4.2; Schro¨dinger, Inc.: Portland, OR, 1991-2000.
(25) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (b) Lee, C.; Yang, W.;
Parr, R. G. Phys. ReV. B 1988, 37, 785.
(26) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 299.
9
J. AM. CHEM. SOC. VOL. 128, NO. 9, 2006 2999

A R T I C L E S
Hedberg et al.
Figure 3. Calculated enantiodetermining migratory-insertion TS of methyl trans-â-methyl cinnamate: (left) sterically and electronically favored â-addition;
(right) sterically and electronically unfavored R-addition.
Figure 4. Calculated enantiodetermining migratory-insertion TS of trans-R-methyl cinnamate: (left) sterically favored, electronically unfavored R-addition;
(right) sterically unfavored, electronically favored â-addition.
i.e., the proton, will face the steric bulk of the ligand. The
catalyst then offers two open sites, located in a trans position
to each other, and a fourth site facing one of the aryl groups of
the phosphine. This fourth site could accommodate smaller
olefin substituents.
Coordination of a trisubstituted olefin into the empty quadrant
model reveals that the olefin will preferentally coordinate to
the catalyst from the Re side, as the opposite coordination mode
will result in unfavorable interactions between one of the large
trans-substituents with the steric bulk of the ligand. This
determines the stereochemical outcome of the reaction and sets
the facial recognition of an electronically neutral trisubstituted
olefin.
Substrates possessing a polarized double bond, such as R,â-
unsaturated esters, add a complicating electronic effect. DFT
calculations with methyl trans-â-methyl cinnamate show a
strong preference for â-addition of the hydride in the migratory
insertion step (13 kcal mol-1).
The migration of the substrate into the Ir-H bond leads to a
tilt of the C-C double bond toward the hydride. For steric
reasons, this means that migratory insertions into an axial
hydride pointing up will be preferred over insertion into an axial
hydride pointing down, since the former leads to a less congested
transition state where the substrate is moving away from the
close proximity of the phenyl ring on the thiazole. For substrate
23 this results in a TS A that is both sterically and electronically
favored over TS B. As seen in Figure 3, â-addition is much
more favored over R-addition, resulting in a lower energy of
the TS due to minimized steric repulsions and a match electronic
polarization of the substrate. In the case of methyl trans-R-
methyl cinnamate, 44, a â-addition of the hydride will require
a migratory insertion step in a iridium-hydride bond pointing
in the same direction as the steric bulk of the thiazole moiety
(Figure 4, right). Thus, there will be a substantial steric
interaction between substrate and catalyst. For this substrate the
steric and electronic effects will be mismatched and result in a
reduced reaction rate and a deterioration of the enantioselectivity.
Competitive experiments between the two substrates revealed
that 23 is indeed hydrogenated 6.7 times faster than 44.
This gives another tool for predicting enantioselectivity and
reactivity of substrates toward asymmetric hydrogenation by
Ir-(P,N) systems. For successful reaction, a steric and electronic
match between substrate and catalyst is necessary. This fact can
be rationalized by the addition of a direction to the enantiode-
9
3000 J. AM. CHEM. SOC. VOL. 128, NO. 9, 2006

Hydrogenation of Olefins with Iridium−Phosphines
A R T I C L E S
Figure 5. Preferred coordination of a polarized substrate (top) and possible facial coordination of substrates 23 and 44 (bottom).
terminating migratory insertion step, in the quadrant selectivity
model.12 A substrate possessing a conjugated electron-withdraw-
ing group (EWG) will show a strong â-preference for the initial
hydride addition (Figure 5).13 Substrate 44 orients into a
sterically matched but electronically mismatched orientation,
hence the lower selectivities obtained with this substrate
compared to 23. Substrate 23 however possesses perfect overlap
for catalysts 2 and 36 (Figure 5) resulting in exceptionally high
selectivity for those catalysts.
of the substituent on the thiazole ring, which is situated in the
blocked quadrant, was also studied, and it was found that its
presence is important for the stereoselectivity of the catalysts.
Decreasing the bulk of this substituent generally results in
decreased ee of the products with the exception of cis-substrates,
which are better accommodated in catalysts having a small
substituent on the thiazole ring.
Acknowledgment. This work was supported by grants from
the SSF SELCHEM graduate research program and Swedish
Research Council (VR).
Conclusions
In conclusion, we have developed a new type of heterocyclic
(P,N) ligand for the asymmetric Ir-catalyzed hydrogenation of
olefins. The results, in terms of enantioselectivites for typically
difficult substrates, are among the best so far reported. The
results obtained for the second-generation thiazole ligands have
allowed a more detailed selectivity model to be derived,
revealing the relation between steric and electronic effects and
their importance for activity and enantioselectivity. The effect
Supporting Information Available: Experimental procedures
for the preparations of the ligands and complexes, hydrogenation
procedures, characterization data, chiral separation data, com-
putational details, and crystallographic data (CIF). This material
is available free of charge via the Internet at http://pubs.acs.org.
JA057178B
9
J. AM. CHEM. SOC. VOL. 128, NO. 9, 2006 3001