Electronic Control of Chiral Quaternary Center Creation in the Intramolecular Asymmetric Heck Reaction

Article abstract of DOI:10.1021/jo049448z

The Boehringer-Ingelheim phosphinoimidazoline (BIPI) ligands were applied to the formation of chiral quaternary centers in the asymmetric Heck reaction. Several different substrates were examined in detail using more than 70 members of this new ligand class. Hammett relationships were determined through systematic variation of the ligand electronics. All substrates showed essentially the same Hammett behavior, where enantioselectivity increased as the ligands were made more electron-deficient. Ligand optimization has led to catalysts which give the highest enantioselectivities reported to date for these difficult systems.

Full text of DOI:10.1021/jo049448z

Electr on ic Con tr ol of Ch ir a l Qu a ter n a r y Cen ter Cr ea tion in th e
In tr a m olecu la r Asym m etr ic Heck Rea ction
Carl A. Busacca,* Danja Grossbach,† Scot J . Campbell, Yong Dong, Magnus C. Eriksson,
Robert E. Harris, Paul-J ames J ones, J i-Young Kim, J on C. Lorenz, Keith B. McKellop,
Erin M. O’Brien, Fenghe Qiu, Robert D. Simpson, Lana Smith, Regina C. So,
Earl M. Spinelli, J ana Vitous, and Chiara Zavattaro
Departments of Chemical Development and Analytical Sciences, Boehringer-Ingelheim Pharmaceuticals,
Inc., 900 Ridgebury Rd., Ridgefield, Connecticut 06877
cbusacca@rdg.boehringer-ingelheim.com
Received April 3, 2004
The Boehringer-Ingelheim phosphinoimidazoline (BIPI) ligands were applied to the formation of
chiral quaternary centers in the asymmetric Heck reaction. Several different substrates were
examined in detail, using more than 70 members of this new ligand class. Hammett relationships
were determined through systematic variation of the ligand electronics. All substrates showed
essentially the same Hammett behavior, where enantioselectivity increased as the ligands were
made more electron-deficient. Ligand optimization has led to catalysts which give the highest
enantioselectivities reported to date for these difficult systems.
In tr od u ction
catalysis. The principal ligand classes which have been
utilized are the PHOX ligands as employed by Pfaltz and
others4a-e and BINAP and its derivatives.4e-k We have
recently reported the application of a new patented ligand
class, the BIPI ligands, to chiral quaternary center
formation in the AHR.5 We report here full details of the
electronic control of asymmetric induction in four differ-
ent substrates in this reaction using the BIPI ligands.
The asymmetric Heck reaction (AHR) is a powerful
method for the creation of tertiary and quaternary
stereocenters via formation of a new C-C bond. Recent
reviews1 document the utility of this reaction in asym-
metric synthesis, and it has proven to be an excellent
method for the assembly of densely functionalized natu-
ral products.2 The AHR has developed into a very
important synthetic tool since the pioneering work of
Shibasaki and Overman.3
Resu lts a n d Discu ssion
The most successful substrate classes have been those
that lead to formation of tertiary stereocenters, particu-
larly the dihydrofurans, dihydropyrroles, and related
cyclic olefins. Quaternary centers are more difficult to
control and remain an important problem in asymmetric
The BIPI ligands (Figure 1) were designed so that gross
electronic tuning could be achieved by simply varying the
nitrogen substituent R4. Alkyl groups in this position lead
to strongly basic systems, while acyl and sulfonyl R4
groups lead to more neutral ligands. This type of elec-
tronic flexibility is essential if one ligand class is to be
* To whom correspondence should be addressed.
† Current address: Schering AG Process Research, D-13342 Berlin.
(1) (a) Link, J . T. In Organic Reactions; Overman, L. E., Ed.;
Wiley: New J ersey, 2002; Vol. 60, pp 157-534. (b) Shibasaki, M.; Vogl,
E. M. In Comprehensive Asymmetric Catalysis; J acobsen, E. N., Pfaltz,
A., Yamamoto, H., Eds.; Springer-Verlag: Heidelberg, 1999; pp 457-
487. (c) Shibasaki, M.; Miyazaki, F. In Handbook of Organopalladium
Chemistry for Organic Synthesis; Negishi, E., Ed.; Wiley: Hoboken,
2002; Vol. 1, pp 1283-1315. (d) Guiry, P. J .; Hennessy, A. J .; Cahill,
J . P. Top. Catal. 1997, 4, 311-326. (e) Shibasaki, M.; Boden, C. D. J .;
Kojima, A. Tetrahedron 1997, 53 (22), 7371-7395.
(2) (a) For reviews, see: Dounay, A. B.; Overman, L. E. Chem. Rev.
2003, 103 (8), 2945-2963. (b) Shibasaki, M.; Kojima, A.; Shimizu, S.
J . Heterocycl. Chem. 1998, 35 (5), 1057-1064. For recent examples,
see: (c) Trost, B. M.; Thiel, O. R.; Tsui, H.-C. J . Am. Chem. Soc. 2003,
125 (43), 13155-13164. (d) Overman, L. E.; Kodanko, J . J . Angew.
Chem., Int. Ed. 2003, 42 (22), 2528-2531. (e) Overman, L. E.; Peterson,
E. A. Tetrahedron 2003, 59 (35), 6905-6919. (f) Mori, M.; Nakanishi,
M.; Kajishima, D.; Sato, Y. J . Am. Chem. Soc. 2003, 125 (32), 9801-
9807. (g) Artman, G. D., III; Weinreb, S. M. Org. Lett. 2003, 5 (9),
1523-1526. (h) Trost, B. M.; Tang, W. Angew. Chem., Int. Ed. 2002,
41 (15), 2795-2797. (i) Overman, L. E.; Lebsack, A. D.; Stearns, B. A.
J . Am. Chem. Soc. 2002, 124 (31), 9008-9009. (j) Shibasaki, M.;
Honzawa, S.; Mizutani, T. Tetrahedron Lett. 1999, 40 (2), 311-314.
(k) Overman, L. E.; Matsuura, T.; Poon, D. J . J . Am. Chem. Soc. 1998,
120 (26), 6500-6503.
(3) (a) Shibasaki, M.; Sato, Y.; Sodeoka, M. J . Org. Chem. 1989, 54,
4738-4739. (b) Overman, L. E.; Carpenter, N. E.; Kucera, D. J . J . Org.
Chem. 1989, 54 (25), 5846-5848. (c) Shibasaki, M.; Sodeoka, M. J .
Synth. Org. Chem. J pn. 1994, 52, 956-957. (d) Overman, L. E. Pure
Appl. Chem. 1994, 66, 1423-1430.
(4) (a) Pfaltz, A.; Loiseleur, O.; Hayashi, M.; Schmees, N. Synthesis
1997, 1338-1345. (b) Loiseleur, O.; Meier, P.; Pfaltz, A. Angew. Chem.,
Int. Ed. Engl. 1996, 35 (2), 200-202. (c) Hallberg, A.; Ripa, L. J . Org.
Chem. 1997, 62 (3), 595-602. (d) Namyslo, J . C.; Kaufmann, D. E.
Chem. Ber. 1997, 130 (9), 1327-1331. (e) Guiry, P. J .; Kiely, D.
Tetrahedron Lett. 2003, 44 (39), 7377-7380. (f) Ozawa, F.; Hayashi,
T. J . Organomet. Chem. 1992, 428, 267-277. (g) Ozawa, F.; Hayashi,
T.; Kubo, A. Tetrahedron Lett. 1992, 33 (11), 1485-1488. (h) Ozawa,
F.; Hayashi, T.; Kobatake, Y. Tetrahedron Lett. 1993, 34 (15), 2505-
2508. (i) Shibasaki, M.; Kondo, K.; Sodeoka, M. J . Org. Chem. 1995,
60 (14), 4322-4323. (j) Overman, L. E.; Poon, D. J . Angew. Chem.,
Int. Ed. Engl. 1997, 36 (5), 518-521. (k) Shibasaki, M.; Ohrai, K.;
Kondo, K.; Sodeoka, M. J . Am. Chem. Soc. 1994, 116 (26), 11737-
11748. (l) Overman, L. E.; Ashimori, A.; Bachand, B.; Poon, D. J . J .
Am. Chem. Soc. 1998, 120 (26), 6477-6487.
(5) (a) Busacca, C. A. U.S. Patent 6 316 620, 2001; European Patent
1218388. (b) Busacca, C. A.; Grossbach, D.; So, R. C.; O’Brien, E. M.;
Spinelli, E. M. Org. Lett. 2003, 5 (4), 595-598. (c) See also the
Supporting Information.
10.1021/jo049448z CCC: $27.50 © 2004 American Chemical Society
Published on Web 07/10/2004
J . Org. Chem. 2004, 69, 5187-5195
5187

Busacca et al.
plots are nevertheless a concise representation of the
observed electronic effects and the ligand optimization
process. For single reactions, curved Hammett plots gen-
erally signify a change in mechanism,8 yet here introduc-
tion of the R4 substituent on nitrogen simply breaks the
original C2 symmetry of the diamine. In short, the two
aryl rings derived from the diamine are now different,
and linear behavior for the sum of the electronic contri-
butions of these two different rings is not expected in any
case. Entries 16-21 were ligands from non-C2-symmetric
diamines and were created to probe whether the proximal
(close to the metal-ligating imine nitrogen) or distal
aromatic ring was more important to asymmetric induc-
tion. We also hoped to uncover the underlying Hammett
behavior of the two individual rings. As shown in Table
1 (and Figure 4), all of the ligands where R2 and R3 are
different give lower enantioselectivities than the sum
curve (R2 ) R3), clustering over a small range of log er
values, and revealing that the underlying Hammett
behavior is definitely nonlinear. Since we had examined
only one substrate at this stage, it was difficult to know
whether this Hammett behavior was actually significant
or largely random, a point which we shall return to
shortly.
The R4 substituent was examined next, as shown in
entries 22-35 of Table 1. Random Hammett behavior
was observed when the N-benzoyl para substituent was
plotted (not shown), and we concluded that this substitu-
ent was too remote to have an effect on enantioselectivity.
The behavior of ligands where either the nitrogen sub-
stituent or the diamine substituent was alkyl (entries
31-35) proved to be quite interesting. These ligands
always gave the opposite enantiomer from those ligands
derived from aryl diamines or from ligands with electron-
withdrawing nitrogen substituents when the absolute
stereochemistry was the same for each series. We thought
that these electronically different ligands might show
different metal chelation, so X-ray quality crystals of
Pd(II) complexes 3 and 4 were grown. The X-ray crystal
structures of these two complexes are shown in Figures
5 and 6. These two metal complexes, which give the
opposite enantiomer from one another in the AHR, are
remarkably similar in their ground-state structures. Both
chelate the imine nitrogen, and both have one axial and
one equatorial phosphine substituent. Both show an
apparent cooperativity between the axial phosphine
substituent and one of the imidazoline C-substituents in
blocking the alpha face. Each structure shows the N-acyl
group to be coplanar with the imidazoline ring, and the
bond lengths and bond angles between the two are very
similar. In summary, the ground-state geometries do not
reveal the origin of the reversal in facial selectivity
observed.
F IGURE 1. BIPI ligands.
used successfully for a variety of different asymmetric
transformations. This can be illustrated by considering
asymmetric hydrogenation and the asymmetric Diels-
Alder cycloaddition as simple examples.
Asymmetric hydrogenation of olefins, perhaps the most
extensively studied area of asymmetric catalysis, is
achieved almost exclusively with electron-rich phosphines
and phosphites.6 These include the very important
DIPAMP, DuPhos, DIOP, and Zhang ligands. Asym-
metric Diels-Alder cycloadditions, by contrast, are best
carried out with neutral or electron-deficient ligands on
Lewis acidic metals.7 No examination of the electronic
effects on asymmetric induction in the AHR had been
reported, so we chose this transformation to apply the
BIPI ligands to. Our approach was simply to determine
if electronic trends in enantioselection existed and, if so,
to then exploit them.
The ligands were prepared as previously described.5
Triflate 1 was the first substrate examined. Table 1
collects the results for the first 46 ligands screened for
this transformation. The phosphine substituent R1 was
systematically varied, while holding R2 and R3 constant
as phenyl, and R4 constant as 2-naphthoyl, as shown in
entries 1-8. As reported, a Hammett plot of these data
proved to be linear with a positive F (Figure 3), which
was a previously unknown relationship for the AHR.
Most importantly, phosphine substituted with 3,5-dif-
luorophenyl (entry 8) gave oxindole 2 in 78% ee, which
is significantly higher enantioselectivity than both (R)-
BINAP (65% ee, (-)) and (S)-t-Bu-PHOX (46% ee, (+))
give for this substrate. A similar Hammett scan for the
R2/R3 substituents was carried out, as shown in entries
9-21. These ligands were prepared from both C2-sym-
metric (R2 ) R3) and non-C2-symmetric diamines. The
Hammett plot for this series is shown in Figure 4. The
Hammett plot for the ligands where R2 ) R3 describes a
curve, which goes through a minimum at the neutral
H-substituent. Both electron-rich and electron-poor mem-
bers give increased enantioselectivity. The best substitu-
tion for both enantioselectivity and yield was found with
the 3,5-difluoro system.
The correlations observed throughout this work are
essentially “qualitative” in nature, since we are examin-
ing the product of multiple reactions (the steps in the
catalytic cycle) rather than the single reaction examined
in classical studies of electronic effects. The Hammett
We were unable to grow high quality crystals of 5 and
6, so we decided to investigate their metal chelation by
(8) (a) Cho, B. R.; Kim, N. S.; Kim, Y. K.; Son, K. H. J . Chem. Soc.,
Perkin Trans. 2 2000, 7, 1419-1423. (b) Kevill, D. K.; Wang, W. K. F.
J . Chem. Soc., Perkin Trans. 2 1998, 12, 2631-2638. (c) Bauld, N. L.;
Yueh, W. J . Chem. Soc., Perkin Trans. 2 1996, 8, 1761-1766. (d)
Petnehazy, I.; Clementis, G.; J aszay, Z. M.; Toke, L.; Hall, C. D. J .
Chem. Soc., Perkin Trans. 2 1996, 11, 2279-2284. (e) Wayner, D. D.
M.; Huang, Y. J . Am. Chem. Soc. 1994, 116 (5), 2157-2158. (f) Drago,
R. S.; Zoltewicz, J . A. J . Org. Chem. 1994, 59 (10), 2824-2830. (g)
Richard, J . P.; Amyes, T. L. J . Am. Chem. Soc. 1990, 112 (26), 9507-
9512.
(6) For reviews, see: (a) Zhang, X.; Tang, W. Chem. Rev. 2003, 103
(8), 3029-3069. (b) Brown, J . M. In Comprehensive Asymmetric
Catalysis; J acobsen, E. N., Pfaltz, A., Yamamoto, H., Eds.; Springer-
Verlag: Heidelberg, 1999; pp 121-182. (c) Halterman, R. L. In
Comprehensive Asymmetric Catalysis; J acobsen, E. N., Pfaltz, A.,
Yamamoto, H., Eds.; Springer-Verlag: Heidelberg, 1999; pp 183-195.
(7) (a) Evans, D. A.; J ohnson, J . S. In Comprehensive Asymmetric
Catalysis; J acobsen, E. N., Pfaltz, A., Yamamoto, H., Eds.; Springer-
Verlag: Heidelberg, 1999; pp 1177-1235.
5188 J . Org. Chem., Vol. 69, No. 16, 2004

Intramolecular Asymmetric Heck Reaction
TABLE 1. Tr ifla te 1, Liga n d Scr een in g Set
entry
R1
R2
R3
R4
base
PMPa
PMP
PMP
PMP
PMP
PMP
PMP
PMP
PMP
PMP
PMP
PMP
PMP
PMP
PMP
PMP
PMP
PMP
PMP
PMP
PMP
PMP
PMP
PMP
PMP
PMP
PMP
PMP
PMP
PMP
PMP
yield, %
ee, %
1
2
3
4
5
6
7
8
9
4-OMe
4-Me
3,5-Me2
(R)-Ph
(R)-Ph
(R)-Ph
(S)-Ph
(R)-Ph
(R)-Ph
(R)-Ph
(S)-Ph
(R)-Ph
(R)-Ph
(R)-Ph
(S)-Ph
(R)-Ph
(R)-Ph
(R)-Ph
(S)-Ph
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
1-naphthoyl
4-NMe2-Bz
4-OMe-Bz
4-Me-Bz
15
15
82
68
68
30
25
39
13
25
85
61
17
12
93
75
73
70
86
65
57
85
25
27
81
76
71
88
90
83
17
16
18
21
88
25
28
26
51
48
51
45
NR
NR
38
50
39.9 (-)
54.5 (-)
49.9 (-)
44.6 (+)
53.3 (-)
66.1 (-)
72.9 (-)
78.1 (+)
51.6 (+)
45.2 (-)
46.7 (+)
48.3 (+)
61.5 (+)
64.6 (-)
62.7 (+)
41.4 (+)
40.8 (+)
58.2 (+)
47.8 (+)
40.9 (+)
42.6 (+)
40.6 (+)
37.0 (-)
47.6 (+)
32.7 (-)
42.2 (+)
40.1 (-)
33.7 (-)
41.9 (-)
36.6 (+)
20.5 (-)b
11.9 (-)b
36.7 (-)
29.2 (+)
29.1 (-)
80.6 (-)
78.2 (-)
74.9 (-)
79.3 (-)
78.8 (-)
76.4 (-)
79.0 (-)
NA
H
4-F
4-Cl
4-CF3
3,5-F2
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
(S)-(4-OMePh)
(R)-(4-MePh)
(S)-(3,5-Me2Ph)
(S)-(4-FPh)
(S)-(4-ClPh)
(R)-(4-CF3Ph)
(S)-(3,5-F2Ph)
(S)-Ph
(S)-(4-OMePh)
(S)-Ph
(S)-(4-OMsPh)
(S)-Ph
(S)-(4-OMePh)
(R)-(4-MePh)
(S)-(3,5-Me2Ph)
(S)-(4-FPh)
(S)-(4-ClPh)
(R)-(4-CF3Ph)
(S)-(3,5-F2Ph)
(S)-(4-OMePh)
(S)-Ph
(S)-(4-OMsPh)
(S)-Ph
(S)-(4-CO2MePh)
(S)-Ph
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
(S)-(4-CO2MePh)
(S)-Ph
(S)-Ph
(R)-Ph
(S)-Ph
(R)-Ph
(S)-Ph
(R)-Ph
(R)-Ph
(R)-Ph
(R)-Ph
(S)-Ph
(R)-Ph
(S)-Ph
(R)-Ph
(R)-Ph
(R)-Ph
(S)-Ph
Bz
4-Cl-Bz
4-OCF3-Bz
4-CF3-Bz
4-CN-Bz
methyl
benzyl
acetyl
acetyl
acetyl
(S)-Ph
(S)-Ph
(S)-Ph
(S)-Ph
(S)-Ph
PMP
PMP
PMP
PMP
PMP
PMP
PMP
H
H
H
CDAc
CDAc
(R)-(CH2)4-
(S)-t-Bu
(R)-(CH2)4-
(S)-t-Bu
4-Cl
4-Cl
4-Cl
4-Cl
4-Cl
4-Cl
4-Cl
4-Cl
4-Cl
3,5-F2Ph
3,5-F2Ph
(R)-3,5-F2Ph
(R)-3,5-F2Ph
(R)-3,5-F2Ph
(R)-3,5-F2Ph
(R)-3,5-F2Ph
(R)-3,5-F2Ph
(R)-3,5-F2Ph
(R)-3,5-F2Ph
(R)-3,5-F2Ph
(R)-3,5-F2Ph
(R)-3,5-F2Ph
(R)-3,5-F2Ph
(R)-3,5-F2Ph
(R)-3,5-F2Ph
(R)-3,5-F2Ph
(R)-3,5-F2Ph
(R)-3,5-F2Ph
(R)-3,5-F2Ph
(R)-3,5-F2Ph
(R)-3,5-F2Ph
(R)-3,5-F2Ph
(R)-3,5-F2Ph
2-naphthoyl
4-OMe-Bz
4-CF3-Bz
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
(R)-CEPd
(S)-CEPe
(R,S)-Pinf
(S,R)-Ping
PSh
Quini
PMP
NA
87.6 (-)
86.8 (-)
(S)-CEPe
a
b
d
Pentamethylpiperidine. DMA as solvent. c (1R,2S,3R)-camphordiamine. (R)-1-(1′-cyclohexylethyl)pyrrolidine. e (S)-1-(1′-cyclohexyl-
ethyl)pyrrolidine. f (R,R,R,S)-pinanepyrrolidine. (S,S,S,R)-pinanepyrrolidine. Proton Sponge. Quinuclidine. Chiralcel OD, 8.70 min
g
h
i
(+), 9.8 min (-).
means of NMR. Isotopically labeled ligands were pre-
pared as previously described by first synthesizing 1,2-
diphenylethylenediamine from benzil using 15NH4OAc.
Imidazoline formation and nitrogen substitution com-
pleted the synthesis. We also introduced 13C-labeled
nitrogen substituents by using 13CH3I and CH313COCl,
respectively. The isotopically labeled free ligands and
Pt(II) complexes (7 and 8) were then examined by 15N,
13C, and 31P NMR. We chose Pt(II) rather than Pd(II)
to avoid the quadrupolar line broadening observed with
the latter metal. The NMR data are summarized in
Figure 2.
J . Org. Chem, Vol. 69, No. 16, 2004 5189

Busacca et al.
F IGURE 4. Diamine Hammett plot, oxindole 2.
F IGURE 2. Isotopically labeled species. 5. 15N NMR δ:
-145.52 (N-im), -275.31 (1J N-C ) 9.3 Hz, N-Me). 31P NMR δ:
-11.13. 13C NMR δ: 33.38 (CH3, dd, J ) 3.3, 9.2 Hz). 1H NMR
δ: 4.87 (CHN-im), 4.23 (CHN-Me). 7. 15N NMR δ: -249.54
(N-im) -276.06 (1J N-C ) 9.7 Hz, N-Me). 31P NMR δ: 3.85
1
(1J P-Pt ) 3791 Hz). 13C NMR δ: 38.01 (CH3, d, J C-N ) 10.0
1
Hz). H NMR δ: 6.12 (CHN-im), 4.31 (CHN-Me). 6. 15N NMR
δ: -119.98 (N-im), -202.26 (1J N-C ) 9.3 Hz, N-Ac). 31P NMR
δ: -11.95. 13C NMR δ: 168.22 (CO, d, 1J C-N ) 10.3 Hz), 24.77
1
(CH3, dd, J ) 5.6, 52.7 Hz). H NMR δ: 5.22 (CHN-Ac), 5.08
(CHN-im). 8. 15N NMR δ: -203.29 (N-im), -214.52 (1J N-C
)
9.7 Hz, N-Ac). 31P NMR δ: 3.71 (1J P-Pt ) 3665 Hz). 13C NMR
1
δ: 169.51 (CO, d, J C-N ) 7.5 Hz), 24.14 (CH3, dd, J ) 8.8,
1
53.9 Hz). H NMR δ: 6.77 (CHN-im), 5.13 (CHN-Ac).
F IGURE 5. ORTEP plot of complex 3. Selected bond lengths
(Å) and angles (deg): Pd-P, 2.232; Pd-N(1), 2.046; P-Pd-
N(1), 86.27°; P-Pd-Cl(1), 93.36°; N-Pd-Cl(2), 89.54°; Cl(1)-
Pd-Cl(2), 91.87°.
methine imidazoline protons proximal to the chelating
nitrogen in both complexes are strongly deshielded (ca.
1.5 ppm) relative to the free ligands, while the distal
methines are essentially unchanged. In contrast, both
proximal and distal methine 13C resonances are mini-
mally affected (1-3 ppm upfield shift) by complexation.
The NMR data for all structures examined thus show
that only the imine nitrogens are involved in metal
chelation.
An examination of entries 33-35 of Table 1 provides
critical insight into the requirements for asymmetric
induction in this substrate. The two most sterically
hindered diamines used were bis-tert-butyl ethylenedi-
F IGURE 3. Phosphine Hammett plot, oxindole 2.
A comparison of the NMR data for free ligand 5 and
platinum complex 7 is informative. The N-Me resonance
appears as a doublet in 15N due to coupling with the
attached 13C nucleus. This 15N resonance shifts less than
one ppm upon complexation, while the imine resonance
moves 104 ppm upfield in the complex. Similar shieldings
of late transition metal bound nitrogen resonances have
been previously observed.9 This unambiguously estab-
lishes the chelating nitrogen as the imine. Chelation of
the imine nitrogen in platinum complex 8 was also
determined by comparison of the 15N data for the free
(6) and bound species.
(9) (a) Schenetti, L.; Mucci, A.; Longato, B. J . Chem. Soc., Dalton
Trans. 1996, 3, 299-303. (b) Bissinger, H.; Beck, W. Z. Naturforsch.,
Teil B 1985, 40B (4), 507-511. (c) Longato, B.; Schenetti, L.; Bandoli,
G.; Dolmella, A.; Trovo, G. Inorg. Chem. 1994, 33 (14), 3169-3176. (d)
Longato, B.; Pasquato, L.; Mucci, A.; Schenetti, L. Eur. J . Inorg. Chem.
2003, 1, 128-137.
Inspection of Figure 2 shows that the N-imine 15N
signal moves upfield by 82 ppm in this platinum complex
(8). The phosphorus resonances are deshielded by ∼15
ppm in both complexes relative to the free ligands. The
5190 J . Org. Chem., Vol. 69, No. 16, 2004

Intramolecular Asymmetric Heck Reaction
SCHEME 1. Syn th esis of Ch ir a l Ter tia r y Am in es
11 was used with our most optimized ligand, as shown
in entry 46. Isolated yield was once again improved, and
the enantioselectivity changed by less than 1% relative
to the use of PMP as base. These small changes in ee
are about the reproducibility in enantioselectivity from
one experiment to the next. Choice of base can clearly
improve catalytic efficiency in these systems, however.
In all examples from Table 1 (as well as all following
Heck reaction tables), the reaction yield based on recov-
ered starting material was nearly quantitative, since
unreacted triflate could be isolated in all cases of
incomplete conversion. No degradation of starting mate-
rial was observed, although a general trend of lower
reaction conversion was found with strongly electron-rich
or electron-deficient ligands. This trend could be partially
offset by meta disposition of ligand aryl substituents,
however. This behavior is well illustrated by comparing
entries 14 and 15 of Table 1. The p-trifluoromethyl and
m-difluoro substituted ligands give nearly identical enan-
tioselectivities (65% vs 63%) as would be expected from
their similar Hammett values. However, the meta-
disubstituted ligand furnishes oxindole 2 in high yield
(93%), while the para-substituted system gives only 13%
product after the same 18 h reaction time.
F IGUR E 6. ORTEP plot of complex 4. Selected bond
lengths (Å) and angles (deg): Pd-P, 2.231; Pd-N(1), 2.083;
P-Pd-N(1), 86.8°; P-Pd-Cl(2), 91.5°; N-Pd-Cl(1), 91.3°;
Cl(1)-Pd-Cl(2), 90.7°.
amine and camphordiamine, used to make the ligands
in entries 33 and 35. Each gives only very modest
enantioselectivity (37%, 29% ee). The least sterically
hindered diamine used was 1,2-cyclohexanediamine, in
entry 34. This ligand furnished oxindole 2 also with 29%
ee. Clearly there is little steric contribution to asymmetric
induction in this substrate, while electronic effects (e.g.,
Figure 3) are operative. We completed the initial screen-
ing set for substrate 1 by combining the best phosphines
(4-chlorophenyl and 3,5-difluorophenyl) and diamine
substitution (3,5-difluorophenyl) and varied R4 between
three electron-withdrawing groups. The chlorophosphine
results (entries 36-38) gave 2 with 75-81% ee, with
2-naphthoyl as the optimum R4 group. The difluorophenyl
phosphine result (entry 45) gave the highest enantiose-
lectivity observed, 88% ee.
We next examined triflate 14, our second substrate,
which was prepared in a manner analogous to tri-
flate 1.5b,c The results of ligand screening are collected
in Table 2.
The isolated yield of 2 after 18 h for most of the
electron-deficient ligand systems were reduced relative
to their more neutral analogues. We therefore undertook
a screening of bases, since choice of base has been
reported to affect enantioselectivity and yield.4f-h,l,10 To
the best of our knowledge, chiral bases have never been
applied to the asymmetric Heck reaction, so we prepared
the four chiral amines 10-13 as shown in Scheme 1 for
evaluation as well. The commercially available optically
pure primary amines were converted to their pyrrolidine
derivatives with 1,4-dibromobutane using the procedure
of Li.11 A total of six additional bases, as shown in entries
39-44, were evaluated. Both Proton Sponge and quinu-
clidine led to a catalytically inactive system (entries
43-44). The enantiomeric bases 10-13 were evaluated
in four side-by-side experiments (entries 39-42). All four
bases provided higher isolated yields of 2 after our 18 h
reaction time, yet there was no significant effect on
enantioselectivity, as all furnished 2 in 76-79% ee. Base
Two striking observations were immediately made
when comparing Table 2 to Table 1: The second sub-
strate gives much higher isolated yields accompanied by
much lower enantioselectivity than the first substrate.
The phosphine and diamine Hammett plots for the
second substrate are shown in Figures 7 and 8, respec-
tively.
Figure 7 shows a diminished linear correlation relative
to the first substrate, and the ee’s span a smaller range.
The overall trend is the same, however, with a posi-
tive F and 3,5-difuorophenyl as the best phosphine
substitution. The diamine Hammett plot in Figure 8 is
remarkably similar to that of the first substrate (Fig-
ure 4). The sum curve defining the R2 ) R3 ligands again
goes through a minimum near the neutral hydrogen
substituent, and both more electron-rich and electron-
poor species give higher enantioselectivity. Even the
relative position of the ligands derived from the non-C2-
symmetric diamines are the same as the first substrate,
once again falling below the sum curve, and showing the
same rank ordering of the proximal and distal methoxy
pair, where the largest difference in log er was seen. In
agreement with the phosphine data for this substrate,
(10) Lloyd-J ones, G. C.; Slatford, P. A. J . Am. Chem. Soc. 2004, 126
(9), 2690-2691.
(11) Li, Z.; Upadhyay, V.; DeCamp, A. E.; DiMichele, L.; Reider, P.
J . Synthesis 1999, Suppl. 1, 1453-1458.
J . Org. Chem, Vol. 69, No. 16, 2004 5191

Busacca et al.
TABLE 2. Tr ifla te 14, F ir st Liga n d Scr een in g Set
entry
R1
R2
R3
R4
yield, %
ee,a
%
1
2
3
4
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
4-OMe
4-Me
3,5-Me2
H
4-Cl
4-CF3
3,5-F2
H
H
H
H
H
H
H
H
H
(R)-Ph
(R)-Ph
(R)-Ph
(S)-Ph
(R)-Ph
(R)-Ph
(S)-Ph
(S)-(4-OMePh)
(R)-(4-MePh)
(S)-(3,5-Me2Ph)
(S)-(4-ClPh)
(R)-(4-CF3Ph)
(S)-(3,5-F2Ph)
(S)-Ph
(S)-(4-OMePh)
(S)-Ph
(S)-(4-OMsPh)
(S)-Ph
(S)-(4-CO2MePh)
(S)-t-Bu
(R)-(3,5-F2Ph)
(R)-(3,5-F2Ph)
(R)-Ph
(R)-Ph
(R)-Ph
(S)-Ph
(R)-Ph
(R)-Ph
(S)-Ph
(S)-(4-OMePh)
(R)-(4-MePh)
(S)-(3,5-Me2Ph)
(S)-(4-ClPh)
(R)-(4-CF3Ph)
(S)-(3,5-F2Ph)
(S)-(4-OMePh)
(S)-Ph
(S)-(4-OMsPh)
(S)-Ph
(S)-(4-CO2MePh)
(S)-Ph
(S)-t-Bu
(R)-(3,5-F2Ph)
(R)-(3,5-F2Ph)
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
2-naphthoyl
acetyl
65
60
71
98
89
97
89
46
87
86
81
79
95
85
46
64
57
84
86
82
81
67
2.3 (R)
10.9 (R)
16.2 (R)
6.6 (S)
17.7 (R)
17.7 (R)
31.1 (S)
13.0 (S)
0.6 (R)
4.5 (S)
25.9 (S)
32.6 (R)
31.4 (S)
5.4 (S)
7.8 (S)
23.8 (S)
8.4 (S)
13.9 (S)
13.0 (S)
50.9 (R)
55.9 (R)
77.5 (R)
H
H
H
H
4-Cl
3,5-F2
2-naphthoyl
2-naphthoyl
a
Chiralcel OD, 13.8 min (S)-(-), 17.6 min (R)-(+).
F IGURE 7. Phosphine Hammett plot, oxindole 15.
the Hammett plot is “compressed”: the shape is the
same, yet the range of ee’s spanned is reduced. Fortu-
nately, combining the optimized phosphine and di-
amine substitution patterns still leads to a ligand with
good enantioselectivity, as shown in entry 23 of Table
2. This optimized ligand furnishes oxindole 15 in 78%
ee and 67% yield after 18 h. It seems the synergistic
effect between the optimized phosphine and diamine
substituents is more pronounced for this substrate
than for triflate 1. We observed 23% ee (S) using
(R)-t-Bu-PHOX for this substrate, as determined by chiral
HPLC. Overman has reported 66% ee (R) for this sub-
strate with (R)-BINAP, using chiral shift reagent NMR
F IGURE 8. Diamine Hammett plot, oxindole 15.
determination of enantioselection.4l We observed 45% ee
using (R)-BINAP with chiral HPLC analysis for this
substrate.
The increased catalytic efficiency for this ketal sub-
strate is quite intriguing, since it applies to all ligands,
irrespective of their electronics. Both the nitrogen sub-
stituent and the olefin were changed relative to the first
substrate, however, so we needed to prepare another
substrate to probe which structural change was respon-
sible for the increased yield. As shown in Table 3,
N-methyl triflate 16 (prepared by methods analogous to
5192 J . Org. Chem., Vol. 69, No. 16, 2004

Intramolecular Asymmetric Heck Reaction
TABLE 3. Tr ifla te 16, Selected Liga n d Scr een in g
TABLE 4. Tr ifla te 18, Selected Liga n d Scr een in g
entry
R1
R2
(S)- Ph
4-OMe (R)-Ph
4-Cl (R)-Ph
R3
(S)- Ph
(R)-Ph
(R)-Ph
(R)-Ph
(S)- Ph
yield, % ee,a
%
entry
R1
R2
R3
yield (%) eea, %
1
2
3
4
5
6
7
8
9
10
11
12
13
H
83
70
85
71
88
44
83
51
58
68
62
25
11
21.4 (-)
11.4 (+)
39.4 (+)
46.9 (+)
29.8 (-)
15.7 (-)
23.6 (+)
32.0 (-)
34.1 (-)
35.0 (-)
35.6 (+)
41.5 (+)
50.1 (-)
1
2
3
4
H
(S)-4-ClPh
(R)-Ph
(S)-4-ClPh
(R)-Ph
(S)-Ph
68
80
99
53
52.1 (R)
62.7 (S)
69.9 (R)
82.8 (S)
4-Cl
3,5-F2 (S)-Ph
3,5-F2 (R)-3,5-F2Ph (R)-3,5-F2Ph
4-CF3 (R)-Ph
3,5-F2 (S)- Ph
a
H
H
H
H
H
(S)-4-OMePh (S)-4-OMePh
Chiralpak AD, 7.70 min (S)-(+), 8.97 min (R)-(-).
(R)-4-MePh
(S)-4-FPh
(S)-4-ClPh
(R)-4-MePh
(S)-4-FPh
(S)-4-ClPh
those already described5b,c) was then screened against a
select group of ligands.
(S)-3,5-F2Ph (S)-3,5-F2Ph
3,5-F2 (R)-3,5-F2Ph (R)-3,5-F2Ph
4-Cl (R)-3,5-F2Ph (R)-3,5-F2Ph
4-CF3 (S)-3,5-F2Ph (S)-3,5-F2Ph
The yields of oxindole 17 found with all four electron-
deficient ligands for this N-methyl ketal substrate were
significantly higher than the N-methyl-gem-dimethyl
substrate 1 and behaved like ketal substrate 14 in terms
of catalytic efficiency. The enantioselectivities observed
with substrate 16, however, were much higher than for
ketal 14, particularly for the ligands optimized in only
one region (Table 3, entries 1-3). We were able to make
two critical conclusions from these data. First, the ketal
moiety of the substrates imparts increased catalytic
efficiency, without affecting enantioselectivity. Second,
the N-benzyl group tends to give lower enantioselectivi-
ties than the analogous N-methyl Heck substrates. We
believe this could have important implications in the use
of the asymmetric Heck reaction for synthesis of complex
tragets. As seen here, the ketal and nitrogen substituent
of the substrate might play the role of protecting groups
in a total synthesis. Their judicious choice prior to the
asymmetric Heck reaction could allow for increased
enantioselectivity without sacrifice of catalytic efficiency.
Several lines of evidence generated here point to
facilitated â-hydride elimination as the source of the
greater catalytic efficiency: (1) Base choice can have
dramatic effects on turnover, as shown in Table 1. (2)
No significant differences in enantioselection were ob-
served between two pairs of enantiomeric strong bases,
suggesting, as expected, that the step they are involved
in is not the enantiodifferentiating step. (3) The ketal
substrates give better turnover than their alkyl analogue,
irrespective of the nitrogen substituent. (4) The increased
catalytic efficiency seen with the ketals is seen over the
entire range of ligand electronics, for all ligand members.
Since electronic effects on enantioselection are operative,
it is unlikely that the source of increased turnover is
electronic in origin.
a
Chiralcel OD, 11.40 min (+), 14.9 min (-).
(S)-BINAP, using chiral shift reagent NMR determina-
tion of enantioselection.4l We observed ∼40% ee for this
substrate using (S)-BINAP with chiral HPLC analysis
of the product. As was observed for the first substrates,
the optimized BIPI ligands give substantially higher
enantioselectivities with substrate 16 than do BINAP or
t-Bu-PHOX.
Our final substrate examined for the asymmetric Heck
reaction was triflate 18 bearing the protected tetrahy-
dropyridine functionality.5b,c We chose this substrate as
a reasonable model for a number of natural products
which possess the spiropiperidine functionality.12 Since
the olefin portion of this triflate was significantly differ-
ent than the other substrates, we again screened against
a sufficient number of ligands to allow generation of the
phosphine and diamine Hammett plots. The data are
collected in Table 4, and the Hammett plots are shown
in Figures 9 and 10. The Hammett behavior for this
substrate differs from all of the other substrates in at
least three ways: (1) 3,5-difluorophosphine substitution
is not the most enantioselective, (2) the diamine Ham-
mett shows a general increase in enantioselectivity with
increasing Σσ without the clear minimum in the plot
previously seen, and (3) ligands optimized in both regions
(entries 11-13, Table 4) do not show the dramatic in-
creases in enantioselectivity previously observed. We ex-
amined (R)-BINAP/n-BuNBr and (R)-t-Bu-PHOX against
this substrate and initially found no reaction with either
ligand after 24 h at 95 °C. Triflate 18 could be recovered
quantitatively from the PHOX experiment, while 18 was
consumed with BINAP. The BINAP experiment was then
Application of our optimized ligand (Table 3, entry 4)
to substrate 16 furnished 17 in 83% ee and 53% yield
after 18 h. We observed 23% ee (S) for (R)-t-Bu-PHOX
for this substrate, as determined by chiral HPLC. Over-
man has reported 71% ee (S) for this substrate with
(12) (a) Ariza, M. R.; Larsen, T. O.; Petersen, B. O.; Duus, J . O.;
Christophersen, C.; Barrero, A. F. J . Nat. Prod. 2001, 64 (12), 1590-
1592. (b) Larsen, T. O.; Frydenvang, K.; Frisvad, J . C.; Christophersen,
C. J . Nat. Prod. 1998, 61 (9), 1154-1157.
J . Org. Chem, Vol. 69, No. 16, 2004 5193

Busacca et al.
TABLE 5. Tr ifa te 1, R5 Scr een in g Set
entry
R5
yield, %
ee,b %
1
2
3
4
5
6
7
8
9
H
3-F
4-F
5-F
68
15
15
46
73
44.6 (+)
27.3 (+)
52.8 (+)
52.8 (+)
26.3 (+)
55.6 (+)
NA
NA
44.0 (+)
34.0 (+)
6-F
5-Cl
4,5-F2
5-CN
4-Ara
5-Ara
50
NR
NR
13
F IGURE 9. Phosphine Hammett plot, oxindole 19.
10
14
a
b
Ar ) 4′-CF3C6H4. Chiralcel OD 8.7 min (+), 9.8 min (-).
maintaining the standard phenyl substitution at R1, R2,
and R3 and holding R4 constant as 2-naphthoyl, our
optimum R4 group. We used trifate 1 to screen the R5
substituents, and the results are collected in Table 5.
We chose electron-withdrawing groups for this ligand
screen, reasoning that the electronics of the benzo ring
would be similar to the other aryl groups on phosphorus,
where those types of groups were preferred. The most
striking result is the generally reduced yields for these
ligands. Entry 3 with the 4-fluoro ligand gave 2 in only
15% isolated yield. It is interesting that four m-fluorines
are well tolerated in catalysis when they are on the
pendant aryl groups on phosphorus, yet a single m-
fluorine on the benzo ring nearly stops catalysis com-
pletely. This same trend was generally observed through-
out, and two fluorines or a cyano group do turn off
catalysis. The ligand that gives the best yield at 73%
(6-fluoro, entry 5) also gives the lowest enantioselectivity
at 26%. This ligand showed rotamers in the NMR, and
may populate a conformation that is good for turnover
and bad for enantioselection. The biphenyl-type ligands
(entries 9 and 10) were created to “mimic” a halogen, as
strongly electron-deficient aromatic rings are roughly
equivalent to a fluoro or chloro substituent directly
attached to the ring in terms of σp Hammett param-
eters.13 Their application here was quite disappointing,
however, furnishing the product in poor yield and enan-
tioselectivity. The best ligand from this screening set was
the 5-chloro derivative, which showed an increase in ee
of 11% relative to the unsubstituted system, although the
isolated yield was again somewhat diminished. We
therefore prepared5c ligand 20 possessing four 3,5-dif-
luorophenyl groups (Figure 11), the 5-chloro benzo sub-
F IGURE 10. Diamine Hammett plot, oxindole 19.
repeated, and aliquots were withdrawn at 2 h intervals.
No cyclization product 19 was ever formed, yet the tri-
flate was consumed in about 6 h, generating polar
impurities. Triflate 18 is clearly unstable under these
reaction conditions. The PHOX experiment was then
repeated at 120 °C, and enecarbamate oxindole 19
could then be isolated in 54% yield, with an ee of
46.2% (+). The optimized BIPI ligands (entries 4 and
13) are thus somewhat more enantioselective than the
PHOX ligand, although only the p-trifluoromethyl ligand
(entry 4) is practical in terms of yield. It is interesting to
note that the same rank ordering of phosphine substit-
uents (p-CF3 > p-Cl > m-F2) with respect to enantiose-
lectivity observed in the phosphine Hammett plot of
Figure 9 is maintained in the optimized ligands, entries
11-13. It is clear that tetrahydropyridines such as 18
are still quite challenging substrates for the asymmetric
Heck reaction.
We wished to briefly examine the one ligand region not
previously studiedsthe central benzo ring. Using the
nomenclature previously introduced, this region was
designated R5. We decided to vary only R5 first, while
(13) (a) Hansch, C.; Leo, A.; Taft, R. W. Chem. Rev. 1991, 91 (2),
165-195. (b) Hansch, C.; Leo, A.; Unger, S. H.; Kim, K. H.; Nikaitani,
D.; Lien, E. J . J . Med. Chem. 1973, 16 (11), 1207-1216. (c) Hansch,
C.; Leo, A. Substituent Constants for Correlation Analysis in Chemistry
and Biology; Wiley: New York, 1979.
5194 J . Org. Chem., Vol. 69, No. 16, 2004

Intramolecular Asymmetric Heck Reaction
therefore likely that the triflate anion remains closely
associated with cationic Pd(II) under these conditions.
A possible explanation for increased enantioselectivity
in the AHR with electron-deficient ligands is therefore
formation of a pentacoordinate transition state that is
inherently more stereoselective than the square planar
system. There is evidence14b,e that square planar and tbp
M(II) complexes are in equilibrium, and this suggests
that fluxional behavior might operate in the catalytic
cycle. The optimized BIPI ligands, containing electron-
deficient P- and N-termini, might simply spend more time
in the pentacoordinate geometry prior to olefin insertion
than the more electron-rich ligands do.
F IGURE 11. Ligand 20.
The relief of steric compression in the ligand plane that
can lead to pentacoordinate geometry is considered to be
more dominant than electronic effects in this geometric
reorganization. It might therefore be asked why sterically
encumbered ligands such as those derived from bis-tert-
butylethylenediamine or camphordiamine in the present
work give only modest enantioselectivity in the AHR. The
vast majority of the complexes for which tbp geometry
has been observed are N,N-ligands forming a five-
membered chelate, such as 2,9-dimethyl-1,10-phenan-
throline, and 2-iminopyridines. The BIPI and PHOX
ligands form six-membered chelates, while BINAP forms
a seven-membered chelate. The driving force to reduce
the interligand contacts of square planar geometry by
geometric change might therefore be attenuated in these
larger chelates, leading to default electronic control of
geometric reorganization. This model would also explain
the superiority of nonpolar solvents, as pentacoordinate
geometry would likely be disfavored in ionizing solvents.
Overall, this suggests the possibility for superior ligand
design in a future system possessing both electron-
deficiency and a five-membered or smaller metal chelate.
F IGURE 12. Tbp geometry.
stituent, and 2-naphthoyl at R4. This ligand was then
screened against Heck substrates 1 and 14. In both cases,
no product was formed, even at elevated temperature. It
seems likely that this highly electron-deficient ligand has
exceeded an electronic limit required for a competent
catalyst in the AHR.
Or igin of En a n tioselection . It is not possible to state
with certainty why electron-deficient ligands lead to
higher enantioselectivity in the AHR, yet the results
obtained with the BIPI ligands do allow for speculation
on the origin of enantioselection. Overman14a has cor-
rectly pointed out the possible role of pentacoordinate Pd
geometry for this transformation. Numerous pentacoor-
dinate Pd(II) and Pt(II) alkene complexes have been
characterized crystallographically.14b,c There are two
primary forces14b,d,e which drive the change from square
planar to trigonal bipyramidyl (tbp) geometry in these
systems: (1) steric compression in the ligand plane and
(2) electron-withdrawing substituents on the metal. This
latter effect applies both to the ligand14b,e and to the
olefin.14b,d Current crystallographic data strongly suggest
that the geometry shown in the simplified structure of
Figure 12 above would be most likely, where the biden-
tate ligand and the olefin are in the equatorial plane,
while the aryl substituent and the triflate occupy the
apical positions. It is important to remember that the
high asymmetric induction observed with the BIPI
ligands occurs only in nonpolar solvents, and enantiose-
lection is dramatically eroded in polar media. It is
Con clu sion
We have explored the electronic requirements for
enantioselection in the creation of chiral quaternary
centers in the intramolecular asymmetric Heck reaction
using a new class of electronically tunable ligands. It has
been found for the first time that stereoselectivity
increases with decreasing phosphine electron density.
The electronic dependence of stereoselection on the
embedded diamine is more complex, yet is also maxi-
mized with decreased diamine electron density. By com-
bining the best phosphorus and nitrogen substitution
patterns as relates to enantioselectivity, optimized ligands
have been created that give the highest enantioselectivi-
ties reported to date for the most challenging types of
asymmetric Heck reactions. The BIPI ligands are cur-
rently under broad investigation for electronic effects in
asymmetric catalysis.
(14) (a) Ashimori, A.; Bachand, B.; Calter, M. A.; Govek, S. P.;
Overman, L. E.; Poon, D. J . J . Am. Chem. Soc. 1998, 120 (26), 6488-
6499. (b) Albano, V. G.; Natile, G.; Panunzi, A. Coord. Chem. Rev. 1994,
133, 67-114. (c) Garrone, R.; Romano, A. M.; Santi, R.; Millini, R.
Organometallics 1998, 17 (20), 4519-4522. (d) Fanizzi, F. P.; Maresca,
L.; Nayile, G.; Lanfranchi, M.; Tiripicchio, A.; Pacchioni, G. J . Chem.
Soc., Chem. Commun. 1992, 333-335. (e) Groen, J . H.; Delis, J . G. P.;
van Leeuwen, P. W. N. M.; Vrieze, K. Organometallics 1997, 16, 6(1),
68-77.
Su p p or tin g In for m a tion Ava ila ble: Experimental in-
formation containing general and specific procedures, char-
acterization of all new compounds, and crystallographic data
for complexes 3 and 4. This material is available free of charge
via the Internet at http://pubs.acs.org.
J O049448Z
J . Org. Chem, Vol. 69, No. 16, 2004 5195