Enantioselective hydrogenations of arylalkenes mediated by [Ir(cod)(JM-phos)]+ complexes

Article abstract of DOI:10.1002/1521-3765(20011217)7:24<5391::AID-CHEM5391>3.0.CO;2-1

Phosphine oxazoline ligands 1a-j were converted to the corresponding [Ir(cod)(phosphine oxazoline)]⁺ complexes 2a-j. X-ray diffraction analyses of complexes 2b, 2h, 2i, and 2j were performed. The tert-butyl-, 1,1-diphenylethyl-, and phenyl-oxazoline complexes (2b, 2h, and 2i, respectively) had typical square planar metal environments with chair-like metallocyclic rings. However, the 3,5-di-tert-butylphenyl oxazoline complex 2j was distorted toward a tetrahedral metal geometry. This library of complexes was tested in asymmetric hydrogenations of several arylalkenes. High enantioselectivities and conversions were observed for some substrates. A possible special role for the HPh₂C-oxazoline substituent in asymmetric hydrogenations was identified and is discussed. In attempts to rationalize why high enantioselectivities were not observed for some alkenes, a series of deuterium labeling experiments were performed to probe for competing reactions that occurred prior to the hydrogenation step. Double bond migrations were inferred for several substrates, and this is a significant complication in asymmetric hydrogenations of arylalkenes that had not been discussed prior to this study. A mechanistic rationale is proposed involving competing double bond migration for some but not all substrates. Appreciation of this complication will be valuable in further studies aimed at optimization of enantioselection in asymmetric hydrogenations of unfunctionalized alkenes. Wiley-VCH Verlag GmbH, 2001.

Full text of DOI:10.1002/1521-3765(20011217)7:24<5391::AID-CHEM5391>3.0.CO;2-1

FULL PAPER
Enantioselective Hydrogenations of Arylalkenes Mediated by
‡
[Ir(cod)(JM-Phos)] Complexes
Duen-Ren Hou,[a] Joseph Reibenspies,[a] Thomas J. Colacot,[b] and Kevin Burgess*[a]
Abstract: Phosphine oxazoline ligands
1a j were converted to the correspond-
ing [Ir(cod)(phosphine oxazoline)]
kenes. High enantioselectivities and
conversions were observed for some
substrates. A possible special role for
the HPh2C-oxazoline substituent in
asymmetric hydrogenations was identi-
fied and is discussed. In attempts to
rationalize why high enantioselectivities
were not observed for some alkenes,a
series of deuterium labeling experiments
were performed to probe for competing
reactions that occurred prior to the
hydrogenation step. Double bond mi-
grations were inferred for several sub-
strates,and this is a significant compli-
cation in asymmetric hydrogenations of
arylalkenes that had not been discussed
prior to this study. A mechanistic ration-
ale is proposed involving competing
double bond migration for some but
not all substrates. Appreciation of this
complication will be valuable in further
studies aimed at optimization of enan-
tioselection in asymmetric hydrogena-
tions of unfunctionalized alkenes.
‡
complexes 2a j. X-ray diffraction anal-
yses of complexes 2b, 2h, 2i,and 2j
were performed. The tert-butyl-,1,-
diphenylethyl-,and phenyl-oxazoline
complexes (2b, 2h,and 2i,respectively)
had typical square planar metal environ-
ments with chair-like metallocyclic rings.
However,the 3,5-di- tert-butylphenyl ox-
azoline complex 2j was distorted toward
a tetrahedral metal geometry. This li-
brary of complexes was tested in asym-
metric hydrogenations of several arylal-
Keywords: alkenes
¥
asymmetric
catalysis ¥ homogeneous catalysis ¥
hydrogenation ¥ N ligands ¥ P li-
gands
The first encouraging data reported for asymmetric hydro-
genations of unfunctionalized alkenes featured cyclopenta-
dienyl complexes of titanium and zirconium. Titanocene
dichloride complexes prepared from camphor-derived,chiral
cyclopentadienyl ligands were used as catalyst precursors for
hydrogenation of 2-phenylbut-1-ene and 2-ethyl-1-hexene.
The best optical purity was 34%,only 10 turnovers occurred.
In order to obtain these data the reaction was run at
À758C.[1,2] Moreover,the enantiomeric excess ( ee) values
were determined by polarimetry and the intrinsic experimen-
tal error for that method is high. Similarly,others have since
used zirconium catalysts functionalized by the Brintzinger
bis(tetrahydroindenyl) system[3] and reported ee values of up
to 65% based on optical rotations.[4,5] Enantiomers of many
alkanes with aromatic substituents are now resolvable on
modern chiral HPLC columns,and,more recently,this
technique was used to analyze hydrogenation reactions
mediated by Brintzinger×s titanocene catalysts. Buchwald
and co-workers examined nine substrates in this way. Eight of
the nine ee values reported were in the 83 99% range.[6] All
these substrates were trisubstituted alkenes. The catalyst
loading required,2 5 mol%,was higher than ideal. Similar
ziconocene catalysts were used for hydrogenations of tetra-
Introduction
Development of effective asymmetric hydrogenation reac-
tions of unfunctionalized alkenes has been slow relative to
some other reactions catalyzed by organometallic complexes.
This is partly due to analytical difficulties. Alkanes do not
have functional groups which interact strongly with chiral
supports so their optical purities tend to be difficult to
ascertain via chromatographic methods. Probably more sig-
nificant is the fact that,for unfunctionalized alkenes,there is
little else but steric forces available to be manipulated to
achieve enantioface differentiation since polarity and hydro-
gen bonding effects are unlikely to be significant.
[a] Prof. K. Burgess,Dr. D.-R. Hou,Dr. J. Reibenspies
Department of Chemistry,Texas A & M University
P.O. Box 30012,College Station,TX 77842 (USA)
Fax: (979)845-8839
E-mail: burgess@tamu.edu
[b] Dr. T. J. Colacot
Johnson Matthey,Catalysts and Chemicals Division
2001 Nolte Drive,West Deptford,NJ 08066-1742 (USA)
Supporting information for this article is available on the WWW under
http://www.wiley-vch.de/home/chemistry/ or from the author: Tabulat-
ed details of the screens performed using different complexes in
Reactions (2) (6). Crystallographic data for the X-ray structural
analyses are available from the author.
[7]
substituted alkenes,but the results were more variable.
Pinene-derived cyclopentadienyl ligands have also been
Chem. Eur. J. 2001, 7,No. 24
¹ WILEY-VCH Verlag GmbH,D-69451 Weinheim,2001
0947-6539/01/0724-5391 $ 17.50+.50/0
5391

K. Burgess et al.
FULL PAPER
transformed into titanocene and zirconocene complexes and
tested as hydrogenation catalysts,but the data reported was
less encouraging; ee values of up to 69% (measured by
polarimetry) were reported for two 1,1-disubstituted alkenes.
‡
À
Crabtree×s catalyst,[Ir(cod)(py)PCy 3] PF6 ,has some
unique properties with respect to hydrogenation of alkenes.[8]
Mechanistic studies[9] indicate that oxidative addition of
hydrogen to this complex takes place with a trans-orientation
relative to the pyridine ligand.[10] Solvent or alkene then
coordinates and the COD ligand is removed by hydrogena-
tion. Coordination of two alkene units to the iridium,then
‡
transformation of the resulting [IrH2(py)(PCy3)(alkene)2]
into hydrogenation products is apparently a prevalent path-
way for many alkene substrates.[11] Relief of unfavorable steric
interactions may drive the hydrogenation step,and this
accounts for the fact that Crabtree×s catalyst hydrogenates
tri- and tetrasubstituted alkenes which are not converted by
closely related rhodium complexes.
Pfaltz and co-workers prepared complexes of the type
‡
[Ir(cod)(A)] (A ˆ the phosphine oxazoline shown below)
‡
and,recognizing their similarity to [Ir(cod)(py)PCy 3] ,inves-
tigated these as catalysts in hydrogenation reactions of
14]
trisubstituted alkenes.[12
Later,they also investigated
reactions of the corresponding complexes prepared from the
phosphite/diazaphospholidine oxazolines B. The enantiose-
lectivities they obtained were comparable to the best results
using Brintzinger×s titanocene catalysts,but the conversions
were higher and less catalyst was required (0.1 1.0 mol%
versus ꢀ5 mol%).
factors which are difficult to control by simple modifications
of reaction conditions and/or ligand structure. The conclu-
sions are a concern for all analogues of Crabtree×s catalyst:
Pfaltz×s,ours and others that may be contemplated. Conse-
quently,while the yields and enantioselectivities reported
here are comparable but not superior to these reported earlier
by Pfaltz,these studies lay foundations for further advances in
the area by delineating obstacles which must be overcome to
attain highly enantioselective hydrogenations of unfunction-
alized alkenes.
We recently reported syntheses of the phosphine oxazoline
ligands 1,and commented on them in relation to ligands
A.[15,16] They have a curved shape,and they tend to form
complexes wherein the R-functionality projects to an orien-
tation nearer the metal than the corresponding substituent in
ligands A. Moreover,there is more scope for varying the R
substituent in the phosphine oxazoline ligands 1 with respect
to electronic and topographical diversity. None of these
considerations mean that ligands 1 must be superior to the
structures A,but they do imply that complexes of these should
be sufficiently different to make the comparison interesting.
This paper describes experiments which were designed to
compare ligands A and 1 in asymmetric hydrogenations of
arylalkenes. The motivation for these studies was to find
superior ligands for asymmetric hydrogenations of aryl
alkenes. Implicit in this approach is the assumption that
optimization of enantioselectivities in this process is only a
function of ligand design and appropriate modification of
conditions. In fact,data that emerged from these studies
indicates that line of reasoning is too superficial. Specifically,
deuterium labeling experiments demonstrated diminished
enantioselectivities for some substrates can be attributed to
Results and Discussion
The catalyst library: The catalysts required for this work were
prepared by Reaction (1). These materials were isolated as
orange-yellow solids by crystallization,and they are not
particularly air-sensitive.
ꢁ1†
Structural variance in the catalyst library as a function of the
different R groups was probed by single crystal X-ray studies
of compounds 2b, 2h, 2i,and 2j (Figure 1). The structures of
the tert-butyl (2b),11,-diphenylethyl ( 2h),and phenyl ( 2i)
complexes are very similar. They have typical square-planar
structures in which the metallocyclic ring adopts a chair-like
5392
¹ WILEY-VCH Verlag GmbH,D-69451 Weinheim,2001
0947-6539/01/0724-5392 $ 17.50+.50/0
Chem. Eur. J. 2001, 7,No. 24

Enantioselective Hydrogenations
5391 5400
conformation. Complex 2h was the only di(2-methylphenyl)-
phosphino derivative examined (the rest were Ph2P ligands)
but this appeared to make no significant difference to the
structure: They all crystallized with an edge-on pseudo-axial
P-aryl group,while they each had a less encumbered,pseudo-
equatorial aryl group which adopted edge- or face-on
orientations. The 3,5-di-tert-butylphenyl complex 2j had a
slightly different solid state structure. It appears that inter-
actions of the 3,5-di-tert-butylphenyl substituent with the
COD ligand cause the latter to be twisted out of the square
plane,so much,in fact,that the complex more closely
resembles a tetrahedral structure. We conclude from these
observations that within the series of complexes that were
examined by X-ray crystallography,ligand 1j imposes the
most serious steric demands on the metal. In the catalysis
experiments,complex 2j was one of the more enantioselective
catalysts (see below).
Hydrogenation reactions: These reactions were performed in
parallel by placing several tubes in a Parr autoclave and
running the hydrogenations simultaneously. Critical data for
hydrogenation of E-1,2-diphenylpropene are shown in Fig-
ure 2,and Table 1 gives more details. The complex with the
smallest R1 substituent,the methyl-oxazoline complex 2a,
gave a high yield of product but the enantiomeric excess was
modest (63%). Substitution of the methyl group with a tert-
butyl decreased the yield but increased the enantioselectivity.
Complexes 2a and 2b both have diphenylphosphino P
centers,whereas the tert-butyl complex 2c has a di(2-
methylphenyl)phosphino group. This change resulted in a
lower yield but a higher enantioselectivity. Complex 2d (R1 ˆ
2
1-Ad,R ˆ Ph) gave a very poor yield and a reasonable
enantioselectivity (slightly increased yield was obtained when
the reaction was run at À58C under 70 bar H2). However,
when the R1 substituent was very large,as in complex 2e
(R1 ˆ CPh3),both the yield and the enantioselectivity of the
reaction were greatly diminished.
Figure 2. Enantioselectivity and yield data as a function of the ligand used
in the reactions shown. [a] 0.1 mol% of catalyst was used.
We hypothesized that a diphenylmethyl R1 substituent
might be advantageous for the following reasons. A diphe-
nylmethyl substituent is large and non-spherical and,unlike
the CPh3 group it can adopt conformations that accommodate
the substrates on the metal by placing the two phenyls away
from the metal. Moreover,the two phenyl groups could adopt
an edge-face conformation (Figure 3a) reminiscent of that in
2
complexed,optically active,C
symmetric phosphines and
some other ligand types (Figure 3b).[17] We were unable to
crystallize the HCPh2 derivative 2g for X-ray crystallographic
analysis. However,the solid state structure of complex 2h
(R1 ˆ MeCPh2) shows that its phenyl groups adopt a con-
formation which resembles the edge-face motif (Figure 3c).
Complex 2 f (R1 ˆ CHPh2, R2 ˆ Ph) gave a higher enantio-
selectivity than the corresponding methyl (2a), tert-butyl
(2b),and 1-adamantyl complexes ( 2d),thus supporting the
hypothesis outlined above. We then looked for another factor
that could be changed to further increase the enantioselec-
tivity. The observation that the diphenylphosphino coordinat-
ing group in complex 2b gave an inferior result to the di(2-
methylphenyl)phosphino complex 2c in the R1 ˆ tert-butyl
series implied that this change might be more generally
Figure 1. Top and side views of a) [(cod)Ir1b]BARF; b) [(cod)Ir1i]PF6;
c) [(cod)Ir1j]PF6; and d) [(cod)Ir1h]BARF. All diagrams generated from
X-ray coordinates and presented in Chem3D. Counterions are omitted for
clarity throughout.
Chem. Eur. J. 2001, 7,No. 24
¹ WILEY-VCH Verlag GmbH,D-69451 Weinheim,2001
0947-6539/01/0724-5393 $ 17.50+.50/0
5393

K. Burgess et al.
FULL PAPER
Table 1. Enantioselective hydrogenation of E-1,2-diphenylpropene.
H2, complex 2
Ph
Ph
Ph
Ph
CH2Cl2
Complex
cation
Anion
Catalyst
equiv [mol%]
T [8C]
t [h]
H2
Yield[a] [%]
ee[b] [%]
pressure [bar]
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
2a
2b
2c
2d
2e
2 f
2g
2h
2i
2i
2i
2j
2b
2b
2b
2b
2b
2d
BARF
BARF
BARF
BARF
BARF
BARF
BARF
BARF
BARF
0.2
0.1
0.1
0.2
0.1
0.2
0.2
0.2
0.1
0.1
0.1
0.2
0.1
0.2
0.4
0.2
0.2
0.2
25
25
25
25
25
25
25
25
25
25
25
25
À 5
À 5
À 5
À 5
À 5
À 5
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
20
2
50
50
50
50
59
50
50
50
50
50
50
50
50
50
50
70
70
70
99
55
40
10
6
71
99
12
89
5
96
55
37
37
35
45
42
58
63
75
89
84
5
87
95
14
79
28
79
93
90
87
88
90
89
82
À
PF6
BARF[c]
BARF[c]
BARF
BARF
BARF
BARF
BARF
BARF
[a] GLC yield. [b] Enantiomeric excess was determined by GLC (1008C,retention time: t1 ˆ 117.6 min, t2 ˆ 120.9 min). [c] 1 mol% of NaBARF {BARF ˆ
tetrakis[3,5-bis(trifluoromethyl)phenyl]borate} was added.
evident when complex 2h (R1 ˆ CMePh2, R2 ˆ 2-MeC6H4)
was tested. Substitution of the methine hydrogen with a
methyl group in the R1 substituent caused a marked reduction
in the catalytic effectiveness in terms of both yield and
enantiomeric excess.
Complex 2i is different to 2a 2h insofar as the R1 phenyl-
substituent in this complex is planar. In the hydrogenation
reaction,complex 2i gave a good yield and a moderately good
enantiomeric excess,and complex 2j gave slightly better data
(Table 2,entries 16 and 17).
Pfaltz and co-workers had discovered that the BARF
counterion[18,19] was critical in their hydrogenation studies for
obtaining good yields. They have measured conversion as a
function of reaction time and from this deduced that catalyst
deactivation was a serious problem for other counterions such
À
as PF6 . The same factors are operative in the reactions
studied here. For instance,if the hexafluorophosphate salt of
cation 2i was used,then the yield dropped precipitously
À
(Table 1,compare entries 9 and 10). The presence of PF is
6
not intrinsically detrimental to this reaction because when 10
equivalents of NaBARF were added to a reaction mediated
À
by 2iPF6 ,the yield was slightly better than observed for the
corresponding BARF complex,and the enantioselectivity was
the same (entries 11 and 9). When the PF6À salt of complex 2j
was screened using 10 equivalents of NaBARF in the same
way,a good enantioselectivity was obtained but the yield was
only 55%.
Figure 3. a) Anticipated edge-face orientation of phenyl groups in com-
plex 2g; b) edge-face orientation of phenyl groups in a typical C2-
symmetric chiral phosphine complex; and,c) Chem3D space filling model
of the [Ir1h] fragment of the complex 2h generated from X-ray
crystallographic coordinates.
Yields and enantioselectivities in these reactions depend on
several variables; this was illustrated in a series of experi-
ments involving complex 2b. When the experiment in entry 2
was repeated but at À58C rather than 258C,the enantiose-
lectivity increased by 15%,but the yield fell by 18% (Table 1,
entries 2 and 13). Catalyst concentration over the range 0.1
0.4 mol% had little effect at À58C (entries 13 15). Both the
‡
beneficial. Indeed,complex 2g (R1 ˆ CHPh2, R2 ˆ 2-MeC6H4)
gave the best data under these conditions. The extreme
sensitivity of the catalysis to the R1 substituent became
5394
¹ WILEY-VCH Verlag GmbH,D-69451 Weinheim,2001
0947-6539/01/0724-5394 $ 17.50+.50/0
Chem. Eur. J. 2001, 7,No. 24

Enantioselective Hydrogenations
5391 5400
yields and the enantioselectivities marginally increased when
the pressure was increased from 50 to 70 bar (entries 14 and
16). Prolonged reaction times,however,did not increase the
yield (entries 16 and 17).
Figure 4 and Table 2 summarize the data for hydrogenation
of a similar,but more electron-rich substrate, E-2-(methoxy-
phenyl)-1-phenylpropene. The methyl-oxazoline complex 2a
gave a good yield and enantioselectivity considering this
Comparison of entries 2 and 3 shows that the reaction system
is extremely sensitive to extraneous coordinating groups. A
small amount of acetone was deliberately added in the second
of these experiments,and both the yield and the enantiose-
lectivity decreased. Similar observations were made by
Crabtree for his catalyst.[20] Entries 7 and 8 describe identical
experiments except that the reaction time was longer in the
latter case. Comparable data were obtained implying that the
catalyst had ceased to function after approximately 2 h in both
experiments. Collectively,these observations indicate that the
size of the substituents R1 and R2 influence the catalyst
stability. Indeed,the more hindered di(2-methylphenyl)phos-
phino-complex 2c gave better conversions than the diphenyl-
phosphino-derivative 2b. The adamantyl complex 2d gave
similar enantioselectivities to the closely related tert-butyl
complex 2b but lower yields. Just as in the last set of
experiments,the triphenylmethyl oxazoline complex 2e gave
very poor conversions and enantioselectivities,but the
diphenylmethyl complex 2 f gave much better data (entries 12
and 13). Superior to that,and the best result in the series,the
corresponding di(2-methylphenyl)phosphino complex 2g
gave good conversion and enantioselectivity (entry 14) at
room temperature and a lower pressure than in the second
best experiment (entry 9). Again,the 11, -diphenylethyl com-
plex 2h did not perform as well as the corresponding
diphenylmethyl complex 2g (entries 14 and 15). The two
complexes with relatively flat R1 substituents, 2i and 2j,did
not perform as well as the ones with more three dimensional
groups (entries 16 and 17).
Figure 4. Enantioselectivity and yield data as a function of the ligands used
in the reaction shown. [a] 0.1 mol% of catalyst was used.
substituent is so small,and that the reaction was run at 25 8C
(Table 2,entry 1). Complex 2b has the larger, tert-butyl,
substituent,and higher enantioselectivities were observed for
this (up to 91%,entry 7). However,for both catalysts 2a and
2b,it was difficult to drive the reactions to completion.
3-Acetoxy-2-methylphenylpropene was screened by a strat-
egy which was similar to those outlined above. Details of these
experiments are given in the Supporting Information. The
best data are shown in Reaction (2). Two complexes gave
Table 2. Enantioselective hydrogenation of 1-phenyl-2-(4-methoxyphe-
nyl)propene.
H2, complex 2
Ph
Ph
Ar
Ar
CH2Cl2
Ar = C6H4-4-OMe
Complex
cation
Catalyst
equiv
(mol%)
T [8C]
t [h]
H2
pressure
[bar]
Yield[a]
[%]
ee[b]
[%]
ꢁ2†
1
2
3c
4
5
6
7
8
9
2a
2b
2b
2b
2b
2b
2b
2b
2c
2c
2d
2e
2 f
2g
2h
2i
0.3
0.1
0.1
0.1
0.2
0.4
0.2
0.2
0.2
0.3
0.2
0.3
0.3
0.3
0.3
0.3
0.3
25
25
25
À 5
À 5
À 5
À 5
À 5
À 5
25
25
25
25
25
25
25
25
2
2
2
2
2
2
2
20
2
2
2
2
2
2
2
2
2
50
50
50
50
50
50
70
70
70
50
50
50
50
50
50
50
50
85
37
15
50
64
61
63
62
92
83
6
78
85
68
88
84
78
91
88
94
92
80
0
84
93
20
87
93
moderately good results. The tert-butyl complex 2b gave the
highest enantioselectivity in the series,but only 53% con-
version to product,whereas the phenyloxazoline complex 2i
gave 15% less enantioselectivity but a good yield. The
corresponding allylic alcohol behaved similarly [Reac-
tion (3)].
10
11
12
13
14
15
16
17
1
96
99
20
48
65
Unlike the data obtained for the first two substrates,the
results shown above are significantly inferior to that reported
by Pfaltz.[12] Two extreme modifications were attempted to
improve the performance of our catalysts,but neither worked.
For instance,hydrogenations using the phenyl-oxazoline
2j
[a] GLC yield. [b] The enantiomeric excess was determined by GLC
(1308C,retention time: t1 ˆ 101.7 min, t2 ˆ 105.0 min). [c] Acetone (0.2 mL,
0.4 equiv catalyst) was added.
Chem. Eur. J. 2001, 7,No. 24
¹ WILEY-VCH Verlag GmbH,D-69451 Weinheim,2001
0947-6539/01/0724-5395 $ 17.50+.50/0
5395

K. Burgess et al.
FULL PAPER
ꢁ6†
ꢁ3†
propene derivatives are the easiest to effect. Lesser enantio-
selectivities for the other substrates probably reflect non-ideal
enantiodiscrimination by the catalyst. However,we became
suspicious that another contributing factor might be oper-
ative: double bond migrations at rates that are competitive
with the desired hydrogenation reactions. To check this
possibility,a series of deuteration experiments were per-
formed. Some typical data from these experiments are shown
in Figure 5.
The first experiment did not provide evidence for double
bond migration. Thus when E-2-(4-methoxyphenyl)-2-butene
was deuterated under the conditions shown in Reaction (7),
deuterium atoms were detected only in the positions where
the double bonds were previously located. GC/MS experi-
ments indicated that only two deuteriums were incorporated
into the product.
complex 2iBARF in the presence of Cs2CO3 or NEt3[21] gave
low yields of product with poor enantioselectivities (<8%). A
reaction using complex 2cBARF in the ionic liquid[22 27] N,N'-
ethylmethylimidazolium tetrafluoroborate gave no conver-
sion to product.
Reactions (4) and (5) show our two best results for hydro-
genation of 2-aryl-2-butene derivatives. On an absolute scale,
the conversions and enantioselectivities were good but not
excellent. The enantioselectivities obtained in Reaction (4)
were better than those reported for iridium complexes of
ligand A,[12,13] but not as good as those reported for ligand B[28]
for the same substrate. Similar trends were observed for the
corresponding trans-alkene substrate [Reaction (5)]. The
absolute configuration of the hydrogenation products in
Reactions (4) and (5) were opposite; Pfaltz and co-workers
observed the same phenomenon in their studies.[12]
ꢁ7†
Deuteration of the cis-substrate, Z-2-(4-methoxyphenyl)-2-
butene,gave a different result to deuteration of the trans-
substrate [Reactions (8) and (7)]. The distribution of deute-
rium atoms in the product implies that extensive double bond
migration occurred before the hydrogenation step. Interest-
ingly,the reaction did not go to completion and the recovered
starting material had deuterium incorporated at the methyl
group shown,but the Z-stereochemistry of the double bond
was retained and no double bond migration product was
observed. GC/MS analyses indicated that the reduced product
contained 2 4 deuterium atoms. More deuterated starting
material was observed when the reaction was performed
under only 1 atm of D2 [Reaction (9)].
ꢁ4†
ꢁ5†
ꢁ8†
ꢁ9†
Finally,two 11,-disubstituted alkenes were investigated,and
the best data are shown in Reaction (6). Interestingly,the
enantioselectivities increased slightly when the reactions were
run at 608C rather than À5 or 258C (see Supporting
Information). In absolute terms,the ee values shown are not
high. However,1,-disubstituted alkenes are notoriously
difficult to hydrogenate with high enantioselectivities,and
the data shown here compare favorably with those reported
for other catalysts.
Deuteriumlabeling experiments : The data presented above
prove that enantioselective hydrogenations of 1,2-diaryl
Deuteration of 2-(4-methoxyphenyl)-1-butene also showed
that extensive double bond migration had occurred [Reac-
5396
¹ WILEY-VCH Verlag GmbH,D-69451 Weinheim,2001
0947-6539/01/0724-5396 $ 17.50+.50/0
Chem. Eur. J. 2001, 7,No. 24

Enantioselective Hydrogenations
5391 5400
tion (10)]. The reduced product containing 2 4 deuterium
atoms was detected in GC/MS experiments. Under 1 atm of
D2 this more reactive substrate was almost totally converted
to product,also with scrambling of the deuterium label
(Reaction 11).
ꢁ10†
ꢁ11†
Two interesting questions arise from the deuterium incorpo-
ration experiments. First,how do the double bond migration
reactions occur? Second,why do they occur for the cis-alkene
shown in Reaction (8) but not the trans-isomer in Reac-
tion (7)? The data collected does not allow us to answer these
questions with complete certainty,but the working hypothesis
expressed in Scheme 1 accounts for all the observations made.
We propose the labeling in Reactions (8) and (9) is consistent
with formation of an allyl intermediate C. Insertion,and
deuteration/hydrogenation of the double bond accounts for
the label distribution observed (Scheme 1a). Similarly,in-
volvement of the p-allyl complex C explains the label
distribution observed in Reaction (10) (Scheme 1b). How-
ever,if this proposal is valid,then there must be a factor that
accounts for lack of scrambling via a similar intermediate, p-
allyl D,in Reaction (7). Complexes C and D are stereo-
isomers,and D is likely to be the least stable of the two.
Therefore,Hammond postulate leads to the conclusion that D
will form more slowly than C,and we propose that this
difference is such that formation of C is competitive with the
hydrogenation reaction whereas D is not.
Conclusion
Double bond migration competing with the hydrogenation
reactions is an important,though unwelcome,observation. If
a substrate partially isomerizes before it reacts,then asym-
metric reduction of both alkenes must be controlled to obtain
high enantioselectivities. This is a difficult thing to achieve.
Consequently,modifications to improve the enantioselectiv-
ities of iridium-mediated hydrogenations of the type described
here must involve methods to suppress double bond migration
and/or enhance the relative rate of the direct reduction step. The
importance of double bond migration in these hydrogenations
mediated by chiral derivatives of Crabtree×s catalyst has not
been noted before this work,even though similar effects have
been thoroughly investigated for hydrogenation of allylic
alcohols mediated by ruthenium-BINAP catalysts.[13,29,30]
Figure 5. Typical data from deuterium labeling studies: a) 1H NMR
spectrum of 2-(4-methoxy)phenylbutane; b) 2H NMR of the crude material
from the reaction of 50 bar D2 with E-2-(4-methoxy)phenyl-2-butene;
c) 2H NMR of the crude material from the reaction of 50 bar D2 and 2-(4-
methoxy)phenylbut-1-ene; d) 2H NMR of the crude material from the
reaction of 50 bar D2 and Z-2-(4-methoxy)phenyl-2-butene.
Chem. Eur. J. 2001, 7,No. 24
¹ WILEY-VCH Verlag GmbH,D-69451 Weinheim,2001
0947-6539/01/0724-5397 $ 17.50+.50/0
5397

K. Burgess et al.
FULL PAPER
a
HCAr2 groups at critical positions was helpful here,and
may be useful in other cases. Development of asymmetric
hydrogenation reactions to give uniformly high enantiomeric
excesses for a variety of alkene substrates will hinge on careful
consideration of issues such as these in conjunction with
modifications to suppress competing double bond migration.
D(H)
Ir
D(H)
IrD(H)
Ir-D(H)
Ar
Ar
Ar
Ar
Ar
fast
C
direct addition
of H2 or D2
addition
of H2 or D2
D(H)
(H)D
D(H)
D(H)
D(H)
Ar
Ar
Experimental Section
General procedures: High field NMR spectra were recorded on Varian
b
2
Unity Plus 300 (1H at 300 MHz, 13C at 75 MHz, H at 46 MHz,and 31P at
D(H)
Ir
D(H)
Ir-D(H)
121 MHz) spectrometers. Chemical shifts of 1H and 13C spectra are
referenced to the NMR solvents; 31P spectra are referenced to H3PO4
(85%) external standard (d ˆ 0). Melting points are uncorrected. Optical
rotations were measured on Jasco DIP-360 digital polarimeter. Flash
chromatography was performed using silica gel (230 600 mesh). Thin-
layer chromatography was performed on glass plates coated with silica gel
60 F254 (E. Merck,Darmstadt,Germany). Elemental analyses were
performed by Atlantic Microlab,Norcross,GA. CH 2Cl2 was distilled over
CaH2 and THF over Na/benzophenone. Other solvents were purchased
from commercial suppliers and were used without further purification.
Chloro-1,5-cyclooctadiene iridium(i) dimer was provided by Johnson
Matthey. E-2-(4-Methoxyphenyl)-1-phenylpropene,[32] E- and Z-2-(4-me-
thoxyphenyl)-2-butene,[33] 2-phenyl-1-butene,and 2-(4-methoxyphenyl)-1-
butene were prepared according to literature procedures. Enantioselectiv-
ities were deduced through GC on a chiral support.[34]
IrD(H)
Ar
Ar
fast
C
direct addition
of H2 or D2
addition
of H2 or D2
(H)D
D(H)
D(H)
(H)D
D(H)
Ar
Ar
c
D(H)
Ir
D(H)
IrD(H)
Ir-D(H)
The absolute configurations of the products were confirmed in some cases,
and inferred in others as outlined below. The product in Table 1 and in
Reaction (6) where Ar' ˆ Ph is the same,and its the absolute configuration
was confirmed by comparing optical rotations with those reported by
Buchwald.[6] The absolute configuration of the product in the experiments
listed in Table 2 and in Reactions (4) (6) (Ar' ˆ C6H4-4-OMe) was
inferred from comparing the GC retention factors of this product with
that in Table 1; this is likely to be correct since the two alkanes are so
similar (differ only by a para-methoxy group on one aryl).
Ar
Ar
slow
D
direct addition
of H2 or D2
addition
of H2 or D2
D(H)
D(H)
(H)D
D(H)
D(H)
Ar
Ar
not detected
Deuterium gas under high pressure was purchased from Praxair Inc,
Danbury,CT. All the ligands 1 were prepared by the method reported
previously,[16] and spectroscopic data for the particular derivatives that have
not been discussed before (1a, 1c, 1 f, 1g and 1h) are given in the following
sections.
Scheme 1. Working hypothesis for the incorporation of deuterium in the
labeling experiments: a) for Reaction (8); b) for Reaction (10); and c) for
Reaction (7).
(S)-2-Methyl-4-[(diphenylphosphino)ethyl]oxazoline (1a): Rf ˆ 0.27 (ethyl
acetate/hexane 3:7 v/v); [a]2D4 ˆ À87.6 (c ˆ ‡3.4,CHCl 3); 1H NMR (CDCl3,
300 MHz): d ˆ 7.48 7.43 (m,4H),7.37 7.33 (m,6H),4.30 (dd, J ˆ 9,8 Hz,
1H),4.16 (m,1H),3.79 (dd, J ˆ 8,8 Hz,1H),2.31 2.21 (m,1H),2.12 2.02
(m,1H),1.99 (s,3H),1.74 1.60 (m,2H); 13C NMR (CDCl3,75 MHz): d ˆ
Double bond migration reactions have been observed in
hydrogenation/deuteration of pinene derivatives using Crab-
tree×s catalysts itself.[31]
The possibility of competing double bond migration
reactions also makes the data from the enantioselective
reactions harder to compare. If one ligand gives a higher
enantioselectivity than another,we cannot be sure that this is
because the design of the first ligand is more conducive to
asymmetric induction in the desired reaction rather than
diminished competing migration reactions.
The yield and enantioselectivity data in this paper does
prove that ligands 1 in complexes 2 can act as highly
enantioselective hydrogenation catalysts. Optimization of
these involves the topography of the R1 substituent. If R1 is
too large,or has an inappropriate shape,the rate of the
desired hydrogenation process might be suppressed relative to
catalyst degradation,adversely affecting the conversions.
Substitution of the phosphine Ph substituents with 2-MeC6H4,
however,tends to increase the enantioselectivities without
decreasing conversions. The strategy expressed in Figure 2 for
optimization of enantioselectivities by incorporation of
164.7,132.8,132.6,128.6 128.4,72.3,67.2 (d,
JC,P ˆ 13.5 Hz),32.1 (d, JC,P ˆ
16.5 Hz),24.2 (d, JC,P ˆ 11.5 Hz),13.9; 31P NMR (CDCl3,121 MHz): d ˆ
‡
À15.46; HRMS: m/z: calcd for C18H21NOP: 298.13608 [M‡H] ; found:
298.13623.
(S)-2-tert-Butyl-4-[(bis(2-tolyl)phosphino)ethyl]oxazoline (1c): Rf ˆ 0.32
(ethyl acetate/hexane 1:9 v/v); [a]2D4 ˆ À79.6 (c ˆ 1.0,CHCl 3); 1H NMR
(CDCl3,300 MHz): d ˆ 7.30 7.20 (m,4H),7.20 7.15 (m,4H),4.28 (dd, J ˆ
9,8 Hz,1H),4.19 (m,1H),3.88 (dd,
2.12 (m,1H),2.03 1.99 (m,1H),1.76 1.33 (m,2H),1.27 (s,9H);
J ˆ 6,8 Hz,1H),2.46 (m,6H),2.18
13C NMR
(CDCl3,75 MHz): d ˆ 173.8,142.4 141.9,136.8 136.4,131.0,130.0,128.3,
125.9,71.9,66.6 (d, JC,P ˆ 13.5 Hz),33.0,32.0 (d, JC,P ˆ 17.5 Hz),27.8,22.5 (d,
JC,P ˆ 12.0 Hz),21.0 (d, JC,P ˆ 21.5 Hz); 31P NMR (CDCl3,121 MHz): d ˆ
‡
À37.6; HRMS: m/z: calcd for C23H31NOP: 368.21433 [M‡H] ; found:
368.21308.
(S)-2-Diphenylmethyl-4-[(diphenylphosphino)ethyl]oxazoline (1 f): Rf ˆ
0.78 (ethyl acetate/hexane 3:7 v/v); [a]2D4 ˆ À57.5 (c ˆ 4.4,CHCl 3);
1H NMR (CDCl3,300 MHz): d ˆ 7.50 7.37 (m,4H),7.36 7.26 (m,16H),
5.32 (s,0.5H),5.15 (s,0.5H),4.38 4.25 (m,2H),3.94 (m,1H),2.24 2.19
(m,1H),2.10 2.02 (m,1H),1.78 1.68 (m,2H);
13C NMR (CDCl3,
75 MHz): d ˆ 166.8,139.3,132.9,132.6,132.4,128.6 128.4,127.0,72.2,66.6
(d, JC,P ˆ 14.0 Hz),50.9,31.9 (d, JC,P ˆ 17.0 Hz),23.6 (d, JC,P ˆ 11.5 Hz); 31P
5398
¹ WILEY-VCH Verlag GmbH,D-69451 Weinheim,2001
0947-6539/01/0724-5398 $ 17.50+.50/0
Chem. Eur. J. 2001, 7,No. 24

Enantioselective Hydrogenations
5391 5400
NMR (CDCl3,121 MHz): d ˆ À15.5; HRMS: m/z: calcd for C30H29NOP:
0.043 mmol,43%) was produced as
a
yellow solid. M.p. 81 84 8C
‡
450.19868 [M‡H] ; found: 450.19874.
(decomp); [a]2D4 ˆ À15.28 (c ˆ 1.7,CDCl 3); 1H NMR (CDCl3,300 MHz):
d ˆ 7.76 7.29 (m,20H),7.21 7.14 (m,2H),4.88 (m,1H),4.39 4.33 (m,
2H),4.26 (t, J ˆ 9 Hz,1H),4.05 (dd, J ˆ 9,3 Hz,1H),3.99 (m,1H),2.94
(m,1H),2.86 2.68 (m,2H),2.49 (m,1H),2.40 (s,1H),2.37 2.33 (m,2H),
2.19 2.05 (m,2H),1.98 1.96 (m,3H),1.86 (s,1H),1.79 (s,5H),1.72 (s,
(S)-2-Diphenylmethyl-4-[(bis(2-methylphenyl)phosphino)ethyl]oxazoline
(1g): Rf ˆ 0.69 (ethyl acetate/hexane 3:7 v/v); [a]2D4 ˆ À70.6 (c ˆ 5.7,
CHCl3); 1H NMR (CDCl3,300 MHz): d ˆ 7.80 7.30 (m,8H),7.29 7.15
(m,10H),5.17 (s,1H),4.40 4.31 (m,2H),4.16 (m,1H),3.96 (dd,
6 Hz,1H),2.46 (s,6H),2.14 2.10 (m,1H),2.02 (m,1H),1.82 1.74 (m,
J ˆ 7,
2H),1.66 1.62 (m,1H),1.57 1.38 (m,5H),1.31 1.29 (m,2H);
31P NMR
(CDCl3,121 MHz): d ˆ 6.9; elemental analysis calcd (%) for C67H56BF24Ir-
2H); 13C NMR (CDCl3,75 MHz): d ˆ 167.4,142.5 141.9,139.3,136.4,
NOP: C 50.90,H 3.57,N 0.89; found: C 51.64,H 3.72,N 0.84.
131.0,129.9,128.6 128.4,127.1,126.0,72.2,66.7 (d,
JC,P ˆ 13.5 Hz),50.1,
32.9 (d, JC,P ˆ 18 Hz),22.6 (d, JC,P ˆ 11.5 Hz),21.1 (d, JC,P ˆ 21 Hz); 31P
Complex 2e: This compound was prepared by the same method used for
compound 2b,but beginning with 1e (75 mg,0.142 mmol) and chloro-(1,5-
cyclooctadiene)iridium(i) dimer (48 mg,0.071 mmol). Complex 2e (94 mg,
0.055 mmol,39%) was produced as a dark yellow solid. M.p. 177 179 8C
(decomp); [a]2D4 ˆ À58.28 (c ˆ 2.3,CDCl 3); 1H NMR (CDCl3,300 MHz):
d ˆ 7.76 7.17 (m,32H),6.99 6.89 (m,5H),4.95 (m,1H),4.75 (m,1H),
4.58 (t, J ˆ 9 Hz,1H),4.41 (dd, J ˆ 10,4 Hz,1H),3.43 (m,2H),2.83 (m,
1H),2.66 2.53 (m,2H),2.43 2.36 (m,1H),2.24 2.13 (m,2H),2.01 (m,
NMR (CDCl3,121 MHz): d ˆ À37.3; LSIMS: m/z: calcd for C32H33NOP:
‡
478 [M‡H] ; found: 478.
(S)-2-(1,1-Diphenyl)ethyl-4-[(bis(2-methylphenyl)phosphino)ethyl]oxazo-
line (1h): Rf ˆ 0.82 (ethyl acetate/hexane 3:7 v/v); [a]2D4 ˆ À68.7 (c ˆ 9.0,
1
CHCl3); H NMR (CDCl3,300 MHz): d ˆ 7.40 7.20 (m,18H),4.39 4.30
(m,2H),3.98 (m,1H),2.47 (s,6H),2.15 2.13 (m,1H),2.10 (s,3H),2.10
2.03 (m,1H),1.82 1.77 (m,2H); 13C NMR (CDCl3,75 MHz): d ˆ 170.7,
144.8,142.5 141.9,136.8 136.4,131.1,129.9,128.7 128.4,127.9,126.6,
126.0,72.2,66.6 (d, JC-,P ˆ 13.5 Hz),50.3,31.9 (d, JC,P ˆ 18 Hz),28.4,22.5 (d,
JC,P ˆ 11.5 Hz),21.2 (d, JC,P ˆ 21 Hz); 31P NMR (CDCl3,121 MHz): d ˆ
1H),1.87 (m,1H),1.74 (m,1H),1.28 1.17 (m,2H),1.15 1.04 (m,2H);
31P
NMR (CDCl3,121 MHz): d ˆ À1.5; elemental analysis calcd (%) for
C76H56BF24IrNOP: C 54.04,H 3.34,N 0.83; found: C 54.06,H 3.39,N 0.79.
‡
À37.3; LSIMS: m/z: calcd for C33H35NOP: 490 [M‡H] ; found: 490.
Complex 2 f: This compound was prepared by the same method used for
compound 2b,but beginning with 1 f (40 mg,0.089 mmol) and chloro-(15, -
cyclooctadiene)iridium(i) dimer (30 mg,0.044 mmol). Complex 2 f (68 mg,
0.043 mmol,48%) was produced as an orange-red solid. M.p. 78 81 8C
(decomp); [a]2D4 ˆ À14.28 (c ˆ 2.6,CDCl 3); 1H NMR (CDCl3,300 MHz):
d ˆ 7.74 (m,8H),7.63 7.29 (m,17H),7.24 7.18 (m,3H),7.10 7.06 (m,
General procedure for preparation of complex 2b (analogously 2a 2j):[36]
Ligand 1b (48.0 mg,0.143 mmol) and chloro-(1,5-cyclooctadiene)iridium( i)
dimer (48.2 mg,0.0717 mmol) were dissolved in CH 2Cl2 (3 mL) in a 10 mL
flask equipped with a condenser and a stir bar. The solution was refluxed
under N2 for 1 h. After the orange-red solution was cooled to room
temperature,NaBARF [18] (195 mg,0.22 mmol) was added followed by H 2O
(2 mL),and the resulting two-phase mixture was stirred vigorously for
15 min. The layers were separated and the aqueous layer was extracted with
CHCl3 (2 Â 10 mL). The combined organic extracts were washed with
water and evaporated. The residue was re-dissolved in EtOH (1.5 mL) and
crystallized by the slow addition of H2O to give 2b (133 mg,0.0887 mmol,
62%) as a yellow-orange solid. M.p. 162 1658C (decomp); [a]2D4 ˆ À54.28
(c ˆ 0.5,CDCl 3); 1H NMR (CDCl3,300 MHz): d ˆ 7.70 (m,8H),7.61 7.55
(m,10H),7.51 7.33 (m,2H),7.11 7.07 (m,2H),4.81 (m,1H),4.37 (m,
1H),4.27 4.18 (m,2H),4.04 (dd, J ˆ 9,3 Hz,1H),3.91 (m,1H),2.88 2.72
(m,3H),2.48 2.43 (m,1H),2.32 2.23 (m,2H),2.15 1.91 (m,4H),1.76
1.59 (m,2H),1.42 1.39 (m,1H),1.10 (s,9H); 13C NMR (CDCl3,75 MHz):
d ˆ 180.1,162.7,162.0,161.4,160.7,134.7,132.5,131.3,131.2 126.8,126.3,
122.7,119.1,117.5,94.8 (d, J ˆ 8 Hz),86.5 (d, J ˆ 17 Hz),72.5,67.8 (d, J ˆ
3 Hz),65.8,64.9,36.7 (d, J ˆ 5 Hz),33.6,30.1,28.6 (d, J ˆ 2 Hz),28.5,25.5
(d, J ˆ 3 Hz),23.8 (d, J ˆ 34 Hz); 31P NMR (CDCl3,121 MHz): d ˆ 7.48;
elemental analysis calcd (%) for C61H50BF24IrNOP: C 48.75,H 3.35,N 0.93;
found: C 48.80,H 3.39,N 0.87.
3H),6.48 6.46 (m,1H),5.43 (s,1H),4.86 (br,1H),4.65 (t,
J ˆ 9Hz,1H),
4.40 (m,1H),4.27 (m,1H),4.18 (m,1H),4.10 (m,1H),2.98 2.87 (m,2H),
2.80 2.50 (m,1H),2.57 2.45 (m,2H),2.39 2.17 (m,3H),1.96 1.70 (m,
3H),1.50 1.29 (m,2H); 31P NMR (CDCl3,121 MHz): d ˆ 9.9; elemental
analysis calcd (%) for C70H52BF24IrNOP ‡ H2O: C 51.54,H 3.34,N 0.86;
found: C 51.32,H 3.58,N 0.83.
Complex 2g: This compound was prepared by the same method used for
compound 2b,but beginning with 1g (55 mg,0.115 mmol) and chloro-(1,5-
cyclooctadiene)iridium(i) dimer (39 mg,0.058 mmol). Complex 2g (107 mg,
0.065 mmol,57%) was produced as an orange-red solid. M.p. 76 79 8C
(decomp); [a]2D4 ˆ À26.88 (c ˆ 3.0,CDCl 3); 1H NMR (CDCl3,300 MHz):
d ˆ 7.80 7.10 (m,24H),6.57 (br,2H),5.65 (s,1H),4.98 (m,1H),4.64 (m,
2H),4.40 4.30 (m,1H),4.29 4.21 (m,1H),4.04 (br,1H),3.88 (br,1H),
3.11 2.82 (m,4H),2.50 (m,1H),2.42 2.34 (m,2H),2.21 2.14 (m,1H),
1.97 1.85 (m,3H),1.70 1.51 (br,3H),1.48 1.26 (m,2H);
31P NMR
(CDCl3,121 MHz): d ˆ 19.6 (br),11.7,0.4 (br); elemental analysis calcd
(%) for C72H56BF24IrNOP ‡ H2O: C 52.12,H 3.52,N 0.84; found: C 51.82,
H 3.42,N 0.91. The NMR spectra of this complex appear to be broadened
by a slow exchange mechanism.
Complex 2a: This compound was prepared by the same method used for
2b,but beginning with 1a (51 mg,0.17 mmol) and chloro-(15,-cycloocta-
Complex 2h: This compound was prepared by the same method used for
compound 2b,but beginning with 1h (80 mg,0.16 mmol) and chloro-(15, -
diene)iridium(i) dimer (58 mg,0.086 mmol). Complex
2a (136.5 mg,
0.094 mmol,55%) was produced as an orange solid. M.p. 138 141 8C
(decomp); [a]2D4 ˆ ‡77.68 (c ˆ 0.5,CDCl 3); 1H NMR (CDCl3,300 MHz):
d ˆ 7.78 7.21 (m,20H),7.13 7.06 (m,2H),4.88 (m,1H),4.38 4.20 (m,
2H),4.15 (m,1H),4.02 (dd, J ˆ 9,3 Hz,1H),3.90 (m,1H),2.94 2.69 (m,
3H),2.54 2.23 (m,4H),2.17 (s,3H),2.15 1.94 (m,2H),1.78 1.68 (m,
2H),1.59 1.46 (m,2H); 31P NMR (CDCl3,121 MHz): d ˆ 10.9; elemental
analysis calcd (%) for C58H44BF24IrNOP ‡ H2O: C 47.10,H 3.14,N 0.95;
found: C 46.98,H 2.98,N 1.04.
cyclooctadiene)iridium(i) dimer (55 mg,0.080 mmol). Complex
2h
(160 mg,0.097 mmol,60%) was produced as an orange-red solid. M.p.
173 1768C (decomp); [a]D24 ˆ À87.38 (c ˆ 2.6,CDCl ); 1H NMR (CDCl3,
3
300 MHz): d ˆ 8.01 7.10 (m,27H),6.77 (br,1H),6.52 (br,1H),6.09 (br,
1H),4.62 (br,1H),4.50 4.39 (2H),4.27 (1H),4.10 (1H),3.51 (1H),3.20
(m,2H),2.98 (3H),2.82 (3H),2.36 (2H),2.22 (m,5H),1.98 1.90 (m,2H),
1.72 (2H),1.58 (1H),1.36 1.25 (m,2H); 31P NMR (CDCl3,121 MHz): d ˆ
20.6 (br),11.1 (br), À0.4; elemental analysis calcd (%) for C73H58BF24Ir-
NOP: C 52.97,H 3.53,N 0.85; found: C 53.12,H 3.51,N 0.89. The NMR
spectra of this complex appear to be broadened by a slow exchange
mechanism.
Complex 2c: This compound was prepared by the same method used for
compound 2b,but beginning with 1c (68 mg,0.19 mmol) and chloro-(15, -
cyclooctadiene)iridium(i) dimer (63 mg,0.094 mmol). Complex 2c (196 mg,
0.128 mmol,68%) was produced as an orange solid. M.p. 169 171 8C
(decomp); [a]2D4 ˆ À16.38 (c ˆ 3.0,CDCl 3); 1H NMR (CDCl3,300 MHz):
d ˆ 8.05 (br,1H),7.74 (s,8H),7.55 (s,4H),7.50 7.20 (m,7H),4.79 (br,
1H),4.38 (m,2H),4.36 (m,1H),3.97 (m,2H),2.98 (br,3H),2.71 (br,1H),
2.48 (br,1H),2.26 (m,3H),2.22 1.88 (m,8H),1.53 (br,9H),1.50 1.25
Complex 2i: This compound was prepared by the same method used for
compound 2b,but beginning with 1i (54 mg,0.15 mmol) and chloro-15,-
cyclooctadiene)iridium(i) dimer (50 mg,0.075 mmol). Complex 2i (149 mg,
0.098 mmol,65%) was produced as an orange-yellow solid. M.p. 149
1518C (decomp); [a]D24 ˆ ‡10.48 (c ˆ 3.0,CDCl 3); 1H NMR (CDCl3,
300 MHz): d ˆ 8.58 8.54 (m,2H),7.75 7.39 (m,23H),7.21 7.14 (m,
2H),4.85 (m,1H),4.65 4.55 (m,2H),4.30 4.24 (m,2H),3.76 (m,1H),
3.02 2.96 (m,1H),2.90 2.81 (m,2H),2.66 2.57 (m,1H),2.47 2.31 (m,
2H),2.24 2.18 (m,2H),2.04 1.99 (m,1H),1.89 1.59 (m,2H),1.59 1.49
(m,2H); 31P NMR (CDCl3,121 MHz): d ˆ 9.2; elemental analysis calcd
(%) for C63H46BF24IrNOP: C 49.68,H 3.04,N 0.92; found: C 49.69,H 3.15,
N 1.07.
(m,2H),1.07 (br,1H);
31P NMR (CDCl3,121 MHz): d ˆ 21.5 (br),11.5
(br),0.3; elemental analysis calcd (%) for C 63H54BF24IrNOP: C 49.42,H
3.55,N 0.91; found: C 49.69,H 3.68,N 1.00. The NMR spectra of this
complex appear to be broadened by a slow exchange mechanism.
Complex 2d: This compound was prepared by the same method used for
compound 2b,but beginning with 1d (42 mg,0.10 mmol) and chloro-(15, -
cyclooctadiene)iridium(i) dimer (34 mg,0.50 mmol). Complex 2d (69 mg,
Chem. Eur. J. 2001, 7,No. 24
¹ WILEY-VCH Verlag GmbH,D-69451 Weinheim,2001
0947-6539/01/0724-5399 $ 17.50+.50/0
5399

K. Burgess et al.
FULL PAPER
Complex 2j: This compound was prepared by the same method used for
compound 2b,but beginning with 1j (58 mg,0.123 mmol),chloro-(15,-
cyclooctadiene)iridium(i) dimer (41 mg,0.061 mmol) and ammonia hexa-
fluorophosphate (400 mg,2.48 mmol). Complex 2j (65 mg,0.071 mmol,
58%) was produced as an orange-red solid. M.p. 206 2088C (decomp);
[a]2D4 ˆ ‡41.08 (c ˆ 0.5,CDCl 3); 1H NMR (CDCl3,300 MHz): d ˆ 8.09 (s,
1H),7.78 7.71 (m,3H),7.63 7.40 (m,7H),7.37 7.25 (m,2H),5.11 (m,
1H),4.94 4.87 (m,2H),4.37 (dd, J ˆ 4 Hz, J ˆ 9Hz,1H),3.97 (m,1H),
3.82 (m,1H),3.29 3.23 (m,2H),2.99 2.90 (m,2H),2.73 2.55 (m,2H),
2.55 2.42 (m,1H),2.42 2.35 (m,2H),2.21 2.02 (m,3H),1.87 1.79 (m,
[2] R. L. Halterman,K. P. C. Vollhardt, Organometallics 1988, 7,883
892.
[3] F. R. W. P. Wild,J. Zsolnai,G. Huttner,H. H. Brintzinger, J. Organo-
met. Chem. 1982, 232,233 247.
[4] R. Waymouth,P. Pino, J. Am. Chem. Soc. 1990, 112,4911 4914.
[5] R. B. Grossman,R. A. Doyle,S. L. Buchwald, Organometallics 1991,
10,1501 1505.
[6] R. D. Broene,S. L. Buchwald, J. Am. Chem. Soc. 1993, 115,12569
12570.
[7] M. V. Troutman,D. H. Appella,S. L. Buchwald, J. Am. Chem. Soc.
1999, 121,4916 4917.
1H),1.72 1.58 (m,5H),1.40 1.25 (m,13H);
31P NMR (CDCl3,
121 MHz): d ˆ 10.9, À143.6 (JP,F ˆ 710 Hz); elemental analysis calcd (%)
[8] R. H. Crabtree,G. E. Morris, J. Organomet. Chem. 1977, 135,395
403.
for C39H50F6IrNOP2: C 51.08,H 5.50,N 1.53; found: C 50.54,H 5.67,N 1.52.
[9] R. Crabtree, Acc. Chem. Res. 1979, 12,331 337.
[10] R. H. Crabtree,R. J. Uriarte, Inorg. Chem. 1983, 22,4152 4154.
[11] R. H. Crabtree,P. C. Demou,D. Eden,J. M. Mihelcic,C. A. Parnell,
J. M. Quirk,G. E. Morris, J. Am. Chem. Soc. 1982, 104,6994 7001.
[12] A. Lightfoot,P. Schnider,A. Pfaltz, Angew. Chem. 1998, 110,3047
3050; Angew. Chem. Int. Ed. 1998, 37,2897 2899.
[13] D. G. Blackmond,A. Lightfoot,A. Pfaltz,T. Rosner,P. Schnider,N.
Zimmermann, Chirality 2000, 12,442 449.
[14] G. Helmchen,A. Pfaltz, Acc. Chem. Res. 2000, 33,336 345.
[15] D.-R. Hou,K. Burgess, Org. Lett. 1999, 1,1745 1747.
[16] D.-R. Hou,J. Reibenspies,K. Burgess, J. Org. Chem. 2001, 66,206
215.
[17] J. K. Whitesell, Chem. Rev. 1989, 89,1581 1590.
[18] N. Hishida,N. Takada,M. Yoshimura,T. Sonoda,H. Kobayashi, Bull.
Chem. Soc. Jpn. 1984, 57,2600 2604.
[19] M. Brookhart,B. Grant,A. F. Volpe, Organometallics 1992, 11,3920
3922.
Alternative procedure for the preparation of the complexes: Recent work
has shown the following procedure can give improved yields of product
complexes. The ligand (0.5 mmol) was dissolved in CH2Cl2 (10 mL) in a
glove box was added on to a stirred solution of [Ir(cod)Py2]PF4 (0.302 g,
0.5 mmol). The yellow reaction mixture was stirred for 2 h at 258C. The 31
P
NMR spectrum of a small aliquot indicated the complete conversation of
the reaction mixture,as evidenced by the formation of a single peak at
around d ˆ 9.0. Solvent from the reaction mixture was then concentrated to
a quarter of its original volume under reduced pressure. To this was added
anhydrous Et2O (30 mL),and stirred for another 10 minutes to obtain an
orange yellow crystalline precipitate,which was filtered,washed with
diethyl ether and dried under vacuum. The filtrate on cooling also gave a
second crop. The combined yield of the products tend to be in the 90 98%
range.
Typical procedure for the enantioselective hydrogenation: Catalyst 2b
(1.1 mg,0.0007 mmol,0.2 mol%),
trans-1,2-diphenylpropene (60 mg,
0.31 mmol) and CH2Cl2 (120 mL) were added to a vial with a stir bar.
The vial was capped with a septum equipped with a needle outlet and put
into a hydrogen autoclave. The autoclave was sealed and pressurized to
50 bar with H2,and the mixture with stirred for 2 h. The solution was passed
through a short silica gel plug (20% EtOAc/hexanes),then the eluent was
analyzed by GC (1008C; retention time, t1 ˆ 115.3 min, t2 ˆ 118.5 min using
a chiral column prepared by Vigh et al.;[34] (30.7 m  0.25 mm,30% b-tert-
butyldimethylsilyl cyclodextrin derivative in OV-1701-vi of 0.25 mm film
thickness). Control experiments confirmed that reactions performed in
parallel and on the small scale described above give the same results as ones
on a larger scale occupying the whole autoclave. Absolute configuration of
the product in Table 1 was determined by comparison of optical rotations
with those reported by Buchwald.[6] The absolute configurations of the
product shown in Table 2 was not determined but was tentatively assigned
by analogy with the data presented in Table 1. The absolute stereo-
chemistries of the products in Reactions (2) and (3) were not determined.
The absolute configuration of the 2-phenylpropane formed in Reaction (6)
was determined by comparison of optical rotations with those reported by
Paquette et al.,[35] and the absolute configurations of the other products in
Reactions (4) (6) were assumed by analogy with that work.
[20] R. H. Crabtree,M. W. Davis, J. Org. Chem. 1986, 51,2655 2661.
[21] R. Kuwano,K. Sato,T. Kurokawa,D. Karube,Y. Ito,
J. Am. Chem.
Soc. 2000, 122,7614 7615.
[22] W. Chen,L. Xu,C. Chatterton,J. Xiao, Chem. Commun. 1999,1247
1248.
[23] J. H. Davis Jr.,K. J. Forrester, Tetrahedron Lett. 1999, 40,1621 1622.
[24] C. W. Lee, Tetrahedron Lett. 1999, 40,2461 2464.
[25] L. Green,I. Hemeon,R. D. Singer, Tetrahedron Lett. 2000, 41,1343
1346.
[26] L. Xu,W. Chen,J. Xiao, Organometallics 2000, 19,1123 1127.
[27] C. E. Song,E. J. Roh, Chem. Commun. 2000,837 838.
[28] R. Hilgraf,A. Pfaltz, Synlett 1999,1814 1816.
[29] Y. Sun,C. LeBlond,J. Wang,D. G. Blackmond,
J. Am. Chem. Soc.
1995, 117,12647 12648.
[30] Y. Sun,J. Wang,C. LeBlond,R. A. Reamer,J. Laquidara,J. R. Sowa-
Jr.,D. G. Blackmond, J. Organomet. Chem. 1997, 548,65 72.
[31] J. M. Brown,A. E. Derome,G. D. Hughes,P. K. Monaghan, Aust. J.
Chem. 1992, 45,143 153.
[32] H. Rottendorf,S. Sternhell,J. R. Wilmshurst, Aust. J. Chem. 1965, 18,
1759 1773.
[33] A. P. Uijttewaal,F. L. Jonkers,A. v. d. Gen, J. Org. Chem. 1978, 43,
3306 3311.
[34] A. Shitangkoon,G. Vigh, J. Chromatography A 1996, 738,31 42.
Acknowledgements
[35] L. A. Paquette,M. R. Sivik,E. I. Bzowej,K. J. Stanton,
Organo-
We thank Ms. Xiuhua Cui and Dr. Mark T. Powell for preparation of
intermediates in the syntheses of ligands 2g and 2h,Benjamin S. Lane for
help with the GC/MS measurements,and Dr. Bob Taylor at A&M for help
with deuterium NMR experiments. We also thank Dr. Richard Teichman at
Johnson Matthey for valuable discussions. Support for this work was
provided by Johnson Matthey and by The Robert A. Welch Foundation.
metallics 1995, 14,4865 4878.
[36] Crystallographic data (excluding structure factors) for the structures
reported in this paper have been deposited with the Cambridge
Crystallographic Data Centre as supplementary publication no.
CCDC-166174 (2j),-166175 ( 2b),-166176 ( 2i),and -166177 ( 2h).
Copies of the data can be obtained free of charge on application to
CCDC,12 Union Road,Cambridge CB21EZ,UK (fax: ( ‡44)1223-
336-033; e-mail: deposit@ccdc.cam.ac.uk).
[1] R. K. Halterman,K. P. C. Vollhardt,M. E. Welker, J. Am. Chem. Soc.
1987, 109,8105 8107.
Received: May 2,2001 [F3229]
5400
¹ WILEY-VCH Verlag GmbH,D-69451 Weinheim,2001
0947-6539/01/0724-5400 $ 17.50+.50/0
Chem. Eur. J. 2001, 7,No. 24