A bowl-shaped phosphine as a ligand in rhodium-catalyzed hydrosilylation: Rate enhancement by a mono(phosphine) rhodium species

Article abstract of DOI:10.1021/om0503491

Bowl-shaped phosphine (BSP) ligands markedly accelerated the rhodium-catalyzed hydrosilylation of ketones and ketimines as compared with the effect of conventional phosphine ligands such as PPh₃ and P(t-Bu)₃. ³¹P NMR study

Full text of DOI:10.1021/om0503491

3468
Organometallics 2005, 24, 3468-3475
A Bowl-Shaped Phosphine as a Ligand in
Rhodium-Catalyzed Hydrosilylation: Rate Enhancement
by a Mono(phosphine) Rhodium Species
Osamu Niyomura,† Tetsuo Iwasawa, Naoki Sawada, Makoto Tokunaga,
Yasushi Obora, and Yasushi Tsuji*
Catalysis Research Center and Division of Chemistry, Graduate School of Science,
Hokkaido University, CREST, Japan Science and Technology Corporation (JST),
Sapporo 001-0021, Japan
Received May 4, 2005
Bowl-shaped phosphine (BSP) ligands markedly accelerated the rhodium-catalyzed
hydrosilylation of ketones and ketimines as compared with the effect of conventional
phosphine ligands such as PPh3 and P(t-Bu)3. 31P NMR study of a mixture of [RhCl(C2H4)2]2
and phosphines at various P/Rh ratios revealed that coordination of BSP to the rhodium
metal was successfully regulated, and the resultant rhodium complex bearing only one
phosphine ligand (a mono(phosphine) rhodium species) was responsible for the acceleration.
Structural comparison between BSP and the conventional phosphines was carried out using
HF/6-31G(d)-CONFLEX/MM3-optimized structures. The mono(phosphine) rhodium complex
having 1,5-cyclooctadiene as a ligand was isolated, and its X-ray molecular structure was
determined.
Introduction
Phosphines are one of the most important ligands in
a homogeneous transition-metal-catalyzed reaction.1
Therefore, a wide variety of phosphines have been
designed and synthesized to realize highly active and
selective catalytic reactions. As a ligand, the steric
(bulkiness) and electronic effects (basicity) of phosphines
are particularly important. Recently, very bulky and
basic phosphines such as P(t-Bu)3 and tricyclohexyl-
phosphine (PCy3) were found to be effective ligands in
several transition-metal-catalyzed reactions.2 On the
other hand, as a new class of phosphines, the first bowl-
shaped3 phosphine (BSP), tris(2,2′′,6,6′′-tetramethyl-m-
terphenyl-5′-yl)phosphine (1; Figure 1A), was recently
synthesized.4 Parts A and E of Figure 2 show structures
of 1 and P(t-Bu)3 which were optimized by ab initio HF/
† Present address: Division of Chemistry, College of Liberal Arts
and Sciences, Kitasato University.
(1) (a) Homogeneous Catalysis with Metal Phosphine Complexes;
Pignolet, L. H., Ed.; Plenum: New York, 1983. (b) Brandsma, L.;
Vasilevsky, S. F.; Verkruijsse, H. D. Applications of Transition Metal
Catalysts in Organic Synthesis; Springer: Berlin, 1999.
(2) (a) Littke, A. F.; Dai, C.; Fu, G. C. J. Am. Chem. Soc. 2000, 122,
4020-4028. (b) Kirchhoff, J. H.; Dai, C.; Fu, G. C. Angew. Chem., Int.
Ed. 2002, 41, 1945-1947. (c) Littke, A. F.; Fu, G. C. Angew. Chem.,
Int. Ed. 2002, 41, 4176-4211. (d) Kataoka, N.; Shelby, Q.; Stambuli,
J. P.; Hartwig, J. F. J. Org. Chem. 2002, 67, 5553-5556. (e) Walker,
S. D.; Barder, T. E.; Martinelli, J. R.; Buchwald, S. L. Angew. Chem.,
Int. Ed. 2004, 43, 1871-1876.
(3) (a) Maverick, E.; Cram, D. J. In Comprehensive Supramolecular
Chemistry; Vo¨gtle, F., Ed.; Elsevier: Oxford, U.K., 1996; Vol. 2, pp
367-418. (b) Goto, K.; Okumura, T.; Kawashima, T. Chem. Lett. 2001,
1258-1259. (c) Goto, K.; Nagahama, M.; Mizushima, T.; Shimada, K.;
Kawashima, T.; Okazaki, R. Org. Lett. 2001, 3, 3569-3572. (d) Naiki,
M.; Shirakawa, S.; Kon-i, K.; Kondo, Y.; Maruoka, K. Tetrahedron Lett.
2001, 42, 5467-5471. (e) Ohzu, Y.; Goto, K.; Kawashima, T. Angew.
Chem., Int. Ed. 2003, 42, 5714-5717.
(4) (a) Goto, K.; Ohzu, Y.; Sato, H.; Kawashima, T. Abstr. Pap. 15th
Int. Conf. Of Phosphorous Chemistry (Sendai, Japan) 2001, 236. (b)
Goto, K.; Ohzu, Y.; Sato, H.; Kawashima, T. Phosphorus, Sulfur, Silicon
Relat. Elem. 2002, 177, 2179.
Figure 1. Bowl-shaped phosphines.
10.1021/om0503491 CCC: $30.25 © 2005 American Chemical Society
Publication on Web 06/07/2005

Rate Enhancement by a Mono(phosphine) Rhodium Species
Organometallics, Vol. 24, No. 14, 2005 3469
Table 1. Rhodium-Catalyzed Hydrosilylation of
Cyclohexanone with Various Ligandsa
entry
ligand
P/Rh
yield, %b
1
1
1
1
2
3
4
4
4
1
2
3
2
76
97
85
93
92
30
40
<1
27
9
<1
22
32
25
<2
31
<2
2
3
4
5
6
1
2
3
1
2
3
2
7
8
9
PPh3
10
11
12
13
14
15
16
17
P(2-furyl)3
P(o-tol)3
PMes3
PEt3
P(t-Bu)3
PCy3
a Conditions: cyclohexanone (1 mmol), silane (1.2 mmol),
[RhCl(C2H4)2]2 (5.0 µmol), ligand (0.01-0.03 mmol), benzene (1
mL), at room temperature for 3 h. b Determined by GC.
catalytic reactions with BSP as a ligand.8 In this paper,
we carried out the rhodium-catalyzed hydrosilylation9
with BSP as a ligand and found remarkable rate
enhancement10 owing to a mono(phosphine) rhodium
species. Here, we describe a full detail and reaction
mechanism for the rate enhancement.
Figure 2. Optimized structures (top view) of 1-4 and P(t-
Bu)3.
6-31G(d) calculations5 using initial structures optimized
by CONFLEX6a/MM3.6b,c As shown in Figure 2A,E, both
1 and P(t-Bu)3 are obviously very bulky. However, the
nature of the bulkiness is quite different between these
two phosphines. The bulkiness of the bowl-shaped
phosphine (1) occurs on the periphery of the phosphine
(at the rim of the bowl) with substantial empty space
around the phosphorus atom.7 In contrast, P(t-Bu)3 has
severe steric congestion within close proximity of the
phosphorus atom. Owing to this notable feature, BSP
ligands may bring about a unique catalytic environment
in homogeneous transition-metal catalysis, which can-
not be realized with conventional phosphines.
Results
Several BSP species were synthesized. They are tris-
(m-terphenyl)phosphine derivatives having substituents
at the 2,2′′,6,6′′- (1) and 2,2′′,4,4′′,6,6′′-positions (2, 3),
along with a derivative without substituents at these
positions (4) (Figure 1). They were synthesized straight-
forwardly by lithiation of the corresponding m-terphenyl
bromides followed by a reaction with PCl3. The struc-
tures of 1-4 were optimized by HF/6-31G(d)-CONFLEX/
MM3 calculations, as shown in Figure 2A-D, which
shows that all of the phosphines 1-4 have bowl-shaped
structures with the phosphorus atom at the bottom of
the bowl.
With BSP as a ligand, we started a transition-metal-
catalyzed reaction, since there were no precedents for
Hydrosilylation of Ketones and Ketimines. To
examine the efficacy of a BSP as a ligand, the rhodium-
catalyzed hydrosilylation of cyclohexanone with HSiMe2-
Ph as the silane was carried out. The bowl-shaped
phosphines 1-4 and other conventional phosphines
were employed as the ligand at a P/Rh ratio of 2 in the
presence of 0.5 mol % [RhCl(C2H4)2]2 (Table 1). The
reactions were carried out at room temperature for 3 h,
and yields of the products at that time were determined
after desilylation. The use of 1, bearing methyl substit-
uents at the 2,2′′,6,6′′-positions, gave cyclohexanol in
(5) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A.,
Jr.; Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.;
Daniels, A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.;
Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo,
C.; Clifford, S.; Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.;
Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.; Stefanov, B. B.; Liu, G.;
Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R. L.;
Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.;
Gonzalez, C.; Challacombe, M.; Gill, P. M. W.; Johnson, B. G.; Chen,
W.; Wong, M. W.; Andres, J. L.; Head-Gordon, M.; Replogle, E. S.;
Pople, J. A. Gaussian 98, revision A.7; Gaussian, Inc.: Pittsburgh, PA,
1998.
(6) (a) Goto, H. J. Am. Chem. Soc. 1989, 111, 8950-8951. (b) Lii,
J.-H. Allinger, N. L. J. Am. Chem. Soc. 1989, 111, 8566-8575. (c) Lii,
J.-H. Allinger, N. L. J. Am. Chem. Soc. 1989, 111, 8576-8582.
(7) Recently we prepared pyridine ligands bearing a 2,3,4,5-tet-
raphenylphenyl moiety at the 3-position. The pyridine ligands have
steric congestion at remote positions from the coordination center
(nitrogen). This “long-range steric effect” of the pyridine ligands
successfully suppresses Pd black formation in air oxidation of alco-
hols: Iwasawa, T.; Tokunaga, M.; Obora, Y.; Tsuji, Y. J. Am. Chem.
Soc. 2004, 126, 6554-6555.
(8) Palladium complexes bearing the bowl-shaped phosphines were
reported: (a) Matsumoto, T.; Kasai, T. Tatsumi, K. Chem. Lett. 2002,
346-347. (b) Ohzu, Y.; Goto, K.; Kawashima, T. Angew. Chem., Int.
Ed. 2003, 42, 5714-5717.
(9) Ojima, I. In The Chemistry of Organic Silicon Compounds; Patai,
S., Rappoport, Z., Eds.; Wiley: Chichester, U.K., 1989; Chapter 25.
(10) A portion of this work was preliminarily reported: Niyomura,
O.; Tokunaga, M.; Obora, Y.; Iwasawa, T.; Tsuji, Y. Angew. Chem.,
Int. Ed. 2003, 42, 1287-1289.

3470 Organometallics, Vol. 24, No. 14, 2005
Niyomura et al.
Table 2. Hydrosilylation of Ketones and
97% yield (entry 2). Similarly, 2 and 3 having methyl
and/or tert-butyl groups at the 2,2′′,4,4′′,6,6′′-positions
also afforded the product at P/Rh ) 2 in 93% and 92%
yields, respectively (entries 4 and 5). However, the
phosphine without the substituents at these positions
(4) was not efficient in affording the product, giving 40%
yield after 3 h (entry 7). It is noteworthy that these
structurally comparable ligands (1-3 and 4) show
different efficiencies in the reaction. The methyl sub-
stituents at the 2,2′′,6,6′′-positions of 1-3 apparently
play a critical role in realizing the high catalytic activity,
even though these methyl substituents are rather
distant from the phosphorus atoms of 1-3.
Ketiminesa,b
The ligands 1-4 are classified as triarylphosphines.
Therefore, several conventional triarylphosphines were
employed in the reaction for comparison. As a result,
PPh3 and P(2-furyl)311 at P/Rh ) 2 afforded the products
in only 9% and 22% yields, respectively (entries 10 and
12). Furthermore, the typical bulky triarylphosphines
P(o-tolyl)311 and P(Mes)311 afforded the product in 32%
and 25% yields, respectively (entries 13 and 14). Use of
the basic trialkylphosphine PEt3 as the ligand provided
the product in <2% yield (entry 15), and representative
bulky basic trialkylphosphines such as P(t-Bu)3 and
PCy3 were also less effective and resulted in 31% and
<2% yields, respectively (entries 16 and 17). Thus, these
results clearly indicated that 1-3, bearing methyl
substituents at the 2,2′′,6,6′′-positions, specifically en-
hance the catalytic activity.
In the reaction, the P/Rh ratio has a significant effect
on the catalytic activity (Table 1). When 1 was employed
as the ligand, the ratio P/Rh ) 1-3 did not change the
yields substantially (entries 1-3). In contrast, with 4
and PPh3 as the ligands, the ratio affected the yield
considerably (entries 6-8 and entries 9-11). The most
evident difference was observed at P/Rh ) 3: for this
ratio, the catalyst system with 1 was still active in
affording the product in 85% yield (entry 3), whereas
with 4 and PPh3 almost no catalytic activity was
observed (entries 8 and 11). In this way, even though
excess 1 did not affect the catalytic activity, excess 4
and PPh3 deactivated the catalyst significantly.
Even the reactions with 4 and the conventional
ligands were very sluggish at P/Rh ) 2 (entries 7, 10,
and 12-17); these catalysts were still living for a long
time, and the product was obtained in good yields
(>70%) with a much longer reaction time (40-500 h).
Accordingly, the diverse yields in Table 1 were caused
by different reaction rates of the hydrosilylation. Thus,
a kinetic study was carried out at 24 °C with 1, 4, PPh3,
and P(o-tol)3 as the representative ligands under the
same reaction conditions as in Table 1. The reaction
rates with these triarylphosphines showed a first-order
dependence on both the ketone and silane concentra-
tions: rate ) kobs[ketone][silane]. The observed rate
constants (kobs) with 1, 4, PPh3, and P(o-tol)3 were 4.79
( 0.36, 0.154 ( 0.004, 0.0031 ( 0.0013, and 0.169 (
0.003 mol-1 dm3 h-1, respectively. The results indicated
that the catalyst system with 1 realized a remarkable
rate enhancement, in which reactions that were 154,
31, and 28 times faster as compared with those with
PPh3, 4, and P(o-tol)3, respectively, were observed.
a Conditions: ketone (1 mmol), silane (1.2 mmol), [RhCl(C2H4)2]2
(5 µmol), ligand (20 µmol), in benzene (1 mL) at room temperature.
b Conditions: imine (0.25 mmol), silane (0.60 mmol), [RhCl(C2H4)2]2
(3.75 µmol), ligand, in toluene (0.5 mL) at 60 °C. c Yield of alcohol
and amine by GC after the desilylation with HCl/MeOH. d At 50
°C. e Product ratio (neomenthol/menthol). f [Rh(COD)2]BF4 (10
µmol) was used instead of [RhCl(C2H4)2]2 in CH2Cl2 as solvent.
g [RhCl(C2H4)2]2 (2.5 µmol). h Conditions: imine (1 mmol), silane
(1.2 mmol), [RhCl(C2H4)2]2 (5 µmol), ligand (20 µmol), in toluene
(1 mL) at 40 °C.
The rate enhancement was further confirmed in the
hydrosilylation of various silanes and ketones (Table 2).
As the silane, like HSiMe2Ph in Table 1, HSiEt3 afforded
cyclohexanol in much higher yield (81%) with 1 (entry
1) than with 4 (36% yield, entry 2) and PPh3 (13% yield,
entry 3). With HSiMePh2, 1 also provided the product
in higher yield (97%) than with 4 (37% yield) and PPh3
(27% yield) (entries 4-6). Furthermore, the distinct rate
enhancement with 1 as compared with 4 and PPh3 was
also observed in the hydrosilylation of various ketones
such as acetophenone (yields of the product: 96% with
1, 15% with 4, 15% with PPh3, entries 7-9), 2-octanone
(yields of the product: 91% with 1, 41% with 4, 31%
with PPh3, entries 10-12), and (-)-menthone (yields of
the product: 92% with 1, 8% with 4, 12% with PPh3,
entries 13-15). As the catalyst precursor, the cationic
rhodium complex [Rh(COD)2]BF4 (COD ) 1,5-cyclo-
octadiene) also showed a rate enhancement with 1
(entries 16-18).
(11) P(2-furyl)3, P(o-tolyl)3, and P(Mes)3 denote tri-2-furylphosphine,
tri-o-tolylphosphine, and tris(2,4,6-trimethylphenyl)phosphine, respec-
tively.

Rate Enhancement by a Mono(phosphine) Rhodium Species
Organometallics, Vol. 24, No. 14, 2005 3471
Table 3. Hydrosilylation of
2-Methylbenzaldehydea
Table 4. Hydrosilylation of
Benzaldehyde-N-phenyliminea
entry
P/Rh
2
ligand
yield, %b
entry
P/Rh
2
ligand
yield, %b
1
2
3
4
5
6
1
4
79
0
1
2
3
4
5
6
1
4
90
90
82
50
40
7
PPh3
1
78
85
0
PPh3
1
4
4
4
4
PPh3
<1
PPh3
a Conditions: imine (2.5 mmol), silane (3.0 mmol), [RhCl(C2H3)2]2
(2.5 µmol), ligand (10 or 20 µmol), at room temperature for 8 h.
b Yield of amine by GC after desilylation with HCl/MeOH.
a Conditions: aldehyde (1 mmol), silane (1.2 mmol), [RhCl-
(C2H4)2]2 (5 µmol), ligand (20 or 40 µmol), benzene (1 mL) at room
temperature for 6 h. b Yield of alcohol by GC after desilylation with
HCl/MeOH.
Table 5. Hydrosilylation of 1-Hexenea
Since the hydrosilylation of ketones is exceedingly
accelerated with 1 as the ligand, the hydrosilylation of
ketimines was also examined with 1, 4, and PPh3 as
the ligands (Table 2, entries 19-33). In the hydrosily-
lation of benzophenone-N-phenylimine with 1 as the
ligand, the corresponding amine was obtained in 88%
yield at P/Rh ) 2 after the desilylation (entry 20).
Similar to the hydrosilylation of ketones (Table 1 and
entries 1-18 in Table 2), 4 and PPh3 afforded the
hydrosilylated product in only low yields, 6% and 24%,
respectively, at P/Rh ) 2 (entries 23 and 26). Moreover,
in the hydrosilylation of acetophenone-N-phenylimine
(entries 28-30) and cyclohexanone-N-phenylimine (en-
tries 31-33), 1 was also a far more efficient ligand than
4 and PPh3. Notably, as observed with ketones (Table
1), excess 1 did not affect the catalytic activity (entries
21), while excess 4 and PPh3 suppressed the catalyst
activity considerably (entries 24 and 27).
Hydrosilylation of Aldehydes, Aldimines, and
Alkenes. Furthermore, hydrosilylation of aldehydes
was attempted. The representative results with 2-me-
thylbenzaldehyde are shown in Table 3. At the ratio of
P/Rh ) 2, the yields with 1, 4, and PPh3 were high:
90%, 90%, and 82%, respectively (entries 1-3). At P/Rh
) 4, however, the yield with PPh3 was reduced to 7%
(entry 6), while the yields with 1 and 4 were still
moderate (entries 4 and 5). Other various aldehydes
showed more or less similar behavior in the reaction.
Thus, the ligands somehow influence the hydrosilylation
of aldehydes, but the distinct rate enhancement with 1
such as observed with ketones and ketimines was not
evident.
With respect to aldimine as the substrate, the hy-
drosilylation of benzaldehyde-N-phenylimine was car-
ried out (Table 4). At P/Rh ) 2, 1 and PPh3 afforded
the corresponding amine in high yields after the desi-
lylation (entries 1 and 3), but 4 did not provide the
product (entry 2). With excess ligand (P/Rh ) 4), 1 did
not suppress the reaction (entry 4), but almost no
catalytic activity was observed with excess 4 and PPh3
(entries 5 and 6).
Finally, the hydrosilylation of 1-hexene was carried
out (Table 5). At P/Rh ) 2, no particular rate enhance-
ment associated with 1 was evident (entries 1-3).
However, again, with excess ligand at P/Rh ) 4, no
catalytic activity appeared with 4 and PPh3, while
excess 1 did not affect the catalytic reaction.
entry
P/Rh
2
ligand
yield, %b
1
2
3
4
5
6
1
4
93
70
87
86
0
PPh3
1
4
4
PPh3
3
a Conditions: 1-hexene (1 mmol), silane (1.2 mmol), [RhCl-
(C2H4)2]2 (5 µmol), ligand (20 or 40 µmol), toluene (1 mL) at room
temperature for 1 h. b By GC.
In this manner, in the hydrosilylation of aldehydes,
aldimines, and 1-hexene, evident rate enhancement
with 1 was not observed. However, in these reactions,
the behavior of excess ligand is quite different. Espe-
cially, in the hydrosilylation of the aldimine (Table 4)
and 1-hexene (Table 5), excess 1 did not suppress the
catalyst activity at all, while excess 4 and PPh3 almost
inhibit the catalytic activity.
Discussion
The BSP ligands 1-3 gave distinct rate enhance-
ments in the hydrosilylation of ketones and ketimines.
However, the structurally comparable 4 and the con-
ventional ligands listed in Table 1 are not effective.
Basicity and Cone Angle. To elucidate the rationale
for the ligands (1-3) to enhance the catalytic activity,
two critical parameters, basicity and cone angle, were
examined. In this study, the basicity of the phosphines
was evaluated by two established methods. One is
theoretical calculation (HF/6-31G(d) level) of the mo-
lecular electrostatic potential minimum (Vmin) according
to the method of Koga et al.:12 more negative Vmin values
1
indicate more basic phosphines. The other is JP-Se
coupling constants of the corresponding phosphine
selenides SedPR3, which are convenient, but less reli-
able, measures of the basicity of the parent phosphines
PR3: smaller coupling constants correspond to higher
basicities.13 These two values are listed in Table 6. For
the steric effect, the cone angle is often employed as a
(12) Suresh, C. H.; Koga, N. Inorg. Chem. 2002, 41, 1573-1578.
(13) (a) Allen, D. W.; Taylor, B. F. J. Chem. Soc., Dalton Trans. 1982,
51-54. (b) Socol, S. M.; Verkade, G. J. Inorg. Chem. 1984, 23, 3487-
3493. (c) Keay, B. A.; Andersen, N. G. Chem. Rev. 2001, 101, 997-
1030. (d) Alyea, E. C.; Malito, J. Phosphorus, Sulfur, and Silicon Relat.
Elem. 1989, 46, 175-181.

3472 Organometallics, Vol. 24, No. 14, 2005
Niyomura et al.
Table 6. Basicities and Cone Angles of the
Phosphines
Table 7. 31P Resonances of Rh Complexes Bearing
1 or 4 as the Liganda
basicity
corresponding PPh3
entry
1
complex, 31P resonance
complex, 31P resonance
1JP-Se
Hz
, cone angle, hydrosilylation
entry
ligand
Vmin
deg
yield. %d
trans-RhCl(C2H4)2(1)
53.0 (d, 1JP-Rh ) 186)
trans-RhCl(C2H4)(1)2
33.5 (d, 1JP-Rh ) 130)
trans-RhCl(C2H4)2(4)
57.2 (d, 1JP-Rh ) 187)
[RhCl(4)2]2
trans-RhCl(C2H4)2(PPh3)
53.3 (d, 1JP-Rh ) 184)16a
trans-RhCl(C2H4)(PPh3)2
35.4 (d, 1JP-Rh ) 129)16a
see entry 1
1
2
3
4
5
6
7
8
9
1
2
3
4
-41.0
-42.8
-42.5
-41.6
770
741
743
766
205
210
223
193
168
152
218
243
166
188
181
97
93
92
40
27
22
32
25
<2
<2
31
2
3
4
PPh3
-42.2 735a
[RhCl(PPh3)2]2
P(2-furyl)3 -38.5 793b
56.9 (dd, J ) 191, 51)
61.0 (dd, J ) 193, 51)
RhCl(COD)(1)
51.9 (d, 1JP-Rh ) 193)16a
P(o-tol)3
PMes3
PEt3
-44.4
-44.6
732
c
-
5
6
7
RhCl(COD)(PPh3)
31.4 (d, 1JP-Rh ) 152)16b
see entry 5
-49.8 705a
30.5 (d, 1JP-Rh ) 153)
RhCl(COD)(4)
10 PCy3
11 P(t-Bu)3
-54.4
-54.4
674
686
34.2 (d, 1JP-Rh ) 153)
RhCl(4)3
a Data in ref 13a. b Data in ref 13b. c Trimesitylphosphine
selenide is not available.13d d Yields of the hydrosilylation of
cyclohexanone in Table 1.
RhCl(PPh3)3
39.5 (dd, J ) 140, 39)
54.9 (dt, J ) 186, 39)
31.5 (dd, J ) 142, 38)16c
48.0 (dt, J ) 189, 38)
a In benzene-d6 at room temperature; resonances given in ppm
and J values given in Hz.
good criterion. Originally, however, Tolman measured
them on “folded-back” (most-packed) CPK molecular
models to minimize arbitrariness.14 Consequently, they
are appreciably smaller than those of the real phos-
phines. In the present study, for more accurate discus-
sion, we measured the cone angles according to Tolman’s
definition using structures optimized by HF/6-31G(d)-
CONFLEX/MM3 calculations (Table 6).
Scheme 1. 31P NMR Study on a Mixture of
[RhCl(C2H4)2]2 and the Bowl-Shaped Phosphines
(P)
The basicity and/or the cone angle have so far suc-
cessfully rationalized the influence of monodentate
phosphines in a wide variety of homogeneous catalysis.
However, in the present study, the effective (1-3,
entries 1-3 in Table 6) and ineffective (4, entry 4) BSP
ligands have essentially similar basicities comparable
to that of PPh3 (entry 5). They are within the range
expected for the conventional triarylphosphines (entries
5-8), which did not accelerate the hydrosilylation of
ketones at all (Table 1). Therefore, a simple explanation
with basicity is not operative. As for the steric effect,
the cone angles of the effective (1) and ineffective ligands
(4) are comparable. Furthermore, a particular range of
the cone angles for the effective ligands cannot be
determined. Thus, neither the basicity nor the cone
angles listed in Table 6 showed evident correlations with
the aforementioned efficacy of the ligands.
results are shown in Scheme 1A. With P/Rh ) 1, only
the mono(phosphine) rhodium complex trans-RhCl-
(C2H4)2(1) formed, whose 31P resonance is in excellent
agreement with that of the corresponding PPh3 com-
plex16a (Table 7, entry 1). At P/Rh ) 2, still the mono-
(phosphine) rhodium complex was the predominant
(90% among rhodium-phosphine complexes) species
with only a small amount of the bis(phosphine) rhodium
complex, trans-RhCl(C2H4)(1)2, which also coincides
with the 31P resonance of the corresponding PPh3
complex16a (Table 7, entry 2). Surprisingly, even at P/Rh
) 3, the mono(phosphine) complex remained dominant
(90%) with a greater amount of free (uncoordinated) 1
at -8.4 ppm. Here, no tris(phosphine) complex formed
at all. Thus, with 1, over the range of P/Rh ) 1-3, the
mono(phosphine) rhodium complex was the prevailing
(>90%) species (Scheme 1A).
31P NMR Study on a Mixture of [RhCl(C2H4)2]2
and 1 or 4. Since the excess ligands affect the catalytic
reaction in a different way, as observed in Tables 1-5,
the coordination behavior of the phosphines toward the
catalyst precursor, [RhCl(C2H4)2]2, was examined. 31P
NMR spectra are most diagnostic of the coordination,
since a phosphine on a Rh (103Rh, I ) 1/2, 100%)15 shows
1
informative JP-Rh coupling and the 31P resonance can
be unambiguously assigned according to the corre-
sponding PPh3 complexes reported in the literature.16
The 31P resonances of the Rh complexes bearing 1 or 4
are compared with the corresponding PPh3 complexes
in Table 7. At first, 1 and [RhCl(C2H4)2]2 were mixed at
various P/Rh ratios for 2 h at room temperature, and
31P NMR spectra of the mixture were measured. The
On the other hand, with 4 as BSP, the same 31P NMR
measurement was carried out (Scheme 1B). In contrast,
the coordination behavior of 4 was quite different from
that of 1. At P/Rh ) 1, a much smaller amount (17%) of
the mono(phosphine) rhodium complex trans-RhCl-
(C2H4)2(4) (Table 7, entry 3) formed, and the bis-
(14) Tolman, C. A. Chem. Rev. 1977, 77, 313-348.
(15) Brevard, C.; Granger, P. Handbook of High-Resolution Multi-
nuclear NMR; Wiley-Interscience: New York, 1981; p 158.
(16) (a) Naaktgeboren, A. J.; Nolte, R. J. M.; Drenth, W. J. Am.
Chem. Soc. 1980, 102, 3350-3354. (b) Elsevier, C. J.; Kowall, B.;
Kragten, H. Inorg. Chem. 1995, 34, 4836-4839. (c) Brown, T. H.;
Green, P. J. J. Am. Chem. Soc. 1970, 92, 2359-2362.

Rate Enhancement by a Mono(phosphine) Rhodium Species
Organometallics, Vol. 24, No. 14, 2005 3473
Scheme 2. 31P NMR Study on a Mixture of
[RhCl(COD)]2 and the Bowl-Shaped Phosphines
(P)
Figure 3. ORTEP drawing of [RhCl(COD)(1)] with ther-
mal ellipsoids at the 50% probability level. Hydrogen atoms
are omitted for clarity.
(phosphine) rhodium complex [RhCl(4)2]2 was the main
species (83% as a mononuclear rhodium complex). 31P
resonances of [RhCl(4)2]2 appeared as two inequivalent
resonances, possibly due to a distortion caused by the
large steric size of 4, while the corresponding PPh3
complex showed equivalent 31P resonances (Table 7,
entry 4). However, the formation of [RhCl(4)2]2 was
unambiguously identified by an FD mass spectrum of
the reaction mixture (m/z 3152, M+, 100%; m/z 1576,
Table 8. Summary of Crystal Structure Data for
the Complex RhCl(COD)(1)a
formula
C76H77Cl7PRh
1372.49
monoclinic
-160
fw
cryst syst
data collecn T, °C
space group
P21/c (No. 14)
16.4206(9)
19.2472(9)
22.0247(11)
105.540(2)
6706.5(6)
4
a, Å
b, Å
c, Å
M
2+, 7%: see the Supporting Information). At P/Rh )
â, deg
V, Å3
3, only the bis(phosphine) rhodium complex was af-
forded, along with a considerable amount of free 4 (33%).
As for PPh3, the reaction with [RhCl(C2H4)2]2 afforded
a considerable amount of the tris(phosphine) complex,
[RhCl(PPh3)3],16c even at P/Rh ) 2. Thus, the 31P NMR
measurements at the different P/Rh ratios revealed that
the coordination behaviors of 1, 4, and PPh3 toward the
catalyst precursor [RhCl(C2H4)2]2 were quite different.
Using 1 as the ligand, the mono(phosphine) rhodium
species was predominant even at P/Rh ) 3.
Z
dcalcd, g cm-3
1.359
cryst size, mm
habit
0.27 × 0.17 × 0.07
prismatic
5.99
µ, cm-1
transmissn factors
no. of intensities (unique, Ri)
no. of intensities >3.00σ(I)
no. of params
Rb
0.9458-1.0000
66 233 (15 254, 0.053)
9936
801
0.067
b
Rw
0.096
GOF
0.836
The BSP ligand 1 realized a far better catalytic
activity in the hydrosilylation of ketones and ketimines
as compared with that of 4 and PPh3 (Table 2). Fur-
thermore, unlike 4 and PPh3, 1 did not lower the
catalytic activity even at high P/Rh ratios (Tables 3 and
4). Thus, the mono(phosphine) complex RhCl(C2H4)2 (1)
might be responsible for the high catalytic activity.
Therefore, isolation of the mono(phosphine) ethylene
complex was attempted. However, all the trials to isolate
RhCl(C2H4)2(1) were unsuccessful, since the ethylene
ligand is highly labile. Accordingly, we employed [RhCl-
(COD)]2 (COD ) 1,5-cyclooctadiene) in place of [RhCl-
(C2H4)]2 in the reaction with 1 to isolate a mono-
(phosphine) COD complex.
a Conditions: radiation, graphite-monochromated, Mo KR, λ )
0.710 70 Å; scan type ω-2θ; 2θmax ) 55.0°. b R ) ∑||Fo| - |Fc||/
∑|Fo|; Rw ) [∑w(|Fo| - |Fc|)2/∑wFo2]1/2, w ) [0.0010Fo2 + 3.0000σ(Fo2)
+ 0.5000]-1
.
obtained exclusively even in the presence of excess 1.
The mono(phosphine) complex RhCl(COD)(1) was iso-
lated successfully from the solution, and single crystals
suitable for X-ray analysis were obtained by slow
diffusion of hexane into a CHCl3 solution of the complex
at room temperature. An X-ray crystallographic analysis
was carried out, and the molecular structure of RhCl-
(COD)(1) is shown in Figure 3. The crystallographic data
and selected distances and angles are given in Tables 8
and 9, respectively. Figure 3 clearly indicates that the
complex bears only one phosphine ligand (1) and the
rhodium is settled in the bowl made by 1. The isolated
RhCl(COD)(1) showed the same 31P resonance as ob-
served in Scheme 2A (entry 5, Table 7). Furthermore,
this isolated mono(phosphine) complex was active as the
catalyst (5 mol %) in the hydrosilylation of cyclohex-
anone with HSiMe2Ph at room temperature for 5 h to
Isolation of the Mono(phosphine) Complex RhCl-
(COD)(1). As shown in Scheme 2A, the similar reaction
of [RhCl(COD)]2 with 1 at P/Rh ) 1 in benzene-d6 at 60
°C for 1 h afforded the mono(phosphine) complex RhCl-
(COD)(1), as judged by the 31P resonance16b (Table 7,
entry 5). Surprisingly, even at P/Rh ) 2, 3, only the 31
P
resonances of the mono(phosphine) complex and free 1
appeared, indicating the mono(phosphine) complex was

3474 Organometallics, Vol. 24, No. 14, 2005
Niyomura et al.
1B and 2B). Even though the deeper bowl ligand keeps
out the other identical deeper ligand, there is sufficient
space for a catalytic reaction around the metal (catalyst)
center of the resultant mono(phosphine) complex (Fig-
ure 3). This unique catalytic environment of the mono-
(phosphine) complex realizes the pronounced rate en-
hancement in the hydrosilylation reaction of ketones
and ketimines (Tables 1 and 2). Conventional sterically
bulky phosphine ligands such as P(t-Bu)3 and PCy3 are
also known to generate low-coordinated (mainly bis)
phosphine rhodium species17 by steric congestion in the
close proximity of the phosphorus atom of the phos-
phines. However, such steric hindrance near a catalytic
center will obstruct coordination sites for coming sub-
strate molecules. Consequently, overall catalytic activity
with these ligands is lowered, as shown in Table 1.
In conclusion, the distinct rate enhancements of the
hydrosilylation of ketones and ketimines are realized
by utilizing BSP in the presence of a rhodium catalyst.
The coordination of the deeper bowl phosphine is
successfully regulated and the resultant mono(phos-
phine) species with a substantial space around the
catalyst center would be responsible for the rate en-
hancement.
Figure 4. Diameters, depths, and dihedral angles of the
bowl-shaped phosphines.
Table 9. Selected Bond Distances and Angles for
RhCl(COD)(1)
Distances (Å)
P(1)-C(1)
1.824(6)
1.817(6)
1.837(6)
2.285(1)
2.365(2)
2.208(7)
2.229(6)
2.115(7)
2.131(7)
C(67)-C(68)
C(68)-C(69)
C(69)-C(70)
C(70)-C(71)
C(71)-C(72)
C(72)-C(73)
C(73)-C(74)
C(74)-C(67)
1.38(1)
1.52(1)
1.53(1)
1.514(9)
1.41(1)
1.525(9)
1.55(1)
1.52(1)
P(1)-C(23)
P(1)-C(45)
P(1)-Rh(1)
Rh(1)-Cl(1)
Rh(1)-C(67)
Rh(1)-C(68)
Rh(1)-C(71)
Rh(1)-C(72)
Angles (deg)
Rh(1)-P(1)-C(1)
113.7(2)
P(1)-Rh(1)-C(67) 159.2(2)
P(1)-Rh(1)-C(68) 164.4(2)
Rh(1)-P(1)-C(23) 110.9(2)
Rh(1)-P(1)-C(45) 116.6(2)
Experimental Section
P(1)-Rh(1)-C(71)
93.0(2)
96.1(2)
P(1)-Rh(1)-Cl(1)
88.04(6) P(1)-Rh(1)-C(72)
General Considerations. All reactions were performed
under an argon atmosphere using Schlenk techniques. Sol-
vents were dried and purified prior to use by standard
methods.18 1H NMR (400.13 MHz), 13C NMR (100.61 MHz),
and 31P NMR spectra (161.98 MHz) were recorded on a Bruker
ARX 400 instrument. The 1H NMR data are referenced relative
Dihedral Angles (deg)
C(2)-C(3)-C(7)-C(8)
89.1(2)
78.0(2)
72.8(3)
66.4(3)
73.4(2)
81.7(2)
C(4)-C(5)-C(15)-C(20)
C(24)-C(25)-C(29)-C(34)
C(26)-C(27)-C(37)-C(42)
C(46)-C(47)-C(51)-C(56)
C(48)-C(49)-C(59)-C(64)
to the residual protiated solvent (7.26 ppm) in CDCl3. The 13
C
NMR chemical shifts are reported relative to CDCl3 (77.0 ppm).
The 31P NMR data are given relative to external 85% H3PO4.
GC analysis was carried out using an Agilent GC 6850
equipped with an Agilent HP-1 column (length 30 m, 0.32 mm
afford the product in 92% yield. In contrast, in the
reaction of 4 with [RhCl(COD)]2 at 60 °C (Scheme 2B),
a considerable amount (23% and 47%, respectively, as
a Rh complex) of the Wilkinson-type complex RhCl(4)3
formed at P/Rh ) 2, 3 in addition to the mono-
(phosphine) complex (Table 7, entry 6). Here, RhCl(4)3
showed the same characteristic 31P resonances (AB2
i.d.). TLC analyses were performed on Merck silica gel 60 F254
.
Column chromatography was carried out with silica gel (Wako
Wakogel C-200). FD mass spectra were recorded on a JEOL
JMS-SX102A instrument at the GC-MS & NMR Laboratory
of the Faculty of Agriculture, Hokkaido University. Elemental
analysis was performed at the Center for Instrumental
Analysis of Hokkaido University.
16c
system) as RhCl(PPh3)3 (Table 7, entry 7), and the
formation of RhCl(4)3 was further confirmed by an FD
General Procedure for the Rhodium-Catalyzed Hy-
drosilylation. A typical reaction procedure in Tables 1-5 is
as follows. Under an argon atmosphere, a phosphine ligand
and a rhodium complex were placed in a 20 mL Schlenk flask.
Anhydrous, degassed solvent was added by a syringe, and the
mixture was stirred at room temperature for 2 h. Then, a
substrate, tridecane (as an internal standard, 0.25 equiv with
respect to the substrate), and a silane were added in this order
by syringe. After the reaction, hydrolysis was performed by
adding 1% HCl/MeOH and yields of the products were deter-
mined by GC analysis relative to the internal standard
(tridecane).
mass spectrum, m/z 2294 (M+).
Structural Difference between 1 and 4. The 31P
NMR study revealed that 1 favored the mono(phos-
phine) complexes, while 4 provided the bis- and/or tris-
(phosphine) species (Schemes 1 and 2). To explain this
quite different coordination behavior, the optimized
structures of 1 and 4 shown in Figure 2A,D were
examined in detail. The diameters (l in Figure 4) of the
bowls (1.99 nm for 1 and 1.95 nm for 4) are similar.
However, due to methyl substituents at the 2,2′′,6,6′′-
positions of 1, the dihedral angles (θ) between the
phosphinated phenyl ring (ring A) and the phenyl ring
in the meta positions (ring B) are quite different: 85.3
( 0.5° for 1 and 44.7 ( 0.3° for 4. Due to the large
dihedral angles of 1, a rim of the bowl comes up, and
consequently 1 becomes a much deeper bowl than 4:
depths (d in Figure 4) of the bowls are 0.208 nm for 1
and 0.132 nm for 4. The deeper bowl ligand would
exclude the other identical deeper ligand on coordination
(Schemes 1A and 2A), whereas two or three shallower
bowls (4) can coordinate on the same metal (Schemes
Synthesis of the BSP. Although 14,8 and 410 have already
appeared in the literature, a detailed preparation method and
spectral data of 1 have not been reported. Therefore, those data
of 1 are described as well as those of the novel phosphines 2
and 3.
(17) (a) Yoshida, T.; Otsuka, S. Inorg. Chim. Acta 1978, 29, L257-
L259. (b) Waal, D. J. A. D.; Robb, W. Inorg. Chim. Acta 1978, 26, 91-
96. (c) Gusev, D. G.; Bakhmutov, V. I.; Grushin, V. V.; Vol’pin, M. E.
Inorg. Chim. Acta 1990, 175, 19-21. (d) van Gaal, H. L. M.; van den
Bekerom, F. L. A. J. Organomet. Chem. 1977, 134, 237-248.
(18) Armarego, W. L. F.; Perrin, D. D. Purification of Laboratory
Chemicals, 4th ed.; Butterworth-Heinemann: Oxford, U.K., 1997.

Rate Enhancement by a Mono(phosphine) Rhodium Species
Organometallics, Vol. 24, No. 14, 2005 3475
Tris(2,2′′,6,6′′-tetramethyl-m-terphenyl-5′-yl)phos-
phine (1). To a solution of 5′-bromo-2,2′′,6,6′′-tetramethyl-m-
terphenyl19 (2.05 g, 5.61 mmol) in 33 mL of THF was added
3.53 mL (5.61 mmol) of n-butyllithium (1.59 M solution in
n-hexane) dropwise over 15 min at -78 °C. The solution was
stirred at that temperature for an additional 1 h, and
phosphorus trichloride (0.257 g, 1.87 mmol) in toluene (5 mL)
was added dropwise over 20 min to the cooled solution. The
reaction mixture was further stirred at -78 °C for 1.5 h and
at room temperature for 15 h. The solvent and all the volatiles
were removed in vacuo, and then degassed toluene (32 mL)
and water (8 mL) were added to the residue with vigorous
stirring. The organic layer was separated and dried over
anhydrous MgSO4. The solution was filtered under an argon
atmosphere, and volatiles were evaporated in vacuo to give
crude products as a viscous material. The product, 1, was
obtained in 71% yield (1.18 g) as white microfine crystals by
recrystallization from CHCl3/MeOH (1/3). 1H NMR (CDCl3):
[RhCl(C2H4)2]2 (3.9 mg, 0.01 mmol) and phosphine (0.02-0.06
mmol). The tube was evacuated and filled with argon, and
degassed C6D6 (0.7 mL) was added under an argon flow to
afford a homogeneous solution. The 31P NMR spectrum was
measured after 2 h at room temperature.
31P NMR Study on a Mixture of [RhCl(COD)]2 and the
Phosphine (Scheme 2). In a NMR sample tube were placed
[RhCl(COD)]2 (4.9 mg, 0.01 mmol) and phosphine (0.02-0.06
mmol). The tube was evacuated and filled with argon, and
degassed C6D6 (0.7 mL) was added. The tube was placed in
an oil bath at 60 °C for 1 h, and the 31P NMR spectrum was
measured at room temperature.
Isolation of RhCl(COD)(1). To a 20 mL Schlenk flask
charged with [RhCl(COD)]2 (2.5 mg, 5 µmol) and ligand 1 (8.9
mg, 10 µmol) was added the degassed benzene (0.39 mL) under
an argon atmosphere. After it was stirred for 1 h at 60 °C, the
mixture was cooled to room temperature and concentrated in
vacuo to give yellow solids. Single crystals of RhCl(COD)(1)
suitable for X-ray structural analysis were obtained by slow
3
4
δ 7.13 (dd, JPH ) 7.5 Hz, JHH )1.4 Hz, 6H), 7.12-7.09 (m,
4
6H), 7.04-7.00 (m, 12H), 6.88 (t, JHH ) 1.4 Hz, 3H), 1.87 (s,
36H). 13C NMR (CDCl3): δ 141.6 (d, 3JCP ) 6.9 Hz), 141.1, 137.7
diffusion of hexane into a CHCl3 solution of RhCl(COD)(1). 31
P
1
2
(d, JCP ) 12.7 Hz), 135.6, 132.5, (d, JCP ) 18.9 Hz), 130.1,
127.1, 127.0, 20.7. 31P NMR (CDCl3): δ -7.0. FD-MS: m/z 886
(M+). Mp: 271.5-273.0 °C. Anal. Calcd for C66H63P: C, 89.35;
H, 7.16. Found: C, 89.42; H, 7.40.
1
NMR (162 MHz, C6D6): δ 30.5 (d, JP-Rh ) 153 Hz). FD-MS:
m/z 1133 ([RhCl(COD)(1)]+). Anal. Calcd for C74H75ClPRh: C,
78.40; H, 6.67; Cl, 3.27. Found: C, 78.34; H, 6.85; Cl, 3.14.
The isolated complex was catalytically active. To a flask
charged with the isolated RhCl(COD)(1) (11.3 mg, 0.01 mmol)
was added degassed benzene (1 mL), and the solution was
stirred for 10 min. Then, cyclohexanone (19.6 mg, 0.2 mmol)
and dimethylphenylsilane (32.7 mg, 0.24 mmol) were added
to the solution, and the reaction was carried out for 5 h at
room temperature. Cyclohexanol was obtained in 92% (GC)
yield after the desilylation with 2 M HCl/MeOH (2 mL).
Tris(2,2′′,4,4′′,6,6′′-hexamethyl-m-terphenyl-5′-yl)phos-
phine (2). In the same manner as 1, the phosphine 2 was
prepared from 5′-bromo-2,2′′,4,4′′,6,6′′-hexamethyl-m-terphe-
nyl20 (2.20 g, 5.61 mmol). By recrystallization from CHCl3/
MeOH (1/6), 2 was afforded in 60% yield (1.09 g) as a white
1
3
4
powder. H NMR (CDCl3): δ 7.13 (dd, JPH ) 7.6 Hz, JHH
)
1.5 Hz, 6H), 6.98-6.97 (m, 3H), 6.90 (s, 12H), 2.31 (s, 18H),
1.86 (s, 36H). 13C NMR (CDCl3): δ 141.4 (d, JCP ) 6.9 Hz),
3
138.4, 137.5 (d, 1JCP ) 12.3 Hz), 136.4, 135.5, 132.7 (d, 2JCP
)
X-ray Structure Determination of RhCl(COD)(1)‚
2CHCl3. The data were collected with Mo KR radiation (λ )
0.710 70 Å) at -160 °C on a Rigaku Saturn CCD area detector
to a maximum 2θ value of 55.0°. The structure was solved by
direct methods using the program SIR9222 and expanded using
Fourier techniques. Non-hydrogen atoms, except for those of
two disordered CHCl3 molecules, were refined anisotropically.
Hydrogen atoms were refined using the riding model. The final
cycle of full-matrix least-squares refinement on F was based
on 9936 observed reflections (I > 3.0σ(I)) and 801 variable
parameters. A Sheldrick weighting scheme was used. Neutral
atom scattering factors were taken from Cromer and Waber.23
Anomalous dispersion effects were included in Fc.24 All calcula-
tions were performed using the CrystalStructure crystal-
lographic software package (version 3.6.0).25
18.8 Hz), 130.6, 127.9, 21.0, 20.6. 31P NMR (CDCl3): δ -6.3.
FD-MS: m/z 971 (M+, 100%). Anal. Calcd for C72H75P: C,
89.03; H, 7.78. Found: C, 88.59; H, 7.67.
5′-Bromo-4,4′′-di-tert-butyl-2,2′′,6,6′′-tetramethyl-m-ter-
phenyl. This new compound was prepared from (4-tert-butyl-
2,6-dimethylphenyl)magnesium bromide and 1-iodo-2,4,6-
tribromobenzene in the same manner as the preparation of
5′-bromo-2,2′′,4,4′′,6,6′′-hexamethyl-m-terphenyl.20 1H NMR
(CDCl3): δ 7.36 (s, 2H), 7.17 (s, 4H), 6.98 (s, 1H), 2.15 (s, 12H),
1.40 (s, 18H). 13C NMR (CDCl3): δ 150.6, 143.8, 137.9, 135.7,
130.9, 129.7, 124.5, 122.8, 34.8, 31.9, 21.6. FD-MS: m/z 478
(M+), 476 (M+). Anal. Calcd for C30H37Br: C, 75.46; H, 7.81;
Br, 16.73. Found: C, 75.09; H, 7.76; Br, 16.78.
Tris(4,4′′-di-tert-butyl-2,2′′,6,6′′-tetramethyl-m-terphe-
nyl-5′-yl)phosphine (3). The phosphine 3 was prepared from
5′-bromo-4,4′′-di-tert-butyl-2,2′′,6,6′′-tetramethyl-m-terphe-
nyl (700 mg, 1.46 mmol) in the same manner as 1. By
recrystallization (twice) from CHCl3/toluene/MeOH (1/1/8), 3
was obtained in 50% yield (296 mg) as a white powder. 1H
NMR (CDCl3): δ 7.10 (dd, 3JPH ) 7.6 Hz, 4JHH ) 1.5 Hz, 6H),
7.04 (s, 12H), 6.90-6.91 (m, 3H), 1.87 (s, 36H), 1.33 (s, 54H).
13C NMR (CDCl3): δ 141.8 (d, 3JCP ) 6.9 Hz), 138.9, 137.9 (d,
1JCP ) 12.5 Hz), 135.6, 133.2, (d, 2JCP ) 18.7 Hz), 130.9, 124.5,
34.7, 31.8, 21.4. 31P NMR (CDCl3): δ - 6.6. FD-MS: m/z 1224
(M+). HRMS (FD): m/z calcd for C90H111P 1222.8423, found
1222.8430. Anal. Calcd for C90H111P: C, 88.33; H, 9.14.
Found: C, 87.77; H, 9.32.
Supporting Information Available: A figure giving the
FD mass spectrum of [RhCl(4)2]2 and crystallographic data for
RhCl(COD)(1) (as a CIF file). This material is available free
of charge via the Internet at http://pubs.acs.org.
OM0503491
(21) Tashiro, M.; Yamato, T. J. Chem. Soc., Perkin Trans. 1 1979,
176-179.
(22) Altomare, A.; Cascarano, G.; Giacovazzo, C.; Guagliardi, A.;
Burla, M.; Polidori, G.; Camalli, M. J. Appl. Crystallogr. 1994, 27, 435.
(23) Cromer, D. T.; Waber, J. T. In International Tables for X-ray
Crystallography; Kynoch Press: Birmingham, U.K., 1974; Vol. IV.
(24) Ibers, J. A.; Hamilton, W. C. Acta Crystallogr. 1964, 17, 781.
(25) (a) Crystal Structure Analysis Package; Rigaku and Rigaku/
MSC (2000-2004), 9009 New Trails Dr., The Woodlands, TX 77381.
(b) Watkin, D. J.; Prout, C. K.; Carruthers, J. R.; Betteridge, P. W.
Chemical Crystallography Laboratory, Oxford, U. K.
31P NMR Study on a Mixture of [RhCl(C2H4)2]2 and the
Phosphine (Scheme 1). In a NMR sample tube were placed
(19) Vinod, T. K.; Hart, H. J. Org. Chem. 1991, 56, 5630-5640.
(20) Ritleng, V.; Yandulov, D.; Weare, W. W.; Schrock, R. R.; Hock,
A. S. Davis, W. M. J. Am. Chem. Soc. 2004, 126, 6150-6163.