Self-assembled monolayers of compact phosphanes with alkanethiolate pendant groups: Remarkable reusability and substrate selectivity in Rh catalysis

Article abstract of DOI:10.1002/anie.200800884

(Chemical Equation Presented) Selective surface: A self-assembled monolayer consisting of a compact trialkylphosphane with an alkanethiolate pendant group was prepared on a gold surface. The rhodium complex of this system (see picture) showed high catalyt

Full text of DOI:10.1002/anie.200800884

Angewandte
Chemie
DOI: 10.1002/anie.200800884
Heterogeneous Catalysis
Self-Assembled Monolayers of Compact Phosphanes with
Alkanethiolate Pendant Groups: Remarkable Reusability and
Substrate Selectivity in Rh Catalysis**
Kenji Hara,* Ryuto Akiyama, Satoru Takakusagi, Kohei Uosaki, Toru Yoshino, Hiroyuki Kagi,
and Masaya Sawamura*
Building-up an assembly of designed molecules on a well-
defined surface would be an interesting approach towards the
development of highly efficient catalysts. Numerous studies
have reported on the modification of a gold surface with
alkanethiolate molecules to afford self-assembled monolayers
(SAMs) with functions that depend on the design of the
terminal moiety.[1,2] However, attempts to utilize the SAM-
modified gold surface to prepare catalysts for organic syn-
thesis have not been fully established.[3–6]
The preparation of the phosphane monolayer 1 and its
complexation with rhodium is illustrated in Scheme 1.
Herein, we report the modification of gold surfaces with a
caged, compact trialkylphosphane (SMAP)[7] which bears an
alkanethiolate pendant group. These surfaces ([Au]-SMAP, 1)
were utilized in a chip form for rhodium-catalyzed dehydro-
genative alcohol silylation. Compared with the corresponding
surface catalyst bearing the conventional Ph2P-type coordi-
nating groups[5,6] and with homogeneous Rh catalysts, the
Rh catalyst [Au]-SMAP-Rh (2) prepared from
1 was
extremely robust and highly efficient in terms of reusability.[8]
[*] Prof. K. Hara
Catalysis Research Center
Hokkaido University
Sapporo 001-0021 (Japan)
Scheme 1. Preparation of [Au]-SMAP (1) and [Au]-SMAP-Rh (2).
Fax: (+81)11-706-9136
E-mail: hara@cat.hokudai.ac.jp
Homepage: http://www.cat.hokudai.ac.jp/fukuoka/e_top.htm
Accordingly, Ph-SMAP sulfide (3) was transformed in five
steps into the phosphane 4 with a decanethiol pendant group.
Immersion of the gold surface (evaporated on glass) in a
1.0 mm solution of 4 in EtOH for 18 h, followed by washing
with EtOH, afforded [Au]-SMAP (1; 130.1 eV in the P 2p
region of the X-ray photoelectron spectrum). Complexation
of 1 with rhodium to obtain [Au]-SMAP-Rh (2) was carried
out by immersion of the phosphanated chip 1 into a 5.0 mm
solution of [{RhCl(C2H4)2}2] in benzene for 15 minutes
followed by washing with EtOH.
Attachment of the Rh atoms was confirmed by surface
analysis of 2 by XPS, which showed a signal at 307.6 eV in the
Rh 3d5/2 region and a signal at 198.6 eV in the Cl 2p region.
Analysis in the P 2p region showed a signal at 131.4 eV, which
is significantly higher in energy than that of the P atom of 1 (at
130.1 eV), thus suggesting the coordination of the P atom to
the Rh center. The P/S/Rh/Cl elemental ratio of 2 was
calculated to be 1:1:0.8:0.7. Accordingly, 2 is considered to
exist in the mono(phosphane)–rhodium form (P/Rh 1:1).[9]
The concentration of Rh atoms (per geometric surface
area) in 2 was determined to be (0.63 Æ 0.02) nmolcmÀ2 by
inductively coupled plasma mass spectrometry (ICP-MS).
R. Akiyama, Dr. S. Takakusagi, Prof. K. Uosaki, Prof. M. Sawamura
Department of Chemistry
Faculty of Science
Hokkaido University
Sapporo 060-0810 (Japan)
Fax: (+81)11-706-3749
E-mail: sawamura@sci.hokudai.ac.jp
Homepage:
http://barato.sci.hokudai.ac.jp/~orgmet/indexEng1.html
T. Yoshino, Prof. H. Kagi
Geochemical Laboratory
Graduate School of Science
The University of Tokyo
Hongo, Tokyo 113-0033 (Japan)
[**] Prof. K. Shimazu, Prof. W. Chun, and Dr. K. Nakata (Hokkaido
University) are acknowledged for XPS measurements. This work
was financially supported by PRESTO (JST) (to M.S.), Grant-in-Aids
for Scientific Research (B) (to M.S.), Grants-in-Aid for Scientific
Research on Priority Areas (to M.S. and K.H.), and the Global COE
Program (to K.H.; Project No. B01: Catalysis as the Basis for
Innovation in Materials Science) from MEXT (Japan).
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2008, 47, 5627 –5630
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5627

Communications
Table 1: Rhodium-catalyzed dehydrogenative silylation of EtOH.[a]
Although we have not been successful in direct quantification
of the density of P atoms in 1 and 2, the density of the Au(111)
single crystal surface modified with the phosphane sulfide
derivative of 4 was accurately determined by electrochemical
reductive desorption of the attached thiolate molecules.[10]
The analysis indicated that the surface density of the thiolates,
which should be identical to the density of P atoms, is
0.69 nmolcmÀ2 (4.2 moleculesnmÀ2; the corresponding den-
sity for the linear alkanethiolate SAM on Au(111) is
4.6 moleculesnmÀ2), which strongly suggests the formation
of a monolayer in the closest packing arrangement (see the
Supporting Information).
Surface modification of a decanethiol derivative bearing a
conventional Ph2P-type coordinating group to afford [Au]-
Ph2P-Rh was also conducted for comparison. ICP-MS analysis
showed a density of Rh atoms (per geometric surface area) of
(0.71 Æ 0.06) nmolcmÀ2, which is almost identical to the value
(0.63 nmolcmÀ2) for 2.
We first examined [Au]-SMAP-Rh (2) for its catalytic
activity towards the hydrosilylation of ketones. Upon reaction
with cyclohexanone and Me2PhSiH (1.1 equiv) in hexane
(ketone/Rh 75000:1), 2 showed exceptionally high catalytic
activity, with a turnover number (TON) of 9800 over 30 h at
258C.[11] In contrast, the corresponding homogeneous con-
ditions, consisting of a mixture of [{RhCl(C2H4)2}2] and Ph-
SMAP (5) (Rh/P 1:1) in CH2Cl2 (ketone/Rh 100:1), afforded a
TON of 51 under otherwise identical condi-
Entry
Catalyst
t [h] Yield [%] TON[b] Total TON
1
2
3
4
[Au]-SMAP-Rh (2)[c]
2nd run
3rd run
16
16
16
16
16
30
16
80
81
78
73
36
37
9
60000
61000
59000
55000
15000
16000
3900
121000
180000
235000
4th run
5[d]
6[d]
7[d]
[{RhCl(C2H4)2}2]
[{RhCl(C2H4)2}2]
5/[{RhCl(C2H4)2}2]
(P/Rh 1:1)
8[d]
5/[{RhCl(C2H4)2}2]
(P/Rh 1:1)
30
9
3900
9[e]
[Au]-Ph2P-Rh[c]
2nd run
16
16
45
7
30000
4700
10[e]
34700
[a] The reaction was carried out with Me2PhSiH (12 mmol), EtOH
(1.2 equiv), and catalyst (Me2PhSiH/Rh 75000:1) in hexane (0.12 mL) at
258C. [b] Catalyst turnover number. [c] A gold surface with dimensions of
5”5 mm2 was used. [d] The reaction was carried out in CH2Cl2.
Me2PhSiH/Rh 43000:1. [e] Me2PhSiH/Rh 67000:1.
To obtain further information on the Rh species after the
catalytic reaction we analyzed 2 by XPS. The binding energy
of Rh 3d was 308.0 eV, which is much higher than that of
metallic Rh (typically, at 306.0–307.5 eV). The binding energy
of P 2p after the reaction was similar to that before the
reaction and much higher than free, noncoordinating phos-
phane (at 130.1 eV). The P/Rh elemental ratio was calculated
to be 1:0.8. On the basis of these XPS results, it is unlikely that
the major component of the Rh species consists of nano-
tions. Moreover, 2 showed good reusability
(TON of 25000 over three successive runs).
However, its applicability to other ketones was
very limited. This result was in sharp contrast
with that of the related silica-supported SMAP
ligand [silica]-SMAP,[12] which was applicable
particles.[9,15] It is probable that the P Rh bond is maintained
in [Au]-SMAP-Rh (2) during the catalysis.
À
to a broad range of sterically hindered ketones. The narrow
substrate scope of the hydrosilylation catalyzed by 2 suggests
a severely crowded catalytic environment.
The use of [{RhCl(C2H4)2}2] in the solution phase (a
homogeneous control) showed a TON of 15000, which was
only a quarter of that obtained with 2 (Table 1, entry 5). The
addition of SMAP 5 to the Rh complex (P/Rh 1:1) caused a
further decrease in the activity (Table 1, entry 7). In addition
to the moderate effect of the immobilization on activity
enhancement, a more pronounced effect was observed in
terms of the catalyst lifetime and reusability. In fact, the
homogeneous catalyst systems completely lost their activities
within 16 h, and afforded no further conversion after pro-
longed reaction times (Table 1, entries 6 and 8).
The same reaction was also catalyzed by [Au]-Ph2P-Rh,
which resulted in a yield of 45% and a TON of 30000 for the
first use. Interestingly, the yield for the second use of [Au]-
Ph2P-Rh dropped to 7%, which indicates that the SMAP
structure was critically important for the robustness and
reusability of the catalytically active surfaces. Furthermore,
[Au]-Ph2P-Rh was found to be unstable under the reaction
conditions. XPS analysis of the [Au]-Ph2P-Rh after a catalytic
reaction showed significant decomposition of the catalyst
monolayer. Although the signal intensity of the S atom
remained unchanged, only a trace amount of P atoms was
detected.
Thus, we turned our attention to the dehydrogenative
silylation of alcohols with a hydrosilane,[13,14] with the expect-
ation that this reaction would be less sterically demanding
(Table 1). Accordingly, a single chip of [Au]-SMAP-Rh (2; 5 ”
5 mm2) was placed at the bottom of a glass screw-capped test
tube containing a solution of Me2PhSiH (12 mmol) and EtOH
(1.2 equiv) in hexane (0.12 mL). The substrate/catalyst ratio
(S/C) under these conditions was 75000:1, based on the
surface density of Rh atoms. The reaction was conducted at
258C without stirring. After 16 h, 80% of the Me2PhSiH was
converted into the silyl ether. This conversion corresponds to
a TON of 60000 (Table 1, entry 1).
The platelike shape of the catalyst means that recycling
can be readily achieved by physically transferring the used
catalyst chip into the reaction vessel for the next run. The
TON over four successive runs reached a total of 235000
(Table 1, entries 1–4). The activity of 2 was maintained during
repeated uses, thus indicating that the active species was not
released into the solution phase, but remained on the catalyst
surface without significant loss of activity. ICP-MS analysis of
the solution phase after a catalytic reaction indicated that
only less than 0.5% of the Rh atoms had leached out from the
Au surface.
The range of substrates compatible with the dehydrogen-
ative silylation catalyzed by 2 (5 ” 5 mm2, S/C 75000:1, 258C,
5628
www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 5627 –5630

Angewandte
Chemie
16 h) is shown in Table 2. The catalyst promoted the reaction
of the longer linear aliphatic alcohol 6a with a TON of 59000
(Table 2, entry 1). The silylation of g-branched alcohol 6b and
benzyl alcohol (6c) also proceeded smoothly to afford the
corresponding silyl ethers (Table 2, entries 2 and 3). In the
reaction with b-branched alcohol 6d, however, the reaction
rate was slightly retarded, and gave a TON of 12000 (Table 2,
entry 4).
for the highly selective homogeneous catalyst DTBM-xant-
phos-Cu(O-tBu) (Table 3, entry 8).[14] Three successive runs
were catalyzed without any decrease in the selectivity, and
afforded a total TON of 335000 with greater than 99.5%
selectivity (Table 3, entries 2–4).
In contrast, the homogeneous counterparts using either
[{RhCl(C2H4)2}2] alone or a mixture of [{RhCl(C2H4)2}2] and
Ph-SMAP (5; Rh/P 1:1) catalyzed the competitive reaction
with fairly low selectivities (Table 3, entries 5 and 6). The
same reaction using [Au]-Ph2P-Rh, also resulted in low
selectivity (Table 3, entry 7). The highly selective nature of
[Au]-SMAP-Rh (2) is attributable to the steric congestion of
the catalytic environment, which exists in a densely packed
rhodium–phosphane assembly.
Enlargement of the catalyst chip was expected to allow for
a larger scale reaction. Thus, in the presence of a [Au]-SMAP-
Rh (2) chip of dimensions 25 ” 25 mm2 (3.9 nmol Rh), the
reaction of 39 mg (0.30 mmol) of 1-octanol, 39 mg
(0.30 mmol) of 2-octanol, and 41 mg (0.30 mmol) of
Me2PhSiH gave the primary silyl ether 6a with greater than
99.5% selectivity and in 81% yield after 16 h, which
corresponds to a TON of 62000. For practical purposes,
chemical engineering methodologies, such as microfluidic
systems, are expected to enhance the efficiency of our
catalytic system.[16] Another option for a scale-up technique
would be immobilization of the rhodium–phosphane mono-
layer on gold particles.[3,6]
Table 2: Dehydrogenative silylation of primary alcohols with Me2PhSiH
catalyzed by [Au]-SMAP-Rh (2).[a]
Entry
Alcohol
Yield [%]
TON[b]
1
2
3
4
CH3(CH2)7OH (6a)
(CH3)2CHCH2CH2OH (6b)
PhCH2OH (6c)
79
76
89
16
59000
57000
68000
12000
CH3CH2CH(CH3)CH2OH (6d)
[a] The reaction was carried out with alcohol (12 mmol), Me2PhSiH
(1.2 equiv), and 2 (5”5 mm2; alcohol/Rh 75000:1) in hexane (0.12 mL)
at 258C for 16 h. [b] Catalyst turnover number.
An additional characteristic feature of the silylation
catalyzed by 2 is the extremely high selectivity for primary
over secondary alcohols: no conversion was observed in the
reaction of secondary alcohols such as 2-octanol, cyclohex-
anol, and 1-phenylethanol. The intrinsic selectivity was
further examined under competitive reaction conditions
(Table 3). [Au]-SMAP-Rh (2) catalyzed the conversion of
primary alcohol 6a into silyl ether 7a without the formation
of the silylated product 7e from the secondary alcohol
(Table 3, entry 1). This selectivity is higher than that observed
In summary, the phosphane-functionalized gold surface
[Au]-SMAP (1), which consists of a caged, compact trialkyl-
phosphane (SMAP) with an alkanethiolate pendant group,
was developed. The rhodium-complexed surface [Au]-
SMAP-Rh (2) showed high catalytic activity and reusability,
as well as a unique selectivity in the dehydrogenative
silylation of alkanols. The application of this system to other
transition-metal-catalyzed reactions is under investigation.
Table 3: Catalytic dehydrogenative silylation of primary alcohols in the
presence of secondary alcohols.[a]
Received: February 23, 2008
Revised: April 15, 2008
Published online: June 23, 2008
Keywords: heterogeneous catalysis · monolayers ·
phosphane ligands · rhodium · surface chemistry
.
Entry Catalyst
Yield [%] TON[b] Total TON 7a/7e
1
[Au]-SMAP-Rh (2)[c]
60
58
55
50
7
90000
87000
83000
75000
6000
>99.5:0.5
177000 >99.5:0.5
260000 >99.5:0.5
335000 >99.5:0.5
71:29
2
2nd run
3
3rd run
[1] For a review, see: a) A. Ulman, Chem. Rev. 1996, 96, 1533; b) A.
Ulman, Thin films: Self-assembled Monolayers of Thiols,
Academic Press, San Diego, 1998; c) S. Flink, F. C. J. M.
van Veggel, D. N. Reinhoudt, Adv. Mater. 2000, 12, 1315;
d) J. J. Gooding, F. Mearns, W. Yang, J. Liu, Electroanalysis
2003, 15, 81; e) J. C. Love, L. A. Estroff, J. K. Kriebel, R. G.
Nuzzo, G. M. Whitesides, Chem. Rev. 2005, 105, 1103.
4
4th run
5[d]
6[d]
[{RhCl(C2H4)2}2]
5/[{RhCl(C2H4)2}2]
(P/Rh 1:1)
3
2600
72:28
7
[Au]-Ph2P-Rh[c,e]
DTBM-xantphos-Cu
43
92
58000
46
81:19
98:2
8[f]
[a] The reaction was carried out with Me2PhSiH (24 mmol), 6a
(1.0 equiv), 6e (1.0 equiv), and catalyst in hexane (0.24 mL) at 258C
for 16 h. Me2PhSiH/catalyst 150000:1. [b] Catalyst turnover number.
[c] A gold surface with dimensions of 5”5 mm2 was used. [d] The
reaction was carried out in CH2Cl2. Me2PhSiH/catalyst 86000:1.
[e] Me2PhSiH/catalyst 130000:1. [f] The reaction was carried out with a
copper complex of a xanthene-based diphosphane. Reaction conditions:
1-decanol, 2-decanol, Me2PhSiH, Me2PhSiH/Cu 50:1, in toluene, 228C,
7 h; see reference [14]. DTBM-xantphos=4,5-bis[di(3,5-di-tert-butyl-4-
methoxyphenyl)phosphino]-9,9-dimethylxanthene.
[2] For selected studies on functional surfaces, see: a) K. Uosaki, Y.
Sato, H. Kita, Langmuir 1991, 7, 1510; b) L. Häussling, B.
Michel, H. Ringsdorf, H. Rohrer, Angew. Chem. 1991, 103, 568;
Angew. Chem. Int. Ed. Engl. 1991, 30, 569; c) G. P. López, M. W.
Albers, S. L. Schreiber, R. Carroll, E. Peralta, G. M. Whitesides,
J. Am. Chem. Soc. 1993, 115, 5877; d) K. Uosaki, T. Kondo, X.-Q.
Zhang, M. Yanagida, J. Am. Chem. Soc. 1997, 119, 8367; e) M.
Abe, T. Michi, A. Sato, T. Kondo, W. Zhou, S. Ye, K. Uosaki, Y.
Sasaki, Angew. Chem. 2003, 115, 3018; Angew. Chem. Int. Ed.
2003, 42, 2912.
Angew. Chem. Int. Ed. 2008, 47, 5627 –5630
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 5629

Communications
[3] a) M. Bartz, J. Kꢀther, R. Seshadri, W. Tremel, Angew. Chem.
the rigid structure of the phosphane moiety, it is unlikely that the
formation of the Rh(P)2- or Rh(P)3-type complexes is significant.
In addition, numerous recent studies revealed that Rh(P)-type
species exhibited high catalytic activity.[12] For details of the XPS
analysis of 2, see the Supporting Information.
1998, 110, 2646; Angew. Chem. Int. Ed. 1998, 37, 2466; b) H. Li,
Y.-Y. Luk, M. Mrksich, Langmuir 1999, 15, 4957; c) K. Mar-
ubayashi, S. Takizawa, T. Kawakusu, T. Arai, H. Sasai, Org. Lett.
2003, 5, 4409; d) B. S. Lee, S. K. Namgoong, S.-G. Lee, Tetrahe-
dron Lett. 2005, 46, 4501; e) F. Ono, S. Kanemasa, J. Tanaka,
Tetrahedron Lett. 2005, 46, 7623.
[10] a) C. A. Widrig, C. Chung, M. D. Porter, J. Electroanal. Chem.
1991, 310, 335; b) T. Sumi, H. Wano, K. Uosaki, J. Electroanal.
Chem. 2003, 550, 321; c) H. Wano, K. Uosaki, Langmuir 2001,
17, 8224, and references therein.
[11] Optimization of the solvent showed that hexane was the best
among those examined.
[4] For studies on the immobilization of Rh on gold surfaces through
N-heterocyclic carbene moieties, see: a) K. Hara, K. Iwahashi, Y.
Kanamori, S. Naito, S. Takakusagi, K. Uosaki, M. Sawamura,
Chem. Lett. 2006, 35, 870; b) K. Hara, K. Iwahashi, S. Takaku-
sagi, K. Uosaki, M. Sawamura, Surf. Sci. 2007, 601, 5127.
[5] H. Gao, R. J. Angelici, Can. J. Chem. 2001, 79, 578.
[6] T. Belser, M. Stꢁhr, A. Pfaltz, J. Am. Chem. Soc. 2005, 127, 8720.
[7] SMAP = silicon-constrained monodentate alkylphosphane:
a) A. Ochida, K. Hara, H. Ito, M. Sawamura, Org. Lett. 2003,
5, 2671; b) A. Ochida, S. Ito, T. Miyahara, H. Ito, M. Sawamura,
Chem. Lett. 2006, 35, 294.
[8] For a related study, which deals with a palladium–bisoxazoline
monolayer catalyst anchored on a Si(111) surface, see: K. Hara,
S. Tayama, H. Kano, T. Masuda, S. Takakusagi, T. Kondo, K.
Uosaki, M. Sawamura, Chem. Commun. 2007, 4280.
[9] The presence or absence of Rh(P)2- or Rh(P)3-type complexes
cannot be concluded solely from the obtained XP spectra.
However, considering the elemental ratio (Rh/P) determined by
XPS, the limited mobility in such a high density monolayer, and
[12] G. Hamasaka, A. Ochida, K. Hara, M. Sawamura, Angew. Chem.
2007, 119, 5477; Angew. Chem. Int. Ed. 2007, 46, 5381.
[13] For selected studies on transition-metal catalysts, see: a) I.
Ojima, T. Kogure, M. Nihonyanagi, H. Kono, S. Inaba, Chem.
Lett. 1973, 501; b) M. P. Doyle, K. G. High, V. Bagheri, R. J.
Pieters, P. J. Lewis, M. M. Pearson, J. Org. Chem. 1990, 55, 6082;
c) X.-L. Luo, R. H. Crabtree, J. Am. Chem. Soc. 1989, 111, 2527.
[14] H. Ito, A. Watanabe, M. Sawamura, Org. Lett. 2005, 7, 1869.
[15] We assume the ethylene ligands are very labile and easily
released or hydrosilylated at the first stage of the catalytic
reaction. We also conducted AFM studies of the catalyst surface
at the end of the first catalytic reaction. The flatness of the AFM
image was the same as that of the bare gold surface. We could
not find any particles under a noise level of 0.3 nm.
[16] J. Kobayashi, Y. Mori, S. Kobayashi, Chem. Asian J. 2006, 1, 22.
5630
www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 5627 –5630