Immobilization of rhodium complexes at thiolate monolayers on gold surfaces: Catalytic and structural studies

Article abstract of DOI:10.1021/ja0500714

Chiral rhodium-diphosphine complexes have been incorporated into self-assembled thiolate monolayers (SAMs) on gold colloids. Catalysts of this type are of interest because they combine properties of homogeneous and heterogeneous systems. In addition, it s

Full text of DOI:10.1021/ja0500714

Published on Web 05/25/2005
Immobilization of Rhodium Complexes at Thiolate Monolayers
on Gold Surfaces: Catalytic and Structural Studies
Thomas Belser,† Meike Sto¨hr,*,‡ and Andreas Pfaltz*,†
Contribution from the Department of Chemistry, St. Johanns-Ring 19, and Department of
Physics, Klingelbergstrasse 82, UniVersity of Basel, 4056 Basel, Switzerland
Received January 6, 2005; E-mail: Meike.Stoehr@unibas.ch; Andreas.Pfaltz@unibas.ch
Abstract: Chiral rhodium-diphosphine complexes have been incorporated into self-assembled thiolate
monolayers (SAMs) on gold colloids. Catalysts of this type are of interest because they combine properties
of homogeneous and heterogeneous systems. In addition, it should be possible to influence the catalytic
properties of the metal center by the neighboring thiolate molecules. Colloids with a diameter of ca. 3 nm,
coated with a mixed monolayer of n-octanethiolates and thiolates with chiral rhodium-PYRPHOS end groups,
were studied as hydrogenation catalysts. With methyl R-acetamido-cinnamate as substrate, virtually the
same enantioselectivities (up to 93% ee) and full conversion were obtained as with the corresponding
homogeneous [Rh(COD)(PYRPHOS)]BArF catalyst. The colloids were easily recovered by filtration and
reused as catalysts three times without loss of enantioselectivity. STM studies of analogous SAMs on
Au(111) gave a detailed picture of the structure and dynamics of mixed monolayers of this type. The STM
images showed that the catalyst-bearing thiolates are distributed statistically on the surface and that the
ordered structure of the n-octanethiolate SAM can be retained during incorporation of the catalyst-bearing
thiols using the place-exchange methodology.
Introduction
complexes to self-assembled monolayers (SAMs) of alkanethi-
olates on gold colloids. The catalytic properties of these
Both homogeneous and heterogeneous catalysts have their
specific advantages and disadvantages. Heterogeneous catalysts
can be easily recovered and recycled and are therefore ideal
for industrial processes. However, our understanding of how
such catalysts work is still insufficient, making the development
of new catalysts very challenging. Moreover, no general methods
are available for the preparation of defined metal surfaces with
particular catalytic properties. In contrast to heterogeneous
catalysis, the mechanism of homogeneous catalytic reactions
is often known in detail. By means of a proper ligand, the
reactivity and selectivity of a catalytic metal center can be tuned.
However, separation from the reaction mixture and recycling
is generally difficult.
To overcome this problem, homogeneous catalysts have been
immobilized on solid supports such as organic polymers or
inorganic solids.1 Ideally, the support allows easy separation
and recycling of the catalyst, but does not affect the catalytic
process. Interesting extensions of this concept were reported
by Tremel,2 Mrksich,3 and Sasai,4 who attached transition-metal
functionalized colloids were similar to those of the correspond-
ing homogeneous catalysts.
SAMs of this type offer unique opportunities in catalysis,
which have not been fully exploited yet. As illustrated in Figure
1, they can combine the properties of heterogeneous and
homogeneous catalysts as well as enzymes. In contrast to
conventional supported catalysts, the metal centers (M) can be
embedded in an ordered monolayer of thiolate spacers (A),
which can interact with the substrate (S) and in this way
influence the reactivity and selectivity of a catalytic process.5
Depending on the chain length of the neighboring thiolates, two
different classes of catalysts seem possible.
In one case a convex catalytic site is formed, resembling a
homogeneous metal catalyst. Using appropriately functionalized
neighboring thiolates, it should be possible to create specific
binding sites for the substrate (e.g., hydrogen-bond donors or
ionic groups). In addition, by separating the metal centers, the
formation of aggregates between two or more metal complexes
should be prevented. For reactions suffering from catalyst
deactivation by aggregation, this should have a beneficial effect.
If the metal centers are located inside the monolayer formed
by the neighboring thiolates, the active site has a concave
structure similar to an enzyme. In this way catalysts can be
constructed that should behave like hybrids of homogeneous
and enzymatic catalysts. When the catalytic site (M) is sur-
† Department of Chemistry.
‡ Department of Physics.
(1) (a) Pugin, B.; Blaser, H.-U. In ComprehensiVe Asymmetric Catalysis;
Jacobsen, E. N.; Pfaltz, A.; Yamamoto, H., Eds.; Springer: Berlin, 1999;
Vol. III, pp 1367-1375. (b) Leadbeater, N. E.; Marco, M. Chem. ReV.
2002, 102, 3217. (c) Fan, Q.-H.; Li, Y.-M.; Chan, A. S. C. Chem. ReV.
2002, 102, 3385. (d) Song, C. E.; Lee, S. Chem. ReV. 2002, 102, 3495. (e)
McMorn, P.; Hutchins, G. H. Chem. Soc. ReV. 2004, 33, 108.
(2) Bartz, M.; Ku¨ther, J.; Seshadri, R.; Tremel, W. Angew. Chem. 1998, 110,
2646.
(3) Li, H.; Luk, Y.-Y.; Mrksich, M. Langmuir 1999, 15, 4957.
(4) Marubayashi, K.; Takizawa, S.; Kawakusu, T.; Arai, T.; Sasai, H. Org.
Lett. 2003, 23, 4409.
(5) For a similar approach based on bifunctionalized mesoporous silica, see:
Huh, S.; Chen, H.-T.; Wiench, J. W.; Pruski, M.; V. Lin, S.-Y. J. Am.
Chem. Soc. 2004, 126, 1010.
9
8720
J. AM. CHEM. SOC. 2005, 127, 8720-8731
10.1021/ja0500714 CCC: $30.25 © 2005 American Chemical Society

Rh Complexes at Thiolate Monolayers on Au Surfaces
A R T I C L E S
Figure 1. Comparison of homogeneous, heterogeneous, and enzymatic catalysts (left) and schematic representations of metal catalysts attached to thiolate
monolayers on gold surfaces (right).
rounded by apolar alkyl chains, a selective reaction of nonpolar
substrates (S) in the presence of polar substrates (SP) should be
possible. By modification of the neighboring thiolates, the
environment of the metal center can be readily changed and, if
desired, specific binding sites for the substrate can be introduced.
To explore these possibilities, we decided to prepare mixed
SAMs on gold surfaces containing alkanethiolates with chiral
rhodium catalysts as terminal groups.6 As a test reaction we
chose the rhodium-catalyzed asymmetric hydrogenation of
enamides, which is mechanistically well understood and,
therefore, ideal for this purpose.7-9
monolayer of alkanethiolate chains bearing chiral rhodium
complexes as end groups (Scheme 1). The chiral diphosphine
ligand PYRPHOS was chosen,10 since it can be easily attached
to a linker through an amide bond, as shown in structure 1. To
functionalize the gold surface, we treated the gold particles in
dichloromethane at 23 °C with disulfide 1, which was readily
prepared from 11-(10′-carboxydecyldisulfanyl)undecanoic acid
via the acid chloride (see Supporting Information). The diphos-
phine-covered particles 2 were then treated with [Rh(COD)-
11,12
Cl]2 (COD ) 1,5-cyclooctadiene) and TlBArF
(BArF )
tetrakis[3,5-bis(trifluoromethyl)phenyl]borate) to give the func-
tionalized particles 3.
We herein report the preparation and catalytic studies of
mixed monolayers of alkanethiolates and thiolates functionalized
with rhodium-diphosphine complexes. We also present the
results of STM analyses of SAMs on (111)-Au surfaces, which
reveal a detailed picture of the structure and dynamics of mixed
monolayers of this type.
The surface coverage with thiolate molecules, typically
obtained under these conditions, was determined by means of
disulfide 4 bearing p-nitrophenyl ester end groups. After
immobilization, the p-nitrophenyl ester groups were hydrolyzed
by treatment under basic conditions (dioxane/1 M NaOH(aq)
solution 1:1 (v/v)) and the concentration of p-nitrophenolate was
measured by UV/vis spectroscopy. After an incubation time of
24 h using a disulfide concentration of 0.05 M in dichlo-
romethane at 23 °C, the thiolate content of the gold particles
was found to be 1.87 × 10-6 mol p-nitrophenyl ester per gram
gold particles.
Results and Discussion
Preliminary Investigations with Gold Particles. Initial
experiments were carried out using gold particles with an
average diameter of 0.3-3 µm, which were covered with a
(6) Incorporation of a Rh catalyst into LB films: To¨llner, K.; Popovitz-Biro,
R.; Lahav, M.; Milstein, D. Science 1997, 278, 2100.
(7) Landis, C. R.; Halpern, J. J. Am. Chem. Soc. 1987, 109, 1746.
(8) (a) Brown, J. M. In ComprehensiVe Asymmetric Catalysis; Jacobsen, E.
N.; Pfaltz, A.; Yamamoto, H., Eds.; Springer: Berlin, 1999; Vol. I, pp
121-182. Gridnev, I. D.; Imamoto, T. Acc. Chem. Res. 2004, 37, 633. (b)
Gridnev, I. D.; Imamoto, T. Acc. Chem. Res. 2004, 37, 633.
(9) (a) Landis, C. R.; Hilfenhaus, P.; Feldgus, S. J. Am. Chem. Soc. 1999,
121, 8741. (b) Feldgus, S.; Landis, C. R. J. Am. Chem. Soc. 2000, 122,
12714.
The functionalized gold particles 3 containing rhodium-
PYRPHOS end groups were tested as catalysts in the hydro-
9
J. AM. CHEM. SOC. VOL. 127, NO. 24, 2005 8721

A R T I C L E S
Belser et al.
Scheme 1. Synthesis of Functionalized Gold Particles 3a
a (a) Gold particles, dichloromethane, 23 °C, 24 h. (b) [Rh(COD)Cl]2, TlBArF,11,12 dichloromethane.
Scheme 2. Asymmetric Hydrogenation of Methyl R-Acetamidocinnamate
genation of methyl R-acetamidocinnamate (Scheme 2). The
reactions were carried out in dichloromethane under 60 bar of
hydrogen pressure. For 1 mmol of olefin, 11 g of gold particles
3 were used, which was equivalent to approximately 1-2 mol
% of rhodium catalyst. For comparison, the analogous homo-
(10) (a) Nagel, U. Angew. Chem. 1984, 96, 425. (b) Nagel, U.; Kinzel, E.;
Andrade, J.; Prescher, G. Chem. Ber. 1986, 119, 3326.
(11) Brookhart, M.; Grant, B.; Volpe, A. F. Organometallics 1992, 11, 3920.
(12) Hughes, R. P.; Lindner, D. C.; Rheingold, A. L.; Yap, G. P. A. Inorg.
Chem. 1997, 36, 1726.
9
8722 J. AM. CHEM. SOC. VOL. 127, NO. 24, 2005

Rh Complexes at Thiolate Monolayers on Au Surfaces
A R T I C L E S
Scheme 3. Apolar and Polar Alkanethiolate-Protected Gold Colloids
geneous reaction was carried out under the same conditions,
using 2 mol % of the rhodium complex derived from N-
octanoyl-PYRPHOS as catalyst.
darkening of the toluene phase. After stirring for 12 h, the
toluene layer was separated, concentrated, and diluted with
ethanol. The thiolate-protected colloids were isolated by pre-
cipitation at -60 °C. In addition to 6-(adamant-1-yl)hexanethiol,
n-octanethiol, and n-dodecanethiol, 11-aminoundecane-1-thiol
or 11-mercaptoundecan-1-ol was used as polar thiolate to protect
the colloids. The apolar colloids 6-8 were purified by precipita-
tion from toluene with ethanol, whereas the polar colloids 9
and 10, which precipitated from the reaction mixture, were
isolated by filtration and washed several times with diethyl ether
to remove any residual thiols (Scheme 3).
The immobilized rhodium-PYRPHOS catalyst proved to be
active and showed virtually the same enantioselectivity as the
analogous homogeneous catalyst. The gold particles could be
recycled, giving the same ee values with only slightly reduced
yields in the second run. In the third run, however, low yields
were obtained due to deactivation of the catalyst. The reaction
solution in these experiments was colorless and did not contain
any phosphorus compounds according to 31P NMR analysis,
ruling out catalyst leaching under the reaction conditions.
Deactivation is a common phenomenon observed upon recycling
of immobilized catalysts, which may be caused, for example,
by undesired reactions of the highly reactive Rh centers with
oxygen or other impurities during catalyst recovery.1
Our results demonstrate that rhodium-diphosphine hydrogena-
tion catalysts can be immobilized on thiolate monolayers
attached to gold surfaces and that the immobilized complexes
retain their catalytic activity and enantioselectivity. However,
for full conversion, a relatively large quantity of gold particles
(11 g/mmol of substrate) has to be used. Experiments with mixed
monolayers consisting of functionalized thiolates surrounded by
an excess of nonfunctionalized spacers would require even larger
amounts of gold, making such studies impractical. Therefore,
we decided to investigate immobilization on gold colloids, which
are 2-3 orders of magnitude smaller than the functionalized
gold particles 3 (2-10 nm vs 300-3000 nm) with far higher
surface area per unit mass.
The colloids were characterized by elemental analysis and
TEM, which visualized the gold core. From the results, the
molecular weight and thiolate surface coverage were estimated
on the basis of a structural model of the gold core shape. The
spherical model employed approximates the gold core as a
sphere, covered with a skin of hexagonally close-packed gold
atoms.14-17 According to this model, colloid 7 with an average
diameter of 3.48 ( 0.61 nm contains 1298 gold atoms and 211
adsorbed octanethiolates on the surface. Gold colloids (8-10)
with sterically nondemanding end groups gave similar results,
whereas the gold colloid 6 with an average diameter of 2.32 (
0.46 nm, bearing bulky adamantyl end groups, contained only
385 gold atoms and 68 adsorbed thiolates. Thermogravimetric
analysis and XPS measurements showed that the S-Au bonds
are stable up to 150 °C. The colloids were also stable under the
conditions of the hydrogenation reactions described below (4
h, 23 °C at 60 bar of H2 with mechanical stirring). Leaching of
adsorbed thiolates after 24 h at 100 bar of H2 was less than
0.2% (GC analysis of the reaction solution). Thus, these
protected gold colloids served as convenient precursors for the
incorporation of rhodium-diphosphine complexes into the thi-
olate monolayer.
Functionalized Gold Colloids. A series of n-alkanethiolate-
protected gold colloids with diameters between 2.3 and 3.5 (
0.6 nm were prepared using a modified procedure of Murray
and co-workers.13 An aqueous solution of hydrogen tetrachlo-
roaurate was extracted into toluene and metathesized to the
tetraoctylammonium salt in the presence of a toluene solution
of tetraoctylammonium bromide. To this organic solution was
added n-alkanethiol, followed by an aqueous solution of sodium
borohydride. Reduction of the gold salt and concomitant
derivatization by n-alkanethiol was evidenced by prompt
Various methods have been described for introducing func-
tional components into SAMs of thiolates on gold surfaces.18
(14) D’Agostino, G.; Pinto, A.; Mobilio, S. Phys. ReV. B 1993, 48, 14447.
(15) Sellers, H.; Ulman, A.; Shnidman, Y.; Eilers, J. E. J. Am. Chem. Soc. 1993,
115, 9389.
(16) Whetten, R. L.; Gelbart, W. M. J. Phys. Chem. 1994, 98, 3544.
(17) Leff, D. V.; Ohara, C.; Heath, J. R.; Gelbart, W. M. J. Phys. Chem. 1995,
99, 7036.
(13) Hostetler, M. J., et al. Langmuir 1998, 14, 17.
9
J. AM. CHEM. SOC. VOL. 127, NO. 24, 2005 8723

A R T I C L E S
Belser et al.
In our case, we chose the place-exchange reaction19,20 for the
preparation of mixed monolayers containing the chiral rhodium
catalysts, because of its simplicity and versatility. Moreover,
this method allows functionalization of colloids without altering
their dimensions.
catalyst. The polar colloids 16 and 17 with amino and hydroxy
end groups proved to be less reactive and less enantioselective
(Table 1). Although no satisfactory explanation for the observed
negative effects of the amino and hydroxy groups can be given
at this stage, the results show that the spacer end groups can
strongly influence the performance of the catalyst.
After separation of the colloids, the reaction solutions were
analyzed by 31P NMR spctroscopy. The lack of any 31P signals
implied that no catalyst leaching was occurring. This was also
confirmed by elemental analysis (Rh, P, S). The particle size
of the recovered colloids was checked by TEM and found to
be the same as before the hydrogenation reaction.
Furthermore, recycling of the functionalized gold colloids 13
was investigated. After the reaction, the colloids were treated
immediately with 10 equiv of COD to restore the original
rhodium complex. The colloids could be readily recovered by
filtration using centrifugal filter units with a cutoff molecular
size of 50 kD and were successfully used again as catalysts.
The enantioselectivities remained constant during three recycling
steps, while conversion decreased moderately from 99% to 74%
in the fourth cycle (standard reaction time: 4 h). However, if
the reaction time was increased to 12 h, almost full conversion
could be obtained even in the third cycle (96% vs 86% after 4
h).
STM Investigations. To obtain detailed structural information
on the arrangement of the ionic catalysts 12 embedded in an
n-octanethiolate matrix and to get a better understanding of the
exchange process itself, STM measurements were conducted.
An important question was whether the catalyst molecules were
clustered or distributed statistically on the gold surface. For the
STM investigations (111)-oriented gold films were chosen, as
it has been found by different analytical techniques such as XRD
(X-ray diffraction), XPS (X-ray photoelectron spectroscopy),
IR spectroscopy, and solid-state NMR spectroscopy2,21 that the
properties of planar surfaces and gold colloids covered with a
SAM are very similar (for an exception see ref 22). Especially,
the Au-S bonding and chain ordering are essentially the same
for both systems. Nevertheless, there are also differences that
should be kept in mind: the curvature of the colloid surface
and the higher concentration of defectlike sites on the colloids,
which explains the higher density of alkanethiolates on gold
colloids compared to flat surfaces. However, mechanistic studies
of place-exchange reactions at colloid and Au(111) surfaces have
shown that, although the reaction rates may differ, the basic
mechanism is the same for the two systems.23 Thus we assume
that STM results obtained on flat surfaces after the exchange
reaction are also relevant for the corresponding gold colloids.
For our studies, an upright standing arrangement of the
thiolates was required, to obtain comparable initial conditions
Initial experiments were carried out with the p-nitrophenyl-
ester-bearing thiol 5, to determine the extent of thiolate exchange
in relation to the reaction time under specific conditions. After
incubation of the n-octanethiolate-protected colloids 7 in a
solution of 5 in dichloromethane, the resulting functionalized
colloids were separated by filtration with a solvent-resistant
stirred cell and isolated as a black solid after precipitaion from
ethanol/hexanes at -60 °C. The ratio between functionalized
thiolates and n-octanethiolates in the SAMs could be deduced
1
from H NMR spectra of the colloids. The concentration of
nitrophenyl esters per gram of gold was determined by UV/vis
spectroscopy after basic hydrolysis. After incubation in a 1.5
mM solution of thiol in dichloromethane for 24 h at 23 °C, a
ratio of 4.4:1 between n-octanethiolate and the p-nitrophenyl-
oxycarbonyl-substituted thiolate and a concentration of 0.15
mmol of nitrophenylester per gram of gold colloids was found.
Longer incubation times or higher concentration of thiol 5
resulted in lower n-octanethiolate/functionalized thiolate ratios
(an incubation time of 48 h at 23 °C resulted in a 2:1 ratio
between n-octanethiolate and the p-nitrophenyloxycarbonyl-
substituted thiolate). Virtually identical results were obtained
in dichloromethane or ethanol.
Exchange with the rhodium-PYRPHOS-bearing thiol 12 was
carried out in the same manner. The resulting functionalized
gold colloids 13-17 were purified by precipitation from ethanol
with hexanes and analyzed by TEM and 1H NMR spectroscopy.
The TEM studies revealed that the size of the gold core did not
significantly change during the place-exchange reaction. The
content of functionalized thiolates 12 in the mixed SAMs,
determined by 1H NMR spectroscopy, was strongly dependent
on the nature of the thiolates in the precursor colloids 6-10
and the incubation time. An incubation time of 48 h at 23 °C
led to a 1:1-2:1 ratio between rhodium-PYRPHOS-containing
thiolates and the surrounding spacer thiolates. For our catalytic
studies, we wanted to use SAMs consisting of rhodium
complexes separated by excess spacer thiolates. Therefore, we
reduced the incubation time to 24 h, resulting in a 1:3-1:8
rhodium-PYRPHOS/spacer ratio (Scheme 4).
The functionalized gold colloids 13-17 were tested as
catalysts in the hydrogenation of methyl R-acetamidocinnamate.
The apolar colloids 13-15 with n-octanethiolate, 6-(adamant-
1-yl)hexanethiolate, and n-dodecanethiolate as spacers were all
found to be highly active catalysts, giving high conversion and
good enantioselectivities similar to the corresponding homoge-
neous reaction using [Rh(COD)(octanoyl-PYRPHOS)]BArF as
(21) (a) Whetten, R. L.; Khoury, J. T.; Alvarez, M. M.; Murthy, S.; Vezmar, I.;
Wang, Z. L.; Stephens, P. W.; Cleveland, C. L.; Luedtke, W. D.; Landman,
U. AdV. Mater. 1996, 8, 428. (b) Hostetler, M. J.; Stokes, J. J.; Murray, R.
W. Langmuir 1996, 12, 3604. (c) Badia, A.; Cuccia, L.; Demers, L.; Morin,
F.; Lennox, R. B. J. Am. Chem. Soc. 1997, 119, 2682. (d) Bourg, M.-C.;
Badia, A.; Lennox, R. B. J. Phys. Chem. B 2000, 104, 6562. (e) Li, X.-M.;
Huskens, J.; Reinhoudt, D. N. J. Mater. Chem. 2004, 14, 2954.
(22) Jackson, A. M.; Myerson, J. W.; Stellacci, F. Nat. Mater. 2004, 3, 330. In
this case, mixtures of alkanethiolates and thiolates with carboxylic acid or
arylamino headgroups were found to form phase-separated ordered domains
on colloids, whereas disordered domains were observed on flat surfaces.
In our case, the functionalized thiolates showed no apparent tendency to
aggregate.
(18) Brust, M.; Fink, J.; Bethell, D.; Schiffrin, D. J.; Kiely, C. J. Chem. Soc.,
Chem. Commun. 1995, 1655.
(19) Hostetler, M. J.; Green, S. J.; Stokes, J. J.; Murray, R. W. J. Am. Chem.
Soc. 1996, 118, 4212.
(23) (a) Hostetler, M. J.; Templeton, A. C.; Murray, R. W. Langmuir 1999, 15,
5, 3782. (b) Donkers, R. L.; Song, Y.; Murray, R. W. Langmuir 2004, 20,
4703. (c) Ionita, P.; Caragheorgheopol, A.; Gilbert, B. C.; Chechik, V.
Langmuir 2004, 20, 11536.
(20) Ingram, R. S.; Hostetler, M. J.; Murray, R. W. J. Am. Chem. Soc. 1997,
119, 9175.
9
8724 J. AM. CHEM. SOC. VOL. 127, NO. 24, 2005

Rh Complexes at Thiolate Monolayers on Au Surfaces
A R T I C L E S
Scheme 4. Synthesis of Functionalized Gold Colloids 13-17a
a
(a) [Rh(COD)Cl]2, TlBArF,11,12 CH2Cl2, 52%. (b) Gold colloids (2 mg/mL), 1.5 mM solution of rhodium complex 12, CH2Cl2, 24 h, 23 °C.
and reproducible experiments and to ensure that an exchange
reaction was taking place rather than an adsorption process,
which could occur if the thiolate molecules were lying parallel
to the surface. Therefore, the different experimental parameters
required for obtaining the desired upright SAM were investi-
gated in more detail using n-octanethiol.
Comparing different solvents (see Experimental Section)
while the immersion time, temperature, and thiol concentration
were kept constant, we found that ethanol and dichloromethane
gave the best results. The ordering of the SAMs was reached
faster in either ethanol or dichloromethane in comparison to
acetone or diethyl ether. Furthermore, the concentration, immer-
sion time, and temperature were varied for ethanol and dichlo-
romethane. If the concentration was in the mM range, an
immersion time of at least 24 h was needed to form an upright
SAM. By increasing the concentration to 0.1 M, the immersion
time could be reduced to 8 h. Finally, the solution temperature
was increased from 23 to 55 °C, resulting in larger domains
and fewer etch pits.24 On the basis of these results, a thiol con-
centration of 0.1 M in ethanol or dichloromethane and an immer-
sion time of 8 h at 55 °C were chosen as standard conditions
for subsequent experiments. Figure 2 shows a typical STM
image for the upright standing hexagonal (x3×x3)R30 thiolate
arrangement on Au(111) obtained under these conditions.25,26
When the sample was kept under ambient conditions for
several days, a different arrangement of n-octanethiolate on Au-
(111) was observed (Figure 3). To date, a structure of this type
was reported only once and was described as the thermody-
(24) Delamarche, E.; Michel, B.; Gerber, C.; Anselmetti, D.; Gu¨ntherodt, H.-
J.; Wolf, H.; Ringsdorf, H. Langmuir 1994, 10, 2869.
(25) Widrig, C. A.; Alves, C. A.; Porter, M. D. J. Am. Chem. Soc. 1991, 10,
2805.
(26) Poirier, G. E.; Tarlov, M. J. Langmuir 1994, 10, 2853.
9
J. AM. CHEM. SOC. VOL. 127, NO. 24, 2005 8725

A R T I C L E S
Belser et al.
Table 1. Hydrogenation of Methyl R-Acetamidocinnamate with
Functionalized Gold Colloidsa
colloida
solvent
spacer/[Rh]b
conversion (%)c
ee (%)d
13 (30 mg)
13 (30 mg)
14 (50 mg)
14 (50 mg)
15 (60 mg)
15 (60 mg)
16 (60 mg)
16 (60 mg)
17 (60 mg)
17 (60 mg)
CH2Cl2
EtOH
CH2Cl2
EtOH
CH2Cl2
EtOH
CH2Cl2
EtOH
1:3
1:3
1:6
1:6
1:8
1:8
1:5
1:5
1:6
1:6
>99 [>99]
84 [61]
>99 [>99]
91 [61]
>99 [>99]
81 [61]
94 [>99]
20 [61]
93 [93]
87 [88]
93 [93]
88 [88]
92 [93]
87 [88]
86 [93]
82 [88]
82 [93]
68 [88]
CH2Cl2
EtOH
32 [>99]
12 [61]
a Conditions: 110 mg (0.50 mmol) of substrate in 1 mL of solvent, ca.
1 mol % of [Rh], 23 °C, 4 h, 60 bar of H2, stirring rate 600 rpm.
b Determined by 1H NMR analysis. c Determined by GC (Restek Rtx-1701);
the values in brackets correspond to the homogeneous reaction using
[Rh(COD)(n-octanoyl-PYRPHOS)]BArF. d Determined by HPLC (Daicel
Chiralcel OD-H); the values in brackets correspond to the homogeneous
reaction using [Rh(COD)(n-octanoyl-PYRPHOS)]BArF.
Figure 3. STM image of n-octanethiolates on Au(111) in the p(6×x3)
arrangement (40 nm × 40 nm, -0.95 V, 37 pA). The gold substrate was
immersed for 8 h in 0.1 M n-octanethiol solution in ethanol at 23 °C and
stored under ambient conditions for 5 days before the image was recorded.
18 rather than metal catalyst 12 was chosen in order to
circumvent potential interactions of the tip with the ionic
rhodium complex or the counterion and to gain information on
the exchange reaction with the binary system. It has been shown
that exchange reactions of this type are fast in the initial phase,
but then slow significantly before they gradually reach equi-
librium after 1-2 days.29-31
Initially, conditions similar to those reported in the literature
were used, with immersion times of the n-octanethiolate-coated
sample in the solution of 18 between 24 and 48 h at 23 °C.
Analysis of the resulting mixed SAMs with STM (not shown)
showed a statistical distribution of protrusions, which could be
ascribed to the functionalized thiolates 18. The protrusions were
surrounded by noisy regions, indicating that the ordered upright
standing arrangement of the n-octanethiolate SAM had vanished.
These findings suggested that a significant amount of n-
octanethiolate was exchanged with ligand 18, but in addition, a
certain amount of thiolates were desorbed into the solution
without being replaced by ligands 18, resulting in partial
destruction of the ordered SAM.32 Consequently, n-octanethiol
was added to the solution of 18 to suppress desorption of
thiolates, and the immersion time was reduced to 4 h. Under
these conditions, the upright-standing ordered n-octanethiolate
matrix was conserved, while the exchange reaction could take
place.
Figure 2. STM image of n-octanethiolates on Au(111) in the (x3×x3)-
R30 arrangement (8 nm × 8 nm, -1 V, 16 pA). The gold substrate was
immersed for 8 h in 0.1 M n-octanethiol solution in ethanol at 55 °C.
namically favored arrangement.27 This structural change is
explained by the transformation of the hexagonal (x3×x3)-
R30 arrangement into the (6×x3) arrangement, which involves
both a motion of the sulfur headgroups to different sites on the
Au(111) surface and a change of the tilt angle of the alkyl
chains. In our case, the (6×x3) structure already emerged after
5 days compared to 6 months, as reported by Noh and
co-workers.27 In an earlier publication, a similar transformation
of the (x3×x3)R30 into a (5x3×x3) structure was reported.28
This structure appeared after storing the sample for ∼1 week
in air. These findings show that the hexagonal (x3×x3)R30
arrangement, initially formed, is not stable over longer time
periods and thus does not correspond to the thermodynamically
favored structure.
A characteristic image of a sample prepared using this
protocol is shown in Figure 4. The functionalized thiolates 18
Exchange Reaction with Ligand 18. For the first exchange
experiments with n-octanethiolate SAMs, the metal-free thiolate
(29) Chidsey, C. E. D.; Bertozzi, C. R.; Putvinski, T. M.; Mujsce, A. M. J.
Am. Chem. Soc. 1990, 112, 4301.
(27) Noh, J.; Hara, M. Langmuir 2002, 18, 1953.
(28) Scho¨nenberger, C.; Jorritsma, J.; Sondag-Huethorst, J. A. M.; Fokkink, L.
G. J. J. Phys. Chem. 1995, 99, 3259.
(30) Collard, D. M.; Fox, M. A. Langmuir 1991, 7, 1192.
(31) Chung, C.; Lee, M. J. Electroanal. Chem. 1999, 468, 91.
(32) Schlenoff, J. B.; Li, M.; Ly, H. J. Am. Chem. Soc. 1995, 117, 12528.
9
8726 J. AM. CHEM. SOC. VOL. 127, NO. 24, 2005

Rh Complexes at Thiolate Monolayers on Au Surfaces
A R T I C L E S
thiolate 18 (see above), the ordered SAM was retained under
these conditions. Again, a statistical distribution of the func-
tionalized thiolates was observed. Figure 5 shows a time lapse
series of STM images of a sample prepared in this way. The
six images, which show the same area of the SAM, were taken
at time intervals of 4 min. The bright protrusion represents one
catalyst molecule 12 (annotated in red). It appears that in the
sequences of images (a)-(f) the molecule is moving to the top;
however, what is seen is in fact thermal drifting of the sample.
Relative to the black arrow that indicates the same position in
all six images, thiolate 12 does not move.
The catalyst molecules appear to overtop the n-octanethiolate
SAM by 2-4 Å, whereas for the headgroup a diameter of 1-3
nm is measured. As observed for the analogous metal-free
thiolate 18, the value for the height difference is substantially
smaller than the value expected for a static stretched geometry
of the thiolate chain of 12 (1.2 nm). The values for both the
diameter and the height are more disperse for the rhodium
complex 12 than for thiolate 18. In addition, scanning over 12
often resulted in noisy areas in the STM image. Before
considering possible explanations for these observations, a more
detailed analysis of Figure 5 should be made. The appearance
of rhodium complex 12 changes from image to image with
regard to diameter, height, and noise. In image (e) no protrusion
is observed, although 12 must still be in the same place as
confirmed by the successive image (f). Such a blinking behavior
was also observed for other catalyst molecules 12 of the same
sample and for different samples prepared in the same manner.
A similar phenomenon was reported for phenylene-ethynylene
oligomers embedded in a thiolate matrix by Donhauser and co-
workers.35 They ascribed this blinking to a switching in
conductance caused by density changes of the surrounding
matrix. In our case we interpret the blinking in the following
way.
Figure 4. STM image of ligands 18 embedded in an n-octanethiolate matrix
on Au(111) (33 nm × 33 nm, 1.2 V, 18 pA). The ligands appear as
protrusions, and the height difference between the n-octanethiolates and
the ligand 18 amounts to about 1.5-2.5 Å.
are distributed statistically and appear as bright protrusions with
heights in the range 1.5-2.5 Å above the n-octanethiolate SAM.
These values are considerably shorter than the theoretical value
for the height difference (1.2 nm) if thiolate 18 exists in a
stretched conformation. However, realistically much shorter
values than 1.2 nm are expected, as the alkyl chain and the
large headgroup are flexible and can bend to the side, placing
themselves on top of neighboring n-octanethiolates.
Furthermore, it has to be kept in mind that STM does not
show the real topography of a sample, especially for organic
adsorbates on conducting substrates. By keeping the conductance
constant as performed during constant current tunneling mode,
the STM image displays the local density of states (LDOS) of
the sample. Interpretation of the resulting image as a real
topographical representation is only possible if the LDOS is
uniform over the entire sample surface. Since we are dealing
with two strongly different organic molecules immobilized on
a metal surface, the LDOS above an n-octanethiolate and thiolate
18 are expected to differ substantially. For the headgroups of
18 an average diameter of about 2 nm was measured. It should
be noted that the measured values for the individual molecules
in the SAM varied between 1.2 and 2.6 nm and, moreover, the
diameter for the same molecule was found to change from one
image to the next. The observed deviation of the measured
diameter from the theoretical value (1.2 nm) is not unexpected
because the size of protrusions is always overestimated by
scanning probe microscopy due to the geometric convolution
of the tip shape and the adsorbate shape.33,34 Moreover, tilting
of the headgroups is also likely to contribute to a seemingly
enlarged diameter. Hence, we conclude that each protrusion
represents a single molecule of thiolate 18.
(i) So far, the ionic character of 12 has not been taken into
account. In general, the role of the counterion on the STM image
contrast is unclear because its position on the surface cannot
be determined precisely. We assume that it must reside in close
vicinity to the cationic rhodium complex to compensate its
charge. If the counterion actually contributes to the image
contrast, the observed diameter and/or height of 12 will vary
depending on the position of the counterion. In addition, it is
reasonable to assume that the ion pair is flexible with the
counterion moving around in the vicinity of the rhodium
complex. This movement could be at least partially responsible
for the observed noisy tunneling current.
(ii) As discussed for thiolate 18, the alkyl chain and the
cationic headgroup of 12 are flexible and can tilt. If a repulsive
interaction between tip and headgroup is assumed, the head will
avoid the tip during scanning. As a consequence, this may result
in a noisy tunneling current with no visible protrusion in the
STM image. The interaction between the tip and the headgroup
will depend on the electronic properties of the tip, which are
strongly influenced by the configuration and material of the
foremost atoms. These properties can alter considerably if the
tip is picking up or releasing adsorbates during scans. In
addition, the interaction will depend on the specific arrangement
Exchange Reaction with Rhodium Catalyst 12. For the
exchange reaction of the n-octanethiol-covered gold substrates
with the rhodium(diphosphine)-containing thiolate 12 an im-
mersion time between 1 and 2 h was chosen. No additional
n-octanethiol was added in order to avoid possible reactions
with the rhodium complex. In contrast to the experiments with
(35) Donhauser, Z. J.; Mantooth, B. A.; Kelly, K. F.; Bumm, L. A.; Monnell,
J. D.; Stapleton, J. J.; Price, D. W.; Rawlett, A. M.; Allara, D. L.; Tour, J.
M.; Weiss, P. S. Science 2001, 292, 2303.
(33) Klyachko, D.; Chen, D. M. Surf. Sci. 2000, 446, 98.
(34) G. Mark, I.; L. Biro, P.; Gyulai, J. Phys. ReV. B 1998, 58, 12645.
9
J. AM. CHEM. SOC. VOL. 127, NO. 24, 2005 8727

A R T I C L E S
Belser et al.
Figure 5. Time lapse series of STM images of the same sample area showing one catalyst 12 (encircled in red) embedded in the n-octanethiolate matrix.
The n-octanethiolate-covered Au(111) sample was immersed for 2 h at 23 °C in 0.1 mM catalyst 12 solution in ethanol. All images are 30 nm × 30 nm,
and the tunneling parameters are -0.9 V and 13 pA. The time difference between two images is 4 min. The black arrows mark the same position on the
sample. Because of thermal and piezoelectric drift, the imaged sample area shifts toward the bottom. This shift was corrected between image (d) and (e).
of the anion and the cationic headgroup, which is continuously
changing. In conclusion, we assume that the observed blinking
behavior of the functionalized thiolate 12 originates from
structural fluctuations of the flexible ionic headgroup, which
result in a different interaction with the tip from one scan to
the other. The aforementioned explanation of Donhauser and
co-workers35 seems less likely in our case because it would
predict similar behavior for the structurally related thiolate 18,
which, however, shows no blinking phenomenon.
monolayers on gold colloids by the place-exchange reaction.
Our results show that these functionalized colloids, coated with
a mixed monolayer of n-octanethiolates and thiolates with chiral
rhodium-diphosphine end groups, possess catalytic properties
comparable to analogous homogeneous rhodium catalysts. In
the hydrogenation of methyl R-acetamidocinnamate, the same
enantioselectivities (up to 93% ee) and full conversion could
be obtained as with the corresponding homogeneous [Rh(COD)-
(PYRPHOS)]BArF catalyst. The colloids could be easily
recovered by filtration and reused as catalysts three times
without loss of enantioselectivity. STM studies of analogous
mixed SAMs on Au(111), prepared in the same manner, showed
that the catalyst-bearing thiolates are distributed statistically on
Conclusions
We have developed a method for the incorporation of chiral
rhodium-diphosphine complexes into self-assembled thiolate
9
8728 J. AM. CHEM. SOC. VOL. 127, NO. 24, 2005

Rh Complexes at Thiolate Monolayers on Au Surfaces
A R T I C L E S
Disulfide 4. Oxalyl chloride (1.50 mL, 17.7 mmol) was added at 0
°C to a solution of 11-(10′-carboxydecyldisulfanyl)undecanoic acid
(2.20 g, 5.00 mmol) in freshly distilled CH2Cl2 (80 mL). The solution
was stirred for 12 h at 23 °C. The solvent was concentrated under
reduced pressure and the residue redissolved in freshly distilled CH2Cl2
(50 mL), which was then added at 0 °C to a solution of p-nitrophenol
(1.40 g, 10.0 mmol) and pyridine (0.80 mL, 10.0 mmol) in freshly
distilled CH2Cl2 (100 mL) and stirred for 12 h at 23 °C. The resulting
reaction mixture was washed with a HCl solution (0.1 M, 100 mL)
and water (100 mL). The organic phase was dried over MgSO4, filtered,
and concentrated under reduced pressure. Purification of the crude
material by flash chromatography, eluting with 20% AcOEt in hexanes,
gave 85% yield of the desired disulfide 4 (2.91 g). Mp: 58-60 °C. Rf
) 0.48 (hexanes/AcOEt, 5:1). 1H NMR (400.1 MHz, CD2Cl2, 295 K):
δ 1.30 (sb, 16 H, CH2), 1.37 (mc, 8 H, CH2), 1.65 (quin, J ) 7.5 Hz,
4 H, SCH2CH2), 1.73 (quin, J ) 7.5 Hz, 4 H, COCH2CH2), 2.58 (t, J
) 7.5 Hz, 4 H, COCH2), 2.67 (t, J ) 7.5 Hz, 4 H, SCH2), 7.27 (d, J
) 9.1 Hz, 4 H, CO2CCHAr), 8.24 (d, J ) 9.1 Hz, 4 H, NO2CCHAr).
13C{1H} NMR (100.6 MHz, CD2Cl2, 295 K): δ 24.8 (CH2), 28.5 (CH2),
29.1 (CH2), 29.3 (CH2), 29.3 (CH2), 29.3 (CH2), 29.4 (CH2), 29.5 (CH2),
34.3 (CH2), 39.2 (SCH2), 122.6 (CO2CCHAr), 125.2 (NO2CCHAr), 145.3
(NO2CAr), 155.7 (CO2CAr), 171.4 (CO2). IR (KBr): ν (cm-1) ) 3100w,
2923s, 2851s, 1756s, 1620m, 1593m, 1525s, 1490m, 1471m, 1412w,
1344s, 1283m, 1208s, 1124sb, 1012m, 923m, 866m, 717m, 495m. MS
(FAB): m/z (rel int %) 676 (M+, 23), 538 (26), 399 (7), 232 (7), 171
(25), 140 (53), 115 (7), 101 (19), 97 (15), 87 (54), 83 (37), 69 (60), 55
(100), 41 (60). Anal. Calcd for C34H48N2O8S2: C, 60.33; H, 7.15; N,
4.14; O, 18.91. Found: C, 60.44; H, 7.05; N, 4.03; O, 18.84.
the surface and that the ordered structure of the n-octanethiolate
SAM can be retained during incorporation of the functionalized
thiols. We are currently varying the structure of the components
of the mixed SAMs, to study the influence of the neighboring
thiolate molecules on the reactivity and enantioselectivity of
the rhodium catalyst. In the systems investigated up to now,
the thiolate linker attached to the rhodium complex extends
beyond the surface of the SAM and, therefore, is exposed to
the solvent. It will be interesting to see how systems in which
the rhodium center is located inside the SAM behave.
Experimental Section
General Methods. Reactions with air- or moisture-sensitive com-
pounds were performed under argon using standard Schlenk techniques
or under purified N2 in a MBraun glovebox. Glassware was oven-dried
and flame-dried prior to use. All chemicals were purchased from Fluka
or Aldrich. CH2Cl2 was dried over CaH2 and THF over Na/K and freshly
distilled under a stream of nitrogen prior to use. Optical rotations were
measured in a Perkin-Elmer polarimeter 341, sodium lamp, 1 dm cuvette
length, c in g/100 mL. 1H, 13C, 19F, and 31P NMR spectra were recorded
in CD2Cl2 or CDCl3 on a Bruker Avance 400 or 500 MHz spectrometer,
and coupling constants are reported in Hz. Chemical shifts are given
in ppm relative to TMS. IR spectra: 1600 Perkin-Elmer Series FT-IR
spectrometer. FAB mass spectra: Finnigan MAT 312. EI mass
spectra: VG 70 SE. Elemental analyses were carried out on Leco CHN-
900 and Leco RO-478 instruments. Chromatographic purifications were
performed by flash chromatography using silica gel (Merck 0.040-
0.063 mm). Yields for final products refer to isolated products and are
the average of three runs. GC analyses: Carlo Erba HRGC Mega2
Series MFC 800 (column: Restek Rtx-1701; 0.25 mm, 30 m, 60 kPa
He). HPLC analyses: Shimadzu VP-system (column: Daicel OD-H;
4.6 × 250 mm).
4-Nitrophenyl 11-Mercaptoundecanoate (5). 1-Ethyl-3-(3′-dimeth-
ylaminopropyl)carbodiimide hydrochloride (3.16 g, 16.5 mmol) and
N-hydroxybenzotriazole (2.23 g, 16.5 mmol) were added to a solution
of 11-mercaptoundecanoic acid (3.00 g, 13.7 mmol) in freshly distilled
CH2Cl2 (150 mL) at 0 °C. After 5 min, p-nitrophenol (1.91 g, 13.7
mmol) and pyridine (4.00 mL, 49.5 mmol) were added, and the reaction
mixture was stirred for 6 h at 23 °C. The resulting solution was washed
with a HCl solution (1 M, 100 mL) and water (100 mL). The organic
phase was dried over MgSO4, filtered, and concentrated under reduced
pressure. Purification of the crude material by flash chromatography,
eluting with 25 vol % AcOEt in hexanes, gave 63% yield of
4-nitrophenyl 11-mercaptoundecanoate 5 (2.90 g). Mp: 38-40 °C. Rf
Disulfide 1. Oxalyl chloride (1.50 mL, 17.7 mmol) was added at 0
°C to a solution of 11-(10′-carboxydecyldisulfanyl)undecanoic acid
(2.20 g, 5.00 mmol) in freshly distilled dichloromethane (80 mL). The
solution was stirred for 12 h at 23 °C. The solvent was concentrated
under reduced pressure and the residue redissolved in freshly distilled
dichloromethane, which was then added at 0 °C to a solution of
PYRPHOS10 (4.60 g, 10.5 mmol) and triethylamine (3.00 mL, 21.6
mmol) in freshly distilled CH2Cl2 (100 mL) and stirred for 12 h at 23
°C. The resulting reaction mixture was washed with a NaOH solution
(0.1 M, 100 mL) and water (100 mL). The organic phase was dried
over MgSO4, filtered, and concentrated under reduced pressure.
Purification of the crude material by flash chromatography, eluting with
50% AcOEt in hexanes, gave 72% yield of the desired disulfide 1
(4.60 g). Mp: 46-48 °C. [R]2D0 +80.1 (c 0.85, CHCl3). Rf ) 0.70
(hexanes/AcOEt, 1:1). 1H NMR (400.1 MHz, CD2Cl2, 295 K): δ 1.26
(sb, 6 H, CH2), 1.29 (sb, 4 H, CH2), 1.37 (mc, 2 H, CH2CH2CH2S),
1.52 (mc, 2 H, COCH2CH2), 1.66 (quin, J ) 7.4 Hz, 2 H, CH2CH2S),
2.08 (t, J ) 7.6 Hz, 2 H, COCH2), 2.68 (t, J ) 7.4 Hz, 2 H, CH2S),
2.87 (mc, 1 H, PCH), 2.95 (mc, 1 H, PCH), 3.33 (t, J ) 11.8 Hz, 1 H,
NCH2), 3.55 (t, J ) 13.0 Hz, 1 H, NCH2), 3.81 (mc, 1 H, NCH2),
3.89 (mc, 1 H, NCH2), 7.12-7.17 (m, 4 H, CHPh), 7.20-7.27 (m, 4 H,
CHPh), 7.32-7.40 (m, 12 H, CHPh). 13C{1H} NMR (100.6 MHz, CD2-
Cl2, 295 K): δ ) 24.9 (COCH2CH2), 28.6 (CH2CH2CH2S), 29.3 (CH2),
29.3 (CH2), 29.4 (CH2), 29.5 (CH2), 29.5 (CH2), 29.5 (CH2), 34.6
(COCH2), 37.2 (mc, JP,C ) 14 Hz, PCH), 39.1 (mc, JC,P ) 14 Hz, PCH),
39.2 (CH2S), 47.9 (mc, NCH2), 48.7 (mc, NCH2), 128.6-129.6 (m,
CHPh-ortho and para), 133.4-133.8 (m, CHPh-meta), 135.9-136.7 (m,
1
) 0.55 (hexanes/AcOEt, 4:1). H NMR (400.1 MHz, CDCl3, 295 K):
δ 1.32 (t, J ) 7.4 Hz, 1 H, SH), 1.35 (mc, 12 H, CH2), 1.59 (quin, J
) 7.4 Hz, 2 H, HSCH2CH2), 1.74 (quin, J ) 7.4 Hz, 2 H, COCH2CH2),
2.50 (q, J ) 7.4 Hz, 2 H, HSCH2), 2.58 (t, J ) 7.4 Hz, 2 H, COCH2),
7.26 (d, J ) 9.1 Hz, 2 H, CO2CCHAr), 8.25 (d, J ) 9.1 Hz, 2 H, NO2-
CCHAr). 13C{1H} NMR (100.6 MHz, CDCl3, 295 K): δ 24.6 (CH2),
24.7 (CH2), 28.4 (CH2), 29.0 (CH2), 29.1 (CH2), 29.2 (CH2), 29.3 (CH2),
29.4 (CH2), 34.0 (CH2), 34.3 (CH2), 122.5 (CO2CCHAr), 125.2
(NO2CCHAr), 145.2 (NO2CAr), 155.5 (CO2CAr), 171.3 (CO2). IR
(KBr): ν (cm-1) 3117w, 3083w, 2921s, 2851s, 1754s, 1620m, 1592m,
1535m, 1487m, 1413w, 1348s, 1202m, 1138s, 1010w, 923m, 860m,
716m, 492m. MS (FAB): m/z (rel int %) 340 ([M + H]+, 7), 201
(100), 140 (43), 123 (23), 97 (16), 83 (44), 69 (51), 55 (90), 41 (64).
Anal. Calcd for C17H25NO4S: C, 60.15; H, 7.42; N, 4.13; O, 18.85.
Found: C, 60.17; H, 7.45; N, 4.13; O, 18.92.
Thiol 11. 1-Ethyl-3-(3′-dimethylaminopropyl)carbodiimide hydro-
chloride (0.96 g, 5.00 mmol) and N-hydroxybenzotriazole (0.68 g, 5.00
mmol) were added to a solution of 11-mercaptoundecanoic acid (1.15
g, 5.00 mmol) in freshly distilled CH2Cl2 (80 mL) at 0 °C. After 5
min, PYRPHOS10 (1.85 g, 4.20 mmol) and triethylamine (1.50 mL,
10.8 mmol) were added, and the reaction mixture was stirred for 6 h
at 23 °C. The resulting solution was washed with a HCl solution (1 M,
100 mL) and water (100 mL). The organic phase was dried over
MgSO4, filtered, and concentrated under reduced pressure. Purification
of the crude material by flash chromatography, eluting with 50% diethyl
C
Ph-ipso), 171.5 (CO). 31P{1H} NMR (162.0 MHz, CD2Cl2, 295 K): δ
-12.5 (d, J ) 8 Hz), -12.3 (d, J ) 8 Hz). IR (KBr): ν (cm-1) 3051w,
2923s, 2850m, 1644s, 1433s, 1197m, 1116m, 1095m, 743m, 696s,
506m. MS (FAB): m/z (rel int %) 1277 (M+, 2), 638 (12), 456 (10),
262 (11), 254 (9), 201 (23), 185 (98), 108 (12), 68 (100), 55 (27),
41(24).
9
J. AM. CHEM. SOC. VOL. 127, NO. 24, 2005 8729

A R T I C L E S
Belser et al.
ether in n-pentane, gave 86% yield of thiol 11 (2.30 g). Mp: 92-93
°C. [R]2D0 +125.8 (c 0.93, CHCl3). Rf ) 0.36 (n-pentane/diethyl ether,
([M - BArF]+, 22), 738 (24), 540 (18), 474 (17), 395 (22), 317 (42),
286 (53), 255 (26), 241 (46), 210 (100), 183 (30), 103 (37), 89 (26),
77 (40), 68 (58), 55 (30), 39 (38). Anal. Calcd for C79H71NOBF24P2-
RhS: C, 55.36; H, 4.18; N, 0.82; O, 0.93. Found: C, 55.25; H, 4.12;
N, 0.81; O, 1.02.
1
1:1). H NMR (500.1 MHz, CD2Cl2, 295 K): δ 1.27 (sb, 12 H, CH2),
3
1.37 (t, JH,H ) 7.6 Hz, 1 H, SH), 1.53 (mc, 2 H, CH2), 1.60 (quin,
3JH,H ) 7.3 Hz, 2 H, CH2CH2SH), 2.09 (t, 3JH,H ) 7.7 Hz, 2 H, COCH2),
3
2.52 (dt, JH,H ) 7.3 Hz, 7.6 Hz, 2 H, CH2SH), 2.87 (mc, 1 H, PCH),
Thiol 18. 1-Ethyl-3-(3′-dimethylaminopropyl)carbodiimide hydro-
chloride (0.91 g, 4.77 mmol) and N-hydroxybenzotriazole (0.64 g, 4.77
mmol) were added to a solution of 11-mercaptoundecanoic acid (1.00
g, 4.70 mmol) in freshly distilled CH2Cl2 (100 mL) at 0 °C. After 5
min, (3R,4R)-3,4-bis(diphenylphosphorothioyl)pyrrolidine (2.00 g, 3.97
mmol) and triethylamine (1.80 mL, 12.8 mmol) were added, and the
reaction mixture was stirred for 6 h at 23 °C. The resulting solution
was washed with a saturated NH4Cl solution (100 mL) and water (100
mL). The organic phase was dried over MgSO4, filtered, and concen-
trated under reduced pressure. Purification of the crude material by
flash chromatography, eluting with 30% diethyl ether in n-pentane, gave
80% yield of thiol 18 (2.24 g). Mp: 62-64 °C. [R]2D0 +26.3 (c 0.83,
CHCl3). Rf ) 0.26 (n-pentane/diethyl ether, 1:2). 1H NMR (500.1 MHz,
CDCl3, 295 K): δ 1.16-1.29 (m, 10 H, CH2), 1.33 (t, J ) 7.6 Hz, 1
H, SH), 1.31-1.39 (m, 2 H, CH2), 1.48 (quinb, J ) 7.0 Hz, 2 H, CH2),
1.59 (quin, J ) 7.4 Hz, 2 H, CH2CH2SH), 1.93 (mc, 2 H, COCH2),
2.51 (mc, J ) 7.1 Hz, 7.6 Hz, 2 H, CH2SH), 3.55-3.64 (m, 1 H, PCH),
3.74-3.83 (m, 2 H, NCH2), 3.84-3.93 (m, 1 H, PCH), 3.96-4.10 (m,
2 H, NCH2), 7.21-7.28 (m, 4 H, CHPh), 7.36-7.53 (m, 8 H, CHPh),
7.60-7.66 (m, 2 H, CHPh), 7.71-7.76 (m, 2 H, CHPh), 7.79-7.87 (m,
4 H, CHPh). 13C{1H} NMR (125.8 MHz, CDCl3, 295 K): δ 24.9 (CH2),
25.0 (CH2), 28.7 (CH2), 29.4 (CH2), 29.6 (CH2), 29.7 (CH2), 29.7 (CH2),
29.8 (CH2), 34.4 (CH2), 34.7 (CH2), 38.7 (d, JC,P ) 54 Hz, PCH), 39.9
(d, JC,P ) 54 Hz, PCH), 48.1 (s, NCH2), 48.4 (s, NCH2), 129.1-129.3
3
2.94 (mc, 1 H, PCH), 3.34 (t, JH,H ) 11.9 Hz, 1 H, NCH2), 3.67 (t,
3JH,H ) 13.0 Hz, 1 H, NCH2), 3.90 (mc, 1 H, NCH2), 3.94 (mc, 1 H,
NCH2), 7.13-7.19 (m, 4 H, CHPh), 7.21-7.28 (m, 4 H, CHPh), 7.32-
7.45 (m, 12 H, CHPh). 13C{1H} NMR (125.8 MHz, CD2Cl2, 295 K): δ
24.7 (CH2), 24.9 (CH2), 28.5 (CH2), 29.1 (CH2), 29.5 (CH2), 29.5 (CH2),
31.0 (CH2), 31.0 (CH2), 34.2 (CH2), 34.8 (CH2), 37.3 (mc, PCH), 39.2
(mc, PCH), 47.9 (mc, NCH2), 48.8 (mc, NCH2), 128.6-129.6 (m,
CHPh-ortho
para), 133.5-133.7 (m, CHPh-meta), 135.9-136.8 (m,
and
C
Ph-ipso), 171.5 (CO).31P{1H} NMR (202.5 MHz, CD2Cl2, 295 K): δ
-12.5 (d, JP,P ) 8 Hz), -12.3 (d, JP,P ) 8 Hz). IR (KBr): ν (cm-1
)
3
3
3053w, 2926s, 2847s, 1643s, 1432s, 1332m, 1270w, 1246w, 1214m,
1197w, 1090m, 1068m, 1025m, 999m, 739s, 702s, 646m, 543w, 513m,
482m. MS (EI): m/z (rel int %) 639 (M+, 16), 530 (13), 371 (11), 268
(13), 262 (64), 236 (7), 185 (47), 108 (16), 68 (100), 55 (12). Anal.
Calcd for C39H47NOP2S: C, 73.21; H, 7.40; N, 2.19; Found: C, 73.12;
H, 7.36; N, 2.15.
Synthesis of Rhodium Complex 12. [Rh(COD)Cl]2 (0.89 g, 1.80
11,12
mmol) and TlBArF
(3.84 g, 3.60 mmol) were dissolved in freshly
distilled CH2Cl2 (50 mL) and stirred for 1 h at 23 °C. The precipitated
TlCl was filtered off and washed with freshly distilled CH2Cl2 (10 mL).
To the resulting orange solution was added at 0 °C slowly a solution
of thiol 11 (2.30 g, 3.60 mmol) in freshly distilled CH2Cl2 (20 mL)
and stirred for 1 h at 23 °C. The solvent was concentrated at reduced
pressure. The residue was purified by chromatography on silica gel,
eluting with CH2Cl2 to provide the desired orange complex 12 (3.18
g) in 52% yield. Mp: 114-116 °C. [R]2D0 -25.8 (c 0.24, CHCl3). Rf )
0.39 (CH2Cl2). 1H NMR (500.1 MHz, CD2Cl2, 295 K): δ 1.19 (mc, 10
(m, CHPh-meta), 129.6-131.3 (m, CPh-ipso), 131.7-132.4 (m, CHPh
and
-ortho
para), 171.5 (CO). 31P{1H} NMR (202.5 MHz, CDCl3, 295 K): δ
50.8 (d, J ) 41 Hz), 51.0 (d, J ) 41 Hz). IR (KBr): ν (cm-1) 2924s,
2851m, 1639s, 1436s, 1100s, 750m, 713s, 692s, 608m, 520m, 489m.
MS (FAB): m/z (rel int %) 704 ([M + H]+, 9), 268 (8), 217 (16), 136
(9), 68 (100), 57 (13), 41 (10). Anal. Calcd for C39H47NOP2S3: C,
66.54; H, 6.73; N, 1.99; O, 2.27. Found: C, 66.45; H, 6.62; N, 1.92;
O, 2.35.
3
H, CH2), 1.33 (quin, JH,H ) 7.3 Hz, 2 H, CH2CH2CH2SH), 1.36 (t,
3JH,H ) 7.3 Hz, 1 H, SH), 1.40 (quin, 3JH,H ) 7.3 Hz, 2 H, COCH2CH2),
1.63 (quin, 3JH,H ) 7.3 Hz, 2 H, CH2CH2SH), 1.94 (mc, 2 H, COCH2),
2.13 (mc, 4 H, CH2-COD), 2.43 (mc, 2 H, CH2-COD), 2.54 (mc, 2 H,
CH2-COD), 2.65 (q, 3JH,H ) 7.3 Hz, 2 H, CH2SH), 2.81 (mc, 1 H, NCH2),
2.94 (mc, 1 H, NCH2), 3.02 (mc, 1 H, PCH), 3.12 (mc, 1 H, PCH), 3.59
(mc, 1 H, NCH2), 3.83 (mc, 1 H, NCH2), 4.52 (mc, 1 H, CHCOD), 4.56
(mc, 1 H, CHCOD), 5.20 (mc, 1 H, CHCOD), 5.23 (mc, 1 H, CHCOD), 7.42
(mc, 4 H, CHPh-ortho), 7.55 (mc, 4 H, CHPh-meta), 7.56 (sbr, 4 H,
CHPh-para-BArF), 7.58 (mc, 2 H, CHPh-para), 7.64 (mc, 4 H, CHPh-meta),
7.70 (mc, 2 H, CHPh-para), 7.75 (sb, 8 H, CHPh-ortho-BArF), 7.92 (mc, 4
H, CHPh-ortho). 13C{1H} NMR (125.8 MHz, CD2Cl2, 295 K): δ 24.7
(COCH2CH2), 28.6 (CH2-COD), 28.8 (CH2CH2CH2SH), 29.4 (CH2CH2-
SH), 29.5 (CH2), 29.6 (CH2), 29.7 (CH2), 29.7 (CH2), 29.8 (CH2), 32.2
(CH2-COD), 34.3 (COCH2), 39.3 (CH2SH), 41.7 (ddd, J ) 2 Hz, 18
Hz, 32 Hz, PCH), 43.2 (ddd, J ) 2 Hz, 19 Hz, 32 Hz, PCH), 43.4 (mc,
JP,C ) 4 Hz, 17 Hz, NCH2), 44.4 (dd, JP,C ) 4 Hz, 16 Hz, NCH2), 98.6
(dd, JP,C ) 7 Hz, 8 Hz, CHCOD), 98.7 (dd, JP,C ) 7 Hz, 8 Hz, CHCOD),
104.8 (dd, JP,C ) 7 Hz, 9 Hz, CHCOD), 104.9 (dd, JP,C ) 7 Hz, 9 Hz,
General Procedure for the Synthesis of Functionalized Gold
Particles. Gold particles and all glassware used for the preparation
and storage of gold particles was treated with aqua regia, rinsed with
deionized water, cleaned in a bath consisting of seven parts of
concentrated sulfuric acid and three parts of 30% hydrogen peroxide
(piranha solution), rinsed again with deionized water, and dried for 12
h at 120 °C. All reactions were carried out under argon.
Freshly cleaned gold particles (0.50 g) were transferred into a 10
mL gastight SGE syringe (fitted with a push button luer lock valve
and an integrated filter), containing a solution of the desired disulfide
(0.05 mmol) in freshly distilled CH2Cl2 (4 mL). The syringe was shaken
(80 rpm) under argon for 24 h at 23 °C. The functionalized gold
particles were filtered in the syringe under argon, washed with freshly
distilled CH2Cl2 (8 × 5 mL), and dried in vacuo for 1 h to give the
functionalized gold particles.
[Rh(COD)Cl]2 (0.10 g, 0.20 mmol) and TlBArF
11,12
3
1
(0.43 g, 0.40
CHCOD), 117.9 (sept, JF,C ) 4 Hz, CHPh-para-BArF), 124.9 (q, JF,C
)
)
272 Hz, CF3-BArF), 126.2 (d, 1JP,C ) 42 Hz, CPh-ipso), 126.3 (d, 1JP,C
mmol) were dissolved in freshly distilled CH2Cl2 (10 mL) and stirred
for 1 h at 23 °C. The precipitated TlCl was filtered off and washed
with freshly distilled CH2Cl2 (2 mL). The resulting orange solution
was transferred under argon into a 10 mL gastight SGE syringe (fitted
with a push button luer lock valve and an integrated filter), containing
the functionalized gold particles 2. The syringe was shaken (80 rpm)
under argon for 12 h at 23 °C. The functionalized gold particles were
filtered in the syringe under argon, washed with freshly distilled CH2Cl2
(8 × 5 mL), and dried in vacuo for 1 h to give the functionalized gold
particles 3, which were immediately used for the asymmetric hydro-
genation of methyl R-acetamidocinnamate.
43 Hz, CPh-ipso), 127.7 (d, 1JP,C ) 44 Hz, CPh-ipso), 127.8 (d, 1JP,C ) 44
2
3
Hz, CPh-ipso), 129.2 (q, JF,C ) 31 Hz, JB,C ) 3 Hz, CCF3), 130.4 (d,
3JP,C ) 12 Hz, CHPh-meta), 130.5 (d, JP,C ) 12 Hz, CHPh-meta), 130.6
3
(d, 3JP,C ) 16 Hz, CHPh-meta), 130.7 (d, 3JP,C ) 16 Hz, CHPh-meta), 132.2
4
4
(d, JP,C ) 2.4 Hz, CHPh-para), 133.8 (d, JP,C ) 2.4 Hz, CHPh-para),
135.2 (mb, CHPh-ortho-BArF), 136.9 (d, 2JP,C ) 21 Hz, CHPh-ortho), 137.0
2
1
(d, JP,C ) 21 Hz, CHPh-ortho), 162.1 (q, JB,C ) 50 Hz, CPh-ipso-BArF),
172.1 (CO). 31P{1H} NMR (202.5 MHz, CD2Cl2, 295 K): δ 30.1 (dd,
1JP,Rh ) 149 Hz, 2JP,P′ ) 28 Hz, P), 31.1 (dd, 1JP,Rh ) 149 Hz, 2JP,P′
)
28 Hz, P′). 19F{1H} NMR (470.6 MHz, CD2Cl2, 295 K): δ -63.9 (s).
IR (KBr): ν (cm-1) 2929m, 1653s, 1430mb, 1356s, 1278s, 1127sb,
889m, 839w, 745m, 694mb, 522m. MS (FAB): m/z (rel int %) 850
General Procedure for the Synthesis of n-Alkanethiolate-
Protected Gold Colloids. All glassware used for the preparation and
9
8730 J. AM. CHEM. SOC. VOL. 127, NO. 24, 2005

Rh Complexes at Thiolate Monolayers on Au Surfaces
A R T I C L E S
storage of colloidal Au was treated with aqua regia, rinsed with
deionized water, cleaned in a bath consisting of seven parts of
concentrated sulfuric acid and three parts of 30% hydrogen peroxide
(piranha solution), rinsed again with deionized water, and dried for 12
h at 120 °C. All reactions were carried out under argon in the dark.
To a vigorously stirred solution (1200 rpm) of HAuCl4‚(H2O)x (3.00
g, ∼49% Au, 1 equiv) in deionized H2O (240 mL) was added a solution
of tetraoctylammoniumbromide (4.50 g, 1.1 equiv) in freshly distilled
toluene (165 mL). The yellow color of the gold acid solution quickly
hood and the reaction solution treated immediately with 10 equiv of
neat COD. The solution was diluted with 10 mL of ethyl acetate and
transferred to a Millipore Centriplus centrifugal filter device with a
cutoff molecular size of 50 kD. After centrifugation (3000 rpm) at 23
°C for 60 min, the filtrate containing the product was removed. The
remaining colloid solution was diluted with ethyl acetate and centrifu-
gated again. This procedure was repeated twice in order to remove
any remaining product. The concentrated colloidal solution was
transferred to a flask, concentrated to near dryness, and treated with
hexanes to induce precipitation. The resulting suspension was cooled
to -78 °C for 2 h. The black precipitate, which had formed, was isolated
by filtration and washed with hexanes. The colloids were then dried
under vacuum and reused as catalysts.
STM Investigations. Au(111) Substrate. The gold films were
obtained by epitaxial growth of gold onto mica at a base pressure of 5
× 10-7 mbar while the substrate was held at 350 °C, similar to the
procedure described in the literature.36 Shortly before the preparation
of the SAMs the gold substrates were either heated to 500 °C for 2 h
in an oven or flame-annealed to remove contaminants.37 The latter
process turned out to be more reliable to give clean and smooth gold
surfaces.
SAM Preparation. The gold films were immersed in a solution of
n-octanethiols under argon atmosphere, whereas the immersion time
(ranging from 3 to 24 h), the solvent (ethanol, dichloromethane, acetone,
toluene, and diethyl ether), the concentration (ranging from 1 mM to
0.1 M), and the solution temperature (ranging from 23 to 60 °C) were
varied in order to determine the best conditions for obtaining upright
standing n-octanethiolate monolayers in the well-known (x3×x3)-
R30 hexagonal arrangement. Upon removal from solution, the samples
were carefully rinsed with pure solvent.
Exchange Reaction. Only samples that showed the hexagonal
(x3×x3)R30 n-octanethiolate ordering, which was checked with STM,
were used to perform the exchange reaction. For this purpose, the
samples were immersed in a 0.1-2 mM solution of either the ligand
18 or the ionic catalyst 12. As solvent, ethanol or dichloromethane
were used, while the immersion time was varied between 1 and 48 h.
Long immersion times (∼8 h) led to disruption of the hexagonal
n-octanethiolate ordering by desorption of n-octanethiol. However,
addition of n-octanethiol to the ligand 18 solution prevented this effect.
For the exchange reaction with catalyst 12, no n-octanethiol was added
to the solution.
-
disappeared and the toluene phase became orange-red as the AuCl4
was transferred into it. The mixture was shaken several times until the
aqueous phase became colorless. The organic phase was isolated, and
n-octanethiol (0.26 mL, 0.20 equiv) was added to the solution. The
resulting solution was cooled to 0 °C. Under vigorous stirring (1200
rpm) was subsequently added NaBH4 (3.10 g, 11 equiv) in deionized
water (200 mL). The colloids were transformed instantaneously, as
witnessed by the color change of the solution from red to black. The
very dark organic phase was further stirred for 12 h at 23 °C. The
organic phase was collected and the solution concentrated to 20 mL at
reduced pressure (the water bath should not exceed 30 °C to prevent
partial product decomposition). The black product was suspended in
ethanol (300 mL) and briefly sonicated to ensure complete dissolution
of byproducts. The mixture was kept at -60 °C for 48 h to effect
complete precipitation. The colloids were collected on a glass filtration
frit and washed twice with ethanol (200 mL) and twice with acetone
(200 mL). The black material was dried in vacuo for 1 h to give
n-octanethiol-passivated gold nanoparticles (1.84 g), free of significant
residual n-octanethiol. The absence of free n-alkanethiol was confirmed
by 1H NMR spectroscopy (CD2Cl2, no R-CH2 resonance at ∼2.5 ppm)
and IR spectroscopy (no νasym(SH) band in the region of 2576 cm-1).
On the basis of elemental analysis (C 7.15%, H 1.31%, S 2.38%, Au
90.1%), TEM (3.48 ( 0.61 nm), and Leff’s model,17 the colloids
contained an average number of 1298 gold atoms/core and 211 adsorbed
n-octanethiolate molecules.
Exchange Reaction on n-Alkanethiolate-Protected Gold Colloids.
To a solution of n-alkanethiol-protected gold colloids (0.80 g, ∼0.6
mmol adsorbed n-alkanethiol, 1 equiv) in freshly distilled CH2Cl2 (400
mL) was added a solution of functionalized thiol (0.6 mmol, 1 equiv)
in freshly distilled CH2Cl2 (100 mL), and the mixture was stirred (80
rpm) for 24-48 h at 23 °C. The solution was concentrated to 50 mL
under reduced pressure (the water bath should not exceed 30 °C to
prevent partial product decomposition), diluted with ethanol (300 mL),
and briefly sonicated to ensure complete dissolution of byproducts. The
colloids were collected by filtration with a solvent-resistant stirred cell
(Millipore, 76 mm disk filter and a ultrafiltration membrane (regenerated
cellulose, PLTK, NMWL: 30′000)) and washed with ethanol until the
filtrate was colorless, indicating removal of all excess free thiol 12.
The black solution remaining in the stir cell was concentrated to 50
mL, suspended in hexanes (300 mL), and briefly sonicated to ensure
complete dissolution of byproducts. The mixture was kept at -60 °C
for 48 h to effect complete precipitation. The colloids were collected
on a glass filtration frit and washed twice with hexanes (200 mL). The
black material was dried in vacuo for 1 h to give the functionalized
STM Measurements. The STM measurements were carried out at
ambient conditions with a Nanoscope IIIa (Digital Instruments Inc.,
Santa Barbara, CA) equipped with a low-current converter. The tips
were mechanically cut from a Pt/Ir wire (90:10). As typical scan
parameters, (0.9 V and 2 pA were used for the bias voltage and the
tunneling current, respectively.
Acknowledgment. Financial support by the Swiss National
Science Foundation and the National Center of Competence in
Research (NCCR) “Nanoscale Science” is gratefully acknowl-
edged. M.S. acknowledges support from the German Academy
of Natural Scientists Leopoldina under the grant number BMBF-
LPD 9901/8-86.
1
gold nanoparticles, which were analyzed by H NMR and TEM.
Hydrogenation. The rhodium colloid catalyst (5 µmol, 1 mol %)
and methyl R-acetamidocinnamate (110 mg, 0.50 mmol) were added
to a borosilicate glass vial (30 × 60 mm) in the glovebox. After addition
of the appropriate solvent (1 mL) and a stirring bar (6 × 15 mm), the
vial was placed in an autoclave, which was pressurized with hydrogen
at 60 bar. The reaction mixture was stirred (600 rpm) for 4 h at 23 °C.
After the reaction, the hydrogen pressure was released, the solvent was
removed, and the residue was purified on a short silica plug to remove
the catalyst (6 × 20 mm; eluting with ethyl acetate). The product was
dissolved in ethanol and used directly for HPLC and GC analysis.
Catalyst Recycling. The autoclave was slowly opened in the fume
Supporting Information Available: Full details of ligand/
colloid synthesis and characterization as well as complete ref
13. This material is available free of charge via the Internet at
http://pubs.acs.org.
JA0500714
(36) DeRose, J. A.; Thundat, T.; Nagahara, L. A.; Lindsay, S. M. Surf. Sci.
1991, 265, 102.
(37) Dishner, M. H.; Ivey, M. M.; Gorer, S.; Hemminger, J. C.; Feher, F. J. J.
Vac. Sci. Technol. A 1998, 16, 329.
9
J. AM. CHEM. SOC. VOL. 127, NO. 24, 2005 8731