Characterization and evaluation of Ag-Pt/SiO2 catalysts prepared by electroless deposition

Article abstract of DOI:10.1016/j.jcat.2007.12.006

A series of Ag-Pt/SiO₂ catalysts have been prepared by the electroless deposition of Ag onto a Pt/SiO₂ catalyst. Results indicate that Ag deposition does not readily occur on the SiO₂ support, but is essentially restricted

Full text of DOI:10.1016/j.jcat.2007.12.006

Journal of Catalysis 254 (2008) 131–143
www.elsevier.com/locate/jcat
Characterization and evaluation of Ag–Pt/SiO2 catalysts prepared
by electroless deposition
Melanie T. Schaal, Anna C. Pickerell, Christopher T. Williams, John R. Monnier ∗
Department of Chemical Engineering, University of South Carolina, 2C02 Swearingen Engineering Center, 301 S. Main Street, Columbia, SC 29208, USA
Received 29 September 2007; revised 12 December 2007; accepted 17 December 2007
Abstract
A series of Ag–Pt/SiO catalysts have been prepared by the electroless deposition of Ag onto a Pt/SiO catalyst. Results indicate that Ag
2
2
deposition does not readily occur on the SiO support, but is essentially restricted to the Pt surface. The Ag–Pt catalysts have been characterized
2
12
13
by FTIR of adsorbed CO and CO, which suggest that Ag is preferentially located on Pt(111) sites rather than the more coordinatively-
unsaturated corner and edge sites. Hydrogenation of 3,4-epoxy-1-butene (EpB), a multifunctional olefin, was chosen as a probe reaction. Reaction
data indicate that ED-derived catalysts provide a more targeted Ag placement on the Pt surface than traditional incipient wetness catalysts. EpB
conversion increased dramatically (∼3×) with the addition of sub-monolayer coverages of Ag on the Pt surface. Conversely, activity for propylene
hydrogenation decreased with increasing Ag coverages. The enhanced EpB hydrogenation appears to be due to a Ag-induced decrease in EpB
adsorption energy on surface Pt sites.
© 2007 Elsevier Inc. All rights reserved.
Keywords: Platinum; Silver; Bimetallic; Catalyst preparation; Electroless deposition; 3,4-epoxy-1-butene; EpB; Hydrogenation
1. Introduction
ducible methods are needed for the preparation of truly bimetal-
lic catalysts.
Over the last several decades bimetallic or multimetallic
catalysts have replaced monometallic catalysts in industrial
processes due to the beneficial activity, selectivity, and sta-
bility modifications that may be achieved by the inclusion of
additional metals [1–3]. These bimetallic and multimetallic
catalysts are often prepared on an industrial scale using ei-
ther successive impregnation or co-impregnation methods [4].
However, these traditional methodologies frequently provide
inadequate control over metal placement and accordingly yield
catalysts containing both isolated, monometallic particles and
bimetallic particles with varying compositions [5–7]. This com-
plex mixture of particles results in poor control of the final
catalyst performance and makes any definitive correlations be-
tween catalyst performance, catalyst characterization, and cat-
alyst composition virtually impossible. Therefore, new, repro-
One possible alternative to traditional preparative techniques
is electroless deposition (ED), wherein a controlled chemical
reaction is used to deposit a metal salt onto a catalytic metal
site that has been activated by a reducing agent. The process
may be either catalytic (deposition of the metal salt [metal A+]
from solution onto the pre-existing metal [metal B0] on the sup-
port) or autocatalytic (deposition of the metal salt [metal A+]
onto the just reduced, deposited metal [metal A0]), depend-
ing on the nature of the activated metal site [8]. In principle,
this ED process should result in the selective deposition of the
secondary metal onto a metallic surface with no formation of
isolated crystallites of the secondary metal on the catalyst sup-
port. Perhaps most importantly, ED is an industrially-relevant
process [8,9] that can be applied to a wide range of metals [8].
A more detailed description of this process has been published
elsewhere [10].
Once bimetallic catalysts are prepared, they can be eval-
uated using selective probe reactions to correlate changes in
catalyst performance with catalyst composition. In the present
study, hydrogenation of 3,4-epoxy-1-butene (EpB) was selected
*
Corresponding author. Fax: +1 803 777 8265.
E-mail addresses: mrtimmon@engr.sc.edu (M.T. Schaal),
pickerea@engr.sc.edu (A.C. Pickerell), willia84@engr.sc.edu (C.T. Williams),
monnier@engr.sc.edu (J.R. Monnier).
0021-9517/$ – see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2007.12.006

132
M.T. Schaal et al. / Journal of Catalysis 254 (2008) 131–143
a prerequisite for adsorption on Pt surfaces, in good agreement
with recent data from Loh and Medlin [20]. Further, this ten-
dency of gas phase EpB to inhibit readsorption and reaction was
also noted for deoxygenation and ring-opened products [11].
EpB hydrogenation has also been studied over Pd [11],
Rh [17], Cu [19], and Cu–Pd [10] catalysts. In general, eval-
uation trends of these heterogeneous catalysts indicate that Pd
favors C–O hydrogenolysis of the epoxy group due to the sta-
bility of the resulting π-allylic intermediates on Pd sites [16],
while Cu0 exhibits high selectivity to 1,3-butadiene by deoxy-
genation of EpB [19]. Pt and Rh are more selective for olefin
hydrogenation, although Pt appears to be less selective than
Rh at comparable conditions [16,17]. These results suggest that
EpB is chemisorbed and activated quite differently for these dif-
ferent metallic catalysts [11].
In this study a family of catalysts has been synthesized to
illustrate the effectiveness of the electroless deposition method
to tune bimetallic surfaces for different catalytic purposes. The
levels of Ag deposition have been intentionally limited to sub-
monolayer coverages on the Pt surface, as verified by selective
H2 chemisorption on the Pt component of the Ag–Pt bimetallic
surface. The evaluation results of the ED-derived Ag–Pt/SiO2
catalysts are compared to those obtained for catalysts prepared
using traditional incipient wetness methods. Although the same
general trends are observed, the ED-derived catalysts exhibit
more dramatic changes in activity and selectivity, suggesting
a more targeted and efficient Ag placement on the Pt surface,
which is supported by catalyst characterization data.
Scheme 1. Schematic of select reaction pathways for EpB hydrogenation taken
from literature [10–15].
as a probe reaction. EpB contains two distinctly different func-
tional groups, a C=C bond and an epoxy group, which can
undergo simple hydrogenation and/or hydrogenolysis (cleavage
of a C–C or C–O bond), respectively, to form a wide variety of
reaction products [11]. Selective EpB hydrogenation reaction
products formed from C=C hydrogenation and (C(#3)–O) hy-
drogenolysis/hydrogenation reactions (where C(#3) refers to the
#3 carbon in the EpB structure) that have been reported [10–15]
are summarized in Scheme 1 [10]. However, additional prod-
ucts may be formed from deoxygenation, C–C hydrogenolysis,
or from epoxide C(#4)–O bond cleavage [11,16]. Butylene ox-
ide (BO), formed by the selective hydrogenation of the C=C
bond, is perhaps the most desired product as it can readily be
used to manufacture of a variety of chemicals including poly-
ethers, alkylene glycols, epoxy resins, and fuel additives [17].
However, hydrogenation of epoxides using Group VIII metals
typically results in hydrogenolysis of the C–O bond [16]. Two
different C–O bonds may be cleaved. Ring opening at the more
sterically hindered, allylic C(#3)–O bond is thermodynamically
favored, forming a terminal alkoxy intermediate which results
in the production of both unsaturated and saturated primary al-
cohols and aldehydes [16,18,19]. Alternatively, cleavage of the
C(#4)–O bond ultimately forms methyl vinyl ketone (3-buten-2-
one) and secondary alcohols [11].
The hydrogenation of EpB over Pt/SiO2 was studied by
Bartók et al. in a gas phase recirculation reactor at 0 and 28 ◦C
(20 kPa H2, 1.33 kPa EpB). At 0 ◦C C4 hydrocarbon (1-butene
and butane) and butylene oxide (BO) yields were very sim-
ilar, although after 35 min online the HC yield was higher,
indicating that deoxygenation was the dominant reaction. How-
ever, at 28 ◦C, BO was the major product, followed by C4 HCs
and trace amounts of n-butyraldehyde; BO yield increased with
both increasing temperature and with time online in the recir-
culation reactor indicating that it is a stable end product over
Pt/SiO2 at these conditions. Thus, although BO could undergo
hydrogenolysis over Pt/SiO2, the rate for BO conversion was
much lower than for EpB hydrogenation; in other words, BO
did not readily readsorb and react on the Pt/SiO2 surface in the
presence of gas phase EpB [11], suggesting that a C=C bond is
2. Experimental
2.1. Catalyst preparation
The electroless deposition of Ag on Pt/SiO2 was con-
ducted using an aqueous bath of 17 Mꢀ deionized (DI) water,
AgNO3 (Sigma–Aldrich 99.9999%) as the Ag+ source, NaNO3
(Mallinckrodt AR grade) as the ionic strength adjustor, HCHO
(Sigma Aldrich, 37 wt% ACS reagent) as the reducing agent,
and NaOH (EM pellets, 97% assay) or HNO3 (70% J.T. Baker)
to adjust pH. The base Pt/SiO2 (1.67 wt%; dispersion ∼3.7%
by H2 chemisorption) catalyst was supplied by BASF Cata-
lysts LLC. During the Ag electroless deposition, the [Ag+]
was monitored using a Cole–Parmer Ag+ ion selective elec-
trode. NaNO3 was added to the ED baths in the recommended
quantities to keep the ionic strength of the solution constant for
optimum probe sensitivity. Liquid samples were taken from the
bath before the catalyst addition, immediately preceding sample
filtration, and after washing the catalyst (filtrate sample). Nitric
acid (∼3 vol%) was added to the withdrawn sample aliquot in
order to ensure complete solubilization of the Ag+ species. The
solutions were then evaluated using atomic absorption (Perkin
Elmer, Model 300) to determine the Ag+ concentration, which
in turn was used to determine the Ag loadings on the cata-
lysts.
Using this methodology, a series of Ag–Pt/SiO2 catalysts
with varying Ag weight loadings was prepared. The amount of
available Ag+ ions in solution was limited in order to ensure

M.T. Schaal et al. / Journal of Catalysis 254 (2008) 131–143
133
that the desired loadings were attained at ∼100% deposition of
Ag+. The HCHO concentrations were selected to give an ini-
tial molar ratio of HCHO:Ag+ = 2:1. The 200 mL aqueous ED
baths contained the calculated amounts of AgNO3 and HCHO
as well as 0.1 M NaNO3, and NaOH to give pH 9. The bath pH
was maintained at 9 0.15 by dropwise additions of NaOH so-
lution. Room temperature deposition was allowed to proceed
for two hours after the addition of 1.0 g of Na+-exchanged
Pt/SiO2 to ensure quantitative deposition of Ag. The slurry was
then filtered and washed, and the catalyst was allowed to dry as
a filtrate cake.
Infrared isotopic 13CO studies were conducted to determine
singleton frequencies of CO adsorption. The desired flow rate
of 13CO (99% purity, Cambridge Isotopes) was controlled using
an Orion Syringe pump (5 mL and 50 mL syringes), while the
required flow rates of 12CO and He diluent were added using
mass flow controllers to give a total concentration of 12CO and
13CO equal to 1%. All isotopic IR spectra were deconvoluted
using Peak Solve software. The spectra containing 12CO and
13CO gas mixtures were deconvoluted using the appropriate
peaks and associated FWHM identified in the 100% 12CO and
100% 13CO spectra. Constraints were set such that the FWHM
was held constant while the peak heights and peak maxima
were allowed to vary to account for the varying degrees of CO
coupling and isotopic CO concentration. Care was taken to en-
sure that consistent, physically meaningful shifts were attained
for all peaks.
The Na+-exchanged Pt/SiO2 catalyst was also exposed to
a similar ED bath (except that it contained no AgNO3) to de-
termine if exposure to HCHO and pH 9 conditions caused
changes in the catalytic properties of the base Pt/SiO2 cata-
lyst. Steady-state evaluation results for both the treated and
untreated Pt/SiO2 samples were similar, indicating that neither
HCHO nor pH 9 exposures had any significant effect on cat-
alyst performance. For consistency, however, the ED-derived
catalysts were compared to the treated Pt/SiO2 sample, while
the bimetallic catalysts prepared by incipient wetness methods
were compared to an untreated Pt/SiO2 catalyst.
For comparison, a series of catalysts was prepared using tra-
ditional incipient wetness (IW) methods to deposit Ag (AgNO3
Sigma–Aldrich 99.9999%) on the freshly reduced Pt/SiO2 cata-
lyst. After impregnation these samples were placed in a vacuum
oven at ∼115 ◦C until dry and then re-reduced at 200 ◦C un-
der flowing H2 for one hour. A 2.0 wt% Ag/SiO2 sample was
also prepared by incipient wetness, followed by drying and re-
duction at 300 ◦C for 2 h under flowing H2. Atomic absorption
analyses of the IW solutions were used to determine the Ag
weight loadings.
2.3. EpB hydrogenation
Catalysts were evaluated using 3,4-epoxy-1-butene (EpB)
hydrogenation as the probe reaction. Approximately 0.3 g of
catalyst (particle diameter ∼125–500 µm) were loaded into a
0.375 in. O.D. fixed-bed, gas phase reactor which was config-
ured as has been described previously [10]. Briefly, the catalyst
was reduced in situ for 1 h at 200 ◦C in 20% H2/balance He.
The 100 ◦C, atmospheric pressure reaction was conducted us-
ing a feed stream of 2.5% EpB, 20% H2, and the balance He
with an exit GHSV = 10,800 h−1 (∼100 sccm). Both reaction
feed and products were evaluated using an on-line Hewlett–
Packard 5890 gas chromatograph. Reaction conversion calcu-
lations were performed based on the sum of products formed
(as determined using the FID detector) and the nmoles of EpB
in the feed stream. The latter was determined using the ther-
mal conductivity (TCD) detector to avoid the nonlinear behav-
ior that can occur when high concentrations are evaluated with
flame ionization detectors (FID). Pressure corrections were im-
plemented in the feed stream results to account for varying
pressure drops in the reactor.
2.2. Catalyst characterization
Hydrogen chemisorption was performed using a Quan-
tochrome Autosorb I to determine the concentration of exposed
Pt surface sites following Ag deposition. Prior to chemisorp-
tion, samples were reduced in flowing H2 at 200 ◦C for 4 h and
then evacuated at 200 ◦C for six hours.
Due to its structure and multiple bonding sites, EpB strongly
adsorbs to most metal surfaces [21], including Pt [11]. Conse-
quently, catalytic reactions involving EpB as either a reactant or
product tend to show deactivation for the first several hours on
line [22]. To eliminate the transient effects of deactivation, all
reaction data reported are based on catalyst activity after expo-
sure to reaction conditions for 15 h.
The catalysts were characterized using transmission Fourier
Transform Infrared Spectroscopy (FTIR) by pressing ∼0.015 g
of ground sample into pellets approximately 0.5 in. diame-
ter. Spectra were collected using a Thermo Electron model
4700 spectrometer with a liquid nitrogen-cooled MCT detec-
tor. All experiments were conducted in an externally-heated
cylindrical sample cell as described previously [10]. Briefly,
a total gas flow of ∼70 mL/min entered the cell in front of
the pellet and exited behind the pellet. The samples were re-
duced in H2 for 1 h at 200 ◦C, cooled to RT in He, exposed
to 1% CO in He, and then flushed with pure He to remove ph-
ysisorbed and gas phase CO. All spectra were referenced to
initial background spectra taken in He prior to CO exposure.
Finally, spectra were digitally processed to remove noise due
to atmospheric water, although the peak shapes were carefully
maintained.
2.4. Kinetic study
Detailed kinetic experiments for EpB hydrogenation were
conducted in a 0.25 in. (O.D.) stainless steel reactor to ensure
more isothermal reaction conditions. In order to maintain differ-
ential conversion levels, EpB conversions were typically kept
<10% by varying the total flowrate and by reducing the mass
of catalyst being evaluated. In these cases, silica diluent was
used to maintain adequate dimensions of the packed catalyst
bed. Finally, catalysts used in the reaction order studies were

134
M.T. Schaal et al. / Journal of Catalysis 254 (2008) 131–143
evaluated to ensure that both internal and external mass transfer
limitations were eliminated.
port. However, it has been shown that Ag(NH3)+2 species are
electrostatically-attracted to SiO2 and can be subsequently re-
duced to form Ag/SiO2 [28]. This surface charge-based interac-
tion is expected since the low point of zero charge (PZC) that is
generally observed for silica (between 1.5 and 3 [5]) is known to
cause deprotonation of surface hydroxyl groups. The resulting
negatively charged surface typically adsorbs positively charged
species from solution when the pH >7 [5,29]. In order to pre-
vent (or at least suppress) the attraction of Ag+ to the SiO2
support in basic solutions, the hydroxyl hydrogens on silica
were ion exchanged with Na+ before exposure to the ED bath.
Sodium ion exchange was achieved by soaking SiO2 or Pt/SiO2
in a large excess of 1 M NaNO3 for approximately 5 days and
then filtering and briefly washing with DI H2O. Sodium nitrate
was chosen as the exchange agent since it was already used to
maintain constant ionic strength in the ED bath. Further, the
presence of Na+ in the ED bath should maintain the Na+ con-
centration on the SiO2 surface during the ED process.
2.5. Propylene hydrogenation
Propylene hydrogenation experiments were conducted using
approximately 0.03 g of sieved (75–150 µm) catalyst mixed
with SiO2 diluent in a thermostatically-jacketed 0.25ꢀꢀ (O.D.)
stainless steel reactor. Reaction temperature was 25–26 ◦C and
feed composition was 5% propylene, 20% H2, balance He with
GHSV of ∼100,000 h−1 (100 sccm) for all catalysts. Feed and
product compositions were evaluated using an in-line TCD de-
tector.
3. Results and discussion
3.1. ED bath development
Aqueous ED baths are usually composed of a metal salt,
a reducing agent, and, optionally, a stabilizing agent or a com-
plexing agent. The complexing agent and stabilizing agent are
added to improve bath stability, which also results in a corre-
sponding decrease in the rate of deposition [8]. The baths in
this study were used for relatively short periods of time, and
were found to be stable in this time frame in the absence of
both complexing and stabilizing agents; thus, no stabilizing or
complexing agents were used. However, it should be noted that
if the concentrations of the reducing agent and the metal salt
are increased above those used in this study, the bath becomes
unstable even at short times. In those cases, the bath must be
stabilized using a stabilizing agent, a complexing agent, or pos-
sibly by maintaining the bath at a lower pH or a lower temper-
ature. Indeed, silver ED baths are inherently unstable [8], and
therefore the deposition solution must be carefully controlled to
prevent spontaneous metal reduction/precipitation.
To test this hypothesis, a silver ED bath was prepared and
exposed to a SiO2 sample. Fig. 1 shows that the Ag+ concentra-
tion drops to ∼0 ppm almost instantaneously when using SiO2
that was not ion exchanged with NaNO3 prior to bath exposure.
In contrast, the Ag+ concentration remained relatively constant
for the ion-exchanged SiO2 sample throughout the 120 min
deposition process (see Fig. 1 inset). However, when NaNO3-
exchanged Pt/SiO2 was added to an identical ED bath, the Ag+
concentration decreased in a controlled manner, indicating the
silver was deposited on the catalytically active sites (Pt0 or
Ag0), as observed by others [30,31]. It is worth noting that the
need for Na+ ion exchange may be eliminated by choosing a
Formaldehyde was chosen as the reducing agent since it is
relatively mild and provides ‘clean’ deposits since it contains no
heteroatoms [23]. Standard redox potentials reported by Barker
[24] indicate that the reduction potential of formaldehyde in-
creases with increasing pH [24,25]. Consequently, NaOH was
added to the ED bath to increase the rate of deposition. How-
ever, excessive alkaline conditions may lead to both silica de-
composition and the formation of insoluble silver salts such as
Ag2O (Ksp = 3.8 × 10−16) [26] or AgOH (Ksp = 2.0 × 10−8
)
[27]. Therefore, the pH was maintained at ∼9.0 to optimize the
HCHO reduction potential while maintaining both silica and
Ag+ stability. Continuous maintenance of pH (by addition of
NaOH solution) during the ED experiments was required since
the pH of the solution decreased with deposition time, consis-
tent with the observations of Chou et al. [25]. The decrease in
pH occurred even in the absence of Ag+ deposition, suggest-
ing that the reaction of HCHO with OH− to give the methylene
glycol anion (HOCH2O−) was occurring.
+
Fig. 1. The effect of Na exchange (using NaNO ) on SiO and Pt/SiO for
3
2
2
3.2. Verification of ED process
+
the room temperature deposition of Ag from the ED solution. ED bath: initial
+
[HCHO] = 464 µmol/L, initial [Ag ] = 232 µmol/L, [NaNO ] = 0.1 mol/L,
3
The formation of true bimetallic catalysts requires that ED
occurs only on the catalytic Pt sites and not on the SiO2 sup-
solution volume = 100 mL, catalyst mass = 0.50 g, and pH 9 held constant by
dropwise NaOH addition.

M.T. Schaal et al. / Journal of Catalysis 254 (2008) 131–143
135
Fig. 2. Hydrogen uptake for Pt/SiO samples with various Ag loadings: (A) results for Ag–Pt/SiO catalysts prepared by electroless deposition; (B) results for
2
2
Ag–Pt/SiO catalysts prepared by incipient wetness.
2
metal salt that is an anion in solution when the pH > PZC or a
cation when the pH < PZC [5].
dispersion) and the further reduction in H2 uptake after the ad-
dition of Ag (which does not chemisorb H2 at 40 ◦C) [32,33],
the data attained are near the lower limit of sensitivity for the
instrument, and thus contain a substantial degree of variabil-
ity. Fig. 2 shows the chemisorption data for the ED (A) and
IW-derived (B) samples. Even with the scatter in the data, it is
evident that in both cases Ag addition leads to a decrease in H2
uptake, indicating that Ag is being deposited on the Pt surface.
For a basis of comparison, 0.034 wt% Ag corresponds to the ad-
dition of one theoretical monolayer assuming monodisperse Ag
coverage. The observation that there is still an uptake of H2 af-
ter deposition of 0.23 wt% Ag (∼6.7 theoretical ML of Ag)
on both the ED and IW samples indicates that the entire Pt
surface is not covered by Ag. For the samples prepared by elec-
troless deposition this supports the existence of both catalytic
(Pt0-catalyzed) and autocatalytic (Ag0-catalyzed) processes in
which Ag is deposited on Pt and Ag, respectively. It is also
evident from Fig. 2 that the hydrogen uptake decreases more
steeply with Ag loading for the ED derived samples than for
the corresponding IW catalysts. This suggests that the former
approach provides a more targeted placement of Ag on the Pt
surface.
3.3. Galvanic displacement
The possibility of Ag+-induced oxidation of Pt and corre-
sponding reduction of Ag (i.e. galvanic displacement) was in-
vestigated by withdrawing periodic liquid samples from an ED
bath that did not contain a reducing agent. No Pt was detected
using AA analysis, indicating that if galvanic displacement does
occur, it is very slow. This absence of galvanic displacement
is expected since standard reduction potentials [24,26] indicate
that it is not thermodynamically favored for the ED bath com-
ponents used in this study.
3.4. Kinetics of electroless deposition
Ag deposition on Pt was found to be essentially first order
for the amount of catalyst added to the solution (supplementary
Fig. S1), fractional order for formaldehyde, and near zero order
for initial Ag+ concentration and OH− concentration (Figs. S2
and S3). Thus, the rate limiting step for Ag+ deposition onto
Pt involves the activation of HCHO (or HOCH2O−) on the
Pt/SiO2 catalyst; the near zero order dependency for Ag+ con-
centration indicates that the reduction step of Ag+ → Ag0 oc-
curs much faster than the activation of HCHO (or HOCH2O−).
These kinetic data provide proof that the electroless deposition
of Ag from solution is a kinetically controlled phenomenon and
is not simply the result of spontaneous solution decomposition.
3.6. FTIR data
Fig. 3 shows the FTIR spectra for monometallic Ag/SiO2,
Pt/SiO2, and 0.23 wt% Ag deposited by electroless deposi-
tion on Pt/SiO2. Three distinct peaks are evident in the Pt/SiO2
spectrum. Two of these peaks lie in the 2050–2100 cm−1 re-
gion and can be attributed to linear, or terminally, bonded
CO on Pt [34,35]. The dominant, linearly bonded CO peak
at ∼2090 cm−1 is assigned to CO on Pt(111), in good agree-
ment with the data of Cruz and Sheppard [36] and Greenler
and Bunch [37]. The distinct shoulder at ∼2052 cm−1 is some-
what unusual for CO–Pt/SiO2 spectra; however, Cruz et al. [36]
3.5. Chemisorption
Hydrogen chemisorption was conducted on a series of Ag–
Pt/SiO2 catalysts. Due to the low dispersion of Pt/SiO2 (3.7%

136
M.T. Schaal et al. / Journal of Catalysis 254 (2008) 131–143
The substantial contribution of the (111) plane is expected
for the large Pt crystallites present in the Pt/SiO2 catalysts in
this study. The statistical results of Van Hardeveld and Hartog
[41] for metals like Pt that crystallize in the FCC structure indi-
cate that for the two most stable structures existing in pseudo-
spherical geometries, the majority of the surface exists as (111)
sites. For Pt particles of ∼305 Å diameter (based on 3.7% dis-
persion) the Pt surface should contain approximately 78–87%
(111) sites.
A common absorbance scale was used in Fig. 3 in order
to more clearly illustrate the differences between the spectra.
From the data it is clear that CO does not adsorb on Ag at room
temperature, in agreement with observations of Rodriguez et al.
[42]. Furthermore, based on the observed decrease in IR ab-
sorbance for the Ag-modified Pt sample, it is apparent that Ag
addition results in a sharp decrease in CO adsorption on Pt due
to Ag coverage of the surface Pt sites. However, CO adsorption
is still observed even after several theoretical ML of Ag (assum-
ing monodisperse Ag coverage) have been deposited, indicating
that there must be three dimensional aggregates of Ag atoms on
the Pt surface. This is consistent with the earlier conclusion that
both catalytic and autocatalytic deposition of Ag occur during
electroless deposition. Further analysis of Fig. 3 also reveals
that the addition of Ag to Pt/SiO2 causes a significant modifica-
tion in the sites for CO adsorption. The bridging CO adsorption
peak is completely eliminated, and the peak that is assigned to
linearly bonded CO is shifted to even lower frequencies.
Figs. 4A and 4B show FTIR spectra for a more complete
series of ED and IW bimetallic catalysts, respectively, which
have been scaled to clarify the overall trends. The representa-
tive spectra shown here are based on several IR analyses for
each sample. Although the same general trends are seen in
both the ED and IW-derived catalysts, respectively, there are
some notable differences. In general, the ED spectra undergo a
smoother, more continuous transition in the 2000–2100 cm−1
region with increasing levels of Ag deposition. In addition,
at higher Ag weight loadings the ED samples show a greater
Fig. 3. Transmission FTIR spectra of CO adsorption on monometallic Pt/SiO ,
2
Ag/SiO and bimetallic Ag–Pt/SiO (6.7 theoretical ML).
2
2
observed a similar shoulder at ∼2053 cm−1 for low coverages
of CO on a Pt surface that was given a different pretreatment
procedure than those commonly used for in-situ FTIR stud-
ies. Other investigators have observed three distinct peaks in
the linear CO region and assigned the low frequency peak be-
tween 2045 and 2065 cm−1 [37,38] to linear CO adsorption on
stepped or kinked Pt sites, or isolated Pt atoms on low coordi-
nation sites in general [36–39].
The peak at ∼1845 cm−1 for Pt/SiO2 has been attributed
to 2-fold bridged CO on Pt(111) [36]. However, due to the
broad nature of the peak (1720–1900 cm−1) and the low acti-
vation barrier between two-fold and three-fold bridged species
on Pt(111) [40], there may also be contributions from the 3-fold
bridging species that has been observed at ∼1800 cm−1 [36].
Fig. 4. FTIR spectra for CO adsorption on Pt/SiO and a series of bimetallic Ag–Pt/SiO catalyst prepared using (A) electroless deposition and (B) incipient wetness.
2
2

M.T. Schaal et al. / Journal of Catalysis 254 (2008) 131–143
137
12
13
Fig. 5. Adsorption of different concentrations of CO and CO on Pt/SiO . Total CO concentration = 1.0 vol%: (A) FTIR absorbance spectra; (B) singleton
2
frequency plots for the deconvoluted spectra in (A).
downshift in frequency than the corresponding IW samples. For
example, the ∼0.46 wt% Ag–Pt/SiO2(IW) spectrum exhibits a
single peak at 2038 cm−1 while an ED sample with almost half
the Ag loading (∼0.23 wt% Ag) has a corresponding peak at
2035 cm−1. This suggests again that ED provides a more tar-
geted placement of Ag on Pt than IW methods. Nevertheless,
both families of ED and IW catalysts displayed a considerable
down-shift in linear CO–Pt frequency. This significant influence
of Ag on the frequency of CO–Pt absorbance has been observed
by others [30,43] who largely attributed this down shift to a re-
duction in dipole–dipole coupling of adsorbed CO molecules.
The high frequency linearly-bonded CO peak decreases
steeply with increasing silver loading. At low silver loadings
there is a decrease in both the Pt(111) peak and low coordi-
nation Pt (presumably corner/edge) peak intensities, indicating
that Ag deposition on both types of sites reduces carbonyl
dipole–dipole coupling, as reported by others [30,43]. How-
ever, the high weight loading Ag sample spectra only have one
peak, and it is below the 2063–2070 cm−1 singleton frequency
reported for CO linearly bonded to Pt(111) [44–47]. The low
frequencies of the high Ag loading samples therefore suggest
that the Pt(111) sites have been essentially covered by Ag.
Isotopic 13CO studies were conducted in order to better un-
derstand the influence of Ag addition for CO adsorption onto
Pt/SiO2. The singleton frequency for the various absorption
peaks, which is extracted from these isotopic studies, provides
information about the presence of electronic and/or dipole–
dipole shifts [35,36,46,48]. Spectra collected using various ra-
tios of 12CO and 13CO on Pt/SiO2 are shown in Fig. 5A. In
order to determine the singleton frequency, the spectra were
deconvoluted and the resulting peak positions were plotted as
wavenumber versus 12CO% and then extrapolated to 0% 12CO
as shown in Fig. 5B. The singleton frequency of the Pt(111)
sites was determined to be ∼2052 cm−1, which is slightly lower
than the 2063–2070 cm−1 values reported by others [44–47].
However, since the singleton frequency has been reported to
decrease with decreasing Pt particle size [35], some variations
between supported Pt particles and Pt(111) crystals should be
expected. The low frequency, linear shoulder gave a singleton
frequency of 2023 6 cm−1. Borovkov et al. observed a simi-
lar singleton frequency at ∼2038 cm−1 and assigned it to corner
and edge sites [49]. Additionally, Ménorval et al. reported sin-
gleton frequencies of 2032 and 2067 cm−1 for 99% dispersed
and 9% dispersed Pt/Al2O3 samples, respectively. As the dis-
persion increases, so does the number of step/corner sites, again
suggesting that the observed low frequency shoulder is due to
corner/edge sites.
Regardless of the specific peak assignments it is clear that
the 2035 cm−1 peak observed for the 0.23 wt% Ag–Pt/SiO2
(ED) sample cannot be attributed to reduced dipole coupling
interactions of the 2090 cm−1 CO–Pt/SiO2 peak. Again, this in-
dicates that at high Ag loadings, the Pt(111) sites are essentially
covered by Ag leaving only corner/edge sites available for CO
adsorption. However, it is also possible that Ag may not initially
be deposited on Pt(111) sites but instead may migrate to occupy
those sites at the elevated pretreatment temperatures. For exam-
ple, Röder et al. studied vapor deposited Ag on Pt(111) using
STM and observed that when submonolayer Ag coverages were
annealed at ∼350 ◦C, Ag–Pt mixing occurred at step edges and
Ag diffused into the Pt(111) terrace [50]. To test this hypoth-
esis, Ag placement was studied at different catalyst pretreat-
ment temperatures before FTIR experiments with adsorbed CO
were conducted. A ∼0.1 wt% Ag–Pt/SiO2 (prepared by ED)
was sequentially pretreated with H2 at RT, 100 ◦C, 150 ◦C, and
200 ◦C, and then flushed with flowing He prior to room temper-
ature exposure to 1% CO. The CO–Pt peak intensity increased
with increasing the reduction temperature (up to 150 ◦C). There
was also a slight increase in the frequency of the CO–Pt peak
at elevated reduction temperatures that may have been caused
by increased dipole–dipole coupling (resulting from more CO
adsorption on the more fully reduced Pt surface). Alternatively,
the increased frequency may have been caused by migration

138
M.T. Schaal et al. / Journal of Catalysis 254 (2008) 131–143
12
13
Fig. 6. Adsorption of varying ratios of CO and CO on 0.1 wt% Ag–Pt/SiO : (A) FTIR absorbance spectra; (B) singleton frequency plots for the deconvoluted
2
spectra in (A).
of Ag from Pt(111) to step/corner sites; in fact, higher tem-
peratures may result in Ag migration from Pt(111) sites to
Pt step/corner sites. Regardless, it appears that reduction up
to 200 ◦C does not lead to Ag migration from step/corner sites
to Pt(111) sites. At this point it is not clear why the Ag appears
to preferentially deposit on the Pt(111) sites while leaving the
low coordination Pt sites available for CO adsorption. However,
this unexpected preferential deposition onto (111) sites as op-
posed to corner/edge sites has been observed for the deposition
Pt onto Pd using direct redox methods [51], which is considered
a form of electroless deposition [8].
In order to better understand the influence of Ag on Pt, the
singleton frequency for Ag–Pt/SiO2 was also investigated us-
ing the 0.1 wt% Ag–Pt/SiO2 (ED) sample. The FTIR spectra
shown in Fig. 6A gave a singleton frequency of ∼2030 cm−1
(Fig. 6B). The singleton frequencies observed for Ag–Pt/SiO2
(2030 3 cm−1) versus that for corner/edge sites on Pt/SiO2
(2023 6 cm−1) suggest that there may be an electronic effect,
Fig. 7. Effect of Ag addition on EpB hydrogenation activity. Based on H
although the inherent error in the deconvolution of multiple,
overlapping peaks prevents a definitive conclusion. Future ex-
periments with C18O could significantly reduce/eliminate this
problem. Nevertheless, it appears that Ag essentially dilutes (by
coverage) the Pt surfaces of the supported catalysts and disrupts
carbonyl coupling, indicating intimate contact between the Ag
and Pt atoms. Furthermore, this effect is more pronounced for
the ED samples than the corresponding samples prepared by
IW methods.
2
chemisorption, one monolayer (assuming monodisperse coverage) of Ag on
Pt corresponds to ∼0.034 wt% Ag. All catalysts were evaluated using the same
catalyst weights and flowrates. Thus, EpB conversion values can be used to di-
rectly compare activities.
tivity for EpB isomerization due to acid-catalyzed conversion
to form 2-butenal (crotonaldehyde). Reported catalyst perfor-
mance was corrected for this background activity.
The conversion of EpB as a function of silver weight loading
is summarized in Fig. 7. The Pt/SiO2 catalyst shows a con-
version of ∼15%, while the low Ag weight loading bimetal-
lic catalysts derived by both ED and IW methodologies ex-
hibit maxima in EpB conversion at relatively low Ag loadings.
Specifically, maximum EpB conversions were observed at ap-
proximately 0.017 wt% for the ED sample and between 0.014%
and 0.035 wt% Ag for the IW catalysts. Hydrogenation activ-
ity declines at higher Ag loadings due to Ag coverage of ac-
3.7. EpB hydrogenation
Before evaluation of the Pt and Ag–Pt catalysts, the activities
of the SiO2 support and a 2 wt% Ag/SiO2 were evaluated. As
expected, they were found to be virtually inactive for EpB hy-
drogenation at the reaction conditions used in this study, consis-
tent with the results of others for Ag-catalyzed hydrogenation
studies [52]. The SiO2 support showed very low levels of ac-

M.T. Schaal et al. / Journal of Catalysis 254 (2008) 131–143
139
tive, surface Pt sites. Catalysts prepared by ED show a slightly
higher maximum catalytic activity with ∼45% conversion com-
pared to the ∼38% achieved for the IW catalysts. Further,
a comparison of the activity curves in Fig. 7 indicates that
preferential deposition of Ag on the Pt surface is improved in
ED-derived catalysts. For IW samples, the gradual response in
catalytic activity to higher Ag loadings suggests deposition is
occurring on both the SiO2 support and the Pt surface. Note that
the activities are compared as percent conversion, rather than
TOF values, which would exaggerate these differences even
more. The purpose is to direct attention to the overall higher
activities of optimum Ag–Pt compositions rather than activity
enhancements due to lower concentrations of active Pt sites.
The observed synergistic effect of Ag on Pt was not ex-
pected based on trends of Ag-containing hydrogenation cata-
lysts [52–54]. Bimetallic Ag–Pt catalysts prepared by sequen-
tial organometallic deposition on xerogel SiO2 showed a sharp
decrease in 1-hexene hydrogenation activity compared to the
monometallic Pt catalysts [52]. Further, Ag–Pt/NaY zeolite cat-
alysts [54] prepared by ion exchange methods gave significantly
lower TOF values (based on Pt surface sites) for ethane hy-
drogenolysis as silver loadings increased. However, the addi-
tion of Ag to Pt/SiO2 has been reported to increase product
selectivity and catalyst lifetime for the dehydrogenation of do-
decane to dodecene, relative to unmodified Pt/SiO2 [55]. Addi-
tionally, Ag–Pt/C catalysts have been evaluated for the oxygen
reduction reaction (ORR) in PEM and AFC based fuel cells;
while these bimetallic catalysts were not as active as unmod-
ified Pt/C [56], they exhibited longer lifetimes [57]. Thus, in
some cases Ag addition was found to improve catalyst stabil-
ity and modify product selectivity; however, before the results
described in this study, the catalytic hydrogenation activity of
unmodified Pt catalysts reportedly decreased upon Ag addition.
Since addition of Ag to Pt/C has been found to increase cat-
alyst stability and reduce deactivation [57], the effect of Ag
on catalytic deactivation was evaluated. The results, which are
summarized in Fig. 8, show that addition of Ag lowered cata-
lyst deactivation for the ED series of catalysts when compared
to their monometallic Pt/SiO2 counterpart. However, the de-
creased deactivation does not appear to be solely responsible for
the activity enhancement. Fig. 9 shows the initial EpB hydro-
genation conversion, which was determined by extrapolating
to zero time on-line. Even at initial conditions, the conversion
is ∼1.5 times greater for the ∼0.017 wt% Ag–Pt/SiO2 sam-
ple than for the treated Pt/SiO2, suggesting the reduced rate of
deactivation is not solely responsible for the activity enhance-
ment.
Fig. 8. Effect of Ag addition on the normalized deactivation for ED derived
Ag–Pt/SiO catalysts.
2
Fig. 9. Initial EpB hydrogenation conversion was determined by extrapolating
the conversion to zero time on-line. Again, all catalysts were evaluated using
the same catalyst weights and flowrates. Thus, EpB conversion values can be
used to directly compare activities.
genation with Ag addition appears to be related to the additional
functional groups present in EpB.
Since the enhanced hydrogenation activity was observed
for EpB but not for simpler olefins, EpB hydrogenation reac-
tion kinetics were measured for Pt/SiO2 and 0.014 wt% Ag–
Pt/SiO2. As shown in Fig. 10A, the reaction orders in hydro-
gen were found to be very similar for both the monometal-
lic and bimetallic catalysts. The positive dependence in hy-
drogen concentration is expected since the strong adsorption
of EpB limits the concentration of surface hydrogen. Indeed,
from Fig. 10B it is apparent that the EpB reaction is essen-
tially zero order (0.06 0.09) over Pt/SiO2, consistent with
a strong EpB–Pt interaction. However, the EpB reaction or-
To better understand the nature of the enhancement in Pt
hydrogenation activity by Ag, the ED-derived Ag–Pt catalysts
were evaluated for propylene hydrogenation. The addition of
Ag to Pt/SiO2 decreased the activity for propylene hydro-
genation; the decreased activity was proportional to Ag cov-
erage on the active Pt surface (Fig. S4). The conversion de-
creased from 17% over Pt/SiO2 to ∼6% with the addition of
∼0.03 wt% Ag, consistent with the results of Campostrini et al.
for organometallic-derived Pt-Ag bimetallic catalysts supported
on silica [53]. Thus, the observed enhancement for EpB hydro-

140
M.T. Schaal et al. / Journal of Catalysis 254 (2008) 131–143
Fig. 10. Kinetic dependency of (A) H and (B) EpB on the rate of EpB hydrogenation for Pt/SiO and 0.014 wt% Ag–Pt/SiO catalysts.
2
2
2
der increased to approximately 0.5 order (0.47 0.05) for the
Ag–Pt/SiO2 sample (Fig. 10B), indicating a decrease in EpB
surface coverage presumably due to a decrease in the adsorp-
tion energy of EpB on Pt. Interestingly, the Arrhenius plots
(Fig. S5) show that the addition of Ag does not modify the ac-
tivation energy for the reaction (Pt/SiO2 = 7.4 0.5 kcal/mol;
Ag–Pt/SiO2 = 7.8 0.7 kcal/mol), indicating that Ag does not
change the rate limiting step for this reaction. The similar en-
ergetics also suggest that the electronic interaction between Ag
and Pt must be relatively small.
Bartok et al. have reported that EpB may interact with Pt at
both the epoxide oxygen and the C=C bond [11]. This multi-
site interaction causes EpB to be strongly bound to Pt [20]. One
possible explanation for the enhanced hydrogenation activity
supported by both the kinetics and the deactivation experiments
is a Ag-induced reduction in EpB adsorption strength on Pt
sites. If adsorption of EpB can be restricted to either the C=C
bond or the epoxy group, the lower strength of EpB adsorp-
tion should result in higher activity. Such a modification in the
EpB–Pt interaction could be caused by electronic, ensemble, or
bifunctional effects. The deposition of small amounts of Ag on
the Pt surface (primarily on the (111) facets as shown by FTIR
results) could dilute the active sites into Pt ensembles that hin-
der the simultaneous adsorption of EpB at both the C=C and
epoxy moieties. Alternatively, there may be an electronic ef-
fect with electron transfer between Ag and Pt. Although there
is some disagreement in the literature regarding the direction
of d-orbital electron transfer [56,58], Lima et al. [56] found
that for Ag–Pt catalysts supported on carbon, there was elec-
tron transfer from Ag to the Pt 5d orbitals. The increase in Pt
electron density of the bonding 5d orbitals should lower the
electrophilicity of Pt and reduce the interaction between the Pt
and the electron-rich oxygen of the epoxide ring. Finally, theo-
retical and surface science studies [20] suggest that the epoxide
end of EpB preferentially interacts with Ag instead of Pt to form
an oxametallacycle intermediate. The ring-opened oxametalla-
cycle is reversible on the Ag sites and may reform the epoxy
Fig. 11. Effect of Ag loading on the selectivity. The legend for compound iden-
tification is as follows: BO: butylene oxide; C=C unsaturated: crotonaldehyde,
crotyl alcohol, 3-buten-1-ol; C unsat HC: butene, butadiene; C sat HC: bu-
4
4
tane.
moiety upon desorption [59]. Thus, the reduced Pt–EpB in-
teraction could result from a bifunctional effect in which Ag
protects the epoxide ring from irreversible hydrogenolysis on
the Pt sites (e.g., aldehyde formation) and subsequent hydro-
genation.
If modifications in the mode of EpB adsorption can be at-
tributed to Ag, then changes in selectivity may also be observed.
The data in Fig. 11 show the effect of electrolessly-deposited
Ag on product selectivity. Monometallic Pt/SiO2 shows mod-
erate (∼36%) selectivity to butylene oxide (BO), consistent
with the results of Bartok [11] for the Pt-catalyzed hydrogena-
tion of EpB. However, upon addition of very low amounts of
Ag, the BO selectivity increases to ∼55%, while the selectiv-
ity to ring-opened C=C unsaturated products (cis- and trans-
crotonaldehyde, 3-buten-1-ol, and crotyl alcohol) decreases.

M.T. Schaal et al. / Journal of Catalysis 254 (2008) 131–143
141
Fig. 12. Rate of product formation as a function of silver loading for (A), ED-derived catalysts and (B), IW derived catalysts. Fig. 11 caption shows product
abbreviations.
These results suggest that even small amounts of Ag modify
the Pt(111) sites (based on FTIR data) such that hydrogena-
tion of the C=C bond is promoted. As stated earlier, this may
result from ensemble effects wherein the number of contiguous
Pt(111) sites becomes too small to accommodate both C=C ad-
sorption and C–O hydrogenolysis. Alternatively, there maybe a
bifunctional effect, in which Ag essentially protects the epoxide
ring by reversible adsorption. Finally, the dramatic increase of
C4 hydrocarbons for the highest Ag-modified Pt catalysts also
suggests that the more reactive corner and edge sites are active
for deoxygenation of EpB or EpB-derived intermediates.
The data in Fig. 12 summarize the effects of Ag addition
on the rates of product formation for both for ED and IW cata-
lysts. The same general trends are observed for catalysts derived
using both methodologies, although the ED catalysts undergo
more dramatic changes, even at lower levels of Ag addition. Ini-
tially, the rates of product formation for all products increase,
though the rate enhancement is most dramatic for BO. It is also
apparent from Fig. 12 that when Ag loading exceeds ∼0.015–
0.03 wt%, the overall rates of product formation decrease due
to the expected Ag ‘poisoning’ of the active Pt sites. However,
the rates of formation of unsaturated, oxygenated species are
much less affected than formation of the saturated product BO.
This is more clearly seen in Fig. 11 which shows that selectivity
for these species increases at higher Ag loadings.
especially for C–C bond cleavage, when compared to highly
coordinated Pt atoms in (111) planes. Theoretical calculations
have also demonstrated that step sites are more active than ter-
race sites for N2, CO, NO, and O2 cleavage [62]. Therefore, the
observed increase in hydrogenolysis products at Ag loadings
ꢀ0.1 wt% Ag is consistent with the FTIR-based conclusion that
only low coordination Pt sites are available.
As seen in the EpB hydrogenation reaction scheme (Sche-
me 1), n-butanol and n-butyraldehyde are the end products re-
sulting from the hydrogenation or isomerization of hydrogenol-
ysis reaction products (crotonaldehyde, 3-buten-1-ol, crotyl al-
cohol). Despite the observed increase in hydrogenolysis at high
Ag loadings, the rates of n-butanol and n-butyraldehyde for-
mation (Fig. 12) are significantly reduced at ꢀ0.1 wt% Ag
loading. The concurrent increase in hydrogenolysis reactions
with lower rates of C=C hydrogenation reactions suggests that
these reactions occur more readily on Pt(111) sites, while cor-
ner/edge sites favor the formation of hydrogenolysis products.
The overall reduction in C=C hydrogenation capabilities for
samples having >0.1 wt% Ag loading is consistent with the dis-
appearance of a distinctive bridged CO FTIR peak on Pt(111)
crystalline planes during CO adsorption.
4. Conclusions
The increase in hydrogenolysis products at high Ag loadings
is somewhat unexpected since hydrogenolysis reactions typi-
cally require relatively large ensembles [60]. In fact, the antic-
ipated decrease in ethane hydrogenolysis activity has been ob-
served for Ag-modified, Pt/NaY catalysts [54]. However, FTIR
analysis of these ED/IW samples indicate that Ag is preferen-
tially located on the Pt(111) planes; thus, at high Ag loadings,
only low coordination Pt sites, such as corners and edges, are
available for reaction. Interestingly, Somorjai and Blakely [61]
reported that low coordination Pt sites, such as those present
at corners, edges, and steps, were more active in general, but
The electroless deposition of Ag onto Pt is a kinetically-
controlled (not spontaneous) process that is catalyzed by Pt, and
therefore does not readily take place on the silica support. The
amount of Pt/SiO2 catalyst and the concentration of formalde-
hyde in the bath were found to be more kinetically important
than the concentration of Ag+ and OH− for electroless deposi-
tion, indicating that formaldehyde activation is the rate limiting
step. Chemisorption and FTIR studies confirm the deposition
of Ag onto Pt and indicate that the autocatalytic electroless de-
position process (Ag+ deposition onto Ag0) occurs at higher
Ag coverages. Catalyst characterization results suggest that ED

142
M.T. Schaal et al. / Journal of Catalysis 254 (2008) 131–143
provides a more controlled and more intimate contact between
the Ag and Pt than IW methods. Furthermore, infrared studies
show that the deposited Ag is preferentially located on Pt(111)
sites, leaving corner and edge sites available for CO adsorption.
Bimetallic Ag–Pt catalysts prepared by both electroless de-
position and incipient wetness methods gave the same general
trends in catalytic activity and selectivity for the hydrogenation
of 3,4-epoxy-1-butene (EpB). However, the catalysts prepared
by ED methods were more responsive to changes of Ag load-
ing, indicating that ED provided a targeted placement of Ag on
the Pt surface.
[3] W. Sachtler, Appl. Surf. Sci. 19 (1984) 167.
[4] J.T. Wrobleski, M. Boudart, Catal. Today 15 (1992) 349.
[5] G. Ertl, H. Knozinger, J. Weitkamp (Eds.), Handbook of Heterogeneous
Catalysis, VCH, Weinheim, 1997, p. 13, 35–38, 191–195, 257–261, 365–
367.
[6] L. Guczi, Stud. Surf. Sci. Catal. 29 (1986) 547.
[7] B. Chandler, L. Pignolet, Catal. Today 65 (2001) 39.
[8] S. Djokic, Mod. Aspect Electroc. 35 (2002) 51.
[9] I. Ohno, Mater. Sci. Eng. A 146 (1991) 33.
[10] M.T. Schaal, A.Y. Metcalf, J.H. Montoya, J.P. Wilkinson, C.C. Stork, C.T.
Williams, J.R. Monnier, Catal. Today 123 (2007) 142.
[11] M. Bartok, A. Fasi, F. Notheisz, J. Catal. 175 (1998) 40.
[12] R. Disselkkamp, K.M. Judd, T.R. Hart, C.H.F. Peden, G.J. Posakony, L.J.
Bond, J. Catal. 221 (2004) 347.
EpB conversion over Pt/SiO2 was greatly enhanced (∼3×)
by the addition of Ag, which is not active for EpB hydrogena-
tion. The increased activity is clearly linked to the modification
of the Pt surface by the addition of small amounts of Ag. This
modification appears to involve the Ag-induced reduction of
Pt–EpB interaction due to ensemble and/or bifunctional effects
since any electronic effects present in this system appear to be
relatively small. Product distribution was also modified with
Ag addition. The results indicate that BO formation is favored
on the Ag-modified Pt(111) sites. In general, hydrogenolysis
product formation appears to be favored on low coordination Pt
sites, such as corners and edges.
[13] M. Englisch, V. Ranade, J. Lercher, Appl. Catal. A Gen. 163 (1997) 111.
[14] H. Fujitsu, E. Matsumura, S. Shirahama, K. Takeshita, I. Mochida,
J. Chem. Soc. Perkin Trans. 1 855 (1982) 855.
[15] J. Simonik, L. Beranek, J. Catal. 24 (1972) 348.
[16] P. Rylander, Catalytic Hydrogenation over Platinum Metals, Academic
Press, New York, 1967, pp. 107, 250–252, 476–483.
[17] J. Monnier, B. Roberts, D. Hitch, US Patent 6,180 559 (2001), to Eastman
Chemical Company.
[18] W. Brown, C. Foote, Organic Chemistry, Harcourt College Publishers,
New York, 2002, pp. 257, 284–285, 410.
[19] M. Bartok, A. Fasi, F. Notheisz, J. Mol. Catal. A Chem. 135 (1998) 307.
[20] A. Loh, W. Medlin, J. Am. Chem. Soc (2007), submitted for publication.
[21] J.W. Medlin, Doctoral Thesis, U. Del. Dept. Chem. Engr., Newark, 2001,
pp. 80–117.
Overall, it has been demonstrated that ED is a feasi-
ble method for the preparation of bimetallic catalysts. This
methodology promotes the controlled deposition of even small
amounts of a second metal on the surface of a pre-existing
metal to give a more targeted placement of the metal than that
obtained using traditional methodologies. Further, EpB hydro-
genation is an informative probe reaction since it contains two
distinctly different functional groups, a C=C bond and a reac-
tive epoxy group.
[22] J.R. Monnier, Stud. Surf. Sci. Catal. 110 (1997) 135.
[23] C. Rao, D. Trivedi, Coordin. Chem. Rev. 249 (2005) 613.
[24] B.D. Barker, Surf. Technol. 12 (1981) 77.
[25] K. Chou, Y. Lu, H. Lee, Mater. Chem. Phys. 94 (2005) 429.
[26] D. Harris, Quantitative Chemical Analysis, W.H. Freeman and Co., New
York, 1999, pp. 337–363, AP13, AP34-35.
[27] J.A. Dean (Ed.), Lange’s Handbook of Chemistry, McGraw–Hill, Inc.,
New York, 1992, pp. 8.10.
[28] K.P. de Jong, J.W. Geus, Appl. Catal. 4 (1982) 41.
[29] J. Park, J.R. Rebalbuto, J. Colloid Interface Sci. 175 (1995) 239.
[30] K.P. de Jong, B.E. Bongenaar-Schlenter, G.R. Meima, R.C. Verkerk,
M.J.J. Lammers, J.W. Geus, J. Catal. 81 (1983) 67.
[31] K.P. de Jong, J.W. Geus, Stud. Surf. Sci. Catal. 16 (1983) 111.
[32] C. Mijoule, V. Russier, Surf. Sci. 254 (1991) 329.
[33] S. Yuvaraj, S.-C. Chow, C.-T. Yeh, J. Catal. 198 (2001) 187.
[34] J. Freysz, J. Saussey, J. Lavalley, P. Bourges, J. Catal. 197 (2001) 131.
[35] P. Fanson, N. Delgass, J. Lauterbach, J. Catal. 204 (2001) 35.
[36] C. Cruz, N. Sheppard, Spectrochim. Acta 50A (1994) 271.
[37] R. Greenler, K. Burch, Surf. Sci. 152/153 (1985) 338.
[38] J. Rasko, J. Catal. 217 (2003) 478.
Acknowledgments
This research project was supported financially by the Uni-
versity of South Carolina Nanocenter, NSF Grant No. 0456899,
and by an NSF Graduate Research Fellowship for MTS. The
authors also gratefully acknowledge BASF Catalysts LLC for
supplying the Pt/SiO2 base catalyst and Eastman Chemical
Company for supplying the EpB used in this study. Further,
the authors would like to thank Dr. Will Medlin for his in-
sightful discussions and Ms. Carol Stork for valuable analytical
assistance. Finally, JRM would like to acknowledge Dr. Joseph
Yudelson, his former laboratory head at Kodak Research Labo-
ratories, for introducing him to the concept of electroless depo-
sition.
[39] R. Brandt, M. Hughes, L. Bourget, K. Truszkowska, R. Greenler, Surf.
Sci. 286 (1993) 15.
[40] B. Hayden, A. Bradshaw, Surf. Sci. 125 (1983) 787.
[41] R. Van Hardeveld, F. Hartog, Surf. Sci. 15 (1969) 189.
[42] J. Rodriguez, C. Truong, D. Goodman, Surf. Sci. Lett. 271 (1992) L331.
[43] F. Gauthard, F. Epron, J. Barbier, J. Catal. 220 (2003) 182.
[44] A. Crossley, D.A. King, Surf. Sci. 95 (1980) 131.
[45] R.A. Shigeishi, D.A. King, Surf. Sci. Lett. 75 (1978) L397.
[46] A. Crossley, D.A. King, Surf. Sci. 68 (1977) 528.
[47] C.W. Olsen, R.I. Masel, Surf. Sci. 201 (1988) 444.
[48] C. Mihut, Synthesis and Characterization of Bimetallic Platinum–Gold
Catalysts, Doctoral Thesis, USC Dept. Chem. Engr., Columbia, 2002,
pp. v–vii, 67–70, 123–124, 152–153, 174–179.
Supplementary material
The online version of this article contains additional supple-
mentary material.
[49] V. Borovkov, S. Kolesnikov, V. Kovalchuk, J. d’Itri, J. Phys. Chem. B 109
(2005) 19772.
Please visit DOI: 10.1016/j.jcat.2007.12.006.
[50] H. Roder, R. Schuster, H. Brune, K. Kern, Phys. Rev. Lett. 71 (1993) 2086.
[51] C. Micheaud, M. Guerin, P. Marecot, C. Geron, J. Barbier, J. Chim.
Phys. 93 (1996) 1394.
References
[1] M. Kuhn, J. Rodriguez, J. Catal. 154 (1995) 355.
[2] T. Miyake, A. Tetsuo, Appl. Catal. A Gen. 280 (2005) 47.
[52] R. Campostrini, G. Carturan, S. Dirè, P. Scardi, J. Mol. Catal. 53 (1989)
L13.

M.T. Schaal et al. / Journal of Catalysis 254 (2008) 131–143
143
[53] R. Campostrini, G. Carturan, J. Mol. Catal. 78 (1993) 169.
[54] R. Ryoo, C. Pak, S. Cho, Jpn. J. Appl. Phys. 32 (1992) 475.
[55] J. Völter, Stud. Surf. Sci. Catal. 27 (1986) 337.
[59] J.W. Medlin, M.A. Barteau, J.M. Vohs, J. Mol. Catal. A Chem. 163 (2000)
129.
[60] J.H. Sinfelt, Acc. Chem. Res. 10 (1977) 15.
[56] F. Lima, C. Sanches, E. Ticianelli, J. Electrochem. Soc. 152 (2005) A1466.
[57] M. Chatenet, M. Aurousseau, R. Durand, F. Andolfatto, J. Electrochem.
Soc. 150 (2003) D47.
[61] G.A. Somorjai, D.W. Blakely, Nature 258 (1975) 580.
[62] J.K. Nørskov, T. Bligaard, A. Logadottir, S. Bahn, L.B. Hansen, M. Bol-
linger, H. Bengaard, B. Hammer, Z. Sljivancanin, M. Mavrikakis, Y. Xu,
S. Dahl, C.J.H. Jacobsen, J. Catal. 209 (2002) 275.
[58] J. Rodriguez, M. Kuhn, J. Phys. Chem. 98 (1994) 11251.