Surface-Bound Ligands Modulate Chemoselectivity and Activity of a Bimetallic Nanoparticle Catalyst

Article abstract of DOI:10.1021/acscatal.5b00262

"Naked" metal nanoparticles (NPs) are thermodynamically and kinetically unstable in solution. Ligands, surfactants, or polymers, which adsorb at a particle's surface, can be used to stabilize NPs; however, such a mode of stabilization is undesirable for catalytic applications because the adsorbates block the surface active sites. The catalytic activity and the stability of NPs are usually inversely correlated. Here, we describe an example of a bimetallic (PtFe) NP catalyst stabilized by carboxylate surface ligands that bind preferentially to one of the metals (Fe). NPs stabilized by fluorous ligands were found to be remarkably competent in catalyzing the hydrogenation of cinnamaldehyde; NPs stabilized by hydrocarbon ligands were significantly less active. The chain length of the fluorous ligands played a key role in determining the chemoselectivity of the FePt NP catalysts. (Chemical Presented).

Full text of DOI:10.1021/acscatal.5b00262

Letter
pubs.acs.org/acscatalysis
Surface-Bound Ligands Modulate Chemoselectivity and Activity of a
Bimetallic Nanoparticle Catalyst
Khanh B. Vu,†,‡ Konstantin V. Bukhryakov,†,‡ Dalaver H. Anjum,§ and Valentin O. Rodionov*,†
§
†Division of Physical Sciences and Engineering and KAUST Catalysis Center, KAUST Advanced Nanofabrication Imaging and
Characterization Core Laboratory King Abdullah University of Science and Technology (KAUST), Thuwal 23955-6900, Kingdom of
Saudi Arabia
S
* Supporting Information
ABSTRACT: “Naked” metal nanoparticles (NPs) are thermodynami-
cally and kinetically unstable in solution. Ligands, surfactants, or
polymers, which adsorb at a particle’s surface, can be used to stabilize
NPs; however, such a mode of stabilization is undesirable for catalytic
applications because the adsorbates block the surface active sites. The
catalytic activity and the stability of NPs are usually inversely correlated.
Here, we describe an example of a bimetallic (PtFe) NP catalyst stabilized
by carboxylate surface ligands that bind preferentially to one of the metals
(Fe). NPs stabilized by fluorous ligands were found to be remarkably
competent in catalyzing the hydrogenation of cinnamaldehyde; NPs stabilized by hydrocarbon ligands were significantly less
active. The chain length of the fluorous ligands played a key role in determining the chemoselectivity of the FePt NP catalysts.
KEYWORDS: cinnamaldehyde, selective hydrogenation, ligand effects, bimetallic nanoparticles, fluorous chemistry
ransition metal nanoparticles (NPs) and nanoclusters are
1
Ta privileged class of metal colloids. Their high surface
areas and the strong correspondence between particle sizes/
shapes and surface chemistries make these species uniquely
suited for applications in catalysis.2−5 Small, catalytically active
NPs are also some of the more unstable colloids because of
their tendency to aggregate. To address this problem, NP
catalysts are typically arrayed on solid supports,1,6,7 which may
also act as synergistic cocatalysts.3,5,8
NPs and metal clusters can also be stabilized and solubilized
with the aid of surface adsorbates, which range from simple
surfactants and amphiphilic copolymers to strongly binding/
soft ligands, such as thiols.9 Naturally, the adsorbate-stabilized
NPs lose some or all of their catalytic prowess.4,9 The metal
surface accessibility is limited compared with “naked” NPs, and
the catalytic sites are easily poisoned by thiols and amines.
Before such NPs can be used as catalysts, a variety of harsh
surface treatments are usually applied to remove the
adsorbates.9,10
Here, we present an approach to dispersible, catalytically
active, adsorbate-stabilized NPs. Our design is based on
bimetallic NPs7,11 consisting of metals that feature significantly
Figure 1. Selectively protected bimetallic NPs. (A) FePt NPs with
different affinities toward a specific class of adsorbates. Thus,
ligand-free Pt sites. (B) TEM image of C10F−Fe0.25Pt0.75 NPs. (C)
Carboxylic acid ligands used in this study.
with judicious choice of metal composition and surface ligand
chemistry, the particle aggregation could be prevented while the
catalytic activity of the unprotected “islands” of the non-
adsorbing metal could be preserved (Figure 1A). This approach
was applied to FePt NPs12 stabilized with perfluorinated
carboxylic acids, which are expected to bind to Fe almost
exclusively.13 We were able to preserve the catalytic activity of
Pt while gaining the capability for fluorous-biphasic recycling of
Received: February 7, 2015
Revised: March 16, 2015
© XXXX American Chemical Society
2529
DOI: 10.1021/acscatal.5b00262
ACS Catal. 2015, 5, 2529−2533

ACS Catalysis
Letter
a
Table 1. Dependence of the Chemoselectivity and Activity of the FePt NP Catalysts on the Nature of Adsorbates
b
b
b
b
c
c
entry
catalyst
OA−Pt
time, min
conversion, %
2, %
3, %
4, %
TON
TOF, h−1
1
2
60
180
180
180
90
14.9
15.0
17.8
25.1
12.0
22.8
43.1
44.1
71.1
72.5
16.2
47.8
52.0
24.8
72.1
85.1
91.8
89.7
94.0
64.1
55.8
32.7
31.3
40.6
18.6
9.9
28.0
19.5
16.8
34.7
9.3
18
18
22
31
15
28
53
54
87
89
18
6
C5H−Fe0.25Pt0.75
C10H−Fe0.25Pt0.75
C18H−Fe0.25Pt0.75
C5F−Fe0.25Pt0.75
C8F−Fe0.25Pt0.75
C10F−Fe0.25Pt0.75
C18F−Fe0.25Pt0.75
C10F−Fe0.33Pt0.67
C10F−Fe0.13Pt0.87
3
7
4
10
10
19
35
36
87
89
5
6
90
5.0
7
90
4.4
3.8
8
90
5.2
5.1
9
60
1.5
4.5
10
60
11.8
24.1
a
Conditions: 0.1 mL of perfluorodecalin, 0.5 mL of n-hexane, H2 (1 atm), 50 °C, cinnamaldehyde (0.16 mmol), catalyst (1.3 × 10−3 mmol Pt, 0.81
mol %). Determined by NMR and GC/MS. Turnover number (TON) and turnover frequency (TOF) calculated with respect to the overall Pt
b
c
loading.30
the NP catalyst.14 Serendipitously, we discovered that the
perfluorinated adsorbates exert a significant influence over both
the activity and chemoselectivity of the NP catalyst.
of the fct structure, indicated that Fe0.13Pt0.87, Fe0.25Pt0.75, and
Fe0.33Pt0.77 feature disordered fcc structures with random
distribution of Fe and Pt in the unit cell. This is unsurprising
because the fcc-to-fct transformation in FePt NPs typically
requires high temperature annealing.18
The size of OA-protected FePt NPs was in the range of 3.2−
3.6 nm (Figure 1B, and SI Figures S6, S8, and S10), largely
independent of the Fe/Pt ratio. Pure Pt nanoparticles prepared
under the same conditions were ill-defined and aggregated
(Figure S3). This confirms that OA has a poor affinity for the
surface of Pt and is of limited use as a capping adsorbate in this
case. Much larger FePt nanocrystals were obtained in the
absence of OA, confirming the crucial role of the surface-bound
OA in the stabilization of the FePt bimetallic particles during
synthesis.19
The OA adsorbate could be conveniently exchanged for a
range of other carboxylates, either perfluorinated or hydro-
carbon. For the fluorous carboxylic acids, the exchange was
performed in a mixture of dichloromethane and acetone. After
incubation with the fluorous ligands, the NPs were extracted
into perfluorodecalin (PFD). The material was purified from
excess carboxylic acid by centrifugation and redispersed in PFD.
The exchange procedure for the hydrocarbon carboxylic acids
was more involved because the NPs could not be extracted into
a fluorous solvent. Several centrifugation−redispersion steps
were necessary to purify the catalyst from the excess adsorbate.
Ligand exchange was not possible for OA−Pt NPs. As could be
expected, these particles could not be redispersed in fluorous
solvents after treatment with perfluorodecanoic acid. No
fluorous acid could be detected on their surfaces by FTIR
(Figure S13 B, SI). For convenience, the fluorous and
hydrocarbon carboxylate ligands will be referred to as CXF
and CXH respectively, where X is the length of the main carbon
chain (Figure 1C). The corresponding NPs will be designated
CXF−FexPty and CXH−FexPty.
The FePt NPs used in this study were synthesized by the
simultaneous decomposition of Fe(CO)5 and reduction of
Pt(acac)2 in dioctyl ether by 1,2-hexadecanediol. In a departure
from the previous reports of this procedure,13,15 we used only
oleic acid (OA) stabilizer, foregoing the use of oleylamine as
the Pt capping ligand. We reasoned that in the absence of an
amine, Pt “islands” will remain ligand-free. The NPs thus
prepared were purified by centrifugation and redispersed in n-
hexane. Although more prone to aggregation than the NPs
stabilized by both amine and OA ligands, the OA-stabilized
NPs were reasonably stable. Significant precipitation from
hexane was typically observed only after several days to weeks,
depending on the fraction of Fe, and we took care to perform
all subsequent surface treatment or catalysis experiments with
freshly prepared materials. To explore the mode of interaction
of OA with the surfaces of pure Pt and FePt NPs, we obtained
FTIR spectra of representative samples (Figure S13 A, SI). For
C10F−Fe0.25Pt0.75 NPs, absorption bands at 2916, 2848, 1525,
and 1399 cm−1 were observed. These bands are characteristic of
OA adsorbed on a metal surface (Table S6, SI).16 No such
bands were observed for pure Pt NPs, indicating that the
interaction of OA and Pt is weak, and any physisorbed OA is
removed from the NP surfaces through simple washing/
reprecipitation.
The Fe/Pt incorporation ratio could be tuned by changing
the Fe(CO)5/Pt(acac)2 feed ratio. Because of the inevitable loss
of the volatile carbonyl, the Fe/Pt ratio was determined by
inductively coupled plasma optical emission spectroscopy. For
convenience, the FePt NP materials will be designated FexPty,
where x and y are the molar fractions of the respective metals.
Particles ranging from pure Pt to Fe0.33Pt0.67 have been
prepared. To ascertain the nature of the metal packing in the
NPs, they were characterized by powder X-ray diffraction
(PXRD) (Figures S14 and S15, SI). The lattice constants
obtained for the FePt materials through Rietveld refinement
were smaller than that of Pt. This conclusion was further
supported by selected area electron diffraction data (Figure S16
and Table S7, SI), suggesting a Pt-like alloy structure with
randomly inserted Fe atoms.17 The absence of peaks at 24° and
33°, which are characteristic of (001) and (110) lattice planes
The ligand exchange had no effect on the morphology of the
FePt NPs, as observed by TEM (Figures S5 and S6, SI).
However, both solubility and dispersion stability of the
materials were significantly modified (Figure S1, SI). The
effect was strongly dependent on both the nature of the
carboxylic acid and the content of Fe in the NPs. Ligands with
shorter carbon chains, such as C5F and C5H, predictably led to
materials with poor colloidal stability. NPs stabilized with
2530
DOI: 10.1021/acscatal.5b00262
ACS Catal. 2015, 5, 2529−2533

ACS Catalysis
Letter
Figure 2. Product selectivity as a function of cinnamaldehyde conversion: (A) OA−Pt, (B) C10F−Fe0.13Pt0.87, and (C) C10F−Fe0.33Pt0.67
.
Conditions: 0.1 mL of perfluorodecalin; 0.5 mL of n-hexane; H2 (1 atm); 50 °C; cinnamaldehyde (0.16 mmol); Pt loading (mol %), 19.2 (OA−Pt),
2.3 (C10F−Fe0.13Pt0.87), 0.62 (C10F−Fe0.33Pt0.67). The solid traces are provided for visual guidance only.
longer-chain adsorbates and Fe fractions over 0.25 were
generally at least as stable as the parent OA-stabilized NPs.
The attained stability was especially impressive for C10F and
C18F ligands. The C10F−Fe0.25Pt0.75 NPs were stable in PFD
solution for at least 2 weeks (Figure S2, SI); however, even the
long-chain fluorous C10F was not sufficient to stabilize the low-
Fe Fe0.13Pt0.87: although the NPs could be dispersed in PFD,
significant precipitation was observed after 1 day. This
observation provides further support for our selective
adsorption hypothesis.
To explore the catalytic properties of the adsorbate-stabilized
FePt NPs, we chose the hydrogenation of cinnamaldehyde as a
model reaction. When catalyzed by Pt, this reaction leads to a
range of products, with the hydrogenation of the CC bond
favored both thermodynamically (by ∼35 kJ·mol−1) and
kinetically.20 The selectivity of the reaction can be adjusted in
favor of CO hydrogenation by performing it in the presence
of metal salts,21 under basic conditions,22 or on hindered solid
supports.23 Bimetallic catalysts,24 especially CoPt25 and FePt26
at high pressure of H2,27 can be CO-selective. The selectivity
can also be imparted by sterically hindered capping ligands.28 A
recent report describes a CO-selective hydrogenation
achieved through specific noncovalent interactions between
cinnamaldehyde and the surface capping ligands.29
We performed cinnamaldehyde hydrogenations in a mixture
of hexane and perfluorodecalin at 50 °C and 1 atm of H2.
Because this solvent mixture phase-separates at lower temper-
atures, all the CXF-protected NP catalysts could be
conveniently recovered and recycled simply by separating the
fluorous phase containing the NPs. The colloidal stability of the
catalysts stabilized by longer-chain (>C10) CXF ligands was
exceptional. The phase mixing/separation could be repeated at
least 10 times for C18F- Fe0.25Pt0.75 without inducing NP
aggregation.
For all the FePt NP catalysts, the selectivity and reactivity
could be readily tuned by changing the length and nature of the
carboxylic acid chains and the Fe content of the catalyst (Table
1 and Figure 2). The parent OA−Pt catalyst, while prone to
aggregation, was active under the mild conditions employed,
converting 14.9% of cinnamaldehyde in 1 h. As expected, 3-
phenylpropanal 3 and 3-phenyl-1-propanol 4 were the major
products (Table 1, entry 1). The CXH−Fe0.25Pt0.75 catalysts
were somewhat less active than pure Pt (Table 1, entries 2−4).
The chemoselectivity of these catalysts was shifted in favor of
the CO hydrogenation product. This modest change in
chemoselectivity was in line with what could generally be
expected from a FePt catalyst26 and attributable to the surface
ensemble effect produced through Fe substitution on fcc-
packed Pt surface. The NPs stabilized with C18H, the longest
chain in the series, were both more active and less selective than
either the C5H- or C10H- stabilized catalysts. This suggests
that the steric effect of the surface CXH ligands and the surface
ensemble effect may be antagonistic.
The change from hydrocarbon-based to fluorous adsorbates
effected a dramatic change in the activity and selectivity of the
FePt catalysts. All the CXF-capped NPs showed an extremely
high preference for CO hydrogenation (Table 1, Entries 5−
10). It is important to note that the rate of the hydrogenation
was markedly accelerated by the longer-chain C8F−C18F
adsorbates (Table 1, entries 6−10). Quantitative conversion of
cinnamaldehyde could be achieved over the C10F−Fe0.33Pt0.67
catalyst after ∼1.3 h (Figure S29, SI). Both the activity and
selectivity of the CXF-capped catalysts were strongly correlated
with the Fe content of the NPs (Table 1, entries 7, 9, and 10).
The relatively low-Fe C10F−Fe0.13Pt0.87 catalyst was markedly
less selective than either C10F−Fe0.25Pt0.75 or C10F−
Fe0.33Pt0.67
.
Next, we explored the dependence of product selectivity on
the degree of conversion for a number of NP catalysts (Figure
2). The pure Pt catalyst was not especially selective. 3-
Phenylpropanal and the secondary product 3-phenyl-1-
propanol were dominant, whereas cinnamyl alcohol all but
disappeared from the reaction mixture at higher conversions
(Figure 2A). For C10F−Fe0.13Pt0.87, the selectivity was shifted
markedly in favor of cinnamyl alcohol (Figure 2B). Another
notable change compared with the pure Pt was the decrease in
abundance of 3-phenylpropanol, especially at lower conver-
sions. The change in selectivity was even more dramatic for
C10F−Fe0.33Pt0.67 (Figure 2C). This catalyst formed cinnamyl
alcohol almost exclusively until a very high degree of
conversion (>90%). Only then could a minor amount of 3-
phenyl-1-propanol be detected in the reaction mixture,
indicating that this product is formed through slow hydro-
genation of cinnamyl alcohol. Notably, this remarkable
improvement in selectivity was achieved concomitantly with
an increase in the reaction rate and turnover frequency.
As expected, the FePt catalysts were readily poisoned by
either amines or thiols (section 3, SI). Addition of
(perfluorooctyl)propylamine to C10F−Fe0.25Pt0.75 led to a
substantial reaction rate decrease (by half), and an analogous
thiol completely deactivated the catalyst. Because both amines
and thiols have a high affinity for Pt surfaces, we believe the
poisoning affects only the Pt islands that are free from the
carboxylic acid adsorbate.
The dramatic rate and selectivity differences between the
FePt catalysts can be readily explained by considering the steric
and nanoenvironment effects exerted by the surface-bound
ligands. The adsorbed carboxylic acids are expected to provide a
degree of site isolation to the Pt sites on the NP surface.
Longer-chain ligands constrain the geometry around these sites,
2531
DOI: 10.1021/acscatal.5b00262
ACS Catal. 2015, 5, 2529−2533

ACS Catalysis
Letter
(4) Narayanan, R.; El-Sayed, M. A. J. Phys. Chem. B 2005, 109,
12663−12676.
permitting only the head-on approach of the substrate. With a
sufficient density of the surface ligands, a sideways approach of
cinnamaldehyde to Pt needed for the CC hydrogenation can
be completely prevented. Thus, a preference for the CO
hydrogenation is realized.29 The perfluorinated carboxylic acids
are more rigid/rodlike than their hydrocarbon counterparts.31
Thus, the geometric confinement effect is much more
pronounced for the CXF-protected NPs. As the carboxylate
ligands bind preferentially to Fe, the density of the surface-
bound chains and the degree of catalytic site confinement
increase with a rise in the Fe content. Thus, catalysts containing
more Fe are more CO-selective. We also expect that the
“everything-phobic” fluorous ligands32 surrounding the active
sites will facilitate the desorption of the inactive reaction
intermediates33 and discourage the unproductive binding of
products and substrates.
In conclusion, we explored the catalytic activity of a range of
FePt bimetallic NPs stabilized by carboxylate ligands. These
ligands have a high affinity for only one of the metals (Fe). The
NPs stabilized by fluorous ligands were remarkably active in
catalyzing the hydrogenation of cinnamaldehyde. The chemo-
selectivity of the catalysis was significantly altered with respect
to the parent Pt catalyst. Longer-chain fluorous carboxylic acid
ligands shifted the selectivity almost entirely in favor of CO
hydrogenation, with a concomitant increase in the reaction rate.
These effects could be readily explained by considering the
isolation/crowding of the Pt active sites by the rigid fluorous
ligands. The fluorous-stabilized bimetallic NPs had excellent
colloidal stability and were readily recyclable. Investigations of
other bimetallic NP catalysts selectively functionalized with
surface adsorbates are under way in our laboratories.
(5) Astruc, D.; Lu, F.; Aranzaes, J. R. Angew. Chem., Int. Ed. 2005, 44,
7852−7872.
(6) White, R. J.; Luque, R.; Budarin, V. L.; Clark, J. H.; Macquarrie,
D. J. Chem. Soc. Rev. 2009, 38, 481−494.
(7) Thomas, J. M.; Johnson, B. F. G.; Raja, R.; Sankar, G.; Midgley, P.
A. Acc. Chem. Res. 2002, 36, 20−30.
́
(8) (a) Kralik, M.; Biffis, A. J. Mol. Catal. A: Chem. 2001, 177, 113−
138. (b) Green, I. X.; Tang, W.; Neurock, M.; Yates, J. T. Acc. Chem.
Res. 2013, 47, 805−815. (c) McEntee, M.; Tang, W.; Neurock, M.;
Yates, J. T. J. Am. Chem. Soc. 2014, 136, 5116−5120. (d) Corma, A.;
Garcia, H. Chem. Soc. Rev. 2008, 37, 2096−2126.
(9) Niu, Z.; Li, Y. Chem. Mater. 2013, 26, 72−83.
(10) (a) Li, D.; Wang, C.; Tripkovic, D.; Sun, S.; Markovic, N. M.;
Stamenkovic, V. R. ACS Catal. 2012, 2, 1358−1362. (b) Ansar, S. M.;
Ameer, F. S.; Hu, W.; Zou, S.; Pittman, C. U.; Zhang, D. Nano Lett.
2013, 13, 1226−1229. (c) Crespo-Quesada, M.; Andanson, J.-M.;
Yarulin, A.; Lim, B.; Xia, Y.; Kiwi-Minsker, L. Langmuir 2011, 27,
7909−7916. (d) Solla-Gullon, J.; Montiel, V.; Aldaz, A.; Clavilier, J. J.
Electrochem. Soc. 2003, 150, E104−E109.
(11) Toshima, N.; Yonezawa, T. New J. Chem. 1998, 22, 1179−1201.
(12) Sun, S. H. Adv. Mater. 2006, 18, 393−403.
(13) Ofir, Y.; Samanta, B.; Arumugam, P.; Rotello, V. M. Adv. Mater.
2007, 19, 4075−4079.
́ ́
(14) (a) Horvath, I. T.; Rabai, J. Science 1994, 266, 72−75.
(b) Chechik, V.; Crooks, R. M. J. Am. Chem. Soc. 2000, 122, 1243−
1244.
(15) Sun, S. H.; Murray, C. B.; Weller, D.; Folks, L.; Moser, A. Science
2000, 287, 1989−1992.
(16) (a) Wu, N. Q.; Fu, L.; Su, M.; Aslam, M.; Wong, K. C.; Dravid,
V. P. Nano Lett. 2004, 4, 383−386. (b) Klokkenburg, M.; Hilhorst, J.;
Erne, B. H. Vib. Spectrosc. 2007, 43, 243−248.
(17) Howard, L. E. M.; Nguyen, H. L.; Giblin, S. R.; Tanner, B. K.;
Terry, I.; Hughes, A. K.; Evans, J. S. O. J. Am. Chem. Soc. 2005, 127,
10140−10141.
(18) Nguyen, H. L.; Howard, L. E. M.; Stinton, G. W.; Giblin, S. R.;
Tanner, B. K.; Terry, I.; Hughes, A. K.; Ross, I. M.; Serres, A.; Evans, J.
S. O. Chem. Mater. 2006, 18, 6414−6424.
(19) Samia, A. C. S.; Schlueter, J. A.; Jiang, J. S.; Bader, S. D.; Qin, C.-
J.; Lin, X.-M. Chem. Mater. 2006, 18, 5203−5212.
(20) (a) Gallezot, P.; Richard, D. Catal. Rev. Sci. Eng. 1998, 40, 81−
126. (b) Yuan, Y.; Yao, S.; Wang, M.; Lou, S.; Yan, N. Curr. Org. Chem.
2013, 17, 400−413. (c) Mohr, C.; Claus, P. Sci. Prog. 2001, 84, 311−
334.
ASSOCIATED CONTENT
* Supporting Information
The following files are available free of charge on the ACS
■
S
Publications website at DOI: 10.1021/acscatal.5b00262.
Additional experimental details and characterization
(TEM, EDS, SAED, PXRD, 1H NMR, GC/MS
chromatograms) (PDF)
Video of catalyst separation/recycling (AVI)
(21) (a) Tuley, W. F.; Adams, R. J. Am. Chem. Soc. 1925, 47, 3061−
3068. (b) Galvagno, S.; Donato, A.; Neri, G.; Pietropaolo, R.;
Pietropaolo, D. J. Mol. Catal. 1989, 49, 223−232.
AUTHOR INFORMATION
Corresponding Author
*Phone: +966-12-8084592. E-mail: valentin.rodionov@kaust.
edu.sa.
■
(22) Bhogeswararao, S.; Srinivas, D. J. Catal. 2012, 285, 31−40.
(23) (a) Blackmond, D. G.; Oukaci, R.; Blanc, B.; Gallezot, P. J.
Catal. 1991, 131, 401−411. (b) Guo, Z.; Xiao, C.; Maligal-Ganesh, R.
V.; Zhou, L.; Goh, T. W.; Li, X.; Tesfagaber, D.; Thiel, A.; Huang, W.
ACS Catal. 2014, 4, 1340−1348.
Author Contributions
‡K.B.V. and K.V.B. contributed equally.
Notes
(24) Qi, X.; Axet, M. R.; Philippot, K.; Lecante, P.; Serp, P. Dalton
Trans. 2014, 43, 9283−9295.
The authors declare no competing financial interest.
(25) Tsang, S. C.; Cailuo, N.; Oduro, W.; Kong, A. T. S.; Clifton, L.;
Yu, K. M. K.; Thiebaut, B.; Cookson, J.; Bishop, P. ACS Nano 2008, 2,
2547−2553.
ACKNOWLEDGMENTS
■
The authors are grateful to Prof. J.-M. Basset for helpful
discussions. This research was supported by King Abdullah
University of Science and Technology.
(26) da Silva, A. B.; Jordao, E.; Mendes, M. J.; Fouilloux, P. Appl.
̃
Catal., A 1997, 148, 253−264.
(27) (a) Goupil, D.; Fouilloux, P.; Maurel, R. React. Kinet. Catal. Lett.
1987, 35, 185−193. (b) Liu, Z.; Tan, X.; Li, J.; Lv, C. New J. Chem.
2013, 37, 1350−1357.
REFERENCES
■
(28) Wu, B. H.; Huang, H. Q.; Yang, J.; Zheng, N. F.; Fu, G. Angew.
Chem., Int. Ed. 2012, 51, 3440−3443.
(1) Roucoux, A.; Schulz, J.; Patin, H. Chem. Rev. 2002, 102, 3757−
3778.
(29) Kahsar, K. R.; Schwartz, D. K.; Medlin, J. W. J. Am. Chem. Soc.
2013, 136, 520−526.
(2) (a) Bell, A. T. Science 2003, 299, 1688−1691. (b) Haruta, M.
Chem. Rec. 2003, 3, 75−87.
(3) Jia, C.-J.; Schuth, F. Phys. Chem. Chem. Phys. 2011, 13, 2457−
2487.
(30) Umpierre, A. P.; de Jesus
1413−1418.
́
, E.; Dupont, J. ChemCatChem 2011, 3,
2532
DOI: 10.1021/acscatal.5b00262
ACS Catal. 2015, 5, 2529−2533

ACS Catalysis
Letter
(31) Eaton, D. F.; Smart, B. E. J. Am. Chem. Soc. 1990, 112, 2821−
2823.
(32) Kirsch, P. Modern Fluoroorganic Chemistry: Synthesis, Reactivity,
Applications; Wiley-VCH: Weinheim, 2005; pp 1−23.
(33) Wang, H.; Krier, J. M.; Zhu, Z.; Melaet, G.; Wang, Y.; Kennedy,
G.; Alayoglu, S.; An, K.; Somorjai, G. A. ACS Catal. 2013, 3, 2371−
2375.
2533
DOI: 10.1021/acscatal.5b00262
ACS Catal. 2015, 5, 2529−2533