Rhodium nanoparticle catalysts stabilized with a polymer that enhances stability without compromising activity

Article abstract of DOI:10.1039/c0cc04641h

Rh NPs coated with a PVP-derived polymer, that combines several distinct protective interactions, exhibit superior thermal and catalytic stability compared to analogue NPs coated with PVP.

Full text of DOI:10.1039/c0cc04641h

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2011, 47, 2529–2531
www.rsc.org/chemcomm
COMMUNICATION
Rhodium nanoparticle catalysts stabilized with a polymer that enhances
stability without compromising activityw
Ning Yan,* Yuan Yuan and Paul J. Dyson*
Received 27th October 2010, Accepted 14th December 2010
DOI: 10.1039/c0cc04641h
Rh NPs coated with a PVP-derived polymer, that combines
several distinct protective interactions, exhibit superior thermal
and catalytic stability compared to analogue NPs coated
with PVP.
Metal nanoparticles (NPs) are thermodynamically unstable
and therefore require stabilizers in both their preparation and
application. Stabilizers can be categorized according to the
Fig. 1 Structures of the polymers used as NP stabilizers in this study.
type of bond/interaction formed between the stabilizer and
the NP surface, viz. steric, electrostatic and coordinative
1
interactions. Manipulating one or more of these interactions
hydrogenation reactions. Furthermore, the new systems were
compared with NPs stabilized by PVP, the ‘benchmark’
stabilizer.
should result in the formation of more efficient stabilizers that
ultimately will afford NPs with superior properties to those in
7
Polymer 1 is known, but as far as we are aware has not
2
current use. In NP catalysis it is imperative to strike a balance
been used as a NP stabilizer. A slightly modified procedure
was used to prepare 1 in this study. N-Vinyl-2-pyrrolidone
(NVP) was reacted with lithium diisopropylamide (LDA) and
then with ethyl chloroformate to afford 3,3-di(ethoxycarbonyl)-
1-vinylpyrrolidin-2-one (DEVP). Free radical polymerization
of DEVP yielded the corresponding polymer poly-3,3-di-
(ethoxycarbonyl)-1-vinylpyrrolidin-2-one (PDEVP), which upon
3
between catalytic activity and stability since, in many cases,
4
the two factors are counter-productive. A general strategy for
the preparation of stable NPs, without significantly sacrificing
their catalytic activity, was recently proposed that is based on
the design of stabilizers that incorporate several weak stabiliza-
5
tion mechanisms. In principle, each individual functionality
may contribute to the stability of the NPs without hindering
their overall catalytic activity.
hydrolysis in NaOH/ethanol/H
2
O mixture gave 1 as the only
) of PDEVP, based
product. The average molecular weight (M
n
To test this hypothesis modified polyvinylpyrrolidone (PVP)
systems that include functionalities which provide all the types
of interactions were prepared and studied. PVP was selected as
it is the most widely used polymer for stabilizing NP catalysts,
combining a weakly coordinating amide functionality with
on GPC analysis, is 5.9 kDa, corresponding to ca. 230 repeating
units per molecular chain. The structures of both PDEVP and
1 were verified by IR and NMR spectroscopy (see ESIw). TGA
analysis (see ESIw, Fig. S1a) shows that the decomposition
temperatures of 1 under both an inert and oxygen containing
atmosphere exceed 300 1C, similar to that of PVP or alkyl
6
steric protection. The synthesis of stabilizers that are superior
8
to PVP, i.e. providing NP catalysts with similar activity but
higher stability, has proven to be very challenging. We postulated
that the introduction of carboxylate group into PVP could
further increase NP stability, by weak coordination of the
carboxylate group with the metal surface, and additionally via
electrostatic stabilization with the formation of a protective
electrical double layer. Accordingly, polymers 1 and 2 (Fig. 1)
were prepared and evaluated as stabilizers for Rh NPs
in organic-aqueous and organic-ionic liquid (IL) biphasic
modified PVPs. Polymer 1 is soluble in water, slightly soluble
in alcohol, and insoluble in other common solvents. The
+
solubility of 1 can be tuned by replacement of Na cation
with other cations. For example, 2, corresponding to the
1-methyl-3-octylimidazolium (omim) salt of 1, is readily obtained
by cation exchange, and exhibits good solubility in both polar
organic solvents and ILs. The TGA curve for 2 indicates a two-
stage weight loss process, corresponding to the decomposition of
omim and polymer chain, respectively (see ESIw, Fig. S1b).
NPs coated with 1 or PVP were prepared in water from
3 4
RhCl by reduction with NaBH in the presence of the polymer
Institut des Sciences et Inge´nierie Chimiques,
(denoted as 1–Rh and PVP–Rh, respectively, see ESIw for full
details). For both samples the resulting NP solutions were
black and transparent. Micrographs of the Rh NPs are shown
in Fig. 2 (and HRTEM is provided in Fig. S2a–b in ESIw).
Relatively narrow unimodal size distributions with a diameter
Ecole Polytechnique Fe´de´rale de Lausanne (EPFL), CH-1015
Lausanne, Switzerland. E-mail: ning.yan@epfl.ch, paul.dyson@epfl.ch
w Electronic supplementary information (ESI) available: Experimental
details, TGA analysis, additional (HR)TEM images, XPS and IR
spectra. See DOI: 10.1039/c0cc04641h
This journal is c The Royal Society of Chemistry 2011
Chem. Commun., 2011, 47, 2529–2531 2529

View Article Online
NP precipitation is observed after reaction, which is not the
case for the catalyst solution containing 1–Rh. Evaluation of
the long term catalytic stability of 1–Rh relative to PVP–Rh
was undertaken at a reaction temperature of 60 1C. Five
batches were carried out, and in the first run a quantitative
conversion was obtained with PVP–Rh and a conversion of
9
3% was reached with 1–Rh. Upon recycling the activity of
PVP–Rh decreases with a conversion of o70% after five
batches whereas the conversion with 1–Rh remains essentially
constant and always exceeds 90%. The decrease in activity of
PVP–Rh may be correlated to NP aggregation which is
observed visually by the formation of metallic deposits and
also by TEM (Fig. 2c and ESIw, Fig. S6). TEM analysis
revealed that PVP–Rh forms spherical superstructures ranging
from 40–100 nm. In contrast, 1–Rh remains essentially
unchanged. With phenol, a substrate that is more readily
Fig. 2 TEM images of Rh NPs. (a) PVP–Rh; (b) 1–Rh; (c) PVP–Rh
after 5 batches; (d) 1–Rh after 5 batches.
11
hydrogenated than toluene, a similar trend was observed in
recycling experiments. 1–Rh maintained its activity over
15 batches (after which experiments were stopped) whereas
PVP–Rh was completely deactivated after 12 batches
(ESIw, Fig. S7). In fact, the difference in the performance between
1–Rh and PVP–Rh is more significant when the polymer : metal
is reduced. As an example, when the ratio is reduced to 2,
PVP–Rh can no longer stabilize Rh NPs effectively and a
conversion of only 7.6% is obtained after three batches for the
hydrogenation of toluene, whereas 1–PVP can be recycled for at
least 8 times without any loss of activity (ESIw, Fig. S8). Being
able to operate at low stabilizer ratios is advantageous since PVP
coated NPs require a large excess relative to the metal in order to
achieve efficient stabilization.
of ca. 4 nm were observed for both 1 and PVP protected Rh
NPs. The thermal stability of the Rh NP solutions was studied
À1
by heating them at a heating rate of 1 1C min to 200 1C and
then maintaining the solutions at this temperature for 2 hours;
PVP protected NPs do not generally survive at temperatures
9
higher than 180 1C. As the temperature was increased above
1
60 1C a black precipitate was observed for the PVP–NPs and
at the end of the heating period the majority of the NPs have
decomposed (Fig. 3). In contrast, for the 1–NPs no precipitate
was observed and based on TEM the size of Rh NPs remained
unchanged (ESIw, Fig. S2c), indicating that the thermal stability
offered by 1 is superior to that of PVP. A second method was
used to prepare Rh NPs based on coating preformed (naked)
1–Rh was evaluated as a catalyst for the hydrogenation of
other aromatic substrates under the same conditions (Table 1),
demonstrating the utility of the nanocatalyst system. Even
for relatively challenging substrate such as propylbenzene a
conversion of 69% was achieved, representing a turnover
1
0
Rh NPs, synthesized in glycol, with PVP or 1. This method
ensures that the size and shape of Rh NPs for the two samples
is the same. The same heat treatment to that described above
applied to these NPs and essentially the same observations
were made, i.e. the PVP protected Rh NPs undergo extensive
precipitation whereas those stabilized by 1 remain essentially
largely unchanged (ESIw, Fig. S3).
À1
frequency of 170 h . This value is higher than that observed
À1 12
for other Rh NPs that range from 7 to 80 h
,
although the
comparison should be treated with caution due to the different
conditions applied. Rh NPs protected by 2, i.e. 2–Rh, are
highly soluble in ILs and these solutions were also found to be
catalytically active under biphasic conditions (Table 1).
The stability of PVP–Rh and 1–Rh under catalytic condi-
tions was also investigated with toluene hydrogenation used as
a model reaction under biphasic conditions. Unfortunately,
both PVP–Rh and 1–Rh deactivated at 200 1C under catalytic
condition, although 1–Rh is highly thermally stable at this
temperature. In the case of 1–Rh, however, higher conversion
of toluene was indeed observed (15% vs. 6%), suggesting
X-Ray photoelectron spectroscopy (XPS) was used to study
the interactions between the surface of the Rh NPs and the
polymers. XPS spectra of the C 1s and N 1s for the two
samples are very similar (ESIw, Fig. S4), indicating that C and
N atoms on the two polymers are not responsible for the
differences in stability between the two Rh NPs. For O 1s
spectra, however, considerable differences are observed. Compared
to PVP that has a single peak for the amide (N–CQO) oxygen
atom at 531.3 eV, the PVP–Rh NPs contain a peak at 532.0 eV,
implying a reduction of electron density of the oxygen, due to
the interaction between PVP and Rh NPs. This new peak is
also observed in 1–Rh NPs, which is not unexpected due to the
presence of amide group. Additionally, a weak O ls signal at
1
–Rh is somewhat more stable than PVP–Rh under forcing
catalytic condition. At 100 1C PVP–Rh achieved a toluene
conversion of 93%, slightly more active than 1–Rh (88%), but
5
35.5 eV indicates the presence of surface absorbed carboxylic
1
3
acid species.
IR spectroscopy confirms the interaction
between the carboxylic acid group in 1 with the Rh NP surface
(ESIw, Fig. S5), and provides a rationale for the greater
stability of 1 compared to PVP.
Fig. 3 Thermal stability of PVP–Rh and 1–Rh; (a) before heating;
(b) after heating at 200 1C for 2 h. Arrows indicate the formation of
metallic deposits.
2
530 Chem. Commun., 2011, 47, 2529–2531
This journal is c The Royal Society of Chemistry 2011

View Article Online
Table 1 Catalytic performance of 1–Rh and 2–Rh NPs in arene hydrogenation
Substrate
Phenol
Stab.
Solvent
Water
Con. (%)
100
Sel. (%)
1
Cyclohexanol
9.8
Cyclohexanone
0.1
9
—
Anisole
Styrene
1
1
Water
Water
100
100
Ethylbenzene
4
Ethylcyclohexane
55.7
4.2
Ethyl-benzene
n-Propyl-benzene
Toluene
1
1
2
2
Water
Water
[bmim][BF ]
4
99
69
95
97
Ethylcyclohexane
Propylcyclohexane
Methylcyclohexane
Methylcyclohexane
Toluene
2 4
[C OHmim][BF ]
Reaction conditions: reactant (0.5 mmol), 1 Â 10À3 mmol of Rh NPs (Rh : stabilizer = 1 : 20) in 1 ml water or IL, H
(20 atm), 60 1C, 2 h.
2
Dr V. Laporte and S. Katsyuba are thanked for helpful
discussions on the XPS data and IR spectra, respectively.
Notes and references
1
2
A. Roucoux, J. Schulz and H. Patin, Chem. Rev., 2002, 102, 3757.
For examples see: (a) C. Chiappe, D. Pieraccini, D. Zhao, Z. Fei
and P. J. Dyson, Adv. Synth. Catal., 2006, 348, 68; (b) I. Favier,
M. Gomez, G. Muller, D. Picurelli, A. Nowicki, A. Roucoux and
´
J. Bou, J. Appl. Polym. Sci., 2007, 105, 2772; (c) D. D. Ankola,
K. Ravi, F. Chiellini and R. Solaro, Macromolecules, 2009, 42,
7
388; (d) G. V. Ramesh, S. Porel and T. P. Radhakrishnan, Chem.
Soc. Rev., 2009, 38, 2646; (e) J. Dupont and J. D. Scholten, Chem.
Soc. Rev., 2010, 39, 1780.
D. Zhao, M. Wu, Y. Kou and E. Min, Catal. Today, 2002, 74, 157.
Fig. 4 Zeta potential of PVP–Rh, 1–Rh and PAA–Rh NPs at
pH = 7.
3
1
4
4 (a) Y. Li and M. A. El-Sayed, J. Phys. Chem. B, 2001, 105, 8938;
b) A. Roucoux, J. Schulz and H. Patin, Adv. Synth. Catal., 2003, 345,
22; (c) C. A. Stowell and B. A. Korgel, Nano Lett., 2005, 5, 1203.
5 N. Yan, X.-C. Xiao and Y. Kou, Coord. Chem. Rev., 2010, 254, 1179.
Electrostatic stabilization was assessed by zeta potential
measurements (Fig. 4), and as expected, the Rh NPs stabilized
by PVP have a low overall charge, indicated by a zeta potential
of À20 mV. The negative charge might originate from absorbed
(
2
6
D. Astruc, F. Lu and J. R. Aranzaes, Angew. Chem., Int. Ed., 2005,
4, 7852.
M. Bencini, E. Ranucci, P. Ferruti, C. Oldani, E. Licandro and
4
3
anions such as chloride from the NP precursor (RhCl ) or
7
electron donation from PVP, as demonstrated in the case
S. Maiorana, Macromolecules, 2005, 38, 8211.
1
5
of a Au NP–PVP system. Compared to PVP–Rh, a highly
negative value (À64 mV) is observed for 1–Rh, indicative of an
electrical double-layer around the NPs, and resulting in a
Columbic repulsion between them. In general, NPs with zeta
potentials higher than Æ60 mV are regarded as being highly
8 (a) N. Yan, J. Zhang, Y. Tong, S. Yao, C. Xiao, Z. Li and Y. Kou,
Chem. Commun., 2009, 29, 4423; (b) N. Yan, J. Zhang, Y. Yuan,
G. Chen, P. J. Dyson, Z. Li and Y. Kou, Chem. Commun., 2010,
4
6, 1631.
9 C. Zhao, W. Gan, X. Fan, Z. Cai, P. J. Dyson and Y. Kou,
J. Catal., 2008, 254, 244.
10 Y. Wang, J. Ren, K. Deng, L. Gui and Y. Tang, Chem. Mater.,
16
stable, and NPs stabilized by carboxylate-containing polymers
2
000, 12, 1622.
1
7
tend to have zeta potentials ranging from À40 to À60 mV.
For comparison, polyacrylic acid (PAA) coated Rh NPs were
prepared in water by NaBH reduction (PAA–Rh) and their
1
1 S. Nishimura, Handbook of Heterogeneous Catalytic Hydrogenation
for Organic Synthesis, John Wiley & Sons, 2001, p. 427.
1
2 (a) J. Schulz, A. Roucoux and H. Patin, Chem.–Eur. J., 2000, 6, 618;
4
(
b) A. Roucoux, J. Schulz and H. Patin, Adv. Synth. Catal., 2003, 345,
22; (c) P. Migowski and J. Dupont, Chem.–Eur. J., 2006, 13, 32;
(d) C. Zhao, H. Wang, N. Yan, C. Xiao, X. Mu, P. J. Dyson and
Y. Kou, J. Catal., 2007, 250, 33; (e) B. Leger, A. Denicourt-Nowicki,
A. Roucoux and H. Olivier-Bourbigou, Adv. Synth. Catal., 2008, 350,
53; (f) B. Leger, A. Denicourt-Nowicki, H. Olivier-Bourbigou and
A. Roucoux, Inorg. Chem., 2008, 47, 9090.
13 K. W. Wulser and M. A. Langell, Catal. Lett., 1992, 15, 39.
zeta potential was found to be À49 mV, in agreement with the
2
1
6
literature values. The reason for the more negative zeta
potential of 1–Rh compared to PAA–Rh is not clear, but is
likely to be due to combined electron donation from both the
amide and the carboxylate group.
´
1
´
In summary, we have shown that Rh NPs coated by 1 exhibit
superior thermal and catalytic stability to PVP coated NPs. The
enhanced stability may be attributed to the presence of
the carboxylate group that may provide weakly coordinate to
the NP surface and an electrical double-layer that helps prevent
aggregation. Combined these additional interactions result in
NPs that can be recycled and reused without loss of activity.
N. Yan thanks the EU for Marie Curie International
Incoming Fellowship (Number: 252125-TCPBRCBDP).
Prof. J. A. Hubbell and Dr A. Vandervlies are acknowledged
for assistance in zeta potential measurement. Prof. H. A. Klok
and Sanhao Ji are thanked for carrying out the GPC analysis.
1
4 R. G. Finke, in Metal Nanoparticles: Synthesis, Characterization
and Applications, ed. D. L. Feldheim and C. A. Foss, Jr., Marcel
Dekker, New York, 2002, ch. 2, p. 17–54.
1
1
5 H. Tsunoyama, N. Ichikuni, H. Sakurai and T. Tsukuda, J. Am.
Chem. Soc., 2009, 131, 7086.
6 ‘‘Zeta Potential of Colloids in Water and Waste Water’’, ASTM
Standard D 4187, American Society for Testing and Materials,
1985.
17 For examples see: (a) S. Chen and K. Kimura, Langmuir, 1999, 15,
1075; (b) K. Yoosaf, B. I. Ipe, C. H. Suresh and K. G. Thomas,
J. Phys. Chem. C, 2007, 111, 12839; (c) J. Hang, L. Shi, X. Feng
and L. Xiao, Powder Technol., 2009, 192, 166; (d) K. Mori,
A. Kumami, M. Tomonari and H. Yamashita, J. Phys. Chem. C,
2009, 113, 16850.
This journal is c The Royal Society of Chemistry 2011
Chem. Commun., 2011, 47, 2529–2531 2531