Topics in Catalysis

, Volume 47, Issue 1, pp 15–21

Some Aspects of Colloidal Nanoparticle Stability, Catalytic Activity, and Recycling Potential

Authors

  • Radha Narayanan
    • USTAR Center for Nanobiosensors, Department of ChemistryUniversity of Utah
    • Laser Dynamics Laboratory, School of Chemistry and BiochemistryGeorgia Institute of Technology
Original Paper

DOI: 10.1007/s11244-007-9029-0

Cite this article as:
Narayanan, R. & El-Sayed, M.A. Top Catal (2008) 47: 15. doi:10.1007/s11244-007-9029-0

Abstract

In this review article, we examine many important aspects of the nanocatalysis field such as size and shape dependent nanocatalysis, the stability of nanoparticles during its catalytic function, and their recycling potential. We provide an overview of some of the work in the literature pertinent to these topics and also discuss some of our own work in these important areas. Some examples of how the catalytic activity is affected by the size of the nanoparticles are discussed as well as how the catalytic process affects the nanoparticle size after its catalytic function. The synthesis of platinum nanoparticles of different shapes is surveyed and the dependence of nanoparticle shape on the catalytic activity is discussed. In addition, changes in the nanoparticle shape and resulting changes in the catalytic activity are also discussed. The recycling potential of the metal nanocatalysts is also highlighted. In addition, a simple examination of the mechanism of nanocatalysis is discussed.

Keywords

NanocatalysisTransition metal nanoparticlesCatalytic activityElectron transfer reactionSuzuki cross-coupling reactionNanoparticles

Introduction

The nanocatalysis field is a highly active one with close to 4,300 publications up to 2006. In the year 2006 alone, there have been over 500 publications so far. The statistical results on the number of publications in the nanocatalysis field from 1996 to 2005 are obtained using the SciFinder Scholar search engine and is shown in Fig. 1. There is definitely an exponential growth in the number of publications in this field during the past decade. It is also worth noting that there have been over 600 patents related to the use of nanoparticles as catalysts. There have been numerous review articles written on the use of colloidal metal nanoparticles as catalysts (homogeneous catalysis) [18] as well as metal nanoparticles supported on various substrates (heterogeneous catalysis) [921].
https://static-content.springer.com/image/art%3A10.1007%2Fs11244-007-9029-0/MediaObjects/11244_2007_9029_Fig1_HTML.gif
Fig. 1

Statistics on the number of publications in the nanocatalysis field from 1995 to 2005 (results obtained using SciFinder Scholar)

In this short review article, we highlight some of the work in different aspects of the nanocatalysis field such as size-dependent nanocatalysis, shape-dependent nanocatalysis, stability of nanoparticles after the catalytic process, and recycling potential of nanocatalysts. We provide an overview of the work in these areas and also discuss some of our own work in these important areas in the nanocatalysis field.

Effect of Nanoparticle Size on Catalytic Activity and the Effect of the Catalytic Process on the Nanoparticle Size

The effect of the nanocatalyst size on the catalytic activity has been studied extensively in the literature. It is well-known that the reaction rate per unit catalyst surface area can vary with the nanoparticle size, and that decreasing the nanoparticle size does not always result in an increased reaction rate per unit mass of the metal nanoparticles [2227]. For example, in the case of the hydrogenation of 2,4-dinitrotoluene to form 2,4-diaminotoluene, it is shown that the catalytic activity is higher for larger palladium nanoparticles since the ability to form the palladium-β-hydride phase controls the hydrogenation activity and the formation of this phase increases with larger palladium nanoparticles [22]. In the case of the NO–CO reaction, the catalytic activity increases with decreasing nanoparticle size, but the smallest nanoparticles are found to be less catalytically active due to increased proportion of strongly bound nitrogen species that tends to cover the particle surface and decrease the amount of available sites for NO dissociation [27].

The effect of the palladium nanoparticle size on the catalytic activity of the Suzuki reaction between phenylboronic acid and iodobenzene to form biphenyl was studied by Li and El-Sayed [28]. The four different sizes of the palladium nanoparticles investigated are 3.0, 3.9, 5.2, and 6.6 nm. The catalytic activity of the Pd nanoparticles is measured by the turnover frequency (TOF) as a function of the total number of surface atoms as well as just the vertex and edge surface atoms. When all of the surface atoms are considered, the TOF varies in the following order: 3.9 nm > 3.0 nm > 5.2 nm > 6.6 nm. When only the vertex and edge atoms are considered, the dependence of the particle size on the catalytic activity disappears. This confirms that the vertex and edge atoms are the active centers for catalysis. In the case of the 3.0 nm Pd nanoparticles, the lower catalytic activity observed could be due to the stronger adsorption of the reaction intermediates on the particle surface, in which the strongly adsorbed species act as a poison to the active sites of the nanoparticles and therefore results in a lower reaction rate.

One important question to assess is how the catalytic process affects the nanoparticles during the course of the catalytic process. The Suzuki cross-coupling reaction is a relatively harsh one since it takes place at 100 °C for 12 h. We have conducted studies to determine what happens to the nanoparticles after catalyzing the Suzuki reaction by using Transmission electron microscopy (TEM) [29]. As can be seen in Fig. 2c–d, we found that the spherical PVP capped palladium nanoparticles grew larger after the first cycle of the Suzuki reaction due to the Ostwald ripening processes that take place during the harsh reaction conditions. After the second cycle of the Suzuki reaction, it is observed that the larger palladium nanoparticles that formed have aggregated and precipitated out of solution and only the smaller nanoparticles remain in solution as can be seen in Fig. 2e–f. In the case of the tetrahedral PVP capped platinum nanoparticles used as catalysts for the Suzuki reaction [30], we observed that there is a transformation in the nanoparticle shape from tetrahedral to similarly sized spherical nanoparticles that also takes place during the first cycle. During the second cycle, the transformed spherical nanoparticles continue to grow larger in size.
https://static-content.springer.com/image/art%3A10.1007%2Fs11244-007-9029-0/MediaObjects/11244_2007_9029_Fig2_HTML.gif
Fig. 2

TEM images and Gaussian fits of the size distributions of the spherical PVP-Pd nanoparticles [29] before the Suzuki reaction (a, b), after the first cycle of the Suzuki reaction (c, d), and after the second cycle of the Suzuki reaction (e, f)

Effect of Stabilizers on the Catalytic Activity and the Effect of the Catalytic Process on the Nanoparticle Size

There have been numerous types of stabilizers that have been used as capping agents for platinum and palladium nanoparticles. Some common types of stabilizers include block copolymers [3134], dendrimers [3538], polymers [3941], etc. It is well known that the type of stabilizer that is used to cap the nanoparticles affects the stability of the nanoparticles and in turn affects its catalytic activity. As a result, it would be interesting to compare the catalytic activity of nanoparticles capped with different stabilizers for catalyzing the same reaction to determine the best stabilizer to use for that particular reaction.

Li and El-Sayed have compared the effect of using PVP, polystyrene-b-poly(sodium acrylate) block copolymer, and PAMAM-OH dendrimers of different generations as stabilizers for catalyzing the Suzuki cross-coupling reaction between phenylboronic acid and iodobenzene to form biphenyl [42]. HPLC was used to monitor the kinetics of the Suzuki reaction. In the case of dendrimer-Pd nanoparticles of different generations, it was determined that Generation 3 dendrimer-Pd nanoparticles are efficient catalysts. Generation 2 dendrimers were not found to be effective stabilizers since the nanoparticles precipitated out of solution after the Suzuki reaction. In the case of Generation 4 dendrimer-Pd nanoparticles, the catalytic activity was lower than that of Generation 3 dendrimer-Pd nanoparticles since higher generation dendrimers encapsulate the nanoparticles more strongly, resulting in a smaller amount of free sites available for catalysis. The catalytic activity of the block copolymer-Pd nanoparticles was found to be comparable to that of the PVP-Pd nanoparticles. Overall, PAMAM-OH Generation 3 dendrimer-Pd nanoparticles, PVP-Pd nanoparticles, and block-copolymer-Pd nanoparticles were all found to be efficient catalysts for the Suzuki reaction between phenylboronic acid and iodobenzene.

Narayanan and El-Sayed compared the stability of PVP-Pd and PAMAM-OH Generation 4 dendrimer-Pd nanoparticles after catalysis of the Suzuki cross-coupling reaction [43]. We found that when PAMAM-OH Generation 4 dendrimer-Pd nanoparticles are used as catalysts for the Suzuki cross-coupling reaction, the size of the nanoparticles continued to grow larger even during the second cycle of the Suzuki reaction as can be seen in Fig. 3. In the case of the PVP-Pd nanoparticles as catalysts, the nanoparticles precipitated and aggregated out of solution after the second cycle of the reaction and can be seen visually at the bottom of the flask. This result shows that the PAMAM-OH Generation 4 dendrimer-Pd nanoparticles are more stable than the PVP-Pd nanoparticles since higher generation dendrimers strongly encapsulate the nanoparticles. As a result, the PAMAM-OH Generation 4 dendrimer-Pd nanoparticles can undergo a greater amount of cycles of the Suzuki reaction before they finally aggregate and precipitate out of solution.
https://static-content.springer.com/image/art%3A10.1007%2Fs11244-007-9029-0/MediaObjects/11244_2007_9029_Fig3_HTML.gif
Fig. 3

TEM images and Gaussian fits of the size distributions of the PAMAM-OH Generation 4 dendrimer-Pd nanoparticles [43] before the Suzuki reaction (a, b), after the first cycle of the Suzuki reaction (c, d), and after the second cycle of the Suzuki reaction (e, f)

Recycling Potential of Nanocatalysts

The recycling potential is another important aspect in the development of nanocatalysts since one wants catalysts that can be used several times before losing its stability and catalytic activity. There have been several studies in the literature investigating the recycling potential of various nanocatalysts. The recycling potential studies have been conducted for a variety of reactions such as the Suzuki reaction [35, 4446], Heck reaction [4648], hydrogenations [35, 46, 49, 50], and many other kinds of reactions.

We have investigated the recycling potential of several different kinds of nanocatalysts such as colloidal PVP-Pd nanoparticles [29], colloidal dendrimer-Pd nanoparticles [43], and PVP-Pd nanoparticles supported on activated carbon [51]. The colloidal PVP-Pd nanoparticles were found to have poor recycling potential since the biphenyl yield during the second cycle was very low [29]. This is logical since it was also observed that after the second cycle of the Suzuki reaction, there were only small nanoparticles left in solution since the larger nanoparticles that formed during the first cycle had aggregated and precipitated out of solution. It was observed that the dendrimer-Pd nanoparticles had better recycling potential compared to the PVP-Pd nanoparticles and this is because the PAMAM-OH Generation 4 dendrimer strongly encapsulates the nanoparticles [43].

In the case of the PVP-Pd nanoparticles supported on activated carbon, we observed that the nanoparticles could be recycled for five consecutive cycles of the Suzuki cross-coupling reaction between phenylboronic acid and iodobenzene to form biphenyl [51]. The high recycling potential could be due to the presence of a large amount of the carbon support around the palladium nanoparticles, which further stabilizes the particles. As a result, the active sites that are available for catalysis are more stable.

Shape Dependent Nanocatalysis and the Effect of the Catalytic Process on the Nanoparticle Shape

The synthesis of tetrahedral and cubic shaped platinum nanoparticles was first reported in Science in 1996 by the El-Sayed group [52]. The synthesis involved the hydrogen reduction method and uses the potassium tetrachloroplatinate as the precursor platinum salt and sodium polyacrylate as the polymeric capping agent. Cubic platinum nanoparticles have also been synthesized using oxalate as the capping agent [53]. In addition, tetrahedral platinum nanoparticles have also been synthesized using PVP as the capping agent [54]. An RNA-mediated growth of hexagonal shaped platinum nanoparticles has been demonstrated [55]. Also, snow-like platinum nanoparticles have been synthesized for the first time [56].

There have been very few reports on the use of shaped metal nanoparticles as catalysts for reactions. Li and El-Sayed have used truncated octahedral platinum nanoparticles as catalysts for the electron transfer reaction [57]. Oxalate-stabilized cubic platinum nanoparticles have been used as catalysts for the decomposition of oxalate [53]. Prior to our work, there has not been any systematic comparison of the catalytic activity of different shapes of metal nanoparticles.

Narayanan and El-Sayed have synthesized tetrahedral, cubic, and spherical shaped platinum nanoparticles and used them as catalysts for the electron transfer reaction between hexacyanoferrate (III) ions and thiosulfate ions to form hexacyanoferrate (II) ions and tetrathionate ions [58]. The activation energy was calculated from the Arrhenius plot obtained by following the kinetics of the reaction via absorbance spectroscopy at four different temperatures. It was determined that the tetrahedral platinum nanoparticles are the most catalytically active since they have the lowest activation energy. In the case of the cubic platinum nanoparticles, the activation energy is the highest and as result, are the least catalytically active. The spherical platinum nanoparticles are “near spherical” in shape and are found to have an intermediate activation energy and as a result its catalytic activity is intermediate to that of the tetrahedral and cubic platinum nanoparticles.

We also found that the catalytic activity is correlated with the fraction of surface atoms located on the corners and edges and this relationship [59] is illustrated in Fig. 4. As a result, the tetrahedral platinum nanoparticles have the highest fraction of surface atoms on its corners and edges and is the most catalytically active. The cubic platinum nanoparticles have the lowest fraction of surface atoms on its corners and edges and are the least catalytically active. The spherical nanoparticles are treated as “near spherical” ones and based on the cubo-octahedral model, it is determined that they have an intermediate fraction of surface atoms on its corners and edges and their catalytic activity is also intermediate to that of the tetrahedral and cubic platinum nanoparticles.
https://static-content.springer.com/image/art%3A10.1007%2Fs11244-007-9029-0/MediaObjects/11244_2007_9029_Fig4_HTML.gif
Fig. 4

Correlation of the fraction of surface atoms on the corners and edges with the catalytic activity of the platinum nanoparticles [59]. Tetrahedral platinum nanoparticles have the highest fraction of surface atoms on its corners and edges and are the most catalytically active

It is worth noting that the dependence of the nanoparticle shape on the catalytic activity is observed during the very early stages of the electron transfer reaction (first 40 min of the reaction). During the entire course of the reaction (2 days), it is observed that distortions in the corners and edges of both the tetrahedral and cubic nanoparticles take place [60]. The rate of shape change was found to be faster for the tetrahedral nanoparticles compared to the cubic platinum nanoparticles [61]. In addition, the activation energies for both kinds of nanoparticles strive toward that of the spherical nanoparticles [60]. This is logical since the spherical nanoparticles are the most thermodynamically stable ones.

Effect of Individual Reactants on the Nanoparticle Size and Shape

In order to fully understand what causes the changes in the size and shape of the palladium and platinum nanoparticles during the course of their catalytic function, a detailed examination on the effect of the individual reactants on the size and shape of the metal nanoparticles is conducted. We carried out these kinds of studies for both the Suzuki cross-coupling reaction between phenylboronic acid and iodobenzene to form biphenyl as well as the electron transfer reaction between hexacyanoferrate (III) ions and thiosulfate ions to form hexacyanoferrate (II) ions and tetrathionate ions.

In the case of the Suzuki reaction, we determined that in the presence of phenylboronic acid alone, the size of both the spherical PVP-Pd and spherical dendrimer-Pd nanoparticles does not change much [29]. In the case of tetrahedral PVP-Pt nanoparticles as catalysts, it is observed that their shape is maintained in the presence of phenylboronic acid [30]. However, in the presence of iodobenzene, the spherical palladium nanoparticles grow larger in size due to the Ostwald ripening processes and the tetrahedral platinum nanoparticles undergo a shape transformation into spherical platinum nanoparticles. Based on the results of these studies, we proposed that phenylboronic acid binds to the nanoparticle surface via the O-group present in the deprotonated form due to the presence of sodium acetate base in the reaction conditions.

We have also obtained spectroscopic evidence for this binding process via infrared spectroscopy. From the FTIR studies, we have shown that the palladium nanoparticles bind to the phenylboronate anion (deprotonated form of phenylboronic acid in the presence of the sodium acetate base) via a bridged binding mode [62]. It was observed that there is a shifted B–O band which occurs due to the phenylboronate anion binding to the palladium nanoparticles and there is no corresponding free B–O band (unshifted band) indicating that the binding occurs via a bridged binding mode. Other evidence for the phenylboronate anion binding to the nanoparticle surface includes shifts in the B–C stretching mode and shifts in the out-of-plane phenyl C–C ring deformation band. As a result, if catalysis takes place on the nanoparticle surface, the mechanism of the Suzuki reaction catalyzed with colloidal palladium nanoparticles involves the phenylboronate anion binding to the nanoparticle surface and reacting with iodobenzene in solution.

In the case of the electron transfer reaction, we have observed that there are no changes in the size or shape of the platinum nanoparticles in the presence of thiosulfate ions alone [63]. This suggests that perhaps the thiosulfate ions bind to the nanoparticle surface and acts as a capping agent. When the nanoparticles are exposed to the hexacyanoferrate (III) ions alone, it is observed that there is a reduction in the size of spherical platinum nanoparticles and there are distortions in the corners and edges of the tetrahedral and cubic platinum nanoparticles. As a result, it is possible that the hexacyanoferrate ions attack platinum atoms on the nanoparticle surface and dissolves them and then forms platinum complexes in solution via the CN ligand, that acts as a catalyst.

We have conducted Raman studies to obtain spectroscopic evidence for the mechanism of binding that takes place [64]. We have observed Raman bands associated with the Pt(CN)4−2 cyanide type complexes, which provides evidence that the hexacyanoferrate ions does form complexes with the dissolved platinum atoms that are in solution. In addition, shifts in the SS stretching frequency provide evidence for the thiosulfate ions binding to the nanoparticle surface via the S anion. As a result, the overall mechanism of the electron transfer reaction catalyzed with platinum nanoparticles could involve the thiosulfate ions binding to the nanoparticle surface and reacting with the hexacyanoferrate (III) ions in solution if catalysis occurs on the nanoparticle surface.

Conclusions

The nanocatalysis revolution is certainly a growing one with an exponential increase in the number of publications during the past decade. Many important aspects involved in the use of nanocatalysts include the effect of the size and shape on the catalytic activity, the effect of stabilizers on the catalytic activity, the effect of the reaction conditions on the stability of the nanoparticles after its catalytic function, the recycling potential of the catalysts, and the effect of individual reactants on the nanoparticle size and shape. It has been shown that catalytic activity does not always increase with decreasing size and changes in the nanoparticle size takes place after the catalytic process. Catalysis is shown to be shape-dependent during the early stages of a reaction, and during the entire course of the reaction, distortions in the shape take place and the activation energies strive toward that of spherical nanoparticles. Stabilizers which bind more strongly to the nanoparticle surface result in less active sites available for the catalytic process and there is a tradeoff between stable nanoparticles and highly active nanocatalysts. The recycling potential is found to be higher for nanoparticles which are capped with strong stabilizers. Based on these factors, new directions can be fashioned in the quest for designing new and better catalysts that are stable during the catalytic process, have high catalytic activity, and have good recycling potential.

Acknowledgments

We thank NSF (CHE #0554668) for funding. We also thank the Georgia Tech Electron Microscopy Center for the TEM facilities that we used to characterize our nanoparticles. We also thank Dr. Gary Schuster’s group for the use of their HPLC instrument for the kinetic studies.

Copyright information

© Springer Science+Business Media, LLC 2008