Surface Science Approach to the Molecular Level Integration of the Principles in Heterogeneous, Homogeneous, and Enzymatic Catalysis
- 283 Downloads
Heterogeneous, homogeneous, and enzymatic catalysis have generally been treated and studied as three separate fields. However all three fields have many aspects that unify them, therefore it is useful to study catalysts from each field in similar manners. Heterogeneous catalysts have been studied extensively under reaction conditions to monitor dynamic changes that occur during catalytic reactions, their atomic and molecular structure, and composition and oxidation state with high spatial and time resolution. The techniques used to monitor these catalysts include sum frequency generation vibrational spectroscopy, high pressure scanning tunneling microscopy, and ambient pressure X-ray photoelectron spectroscopy. In order to use these techniques to study homogeneous catalysts and enzymes under reaction conditions, we have heterogenized homogeneous catalysts by encapsulating small metal clusters in dendrimers and immobilized enzymes through the use of DNA tethers. By studying all three fields under reaction conditions with the same techniques we aim to show that heterogeneous, homogeneous, and enzymatic catalysts all behave similarly at the molecular level.
KeywordsCatalysis Heterogeneous Homogeneous Enzyme Nanoparticles Surface chemistry
Catalysis has long been separated into three distinct fields—heterogeneous, homogeneous, and enzymatic—all studied independently. Heterogeneous catalysis operates with the catalyst in a different phase than the reactants and products (e.g. solid–liquid, solid–gas); homogeneous catalysis has both catalyst and reactants/products in the same phase (almost exclusively liquid); enzymatic catalysis uses the active site of proteins to carry out the catalysis needed for biological systems.
Instead of treating each field as its own distinct discipline, we aim to unify all three types of catalysis and show that they all behave similarly on the molecular scale. All three fields have the capability to carry out the same broad class of reactions. Such as the oxidation of primary alcohols where bimetallic Au–Pd nanoparticles supported on TiO2 (heterogeneous), a Cu–azo complex (homogeneous), and horse liver alcohol dehydrogenase (enzyme) all have an affinity for the oxidation of benzyl alcohol [1, 2, 3]. Other examples of catalysts from each field catalyzing the same reaction can be seen in Table 1.
Au–Pd bimetallic nanoparticles on TiO2
Zn based coordination network
Bulky palladium hydride complex
Bulky diarlyammonium arenesulfonate
Peptidyl prolyl cis–trans isomerase
In most cases, the active catalyst, whether heterogeneous, homogeneous, or enzyme, is under 10 nm in scale (Fig. 1). As such, all of these catalysts could be considered nanoparticles. Besides the connection in their sizes, it has also been shown that the oxidation states of noble metal nanoparticles, including Pt and Rh, increase with decreasing sizes [10, 11]. When the nanoparticles are sufficiently small, their oxidation states eventually approach that of the metal complexes, which are extensively used as catalysts in the homogeneous catalysis.
As in the case of heterogeneous catalysis, the development of colloidal chemistry provides enormous possibilities towards producing various metal and metal oxide nanoparticles. By carefully controlling the precursors, surfactants, solvents, reducing agents, and the reaction temperature used during the synthesis, a wide range of nanoparticles with well-defined sizes, shapes, and compositions could be obtained [12, 13, 14].
The synthesized nanoparticles, either deposited on a two-dimensional surface or encapsulated into a three-dimensional porous material, are utilized in heterogeneous catalysis (Fig. 2). Following this approach, heterogenizing homogeneous and enzyme catalysts are also actively sought after, which will be discussed in detail in later sections.
1.1 Characterization of Nanoparticle Catalysts Under Reaction Conditions
To truly understand these nanoparticle catalysts it is not enough to study the catalyst before and after the reaction, but to actually follow the catalyst while the reaction is ongoing. For example, it has been demonstrated repeatedly that the catalysts undergo structure reconstruction and changes in oxidation states upon contact with the reaction atmosphere, and that the active sites are often formed in-situ under reaction conditions [15, 16, 17, 18]. In addition, since the catalytic reactions often take place on the surface of the nanoparticle catalysts, characterization techniques equipped with superior surface sensitivities would be extremely powerful. Much of the work that has been done in studying catalysis under reaction conditions has focused on heterogeneous catalysis due to limitations of the spectroscopic and microscopic techniques used. These techniques include: sum frequency generation (SFG) vibrational spectroscopy, high pressure scanning tunneling microscopy (STM), ambient pressure X-ray photoelectron spectroscopy (AP-XPS), and nanodiode hot electron detection.
SFG is an inherently surface sensitive spectroscopic technique, making it a particularly good tool for studying adsorbed species under catalytic reaction conditions [19, 20, 21, 22]. SFG requires the spatial and temporal overlap of a fixed wavelength visible beam and a variable IR beam (Fig. 3a). When the IR frequency matches a vibrational mode of the surface there is a resulting outgoing beam that has a frequency equal to the sum of the frequencies of the two incoming beams (Fig. 3b). During ethylene hydrogenation reactions on Pt(111) we have observed several molecular species on the surface of the catalysts . These include ethylidyne bound perpendicularly to the surface and ethylene bound parallel to the surface via either π-bonded or di-σ-bonded (Fig. 3c). Knowing the orientation and bonding of these molecules makes it possible to determine, under reaction conditions, the molecular details of the mechanisms of these reactions.
Using high pressure STM it is seen that the species adsorbed on a solid catalyst are mobile, not just stuck in one active site [23, 24, 25]. Studies of cyclohexene hydrogenation on Pt(111) show streaky, diffuse STM images while the reaction is ongoing (Fig. 4a, b). However, upon the addition of carbon monoxide (which poisons any catalytic turnover) clear, ordered structures are seen (Fig. 4c, d). This suggests that when the catalyst is active and the reaction is ongoing the adsorbed species move faster than the tip of the STM (100 Å/ms) .
Ambient pressure XPS can be used to determine the surface composition and oxidation states of bimetallic nanoparticles under reaction relevant conditions rather than the ultra-high vacuum conditions typically needed for XPS (Fig. 5a). In the presence of an oxidizing gas (nitric oxide) 15 nm rhodium-palladium nanoparticles preferentially segregate rhodium to the surface, with ~ 94% of the Rh being in the oxide form. Upon the addition of a reducing gas (carbon monoxide) to mimic the reaction conditions, the surface composition becomes much more equal and about 76% of the remaining surface Rh are reduced to its metallic state. This trend continues through cycles of oxidizing and reducing conditions (Fig. 5b) .
Besides the characterization techniques that probe the surface structure and composition as well as the adsorbed species, the detection of hot electrons generated during exothermic catalytic reactions with nanodiodes grants access to the charge transfer process on the catalytic interfaces during the reaction [27, 28]. For example, the turnover rates of H2 and CO oxidation were shown to be directly correlated to the strength of the chemicurrent detected at the metal-support interface, which provides important insights on the essence of the strong metal-support interaction (SMSI) effect .
1.2 Heterogenization of Homogeneous and Enzymatic Catalysts
In order to be able to use these surface based techniques to study homogeneous and enzymatic catalysts we have immobilized catalysts from both fields onto solid supports. This includes small metal clusters in dendrimers supported onto mesoporous supports to study homogenous catalysis and enzymes tethered to glass slides to study enzymatic catalysis.
Metal clusters contain two to several hundred atoms. Theoretical work by Nørskov’s group suggests that above 560 atoms (~ 2.7 nm), the surface chemical activity of Au clusters transforms from a molecular to bulk behavior . For instance, below that size, finite-size effects are observable, and facets become small enough for the charge redistribution upon oxygen adsorption to start reaching the edge of the particle (Fig. 6a). Additionally, we have shown that the oxidation state of platinum nanoparticles changes with the size of the particle. As the particle goes from 1.5 to 0.8 nm, the platinum changes from being almost entirely metallic to being primarily oxidized  (Fig. 6b). These two studies show that the electronic structure of metal nanoparticles is dependent on the size of the particle.
We can synthesize the small nanoclusters in dendrimers, a class of uniform polymers with the connectivity of fractal trees and generally radial symmetry. The dense multivalency, shape persistence, and structural uniformity of dendrimers render them versatile scaffolds for the synthesis and stabilization of small nanoclusters. These dendrimer-encapsulated metal clusters (DEMCs) are then loaded into mesoporous silica to generate solid catalysts (Fig. 7). These catalysts can catalyze reactions that had been challenging for conventional heterogeneous catalysts (e.g. aldol reactions and π-bond activation). In addition, supported DEMCs also show high activities for typical heterogeneous reactions, including hydrogenation and alkane hydroisomerization, which are listed in Table 2 [17, 32, 33, 34, 35, 36, 37, 38]. Critically, studies confirmed that these supported DEMCs are truly heterogeneous and stable against leaching. Thus, we have heterogenized homogeneous catalysts, which can be defined as the modification of homogeneous catalysts to render them in a separable (solid) phase from the starting materials and products.
A list of reactions studied with supported DEMCs in our group
Dendrimer-encapsulated Au clusters are able to catalyze the Hayashi–Ito aldol reaction of methyl isocyanoacetate (MI) and benzaldehydes, a classic homogeneous Au(I)-catalyzed reaction . The Au clusters showed higher turnover rates than the homogeneous equivalents, yet they allowed for tunable selectivity to the products in the trans or cis configurations. When these Au clusters are loaded in mesoporous silica with relatively large pores, the selectivity to the trans product could be as high as 100%, while the cis product selectivity is higher when the pore size was limited. Another example is the ring-opening of cyclopropane derivatives under hydrogen, a model reaction of C–C bond activation . Dendrimer-encapsulated rhodium clusters are able to catalyze the ring-opening of phenylcyclopropane at room temperature with 100% selectivity to the linear product, while the Wilkinson’s catalyst, a classic homogeneous catalyst for this reaction, gives much lower turnover rates with a mixture of linear and branched products. The differences of activities and selectivity of the clusters and typical homogeneous catalysts are attributed to different reaction mechanisms.
To heterogenize enzymes, we have begun studying DNA directed immobilization (DDI) [39, 40, 41]. In this method a short (20–25 base) DNA strand is conjugated to an enzyme. The complementary strand is attached to a solid surface. The two strands can then hybridize with each other giving an immobilized enzyme system (Fig. 8a). Specifically, in the method we use, the enzyme is conjugated to the DNA strand at the N-terminus and no other locations. This allows for control in the orientation of the immobilized enzymes.
We have used DDI to immobilize aldolase onto glass slides . Aldolase is an enzyme central to the glycolysis pathway that catalyzes the C–C bond breaking reverse aldol reaction. Specifically aldolase catalyzes the conversion of fructose 1,6-bisphosphate into dihydroxyacetone phosphate and glyceraldehyde 3-phosphate (Fig. 8b). Upon immobilization aldolase retains its activity and there is a minimal activity coming from any non-specifically adsorbed enzyme (Fig. 8c). Additionally, the immobilized aldolase is easily reusable and is more stable than the free enzyme in solution through multiple cycles (Fig. 8d). The ordering of the enzymes achieved by this immobilization technique is required for studying the enzyme using SFG vibrational spectroscopy.
Now that we have shown the ability to heterogenize both homogeneous and enzymatic catalysts, we are working on using the surface techniques typically reserved for studying heterogeneous catalysis to understand these other fields. By using these same techniques to study catalysts on the molecular level from each field while they are catalyzing the same reactions, we hope to unify all three fields of catalysis.
The work shown in this article was supported by the Director, Office of Science, Office of Basic Energy Sciences, Chemical Sciences, Geosciences and Biosciences Division of the US Department of Energy under Contract DE-AC02-05CH11231.
- 7.Sejidov FT, Mansoori Y, Goodarzi N (2005) J Mol Catal A: Chem 240(1–2):186–190Google Scholar