Single atom catalysis is a relatively new form of heterogeneous catalysis. While single atom catalysts probably are already used in a lot of catalysis, their identification and characterization has only recently become common place. As we now have the ability to synthesis relatively pure systems consisting of single atoms and then to characterize them, there are many interesting questions that we can answer about them. In this work we will use a combination of several different types of characterizations such as kinetic measurements, diffuse reflectance infrared Fourier transform spectroscopy, extended x-ray absorption fine structure, and many more to better understand how single atoms react and how we can attempt to make such systems more active. The work here is primarily based around Ir single atoms and/or small clusters on three different supports MgAl2O4, TiO2, and CeO2. In each of these cases we attempt to understand how the Ir and the support catalytically oxidize CO into CO2 through a kinetic, and if possible, mechanistic study. Through these mechanistic studies we attempt to isolate the most important parameters of the catalyst so that we can create a more active catalyst. There are, of course, many different ways that we can use this information. The most obvious is by changing the catalyst support, but as the breadth of the research presented here will show, we can also optimize catalytic activity through using mixtures of single atoms with larger species as well as by changing the nuclearity of the said species, i.e., we can increase activity by controlling the size of the catalysts. However, in order to be able to control the activity in this way, we must 1) know how the size affects the activity and 2) know how the reaction conditions affect the size, i.e., we must establish the catalyst size is stable during reaction. Each of these topics are discussed to some extent here. Additionally, we also discuss how different sites of single atoms on the same support might differ and we show that we can create such different sites. On the whole, we have studied single atom and small cluster catalysis in many different directions based on systems of Ir for CO oxidation. This work is also performed with the intent to compare these Ir systems to similar systems of Rh, Pt, Pd, etc. However here we will only discuss the Ir pieces. / Doctor of Philosophy / In this work we study various properties of Ir single atom and subnanometer cluster catalysts for CO oxidation in hopes that we might be able to design a better catalyst with this information. A catalyst is a substance that facilitates a chemical reaction but is not consumed. For this work we will be considering the reaction of carbon monoxide (CO), which is a common pollutant and highly toxic gas, with O2 to create CO2, a much less dangerous pollutant. Our catalyst thus makes this reaction happen much faster and thus allows us to remove CO from exhaust streams, such as car exhaust, better. A single atom catalyst is a catalyst that is primarily a single atom on a metal oxide support. A subnanometer cluster catalyst is thus a catalyst that is smaller than one nanometer (0.00000004 inches). These are typically 10-20 atoms grouped together. This size is interesting as it is bigger than a single atom, but it is still much smaller than a classical catalyst nanoparticle and is thus controlled or dominated by different properties. In this work we will look at how different characteristics of the singe atom and cluster catalysts affect how good of a catalyst it is. The first is how the amount of single atoms and nanoparticles affect the overall activity of the catalyst. This study will tell us what the best mixture of single atoms is. The second study is how small clusters of Ir/MgAl2O4 react differently than single atoms and large nanoparticles. This tells us what the best size for Ir/MgAl2O4 catalysts are. The third study tells us how Ir/TiO2 single atom catalysts react which is useful when compared to Ir/MgAl2O4 and Ir/CeO2 (Chapter 7). The combination of single atom studies then allows us to make predictions on which supports (apart from Ir/MgAl2O4, Ir/TiO2, and Ir/CeO2) will be the best for CO oxidation. The fourth study compares different single atoms (all of Ir/TiO2) and shows how they behave differently, this is another possibility to increase the effectiveness of the catalyst. The fifth study discusses how different conditions affect the size of the Ir/TiO2 catalysts. Specifically, whether they exist as single atoms, subnanometer clusters, or larger clusters. All of these different studies represent another way that we can potentially increase catalytic activity and hopefully will allow our group, or another group to create even more active catalysts.

A range of computational and experimental techniques have been applied to the study of four metal cluster systems. Decorated rhodium clusters Rh n O m (N 2 O) + ( n = 4 − 8, m = 0 − 2) have been investigated both experimentally by IR-MPD and computationally using DFT. The effect of cluster size as well as oxygen coverage on the spectroscopy of the N 2 O bend are analyzed. The infrared-induced decomposition of N 2 O on Rh n O + m is observed on all cluster sizes, with marked differences as a function of size and oxygen coverage, particularly in the case of Rh 5 (N 2 O) + . The oxidation of CO was studied on the surface of small platinum cluster cations Pt n O + m ( n = 3 − 7, m = 2 , 4) by IR-MPD at 400 – 2100cm −1 . Spectroscopically, oxygen is found to be bound both dissociatively and molecularly on the cluster surface, while the CO band is found to red shift in cluster size, and blue shift with oxygen coverage. Oxidation of CO proceeds on all cluster sizes, with a constant branching ratio of 40% : 60% . DFT calculations identified key stationary points and barriers on the Pt 4 O 2 CO + reaction pathway. The one-colour Ta 2 photodissociation is studied by photoionization and VMI in the range 23500 – 24000cm −1 , finding clear evidence of a fragmentation process producing Ta , which is interpreted as fragmentation of cationic Ta + 2 at the two photon level. A majority of the observed channels produce either atomic ( Ta( 4 F 3/2 ) ) or cationic ( Ta + ( 5 F 1 ) ) ground state. An improved value for the dissociation energy D 0 ( Ta + 2 ) is obtained, in agreement with computational predictions. The anisotropies observed show weak evidence of a perpendicular transition being involved in the photodissociation process. Finally, the photodissociation dynamics of Cu 2 are studied by spectroscopy in the range 36000 – 38200cm −1 as well as VMI. Clear evidence for resonant photolysis of Cu 2 is obtained, as a result of both direct dissociation of the Cu + 2 2 Π ion state as well as dissociation of doubly excited Cu 2 states, which leads to a determination of dimer dissociation energies. Finally, the production of Cu + 2 is interpreted as evidence of photolysis of Cu 3 , from which a Cu 3 dissociation energy is derived.