Table of Contents

Nanoscale Fabrication for Enhanced Properties D. R. Baer and Y. Liang Supported by PNNL Laboratory Directed Research and Development. This research project investigates the formation and properties of nanoscale functional structures with the ultimate objective of integrating active nanometer-sized components into "smart" microtechnology devices. The focus of this project is the creation of ordered nano-sized clusters with a potential for catalytic studies. During FY99, we investigated using a self-assembled process to form organized Pt nanoclusters on TiO2 surfaces. Compared to the lithography methods, self-assembly has the advantage of higher yield and lower cost. Our results show that Pt preferentially interacts with the (1 x 2) phase on TiO2 surfaces and forms organized strings of nanoclusters along the step edges of the (1 x 2) domains on TiO2 surfaces. This effort on nanoscale component development uses a range of new experimental tools recently made available in the EMSL. These tools enhance our ability to work in the nanometer dimension. We have demonstrated the general ability to create the desired types of surfaces, and form organized strings of Pt nanoclusters on these surfaces using the self-assembled process. Creation of TiO2 Surfaces with Different Domains TiO2 is an important catalyst support material. A flat well-defined surface is needed to enable detailed characterization of the properties of metal clusters attached to this surface. It is well established that the TiO2 (110) surface can form two different structures, i.e., (1 x 1) and (1 x 2). However, recent work using the scanning tunneling microscope (STM) has shown that these structures can exist in different domains on a single surface. Because of their different geometric and electronic structures, one would expect that these two domains exhibit different interactions with metal clusters (supported metal catalysts) and thus result in different catalyst formations. During FY98, we demonstrated that Rh grew along the step edges of TiO2(1 x 1) domain. One objective in FY99 was to create better defined steps and surface structures on TiO2 (110) surfaces. We were able to obtain well-characterized TiO2 (1 x 1) surfaces and surfaces with both (1 x 1) and (1 x 2) domains (as shown in Figure 3.1a). These were ideal substrates for examination of the interaction of Pt with these two different domains. We used STM and low energy electron diffraction (LEED) to examine atomic structures and symmetry of TiO2 (110) surfaces under different process conditions. Results showed that after annealing the TiO2 (110) surfaces at 800°C or higher temperatures for an extended period time, the surface underwent the following phase transitions: (1 x 1) phase ® mixed (1 x 1) and (1 x n) phases ® mixed (1 x 1) and (1 x 2) phases ® (1 x 2) phase. Here (1 x n) phase represents a TiO2 (110) surface covered by line defects with various atomic spacings. Figure 3.1b shows an STM image and the related LEED pattern of a TiO2(110) surface annealed at 850°C. While the LEED shows that the surface exhibits mixed (1 x 1) and (1 x 2) domains, the STM image reveals that (1 x 1) and (1 x 2) domains alternate on the surface with distinct boundaries between the two. Because different catalysts may have different interactions with these domains or their associated step edges, a significant implication of these alternating domains is that one could create an organized, multi-component catalyst system supported on TiO2(110) surfaces. Figure 3.1a. A 400 Å x 400 Å STM image showing a TiO2(110) surface consists of alternating (1 x 1) and (1 x 2) domains and (1 x 2) phases. Figure 3.1b. A LEED pattern showing a TiO2(110) surface has a mixed (1 x 1). Pt Nano-Clusters and Strings on TiO2 Surfaces We used vapor deposition method to create Pt nanoclusters on TiO2 surfaces. By controlling the amount of Pt deposited on the surface, we were able to control the average size of Pt clusters on the surfaces. Figure 3.2 is an STM image of Pt deposited on a TiO2(110) (1 x 1) surface. The result shows that Pt clusters are randomly distributed on the TiO2 (1 x 1) surface. Figure 3.2. A 400 Å x 400 Å STM image showing that Pt nanoclusters randomly distributed on the TiO2(110) (1 x 1) phase. In contrast, strings of Pt clusters along the step edges of TiO2 (1 x 2) domains were found on surfaces with alternating (1 x 1) and (1 x 2) domains, as shown in Figure 3.3. The fact that Pt clusters preferentially formed along the edges of (1 x 2) domains suggests a stronger interaction of Pt with the TiO2 (1 x 2) domain than (1 x 1) domain. Figure 3.3. A 400 Å x 400 Å STM image showing formation of strings of Pt nanoclusters along the step edges of (1 x 2) domains. In addition to the preferential growth of Pt nanoclusters along the steps of the (1 x 2) phase, under the same amount of Pt deposition, the average size and distribution of Pt clusters is smaller and narrower on the TiO2 (1 x 2) domain than that of the (1 x 1) domain. Figure 3.4 shows the size distribution of Pt clusters along the step edges of (1 x 2) domains. The average Pt cluster size is 1.4 nm as compared to ~2.5 nm on a (1 x 1) domain. Our current effort is to test the reactivity of Pt/TiO2 (1 x 1) and (1 x 2) using temperature program desorption method. Figure 3.4. Size distribution of Pt nanoclusters stabilized by the steps of (1 x 2) domains. William R. Wiley Environmental Molecular Sciences Laboratory Feedback: webmaster@emsl.pnl.gov Revised: June 12, 2001 Security & Privacy PNNL-13147