“In our work, we were motivated by the goal of rational catalyst design,” explains co-team leader Ping Liu. “The atomic structure of nanoparticles used in catalysis has a huge impact on how well they perform – in other words, how they can enhance desired reactions and suppress unwanted ones. In catalysis applications, these nanoparticles are grown on powders of a secondary material. So to design a catalyst for a target reaction, we first need to understand how to control the atomic-scale structure of one type of nanoparticle as it is grown on the surface of a different one.”

In their studies, Liu and colleagues used density functional theory (DFT)-based calculations to study how ceria (CeO2) grows on two facets of anatase titania (TiO2) – the low energy (101) facet and the higher energy (112) one. Recent experiments – done by another Brookhaven research team led by José Rodriguez – showed that a diverse array of ceria morphologies, ranging from isolated small clusters to lines and plates, and in some cases well-ordered 3D nanoparticles, could form.

“Why do these different structures grow?” asks Liu. “And what is important about the structure of the different TiO2 facets that affects ceria growth?”

Bottom-up approach calculations are novel

“Thanks to DFT, we identified several factors that affect ceria growth morphology,” says co-team leader Mark Hybertsen. “We think the most important effect on growth is the way the CeO2 and TiO2 interact with each other at the interface between them. Sometimes these optimized structures can form a ‘template’ that supports the growth of a nanoparticle, but in other cases they are not compatible with the ceria crystal structure and the growth terminates with a line or plate-like structure just one layer thick, no more.”

This bottom-up approach to understanding catalyst nanoparticle growth modes is novel, he adds, and advances our understanding of how these particles can be grown.

“In their experiments, our colleagues Rodriguez et al. deposited Ce atoms on the surface of the titania particles through a wet chemical process and then exposed the samples to an oxygen atmosphere at high temperatures,” explains Hybertsen. “We have now developed a bottom-up method to see how the cerium atoms and oxygen could combine to find minimum energy structures at each step along the way,” he tells nanotechweb.org.

Finding optimal structures

“To do this, we first added a single Ce atom on the target TiO2 facet. We then examined how an oxygen molecule adsorbed on that Ce atom and found that the two oxygen atoms then separated to form a compact CeO2 unit. As we repeated this procedure, we saw that the lowest energy scenario was when the Ce and O2 combined together to form CeO2 units on the surface. From there, we added new CeO2 units gradually to try and build up lines, plates and 3D structures.”

For each new structure, the team used the energy required as a guide to find the optimal structures. “In this way, diverse ceria nanostructures naturally emerged in a way that greatly depended on the detailed structure of the different facets of the anatase-TiO2 support.”

A step forward to improving catalysts

The new studies could be extended to look into other catalytically important structures on oxide support particles, says Liu. “For example, transition metal clusters are crucial in catalysis and understanding their atomic-scale structure is very important.”

The results may also point to new ways of growing catalytically active anatase-TiO2 films with a (112) free surface, for example, he adds.

“Our work is just a first step forward to improving catalysts,” says Liu. “From the experiments done by our colleagues, we know that adding platinum to ceria nanostructures on the titania support, for instance, makes an excellent catalyst for converting CO2 into something useful. But how does this work? With our computer models for the atomic-scale structure of ceria on titania, we can systematically add in Pt atoms to figure out where they want to sit on the ceria lattice. And we can answer questions like: do they remain isolated or form clusters?

“We aim to discover the best platinum structures here, find out how they make ceria so effective and then ideally find ways to improve them.”

The researchers report their work in Nano Letters DOI: 10.1021/acs.nanolett.6b04218.