Scientists uncover detailed structure of the aluminum oxide surface

Aluminum oxide (Al2O3), also referred to as alumina, corundum, sapphire, or ruby, is an exceptional insulator widely used across various applications such as electronic components, catalyst support material, and chemically resistant ceramics.

Understanding the precise surface atom arrangement is crucial for comprehending the chemical reactions occurring on this material, particularly in catalytic processes. While the atoms inside the material have a fixed arrangement leading to characteristic crystal shapes, the surface structure differs from the crystal interior.

Despite being a strongly insulating material, researchers from TU Wien and the University of Vienna have recently unraveled the complex structure of the Al2O3 surface, a puzzle that had eluded precise determination for over half a century. The research findings by Jan Balajka and Ulrike Diebold’s group have been published in the prestigious journal Science, marking a significant step forward in surface science.

The innovative use of noncontact atomic force microscopy (ncAFM) allowed the research team to meticulously analyze surface structures. This advanced method involves capturing high-resolution images by scanning a sharp tip mounted on a quartz tuning fork in close proximity to the surface. The frequency of the tuning fork adjusts as the tip interacts with surface atoms without making physical contact with the material.

“In a ncAFM image, one can see the location of atoms, but not their chemical identity,” explained Johanna Hütner, who performed the experiments. “We overcame the lack of chemical sensitivity by precisely controlling the tip. Attaching a single oxygen atom to the tip apex allowed us to distinguish between oxygen and aluminum atoms on the surface. The oxygen atom on the tip is repelled from other oxygen atoms at the surface and attracted to aluminum atoms of the Al2O3 surface. Mapping the local repulsion or attraction enabled us to visualize the chemical identity of each surface atom directly.”

The researchers discovered that the surface undergoes rearrangement, allowing aluminum atoms to penetrate and form chemical bonds with the deeper oxygen atoms, effectively stabilizing the structure. Surprisingly, the numerical ratio of aluminum to oxygen atoms remains unchanged.

To achieve this, machine learning methods were used to optimize the three-dimensional model of the aluminum oxide surface, with the primary challenge being to align the restructured surface with the underlying crystal.

“The structure is very complex, resulting in a vast number of possibilities for how the experimentally inaccessible atoms below the surface could be arranged. The state-of-the-art machine learning algorithms combined with conventional computational methods allowed us to examine numerous possibilities and create the stable three-dimensional model of the aluminum oxide surface,” states Andrea Conti, who carried out the computational modeling.

“Through the collaborative effort of experimental and computational research, we not only tackled a long-standing mystery by determining the detailed structure of this enigmatic insulator but also discovered structure design principles applicable to an entire class of materials. Our results pave the way for advancements in catalysis, material science, and other fields,” says Jan Balajka, who led the research.

Journal reference:

1. Johanna I. Hütner, Andrea Conti, David Kugler, Florian Mittendorfer, Georg Kresse, Michael Schmid, Ulrike Diebold, Jan Balajka. Stoichiometric reconstruction of the Al2O3(0001) surface. Science, 2024; DOI: 10.1126/science.adq4744