Project

Defect chemistry and optical properties of lutetium aluminum garnet

Point defects play a central role in determining the physical and chemical properties of functional materials. Depending on their nature and concentration, they can enable desirable processes such as electrical and ionic transport or catalytic activity, while also causing detrimental effects including phase instability, segregation, and non-radiative recombination. Understanding defect formation, stability, interactions, and mobility is therefore essential for predicting and controlling material performance. In this project, these aspects are investigated in lutetium aluminum garnet (LuAG), an oxide of major technological relevance for optical applications, including scintillators and solid-state lasers. Energies of defect formation, defect migration and optical properties are calculated using density functional theory (DFT).

Project Details

Project term

September 1, 2025–August 31, 2026

Affiliations

RWTH Aachen University

Institute

Institute of Physical Chemistry

Principal Investigator

‎Dr. Steffen Neitzel-Grieshammer

Methods

All DFT calculations were performed using VASP within density functional theory (DFT), employing the PAW method and a plane-wave basis set. The conventional 160-atom LuAG unit cell was used with a 500 eV plane-wave cutoff. PBEsol was applied for structural relaxations, while the HSE06 hybrid functional was used for single point calculations where an improved description of the electronic structure is required.

In Subproject 1, formation energies of intrinsic vacancies, interstitials, and antisite defects were determined. Possible interstitial sites were screened based on structural criteria, and defect concentrations were estimated.

Subproject 2 addressed defect migration using NEB and CI-NEB calculations. Migration barriers were determined for vacancy and interstitial mechanisms.

In Subproject 3, machine-learning potentials (MLPs) were fitted to DFT molecular-dynamics data, validated against results from Subprojects 1 and 2, and subsequently employed in larger-scale simulations to identify migration mechanisms.

Subproject 4 investigated electronic and optical properties using PBEsol and HSE06 band-structure calculations and dielectric functions obtained from linear-response theory in the independent particle approach.

Results

The calculations using PBEsol and HSE06 reproduce the experimental crystal structure well, with lattice parameters deviating by only about 0.5%. As expected, PBEsol substantially underestimates the band gap, yielding 5.0 eV, whereas HSE06 predicts 6.9 eV, in much better agreement with the experimental range of 7–8 eV. The valence band is dominated by O-p states, while Lu-5d states make the largest contribution to the conduction band. The calculated optical response shows the onset of absorption below approximately 180 nm.

The defect calculations show that intrinsic defect formation in stoichiometric LuAG is generally energetically unfavorable, resulting in very low equilibrium defect concentrations. Among the considered Frenkel, Schottky, and antisite disorders, antisite defects have the lowest formation energies. In particular, Lu occupying octahedral Al sites is the most favorable antisite configuration. Even so, the calculated intrinsic antisite concentration at 1000 K is only about 10e-17 / ccm. Under non-stoichiometric, Lu2O3- or Al2O3-rich conditions, formation of Lu-on-Al antisites becomes substantially more favorable, with calculated formation energies between about 0.9 and 1.3 eV.

The electronic-structure calculations indicate that antisite defects do not cause major changes in the density of states. In contrast, oxygen vacancies introduce an additional electronic state just below the conduction-band minimum, suggesting that these defects may act as electron traps and therefore potentially influence optical performance. No pronounced changes in the dielectric function were observed for the investigated defective cells.
NEB calculations reveal a strong contrast between cation and oxygen mobility. Cation migration barriers are generally above 1 eV, with barriers of 2.75 eV for Lu vacancy migration and 1.38–2.79 eV for Al vacancy migration. Oxygen defects are considerably more mobile: three-dimensional vacancy transport is possible with a rate-limiting barrier of 0.77 eV, while oxygen interstitial migration exhibits particularly low barriers of about 0.40 eV for both direct and concerted mechanisms. These values are comparable to, or even lower than, those of established oxide-ion conductors. Nevertheless, the low intrinsic defect concentrations limit overall ionic transport in undoped LuAG, indicating that suitable doping would be necessary to achieve significant ionic conductivity.

Discussion

DFT calculations describe the electronic and structural properties of LuAG well, especially using the hybrid HSE06 function.
The calculations reveal, that the concentration of intrinsic defects in LuAG is negligible, even at high temperatures. Considerable amounts of defects are only found under non-stoichiometric conditions or due to dopants/impurities.

Another important finding is the low migration barrier for oxide interstitials which would make the material a good ionic conductor if a considerable number of these defects is present. As the concentration of intrinsic defects is negligible, only doping could lead to a sufficient amount of interstitials. Nevertheless, this is a possible route to new oxide ion conductors which has to be pursued in future work.

A machine-learning potential was successfully trained and tested in molecular dynamics simulations. Nevertheless, further simulations and refitting of the potential have to be performed to ensure the quality of the results and further investigate the movement of ions in the lattice.

Further investigation of the effect of dopants on the optical and transport properties of the material have not been conducted due to time limitations. Results of this project have been submitted to Solid State Ionics for publication.

Additional Project Information

DFG classification: 302-02 Physical Chemistry of Solids and Surfaces, Material Characterisation
Software: VASP
Cluster: CLAIX

Publications

A first-principles study of intrinsic defect formation and migration in lutetium aluminium garnet,
Steffen Neitzel-Grieshammer,
Solid State Ionics, 2026