Efficient Modelling of Anharmonicity and Quantum Effects in PdCuH$_2$ with Machine Learning Potentials

Francesco Belli,Eva Zurek

2024-06-19

Abstract:Quantum nuclear effects and anharmonicity impact a wide range of functional materials and their properties. One of the most powerful techniques to model these effects is the Stochastic Self-Consistent Harmonic Approximation (SSCHA). Unfortunately, the SSCHA is extremely computationally expensive, prohibiting its routine use. We propose a protocol that pairs machine learning interatomic potentials, which can be tailored for the system at hand via active learning, with the SSCHA. Our method leverages an upscaling procedure that allows for the treatment of supercells of up to thousands of atoms with practically minimal computational effort. The protocol is applied to PdCuH$_x$ ($x = 0-2$) compounds, chosen because previous experimental studies have reported superconducting critical temperatures, $T_\text{c}$s, as high as 17~K at ambient pressures in an unknown hydrogenated PdCu phase. We identify a $P4/mmm$ PdCuH$_2$ structure, which is shown to be dynamically stable only upon the inclusion of quantum fluctuations, as being a key contributor to the measured superconductivity. For this system, our methodology is able to reduce the computational expense for the SSCHA calculations by $\sim$96\%. The proposed protocol opens the door towards the routine inclusion of quantum nuclear motion and anharmonicity in materials discovery.

Superconductivity

What problem does this paper attempt to address?

The paper aims to address the efficient modeling of palladium-copper hydride (PdCuH₂) under the influence of quantum nuclear effects and anharmonicity. Specifically:

1. **Computational Cost Issue**: Traditional self-consistent harmonic approximation (SSCHA), while powerful, is extremely computationally expensive, limiting its widespread application. The paper proposes a new method that combines machine learning potentials (MLP) with SSCHA, significantly reducing computational costs through an active learning process.

2. **Material Properties Study**: The paper focuses particularly on the PdCuH₂ compound because previous studies have suggested that this material may have a superconducting critical temperature (Tc) as high as 17K at ambient pressure. The study finds that only when quantum fluctuations are considered does the P4/mmm phase of PdCuH₂ exhibit dynamic stability, which is crucial for its superconductivity.

3. **Structure Prediction**: Through crystal structure prediction (CSP) algorithms, the paper also discovers a new I41/amd phase of PdCuH₂, which has an energy approximately 18meV/atom lower than the previously known P4/mmm phase. This suggests that the new structure could be key to the superconductivity observed in actual experiments.

In summary, the main goal of the paper is to achieve efficient and accurate simulation of the palladium-copper hydride system by combining machine learning potentials with SSCHA, thereby better understanding its superconductivity and other physical properties.