Better and faster design of light-emitting organic materials with machine learning and quantum computing

Over the past decade, organic luminescent materials have been recognized by academia and industry as promising components for lightweight, flexible and versatile optoelectronic devices such as OLED displays. However, finding adequately efficient materials is a challenge.

To address this challenge, a joint research team has developed a novel approach that combines a machine learning model with a quantum-classical computational molecular design to accelerate the discovery of efficient OLED emitters. This research was published May 17 in Intelligent calculationa partner scientific journal.

The optimal OLED emitter discovered by the authors using this “hybrid quantum-classical procedure” is a deuterated derivative of Alq and is highly efficient in emitting light and synthesizable.

Deuterated OLED emitters are organic materials in which the hydrogen atoms are replaced with deuterium atoms in the emitter molecules. While they have the potential to emit light very efficiently, the design of such deuterated OLED emitters poses a computational challenge. This challenge arises from the need to optimize the position of deuterium atoms in emitter molecules, which requires calculations to be performed from scratch.

The new workflow, which involves both a classical computer and a quantum computer, speeds up these calculations. First, quantum chemistry calculations are performed on a classical computer to obtain the “quantum efficiencies” of a set of deuterated Alq molecules. This data on the light emitting efficiencies of different molecules is used to create training and test datasets for building a machine learning model to predict the quantum efficiencies of various deuterated Alq molecules.

Next, the machine learning model is used to construct an energy function of the system, known as the Hamiltonian. Quantum optimization is then performed on a quantum computer using two quantum variational optimization algorithms: the quantum variational autosolver (VQE) and the quantum approximate optimization algorithm (QAQA) — to help machine learning discover molecules with efficiencies optimal quantum. During the quantum optimization process a synthetic constraint is introduced to ensure that the optimized molecule is synthesizable.

To improve the prediction accuracy on quantum devices, the authors adopted a noise-resistant technique called recursive variable probabilistic elimination (RPVE) and were able to ‘find the optimal deuterated molecule with extremely high precision using a quantum device’. Furthermore, they point out that combining this new noise-resistant technique with the two chosen quantum optimization algorithms could achieve a quantum advantage for computations on short-term quantum devices.

Overall, the authors expect their approach, which combines quantum chemistry, machine learning and quantum optimization, to create ‘new opportunities to generate and optimize key molecules for material computing’.