To develop new chemical products and medications it helps to know how the electronic characteristics of certain molecules are calculated as the structure determines chemical reactions and bonds. However, an exact calculation is, in the rule, not guaranteed. ¡°We are reliant on approximations,¡± explains Prof Jakob Kottmann, professor of quantum algorithms at the ΰµÂ¹ú¼Ê_ΰµÂ¹ú¼Ê1946$ÓéÀÖappÓÎÏ· of Augsburg.

Such estimates can be made through the simulation of molecules. ¡°Since such molecules are quantum systems, we hope that quantum computers will offer a natural approach for calculating their properties,¡± says Kottmann.

Improving calculation models

This is where the project ¡°Verifizierbare Variationelle Quantenalgorithmen¡± (VeriVaQ) comes in: ¡°We are developing methods and software that can be used to reliably assess how accurate the results of calculated approximations are,¡± explains project coordinator Kottmann, who was appointed to the ΰµÂ¹ú¼Ê_ΰµÂ¹ú¼Ê1946$ÓéÀÖappÓÎÏ· of Augsburg three years ago as part of the Bavarian High-Tech Agenda and is a member of the Centre for Advanced Analytics and Predictive Sciences (CAAPS) at the ΰµÂ¹ú¼Ê_ΰµÂ¹ú¼Ê1946$ÓéÀÖappÓÎÏ· of Augsburg.

¡°We want to find out which molecules can be precisely described, and which methods or molecules are associated with high levels of uncertainty,¡± explains Kottmann. This is crucial for refining calculation models and forms the basis for experimental approaches in chemistry and materials science. ¡°Our goal is to make algorithms for quantum computers more reliable,¡± says Kottmann.

Funding in the millions

Kottmann (ΰµÂ¹ú¼Ê_ΰµÂ¹ú¼Ê1946$ÓéÀÖappÓÎÏ· of Augsburg) is collaborating with Prof Mariami Gachechiladze's working group at TU Darmstadt and with Dr Mathias Winkel from the AI Research Team at the pharmaceutical company Merck in the ¡°VeriVaQ¡± project. The project aims to develop algorithms directly targeted at chemically relevant molecules.

For this, the researchers are relying on variational quantum computing. Variational algorithms compute an upper bound to the ground state energy of an electronic system through minimization. In combination with semidefinite programming techniques, a corresponding lower bound can be computed. This allows tackling the problem from both directions and arriving at a more precise estimate in the form of a confidence interval.

The project is being funded for three years with €1.3 million from the Federal Ministry of Research, Technology and Space (BMFTR) as part of the ¡°Applied Quantum Information Processing¡± funding programme.

cg