PhD in chemistry (M/F) – Chemically accurate excited states

The recruited PhD student will be involved in the below-described objectives ii, iii, and iv and will therefore apply a large panel of excited-state methods (TD-DFT, ADC(n), CCn, CAS-PT2, etc.) both to define reference values and to benchmark lower-order methods. In more details, the following tasks will be carried out in parallel:

1. The PhD student will extend the current database (see, J. Chem. Theory Comput.2025, 21, 8010−8033) by obtaining highly-accurate reference values for properties not yet available, including: a) oscillator strengths and excited-state dipole moments for both organic and inorganic compounds; b) core excitations and ionization-related energies for
diverse molecules; and c) potential energy surfaces for the few lowest electronic states of compact-size molecules and compounds representative of typical photochemical reactions. In each case, the PhD candidate will be in charge of defining the « best » approach for the target property and system. 

2. A second task, related to objective iii that will be tackled by the PhD student is to benchmark effective electronic structure methods for the above-described properties. In more details, this will require the development of an automated approach that allows identifying the excited-states across several levels of theory, which will greatly facilitate the subsequent benchmarks, especially at the TD-DFT level. Thanks to this, the PhD candidate will perform extensive electronic structure benchmarks considering a wide panel of methods, including less conventional schemes such as various flavours of BSE/GW, unitary CC ansatzes, orbital-optimized DFT, spin-flip schemes, etc.

3. The last aspect in which the PhD candidate will be involved is the communication of the results, a cornerstone of the project. This will be achieved by participating to the maintenance and development of the Git repository that will serve as a comprehensive platform for data extraction, analysis, and version control, providing access to highly curated
datasets in a machine learning-friendly format.

This PhD is part of the CARES project (Chemically Accurate Excited States), funded by the French National Agency (ANR). CARES performed in collaboration between Nantes and Toulouse’s theoretical teams is a quantum chemistry project aiming to provide a large and diverse database of chemically-accurate reference values for excited-state calculations. Building on the experience and collaboration of the two groups (QUEST database), the objectives of CARES are: i) to develop a new selected CI algorithm allowing reaching chemically-accurate results for large compounds; ii) to extend the currently-available database to properties relevant for both core spectroscopies and dynamic studies; iii) to benchmark a large panel of low-order popular and emerging methods including density-based and wavefunction-based theories; iv) to provide a free and user-friendly webinterface to download all key results, in particular the reference data.

• A recent Master’s degree (or equivalent) in Chemical Physics, Physical Chemistry, Theoretical Chemistry, Theoretical Physics, or a related discipline.
• Experience with recognized computational chemistry software (e.g., Gaussian, Dalton, TurboMole) is highly desirable.
• Experience in Time-Dependent Density Functional Theory (TD-DFT) and our wavefunction methods for excited states (CC, ADC, CAS, etc.) are desirable assets.
• Strong analytical and problem-solving skills, with a proactive and motivated mindset.
• Ability to work both independently and collaboratively within a research team.
• Excellent communication skills, both written and verbal.