Contacts:
Supervisor(s):
Prof. Filippo Lipparini and Prof. Claudio Amovilli
Title of the PhD project:
Scalable approaches to static and dynamic electronic correlation.
Abstract of the PhD project:
An accurate theoretical description of a molecular system requires a rigorous treatment of both static and dynamic correlation. This is a far from trivial problem, especially due to the elevated – sometimes insurmountable – computational cost exhibited by accurate methods.
Handling static correlation requires a multireference description, which can be provided by the CASSCF method. During my PhD I will develop and implement new and efficient numerical realizations of such method by coupling a treatment of the molecular orbitals optimization based on the Cholesky Decomposition of the two-electron integrals with an approximate Full CI solver, such as the Density Matrix Renormalization Group (DMRG) Full CI solver. The reliability, numerical stability and performance of such an approach will be thoroughly tested and applied to the description of structures and properties of complex molecular systems in their ground and excited state.
CASSCF methods provide a reliable and qualitatively correct description of strongly correlated systems. To achieve a quantitative description, the treatment of dynamic correlation is further required. The strategy developed can be used as a starting point for further Quantum Monte Carlo calculations. In particular, the CASSCF orbitals and most significant determinants in the CI expansion, together with an explicitly correlated Jastrow function, can provide a guess for a Variational Quantum Monte Carlo (VMC) calculation. The latter, can be further refined using Diffusion Monte Carlo (DMC).
A paramount aspect of this PhD project will be the development of scalable codes, that can be deployed on modern parallel computers, such as the ones present in national computing centers (e.g., CINECA). The combination of Cholesky Decomposition, DMRG, and QMC is well suited for such a purpose, as it allows to write all parts of the code in a way that can be distributed among many CPUs.
Publications:
• Nottoli, T., Giannì, I., Levitt, A. & Lipparini, F. A robust, open-source implementation of the locally optimal block preconditioned conjugate gradient for large eigenvalue problems in quantum chemistry. Theoretical chemistry accounts (2023)
• Alessandro, R., Giannì, I., Pes, F., Nottoli, T. & Lipparini, F. Linear Response Equations Revisited: A Simple and Efficient Iterative Algorithm. Journal of Chemical Theory and Computation (2023)
Oral communications at conferences:
...
Poster communications at conferences:
• Giannì I., Greiner J., Nottoli T., Eriksen J., Lipparini F. Combining CASSCF with MBE-FCI to treat large active spaces - Chemistry for the Future International Conference; Università di Pisa; Dipartimento di Chimica e Chimica Industriale, Via Giuseppe Moruzzi, 13 – Pisa, June 28 – June 30, 2023
• Giannì I., Greiner J., Nottoli T., Eriksen J., Lipparini F. Combining CASSCF with MBE-FCI to treat large active spaces - VIII Congresso Nazionale della Divisione di Chimica Teorica e Computazionale della SCI, Scuola Superiore Normale, Pisa, September 20 – September 22, 2023
Other achievements:
...
