State-averaged Monte Carlo configuration interaction applied to electronically excited states

Area of Science:

• Quantum Chemistry
• Computational Chemistry
• Electronic Structure Theory

Background:

• Calculating excited electronic states is crucial for understanding molecular properties and reactions.
• Traditional methods often struggle with efficiency and stability for excited state calculations.
• The full configuration interaction (FCI) method provides high accuracy but is computationally prohibitive for larger systems.

Purpose of the Study:

• To introduce and validate a new computational method, state-averaging Monte Carlo configuration interaction (SA-MCCI), for stable and efficient excited state calculations.
• To demonstrate the accuracy of SA-MCCI in reproducing potential energy curves, including crossings and avoided crossings, for various molecular systems.
• To assess the applicability of SA-MCCI for calculating vertical excitations and oscillator strengths in small organic molecules.

Main Methods:

• Implementation of state-averaging into the Monte Carlo configuration interaction (MCCI) method.
• Calculation of potential energy curves for H3, LiF, and CH2 using SA-MCCI.
• Comparison of SA-MCCI results with full configuration interaction (FCI) calculations and experimental data where available.
• Application of SA-MCCI to vertical excitations and oscillator strengths for small organic molecules, comparing with CASPT2 values.

Main Results:

• SA-MCCI accurately reproduces excited potential curves for H3, including ground state crossings, using a fraction of the FCI space.
• Potential curves for LiF calculated with SA-MCCI show excellent agreement with FCI results, correctly capturing the avoided crossing.
• SA-MCCI successfully identifies the seam of conical intersections for CH2 and accurately computes potential curves for its first three triplet states.
• Vertical excitation energies and oscillator strengths for small organic molecules calculated by SA-MCCI show good agreement with CASPT2 values.

Conclusions:

• State-averaging Monte Carlo configuration interaction (SA-MCCI) offers a stable and efficient approach for calculating excited electronic states.
• The method demonstrates high accuracy in reproducing key features of potential energy surfaces, such as crossings and conical intersections.
• SA-MCCI shows promise for accurate predictions of excitation energies and related properties in various molecular systems.