Spin-Flip Configuration Interaction for Strong Static Correlation in Quantum Electrodynamics

Braden M. Weight, Zheng Pei, Sergei Tretiak · 2026

Read Spin-Flip Configuration Interaction for Strong Static Correlation in Quantum Electrodynamics by Braden M. Weight, Zheng Pei, Sergei Tretiak (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-18228.txt.

Introduction

Weight, Pei, and Tretiak extend spin-flip configuration interaction to cavity quantum electrodynamics. Strong static correlation breaks single-determinant methods near bond breaking and conical intersections. Cavity photons add control but also new complexity for such systems. Their QED-SF-CIS method adds quantized photons to the spin-flip scheme. The double excitation space must be kept to describe singlet states with cavity photons, they show.

Catalog record

CREATORBraden M. Weight, Zheng Pei, Sergei Tretiak
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE52131
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.18228

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Spin-Flip Configuration Interaction for Strong Static Correlation in Quantum Electrodynamics Braden M. Weight, Zheng Pei, Sergei Tretiak Abstract In computational chemistry of molecular materials, strong static correlation effects appear when electronic states, often involving the ground state, become quasi-degenerate, as occurs, for example, in bond-breaking processes. Such situations present significant challenges for accurate theoretical treatment. In these regimes, many-body methods involving a single-determinant description, such as Hartree-Fock theory and its time-dependent extension, fail to reproduce the correct topology of the ground and excited state potential energy surfaces ( e.g. , near conical intersections). When strongly correlated electronic systems are further strongly coupled to a quantized radiation field within the framework of non-relativistic cavity quantum electrodynamics, an additional photonic degree of freedom introduces both new complexity and new opportunities to control. Excited cavity photons can modify bond-breaking processes and enable tunability of geometrical and spin-phase transitions, for instance, in organometallic complexes. To overcome this bottleneck, in this work, we extend the well-studied spin-flip configuration interaction singles (SF-CIS) approach to explicitly include quantized cavity photons leading to QED-SF-CIS method. We derive the spin-flip Hamiltonian and find that the double excitation subspace of the system (single with respect to electronic excitation) must be included in the configurations to properly describe singlet electronic states interacting with cavity photons. We then illustrate, through representative molecular examples, how cavity coupling can provide additional tunability in bond-breaking processes. We finally generalize this approach to include higher numbers of photonic excitations, which are required in the strong coupling regime. I Introduction Treatment of strongly correlated systems has long been at the forefront of molecular and solid state science. While such systems give rise to rich and novel physical phenomena, paradigmatic models, e.g. , Hubbard model and similar, are notoriously difficult to solve. Even for a moderate number of degrees of freedom, exact solutions quickly become intractable. For realistic systems containing 100s or 1000s of electrons, no attempt can be made to solve them exactly. This situation is further complicated when strongly interacting matter is coupled to a quantized electromagnetic field under strong light–matter coupling conditions, forming hybrid light–matter states known as polaritons. The inclusion of photonic degrees of freedom also requires strict quantum mechanical treatment and could yield increased (or decreased) electron-electron correlation mediated by the photonic degree of freedom. 17 ; 30 ; 21 ; 28 Electronic correlation in molecules is commonly classified into two well-accepted types, often referred to as static and dynamic correlations. In the case of static correlation (the main focus of this work), the correlation stems from nearly degenerate electronic configurations. Common examples for this type of correlation are bond dissociation models, e.g. , H 2 , LiF, F 2 , Cr 2 , etc., rotations, e.g. , ethylene dihedral twisting, and bond formation in chemical reactions, e.g. , Diels-Alder reactions. There are many tools available for simulating strongly correlated materials. The most obvious solution is to use multi-reference methods, such as MC-SCF ( e.g. , CASSCF) and its perturbative extensions ( e.g. , CASPT2). However, even for nominally single-reference methods, there are strategies to treat strong static correlation. One such approach is the so-called spin-flip method, 11 ; 24 ; 2 where the reference is calculated at an adjacent multiplicity ( e.g. , triplet) and spin-flipping excitations are performed to obtain a different multiplicity ( e.g. , singlet). This method has proven to be powerful in overcoming static correlation that is mediated by nearly degenerate electronic configurations of the same multiplicity because the ground and excited states are treated on the same footing ( i.e. , as excited states of a different reference). Therefore, the topology of the ground and excited states is immediately corrected. This feature is especially important in nonadiabatic dynamics, where conical intersections of S 0 /S 1 often pose unique challenges for simulation of no-radiative relaxation due to the incorrect topology near these conical intersections. II Methods II.1 Spin-Flip Configuration Interaction Singles (SF-CIS) In the realm of single reference methods for calculating electronically excited states, the CIS approach is an approximation of the full time-dependent Hartree-Fock (TDHF) equations in that the de-excitation matrix, often denoted as \bm{B} , is neglected. The TDHF equations may relate to the first order random phase approximation (RPA). For time-dependent density functional theory (TDDFT)-based response methods, the corresponding nomenclature is the Tamm-Dancoff Approximation (TDA), which serves as the TDDFT analogue of CIS. In either case, it is generally well-accepted that the CIS for TDHF (or TDA for TDDFT) is often preferable for a variety of practical reasons, such as reduced numerical cost, the alleviation of the triplet instability problem and real eigenvalues via the use of a Hermitian eigenvalue problem. 3 ; 33 The full CIS Hamiltonian can be written in spin-space matrix blocks as, \displaystyle\hat{H}_{\mathrm{CIS}}=\begin{bmatrix}0&0&0&0&0\\ 0&{\bf A}_{\alpha\alpha,\alpha\alpha}&{\bf A}_{\alpha\alpha,\beta\beta}&0&0\\ 0&{\bf A}_{\beta\beta,\alpha\alpha}&{\bf A}_{\alpha\alpha,\alpha\alpha}&0&0\\ 0&0&0&{\bf A}_{\alpha\beta,\alpha\beta}&0\\ 0&0&0&0&{\bf A}_{\beta\alpha,\beta\alpha}\end{bmatrix} (1) where the reference energy has already been subtracted from the diagonal elements as \hat{H}_{\mathrm{CIS}}-E_{\mathrm{HF}}\rightarrow\hat{H}_{\mathrm{CIS}} . From left to right (or top to bottom), the basis configurations (in spin blocks) of the matrix are written as |\mathrm{HF}\rangle , |\Phi^{\alpha}_{\alpha}\rangle , |\Phi^{\beta}_{\beta}\rangle , |\Phi^{\alpha}_{\beta}\rangle , |\Phi^{\beta}_{\alpha}\rangle , where |\Phi_{i_{\alpha}}^{a_{\beta}}\rangle\equiv\hat{c}^{\dagger}_{a_{\beta}}\hat{c}_{i_{\alpha}}|\mathrm{HF}\rangle is a Slater determinant constructed by a single substitution on the HF reference determinant of occupied orbital i_{\alpha} with spin \alpha by a virtual orbital a of spin \beta , denoted by the fermionic creation ( \hat{c}^{\dagger}_{a_{\beta}}) and annihilation ( \hat{c}_{i_{\alpha}} ) operators. The subsequent matrix elements (in chemist’s notation) of each block are written as, \displaystyle A_{i_{\alpha}a_{\alpha},j_{\alpha}b_{\alpha}}=(\epsilon_{a_{\alpha}}-\epsilon_{i_{\alpha}})\delta_{i_{\alpha}j_{\alpha}}\delta_{a_{\alpha}b_{\alpha}} \displaystyle~~~~~~~~~~~~~~~~+(i_{\alpha}a_{\alpha}|j_{\alpha}b_{\alpha})-(i_{\alpha}j_{\alpha}|a_{\alpha}b_{\alpha}) (2) \displaystyle A_{i_{\beta}a_{\beta},j_{\beta}b_{\beta}}=(\epsilon_{a_{\beta}}-\epsilon_{i_{\beta}})\delta_{i_{\beta}j_{\beta}}\delta_{a_{\beta}b_{\beta}} \displaystyle~~~~~~~~~~~~~~~~+(i_{\beta}a_{\beta}|j_{\beta}b_{\beta})-(i_{\beta}j_{\beta}|a_{\beta}b_{\beta}) (3) \displaystyle A_{i_{\alpha}a_{\alpha},j_{\beta}b_{\beta}}=(i_{\alpha}a_{\alpha}|j_{\beta}b_{\beta}) (4) \displaystyle A_{i_{\beta}a_{\beta},j_{\alpha}b_{\alpha}}=(i_{\beta}a_{\beta}|j_{\alpha}b_{\alpha}) (5) \displaystyle A_{i_{\alpha}a_{\beta},j_{\alpha}b_{\beta}}=(\epsilon_{a_{\beta}}-\epsilon_{i_{\alpha}})\delta_{i_{\alpha}j_{\alpha}}\delta_{a_{\beta}b_{\beta}} \displaystyle~~~~~~~~~~~~~~~~~~~~~~-(i_{\alpha}j_{\alpha}|a_{\beta}b_{\beta}) (6) \displaystyle A_{i_{\beta}a_{\alpha},j_{\beta}b_{\alpha}}=(\epsilon_{a_{\alpha}}-\epsilon_{i_{\beta}})\delta_{i_{\beta}j_{\beta}}\delta_{a_{\alpha}b_{\alpha}} \displaystyle~~~~~~~~~~~~~~~~~~~~~~-(i_{\beta}j_{\beta}|a_{\alpha}b_{\alpha}) (7) Note that in the CIS approximation of TDHF, the spin-flipping blocks {\bf A}_{\alpha\beta,\alpha\beta} and {\bf A}_{\beta\alpha,\beta\alpha} are decoupled. 6 Thus, this matrix block can be constructed and solved independently of all other blocks. Thus, the SF-CIS Hamiltonian is simply, \hat{H}^{\mathrm{SF-CIS}}_{\alpha\rightarrow\beta}={\bf A}_{\alpha\beta,\alpha\beta} (8) or \hat{H}^{\mathrm{SF-CIS}}_{\beta\rightarrow\alpha}={\bf A}_{\beta\alpha,\beta\alpha}. (9) II.2 Quantum Electrodynamics Hartree-Fock (QED-HF) The non-relativistic Pauli-Fierz Hamiltonian can be written as, \hat{H}_{\mathrm{PF}}=\hat{H}_{\mathrm{el}}+\omega_{\mathrm{c}}\hat{a}^{\dagger}\hat{a}+\sqrt{\frac{\omega_{\mathrm{c}}}{2}}\vec{\lambda}\cdot\vec{\hat{d}}(\hat{a}^{\dagger}+\hat{a})+\frac{1}{2}(\vec{\lambda}\cdot\vec{\hat{d}})^{2}, (10) where \hat{H}_{\mathrm{el}} is the standard electronic Hamiltonian, composed of the one- and two-body terms, in the Born-Oppenheimer approximation. The light-matter coupling strength is defined as \vec{\lambda}=\sqrt{\frac{1}{\epsilon\mathcal{V}}}~\vec{e}, (11) where \epsilon is the dielectric constant of the media, \mathcal{V} is the cavity mode volume, and \vec{e} is the electric polarization unit vector of the cavity mode. Often, one chooses the the coherent state basis for the cavity photons 8 , which shifts the molecular dipole operator by a constant via unitary transformation of the Hamiltonian with \hat{U}^{\mathrm{CS}}=e^{-\frac{\vec{\lambda}x}{\sqrt{2\omega_{\mathrm{c}}}}\cdot\langle\vec{\hat{d}}\rangle(\hat{a}^{\dagger}+\hat{a})} , taken to be the expectation value of the molecular dipole in the reference state ( i.e. , often HF or DFT ground state), resulting in \hat{H}_{\mathrm{PF}}^{\mathrm{CS}}=\hat{H}_{\mathrm{el}}+\omega_{\mathrm{c}}\hat{a}^{\dagger}\hat{a}+\sqrt{\frac{\omega_{\mathrm{c}}}{2}}\vec{\lambda}_{\mathrm{c}}\cdot\vec{\Delta}{\hat{d}}(\hat{a}^{\dagger}+\hat{a})+\frac{1}{2}\vec{\lambda}_{\mathrm{c}}^{2}\cdot(\Delta\vec{\hat{d}})^{2}, (12) where \Delta\vec{\hat{d}}=\vec{\hat{d}}-\langle\vec{\hat{d}}\rangle_{\mathrm{HF}} with \langle\vec{\hat{d}}\rangle_{\mathrm{HF}}\equiv\langle\mathrm{HF}|\vec{\hat{d}}|\mathrm{HF}\rangle . It should be noted that more sophisticated representations for the photonic degrees of freedom have been developed. 13 ; 18 ; 16 In the case of HF theory, the energy is then calculated by expectation value of a single determinant composed of the tensor product of the Slater determinant and the cavity vacuum state as, \displaystyle E_{\mathrm{QED-HF}}^{\mathrm{CS}} \displaystyle=\langle\mathrm{HF},0|\hat{H}_{\mathrm{PF}}^{\mathrm{CS}}|\mathrm{HF},0\rangle \displaystyle=\langle\mathrm{HF}|\hat{H}_{\mathrm{el}}|\mathrm{HF}\rangle\langle 0|0\rangle+\langle 0|\omega_{\mathrm{c}}\hat{a}^{\dagger}\hat{a}|0\rangle \displaystyle+\sqrt{\frac{\omega_{\mathrm{c}}}{2}}\vec{\lambda}\cdot\langle\mathrm{HF}|\Delta\vec{\hat{d}}|\mathrm{HF}\rangle\langle 0|\hat{a}^{\dagger}+\hat{a}|0\rangle \displaystyle+\frac{1}{2}\vec{\lambda}\cdot\langle\mathrm{HF}|(\Delta\vec{\hat{d}})^{2}|\mathrm{HF}\rangle\cdot\vec{\lambda}~\langle 0|0\rangle \displaystyle=E_{\mathrm{HF}}+E^{\mathrm{QED}}_{\mathrm{RHF}}. (13) with E^{\mathrm{QED}}_{\mathrm{RHF}}=\frac{1}{2}\vec{\lambda}\cdot\langle\mathrm{HF}|(\Delta\vec{\hat{d}})^{2}|\mathrm{HF}\rangle\cdot\vec{\lambda} is the residual dipole self-energy (DSE) term after the coherent state transformation. Note that the bilinear term does not explicitly appear in the

Read the complete text