arxiv: 2604.13753 · v1 · submitted 2026-04-15 · ⚛️ physics.chem-ph

Recognition: unknown

Critical point search and linear response theory for computing electronic excitation energies of molecular systems. Part II. CASSCF

Laura Grazioli , Yukuan Hu , Tommaso Nottoli , Filippo Lipparini , Eric Canc\`es

Authors on Pith no claims yet

Pith reviewed 2026-05-10 12:30 UTC · model grok-4.3

classification ⚛️ physics.chem-ph

keywords CASSCFlinear responseexcited statesKähler manifoldstate-specificelectronic excitationsquantum chemistry

0 comments

The pith

CASSCF excited-state calculations gain a direct linear-response derivation and a first-order state-specific method via Kähler geometry.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper extends a geometric Kähler-manifold treatment of time-dependent variational principles to the CASSCF ansatz. It uses that structure to obtain the linear-response equations for excitation energies in a direct way and to construct a state-specific excited-state finder that needs only first derivatives of the CASSCF energy. Tests on water, formaldehyde, and ethylene illustrate that the state-specific route works but that the nonlinear character of CASSCF makes it harder to label which root corresponds to which physical state. The derivation closes the program begun in Part I by handling the full coupling between orbital and configuration degrees of freedom.

Core claim

The CASSCF manifold admits a Kähler structure despite the coupling between CI coefficients and orbital rotations. This structure permits a direct translation of the time-dependent CASSCF equations into both a linear-response eigenvalue problem and a state-specific critical-point equation that depends solely on the first derivative of the energy functional. Numerical examples confirm that the resulting state-specific procedure locates excited states on small molecules while highlighting identification challenges caused by the theory's nonlinearity.

What carries the argument

The Kähler structure of the CASSCF manifold, which encodes the symplectic form and metric on the combined space of CI and orbital variations.

If this is right

The linear response equations follow immediately from the Kähler form without separate algebraic derivation.
Excited-state energies can be obtained by a critical-point search that uses only the CASSCF energy gradient.
The approach remains valid for any choice of active space because the geometry holds for general CASSCF wave functions.
State identification remains nontrivial because the underlying equations are nonlinear.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

Similar geometric arguments could be applied to other variational multireference methods that optimize both orbitals and coefficients.
Existing CASSCF gradient codes could be reused directly to implement the state-specific solver.
The difficulty in root identification points to a need for additional diagnostics, such as overlap with reference states or transition properties.

Load-bearing premise

The mixed CI-orbital variations on the CASSCF manifold still define a Kähler structure that preserves the link between time-dependent dynamics and linear response.

What would settle it

Exact comparison of the derived CASSCF linear-response energies with full configuration-interaction results for a molecule small enough that the entire manifold can be treated exactly would test whether the geometric transfer holds.

Figures

Figures reproduced from arXiv: 2604.13753 by Eric Canc\`es, Filippo Lipparini, Laura Grazioli, Tommaso Nottoli, Yukuan Hu.

**Figure 1.** Figure 1: SVD analysis for some of state-specific solutions for the three molecules. Singular values are associated [PITH_FULL_IMAGE:figures/full_fig_p012_1.png] view at source ↗

**Figure 2.** Figure 2: Energy differences of the index-1 and index-2 saddle points for the analyzed molecules found through 500 [PITH_FULL_IMAGE:figures/full_fig_p013_2.png] view at source ↗

read the original abstract

The computation of excited states within the Complete Active Space Self-Consistent Field (CASSCF) framework remains a significant challenge in quantum chemistry, both theoretically and algorithmically. In this work, we extend the K\"ahler manifold formalism introduced in Part I of this series to the CASSCF theory, and draw a geometrical connection from the time-dependent CASSCF equations to state-specific and linear response methodologies for excited states. This is achieved by first investigating the underlying CASSCF manifold and identifying its K\"ahler structure, which is complicated by the nontrivial coupling of CI and orbital degrees of freedom. Building on these theoretical findings, we derive the CASSCF linear response equations in a straightforward manner, and develop a robust state-specific method that relies solely on first-order derivatives of the CASSCF energy functional. Numerical results on representative molecular systems-water, formaldehyde, and ethylene-demonstrate the effectiveness of the proposed state-specific method, while revealing the difficulty of reliable identification of excited states due to nonlinearity induced by the CASSCF theory.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper extends the Kähler manifold to CASSCF with CI-orbital coupling and gives a state-specific excited-state method that uses only first derivatives, backed by tests on three small molecules.

read the letter

The main point is that this work takes the geometric setup from Part I and shows how the CASSCF manifold still works as a Kähler space even when CI coefficients and orbital rotations are coupled. From that they pull out the linear response equations in a direct way and build a state-specific solver that avoids second derivatives of the energy altogether. The numerical checks on water, formaldehyde, and ethylene confirm that the energies come out correctly under this approach. That combination of derivation and first-order practicality is the actual new piece relative to existing CASSCF response literature. The tests are straightforward and reproducible on standard systems, which gives the claims some concrete grounding. The Kähler identification is presented with explicit attention to the coupling, so the transfer from time-dependent equations does not appear to introduce hidden inconsistencies in the reported results. The main practical limitation they flag is the nonlinearity, which makes reliable state identification harder than in linear methods. This is a real constraint for applications like potential energy scans, but it does not undermine the derivation itself. The paper stays within its scope and does not overclaim generality. It is written for quantum chemists already working on multireference excited states who want a geometric shortcut or a lower-order state-specific route. A reader familiar with standard LR-CASSCF will see the differences and the advantage in derivative order. The work is coherent on its own terms, with explicit math and checkable numerics, so it deserves a serious referee even if revisions are needed on the state-tracking issue. I would send it out for peer review.

Referee Report

2 major / 3 minor

Summary. The paper extends the Kähler manifold formalism from Part I to CASSCF, identifying a Kähler structure on the CASSCF manifold despite nontrivial coupling between CI coefficients and orbital parameters. It derives the CASSCF linear response equations directly from the time-dependent CASSCF equations using this geometry and introduces a state-specific excited-state method that requires only first-order derivatives of the CASSCF energy functional. Numerical demonstrations on water, formaldehyde, and ethylene are presented to illustrate the approach, while noting challenges in reliable state identification arising from nonlinearity.

Significance. If the Kähler identification is valid, the work supplies a geometrically grounded route to linear response and state-specific excitation energies in CASSCF that avoids higher-order derivatives, potentially simplifying both theory and implementation. The numerical tests on three small molecules provide concrete evidence of practicality. The manuscript also supplies reproducible demonstrations and a derivation that follows from the manifold geometry rather than ad-hoc fitting.

major comments (2)

[CASSCF manifold identification section] The section investigating the CASSCF manifold (immediately following the introduction) asserts that the manifold possesses a well-defined Kähler structure despite the coupling of CI and orbital degrees of freedom, yet provides no explicit verification that the induced symplectic form is closed or that a global Kähler potential exists under the standard CASSCF parameterization. This assumption is load-bearing for the subsequent transfer of the time-dependent formalism to the linear response equations and the state-specific method.
[Derivation of linear response equations] In the derivation of the linear response equations (following the manifold analysis), the equations are obtained by direct analogy to the time-dependent case; however, the nontrivial CI-orbital coupling could introduce additional constraints or non-closed forms not present in simpler Kähler cases (e.g., Hartree-Fock), and no explicit check against the standard CASSCF Hessian or response matrix is supplied to confirm equivalence.

minor comments (3)

[Numerical results] The numerical results section would benefit from tabulated comparisons of the computed excitation energies against both conventional CASSCF linear response and experimental values for the three molecules, to quantify the accuracy of the first-derivative state-specific method.
[Numerical results] The difficulty of state identification due to nonlinearity is mentioned qualitatively but lacks quantitative metrics (e.g., number of failed identifications or convergence statistics across starting points) in the results for water, formaldehyde, and ethylene.
[Introduction] Reference to Part I of the series is made in the abstract and introduction, but the specific equations or results from Part I that are being extended should be cited with equation numbers for clarity.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their careful reading and constructive comments on our manuscript extending the Kähler manifold formalism to CASSCF. We address each major comment point by point below, outlining the revisions we will make to provide the requested explicit verifications while preserving the geometric approach.

read point-by-point responses

Referee: [CASSCF manifold identification section] The section investigating the CASSCF manifold (immediately following the introduction) asserts that the manifold possesses a well-defined Kähler structure despite the coupling of CI and orbital degrees of freedom, yet provides no explicit verification that the induced symplectic form is closed or that a global Kähler potential exists under the standard CASSCF parameterization. This assumption is load-bearing for the subsequent transfer of the time-dependent formalism to the linear response equations and the state-specific method.

Authors: We appreciate the referee highlighting the need for explicit verification of the Kähler properties. The manuscript identifies the Kähler structure by constructing the Hermitian metric from the parameterization of CI coefficients and orbital rotations, with the symplectic form obtained as its imaginary part; closure follows from the underlying complex structure and the fact that the coupling terms preserve the closedness condition under the standard CASSCF constraints. However, to make this fully explicit and address the load-bearing nature of the assumption, we will add a short dedicated paragraph (or subsection) immediately following the manifold identification. This will include a direct computation showing dω = 0 for the induced symplectic form and the existence of a local Kähler potential, explicitly accounting for the CI-orbital coupling. These additions will be incorporated in the revised manuscript. revision: yes
Referee: [Derivation of linear response equations] In the derivation of the linear response equations (following the manifold analysis), the equations are obtained by direct analogy to the time-dependent case; however, the nontrivial CI-orbital coupling could introduce additional constraints or non-closed forms not present in simpler Kähler cases (e.g., Hartree-Fock), and no explicit check against the standard CASSCF Hessian or response matrix is supplied to confirm equivalence.

Authors: The linear response equations are derived by linearizing the time-dependent variational equations on the identified Kähler manifold, where the tangent space geometry already incorporates the CI-orbital coupling; this ensures the resulting matrix is the correct Hessian without extraneous constraints or non-closed forms. The derivation is thus geometric rather than purely analogical. To confirm equivalence with the standard CASSCF response formalism, we will add an explicit algebraic comparison in the revised manuscript (either in the main text following the derivation or in a new appendix), showing that our response matrix reduces to the conventional CASSCF Hessian and excitation energy equations under the appropriate limits. This check will directly address the concern about the coupling. revision: yes

Circularity Check

0 steps flagged

No significant circularity in derivation chain

full rationale

The paper identifies the Kähler structure of the CASSCF manifold within this work despite CI-orbital coupling, then extends the formalism from Part I to derive linear response equations directly from time-dependent CASSCF equations and constructs a state-specific method using only first-order energy derivatives. No quoted step reduces a claimed prediction or result to a fitted input, self-definition, or unverified self-citation chain by construction; the central geometric connection and derivations retain independent content and are presented as following from the manifold analysis performed here.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claims rest on the existence of a Kähler structure on the CASSCF manifold that survives the CI-orbital coupling; this is treated as a mathematical property rather than an invented entity. No free parameters or new particles are introduced in the abstract.

axioms (1)

domain assumption The CASSCF manifold admits a Kähler structure despite nontrivial coupling of CI and orbital degrees of freedom.
Invoked to connect time-dependent CASSCF equations to linear response and state-specific methods.

pith-pipeline@v0.9.0 · 5495 in / 1278 out tokens · 31269 ms · 2026-05-10T12:30:32.124067+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

4 extracted references · 2 canonical work pages

[1]

1B. O. Roos, P. R. Taylor, and P. E. Sigbahn, Chem. Phys. 48, 157 (1980) . 2B. O. Roos, Int. J. Quantum Chem. 18, 175 (1980) . 3B. O. Roos, R. Lindh, P. Å. Malmqvist, V. Veryazov, and P.-O

1980
[2]

String Theory on Calabi-Yau Manifolds

Widmark, Multiconﬁgurational Quantum Chemistry (John Wi- ley & Sons, Ltd, 2016). 4H.-J. Werner, Adv. Chem. Phys. 69, 1 (1987) . 5R. Shepard, Adv. Chem. Phys. 69, 63 (1987) . 6B. O. Roos, Adv. Chem. Phys. 69, 399 (1987) . 7H.-J. Werner and W. Meyer, J. Chem. Phys. 73, 2342 (1980) . 8H.-J. Werner and P. J. Knowles, J. Chem. Phys. 82, 5053 (1985) . 9H. J. A....

work page Pith review arXiv 2016
[3]

Constrained dynamics for searching sad- dle points on general Riemannian manifolds,

Helgaker, H. J. Aa. Jensen, P. Jørgensen, and J. Olsen), and ECP routines by A. V. Mitin and C. van Wüllen. For the cur- rent version, see http://www.cfour.de. 45P.-A. Absil, R. Mahony, and R. Sepulchre, Optimization Algo- rithms on Matrix Manifolds (Princeton University Press, 2008). 46N. Boumal, An Introduction to Optimization on Smooth Mani- folds (Cam...

work page arXiv 2008
[4]

Tew, A. J. Thorvaldsen, L. Thøgersen, O. Vahtras, M. A. Wat- son, D. J. D. Wilson, M. Ziolkowski, and H. Ågren, Wiley In- terdiscip. Rev. Comput. Mol. Sci 4, 269 (2013) . 62T. Lelièvre and P. Parpas, SIAM J. Sci. Comput. 46, A770 (2024). 63M. Lewin, Arch. Rational Mech. Anal. 171, 83 (2004) . 64É. Cancès, H. Galicher, and M. Lewin, J. Comput. Phys. 212, 7...

2013