betaFeedback

WorksPhysics and Astronomy

Dysfunctional Functionals and their Therapies in Time-Dependent Density Functional Theory

Abstract

The exchange-correlation functional in Time-Dependent Density Functional Theory (TDDFT) is known to have dependence on past densities and initial states, and yet this memory-dependence is ignored in practical calculations which almost always use an adiabatic approximation. We discuss the dysfunction that arises due to this deficiency and point to therapies that go beyond the adiabatic approximation. Still, adiabatic TDDFT has had success in a wide range of systems for response properties and for fully non-equilibrium dynamics, and why this is so is only partially understood. This, together with the rigorous foundation of the exact theory and its favorable system-size scaling, provides tremendous impetus for continuing to strive towards improved approximations to the exact functionals, and a deeper understanding of the impact of approximations in applications.

Introduction

A century after the publication of the Schrödinger equation, we may step back and marvel at the fundamental understanding and technological developments that the mathematical formulation of non-relativistic quantum mechanics has provided, with novel subfields and applications continuing to emerge today. In practise, the equation would be far less impactful without the formulation of reduced-variable theories: the description of even the smallest relevant systems typically involve many electrons and many nuclei that solving Schrödinger’s equation not only pushes computational limits but also the many-variable wavefunction gives us far more information than what we could possibly want. Considering the description of the electronic subsystem, density functional methods are the most widely-used, due to their balance between accuracy and efficiency while resting on a rigorous foundation. For matter interacting with light or otherwise disturbed from equilibrium, time-dependent density functional theory (TDDFT) was formulated almost 40 years ago [1, 2, 3], reviewed in more than ten reviews in the past ten years [4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16]. TDDFT is based on proving a one-to-one map between the time-evolving one-body density of a many-body system and the external potential, for a fixed initial wavefunction.

The proof itself is both a beauty and a beast: the beauty is that it is based on Heisenberg’s equation of motion for the current-density, making it relatively straightforward to understand after even one course in quantum mechanics. However, a beast hides inside because of implied time-analyticity properties that do not typically hold in the presence of spatial singularities such as the Coulomb potential. This has triggered the development of subsequent proofs attempting to avoid this requirement [17, 18, 19, 20, 21]; most notably the recent work of Ref. [21].

TDDFT operates by solving a set of coupled single-particle non-linear Schrödinger equations, mimicking the interacting electronic system by a non-interacting Kohn-Sham (KS) one, with all the quantum many-body effects hiding in the exchange-correlation (xc) potential, vxc[n;Ψ(0),Φ(0)](r,t)v_{\mathrm{xc}}[n;\Psi(0),\Phi(0)](\mathbf{r},t). At time tt and position r\mathbf{r}, this potential is a functional of the one-body density n(r′,t′)=N∑σ1′,…,σN′∫d3r2⋯d3rN∣Ψ(r′,σ1′,r2σ2′…rNσN′,t′)∣2n(\mathbf{r}',t') = N \sum_{\sigma'_1,\ldots,\sigma'_N} \int d^3\mathbf{r}_2\cdots d^3\mathbf{r}_N \lvert\Psi(\mathbf{r}',\sigma'_1,\mathbf{r}_2\sigma'_2\ldots\mathbf{r}_N\sigma'_N,t')\rvert^2 at all points in space r′\mathbf{r}' and all past times t′<tt' < t, the initial many-body interacting state Ψ(0)\Psi(0) and the initial KS state Φ(0)\Phi(0). The theorems state that with the use of the exact functionals, all observables of the true interacting electronic system can in principle be extracted from that of the KS system [1, 2, 3, 4, 16]. This dependence on the past (memory-dependence) is a price to pay for the reduced description, as is the non-linearity; the many-body Schrödinger equation is linear, and needs only the instantaneous wavefunction in order to propagate forward.

While the non-linearity itself has not appeared to pose any numerical problem, practical approximations that contain appropriate memory-dependence remain elusive. The usual approach is to completely neglect memory and use the xc potential for a ground state of the same instantaneous density, vxcgs[n(t)](r,t)v_{\mathrm{xc}}^{\mathrm{gs}}[n(t)](\mathbf{r},t), known as the adiabatic functional approximation. In this Perspective, we revisit both the success and the dysfunction that can arise from such a functional approximation, and the therapies that have been put forward to go beyond the adiabatic approximation.

Pleasure and Pain: Successes and Failures of Adiabatic TDDFT

Adiabatic TDDFT is now a central part of many codes (see Figure 1) where it has enabled the quantum mechanical study of complex and fascinating phenomena, that would be computationally impossible otherwise. In what is often called real-time TDDFT, the KS equations are propagated in real-time (TDKS):

Figure 1. Codes with TDDFT capability, either real-time TDKS propagation, or linear response, or both [35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71] have widened the range of scientists who can run TDDFT calculations, providing truly valuable tools for both experimentalists and coding-shy theorists.

(−∇22+vS(r,t))ϕi(r,t)=i∂tϕi(r,t)(1)\left(-\frac{\nabla^2}{2}+v_{\mathrm{S}}(\mathbf{r},t)\right)\phi_i(\mathbf{r},t)=i\partial_t\phi_i(\mathbf{r},t) \tag*{(1)}

where vS(r,t)=vext(0)(r)+vextapp(r,t)+vH[n](r,t)+vXC[n;Ψ(0),Φ(0)](r,t)v_{\mathrm{S}}(\mathbf{r},t)=v_{\mathrm{ext}}^{(0)}(\mathbf{r})+v_{\mathrm{ext}}^{\mathrm{app}}(\mathbf{r},t)+v_{\mathrm{H}}[n](\mathbf{r},t)+v_{\mathrm{XC}}[n;\Psi(0),\Phi(0)](\mathbf{r},t) is the sum of the static external potential due to nuclei, an applied, possibly time-dependent, potential such as a laser field, the Hartree potential representing the classical Coulomb potential of the density, and xc potential, typically approximated by an adiabatic approximation. The resulting instantaneous dependence of the Hamiltonian on the density n(r,t)=∑iN∣ϕi(r,t)∣2n(\mathbf{r},t)=\sum_{i}^{N}|\phi_i(\mathbf{r},t)|^2 simplifies the computational implementation of the equations compared to if a memory-dependent functional was used, allowing similar propagators to be used as for the linear TDSE, with the nonlinearity entering through predictor-corrector approach for example [22, 23, 24]. In the limit of weak perturbations around a ground state, usually Eq. (1) is instead cast in frequency-space through response equations derived from time-dependent perturbation theory. Linear response gives access to the excitation energies and properties related to the ground-to-excited transition densities of the system (e.g. oscillator strengths) [25, 26, 27].

Excited-state densities are available from considering time-independent perturbation theory of the excitation energies computed from linear response [28, 29, 30], which could equivalently be framed as a variational problem. Quadratic response has been formulated for properties related to transition-densities between excited states (e.g. as needed for their non-adiabatic couplings, hyperpolarizabilities, two-photon absorption) [31, 32, 33, 34].

In the following, we will discuss how the adiabatic approximation gives both pleasure and pain in these two regimes, considering first linear response and then non-perturbative.

Linear response regime

By far, the dominant applications have been in the linear response regime, to compute the optical spectra of molecules and materials, as well as in the computation of potential energy surfaces and their couplings in photochemical applications [25, 26]. Simulations treating an unprecedented number of electrons and excitations explicitly have been enabled, e.g. the green fluorescent protein complex of 4353 atoms with 40,518 atomic orbitals in Ref. [72]. High-energy-density materials have been studied, without the need for arbitrary partitionings [73]; for large systems, spectra and response properties can arguably be obtained most efficiently by real-time propagation under a weak perturbation, and the same propagation framework can also be used with stronger fields. Quadratic response TDDFT for molecules for excited state absorption has been recently benchmarked [74].

Turning to the role of memory in these calculations, let us consider first where the xc functional enters these calculations. The xc functional approximation involves vxc[n;Ψ(0),Φ(0)](r,t)v_{\mathrm{xc}}[n;\Psi(0),\Phi(0)](\mathbf{r},t) in a domain where the time-dependence in vextapp(r,t)v_{\mathrm{ext}}^{\mathrm{app}}(\mathbf{r},t) is weak such that only first-order responses to it are considered, and the initial states Ψ(0),Φ(0)\Psi(0),\Phi(0) are respectively the interacting and KS ground-state wavefunctions of vext(0)(r)v_{\mathrm{ext}}^{(0)}(\mathbf{r}) and the KS counterpart vs(0)(r)v_{\mathrm{s}}^{(0)}(\mathbf{r}) (and by Hohenberg-Kohn theorem [75] no explicit initial-state dependence is needed here). While this implies a single choice is needed for the approximation for vxc[n](r,t)v_{\mathrm{xc}}[n](\mathbf{r},t), it is important to realize that in practise the xc functional usually enters in two distinct places in response calculations. That is, in practise, one separates the ground-state calculation from the perturbative one, so that the xc functional appears first in the choice of the ground-state functional out of which the KS response is calculated (e.g. molecular orbitals and their orbital energies, or in the case of extended systems, the density-density response function of the KS ground state), and second in the dynamical response, embodied in the functional derivative in frequency-space fxc[ng.s.](r,r′,ω)f_{\mathrm{xc}}[n^{\mathrm{g.s.}}](\mathbf{r},\mathbf{r}',\omega). Formally, the two choices (vxcg.s.[ng.s.]v_{\mathrm{xc}}^{\mathrm{g.s.}}[n^{\mathrm{g.s.}}] and fxc[ng.s.]f_{\mathrm{xc}}[n^{\mathrm{g.s.}}]) should be consistent with a single choice for the perturbative-domain vxcv_{\mathrm{xc}}, but often they are considered separately. This can be for practical reasons: For example, one might take a higher-level approximation for vxcg.s.[ng.s.]v_{\mathrm{xc}}^{\mathrm{g.s.}}[n^{\mathrm{g.s.}}] than for the dynamical part fxcf_{\mathrm{xc}}, especially if the sophistication in the former is such that its functional derivative has not been computed or implemented. For semiconductors and insulators, xc kernels for excitonic spectra are more often than not developed independently of the ground-state approximation [76, 77], and the band-gap is often evaluated as a ground-state problem [78], needing a derivative discontinuity or an orbital functional treated with generalized KS, but it could equally be viewed as an excitation property [3].

For the question of memory-dependence, it is certainly useful to think of the ground-state and response parts separately, and away from ω=0\omega= 0 there would be no conflict in these arising from an underlying vXC[n](r,t)v_{\mathrm{XC}}[n](\mathbf{r},t). Time-non-locality translates to ω\omega-dependence in fXC(r,r′,ω)f_{\mathrm{XC}}(\mathbf{r},\mathbf{r}',\omega), and ignoring this dependence, as an adiabatic approximation does, places limitations on the accuracy when multiple excitations are involved as we will now discuss.

Linear response suggests single-photon processes, which in a non-interacting picture imply a response that involves only single excitations. Correspondingly, the KS response function is composed of single transitions from an occupied to an unoccupied orbital of the KS system. On the other hand, excitations of the true interacting system typically involve linear combinations of single, double, and higher excitations, relative to some non-interacting reference expansion, and moreover the correlated nature of the ground-state means one cannot clearly delineate ‘occupied’ and ‘unoccupied’ orbitals in the first place, except in the weak-correlation limit [79]. When applying the TDDFT linear response formalism to the non-interacting KS system, the xc kernel fxcf_{\mathrm{xc}} mixes contributions from the KS single excitations, but any adiabatic approximation to it is blind to double and higher multiple excitations [79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93]. Panel a) of Figure 2, and the discussion in the caption, compare the true, exact KS, and adiabatically-exact TDDFT spectrum for an exactly-solvable model system, showing the consequences of failing to account for double-excitation character.

Dysfunction from neglect of memory-dependence in the adiabatic approximation

Figure 2. Dysfunction from neglect of memory-dependence in the adiabatic approximation for the asymmetric Hubbard dimer, even when the exact ground-state functional is used (adiabatically-exact (AE)). a) Linear response regime: Excitation frequencies ν\nu (top panel) and a measure of relative oscillator strength ff for the second excitation (lower panel), in units of uu, for dimensionless Hubbard parameters u=U/2t=1u = U/2t = 1 (left) and 5 (right), as a function of scaled site potential difference x/u=Δv/ux/u = \Delta v/u, with exact (black), KS (blue), and AE (pink) [110]. Note the scaled site potential scans over the Mott-Hubbard regime x/u<1x/u < 1 to the charge-transfer regime x/u>1x/u > 1, and all values shown for xx in the case of u=1u = 1 and for x/u>1x/u > 1 for u=5u = 5 are characterized as weakly-correlated regimes [110, 111]. The AE excitation energy provides an excellent approximation to the true lowest excitation energy in these weakly-correlated regimes, providing a useful correction of the bare KS value. However the AE completely misses the second excitation over the entire range of parameters, and likewise its relative oscillator strength, due to the inability of AE to capture contributions due to double excitations. It also performs poorly in the strongly-correlated regime (x/u<1x/u < 1, u=5u = 5) for either excitation, since static correlation means that the excitations have a significant contribution from the double-excitation. While the adiabatic approximation is good for states dominated by single-excitation character, memory-dependence included in a frequency-dependent kernel is necessary to capture double-excitation character [112, 113] and ’dressed’ TDDFT approximations have been developed for this purpose [79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93]. b) Non-equilibrium dynamics: Site occupation density difference (left) and xc potential difference (right), for dynamics under a π\pi-pulse resonant with the lowest excitation as a function of time, for parameters t=1/2t = 1/2, δvext=−15\delta v_{\mathrm{ext}} = -15 exact (orange), and AE (purple). The top panel, with U=10U = 10 is in the charge-transfer regime, while the bottom panel with U=20U = 20 is in the Mott-Hubbard regime. While the system remains close to the ground-state in the initial stages, the adiabatic approximation works well but as soon as it the system begins to finitely populate the excited state, the adiabatic approximation fails dramatically in all cases. The exact xc potential develops a stable step feature as time evolves in both cases, that serves to hold the transferred charge in place, in addition to oscillating steps (and in the real-space case, peaks), that are completely missed by any adiabatic approximation [114, 115, 116, 117, 118]. While practical non-adiabatic solutions exist for fxc[ngs.]f_{\mathrm{xc}}[n^{\mathrm{gs.}}] in the response regime, developing approximations for non-adiabatic vxc[n;Ψ(0),Φ(0)]v_{\mathrm{xc}}[n;\Psi(0),\Phi(0)] remain a challenge. Ref. [119, 120] offer a way out through a formulation that uses response xc quantities to simulate non-equilibrium dynamics.

States of double excitation character appear even in lower parts of an excitation spectrum however [94], and their propensity only increases when considering a full landscape of nuclear geometries encountered during photo-induced dynamics [95]. In extended systems, while plasmons are typically well-captured in even the simplest adiabatic approximation, ALDA, double- and higher plasmons are simply absent [93, 96]. The lack of double-excitations is likely responsible for the poor predictions of thermally activated delayed fluorescence, predicting incorrect signs of singlet-triplet gaps [97]. For strongly correlated systems, where a single-determinant ground state description is qualitatively inadequate for the true ground state, double excitations in the KS system are important for every excitation [79]: even if an extremely accurate ground-state functional was found such that the KS single-determinant ground state yields an accurate ground state density, static correlation means that all excitations out of the KS ground state would be near a double excitation where a second electron is excited to a near-degenerate orbital [79]. With usual ground state approximations, the ground-state description is itself very poor in these situations, which plagues conical intersections involving them [95, 98, 99].

There are other instances where the lack of frequency-dependence in the xc kernel is responsible for significant errors, which could be interpreted in terms of missing multiple excitations but are often not discussed in that context because of their nesting in continuous spectra: the excitonic Rydberg series in solids [100], the opening of the KS band-gap in semiconductors [3, 101, 102], and dissipation in extended systems [103, 104]. Frequency-dependence is also how the exact kernels overcome unphysical divergences in quadratic response TDDFT [105, 106, 107, 108, 109].

Not all the errors in the linear response regime can be attributed to the adiabatic approximation however; sometimes the spatial-dependence is at fault. We refer the reader to a short discussion in Sec. III and to the recent reviews cited in the introduction for a more complete details of the pleasures and pains of linear response TDDFT.

Fully non-perturbative regime

Going beyond linear response and allowing strong fields reveals a wealth of physics and phenomena: attosecond-scale time-resolved dynamics of molecules and materials, charge migration, dielectric breakdown in insulators, radiation chemistry, stopping power of projectiles, high-harmonic generation, ultrafast demagnetization and optically-induced interlattice spin transfer, Floquet band engineering, and its optimal control, biomimetic photovoltaic candidates; we refer the reader to the reviews over the past ten years and references therein [4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16]. Again TDDFT allows simulations at unprecedented scales, e.g. Ref. [121] studied conductivity in a disordered aluminium system that treated almost 60,000 electrons explicitly. Adopting a term from Ref. [122], it is fair to say that TDDFT has provided usefully accurate input in these applications, illuminating physical mechanisms and interpretations.

But, stepping back, one might wonder why do such calculations work?! This is not well-understood; the performance of the adiabatic approximation is much better characterized in the response regime. Various reasons have been pondered but none are entirely satisfactory. For example, could it be that in many of these applications vextapp(r,t)v_{\mathrm{ext}}^{\mathrm{app}}(\mathbf{r},t) dominates and one begins in a ground-state, so that the role of xc is really only needed to compensate self-interaction in vH(r,t)v_{\mathrm{H}}(\mathbf{r},t), which adiabatic functionals approximately do to varying degrees of effectiveness? Even if this is the case sometimes, it should also not be read to mean that an adiabatic self-interaction-free functional would give problem-free results e.g. Ref. [123]. Could it instead be that the observable of interest is a relatively insensitive quantity, and the experimental reference is not so well-resolved? Most

vxcAE[n;Ψ(0),Φ(0)](r,t)=vxcexact[n(t)](r)v_{\mathrm{xc}}^{\mathrm{AE}}[n;\Psi(0),\Phi(0)](\mathbf{r},t)=v_{\mathrm{xc}}^{\mathrm{exact}}[n(t)](\mathbf{r})

observables are spatially-integrated; sub-Angstrom spatial resolution in time-resolved experiments on the electronic time-scale exist but are challenging. Could it be due to a judicious choice of initial KS state in cases where the simulation does not begin in a ground state? That is, if the KS initial state has a similar configuration to that of the excited state, then one might be able to argue that remaining correlation could be similar to that of a ground state, and so the adiabatic approximation might provide reasonable results even for this far-from-ground state situation. Again this is not entirely satisfactory since different initial configurations reproduce the same density for excited or non-stationary states and have vastly different exact xc potentials even though the adiabatic ones are identical [124, 125]; further, the configuration changes significantly over time as reflected by the rank of the interacting one-body density matrix while that of the KS system must remain constant, and so any advantage from a judicious pick of the KS initial state would be lost over time [126]. Finally, could it be that, through its lack of any memory-dependence, the adiabatic approximation ironically satisfies in a zeroth order way exact conditions (see Sec. III) to do with memory?

In fact, adiabatic TDDFT does not always yield usefully accurate results in the non-equilibrium regime; disagreements with experiment make it into published papers less often than agreements in successful applications, which is unfortunate as it would help us to understand why it does work when it does. Sometimes, more sophisticated functionals perform worse than simpler ones [127]. Some failures which lead to unphysical results are well-documented and understood. For example, resonantly-driven dynamics targeting a single-excitation of a closed-shell system [114, 128] cannot be captured. At the root of the problem is the inherently non-ground-state nature of the evolving state: xc correlation that pretends such a state is a ground state does not capture step and peak features of the exact xc potential, which depend on the history of the density and the initial states [114, 129]. It is perhaps not surprising such dynamics is so troublesome for TDKS when one thinks what the non-interacting KS system is trying to do: a minimal description of a single excitation involves two orbitals, while the KS system starting in doubly-occupied orbitals must remain with doubly-occupied orbitals for all time due to the one-body nature of the KS potential and so, one orbital must describe at the same time both an electron that stays and one that gets excited.

A striking example is laser-driven charge-transfer dynamics. Although optimally-tuned range-separated hybrids [130, 131, 132, 133, 134] provide a reliable tool to obtain excitation energies of long-range charge transfer nature from TDDFT, applying these functionals, or any adiabatic functional, to simulate laser-driven dynamics from the ground to the charge-transfer excited state generally fails [115, 118, 119, 135, 136]. Studies of model systems show that, when the charge-transfer occurs out of a closed-shell fragment, not only non-adiabatic step and peak structures evolve in time in the exact xc potential, but also a gradual step feature develops in time that stabilizes the charge-transfer state [115, 118]. The latter is completely missed by most adiabatic functionals and only partially captured with the adiabatically-exact, vxcA=vxcAEv_{\mathrm{xc}}^{\mathrm{A}} = v_{\mathrm{xc}}^{\mathrm{AE}}, which inputs the instantaneous density into the exact ground-state functional. Charge begins to transfer but ultimately turns back, even when vxcA=vxcAEv_{\mathrm{xc}}^{\mathrm{A}} = v_{\mathrm{xc}}^{\mathrm{AE}} is used. Panel b) in Figure 2 gives an example of this for a very simple system, the asymmetric Hubbard dimer, that nonetheless contains the same physics in this case; the top panels (charge-transfer regime) corresponding to charge-transfer between closed-shell fragments, and the bottom panels (Mott-Hubbard regime) corresponding to charge-transfer between open-shell fragments.

As further discussed in the caption of Figure 2, charge-transfer between open-shell fragments of a closed-shell system is instead problematic from the very start, because of static correlation: the KS HOMO and LUMO of the closed-shell system are delocalized over both fragments (like bonding and antibonding orbitals). Semilocal functional approximations yield unphysical fractional charges in the ground-state, while the exact ground-state functional has a step feature that re-aligns the two HOMOs [118]. We note these concerns do not necessarily apply with such ferocity to other types of charge transfer, e.g. from injection of an electron, or driven by nuclear motion [118]; adiabatic TDDFT with range-separated hybrid functionals tends to be more successful in these cases.

The failure to capture resonant driving in general can be viewed as a consequence of spurious peak-shifting [137]: with an adiabatic approximation, the resonant frequencies of a system that has been driven away from a ground-state spuriously change in time even when that driving field has been turned off. Depending on the size of this shift, it can muddy the interpretation of time-resolved spectra in pump-probe experiments, for example, and it has been found to be noticeable in model systems as well as small molecules [5, 136, 137, 138]. Although it has not been explicitly shown, the steps and peaks in the exact xc potential continually bring the otherwise drifting resonant frequencies back to their fixed positions.

Therapies, but not always tried and true

Approximations going beyond the adiabatic approximations were proposed and tested soon after the birth of the Runge-Gross theorem, branching out in several directions as illustrated in Figure 3. While some have had success for specific types of problems particularly in the linear response regime, a fully non-equilibrium approximation that has essential features of the exact vxcv_{\mathrm{xc}} has so far remained elusive. An ideal goal that the universality of the theorems would encourage us to aim for is an all-purpose functional approximation, incorporat-

Figure 3. Approximations developed for going beyond the adiabatic approximation, rising up from earlier ones to more recent ones as one climbs the tree. More details are given in Ref. [12]. The most robust approximations are watered by exact conditions such as GTI and ZFT (see text). The first, GK [139, 140] is a frequency-dependent LDA; its violation of GTI and ZFT led to the idea of local memory following the fluid element in DBG [141] and a generalization to current-density functionals in TD CDFT [142, 143] with the VK approximation [143, 144, 145], with later extensions to a GGA form (TV-GGA) [146] and a meta-GGA (TVT-mGGA) [147]. Translationally-invariant (TI) memory actions were developed in Refs. [148, 149], and time-dependent deformation functional theory TDdefFT [150, 151, 152, 153]. On the other side, orbital functionals, with implicit non-local density-dependence, have been explored with TDOEP formalized in Ref. [154], including exact exchange (EXX) [155, 156, 157, 158], and self-interaction corrected LDA (SIC) [159, 160]. These, and other ground-state energy functionals of orbitals Exc[{ϕ}]E_{\mathrm{xc}}[\{\phi\}], have been applied also within a generalized KS scheme (GKS) [161, 162, 163], Many-body perturbation theory (MBPT) was mapped to DFT via Sham-Schlüter (TDSS) [164, 165], and through conserving Φ\Phi or Ψ\Psi-derivable approximations [166]. Relatedly, Bethe-Salpeter-Equation-contracted (BSE-c) xc kernels were found [77, 167, 168, 169, 170]; in particular, frequency-dependence arises from contracting over spatial indices [170]. DMxc are density-matrix coupled approaches [171], and ML represent machine-learning approaches [172, 173, 174, 175] (see text). Turning to approximations for targetted purposes: DTDA is a dressed Tamm-Dancoff-derived frequency-dependent kernel capturing double-excitation energies [79, 80, 81, 82, 84, 87, 88] while DTDDFT recovers also their oscillator strengths [90, 91, 92]. The ‘Pure’ functional [102] contains an implicit frequency-dependence that opens the KS gap in semiconductors and insulators. SIAM derives a non-adiabatic kernel for the Anderson impurity model [176, 177] and DMFTxc for the Hubbard model [178, 179]. Alternate TD(C)DFT-based approaches float through, capturing non-adiabatic effects in different ways: Quantum continuum mechanics (QCM) [180, 181, 182], RR-TDDFT [119] (see text) and Geometric-TDDFT [183, 184] (see text).

:::`

ing known exact conditions associated with spatial and time dependence of the density and dependence on the initial states. At the same time there have been new related approaches describing processes beyond linear response within which the functionals may be more amenable to simple approximations [119, 183, 184], while at another extreme, machine-learned functionals offer an alternative perspective [172, 173, 174].

Figure 3 shows the branching of different proposed approximations that go beyond the adiabatic approximation. The caption contains only a short description and some relevant references, and we refer the reader to a recent review [12] for a more detailed description of the different approaches that were proposed prior to about 2022. In the following, we go into more detail for the advances made since then.

Before doing so, we make some general observations. First, just as in ground-state DFT, exact conditions form a trusty guide for robust approximations. TDDFT inherit some from the ground-state (e.g. self-interaction-free), but there are several known exact conditions without any ground-state counterpart. These have been highlighted in Figure 4 of Ref. [12] and its associated discussion, and include the zero-force theorem (ZFT) and generalized translational invariance (GTI) [32, 185, 186, 187], symmetry and reciprocity relations [188], causality [189, 190], the memory condition intertwining initial-state- and history-dependence [129, 191] (which still plays a role in the structure of the history-dependence even when the initial states are ground states), behavior under uniform coordinate and time scalings [192], constant resonant frequencies for a given transition in absence of external time-dependence [136, 138, 193, 194], and the reduction to the ground-state functionals in the adiabatic limit for systems with a gap [190]. While the severity of the consequences of the violation of these conditions depends on the situation, the particularly egregious ones yield unphysical and unstable predictions, and can be held responsible for failures, e.g. violation of ZFT leads to self-excitation [195, 196, 197], and spurious peak-shifting leads to failures of resonantly-driven processes [136, 193, 194].

A second observation is that not all the errors in adiabatic TDDFT predictions are the fault of the adiabatic approximation per se: sometimes it is that the spatial-dependence on the density is too short-ranged in either or both of the ground-state xc potential or the xc kernel. For examples, for even the lowest excitonic peak in a semiconductor, one needs an xc kernel with 1/q21/q^{2} dependence [76, 167, 168], for Rydberg excitations in a molecule the ground-state xc potential needs a non-local dependence on the ground-state density to yield a −1/r-1/r tail [112, 198], and long-range charge-transfer excitations require non-local density dependence such as provided by exact-exchange in the ground-state xc functional in both pure- and generalized-KS approaches [118, 132, 199], additionally non-local spatial density-dependence and frequency-dependence in the xc kernel for pure density-functional approaches [79, 158, 200]. Looked at from a different angle however, if a functional does have time non-local dependence, errors from inadequate spatial non-locality cannot be completely disentangled from history-dependence [145, 188]: a time-non-local approximation that is spatially local violates the ZFT, ruling out any local, or even semi-local, density approximation with memory.

The examples above all fall within linear response, which brings us to our third observation: there is no guarantee that a functional that performs well in the linear response regime performs well for fully non-equilibrium processes simulated by the TDKS equations. For example, while range-separated hybrids can be tuned to accurately reproduce charge-transfer excitation energies [130, 131, 132, 133, 134], evaluating them on a fully non-equilibrium system introduces severe errors that prevent even a qualitative description of charge-transfer dynamics [115, 118, 119, 135].

A fourth observation relates to initial-state dependence. Even when beginning in ground-states which are functionals of the initial density by the Hohenberg-Kohn theorem, initial-state dependence still structures the history-dependence, due to the memory condition [191]. But the functional-dependence is not on all 3N3N complex degrees of freedom of the initial states, but rather just on the initial interacting one-body reduced density matrix (RDM) and diagonal of the two-body RDM, and the initial one-body RDM of the chosen KS initial state. This can be seen from the exact expression for the time-dependent xc potential [171, 201, 202].

More Modern Therapies

Machine-learned functionals (2020 onwards) Naturally, artificial intelligence explorations have extended to the development of improved memory-dependent functionals. Thankfully these efforts have so far remained outside the “AI empires” [203] avoiding the highly problematic operating procedures that exist in these companies at the topmost levels. Instead, neural networks have been developed in-house codes that attempt to learn the xc potential for a given prescribed density evolution. The one-dimensional electron-hydrogen scattering example [204, 205] has emerged as a popular and challenging test case, due to the clear evidence of history and initial-state dependence in yielding accurate dynamics here. The problem for vxcv_{\mathrm{xc}} is difficult from the very start when the initial KS state is chosen as a single Slater determinant, since this state has such a different structure than the minimally two configuration interacting state (one electron in a target ground-state and one electron in a moving Gaussian wavepacket). The neural net construction in Ref.[172] learned vxcv_{\mathrm{xc}} from input densities, and succeeded in reproducing step and peak features of the correlation potential, showing that the learned functional including memory through a Gaussian-weighted history feature gave better dynamics than a trained functional without memory on related scattering problems not used in the training. Instead, by viewing the xc functional as a control that guides TDKS propagation, one can avoid the need for accurate a priori input data for vXCv_{\mathrm{XC}}, an idea that was first demonstrated in Ref. [173] and recently extensively studied and expanded in Ref. [174], including to beyond one-dimensional models. In fact, two pathways are pursued in Ref. [174]: In one, inversion is carried out first using a combination of PDE-solving and adjoint-based optimization to learn the best pointwise vXC(r,t)v_{\mathrm{XC}}(r,t) from a given n(r,t)n(r,t) input, without a neural net. The pairs of density-potential functions are then used to train a neural net in a similar spirit to Ref. [172], but the refinement of the vXCv_{\mathrm{XC}} resulting from the first step greatly improves the accuracy of the training data. In a second pathway, the adjoint-based neural-network model combined with the TDKS-constrained optimization is treated to directly learn vXCv_{\mathrm{XC}}, avoiding the inversion step, and it can be used to learn its density-functional-dependence. Again their functional generalized well to parameters outside the training set for two-electron scattering problems now in two dimensions. Ref. [174] pointed out the need for high quality numerical propagation methods to generate accurate time-dependent densities as input training data; learning the time-dependent vXCv_{\mathrm{XC}} is very sensitive to this. An alternate approach that avoids the TDKS constraint, and in fact avoids time-stepping altogether, but instead encodes a Hamiltonian-consistency is the “three-head neural architecture” presented in Ref. [175]. This includes training for symplectic structure together with a memory-dependence as a way to ensure long-time stability, energy conservation, and non-growing errors from discretization. The extent of history-dependence is learned from the data instead of prescribing it. Incorporating Hamilton’s equations into the neural network architecture to learn the potential was explored in the classical case in Ref. [206] and had been previously applied to TDKS in Ref. [207] but for predicting adiabatic TDDFT propagation only. In Ref. [175] the three heads function as a coupled system, with the learning of the symplectic structured dynamics and memory-based potential reinforcing each other.

Response-Reformulated TDDFT (2024 onwards)

Noting that the adiabatic functionals are, by definition, more suited to be evaluated on domains close to a ground-state, and further that functional developments beyond adiabatic have been far more successful in the linear response regime, Ref. [119] proposed a “response-reformulated TDDFT” (RR-TDDFT) where xc enters only for ground-state, linear and quadratic response quantities, and yet can describe fully non-equilibrium dynamics. The structure of RR-TDDFT is similar to that of time-dependent configuration interaction, in the sense dynamics enters only in propagation of coefficients of fully-interacting wavefunctions, but in RR-TDDFT, these wavefunctions are never found, and the only ingredients in the propagation are the externally applied field and static electronic structure quantities: ground and excited state energies and couplings between the excited states, all of which are formally exactly available from ground-state DFT, linear and quadratic response TDDFT. The key point is that separation of the time-evolution from the electronic structure ingredients means xc functionals are required to be evaluated only near a ground-state, where the adiabatic approximation performs far better than when these functionals are evaluated on fully non-equilibrium system as they would be in TDKS. Ref. [119] demonstrated that RR-TDDFT can accurately describe resonantly-driven dynamics, while TDKS completely fails, even when both methods use the same approximate functional. RR-TDDFT can exploit improved functionals developed in the response regime for describing fully non-equilibrium dynamics, as was shown in Ref. [120] where dynamics reaching states of double excitation could be reliably predicted, and it can resolve puzzling predictions of the standard TDKS propagation that arise from the incorrect nonlinearity of the xc potential when there is a lack of memory, such as the molecular dipole instability after an ultrafast XUV pulse found in adiabatic TDKS propagation [123]. The scaling depends more heavily on the number of excited states anticipated to be involved in the dynamics, in particular in computing their interstate couplings; whether the method could be applied to extended systems is unclear without further approximations.

Geometric TDDFT (2026 onwards)

A fundamentally new formulation was recently presented in Ref. [183, 184], with new xc concepts and objects, and indications are that their non-adiabatic features are ‘smaller’ than those in the conventional TDDFT. The original formulation of TDDFT could be viewed as resulting from the wavefunction variational principle performed on a constrained manifold of wavefunctions satisfying a target time-dependent density (not to be confused with the variational principle with respect to the density [189, 208]), such that the residue of the unconstrained TDSE belongs to the normal space NΨ(t)\mathcal{N}_{\Psi(t)} of possible wavefunctions that have this target density [184]. Instead of using the variational principle, geometric TDDFT minimizes the norm of the residue of the unconstrained TDSE for wavefunctions evolving on the manifold of target density, leading to the residue instead belonging to iNΨ(t)(t)i\mathcal{N}_{\Psi(t)}(t). This results in a Schrödinger equation with a complex but norm-conserving potential, W~(t)\widetilde{W}(t). This complex potential can be interpreted as replacing the large peaks and step structures of conventional TDDFT with sources and sinks, locally changing the density to match the target, and a corresponding Geometric Kohn-Sham system can be established. Examples on model systems show that the non-adiabatic components of this complex potential tend to be much smaller than those of the conventional TDDFT. How to approximate this potential and how it performs on a wide range of nonequilibrium situations remains to be seen but certainly offers a fresh and promising approach.

Outlook

Progress has been made in understanding where the adiabatic functionals do and do not work, particularly in the linear response regime. Correspondingly, building improved non-adiabatic functionals has so far been simpler in the response regime, as well as improved adiabatic functionals with better spatial-dependence. To illustrate the relevance of such developments, one could consider photochemistry applications, typically modelled by nuclear trajectories or wavepackets moving on coupled electronic potential energy surfaces, with the nuclear rearrangement leading to photo-induced chemical reactions and their control. As highlighted by a recent exciting “calibration exercise” [209], accurate electronic structure quantities play a key role in the outcome of the excited-state dynamics, more critically so than the method used for the nuclear motion. Given the changing character of electronic states over the landscape of nuclear configurations, improved TDDFT functionals in the linear response are still needed. Further, the RR-TDDFT framework now makes these developments applicable also for simulating fully non-equilibrium processes. Fresh approaches such as Geometric-TDDFT and machine-learning discussed here are also promising, and there are certainly other budding ideas being proposed, including using the density’s second time-derivative as basic variable [210], and coupling-constant integral transforms [126, 211]. Further, the role of the breaking of CTCT-symmetry in approximate non-linear theories brings a new perspective [212], that may lead to a paradigm shift in the development of new approaches.

Even if a robust approach to xc in all regimes is found, a remaining problem is how to model functionals for observables not directly related to the density. A particularly relevant example is the current-density: even with the exact xc functional, the KS system in TDDFT only reproduces the longitudinal part of the interacting system’s current exactly, and an extra functional would be needed to correct the rotational part of the KS current to that of the true system. Instead, the KS system of TDCDFT in principle reproduces the rotational component directly [142, 213], and further allows external magnetic fields. But propagating with the generally spatially-dependent xc vector potential that these calculations require render the computational problem challenging, and to treat non-metallic systems, functionals other than VK and its relatives would be desirable to avoid [143, 146, 147] unphysical linewidths and decay in systems with a gap [214].

The task is challenging but also exciting and creative. Given the promise of approximate TDDFT, with its favorable system-size scaling on top of a rigorous foundation, it is certainly worth developing further therapies, including improved functionals and alternate related approaches, to build on the results and insights it has already provided.

Funding Information

This work was partially supported by the U.S. Department of Energy, Office of Science, Office of Advanced Scientific Computing Research and Office of Basic Energy Sciences, Scientific Discovery through Advanced Computing (SciDAC) program under Award Number DE-SC0026088.

Acknowledgments

The author is grateful to Harish Bhat for very helpful comments on the manuscript.

Data Availability

Data sharing is not applicable to this article as no new data were created or analyzed in this study.

References

  1. [1]E. Runge and E. K. U. Gross, Density-functional theory for time-dependent systems, Phys. Rev. Lett. 52, 997 (1984).
  2. [2]C. A. Ullrich, Time-dependent density-functional theory: concepts and applications (Oxford University Press, 2011).
  3. [3]E. K. Gross and N. T. Maitra, Introduction to TDDFT, in Fundamentals of time-dependent density functional theory, Vol. 837, edited by M. A. Marques, N. T. Maitra, F. M. Nogueira, E. K. Gross, and A. Rubio (Springer, 2012) pp. 53–97.DOI
  4. [4]N. T. Maitra, Perspective: Fundamental aspects of time-dependent density functional theory, The Journal of Chemical Physics 144, 220901 (2016), https://doi.org/10.1063/1.4953039.
  5. [5]M. Provorse and C. M. Isborn, Int. J. Quant. Chem. 116, 739 (2016).
  6. [6]J. J. Goings, P. J. Lestrange, and X. Li, Real-time time-dependent electronic structure theory, WIREs Computational Molecular Science 8, e1341 (2018), https://wires.onlinelibrary.wiley.com/doi/pdf/10.1002/wcms.1341.
  7. [7]X. Li, N. Govind, C. Isborn, A. E. DePrince III, and K. Lopata, Real-time time-dependent electronic struc- ture, theory, Chem. Rev. 120, 9951 (2020), pMID: 32813506.
  8. [8]S. A. Sato, First-principles calculations for attosecond electron dynamics in solids, Computational Materials Science 194, 110274 (2021).
  9. [9]P. You, D. Chen, C. Lian, C. Zhang, and S. Meng, First-principles dynamics of photoexcited molecules and materials towards a quantum description, WIREs Computational Molecular Science 11, e1492 (2021), https://wires.onlinelibrary.wiley.com/doi/pdf/10.1002/wcms.1492DOI
  10. [10]R. Sarkar, M. Boggio-Pasqua, P.-F. Loos, and D. Jacquemin, Benchmarking td-dft and wave function methods for oscillator strengths and excited-state dipole moments, Journal of Chemical Theory and Computation 17, 1117 (2021).DOI
  11. [11]J. M. Herbert, Visualizing and characterizing excited states from time-dependent density functional theory, Phys. Chem. Chem. Phys. 26, 3755 (2024).
  12. [12]L. Lacombe and N. T. Maitra, Non-adiabatic approximations in time-dependent density functional theory: progress and prospects, npj Comput. Mater. 9, 124 (2023).
  13. [13]J. Xu, T. E. Carney, R. Zhou, C. Shepard, and Y. Kanai, Real-time time-dependent density functional theory for simulating nonequilibrium electron dynamics, Journal of the American Chemical Society 146, 5011 (2024), pMID: 38362887, https://doi.org/10.1021/jacs.3c08226.
  14. [14]S. A. Sato, H. Hübener, U. De Giovannini, and A. Rubio, Technical review: Time-dependent density functional theory for attosecond physics ranging from gas-phase to solids, npj Computational Materials 11, 233 (2025).DOI
  15. [15]R. D. Amrutkar, P. S. Indaji, D. B. Chaudhari, N. A. Waghmode, and M. A. Jadhav, Advancements in td-dft: A comprehensive review, Current Drug Research Reviews 17, 1 (2025).DOI
  16. [16]C. A. Ullrich, A snapshot of time-dependent density-functional theory, APL Computational Physics 1, 020901 (2025).
  17. [17]N. T. Maitra, T. N. Todorov, C. Woodward, and K. Burke, Density-potential mapping in time-dependent density-functional theory, Phys. Rev. A 81, 042525 (2010).
  18. [18]S. Fournais, J. Lampart, M. Lewin, and T. O. Sørensen, Coulomb potentials and taylor expansions in time-dependent density-functional theory, Phys. Rev. A 93, 062510 (2016).DOI
  19. [19]M. Ruggenthaler and R. van Leeuwen, Global fixed-point proof of time-dependent density-functional theory, Europhys. Lett. 95, 13001 (2011).DOI
  20. [20]M. Ruggenthaler, M. Penz, and R. van Leeuwen, Existence, uniqueness, and construction of the density-potential mapping in time-dependent density-functional theory, J. Phys. Condens. Matter 27, 203202 (2015).DOI
  21. [21]A. B. Lauritsen, M. Lewin, and J. Oldenburg, Time-dependent density functional theory with coulomb interactions (2026), arXiv:2609.08983 [cond-mat.mtrl-sci].arxiv.org/abs/2609.08983
  22. [22]A. Gomez Pueyo, M. A. L. Marques, A. Rubio, and A. Castro, Propagators for the time-dependent kohn-sham equations: Multistep, runge-kutta, exponential runge-kutta, and commutator free magnus methods, Journal of Chemical Theory and Computation 14, 3040 (2018).
  23. [23]Y. Zhu and J. M. Herbert, Self-consistent predictor/corrector algorithms for stable and efficient integration of the time-dependent kohn-sham equation. The Journal of Chemical Physics 148, 044117 (2018).DOI
  24. [24]A. Gomez Pueyo, S. Blanes, and A. Castro, Propagators for quantum-classical models: Commutator-free magnus methods, Journal of Chemical Theory and Computation 16, 1420 (2020).DOI
  25. [25]M. Petersilka, U. J. Gossmann, and E. K. U. Gross, Excitation energies from time-dependent density-functional theory, Phys. Rev. Lett. 76, 1212 (1996).
  26. [26]M. Casida, Time-dependent density functional response theory for molecules, in Recent Advances in Density Functional Methods, Part I, edited by D. Chong (World Scientific, Singapore, 1995).
  27. [27]R. Bauernschmitt and R. Ahlrichs, Treatment of electronic excitations within the adiabatic approximation of time dependent density functional theory, Chem. Phys. Lett. 256, 454 (1996).DOI
  28. [28]F. Furche and R. Ahlrichs, Adiabatic time-dependent density functional methods for excited state properties, The Journal of Chemical Physics 117, 7433 (2002).
  29. [29]F. Furche, On the density matrix based approach to time-dependent density functional response theory, The Journal of Chemical Physics 114, 5982 (2001).
  30. [30]A. Baranova and N. T. Maitra, Excited-State Densities from Time-Dependent Density Functional Response Theory, Journal of Chemical Theory and Computation 10.1021/acs.jctc.5c00909 (2025).
  31. [31]G. Senatore and K. Subbaswamy, Nonlinear response of closed-shell atoms in the density-functional formalism, Physical Review A 35, 2440 (1987).
  32. [32]E. K. U. Gross, J. F. Dobson, and M. Petersilka, Density functional theory of time-dependent phenomena, in Density Functional Theory II: Relativistic and Time Dependent Extensions, edited by R. F. Nalewajski (Springer Berlin Heidelberg, Berlin, Heidelberg, 1996) pp. 81–172.
  33. [33]P. Salek, O. Vahtras, T. Helgaker, and H. Ågren, Density-functional theory of linear and nonlinear time-dependent molecular properties, The Journal of chemical physics 117, 9630 (2002).
  34. [34]S. M. Parker and F. Furche, Response theory and molecular properties, in Frontiers of Quantum Chemistry, edited by M. J. Wójcik, H. Nakatsuji, B. Kirtman, and Y. Ozaki (Springer Singapore, Singapore, 2018) pp. 69–86.
  35. [35]A. Castro, H. Appel, M. Oliveira, C. A. Rozzi, X. Andrade, F. Lorenzen, M. A. L. Marques, E. K. U. Gross, and A. Rubio, octopus: a tool for the application of time-dependent density functional theory, physica status solidi (b) 243, 2465 (2006).
  36. [36]M. Noda, S. A. Sato, Y. Hirokawa, M. Uemoto, T. Takeuchi, S. Yamada, A. Yamada, Y. Shinohara, M. Yamaguchi, K. Iida, I. Floss, T. Otobe, K.-M. Lee, K. Ishimura, T. Boku, G. F. Bertsch, K. Nobusada, and K. Yabana, Salmon: Scalable ab-initio light-matter simulator for optics and nanoscience, Computer Physics Communications 235, 356 (2019).DOI
  37. [37]Q. Sun, M. R. Hermes, X. Wu, H. Zhai, X. Zhang, A. M. Ahmed, J. J. Aucar, O. J. Backhouse, S. Banerjee, P. Bao, N. A. Bogdanov, K. Bystrom, F. Chapoton, N.-Y. Chen, I. Y. Chernyshov, H. S. Clifford, S. Cohen-Janes, Z.-H. Cui, Y. D. Damour, N. Dattani, L. B. Dittmer, S. Ehlert, J. J. Eriksen, F. A. Evangelista, S. A. Ewing, A. Farahvash, K. Focke, Y. Gao, K. E. Gasperich, N. Gillispie, J. Greiner, M. R. Hennefarth, J. Hermann, C. Hillenbrand, J. Huhtasalo, B. Ibrahim, B. Jangid, A. N. Javaremi, A. J. Jenkins, Y. Jin, D. S. King, D. P. Kooi, J. S. Kurian, H. R. Larsson, B. T. G. Lau, S. Lee, S. Lehtola, C. Li, H. Li, J. Li, R. Li, S. Li, A. O. Lykhin, A. Mahajan, N. Mauger, P. d. Mazo-Sevillano, J. Moussa, K. Nakano, V. A. Neufeld, L. Peng, H. Q. Pham, P. Pinski, P. Pokhilko, Z. Pu, Y. Qian, S. J. Quiton, W. T. Schulze, T. R. Scott, A. Seal, J. D. Serna, J. E. T. Smith, K. E. Smyser, T. Stahl, C. Sun, K. J. Sung, E. Trushin, S. Upadhyay, E. A. Vo, T. Vogels, S. Wang, T. Wang, X. Wang, X. Wang, Y. Wang, M. Williamson, J. Yang, H.-Z. Ye, C.-N. Yeh, H. Yu, J. Yu, V. W.-z. Yu, C. Zhang, D. Zhang, Y. Zhang, Z. Zhao, Z. Zhou, A. J. Zhu, T. Zhu, T. C. Berkelbach, L. Gagliardi, S. Sharma, A. V. Sokolov, and G. Kin-Lic Chan, The python simulations of chemistry framework: 10 years of an open-source quantum chemistry project, The Journal of Chemical Physics 165, 102502 (2026).
  38. [38]TURBOMOLE V7.2 2017, a development of University of Karlsruhe and Forschungszentrum Karlsruhe GmbH, 1989-2007, TURBOMOLE GmbH, since 2007; available from http://www.turbomole.com (2017).
  39. [39]J. C. A. Prentice, J. Aarons, J. C. Womack, A. E. A. Allen, L. Andrinopoulos, L. Anton, R. A. Bell, A. B. Bhandari, G. A. Bramley, R. J. Charlton, R. J. Clements, D. J. Cole, G. Constantinescu, F. Corsetti, S. M.-M. Dubois, K. K. B. Duff, J. M. Ecesártin, A. Greco, Q. Hill, L. P. Lee, E. Linscott, D. D. O’Regan, M. J. S. Phipps, L. E. Ratcliff, A. R. Serrano, E. W. Tait, G. Teobaldi, V. Vitale, N. Yey,ung, T. J. Zuehlsdorff, J. Dziedzic, P. D. Haynes, N. D. M. Hine, A. A. Mostofi, M. C. Payne, and C.-K. Skylaris, The onetep linear-scaling density functional theory program, The Journal of Chemical Physics 152, 174111 (2020).DOI
  40. [40]M. Raya-Moreno, A. Buccheri, N. A. Dasch, N. Farahani, I. Gonzalez Oliva, A. Gulans, S. Hong, M. Hossain, H. Kleine, M. Kuban, S. Lubeck, B. Mauer, P. Pavone, F. Peschel, D. Popova-Gorelova, L. Qiao, E. Richter, S. Rigamonti, R. R. Pela, M. Schebek, K. Sinha, D. T. Speckhard, J. Stutz, S. Tillack, D. Tümakov, J. Užulis, M. Voiculescu, C. Vona, M. Yang, and C. Draxl, An exciting approach to theoretical spectroscopy, Advanced Science 13, e76167 (2026), https://advanced.onlinelibrary.wiley.com/doi/pdf/10.1002/advs.76167.
  41. [41]A. García, N. Papior, A. Akhtar, E. Artacho, V. Blum, E. Bosoni, P. Brandimarte, M. Brandbyge, J. I. Cerdá, F. Corsetti, R. Cuadrado, V. Dikan, J. Ferrer, J. Gale, P. García-Fernández, V. M. García-Suárez, S. García, G. Huhs, S. Illera, R. Korytár, P. Koval, I. Lebedeva, L. Lin, P. López-Tarifa, S. G. Mayo, S. Mohr, P. Ordejón, A. Postnikov, Y. Pouillon, M. Pruneda, R. Robles, D. Sánchez-Portal, J. M. Soler, R. Ullah, V. W.-z. Yu, and J. Unquera, Siesta: Recent developments and applications, The Journal of Chemical Physics 152, 204108 (2020).
  42. [42]M. Valiev, E. Bylaska, N. Govind, K. Kowalski, T. Straatsma, H. V. Dam, D. Wang, J. Nieplocha, E. Apra, T. Windus, and W. de Jong, Nwchem: A comprehensive and scalable open-source solution for large scale molecular simulations, Computer Physics Communications 181, 1477 (2010).
  43. [43]J. M. H. Olsen, S. Reine, O. Vahtras, E. Kjellgren, P. Reinholdt, K. O. Hjorth-Jensen, X. Li, J. Cukras, M. Ringholm, E. D. Hedegård, R. Di Remigio, N. H. List, R. Faber, B. N. Cabral Tenorio, R. Bast, T. B. Pedersen, Z. Rinkevičius, S. P. A. Sauer, K. V. Mikkelsen, J. Kongsted, S. Coriani, K. Ruud, T. Helgaker, H. J. A. Jensen, and P. Norman, Dalton project: A python platform for molecular- and electronic-structure simulations of complex systems, The Journal of Chemical Physics 152, 214115 (2020).
  44. [44]J. Shao, Z. Gan, E. Epifanovsky, A. T. Gilbert, M. Wormit, J. Kussmann, A. W. Lange, A. Behn, J. Deng, X. Feng, D. Ghosh, M. Goldey, P. R. Horn, L. D. Jacobson, I. Kaliman, R. Z. Khaliliullin, T. Kuś, A. Landau, J. Liu, E. I. Pronoy, V. M. Rhee, R. M. Richard, M. A. Rohrdanz, R. P. Steele, E. J. Sundstrom, H. L. W. III, P. M. Zimmerman, D. Zuev, B. Albrecht, E. Alguire, B. Austin, G. J. O. Beran, V. A. Bernard, E. Berquist, K. Brandhorst, K. B. Bravaya, S. T. Brown, D. Casanova, C.-M. Chang, Y. Chen, S. H. Chien, K. D. Closser, D. L. Critenden, M. Diedenhofen, R. A. Di. Jr., H. Do, A. D. Duttoi, R. G. Edgar, S. Fatehi, L. Fusti-Molnar, A. Ghysels, A. Golubeva-Zadorozhnaya, J. Gomes, M. W. Hanson-Heine, P. H. Harbach, A. W. Hauser, E. G. Hohenstein, Z. C. Holden, T.-C. Jagau, H. Ji, B. Kaduk, K. Khistyayev, J. Kim, J. Kim, R. A. King, P. Klunzinger, D. Koskenov, T. Kowalczyk, C. M. Krauter, K. U. Lao, A. D. Laurent, K. V. Lawler, S. V. Levchenko, C. Y. Lin, F. Liu, E. Livshits, R. C. Lochan, A. Luenser, P. Manohar, S. F. Manzer, S.-P. Mao, N. Madrossian, A. V. Marenich, S. A. Maurer, N. J. Mayhall, E. Neuscamman, C. M. Oana, R. Olivares-Amaya, D. P. O’Neill, J. A. Parkhill, T. M. Perrine, R. Peverati, A. Prociuk, D. R. Rehn, E. Rosta, N. J. Russ, S. M. Sharada, S. Sharma, D. W. Small, A. Sodt, T. Stein, D. Stück, Y.-C. Su, A. J. Thom, T. Tsuchimochi, V. Vanovschi, L. Vogt, O. Vydrov, T. Wang, M. A. Watson, J. Wenzel, A. White, C. E. Williams, J. Yang, S. Yeganeh, S. R. Yost, Z.-Q. You, I. Y. Zhang, X. Zhang, Y. Zhao, B. R. Brooks, G. K. Chan, D. M. Chipman, C. J. Cramer, W. A. G. III, M. S. Gordon, W. J. Hehre, A. Klamt, H. F. S. III, M. W. Schmidt, C. D. Sherrill, D. G. Truhlar, A. Warshel, X. Xu, A. Aspuru-Guzik, R. Baer, A. T. Bell, N. A. Besley, J.-D. Chai, A. Dreuw, B. D. Dunietz, T. R. Furlani, S. R. G. Gawaltney, C.-P. Hsu, Y. Jung, J. Kong, D. S. Lambrecht, W. Liang, C. Ochsenfeld, V. A. Rasolov, L. V. Slipchenko, J. E. Subotnik, T. V. Voorhis, J. M. Zerbet, A. I. Krylov, P. M. Gill, and M. Head-Gordon, Advances in molecular quantum chemistry contained in the q-chem 4 program package, Molecular Physics 113, 184 (2015), https://doi.org/10.1080/00268976.2014.952696.
  45. [45]The Elk Code, http://elk.sourceforge.net/.
  46. [46]H.-J. Werner, P. J. Knowles, F. R. Manby, J. A. Black, K. Doll, A. Heßelmann, D. Kats, A. Köhn, T. Korona, D. A. Kreplin, Q. Ma, I. Miller, Thomas F., A. Mitrushchenkov, K. A. Peterson, I. Polyak, G. Rauhut, and M. Sibaev, The molpro quantum chemistry package, The Journal of Chemical Physics 152, 144107 (2020).
  47. [47]X. Gonze, B. Amadon, G. Antonius, F. Arnardi, L. Baguet, J.-M. Beuken, J. Bieder, F. Bottin, J. Bouchet, E. Bousquet, N. Brouwer, F. Bruneval, G. Brunin, T. Cavignac, J.-B. Charraud, W. Chen, M. Côté, S. Cottenier, J. Denier, G. Geneste, P. Ghosez, M. Giantomassi, Y. Gillet, O. Gingras, D. R. Hamann, G. Hautier, X. He, N. Helbig, N. Holzwarth, Y. Jia, F. Jollet, W. Lafargue-Dit-Hauret, K. Lejaeghere, M. A. Marques, A. Martin, C. Martins, H. P. Miranda, F. Naccarato, K. Persson, G. Petretto, V. Planes, Y. Pouillon, S. Prokhorenko, F. Ricci, G.-M. Rignanese, A. H. Romero, M. M. Schmitt, M. Torrent, M. J. van Setten, B. Van Troeye, M. J. Verstraete, G. Zérah, and J. W. Zwanziger, The abinitproject: Impact, environment and recent developments, Computer Physics Communications 248, 107042 (2020).
  48. [48]T. D. Kühne, M. Iannuzzi, M. Del Ben, V. V. Rybkin, P. Seewald, F. Stein, T. Laino, R. Z. Khaliullin, O. Schütt, F. Schiffmann, D. Golze, J. Wilhelm, S. Chulkov, M. H. Bani-Hashemian, V. Weber, U. Börstnik, M. Taillefumier, A. S. Jakobovits, A. Lazzaro, H. Pabst, T. Müller, R. Schade, M. Guidon, S. Andermatt, N. Holmberg, G. K. Schenter, A. Hehn, A. Bussy, F. Belleflamme, G. Tabacchi, A. Giblö, M. Lass, I. Bethune, C. J. Mundy, C. Plessl, M. Watkins, J. VandeVondele, M. Krack, and J. Hutter, Cp2k: An electronic structure and molecular dynamics software package - quickstep: Efficient and accurate electronic structure calculations, The Journal of Chemical Physics 152, 194103 (2020).DOI
  49. [49]M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, G. A. Petersson, H. Nakatsuji, X. Li, M. Caricato, A. V. Marenich, J. J. Bloino, B. G. Janesko, and R. Gomperts..., Gaussian 16, Revision C.01, Gaussian, Inc., Wallingford, CT (2016).
  50. [50]G. Geudtner, P. Calaminici, J. Carmona-Espíndola, J. M. del Campo, V. D. Domínguez-Soria, R. F. Moreno, G. U. Gamboa, A. Gourson, A. M. Köster, J. U. Reveles, T. Mineva, J. M. Vásquez-Pérez, A. Vela, B. Zúñiga-Gutierrez, and D. R. Salahub, demon2k, WIREs Computational Molecular Science 2, 548 (2012), https://wires.onlinelibrary.wiley.com/doi/pdf/10.1002/wcms.98.
  51. [51]E. J. Baerends, N. F. Aguirre, N. D. Austin, J. Autschbach, F. M. Bickelhaupt, R. Bulo, C. Capelli, A. C. T. van Duin, F. Egidi, C. Fonseca Guerra, A. Förster, M. Franchini, T. P. M. Goumans, T. Heine, M. Hellström, C. R. Jacob, L. Jensen, M. Krykunov, E. van Lenthe, A. Michalak, M. M. Mitoraj, J. Neugebaur, V. P. Nicu, P. Philipsen, H. Ramanantoanina, R. Rüger, G. Schreckbach, M. Stener, M. Swart, J. M. Thijssen, T. Trnka, L. Visscher, A. Yakovlev, and S. van Gisbergen, The amsterdam modeling suite, The Journal of Chemical Physics 162, 162501 (2025).DOI
  52. [52]V. Blum, R. Gehrke, F. Hanke, P. Have, V. Havu, X. Ren, K. Reuter, and M. Scheffler, Ab initio molecular simulations with numeric atom-centered orbitals, Computer Physics Communications 180, 2175 (2009).
  53. [53]J. J. Mortensen, A. H. Larsen, M. Kuisma, A. V. Ivanov, A. Taghizadeh, A. Peterson, A. Haldar, A. O. Dohn, C. Schäfer, E. O. Jónsson, E. D. Hermes, E. A. Nilsson, G. Kastlunger, G. Levi, H. Jonsson, H. Hakkinen, J. Fojt, J. Kangsabanik, J. Sødequist, J. Lehtomaki, J. Heske, J. Enkovaara, K. T. Winther, M. Dulak, M. M. Melander, M. Ovesen, M. Louhivuori, M. Walter, M. Gjerdring, O. Lopez-Acevedo, P. Erhart, R. Wambier, R. Wurdemann, S. Kaappa, S. Latini, T. M. Boland, T. Bligaard, T. Skovhus, T. Susi, T. Maxson, T. Rossi, X. Chen, Y. L. A. Schmerwitz, J. Schiotz, T. Olsen, K. W. Jacobsen, and K. S. Thygesen, Gpaw: An open python package for electronic structure calculations, The Journal of Chemical Physics 160, 092503 (2024).
  54. [54]C. Lian, M.-X. Guan, S. Hu, J. Zhang, and S. Meng, Photoexcitation in solids: First-principles quantum simulations by real-time tddft, Advanced Theory and Simulations 1, 1800055 (2018).DOI
  55. [55]F. Neese, Software update: The orca program system—version 6.0, Wiley Interdiscip. Rev. Comput. Mol. Sci. 15, e70019 (2025).DOI
  56. [56]D. B. Williams-Young, A. Petrone, S. Sun, T. F. Stetina, P. Lestrange, C. E. Hoyer, D. R. Nascimento, L. Koulias, A. Wildman, J. Kasper, J. J. Goings, F. Ding, A. E. DePrince III, E. F. Valeev, and X. Li, The chronus quantum software package, Wiley Interdisciplinary Reviews: Computational Molecular Science , e1436 (2019).
  57. [57]F. Bruneval, T. Rangel, S. M. Hamed, M. Shao, C. Yang, and J. B. Neaton, molgw 1: Many-body perturbation theory software for atoms, molecules, and clusters, Computer Physics Communications 208, 149 (2016).
  58. [58]X. Andrade, C. D. Pemmaraju, A. Kartsev, X. J. Xiao, A. Lindenberg, S. Rajpurohit, L. Z. Tan, T. Ogitsu, and A. A. Correa, Inq, a modern gpu-accelerated computational framework for (time-dependent) density functional theory, Journal of Chemical Theory and Computation 17, 7447 (2021).
  59. [59]B. Hourahine, M. Berdakin, J. A. Bich, F. P. Bonafé, C. Camacho, Q. Cui, M. Y. Deshaye, G. Díaz Miron, S. Ehlert, M. Elstner, T. Frauenheim, N. Goldman, R. A. González León, T. van der Heide, S. Irle, T. Kowalczyk, T. Kubar, L. S. Lee, C. R. Lien-Medrano, A. Maryewski, T. Melson, S. K. Min, T. Niehaus, A. M. N. Niklasson, A. Pecchia, K. Reuter, C. G. Sanchez, C. Scheurer, M. A. Senft, P. V. Stishenko, V. Q. Vuong, and B. Aradi, Recent developments in dfbt+, a software package for efficient atomistic quantum mechanical simulations, The Journal of Physical Chemistry A 129, 5373 (2025).DOI
  60. [60]D. Sangalli, A. Ferretti, H. Miranda, C. Attaccalite, I. Marri, E. Cannuccia, P. M. Melo, M. Marsili, F. Paleari, A. Marrazzo, et al., Many-body perturbation theory calculations using the yambo code, Journal of Physics: Condensed Matter 31, 325902 (2019).DOI
  61. [61]P. Dinh, M. Vincendon, F. Coppens, E. Suraud, and P.-G. Reinhard, Quantum dissipative dynamics (qdd): A real-time real-space approach to far-off-equilibrium dynamics in finite electron systems, Computer Physics Communications 270, 108155 (2022).
  62. [62]G. M. J. Barca, C. Bertoni, L. Carrington, D. Datta, N. De Silva, J. E. Deustua, D. G. Fedorov, J. R. Gour, A. O. Gunina, E. Guidez, T. Harville, S. Irle, J. Ivanic, K. Kowalski, S. S. Leang, H. Li, W. Li, J. J. Lutz, I. Magoulas, J. Mato, V. Mironov, H. Nakata, B. Q. Pham, P. Piecuch, D. Poole, S. R. Pruitt, A. P. Rendell, L. B. Roskop, K. Ruedenberg, T. Sattasathuchana, M. W. Schmidt, J. Shen, L. Slipchenko, M. Sosonkin, V. Sundriyal, A. Tiwari, J. L. Galvez Vallejo, B. Westheimer, M. Wloch, P. Xu, F. Zahariev, and M. S. Gordon, Recent developments in the general atomic and molecular electronic structure system, The Journal of Chemical Physics 152, 154102 (2020).DOI
  63. [63]A. Degomme, T. Deutsch, and L. Genovese, Wavelet-based linear-response time-dependent density-functional theory, Chemical Physics 402, 112 (2012).
  64. [64]R. M. Parrish, L. A. Burns, D. G. A. Smith, A. C. Simonett, I. DePrince, A. Eugene, E. G. Hohenstein, U. Bozkaya, A. Y. Sokolov, R. Di Remigio, R. M. Richard, J. F. Gonthier, A. M. James, H. R. McAlexander, A. Kumar, M. Saitow, X. Wang, B. P. Pritchard, P. Verma, I. Schaefer, Henry F., K. Patkowski, R. A. King, E. F. Valeev, F. A. Evangelista, J. M. Turney, T. D. Crawford, and C. D. Sherrill, Psi4 1.1: An open-source electronic structure program emphasizing automation, advanced libraries, and interoperability, Journal of Chemical Theory and Computation 13, 3185 (2017).
  65. [65]A. D. Boechatov, E. Harder, T. F. Hughes, J. R. Greenwood, D. A. Braden, D. M. Philipp, D. Rinaldo, M. D. Halls, J. Zhang, and R. A. Friesner, Jaguar: A high-performance quantum chemistry software program with strengths in life and materials sciences, International Journal of Quantum Chemistry 113, 2110 (2013), https://onlinelibrary.wiley.com/doi/pdf/10.1002/qua.24481.DOI
  66. [66]M. C. Rohan, V. A. Suarez, M. C. Philp, and J. S. Kretchmer, Tides: A time-dependent electronic structure code for real-time electron and spin dynamics, The Journal of Chemical Physics 164, 234117 (2026).DOI
  67. [67]L. Yang, J. Yang, and W. Hu, Kssolv toolbox: A matlab graphical user interface for plane-wave density functional theory calculations, Journal of Chemical Theory and Computation 22, 5579 (2026).DOI
  68. [68]J. Jakowski, W. Lu, E. Briggs, D. Lingerfelt, B. G. Sumpter, P. Ganesh, and J. Bernholc, Simulation of 24,000 electron dynamics: Real-time time-dependent density functional theory (tddft) with the real-space multigrids (mg), Journal of Chemical Theory and Computation 21, 10.1021/acs.jctc.4c01241 (2025).DOI
  69. [69]G. Kresse, Ab initio simulations of materials using vasp: Density-functional theory and beyond, Journal of Computational Chemistry 29, 204 (2008).
  70. [70]Y. Chen, M. S. Okay, and B. M. Wong, Mister-t: An open-source software package for quantum optimal control of multi-electron systems on arbitrary geometries, Computer Physics Communications 302, 109248 (2024).DOI
  71. [71]S. Seritan, C. Bannwarth, B. S. Fales, E. G. Hohenstein, C. M. Isborn, S. I. L. Kokkila-Schumacher, X. Li, F. Liu, N. Luehr, J. W. Snyder Jr., C. Song, A. V. Titov, I. S. Ulitsmsev, L.-P. Wang, and T. J. Martínez, Terachem: A graphical processing unit–accelerated electronic structure package for large-scale ab initio molecular dynamics, WIREs Computational Molecular Science 11, e1494 (2021), https://wires.onlinelibrary.wiley.com/doi/pdf/10.1002/wcms.1494.DOI
  72. [72]I. Kim, D. Jeong, W.-J. Son, H.-J. Kim, Y. M. Rhee, Y. Jung, H. Choi, J. Yim, I. Jang, and D. S. Kim, Kohn–sham time-dependent density functional theory with tamm–dancoff approximation on massively parallel gpus, npj Computational Materials 9, 81 (2023).DOI
  73. [73]A. Kononov, M. Nguyen, and A. D. Baczweski, Real-time time-dependent density functional theory for high-energy density physics, Physics of Plasmas 33, 043905 (2026).
  74. [74]J. Sirucek, B. Le Guennic, Y. Damour, P.-F. Loos, and D. Jacquemin, Excited-state absorption: Reference oscillator strengths, wave function, and tddft benchmarks, Journal of Chemical Theory and Computation 21, 4688 (2025).DOI
  75. [75]P. Hohenberg and W. Kohn, Inhomogeneous electron gas, Phys. Rev. 136, B864 (1964).
  76. [76]Y.-M. Byun, J. Sun, and G. A. Ullrich, Time-dependent density-functional theory for periodic solids: assessment of excitonic exchange–correlation kernels, Electronic Structure 2, 023002 (2020).DOI
  77. [77]S. Botti, A. Schindlmayr, R. D. Sole, and L. Reining, Time-dependent density-functional theory for extended systems, Reports on Progress in Phys. 70, 357 (2007).
  78. [78]P. Borlido, T. Aull, A. W. Huran, F. Tran, M. A. L. Marques, and S. Botti, Large-scale benchmark of exchange–correlation functionals for the determination of electronic band gaps of solids, Journal of Chemical Theory and Computation 15, 5069 (2019).DOI
  79. [79]N. T. Maitra, Double and charge-transfer excitations in time-dependent density functional theory, Annual Review of Physical Chemistry 73, 117 (2022), pMID: 34910562, https://doi.org/10.1146/annurev-physchem-082720-124933.
  80. [80]N. T. Maitra, F. Zhang, R. J. Cave, and K. Burke, Double excitations within time-dependent density functional theory: Linear response, J. Chem. Phys. 120, 5932 (2004).
  81. [81]R. J. Cave, F. Zhang, N. T. Maitra, and K. Burke, A dressed TDDFT treatment of the 21ag states of butadiene and hexatriene, Chem. Phys. Lett. 389, 39 (2004).DOI
  82. [82]M. E. Casida, Propagator corrections to adiabatic time-dependent density-functional theory linear response theory, J. Chem. Phys. 122, 054111 (2005).
  83. [83]G. Mazur and R. Włodarczyk, Application of the dressed time-dependent density functional theory for the excited states of linear polyenes, J. Comput. Chem. 30, 811 (2009).
  84. [84]G. Mazur, M. Makowski, R. Włodarczyk, and Y. Aoki, Dressed TDDFT study of low-lying electronic excited states in selected linear polyenes and diphenylpolyenes, Int. J. Quant. Chem. 111, 819 (2011).DOI
  85. [85]P. Romaniello, D. Sangalli, J. A. Berger, F. Sottile, L. G. Molinari, L. Reining, and G. Onida, Double excitations in finite systems, J. Chem. Phys. 130, 044108 (2009).
  86. [86]D. Sangalli, P. Romaniello, G. Onida, and A. Marini, Double excitations in correlated systems: A many-body approach, The Journal of Chemical Physics 134, 034115 (2011), https://doi.org/10.1063/1.3518705.
  87. [87]M. Huix-Rotllant, A. Ipatov, A. Rubio, and M. E. Casida, Assessment of dressed time-dependent density-functional theory for the low-lying valence states of 28 organic chromophores, Chem. Phys. 391, 120 (2011).DOI
  88. [88]M. Casida and M. Huix-Rotllant, Progress in time-dependent density-functional theory, Ann. Rev. Phys. Chem. 63, 287 (2012).
  89. [89]M. E. Casida and M. Huix-Rotllant, Many-body perturbation theory (MBPT) and time-dependent density-functional theory (TD-DFT): MBPT insights about what is missing in, and corrections to, the TD-DFT adiabatic approximation, in Density-Functional Methods for Excited States, edited by N. Ferré, M. Filatov, and M. Huix-Rotllant (Springer International Publishing, Cham, 2016) pp. 1–60.
  90. [90]D. B. Dar and N. T. Maitra, Oscillator strengths and excited-state couplings for double excitations in time-dependent density functional theory, The Journal of Chemical Physics 159 (2023).
  91. [91]D. B. Dar and N. T. Maitra, Capturing the elusive curve-crossing in low-lying states of butadiene with dressed tddft, The Journal of Physical Chemistry Letters 16, 703 (2025).
  92. [92]D. B. Dar and N. T. Maitra, Correction to “capturing the curve-crossing in low-lying states of butadiene with dressed tddft”, The Journal of Physical Chemistry Letters 17, 5625 (2026).
  93. [93]S. Gao, Nonlinear response of metal nanoparticles: Double plasmon excitation and electron transfer, The Journal of Chemical Physics 142, 234701 (2015).DOI
  94. [94]K. Kossowski, M. Boggio-Pasqua, P.-F. Loos, and D. Jacquemin, Reference energies for double excitations: Improvement and extension, Journal of Chemical Theory and Computation 20, 5655 (2024).DOI
  95. [95]B. G. Levine, C. Ko, J. Quenneville, and T. J. Martinez, Conical intersections and double excitations in time-dependent density functional theory, Mol. Phys. 104, 1039 (2006).
  96. [96]M. Panholzer, M. Gatti, and L. Reining, Nonlocal and nonadiabatic effects in the charge-density response of solids: A time-dependent density-functional approach, Phys. Rev. Lett. 120, 166402 (2018).
  97. [97]P. de Silva, Inverted singlet–triplet gaps and their relevance to thermally activated delayed fluorescence, The Journal of Physical Chemistry Letters 10, 5674 (2019).
  98. [98]E. Tapavicza, I. Tavernelli, U. Rothlisberger, C. Filippi, and M. E. Casida, Mixed time-dependent density-functional theory/classical trajectory surface hopping study of oxirane photochemistry, J. Chem. Phys. 129, 124108 (2008).DOI
  99. [99]J. T. Taylor, D. J. Tozer, and B. F. E. Curchod, On the topological phase around conical intersections with tamm–dancoff linear-response time-dependent density functional theory, The Journal of Physical Chemistry A 128, 5314 (2024).DOI
  100. [100]D. R. Gulevich, Y. V. Zhumagulov, V. K. Kozin, and I. V. Tokatly, Excitonic effects in time-dependent density functional theory from zeros of the density response, Phys. Rev. B 107, 165118 (2023).
  101. [101]G. Giuliani and G. Vignale, Quantum Theory of the Electron Liquid, Masters Series in Physics and Astronomy (Cambridge University Press, 2005).
  102. [102]S. Cavo, J. A. Berger, and P. Romaniello, Accurate optical spectra of solids from pure time-dependent density functional theory, Phys. Rev. B 101, 115109 (2020).
  103. [103]R. D’Agosta and G. Vignale, Relaxation in time-dependent current-density-functional theory, Phys. Rev. Lett. 96, 016405 (2006).
  104. [104]I. D’Amico and C. A. Ullrich, Dissipation through spin coulomb drag in electronic spin transport and optical excitations, Phys. Rev. B 74, 121303 (2006).DOI
  105. [105]S. M. Parker, S. Roy, and F. Furche, Unphysical divergences in response theory, The Journal of Chemical Physics 145, 134105 (2016), http://dx.doi.org/10.1063/1.4963749.
  106. [106]Z. Li and W. Liu, First-order nonadiabatic coupling matrix elements between excited states: A lagrangian formulation at the cis, rpa, td-hf, and td-dft levels, The Journal of Chemical Physics 141, 014110 (2014).DOI
  107. [107]Q. Ou, G. D. Bellchambers, F. Furche, and J. E. Subotnik, First-order derivative couplings between excited states from adiabatic TDDFT response theory, J. Chem. Phys. 142, 064114 (2015).
  108. [108]X. Zhang and J. M. Herbert, Analytic derivative couplings in time-dependent density functional theory: Quadratic response theory versus pseudo-wavefunction approach, J. Chem. Phys. 142, 064109 (2015).
  109. [109]D. Dar, S. Roy, and N. T. Maitra, Curing the divergence in time-dependent density functional quadratic response theory, The Journal of Physical Chemistry Letters 14, 3186 (2023).
  110. [110]D. J. Carrascal, J. Ferrer, N. Maitra, and K. Burke, Linear response time-dependent density functional theory of the hubbard dimer, The European Physical Journal B 91, 142 (2018).
  111. [111]D. J. Carrascal, J. Ferrer, J. C. Smith, and K. Burke, The hubbard dimer: a density functional case study of a many-body problem, J. Phys. Condens. Matter 27, 393001 (2015).
  112. [112]D. J. Tozer, R. D. Amos, N. C. Handy, B. O. Roos, and L. Serrano-Andres, Does density functional theory contribute to the understanding of excited states of unsaturated organic compounds?, Mol. Phys. 97, 859 (1999).
  113. [113]D. J. Tozer and N. C. Handy, On the determination of excitation energies using density functional theory, Phys. Chem. Chem. Phys. 2, 2117 (2000).
  114. [114]P. Elliott, J. I. Fuks, A. Rubio, and N. T. Maitra, Universal dynamical steps in the exact time-dependent exchange-correlation potential, Phys. Rev. Lett. 109, 266404 (2012).
  115. [115]J. I. Fuks, P. Elliott, A. Rubio, and N. T. Maitra, Dynamics of charge-transfer processes with time-dependent density functional theory, J. Phys. Chem. Lett. 4, 735 (2013).
  116. [116]J. I. Fuks and N. T. Maitra, Charge transfer in time-dependent density-functional theory: Insights from the asymmetric hubbard dimer, Phys. Rev. A 89, 062502 (2014).
  117. [117]J. Ramsden and R. Godby, Exact density-functional potentials for time-dependent quasiparticles, Phys. Rev. Lett. 109, 036402 (2012).
  118. [118]N. T. Maitra, Charge transfer in time-dependent density functional theory, Journal of Physics: Condensed Matter 29, 423001 (2017).
  119. [119]D. B. Dar, B. Baranova, and N. T. Maitra, Reformulation of time-dependent density functional theory for nonperturbative dynamics: The rabi oscillation problem resolved, Phys. Rev. Lett. 133, 096401 (2024).
  120. [120]D. Ray, A. Baranova, D. B. Dar, and N. T. Maitra, Harnessing dressed time-dependent density functional theory for the nonperturbative regime: Electron dynamics with double excitations, APS Open Sci. 1, 1?00014 (2026).
  121. [121]E. W. Draeger, X. Andrade, J. A. Gunnells, A. Bhatle, A. Schleife, and A. A. Correa, Massively parallel first-principles simulation of electron dynamics in materials, J. Parallel Distrib. Comput. 106, 205 (2017).
  122. [122]D. Rappoport, N. R. M. Crawford, F. Furche, and K. Burke, Which functional should i choose?, in Computational Inorganic and Bioinorganic Chemistry (Wiley, Chichester, 2008) pp. 1–25.
  123. [123]D. B. Dar, D. Ray, and N. T. Maitra, Post-pulse dipole instability in adiabatic time-dependent density functional theory: Fact versus artifact, Phys. Rev. A 114, 022830 (2026).
  124. [124]J. I. Fuks, S. Nielsen, M. Ruggenthaler, and N. Maitra, Time dependent density functional theory beyond kohn-sham slater determinants, Phys. Chem. Chem. Phys. 18, 20976 (2016).
  125. [125]P. Elliott and N. T. Maitra, Propagation of initially excited states in time-dependent density-functional theory, Phys. Rev. A 85, 052510 (2012).
  126. [126]L. Lacombe and N. T. Maitra, Developing new and understanding old approximations in TDDFT, Faraday Discuss. 224, 382 (2020).
  127. [127]O. Neufeld, N. Tancogne-Dejean, and A. Rubio, Benchmarking functionals for strong-field light-matter interactions in adiabatic time-dependent density functional theory, The Journal of Physical Chemistry Letters 15, 7254 (2024).
  128. [128]M. Ruggen­thaler and D. Bauer, Rabi oscillations and few-level approximations in time-dependent density functional theory, Phys. Rev. Lett. 102, 233001 (2009).
  129. [129]D. Dar, L. Lacombe, and N. T. Maitra, The exact exchange–correlation potential in time-dependent density functional theory: Choreographing electrons with steps and peaks, Chem. Phys. Rev. 3, 031307 (2022).
  130. [130]R. Baer, E. Livshits, and U. Salzner, Tuned range-separated hybrids in density functional theory, Ann. Rev. Phys. Chem. 61, 85 (2010).
  131. [131]L. Kronik, T. Stein, S. Refaely-Abramson, and R. Baer, Excitation gaps of finite-sized systems from optimally tuned range-separated hybrid functionals, J. Chem. Theory and Comput. 8, 1515 (2012).
  132. [132]T. Stein, L. Kronik, and R. Baer, Reliable prediction of charge transfer excitations in molecular complexes using time-dependent density functional theory, Journal of the American Chemical Society 131, 2818 (2009).
  133. [133]A. Karolewski, L. Kronik, and S. Kümmel, Using optimally tuned range separated hybrid functionals in ground-state calculations: Consequences and caveats, J. Chem. Phys. 138, 204115 (2013).
  134. [134]T. Körzdörfer and J.-L. Brédas, Organic electronic materials: Recent advances in the dft description of the ground and excited states using tuned range-separated hybrid functionals, Accounts of Chemical Research 47, 3284 (2014), pMID: 24784485, http://dx.doi.org/10.1021/ar500021t.
  135. [135]S. Raghavachari and M. Nest, Critical examination of explicitly time-dependent density functional theory for coherent control of dipole switching, J. Chem. Theory and Comput. 7, 2492 (2011).
  136. [136]J. I. Fuks, K. Luo, E. D. Sandoval, and N. T. Maitra, Time-resolved spectroscopy in time-dependent density functional theory: An exact condition, Phys. Rev. Lett. 114, 183002 (2015).
  137. [137]R. Ramakrishnan and M. Nest, Control and analysis of single-determinant electron dynamics, Phys. Rev. A 85, 054501 (2012).
  138. [138]B. F. Habenicht, N. P. Tani, M. R. Prov­orse, and C. M. Isborn, J. Chem. Phys. 141, 184112 (2014).
  139. [139]E. K. U. Gross and W. Kohn, Local density-functional theory of frequency-dependent linear response, Phys. Rev. Lett. 55, 2850 (1985).
  140. [140]E. K. U. Gross and W. Kohn, Time-dependent density functional theory, Adv. Quantum Chem. 21, 255 (1990).
  141. [141]J. F. Dobson, M. J. Bünner, and E. K. U. Gross, Time-dependent density functional theory beyond linear response: An exchange-correlation potential with memory, Phys. Rev. Lett. 79, 1905 (1997).
  142. [142]S. K. Ghosh and A. K. Dhara, Density-functional theory of many-electron systems subjected to time-dependent electric and magnetic fields, Phys. Rev. A 38, 1149 (1988).
  143. [143]G. Vignale and W. Kohn, Current-dependent exchange-correlation potential for dynamical linear response theory, Phys. Rev. Lett. 77, 2037 (1996).
  144. [144]G. Vignale, C. A. Ullrich, and S. Conti, Time-dependent density functional theory beyond the adiabatic local density approximation, Phys. Rev. Lett. 79, 4878 (1997).
  145. [145]G. Vignale, Time-dependent current density functional theory, in Fundamentals of Time-Dependent Density Functional Theory, edited by M. A. Marques, N. T. Maitra, F. M. Nogueira, E. K. U. Gross, and A. Rubio (Springer Berlin Heidelberg, 2012) pp. 457–469.
  146. [146]J. Tao and G. Vignale, Time-dependent density-functional theory beyond the local-density approximation, Phys. Rev. Lett. 97, 036403 (2006).
  147. [147]J. Tao, G. Vignale, and I. V. Tokatly, Time-dependent density functional theory: Derivation of gradient-corrected dynamical exchange-correlation potentials, Phys. Rev. B 76, 195126 (2007).
  148. [148]Y. Kurzweil and R. Baer, Time-dependent exchange-correlation current density functionals with memory, J. Chem. Phys. 121, 8731 (2004), https://doi.org/10.1063/1.1802793.
  149. [149]Y. Kurzweil and R. Baer, Quantum memory effects in the dynamics of electrons in gold clusters, Phys. Rev. B 73, 075413 (2006).
  150. [150]I. V. Tokatly and O. Pankratov, Local exchange-correlation vector potential with memory in time-dependent density functional theory: The generalized hydrodynamics approach, Phys. Rev. B 67, 201103 (2003).
  151. [151]I. V. Tokatly, Quantum many-body dynamics in a lagrangian frame: I. Equations of motion and conservation laws, Phys. Rev. B 71, 165104 (2005).
  152. [152]I. V. Tokatly, Quantum many-body dynamics in a lagrangian frame: II. Geometric formulation of time-dependent density functional theory, Phys. Rev. B 71, 165105 (2005).
  153. [153]I. V. Tokatly, Time-dependent deformation functional theory, Phys. Rev. B 75, 125105 (2007).
  154. [154]C. A. Ullrich, U. J. Grossmann, and E. K. U. Gross, Time-dependent optimized effective potential, Phys. Rev. Lett. 74, 872 (1995).
  155. [155]A. Görling, Exact exchange-correlation kernel for dynamic response properties and excitation energies in density-functional theory, Phys. Rev. A 57, 3433 (1998).
  156. [156]M. Hellgren and U. von Barth, Linear density response function within the time-dependent exact-exchange approximation, Phys. Rev. B 78, 115107 (2008).
  157. [157]M. Hellgren and U. von Barth, Exact-exchange kernel of time-dependent density functional theory: Frequency dependence and photoabsorption spectra of atoms, J. Chem. Phys. 131, 044110 (2009).
  158. [158]M. Hellgren and E. K. U. Gross, Discontinuous functional for linear-response time-dependent density-functional theory: The exact-exchange kernel and approximate forms, Phys. Rev. A 88, 052507 (2013).
  159. [159]C. A. Ullrich, P.-G. Reinhard, and E. Suraud, Electron dynamics in strongly excited sodium clusters: a density-functional study with self-interaction correction, J. Phys. B: At. Mol. Opt. Phys 31, 1871 (1998).
  160. [160]D. Hofmann, T. Körzdörfer, and S. Kümmel, Kohn-sham self-interaction correction in real time, Phys. Rev. Lett. 108, 146401 (2012).
  161. [161]A. Seidl, A. Görling, P. Vogl, J. A. Majewski, and M. Levy, Generalized Kohn-Sham schemes and the band-gap problem, Phys. Rev. B 53, 3764 (1996).
  162. [162]A. Görling and M. Levy, Hybrid schemes combining the Hartree–Fock method and density-functional theory: Underlying formalism and properties of correlation functionals, The Journal of Chemical Physics 106, 2675 (1997), http://dx.doi.org/10.1063/1.473369.
  163. [163]R. Garrick, A. Natan, T. Gould, and L. Kronik, Exact generalized Kohn-Sham theory for hybrid functionals, Phys. Rev. X 10, 021040 (2020).
  164. [164]R. van Leeuwen, The Sham-Schlüter equation in time-dependent density-functional theory, Phys. Rev. Lett. 76, 3610 (1996).
  165. [165]I. V. Tokatly and O. Pankratov, Many-body diagrammatic expansion in a kohn-sham basis: Implications for time-dependent density functional theory of excited states, Phys. Rev. Lett. 86, 2078 (2001).
  166. [166]U. von Barth, N. E. Dahlen, R. van Leeuwen, and G. Stefanucci, Conserving approximations in time-dependent density functional theory, Phys. Rev. B 72, 235109 (2005).
  167. [167]L. Reining, V. Olevano, A. Rubio, and G. Onida, Excitonic effects in solids described by time-dependent density-functional theory, Phys. Rev. Lett. 88, 066404 (2002).
  168. [168]F. Bruneval, F. Sottile, V. Olevano, R. Del Sole, and L. Reining, Many-body perturbation theory using the density-functional concept: Beyond the GW approximation, Phys. Rev. Lett. 94, 186402 (2005).
  169. [169]R. Stubner, I. V. Tokatly, and O. Pankratov, Excitonic effects in time-dependent density-functional theory: An analytically solvable model, Phys. Rev. B 70, 245119 (2004).
  170. [170]M. Gatti, V. Olevano, L. Reining, and I. V. Tokatly, Transforming nonlocality into a frequency dependence: A shortcut to spectroscopy, Phys. Rev. Lett. 99, 057401 (2007).
  171. [171]L. Lacombe and N. T. Maitra, Density-matrix coupled time-dependent exchange-correlation functional approximations, J. Chem. Theory Comput. 15, 1672 (2019).
  172. [172]Y. Suzuki, R. Nagai, and J. Haruyama, Machine learning exchange-correlation potential in time-dependent density-functional theory, Phys. Rev. A 101, 050501(R) (2020).
  173. [173]H. S. Bhat, K. Collins, P. Gupta, and C. M. Isborn, Dynamic learning of correlation potentials for a time-dependent kohn-sham system, in Proceedings of The 4th Annual Learning for Dynamics and Control Conference, Proceedings of Machine Learning Research, Vol. 168, edited by R. Firoozi, N. Mehr, E. Yel, R. Antonova, J. Bohg, M. Schwager, and M. Kochenderfer (PMLR, 2022) pp. 546–558.
  174. [174]H. S. Bhat and C. M. Isborn, Learning memory-dependent neural network correlation potentials for accurate electron dynamics (2026), arXiv:2609.04527 [physics.chem-ph].arxiv.org/abs/2609.04527
  175. [175]L. Wang, J. Zhou, L. Li, and D. D. Zeng, A three-head hamiltonian-consistent neural network for nonadiabatic time-dependent density functional theory dynamics, Machine Learning: Science and Technology 7, 045074 (2026).
  176. [176]N. Dittmann, J. Splettstoesser, and N. Helbig, Nonadiabatic dynamics in single-electron tunneling devices with time-dependent density-functional theory, Phys. Rev. Lett. 120, 157701 (2018).
  177. [177]N. Dittmann, N. Helbig, and D. M. Kennes, Dynamics of the Anderson impurity model: Benchmarking a nonadiabatic exchange-correlation potential in time-dependent density-functional theory, Phys. Rev. B 99, 075417 (2019).
  178. [178]V. Turkowski and T. S. Rahman, Nonadiabatic exchange-correlation kernel for strongly correlated materials, J. Phys. Condens. Matter 29, 455601 (2017).
  179. [179]S. R. Acharya, V. Turkowski, G. P. Zhang, and T. S. Rahman, Ultrafast electron correlations and memory effects at work: Femtosecond demagnetization in ni, Phys. Rev. Lett. 125, 017202 (2020).
  180. [180]J. Tao, X. Gao, G. Vignale, and I. V. Tokatly, Linear continuum mechanics for quantum many-body systems, Phys. Rev. Lett. 103, 086401 (2009).
  181. [181]X. Gao, J. Tao, G. Vignale, and I. V. Tokatly, Continuum mechanics for quantum many-body systems: Linear response regime, Phys. Rev. B 81, 195106 (2010).
  182. [182]T. Gould, G. Jansen, I. V. Tokatly, and J. F. Dobson, Quantum continuum mechanics made simple, J. Chem. Phys. 136, 204115 (2012), https://doi.org/10.1063/1.4721269.
  183. [183]É. Cancès, T. Duez, J. van Gog, A. B. Lauritsen, M. Lewin, and J. Toulouse, Geometric time-dependent density functional theory, Phys. Rev. Lett. 136, 256401 (2026).
  184. [184]É. Cancès, T. Duez, J. van Gog, A. B. Lauritsen, M. Lewin, and J. Toulouse, Geometric theory of constrained schrodinger dynamics with application to time-dependent density-functional theory on a finite lattice, Phys. Rev. A 113, 062222 (2026).
  185. [185]J. F. Dobson, Harmonic-potential theorem: Implications for approximate many-body theories, Phys. Rev. Lett. 73, 2244 (1994).
  186. [186]G. Vignale, Center of mass and relative motion in time dependent density functional theory, Phys. Rev. Lett. 74, 3233 (1995).
  187. [187]G. Vignale, Sum rule for the linear density response of a driven electronic system, Phys. Lett. A 209, 206 (1995).
  188. [188]G. Vignale and W. Kohn, Current-density functional theory of linear response to time-dependent electromagnetic fields, in Electronic Density Functional Theory: Recent Progress and New Directions, edited by J. F. Dobson, G. Vignale, and M. P. Das (Springer US, Boston, MA, 1998) pp. 199–216.
  189. [189]R. van Leeuwen, Causality and symmetry in time-dependent density-functional theory, Phys. Rev. Lett. 80, 1280 (1998).
  190. [190]L. O. Wagner, Z.-h. Yang, and K. Burke, Exact conditions and their relevance in TDDFT, in Fundamentals of Time-Dependent Density Functional Theory, edited by M. A. Marques, N. T. Maitra, F. M. Nogueira, E. Gross, and A. Rubio (Springer, Berlin, Heidelberg, 2012) pp. 101–123.
  191. [191]N. T. Maitra, K. Burke, and C. Woodward, Memory in time-dependent density functional theory, Phys. Rev. Lett. 89, 023002 (2002).
  192. [192]P. Hessler, J. Park, and K. Burke, Several theorems in time-dependent density functional theory, Phys. Rev. Lett. 82, 378 (1999).
  193. [193]K. L. J. Fuks, and N. T. Maitra, Studies of spuriously shifting resonances in time-dependent density functional theory, The Journal of Chemical Physics 145, 044101 (2016), http://dx.doi.org/10.1063/1.4955447.
  194. [194]S. Raghunathan and M. Nest, The lack of resonance problem in coherent control with real-time time-dependent density functional theory, J. Chem. Theory and Comput. 8, 806 (2012).
  195. [195]M. Mundt, S. Kümmel, R. van Leeuwen, and P.-G. Reinhard, Violation of the zero-force theorem in the time-dependent krieger-li-iafrate approximation, Phys. Rev. A 75, 050501 (2007).
  196. [196]Y. Kurzweil and R. Baer, Adapting approximate-memory potentials for time-dependent density functional theory, Phys. Rev. B 77, 085121 (2008).
  197. [197]J. Sun, C.-W. Lee, A. Kononov, A. Schleife, and C. A. Ullrich, Real-time excitation dynamics with time-dependent density-functional theory, Phys. Rev. Lett. 127, 077401 (2021).
  198. [198]A. Wasserman, N. T. Maitra, and K. Burke, Accurate Rydberg excitations from the local density approximation, Phys. Rev. Lett. 91, 26300 (2003).
  199. [199]S. Kümel, Charge-transfer excitations: A challenge for time-dependent density functional theory that has been met, Advanced Energy Materials 7, 1700440.
  200. [200]M. Hellgren and E. K. U. Gross, Discontinuities of the exchange-correlation kernel and charge-transfer excitations in time-dependent density-functional theory, Phys. Rev. A 85, 022514 (2012).
  201. [201]R. van Leeuwen, Mapping from densities to potentials in time-dependent density-functional theory, Phys. Rev. Lett. 82, 3863 (1999).
  202. [202]K. Luo, J. I. Fuks, E. D. Sandoval, P. Elliott, and N. T. Maitra, Kinetic and interaction components of the exact time-dependent correlation potential, J. Chem. Phys 140, 18A515 (2014).
  203. [203]K. Hao, Empire of AI: Inside the Reckless Race for Total Domination (Allen Lane, 2025).
  204. [204]Y. Suzuki, L. Lacombe, K. Watanabe, and N. T. Maitra, Exact time-dependent exchange-correlation potential in electron scattering processes, Physical review letters 119, 263401 (2017).
  205. [205]L. Lacombe, Y. Suzuki, K. Watanabe, and N. T. Maitra, Electron scattering in time-dependent density functional theory, Eur Phys J B 91, 96 (2018).
  206. [206]H. S. Bhat, Learning and interpreting potentials for classical hamiltonian systems, in Machine Learning and Knowledge Discovery in Databases, edited by P. Cellier and K. Driessens (Springer International Publishing, Cham, 2020) pp. 217–228.
  207. [207]J. Yang and J. Whitfield, Machine-learning kohn–sham potential from dynamics in time-dependent kohn–sham systems, Machine Learning: Science and Technology 4, 035022 (2023).
  208. [208]G. Vignale, Real-time resolution of the causality paradox of time-dependent density-functional theory, Phys. Rev. A 77, 062511 (2008).
  209. [209]J. Janoš, N. H. List, A. J. Orr-Ewing, J. Suchan, M. Barbatti, O. Bennett, M. Brady, J. Carmona-García, R. Crespo-Otero, J. Eng, O. J. Fajen, M. Garavelli, S. Gómez, A. E. Green, F. J. Hernández, D. Hollas, L. Hutton, L. M. Ibele, A. Kirrander, Z. Lan, Y. Lassmann, J. E. Lawrence, B. G. Levine, D. V. Makhov, J. R. Mannouch, X. Miao, R. Mitrić, S. M. Parker, T. J. Penfold, J. Peng, J. O. Richardson, D. Shalashilin, P. Slavíček, K. E. Spinlove, P. Vindel-Zandbergen, F. Agostini, S. Bonella, T. J. Martínez, G. A. Worth, and B. F. E. Curchod, Perspective on a challenge: Predicting the photochemistry of cyclobutanone, The Journal of Chemical Physics 165, 020902 (2026).
  210. [210]W. Tarantino and C. A. Ullrich, A reformulation of time-dependent kohn–sham theory in terms of the second time derivative of the density, J. Chem. Phys. 154, 204112 (2021), https://doi.org/10.1063/5.0039962.
  211. [211]A. Görling, Time-Dependent Kohn-Sham formalism, Phys. Rev. A 55, 2630 (1997).
  212. [212]F. Furche and G. P. Chen, Ct symmetry in nonlinear time-dependent electronic structure theories, The Journal of Chemical Physics 163, 174119 (2025).
  213. [213]G. Vignale, Mapping from current densities to vector potentials in time-dependent current density functional theory, Phys. Rev. B 70, 201102 (2004).
  214. [214]C. A. Ullrich and K. Burke, Excitation energies from time-dependent density-functional theory beyond the adiabatic approximation, J. Chem. Phys. 121, 28 (2004).

Paper details

vlnc:W_69mnf4fm
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents