Time-Dependent Coupled Channels Wave-Packet

• TDCCWP is a quantum dynamics method that propagates time-dependent wave-packets under a coupled channels framework to simulate multichannel reaction dynamics.
• It employs propagation algorithms such as split-operator, Chebyshev, and Crank–Nicolson to balance computational efficiency and accuracy.
• Observables including transmission coefficients, cross sections, and reaction rates are extracted from the propagated state, serving as precise benchmarks for validation.

Searching arXiv for recent TDCCWP-related papers and the specific IDs provided. Time-Dependent Coupled Channels Wave-Packet (TDCCWP) denotes a quantum dynamical approach that utilizes a time-dependent wave-packet in a coupled-channels framework to simulate dynamics on multiple coupled potential surfaces or reaction channels. In the formulations represented here, the method propagates a wave packet under a matrix-valued Hamiltonian whose diagonal terms describe channel-resolved motion and whose off-diagonal terms encode inter-channel couplings, then extracts observables such as transmission coefficients, cross sections, and reaction rates from the propagated state. TDCCWP appears in molecular quantum dynamics, neutron capture, and electron–molecule collision theory, and it is closely connected to split-operator propagation, Crank–Nicolson propagation, modified Chebyshev propagation, and exact or benchmark time-domain solutions for reduced coupled-state models (Lightfoot et al., 15 Sep 2025, Nkambule et al., 9 Jun 2026, Aharonovich et al., 2016, Rajendran et al., 2018).

1. Formal definition and governing equations

At its core, TDCCWP is the direct solution of a time-dependent Schrödinger equation for a vector-valued wave packet. In the most compact representation, the nuclear dynamics on coupled electronic states are described by

$i\frac{\partial}{\partial t}\Psi(R,t)=H\Psi(R,t),$

with $R$ the internuclear distance and $H$ a channel-coupled Hamiltonian matrix (Nkambule et al., 9 Jun 2026). In a two-state form, the coupled time-dependent Schrödinger equation can be written as

$$i\hbar \frac{\partial}{\partial t} \begin{pmatrix} \Psi_1(x,t)\ \Psi_2(x,t) \end{pmatrix} = \begin{pmatrix} \hat H_1 & k_0\delta(x)\ k_0\delta(x) & \hat H_2 \end{pmatrix} \begin{pmatrix} \Psi_1(x,t)\ \Psi_2(x,t) \end{pmatrix},$$

where the non-diagonal matrix element supplies the channel coupling (Rajendran et al., 2018).

In nuclear-reaction applications, the coupled-channels Hamiltonian is written in terms of kinetic, centrifugal, nuclear, absorptive, and coupling contributions. For the neutron–$^{188}$Os problem, the uncoupled operator is

$$\hat H_0= -\frac{\hbar^2}{2\mu}\frac{d^2}{dr^2} +\frac{\hbar^2 l(l+1)}{2\mu r^2} +V_N(r)+iW(r),$$

and the full coupled-channels Hamiltonian for the ground and first excited target states takes the matrix form

$$\hat{H} |\psi\rangle = \begin{pmatrix} \hat{H}_0 + V_{0-0} & V_{0-2} \ V_{2-0} & \hat{H}_0 + V_{2-2} + \epsilon_2 \end{pmatrix} \begin{pmatrix} |\psi_0\rangle \ |\psi_2\rangle \end{pmatrix}.$$

Here the off-diagonal terms encode excitations and de-excitations between target states (Lightfoot et al., 15 Sep 2025).

In electron–molecule collision problems, each diagonal channel may include a potential energy curve, an autoionization width, and a rotational correction,

$$H_{ij}(R)= \left( -\frac{1}{2\mu}\frac{d^2}{dR^2} +V_i(R)-i\frac{\Gamma_i(R)}{2} +V_{\mathrm{rot}}(R)\delta_{ij} \right)\delta_{ij} +H_{ij}^{\mathrm{coup}}(R),$$

with off-diagonal rotational coupling for $\Delta\Lambda=\pm 1$ transitions such as $\Sigma$–$R$0 and $R$1–$R$2 (Nkambule et al., 9 Jun 2026).

These formulations show that the term “channel” is used broadly: it may denote electronic states, rovibrational sectors, target excitations, or other coupled subspaces. This suggests that TDCCWP is best understood as a propagation framework for matrix-valued quantum dynamics rather than a single fixed Hamiltonian model.

2. Propagation algorithms and computational structure

A central numerical issue in TDCCWP is the cost of propagating a wave packet under a coupled Hamiltonian. A widely used baseline is the Strang split-operator approximation,

$R$3

which is second-order accurate in time, with the kinetic term efficiently propagated using Fast Fourier Transforms (Aharonovich et al., 2016). When $R$4 is matrix-valued on multiple coupled potential surfaces, the potential step becomes expensive because it requires matrix exponentiation at each grid point and time step.

An important optimization is split potential propagation, which decomposes the potential matrix into a diagonal space-dependent part $R$5 and an off-diagonal time-dependent coupling-field $R$6,

$R$7

and then applies a second Strang splitting,

$R$8

The diagonal exponential is trivial at each grid point, whereas the off-diagonal matrix needs to be exponentiated only once per time step, independent of the spatial grid (Aharonovich et al., 2016).

Other TDCCWP implementations use different propagators. The neutron-capture study employs a modified Chebyshev propagator for the coupled-channels wave packet, including complex absorption (Lightfoot et al., 15 Sep 2025). The HeH$R$9 isotopologue study uses a Crank–Nicolson scheme for stable propagation in the presence of absorbing boundary conditions and complex potentials (Nkambule et al., 9 Jun 2026). A 2025 proof-of-principle study presents Rothe’s method for propagating linear combinations of explicitly correlated Gaussians and reports that the approach very closely reproduce the virtually exact results of grid-based propagation for two model systems, providing further evidence that ECGs constitute a viable alternative to purely grid-based simulations of coupled nuclear-electronic dynamics driven by intense laser pulses (Woźniak et al., 12 Mar 2025).

A plausible implication is that TDCCWP is methodologically defined more by direct time propagation of coupled wave-packet amplitudes than by any single integrator. Split-operator, Chebyshev, Crank–Nicolson, and Rothe formulations all occupy this broader numerical space.

3. Initialization, coupling structure, and representation choice

The initial state in TDCCWP is often a Gaussian wave packet. In the neutron-capture formulation, the incident neutron is represented as

$H$0

where $H$1 is the initial position, $H$2 the wave-packet width, and $H$3 the incident wavenumber (Lightfoot et al., 15 Sep 2025). In the exact two-state point-coupling problem, the initial wave packet in state 1 is also Gaussian, centered at $H$4 with mean momentum $H$5, while state 2 is initially unoccupied (Rajendran et al., 2018).

A distinctive extension of TDCCWP is the inclusion of temperature through the initial state. For neutron capture on $H$6Os, thermal effects are incorporated by constructing a mixed initial state with Boltzmann factors,

$H$7

or, in the two-state case,

$H$8

with normalization applied (Lightfoot et al., 15 Sep 2025). This embeds temperature at the wave-function level rather than only through a Maxwell-Boltzmann folding at the end.

The coupling structure depends on the physical problem. In molecular surface dynamics, the off-diagonal term may be a time-dependent coupling field, often nearly independent of the spatial grid, while the diagonal terms represent the potential energy surfaces (Aharonovich et al., 2016). In nuclear coupled-channels calculations, the couplings derive from deformation of the nuclear potential and connect target states such as $H$9 and $$i\hbar \frac{\partial}{\partial t} \begin{pmatrix} \Psi_1(x,t)\ \Psi_2(x,t) \end{pmatrix} = \begin{pmatrix} \hat H_1 & k_0\delta(x)\ k_0\delta(x) & \hat H_2 \end{pmatrix} \begin{pmatrix} \Psi_1(x,t)\ \Psi_2(x,t) \end{pmatrix},$$0 (Lightfoot et al., 15 Sep 2025). In HeH$$i\hbar \frac{\partial}{\partial t} \begin{pmatrix} \Psi_1(x,t)\ \Psi_2(x,t) \end{pmatrix} = \begin{pmatrix} \hat H_1 & k_0\delta(x)\ k_0\delta(x) & \hat H_2 \end{pmatrix} \begin{pmatrix} \Psi_1(x,t)\ \Psi_2(x,t) \end{pmatrix},$$1, rotational coupling explicitly transfers population among $$i\hbar \frac{\partial}{\partial t} \begin{pmatrix} \Psi_1(x,t)\ \Psi_2(x,t) \end{pmatrix} = \begin{pmatrix} \hat H_1 & k_0\delta(x)\ k_0\delta(x) & \hat H_2 \end{pmatrix} \begin{pmatrix} \Psi_1(x,t)\ \Psi_2(x,t) \end{pmatrix},$$2, $$i\hbar \frac{\partial}{\partial t} \begin{pmatrix} \Psi_1(x,t)\ \Psi_2(x,t) \end{pmatrix} = \begin{pmatrix} \hat H_1 & k_0\delta(x)\ k_0\delta(x) & \hat H_2 \end{pmatrix} \begin{pmatrix} \Psi_1(x,t)\ \Psi_2(x,t) \end{pmatrix},$$3, and $$i\hbar \frac{\partial}{\partial t} \begin{pmatrix} \Psi_1(x,t)\ \Psi_2(x,t) \end{pmatrix} = \begin{pmatrix} \hat H_1 & k_0\delta(x)\ k_0\delta(x) & \hat H_2 \end{pmatrix} \begin{pmatrix} \Psi_1(x,t)\ \Psi_2(x,t) \end{pmatrix},$$4 symmetries (Nkambule et al., 9 Jun 2026).

Representation choice is likewise central. The HeH$$i\hbar \frac{\partial}{\partial t} \begin{pmatrix} \Psi_1(x,t)\ \Psi_2(x,t) \end{pmatrix} = \begin{pmatrix} \hat H_1 & k_0\delta(x)\ k_0\delta(x) & \hat H_2 \end{pmatrix} \begin{pmatrix} \Psi_1(x,t)\ \Psi_2(x,t) \end{pmatrix},$$5 study treats nuclear dynamics in both adiabatic and strictly diabatic representations. In the adiabatic basis, nonadiabatic effects are encoded in second-order derivative couplings; in the diabatic basis, those derivative terms vanish by construction and are replaced by direct electronic couplings (Nkambule et al., 9 Jun 2026). This suggests that TDCCWP is compatible with either representation, provided the coupling topology is encoded consistently in the Hamiltonian.

4. Extraction of observables

TDCCWP calculations usually infer observables from the propagated wave packet after the interaction region has been traversed. In the neutron-capture application, transmission and reflection are obtained after the reflected wave packet returns to its original location, with

$$i\hbar \frac{\partial}{\partial t} \begin{pmatrix} \Psi_1(x,t)\ \Psi_2(x,t) \end{pmatrix} = \begin{pmatrix} \hat H_1 & k_0\delta(x)\ k_0\delta(x) & \hat H_2 \end{pmatrix} \begin{pmatrix} \Psi_1(x,t)\ \Psi_2(x,t) \end{pmatrix},$$6

and the temperature-dependent cross section given by

$$i\hbar \frac{\partial}{\partial t} \begin{pmatrix} \Psi_1(x,t)\ \Psi_2(x,t) \end{pmatrix} = \begin{pmatrix} \hat H_1 & k_0\delta(x)\ k_0\delta(x) & \hat H_2 \end{pmatrix} \begin{pmatrix} \Psi_1(x,t)\ \Psi_2(x,t) \end{pmatrix},$$7

The associated reaction rate is computed from a velocity-weighted average of the cross section (Lightfoot et al., 15 Sep 2025).

In the HeH$$i\hbar \frac{\partial}{\partial t} \begin{pmatrix} \Psi_1(x,t)\ \Psi_2(x,t) \end{pmatrix} = \begin{pmatrix} \hat H_1 & k_0\delta(x)\ k_0\delta(x) & \hat H_2 \end{pmatrix} \begin{pmatrix} \Psi_1(x,t)\ \Psi_2(x,t) \end{pmatrix},$$8 study, the energy-resolved outgoing flux is extracted by Fourier transforming the asymptotic wave packet,

$$i\hbar \frac{\partial}{\partial t} \begin{pmatrix} \Psi_1(x,t)\ \Psi_2(x,t) \end{pmatrix} = \begin{pmatrix} \hat H_1 & k_0\delta(x)\ k_0\delta(x) & \hat H_2 \end{pmatrix} \begin{pmatrix} \Psi_1(x,t)\ \Psi_2(x,t) \end{pmatrix},$$9

followed by

$^{188}$0

The total cross section is then obtained by summing over resonant states (Nkambule et al., 9 Jun 2026).

The exact two-state Dirac delta model yields a different kind of output: full analytic wave functions in the time domain for both states, not only transition probabilities. The state-1 solution is separated into a free Gaussian contribution and a scattered term, while the state-2 solution is given as an exact integral expression in time (Rajendran et al., 2018). That model is especially useful where transient dynamics, interference, and the complete spatial-temporal structure of the wave packet matter.

Problem class	TDCCWP ingredients	Output
Neutron capture on $^{188}$1Os	Gaussian wave packet, coupled channels, absorptive potential, temperature-dependent initial state	Transmission coefficients, cross sections, reaction rates
HeH$^{188}$2 isotopologues	23 coupled electronic states, rotational coupling, adiabatic/diabatic propagation, local complex potential	DR and RIP cross sections
Two-state point coupling	Gaussian initial packet, Dirac delta non-diagonal coupling, exact time-domain solution	Full wave functions for both states

These output strategies show that TDCCWP is not restricted to a single observable. Depending on the application, it can target survival amplitudes, capture probabilities, dissociation cross sections, or fully resolved wave-function dynamics.

5. Accuracy, efficiency, and benchmark status

A recurrent result in TDCCWP research is that numerical acceleration must be balanced against controllable splitting error. For split potential propagation, both the standard split-operator method and the split-potential method are formally second-order accurate, and the additional error from potential splitting has leading order $^{188}$3, of the same order as the primary kinetic/potential split. The paper reports that the overlap of wave packets propagated with the standard and split-potential methods is virtually identical across a range of time steps and interaction strengths, and that even for strong coupling fields the split-potential error remains comparable to or smaller than the primary split-operator error (Aharonovich et al., 2016).

The computational gain is explicit. For a 2-surface system on a spatial grid of 512 points over $^{188}$4 time steps, the potential propagation time is reduced by approximately $^{188}$5 ($^{188}$6 s $^{188}$7 $^{188}$8 s), and the overall simulation time is reduced by approximately $^{188}$9, with the kinetic FFT step then becoming the dominant cost (Aharonovich et al., 2016). The same work states that the benefit scales even more for three or more coupled surfaces.

Benchmarking also enters through analytically solvable models. The exact time-domain solution for a Gaussian wave packet in two potential curves coupled at a point is explicitly identified as a benchmark result for testing numerical and approximate time-dependent coupled-channel wave-packet algorithms (Rajendran et al., 2018). Because it provides the complete time-domain wave function for arbitrary delta coupling strength, wave-packet width, mean momentum, and potential separation, it serves as a stringent reference for transient coupled-channel propagation.

A complementary benchmark perspective appears in the Rothe-propagation study, which shows that linear combinations of explicitly correlated Gaussians very closely reproduce the virtually exact results of grid-based propagation for the model systems considered (Woźniak et al., 12 Mar 2025). This does not replace grid-based TDCCWP, but it indicates that basis-set formulations can reproduce the same dynamics with high fidelity in selected non-Born-Oppenheimer settings.

6. Applications, scope, and debated interpretations

TDCCWP has been applied to materially different physical regimes. In nuclear astrophysics, the method is used for neutron capture on $$\hat H_0= -\frac{\hbar^2}{2\mu}\frac{d^2}{dr^2} +\frac{\hbar^2 l(l+1)}{2\mu r^2} +V_N(r)+iW(r),$$0Os with a many-body nuclear potential obtained from Hartree-Fock and fitted with a Woods-Saxon form (Lightfoot et al., 15 Sep 2025). In that application, increasing the environmental temperature leads to a decrease in capture or transmission coefficients and hence the cross section, and the reaction rates decrease by up to $$\hat H_0= -\frac{\hbar^2}{2\mu}\frac{d^2}{dr^2} +\frac{\hbar^2 l(l+1)}{2\mu r^2} +V_N(r)+iW(r),$$1 for the highest incident energies studied (Lightfoot et al., 15 Sep 2025).

In electron–molecule collision dynamics, a time-dependent wave-packet study of HeH$$\hat H_0= -\frac{\hbar^2}{2\mu}\frac{d^2}{dr^2} +\frac{\hbar^2 l(l+1)}{2\mu r^2} +V_N(r)+iW(r),$$2 isotopologues includes 23 coupled electronic states, $$\hat H_0= -\frac{\hbar^2}{2\mu}\frac{d^2}{dr^2} +\frac{\hbar^2 l(l+1)}{2\mu r^2} +V_N(r)+iW(r),$$3, $$\hat H_0= -\frac{\hbar^2}{2\mu}\frac{d^2}{dr^2} +\frac{\hbar^2 l(l+1)}{2\mu r^2} +V_N(r)+iW(r),$$4, and $$\hat H_0= -\frac{\hbar^2}{2\mu}\frac{d^2}{dr^2} +\frac{\hbar^2 l(l+1)}{2\mu r^2} +V_N(r)+iW(r),$$5 symmetries, both adiabatic and strictly diabatic representations, and explicit rotational coupling over collision energies from $$\hat H_0= -\frac{\hbar^2}{2\mu}\frac{d^2}{dr^2} +\frac{\hbar^2 l(l+1)}{2\mu r^2} +V_N(r)+iW(r),$$6 to $$\hat H_0= -\frac{\hbar^2}{2\mu}\frac{d^2}{dr^2} +\frac{\hbar^2 l(l+1)}{2\mu r^2} +V_N(r)+iW(r),$$7 eV (Nkambule et al., 9 Jun 2026). The inclusion of a large manifold of resonant states and rotational couplings significantly enhances the dissociative recombination cross section relative to earlier theoretical studies. In the diabatic representation, $$\hat H_0= -\frac{\hbar^2}{2\mu}\frac{d^2}{dr^2} +\frac{\hbar^2 l(l+1)}{2\mu r^2} +V_N(r)+iW(r),$$8 states dominate the recombination dynamics, whereas in the adiabatic representation $$\hat H_0= -\frac{\hbar^2}{2\mu}\frac{d^2}{dr^2} +\frac{\hbar^2 l(l+1)}{2\mu r^2} +V_N(r)+iW(r),$$9 and $$\hat{H} |\psi\rangle = \begin{pmatrix} \hat{H}_0 + V_{0-0} & V_{0-2} \ V_{2-0} & \hat{H}_0 + V_{2-2} + \epsilon_2 \end{pmatrix} \begin{pmatrix} |\psi_0\rangle \ |\psi_2\rangle \end{pmatrix}.$$0 contribute significantly at low collision energies. For resonant ion-pair formation, two different diabatization schemes yield systematically larger cross sections than previous models, and isotopic effects show a clear inverse dependence of cross section magnitude on reduced mass (Nkambule et al., 9 Jun 2026).

A distinct issue concerns barrier transmission and the interpretation of tunneling in wave-packet superposition approaches. A 2022 phase-space analysis of a time-dependent variational approach with superposed wave packets concludes that passage over the barrier can occur due to the high-momentum components in the incoming state corresponding to energies above the barrier height, which is of classical nature and needs to be distinguished from true quantum tunneling (Ono, 2022). The same work reports difficulty in guaranteeing the energy conservation when the energies of different exit channels are individually measured and explicitly disagrees with the conclusion that quantum tunneling was simulated by the model (Ono, 2022).

That controversy does not invalidate TDCCWP as a propagation framework. Rather, it sharpens a methodological point: in coupled-channel wave-packet calculations, transmitted flux is not automatically synonymous with sub-barrier tunneling. Careful analysis of the initial momentum distribution, phase-space evolution, channel-resolved energy accounting, and barrier-thickness dependence is required before assigning a transmission component to genuine quantum tunneling.