1D Matrix Charge Transfer Models

• 1D matrix charge transfer models are one-dimensional systems where charge dynamics and ordering are encoded in finite-dimensional matrices, uniting concepts from MPS representations to matrix-valued Schrödinger equations.
• These models utilize various matrix structures—such as covariance, capacitance, and coupled-cluster matrices—to capture detailed charge transport mechanisms and topological pumping phenomena.
• They offer actionable insights for analyzing correlated lattice phases and designing optoelectronic devices by linking rigorous analytical methods with practical transport and scattering applications.

One-dimensional matrix charge transfer models are one-dimensional systems in which charge motion is organized by an explicit matrix structure at the level of hopping amplitudes, covariance matrices, capacitance matrices, matrix-product tensors, or matrix-valued Schrödinger operators. In the literature considered here, the term encompasses correlated lattice-fermion chains, covariance-matrix-driven auxiliary pumps, bilinear tunnel-junction arrays, effective charge-only Hamiltonians built from many-body Wannier functions, and non-self-adjoint $2\times2$ charge-transfer equations with moving potentials. Taken together, these works suggest that “matrix charge transfer” is less a single model class than a family of 1D constructions in which charge transport, charge ordering, and charge-transfer excitations are encoded in finite-dimensional matrix data (Snyman et al., 3 Jul 2025, Wawer et al., 2020, Walker et al., 2013, Chen et al., 1 May 2025).

1. Conceptual scope and defining structures

In one important usage, a one-dimensional matrix charge transfer model is a tight-binding chain in which charge transfer processes are governed by matrices rather than simple scalar hopping amplitudes, and in which both the ground state and the effective Hamiltonians can be written exactly in matrix-product form. This meaning is developed explicitly for the half-filled spinless $t\text{-}V$ chain, where a correlated charge-insulator phase is recast as a coupled-cluster, matrix-product, and Rokhsar–Kivelson state of a correlated hopping Hamiltonian (Snyman et al., 3 Jul 2025).

A second usage treats the one-body covariance matrix

$$m_{\mu\nu}(k)=\langle \hat c^\dagger_\mu(k)\hat c_\nu(k)\rangle$$

as the central matrix object. In that construction, the covariance matrix of a 1D many-body state defines a fictitious Hamiltonian for an auxiliary fermion chain,

$$H_\text{fict}=\eta\sum_k\sum_{\mu,\nu=1}^p \hat a^\dagger_\mu(k)\,m_{\mu\nu}(k)\,\hat a_\nu(k),$$

and quantized charge transfer in the auxiliary chain is determined by the Chern number of this covariance-matrix Hamiltonian (Wawer et al., 2020).

A third usage is explicitly electrostatic. In bilinear tunnel-junction arrays, charge transfer is governed by a block-capacitance matrix $\mathbf C$, and the Hamiltonian has the quadratic form

$$\mathcal H=\frac{1}{2}\mathbf Q^T\mathbf C^{-1}\mathbf Q-\mathbf Q^T\mathbf V_\text{eff},$$

so the matrix object controlling transport is the inverse capacitance kernel rather than a single-particle hopping matrix (Walker et al., 2013).

A fourth usage arises in PDE and scattering theory. There the model is a time-dependent, matrix-valued Schrödinger equation with multiple moving potentials,

$i\partial_t\vec\psi+\begin{pmatrix}\partial_x^2&0\0&-\partial_x^2\end{pmatrix}\vec\psi+\sum_{j=1}^m V_j(t)\vec\psi=0,$

where each $V_j(t)$ is a $2\times2$ matrix potential associated with a moving center. In that setting, “charge transfer” refers to scattering among several moving localized potentials, especially in linearizations of multi-soliton dynamics (Chen et al., 1 May 2025, Chen et al., 13 Oct 2025).

2. Microscopic lattice Hamiltonians and ordered phases

A canonical correlated example is the half-filled spinless $t\text{-}V$ chain on $t\text{-}V$0 sites,

$t\text{-}V$1

with periodic boundary conditions. For $t\text{-}V$2 the model is a Luttinger liquid, while at half filling and strong repulsion $t\text{-}V$3 it enters a charge-ordered insulating phase with a staggered CDW pattern $t\text{-}V$4 or $t\text{-}V$5. The covariance-matrix spectrum in this phase reveals a post-Hartree–Fock structure: to order $t\text{-}V$6,

$t\text{-}V$7

so the dependence on $t\text{-}V$8 signals emergent four-site disruptions of the underlying staggered state rather than merely two-site alternation. Those disruptions are captured by a compact four-fermion coupled-cluster doubles ansatz, by a bond-dimension-four matrix-product state, and by a Rokhsar–Kivelson state of a quantum tetramer model (Snyman et al., 3 Jul 2025).

Quarter-filled organic charge-transfer systems are commonly modeled by the extended Peierls–Hubbard Hamiltonian

$t\text{-}V$9

with dimerized hopping

$$m_{\mu\nu}(k)=\langle \hat c^\dagger_\mu(k)\hat c_\nu(k)\rangle$$0

at filling $$m_{\mu\nu}(k)=\langle \hat c^\dagger_\mu(k)\hat c_\nu(k)\rangle$$1. In this setting $$m_{\mu\nu}(k)=\langle \hat c^\dagger_\mu(k)\hat c_\nu(k)\rangle$$2, so the characteristic instabilities are $$m_{\mu\nu}(k)=\langle \hat c^\dagger_\mu(k)\hat c_\nu(k)\rangle$$3 and $$m_{\mu\nu}(k)=\langle \hat c^\dagger_\mu(k)\hat c_\nu(k)\rangle$$4. Charge and bond modulations are parameterized as

$$m_{\mu\nu}(k)=\langle \hat c^\dagger_\mu(k)\hat c_\nu(k)\rangle$$5

$$m_{\mu\nu}(k)=\langle \hat c^\dagger_\mu(k)\hat c_\nu(k)\rangle$$6

making the matrix charge-transfer problem one of competing period-four and period-two charge and bond textures in a quarter-filled correlated chain (Rincon et al., 2014).

At the same filling, the Peierls-extended Hubbard model with both SSH-like and Holstein-like couplings,

$$m_{\mu\nu}(k)=\langle \hat c^\dagger_\mu(k)\hat c_\nu(k)\rangle$$7

supports two distinct bond-charge-density waves with the same $$m_{\mu\nu}(k)=\langle \hat c^\dagger_\mu(k)\hat c_\nu(k)\rangle$$8 charge pattern. BCDW2 has the SMWM bond pattern and in general coexists with a large magnitude charge order, whereas BCDW1 has the SWSW$$m_{\mu\nu}(k)=\langle \hat c^\dagger_\mu(k)\hat c_\nu(k)\rangle$$9 pattern and is characterized by weak charge order. The boundary between them is identified in the limit of infinitesimal electron-phonon coupling by the bond-susceptibility ratio

$$H_\text{fict}=\eta\sum_k\sum_{\mu,\nu=1}^p \hat a^\dagger_\mu(k)\,m_{\mu\nu}(k)\,\hat a_\nu(k),$$0

with BCDW1 occurring when $$H_\text{fict}=\eta\sum_k\sum_{\mu,\nu=1}^p \hat a^\dagger_\mu(k)\,m_{\mu\nu}(k)\,\hat a_\nu(k),$$1 (Clay et al., 2016).

A different microscopic realization is the 1D $$H_\text{fict}=\eta\sum_k\sum_{\mu,\nu=1}^p \hat a^\dagger_\mu(k)\,m_{\mu\nu}(k)\,\hat a_\nu(k),$$2 chain with alternating transition-metal $$H_\text{fict}=\eta\sum_k\sum_{\mu,\nu=1}^p \hat a^\dagger_\mu(k)\,m_{\mu\nu}(k)\,\hat a_\nu(k),$$3 and ligand $$H_\text{fict}=\eta\sum_k\sum_{\mu,\nu=1}^p \hat a^\dagger_\mu(k)\,m_{\mu\nu}(k)\,\hat a_\nu(k),$$4 orbitals,

$$H_\text{fict}=\eta\sum_k\sum_{\mu,\nu=1}^p \hat a^\dagger_\mu(k)\,m_{\mu\nu}(k)\,\hat a_\nu(k),$$5

$$H_\text{fict}=\eta\sum_k\sum_{\mu,\nu=1}^p \hat a^\dagger_\mu(k)\,m_{\mu\nu}(k)\,\hat a_\nu(k),$$6

whose noninteracting Bloch Hamiltonian is an explicit $$H_\text{fict}=\eta\sum_k\sum_{\mu,\nu=1}^p \hat a^\dagger_\mu(k)\,m_{\mu\nu}(k)\,\hat a_\nu(k),$$7 matrix in the $$H_\text{fict}=\eta\sum_k\sum_{\mu,\nu=1}^p \hat a^\dagger_\mu(k)\,m_{\mu\nu}(k)\,\hat a_\nu(k),$$8 basis. The charge-transfer energy is

$$H_\text{fict}=\eta\sum_k\sum_{\mu,\nu=1}^p \hat a^\dagger_\mu(k)\,m_{\mu\nu}(k)\,\hat a_\nu(k),$$9

Varying $\mathbf C$0 interpolates between the Mott insulating regime $\mathbf C$1, the charge-transfer-insulating regime $\mathbf C$2, and the negative charge-transfer regime $\mathbf C$3 (Milner et al., 1 Aug 2025).

3. Exact and effective matrix encodings of charge transfer

In the strong-coupling $\mathbf C$4 chain, the dominant post-Hartree–Fock fluctuation is a four-site disruption of the staggered CDW pattern. In a Wannier basis adapted to the CDW order, these disruptions are generated by

$\mathbf C$5

and the coupled-cluster state

$\mathbf C$6

is a thermodynamically extensive superposition of non-overlapping hard-core tetramers. The same state admits an exact matrix-product representation with bond dimension exactly four, and it is also the unique zero-energy ground state of both a correlated-hopping parent Hamiltonian $\mathbf C$7 and an RK parent Hamiltonian for a 1D quantum tetramer model (Snyman et al., 3 Jul 2025).

The covariance-matrix construction gives a different form of exact matrix encoding. The two-chain Hamiltonian

$\mathbf C$8

induces, in the weak-coupling limit, an effective auxiliary-chain Hamiltonian $\mathbf C$9 built from the covariance matrix $$\mathcal H=\frac{1}{2}\mathbf Q^T\mathbf C^{-1}\mathbf Q-\mathbf Q^T\mathbf V_\text{eff},$$0. For non-interacting pumps, the topology of $$\mathcal H=\frac{1}{2}\mathbf Q^T\mathbf C^{-1}\mathbf Q-\mathbf Q^T\mathbf V_\text{eff},$$1 is identical to the TKNN invariant of the original system, while for interacting systems the transported charge in the auxiliary chain defines a topological invariant even when the single-particle TKNN number does not exist. In the Rice–Mele model at half filling the auxiliary chain exhibits a quantized transport of one particle per cycle, and in the extended superlattice Bose–Hubbard model at quarter filling the induced charge pump is fractional, with $$\mathcal H=\frac{1}{2}\mathbf Q^T\mathbf C^{-1}\mathbf Q-\mathbf Q^T\mathbf V_\text{eff},$$2 per physical cycle (Wawer et al., 2020).

A complementary reduction appears in the “charge model” for the half-filled 1D Hubbard and extended Hubbard chains. There the Hilbert space is projected to a subspace $$\mathcal H=\frac{1}{2}\mathbf Q^T\mathbf C^{-1}\mathbf Q-\mathbf Q^T\mathbf V_\text{eff},$$3 spanned by configurations of holons $$\mathcal H=\frac{1}{2}\mathbf Q^T\mathbf C^{-1}\mathbf Q-\mathbf Q^T\mathbf V_\text{eff},$$4, doublons $$\mathcal H=\frac{1}{2}\mathbf Q^T\mathbf C^{-1}\mathbf Q-\mathbf Q^T\mathbf V_\text{eff},$$5, and singly occupied sites, while the spin wave function for each sector with $$\mathcal H=\frac{1}{2}\mathbf Q^T\mathbf C^{-1}\mathbf Q-\mathbf Q^T\mathbf V_\text{eff},$$6 holon–doublon pairs is fixed to the Heisenberg ground state for $$\mathcal H=\frac{1}{2}\mathbf Q^T\mathbf C^{-1}\mathbf Q-\mathbf Q^T\mathbf V_\text{eff},$$7 sites. The projected Hamiltonian

$$\mathcal H=\frac{1}{2}\mathbf Q^T\mathbf C^{-1}\mathbf Q-\mathbf Q^T\mathbf V_\text{eff},$$8

keeps charge fluctuations, including pair creation and annihilation, but replaces detailed spin dynamics by a boundary phase $$\mathcal H=\frac{1}{2}\mathbf Q^T\mathbf C^{-1}\mathbf Q-\mathbf Q^T\mathbf V_\text{eff},$$9 and an overlap amplitude $i\partial_t\vec\psi+\begin{pmatrix}\partial_x^2&0\0&-\partial_x^2\end{pmatrix}\vec\psi+\sum_{j=1}^m V_j(t)\vec\psi=0,$0, taken as $i\partial_t\vec\psi+\begin{pmatrix}\partial_x^2&0\0&-\partial_x^2\end{pmatrix}\vec\psi+\sum_{j=1}^m V_j(t)\vec\psi=0,$1 in the thermodynamic-limit approximation. Many-body Wannier functions localized in the holon–doublon separation $i\partial_t\vec\psi+\begin{pmatrix}\partial_x^2&0\0&-\partial_x^2\end{pmatrix}\vec\psi+\sum_{j=1}^m V_j(t)\vec\psi=0,$2 then produce an effective quasi-tridiagonal Hamiltonian matrix $i\partial_t\vec\psi+\begin{pmatrix}\partial_x^2&0\0&-\partial_x^2\end{pmatrix}\vec\psi+\sum_{j=1}^m V_j(t)\vec\psi=0,$3, so optical excitations reduce to a one-dimensional tight-binding problem in separation space (Ohmura et al., 2019).

Taken together, these constructions show that the “matrix” in 1D charge transfer models may encode the full wavefunction as an MPS tensor network, the one-body correlations as a covariance-matrix Hamiltonian, or a reduced many-body charge basis through an effective hopping matrix. This suggests that matrix structure is a reorganization principle rather than a unique microscopic prescription.

4. Nonequilibrium transport, pumping, and correlation dynamics

Under ultrafast optical driving, the quarter-filled extended Peierls–Hubbard chain exhibits a pronounced selectivity among charge-transfer channels. With a few-cycle vector potential

$i\partial_t\vec\psi+\begin{pmatrix}\partial_x^2&0\0&-\partial_x^2\end{pmatrix}\vec\psi+\sum_{j=1}^m V_j(t)\vec\psi=0,$4

implemented through Peierls phases in the hopping, time-dependent DMRG shows that the $i\partial_t\vec\psi+\begin{pmatrix}\partial_x^2&0\0&-\partial_x^2\end{pmatrix}\vec\psi+\sum_{j=1}^m V_j(t)\vec\psi=0,$5 bond and charge instabilities dominate the photogenerated time-dependent behavior, while the $i\partial_t\vec\psi+\begin{pmatrix}\partial_x^2&0\0&-\partial_x^2\end{pmatrix}\vec\psi+\sum_{j=1}^m V_j(t)\vec\psi=0,$6 instabilities remain largely unaffected. This conclusion holds for large and moderate dimerization, with or without strong nearest-neighbor $i\partial_t\vec\psi+\begin{pmatrix}\partial_x^2&0\0&-\partial_x^2\end{pmatrix}\vec\psi+\sum_{j=1}^m V_j(t)\vec\psi=0,$7, and for resonant as well as nonresonant perturbations. The authors attribute the selectivity mainly to the fact that the current operator, or equivalently the kinetic energy operator, only strongly couples $i\partial_t\vec\psi+\begin{pmatrix}\partial_x^2&0\0&-\partial_x^2\end{pmatrix}\vec\psi+\sum_{j=1}^m V_j(t)\vec\psi=0,$8 states, and they suggest potential applications of charge-transfer systems with slow phononic dynamics as optoelectronic switching devices (Rincon et al., 2014).

In topological pumping problems, the same idea of charge transfer is realized geometrically. A cyclic adiabatic drive $i\partial_t\vec\psi+\begin{pmatrix}\partial_x^2&0\0&-\partial_x^2\end{pmatrix}\vec\psi+\sum_{j=1}^m V_j(t)\vec\psi=0,$9 on an insulating system chain induces a quantized transported charge in an auxiliary chain,

$V_j(t)$0

where $V_j(t)$1 is the Chern number of the covariance-matrix Hamiltonian felt by the auxiliary fermions. In non-interacting systems this reproduces the TKNN invariant of the original Thouless pump; in interacting systems it provides a covariance-matrix-based topological invariant that in certain cases agrees with the many-body Niu–Thouless–Wu invariant (Wawer et al., 2020).

The bilinear tunnel-junction array realizes yet another transport regime. Two parallel 1D rails are coupled only by an inter-rail capacitance $V_j(t)$2, so transport is simulated by kinetic Monte Carlo within the orthodox theory of single-electron tunneling. The full $V_j(t)$3 capacitance matrix generates two interaction lengths $V_j(t)$4, corresponding to symmetric and antisymmetric charge modes. When only one rail is voltage biased, stationary charge states form in the undriven rail; under symmetric bias of both rails, the site at which positive and negative charge carriers recombine can drift throughout the array; and under antisymmetric “escalator” bias, strong inter-rail anticorrelations appear, corresponding to tightly bound dipole transport. Charge densities and auto- and cross-correlation functions are the primary diagnostics of these matrix-governed transport regimes (Walker et al., 2013).

5. Materials, spectroscopy, and experimentally relevant order parameters

The quasi-one-dimensional organic charge-transfer salt $V_j(t)$5-$V_j(t)$6 provides an experimentally explicit realization of a 1D charge-ordered chain. The donor molecules form stacks along the crystallographic $V_j(t)$7-axis, the conduction band is quarter-filled, the system is a charge-ordered Mott insulator at room temperature and ambient pressure, and a charge disproportionation ratio of about $V_j(t)$8 is observed for neighboring molecules. Infrared spectroscopy gives an optical gap of $V_j(t)$9, discussed as domain-wall excitations of a one-dimensional Wigner lattice. Thermal-expansion and dielectric measurements reveal three relaxational processes: a glasslike ethylene-group freezing with $2\times2$0 and a glass temperature of about $2\times2$1; a process with $2\times2$2 and $2\times2$3, consistent with Wigner-lattice domain-wall excitations; and a faster Arrhenius process with $2\times2$4 and $2\times2$5 (Fischer et al., 2018).

In quarter-filled charge-transfer solids described by the Peierls-extended Hubbard model, bond patterns provide experimentally accessible markers of matrix-organized charge transfer. The SMWM pattern identifies BCDW2 and is associated with large charge disproportionation, while the SWSW$2\times2$6 pattern identifies BCDW1 and is associated with weak charge order. This distinction is used to interpret $2\times2$7 and $2\times2$8, whose low-temperature overlap pattern is SMWM and whose large-amplitude charge order places them in the BCDW2 sector (Clay et al., 2016).

The 1D $2\times2$9 chain supplies a spectroscopic classification tool for transition-metal oxides. Density matrix renormalization group calculations of the single-particle spectral function, x-ray absorption spectrum, and dynamical spin structure factor are performed as functions of $t\text{-}V$0 and $t\text{-}V$1, spanning from the Mott insulating regime $t\text{-}V$2 to the negative charge transfer regime $t\text{-}V$3. The model resolves how the orbital content of low-energy states, the relative positions of $t\text{-}V$4- and $t\text{-}V$5-derived features, and the dynamical spin spectral weight evolve across the Zaanen–Sawatzky–Allen classification (Milner et al., 1 Aug 2025).

These examples show that experimentally relevant order parameters are often still one-body or few-body objects—charge disproportionation, bond modulation, optical conductivity, x-ray absorption, dielectric relaxation—but their interpretation depends on an underlying matrix organization of the charge sector.

6. Matrix-valued charge transfer equations and dispersive theory

In the PDE literature, one-dimensional matrix charge transfer models are linear Schrödinger equations with multiple moving potentials. The 2025 analysis develops the scattering theory for the non-self-adjoint $t\text{-}V$6 system

$t\text{-}V$7

with

$t\text{-}V$8

under the assumption that the potentials move at significantly different velocities, even in the presence of unstable modes. The analysis proves the existence of wave operators, asymptotic completeness, and pointwise decay of solutions, without requiring the absence of threshold resonances, and introduces scattering and continuous projections adapted to the moving-center framework (Chen et al., 1 May 2025).

A subsequent refinement removes the large-velocity separation assumption. The model remains the same $t\text{-}V$9 matrix Schrödinger charge-transfer equation arising from linearization of 1D NLS around a multi-soliton background, but the construction of distorted Fourier transforms is adapted to the multi-potential framework in a more precise way. The resulting theory gives scattering decompositions, dispersive estimates, and asymptotic completeness for general one-dimensional charge transfer models with merely distinct ordered velocities $t\text{-}V$00, completing the linear theory needed for analyzing asymptotic stability and collision phenomena for multi-solitons in a general setting (Chen et al., 13 Oct 2025).

In this analytical setting, “matrix charge transfer” has a literal operator-theoretic meaning: the moving centers are encoded by time-dependent $t\text{-}V$01 potentials, the continuous spectrum is resolved by distorted Fourier transforms, and the solution decomposes into a dispersive component together with moving discrete and generalized modes. A plausible implication is that the term “charge transfer model” in analysis and in condensed-matter theory denotes structurally different objects, but in both cases it identifies systems whose nontrivial dynamics are organized by several competing localized channels of transfer.

Across these domains, one-dimensional matrix charge transfer models provide a common language for several phenomena that are otherwise usually separated: charge ordering beyond Hartree–Fock, low-rank tensor-network encodings of correlated insulators, covariance-matrix topological pumping, capacitance-matrix-controlled Coulomb transport, spectroscopic classification of charge-transfer insulators, and dispersive scattering in matrix Schrödinger systems. The shared feature is not a single Hamiltonian form, but the use of finite-dimensional matrix structure to expose the relevant 1D charge-transfer degrees of freedom and to make them analytically or numerically tractable.