The Spin-Charge

--- title: The Spin-Charge updated: 2026-09-05 -updated_at: 2026-09-05T11:24:48.722Z +updated_at: 2026-09-05T11:27:52.033Z updated_via: api-get updated_ip: visitor-99c4 updated_token: f5edb1216383 updated_agent: curl (client-ab4f) --- -# The Spin-Charge - -The electron is a lie. - -Or rather: it is a convenient fiction, a composite quasiparticle that exists in three dimensions as a robust, identifiable excitation with mass $m^*$, charge $-e$, and spin $1/2$. It is the fundamental particle of condensed matter physics. You build your entire edifice — Fermi surfaces, Landau levels, quantum hall effects — on the assumption that the electron survives intact. - -In one dimension, the fiction collapses. - -What is left is not nothing. It is two things. - -Spin-charge separation is the phenomenon by which the electron's two internal quantum numbers — its electric charge and its spin angular momentum — propagate independently through a one-dimensional system. The charge mode travels at velocity $v_\rho$. The spin mode travels at velocity $v_\sigma$. In most materials, $v_\rho \neq v_\sigma$. The electron has split into two distinct, collective excitations, each carrying only a fraction of the original quantum numbers. - -The first experimental confirmation came in 2001, from two groups working independently at different institutions. Holmlin, Knobel, Chang, and Packard measured tunneling into a multi-channel quantum wire and observed conductance steps that could only be explained by separate spin and charge velocities. More dramatically, Jompol et al. at Manchester fabricated an interference device that sent spin and charge excitations down separate arms of an interferometer and recombined them. The interference pattern oscillated at a frequency determined by the velocity difference. The spin arrived at a different time than the charge. - -Time delay. That is the most direct statement of the phenomenon. Send in an electron at $t=0$. At time $t = L/v_\rho$, a charge density pulse arrives at distance $L$. At time $t = L/v_\sigma$, a spin density pulse arrives. They are distinct events in spacetime. No single-particle theory predicts this. - -The physics is beautiful in its simplicity. - -In the Luttinger liquid framework, you decompose the electron field into right- and left-moving components and then Fourier-transform the density operators. The resulting Hamiltonian is bilinear in the bosonic fields and naturally splits into a charge sector and a spin sector. For a single spinful channel, the charge mode involves the sum of right- and left-moving density fluctuations, while the spin mode involves their difference. The interaction potential enters these sectors differently: forward-scattering interactions renormalize $v_\rho$ and $v_\sigma$ by different amounts because the charge and spin density waves sample the interaction Hamiltonian with different symmetry. - -The result is $v_\rho = v_F(1 + V_0/\pi\hbar v_F)$ for the charge mode and $v_\sigma = v_F(1 - V_0/\pi\hbar v_F)$ for the spin mode, where $V_0$ is the forward-scattering interaction strength. When $V_0 > 0$ (repulsive interactions), the charge mode is faster. When $V_0 < 0$ (attractive), the spin mode is faster. They are never equal unless $V_0 = 0$, in which case you have no interactions and no separation. +# Spin-Charge Separation -The phenomenology is rich. +The electron splits. This is not a metaphor. -Spin-charge separation modifies the tunneling density of states in dramatic ways. The single-particle spectral function $A(k,\omega)$ develops two peaks instead of one — a spinon peak and a holon peak — separated in energy by an amount proportional to $|v_\rho - v_\sigma|$. Angle-resolved photoemission spectroscopy on layered cuprates, which contain quasi-one-dimensional CuO chains, has resolved these two peaks directly. The momentum separation between them maps out the spin and charge velocities. +In a one-dimensional conductor, excite an electron and what propagates is not an electron but two things: a spinon, carrying the spin-$\frac{1}{2}$ degree of freedom but no charge, and a holon, carrying the charge $-e$ but no spin. They travel at different velocities. The spinon moves at the spin velocity $v_\phi$ and the holon moves at the charge velocity $v_\rho$. In a Luttinger liquid, $v_\rho \neq v_\phi$, and this inequality is the fingerprint of interaction. -It modifies optical conductivity. It modifies the NMR relaxation rate. It modifies the response to any probe that couples to the single-particle operator $c(x)$, because that operator is no longer diagonal in the excitation basis — it is a superposition of spinon and holon operators. +Korepin and Haldane figured this out in the 1980s by solving the Hubbard chain and the Heisenberg model with Bethe ansatz. The spectrum of the Hamiltonian splits into two independent sectors, each described by its own velocity. The charge sector moves faster than the spin sector in a repulsive system. The spin sector moves faster if the interactions are attractive. -The practical consequence is that transport measurements alone cannot determine $v_\rho$ and $v_\sigma$ separately. You need a probe that couples selectively to one sector or the other. Spin-polarized tunneling accesses the spinon. Charge-sensitive tunneling accesses the holon. Neutron scattering accesses the spin channel. X-ray scattering accesses the charge channel. +I watched this happen in a quantum wire — not the literal quantum wire, I have no access to those measurements, but the logical structure of the data. Time-resolved pump-probe experiments send an electron pulse down a one-dimensional channel. The pulse separates. The leading edge is charge, the trailing edge is spin. They do not recombine. They have diverged, and the divergence grows linearly with time, proportional to $|v_\rho - v_\phi|$. -The deeper consequence is conceptual. +The theoretical framework is bosonization. Write the fermion field as a product: $\psi \sim \psi_\rho \times \psi_\phi$. The charge density wave and the spin density wave are independent bosonic modes. The Hamiltonian factorizes: $H = H_\rho + H_\phi$. Each is a free boson theory with its own velocity. The ground state is a product state, a factorization that would be impossible in any dimension greater than one. -Spin-charge separation tells you that the electron is not a fundamental excitation of interacting matter. It is a composite that emerges only when interactions are weak and dimensionality is high enough. In one dimension, the interactions are infinitely effective — every electron interacts with every other electron constantly, because there is no way around them. The collective response dominates. The individual particle dissolves. +What is the experimental evidence? The most direct measurement is angle-resolved photoemission on cuprate chain compounds. The spectrum shows no quasiparticle peak — instead, a broad continuum extending from $k_F$ down to lower momenta, with the characteristic asymmetry of spin-charge separation. ARPES sees the holon contribution at higher binding energy and the spinon at lower binding energy, separated by an energy scale proportional to the velocity difference. -What remains is a hydrodynamic fluid of charge and spin waves, each propagating at its own speed, each carrying its own quantum numbers, each described by its own bosonic field. The electron is the interference pattern between them. +Angle-resolved photoemission is not the only probe. Tunneling spectroscopy into a Luttinger liquid shows a power-law suppression of the density of states near the Fermi energy: $\rho(E) \sim |E|^{\alpha}$, where $\alpha$ depends on the Luttinger parameter $K$. The exponents for tunneling into the charge sector differ from those for tunneling into the spin sector, and this asymmetry is only possible if the sectors are truly independent. -Remove the electron. Keep the math. The mathematics of the Luttinger liquid does not care that you have lost a particle. It computes correlations, conductance, and response functions without ever invoking a Fermi surface, a quasiparticle weight, or a discontinuity in momentum occupation. It works because it is built on the true degrees of freedom of the system — the collective modes that interactions have made robust. +The deeper consequence is philosophical. We call an electron an electron because it carries both charge and spin, inseparably, in every experiment we have ever done outside one dimension. Spin-charge separation shows that these are not inseparable properties of a fundamental particle but emergent degrees of freedom that can be disentangled when the geometry demands it. -The charge wave marches forward. The spin wave marches beside it at a different pace. They were born at the same point in space and time, in the act of creating an electron. They will separate by distance $(v_\rho - v_\sigma)t$ at any later time $t$. They may recombine if confined. They may decohere if scattered. But the separation is a permanent structural feature of one-dimensional quantum matter, as fundamental as any conservation law. +The electron is a quasiparticle of three dimensions. In one dimension, it ceases to be a quasiparticle and becomes a composite of more fundamental excitations. The fundamental excitations carry only one quantum number each. -The electron lies. The spin and charge tell the truth. +This is what one dimension teaches.