Spin-Current ACF: Transport & Noise Analysis

• Spin-Current ACF is a two-time correlation of spin-current operators used to diagnose transport regimes and coherent many-body dynamics in quantum systems.
• Its formulations vary across contexts—quantum spin chains, tunnel junctions, and quantum dots—providing insights into conductivity, diffusion, and noise spectra.
• Recent digital quantum simulations enable direct measurement of the ACF, revealing transitions among ballistic, superdiffusive, and diffusive transport behaviors.

Spin-current autocorrelation function (ACF) denotes a two-time current-current correlation built from a spin-current observable. In practice, the observable may be a bona fide spin-current operator derived from a continuity equation, a difference of spin-resolved charge currents, a symmetrized or non-symmetrized current-noise correlator, a zero-frequency second cumulant from full counting statistics, or, in mesoscopic fluctuation theory, a covariance of spin-current fluctuations under parameter variation. Its central role is transport diagnosis: current-current correlations are tied to conductivity, Drude weight, diffusion, finite-frequency noise, and coherent many-body dynamics. The term is therefore context-dependent, and the distinction between a genuine spin-current ACF and a spin-density or spin autocorrelation function is fundamental (Lee et al., 30 Jul 2025).

1. Definitions and representational forms

No single formula covers all uses of the spin-current ACF. The most common constructions in the cited literature are summarized below.

Context	Spin-current observable	Representative correlation
Quantum spin chain	$J_r = S_r^X S_{r+1}^Y - S_r^Y S_{r+1}^X$ or $j^\mu = i\sum_r (\vec S_r\times \vec S_{r+1})^\mu$	$C^{\mu\nu}(t)$, $C^{\mu\nu}(\omega)$
Tunnel-junction pure spin current	$\hat I_S=\hat I_\uparrow-\hat I_\downarrow$	$S(\nu)$, then $C(t)$ by Wiener–Khinchin
Spin-resolved counting statistics	$J_\ell^{\rm sp}=J_\ell^+-J_\ell^-$	$J_\ell^\nu J_{\ell'}^{\nu'}$, $S_{\ell\ell'}^{\rm sp}$
Mesoscopic spin Hall fluctuations	$j^\mu = i\sum_r (\vec S_r\times \vec S_{r+1})^\mu$0 in a four-terminal geometry	$j^\mu = i\sum_r (\vec S_r\times \vec S_{r+1})^\mu$1

In the spin-$j^\mu = i\sum_r (\vec S_r\times \vec S_{r+1})^\mu$2 Heisenberg chain in a longitudinal field, the finite-temperature spectral ACF is written as

$j^\mu = i\sum_r (\vec S_r\times \vec S_{r+1})^\mu$3

with the symmetrized form

$j^\mu = i\sum_r (\vec S_r\times \vec S_{r+1})^\mu$4

For the XXZ chain, the local current follows from the continuity equation,

$j^\mu = i\sum_r (\vec S_r\times \vec S_{r+1})^\mu$5

and the spatially averaged ACF is

$j^\mu = i\sum_r (\vec S_r\times \vec S_{r+1})^\mu$6

In these formulations, the ACF is the direct transport observable rather than a proxy (Steinigeweg et al., 2010).

In a tunnel junction driven by spin accumulation, the current operator is spin resolved, the noise spectrum is defined from

$j^\mu = i\sum_r (\vec S_r\times \vec S_{r+1})^\mu$7

and the spin current is

$j^\mu = i\sum_r (\vec S_r\times \vec S_{r+1})^\mu$8

The time-domain ACF is then obtained from the finite-frequency noise after symmetrization and subtraction of the zero-temperature vacuum contribution (Iwakiri et al., 2017).

In spin-resolved full counting statistics, the ACF is represented at zero frequency through second cumulants,

$j^\mu = i\sum_r (\vec S_r\times \vec S_{r+1})^\mu$9

and the net spin-current noise is

$C^{\mu\nu}(t)$0

This is the zero-frequency, time-integrated version of the spin-current ACF (Hu et al., 2024).

A distinct but established usage occurs in ballistic mesoscopic spin Hall transport, where the “ACF” is a covariance under energy and magnetic-field variation,

$C^{\mu\nu}(t)$1

This object is not a time-domain correlator, but it plays the same statistical role for mesoscopic spin-current fluctuations (Ramos et al., 2012).

2. Lattice-spin transport: Drude structure, coherence, and anomalous regimes

In the isotropic spin-$C^{\mu\nu}(t)$2 Heisenberg chain subject to a longitudinal field $C^{\mu\nu}(t)$3,

$C^{\mu\nu}(t)$4

the transverse current ACF is qualitatively richer than a single precession mode. The current operator is

$C^{\mu\nu}(t)$5

with transverse component $C^{\mu\nu}(t)$6 the principal observable. The longitudinal and transverse spectra differ structurally: $C^{\mu\nu}(t)$7

$C^{\mu\nu}(t)$8

Accordingly, the transverse time-domain ACF contains a nondecaying coherent oscillation at the Larmor frequency $C^{\mu\nu}(t)$9, while the regular part develops an additional oscillatory component at $C^{\mu\nu}(\omega)$0 (Steinigeweg et al., 2010).

At high temperature, the transverse ACF is essentially a Larmor oscillation with a rapidly decaying regular contribution. At lower temperature, the regular part becomes strongly asymmetric in frequency and yields a damped higher-frequency mode,

$C^{\mu\nu}(\omega)$1

with $C^{\mu\nu}(\omega)$2 increasing roughly linearly with $C^{\mu\nu}(\omega)$3. In the low-$C^{\mu\nu}(\omega)$4, high-$C^{\mu\nu}(\omega)$5 regime, the asymptotic analytic analysis instead gives a power-law envelope,

$C^{\mu\nu}(\omega)$6

and the frequency shift approaches $C^{\mu\nu}(\omega)$7. The paper attributes the higher-frequency mode not to a one-magnon Larmor process but to a collective many-body transition from one-magnon states to a two-magnon antibound branch near $C^{\mu\nu}(\omega)$8 (Steinigeweg et al., 2010).

The XXZ chain organizes the spin-current ACF directly around transport regimes. With

$C^{\mu\nu}(\omega)$9

the time-integrated current correlator is

$\hat I_S=\hat I_\uparrow-\hat I_\downarrow$0

with

$\hat I_S=\hat I_\uparrow-\hat I_\downarrow$1

The paper states that ballistic transport arises for $\hat I_S=\hat I_\uparrow-\hat I_\downarrow$2, superdiffusive transport at $\hat I_S=\hat I_\uparrow-\hat I_\downarrow$3, and diffusive transport for $\hat I_S=\hat I_\uparrow-\hat I_\downarrow$4. In the near-ballistic regime the ACF spreads in a broad light cone and decays slowly, leaving a positive long-time tail; at the isotropic point the ACF remains positive over accessible times but its integrated scaling follows the KPZ exponent; in the diffusive regime the total current ACF drops below zero and oscillates around zero, leading to a vanishing Drude weight and saturation of $\hat I_S=\hat I_\uparrow-\hat I_\downarrow$5 (Lee et al., 30 Jul 2025).

This suggests that, in spin chains, the spin-current ACF is not merely a noise diagnostic. It also resolves the separation between ballistic Drude contributions, coherent precessional structure, regular many-body modes, and hydrodynamic scaling.

3. Finite-frequency spin-current noise and pure-spin transport

In a tunnel junction driven by spin accumulation $\hat I_S=\hat I_\uparrow-\hat I_\downarrow$6, the spin-current ACF is obtained from finite-frequency spin-current noise in the high-frequency quantum regime. The mean currents are

$\hat I_S=\hat I_\uparrow-\hat I_\downarrow$7

At $\hat I_S=\hat I_\uparrow-\hat I_\downarrow$8, the charge current vanishes while the spin current remains finite if $\hat I_S=\hat I_\uparrow-\hat I_\downarrow$9; this is the paper’s definition of a pure spin current. After symmetrizing the noise spectrum and subtracting the zero-temperature vacuum contribution, the time-domain ACF is

$S(\nu)$0

Its temperature-independent normalized form is

$S(\nu)$1

The oscillation frequency is $S(\nu)$2, or $S(\nu)$3 in ordinary frequency units, and the period is $S(\nu)$4. The paper interprets the sinusoidal ACF as a manifestation of the Pauli exclusion principle in pure spin transport, even in the absence of any net charge current (Iwakiri et al., 2017).

The same paper emphasizes that temperature controls the envelope but not the normalized oscillation frequency. When $S(\nu)$5, the ACF becomes purely thermal and decays monotonically; when $S(\nu)$6, coherent oscillation is superposed on the thermal decay. Experimental visibility requires $S(\nu)$7 and $S(\nu)$8, and the paper gives $S(\nu)$9 and $C(t)$0 as an example where the relevant bandwidth is beyond about $C(t)$1 (Iwakiri et al., 2017).

A different fluctuation-theoretic setting is the mesoscopic spin Hall effect in a ballistic chaotic four-terminal conductor with strong spin-orbit coupling. There the spin-current ACF is an exact covariance,

$C(t)$2

The average spin current vanishes, so the covariance equals $C(t)$3. This correlator is Lorentzian in energy variation, square-Lorentzian in magnetic-field variation, and invariant across the symplectic–unitary crossover because cooperon contributions drop out in the spin sector (Ramos et al., 2012).

These two constructions—finite-frequency time-domain ACF in pure-spin tunneling and parametric covariance in mesoscopic spin Hall transport—show that the spin-current ACF can diagnose either temporal coherence or universal fluctuation structure, depending on the ensemble and the measured variable.

4. Interacting quantum dots: equilibrium spin noise, spin mixing, and counting statistics

In the equilibrium Kondo regime, the spin-current ACF is formulated as symmetrized spin-resolved current noise,

$C(t)$4

The full equilibrium conductance and noise tensors are governed by two universal scaling functions,

$C(t)$5

$C(t)$6

related by fluctuation–dissipation. The central result is that same-spin and opposite-spin correlations behave very differently. At $C(t)$7 and $C(t)$8,

$C(t)$9

so cross-spin noise is strongly suppressed at low frequency. Near $J_\ell^{\rm sp}=J_\ell^+-J_\ell^-$0, the spin-dependent response develops a broad dynamical spin accumulation resonance. At $J_\ell^{\rm sp}=J_\ell^+-J_\ell^-$1, the low-frequency anomaly is controlled by the Korringa rate

$J_\ell^{\rm sp}=J_\ell^+-J_\ell^-$2

and the cross-spin part obeys

$J_\ell^{\rm sp}=J_\ell^+-J_\ell^-$3

again vanishing as $J_\ell^{\rm sp}=J_\ell^+-J_\ell^-$4 in equilibrium when spin is conserved (Moca et al., 2011).

In a two-terminal quantum dot with semiconductor electrodes, the spin-current ACF is embedded in the zero-frequency noise tensor

$J_\ell^{\rm sp}=J_\ell^+-J_\ell^-$5

For charge current in a two-terminal device, current conservation makes the auto-correlation sufficient. For spin current, this generally fails when a transverse magnetic field $J_\ell^{\rm sp}=J_\ell^+-J_\ell^-$6 induces spin mixing. The paper states that the average spin current is conserved but the spin shot noise is not, so $J_\ell^{\rm sp}=J_\ell^+-J_\ell^-$7 in general. As a consequence, both auto-shot noise and cross-shot noise are essential to characterize spin-current fluctuations when $J_\ell^{\rm sp}=J_\ell^+-J_\ell^-$8 is present. In the AB/QD/AB junction, the spin auto-correlation is positive for all possible surface states while the spin cross-shot noise can oscillate between positive and negative values; in the AS/QD/AS junction, the spin cross-shot noise is negative for all studied surface states (Sartipi et al., 2016).

In a noncollinear quantum-dot spin valve, the zero-frequency spin-current ACF is obtained from spin-resolved full counting statistics. The spin current is

$J_\ell^{\rm sp}=J_\ell^+-J_\ell^-$9

and the net spin-current noise is

$J_\ell^\nu J_{\ell'}^{\nu'}$0

The spin-resolved current itself depends on occupations and accumulated dot spin,

$J_\ell^\nu J_{\ell'}^{\nu'}$1

while the exchange magnetic field

$J_\ell^\nu J_{\ell'}^{\nu'}$2

drives spin precession. The paper finds that majority- and minority-spin currents can be strongly autocorrelated despite uncorrelated charge transfer, that dynamical spin blockade yields super-Poissonian majority-spin autocorrelation in the first transport plateau, and that exchange-field-induced precession suppresses spin bunching and produces a double-peak structure in the net spin-current noise. It also states that the spin autocorrelation is highly sensitive to magnetization alignment and can therefore probe the relative angle between ferromagnetic electrodes (Hu et al., 2024).

5. Direct measurement and digital quantum simulation

A major recent development is the direct measurement of the spin-current ACF on superconducting-qubit hardware. For the 40-site XXZ chain, the central observable is

$J_\ell^\nu J_{\ell'}^{\nu'}$3

with local current

$J_\ell^\nu J_{\ell'}^{\nu'}$4

The technical obstacle is that the current operator is a two-site $J_\ell^\nu J_{\ell'}^{\nu'}$5 operator, so unequal-time correlators are expensive in ancilla-based schemes. The paper contrasts the Hadamard test, whose circuit count scales as $J_\ell^\nu J_{\ell'}^{\nu'}$6, with a direct protocol based on nonunitary operations and mid-circuit measurements, whose circuit count scales as $J_\ell^\nu J_{\ell'}^{\nu'}$7 for the full ACF (Lee et al., 30 Jul 2025).

The direct-measurement identities are

$J_\ell^\nu J_{\ell'}^{\nu'}$8

and

$J_\ell^\nu J_{\ell'}^{\nu'}$9

For the current ACF, the real part is reconstructed from two mid-circuit measurement configurations, while the imaginary part uses four unitary circuit variants. The XXZ dynamics are implemented with second-order Suzuki–Trotter decomposition,

$S_{\ell\ell'}^{\rm sp}$0

on IBM’s 156-qubit Heron processor $S_{\ell\ell'}^{\rm sp}$1, with the 40-site chain realized directly on 40 physical qubits (Lee et al., 30 Jul 2025).

The transport diagnostics extracted from the current ACF are the Drude-weight-like quantity

$S_{\ell\ell'}^{\rm sp}$2

and the time-dependent diffusion coefficient

$S_{\ell\ell'}^{\rm sp}$3

For $S_{\ell\ell'}^{\rm sp}$4, the reported values are

$S_{\ell\ell'}^{\rm sp}$5

and

$S_{\ell\ell'}^{\rm sp}$6

respectively. The paper interprets these as near-ballistic, superdiffusive/KPZ, and diffusive behavior, with the diffusive Drude weight effectively vanishing and the KPZ exponent in quantitative agreement with $S_{\ell\ell'}^{\rm sp}$7 (Lee et al., 30 Jul 2025).

This direct-measurement framework materially changes the status of the spin-current ACF. It turns an object that was previously central mainly in theory into one that can be probed on present-day hardware at system sizes where full exact classical simulation is already challenging.

6. Distinctions, ambiguities, and recurrent misconceptions

A recurrent misconception is that any spin autocorrelation function is automatically a spin-current ACF. In the one-dimensional classical Heisenberg model,

$S_{\ell\ell'}^{\rm sp}$8

the evidence for normal spin diffusion comes from the decay of the spin-density autocorrelation and from finite-size scaling,

$S_{\ell\ell'}^{\rm sp}$9

not from a directly computed spin current, current continuity equation in real space, spin-current ACF, or Green–Kubo current formula. The paper explicitly states that it does not define or measure a spin current $j^\mu = i\sum_r (\vec S_r\times \vec S_{r+1})^\mu$00 and does not compute $j^\mu = i\sum_r (\vec S_r\times \vec S_{r+1})^\mu$01 (Bagchi, 2012).

The same distinction appears in mean-field glassy dynamics. In the Sherrington–Kirkpatrick model under asynchronous dynamics, the two-time observable is the spin overlap

$j^\mu = i\sum_r (\vec S_r\times \vec S_{r+1})^\mu$02

or $j^\mu = i\sum_r (\vec S_r\times \vec S_{r+1})^\mu$03 in temporal notation. This is a spin autocorrelation in a disordered Ising system, not a current-current correlator and not a transport ACF (Sakata, 2012).

A deeper controversy concerns the definition of spin current itself in spin-orbit-coupled systems. One formulation begins from

$j^\mu = i\sum_r (\vec S_r\times \vec S_{r+1})^\mu$04

with a relaxation torque $j^\mu = i\sum_r (\vec S_r\times \vec S_{r+1})^\mu$05. Because spin is not conserved, the paper argues that the definition of spin current is not unique and that theories of spin-charge conversion should instead be expressed through correlators of physical observables: spin density–electric current,

$j^\mu = i\sum_r (\vec S_r\times \vec S_{r+1})^\mu$06

electric current–spin density,

$j^\mu = i\sum_r (\vec S_r\times \vec S_{r+1})^\mu$07

and spin density–spin density,

$j^\mu = i\sum_r (\vec S_r\times \vec S_{r+1})^\mu$08

It explicitly states that what is often called spin-current transmission efficiency is essentially the nonuniform component of magnetic susceptibility. In this framework, a pure spin-current ACF is deliberately avoided rather than privileged (Tatara, 2018).

These distinctions delimit the scope of the term. The spin-current ACF is most sharply defined when a conserved or operationally specified spin-current observable exists. Where only spin density is monitored, or where spin current is nonunique because of spin-orbit torque terms, a density correlator or mixed current–spin correlator may be the more faithful object. This suggests that the correct “ACF” for spin transport is not universal across all spin systems, but contingent on conservation laws, measurement protocol, and the operational meaning assigned to spin current.