Green–Kubo Transport Coefficients

• Green–Kubo formulas are defined as time-integrated correlations that compute transport coefficients like diffusivity, viscosity, and conductivity from equilibrium fluctuations.
• The scaling Green–Kubo formulation generalizes the standard approach to address anomalous, aging, and long-range correlated systems by incorporating nonstationary dynamics.
• Quantum and nonequilibrium generalizations employ symmetrized current correlations and specialized density matrices to extend the framework to systems in nonsteady conditions.

The Green–Kubo formulas provide a unifying framework for calculating transport coefficients in equilibrium and near-equilibrium statistical mechanics, relating quantities such as diffusivity, electrical conductivity, thermal conductivity, and viscosity to time-integrated equilibrium correlation functions of appropriate fluxes. They underpin both classical and quantum linear-response theory, but also admit generalizations for anomalous, nonstationary, and strongly nonequilibrium dynamics. Contemporary research has extended the Green–Kubo paradigm to address systems with aging, long-range correlations, quantum statistics, nonreciprocal coupling, and higher-order effects in transport.

1. Classical Green–Kubo Formulation: Transport from Time-Correlation Functions

The canonical Green–Kubo formula expresses a transport coefficient (e.g., diffusivity $D$, viscosity $\eta$, or conductivity $\sigma$) as the infinite-time integral of an equilibrium time-autocorrelation function of the associated microscopic flux. For normal diffusion, consider a particle with position $x(t)$ and velocity $v(t)=dx/dt$, with stationary velocity statistics. The mean-square displacement fulfills

$$\langle x^2(t) \rangle = 2 \int_0^t d\tau\, (t-\tau) \langle v(0) v(\tau) \rangle_s,$$

leading, for large $t$, to the standard Green–Kubo relation for the diffusion coefficient

$$D_1 = \int_0^\infty \langle v(0) v(\tau) \rangle_s d\tau,$$

where $C_v(\tau) = \langle v(0)v(\tau)\rangle$ is the equilibrium velocity autocorrelation function. This formalism generalizes to other transport coefficients, with the appropriate identification of microscopic fluxes (stress tensor for viscosity, current operator for conductivity, etc.) (Dechant et al., 2013).

2. Scaling Green–Kubo Formulation for Anomalous and Aging Transport

If the velocity autocorrelation becomes nonstationary and exhibits scale invariance, the usual Green–Kubo relation fails or diverges. For systems with anomalous diffusion $\langle x^2(t)\rangle \sim 2 D_\nu t^\nu$ ($\nu>1$), the two-time velocity correlation scales as

$$\langle v(t+\tau)v(t)\rangle \sim \mathcal{C} t^{\nu-2} \phi(\tau/t),$$

where $\phi(s)$ is a scaling function. The generalized (scaling) Green–Kubo relation then gives (Dechant et al., 2013)

$$D_\nu = \mathcal{C} \frac{1}{\nu} \int_0^\infty ds\, (1+s)^{-\nu} \phi(s),$$

valid for superdiffusive and aging systems with long-range and/or nonstationary correlations. The scaling function $\phi(s)$ must obey specific power-law bounds to ensure convergence.

Distinct diffusion coefficients may result depending on the initial preparation: $D_\nu$ for a nonstationary (fresh) initial state, $D_{\nu,s}$ for stationary preparations where the velocity autocorrelation achieves (possibly non-integrable) steady-state form. In particular, ballistic or superaging dynamics admit only the scaling formulation.

3. Quantum and Nonequilibrium Generalizations

In quantum systems, the Green–Kubo relations require symmetric (Jordan-product) current–current or stress–stress correlation functions. For instance, the quantum shear viscosity in a fluid with steady shear flow is given by (Matsuoka, 2012):

$$\eta = \frac{V}{k_B T} \int_0^\infty dt\, \frac{1}{2} \left\langle \{ \tilde P_F(t), \tilde P_F(0) \} \right\rangle_{\textrm{eq}},$$

where $V$ is the volume and $\tilde P_F(t)$ is the Heisenberg-picture shear-stress operator.

For quantum current systems around nonequilibrium steady states (NESS), the Green–Kubo formula is generalized to incorporate the McLennan–Zubarev NESS density matrix and time-ordered exponentials summing all orders of external field perturbations (Hayakawa, 2011):

$$\delta \langle A_H(t) \rangle = \int_0^t ds\, F_{\textrm{ex}}(s) \int_0^\beta d\lambda\, \langle T_{\rightarrow} \exp[-\int_0^s d\tau \Omega(\tau)] J_B(s; -i\hbar \lambda) A_H(t) \rangle_0,$$

with $\Omega(\tau)$ encoding instantaneous entropy production. In NESS, the fluctuation theorem emerges naturally as a consequence of normalization, linking entropy production to the same kernel $\Omega(\tau)$ appearing in the nonlinear Green–Kubo response.

4. Dependence on Initial Conditions and Aging Effects

In systems where aging is significant, the measured mean square displacement $\langle \Delta x^2(t) \rangle_{t_0}$ depends on the elapsed time since the last renewal $t_0$, and the scaling Green–Kubo coefficient $D_\nu^{t/t_0}$ interpolates between stationary and nonstationary behaviors:

$$D_\nu^{t/t_0} = \mathcal{C} \int_0^1 dz\, z^{\nu-1} \left[ 1 + \frac{1}{z t/t_0} \right]^{\nu-1} \int_0^{z t/t_0} ds\, (1+s)^{-\nu} \phi(s),$$

recovering $D_\nu$ for $t \gg t_0$ and $D_{\nu,s}$ for $t \ll t_0$ when a stationary velocity correlation exists.

This sensitivity clarifies why anomalous or nonstationary systems can exhibit nonunique transport coefficients, and why the particular time origin and preparation protocol impact observed behavior.

5. Applications to Aging and Anomalous Transport Models

The scaling Green–Kubo formalism resolves various anomalous transport behaviors across several models:

• Nonlinear friction–driven diffusion (cold atoms in optical lattices): Fokker–Planck equations with $F(v) \sim 1/v$ induce heavy-tailed stationary velocity distributions and superdiffusion ($\nu<3$), with explicit formulas for $D_\nu$ and $D_{\nu,s}$ in terms of model parameters and scaling functions (Dechant et al., 2013).
• Fractional Langevin dynamics with long-memory and external noise (active transport): For generalized friction $\gamma(t) \sim t^{\rho-1}$ and external noise exhibiting fractional statistics, the scaling Green–Kubo relation yields $D_\nu$ and $D_{\nu,s}$, subsuming both stationary and aging regimes.
• Lévy walks/blinking quantum dots: Power-law distributed waiting times in renewal processes lead to subballistic or ballistic transport, with the scaling function dictating the form of the correlation and the appropriate $D_\nu$ coefficient depending on the value of the blinking exponent $\mu$.

The unified framework allows for specific evaluation of anomalous coefficients in these disparate contexts, emphasizing its broad applicability.

6. Significance, Generalizations, and Limitations

The scaling and generalized Green–Kubo formulas provide a precise link between transport coefficients and the underlying temporal structure of equilibrium or nonequilibrium fluctuations. In anomalous systems, they replace the standard stationary autocorrelation integral by a scaling form, accommodating nonstationarity and resolving ambiguities in the presence of aging or long-range correlations.

Key points:

• For normal stationary diffusion, the standard Green–Kubo integral yields a unique transport coefficient.
• For scale-invariant, aging, or nonstationary systems, the scaling Green–Kubo relation extracts both the exponent and the appropriate coefficient from the structure of the velocity correlation and initial conditions.
• In quantum and strongly nonequilibrium settings, the Green–Kubo formula requires symmetrization, time-ordering, and (in NESS) the use of generalized density matrices, with fluctuation theorems embedded in the structure.
• Depending on system preparation, multiple effective transport coefficients may coexist, demanding careful experimental and theoretical specification.

These generalizations facilitate rigorous calculation of transport properties in a wide range of complex dynamical regimes and have direct implications for systems ranging from cold-atom diffusion to active matter, quantum dots, and more (Dechant et al., 2013).