Phenomenological Non-Equilibrium Process

• Phenomenological Non-Equilibrium Process is a macroscopic approach that describes irreversible evolution using coarse-grained variables and generalized thermodynamic laws.
• It employs extended state spaces, embedding theorems, and dissipation inequalities to bridge deterministic dynamics with stochastic fluctuations in complex systems.
• The framework underpins modeling in engineering, condensed matter, chemistry, and quantum systems, offering actionable insights into dynamic criticality and process efficiencies.

A phenomenological non-equilibrium process refers to the macroscopic description of irreversible evolution in systems driven away from thermodynamic equilibrium, where only partial knowledge of the microscopic dynamics enters the formalism, and the state is characterized by a finite set of coarse-grained (often experimentally accessible) variables. Unlike microscopic or fully statistical mechanical approaches, phenomenological theories employ extended state spaces, generalized thermodynamic relations, and constitutive laws derived from symmetry, conservation, and minimal assumptions about irreversibility. This framework underpins most practical theoretical modeling of non-equilibrium dynamics in engineering, condensed matter, chemistry, and @@@@1@@@@.

1. Extended State Spaces, Schottky Systems, and Nonequilibrium Variables

A phenomenological account begins by specifying the system of interest (often a discrete "Schottky system"—a single homogeneous body exchanging heat, work, and matter with its environment via a controllable partition) (Muschik, 2020). The full non-equilibrium state space is parametrized as

$Z = (a, n, U, E)$

where $a$ (generalized work variables), $n$ (mole numbers), $U$ (internal energy), and a set of non-equilibrium variables $E$ (e.g., internal variables like strains, reaction coordinates, or dissipative order parameters) capture the departure from equilibrium. Any trajectory $Z(t)$ can be projected orthogonally onto an equilibrium subspace $Z_{\rm eq}=(a,n,U)$, leading to the definition of an "accompanying reversible process" which, while not physically realizable, satisfies vanishing entropy production and allows direct comparison between irreversible and reversible evolution.

In advanced quantum thermodynamics, a similar strategy holds: the reduced description involves modified von Neumann equations for the density operator, with dissipative "propagator" terms acting on populations, thus embedding irreversibility without explicit reference to the underlying microscopic bath details (Muschik, 2022).

2. Nonequilibrium Entropy, The Embedding Theorem, and State-Function Construction

The existence of a well-defined non-equilibrium entropy $S(Z)$, as a state function on the enlarged space $Z=(a,n,U,\Theta,\xi)$ (with $\Theta$ the "contact" temperature and $\xi$ a vector of internal variables) is central to phenomenological non-equilibrium thermodynamics (Muschik, 2020). The embedding theorem states that the rate of entropy change integrated along any irreversible trajectory matches the change calculated along its projected reversible counterpart:

$$\int_{t_1}^{t_2} \dot S(Z(t)) dt = S(PZ(t_2)) - S(PZ(t_1)),$$

which guarantees consistency between non-equilibrium and equilibrium thermodynamics over macroscopic transitions.

The fundamental thermodynamic relation generalizes to

$$\dot S = \frac{1}{\Theta} \dot U - \frac{A}{\Theta} \cdot \dot a - \frac{\mu}{\Theta} \cdot \dot n + \alpha \dot \Theta + B \cdot \dot \xi + \Sigma,$$

where $\Sigma \geq 0$ is the internal entropy production.

In the NEET approach to evolution thermodynamics of solids (Metlov, 2011), a similar extension is built, expressing the internal energy as a function of both standard and non-equilibrium entropy and a vector of defect densities, with the entropy differential containing contributions from both thermal and non-equilibrium "temperatures."

3. Dissipation Inequalities and Generalized Constitutive Laws

A central outcome is the global dissipation inequality, derived independently of explicit entropy references:

$$\int_{t_1}^{t_2} \left[ \left(\frac{1}{\Theta} - \frac{1}{T^*}\right) \dot Q + \left(\frac{\mu}{\Theta} - \frac{\mu^*}{T^*}\right) \cdot \dot n^e \right] dt \geq 0,$$

with $T^*, \mu^*$ denoting the temperature and chemical potentials of external reservoirs. This condition constrains admissible phenomenological trajectories and connects with the second law.

General frameworks such as GENERIC formalism (Öttinger et al., 2020) express time evolution as

$\frac{dx}{dt} = L(x)\nabla E(x) + M(x)\nabla S(x),$

where $L$ encodes reversible Hamiltonian dynamics and $M$ is a positive semi-definite friction operator, with entropy production ensured by the structure of $M$. Both continuous (diffusive) and discontinuous (jump-driven) Markovian noise can be incorporated, with macroscopic transport coefficients measured via Green–Kubo relations or Kramers escape rates, providing a link between stochastic microscopic models and phenomenological evolution equations.

4. Contact Temperature, Reduction to Equilibrium, and the Role of Internal Variables

A salient feature of phenomenological frameworks is the introduction of a non-equilibrium (contact) temperature $\Theta$ as an observable conjugate to entropy, distinct from the equilibrium thermodynamic temperature $T_0(U,n,a)$ (Muschik, 2020). Under the condition that internal entropy production $\Sigma$ does not depend on internal energy $U$,

$\frac{\partial \Sigma}{\partial U} = 0,$

the "contact temperature reduction condition" ensures that $\Theta$ coincides with $T_0$ even out of equilibrium.

This permits temperature to be used as a primitive variable in many irreversible scenarios and validates the consistency between phenomenological non-equilibrium formulations and standard equilibrium thermodynamics. Internal variables $\xi$ (e.g. viscoelastic strain, defect densities, or reaction coordinates) enter both the entropy state function and the evolution equations, serving as a minimal set of slow variables essential to capturing memory, hysteresis, and structural relaxation in complex systems (Huang et al., 2019, Metlov, 2011).

5. Phenomenological Process Dynamics: Cyclic Processes, Efficiency, and Topology

The phenomenological approach yields explicit bounds on the efficiency of generalized cyclic processes. For a system exchanging heat with reservoirs at fixed temperatures $T_H > T_L$, but with the local contact temperatures $\Theta_H(t), \Theta_L(t)$ varying dynamically, the average contact temperatures satisfy

$$\langle \Theta_H \rangle \leq T_H, \qquad \langle \Theta_L \rangle \geq T_L,$$

and the cycle efficiency becomes

$$\eta_{\rm gcp} = 1 - \frac{\langle \Theta_L \rangle}{\langle \Theta_H \rangle} < \eta_{\rm Carnot},$$

with strict inequality except for reversible cycles (Muschik, 2020). This generalization indicates how time-dependent deviations in local intensive fields from their reservoir values systematically reduce the thermodynamic efficiency of real engines.

In stochastic processes violating detailed balance, the nonequilibrium steady state is characterized by persistent probability currents and can exhibit phenomena such as non-convexity in the quasi-potential (nonequilibrium free energy), leading to the occurrence of Lagrangian phase transitions distinctly absent in equilibrium systems (Bertini et al., 2010). These transitions are manifest as caustics in large-deviation rate functions and represent a fundamental phenomenological difference between equilibrium and nonequilibrium fluctuation structures.

Non-equilibrium processes in driven quantum or classical stochastic systems can additionally be characterized by topological invariants (e.g., non-Bloch winding numbers) which predict boundary-localized zero modes, anomalous spectral response, and dynamical crossovers between relaxation mechanisms, as seen in non-Hermitian extensions of reaction–diffusion chains (Li et al., 2023).

6. Statistical Features and Applications in Complex and Critical Systems

Phenomenological, data-driven models for non-equilibrium relaxation in complex systems introduce scale-coupled coefficients, enabling unified description of anomalous diffusion, non-Gaussian statistics, and long-term memory that span both the non-equilibrium and equilibrium regimes (Huang et al., 2019). For instance, the iterative law for force fluctuations between scales produces closed-form anomalous-diffusion exponents and captures the empirical crossover from heavy-tailed to Gaussian statistics as systems equilibrate.

Scaling crossovers in critical non-equilibrium dynamics are described by a generalized Bernoulli equation with time-dependent coefficients, from which universal power-law crossover functions are derived. These functions match the multi-stage relaxation observed in turbulent liquid crystal experiments and reaction-diffusion–limited exciton recombination, highlighting how phenomenological ODEs with time-dependent coefficients can embody the essential physics of dynamic criticality and its crossovers (Li et al., 7 Mar 2025).

7. Outlook: Scope, Limitations, and Generalizability

Phenomenological non-equilibrium process theory offers a mathematically precise, experimentally grounded, and computationally tractable framework for predicting the evolution and efficiency of systems far from equilibrium. It bridges deterministic thermodynamic laws, stochastic fluctuation theory, and extended thermodynamics (with internal variables), remaining agnostic to most microscopic details while encoding correctly all relevant macroscopic constraints and measurable properties.

Its limitations include reliance on a finite set of macroscopic variables (and thus loss of information about microscopic or highly non-Markovian effects), dependence on the local approach to equilibrium (not always justified for strongly coupled or glassy systems), and the necessity to supplement the theory by constitutive relations and coefficients provided by experiment or microscopic simulation.

Nonetheless, this approach—supported by embedding theorems, refined entropy and temperature definitions, explicit dissipation inequalities, and broad extensions to stochastic, quantum, and topological contexts—remains foundational to contemporary non-equilibrium statistical mechanics and engineering thermodynamics (Muschik, 2020, Öttinger et al., 2020, Muschik, 2022, Li et al., 7 Mar 2025, Bertini et al., 2010).