Home Is there a need for an extended phase definition for systems far from equilibrium?
Article Open Access

Is there a need for an extended phase definition for systems far from equilibrium?

  • Fernando Bautista ORCID logo EMAIL logo , Juan Paulo García-Sandoval ORCID logo and Octavio Manero ORCID logo
Published/Copyright: February 13, 2025

Abstract

Phase diagrams out of equilibrium have been the subject of intense research. An essential concept in these diagrams is the phase definition. Currently, that definition is well established for systems at classic thermodynamic equilibrium conditions. However, such phase definition is inadequate for systems that are not at equilibrium in the classic thermodynamic sense, like fluid systems under flowing conditions. Complex fluids may exhibit instabilities like shear-banding flow and a non-equilibrium critical point where banding flow ends. At this point, the fluid undergoes a phase transition from heterogeneous to homogeneous states. An extended thermodynamic space of variables is considered to adequately address this situation, which includes non-conservative variables such as stress and shear rate. Hence, the current phase definition based on conservative variables does not apply to non-equilibrium phase diagrams. On this basis, a broader definition of phase is required. We propose that this broader definition considers the thermodynamic variables space of Extended Irreversible Thermodynamics and the mathematical conditions that ensure compatibility between equilibrium and non-equilibrium conditions for systems where phases are well described by thermodynamic potentials even out of equilibrium.

1 Introduction

Phase diagrams have been an effective tool for describing and categorising different phase transitions in equilibrium [1]. However, non-equilibrium phase transitions have garnered significant research interest in recent decades, leading to the publication of many phase diagrams of systems beyond equilibrium (PDBE) [2], [3], [4], [5], [6], [7]. These diagrams, where each axis corresponds to a conservative or a non-conservative variable, illustrate large, connected regions with unique physical properties called “phases”. It is worth noting that the terms “phase” and “phase diagrams” have been used for non-equilibrium cases, such as shear-induced phase transitions, which have been extensively studied in various fluid systems [8], [9], [10], [11], [12].

After a brief analysis of experimental data, it is possible to classify the information of phase diagrams into two categories based on the type of variables involved. In the first one, we focus on the influence of non-conservative variables on conservative ones. The non-conservative variables are, for instance, stress σ ̄ ̄ , deformation B ̄ ̄ , deformation rate D ̄ ̄ , mass flux J ̲ , electric flux J ̲ e , electric potential V , and structural parameter φ, among others, while the conservative ones are volume V, internal energy U, mass m, charge q, and the intensive variables, such as pressure p, temperature T, chemical potential μ, electric field E , and so forth.

Next, some experimental results are discussed, considering cases where shear effects are especially striking. Classical thermodynamic properties such as p, V, and T are affected in a system initially at equilibrium and subjected to shear flow (first category). Because the conservative variables depend on microstructure and shear flow induces microstructural deformations, shear flow leads to a downward shift of the critical temperature and modification of the coexistence curve with respect to the equilibrium one [13], [14]. Onuki [4] reviewed the influence of shear in near-critical fluids, such as polymer, gels, and surfactant systems. Convective deformations drastically alter the critical behaviour, spinodal decomposition, and nucleation of fluids at near-critical conditions. For instance, convective deformation induces a downward shift of the critical temperature. Such a nonlinear shear regime was examined using the Ginzburg–Landau model [4]. Glasner [15] derived equations for both the solute profile and free energy balance at the interface, demonstrating solute trapping at large growth velocities. A non-equilibrium phase diagram was constructed, including an extensive range of growth conditions. Zhou et al. [7] studied an evaporation process in complex salt-water systems under non-equilibrium states, substantially different from equilibrium solubility diagrams. Helgeson et al. [6] presented direct evidence of flux-concentration coupling in a shear-banding fluid using neutron scattering. The formation of shear bands in worm-like micelles near an equilibrium isotropic-nematic transition appears simultaneously with shear-induced demixing, establishing a difference between the behaviour at equilibrium and beyond it. In the second category, the analysis of a rheological diagram reveals modifications of the non-conservative variables due to changes in the conservative variables. For instance, Landazuri et al. [12] found evidence of the shift of a non-equilibrium critical point induced by shear flow due to changes in conservative variables. Four different micellar solutions were selected since the chosen thermodynamic variables (T, surfactant concentration CD, or salt/-surfactant ratio CS/CD) are easy to control near the non-equilibrium critical point (NECP). In addition, these thermodynamic variables are included in the BMP (Bautista-Manero-Puig) model through the shear-banding intensity parameter [14].

The experimental shear flow curve plotted in a normalised shear stress σ ̄ ̄ as a function of shear rate γ ̇ graph is remarkably similar to the equilibrium liquid–gas phase transition: both have a coexistence region, a spinodal region, two metastable regions, and a critical point. When T or CD increases, the shear-banding flow region (corresponding to a coexistence region) shrinks and disappears at a “critical” temperature or surfactant concentration. Hence, a “critical” line (or surface when adding an extra component such as a particular salt) is defined because polymer-like micellar solutions are at least two-component systems. In addition, due to the flow-concentration coupling, the regions of instability broaden to include regions of positive slope in the constitutive flow curve [8], [9], [12]. Consequently, the experimental critical point shifts downwards in T or CD. Rheo-optical data reveal large fluctuations and increasingly erratic oscillations of the flow birefringence as the NECP is approached. This behaviour is related to flattening the dissipated energy per unit time as the NECP is approached. So, minor variations in the control thermodynamic variable produce large fluctuations and oscillations near the critical point. Martin del Campo [16] presented a methodology to compute and classify shear-induced structural and phase transitions in surfactant/water mixtures. Non-linear rheological experiments, considering CD and T variations were analysed. The parameters of the BMP model, obtained from the fitting of the σ ̄ ̄ versus γ ̇ data, which are functions of CD and T, allow classifying structural and phase transition boundaries. Then, considering the theoretical and experimental data, an improved phase definition must include conservative and non-conservative variables.

Now, for systems far from equilibrium, one encounters two opposing postures when deciding if the concepts of phase, phase diagrams, and phase coexistence apply in the same way as they do for PBDE. On the one hand, the assumption is that non-equilibrium phase coexistence is well-defined for systems in non-equilibrium steady states. Chatterjee et al. [17] argue that an equilibrium-like thermodynamic structure can be extended to non-equilibrium steady states having short-ranged spatial correlations, provided that the systems interact weakly to exchange mass with rates satisfying a balance condition – reminiscent of a detailed balance condition in equilibrium. On the other hand, it is assumed that phase coexistence is not well defined for non-equilibrium statistical systems. Dickman [18] studied theoretically a stochastic particle system known as lattice-driven gas with attractive interactions using the KLS (Katz-Lebowitz-Spohn) and the TTLG (two-temperature lattice gas) models [19], [20], and predicted phase separation in conditions far from equilibrium. Both Dickman and Katz et al. found that violations of universality are observed in the coexistence of the phases, depending upon how the phases coexist. These results suggest that the notion of phase as a state of matter with global properties that depend only on a small set of intensive parameters does not apply far from equilibrium. Furthermore, intensive properties (whose equality would predict the coexistence of phases) also cannot be consistently assigned to coexisting phases.

We formulate the following question: using these concepts in non-equilibrium states is still correct? We believe that it is an ill conception and requires an improved phase definition for systems beyond equilibrium, even for systems near equilibrium.

2 Extended thermodynamic variables

There is no universal choice of state variables when describing a non-equilibrium system. As a result, time evolution equations cannot be based on conservation laws alone. The variables characterising non-equilibrium states tend to decrease with time and vanish when equilibrium is reached. Furthermore, no single set of state variables can be used to describe the internal structure of a system that is out of equilibrium. Hence, the choice of variables depends on the nature of the system. There are several ways to attack this problem: the first approach assumes the validity of the local equilibrium hypothesis – the cornerstone of Classical Irreversible Thermodynamics-which is only strictly valid near equilibrium. The second approach assumes that the set of thermodynamic variables includes, in addition to the conservative variables, “flows” and “driving thermodynamics forces” even for conditions far from equilibrium. This assumption is the basis of the Extended Irreversible Thermodynamics (EIT) formalism, which presumes the existence of a regular and continuous function η E that plays the role of a generalised entropy and depending on a specific set of nonconserved variables, in addition to the ordinary conserved densities [21]. The third approach uses a formalism that does not commit to the choice of state variables, which implies that the language and ideas must be somewhat abstract. This approach is presented by the General Equation for Non-Equilibrium reversible–irreversible coupling (GENERIC) formalism. However, once the study system has been chosen, a set of variables must be used, which includes conservative and non-conservative variables. The GENERIC formalism includes constrictions and properties of equations governing the time evolution of macroscopic systems on mesoscopic levels of description that involve the following building blocks: state variables x, Hamiltonian kinematics L, dissipation potential, energy e x , and entropy (eta-function, η E ). Note that in the GENERIC formalism, the existence of well-defined states is assumed by this equation and includes an extended thermodynamic space of state variables x [22]. Rational Thermodynamics (RT) is similar to EIT but selects several additional variables. In addition to the variables’ instantaneous value, the non-equilibrium state is characterised by its deformation history. Emphasis is placed on axiomatic aspects, with theorems, axioms, and lemmas dominating the explanation. The main objective of RT is to impose restrictions on the form of the constitutive equations by applying formal statements of thermodynamics [23].

In summary, the three more popular theoretical approaches published, that describe non-equilibrium systems (dynamical systems) consider an extended thermodynamic space of state variables, including conservative and non-conservative variables.

3 Evolution of the phase concept

3.1 First published definition

Some representative phase definitions for systems under equilibrium conditions found in the literature are discussed below. In his seminal work, Gibbs defines the term “phase” in relation to the thermodynamic equation of state, which reads [24].

Definition 1.

“In considering the different homogeneous bodies which can be formed out of any set of components substances, it will be convenient to have a term which shall refer solely to composition and thermodynamic state of any such body without regard to its quantity or form. We may call such bodies that differ in composition or state different phases of the matter considered, regarding all bodies that differ only in quantity and form as different examples of the same phase. Phases which can exist together, the dividing surfaces being planes, in an equilibrium which does not depend upon passive resistances to change, we shall coexistent”.

This phase definition establishes a relationship between the concept of phase and thermodynamic state, and justifies the coexistence of phases at equilibrium. Therefore, the phase concept is only well defined at equilibrium, excluding any system outside equilibrium.

3.2 Some examples of highlighted definitions

Definition 2.

“Mechanically separable parts of a system are called phases if one of them, or some constituents thereof, may become part of another in a reversible way” [25].

In this case, the phase definition is centred on the separability of system parts. Let us consider a piece of clear plastic between two polarizers. When light passes through the material and the polarisers are homogeneous, it is observed that the system has no mechanically separable parts. Hence, a unique phase constitutes the system. Conversely, suppose a force (non-conservative variable) is applied to bend the plastic, and the bent piece is observed under polarised light. In that case, there are appreciable differences in the light pattern observed compared to the unaltered plastic piece. However, there is no possibility of separating any portions of the plastic mechanically by isolating sections with homogeneous light patterns. Neither is reversible for the bent plastic piece to “become part of another” phase. Hence, according to this phase definition, the bent plastic piece must be a single phase, which contradicts the previous reasoning. Since there is more than one phase in the bent plastic piece and such phases cannot be separated, this phase definition fails to describe this system.

Definition 3.

“A phase is a region of space, throughout of a material is essentially uniform” [26].

This definition emphasises the importance of uniformity in determining the phases of a material. For instance, an unbent plastic sample mentioned in Definition 2, which exhibits uniformity in the polarised light pattern, would be considered a single phase. However, when the same plastic is bent, its optical properties become non-uniform due to the applied stress. Therefore, Definition 3 cannot describe this system under stress.

Definition 4.

“Phase is defined as a system or part of a homogeneous system with definite boundaries. A phase may be a chemically pure substance or may contain more than one component” [1].

The key elements of this definition are homogeneity and defined boundaries. However, when we consider the analysis of the bent plastic piece presented in Definition 2, we see that homogeneity and defined boundaries are absent. Furthermore, the internal stress in the bent plastic piece relaxes slowly, leading to non-uniform pattern changes over time. Therefore, Definition 4 is inadequate to describe this system under stress. Similarly, when we look at the deformed plastic piece in the context of Definition 2, it is essential to note that a thermodynamic state is only well-defined at equilibrium. As a result, this definition fails to describe the piece mentioned above out of equilibrium.

Definition 5.

“Bulk phase is a set of spatially homogeneous states of matter described by a given fundamental equation with the following properties:

  1. U , V , N 1 , , N k are extensive (additive) parameters conserved in any isolated system (“additive invariants”).

  2. S U , V , N 1 , , N k is a homogeneous function of first degree with respect to all extensive parameters.

  3. S U , V , N 1 , , N k is a smooth (infinitely differentiable) function.” [27]

This definition is closely related to an equation of state, like Gibbs’ definition. However, it is more detailed and explicitly describes the mathematical properties of this equation. It is important to note that this definition cannot describe out-of-equilibrium systems.

Definition 6.

“A thermodynamic phase of a material is an open, connected region in the space of thermodynamic states parametrized by the variables T, μ, and p. The pressure is analytic in T and μ. Hence, p has a convergent power series expansion in a neighborhood about T 0 ; μ 0 that gives its values.” [28]

This phase definition accurately includes the analyticity of functions, resulting in a better and more complete definition of “phase” than those contained in the first five. However, other desirable mathematical requirements to reinforce the definition are absent, such as the convexity upward of the functions, ensuring thermodynamic stability. However, Definition 5 still considers the requirement of analyticity of functions. Since all these functions are defined at equilibrium, the phase coexistence or phase transitions of systems out of equilibrium continue to be ill-defined. Last phase definitions cannot be correctly applied to fluid systems subjected to shear flows out of equilibrium because of the transitions that non-conservative variables induce in such systems. With the definitions provided above, it is impossible to explain phase changes induced by a non-conservative variable, such as, for example, banded flow in complex systems subjected to shear [13], [15], [16], photo induced phase transitions in liquid crystals [29], [30], synthesis under external fields [31], among other cases with different driven external potentials.

At the present century many authors have reported phase transition beyond equilibrium, but they imply a phase definition similar to published recently by Dickman [18]:

Definition 7.

“A non equilibrium phase corresponds to a macroscopic state of a system under a drive, having time independent, reproducible properties that vary smoothly with the drive intensity and the external parameters (such as temperature and chemical potential) associated with the reservoir or reservoirs in contact with the system. If the macroscopic properties depend in a singular manner on the drive or other external parameters, the system is said to suffer a (non equilibrium) phase transition.” [18]

And for phase coexistence Dickman [18] wrote:

“Now consider two systems in NESSs, subject to the same drive and external parameters, but with distinct, spatially uniform macroscopic properties. The systems represent coexisting phases if, when allowed to exchange energy or matter, the net flux of the quantity or quantities they may exchange is zero. Non equilibrium phase coexistence emerges spontaneously at phase separation, in which, varying some external parameter, a homogeneous phase becomes unstable, yielding a new stable steady state containing distinct macroscopic phases separated by a sharp interface. (The coexisting phases are clearly free to exchange particles and energy in this situation.)”

This definition could be applied to non-equilibrium systems and, although recently published, is analogous to definition 2, since it emphasises the importance of uniformity and reproducibility to determine the phases of a material, while it differs from Definitions 1 to 6 because the system must be under the influence of a variable that deviates from equilibrium. However, Definition 7 does not comment at all on the mathematical properties that it must fulfill.

4 Requirements to develop an extended definition of a thermodynamic phase

When describing non-equilibrium systems, most authors use an extended thermodynamic space of state variables which includes both conservative and non-conservative variables. Additionally, “phase” and “phase diagrams” have been used for non-equilibrium cases. However, the current phase definitions are unsuitable for dealing with non-equilibrium systems and PDBE. Therefore, developing an improved phase definition for systems beyond equilibrium, even those near equilibrium, is necessary.

The new definition must exhibit no theoretical inconsistencies concerning the fundamental laws of irreversible thermodynamics – specifically, the entropy inequality σ s > 0. Furthermore, the Gibbs relation of thermodynamic fluxes J α and thermodynamic forces X α α N : α ν expressed as σ s = α J α X α should lead to an explicit equation for entropy evolution production beyond equilibrium, whether the system is near or far from it. According to EIT, X α must be constructed as the gradient of nonconserved potentials. The equation obtained must be an antisymmetric, and multilinear manifold to impart the equation with the characteristic of being univocal. Once the equation to describe a phase exhibits the mathematical properties of equations of state (at equilibrium) and the evolution or dynamical equations (beyond equilibrium), the applicability of such an equation must be tested. Some experimental systems must agree with the extended definition and should be susceptible to being classified using this definition.

5 Proposal of the extended definition of phase

This analysis aims to provide a definition that correctly describes many systems out of equilibrium that depend on conservative and non-conservative variables. Hence, an extended definition of “phase” is proposed. This extended definition must be reduced to Definition 6 and considers the mathematical properties of Definition 5 when the system is at equilibrium.

The analysis of the results cited before concludes that a non-equilibrium phase diagram may be built if an additional orthogonal direction representing the non-conservative variable is added to the equilibrium phase diagram. The path the system follows in this non-equilibrium phase diagram represents the relaxation of the non-conservative variable from one equilibrium point to another. Then, an extended phase definition is necessary and sufficient to build the phase diagram out of equilibrium without internal contradictions. However, a system out of equilibrium depends on conservative and non-conservative variables. So, the variable space must be extended accordingly. Non-conservative variables can be included following the Extended Irreversible Thermodynamics approach. It is important to remark that in the present work we are considering systems where phases are well described by thermodynamic potentials even out of equilibrium, i.e. the variables representing the forcings (e.g. shear or voltage) are included in the potentials. Therefore, extended free energy should be used since, the non-conservative variables must be included. This assumption may impose a limitation for some particular cases of non-equilibrium systems, whose phases cannot always be described through thermodynamic potentials [32]. For instance, in active matter theory, one of the most used field theories is the so-called Active Model B which has a non-gradient term, i.e. a term that cannot be expressed through any free energy [33], while in Markov stochastic processes (e.g. the Brownian motion of an intruder in a complex fluid) the absence of equilibrium is strictly related to the violation of detailed balance, which is mathematically equivalent to the impossibility to express the stationary state in the Boltzmann-form exp β H , as H does not exist and the dynamics is non-gradient [34].

5.1 Extended phase definition

Definition.

A thermodynamic phase of a material is an open, connected region in the space of extended irreversible thermodynamic states parametrised by the intensive variables T , μ i i N : i k , p , and nonconserved variables, X β β N : β ν . The extended thermodynamics potentials are convex upward analytic in T , p , μ i , and X β . Hence, the potentials can be expressed as a convergent power series expansion in a neighbourhood about T , p , μ i ; X β providing determined values.

     A phase is a set of spatially homogeneous states of matter described by a fundamental equation valid even far from equilibrium, possessing the following properties:

  1. U , V , N 1 , , N k are extensive (additive) conserved parameters. In addition, for nonconserved parameters X 1 , , X ν thermodynamic forces are used, which may be extensive or intensive.

  2. σ U , V , N 1 , , N k ; X 1 , , X ν is a homogeneous function of first degree for all parameters.

  3. σ U , V , N 1 , , N k ; X 1 , , X ν is a smooth (infinitely differentiable) function.

5.2 Entropy production rate

The total differential of the function σ given in the proposed extended phase definition seems as

(1) σ = σ e q + σ s = i Q i P i + α J α X α

where P i represents an extensive conservative variable, Q i a conjugate conservative variable, J α a flux, and X α a driving thermodynamic force.

If the flux is a continuous function of the driving thermodynamic forces, the relationship between non-conservative variables and non-equilibrium fluxes can be expanded as a series using the Taylor’s theorem around the equilibrium (X β  = 0) to obtain

(2) J α X β = J α 0 + β J α X β 0 X β + β h α β X 1 , , X ν X β = β L α β X β + β h α β X 1 , , X ν X β

where L α β J α / X β 0 and h α β X 1 , , X ν = 1 2 δ 0 1 1 t 2 J α X β X δ t X 1 , , t X ν X δ d t with lim X 1 , , X ν 0 , , 0 h α β X 1 , , X ν = 0 . Notice that L αβ and h αβ may be also functions of the conjugate variables. The entropy source under non-equilibrium conditions reads.

σ s = α β L α β + h α β X 1 , , X ν X β X α

here, the coefficients h α , β X 1 , , X ν may not be positive defined, but always the entropy inequality σ s  > 0, therefore the matrix L ̂ , whose elements are L ̂ α , β = L α β + h α β X 1 , , X ν , must be positive definite.

The latest phase definition should evolve towards a more comprehensive and accurate definition of the phase concept that encompasses systems that are not in a state of equilibrium.

Consider a faraway equilibrium process in a volume V at rest subjected to time-independent constraints at its surface. The total entropy S produced inside a system during a procedure taking place in volume V subjected to time-independent constraints can be written as

(3) σ s = α J α X α d V = α β L α β + h α β X 1 , , X ν X β X α d V

Considering that the thermodynamic forces X α take the form of gradients of intensive variables X α  = ∇Γ α , the previous equation becomes

(4) σ s = α β L α β Γ α Γ β d V + α β h α β Γ 1 , , Γ ν Γ α Γ β d V .

Taking the time derivative, using the Leibnitz rule (differentiation under the integral sign), and after rearranging, the equation for the rate of production of entropy follows

d σ s d t = α β L α β + L β α Γ β Γ ̇ α d V + α β h α β + h β α + δ h β δ Γ α Γ δ Γ β Γ ̇ α d V × α β δ L α β Γ δ + h α β Γ δ Γ β Γ α Γ ̇ δ d V

Therefore, the stability properties depend on the behaviour of this equation that further depends on the dynamical behaviour of the intensive variables.

5.2.1 Phase rule for non-equilibrium systems

Similar to the process used in equilibrium thermodynamics for a heterogeneous system, the key is the equality of potentials for the coexistence of phases. In the case of non-equilibrium systems, an additional process is the change of structure, and therefore an additional structural potential to the chemical potential in equilibrium must be considered. Two postulates to describe systems composed of multiple phases are introduced:

Postulate 1.

Each homogeneous substance can potentially exist in multiple phases, each one with its fundamental equation.

Postulate 2.

Any heterogeneous substance is composed of homogeneous subsystems representing phases.

Note that, the subsystems can be either different phases or states of the same phase. Consider an open heterogeneous systems composed of π phases described by the fundamental equations,

(5) S n U n , V n , N n , 1 , , N n , k ; X n , 1 , , X n , ν

here, n represents the total number of existing phases. Neglecting the interface contributions between the phases, the additivity of entropy dictates that the total entropy of the system is

(6) S tot = n S n U n , V n , N n , 1 , , N n , k ; X n , 1 , , X n , ν

According to the entropy maximum principle, the necessary and sufficient condition of equilibrium of the system is the maximum of S tot under the following steady state constraints

(7) n U n = c o n s t

(8) n V n = c o n s t

(9) n N n , i = c o n s t , i = 1 , , k

(10) n f n X n , α = c o n s t , α = 1 , , ν

expressing the conservation of energy, volume, and the amount of each chemical component, respectively, while f n Γ n , α is correlated with the constitutive equations for non-conservative variables. This variational problem is solved by using a set of Lagrange multipliers λ u , λ v , λ i , and λ γ,α

(11) max n = 1 ϕ S n λ u n U n λ v n V n i λ i n N n , i α λ γ , α n f n X n , α

The solution is the equality of temperatures, pressures, chemical potentials and structural potentials in all phases,

(12) T 1 = = T π T ,

(13) p 1 = = p π p ,

(14) μ 1 i = = μ π , i μ i , i = 1 , , k

(15) f 1 X 1 , α = = f π X π , α f X α , α = 1 , , ν

Thus, the entire heterogeneous system is described by k + ν + 2 intensive variables (T, p, μ 1, …, μ k ; X 1, …, X ν ). However, these variables are not independent. Each phase must satisfy its equation analog of the Gibbs-Duhem equation, which imposes π constraints

(16) S n d T + V n d p n N n , i d μ n , i α J n , α d X n , α = 0

As a result, the number of independent variations of the heterogeneous system (also called degrees of freedom) are

(17) f = k + 2 + ν π

This is a relation analog to the Gibbs phase rule. The global properties (T, p, μ 1, …, μ k ; X 1, …, X ν ) are referred to as intensities or fields and characterise the phase. They are distinguished from densities (ratios of extensive properties), such as energy, entropy, or the amounts of chemical components per unit volume or particle. Both intensities and densities are local variables that characterise physical points. However, while the intensities are uniform across the heterogeneous system, the densities generally differ in different phases and experience discontinuities across phase boundaries. Notice that the non-equilibrium degrees of freedom increases with ν in comparison with the traditional equilibrium degrees of freedom due to the presence of additional driving forces that can be imposed to the system.

6 Study cases

6.1 Flow of complex fluids

A relatively simple theoretical model of complex fluids that qualitatively captures the key experimental observations has been proposed by Bautista et al. [14], [35]. It was derived using the extended irreversible thermodynamics formalism, that used extended space of parameters, here the stress tensor is the non-conserved variable. The Helmholtz free energy is a homogeneous function of first degree for all parameters, and follow all mathematical properties of the definition given in Section 5.1. This model predicts rather well steady states and transient nonlinear features of complex fluids that exhibit shear-banding flow and the shear-thinning shear-thickening transitions. The authors proposed a stress constitutive equation coupled with an evolution equation of fluidity and a constitutive equation for the mass flux. The fluidity is interpreted as a macroscopic measure of the internal structure of the fluid. The extended thermodynamic potential in this case is the extended Helmholtz free energy, given by

(18) d A = T d s v d P + v φ 2 G 0 φ 0 σ ̲ ̲ : d σ ̲ ̲

with the working variables: entropy (s), temperature (T), molar volume (v), hydrostatic pressure (P), elastic modulus at high frequencies (G 0), the fluidity (φ), and the stress tensor ( σ ̲ ̲ ) , which is composed of shear stresses (σ 12, σ 13, and σ 23), and normal stresses (σ ii ). This potential is convex upward analytic in the conserved variables T, p, and in the non conserved ones, such as σ ̲ ̲ .

From dA, the evolution equations of the model are obtained [14]

(19) d φ d t = 1 λ φ 0 φ 1 + ε I I D 2 I I σ + k 0 1 + ϑ I I D + μ I I σ φ φ σ ̲ ̲ : D ̲ ̲ ,

(20) σ ̲ ̲ + 1 G 0 φ σ ̲ ̲ = 1 φ D ̲ ̲ .

Here, in the context of the BMP model, φ is the fluidity, while φ 0 and φ are the fluidities at zero and high shear rates, respectively. λ is the relaxation time of the structure, k 0 is a kinetic constant, ɛ is the transition shear thinning–thickening parameter, ϑ and μ are the shear-banding intensity parameters in the direction of gradient and in the vorticity direction. σ ̲ ̲ = d σ ̲ ̲ d t L ̲ ̲ σ ̲ ̲ + σ ̲ ̲ L ̲ ̲ T is the convective derivative, L ̲ ̲ is the velocity gradient tensor, and D ̲ ̲ is the symmetric part of the rate of strain tensor.

Figure 1 depicts shear flow curves generated with the BMP model for a 20 wt% Pluronics P103/water system [16]. The curves plotted in a γ ̇ σ T graph resemble a PVT 3D diagram for equilibrium phase transitions. Near the non-equilibrium critical point, the orthogonal projection of the γ ̇ -T plane may be divided into two regions by a continuous curve, which we shall call the “phase boundary” or coexistence curve. For temperatures lower than the critical temperature T c, there is an heterogeneous flow or two-phase region where bands at high ( γ ̇ C 1 ) and low shear rates ( γ ̇ C 2 ) coexist at steady-state and γ ̇ , which denotes the average shear rate of the fluid. Above T c lies the homogeneous region. The phase boundary T γ ̇ is a convex upward analytic function of γ ̇ around the critical point except, perhaps, at γ ̇ = γ ̇ c , the unique point where T γ ̇ achieves its maximum. In the heterogeneous region, the free energy is the sum of free energies of the two bands, plus corrections due to surfaces separating the two phases and at the container walls. When the container is large, or there are a few bands or both, the surface free energy is negligible. Some models account for this contribution necessarily but imply non-linear terms.

Figure 1: 
Scheme of a 3D rheological diagram for systems out of equilibrium (adapted from [16]).
Figure 1:

Scheme of a 3D rheological diagram for systems out of equilibrium (adapted from [16]).

Furthermore, in the orthogonal projection of the γ ̇ -σ plane, the steady-state of the two-phase region is obtained by a double tangent construction similar to Maxwell’s equal areas construction. For stress lower than the non-equilibrium critical point, the γ ̇ -σ plane may be divided into two regions by a continuous curve called the bands’ coexistence curve, which is a “phase boundary.” For shear stress lower than σ c , there is an heterogeneous flow where bands at γ ̇ C 1 and γ ̇ C 2 coexist at steady-state, and γ ̇ denotes the average shear rate of the fluid. Above this σ lies the homogeneous region. Also, in the heterogeneous region, the free energy is the sum of free energies of the two bands, plus corrections due to surfaces separating the two phases and at the container walls. The phase boundary σ γ ̇ is a convex upward analytic function of γ ̇ around the critical point except, perhaps, at γ ̇ = γ ̇ c , the unique point where σ γ ̇ achieves its maximum.

It is noteworthy that the equilibrium thermodynamic properties of a system are entirely determined by the Helmholtz free energy per unit volume A ρ , T as a function of the density ρ and the temperature T, or another thermodynamic potential. As stated before, the necessary and sufficient condition to ensure their validity near the critical point is that the thermodynamic functions are analytical even near the critical point, as stated before. Additionally, they must satisfy the conditions of stability (convexity) [36], [37]. Hence, for non-equilibrium systems to comply with the above requirements, an extended free energy should be used since the non-conservative variables must be included. Note that in the biphasic region (heterogeneous), A ρ , T will be the sum of the free energies of the liquid and vapour phases present, plus the energies due to the surfaces that separate the two phases and on the walls of the container. If the container is not too small, the free energies on the surface are negligible. In addition, by applying the phase rule for non-equilibrium systems at steady state, Eq. (17), for two coexistence phases (π = 2) it is obtained that the number of degrees of freedom is three (f = 3), which is interpreted as an area in the non-equilibrium phase diagram in agree with experimental data. Since the number of species k is two (pluronics and water) and the number of driving forces is in this case one (the shear stress ν = 1). The degrees of freedom are associated with the temperature, the shear stress and the fraction of pluronics in the solution, that is constant and equal to 20 %wt for the particular case depicted in Figure 1. Each of the two coexistence phases are differentiated by a low shear rate ( γ ̇ C 1 ) and a high shear rate ( γ ̇ C 2 ) , respectively, that produce the shear-banding flow. On this case, Eqs. (15) describe the plateau for the same shear stress with two different shear rates, i.e. σ 12 γ ̇ C 1 = σ 12 γ ̇ C 2 = σ plateau .

6.2 Electrical system

In this case, the extended space of parameters is A , P , N i ; I ext , here the electric current, I ext , is the non-conserved variable and its conjugated variable is the voltage, V. Also, the Helmholtz free energy

(21) d A = T d s v d P + I ext d V

is a homogeneous function of first degree for all parameters, and follow all mathematical properties of definition (Section 5.1)

The organic salt θ-(BEDT-TTF)2CsCo(SCN)4 is a material composed of conductive layers of BEDT-TTF (bis(ethylenedithio) – Tetrathiafulvalene) and insulating layers of CsCo(SCN)4 stacked alternatively along an axis. It presents a triangular BEDT-TTF network, where there is one hole for every two molecules, at low temperatures, due to Coulomb repulsion. Hence, charges in the system induce order. It exhibits a giant nonlinear resistance suggesting a coexistence of non-equilibrium phases, namely, a competition between different charge-induced order patterns [38]. Endo et al. [39] studied the electric field driven by a non-equilibrium phase transition and its critical phenomena, which explains this behaviour in the conductor θ-(BEDT-TTF)2CsCo(SCN)4 [38]. The conductivity of this conductor changes discontinuously when the electric field acting on the system varies continuously, that is, the sudden change in resistance evidences a phase transition driven by an electric current. Here, the non-conservative variable is the electric current. Endo et al. [39] found that the system undergoes the non-equilibrium phase transition due to the non-linear J-E characteristics. The authors produced a three-dimensional phase diagram for the non-equilibrium phase transition, examined critical phenomena, and calculated critical exponents of the experimental system and found them to agree with the Landau theory.

Figure 2 depicts experimental data of voltage against external current as a function of temperature for the organic salt θ-(BEDT-TTF)2CsCo(SCN)4. For this organic salt, the curves plotted in a graph, also resemble a PV-T 3D diagram for equilibrium phase transitions, exhibiting a non-equilibrium critical point. Near this point an orthogonal projection of the T-V plane may be divided into two regions by a continuous curve, which we shall call the “phase coexistence line.” For temperatures lower than the critical temperature T c ≈ 6.5 K, a two-phase region where a phase of high electric resistance becomes a phase of low electric resistance and vice verse. In the projection of the IV plane, the current is multivalued, i., e., three values of current correspond a voltage value. Above T c lies the homogeneous region. It is noteworthy that the equilibrium thermodynamic properties of a system are entirely determined by the Helmholtz free energy per unit volume as a function of the voltage, and current density and the temperature T, or another thermodynamic potential. By applying the phase rule for non-equilibrium systems, Eq. (17), here the number of species k is one (organic salt θ-(BEDT-TTF)2CsCo(SCN)4, the electric current is the unique driving force, so ν = 1; also there is a phase of low electric resistance coexisting with a high electric one, and the constant two refers to temperature and pressure. Hence, the number of degrees of freedom is two, interpreted as an area in the non-equilibrium phase diagram that ends at a critical point.

Figure 2: 
Experimental data of voltage against external current as a function of temperature for the organic salt θ-(BEDT-TTF)2CsCo(SCN)4 [38].
Figure 2:

Experimental data of voltage against external current as a function of temperature for the organic salt θ-(BEDT-TTF)2CsCo(SCN)4 [38].

7 Conclusions

A new, theoretically more satisfactory definition of a phase for systems where phases are well described by thermodynamic potentials even out of equilibrium, i.e. the variables representing the forcings (e.g. shear or voltage) are included in the potentials is presented, including cases out of equilibrium and where the equilibrium case is a limit. Many publications deal with phase diagrams, inside and outside equilibrium. A fundamental concept in these diagrams is the phase definition.

Currently, the phase definition is well-defined for systems under equilibrium conditions; however, such phase definition turns out to be inadequate when the system under consideration is beyond equilibrium conditions. In particular, the problem arises in the description of shear-banding flow and the non-equilibrium critical point where the banding flow ends. At this point, the fluid changes from heterogeneous to homogeneous. Here, an extended thermodynamic space of variables is considered. Nevertheless, the current phase definition is based on conservative variables. Consequently, it does not apply to phase diagrams that include non-conservative variables. Therefore, a broader definition of phase is required.

We propose that the broader definition considers the same thermodynamic space of EIT, which includes conservative and non-conservative variables. Notice the advantage of using the EIT approach. When the non-conservative variables of a system out of equilibrium vanish, the broader definition proposed here reduces to the phase definition currently applied for systems at equilibrium.

We have used two far-equilibrium systems to demonstrate the validity and robustness of our proposed extended phase definition. The first system was analysed using the BMP model, which accurately predicts the rheological behaviour of many experimental complex systems. The second system was examined using an electric model system that displays a wide range of behaviour far from equilibrium. Both systems align with the definition of extended phase because their fundamental equations consider conserved and non conserved variables, remain valid even far from equilibrium, and exhibit the properties mentioned in Section 5.1.


Corresponding author: Fernando Bautista, Departamento de Física, Universidad de Guadalajara, Blvd. M. García Barragán 1451, Guadalajara, Jalisco, 44430, México, E-mail: 

Acknowledgments

Fernando Bautista Rico reports financial support was provided byNational Council on Science and Technology. Fernando Bautista Ricoreports a relationship with University of Guadalajara that includes:board membership. If there are other authors, they declare that theyhave no known competing financial interests or personal relationshipsthat could have appeared to influence the work reported in this paper.

  1. Research ethics: The local Institutional Review Board deemed the study exempt from review.

  2. Author contributions: Conceptualization, F.B.; methodology, F.B.; software, J.P.G-S.;validation, F.B.. and J.P.G-S..; formal analysis, F.B. and J.P.G-S. andinvestigation, F.B.; data curation, J.P.G-S. and F.B; writing—originaldraft preparation, F.B.; writing—review and editing, O.M.; visualization,J.P.G-S.; funding acquisition, O.M. All authors have read and agreedto the published version of the manuscript.

  3. Informed consent: Not applicable.

  4. Conflict of interest: The authors state no conflict of interest.

  5. Use of Large Language Models, AI and Machine Learning Tools: The authors state that no Large Language Models, AI, or Machine Learning Tools were used in the design and methodology of this research study." above conflict of interest.

  6. Research funding: This research was funded by Mexican National Council of Science andTechnology (CONACyT) grant number F.B. (SNI-16004), J.P.G-S.(SNI-43658) and O.M. Project IN100623 from PAPIIT-UNAM.

  7. Data availability: The raw data can be obtained on request from the corresponding author.

References

[1] C. J. Adkins, Equilibrium Thermodynamics, Cambridge, Cambridge University Press, 1983.10.1017/CBO9781139167703Search in Google Scholar

[2] H. Geerissen, J. R. Schmidt, and B. A. Wolf, “On the factors governing the pressure dependence of the viscosity of moderately concentrated polymer solutions,” J. Appl. Polym. Sci., vol. 27, no. 4, pp. 1277–1291, 1982, https://doi.org/10.1002/app.1982.070270417.Search in Google Scholar

[3] A. Onuki, “Homogeneous block copolymer systems under shear flow,” J. Chem. Phys., vol. 87, no. 6, pp. 3692–3697, 1987, https://doi.org/10.1063/1.452967.Search in Google Scholar

[4] A. Onuki, “Phase transitions of fluids in shear flow,” J. Phys.: Condens. Matter, vol. 9, no. 29, pp. 6119–6157, 1997. https://doi.org/10.1088/0953-8984/9/29/001.Search in Google Scholar

[5] E. Cappelaere, R. Cressely, and J. P. Decruppe, “Linear and non-linear rheological behaviour of salt-free aqueous CTAB solutions,” Colloids Surf., A: Physicochem. Eng. Aspects, vol. 104, nos. 2–3, pp. 353–374, 1995. https://doi.org/10.1016/0927-7757(95)03332-2.Search in Google Scholar

[6] M. E. Helgeson, M. D. Reichert, Y. T. Hu, and N. J. Wagner, “Relating shear banding, structure, and phase behavior in wormlike micellar solutions,” Soft Matter, vol. 5, no. 20, pp. 3858–3869, 2009. https://doi.org/10.1039/b900948e.Search in Google Scholar

[7] H. Zhou, et al.., “Non-equilibrium state salt-forming phase diagram: utilization of bittern resource in high efficiency,” Chin. J. Chem. Eng., vol. 18, no. 4, pp. 635–641, 2010, https://doi.org/10.1016/s1004-9541(10)60268-6.Search in Google Scholar

[8] H. Rehage and H. Hoffmann, “Shear induced phase transitions in highly dilute aqueous detergent solutions,” Rheol. Acta, vol. 21, no. 4, pp. 561–563, 1982, https://doi.org/10.1007/bf01534347.Search in Google Scholar

[9] J.-F. Berret, G. Porte, and J.-P. Decruppe, “Inhomogeneous shear flows of wormlike micelles: a master dynamic phase diagram,” Phys. Rev. E, vol. 55, no. 2, pp. 1668–1676, 1997. https://doi.org/10.1103/physreve.55.1668.Search in Google Scholar

[10] F. Bautista, et al.., “Experimental evidence of the critical phenomenon and shear banding flow in polymer-like micellar solutions,” J. Non-Newtonian Fluid Mech., vols. 177–178, pp. 89–96, 2012, https://doi.org/10.1016/j.jnnfm.2012.03.006.Search in Google Scholar

[11] J. P. García-Sandoval, O. Manero, F. Bautista, and J. E. Puig, “Inhomogeneous flows and shear banding formation in micellar solutions: predictions of the BMP model,” J. Non-Newtonian Fluid Mech., vols. 179–180, pp. 43–54, 2012, https://doi.org/10.1016/j.jnnfm.2012.05.006.Search in Google Scholar

[12] G. Landázuri, et al.., “On the modelling of the shear thickening behavior in micellar solutions,” Rheol. Acta, vol. 55, no. 7, pp. 547–558, 2016, https://doi.org/10.1007/s00397-016-0933-8.Search in Google Scholar

[13] A. N. Morozov, A. V. Zvelindovsky, and J. G. E. M. Fraaije, “Orientational phase transitions in the hexagonal phase of a diblock copolymer melt under shear flow,” Phys. Rev. E, vol. 61, no. 5, pp. 4125–4132, 2000. https://doi.org/10.1103/physreve.61.4125.Search in Google Scholar PubMed

[14] O. Manero, J. Pérez-López, J. Escalante, J. Puig, and F. Bautista, “A thermodynamic approach to rheology of complex fluids: the generalized bmp model,” J. Non-Newtonian Fluid Mech., vol. 146, no. 1, pp. 22–29, 2007, 3rd Annual European Rheology Conference, https://doi.org/10.1016/j.jnnfm.2007.02.012.Search in Google Scholar

[15] K. Glasner, “Solute trapping and the non-equilibrium phase diagram for solidification of binary alloys,” Physica D: Nonlinear Phenomena, vol. 151, no. 2, pp. 253–270, 2001, https://doi.org/10.1016/s0167-2789(01)00231-7.Search in Google Scholar

[16] A. Martín del Campo and J. P García-Sandoval, “Shear-induced structural and thermodynamic phase transitions in micellar systems,” Euro. Phys. J. E, vol. 40, no. 2, p. 20, 2017, https://doi.org/10.1140/epje/i2017-11508-6.Search in Google Scholar PubMed

[17] S. Chatterjee, P. Pradhan, and P. K. Mohanty, “Zeroth law and nonequilibrium thermodynamics for steady states in contact,” Phys. Rev. E, vol. 91, no. 6, p. 062136, 2015. https://doi.org/10.1103/physreve.91.062136.Search in Google Scholar PubMed

[18] R. Dickman, “Phase coexistence far from equilibrium,” New J. Phys., vol. 18, no. 4, p. 043034, 2016. https://doi.org/10.1088/1367-2630/18/4/043034.Search in Google Scholar

[19] S. Katz, J. L. Lebowitz, and H. Spohn, “Phase transitions in stationary nonequilibrium states of model lattice systems,” Phys. Rev. B, vol. 28, no. 3, pp. 1655–1658, 1983. https://doi.org/10.1103/physrevb.28.1655.Search in Google Scholar

[20] S. Katz, J. L. Lebowitz, and H. Spohn, “Nonequilibrium steady states of stochastic lattice gas models of fast ionic conductors,” J. Stat. Phys., vol. 34, nos. 3–4, pp. 497–537, 1984. https://doi.org/10.1007/bf01018556.Search in Google Scholar

[21] D. Jou, J. Casas-Vazquez, and G. Lebon, Extended Irreversible Thermodynamics, New York, Dordrecht, Heidelberg, London, Springer, 2010.10.1007/978-90-481-3074-0Search in Google Scholar

[22] M. Grmela and H. C. Õttinger, “Dynamics and thermodynamics of complex fluids. I. development of a general formalism,” Phys. Rev. E, vol. 56, no. 6, pp. 6620–6632, 1997. https://doi.org/10.1103/physreve.56.6620.Search in Google Scholar

[23] B. D. Coleman, “Thermodynamics of materials with memory,” Arch. Ration. Mech. Anal., vol. 17, no. 1, pp. 1–46, 1964. https://doi.org/10.1007/bf00283864.Search in Google Scholar

[24] J. W. Gibbs, “On the equilibrium of heterogeneous substances,” Am. J. Sci., vol. s3–16, no. 96, pp. 441–458, 1878. https://doi.org/10.2475/ajs.s3-16.96.441.Search in Google Scholar

[25] G. Antonoff, “What is a phase?,” J. Chem. Educ., vol. 21, no. 4, p. 195, 1944. https://doi.org/10.1021/ed021p195.2.Search in Google Scholar

[26] M. Modell and R. C. Reid, “Thermodynamics and its applications,” in Prentice-Hall International Series in the Physical and Chemical Engineering Sciences, New York, Prentice-Hall, 1974.Search in Google Scholar

[27] T. Frolov and Y. Mishin, “Phases, phase equilibria, and phase rules in low-dimensional systems,” J. Chem. Phys., vol. 143, no. 4, p. 044706, 2015. https://doi.org/10.1063/1.4927414.Search in Google Scholar PubMed

[28] M. Fisher and C. Radin, “Definition of thermodynamic phases and phase transitions,” AIM Workshop Phase Transitions, 2006. Available at: http://www.aimath.org/WWN/phasetransition/Defs16.pdf [accessed: Feb. 2, 2025].Search in Google Scholar

[29] S. K. Prasad, “Photo-stimulated and photo-suppressed phase transitions,” Mol. Cryst. Liq. Cryst., vol. 509, no. 1, pp. 317/[1059]–327/[1069], 2009, https://doi.org/10.1080/15421400903065887.Search in Google Scholar

[30] S. Khajehpour Tadavani and A. Yethiraj, “Tunable hydrodynamics: a field-frequency phase diagram of a non-equilibrium order-to-disorder transition,” Soft Matter, vol. 13, no. 40, pp. 7412–7424, 2017, https://doi.org/10.1039/c7sm01145h.Search in Google Scholar PubMed

[31] H. Takizawa, “Survey of new materials by solid state synthesis under external fields: high-pressure synthesis and microwave processing of inorganic materials,” J. Ceram. Soc. Jpn., vol. 126, no. 6, pp. 424–433, 2018, https://doi.org/10.2109/jcersj2.18036.Search in Google Scholar

[32] G. Auletta, L. Rondoni, and A. Vulpiani, “On the relevance of the maximum entropy principle in non-equilibrium statistical mechanics,” Euro. Phys. J. Spec. Top., vol. 226, no. 10, pp. 2327–2343, 2017, https://doi.org/10.1140/epjst/e2017-70064-x.Search in Google Scholar

[33] J. O’Byrne, A. Solon, J. Tailleur, and Y. Zhao, “An introduction to motility-induced phase separation,” in Out-of-Equilibrium Soft Matter, London, The Royal Society of Chemistry, 2023, pp. 107–150.10.1039/9781839169465-00107Search in Google Scholar

[34] A. Sarracino, D. Villamaina, G. Gradenigo, and A. Puglisi, “Irreversible dynamics of a massive intruder in dense granular fluids,” Europhys. Lett., vol. 92, no. 3, p. 34001, 2010. https://doi.org/10.1209/0295-5075/92/34001.Search in Google Scholar

[35] F. Bautista, J. Soltero, J. Pérez-López, J. Puig, and O. Manero, “On the shear banding flow of elongated micellar solutions,” J. Non-Newtonian Fluid Mech., vol. 94, no. 94, pp. 57–66, 2000, https://doi.org/10.1016/s0377-0257(00)00128-2.Search in Google Scholar

[36] S. Grossmann, Advances In Solid State Physics, Ch. Analytic Properties of Thermodynamic Functions and Phase Transitions, Festkõrpeperproblem 9, Berlin, Heidelberg, Springer, 1969, pp. 207–254.10.1016/B978-0-08-015543-2.50011-5Search in Google Scholar

[37] R. B. Griffiths, “Thermodynamic functions for fluids and ferromagnets near the critical point,” Phys. Rev., vol. 158, no. 1, pp. 176–187, 1967, https://doi.org/10.1103/physrev.158.176.Search in Google Scholar

[38] F. Sawano, et al.., “An organic thyristor,” Nature, vol. 437, no. 7058, pp. 522–524, 2005. https://doi.org/10.1038/nature04087.Search in Google Scholar PubMed

[39] D. Endo, Y. Fukazawa, M. Matsumoto, and S. Nakamura, “Electric-field driven nonequilibrium phase transitions in ads/cft,” J. High Energy Phys., vol. 2023, no. 3, p. 173, 2023. https://doi.org/10.1007/jhep03(2023)173.Search in Google Scholar

Received: 2024-07-17
Accepted: 2025-01-27
Published Online: 2025-02-13
Published in Print: 2025-04-28

© 2025 the author(s), published by De Gruyter, Berlin/Boston

This work is licensed under the Creative Commons Attribution 4.0 International License.

Downloaded on 13.10.2025 from https://www.degruyterbrill.com/document/doi/10.1515/jnet-2024-0063/html
Scroll to top button