REVIEW

published: 02 October 2018

doi: 10.3389/fphy.2018.00097

Mode-Coupling Theory of the Glass

Transition: A Primer
Liesbeth M. C. Janssen*

Theory of Polymers and Soft Matter, Department of Applied Physics, Eindhoven University of Technology, Eindhoven,
Netherlands

Understanding the physics of glass formation remains one of the major unsolved
challenges of condensed matter science. As a material solidifies into a glass, it exhibits
a spectacular slowdown of the dynamics upon cooling or compression, but at the
same time undergoes only minute structural changes. Among the numerous theories
put forward to rationalize this complex behavior, Mode-Coupling Theory (MCT) stands
out as a unique framework that provides a fully first-principles-based description of glass
phenomenology. This review outlines the key physical ingredients of MCT, its predictions,
successes, and failures, as well as recent improvements of the theory. We also discuss
the extension and application of MCT to the emerging field of non-equilibrium active soft
matter.
Keywords: mode-coupling theory, glass transition, molecular hydrodynamics, liquid structure, amorphous solids,
supercooled liquids, colloids, active matter

Edited by:
Jennifer Lynn Ross,
1. INTRODUCTION TO THE PHYSICS OF GLASS FORMATION
University of Massachusetts Amherst,
United States
Glasses are solid materials that lack any long-range structural order, representing a state of matter
that lies somewhere in between a crystalline solid and a disordered liquid. The most common
Reviewed by:
pathway toward a glassy state is by rapidly cooling a liquid to below its melting point–thus entering
Ramon Castañeda-Priego,
Universidad de Guanajuato, Mexico the so-called supercooled regime–, until the liquid’s viscosity η simply becomes so large that it
J. M. Schwarz, stops flowing on any practical time scale [1–3]. The operational definition of the glass transition
Syracuse University, United States temperature Tg is the point where the viscosity exceeds a value of 1012 Pa.s or the structural
*Correspondence: relaxation time τ exceeds 100 s, but most glasses in our everyday lives have a viscosity that is
Liesbeth M. C. Janssen still orders of magnitude higher [4]. Aside from common applications such as window panes
l.m.c.janssen@tue.nl and household items, amorphous solids can be found in, e.g., phase-change memory devices,
pharmaceutical compounds, optical fibers, and wearable electronics, and there is compelling
Specialty section: evidence that even living cells employ glass-like behavior to regulate intra- and intercellular
This article was submitted to processes [5–11]. Curiously, most of the water in the universe is also believed to exist in the glassy
Physical Chemistry and Chemical
state [12].
Physics,
Given the vast abundance and importance of glasses, it may come as a surprise that we still
a section of the journal
Frontiers in Physics understand very little about them. In fact, after decades of intense research, there is still no
consensus on which physical mechanisms underlie the process of glass formation. Unraveling the
Received: 01 June 2018
nature of the glassy state ranks among the “most compelling puzzles and questions facing scientists
Accepted: 17 August 2018
Published: 02 October 2018 today” [13], and Nobel laureate Philip Anderson even called it “the deepest and most interesting
unsolved problem in solid-state theory” [14]. What makes the glass transition so notoriously
Citation:
Janssen LMC (2018) Mode-Coupling
difficult to understand? At the heart of the problem lies the fact that a vitrifying material exhibits
Theory of the Glass Transition: A a spectacular growth of viscosity (or relaxation time) upon cooling or compression, but at the
Primer. Front. Phys. 6:97. same time undergoes only minute structural changes. Thus, at the molecular level, the structure
doi: 10.3389/fphy.2018.00097 of a glass is almost indistinguishable from that of a normal liquid (as probed by, e.g., the radial

Frontiers in Physics | www.frontiersin.org 1 October 2018 | Volume 6 | Article 97

Janssen Mode-Coupling Theory of the Glass Transition: A Primer

FIGURE 2 | Typical fragility plot, showing the logarithm of the viscosity as a

function of inverse temperature 1/T, normalized with respect to the glass
transition temperature Tg . A viscosity of 10−3 Pa.s corresponds to a normal
liquid, while a value of 1012 Pa.s defines a glassy solid. So-called strong glass
formers such as silica exhibit an Arrhenius-type growth of the viscosity upon
cooling, while fragile glass formers such as o-terphenyl show a much steeper
temperature dependence close to Tg . Many materials, including colloidal hard
spheres and confluent cells, fall in between these two extremes.

FIGURE 1 | Schematic picture of the structure of (A) a normal liquid, (B) an such as silica fall in the class of “strong” glass formers, exhibiting
amorphous solid, i.e., a glass, and (C) a crystalline solid. The first panel
an Arrhenius-type (exponential) viscosity growth upon cooling,
highlights four (green) particles which are separated by a distance r from a
(red) reference particle. The right panels illustrate the corresponding radial while “fragile” materials have a viscosity that increases faster
distribution functions g(r), which describe the probability of finding a particle a than an Arrhenius law. It is widely believed that a thorough
distance r away from any reference particle, relative to the ideal-gas case. The understanding of the mechanisms underlying fragility will be key
first peak in g(r) represents the first solvation shell at r ≈ 1d, where d is the to achieving a universal description of the glass transition, but
particle diameter. The dashed black lines indicate the ideal-gas result g(r) = 1.
no theory to date has been able to predict a material’s degree
of fragility from the sole knowledge of its microscopic structure
[20].
While the viscosity already gives an important clue about the
distribution function or the static structure factor), yet their complex behavior of glass-forming materials, the most detailed
viscosities differ by at least fifteen (!) orders of magnitude. This information is contained in the microscopic relaxation dynamics,
is unlike any conventional thermodynamic phase transition, and this will also be the focus of the remainder of this review. A
such as the liquid-to-crystal transition, which is marked common probe of such dynamics is the time-dependent density-
by the appearance of long-range, periodic structural order density correlation function or so-called intermediate scattering
(Figure 1). Nonetheless, it is not unimaginable that some kind function, F(k, t), which probes correlations in particle density
of “amorphous order” emerges during vitrification, albeit in a far fluctuations over a certain wavenumber k and over a time
less obvious way than in the crystallization example. A popular interval t [21]. Simply put, F(k, t) measures to what extent the
hypothesis is that the subtle microstructural changes observed instantaneous molecular configuration of a material will resemble
in supercooled liquids might somehow contain a “hidden” the new configuration a time t later; the wavenumber k designates
growing (and possibly diverging) length scale that accompanies the inverse length scale over which this resemblance is measured.
the transition from liquid to amorphous solid, and indeed a By choosing k as approximately one inverse particle diameter,
large ongoing effort is devoted to identifying such a length scale F(k, t) will thus probe the relaxation dynamics at the molecular
[15–17]. level, while the limit k → 0 describes the macroscopic dynamics.
Another major unresolved piece of the glass puzzle is that not We note that the characteristic relaxation time τ associated with
all materials vitrify in the same manner. More specifically, the F(k, t) is also a measure for the viscosity (with the shear modulus
viscosity growth as a function of inverse temperature can differ as the proportionality factor [4]), and hence F(k, t) also provides
significantly from one material to another. These differences are a means to quantify e.g. the fragility.
captured in an empirical property called “fragility” [1, 18, 19], The behavior of F(k, t) upon cooling thus reveals how the
which characterizes the slope of the viscosity with temperature microscopic relaxation dynamics changes during the vitrification
as a material approaches the glass transition (Figure 2). Materials process [22, 23] (Figure 3). In a normal high-temperature liquid,

Frontiers in Physics | www.frontiersin.org 2 October 2018 | Volume 6 | Article 97

Janssen Mode-Coupling Theory of the Glass Transition: A Primer

FIGURE 3 | Schematic picture of the structure and dynamics in a normal liquid, supercooled liquid, and glass. (A,B) depict a typical trajectory of a particle in the
normal liquid phase and glassy phase, respectively. In the glassy state, particles become trapped in a cage formed by their neighbors. The dashed red line indicates
the typical size of a cage, with a radius of approximately one particle diameter d. (C) Shows the static structure factors S(k) for a glass-forming system of hard spheres
for several packing fractions φ, calculated using the Percus-Yevick approximation. The main peak position of S(k) corresponds to a wavenumber of approximately one
inverse particle diameter, k ≈ 2π/d. Within MCT, the glass transition for this system takes place at φc = 0.516 [24]. (D) Shows typical intermediate scattering
functions F(k, t) as a function of time for k ≈ 2π/d. As the temperature is decreased or the packing fraction is increased, the system becomes more glassy and F(k, t)
decays more slowly.

F(k, t) will decay to zero in a rapid and simple exponential There are several other aspects in the dynamics of supercooled
fashion, since the particles can move around easily and therefore liquids that differ markedly from those seen in ordinary liquids,
quickly lose track of their initial positions. At temperatures in including the emergence of dynamic heterogeneity [26–29]
the supercooled regime, however, F(k, t) shows a more complex and the breakdown of the Stokes-Einstein relation [30, 31].
multi-step relaxation pattern (also see Figure 7): at intermediate Dynamic heterogeneity refers to the fact that structural relaxation
times (the so-called β-relaxation regime), a plateau develops does not take place uniformly throughout the entire material–
during which F(k, t) remains constant, indicating the transient as in a normal liquid–, but rather in clusters of collectively
freezing of particles; only at sufficiently long times will the rearranging particles, while the rest of the supercooled liquid
correlation function fully decay to zero. Notably, this final remains temporarily frozen (Figure 4). The appearance of
decorrelation process (so-called α-relaxation) is not a simple such mobile domains will vary both in space and in time,
exponential decay, as in a normal liquid, but rather a more thus giving rise to non-trivial spatiotemporal fluctuations that
slowly decaying, “stretched” exponential behavior of the form become more pronounced as the glass transition is approached.
exp(−t/τ )β , with 0 < β < 1. As the temperature decreases Dynamic heterogeneity cannot be seen in F(k, t) itself, but
toward the glass transition temperature, the plateau in F(k, t) will rather in the fluctuations of F(k, t) among different particle
extend to increasingly long times, until it finally exceeds the entire trajectories [32, 33]. These fluctuations are encoded in the so-
time window of observation. Thus, at the glass transition, F(k, t) called dynamic susceptibility χ4 (t), whose peak height is a
fails to decorrelate on any practical time scale–implying that measure for the size of the cooperatively rearranging regions.
particles always stay reasonably close to their initial positions–, As a material is supercooled, a growing χ4 (t) thus indicates a
marking the onset of solidity. The final value of the intermediate growing dynamic length scale associated with vitrification, but
scattering function, f (k) = limt→∞ F(k, t), is known as the a true divergence of this length scale–as expected for typical
non-ergodicity parameter [25], and is often used as the order critical phenomena–has not yet been observed [17]. A related
parameter for the glass transition: f (k) = 0 corresponds to the puzzling phenomenon concerns the Stokes-Einstein equation,
liquid state, and f (k) > 0 indicates a solid (Figure 3D). which states that the viscosity (or relaxation time), diffusion

Frontiers in Physics | www.frontiersin.org 3 October 2018 | Volume 6 | Article 97

Janssen Mode-Coupling Theory of the Glass Transition: A Primer

In this review, we focus on one of several theories that seeks

to describe the above complex phenomenology of glass-forming
materials, namely Mode-Coupling Theory (MCT) [36, 37]. This
theory was first put forward by Götze and coworkers in the
1980s [24, 38], and is one of the few frameworks of glassy
dynamics that is based entirely on first principles, starting from
the exact microscopic picture of a correlated liquid. A somewhat
related and more recent theory, the Self-Consistent Generalized
Langevin Equation (SCGLE) approximation [39], will also briefly
be discussed. We outline the key physical ingredients and sketch
of the MCT derivation, its predictions, successes, and failures, as
well as recent improvements and extensions of the theory. Part of
FIGURE 4 | Illustration of dynamic heterogeneity in supercooled liquids. The this work is based on the review by Reichman and Charbonneau
red-colored particles represent mobile particles that have moved further than a [22], which also contains a full derivation of the MCT equations,
certain distance 1r during a time interval 1t, while blue-colored particles
represent particles that have moved less than this distance in the same time
and the review by Szamel [40]; For a detailed discussion of
interval. In the normal liquid phase (A), particle motion occurs rather the original theory, including an extensive treatment of the
homogeneously across the entire sample. Conversely, in a supercooled liquid involved mathematics, we refer to the seminal work of Götze
(B), particle motion occurs heterogeneously in clusters of collectively moving [36]. For an overview of the many other existing theories
particles, and the appearance of such mobile clusters fluctuates both in space
of glass formation, such as free volume theory, Adam-Gibbs
and in time. The figure is based on Weeks et al. [26].
theory, Random First Order Transition (RFOT) theory, dynamic
facilitation and kinetically constrained models, energy landscape
approaches, and geometric frustration, see e.g. Refs. [2, 15, 20,
constant D, and temperature of a liquid are related as Dη/T = 41–43]. A discussion of these alternative theories falls outside
constant. This ratio holds generally for normal liquids, but in the the scope of the present work; here we only mention that, of all
supercooled regime the viscosity increase tends to be stronger the other existing frameworks, RFOT theory is directly related
than the diffusion-constant decrease. This breakdown of Stokes- to MCT at high temperatures, but is further augmented with
Einstein behavior is widely believed to be a manifestation of thermodynamic, Adam-Gibbs-like concepts at low temperatures.
dynamic heterogeneity, but the fundamental origins of both
phenomena remain poorly understood. 2. DERIVATION OF THE MCT EQUATIONS
Finally, we mention another hallmark of glassy dynamics that
is rather general for out-of-equilibrium systems, namely aging 2.1. Preliminaries
[2, 20, 23, 34, 35]. Aging implies that the behavior of a material As already noted in the introduction, MCT provides a purely
depends explicitly on its age, i.e., the structural and dynamical first-principles route toward the description of glassy behavior,
properties may slowly change as time progresses. These changes making it a unique theory that does not rely on any
are commonly a manifestation of the material’s gradual approach phenomenological assumptions. Explicitly, MCT aims to predict
to an equilibrium state. In the supercooled phase, such aging the full microscopic relaxation dynamics of a glass-forming
effects are usually observed after a (small) temperature quench, material–as a function of time, wavenumber, temperature, and
but vanish after a sufficiently long equilibration time. Indeed, density–, using only knowledge of static, time-independent
“properly equilibrated” supercooled liquids that are cooled properties as input. Aside from constants such as the system’s
sufficiently slowly (such that there is ample time for the material temperature and density, the main theory input is the average
to undergo full structural relaxation at a given temperature) microscopic structure of the material. The simplest experimental
behave as ordinary equilibrium liquids in the sense that, e.g., measure of the latter is the static structure factor S(k), which
ergodicity and the fluctuation-dissipation theorem hold. Within can be obtained directly from scattering experiments. This
the glass state, however, ergodicity is broken and the relaxation structure function is related to the radial distribution function
time to reach equilibrium exceeds–by definition–any practical g(r) (Figure 1) through a Fourier transform [21, 44], and thus
time scale. Hence, a glass can be regarded as a supercooled liquid probes–in Fourier space–the likelihood of finding a particle at
that has fallen out of equilibrium, and its properties depend a certain distance r ∼ 2π/k away from any other particle
explicitly on its history. There is a large body of literature (Figure 3D). Formally S(k) is also equivalent to F(k, t = 0).
devoted to aging effects in glasses (see e.g. Ref. [2] and references It must be noted that MCT also admits more intricate three-
therein), but here we only briefly mention that physical aging particle correlation functions as additional structural input, but–
within the glassy state is generally associated with the material’s with the exception of network-forming fluids [45, 46]–the sole
tendency to reach a lower-energy state that–eventually–will knowledge of S(k) generally suffices. Importantly, it is through
correspond to a deeply supercooled (quasi-equilibrium) liquid these structural metrics that MCT knows about the chemical
phase at a temperature below the original Tg . Whether the composition of the material under study. That is, the theory is
equilibrium supercooled liquid branch terminates at a low but able to distinguish between, say, a glass-forming fluid of silica
finite temperature, the so-called Kauzmann temperature TK or Lennard-Jones particles only through their differences in
(TK < Tg ), remains a matter of debate. (wavevector-dependent) structure.

Frontiers in Physics | www.frontiersin.org 4 October 2018 | Volume 6 | Article 97

Janssen Mode-Coupling Theory of the Glass Transition: A Primer

modes,
1
F(k, t) = hρ(−k, 0)ρ(k, t)i, (2)
N
where the brackets denote a canonical ensemble average. At time
t = 0, this correlation function reduces to the static structure
factor,
1
S(k) = hρ(−k, 0)ρ(k, 0)i ≡ F(k, 0), (3)
N
which thus contains information on the static density
distribution of the material, i.e., the average microscopic
FIGURE 5 | Sketch of the MCT equations. The theory seeks to predict the full structure. Note that in an isotropic material, such as a powder
dynamics of the intermediate scattering function F(k, t) for all possible or a “simple” fluid, both S(k) and F(k, t) depend only on the
wavevectors k and all times t. The exact F(k, t) dynamics is governed by a
memory function that, within standard MCT, is approximated as a product of
magnitude of the wavector, k = |k|, but in e.g. the presence
two intermediate scattering functions that probe density correlations at of an external field the full wavevector dependence should be
different wavevectors q and k − q. The couplings among the different considered [47].
wavevectors {k, q, k − q} are determined by the so-called vertices, which
depend explicitly on the static structure factor. Hence, the static structure 2.2. Mori-Zwanzig Projection Formalism
factor must be given as input to the theory, and dynamical information is given In order to obtain an exact equation of motion for F(k, t), we
as output.
make use of the so-called Mori-Zwanzig projection formalism
[48, 49]. The basic idea behind this formalism is to divide the
entire universe into two mutually orthogonal subspaces: one
In the standard formulation of MCT, the theory seeks containing the variables of interest, and one simply containing
to predict the full dynamics of the intermediate scattering “everything else.” The goal is to describe how the dynamics
function F(k, t) of a given material, starting with the exact of the relevant variables evolves over time, in the presence of
equation of motion for F(k, t). Below we sketch the derivation all other “non-interesting” variables. Physically, this idea relies
of this equation, followed by a discussion of the various MCT on a separation of time scales in the dynamics, whereby the
approximations made to solve it. Briefly, the derivation will fast variables are integrated out. Here we will focus mainly on
amount to an exact integro-differential equation for F(k, t) molecular glass-forming fluids, in which case the variables of
(Equation 10) that is governed by an even more complicated interest are the collective density modes of Equation (1) and their
time-dependent correlation (“memory”) function. MCT makes associated current modes
the ad hoc assumption that the latter memory function can be N
approximated as a product of F(k, t) functions, thus yielding a (k · ṙl )eikrl (t) ,
X
j(k, t) ≡ ρ̇(k, t) = i (4)
closed, self-consistent equation (see Figure 5). As described in l=1
section 2.3, the final MCT equation (Equation 12) is reminiscent
of a damped harmonic oscillator, but with a time-dependent where the dots denote time derivatives. Note that in general
damping term that ultimately produces the dramatic dynamic there is no simple recipe for deciding which variables are
slowdown in supercooled liquids. “relevant”; typically we focus on quasi-conserved or “slow”
Let us first define our variables of interest, namely the variables that show some non-trivial time-dependence (unlike
collective density modes, strictly conserved variables that are constant), but which do not
fluctuate too fast either, so as not to be confused with noise.
N
X From Equation (4), it is easy to see that in the limit k → 0,
ρ(r, t) = δ(r − rj (t)), (1) corresponding to very large length scales, the current ρ̇(k, t)
j will vanish and consequently the macroscopic density is strictly
Z conserved. On smaller length scales, however, i.e., k > 0, the local
ρ(k, t) = dreikr ρ(r, t) density will fluctuate as particles move around, and it is these
N
fluctuations–and their time-dependent correlations–that we seek
=
X
eikrj (t) , to probe in F(k, t) and predict with MCT.
j
For convenience we will organize the variables ρ(k, t) and
j(k, t) into a two-component vector A, which thus spans our
where N denotes the total number of particles and rj (t) is the subspace of interest:
position of particle j at time t. The real-space density ρ(r, t) thus 
A (t)
 
ρ(k, t)

simply measures where all particles are located at a given point A(t) ≡ 1 = . (5)
A2 (t) j(k, t)
in time, and ρ(k, t) is the corresponding Fourier transform for
wavevector k. The intermediate scattering function F(k, t) probes Importantly, in this notation, time-dependent correlation
the time-dependent correlations between these collective density functions may now be identified simply as scalar

Frontiers in Physics | www.frontiersin.org 5 October 2018 | Volume 6 | Article 97

Janssen Mode-Coupling Theory of the Glass Transition: A Primer

products between such vector elements, e.g., F(k, t) = For the matrix of correlation functions C(t) we similarly
(1/N)hA1 (0)|A1 (t)i = (1/N)hA∗1 (0)A1 (t)i, where we have find
used the standard bra-ket notation with the asterisk representing Z t
dC(t)
the complex conjugate. We define the full matrix of all possible = i · C(t) − dsK(s) · C(t − s). (10)
scalar products as C(t), with matrix elements dt 0

Here,  is the so-called frequency matrix (the name will

Cαβ (t) ≡ hAα (0)|Aβ (t)i. (6) become apparent later on), which captures the part of the
time derivative of A that remains in the slow subspace as
Note that the first matrix element C11 (t) equals NF(k, t), and time evolves, K(s) is a time-dependent memory function, and
C21 (t) = (N/i)(dF(k, t)/dt). Furthermore, in analogy to ordinary f (t) is the “fast” fluctuating force, which is defined as f (t) =
projections in vector space, we can now use these scalar products ei(1−PA )Lt i(1 − PA )LA(0). That is, f (t) is obtained by first
to define a projection operator PA as removing all the “slow” character from the time derivative
X of A using the complementary projection operator (1 − PA ),
PA = |Aα i[C(0)−1 ]αβ hAβ | (7) and is subsequently propagated in time in the “fast” subspace
α,β orthogonal to A. The memory function K(t) is given by the time-
autocorrelation function of this fluctuating force; physically,
where the sums run over all possible matrix elements. The K(t) represents a dissipative term that ultimately breaks the
projection of some vector X onto A is then given by PA X. conservation of A. In other words, K(t) and f (t) embody how
Such a projection essentially extracts all the “slow” or “relevant” our slow variable A–which at time t = 0 lives strictly in the
character (defined through A) from an arbitrary variable X, slow subspace–will gradually evolve under the influence of the
leaving the remaining part of X orthogonal to A. It is easy to show rest of the universe, e.g. in the presence of “fast” variables such
that PA2 = PA and PA A = A, i.e., the projection of A onto itself as thermal noise. Note that in arriving at Equation (10), we
returns A. This projection formalism, introduced by Zwanzig have used that hA(0)|f (t)i = 0 by construction. Importantly,
and Mori, thus establishes a link between dynamic variables Equations (9) and (10), which are known as the generalized
and standard vector algebra. Without any loss of generality, it Langevin equation and memory equation, respectively, are both
will enable us to separate the full dynamical behavior of our exact.
system into two contributions: i) the dynamics evolving in the
“slow” subspace spanned by A(0), and ii) the dynamics due to all 2.3. Mode-Coupling Theory
remaining “fast” variables, obtained simply by projecting out all Approximations
the slow A-character from the full dynamics. By Equation (10), the difficulty of predicting the full time-
Let us now look explicitly at the time-dependent dynamics dependent dynamics of F(k, t) is now deferred to the the question
of a glass-forming supercooled liquid. For classical fluids that of how the memory function K(t) evolves with time. In general,
obeys Newton’s equation of motion, the time evolution of A(t) there is no rigorous solution for this equation, and hence
can always be formally written as approximations must be made. The main idea behind MCT is
to approximate K(t) in “the simplest non-trivial way” using a
A(t) = eiLt A(0), (8) two-step approach:
1. Approximate the memory function as a four-point
where L is the so-called Liouvillian operator. The definition of L density correlation function. First, using the density modes as
can be found in (e.g., [22]), but here we will not be concerned the main physical variables of interest, the fluctuating force f (t)
with its explicit form; it suffices to know that this operator is projected onto a new basis of products of two density modes,
governs the full dynamics of our variables of interest. Note ρ(k1 , t)ρ(k2 , t), where k1 and k2 run over all possible wavevectors
that for colloidal glass-forming systems undergoing Brownian relevant to our system. Physically, this projection is motivated
rather than Newtonian motion, a similar equation applies when by the fact that for particles interacting through an arbitrary
considering only the density modes in A and replacing the pair potential, such products of densities emerge naturally in
Liouvillian by the so-called Smoluchowski operator [50]. the expression for the fluctuating force [22]. This may seem
While Equation (8) is formally exact, it does not necessarily rather counterintuitive at first, since the fluctuating force is a
yield any new physical insight into the complex time-dependent fast variable while density modes are slow by definition, but it
dynamics of supercooled liquids. Instead, we can rewrite this can be shown by Fourier transformation that, for an n-body
equation through a somewhat lengthy derivation (involving the interaction potential, f (t) always contains products of n density
insertion of the unit matrix operator 1 = PA + 1 − PA and modes [41]. In the standard MCT formulation, it is assumed
separating the time-evolution operator exp (iLt) into a “slow” that the pair densities dominate the fluctuating force entirely, but
component and its orthogonal part) in the following form higher-order generalizations with projections onto an n-density-
[22]: mode basis have also been considered [51, 52]. Mathematically,
the projection onto pair densities also corresponds to the first
dA(t) t non-vanishing component in density space, i.e., “the simplest
Z
= i · A(t) − dsK(s) · A(t − s) + f (t). (9) non-vanishing term,” since a projection onto a single density
dt 0

Frontiers in Physics | www.frontiersin.org 6 October 2018 | Volume 6 | Article 97

Janssen Mode-Coupling Theory of the Glass Transition: A Primer

mode would always give zero by construction [36]. Overall, this 23, 54]), but in general the vertices may also contain higher-order,
approximation brings the memory function K(t), which is the triplet-density correlations [45, 55]. Equation (12) is a closed, self-
time-correlation function of f (t), into the form of a four-point consistent equation, and is subject to the boundary conditions
density correlation function: F(k, 0)/S(k) = 1 and Ḟ(k, 0) = 0.
Let us briefly compare this MCT result with the equation
hρ ∗ (k1 , 0)ρ ∗ (k2 , 0)ei[1−PA ]Lt ρ(k3 , 0)ρ(k4 , 0)i, of motion for a one-dimensional damped harmonic oscillator:
X
K(t) ∼
k1 ,k2 ,k3 ,k4 ẍ + ω2 x + 2ζ ωẋ = 0, where ω is the frequency of the undamped
(11) oscillator and ζ is the damping coefficient. It can be seen that the
MCT equation is rather similar, with x(t) ∼ F(k, t) and 21 =
with the time-propagation operator exp [i(1 − PA )Lt] acting in kB Tk2 /[mS(k)] playing the role of ω2 . Hence, the  matrix is
the fast subspace. referred to as the frequency matrix. The damping coefficient,
2. Factorize four-point correlation functions into two- on the other hand, appears in the MCT equation in the form
point correlation functions. Second, the (unknown) four-point of the memory function KMCT (t) (note the first derivative of
correlation functions in K(t) are further simplified by factorizing F(k, t) in the integrand). Consequently, we may interpret the
them into a product of two two-point correlation functions memory function as a generalized, time-dependent damping,
hρ ∗ (k1 , 0)ρ(k1 , t)i and hρ ∗ (k2 , 0)ρ(k2 , t)i. At the same time, the which will ultimately cause the dynamical slowdown in F(k, t)
operator exp [i[1 − PA ]Lt] is replaced by the normal operator [23].
exp [iLt], since the single density modes ρ(k1 , t) and ρ(k2 , t), To make the comparison with a damped harmonic oscillator
which start out in the slow subspace, would otherwise give a zero more explicit, let us first drop all wavevector dependence in the
contribution. It is important to note that this factorization is an MCT equations and write x(t) = F(k, t) and 21 = ω2 , so
that 2
ad hoc approximation that is not necessarily motivated by any R t the MCT equation of motion becomes ẍ(t) + ω x(t) +
physical insight; rather, it merely serves to produce a “simple” 0 KMCT (s)ẋ(t − s)ds = 0. Such simplified, k-independent
memory function that is not trivially zero. Nonetheless, it can MCT equations are known as schematic MCT models [24,
be shown that the factorization is exact for so-called Gaussian 38]. We may now write the schematic memory function as
variables [53], but density modes in general do not behave as KMCT (t) = ax2 (t), with a denoting a damping factor that
such. represents the effective strength of the vertices at a given
After the second approximation is made, we may then temperature. Note that if the memory function would be a
realize that the factorized two-point density correlation functions simple delta-function aδ(t), the schematic MCT equation would
hρ ∗ (ki , 0)ρ(ki , t)i are, in fact, equal to NF(ki , t) by virtue reduce to a damped harmonic oscillator with a = 2ζ ω. The
of Equation (2). Thus, our full equation of motion for the fact that the true (schematic) MCT memory function contains
intermediate scattering function F(k, t) is now governed by a the non-linear product x2 (t) instead of a simple delta-function,
memory function containing precisely the same function, but for however, has important consequences for the dynamics and gives
many different wavenumbers. After explicitly working out all the rise to a strong feedback effect that is absent in an ordinary
expressions for the frequency matrix  and the (approximate) damped oscillator. Figure 6 shows the solutions x(t) for a one-
memory function K(t), and concentrating on the lower left dimensional damped harmonic oscillator and for the schematic
corner of the correlation matrix C21 (t) in Equation (10), we MCT model as a function of time for different damping factors a.
finally arrive at the full MCT equation [22]: It may be seen that, as the damping a is increased, the harmonic
oscillator undergoes only a moderate change in the dynamics,
d2 F(k, t) kB Tk2
Z t dF(k, t − s) while the MCT solution develops a plateau and exhibits an
+ F(k, t) + dsKMCT (k, s) = 0, orders-of-magnitude dynamical slowdown. At a critical damping
dt 2 mS(k) 0 dt
(12) value of ac = 4, the schematic MCT model predicts an
with the memory function given by ergodicity-breaking transition such that the function x(t) fails
to decay to zero on any time scale–i.e., a glass has formed.
ρkB T
Z Increasing the damping further (a > 4) brings the system
KMCT (k, t) = dq|Vq,k−q |2 F(q, t)F(|k − q|, t). (13) more deeply in the glass phase. The full wavevector-dependent
16π 3 m
MCT equation predicts a similar scenario, in which the memory
Here, kB is the Boltzmann constant, m is the particle mass, ρ is function constitutes a non-linear dynamical feedback mechanism
the bulk density, and the factors (with the damping strength implicitly controlled by small
changes in the static structure factor as the temperature is
Vq,k−q = k−1 [(k · q)c(q) + k · (k − q)c(|k − q|)] (14) decreased), though of course the explicit coupling of density
modes at different wavevectors leads to richer (and more
are referred to as vertices, with c(k) = ρ −1 [1 − 1/S(k)] denoting complicated) behavior than that predicted by schematic MCT
the direct correlation function [21]. These vertices represent (see section 3). Despite the simplicity of schematic MCT
the strength of the coupling between different density modes at models, it is now well established that many predictions
wavevectors q and k − q. In arriving at this equation, we have also of schematic MCT–which can often be derived analytically
assumed that S(k) contains all the relevant microscopic structural (see [38])–are also preserved in the full wavevector-dependent
information (using the so-called convolution approximation [22, version, and hence schematic MCT approaches remain widely

Frontiers in Physics | www.frontiersin.org 7 October 2018 | Volume 6 | Article 97

Janssen Mode-Coupling Theory of the Glass Transition: A Primer

matrix equation that explicitly couples the partial intermediate

scattering functions and partial structure factors for all species.
These MCT extensions and variations, however, will not further
be discussed in this review.
As a final note, we mention that an alternative and more
recent first-principles-based theory has been formulated that
is somewhat related to MCT, namely the Self-Consistent
Generalized Langevin Equation approximation [39]. This SCGLE
theory also starts from the exact generalized Langevin equation
(Equation 9), but employs different and somewhat simpler
approximations to obtain a self-consistent equation for the
dynamics. Instead of projecting the fluctuating force onto pair
density modes, the main assumptions of SCGLE theory are
a Vineyard-like approximate relation between the memory
function of F(k, t) and the memory function of the self part
Fs (k, t), and a Gaussian-like approximation for the memory
function of Fs (k, t) that relates its dynamics to the Brownian
motion of individual particles [39]. Overall, SCGLE theory also
FIGURE 6 | Solutions x(t) for a damped harmonic oscillator, amounts to a closed, self-consistent dynamical equation that
ẍ(t) + ω2 x(t) + aẋ(t) = 0 (dashed lines), and for a schematic MCT equation, requires only simple static properties as input. An important
ẍ(t) + ω2 x(t) + a 0t x2 (t − s)ẋ(s)ds = 0 (solid lines), with ω2 = 1 and a = 2ζ ω.
R
advantage is that the SCGLE theory can be readily extended
Note the logarithmic scale for the time axis. As the damping a is increased, the
to account for non-equilibrium aging effects, such that the
MCT solution exhibits a highly non-linear slowdown and becomes non-ergodic
for damping values a ≥ 4, signaling the formation of a glass. For the damped
dynamics depends explicitly on the waiting time or age of
harmonic oscillator, the solutions for a = 3.99 and a = 4 are indistinguishable the material [69–72]. This is to be contrasted with standard
on the scale of this figure. MCT, which in its standard form applies only to stationary
(quasi-)equilibrium systems and consequently cannot make any
predictions of aging phenomena; the extension of MCT to non-
used to gain better insight into the phenomenology of glassy equilibrium aging systems is technically rather involved [73]. The
materials. SCGLE equations have also been extended to multi-component
While analytic solutions of the full wavevector-dependent systems [74, 75] and to non-spherical particles [76, 77], and are
MCT equation generally do not exist, it is always possible to generally somewhat simpler to use than the MCT equations. For
solve the equation numerically, namely by iteratively making an a more detailed discussion of SCGLE theory [see [39, 72]] and
ansatz for F(k, t) for all k, subsequently constructing the memory references therein.
function, and updating F(k, t) until convergence is reached. See
Fuchs et al. [56] and Flenner and Szamel ([57], appendix) for a
detailed discussion of the numerical algorithm to solve the MCT 3. MODE-COUPLING THEORY
equations. We also note that for systems undergoing Brownian PREDICTIONS
instead of Newtonian dynamics, in which case the Liouvillian
should be replaced by the Smoluchowski operator, MCT yields an The microscopic MCT equation, Equation (12), can be solved
identical equation (with kB T/m being replaced by the diffusion for any glass-forming material at a given bulk density ρ and
constant D) [58]; however, the origin of this similarity is subtle temperature T once the corresponding static structure factor
and rather non-trivial. Moreover, it has also been shown that S(k) = S(k; ρ, T) is known. Thus, MCT predicts the full
this equation applies reasonably well to glass-forming polymer microscopic dynamics given only time-independent information
chains [59], suggesting that MCT captures at least some degree of as input. In order to describe the entire vitrification process
universal dynamical behavior. Finally, we note that MCT-based from liquid to glass, one typically measures S(k) for a series
equations have also been formulated for, e.g., the self-part of of temperatures or densities, and performs a separate MCT
the intermediate scattering function Fs (k, t) = he−ikrj (0) eikrj (t) i calculation for every relevant temperature and density. In this
(i.e., the density correlation function for a single particle j) [24], section, we summarize the main successes and failures of such
the shear relaxation function [60], the dynamics under shear MCT predictions.
flow [61] and in confinement [62–64], microrheology studies
[65, 66], fluids composed of anisotropic particles [67], and multi- 3.1. Successes
component glass-forming systems [57, 60, 68]. In case of the Despite the various approximations made in MCT, the theory
latter, the MCT equations can be derived in a similar manner as gives a remarkable set of accuracte predictions. Firstly, MCT
above, except that the density modes (and their corresponding is indeed capable of predicting a glass transition, which is
currents) are replaced by their partial analogs ρxi (k, t), where xi non-trivial considering that the static structure factor S(k)–the
labels a particle species. For an M-component system, the MCT main theory input–changes only very weakly upon vitrification
equations (Equations 12, 13) will then amount to an M × M (Figure 3C). As mentioned earlier, the relaxation time of the

Frontiers in Physics | www.frontiersin.org 8 October 2018 | Volume 6 | Article 97

Janssen Mode-Coupling Theory of the Glass Transition: A Primer

predicted F(k, t) is used as an indicator for the glassiness: at the experimental glass transition [see [79]]. We will return to this
the glass transition, the relaxation time diverges and F(k, t) fails point in the next subsection. Furthermore, MCT predicts that the
to decay to zero on any time scale. The corresponding non- onset and decay of the β-relaxation regime, i.e., the plateau in
ergodicity parameter f (k) is also often in good quantitative F(k, t) at intermediate times, are described by power laws of the
agreement with the results of computer simulations and form F(k, t) ∼ f + At −a and f − Bt b , respectively, where f is the
experiments (see, e.g., [68, 78]). (constant) plateau height (Figure 7). Sufficiently close to Tc , the
Mathematically, MCT’s ability to predict a glass follows from MCT exponents a and b are related as Ŵ(1 − a)2 / Ŵ(1 − 2a) =
the non-linearity of the equation (by virtue of the product of Ŵ(1 + b)2 / Ŵ(1 + 2b), where Ŵ denotes the Gamma function.
two F(k, t) functions in the memory function), which renders This is an entirely non-trivial and remarkable prediction that is
the theory very sensitive to any small change in structural input. fully consistent with experiments and simulations. For the α-
This non-linearity leads to a feedback mechanism that ultimately relaxation regime, i.e., the final decay of F(k, t) on the liquid side
drives the dramatic dynamical slowdown: upon cooling, S(k) of the transition, MCT predicts a stretched exponential of the
will become slightly larger at certain wavevectors, causing form exp(−t/τ )β , with 0 < β ≤ 1 (Figure 7). This is again
the vertices to increase as well. Consequently, the memory in excellent agreement with experimental and simulation data,
function will become larger and produce a stronger damping and physically arises from the coupling of multiple density-mode
for F(k, t). The resulting slower intermediate scattering function relaxation channels over different length scales, each relaxing
will further strengthen the memory function, slowing down the on its own time scale. Another success of MCT that has been
dynamics even more. This non-linear feedback effect explains verified experimentally is its prediction of a time-temperature
at least qualititatively why the relaxation dynamics can change superposition principle, such that F(k, t) = F̂(k, t/τ (T)), where
so dramatically upon only small changes in the structure and F̂ is a master function and τ (T) is the α-relaxation time.
temperature [23]. Among the other celebrated results of MCT, we mention
A related success of MCT is its prediction of the cage here its qualitative prediction of complex reentrant effects in
effect as a microscopic mechanism for vitrification (Figures 3, sticky hard spheres (particles with a hard repulsive core and
7). Caging refers to the fact that, in a supercooled liquid, short-ranged attractions) [80] and ultrasoft repulsive particles
particles become (transiently) trapped in local cages formed [81], which exhibit glass-fluid-glass and fluid-glass-fluid phases
by their neighboring particles, which in turn are trapped in upon a monotonic increase in attraction stength and density,
their respective cages, preventing them from moving around respectively. In the case of sticky hard spheres, MCT has also
as in a normal liquid. This is the molecular origin of the β- provided a qualitative explanation for the existence of the two
relaxation regime, which is manifested as a plateau in F(k, t). distinct glass phases in terms of different dominant length scales
As long as the material is on the supercooled-liquid side of [80]. Furthermore, the schematic version of MCT [24, 38], which
the transition, the particles will eventually manage to escape is obtained by ignoring all wavevector dependence in Equation
their cages, but at and below the glass transition, the cage effect (12), is rigorously exact for certain classes of spin-glass models
keeps them trapped indefinitely. The only motion in the glassy with queched disorder (so-called p-spin spherical spin glasses),
state then corresponds to a vibrational or rattling motion of
the particles within their confining cages. More mathematically,
the cage effect emerges from MCT by considering that the
most prominent change in S(k) upon supercooling occurs at the
main peak at wavenumber k0 , corresponding to length scales of
approximately one particle diameter. As a consequence, the first
intermediate scattering function that falls out of equilibrium at
the glass transition is F(k0 , t), which in turn drives the freezing
on all other wavevectors. Notably, within MCT, the dominant
structural length scale governing vitrification thus remains on
the order of only one particle diameter, in stark contrast with
conventional critical phenomena that are usually accompanied
by diverging, macroscopic length scales. However, as will be
described in section 4.3, recent work suggests that a diverging
length scale also emerges within an extended (“inhomogeneous”)
version of MCT that is related to the dynamic susceptibility χ4 (t).
Regardless of the molecular details of the material, which are
contained in S(k), MCT also makes several general predictions
FIGURE 7 | Typical MCT prediction for F(k, t) of a supercooled liquid as a
for the relaxation dynamics [22, 23, 36, 38]. Firstly, MCT predicts function of time, for a wavenumber k = k0 that corresponds to the first peak of
that close to the glass transition temperature Tc , the relaxation the static structure factor. At very short times, particles undergo ballistic
time F(k, t) will always diverge as a power law, τ ∼ (T − motion. At intermediate times, particles become transiently trapped in cages
Tc )−γ . Such a functional form is often in good agreement with (β-relaxation) and F(k, t) correspondingly remains approximately constant. Only
at sufficiently long times will particles break free and full relaxation takes place
experiments and simulations in the mildly supercooled regime (α-relaxation).
(using γ as a fit parameter), but generally breaks down closer to

Frontiers in Physics | www.frontiersin.org 9 October 2018 | Volume 6 | Article 97

Janssen Mode-Coupling Theory of the Glass Transition: A Primer

pointing toward a possible deep connection between systems falls outside the scope of the present work, but we refer the
with quenched and self-generated disorder. For a more extensive interested reader to [e.g., [43, 83, 86]] for a recent overview.
overview of MCT results, we refer the reader to Kob [23], and As mentioner earlier, MCT’s prediction of a power-law
especially Götze [36]. divergence of the relaxation time also breaks down in
most experimental and simulated glass-forming systems. More
generally, the fact that MCT always yields a power law, regardless
3.2. Failures of the molecular composition of the material, also implies that
Even though MCT successfully predicts a glass transition,
MCT has essentially no notion of the concept of fragility. At
its most notable failure is that the predicted glass transition
best, an MCT power law may correctly describe the relaxation
temperature Tc occurs at much higher temperatures than the
dynamics of fragile glass formers, but strong glass formers
true experimental value Tg . Thus, the static structure factor for
exhibit a fundamentally different, Arrhenius-type growth of the
which MCT predicts a glassy state corresponds in reality to only
relaxation time. Indeed, an accurate (first-principles) prediction
a mildy supercooled liquid. In practice, the MCT predictions
of the fragility of a material on the sole basis of its microscopic
are often rescaled such that Tc coincides with Tg [68], but even
structure remains a major open challenge in the field [20].
with such a relative comparison, MCT generally fails to accurate
Nonetheless, we note that MCT can predict other properties of
describe the dynamics in the deeply supercooled regime. This
strong glass formers rather accurately, such as the wavevector-
discrepancy is attributed to MCT’s lack of ergodicity-restoring
dependent non-ergodicity parameter in the glassy phase [45].
relaxation mechanisms that keep the experimental system in the
MCT is also generally unable to account for the breakdown
liquid phase well below Tc (Figure 8). Such mechanisms are
of the Stokes-Einstein relation in the deeply supercooled regime.
generally referred to as activated dynamics, and are commonly
This is again attributed to the inherent mean-field character
identified with particles “hopping” out of their local cages to resist
of the theory and the absence of activated hopping dynamics
freezing [82]. MCT fails to account for such hopping motion and
[82]. Moreover, in its standard formulation, MCT does not offer
thus strongly overestimates the degree of caging–a feature that is
an explanation for the emergence of dynamic heterogeneity,
believed to arise from its so-called mean-field nature. In practice,
since MCT only predicts a single F(k, t) for a given wavevector,
the predicted MCT transition at Tc is therefore interpreted as a
density, and temperature, and hence does not give access to
crossover point where the dynamics changes into an activated
correlations in the fluctuations of F(k, t). However, as discussed
form [83]. In section 4, we will return to this point and
in section 4.3, an extension of the theory does allow for the
address recent efforts to incorporate activated dynamics directly
calculation of a quantity related to the dynamic susceptibility
into the theory. We note that activated dynamics may also be
χ4 (t) and a corresponding growing (and ultimately diverging)
incorporated via, e.g., the Random First Order Transition theory,
correlation length scale. Furthermore, despite its mean-field
which is a spin-glass-inspired framework that merges MCT with
character, it was recently shown that MCT does not become
thermodynamics-based concepts [84, 85]. A description of RFOT
exact in the mean-field limit of infinite dimensions for a
system composed of hard spheres [87–89], making it difficult to
rationalize the set of standard-MCT approximations in a simple
physical manner. Moreover, MCT assumes that the material is in
(quasi-)equilibrium, and consequently fails to account for non-
equilibrium aging and protocol-dependent history effects. The
afore-mentioned SCGLE theory does allow for explicit aging
predictions [72], and hence it constitutes an attractive alternative
to MCT in this regard. Finally, since MCT (like SCGLE) is a
purely dynamical theory, it cannot make any statements about
thermodynamic properties such as the entropy. The latter is
believed to also play an important role in the process of glass
formation, and in particular may point toward an underlying
thermodynamic transition that in practice is masked by the
dynamic transition. Nonetheless, it is possible that MCT is
implicitly aware of at least some changes in thermodynamic
properties through changes in the static structure factor [90, 91].

FIGURE 8 | Typical MCT prediction (purple curve) and simulation result (blue
curve) for the dynamical slowdown of a glass-forming material as a function of
4. GOING BEYOND STANDARD
the control parameter. MCT generally predicts that the viscosity or relaxation MODE-COUPLING THEORY
time grows as a power law and diverges at the glass transition temperature Tc
or critical packing fraction φc . In reality, a material tends to remain in the Since standard MCT is not exact, as exemplified by the
supercooled-liquid phase at temperatures well below Tc (or packing fractions drawbacks and failures discussed in the previous section,
above φc ), which is attributed to so-called activated or hopping dynamics
missing in standard MCT. The figure is based on Charbonneau et al. [82].
various attempts have been made in the last few decades to
improve the theory’s predictive power for glassy dynamics.

Frontiers in Physics | www.frontiersin.org 10 October 2018 | Volume 6 | Article 97

Janssen Mode-Coupling Theory of the Glass Transition: A Primer

Below we will summarize the most notable efforts to remedy Newtonian dynamics. Hence, it appears likely that EMCT
at least some of MCT’s problems, including the formulation cannot offer a rigorous, universal remedy for the lack of
of “Extended” MCT (EMCT) and “Generalized” MCT (GMCT) ergodicity-restoring activated dynamics within the standard
to incorporate activated dynamics mechanisms, the potential MCT framework.
of GMCT to account for fragility and dynamic heterogeneity,
and the formulation of “Inhomogeneous” MCT (IMCT)
to predict dynamic susceptibilities. Finally, we also briefly
4.2. “Generalized” Mode-Coupling Theory:
discuss recent generalizations of MCT to a new class of soft Toward an Exact Equation for the Memory
condensed-matter systems referred to as active matter. Such Function
active materials are composed of particles that can undergo An alternative route to rigorously improve MCT was put
autonomous motion through the consumption of energy, and forward by Szamel in 2003 [96]. This approach, referred to
are now emerging as a new paradigm to understand collective as Generalized MCT or GMCT, seeks to systematically avoid
behavior seen in many living systems. The recent realization the second main approximation of standard MCT, i.e., the
that active particles can also vitrify into a glassy state has uncontrolled factorization of the four-point density correlations
spurred the formulation of various MCT frameworks for appearing in the memory function. To this end, a new and
active matter, the development of which will be reviewed in formally exact equation of motion is developed for the four-
section 4.4. point correlation functions themselves (again by applying the
Mori-Zwanzig projection formalism of section 2.2, this time
using the basis of pair densities ρ(k1 , t)ρ(k2 , t) as the “relevant”
4.1. “Extended” Mode-Coupling Theory: variables). The new equation is governed by another memory
Incorporating Couplings to Currents function that, to leading order, is controlled by six-point density
The first attempts to remove the spurious MCT transition at correlation functions, which in turn are dominated by eight-
Tc were proposed by Das and Mazenko in 1986 [92] and by point correlations, etc. Hence, by repeatedly developing a new
Götze and Sjögren in 1987 [93], only a few years after the equation of motion for the new memory function, a hierarchy
original formulation of standard MCT [24, 38]. Das and Mazenko of coupled equations emerges, in which the uncontrolled
employed a field-theoretic description, commonly referred to as factorization approximation may be applied at an arbitrary level
fluctuating nonlinear hydrodynamics, while Götze and Sjögren to close the set of equations. This GMCT scheme thus allows, in
used a projection-based formalism to improve the theory in the principle, for a systematic delay of the closure approximation and,
temperature regime near and below Tc . Both approaches amount notably, remains based entirely on first principles (see Figure 9).
to a perturbative treatment of nonlinear couplings to certain Szamel [96] and Wu and Cao [98] showed that GMCT
current modes that are neglected in the standard formulation hierarchies factorized at the level of six- and eight-point
of MCT, and which cut off the sharp MCT transition such that correlation functions, respectively, indeed bring the predicted
the strict divergence of the relaxation time at Tc is removed. glass transition density systematically closer to the empirical
This “rounding off ” of the MCT transition was interpreted as a value for a system of colloidal hard spheres. More recent work
mechanism for activated or hopping dynamics that would keep [97] also established that the full time-dependent microscopic
the material ergodic, i.e., in the supercooled liquid phase, below dynamics for a quasi-hard-sphere glass former is systematically
Tc . The 2004 review by Das [37] provides an extensive overview improved by GMCT. In fact, fit-parameter-free third-order
of this line of EMCT research. GMCT calculations could achieve full quantitative agreement
However, more recent theoretical studies have argued on for F(k, t) up to the moderately supercooled regime, at densities
general physical grounds that the invoked couplings to currents where standard MCT would already predict a spurious glass
in EMCT cannot provide a satisfactory explanation of activated transition [97]. Furthermore, within a simplified schematic
dynamics, since these couplings should always become negligible (wavevector-independent) GMCT model, Mayer et al. [99]
close to a glass transition [94]. Moreover, Andreanov et al. showed analytically that the sharp MCT glass transition can be
[95] suggested that the fluctuating nonlinear hydrodynamics completely removed when avoiding the closure approximation
approach employs an incorrect treatment of time-reversal altogether, i.e., when applying infinite-order GMCT. Even
symmetry. Another argument that casts doubt on the general though all GMCT studies to date still rely on several
applicability of EMCT is the fact that experimental and approximations–including the neglect of “projected” dynamics
numerical simulation studies have unambiguously established in the memory functions (section 2.3) and the factorization of
that materials obeying Newtonian and Brownian (stochastic) all static correlation functions into products of S(k)’s–, the good
dynamics exhibit the same deviations from standard-MCT agreement so far with computer simulations and experiments,
behavior, despite their differences in microscopic dynamical as well as the apparent convergent behavior of the hierarchy
details. This suggests that the physical mechanisms governing [100], suggest that GMCT offers a promising first-principles path
activated behavior below Tc have a universal origin in both toward systematic MCT improvement. In particular, it appears
molecular (Newtonian) fluids and colloidal (Brownian) systems. that higher-order GMCT captures at least some aspects of
Since the current modes introduced in EMCT cannot be activated dynamics to keep the material ergodic at temperatures
properly defined in Brownian systems [40], the proposed EMCT below Tc , consistent with empirical observations. Importantly,
mechanism may thus only apply to materials undergoing we note that GMCT is applicable to both Newtonian and

Frontiers in Physics | www.frontiersin.org 11 October 2018 | Volume 6 | Article 97

Janssen Mode-Coupling Theory of the Glass Transition: A Primer

notably different from standard MCT, which is mathematically

only capable of predicting power-law growth close to the
transition. It remains to be tested whether the fully microscopic
(wavevector-dependent) version of GMCT will indeed be able
to account for different degrees of fragility, given solely the
static structure factors S(k) (and possibly higher-order static
correlation functions) of strong and fragile materials as input. It
might be tempting to assume that, with increasing closure level,
the GMCT predictions should become more accurate, but let us
reiterate that the current formulation of GMCT still relies on
several approximations, and it is still unclear how the remaining
assumptions ultimately affect the dynamics.
Finally, we note that by construction, higher-order GMCT also
makes microscopic predictions for the (approximate) dynamics
of unfactorized four-point density correlations [97]. Although
these functions are not exactly equivalent to the dynamic
susceptibility χ4 (t), they should nonetheless be able to provide
insight into dynamic heterogeneities, since they essentially
describe particle correlations over two points in time and at least
two points in space. Hence, GMCT may also offer a suitable
starting point to study dynamic heterogeneity, as well as the
breakdown of the Stokes-Einstein relation in supercooled liquids,
from a strictly first-principles perspective. We expect this avenue
of research to be explored in the coming years.

FIGURE 9 | (A) Graphical illustration of the GMCT hierarchy. GMCT seeks to 4.3. “Inhomogeneous” Mode-Coupling
systematically avoid the uncontrolled MCT factorization approximation by
developing a new, and formally exact, equation of motion for the unknown
Theory: A Measure for Dynamic
memory function. This equation in turn is governed by a new memory function, Heterogeneity
which is controlled by another memory function, etc. Standard MCT As noted earlier, standard MCT seeks to describe the “average”
corresponds to the lowest-order (self-consistent) closure of this hierarchy. F(k, t) for a given set of wavevectors and system parameters,
(B) Microscopic GMCT predictions for F(k, t) [97] compared to numerical data
obtained from computer simulations [68] for a system of quasi-hard spheres at
but does not give immediate access to the fluctuations of
packing fraction φ = 0.570 and wavenumber kd = 7.4. These results indicate F(k, t) that are encoded in the dynamic susceptibility χ4 (t) [33].
that the GMCT hierarchy apparently converges and that the theory becomes Hence, standard MCT cannot make direct predictions about
more quantitatively accurate as the closure level is increased. The figure is dynamically heterogeneous behavior, which is generally revealed
adapted from Janssen and Reichman [97] with permission.
as a growing peak in χ4 (t). There is, however, an indirect way
to extract a dynamic susceptibility from MCT by incorporating
an external field into the theory–a framework referred to as
Inhomogeneous MCT or IMCT. The idea of IMCT is to measure
Brownian systems, and therefore also holds the potential to offer the dynamic response of the intermediate scattering function
a more universal picture of glassy dynamics. F(k, t) to changes in the external field; this response amounts
In addition to accounting for some kind of ergodicity- to a three-point dynamic density correlation function χ3 (t).
restoring processes below Tc , GMCT might also provide a The IMCT study of Biroli et al. [47] argues that the induced
suitable framework to describe fragility. The work of Mayer fluctuations by the external field are intimately related to the
et al. [99] revealed that, within their particular schematic spontaneous fluctuations described by χ4 (t), and hence the
model, infinite-order GMCT predicts an exponential growth susceptibility χ3 (t) should behave in a similar manner as the
of the relaxation time, fundamentally distinct from the fragile four-point function χ4 (t).
power-law behavior of standard MCT. In later studies, we Biroli et al. found that χ3 (t) grows upon approaching the
demonstrated that other schematic GMCT models may also dynamical MCT transition, and in fact diverges at the critical
give rise to other functional forms of relaxation-time growth, temperature Tc . Furthermore, a correlation length ξ could
ranging from fragile super-Arrhenius to strong (sub-)Arrhenius be defined–a measure, perhaps, for the size of cooperatively
behavior, depending on the choice of schematic parameters [101]. rearranging particles in the supercooled regime–, that grows as
Although these simplified GMCT models inherently lack any ξ ∼ |T − Tc |−ν with a critical exponent of ν = 1/4. Notably,
wavevector dependence, and therefore cannot make detailed IMCT also predicts that this length scale governs both the α- and
predictions for any structural glass former with a realistic β-relaxation regimes. This suggests that the traditional picture
S(k), they suggest that higher-order GMCT has at least the of caging in the β-regime, commonly interpreted as the rattling
mathematical flexibility to account for different fragilities. This is of particles in local cages formed by their nearest neighbors (see

Frontiers in Physics | www.frontiersin.org 12 October 2018 | Volume 6 | Article 97

Janssen Mode-Coupling Theory of the Glass Transition: A Primer

3.1), is actually more subtle; rather, IMCT implies that these a passive colloid, but with a higher effective diffusion constant.
cages become more and more collective as the MCT transition is This approximation was subsequently used to derive an effective
approached. However, it should be noted that the predictions of Smoluchowski operator for the collective dynamics of a dense
IMCT are not generally in quantitative agreement with empirical ensemble of active particles. In essence, this effective-diffusion
results. For example, numerical simulations for a model glass approach amounts to the removal of explicit rotational degrees of
former composed of Lennard-Jones particles indicate a growth freedom. The resulting MCT approach yields a modified version
of ξ (extracted from the numerical χ4 (t)) with a fitted exponent of Equation (12), in which both the frequency term and the
of ν ≈ 0.5, and suggest that the length scale predicted by IMCT memory function acquire an activity-dependent prefactor. The
does not necessarily describe the true size of the correlated spatial main outcome of this MCT study is that the addition of particle
domains relevant in real glass-forming materials [102]. On the activity can soften (i.e., decrease the non-ergodicity parameter)
other hand, simulations for another model glass former (the so- and eventually melt a passive glass, and shift the glass transition
called Gaussian core model, which is believed to behave more as a toward higher densities. These findings are also in qualitative
mean-field system) have revealed that the predicted IMCT scaling agreement with computer simulations of a similar active material
of χ3 (t) is in good quantitative agreement with the numerical composed of self-propelling Brownian hard particles [107, 108].
χ4 (t) [103], implying that IMCT constitutes at least in some The MCT approach of Farage and Brader was later also extended
sense a suitable mean-field framework for glassy dynamics. The by Ding et al. [116] to mixtures of active and passive particles.
question to what extent, and under which conditions, IMCT A different and more extensive active-matter study was
can offer an accurate description of dynamic heterogeneity, performed by Szamel et al. [110, 117]. Here, the authors
and how the IMCT predictions relate to, e.g., the four-point modeled active particles by an Ornstein-Uhlenbeck stochastic
dynamic correlations emerging from GMCT, still remains to be process, characterized by an effective temperature that quantifies
established. the strength of the active forces, and a persistence time that
describes the duration of persistent self-propelled motion. In
this model, particle motion is thus described as a persistent
4.4. Mode-Coupling Theories for Active random walk. Within their framework, the self-propulsion is first
Matter integrated out before applying the projection-operator method
We end this review with a very recent development in the and MCT-like approximation; this approach essentially assumes
field, namely the study of active matter. Active materials consist that particle positions evolve on a time scale much larger than
of particles that can convert energy into autonomous motion, the time scale needed for reorientation of the activity direction,
rendering them out of thermodynamic equilibrium at the single- somewhat akin to the effective-diffusion assumption of Farage
particle level [104]. Such particle activity can lead to rich and Brader [115]. An important difference between the active
self-organizing behavior, as exemplified in nature by, e.g., the MCT of Szamel et al. and previous MCT studies is that not
collective motion of living cells and the flocking of birds. During only the static structure factor–i.e., static correlations between
the last decade, numerous synthetic active systems have also particle positions–should be given as input to the theory, but
become available [105], spurring the development of theoretical also static correlations between particle velocities. Contrary to
approaches to describe the emergent behavior in these non- the behavior of ABPs, it was found that the incorporation of
equilibrium materials. In particular, it was found that dense activity can both enhance and suppress glass formation: for
active matter can also exhibit properties of supercooled liquids small persistence times, the active fluid relaxes faster than a
and vitrifying colloidal suspensions [6, 8, 10, 11, 106–114], passive system at the same effective temperature, but for large
including slow structural relaxation, dynamic heterogeneity, persistent times the active material becomes more glass-like
varying degrees of fragility, and the ultimate formation of a compared to the passive reference system. This non-monotonic
kinetically arrested, amorphous solid state. dependence of the relaxation time was observed both in the
Here we briefly discuss recent extensions of standard MCT MCT analysis and in computer simulations, and was attributed to
to describe the glassy dynamics in active materials. Since the competition between increasing velocity correlations (which
many synthetic active particles are composed of colloids speed up the dynamics) and increasing structural correlations
undergoing active Brownian motion, all active versions of (which slow down the dynamics) [110]. For sufficiently large
MCT to date are based on the Smoluchowski formalism for persistence times, it was found that the fitted MCT glass
Brownian systems, rather than the Newtonian description for transition temperature increases monotonically with increasing
molecular fluids discussed in section 2. We note, however, that persistence time, suggesting that–at least within this active-
continuum descriptions of active matter, such as those for active matter model–vitrification occurs more easily as the material
liquid crystals, are usually derived from Newtonian-based fluid becomes more active. An MCT-based scaling analysis for this
mechanics [104]. type of active-matter system was later performed by Nandi and
The first MCT approach to active glasses was presented by Gov [118].
Farage and Brader in 2014 [115]. In this work, they considered Feng and Hou [119] subsequently studied a quasi-equilibrium
so-called active Brownian particles (ABPs) that move with a thermal version of the active Ornstein-Uhlenbeck model of
constant self-propulsion speed in a random direction, subject Szamel and co-workers, which additionally accounts for thermal
to translational and rotational Brownian motion. The authors translational noise. Their MCT derivation differs from the
assumed that a single, non-interacting ABP behaves effectively as approach taken by Szamel [117], however: it is valid only

Frontiers in Physics | www.frontiersin.org 13 October 2018 | Volume 6 | Article 97

Janssen Mode-Coupling Theory of the Glass Transition: A Primer

for sufficiently small persistence times (since it relies on a sets an effective kinetic temperature; a coefficient of restitution
perturbative expansion), and does not require explicit velocity ǫ is introduced to account for dissipative particle collisions.
correlation functions to be given as input. Rather, their active- The resulting MCT equations are similar to Equation (12),
MCT dynamics is governed by an averaged diffusion constant except for an explicit ǫ-dependent prefactor in the frequency
D̄ and a non-trivial steady-state structure function S2 (k), which term and memory function. It was found that the critical
both depend on the effective temperature and density of the glass transition systematically shifts to higher densities as the
system, as well as on the persistence time of the active particles. dissipation increases (decreasing ǫ). Furthermore, the increasing
The coefficient D̄ and S2 (k) should both be given as additional dissipation was found to have three noticeable effects in the non-
input to the theory in order to predict F(k, t). It was found that ergodicity parameter at and above the glass transition density: i)
the critical density at which the glass transition takes place shifts correlations at small wave numbers are enhanced, ii) oscillations
to larger values with increasing magnitude of the self-propulsion reflecting the local structure become less pronounced, and iii)
force or effective temperature, and that the critical effective glass the localization length (measured by the inverse of the width of
temperature increases with the persistence time. In the limit the non-ergodicity-parameter peak) decreases. The last finding
of a vanishing persistence time, the theory naturally yields the is a consequence of the glass transition taking place at a higher
expected result for a simple passive Brownian system [119]. density.
Very recently, Liluashvili, et al. [120] formulated the first
MCT for ABPs in which both the translational and rotational
degrees of freedom are treated on an equal footing. That is, 5. CONCLUSIONS AND OUTLOOK
rather than seeking to reduce the active material to a near-
equilibrium system, the rotational degrees of freedom governing This review has sought to provide a brief overview of the
the reorientation of the active forces are now explicitly coupled to main phenomenology of glassy dynamics, and of its theoretical
the translational motion. This approach thus avoids the effective- description using Mode-Coupling Theory–arguably the most
diffusion assumption (which in principle may be valid only at low successful theory of the glass transition that is based entirely
densities and sufficiently long times), and the resulting dynamics on first principles. We have focused mainly on the behavior
now also depends non-trivially on the rotational diffusion of the density correlation function F(k, t) as a probe of the
constant. The only required material-dependent input for this microscopic dynamics associated with vitrification. In the normal
active MCT is the passive-equilibrium static structure factor. An liquid phase, this correlation function rapidly decays to zero,
important outcome of this study is the three-dimensional fluid- but at the glass transition it fails to decay on any practical
glass phase diagram for hard ABPs as a function of packing time scale, marking the onset of rigidity and providing an
fraction, self-propulsion speed, and rotational diffusion constant. order parameter for the transition. Upon approaching the
It was shown that this surface cannot be collapsed onto a single glass transition temperature, several complex features become
line in the two-dimensional plane, highlighting the importance visible in the dynamics, such as a transient plateau and
of treating the rotational degrees of freedom explicitly. Indeed, stretched exponential behavior in F(k, t), a breakdown of
depending on the density of the active material, separate regimes the Stokes-Einstein relation, and the emergence of dynamical
could be identified that are dominated either by translational or heterogeneity–the latter being associated with increasingly large
reorientational motion. As in the study of Farage and Brader fluctuations in F(k, t). Remarkably, during the process of
[115], and in agreement with computer simulations [107, 108], it glass formation, the microscopic structure of the material, as
was also found that activity generally makes hard-sphere systems probed by e.g., the radial distribution function g(r) or static
more fluid-like and consequently shifts the glass transition to structure factor S(k), undergoes only very minor changes,
higher packing fractions. Notably, this active fluidization effect yet the viscosity and dynamic relaxation time increase by
grows monotonously with increasing persistence time or inverse many orders of magnitude. It is this seemingly paradoxical
rotational diffusion constant, in contrast with the findings of discrepancy between structure and dynamics that makes the
Szamel and co-workers [110]. This difference is attributed to the glass transition a notoriously difficult problem in theoretical
absence of thermal Brownian noise in the model of Szamel et al.; physics.
in the limit of infinitely large persistence (vanishing rotational MCT offers a first-principles-based framework to account for
diffusion), active particles can block themselves and produce a at least some aspects of glassy dynamics. Its starting point is the
glassy state, while the finite thermal diffusive motion in ABPs will exact equation of motion for F(k, t); through a series of (partly
make such blocking ineffective [120]. uncontrolled) approximations, MCT subsequently provides a
Finally, we mention another class of non-equilibrium self-consistent equation for F(k, t) that can be solved numerically
materials that is closely related to active fluids, namely driven using only the static structure factor as input. As such, the
granular matter. Such systems can be realized experimentally by theory makes a set of detailed predictions for the full microscopic
placing granular particles on, e.g., an air-fluidized or vibrating relaxation dynamics of a glass-forming material as a function of
bed. Kranz, Sperl, and Zippelius [121–123] developed an MCT time, wavevector, temperature, and density, on the sole basis of
for driven granular spheres, focusing on the role of energy simple structural information. Among its notable successes is the
dissipation (due to inelastic particle collisions) on the dynamics. qualitative prediction of a glass transition, a physically intuitive
In their work, the “activity” is modeled by a driving amplitude picture for glass formation in terms of the cage effect, and the
that gives rise to random particle kicks and that implicitly correct prediction of several highly non-trivial scaling behaviors

Frontiers in Physics | www.frontiersin.org 14 October 2018 | Volume 6 | Article 97

Janssen Mode-Coupling Theory of the Glass Transition: A Primer

in F(k, t). However, MCT is generally not quantitatively accurate, it can be expected that active MCT will also contribute to
and cannot account properly for the concept of fragility, the our understanding of disordered active and living materials
violation of the Stokes-Einstein relation, and the emergence of from a statistical-physics-based and purely first-principles
dynamic heterogeneity. perspective.
The shortcomings of MCT might be remedied using (first- In conclusion, despite the fact that Mode-Coupling
principles-based) extensions of the theory, such as Generalized Theory is not exact, it does provide a suitable–and in some
MCT and Inhomogeneous MCT. The first studies in this cases remarkably accurate–foundation for the study of
direction show that GMCT can indeed offer a more quantitative glassy dynamics in amorphous materials. The theory also
description of the F(k, t) dynamics and can potentially describe offers ample opportunity for new research aimed toward
fragility, while IMCT offers a framework to qualitatively account a complete and ultimately rigorously exact description of
for dynamic heterogeneity. However, GMCT still relies on the glass transition, as well as for the study of emergent
several approximations such as the neglect of certain wavevector- new classes of materials such as active matter. We expect
dependent density correlations, and IMCT provides–just like future work to be directed toward these exciting avenues of
standard MCT–only a mean-field description of glassy dynamics. research.
Hence, more work will be needed to establish how successful
these theoretical approaches are in ultimately achieving a fully AUTHOR CONTRIBUTIONS
correct first-principles description of glassy dynamics.
A more recent addition to the palette of Mode-Coupling The author confirms being the sole contributor of this work and
theories involves the study of non-equilibrium active matter. approved it for publication.
In the last few years, several MCT frameworks have been
developed to describe glassy dynamics in active materials that ACKNOWLEDGMENTS
are composed of self-propelled particles. Not only can these
theories offer new insight into the behavior of dense assemblies It is a pleasure to thank David Reichman, Grzegorz Szamel,
of synthetic active colloids, but they might also shed new Jürgen Horbach, Thomas Voigtmann, Hartmut Löwen, Matthias
light on glassy and jamming phenomena in living cell tissues. Fuchs, Jörg Baschnagel, Jean Farago, Atsushi Ikeda, Peter
Similar to how standard MCT has shaped our understanding Mayer, and Till Kranz for many interesting and enlightening
of passive glass-forming materials over the last few decades, discussions.

REFERENCES 12. Debenedetti PG, Stanley HE. Supercooled and glassy water. Phys Today
(2003) 56:40. doi: 10.1063/1.1595053
1. Debenedetti P, Stillinger F. Supercooled liquids and the glass transition. 13. So much more to know. Science (2005) 309:78–102. doi: 10.1126/science.309.
Nature (2001) 410:259–67. doi: 10.1038/35065704 5731.78b
2. Berthier L, Biroli G. Theoretical perspective on the glass transition and 14. Anderson P. Through the glass lightly. Science (1995) 267:1615.
amorphous materials. Rev Mod Phys. (2011) 83:587–645. doi: 10.1103/ 15. Royall CP, Williams SR. The role of local structure in dynamical arrest. Phys
RevModPhys.83.587 Rep. (2015) 560:1–75. doi: 10.1016/j.physrep.2014.11.004
3. Biroli G, Garrahan JP. Perspective: the glass transition. J Chem Phys. (2013) 16. Karmakar S, Dasgupta C, Sastry S. Length scales in glass-forming liquids
138:12A301. doi: 10.1063/1.4795539 and related systems: a review. Rep Prog Phys. (2016) 79:016601. doi: 10.1088/
4. Zanotto ED. Do cathedral glasses flow? Am J Phys. (1998) 66:392. doi: 10. 0034-4885/79/1/016601
1119/1.19026 17. Albert S, Bauer T, Michl M, Biroli G, Bouchaud J-P, Loidl A, et al.
5. Zhou EH, Trepat X, Park CY, Lenormand G, Oliver MN, Mijailovich SM, Fifth-order susceptibility unveils growth of thermodynamic amorphous
et al. Universal behavior of the osmotically compressed cell and its analogy order in glass-formers. Science (2016) 352:1308–11. doi: 10.1126/science.aa
to the colloidal glass transition. Proc Natl Acad Sci USA. (2009) 106:10632–7. f3182
doi: 10.1073/pnas.0901462106 18. Angell C A. Formation of glasses from liquids and biopolymers. Science
6. Angelini TE, Hannezo E, Trepat X, Marquez M, Fredberg JJ, Weitz DA. (1995) 267:1924–35. doi: 10.1126/science.267.5206.1924
Glass-like dynamics of collective cell migration. Proc Natl Acad Sci USA. 19. Xia X, Wolynes PG. Fragilities of liquids predicted from the random first
(2011) 108:4714–9. doi: 10.1073/pnas.1010059108 order transition theory of glasses. Proc Natl Acad Sci USA. (2000) 97:2990–4.
7. Schötz E-M, Lanio M, Talbot JA, Manning ML. Glassy dynamics in three- doi: 10.1073/PNAS.97.7.2990
dimensional embryonic tissues. J R Soc Interface (2013) 10:20130726 20. Tarjus G. An overview of the theories of the glass transition. In: Berthier
8. Sadati M, Nourhani A, Fredberg JJ, Taheri Qazvini N. Glass-like dynamics in L, Biroli G, Bouchaud JP, Cipelletti L, van Saarloos W, editors. Dynamical
the cell and in cellular collectives. Wiley Interdiscip Rev Syst Biol Med. (2014) Heterogeneities in Glasses, Colloids, and Granular Media. Oxford: Oxford
6:137–49. doi: 10.1002/wsbm.1258 University Press (2011). p. 39–67.
9. Parry BR, Surovtsev IV, Cabeen MT, O’Hern CS, Dufresne ER, Jacobs- 21. Hansen J-P, McDonald IR. Theory of Simple Liquids, Amsterdam: Elsevier
Wagner C. The bacterial cytoplasm has glass-like properties and is fluidized (2013).
by metabolic activity. Cell (2014) 156:183–94. doi: 10.1016/j.cell.2013.11.028 22. Reichman DR, Charbonneau P. Mode-coupling theory. J Stat Mech Theor
10. Bi D, Lopez JH, Schwarz JM, Manning ML. A density-independent rigidity Exp. (2005) 2005:P05013. doi: 10.1088/1742-5468/2005/05/P05013
transition in biological tissues. Nat Phys. (2015) 11:1074. doi: 10.1038/ 23. Kob W. Supercooled liquids, the glass transition, and computer simulations.
NPHYS3471 In: Barrat J-L, Feigelman MV, Kurchan J, Dalibard J. editors. Les Houches
11. Bi D, Yang X, Marchetti MC, Manning ML. Motility-driven glass and 2002 Summer School Session LXXVII: Slow Relaxations Nonequilibrium
jamming transitions in biological tissues. Phys Rev X (2016) 6:021011. Dynamics in Condensed Matter. Berlin: Springer-Verlag (2002).
doi: 10.1103/PhysRevX.6.021011 p. 199–269.

Frontiers in Physics | www.frontiersin.org 15 October 2018 | Volume 6 | Article 97

Janssen Mode-Coupling Theory of the Glass Transition: A Primer

24. Bengtzelius U, Götze W, Sjölander A. Dynamics of supercooled liquids and 51. Schofield J, Lim R, Oppenheim I. Mode coupling and generalized
the glass transition. J Phys C Solid State Phys. (1984) 17:5915–34. doi: 10. hydrodynamics. Physica A (1992) 89:181.
1088/0022-3719/17/33/005 52. van Zon R, Schofield J. Mode-coupling theory for multiple-point and
25. van Megen W, Underwood SM, Pusey PN. Nonergodicity parameters of multiple-time correlation functions. Phys Rev E (2001) 65:011106. doi: 10.
colloidal glasses. Phys Rev Lett. (1991) 67:1586–9. doi: 10.1103/PhysRevLett. 1103/PhysRevE.65.011106
67.1586 53. Zaccarelli E, Foffi G, Sciortino F, Tartaglia P, Dawson KA. Gaussian density
26. Weeks ER, Crocker JC, Levitt AC, Schofield A, Weitz DA. Three- fluctuations and mode coupling theory for supercooled liquids. EPL (2001)
dimensional direct imaging of structural relaxation near the colloidal glass 55:157–63. doi: 10.1209/epl/i2001-00395-x
transition. Science (2000) 287:5453. 54. Jackson HW, Feenberg E. Energy spectrum of elementary excitations in
27. Kegel WK, van Blaaderen A. Direct observation of dynamical heterogeneities Helium II. Rev Mod Phys (1962) 34:686–93. doi: 10.1103/RevModPhys.34.
in colloidal hard-sphere suspensions. Science (2000) 287:290–3. doi: 10.1126/ 686
science.287.5451.290 55. Berthier L, Tarjus G. Nonperturbative effect of attractive forces in viscous
28. Ediger MD. Spatially heterogeneous dynamics in supercooled liquids. Annu liquids. Phys Rev Lett. (2009) 103:170601. doi: 10.1103/PhysRevLett.103.
Rev Phys Chem. (2000) 51:99–128. doi: 10.1146/annurev.physchem.51.1.99 170601
29. Berthier L. Dynamic heterogeneity in amorphous materials. Physics (2011) 56. Fuchs M, Götze W, Hofacker I, Latz A. Comments on the alpha-peak
4:7. doi: 10.1103/Physics.4.42 shapes for relaxation in supercooled liquids. J Phys Condens Matter (1991)
30. Tarjus G, Kivelson D. Breakdown of the Stokes–Einstein relation in 3:5047–71. doi: 10.1088/0953-8984/3/26/022
supercooled liquids. J Chem Phys. (1995) 103:3071–3. doi: 10.1063/1.470495 57. Flenner E, Szamel G. Relaxation in a glassy binary mixture: comparison of
31. Shi Z, Debenedetti PG, Stillinger FH. Relaxation processes in liquids: the mode-coupling theory to a Brownian dynamics simulation. Phys Rev E
variations on a theme by Stokes and Einstein. J Chem Phys. (2013) (2005) 72:031508. doi: 10.1103/PhysRevE.72.031508
138:12A526. doi: 10.1063/1.4775741 58. Szamel G, Löwen H. Mode-coupling theory of the glass transition in
32. Lačević N, Starr FW, Schrøder TB, Glotzer SC. Spatially heterogeneous colloidal systems. Phys Rev A (1991) 44:8215–9. doi: 10.1103/PhysRevA.4
dynamics investigated via a time-dependent four-point density correlation 4.8215
function. J Chem Phys. (2003) 119:7372–87. doi: 10.1063/1.1605094 59. Chong S-H, Aichele M, Meyer H, Fuchs M, Baschnagel J. Structural and
33. Biroli G, Bouchaud J-P. Diverging length scale and upper critical dimension conformational dynamics of supercooled polymer melts: insights from first-
in the Mode-Coupling Theory of the glass transition. Europhys Lett. (2004) principles theory and simulations. Phys Rev E (2007) 76:051806. doi: 10.1103/
67:21–27. doi: 10.1209/epl/i2004-10044-6 PhysRevE.76.051806
34. Struik CEL. Physical Aging in Amorphous Polymers and Other Materials. 60. Nägele G, Bergenholtz J. Linear viscoelasticity of colloidal mixtures. J Chem
Ph.D. thesis, Technische Hogeschool Delft (1977). Phys. (1998) 108:9893. doi: 10.1063/1.476428
35. Berthier L, Biroli G. A statistical mechanics perspective on glasses and aging. 61. Fuchs M, Cates ME. A mode coupling theory for Brownian
In: Meyers RE, editor. Encyclopedia of Complexity and Systems Science. particles in homogeneous steady shear flow. J Rheol. (2009) 53:957.
(Springer) (2009). p. 4209–4240. Available online at: https://www.springer. doi: 10.1122/1.3119084
com/gp/book/9780387758886 62. Lang S, Boan V, Oettel M, Hajnal D, Franosch T, Schilling R. Glass transition
36. Götze, W. Complex dynamics of glass-forming liquids: A Mode-Coupling in confined geometry. Phys Rev Lett. (2010) 105:125701. doi: 10.1103/
Theory. Oxford, UK: Oxford University Press (2009). PhysRevLett.105.125701
37. Das SP. (2004). Mode-coupling theory and the glass transition in supercooled 63. Mandal S, Lang S, Gross M, Oettel M, Raabe D, Franosch T, et al. Multiple
liquids. Rev Mod Phys. 76:785. doi: 10.1103/RevModPhys.76.785 reentrant glass transitions in confined hard-sphere glasses. Nat Commun.
38. Leutheusser E. Dynamical model of the liquid-glass transition. Phys Rev A (2014) 5:4435. doi: 10.1038/ncomms5435
(1984) 29:2765. 64. Franosch T, Lang S, Schilling R. Fluids in extreme confinement. Phys Rev
39. Yeomans-Reyna L, Chávez-Rojo MA, Ramírez-González PE, Juárez- Lett. (2012) 109:240601. doi: 10.1103/PhysRevLett.109.240601
Maldonado R, Chávez-Páez M, Medina-Noyola M. Dynamic arrest within 65. Gazuz I, Puertas AM, Voigtmann T, Fuchs M. Active and nonlinear
the self-consistent generalized Langevin equation of colloid dynamics. Phys microrheology in dense colloidal suspensions. Phys Rev Lett. (2009)
Rev E (2007) 76:041504. doi: 10.1103/PhysRevE.76.041504 102:248302. doi: 10.1103/PhysRevLett.102.248302
40. Szamel G. Mode-coupling theory and beyond: a diagrammatic approach 66. Puertas AM, Voigtmann T. Microrheology of colloidal systems. J Phys
(2013) 2013:012J01. doi: 10.1093/ptep/pts036 Condens Matter (2014) 26:243101. doi: 10.1088/0953-8984/26/24/243101
41. Schilling R. Theories of the structural glass transition. In: Radons G, Just W, 67. Schilling R, Scheidsteger T. Mode coupling approach to the ideal glass
Häussler P, editors. Collective Dynamics of Nonlinear and Disordered Systems transition of molecular liquids: linear molecules. Phys Rev E (1997) 56:2932–
Heidelberg: Springer (2005). p. 171–202. 49. doi: 10.1103/PhysRevE.56.2932
42. Langer JS. Theories of glass formation and the glass transition. Rep Prog Phys. 68. Weysser F, Puertas AM, Fuchs M, Voigtmann T. Structural relaxation of
(2014) 77:042501. doi: 10.1088/0034-4885/77/4/042501 polydisperse hard spheres: comparison of the mode-coupling theory to a
43. Lubchenko V, Wolynes PG. Theory of structural glasses and supercooled Langevin dynamics simulation. Phys Rev E (2010) 82:011504. doi: 10.1103/
liquids. Annu Rev Phys Chem. (2007) 58:235–66. doi: 10.1146/annurev. PhysRevE.82.011504
physchem.58.032806.104653 69. Ramírez-González P, Medina-Noyola M. General nonequilibrium theory of
44. Zhang K. On the concept of static structure factor. arXiv:1606.03610 (2016). colloid dynamics. Phys Rev E (2010) 82:061503. doi: 10.1103/PhysRevE.82.
45. Sciortino F, Kob W. Debye-waller factor of liquid silica: theory and 061503
simulation. Phys Rev Lett. (2001) 86:648–51. doi: 10.1103/PhysRevLett.86. 70. Ramírez-González P, Medina-Noyola M. Aging of a homogeneously
648 quenched colloidal glass-forming liquid. Phys Rev E (2010) 82:061504.
46. Coslovich D. Static triplet correlations in glass-forming liquids: A molecular doi: 10.1103/PhysRevE.82.061504
dynamics study. J Chem Phys. (2013) 138:12A539. doi: 10.1063/1.4773355 71. Sánchez-Díaz LE, Ramírez-González P, Medina-Noyola M. Equilibration
47. Biroli G, Bouchaud J-P, Miyazaki K, Reichman DR. Inhomogeneous Mode- and aging of dense soft-sphere glass-forming liquids. Phys Rev E (2013)
Coupling Theory and Growing Dynamic Length in Supercooled Liquids. 87:052306. doi: 10.1103/PhysRevE.87.052306
Phys Rev Lett. (2006) 97:195701. doi: 10.1103/PhysRevLett.97.195701 72. Mendoza-Méndez P, Lázaro-Lázaro E, Sánchez-Díaz LE, Ramírez-González
48. Zwanzig R. Memory effects in irreversible thermodynamics. Phys Rev. (1961) PE, Pérez-Ángel G, Medina-Noyola M. Crossover from equilibration
124:983–992. doi: 10.1103/PhysRev.124.983 to aging: nonequilibrium theory versus simulations. Phys Rev E (2017)
49. Mori H. Transport, collective motion, and Brownian Motion. Prog Theor 96:022608. doi: 10.1103/PhysRevE.96.022608
Phys. (1965) 33:423–55. doi: 10.1143/PTP.33.423 73. Latz A. Non-equilibrium mode-coupling theory for supercooled liquids and
50. Nägele G. On the dynamics and structure of charge-stabilized suspensions. glasses. J Phys Condens Matter (2000) 12:6353–63. doi: 10.1088/0953-8984/
Phys Rep. (1996) 272:215–372. doi: 10.1016/0370-1573(95)00078-X 12/29/307

Frontiers in Physics | www.frontiersin.org 16 October 2018 | Volume 6 | Article 97

Janssen Mode-Coupling Theory of the Glass Transition: A Primer

74. Juárez-Maldonado R, Medina-Noyola M. Theory of dynamic arrest in 96. Szamel G. Colloidal Glass Transition: beyond mode-coupling theory. Phys
colloidal mixtures. Phys Rev E (2008) 77:051503. doi: 10.1103/PhysRevE.77. Rev Lett. (2003) 90:228301. doi: 10.1103/PhysRevLett.90.228301
051503 97. Janssen LMC, Reichman DR. Microscopic dynamics of supercooled liquids
75. Sánchez-Díaz LE, Lázaro-Lázaro E, Olais-Govea JM, Medina-Noyola M. from first principles. Phys Rev Lett. (2015) 115:205701. doi: 10.1103/
Non-equilibrium dynamics of glass-forming liquid mixtures. J Chem Phys. PhysRevLett.115.205701
(2014) 140:234501. doi: 10.1063/1.4882356 98. Wu J, Cao J. High-order mode-coupling theory for the colloidal glass
76. Elizondo-Aguilera LF, Zubieta Rico PF, Ruiz-Estrada H, Alarcón-Waess transition. Phys Rev Lett. (2005) 95:078301. doi: 10.1103/PhysRevLett.95.
O. Self-consistent generalized Langevin-equation theory for liquids of 078301
nonspherically interacting particles. Phys Rev E(2014) 90:052301. doi: 10. 99. Mayer P, Miyazaki K, Reichman, DR. Cooperativity beyond caging:
1103/PhysRevE.90.052301 generalized mode-coupling theory. Phys Rev Lett. (2006) 97:095702. doi: 10.
77. Cortés-Morales EC, Elizondo-Aguilera LF, Medina-Noyola M. Equilibration 1103/PhysRevLett.97.095702
and aging of liquids of non-spherically interacting particles. J Phys Chem B 100. Janssen LMC, Mayer P, Reichman DR. Generalized mode-coupling theory of
(2016) 120:7975. doi: 10.1021/ACS.JPCB.6B04635 the glass transition: schematic results at finite and infinite order. J Stat Mech
78. Kob, W, Nauroth, M, Sciortino, F. Quantitative tests of mode-coupling Theor Exp. (2016) 2016:054049. doi: 10.1088/1742-5468/2016/05/054049
theory for fragile and strong glass formers. J Non Cryst Solids (2002) 307– 101. Janssen LMC, Mayer P, Reichman DR. Relaxation patterns in supercooled
310:181–7. doi: 10.1016/S0022-3093(02)01457-6 liquids from generalized mode-coupling theory. Phys Rev E (2014)
79. Brambilla G, El Masri D, Pierno M, Berthier L, Cipelletti L, Petekidis G, et al. 90:052306. doi: 10.1103/PhysRevE.90.052306
Probing the equilibrium dynamics of colloidal hard spheres above the mode- 102. Karmakar S, Dasgupta C, Sastry S. Growing length and time scales in glass-
coupling glass transition. Phys Rev Lett. (2009) 102:085703. doi: 10.1103/ forming liquids. Proc Natl Acad Sci USA. (2009) 106:3675. doi: 10.1073/pnas.
PhysRevLett.102.085703 0811082106
80. Pham KN, Puertas AM, Bergenholtz J, Egelhaaf SU, Moussad A, Pusey 103. Coslovich D, Ikeda A, Miyazaki K. Mean-field dynamic criticality and
PN, et al. Multiple glassy states in a simple model system. Science (2002) geometric transition in the Gaussian core model. Phys Rev E (2016)
296:104–6. doi: 10.1126/science.1068238 93:042602. doi: 10.1103/PhysRevE.93.042602
81. Berthier L, Moreno AJ, Szamel G. Increasing the density melts ultrasoft 104. Marchetti MC, Joanny JF, Ramaswamy S, Liverpool TB, Prost J, Rao M, et al.
colloidal glasses. Phys Rev E (2010) 82:060501. doi: 10.1103/PhysRevE.82. Hydrodynamics of soft active matter. Rev Mod Phys. (2013) 85:1143–89.
060501 doi: 10.1103/RevModPhys.85.1143
82. Charbonneau P, Jin Y, Parisi G, Zamponi F. (2014). Hopping and the Stokes– 105. Bechinger C, Di Leonardo R, Löwen H, Reichhardt C, Volpe G, Volpe G.
Einstein relation breakdown in simple glass formers. Proc Natl Acad Sci USA. Active brownian particles in complex and crowded environments. Rev Mod
111:15025. doi: 10.1073/pnas.1417182111 Phys. (2016) 88:045006
83. Biroli G, Bouchaud J-P. The random first-order transition theory of glasses: 106. Henkes S, Fily Y, Marchetti MC. Active jamming: self-propelled soft particles
a critical assessment. arXiv:0912.2542 (2009). at high density. Phys Rev E (2011) 84:040301. doi: 10.1103/PhysRevE.84.
84. Kirkpatrick TR, Thirumalai D. Dynamics of the structural 040301
glass transition and the p-Spin-Interaction Spin-Glass Model. 107. Ni R, Cohen Stuart MA, Dijkstra M. Pushing the glass transition towards
Phys Rev Lett. (1987) 58:2091–2094. doi: 10.1103/PhysRevLett.5 random close packing using self-propelled hard spheres. Nat Commun.
8.2091 (2013) 4:789–845. doi: 10.1038/ncomms3704
85. Kirkpatrick TR, Wolynes PG. Connections between some kinetic and 108. Berthier L. Nonequilibrium glassy dynamics of self-propelled hard disks.
equilibrium theories of the glass transition. Phys Rev A (1987) 35:3072–80. Phys Rev Lett. (2014) 112:220602. doi: 10.1103/PhysRevLett.112.220602
doi: 10.1103/PhysRevA.35.3072 109. Pilkiewicz KR, Eaves JD. Reentrance in an active glass mixture. Soft Matter
86. Kirkpatrick TR, Thirumalai D. Random First Order Theory concepts in (2014) 10:7495–501. doi: 10.1039/C4SM01177E
Biology and Condensed Matter physics. arXiv:1412.5017 (2014). 110. Szamel G, Flenner E, Berthier L. Glassy dynamics of athermal self-propelled
87. Ikeda A, Miyazaki K. Mode-coupling theory as a mean-field description particles: computer simulations and a nonequilibrium microscopic theory.
of the Glass transition. Phys Rev Lett. (2010) 104:255704. doi: 10.1103/ Phys Rev E (2015) 91:062304. doi: 10.1103/PhysRevE.91.062304
PhysRevLett.104.255704 111. Delarue M, Hartung J, Schreck C, Gniewek P, Hu L, Herminghaus S, et al.
88. Schmid B, Schilling R. Glass transition of hard spheres in high dimensions. Self-driven jamming in growing microbial populations. Nat. Phys. (2016)
Phys Rev E (2010) 81:041502. doi: 10.1103/PhysRevE.81.041502 12:762–6. doi: 10.1038/nphys3741
89. Maimbourg T, Kurchan J, Zamponi F. Solution of the dynamics of liquids in 112. Yazdi A, Sperl M. Glassy dynamics of Brownian particles with velocity-
the large-dimensional limit. Phys Rev Lett. (2016) 116:015902. doi: 10.1103/ dependent friction. Phys Rev E (2016) 94:032602. doi: 10.1103/PhysRevE.94.
PhysRevLett.116.015902 032602
90. Banerjee A, Sengupta S, Sastry S, Bhattacharyya SM. Role of structure 113. Berthier L, Flenner E, Szamel G. How active forces influence
and entropy in determining differences in dynamics for glass formers with nonequilibrium glass transitions. New J Phys. (2017) 19:125006.
different interaction potentials. Phys Rev Lett. (2014) 113:225701. doi: 10. doi: 10.1088/1367-2630/aa914e
1103/PhysRevLett.113.225701 114. Janssen LMC, Kaiser A, Löwen H. Aging and rejuvenation of
91. Nandi MK, Banerjee A, Sengupta S, Sastry S, Bhattacharyya SM. Unraveling active matter under topological constraints. Sci Rep. (2017) 7:5667.
the success and failure of mode coupling theory from consideration of doi: 10.1038/s41598-017-05569-6
entropy. J Chem Phys (2015) 143:174504. doi: 10.1063/1.4934986 115. Farage TFF, Brader JM. Dynamics and rheology of active glasses.
92. Das SP, Mazenko GF. Fluctuating nonlinear hydrodynamics and the liquid- arXiv:1403.0928 (2014).
glass transition. Phys Rev A (1986) 34:2265–82. doi: 10.1103/PhysRevA.34. 116. Ding H, Feng M, Jiang H, Hou Z. Nonequilibrium glass transition in
2265 mixtures of active-passive particles. arXiv:1506.02754 (2015).
93. Götze W, Sjögren L. The glass transition singularity. Z Phys B (1987) 117. Szamel G. Theory for the dynamics of dense systems of athermal self-
65:415–27. doi: 10.1007/BF01303763 propelled particles. Phys Rev E (2016) 93:012603. doi: 10.1103/PhysRevE.93.
94. Cates ME, Ramaswamy S. Do current-density nonlinearities cut off the glass 012603
transition? Phys Rev Lett. (2006) 96:135701. doi: 10.1103/PhysRevLett.96. 118. Nandi SK, Gov NS. Nonequilibrium mode-coupling theory for dense
135701 active systems of self-propelled particles. Soft Matter (2017) 13:7609.
95. Andreanov A, Biroli G, and Lefèvre A. Dynamical field theory for glass- doi: 10.1039/c7sm01648d
forming liquids, self-consistent resummations and time-reversal symmetry. 119. Feng M, Hou Z. Mode coupling theory for nonequilibrium glassy
J Stat Mech (2006) 2006:P07008. doi: 10.1088/1742-5468/2006/07/P dynamics of thermal self-propelled particles. Soft Matter (2017) 13:4464.
07008 doi: 10.1039/c7sm00852j

Frontiers in Physics | www.frontiersin.org 17 October 2018 | Volume 6 | Article 97

Janssen Mode-Coupling Theory of the Glass Transition: A Primer

120. Liluashvili A, Onody J, Voigtmann T. Mode-coupling theory for active Conflict of Interest Statement: The author declares that the research was
Brownian Particles. Phys Rev E (2017) 96:062608. conducted in the absence of any commercial or financial relationships that could
121. Kranz WT, Sperl M, Zippelius A. Glass transition for driven granular be construed as a potential conflict of interest.
fluids. Phys Rev Lett. (2010) 104:225701. doi: 10.1103/PhysRevLett.104.2
25701 Copyright © 2018 Janssen. This is an open-access article distributed under the terms
122. Sperl M, Kranz WT, Zippelius A. Single-particle dynamics in dense granular of the Creative Commons Attribution License (CC BY). The use, distribution or
fluids under driving. EPL (2012) 98:28001. reproduction in other forums is permitted, provided the original author(s) and the
123. Kranz WT, Sperl M, Zippelius A. Glass transition in driven granular fluids: copyright owner(s) are credited and that the original publication in this journal
a mode-coupling approach. Phys Rev E (2013) 87:022207. doi: 10.1103/ is cited, in accordance with accepted academic practice. No use, distribution or
PhysRevE.87.022207 reproduction is permitted which does not comply with these terms.

Frontiers in Physics | www.frontiersin.org 18 October 2018 | Volume 6 | Article 97