Critical Point Thermodynamics

Hello! Welcome to the fifth and final lesson of our first module on the thermodynamics of phase transitions.

In our last lesson, we established a crucial link between macroscopic, measurable response functions (like specific heat and susceptibility) and the microscopic world of statistical mechanics. We saw that these functions are second derivatives of the free energy and, through the fluctuation-dissipation theorem, are directly proportional to the magnitude of equilibrium fluctuations. We concluded by noting that these functions exhibit dramatic, singular behavior near a critical point, a clear sign that something profound is happening with the system's fluctuations.

This lesson delves into that singular behavior. We will develop a precise language to describe how various thermodynamic quantities behave as a system approaches a critical point. This language is that of critical exponents, a set of numbers that characterize the power-law divergences and scaling behaviors that are the universal signature of criticality. Understanding these exponents is the first step toward the modern theory of phase transitions.

The Language of Criticality: Critical Exponents

To precisely describe the behavior "near" a critical point, we first need a dimensionless measure of how close we are to the critical temperature, $T_c$. We define the reduced temperature $t$ as:
$t=\frac{T-T_c}{T_c}$
As we approach the critical point, $t\to 0$. The central idea of critical phenomena is that the singular (non-analytic) parts of thermodynamic functions scale as power laws of $t$. The exponents in these power laws are the critical exponents.

Let's define the most common set of these exponents.

As you've just read, we can summarize the key "static" critical exponents as follows. (Note: we use the magnetic system notation $m,h$ for the order parameter and conjugate field, but the concepts apply equally to liquid-gas systems with ${\rho }_L-{\rho }_G$ and $p-p_c$).

• Specific Heat ($C_V$): Describes how the specific heat diverges or shows a cusp.
  $C_V\sim |t{|}^{-\alpha }$
  A positive $\alpha$ indicates a divergence. A special case is $\alpha =0$, which can correspond to a logarithmic divergence (like $C_V\sim \ln |t|$) or a finite cusp.

• Order Parameter ($m$): Describes how the order parameter vanishes as $T$ approaches $T_c$ from below ($t<0$).
  $m\sim (-t{)}^{\beta }(\text{for }t<0,h=0)$

• Susceptibility (${\chi }_T$): Describes how the response function (e.g., magnetic susceptibility or isothermal compressibility) diverges as $T$ approaches $T_c$, typically from above ($t>0$).
  ${\chi }_T\sim t^{-\gamma }(\text{for }t>0,h=0)$

• Critical Isotherm ($m(h)$): Describes the non-linear relationship between the order parameter and the external field exactly at the critical temperature.
  $h\sim |m{|}^{\delta }\mathrm{sgn}(m)(\text{for }t=0)$

Correlations and their Exponents

The divergence of response functions is a macroscopic manifestation of something happening at the microscopic level: fluctuations are becoming correlated over increasingly large distances.

We quantify this with the correlation length, $\xi$. Away from the critical point, the correlation function $G(r)$ between fluctuations at two points separated by a distance $r$ typically decays exponentially, $G(r)\sim e^{-r/\xi }$. As we approach the critical point, $\xi$ diverges, signifying the emergence of long-range order.

• Correlation Length ($\xi$): Describes the divergence of the characteristic length scale of fluctuations.
  $\xi \sim |t{|}^{-\nu }$

Exactly at the critical point, $t=0$, the correlation length is infinite. The system is scale-invariant, meaning it "looks the same" at all length scales. At this point, the exponential decay of correlations is replaced by a much slower power-law decay.

• Correlation Function ($G(r)$): Describes the power-law decay of correlations at the critical point itself.
  $G(r)\sim \frac{1}{r^{d-2+\eta }}(\text{for }t=0,r\to \infty )$
  Here, $d$ is the spatial dimension. The $d-2$ term is motivated by the simplest field theories (Gaussian approximation), so the exponent $\eta$ (the "anomalous dimension") measures the deviation from this simple behavior.

We now have a full set of exponents: $\alpha ,\beta ,\gamma ,\delta ,\nu ,\eta$. These numbers provide a complete quantitative description of the physics at a critical point.

The Astonishing Puzzle of Universality

A natural next question is: what are the values of these exponents? This leads to one of the most profound discoveries in physics. To see why, let's watch the beginning of a lecture by Prof. Mehran Kardar.

The puzzle presented in the video is at the heart of modern condensed matter physics. Systems as different as a fluid of interacting molecules and a quantum fluid of helium atoms exhibit nearly identical critical exponents. This phenomenon is called universality.

It implies that near a critical point, the specific microscopic details of a system—the chemical composition of a fluid, the precise nature of the atomic interactions, the lattice structure of a magnet—become irrelevant. The singular behavior is governed only by two fundamental properties:

1. The spatial dimensionality ($d$).
2. The symmetries of the order parameter (e.g., is it a scalar, like magnetization in one direction, or a vector, like magnetization that can point anywhere in a plane?).

Systems that share these fundamental properties belong to the same universality class and share the same critical exponents.

Let's see this in action by comparing two theories you've encountered: the van der Waals model for a liquid-gas transition and the mean-field theory of the Ising model for a magnet.

This is a remarkable result. Two different physical systems, described by two different approximate theories, yield the same set of critical exponents:

• $\alpha =0$ (a jump in specific heat, not a divergence)
• $\beta =1/2$
• $\gamma =1$
• $\delta =3$

This shows that the van der Waals equation and the mean-field Ising model belong to the same universality class—the mean-field universality class.

However, as the notes and video emphasize, these values do not agree with experiments for real 3D systems (e.g., $\beta \approx 0.32$, not 0.5). This tells us that mean-field theory, which neglects fluctuations, fails to capture the correct physics in lower dimensions where fluctuations dominate. But, strikingly, the experiments show that the 3D liquid-gas transition and the 3D Ising magnet are still in the same universality class, just a different one from mean-field theory!

Deeper Structure: The Scaling Hypothesis

The existence of universality classes suggests that the critical exponents are not just a random collection of numbers. They must be connected by some underlying structure. This is formalized by the scaling hypothesis.

The hypothesis states that near the critical point, the singular part of the Gibbs free energy, $G_s(t,h)$, is a generalized homogeneous function:
$G_s({\lambda }^{a}t,{\lambda }^{b}h)=\lambda G_s(t,h)$
for some scaling exponents $a$ and $b$, and any rescaling factor $\lambda$. This is a mathematical way of expressing the physical idea of scale invariance.

Without diving into the full derivation (which we will do in Module 7), this single assumption has profound consequences: it implies that the six critical exponents are not independent. They are related by a set of equations called scaling relations.

The video shows how this hypothesis leads to scaling laws. Some of the key relations are:

• Rushbrooke Scaling: $\alpha +2\beta +\gamma =2$
• Widom Scaling: $\gamma =\beta (\delta -1)$
• Fisher Scaling: $\gamma =(2-\eta )\nu$
• Josephson Scaling: $d\nu =2-\alpha$

These relations mean that if you can determine any two of the critical exponents, you can calculate all the others! This is a powerful confirmation of the deep internal consistency of the theory of critical phenomena.

Conclusion

This lesson has armed you with the fundamental language used to describe the universal behavior of matter near a continuous phase transition.

Key Takeaways:

• The behavior of thermodynamic quantities near a critical point is described by a set of critical exponents ($\alpha ,\beta ,\gamma ,\delta ,\nu ,\eta$) which characterize power-law scaling with respect to the reduced temperature $t$.
• The phenomenon of universality dictates that systems with the same spatial dimension and order parameter symmetry share the same critical exponents, regardless of their microscopic details.
• Simple mean-field theories (like van der Waals and the mean-field Ising model) correctly predict the existence of a phase transition but yield exponents that are only accurate for dimensions $d\ge 4$. They fail to account for the crucial role of fluctuations in lower dimensions.
• The scaling hypothesis posits that the free energy is a generalized homogeneous function near $T_c$, which leads to scaling relations among the critical exponents. This implies that only two exponents are independent.

Preview of the Next Module:

We have now completed our survey of the general thermodynamic framework for phase transitions. We've seen that simple models like the van der Waals equation can provide a first, albeit imperfect, window into critical phenomena. In the next module, we will begin our deep dive into specific, solvable models by taking a much closer look at the Van der Waals Model of a Liquid-Gas Transition. We will derive the equation of state from first principles, analyze its predictions for phase coexistence, and rigorously extract the mean-field critical exponents we've discussed today.