Critical Points & Phase Transitions — Critical points and phase transitions (conceptual) - Physical Chemistry 1

SECTION 1

Historical Context & Motivation

The study of phase transitions — the transformations matter undergoes between solid, liquid, and gas states — dates to some of the earliest quantitative investigations in thermodynamics. For centuries, scientists understood that water boils and ice melts, but the deeper question of why these transitions occur so abruptly, and whether there exist conditions under which the liquid–gas boundary simply vanishes, remained elusive. The pursuit of this question led to one of the most elegant concepts in physical chemistry: the critical point, a unique set of thermodynamic conditions at which the distinction between liquid and gas ceases to exist entirely.

1822 • Cagniard de la Tour's Discovery

Charles Cagniard de la Tour observed that above a certain temperature and pressure, the meniscus between liquid and vapor in a sealed tube disappeared entirely, providing the first experimental evidence of a critical state.

1869 • Andrews' Isotherms for CO₂

Thomas Andrews systematically mapped pressure–volume isotherms for carbon dioxide, identifying the critical temperature (T_c = 304.2 K) above which no amount of pressure can liquefy the gas.

1873 • Van der Waals Equation of State

Johannes Diderik van der Waals proposed his equation accounting for molecular volume and intermolecular attractions, providing the first theoretical framework that naturally predicted critical behavior and earned him the 1910 Nobel Prize.

1933 • Ehrenfest Classification

Paul Ehrenfest introduced a systematic classification scheme for phase transitions based on the order of the thermodynamic derivative that exhibits a discontinuity, distinguishing first-order from second-order transitions.

1971 • Renormalization Group Theory

Kenneth Wilson developed renormalization group methods to explain universal behavior near critical points, revealing that seemingly different systems share identical critical exponents — work recognized by the 1982 Nobel Prize in Physics.

The central question that motivated a century of research was deceptively simple: what happens at the boundary between two phases, and is it possible for that boundary to end? Understanding critical phenomena has proven foundational not only for chemistry and physics but also for materials science, geology, and even cosmology, where analogous phase transitions occurred in the early universe.

SECTION 2

Core Principles & Definitions

To understand critical points and phase transitions, one must first appreciate how phases are defined thermodynamically and how their stability depends on the chemical potential (μ). A phase is a macroscopically homogeneous region of matter characterized by uniform intensive properties — temperature, pressure, density, and composition. Phase transitions occur when the equilibrium state of a system shifts from one phase to another as external conditions change, driven by the principle that matter spontaneously adopts the phase with the lowest Gibbs energy at constant T and P.

Phase Boundary

A curve in P–T space along which two phases coexist in equilibrium. The chemical potentials of both phases are equal: μ_α = μ_β. The Clausius–Clapeyron equation governs the slope of these curves.

Critical Point

The terminus of the liquid–vapor coexistence curve, defined by T_c, P_c, and V_c. Beyond this point, liquid and gas become indistinguishable — a supercritical fluid.

First-Order Transition

A transition involving a discontinuity in a first derivative of the Gibbs energy (e.g., volume, entropy). Characterized by latent heat absorption or release. Examples: melting, boiling, sublimation.

Second-Order (Continuous) Transition

A transition in which first derivatives of G are continuous, but second derivatives (e.g., heat capacity C_P, compressibility) diverge. No latent heat is exchanged. The critical point itself is the canonical example.

Order Parameter

A measurable quantity that is zero in the disordered phase and nonzero in the ordered phase — for the liquid–gas transition, it is the density difference (ρ_liq − ρ_gas), which approaches zero continuously at the critical point.

✦

KEY TAKEAWAY

Think of the critical point as the summit of a mountain ridge separating two valleys (liquid and gas). A first-order transition is like hiking over the ridge — you must climb up (absorb latent heat) and come down. But at the critical point, the ridge flattens to a saddle, and you can walk from one valley to the other without any elevation change. Beyond the critical point, the two valleys merge into a single open plain — you can no longer define which valley you are in. This is why a supercritical fluid possesses properties intermediate between liquid and gas, with no phase boundary to cross.

SECTION 3

Visual Explanation — The Phase Diagram

The pressure–temperature phase diagram is the most fundamental visual tool for understanding phase equilibria. It maps the regions of stability for each phase in P–T space, delineated by coexistence curves where two phases can exist in equilibrium. Three such curves — sublimation, fusion, and vaporization — meet at the triple point, the unique T and P at which all three phases coexist simultaneously. The vaporization curve, however, does not extend indefinitely; it terminates at the critical point, beyond which the concept of distinct liquid and gas phases loses meaning.

A generic P–T phase diagram showing the three coexistence curves meeting at the triple point (green). The vaporization curve terminates at the critical point (red). The dashed cyan path illustrates how a system can travel continuously from liquid to gas via the supercritical region without ever crossing a phase boundary.

A remarkable consequence of the critical point's existence is that one can convert liquid to gas (or vice versa) without ever boiling. As shown by the dashed cyan path in the diagram, one may heat a liquid above T_c at high pressure, then reduce the pressure while remaining above T_c, and finally cool back below T_c at low pressure — arriving in the gas phase without a single bubble forming. This thought experiment demonstrates that the liquid–gas distinction is not absolute; it is path-dependent and disappears entirely in the supercritical regime.

SECTION 4

Mathematical Framework

The thermodynamic description of phase transitions is rooted in the Gibbs energy G(T, P), whose first and second derivatives with respect to T and P encode all the information needed to characterize the nature of a phase transition. A first-order transition manifests as a discontinuity in the first partial derivatives of G, while a continuous (second-order) transition preserves continuity in those derivatives but introduces divergences in the second derivatives.

GIBBS ENERGY FIRST DERIVATIVES

S = −(∂G/∂T)_P V = (∂G/∂P)_T

At a first-order transition, both S and V exhibit discontinuities: ΔS = Δ_trsH / T (latent heat) and ΔV ≠ 0.

CLAUSIUS–CLAPEYRON EQUATION

dP/dT = Δ_trs H / (T × Δ_trs V)

This equation governs the slope of any first-order phase boundary in the P–T diagram. Δ_trsH is the molar enthalpy of transition and Δ_trsV is the molar volume change.

VAN DER WAALS CRITICAL CONSTANTS

T_c = 8a/(27Rb) P_c = a/(27b²) V_c = 3b

Derived from the van der Waals equation by imposing the conditions (∂P/∂V)_T = 0 and (∂²P/∂V²)_T = 0 simultaneously, defining the inflection point on the critical isotherm. Here, a quantifies intermolecular attraction and b represents the excluded volume per mole.

CRITICAL COMPRESSIBILITY FACTOR

Z_c = P_c V_c / (R T_c) = 3/8 = 0.375 (van der Waals prediction)

Real gases exhibit Z_c values typically between 0.23 and 0.29, revealing the quantitative limitations of the van der Waals model while its qualitative predictions of critical behavior remain remarkably insightful.

The conditions defining the critical point — that both the first and second partial derivatives of pressure with respect to volume vanish — have a deep geometric meaning: the critical isotherm possesses a horizontal inflection point on the P–V diagram. Below T_c, isotherms exhibit the characteristic van der Waals loop with local maxima and minima (physically realized as the two-phase coexistence region). At T_c, this loop collapses to a single inflection point, and above T_c the isotherms become monotonically decreasing — no phase separation is possible.

SECTION 5

Classification of Phase Transitions

The Ehrenfest classification categorizes phase transitions by the lowest-order derivative of the Gibbs energy that shows a discontinuity at the transition point. While modern understanding — particularly of critical phenomena — has revealed some limitations of this strict classification, it remains the standard pedagogical framework and provides considerable conceptual clarity. The diagram below illustrates how the Gibbs energy and its derivatives behave across first-order and continuous transitions.

Comparison of thermodynamic behavior across first-order (left) and continuous/second-order (right) transitions. In a first-order transition, the Gibbs energy shows a kink, entropy jumps discontinuously, and the heat capacity shows a delta-function spike. In a continuous transition, G is smooth, S is continuous, but C_P exhibits a characteristic λ-shaped divergence.

Feature	First-Order	Continuous (2nd-Order)
Latent heat	Present (Δ_trsH ≠ 0)	Absent (Δ_trsH = 0)
Volume change	Discontinuous (ΔV ≠ 0)	Continuous (ΔV = 0)
Entropy change	Discontinuous jump in S	S continuous; slope may change
Heat capacity	Delta-function spike at T_trs	λ-shaped divergence at T_c
Examples	Melting, boiling, sublimation	Liquid–gas critical point, ferromagnetic Curie point, superfluid He-4 (λ-transition)
Order parameter	Jumps discontinuously to zero	Vanishes continuously as T → T_c

Comparison of first-order and continuous phase transitions

📝

NOTE ON MODERN CLASSIFICATION

The strict Ehrenfest scheme, which envisions clean discontinuities in successively higher derivatives, breaks down at the critical point because heat capacity and compressibility do not merely jump — they diverge. Modern theory therefore distinguishes transitions into just two categories: first-order (with latent heat and coexisting phases) and continuous (where the order parameter vanishes smoothly and correlation lengths diverge). This refined view emerged from renormalization group theory and the study of universality classes.

SECTION 6

Worked Example — Critical Constants from Van der Waals Parameters

Let us determine the critical temperature, critical pressure, and critical molar volume for carbon dioxide using its van der Waals parameters: a = 3.592 L² atm mol⁻² and b = 0.04267 L mol⁻¹. We will then compare our predictions with experimental values to assess the model's accuracy.

Predicting Critical Constants for CO₂

1
Step 1 — Recall the Van der Waals Critical Expressions

From the mathematical conditions (∂P/∂V)_T = 0 and (∂²P/∂V²)_T = 0, the critical constants are: T_c = 8a / (27Rb), P_c = a / (27b²), and V_c = 3b.
2
Step 2 — Calculate Critical Temperature

T_c = 8 × 3.592 / (27 × 0.08206 × 0.04267). First compute the denominator: 27 × 0.08206 × 0.04267 = 0.09451 L atm mol⁻¹ K⁻¹. Numerator: 8 × 3.592 = 28.736 L² atm mol⁻². Therefore T_c = 28.736 / 0.09451 = 304.1 K.

T_c = 304.1 K (experimental: 304.2 K — excellent agreement!)
3
Step 3 — Calculate Critical Pressure

P_c = a / (27b²) = 3.592 / (27 × (0.04267)²) = 3.592 / (27 × 0.001821) = 3.592 / 0.04916 = 73.1 atm.

P_c = 73.1 atm (experimental: 72.9 atm — again very close)
4
Step 4 — Calculate Critical Molar Volume

V_c = 3b = 3 × 0.04267 = 0.1280 L mol⁻¹ = 128.0 cm³ mol⁻¹.

V_c = 128.0 cm³ mol⁻¹ (experimental: 94.0 cm³ mol⁻¹ — a 36% overestimate)
5
Step 5 — Evaluate the Model

The van der Waals equation predicts T_c and P_c accurately because these depend on ratios of a and b. The critical volume, however, is significantly overestimated. The predicted compressibility factor Z_c = P_cV_c/(RT_c) = 0.375, while the experimental value for CO₂ is 0.274 — a discrepancy that arises because the van der Waals equation oversimplifies repulsive interactions at high density near the critical point.

Z_c(predicted) = 0.375 vs. Z_c(expt) = 0.274

SECTION 7

Strengths & Limitations of Classical Approaches

Classical thermodynamic treatments of phase transitions, grounded in equations of state and the Gibbs formalism, provide powerful predictive tools but encounter systematic difficulties as conditions approach the critical point. Understanding where these models succeed and fail is essential for knowing when to invoke more sophisticated theories.

Aspect	Strengths	Limitations
Qualitative predictions	Van der Waals equation correctly predicts the existence of a critical point, supercritical fluid region, and continuous vanishing of the order parameter.	Classical models predict mean-field critical exponents (e.g., β = ½) that deviate from experimental values (β ≈ 0.326 for 3D Ising systems).
Corresponding states	The law of corresponding states (reduced variables P/P_c, T/T_c, V/V_c) collapses data for many substances onto universal curves.	Breaks down for polar or associating molecules (H₂O, alcohols) where intermolecular forces differ qualitatively from simple dispersion.
Clausius–Clapeyron	Accurately describes phase boundary slopes far from the critical point using measurable quantities (latent heat, volume change).	Fails at the critical point where ΔV → 0 and ΔH → 0, making dP/dT indeterminate (0/0).
Critical fluctuations	Classical theory provides clear, analytically tractable frameworks for engineering applications.	Near the critical point, density fluctuations span all length scales (critical opalescence), and mean-field theory neglects these correlated fluctuations entirely.

Classical approaches to phase transitions: strengths and limitations

✦

KEY TAKEAWAY

Classical (mean-field) models are like predicting weather from a single barometer reading — they capture the broad trends reliably but miss the violent, correlated turbulence that dominates near extreme conditions. Just as detailed weather forecasting requires accounting for interactions across many length scales, accurately describing critical phenomena requires renormalization group theory, which accounts for fluctuations at every scale simultaneously. This is why mean-field critical exponents are only approximations — the true exponents reflect the deep, scale-invariant structure of cooperative fluctuations.

SECTION 8

Connections to Advanced Theory — Universality & Critical Exponents

One of the most profound discoveries in twentieth-century physics is that seemingly unrelated systems — the liquid–gas transition, the ferromagnetic–paramagnetic transition, and the order–disorder transition in binary alloys — all exhibit identical behavior near their respective critical points. This remarkable observation is known as universality, and it implies that the microscopic details of a system (molecular shape, interaction potential, lattice structure) become irrelevant near criticality. Only the dimensionality of the system and the symmetry of the order parameter matter, assigning each system to a universality class characterized by a common set of critical exponents.

Critical Exponent	Quantity It Describes	Mean-Field Value	3D Ising (Experimental)
β	Order parameter vanishing: \|ρ_l − ρ_g\| ∝ (T_c − T)^β	1/2	0.326
γ	Isothermal compressibility divergence: κ_T ∝ \|T − T_c\|^(−γ)	1	1.237
δ	Critical isotherm shape: \|P − P_c\| ∝ \|ρ − ρ_c\|^δ at T = T_c	3	4.789
α	Heat capacity divergence: C_P ∝ \|T − T_c\|^(−α)	0 (jump)	0.110

Critical exponents: mean-field vs. experimental values for the 3D Ising universality class

The discrepancies between mean-field and experimental critical exponents arise because classical theory assumes each molecule interacts with an average (mean) field produced by all others, neglecting the fact that near T_c the correlation length — the distance over which density fluctuations are coupled — diverges toward infinity. When fluctuations are correlated over macroscopic distances, the mean-field approximation fundamentally breaks down. This is the regime where critical opalescence occurs: density fluctuations reach the wavelength of visible light, scattering it intensely and rendering the fluid milky and opalescent. The renormalization group provides a systematic framework for handling these multi-scale fluctuations, and its predictions for critical exponents match experiment to five or more significant figures — a triumph of modern theoretical physics that extends far beyond phase transitions into quantum field theory and statistical mechanics.

SECTION 9

Practice Problems

Question 1: CONCEPTUAL

Explain why the solid–liquid phase boundary in a P–T diagram does not terminate at a critical point, whereas the liquid–gas boundary does. What is fundamentally different about the structural relationship between solid and liquid compared to liquid and gas?

Show answer

The liquid–gas transition involves a continuous change in a scalar order parameter (density), meaning the two phases differ only quantitatively — both are disordered fluids with different densities. As conditions approach the critical point, this density difference vanishes continuously, and the phases become indistinguishable. By contrast, the solid–liquid transition involves a change in symmetry — solids possess long-range translational order (a crystalline lattice), while liquids do not. You cannot continuously interpolate between 'ordered' and 'disordered' using a single scalar parameter in the same way; the symmetry either is present or it is not. This discrete symmetry-breaking means no critical point terminates the fusion curve, and so the solid–liquid boundary extends indefinitely in principle.

Question 2: BASIC CALCULATION

Using the van der Waals parameters for nitrogen (N₂) — a = 1.390 L² atm mol⁻² and b = 0.03913 L mol⁻¹ — calculate T_c, P_c, and V_c. Compare with the experimental values (T_c = 126.2 K, P_c = 33.5 atm).

Show answer

Using T_c = 8a/(27Rb): T_c = 8 × 1.390 / (27 × 0.08206 × 0.03913) = 11.120 / 0.08672 = 128.2 K (experimental: 126.2 K, 1.6% error). For P_c = a/(27b²): P_c = 1.390 / (27 × 0.001531) = 1.390 / 0.04134 = 33.6 atm (experimental: 33.5 atm, excellent). For V_c = 3b = 3 × 0.03913 = 0.1174 L mol⁻¹ = 117.4 cm³ mol⁻¹. The experimental V_c is 89.2 cm³ mol⁻¹, so the van der Waals equation overestimates V_c by about 32%, consistent with its known limitations in predicting densities near the critical point.

Question 3: INTERMEDIATE

At the critical point for water (T_c = 647 K, P_c = 218 atm), both ΔH_vap and ΔV_vap vanish. Explain how this is consistent with the Clausius–Clapeyron equation. What happens to dP/dT of the vaporization curve as T → T_c?

Show answer

The Clausius–Clapeyron equation gives dP/dT = Δ_vapH / (TΔ_vapV). As T → T_c, both numerator and denominator approach zero, yielding an indeterminate form 0/0. The actual slope dP/dT remains finite and well-defined along the vaporization curve — it equals the slope of the coexistence curve at its terminus, which is roughly 4.5 atm/K for water near the critical point. What this tells us physically is that latent heat and volume change diminish in tandem as the distinction between liquid and vapor fades. Mathematically, L'Hôpital-like analysis (or the fact that ΔH/ΔV → T × dP/dT, which stays finite) confirms consistency. The key insight is that the Clausius–Clapeyron equation describes the phase boundary slope, not the individual quantities in the ratio, and the ratio remains finite even though both quantities vanish.

Question 4: APPLIED

Supercritical CO₂ (scCO₂) is widely used as a green solvent in industrial extraction processes (e.g., decaffeinating coffee). Explain why scCO₂ is favored over liquid solvents from the perspective of phase behavior and critical point properties. The critical constants for CO₂ are T_c = 304.2 K and P_c = 72.9 atm.

Show answer

Supercritical CO₂ is ideal because its critical temperature (304.2 K ≈ 31 °C) is near ambient, so only modest heating is required, and its critical pressure (72.9 atm ≈ 74 bar), while high, is industrially accessible. In the supercritical state, CO₂ possesses liquid-like density (providing strong solvating power) combined with gas-like viscosity and diffusivity (enabling rapid penetration into solid matrices like coffee beans). Crucially, because there is no phase boundary above the critical point, the solvent's properties can be tuned continuously by adjusting temperature and pressure — increasing pressure raises density and solvating power, while depressurization causes dissolved solutes to precipitate cleanly as CO₂ reverts to a gas and escapes without residue. This eliminates the need for energy-intensive distillation steps required by conventional liquid solvents, making the process both efficient and environmentally favorable.

Question 5: CRITICAL THINKING

Consider a hypothetical system in which the mean-field critical exponent β = 1/2 were exactly correct. How would the coexistence curve (a plot of ρ_liq and ρ_gas vs. T near T_c) differ in shape compared to the experimental case where β ≈ 0.326? Sketch or describe both curves qualitatively and discuss what the experimental value of β tells us about the nature of fluctuations near T_c.

Show answer

The order parameter follows |ρ_liq − ρ_gas| ∝ (T_c − T)^β. With β = 1/2 (mean-field), the coexistence curve near T_c would be a parabola — a gently rounded dome. With the experimental β ≈ 0.326 < 1/2, the coexistence curve is flatter near the top and more sharply pointed — the dome has a pronounced cusp-like shape. This flatter top means that ρ_liq and ρ_gas remain distinguishable (differ from each other) over a narrower temperature range below T_c than mean-field theory predicts. Physically, the smaller β reflects the role of large-scale, correlated density fluctuations: these fluctuations effectively 'mix' the two phases more efficiently than mean-field theory anticipates, causing the order parameter to vanish more rapidly as T_c is approached. The deviation of β from 1/2 is thus a direct signature of cooperative phenomena — fluctuations that are not independent but correlated over divergent length scales.

SUMMARY

Lesson Summary

Summary

Phase transitions describe transformations between thermodynamic phases driven by the principle of Gibbs energy minimization. First-order transitions involve latent heat and discontinuities in volume and entropy, as governed by the Clausius–Clapeyron equation. The liquid–gas coexistence curve terminates at the critical point (T_c, P_c, V_c), where the order parameter (density difference) vanishes continuously and the transition becomes second-order (continuous). Beyond this point, matter exists as a supercritical fluid with hybrid liquid–gas properties.

The van der Waals equation provides the simplest theoretical framework predicting critical behavior, yielding critical constants in terms of molecular parameters a and b and a universal Z_c = 3/8. While quantitatively limited, it correctly captures the topology of the phase diagram. Near T_c, critical fluctuations dominate and mean-field theory fails; the experimental critical exponents differ from classical predictions. Universality — the observation that critical exponents depend only on dimensionality and symmetry, not on molecular details — is one of the deepest insights of modern statistical mechanics, explained by renormalization group theory.