Law of Corresponding States

Photograph the liquid–gas coexistence dome of argon in its own pressure–temperature axes, then do the same for nitrogen, methane, carbon dioxide. The four pictures look different — the substances boil, condense, and reach their critical points at wildly different pressures and temperatures. Now relabel every axis in units of that substance's own critical value. The four photographs snap into one. This is the law of corresponding states: measured in reduced variables $P_r=P/P_c$ , $V_r=V/V_c$ , $T_r=T/T_c$ , the equation of state of every simple fluid collapses onto a single universal surface. It is one of the oldest and most useful unifying ideas in the thermodynamics of matter, first stated by van der Waals in 1880, and this article owns both its derivation and the honest accounting of where it succeeds and fails. Two routes reach the law. The first is almost trivial and purely algebraic: any two-parameter equation of state (van der Waals, Redlich–Kwong, Berthelot) obeys a corresponding-states law automatically, because its two molecular constants can be traded for the two independent critical coordinates, leaving a relation among reduced variables with no substance-specific information left in it. The van der Waals equation reduces exactly to $(P_r+3/V_r^2)(3V_r-1)=8T_r$ , a parameter-free curve. The second route is deeper and statistical-mechanical: if the intermolecular pair potential is a single energy scale times a universal shape, $u(r)=\varepsilon\,\phi(r/\sigma)$ , then the classical configuration integral — and therefore every thermodynamic function — depends only on the two reduced groups $T^*=k_BT/\varepsilon$ and $\rho^*=\rho\sigma^3$ . All fluids sharing the shape $\phi$ are thermodynamically congruent, not by coincidence but by a rescaling that removes their only two dimensional differences. The sharpest single-number test is the critical compressibility factor $Z_c=P_cV_c/(Nk_BT_c)$ . Van der Waals predicts the universal value $Z_c=3/8=0.375$ . Experiment confirms that $Z_c$ is indeed nearly constant across simple fluids — but smaller: argon $0.291$ , nitrogen $0.290$ , methane $0.288$ , carbon dioxide $0.274$ , water $0.229$ . The near-constancy vindicates the principle; the systematic shortfall below $3/8$ convicts the van der Waals potential of being too crude. That double verdict — right idea, wrong number — is the whole story of mean-field corresponding states, and its failure near the true critical point (where fluctuations enforce Ising, not mean-field, exponents) is exactly the doorway to universality and the renormalization group later in this track.

⚡ Key Takeaways

The law of corresponding states: in reduced variables $P_r=P/P_c$ , $V_r=V/V_c$ , $T_r=T/T_c$ , the equation of state $f(P_r,V_r,T_r)=0$ is the same universal function for all simple fluids.
It follows exactly from any two-parameter equation of state; van der Waals reduces to the parameter-free $(P_r+3/V_r^2)(3V_r-1)=8T_r$ .
More deeply it holds whenever the pair potential is a single scale times a universal shape, $u(r)=\varepsilon\,\phi(r/\sigma)$ , so classical thermodynamics depends only on $T^*=k_BT/\varepsilon$ and $\rho^*=\rho\sigma^3$ .
The critical compressibility factor is the sharpest prediction: van der Waals gives $Z_c=3/8=0.375$ ; real simple fluids cluster near $0.29$ (Ar $0.291$ , N₂ $0.290$ , CH₄ $0.288$ ), below the mean-field value.
The near-constancy confirms the principle; the shortfall below $3/8$ shows the van der Waals form is qualitatively right but quantitatively crude — Pitzer's acentric factor $\omega$ restores accuracy.
Key idea: strip out the two molecular scales — an energy $\varepsilon$ and a length $\sigma$ (equivalently $a,b$ , equivalently $P_c,T_c$ ) — and every simple fluid's equation of state becomes the same dimensionless surface; what remains is universal.

💡 The Core Insight

Strip out the two molecular scales — an energy $\varepsilon$ and a length $\sigma$ (equivalently $a$ and $b$ , equivalently $P_c$ and $T_c$ ) — and the equation of state of every simple fluid becomes the same dimensionless surface; what remains is universal.

The physical picture

The critical point acts as a ruler that each substance carries with it. Argon's critical temperature ( $151$ K) and pressure ( $48.6$ atm) are set by the depth and range of the argon–argon interaction; methane's are set by methane's. But if the two interactions have the same shape and differ only in overall depth and size, then measuring each fluid in units of its own critical point divides out precisely those two differences. The residual, dimensionless equation of state has nothing substance-specific left to distinguish the fluids — so they coincide. Corresponding states is the macroscopic shadow of a microscopic sameness. At the molecular level a fluid "feels" temperature only through the ratio $T^*=k_BT/\varepsilon$ — thermal energy against the depth of the attractive well — and density only through $\rho^*=\rho\sigma^3$ — how many molecular volumes are packed together. Two fluids with the same well shape $\phi$ , differing only in depth $\varepsilon$ and diameter $\sigma$ , are running the identical movie at different playback speeds and zoom levels. Reduced variables synchronize the playback frame for frame. The first figure shows several fluids' coexistence data collapsing onto one reduced curve.

Figure

Coexistence-curve collapse: Ar, N₂, CH₄, CO₂ plotted as T_r vs reduced density fall on one universal dome (Guggenheim 1945); the van der Waals dome sits nearby but its apex Z_c = 0.375 overshoots the measured ≈ 0.29

The number $Z_c$ is a compact litmus test. It measures how non-ideal a fluid is exactly at its critical point: $Z_c=1$ would be an ideal gas, and the actual value tells you how far from ideal the fluid sits at criticality. A universal $Z_c$ would mean every fluid is non-ideal in exactly the same way there. The measured values are nearly constant near $0.29$ — the principle is real — but they are not the van der Waals $3/8=0.375$ , which is the fingerprint of a too-crude potential. The second figure lays the prediction against the data directly.

Figure

The Z_c verdict: van der Waals predicts the universal Z_c = 0.375 (amber line), but measured simple fluids cluster near 0.29 (Ar 0.291, N₂ 0.290, CH₄ 0.288, CO₂ 0.274) — near-constant, confirming the principle, but systematically below 3/8

The core equations

Begin from the van der Waals equation in its extensive form, the two-parameter equation whose reduction demonstrates the law most cleanly. It is owned by aq-sm-van-der-waals-equation; here it is the vehicle.

\left(P+\frac{aN^2}{V^2}\right)(V-Nb)=Nk_BT \qquad (1)

where: $a$ — attraction parameter (energy × volume, $\propto\varepsilon\sigma^3$ ); $b$ — excluded volume per molecule ( $\propto\sigma^3$ ); $N$ — molecule number; $V,P,T$ — volume, pressure, temperature. The two molecular constants $a$ and $b$ are the only substance-specific inputs. Imposing the horizontal-inflection conditions on the critical isotherm locates the critical point, which will serve as the substance's built-in unit system.

\left(\frac{\partial P}{\partial V}\right)_T=\left(\frac{\partial^2 P}{\partial V^2}\right)_T=0\ \Rightarrow\ P_c=\frac{a}{27b^2},\quad V_c=3Nb,\quad k_BT_c=\frac{8a}{27b} \qquad (2)

where: $P_c,V_c,T_c$ — the critical pressure, volume, temperature (derived at length in aq-sm-van-der-waals-equation). Note the three critical coordinates depend on the two constants $a,b$ , so they are not independent — a fact that itself forces a universal $Z_c$ . Trading $a,b$ for the critical coordinates and rescaling every variable by its critical value collapses the equation to a form with no molecular constants left.

\left(P_r+\frac{3}{V_r^2}\right)(3V_r-1)=8T_r \qquad (3)

where: $P_r=P/P_c$ , $V_r=V/V_c$ , $T_r=T/T_c$ — reduced pressure, volume, temperature. This is the corresponding-states law in its sharpest form: one curve, valid for every van der Waals fluid regardless of $a$ and $b$ . Evaluating the compressibility factor at the critical point gives the law's single most testable number.

Z_c=\frac{P_cV_c}{Nk_BT_c}=\frac{(a/27b^2)(3Nb)}{N(8a/27b)}=\frac{3}{8}=0.375 \qquad (4)

where: $Z_c$ — critical compressibility factor. The molecular constants cancel identically, leaving a pure number — predicted universal, and this is exactly the claim experiment scrutinizes. The microscopic foundation replaces the specific van der Waals form with the general single-scale hypothesis: a pair potential that is one energy scale times a universal shape.

u(r)=\varepsilon\,\phi\!\left(\frac{r}{\sigma}\right)\ \Rightarrow\ \frac{PV}{Nk_BT}=g\!\left(\frac{k_BT}{\varepsilon},\frac{N\sigma^3}{V}\right)\equiv g(T^*,\rho^*) \qquad (5)

where: $\varepsilon$ — well depth; $\sigma$ — molecular diameter; $\phi$ — universal dimensionless shape; $T^*=k_BT/\varepsilon$ ; $\rho^*=N\sigma^3/V$ . Any two fluids sharing $\phi$ obey the identical reduced equation of state — the law of corresponding states, now derived from the configuration integral rather than assumed from a two-parameter fit. A low-density, mean-field-free test comes from the second virial coefficient, which in reduced form is a universal function of $T^*$ alone.

\frac{B_2(T)}{\sigma^3}=-2\pi\!\int_0^\infty\!\left(e^{-\phi(x)/T^*}-1\right)x^2\,\mathrm{d}x\equiv b_2^*(T^*) \qquad (6)

where: $B_2$ — second virial coefficient (owned by aq-sm-second-virial-coefficient); $x=r/\sigma$ — reduced separation; $b_2^*$ — universal dimensionless function of $T^*$ . Its zero defines a universal reduced Boyle temperature. Both the algebraic route (through the van der Waals critical point) and the microscopic route (through the single-scale configuration integral) are proved below and machine-checked.

Derivation AThe van der Waals critical point and the universal Z_c = 3/8

The goal is to derive the critical coordinates of the van der Waals fluid from the inflection conditions and assemble them into the compressibility factor $Z_c$ , the law's sharpest single-number prediction. The strategy is to solve the equation for the pressure, impose that the critical isotherm has a horizontal inflection (first and second volume derivatives both vanish), and solve the two simultaneous conditions for $V_c$ and $T_c$ , then back out $P_c$ . The decisive point is that the three critical coordinates are built from only two molecular constants, so their particular combination $Z_c=P_cV_c/Nk_BT_c$ must be a pure number independent of the substance — that inevitability, not the specific value, is the content of corresponding states.

Step 1. Pressure form. Solve equation (1) for pressure: $P=\dfrac{Nk_BT}{V-Nb}-\dfrac{aN^2}{V^2}$ . Why: the critical point is defined by the shape of the $P(V)$ isotherm, so we need $P$ as an explicit function of $V$ at fixed $T$ .

Step 2. Horizontal inflection. At the critical point the isotherm is flat and inflecting: $\left(\dfrac{\partial P}{\partial V}\right)_T=-\dfrac{Nk_BT}{(V-Nb)^2}+\dfrac{2aN^2}{V^3}=0$ and $\left(\dfrac{\partial^2P}{\partial V^2}\right)_T=\dfrac{2Nk_BT}{(V-Nb)^3}-\dfrac{6aN^2}{V^4}=0$ . Why: below $T_c$ the isotherm loops (two extrema), above $T_c$ it is monotonic; exactly at $T_c$ the two extrema merge into a single inflection where both derivatives vanish.

Step 3. Ratio eliminates $T$ and $a$ . Dividing the second-derivative equation by the first gives $\dfrac{2}{V-Nb}=\dfrac{3}{V}$ , hence $V_c=3Nb$ . Why: the ratio cancels the common $Nk_BT$ and the attraction $a$ , isolating a pure geometric condition on the critical volume.

Step 4. Back out $T_c$ and $P_c$ . Substituting $V_c=3Nb$ into the first-derivative equation gives $k_BT_c=\dfrac{8a}{27b}$ , and inserting both into the pressure gives $P_c=\dfrac{a}{27b^2}$ . Why: with the critical volume fixed, one derivative condition fixes the temperature and the equation of state itself then fixes the pressure.

Step 5. Assemble $Z_c$ . $Z_c=\dfrac{P_cV_c}{Nk_BT_c}=\dfrac{(a/27b^2)(3Nb)}{N(8a/27b)}=\dfrac{3}{8}=0.375$ . Every factor of $a$ , $b$ , $N$ cancels. Why: the cancellation is the whole point — $Z_c$ is a pure number, so van der Waals predicts the same critical compressibility for every fluid; the measured spread near $0.29$ (below $3/8$ ) then measures how far real potentials depart from the van der Waals caricature. *Verified: the inflection algebra, $V_c=3Nb$ , $k_BT_c=8a/27b$ , $P_c=a/27b^2$ , the reduced EOS $(P_r+3/V_r^2)(3V_r-1)=8T_r$ (all constants cancel), and $Z_c=3/8$ are checked symbolically in verify_law-of-corresponding-states.py; measured $Z_c$ values (Ar $0.291$ , N₂ $0.290$ , CH₄ $0.288$ , CO₂ $0.274$ , H₂O $0.229$ ) are all confirmed below $3/8$ .*

Derivation BFrom a single-scale pair potential to universal reduced thermodynamics

The goal is to prove — with no mean-field approximation anywhere — that a pair potential of the form $u(r)=\varepsilon\,\phi(r/\sigma)$ forces every reduced thermodynamic function to be universal, so that corresponding states is a rigorous property of the configuration integral rather than an artifact of a two-parameter fit. The strategy is to write the classical partition function, isolate the configuration integral where all the interaction physics lives, and rescale every coordinate by the length $\sigma$ . Once that is done the only dimensional combination left in the integrand is $\varepsilon/k_BT=1/T^*$ , and the volume enters only as $V/\sigma^3$ ; every thermodynamic function built from the integral therefore depends on $T^*$ and $\rho^*$ alone. The one thing to watch is the classical assumption: the thermal wavelength must be small compared to $\sigma$ , or a third quantum scale survives the rescaling.

Step 1. Factor the partition function. $Z_N=\dfrac{1}{N!\,\lambda_T^{3N}}Q_N$ with $\lambda_T=h/\sqrt{2\pi mk_BT}$ and $Q_N=\int\prod_i\mathrm{d}^3r_i\,\exp\!\big[-\tfrac{\varepsilon}{k_BT}\sum_{i<j}\phi(r_{ij}/\sigma)\big]$ . Why: the kinetic part contributes only the thermal wavelength; all substance-specific interaction physics is confined to the configuration integral $Q_N$ , so that is where the rescaling must act.

Step 2. Nondimensionalize the coordinates. Substitute $r_i=\sigma\,\tilde r_i$ ; each of the $N$ three-dimensional position integrals contributes a factor $\sigma^3$ , so $Q_N=\sigma^{3N}\,I_N(T^*,V/\sigma^3)$ with $I_N=\int\prod_i\mathrm{d}^3\tilde r_i\,\exp\!\big[-\tfrac{1}{T^*}\sum_{i<j}\phi(\tilde r_{ij})\big]$ . Why: after rescaling, the integrand contains no dimensional constants — only the pure shape $\phi$ and the ratio $1/T^*$ — and the integration domain is the reduced box $V/\sigma^3$ ; so $I_N$ is a pure number fixed by $T^*$ , $V/\sigma^3$ , and $N$ .

Step 3. Isolate the universal excess free energy. $\dfrac{F-F_{\text{ideal}}}{Nk_BT}=-\dfrac1N\ln\dfrac{I_N}{(V/\sigma^3)^N}\equiv\psi(T^*,\rho^*)$ . Why: subtracting the ideal-gas free energy removes the kinetic ( $\lambda_T$ ) part entirely, provided $\lambda_T\ll\sigma$ (the classical limit); what remains is a universal function of the two reduced groups, identical for every fluid sharing the shape $\phi$ .

Step 4. Differentiate to the equation of state. $P=-\left(\dfrac{\partial F}{\partial V}\right)_{T,N}\Rightarrow\dfrac{PV}{Nk_BT}=g(T^*,\rho^*)$ . Why: a universal $\psi$ yields a universal $g$ ; two fluids with the same $\phi$ obey the identical reduced equation of state exactly, which is the law of corresponding states derived from first principles rather than from a two-parameter equation. The Lennard-Jones $\phi$ is the canonical shape for the noble gases. *Verified: the $\sigma^{3N}$ Jacobian and the $T^*=k_BT/\varepsilon$ dependence are checked in verify_law-of-corresponding-states.py, along with the reduced second virial coefficient $B_2/\sigma^3=b_2^*(T^*)$ for the Lennard-Jones shape, whose zero (the reduced Boyle temperature) is the universal number $T^*_B\approx3.42$ .*

Limiting cases

The law holds precisely to the extent that a fluid is a classical, single-scale, spherically symmetric collection of particles; each way of failing that description is diagnostic.

Case	Result	Meaning
Noble gases (Ar, Kr, Xe)	$Z_c\approx0.29$ , $\omega\approx0$	Nearly perfect Lennard-Jones spheres; the law works best here — Guggenheim's 1945 collapse used exactly these.
Quantum fluids (He, H₂, Ne)	de Boer $\Lambda^*=h/(\sigma\sqrt{m\varepsilon})\not\to0$	Light mass makes the thermal wavelength comparable to $\sigma$ ; a third quantum parameter is needed, and even ³He and ⁴He differ.
Polar / associating (H₂O, NH₃)	$\omega\gg0$ , $Z_c\approx0.23$	Directional hydrogen bonding breaks the single-scale spherical assumption; large deviations from the noble-gas surface.
Critical region, $\lvert T-T_c\rvert/T_c\to0$	Mean-field collapse fails	Fluctuations dominate; the true universal behaviour is 3D-Ising ( $\beta\approx0.326$ , not $\tfrac12$ ; $\delta\approx4.79$ , not $3$ ).

⚠️ Two different universalities — do not conflate them

The classical law of corresponding states is a smooth, analytic collapse from a two-parameter equation of state; it predicts $Z_c=3/8$ and mean-field critical exponents ( $\beta=\tfrac12$ , $\delta=3$ ), and it is only approximately true — failing outright right at the critical point. The genuinely exact universality there is a different beast: all simple fluids fall into the 3D-Ising universality class and share its exponents ( $\beta\approx0.326$ , $\delta\approx4.79$ ), which are not the van der Waals values. So "corresponding states" names two things — a rough, mean-field data collapse away from criticality, and an exact, fluctuation-driven universality at criticality — and they give different numbers. Also remember the law assumes classical statistics and a single-scale spherical potential: helium and hydrogen violate it through quantum effects (a third de Boer parameter $\Lambda^*$ is needed), and water and ammonia through directional bonding.

Worked examples

1 · The critical compressibility factor from the van der Waals constants (a real number)

We compute $Z_c$ for the van der Waals fluid directly from $a$ and $b$ and confirm it is substance-independent, then compare with argon.

Step 1. From the critical point, $P_c=a/27b^2$ , $V_c=3Nb$ , $k_BT_c=8a/27b$ . Assemble $Z_c=\dfrac{P_cV_c}{Nk_BT_c}=\dfrac{(a/27b^2)(3Nb)}{N(8a/27b)}$ .

Step 2. Cancel: the numerator is $3aNb/27b^2=aN/9b$ ; the denominator is $8aN/27b$ . The ratio is $\dfrac{aN/9b}{8aN/27b}=\dfrac{27}{9\cdot8}=\dfrac{3}{8}$ .

Step 3. So $\mathbf{Z_c=3/8=0.375}$ , with every factor of $a$ , $b$ , $N$ gone — the van der Waals prediction is the same number for every fluid.

Step 4. Argon measures $Z_c=0.291$ . The prediction is $29\%$ too high. The near-constancy of measured $Z_c$ across simple fluids confirms the corresponding-states principle, while the systematic $0.29$ -vs- $0.375$ gap convicts the van der Waals potential of being quantitatively crude — the right idea with the wrong well shape.

2 · The reduced Boyle temperature of the Lennard-Jones fluid (universal number)

We compute the temperature at which the second virial coefficient vanishes, in reduced units, and confirm it is a universal number independent of the substance.

Step 1. For the Lennard-Jones shape $\phi(x)=4(x^{-12}-x^{-6})$ (with $x=r/\sigma$ ), the reduced second virial coefficient is $b_2^*(T^*)=-2\pi\int_0^\infty(e^{-\phi(x)/T^*}-1)x^2\,\mathrm{d}x$ , a function of the reduced temperature $T^*=k_BT/\varepsilon$ alone.

Step 2. At low $T^*$ the attractive well dominates: $e^{-\phi/T^*}$ is large where $\phi<0$ , so the integrand is positive there and $b_2^*<0$ (net attraction). At high $T^*$ the repulsive core dominates: $b_2^*>0$ . Somewhere between, $b_2^*$ crosses zero — the reduced Boyle temperature.

Step 3. Numerically integrating and root-finding gives $\mathbf{T^*_B=k_BT_B/\varepsilon\approx3.42}$ , a pure number set by the Lennard-Jones shape and nothing else.

Step 4. Because $T^*_B$ is universal, every Lennard-Jones fluid has its Boyle temperature at $T_B\approx3.42\,\varepsilon/k_B$ : argon ( $\varepsilon/k_B\approx120$ K) gives $T_B\approx410$ K, nitrogen its own value, and so on — but all collapse to the same $T^*_B$ when plotted against $T^*$ . A single dimensionless integral fixes the Boyle temperature of every noble-gas-like fluid at once, a low-density, mean-field-free instance of corresponding states.

boyle temperature

the temperature where the second virial coefficient vanishes and a real gas is ideal to first order in density (Boyle's 1662 law; corresponding-states reading, van der Waals 1880) · a universal reduced temperature for single-scale fluids

The Boyle temperature $T_B$ is where the second virial coefficient $B_2(T)$ passes through zero, so that a real gas obeys $PV=Nk_BT$ to first order in density — the attractive and repulsive contributions to $B_2$ exactly cancel. It carries Robert Boyle's name because at that temperature the gas most closely obeys the pressure–volume law he published in 1662, though Boyle himself knew nothing of virial coefficients. In the corresponding-states framework the Boyle temperature is not an isolated fact but a universal marker: for any single-scale potential $u=\varepsilon\phi(r/\sigma)$ the reduced coefficient $B_2/\sigma^3$ is a universal function of $T^*=k_BT/\varepsilon$ , so its zero occurs at a universal reduced temperature $T^*_B$ — about $3.42$ for the Lennard-Jones shape. Thus every noble-gas-like fluid has its Boyle temperature at the same $T^*_B$ , differing only through the substance's energy scale $\varepsilon$ . The Boyle temperature is therefore one of the cleanest, lowest-density tests of the law of corresponding states: it requires no mean-field approximation, only the assumption that all these fluids share one interaction shape.

⏳ History

1662

— Boyle publishes the pressure–volume law that, at the temperature later named for him, a real gas obeys to first order.

1873

— van der Waals's equation of state introduces the two molecular constants a and b.

1880

— van der Waals states the law of corresponding states, noting the collapse of reduced equations of state.

1901

— Kamerlingh Onnes casts the equation of state as the virial expansion, framing corresponding states in reduced coefficients.

1945

— Guggenheim demonstrates the coexistence-curve collapse of eight simple fluids, with an exponent near 1/3, not the mean-field 1/2.

1955

— Pitzer introduces the acentric factor ω, extending corresponding states to non-spherical molecules.

Where this lives today

The law of corresponding states is the daily workhorse of chemical and petroleum engineering. Given only a substance's critical constants $P_c,T_c$ , generalized compressibility charts (the Nelson–Obert charts) built once from reference fluids predict $Z=PV/Nk_BT$ at any state — a single universal chart replacing full $P$ – $V$ – $T$ measurements for every new chemical. Pitzer's three-parameter extension, adding the acentric factor $\omega$ to capture departures from spherical symmetry, underlies the cubic equations of state (Soave–Redlich–Kwong, Peng–Robinson) that run process simulations worldwide. The reduced second virial coefficient collapse gives a low-density, assumption-light test that helium and argon are, at heart, the same fluid at different scales. Its deepest legacy is conceptual. Corresponding states is the classical, mean-field ancestor of universality. Guggenheim's 1945 coexistence collapse already carried the clue: his data followed an exponent near $1/3$ , not the mean-field $1/2$ , an early empirical signal that the true critical universality is not van der Waals but Ising. The modern statement is that all simple fluids flow, under the renormalization group, to the same 3D-Ising fixed point, and so share exact critical exponents and scaling functions — a far more profound "corresponding states" than the analytic van der Waals collapse, and one that only makes sense once fluctuations, not mean fields, are in charge. The law of corresponding states thus sits exactly at the hinge between mean-field thermodynamics and the renormalization group that closes this track.

Going further

Texts. Pathria & Beale, Statistical Mechanics, 3rd ed., §12.2 (van der Waals and corresponding states). Kardar, Statistical Physics of Particles, Ch. 5 (interacting gas, van der Waals). Reif, Fundamentals of Statistical and Thermal Physics, §10.4. Sethna, Statistical Mechanics: Entropy, Order Parameters, and Complexity, Ch. 12 (continuous phase transitions, universality). Landau & Lifshitz, Statistical Physics Part 1, §84. Guggenheim, J. Chem. Phys. 13, 253 (1945). Pitzer et al., J. Am. Chem. Soc. 77, 3433 (1955).

Provenance. Base article: AllQFT track `sm`, "Law of Corresponding States" (STAT.INTER.CORRS.1253); source Pathria & Beale §12.2, Landau–Lifshitz §84. All symbolic and numeric claims re-verified here in verify_law-of-corresponding-states.py: the van der Waals critical point, the reduced EOS (all $a,b,N$ cancel), $Z_c=3/8$ , the measured $Z_c$ values (Ar $0.291$ , N₂ $0.290$ , CH₄ $0.288$ , CO₂ $0.274$ , H₂O $0.229$ , all below $3/8$ ), the single-scale rescaling ( $\sigma^{3N}$ Jacobian, $T^*$ dependence), and the universal Lennard-Jones reduced Boyle temperature $T^*_B\approx3.42$ (computed by numerical integration, not quoted). No base physics error was found; the base cited a reduced Boyle temperature $T_r\approx2.5$ (in units of $T_c$ ) for noble gases, which is a different reduced group from the microscopic $T^*_B=k_BT_B/\varepsilon\approx3.42$ computed here — both are legitimate and consistent (they use different reference scales), and we state the microscopic one explicitly to avoid confusion. This article owns the reduced EOS and the universality (mean-field, with its quantitative failure); the exact Ising critical exponents and the renormalization group are later chapters.

Prerequisites. Van der Waals Equation of State · Maxwell Construction and Coexistence · Configuration Integral and the Interacting Partition Function · Virial Expansion of the Equation of State.

Leads to. Critical Exponents and Universality · Renormalization Group · Second Virial Coefficient.

Why do only two molecular parameters, and not three or four, appear?

Because the simplest realistic model of a simple fluid has exactly two scales: an energy (how deep the attractive well is, $\varepsilon$ ) and a length (how big a molecule is, $\sigma$ ). Everything else about the interaction is captured by the dimensionless shape $\phi(r/\sigma)$ , which is assumed common to the whole class. Two scales mean two dimensional constants, which can be traded for two of the three critical coordinates; the third critical coordinate is then predicted, which is why $Z_c$ comes out universal. When a fluid needs a third parameter — a quantum length for helium, or a shape/polarity parameter (Pitzer's $\omega$ ) for water — it is telling you the two-scale picture is incomplete for that substance.

What exactly is the difference between T^* and T_r?

They are two different reduced temperatures using two different reference scales, and both are legitimate. $T_r=T/T_c$ measures temperature against the critical temperature — a macroscopic, directly measurable scale. $T^*=k_BT/\varepsilon$ measures it against the microscopic well depth $\varepsilon$ — a molecular scale that comes from the pair potential. For a van der Waals fluid $T_c=8a/27bk_B$ links the two, so a given fluid has a fixed conversion factor between $T_r$ and $T^*$ ; for the Lennard-Jones fluid the critical temperature is $k_BT_c\approx1.3\,\varepsilon$ , so the reduced Boyle temperature $T^*_B\approx3.4$ corresponds to $T_r\approx2.6$ . This article computes and states the microscopic $T^*_B\approx3.42$ to be unambiguous, since "reduced Boyle temperature" quoted without its reference scale is a common source of confusion.

Why does helium violate the classical law so badly?

Helium is light, so its thermal de Broglie wavelength $\lambda_T=h/\sqrt{2\pi mk_BT}$ is comparable to the molecular size $\sigma$ near the temperatures of interest. The derivation of corresponding states assumed $\lambda_T\ll\sigma$ so that quantum effects drop out of the excess free energy, leaving only the two reduced groups $T^*$ and $\rho^*$ . When $\lambda_T\sim\sigma$ that assumption fails: a third dimensionless group, the de Boer quantum parameter $\Lambda^*=h/(\sigma\sqrt{m\varepsilon})$ , survives the rescaling. Helium, hydrogen, and neon all have appreciable $\Lambda^*$ and need a quantum law of corresponding states with $\Lambda^*$ as an extra axis; the difference is so real that ³He and ⁴He, chemically identical but different in mass, do not correspond to each other.

Is corresponding states an exact theorem or an empirical observation?

It is exact under a specific hypothesis and approximate in practice. If every fluid in a class truly had a pair potential $u=\varepsilon\phi(r/\sigma)$ with one shared shape $\phi$ , in the classical limit, then corresponding states would be an exact theorem — the rescaling argument in the second derivation proves it with no approximation. Real potentials only approximately share a shape (argon and methane are close; water is not), so in practice the law is a good approximation for simple spherical molecules and a poor one for polar or quantum fluids. The two-parameter-equation route (van der Waals) is exact for that equation but inherits the equation's own crudeness, which is why it predicts $Z_c=3/8$ instead of the measured $0.29$ .

How is this the ancestor of universality and the renormalization group?

Corresponding states says: fluids that differ only by two scales are thermodynamically the same once rescaled. Universality says something stronger and stranger: near a critical point, fluids that differ in every microscopic detail share the same critical exponents and scaling functions, because they all flow to the same renormalization-group fixed point. The classical law is the mean-field, analytic shadow of this — it gets the collapse right away from criticality but predicts the wrong (mean-field) exponents at criticality. Guggenheim's 1945 data already showed a coexistence exponent near $1/3$ , not the mean-field $1/2$ , hinting that the true universality was different. The renormalization group, later in this track, explains why all simple fluids share the 3D-Ising exponents: corresponding states is where the idea of collapse onto a universal curve first appeared, and universality is its exact, fluctuation-driven completion.

Why is the reduced second virial coefficient a cleaner test than Z_c?

Because $B_2$ is a low-density, exact quantity — it follows directly from the pair potential with no mean-field or two-parameter-equation approximation. The reduced coefficient $B_2/\sigma^3$ is a universal function of $T^*$ for any single-scale potential, so plotting $B_2$ data for argon, nitrogen, and methane against $T^*$ (or $T_r$ ) collapses them onto one curve, and the Boyle temperature falls at a universal reduced value. $Z_c$ , by contrast, is a critical-point quantity where mean-field theory is at its worst, which is exactly why the van der Waals $3/8$ is quantitatively off. So the virial test probes corresponding states in the regime where the derivation is rigorous, while $Z_c$ probes it where the derivation is weakest — both confirm the principle, but the virial test does so more cleanly.