Law of Corresponding States

The idea: every gas keeps its own ruler

Two gases can look nothing alike in absolute units. Helium turns to liquid only below 5.2 K, while water needs to be heated past 647 K before its liquid and vapor become indistinguishable. Push them to the same absolute temperature — say 300 K — and one is a wispy, almost-ideal gas while the other is dense steam on the edge of condensing. They seem to follow completely different rules.

The law of corresponding states makes a startling claim: that difference is almost entirely a matter of scale. Every fluid has one special point — the critical point, where the liquid and gas phases merge into a single supercritical fluid — defined by a critical temperature Tc, critical pressure Pc, and critical molar volume Vc. Use that point as each substance's personal ruler:

Reduced temperature   Tr = T / Tc
Reduced pressure      Pr = P / Pc
Reduced volume        Vr = V / Vc

Now compare gases not at "300 K" but at "Tr = 1.2" — twenty percent above their own critical temperatures. When you do this, the differences nearly vanish. Nitrogen at Tr = 1.5 and Pr = 2 is compressed by almost exactly the same fraction as argon, methane, or carbon dioxide at Tr = 1.5 and Pr = 2. Plot any of them and the points land on a single shared curve. The substance-specific scale has been divided out, leaving a behavior that is universal.

How the scaling collapses every gas onto one curve

The cleanest way to measure non-ideal behavior is the compressibility factor:

Z = PV / (nRT)

For an ideal gas Z = 1 exactly. When attractions dominate (moderate pressures), molecules pull together and Z dips below 1. When repulsions dominate (very high pressures, where the molecules' own volume matters), Z climbs above 1. The whole story of a real gas is the shape of Z as you change conditions.

The law of corresponding states says that Z is not a function of P and T separately for each gas — it is a single function of the reduced variables, the same function for everyone:

Z = Z(Tr, Pr)      — same surface for (almost) all fluids

That is the entire content of the law in one line. It means a chemical engineer who knows only the critical constants of a brand-new fluid — and nothing else — can read its compressibility off a chart built from argon and nitrogen data, and be right to within a few percent. The chart, in this two-parameter form, is the generalized compressibility (Nelson–Obert) chart, and it works because the chart's axes are Tr and Pr, not T and P.

The van der Waals derivation: where the constants cancel

The law isn't an empirical accident; it falls straight out of any two-parameter equation of state. Take the van der Waals equation, which corrects the ideal gas for molecular attraction (the a term) and finite molecular size (the b term):

( P + a/V² )( V − b ) = RT       (per mole)

The critical point is the inflection where the first and second derivatives of P with respect to V both vanish. Solving those two conditions ties the critical constants directly to a and b:

Vc = 3b        Pc = a / (27 b²)        Tc = 8a / (27 R b)

Now substitute P = Pr·Pc, V = Vr·Vc, T = Tr·Tc back into the equation and watch what happens. Every appearance of a, b, and R collects into the critical constants and cancels. What is left contains no molecule-specific number at all:

( Pr + 3/Vr² )( 3Vr − 1 ) = 8 Tr      ← the reduced equation of state

This single equation has no a, no b, no R. Two fluids with utterly different intermolecular constants obey the identical reduced equation. That is the theorem of corresponding states, proved in two lines: any equation of state with exactly two adjustable constants, scaled by its own critical point, becomes universal.

A direct consequence: the critical compressibility

Zc = Pc·Vc / (R·Tc)

must be the same number for every fluid. Van der Waals predicts Zc = 3/8 = 0.375. The fact that real fluids sit near 0.27–0.29 instead is the first crack in the simple law — and the doorway to the corrections below.

Real critical constants and the Zc spread

Here is where theory meets the lab. The two-parameter law demands a single universal Zc; the measured values show how close reality comes — and which molecules misbehave.

Fluid	Tc (K)	Pc (bar)	Zc	Acentric ω
Helium-4	5.2	2.27	0.301	−0.39
Argon	150.7	48.6	0.291	0.000
Krypton	209.4	55.0	0.288	0.000
Nitrogen	126.2	33.9	0.289	0.039
Methane	190.6	46.0	0.286	0.011
Carbon dioxide	304.1	73.8	0.274	0.224
n-Octane	568.7	24.9	0.259	0.398
Ammonia	405.5	113.5	0.244	0.253
Water	647.1	220.6	0.229	0.344

Read down the Zc column. The simple noble gases argon, krypton and the near-spherical methane and nitrogen cluster tightly at 0.286–0.291 — for these, two parameters are essentially exact. As molecules get long and floppy (n-octane) or strongly polar and hydrogen-bonding (ammonia, water), Zc slides down toward 0.23. The rightmost column shows why: the acentric factor ω, which is near zero for the noble gases, grows large for exactly the fluids whose Zc drifts. The law isn't broken; it just needs to sort fluids along one more axis.

The three-parameter fix: Pitzer's acentric factor

In 1955 Kenneth Pitzer recognized that the deviation from corresponding states tracks one geometric idea: how far a molecule's force field departs from a tiny sphere. He captured it in a single number read off the vapor-pressure curve at one reduced temperature:

ω = − log₁₀( Pr_sat )  − 1        evaluated at Tr = 0.70

Spherical noble gases follow a saturation curve that passes through Pr_sat ≈ 0.1 at Tr = 0.7, which makes ω = 0 by construction. Argon, krypton, xenon: ω ≈ 0. Elongated or polar molecules have higher vapor pressures relative to their critical pressure, so ω rises — methane 0.011, CO₂ 0.224, water 0.344, n-octane 0.398.

Pitzer then wrote the compressibility as a leading universal term plus a first correction weighted by ω:

Z = Z⁰(Tr, Pr) + ω · Z¹(Tr, Pr)

Z⁰ is the "simple-fluid" surface tabulated from argon-like data; Z¹ is a universal correction surface, also tabulated. Plug in a fluid's ω and you recover its compressibility to better than 1–2% for normal substances. This three-parameter corresponding-states correlation is what generalized charts and modern equations of state actually use. The acentric factor is the price of admission for non-spherical molecules onto the universal chart.

Worked example: estimating the density of an unknown gas

Suppose you need the molar volume of nitrogen at T = 189 K and P = 67.8 bar and you have no N₂ data table — only its critical constants from above, Tc = 126.2 K and Pc = 33.9 bar.

Step 1 — reduce the conditions:
   Tr = 189 / 126.2 = 1.50
   Pr =  67.8 / 33.9 = 2.00

Step 2 — read Z from the two-parameter chart at (Tr = 1.50, Pr = 2.00):
   Z ≈ 0.86      (attractions still pulling Z below 1)

Step 3 — solve for molar volume:
   V = Z·R·T / P
     = 0.86 · 0.08314 L·bar·mol⁻¹·K⁻¹ · 189 K / 67.8 bar
     = 0.199 L/mol

Compare ideal gas (Z = 1):
   V_ideal = R·T / P = 0.232 L/mol  →  16% too large

The experimental molar volume of nitrogen at these conditions is about 0.198 L/mol — the corresponding-states estimate nails it to roughly half a percent, while the ideal-gas law is off by 16%. We obtained this without a single nitrogen data point: only its critical temperature and pressure, plus a chart built from other gases. That is the practical magic of the law.

Where the law earns its keep

• Natural-gas pipelines. Field engineers cannot tabulate every methane–ethane–CO₂–N₂ mixture. Instead they compute pseudo-critical Tc and Pc as mole-fraction averages (Kay's rule), reduce by them, and read Z off a generalized chart to get compressibility, density, and flow rates. The entire metering of how much gas crosses a custody-transfer point leans on corresponding states.
• Cubic equations of state. The Soave–Redlich–Kwong (1972) and Peng–Robinson (1976) equations — the engines inside Aspen Plus, HYSYS, and every refinery simulator — are corresponding-states equations: their a(T) and b are fixed entirely by Tc, Pc, and ω. Phase equilibria for distillation columns separating thousands of mixtures all trace back to this law.
• Cryogenics and supercritical processing. Designing a hydrogen or helium liquefier, or a supercritical-CO₂ decaffeination plant operating near Tr ≈ 1, relies on departure functions (enthalpy and entropy corrections) that are themselves tabulated in reduced coordinates — one chart serving every working fluid.
• Estimating data that was never measured. For a newly synthesized refrigerant or solvent, you may know only its boiling point and a rough critical estimate. Corresponding states gives first-pass densities, viscosities (via reduced-property correlations), and fugacities good enough for preliminary design before any experiment is run.

Corresponding states vs the ideal gas and van der Waals

	Ideal gas law	Van der Waals	Corresponding states
Equation	PV = nRT	(P + a/V²)(V − b) = RT	Z = Z(Tr, Pr)
Inputs per fluid	None	a, b (two constants)	Tc, Pc (and ω)
Captures liquefaction	No	Qualitatively, yes	Yes, via the chart
Universal?	Yes, but wrong near Tc	No — a, b differ per fluid	Yes (2-param) / near (3-param)
Critical Zc	1 (no critical point)	0.375 for all fluids	0.27–0.29 measured
Accuracy near Tc	Poor (errors > 20%)	~10–15%	1–2% with ω
Best use	Dilute gases, Pr ≪ 1	Teaching the mechanism	Real-process estimation

The ideal gas law has no critical point and no notion of attraction, so it fails badly anywhere near condensation. Van der Waals introduces the mechanism and even predicts a universal Zc — but the wrong value (0.375). Corresponding states keeps van der Waals's universality while replacing its hard-coded constants with real, measured critical data, which is why it survives in engineering practice where van der Waals alone does not.

Common misconceptions and pitfalls

• "It's an exact law of nature." It is not. It is an approximate consequence of assuming all molecules share one shape of intermolecular potential (a scaled Lennard-Jones well). The two-parameter version is excellent for noble gases and small symmetric molecules and steadily worse for polar, hydrogen-bonding, or quantum (He, H₂) fluids.
• Forgetting to use absolute temperature. Reduced temperature is Tr = T/Tc with both in kelvin. Dividing 25 °C by a critical temperature in kelvin is a classic and silent error.
• Expecting one universal Zc and finding 0.375. The 3/8 figure is the van der Waals prediction, not the measurement. Real fluids cluster near 0.27–0.29, which is why the simple chart is good but the acentric-factor correction exists.
• Mishandling mixtures. A mixture has no real critical point in the single-component sense. You must build pseudo-critical constants (e.g. Kay's mole-fraction rule) before reducing — using a pure-component Tc for a mixture gives nonsense.
• Quantum gases break it from the other end. Helium and hydrogen have negative acentric factors (He ω = −0.39) because zero-point motion makes them more ideal than a classical sphere. They need quantum corrections, not just an ω shift, to rejoin the chart.
• Confusing reduced volume with the chart's volume axis. Generalized charts plot Z against Pr with Tr as a parameter; the "ideal reduced volume" Vr,ideal = V·Pc/(R·Tc) is sometimes used to avoid the implicit-solve for V. Mixing the two conventions for Vr is a frequent source of factor-of-Zc errors.