Graham’s Law states that the rate at which a gas escapes through a small opening is inversely proportional to the square root of its molar mass. A light gas such as helium diffuses about 2.6 times faster than nitrogen under the same conditions, because the ratio of rates equals the square root of the ratio of masses. This single relationship explains isotope enrichment, gas leaks, and the limits of effusion-based separation.

The law works because all gas molecules carry the same average kinetic energy at a given temperature. Lighter molecules must therefore move faster to hold that energy. When they reach a pinhole, the faster ones strike it more often, so more of them pass through per unit time. That is effusion: escape through a hole smaller than the mean free path between collisions. Graham’s Law treats effusion and diffusion alike, though the two obey the same square-root scaling for different reasons.

Thomas Graham published the empirical relationship in 1846 after measuring how different gases passed through plaster of paris. He noticed the rates scaled with the inverse square root of density. Kinetic theory later explained why. The law is a first-order approximation, not an exact description, and its failures are as instructive as its successes.

The square-root mechanism

At a fixed temperature, the molecules of any gas share an average translational kinetic energy of (3/2)kT per molecule, where k is Boltzmann’s constant and T is absolute temperature. Since kinetic energy is ½mv², a molecule with one-quarter the mass must move at twice the speed to carry the same energy. Speed scales as 1/√m, and so does the effusion rate.

Effusion adds a second condition. The hole must be small compared with the mean free path, the average distance a molecule travels between collisions. Under that condition, molecules cross the opening without colliding with each other in transit, so each molecule behaves independently. The flux through the hole is proportional to the number density times the mean molecular speed projected toward the aperture. Density and temperature are common to both gases in a comparison, so the rate ratio collapses to √(M₂/M₁).

Diffusion, mixing driven by concentration gradients, follows the same square-root dependence at the level of molecular speeds. But bulk diffusion also depends on collision frequency between different species. The precise diffusion coefficient carries an extra term for the colliding partners, so Graham’s Law is a scaling rule rather than a complete transport equation.

One detail separates effusion from diffusion cleanly. Effusion is a mass-transport process through a physical barrier; diffusion is mixing without one. The square-root scaling emerges from both, which is why the law gets applied to situations where only one mechanism is actually at work. A plain-language version: lighter gas wins the race to a hole, and the race is unfair by exactly the square root of the mass ratio.

Quantified context: how big the advantage is

The rate ratio between two gases depends only on their molar masses. Helium (4 g/mol) versus nitrogen (28 g/mol) gives √7, about 2.65. Hydrogen (2 g/mol) versus oxygen (32 g/mol) gives √16, exactly 4. Carbon dioxide (44 g/mol) versus nitrogen gives √(44/28), about 1.25. Uranium hexafluoride is the industrial case: the gas holds one uranium atom, and the two isotopes differ by only three neutrons.

Natural uranium is about 0.7 percent uranium-235 and 99.3 percent uranium-238. The hexafluoride molar masses are roughly 349 g/mol and 352 g/mol, giving a single-stage enrichment ratio of √(352/349), about 1.0043. That is a 0.43 percent edge per pass. The first enrichment plants chained thousands of stages to lift the uranium-235 fraction to weapons grade. The math is unforgiving: you need many stages because each one barely changes the composition.

Temperature matters only through the speed constant, and it cancels in ratio comparisons between two gases at the same temperature. If you compare the same gas at two temperatures, the ratio scales as the square root of the absolute temperature ratio. Doubling absolute temperature raises the molecular speed by only √2, about 1.41. That is a modest lever compared with choosing a lighter gas.

Gas pair Molar masses (g/mol) Rate ratio Practical read
H₂ vs O₂ 2 vs 32 4.00 Fourfold edge, the largest common pairing
He vs N₂ 4 vs 28 2.65 Helium leak detection baseline
H₂ vs N₂ 2 vs 28 3.74 Fastest common carrier gas
He vs O₂ 4 vs 32 2.83 Light gas against heavy gas
N₂ vs CO₂ 28 vs 44 1.25 Modest contrast, easy to miss
²³⁵UF₆ vs ²³⁸UF₆ ~349 vs ~352 1.0043 One stage barely moves the needle

Edge cases where the law bends

The square-root rule assumes ideal gas behavior, a single species pair, and no intermolecular forces. Real gases depart from it. Heavy gases with strong attraction, such as sulfur hexafluoride (146 g/mol), show measurable deviations at high pressure. The law also breaks when the aperture is not small relative to the mean free path: that is viscous flow, not effusion, and the square-root scaling disappears.

Mixtures add a second complication. In a gas mixture, the effective molar mass of the escaping stream shifts as the lighter component depletes. The composition of the effusing gas changes over time. A container holding a hydrogen–nitrogen mix loses hydrogen preferentially, so the remaining mix becomes heavier and the flow rate declines. That time course is a fingerprint of effusion.

Graham’s Law is often quoted for leaks, but a real pinhole leak in a pressurized vessel usually sits in a continuum-flow regime where the flow rate depends on pressure, viscosity, and hole geometry. The square-root mass dependence reappears only in the free-molecular regime. A technician checking a helium leak detector should treat the readout as a calibration value, not an exact mass ratio inference.

The professional caveat: Graham’s Law gives ratios, not absolute rates. To compute an actual flux you also need the aperture area, the number density, and the temperature. The law tells you how two gases compare, and it tells you nothing about how fast a single gas crosses a given hole without those additional parameters. That distinction trips up students who try to use it as a complete flow equation.

Check the flow regime before applying the square root: free-molecular flow requires hole size below the mean free path. Expect ratio-only results; absolute flux needs area, density, and temperature. Treat leaking mixtures as time-varying, because the light fraction leaves first. Do not apply the law to liquid transport or to dissolved gases across membranes. Remember that real gases deviate near condensation and at high pressure.

Misconceptions worth clearing

The most common error is treating Graham’s Law as a statement about molecular size. The law tracks mass, not diameter. Two gases with similar masses but different molecular diameters effuse at nearly the same rate, because mass sets the mean speed. Size enters only through the aperture-to-mean-free-path condition that decides whether you are in the effusion regime at all.

A second misconception conflates effusion with diffusion. In diffusion through a tube or across a barrier, the transport also depends on collisions between species, so the rate ratio is only approximately √(M₂/M₁). The square root appears in the Maxwell–Boltzmann speed distribution, from which both processes derive, but the prefactors differ. The law is a scaling relation, and treating it as an exact rate formula overpromises.

A third error is assuming temperature drives the ratio. Between two gases at equal temperature, temperature cancels. Between the same gas at two temperatures, the effect is √(T₂/T₁) — much weaker than a mass change. This is why thermal methods are not the first choice when you want to separate gases by mass; you choose a lighter carrier or a lighter species instead.

The counterintuitive detail: the lightest gas does not always dominate a mixture’s effusion. In a mixture of hydrogen and helium (2 and 4 g/mol), the rate ratio is only about 1.41, smaller than either gas versus a heavy partner. The separation between similar-mass light gases is subtle, and the law's power grows when the mass ratio is large. By contrast, hydrogen versus heavy gases offers the widest contrast and the strongest separation.

Practical actions and thresholds

The law is a design and troubleshooting tool. Use it to predict which gas leaves a vessel first, to choose a tracer gas, and to estimate how many stages a separation needs. The working threshold is the mass ratio: below about 1.05, a single stage separates almost nothing, which is why uranium enrichment needs thousands of passes. At ratios above 2, the separation per stage becomes visually obvious.

For leak detection, choose helium as the tracer when you need sensitivity, because its high effusion rate against air (about 2.7) makes small leaks register faster. For purging a system, nitrogen and argon behave similarly per unit time because their masses are close, so a purge designed for nitrogen will not scale to helium without recalculating. For any separation process, estimate the per-stage enrichment as √(M₂/M₁) and multiply stages accordingly.

Concrete steps: first, identify the molar masses of the two species at issue. Second, compute the ratio √(M₂/M₁) to get the per-pass separation. Third, confirm the aperture or barrier is small enough that effusion rather than viscous flow governs. Fourth, plan the number of stages from the target composition. Fifth, account for the composition drift as the light species depletes. These steps convert the law from a textbook formula into a sizing calculation.

Time point What happens in an effusing mixture
t = 0 Both gases start at their initial fractions
Early stage Light gas leaves fastest; escaping stream enriched in it
Mid stage Remaining container mix grows heavier; flow rate declines
Late stage Light gas nearly depleted; rate approaches heavy-gas value
Completion Effusing total matches original composition weighted by rate

Limits and what remains uncertain

Graham’s Law is exact only for ideal gases in free-molecular flow. Real mixtures, real apertures, and real pressures all introduce corrections that the simple formula ignores. The law says nothing about the absolute rate, the effect of aperture geometry, or the coupling between effusing species in a dense mixture. For those you need kinetic theory transport equations, such as the Chapman–Enskog approach, which the law approximates.

The numbers that make the law useful are stable: molar masses, the square-root ratios, and the ideal-gas limit do not change with policy or market conditions. Industrial isotope enrichment is a well-documented application, but current enrichment capacities and prices are not something this article should assert without a cited source. What can be said with confidence is the mechanism and the scaling, not the present-day market.

A final limit is conceptual. The law describes rates, not equilibria. It tells you how fast gases leave a container, not where they settle. Separation by effusion is a kinetic process, and a system left long enough will re-mix. That is why enrichment plants run continuously rather than in a single batch. The law is a tool for the transient, not a guarantee of a final state.