Graham’s Law of Effusion Calculator

Lighter gases get out faster, and exactly how much faster

Through a pinhole, two gases escape at rates in the inverse ratio of the square roots of their molar masses: r1/r2 = √(M2/M1). Hydrogen leaves four times faster than oxygen. The reason is that at a given temperature both gases carry the same average kinetic energy, so the lighter one must be moving faster. This page gives the rate ratio, the time and volume ratios that follow from it, and the uranium-enrichment application in which a ratio of 1.0043 was made to do industrial work.

Ratio of effusion rates

r1/r2 = √(M2/M1)
Molar masses built from IUPAC standard atomic weights. The two uranium hexafluoride entries differ by only 3 g/mol out of 350 and are there deliberately: picking both is the calculation that built the first atomic bombs and still defines the hardest separation anyone does at scale.
The gas you are comparing gas 1 with. The answer is always the rate of gas 1 divided by the rate of gas 2, so a number above 1 means gas 1 is the faster one, which happens exactly when it is the lighter one.
Only read when gas 1 is set to “something else”. Any consistent mass unit works, because only the ratio matters — but keep both boxes in the same unit.
Only read when gas 2 is set to “something else”. A common exam variant gives you the ratio and one molar mass and asks for the other: square the rate ratio and multiply or divide.
Used only for the time row. Rate and time are reciprocal for the same amount of gas, so t2/t1 = r1/r2: the slower gas takes proportionally longer. A great many questions on this topic are really this one sentence, written in a way that invites you to invert the ratio by mistake.
The ratio does not depend on temperature at all — it cancels, which is the point of Graham’s law. The temperature is here only so the two root-mean-square speeds can be printed as real numbers in metres per second, because seeing that hydrogen molecules average about 1900 m/s at room temperature is worth more than another ratio.
Two identical containers, each leaking through a pinhole. The arrows are the two effusion rates drawn to scale, with the faster gas given the full length and the slower one shortened in the ratio Graham's law gives; the escaped molecules are spaced along each arrow, so the lighter gas's are further apart. Both containers are at the same temperature and pressure, which is the condition the law requires.
3.983971Example

hydrogen against oxygen at the same temperature and pressure

Advertisement

One square root, and the kinetic theory behind it

r1/r2 = √(M2/M1)  ·  t2/t1 = r1/r2  ·  urms = √(3RT/M)  ·  n = log F / log(r1/r2)
r
the rate of effusion: amount of gas escaping per unit time through a hole small enough that molecules leave one at a time, without colliding on the way out. Measured as moles per second, or litres per second, or as the reciprocal of the time taken for a fixed amount — all three give the same ratio
M
the molar mass. Only the ratio of the two appears, so the unit does not matter as long as both are in the same one. The square root is the whole content of the law and the whole source of the mistakes
urms
the root-mean-square molecular speed, √(3RT/M), which is where the square root comes from. At a given temperature every gas has the same average kinetic energy ½mu², so a molecule sixteen times heavier moves four times slower — and the rate at which molecules find a hole is proportional to how fast they are going
n
the number of identical stages needed to multiply the abundance ratio of the two gases by an overall factor F. The base of the logarithm does not matter provided both logarithms use the same one, because it cancels; this page uses base ten. n grows without limit as the per-stage ratio approaches 1, which is the entire engineering problem of isotope separation
effusion vs diffusion
effusion is escape through a hole smaller than the mean free path, with no collisions in the hole; diffusion is one gas spreading through another, with collisions at every step. Graham measured both and found the same square-root dependence, which is why the law carries his name twice, but the clean derivation is the effusion one

Worked example

hydrogen against oxygen at the same temperature and pressure
THE TWO MOLAR MASSES. M(H2) = 2.016 and M(O2) = 31.998 g/mol. Note which way up the law goes: the rate ratio of 1 to 2 uses the mass ratio of 2 to 1, upside down relative to the rates.
THE MASS RATIO. M2/M1 = 31.998 / 2.016 = 15.872 02. Oxygen molecules are about sixteen times heavier.
TAKE THE SQUARE ROOT. r1/r2 = √15.872 02 = 3.983 971. Hydrogen effuses almost exactly four times as fast as oxygen — not sixteen times, which is the mistake the square root exists to prevent.
WHERE THE SQUARE ROOT COMES FROM. At the same temperature both gases have the same mean kinetic energy, ½Mu² = (3/2)RT, so u = √(3RT/M). At 298.15 K that is 1920.7 m/s for hydrogen and 482.1 m/s for oxygen — and 1920.7/482.1 = 3.984, the same number. The rate at which molecules happen upon a hole is proportional to how fast they are travelling, so the speed ratio is the rate ratio.
TURN IT INTO A TIME. Rate and time are reciprocal for a fixed amount of gas, so if 60 s of hydrogen effuses, the same amount of oxygen takes 60 × 3.983 971 = 239.04 s, just under four minutes. In the same 60 s only 1/3.983 971 = 0.251 L of oxygen gets out for every litre of hydrogen.
THE HARD CASE, FOR CONTRAST. Put 235UF6 against 238UF6: √(352.041/349.034) = 1.004 298. One pass changes the mixture by four parts in a thousand. To shift the ratio of the two isotopes a hundredfold takes log 100 / log 1.004 298 = about 1074 identical stages, each one a complete apparatus. The same law, a different number, and an entirely different kind of engineering.
Advertisement

Effusion rates relative to oxygen at the same temperature

GasM / g mol⁻¹Rate relative to O2urms at 298.15 KTime to effuse what O2 does in 60 s
hydrogen, H22.0163.983971920.7 m/s15.06 s
helium, He4.0032.827281363.0 m/s21.22 s
methane, CH416.0431.41227680.9 m/s42.48 s
ammonia, NH317.0311.37070660.8 m/s43.77 s
nitrogen, N228.0141.06874515.2 m/s56.14 s
oxygen, O231.9981.00000482.1 m/s60.00 s
carbon dioxide, CO244.0090.85269411.1 m/s70.37 s
sulfur dioxide, SO264.0580.70676340.7 m/s84.89 s
chlorine, Cl270.900.67180323.9 m/s89.31 s
235UF6349.0340.30278146.0 m/s198.16 s
238UF6352.0410.30148145.3 m/s199.02 s
Molar masses from IUPAC standard atomic weights; every other column computed on this page’s own expressions at 298.15 K. Two things to read off it. The whole table spans a factor of only thirteen in rate while spanning a factor of 175 in molar mass — that is the square root at work. And the last two rows differ in the fourth decimal place, which is the entire basis on which uranium is enriched.

Why 1.0043 was enough to build an industry

QuantityValueComment
M of 235UF6349.034 g/mol235.0439 for the uranium plus 113.9904 for six fluorines
M of 238UF6352.041 g/mol238.0508 plus the same 113.9904. Fluorine has one stable isotope, which is exactly why UF6 is the compound used: none of the mass spread comes from the fluorine
Rate ratio, lighter over heavier1.004 298√(352.041/349.034). One pass enriches by four parts in a thousand
Stages for a hundredfold enrichmentabout 1074log 100 divided by log 1.004 298. Natural uranium is 0.72% 235U; reactor fuel is 3 to 5%, weapons material above 90%
What was actually builtthe K-25 plant, Oak Ridge, 1945Gaseous diffusion through porous barriers rather than effusion through a pinhole — the same square-root law with a smaller per-stage factor. The building covered about 170 000 square metres and the plant drew a large fraction of a power station’s output
What replaced itthe gas centrifugeSeparates on the mass difference rather than on its square root, so the per-stage factor is far larger and the stage count and energy use far smaller. Gaseous diffusion is now obsolete
The molar masses and the first four rows are computed from IUPAC nuclide masses on this page’s own expressions. The historical entries are widely published figures given for orientation rather than as sourced measurements, and the plant’s size and power draw are quoted as orders of magnitude. The chemistry — why the compound is UF6, and why a ratio of 1.0043 forces a cascade of a thousand stages — is the part this page is actually about.

Why speed goes as one over the square root of the mass

Put two gases at the same temperature and they have the same average kinetic energy per molecule. That is what temperature means in kinetic theory, and it is not an assumption about the gases but the definition of thermal equilibrium: if one gas had more energy per molecule, collisions at the boundary would move energy until it did not. So ½m1u1² = ½m2u2², and u1/u2 = √(m2/m1). The heavier molecule carries its share of the energy as mass rather than as speed.

Now make a hole small enough that a molecule passing through it does not collide with anything on the way. The number of molecules finding the hole per second is proportional to how many there are per unit volume and to how fast they are moving. If the two gases are at the same pressure and temperature, the first factor is the same for both, so the rate ratio is just the speed ratio. Graham’s law is that sentence, and everything else on this page is arithmetic on top of it.

The hole size matters, and this is where effusion and diffusion part company. Effusion requires a hole smaller than the mean free path — the average distance a molecule travels between collisions, about 70 nm in air at room conditions — so that the gas leaks out molecule by molecule rather than flowing out as a fluid. Through a large hole you get bulk flow, which is governed by pressure and viscosity and not by molar mass at all, and the square-root law simply does not apply. Diffusion is one gas spreading through another, with collisions at every step; Graham measured that too and found the same square-root dependence, which is why his name attaches to both, though the derivation above is the effusion one.

The classic demonstration is worth picturing, because it makes the direction of the effect concrete. A porous pot full of air is connected to a water manometer and a beaker of hydrogen is inverted over it. Hydrogen rushes in through the pores about four times faster than air rushes out, so the pressure inside rises and the manometer jumps. Take the beaker away and the hydrogen now leaks out faster than air leaks in, the pressure falls, and the manometer swings the other way. The same arithmetic also explains why a helium balloon goes soft in a day while an air-filled one does not: helium is both lighter and smaller, and it works its way through the rubber far faster.

The application that made this law famous is at the opposite extreme of difficulty. Uranium-235 and uranium-238 are chemically indistinguishable, so no chemical method separates them; only their mass differs, by about 1.3%. Converted to the hexafluoride — chosen because fluorine has a single stable isotope, so none of the mass spread is wasted on the fluorine, and because UF6 is a gas just above room temperature — the molar masses are 349.03 and 352.04, and the effusion rate ratio is 1.0043. One pass enriches by four parts in a thousand. The response was to build a cascade of thousands of identical stages, each taking the slightly enriched output of the last, which is what the K-25 plant at Oak Ridge was. The method has since been replaced by the gas centrifuge, which separates on the mass difference itself rather than on its square root and therefore needs far fewer stages for the same job.

One limitation worth stating plainly, because it is easy to miss. Graham’s law compares rates at the same temperature and the same pressure. Change either and the individual rates change, even though the ratio does not; and if the two gases are at different pressures the ratio of rates is not given by the molar masses alone. The law is also a statement about an ideal gas, so at the pressures where the compressibility factor has left 1 it is an approximation like everything else here. For the pressures and volumes themselves rather than the ratio of escape rates, that is the combined gas law and PV = nRT in the physics plugin.

Frequently asked questions

Why is it a square root and not just the mass ratio?

Because what effusion depends on is molecular speed, and speed depends on the square root of the mass. At one temperature every gas has the same mean kinetic energy ½mu², so a sixteen-fold heavier molecule is four times slower, not sixteen times. If you ever find yourself writing a rate ratio equal to a mass ratio, the square root has gone missing.

What is the difference between effusion and diffusion here?

Effusion is escape through a hole smaller than the mean free path, so molecules leave individually without colliding in the opening. Diffusion is one gas mixing through another, with collisions at every step. Both follow the same square-root dependence on molar mass and both are called Graham’s law, but only effusion has a clean one-line derivation from kinetic theory. Through a large hole neither applies: that is bulk flow, driven by the pressure difference, and it does not care about molar mass.

An unknown gas effuses 1.5 times faster than oxygen. What is it?

Square the ratio and divide. rx/rO2 = √(32/Mx) = 1.5, so 32/Mx = 2.25 and Mx = 14.2 g/mol. That is close to methane at 16.04, and the small discrepancy is the sort of thing a real experiment gives. Set gas 2 to oxygen and gas 1 to “something else”, then type molar masses until the headline reads 1.5 — or just do the division, which is quicker.

Does the temperature change the answer?

Not the ratio. √(3RT/M) contains the temperature, but it contains the same temperature for both gases, so it cancels when you divide. Heating both gases speeds both up by the same factor and leaves their ratio untouched. The temperature box on this page exists only so the two individual speeds can be printed; change it and watch the speeds move while the headline does not.

Why uranium hexafluoride rather than uranium metal or an oxide?

Three reasons, all of them chemistry. It has to be a gas at a workable temperature, and UF6 sublimes at 56 °C. Fluorine has only one stable isotope, so every gram of mass difference between the two molecules comes from the uranium and none is wasted diluting the signal — chlorine, with two abundant isotopes, would smear the masses and ruin the separation. And uranium forms that hexafluoride readily and reversibly. Even so the rate ratio is only 1.0043, because the fluorines contribute a third of the molar mass and none of the difference.

Related calculators

References

  1. IUPAC Commission on Isotopic Abundances and Atomic Weights, Standard Atomic Weights (2021 table) and the Atomic Mass Evaluation for single-nuclide masses. The molar masses on this page are built from those values: for example 19F at 18.998 403 u gives UF6 a fluorine contribution of 113.990 4 g mol−1, so 235UF6 is 349.034 and 238UF6 is 352.041 g mol−1. Cited from the published tables; no online copy was fetched while writing.
  2. P. Atkins, J. de Paula and J. Keeler, Atkins’ Physical Chemistry, Oxford University Press. The standard undergraduate treatment of colligative properties, of the van der Waals equation and of the compressibility factor, including the derivation of the critical constants from a and b and the definition of the Boyle temperature. Cited from the printed text; no online copy was fetched.
  3. Chemistry, Textbook for Class XI, Part II, Unit “States of Matter”, NCERT, New Delhi, and the kinetic molecular theory section that gives urms = √(3RT/M). Graham’s law is the direct consequence. Cited from the printed syllabus text; no online copy was fetched while writing this page on 14 November 2026.
  4. CODATA 2022 recommended values of the fundamental physical constants, NIST Standard Reference Database 121. R = 8.314 462 618 153 24 J mol−1 K−1 is exact in the 2019 SI, because the kelvin and the mole are defined through the Boltzmann and Avogadro constants. In litre-atmospheres, R = 8.314 462 618 153 24 / 101 325 = 0.082 057 366 08 L atm mol−1 K−1, and that division is exact too, because the standard atmosphere is defined as exactly 101 325 Pa. The NIST wallet-card PDF would not load when this page was written on 14 November 2026, so the value is quoted from the 2019 SI definitions rather than from a fetched copy of the table.