Mean free path

Take a single molecule of nitrogen sitting inside the air in this room. If you could paint a trail behind it, you would not see a straight arrow shooting across the room — you would see a frantic, zig-zagging scribble, thousands of short straight segments joined at sharp angles. Between one segment and the next, the molecule slams into another molecule and bounces off in a new direction, only to travel a tiny distance before it happens again. That average length of one straight segment — the average distance a molecule manages to cover between two successive collisions — is called the mean free path, and it is the last big idea in the CBSE Kinetic Theory chapter. It sounds like a small, technical quantity, but it answers a question that should already be nagging you. The kinetic theory of gases tells you that air molecules at room temperature are hurtling around at roughly $500\ \text{m/s}$ — faster than a rifle bullet. So why does a whiff of perfume take tens of seconds, not milliseconds, to cross a room? The rifle-bullet speed is real, but it is the speed between collisions, and collisions happen astonishingly often. The mean free path is the number that reconciles the two facts: it turns out to be only about a ten-millionth of a metre, so a molecule changes direction roughly a billion times every second, and its net progress across the room is a slow, drunken stagger rather than a straight dash. This article builds the mean free path formula from a simple geometric picture — a molecule sweeping out a cylinder as it moves — fixes the one subtle correction factor that examiners love to test, and works through the same numerical example the chapter is built around so you can compute $\lambda$ confidently for any gas.

Key takeaways

The mean free path $\lambda$ is the average distance a gas molecule travels between two successive collisions.
A moving molecule of diameter $d$ effectively sweeps out a cylinder of cross-sectional area $\sigma = \pi d^2$ (the collision cross-section); it collides whenever another molecule's centre enters this cylinder.
The formula, accounting for the fact that every other molecule is also moving, is $\lambda = \dfrac{1}{\sqrt2\,\pi d^2 n}$ , where $n = N/V$ is the number density of molecules.
Using $n = P/(k_BT)$ , the same result can be written $\lambda = \dfrac{k_BT}{\sqrt2\,\pi d^2 P}$ — mean free path rises with temperature and falls with pressure at fixed molecular size.
For air at STP, $\lambda \approx 9\times10^{-8}\ \text{m}$ — a few hundred molecular diameters — which is why gases diffuse slowly despite molecules moving at hundreds of metres per second.
The commonest mark-losing slip is dropping the $\sqrt2$ factor, which makes every computed $\lambda$ too large by about $41\%$ .
Key idea: a moving molecule sweeps a cylinder of cross-section $\pi d^2$ , and correcting for the relative motion of all the other moving molecules with a factor of $\sqrt2$ gives the mean free path $\lambda = 1/(\sqrt2\,\pi d^2 n)$ , which grows with temperature and shrinks with pressure and molecular size.

The core insight

The mean free path $\lambda = 1/(\sqrt2\,\pi d^2 n)$ is the average distance a molecule of diameter $d$ travels between collisions in a gas of number density $n$ , the $\pi d^2$ coming from the cylindrical collision cross-section it sweeps out and the $\sqrt2$ correcting for the fact that the molecules it collides with are moving too.

Where it fits in the syllabus

Mean free path is the closing topic of the CBSE Class 11 Kinetic Theory chapter, coming right after the equipartition-of-energy and specific-heat results that this section's prerequisite covers. Having already used kinetic theory to explain pressure and temperature as bulk, statistical outcomes of molecular motion, the chapter turns here to a genuinely molecular question — how far, on average, does one molecule get before it collides with another? The prescribed formula is exactly $\lambda = \dfrac{1}{\sqrt2\,\pi d^2 n}$ , along with Avogadro's number $N_A = 6.022\times10^{23}$ for converting between moles and molecule counts. The full Maxwell–Boltzmann speed distribution, from which a more careful mean free path can in principle be derived, is explicitly out of scope for CBSE 2025-26 — you are expected to know and apply this one formula, understand where each symbol and factor comes from qualitatively, and connect it to pressure and temperature through the ideal gas equation, not to derive it from the full distribution.

The physical picture

Picture one molecule, modelled as a hard sphere of diameter $d$ , flying through a gas in which every other molecule is also a sphere of diameter $d$ . A collision happens the instant the centres of two spheres come within a distance $d$ of each other — not when their surfaces first meet at $d/2$ , but when the two centres are separated by a full diameter $d$ , because that is when the two solid spheres of radius $d/2$ each first touch. So, as our tracked molecule travels, imagine a circle of radius $d$ drawn around its centre, swept forward along its path: any other molecule whose centre lies inside that swept-out cylinder gets hit.

\sigma = \pi d^2

where: $\sigma$ is the collision cross-section, the effective target area (in $\text{m}^2$ ) that the moving molecule presents for a collision, and $d$ is the molecular diameter. Notice this uses the full diameter, not the radius — a frequent source of a factor-of-4 error if you carelessly reach for $\pi r^2$ with $r = d/2$ instead. If the molecule travels a distance $L$ , it sweeps out a cylindrical volume $\sigma L = \pi d^2 L$ . On average, this cylinder should contain exactly one other molecule by the time a collision occurs, so $n\,(\pi d^2 L) \approx 1$ , giving a first estimate $L \approx 1/(\pi d^2 n)$ . This captures the geometry correctly but treats every other molecule as sitting still while only the tracked one moves — in a real gas, all molecules are moving simultaneously, and a proper treatment of the relative speed between two randomly moving molecules brings in an extra factor of $\sqrt2$ in the denominator. Putting this together gives the mean free path:

\lambda = \frac{1}{\sqrt{2}\,\pi d^2 n}

where: $\lambda$ is the mean free path (m), $d$ is the molecular diameter (m), and $n = N/V$ is the number density of molecules (molecules per $\text{m}^3$ ). Physically, this says $\lambda$ shrinks if the gas is denser (larger $n$ , more targets to hit) or if the molecules are bigger (larger $d$ , bigger targets), and it climbs if either falls. Since the ideal gas equation in molecular form is $PV = Nk_BT$ (equivalent to $PV = nRT$ with $R = N_Ak_B$ ), the number density can be written $n = N/V = P/(k_BT)$ , so the mean free path in terms of pressure and temperature is:

\lambda = \frac{k_BT}{\sqrt2\,\pi d^2 P}

where: $k_B = 1.38\times10^{-23}\ \text{J/K}$ is Boltzmann's constant, $T$ is the absolute temperature (K), and $P$ is the pressure (Pa). This second form is the one you reach for whenever a problem gives you $P$ and $T$ instead of a number density directly — heating a gas at fixed pressure makes it less dense, so $\lambda$ grows; compressing a gas at fixed temperature packs molecules closer, so $\lambda$ shrinks.

Figure

A molecule sweeping a cylinder of cross section pi d squared through a field of other molecules, colliding with the ones whose centres lie inside the swept cylinder and missing the ones outside it

Derivation AGetting the mean free path formula from the swept cylinder

The goal of this derivation is to justify, geometrically and dimensionally, the prescribed formula $\lambda = 1/(\sqrt2\,\pi d^2 n)$ , starting from the simple picture of a molecule sweeping out a cylinder and colliding with whatever lies inside it, and then folding in the one physical correction — that the "target" molecules are moving too — that turns the naive estimate into the exact factor of $\sqrt2$ quoted in the NCERT syllabus. No calculus is needed; the whole argument is geometry, an average, and one physically motivated correction factor.

Step 1. Set up the swept cylinder. As the tracked molecule (diameter $d$ ) travels a distance $L$ in a straight line, it collides with any other molecule whose centre lies within a distance $d$ of its own centre's path — because two spheres of diameter $d$ touch when their centres are $d$ apart. Effectively, the molecule sweeps out an imaginary cylinder of cross-sectional area $\sigma = \pi d^2$ and length $L$ , and it collides with (on average) every molecule whose centre falls inside that cylinder.

Step 2. Count the expected number of collisions. If the gas has number density $n$ molecules per unit volume, the swept cylinder of volume $\sigma L = \pi d^2 L$ contains, on average, $n\,\pi d^2 L$ molecules — and each one represents a collision. Setting this equal to one collision defines the distance over which, on average, exactly one collision occurs: $n\,\pi d^2 L = 1$ , so $L = 1/(\pi d^2 n)$ .

Step 3. Recognise the missing physics. This estimate silently assumes the target molecules sit frozen in place while only the tracked molecule moves, so $L$ is really an estimate of the distance travelled relative to a stationary background. In reality every molecule in the gas is moving with comparable speed and in random directions, so what actually matters for how often collisions happen is the relative speed between the tracked molecule and whichever molecule it is about to hit, not its speed relative to the container walls.

Step 4. Bring in the relative-speed correction. For two molecules moving with the same average speed $\bar v$ in random, uncorrelated directions, the average relative speed between them works out (by combining their velocity vectors over all possible relative directions) to be $\sqrt2\,\bar v$ , larger than the speed of either alone by exactly the factor $\sqrt2$ . Because collisions happen at the rate set by this larger relative speed, the true mean free path is the naive estimate divided by this factor: $\lambda = \dfrac{L}{\sqrt2} = \dfrac{1}{\sqrt2\,\pi d^2 n}$ .

Step 5. Read off the final result and check its behaviour. $\lambda = \dfrac{1}{\sqrt2\,\pi d^2 n}$ has units of $\text{m}^2 \times \text{m}^{-3}$ inverted, i.e. metres — correct for a length. It falls as $n$ or $d$ increases (more or bigger targets, more frequent collisions) and is independent of the molecule's own speed, exactly as a purely geometric, statistical quantity should be — the mean free path is about how far you get, not how fast you get there.

Worked examples

Board level - mean free path of nitrogen at STP

Nitrogen gas at STP has number density $n = 2.7\times10^{25}\ \text{m}^{-3}$ and molecular diameter $d = 3.0\times10^{-10}\ \text{m}$ . Find the mean free path.

Step 1. Write down the formula and the values to substitute: $\lambda = \dfrac{1}{\sqrt2\,\pi d^2 n}$ , with $d = 3.0\times10^{-10}\ \text{m}$ and $n = 2.7\times10^{25}\ \text{m}^{-3}$ .

Step 2. Compute $d^2$ : $d^2 = (3.0\times10^{-10})^2 = 9.0\times10^{-20}\ \text{m}^2$ .

Step 3. Compute $\pi d^2$ : $\pi d^2 = 3.1416 \times 9.0\times10^{-20} = 2.827\times10^{-19}\ \text{m}^2$ — this is the collision cross-section.

Step 4. Multiply in $\sqrt2$ and $n$ : $\sqrt2\,\pi d^2 n = 1.414 \times 2.827\times10^{-19} \times 2.7\times10^{25} = 1.414 \times 2.827 \times 2.7 \times 10^{6} = 10.79\times10^{6} = 1.079\times10^{7}\ \text{m}^{-1}$ .

Step 5. Invert to get $\lambda$ : $\lambda = \dfrac{1}{1.079\times10^{7}} \approx 9.27\times10^{-8}\ \text{m}$ . The mean free path of nitrogen at STP is about $9.3\times10^{-8}\ \text{m}$ , roughly 93 nanometres — a few hundred times the molecular diameter itself.

Competitive level - how mean free path scales when temperature doubles at fixed pressure

If the temperature of the same nitrogen gas is doubled while its pressure is kept constant, and molecular diameter $d$ is unchanged, find (a) the factor by which the mean free path changes and (b) the new value of $\lambda$ , using the STP value $\lambda_1 = 9.27\times10^{-8}\ \text{m}$ as the starting point.

Step 1. Choose the pressure–temperature form, since $P$ is now the quantity held fixed: $\lambda = \dfrac{k_BT}{\sqrt2\,\pi d^2 P}$ . With $d$ , $P$ , $k_B$ and the geometric factor all constant, this reduces to a direct proportionality, $\lambda \propto T$ .

Step 2. Apply the proportionality to the doubled temperature: with $T_2 = 2T_1$ , $\dfrac{\lambda_2}{\lambda_1} = \dfrac{T_2}{T_1} = 2$ , so the mean free path doubles.

Step 3. Explain the mechanism behind the ratio, not just quote it: at constant pressure, $PV = Nk_BT$ shows doubling $T$ doubles $V$ for the same $N$ , which halves the number density $n = N/V$ . Since $\lambda \propto 1/n$ , halving $n$ doubles $\lambda$ — the same conclusion reached from the geometric route, a useful cross-check.

Step 4. Compute the numerical value using the STP result as the starting point: $\lambda_2 = 2\lambda_1 = 2\times9.27\times10^{-8} = 1.85\times10^{-7}\ \text{m}$ . Answer: the mean free path doubles, from $9.27\times10^{-8}\ \text{m}$ to about $1.85\times10^{-7}\ \text{m}$ . Timing note: under a minute once you recognise that the question is asking for a ratio, not a fresh calculation from scratch. The trap is reaching for the $n$ -form of the formula and trying to compute a new number density from scratch — the $P,T$ -form makes the proportionality $\lambda \propto T$ (at fixed $P$ ) immediate, and reusing the already-computed STP value turns the whole problem into one multiplication.

Common traps

Board exams repeatedly dock marks on this topic for a small set of recurring slips, each traceable to skipping the physical picture and just memorising the final formula.

Dropping the $\sqrt2$ . The single most common error is writing $\lambda = 1/(\pi d^2 n)$ without the $\sqrt2$ , which overestimates $\lambda$ by a factor of $\sqrt2 \approx 1.41$ (about $41\%$ too large). The $\sqrt2$ is not a rounding convenience — it comes from properly averaging the relative speed of two moving molecules, as shown in the derivation above, and must appear in every substitution.
Using the radius instead of the diameter in the cross-section. The collision cross-section is $\sigma = \pi d^2$ , using the full molecular diameter, because two spheres of diameter $d$ touch when their centres are separated by $d$ , not $d/2$ . Writing $\pi(d/2)^2 = \pi d^2/4$ instead makes every computed $\lambda$ four times too large.
Treating $\lambda$ as a fixed property of a gas. Mean free path is not a constant like molecular diameter — it depends on the state of the gas through $n$ (equivalently, through $P$ and $T$ ). The same gas has a longer $\lambda$ at low pressure or high temperature and a shorter one when compressed or cooled.
Confusing mean free path with molecular speed. A long mean free path does not mean molecules move faster — it means they travel farther, on average, before a collision. The rms speed depends only on $T$ and molar mass $M$ (via $v_{rms} = \sqrt{3RT/M}$ ) and is completely independent of pressure; $\lambda$ , by contrast, depends strongly on pressure through $n$ .
Mixing up units when substituting $P$ and $T$ . In $\lambda = k_BT/(\sqrt2\,\pi d^2 P)$ , $T$ must be in kelvin and $P$ in pascals (SI), with $k_B$ in J/K — plugging in pressure in atmospheres or temperature in Celsius silently gives an answer wrong by orders of magnitude.

You should be able to…

State the mean free path formula $\lambda = \dfrac{1}{\sqrt2\,\pi d^2 n}$ and correctly identify every symbol, including why $d$ is the diameter and not the radius.
Explain, using the swept-cylinder picture, why the collision cross-section is $\pi d^2$ and why the $\sqrt2$ factor appears from relative molecular motion.
Rewrite the formula in terms of pressure and temperature as $\lambda = \dfrac{k_BT}{\sqrt2\,\pi d^2 P}$ using $n = P/(k_BT)$ , and know when to use each form.
Calculate $\lambda$ numerically given either the number density $n$ or the pair $(P, T)$ , together with the molecular diameter $d$ .
Reason qualitatively about how $\lambda$ changes when pressure, temperature, or molecular size are varied, without recomputing from scratch every time.
Distinguish mean free path (a distance between collisions) from molecular speed (independent of pressure) and from collision frequency (a rate, outside the syllabus).

Ask the tutor: Want a walk-through of exactly why the relative-speed averaging gives root two rather than some other factor, or more practice problems mixing the number-density and pressure-temperature forms of the mean free path formula?

Why does the collision cross-section use pi d squared and not pi times the radius squared?

Because a collision is decided by how close the centres of two spheres get, not by the size of one sphere's cross-sectional disc. Each molecule has diameter $d$ , so two molecules first touch (surfaces meeting) exactly when their centres are a distance $d$ apart — not $d/2$ . So the "target" that a moving molecule's centre must miss to avoid a collision is a circle of radius $d$ around every other molecule's centre, giving an area $\pi d^2$ . If you used $\pi(d/2)^2$ instead, you would be answering a different, wrong question: how close do the surfaces of two point-sized markers need to be, which is not how two extended spheres actually collide.

Why does mean free path not depend on how fast the molecule is moving?

Mean free path is a purely geometric and statistical quantity — it asks how much distance a molecule covers, on average, before its path crosses close enough to another molecule's centre to collide, given how densely packed and how large the molecules are. Speed does not appear in $\lambda = 1/(\sqrt2\,\pi d^2 n)$ at all. What speed does affect is how often collisions happen per second (the collision frequency) and hence how quickly the molecule covers that same mean free path — a faster molecule crosses the same average gap in less time, but the gap itself, $\lambda$ , stays fixed by $n$ and $d$ alone.

If air molecules move at about 500 metres per second, why does perfume take so long to cross a room?

Because 500 m/s is the speed between collisions, and at atmospheric density a molecule collides roughly a billion times a second — its mean free path is only about $10^{-7}\ \text{m}$ . So instead of travelling in a straight line, a molecule takes an enormous number of very short, randomly redirected steps, a random walk rather than a straight dash. The straight-line distance a molecule makes progress across a room in a given time grows much more slowly than its actual speed times that time, because most of its motion cancels out in random zig-zags. This is exactly why diffusion (perfume spreading) is so much slower than the blistering molecular speeds kinetic theory predicts, and mean free path is the number that quantifies just how short each straight step really is.

Is the mean free path formula the same as the collision frequency formula?

No, and CBSE only asks for the mean free path itself, not collision frequency, so keep the two ideas separate for your board exam. Mean free path $\lambda$ is a distance — how far, on average, between collisions. Collision frequency is a rate — how many collisions per second — and would be found (outside your syllabus) by roughly dividing a typical molecular speed by $\lambda$ . You are only required to compute and reason about $\lambda$ itself, using the number-density form or the pressure–temperature form; do not bring in collision-frequency formulas from outside the prescribed scope into a board answer.