Maxwell–Boltzmann distribution

Not to be confused with Maxwell–Boltzmann statistics. Maxwell-Boltzmann Molecular Speed Distribution for Noble Gases

Probability density (s/m)

In statistics the Maxwell–Boltzmann distribution is 0.004

a particular probability distribution named after James

0.003
Clerk Maxwell and Ludwig Boltzmann. It was first de- 4 He
20
Ne
fined and used in physics (in particular in statistical me- 0.002 40Ar
132
Xe
chanics) for describing particle speeds in idealized gases
where the particles move freely inside a stationary con- 0.001

tainer without interacting with one another, except for

very brief collisions in which they exchange energy and 0 500 1000 1500 2000 2500

momentum with each other or with their thermal envi- Speed (m/s)

ronment. Particle in this context refers to gaseous parti-

cles (atoms or molecules), and the system of particles is The speed probability density functions of the speeds of a few
assumed to have reached thermodynamic equilibrium.[1] noble gases at a temperature of 298.15 K (25 °C). The y-axis
While the distribution was first derived by Maxwell in is in s/m so that the area under any section of the curve (which
1860 on heuristic grounds,[2] Boltzmann later carried out represents the probability of the speed being in that range) is di-
significant investigations into the physical origins of this mensionless.
distribution.
A particle speed probability distribution indicates which per unit speed, of finding the particle with a speed near v .
speeds are more likely: a particle will have a speed se- This equation is simply the Maxwell distribution√ (given in
lected randomly from the distribution, and is more likely the infobox) with distribution parameter a = kT /m
to be within one range of speeds than another. The distri- . In probability theory the Maxwell–Boltzmann distribu-
bution depends on the temperature of the system and the tion is a chi distribution √
with three degrees of freedom
mass of the particle.[3] The Maxwell–Boltzmann distri- and scale parameter a = kT /m .
bution applies to the classical ideal gas, which is an ide-
The simplest ordinary differential equation satisfied by
alization of real gases. In real gases, there are various
the distribution is:
effects (e.g., van der Waals interactions, vortical flow,
relativistic speed limits, and quantum exchange interac-
tions) that can make their speed distribution different
from the Maxwell–Boltzmann form. However, rarefied kT vf ′ (v) + f (v)(mv 2 − 2kT ) = 0,
gases at ordinary temperatures behave very nearly like an
ideal gas and the Maxwell speed distribution is an excel- √
lent approximation for such gases. Thus, it forms the ba- 2 − m ( m )3/2
f (1) = e 2kT
sis of the Kinetic theory of gases, which provides a sim- π kT
plified explanation of many fundamental gaseous proper- or in unitless presentation:
ties, including pressure and diffusion.[4]

( )
1 Distribution function a2 xf ′ (x) + x2 − 2a2 f (x) = 0,

√
The Maxwell–Boltzmann distribution is the function[5] 2 − 2a12
πe
f (1) = .
a3
√(
m )3 mv 2 Note that a distribution (function) is not the same as the
f (v) = 4πv 2 e− 2kT ,
2πkT probability. The distribution (function) stands for an av-
erage number, as in all three kinds of statistics (Maxwell–
where m is the particle mass and kT is the product of Boltzmann, Bose–Einstein, Fermi–Dirac). With the
Boltzmann’s constant and thermodynamic temperature. Darwin–Fowler method of mean values the Maxwell–
This probability density function gives the probability, Boltzmann distribution is obtained as an exact result.

1
2 3 DERIVATION AND RELATED DISTRIBUTIONS

2 Typical speeds The derivations in this section are along the lines of Boltz-
mann’s 1877 derivation, starting with result known as
The mean speed, most probable speed (mode), and root- Maxwell–Boltzmann statistics (from statistical thermody-
mean-square can be obtained from properties of the namics). Maxwell–Boltzmann statistics gives the aver-
Maxwell distribution. age number of particles found in a given single-particle
microstate, under certain assumptions:[1][6]
• The most probable speed, vp, is the speed most
likely to be possessed by any molecule (of the same
mass m) in the system and corresponds to the maxi-
mum value or mode of f(v). To find it, we calculate where:
the derivative df/dv, set it to zero and solve for v:
• i and j are indices (or labels) of the single-particle
df (v)
=0 micro states,
dv
• Ni is the average number of particles in the single-
which yields:
particle microstate i,

√ √ • N is the total number of particles in the system,

2kT 2RT
vp = = • Ei is the energy of microstate i,
m M
where R is the gas constant and M = NA m is the • T is the equilibrium temperature of the system,
molar mass of the substance. • k is the Boltzmann constant.
For diatomic nitrogen (N2 , the primary component
of air) at room temperature (300 K), this gives vp = The assumptions of this equation are that the particles
422 m/s do not interact, and that they are classical; this means
that each particle’s state can be considered independently
• The mean speed is the expected value of the speed
from the other particles’ states. Additionally, the particles
distribution
are assumed to be in thermal equilibrium. The denomi-
∫ ∞ √ √
8kT 8RT 2 nator in Equation (1) is simply a normalizing factor so
⟨v⟩ = v f (v) dv = = = √ vp that the Ni/N add up to 1 — in other words it is a kind of
0 πm πM π
partition function (for the single-particle system, not the
usual partition function of the entire system).
• The root mean square speed is the second-order
moment of speed: Because velocity and speed are related to energy, Equa-
√ (1) can be used to derive relationships between tem-
tion
(∫ ∞ )1/2 √ √
√ 3kT 3RT perature
3 and the speeds of gas particles. All that is needed
⟨v 2 ⟩ = v 2 f (v) dv = = = is to vdiscover
p the density of microstates in energy, which
0 m M 2
is determined by dividing up momentum space into equal
sized regions.
The typical speeds are related as follows:

√ 3.1 Distribution for the momentum vector

0.886⟨v⟩ = vp < ⟨v⟩ < ⟨v 2 ⟩ = 1.085⟨v⟩.
The potential energy is taken to be zero, so that all en-
ergy is in the form of kinetic energy. The relationship
3 Derivation and related distribu- between kinetic energy and momentum for massive non-
relativistic particles is
tions
The original derivation in 1860 by James Clerk Maxwell
was an argument based on demanding certain symme-
where p2 is the square of the momentum vector p = [px,
tries in the speed distribution function.[2] After Maxwell,
py, pz]. We may therefore rewrite Equation (1) as:
Ludwig Boltzmann in 1872 derived the distribution on
more mechanical grounds by using the assumptions of
his Kinetic theory of gases, and showed that gases should
over time tend toward this distribution, due to collisions
(see H-theorem). He later (1877) derived the distribution where Z is the partition function, corresponding to the
again under the framework of statistical thermodynamics. denominator in Equation (1). Here m is the molecular
3.3 Distribution for the velocity vector 3

mass of the gas, T is the thermodynamic temperature and Since the energy is proportional to the sum of the squares
k is the Boltzmann constant. This distribution of Nᵢ/N of the three normally distributed momentum compo-
is proportional to the probability density function f for nents, this distribution is a gamma distribution; in par-
finding a molecule with these values of momentum com- ticular, it is a chi-squared distribution with three degrees
ponents, so: of freedom.
By the equipartition theorem, this energy is evenly dis-
tributed among all three degrees of freedom, so that the
energy per degree of freedom is distributed as a chi-
[7]
The normalizing constant c, can be determined by rec- squared distribution with one degree of freedom:
ognizing that the probability of a molecule having some
momentum must be 1. Therefore the integral of equation √ [ ]
(4) over all px, py, and pz must be 1. 1 −ϵ
fϵ (ϵ) dϵ = exp dϵ
It can be shown that: πϵkT kT
where ϵ is the energy per degree of freedom. At equi-
librium, this distribution will hold true for any number
of degrees of freedom. For example, if the particles are
Substituting Equation (5) into Equation (4) gives: rigid mass dipoles of fixed dipole moment, they will have
three translational degrees of freedom and two additional
rotational degrees of freedom. The energy in each degree
of freedom will be described according to the above chi-
squared distribution with one degree of freedom, and the
The distribution is seen to be the product of three in- total energy will be distributed according to a chi-squared
dependent normally distributed variables px , py , and distribution with five degrees of freedom. This has impli-
pz , with variance mkT . Additionally, it can be seen cations in the theory of the specific heat of a gas.
that the magnitude of momentum will be distributed
√ as
The Maxwell–Boltzmann distribution can also be ob-
a Maxwell–Boltzmann distribution, with a = mkT .
tained by considering the gas to be a type of quantum
The Maxwell–Boltzmann distribution for the momentum
gas for which the approximation ε >> k T may be made.
(or equally for the velocities) can be obtained more fun-
damentally using the H-theorem at equilibrium within the
Kinetic theory of gases framework. 3.3 Distribution for the velocity vector
Recognizing that the velocity probability density fᵥ is
3.2 Distribution for the energy proportional to the momentum probability density func-
tion by
The energy distribution is found imposing

( )3
3 dp
fv d v = fp d3 v
dv
where d3 p is the infinitesimal phase-space volume of mo-
menta corresponding to the energy interval dE . Making and using p = mv we get
use of the spherical symmetry of the energy-momentum
dispersion relation E = |p|2 /2m , this can be expressed
in terms of dE as
which is the Maxwell–Boltzmann velocity distribution.
The probability of finding a particle with velocity in the
infinitesimal element [dvx, dvy, dvz] about velocity v =
[vx, vy, vz] is
Using then (8) in (7), and expressing everything in terms
of the energy E , we get
fv (vx , vy , vz ) dvx dvy dvz .
√ ( )3/2 ( )
1 −E/kT
√ E the
Like 1 momentum,−E this distribution is seen to be the
fE (E)dE = e 4πm 2mEdE = 2 exp dE
(2πmkT )3/2 product
π kT of three independent
kT normally distributed vari-
ables vx , vy , and vz , but with variance kT m . It can
and finally also be seen that the Maxwell–Boltzmann velocity distri-
bution for the vector velocity [vx, vy, vz] is the product
of the distributions for each of the three directions:
4 6 FURTHER READING

• Maxwell, J.C. (1860) “Illustrations of the dynam-

ical theory of gases. Part I. On the motions and
fv (vx , vy , vz ) = fv (vx )fv (vy )fv (vz ) collisions of perfectly elastic spheres,” Philosophi-
cal Magazine, 4th series, 19 : 19–32.
where the distribution for a single direction is
• Maxwell, J.C. (1860) “Illustrations of the dynami-
cal theory of gases. Part II. On the process of dif-
√ [ ] fusion of two or more kinds of moving particles
m −mvi2 among one another,” Philosophical Magazine, 4th
fv (vi ) = exp .
2πkT 2kT series, 20 : 21–37.

Each component of the velocity vector has a normal dis- [3] University Physics – With Modern Physics (12th Edition),
tribution with mean µvx = µvy = µvz = 0 and standard H.D. Young, R.A. Freedman (Original edition), Addison-
√
deviation σvx = σvy = σvz = kT Wesley (Pearson International), 1st Edition: 1949, 12th
m , so the vector has
Edition: 2008, ISBN 978-0-321-50130-1
a 3-dimensional normal distribution, a particular kind of
multivariate normal distribution,
√ with mean µv = 0 and
[4] Encyclopaedia of Physics (2nd Edition), R.G. Lerner,
3kT
standard deviation σv = m . G.L. Trigg, VHC publishers, 1991, ISBN 3-527-26954-1
(Verlagsgesellschaft), ISBN 0-89573-752-3 (VHC Inc.)
The Maxwell–Boltzmann distribution for the speed fol-
lows immediately from the distribution of the velocity [5] H.J.W. Müller-Kirsten, Basics of Statistical Physics, 2nd
vector, above. Note that the speed is ed., World Scientific (2013),ISBN 978-981-4449-53-3,
Chapter 2.
√
v= vx2 + vy2 + vz2 [6] McGraw Hill Encyclopaedia of Physics (2nd Edition),
C.B. Parker, 1994, ISBN 0-07-051400-3
and the volume element in spherical coordinates
[7] Laurendeau, Normand M. (2005). Statistical thermody-
namics: fundamentals and applications. Cambridge Uni-
versity Press. p. 434. ISBN 0-521-84635-8., Appendix
dvx dvy dvz = v 2 sin θ dv dθ dϕ N, page 434

where ϕ and θ are the “course” (azimuth of the velocity

vector) and “path angle” (elevation angle of the veloc-
ity vector). Integration of the normal probability density 6 Further reading
function of the velocity, above, over the course (from 0
to 2π ) and path angle (from 0 to π ), with substitution of • Physics for Scientists and Engineers – with Modern
the speed for the sum of the squares of the vector com- Physics (6th Edition), P. A. Tipler, G. Mosca, Free-
ponents, yields the speed distribution. man, 2008, ISBN 0-7167-8964-7

• Thermodynamics, From Concepts to Applications

4 See also (2nd Edition), A. Shavit, C. Gutfinger, CRC Press
(Taylor and Francis Group, USA), 2009, ISBN 978-
• Maxwell–Boltzmann statistics 1-4200-7368-3
• Maxwell–Jüttner distribution • Chemical Thermodynamics, D.J.G. Ives, Univer-
• Boltzmann distribution sity Chemistry, Macdonald Technical and Scientific,
1971, ISBN 0-356-03736-3
• Boltzmann factor
• Elements of Statistical Thermodynamics (2nd Edi-
• Rayleigh distribution tion), L.K. Nash, Principles of Chemistry, Addison-
• Kinetic theory of gases Wesley, 1974, ISBN 0-201-05229-6

• Ward, CA & Fang, G 1999, 'Expression for predict-

ing liquid evaporation flux: Statistical rate theory
5 References approach', Physical Review E, vol. 59, no. 1, pp.
429–40.
[1] Statistical Physics (2nd Edition), F. Mandl, Manch-
ester Physics, John Wiley & Sons, 2008, ISBN
• Rahimi, P & Ward, CA 2005, 'Kinetics of Evapo-
9780471915331
ration: Statistical Rate Theory Approach', Int. J. of
[2] See: Thermodynamics, vol. 8, no. 9, pp. 1–14.
 5

7 External links
• “The Maxwell Speed Distribution” from The Wol-
fram Demonstrations Project at Mathworld
6 8 TEXT AND IMAGE SOURCES, CONTRIBUTORS, AND LICENSES

8 Text and image sources, contributors, and licenses

8.1 Text
• Maxwell–Boltzmann distribution Source: https://en.wikipedia.org/wiki/Maxwell%E2%80%93Boltzmann_distribution?oldid=
742522139 Contributors: AxelBoldt, Tobias Hoevekamp, Marj Tiefert, Derek Ross, Bryan Derksen, The Anome, Stokerm, XJaM,
Arvindn, SimonP, FlorianMarquardt, Lir, Michael Hardy, Delirium, Dgrant, Mdebets, Ahoerstemeier, Cyp, Snoyes, Bogdangiusca,
Charles Matthews, Pablo Mayrgundter, DJ Clayworth, Robbot, Pfortuny, Henrygb, Wile E. Heresiarch, Giftlite, Cfp, Awolf002, Gunnar
Larsson, Karl-Henner, Tsemii, Edsanville, Guanabot, Tooto, Euyyn, Svdmolen, Constantine, O18, Wood Thrush, Smalljim, Jamesgriff,
Evgeny, Jérôme, Alansohn, PAR, Bucephalus, Jheald, Linas, Btyner, Tevatron~enwiki, Eteq, Nanite, Salix alba, NeonMerlin, Krash,
Margosbot~enwiki, Krishnavedala, DVdm, Tone, Whosasking, YurikBot, Wavelength, RussBot, Stephenb, Giro720, Rsrikanth05,
Jess Riedel, JoanneB, Croat Canuck, Jaysbro, Infinity0, SmackBot, Bromskloss, Kmarinas86, Bh3u4m, Jjalexand, Oli Filth, Hongooi,
Iwaterpolo, Can't sleep, clown will eat me, Shalom Yechiel, Lostart, Jon Awbrey, Eliyak, John, FrozenMan, Jaganath, Physis, Onionmon,
Harej bot, Mct mht, Boulderinionian, Astrochemist, Hkyriazi, Sytelus, Rolinc, Headbomb, Zé da Silva, Nick Number, Oreo Priest,
JAnDbot, Hamsterlopithecus, Felix116, Magioladitis, ZomBGolth, Dirac66, User A1, TheNoise, CommonsDelinker, Tarotcards, Tow-
erman86, Deltasct, Worp8d, Pasquale.Carelli, Tkriddler, VanishedUserABC, Legoktm, Cwkmail, Maitraz, Paolo.dL, Joldy, Melcombe,
Abmcdonald, Andresj, Niceguyedc, UKoch, LizardJr8, Excirial, Gulmammad, Ykhwong, Qwfp, Truthnlove, Addbot, LaaknorBot,
FutureKingMark, Jasper Deng, Lightbot, Teles, Yobot, Legobot II, AnomieBOT, Citation bot, Gap9551, ChristopherKingChemist,
Mnmngb, Rainald62, FrescoBot, Yasha Azar, Khashishi, Gryllida, Kastchei, EmausBot, WikitanvirBot, JaeDyWolf, Phelimb, SporkBot, L
Kensington, ChuispastonBot, Sjjamsa, ClueBot NG, Nohwave, Frietjes, Rezabot, Helpful Pixie Bot, Bibcode Bot, Gredner, F=q(E+v^B),
Lujunkai, Roshan220195, ChrisGualtieri, Planetree, Hschiret, Jakan Sanjiv, BeyondNormality, Jeffreyfwebb, Asympto, Ggf4t, Martin
Embeh, SamuelHon, CanaDeAlba, Notjustphysics, Crvsensei, Mathxpert, Bender the Bot and Anonymous: 180

8.2 Images
• File:Maxwell-Boltzmann_distribution_cdf.svg Source: https://upload.wikimedia.org/wikipedia/commons/9/98/Maxwell-Boltzmann_
distribution_cdf.svg License: CC0 Contributors: Own work Original artist: Krishnavedala
• File:Maxwell-Boltzmann_distribution_pdf.svg Source: https://upload.wikimedia.org/wikipedia/commons/1/19/Maxwell-Boltzmann_
distribution_pdf.svg License: CC0 Contributors: Own work Original artist: Krishnavedala
• File:MaxwellBoltzmann-en.svg Source: https://upload.wikimedia.org/wikipedia/commons/0/01/MaxwellBoltzmann-en.svg License:
Public domain Contributors: Originally from en.wikipedia; description page is/was here. Original artist: The original uploader was Pdbailey
at English Wikipedia

8.3 Content license

• Creative Commons Attribution-Share Alike 3.0