Non-equilibrium Statistical Physics Tutorial session 2
Tutorial session 2
Discussion on January 8, 15:00-17:00, Room SP F2.04
Motion of an damped Brownian particle in a harmonic potential
Consider a Brownian particle of mass m which is constrained to move in one dimension in a harmonic
potential V (x) = kx2 /2. The harmonic oscillator provides a clear and widely relevant example of Langevin
dynamics in experiments. A prominent example is a Brownian particle in a fluid trapped by optical tweez-
ers, which effectively impose a harmonic potential on a particle that would otherwise undergo free Brownian
motion, see Fig. 1. In this exercise, you will be walked through to investigate the dynamics of a Brownian
Figure 1: An illustration of a Brownian particle constrained by an optical trap undergoing Brownian motion
due to frequent kicks from the solvent particles (blue spheres).
particle constrained by a harmonic trap. The Langevin equation of motion for the particle position is given
by
ẋ 1
ẍ = − − ω02 x + ξ(t), (1)
τm m
where
k m
ω02 = , τm = . (2)
m γ
represent the characteristic frequency of oscillator and inertial timescale, respectively.
Here the random force is a Gaussian random process with moments
⟨ξ(t)⟩ = 0, (3)
⟨ξ(t)ξ(t′ )⟩ = 2γkB T δ(t − t′ ) (4)
a) Fourier transform the equation of motion where Fourier transform of a function F (t) is defined as
Z ∞
F̃ (ω) = dt F (t) eiωt (5)
−∞
Non-equilibrium Statistical Physics Tutorial session 2
˜
and solve for x̃(ω) in terms of the fluctuating force ξ(ω).
b) The spectral density of a quantity is defined as the Fourier transform of its time correlation function:
Z ∞
SF (ω) = dt ⟨F (0)F ∗ (t)⟩eiωt . (6)
−∞
The spectral density is linked to
⟨F̃ (ω)F̃ ∗ (ω ′ )⟩ = 2πδ(ω − ω ′ )SF (ω). (7)
If F is real, then F̃ ∗ (ω) = F̃ (−ω), and
⟨F̃ (ω)F̃ (ω ′ )⟩ = 2πδ(ω + ω ′ )SF (ω). (8)
Given the definitions above, calculate the spectral density of noise ξ(t) and position x(t).
c) Calculate the two-time correlation function ⟨x(t)x(0)⟩ by Fourier transforming back the spectral density
of Sx (ω)
Z ∞
1
⟨x(0)x(t)⟩ = dω Sx (ω)e−iωt . (9)
2π −∞
Hint: The integral can be calculated as a contour integral in the complex plane using the residue
theorem. For a complex function f (z) that is analytic everywhere inside and on a closed contour Γ, except
for a finite number of isolated singularities z1 , z2 , . . . , zn (poles) inside Γ, the residue theorem states
I n
X
f (z) dz = ±2πi Res f, zk , (10)
Γ k=1
where + and − apply for counterclockwise and clockwise integration, respectively. For a function of the
form f (z) = g(z)/h(z) that has h(z0 ) = 0 and h′ (z0 ) ̸= 0 at the pole z0 , the residue has the expression
g(z0 )
Res(f, z0 ) = . (11)
h′ (z0 )
d) Plot Sx (ω)/⟨x2 (0)⟩ as a function of ω/ω0 for 1/τm = 0.2, 1 and 1.5 ω0 . Discuss how the behavior changes
as you go from overdamped to underdamped regime.
Nice to know: The power spectrum of position fluctuations of a Brownian particle trapped by opti-
cal tweezers can be measured in experiments.
e) Bonus: Find the velocity correlation function ⟨v(t)v(0)⟩, from its spectral density using the relation
ṽ(ω) = −iωx̃(ω).