Professional Documents
Culture Documents
Fick's Law and Diffusion
Fick's Law and Diffusion
In the first part of this course we assumed that the state of a cell at a particular point in
time could be described by the concentrations (or number of molecules in the stochastic
case) of key components. These concentrations could change over time but we implicitly
assumed that within the cell these concentrations do not vary as a function of position
inside the cell. In other words, we considered the cell as a well‐stirred biochemical
reactor. For small cells such as Escherichia coli this is often a valid assumption
particularly when we are interested in dynamical phenomena that occur at a much
slower time‐scale than the relatively fast random mixing by diffusion. However in larger
cells such as eukaryotic cells and even sometimes in bacteria we cannot assume that the
cell interior is well‐stirred. As a starting point for building up the mathematical tools
necessary to model spatial inhomogeneous system we use Fick’s laws:
Fick’s first law
Figure 24. Particles randomly
moving along one‐dimension
crossing an area A
Consider Fig. 24 illustrating particles moving along one dimension x. The particles are
randomly moving. Assume there are N(x) particles in the gray region on the left of area
A and N(x+Δx) particles in the gray region on the right of area A. How many particles will
−
1
(N ( x + Δx) − N ( x) ) . The flux of molecules J through the area A during a short time
2
interval τ is defined as:
−
1
(N ( x + Δx) − N ( x) )
J= 2 [X.1]
Aτ
N ( x)
rewriting [X.1] in terms of the concentration C ( x) ≡ results in:
AΔx
(Δx) 2 (C ( x + Δx) − C ( x) ) ∂C ( x)
J =− = −D [X.2]
2τ Δx ∂x
Therefore in the limit where Δx→0 the flux J is proportional to the concentration
gradient and to a constant we define as the diffusion constant. It is not a priori clear why
the ratio D ≡
(Δx )
2
is a constant. This will become more clear later when we treat
τ
diffusion from a more microscopic viewpoint. Equation[X.2] is often called Fick’s first
law.
Figure 25. Derivation of Fick’s
second law.
∂C ( x, t ) ∂ 2C ( x, t )
=D [X.5]
∂t ∂x 2
This tells us that rate of change of the concentration at a specific time and position is
proportional to the curvature (second derivative) of the concentration profile.
Random walks
Let us now consider a one‐dimensional random walk of N particles. A particle always
takes a fixed step size Δx towards the left or right with equal probability. The position of
the particles is denoted by xi(n), where the subscript denotes which particles (i=1…N)
and n denotes the number of steps that particle I took. All particles start their random
walk at x=0: xi(0)=0 for i=1…N. It is easy to show that the position averaged over all
particles is always zero independent of how many steps were taken:
However individual particles are spreading in both the positive and negative x direction.
A convenient way to quantify the spreading is to calculate the variance of the
( )
N N
Var ( x(n) ) = x 2 (n) − 0 = ∑ xi2 (n) =∑ xi (n − 1) ± Δx = x 2 (n − 1) + Δx 2 [X.7]
2
i =1 i =1
As expected the variance is non‐zero. For each extra step the variance grows by Δx2 and
since the variance is zero for n=0 we can write:
During one step the particle moves at a constant velocity v= Δx/τ . This means that a
time t the particle performed n=t/τ steps, therefore the variance is proportional to t:
Δx 2
x 2 ( n) = t ≡ 2 Dt [X.9]
τ
Here we used the definition that we introduced in equation [X.2].
Further reading
Howard C. Berg, Random walks in biology. (Princeton University Press, 1993).