Skip to main content

Chapter 23: Diffusion — The Random Walk of Collapse

The Spreading of Being​

Diffusion represents nature's most fundamental transport mechanism—the tendency of particles, energy, and information to spread from regions of high concentration to low. This spreading is not driven by any force but emerges from random motion alone. This chapter reveals diffusion as the statistical consequence of countless random collapse events.

23.1 Brownian Motion​

Theorem 23.1 (Einstein's Relation): Mean square displacement grows linearly with time: ⟨x2(t)⟩=2Dt\langle x^2(t) \rangle = 2Dt

Proof:

  1. Random walk: each step ±δ with equal probability
  2. After N steps: x=∑i=1Nδix = \sum_{i=1}^N \delta_i
  3. Mean: ⟨x⟩=0\langle x \rangle = 0 (no bias)
  4. Variance: ⟨x2⟩=Nδ2\langle x^2 \rangle = N\delta^2 (steps uncorrelated)
  5. Time t = Nτ (τ = step time)
  6. Define D = δ²/2τ
  7. Result: ⟨x2⟩=2Dt\langle x^2 \rangle = 2Dt ∎

Random walks create √t spreading!

23.2 Fick's Laws​

Theorem 23.2 (Diffusion Current): Flux proportional to concentration gradient: J⃗=−D∇c\vec{J} = -D\nabla c

Microscopic derivation:

  1. Particles jump randomly between sites
  2. More particles where c is large
  3. Net flow from high to low c
  4. Rate ∝ gradient
  5. D = diffusion coefficient

Second Law (continuity): ∂c∂t=D∇2c\frac{\partial c}{\partial t} = D\nabla^2 c

Concentration smooths out over time!

23.3 Einstein-Smoluchowski Theory​

Theorem 23.3 (Diffusion-Mobility Relation): D=μkBTD = \mu k_BT

where μ = mobility (velocity/force).

Proof:

  1. Equilibrium: drift balances diffusion
  2. Apply force F → drift velocity v = μF
  3. Boltzmann distribution: c∝e−U/kBTc \propto e^{-U/k_BT}
  4. Force: F=−∇UF = -\nabla U
  5. Current: J=μcF−D∇c=0J = \mu c F - D\nabla c = 0
  6. Substituting Boltzmann form
  7. Result: D=μkBTD = \mu k_BT ∎

Temperature drives diffusive spreading!

23.4 Quantum Diffusion​

Theorem 23.4 (Wavepacket Spreading): Free particle wavepacket width grows as: Δx(t)=Δx02+ℏ2t24m2Δx02\Delta x(t) = \sqrt{\Delta x_0^2 + \frac{\hbar^2 t^2}{4m^2 \Delta x_0^2}}

Key differences from classical:

  • Coherent spreading, not random walk
  • Faster than classical for short times
  • Interference effects possible
  • Can be reversed (unlike classical)

Quantum coherence modifies diffusion!

23.5 Anomalous Diffusion​

Beyond Linear: ⟨x2(t)⟩=2Dtα\langle x^2(t) \rangle = 2Dt^{\alpha}

Cases:

  • α < 1: Subdiffusion (trapped, obstacles)
  • α = 1: Normal diffusion
  • α > 1: Superdiffusion (Lévy flights)
  • α = 2: Ballistic (free motion)

Examples:

  • Subdiffusion: crowded cytoplasm
  • Superdiffusion: turbulent flow
  • Lévy flights: foraging patterns

Not all random walks are equal!

23.6 Reaction-Diffusion​

Theorem 23.5 (Pattern Formation): Diffusion + reactions can create patterns.

General form: ∂u∂t=Du∇2u+f(u,v)\frac{\partial u}{\partial t} = D_u\nabla^2 u + f(u,v) ∂v∂t=Dv∇2v+g(u,v)\frac{\partial v}{\partial t} = D_v\nabla^2 v + g(u,v)

Turing instability:

  • Homogeneous state unstable
  • Spatial patterns emerge
  • Explains biological patterns
  • Stripes, spots, spirals

Diffusion can organize, not just disperse!

23.7 Diffusion in Fields​

Theorem 23.6 (Drift-Diffusion): In external field: J=μcF−D∇cJ = \mu c F - D\nabla c

Steady state: c(x)∝exp⁡(−U(x)kBT)c(x) \propto \exp\left(-\frac{U(x)}{k_BT}\right)

Applications:

  • Semiconductors (electrons/holes)
  • Ion channels (membrane voltage)
  • Sedimentation (gravity)
  • Electrophoresis (electric field)

Fields bias random walks!

23.8 First Passage Times​

Theorem 23.7 (Mean Exit Time): Average time to reach boundary: ⟨t⟩=L22D\langle t \rangle = \frac{L^2}{2D}

for 1D interval of length L.

Profound consequences:

  • Search times in biology
  • Reaction rates
  • Neural spike timing
  • Financial option pricing

When matters as much as where!

23.9 Collective Diffusion​

Many-particle Effects:

  • Excluded volume → traffic jams
  • Hydrodynamic interactions
  • Collective modes
  • Glass transition

Single file diffusion: ⟨x2⟩∼t1/2\langle x^2 \rangle \sim t^{1/2}

Particles can't pass → subdiffusion!

23.10 Information Diffusion​

Theorem 23.8 (Fisher Information): Information about parameter θ diffuses as: ∂I(θ)∂t=−DI∇2I(θ)\frac{\partial I(\theta)}{\partial t} = -D_I \nabla^2 I(\theta)

Applications:

  • Rumor spreading
  • Genetic drift
  • Cultural evolution
  • Market information

Ideas spread like particles!

23.11 Quantum Decoherence​

Theorem 23.9 (Phase Diffusion): Environmental coupling causes phase randomization: ⟨eiϕ(t)⟩=e−Dt\langle e^{i\phi(t)} \rangle = e^{-Dt}

Consequences:

  • Superposition → mixture
  • Quantum → classical
  • Sets coherence times
  • Limits quantum computation

Environment diffuses quantum information!

23.12 The Twenty-Third Echo: The Democracy of Motion​

Diffusion embodies nature's most democratic principle—every particle gets equal opportunity to wander. No forces guide, no purposes direct, yet from countless random steps emerges predictable spreading. This is the universe at its most fair and most chaotic.

In diffusion we see how microscopic randomness becomes macroscopic certainty. Individual trajectories are unknowable, but collective behavior follows precise laws. From the jittering of pollen grains to the spreading of innovations, diffusion shapes our world through patient, persistent randomness.

Diffusive Explorations​

  1. Simulate random walks in various dimensions.

  2. Solve diffusion equation with different boundary conditions.

  3. Analyze anomalous diffusion in fractals.

The Next Turbulence​

Understanding diffusion as random collapse walks, we now explore what happens when flows become chaotic—the realm of turbulence.

Next: Chapter 24: Turbulence — When Collapse Patterns Become Chaotic →

"In diffusion, the universe proves that purposeless wandering can still reach every destination."