AI summary by TwinMind
Lecture 29: Boltzmann Distribution
MIT OpenCourseWare
49m 35s
Course Structure and Remaining Content
- Finishing statistical thermodynamics coverage today
- Monday lectures will cover reacting systems and oxidation processes (Chapter 11)
- Final exam is non-cumulative
- Game show format planned for review session
Boltzmann Distribution Derivation Continuation
- Continued from previous lecture's derivation using Lagrange multipliers
- Optimized entropy for isolated system with constraints:
- Conservation of energy
- Conservation of mass
- Two Lagrange multipliers: α and β
- Derived distribution function: ni/Ntotal = e^(-βεi/kB)/Q
- Partition function Q = Σ e^(-βεi/kB) over all states
Determination of β Parameter
- Analyzed microscopic reversible process with state changes
- Used entropy change formula: dS = kB Σ ln(ni/Ntotal) dni
- Substituted Boltzmann distribution to get: dS = -β du + kB log(Q) dNtotal
- By construction, dS = 0 for isolated system (dU = 0, dNtotal = 0)
- Compared with classical thermodynamics: dS = (1/T)dU + (p/T)dV - (μ/T)dN
- By coefficient comparison: β = -1/T
Final Boltzmann Distribution
- ni/Ntotal = e^(-Ei/kBT)/Q
- Q = Σ e^(-Ei/kBT) over all states
- Falling exponential form with state energy
- Distribution that maximizes entropy for isolated system with fixed energy levels
- Represents equilibrium distribution of single particle energy levels
Likelihood and Energy Scaling
- Ratio of occupation probabilities: nm/nl = e^(-Δε/kBT)
- Less likely to find particles at higher energy states
- Depends on energy splitting Δε/kBT
- Becomes less likely with increasing energy splitting
- Becomes more likely with increasing temperature
Thermal Energy Scale
- kBT represents thermal energy scale for likely fluctuations
- Boltzmann constant: kB = 1.380×10^-23 J/K = 8.617×10^-5 eV/K
- Room temperature (298K): kBT ≈ 25.7 meV
- Sets energy scale for thermally activated processes
- Foundation for Arrhenius rate equation: k = A·e^(-Ea/kBT)
Maxwell-Boltzmann Distribution
- Distribution of particles with kinetic energy
- Kinetic energy: E = (px² + py² + pz²)/2m
- Combines Boltzmann distribution with density of states
- Density of states in 3D: g(E) = 4π(2m)^(3/2)E^(1/2)
- Final form: f(E) = (2E/π)(1/kBT)^(3/2)·e^(-E/kBT)
- Speed distribution shows:
- Sharp peak at low temperature
- Broader distribution at high temperature
- Mean speed increases with temperature
- Vanishingly small number at zero speed
Ensembles Introduction
- Ensemble: group of people/systems acting together or taken as whole
- Preparation for advanced statistical mechanics courses
Microcanonical Ensemble
- Set of all possible microstates of system with fixed:
- Energy (E)
- Volume (V)
- Particle number (N)
- Isolated system
- Equilibrium at maximum entropy
- Partition function: Q = Σ e^(-Ei/kBT) over single particle states
- Probability of finding particle in state i: P(i) = e^(-Ei/kBT)/Q
Canonical Ensemble
- Set of all possible microstates with fixed:
- Volume (V)
- Particle number (N)
- BUT energy can fluctuate
- System is closed and rigid but can exchange heat energy with surroundings
- Ensemble of systems with diathermal walls
- Entire ensemble is isolated
- Equilibrium at minimum Helmholtz free energy (F)
- Partition function: Z = Σ e^(-Uν/kBT) over system energy states
- Probability of finding system in state ν: P(ν) = e^(-Uν/kBT)/Z
- Equilibrium at fixed temperature (thermal equilibrium)
Partition Functions and Thermodynamic Properties
- Partition functions describe equilibrium and fluctuations
- Can calculate all thermodynamic properties from partition function:
- Mean energy: <U> = d(ln Z)/d(1/kBT)
- Heat capacity: CV = (1/kBT²)d²(ln Z)/d(1/kBT)²
- Entropy: S = <U>/T + kB ln Z
- Helmholtz free energy: F = -kBT ln Z
- Root mean squared energy fluctuation also calculable
- Example: Derive ideal gas law from first principles using partition functions
Limitations of Statistical Approach
- Approach is beautiful, elegant, and powerful but extremely limited
- Can only calculate partition functions for simplest toy problems
- Real systems too complex for analytical solutions
- Example: 1D Ising model (solvable) vs 2D Ising model (Nobel Prize level difficulty)
- 3D real magnets essentially unsolvable analytically
- Classical thermodynamics more practical for engineering applications
- Statistical mechanics provides theoretical foundation but classical approach needed for real-world problems
