Boltzmann Distribution — Definition, Formula & Example
The Boltzmann distribution is a probability distribution that says how likely a system at a fixed temperature is to be found in each of its possible states. Each state's probability falls off exponentially with its energy, and a higher temperature makes the high-energy states more accessible.
Let a system in thermal equilibrium at absolute temperature have discrete states with energies . The Boltzmann distribution assigns state the probability . The normalizing constant is the partition function. Among all distributions with a fixed mean energy, this is the one with maximum entropy.
Key Formula
Where:
- = Probability that the system is in state
- = Energy of state (SI units, joules, or electron-volts, as is conventional in physics)
- = Boltzmann constant, J/K, or about eV/K
- = Absolute temperature in kelvins
- = Partition function, which makes the probabilities sum to 1
Worked Example
Problem: A molecule has three energy levels: 0 eV, 0.05 eV, and 0.10 eV. The temperature makes eV (about 580 K). Find the probability of each level.
Step 1: Divide each energy by to get dimensionless exponents.
Step 2: Compute each Boltzmann factor and add them to get the partition function.
Step 3: Divide each factor by .
Answer: The probabilities are about 0.6652, 0.2447, and 0.0900, and they sum to 1 (up to rounding). The lowest level is times as likely as the highest level. That ratio depends only on the energy gap, not on .
Why It Matters
In chemistry, the Boltzmann factor drives the Arrhenius equation, which explains why reaction rates rise sharply with temperature. The same formula appears in machine learning as the softmax function, which turns scores into probabilities. It also appears in simulated annealing, an optimization method that slowly lowers a temperature parameter.
Common Mistakes
Mistake: Plugging the temperature in degrees Celsius or Fahrenheit into .
Correction: The formula requires absolute temperature. Convert to kelvins first, for example K. Using Celsius gives the wrong value, and at it would divide by zero.
