Electronic and Magnetic Contributions to Heat Capacity

Introduction

Heat capacity can reveal energy levels beyond molecular vibration and crystal phonons. A magnetic field may split spin states; an atom or ion may have a low-lying electronic excitation. As temperature rises, the upper level first begins to populate, causing a heat-capacity increase, but eventually both levels approach their limiting population ratio and the contribution falls again. This broad peak is a Schottky anomaly.

Core explanation

Consider two levels with energies 0 and Δ > 0 and degeneracies g₀ and g₁. Their partition function is q = g₀ + g₁e^(−x), where x = Δ/(kT). The total probability of the upper level is p₁ = g₁e^(−x)/q. Its mean excitation energy is U = Δp₁. Since the energy is either 0 or Δ, its variance is Δ²p₁(1 − p₁), and the heat capacity is C V = kx²p₁(1 − p₁). Degeneracy enters through p₁; it changes both limiting populations and the peak shape.

At low T, x is large and the upper level is scarcely populated, so C V approaches zero. At high T, p₁ approaches g₁/(g₀+g₁), a constant, while x² tends to zero, so C V again approaches zero. Between these limits, population changes rapidly with T and C V reaches a maximum. The temperature of that maximum depends on Δ and the degeneracy ratio; it is not simply T = Δ/k. A measured peak can therefore help estimate a level splitting when other contributions are accounted for.

Magnetic spins in an applied field provide one source of two-level-like behaviour. A low-lying crystal-field excitation of an ion can produce another. For conduction electrons in a simple metal, however, the low-temperature electronic heat capacity is often approximately γT, reflecting excitations near a Fermi surface rather than an isolated two-level Schottky peak. Lattice phonons may contribute a T³ term at low T. Separating these terms from one measured C p curve requires appropriate temperature range and knowledge of magnetic ordering or other transitions.

One must also distinguish constant level splitting from a splitting that changes with field or temperature. For a fixed applied field, the simple two-level formula may be useful; if the level energies themselves change with T, differentiating only the Boltzmann populations omits part of the energy response. Strongly interacting spins may order collectively, producing a sharp phase-transition feature instead of a simple independent-spin Schottky hump.

Step-by-step reasoning

Specify level energies and degeneracies, choose a zero of energy, and form q. Calculate p₁, then U = Δp₁. Differentiate with respect to T or use energy variance to obtain C V. Check that it vanishes at both low and high T. When applying to data, subtract or model lattice and other background heat-capacity contributions before assigning a peak to discrete levels.

Visual explanation

Draw two horizontal levels separated by Δ, with g₀ dots on the lower line and g₁ dots on the upper. Beside them plot C V against T: a low baseline, one broad hump and a fall at high T. On another graph show upper-level population rising from zero toward g₁/(g₀+g₁), flattening as the heat-capacity hump subsides.

Real-world analogy

A theatre has a cheap section and a more costly balcony. At very low willingness to pay, nearly everyone stays below; at intermediate willingness, people move upstairs as conditions change; once the final seating ratio is nearly saturated, changing conditions moves few additional people. The analogy captures changing occupancy, though quantum degeneracy and thermal probabilities are mathematical rather than economic choices.

Real-world example

Low-temperature calorimetry of a paramagnetic material can show a field-dependent broad peak from spin-level splitting. Shifting the magnetic field changes Δ and can shift the peak. If the feature instead comes from a collective magnetic ordering transition, a noninteracting two-level fit may be inappropriate, so field and temperature dependence help distinguish mechanisms.

Why?

Heat capacity is a temperature derivative of mean energy. At very low T the upper level is inaccessible, and at very high T its population ratio has nearly saturated. Only the crossover region changes U strongly with T. The finite energy gap creates the temperature scale; degeneracies control how many states compete at each level.

Common misconception

A Schottky peak does not mean a phase transition necessarily occurred; independent levels can produce a broad maximum without singular thermodynamics. Nor does an electronic heat capacity always follow the two-level formula. Metallic electrons and correlated magnetic systems can have different low-temperature laws.

Worked example

Take g₀ = g₁ = 1 and T such that x = Δ/(kT) = 2. Then p₁ = e⁻²/(1+e⁻²) ≈ 0.119. The heat capacity is C V/k = x²p₁(1−p₁) ≈ 4(0.119)(0.881) ≈ 0.420. At much lower T, e^(−x) suppresses this contribution; at much higher T, x² becomes small.

Quick check

1. Why does the two-level heat capacity approach zero at very high temperature even though the upper level remains populated? Answer: Its upper-level population approaches a constant degeneracy-determined fraction, so mean energy stops changing appreciably with T. Heat capacity, the temperature derivative of that energy, tends to zero.

Exam focus

Include degeneracy factors in q and distinguish level probability from individual-state probability. Use Δ in joules per particle with k or molar energy with R consistently. Check both temperature limits and avoid identifying every peak as a phase transition. When interpreting experiments, consider lattice and conduction-electron backgrounds.

Advanced insight

For equal degeneracies, the peak's location is set by a transcendental condition rather than a simple equality of kT and Δ. Fitting peak shape and field dependence can reveal both splitting and degeneracy, but interactions and distributions of local environments can broaden or distort the ideal curve.

Summary

A two-level system with degeneracies g₀,g₁ has q = g₀ + g₁e^(−Δ/kT), U = Δp₁ and C V = k[Δ/(kT)]²p₁(1−p₁). Its heat capacity forms a broad Schottky peak between low-T freeze-out and high-T population saturation. Real solids may also have phonon, metallic-electron or collective magnetic contributions.

Practice questions

1. If g₁ = 3g₀, what is the high-temperature upper-level population fraction? Answer: As T → ∞, the Boltzmann factor tends to one, so p₁ → g₁/(g₀+g₁) = 3/4. Heat capacity still tends to zero because the fraction stops changing with T. 2. A two-level splitting Δ is doubled while the temperature is fixed. What happens to x? Answer: x = Δ/(kT) doubles. If the system was in the low-T tail, the upper level becomes much less populated and its heat-capacity contribution is further suppressed. 3. What low-T dependence often distinguishes simple metallic electronic C V from a Debye lattice contribution? Answer: A simple metal may have an electronic term approximately proportional to T, while a three-dimensional Debye lattice term is proportional to T³. Their coefficients and validity depend on the material.