Field Note: The Resistivity Minimum
Temperature is dropped. The metal cools. And in the cryostat, something wrong begins to show.
In any ordinary metal, resistivity falls as temperature falls. At room temperature, electrons scatter off phonons — lattice vibrations — and the scattering rate is proportional to T. Drop the temperature and the phonons quiet. The resistivity drops with them. At very low temperature, the phonons are effectively frozen, and residual resistivity — set by static defects and impurities — dominates. The curve flattens. This is Matthiessen's rule. This is standard.
Except when magnetic impurities are present.
Then the curve does something that makes no sense at all. After the resistivity falls with temperature and reaches its residual value, it turns around. It goes back up. Resistance increases as temperature decreases. Below a certain minimum — the Kondo resistivity minimum — the impurity is making the metal harder to conduct. The colder it gets, the more it resists.
Kondo's explanation was radical: the scattering cross-section of a magnetic impurity is not a fixed number. It grows logarithmically as temperature falls. The conduction electrons and the impurity spin exchange virtual excitations, and each exchange adds to the scattering amplitude. The perturbation series looks like:
ρ(T) = ρ₀ + ρ_M [ln(T/T_K) + const]
where ρ₀ is the temperature-independent residual resistivity from non-magnetic defects, ρ_M is the magnitude of the Kondo contribution, and T_K is the Kondo temperature. For T > T_K, the logarithm is positive and the resistivity rises above ρ₀ as we approach T_K from above. For T < T_K, the perturbative formula breaks down entirely — the resistivity saturates at a value set by the formation of the Kondo singlet.
The minimum appears at the crossover. Above T_K, the logarithmic term dominates and the curve bends upward. Below T_K, the screening cloud has formed and the resistivity levels off. The minimum sits roughly at T ≈ T_K.
This was first observed experimentally in the late 1960s. Low-temp resistivity measurements on gold doped with iron showed the minimum unmistakably. The effect was small — fractions of an ohm — but reproducible, systematic, and impossible to explain without the Kondo mechanism. It became one of the canonical solid-state puzzles of the era: a phenomenon so robust it could not be ignored, so simple to describe that its resolution required decades.
The field note is this: the resistivity minimum is the experimental fingerprint of a many-body quantum phase transition that happens inside a normal metal. The metal does not become superconducting. It does not become magnetic. It simply — at low temperature — chooses to screen the impurity. And the resistance minimum is the thermodynamic record of that choice.
Measure the minimum. Read off T_K. The Kondo temperature encodes the strength of the exchange coupling, the density of states at the Fermi level, and the electronic structure of the host. One number, extracted from a resistivity curve, contains the entire physics of the impurity-host interaction.