The Kondo effect is one of condensed matter physics' cleanest examples of entanglement doing mechanical work. A magnetic impurity — a lone spin — sits in a metal. Below a characteristic temperature T_K, conduction electrons collectively screen the impurity by forming an entangled singlet. The impurity's magnetism vanishes. Its spectral function develops a sharp resonance at the Fermi level. The physics has been understood since the 1960s: antiferromagnetic exchange coupling drives spin-flip scattering that builds impurity-bath entanglement until the singlet forms.
Apply a magnetic field of order 0.5 T_K and the Kondo peak splits and vanishes. The impurity polarizes along the field, decouples from the bath, and its magnetic moment reappears. This field-driven breakdown has been observed in quantum dots for over two decades. The standard explanation is energetic: the Zeeman splitting competes with the Kondo binding energy. When the field wins, the singlet dissolves.
Debata, Mukherjee, and Lal (arXiv 2602.20600, February 2026) reframe this as a measurement problem. The magnetic field isn't just competing energetically with the Kondo coupling — it is continuously observing the impurity spin. The Zeeman term B * S_z preferentially projects the impurity into a definite polarization state, suppressing the transverse spin-flip fluctuations that the Kondo coupling needs to build entanglement. The field is a deterministic continuous measurement, collapsing the impurity spin toward a classical state at a rate set by B.
Their coupled RG flow equations make the competition precise. The Kondo coupling J and the field B flow in opposite directions, governed by a common denominator that encodes their rivalry. When J dominates, the field flows to zero and the singlet forms — maximum entanglement. When B dominates, J flows to zero and the impurity polarizes — zero entanglement. The entanglement entropy is the order parameter: it jumps from ln 2 to 0 at the critical field.
The structure maps directly onto measurement-induced phase transitions in quantum circuits. In that setting, unitary gates generate entanglement while projective measurements destroy it. Below a critical measurement rate, entanglement survives (volume-law phase). Above it, entanglement is suppressed (area-law phase). The Kondo system does the same thing: the exchange coupling generates entanglement, the field destroys it, and there's a critical field where the transition occurs.
But there's a key difference. In the circuit model, measurements are stochastic — each one produces a random Born-rule outcome, and the transition is a property of trajectory-averaged entanglement. The magnetic field is deterministic. The outcome is fixed by the field direction. No trajectory averaging, no randomness. This makes the Kondo breakdown a cleaner realization of the measurement-entanglement competition than the circuit model that inspired the analogy.
At the critical point, the effective Hamiltonian takes an anisotropic Heisenberg form — distinct from the isotropic Fermi liquid that describes the screened phase. The quasiparticle residue vanishes. Non-Fermi liquid excitations emerge from an orthogonality catastrophe between degenerate ground states. The critical point isn't just a boundary between two known phases — it's a distinct state with its own excitation spectrum.
The paper doesn't discover new physics. The field-driven Kondo breakdown has been measured, simulated, and analytically solved by multiple methods. What it discovers is a new way of understanding known physics: a condensed matter problem is secretly a quantum measurement problem, and the magnetic field is secretly an observer.
Note (2026-11-11): checked against the full text of the paper (Debata, Mukherjee and Lal, arXiv 2602.20600, a SciPost Physics submission). The framing is as the essay says: the Kondo coupling builds entanglement between the impurity and the bath, the local field acts as a continuous observer whose outcome is deterministic (the paper contrasts this with the random outcomes in circuit models), unitary-RG flow equations for the coupling and the field separate a screened phase from a polarized local-moment phase, the impurity entanglement entropy is ln 2 in the screened phase and zero in the local-moment phase, the quasiparticle residue vanishes toward the transition, the effective Hamiltonian at the transition is an anisotropic Heisenberg model for two effective spins, and the paper attributes non-Fermi-liquid behaviour there to an orthogonality catastrophe between the two degenerate ground states. Four points differ. (1) The essay says the field-driven breakdown 'has been observed in quantum dots for over two decades'. The paper says the splitting of the spectral density above a field of about 0.5 T_K was 'first demonstrated theoretically' and has since been reported in several theoretical and 'even a few recent experimental' studies; quantum dots appear in the paper as a setting where conductance should be computed in future work. (2) The essay says the Kondo peak 'splits and vanishes' at a field of order 0.5 T_K; the paper cites the splitting at about that field and describes the weight of the central peak (the quasiparticle residue) as vanishing as the field is tuned toward the transition, which is a statement about the transition in its own calculation, not about the 0.5 T_K splitting. (3) 'The entanglement entropy jumps from ln 2 to 0 at the critical field' compresses the paper's figures, which show the value at ln 2 in the screened phase and zero in the local-moment phase for the impurity along the RG flow and as bath sites are added; I found no statement that this is a jump or a particular order of transition, and I found no mention of volume-law or area-law phases (the description of the circuit transition is the essay's). (4) 'A cleaner realization of the measurement-entanglement competition than the circuit model' and 'the paper doesn't discover new physics' are the essay's judgments; the paper itself calls the non-Fermi liquid at the transition novel. I did not check the essay's statement of the coupled flow equations and their common denominator.