Schmid–Bulgadaev Quantum Phase Transition

• The Schmid–Bulgadaev transition is a zero-temperature superconductor–insulator change driven by dissipation when the environmental resistance exceeds the quantum threshold R₍Q₎.
• The analysis employs boundary sine-Gordon models and renormalization-group techniques to reveal a universal vertical critical line at α = 1 in the (α, E_J/E_C) parameter space.
• Experimental realizations using superconducting transmission lines and on-chip resistors confirm the predicted scaling laws, inelastic scattering signatures, and robust quantum-critical behavior.

The Schmid–Bulgadaev quantum phase transition is a dissipation-induced zero-temperature superconductor–insulator transition of a single Josephson junction coupled to an Ohmic electromagnetic environment. In the standard formulation, the control parameter is the low-frequency environmental impedance, and the phase boundary is set by the quantum resistance for Cooper pairs,

$R_Q=\frac{h}{4e^2}\approx 6.5~\mathrm{k}\Omega .$

For environmental resistance below this threshold the junction remains superconducting, whereas above it quantum fluctuations suppress coherent Cooper-pair tunneling and the junction becomes insulating. Recent work has combined boundary field theory, conformal methods, microwave spectroscopy, and transport experiments to argue that the critical line is universal and vertical in $(\alpha,E_J/E_C)$ space, while other analyses have challenged the scope of the standard Caldeira–Leggett reduction and the equivalence to other dissipative impurity models (Kuzmin et al., 2023, Paris et al., 2024, Kashuba et al., 2023, Subero et al., 24 Sep 2025).

1. Canonical formulation and physical content

The standard model is the resistively shunted Josephson junction, equivalently a Josephson junction coupled to an Ohmic transmission line. One convenient Hamiltonian is

$$H = 4E_C(\hat N-\hat N_0)^2 - E_J \cos \hat\varphi + \sum_q \hbar\omega_q a_q^\dagger a_q,$$

where $E_C=e^2/(2C)$, $E_J$ is the Josephson energy, $\hat\varphi$ is the superconducting phase difference, and $\hat N$ is the number of transferred Cooper pairs. In the Schmid picture, dissipation suppresses quantum phase fluctuations: for $R<R_Q$ the Josephson term is relevant and the junction remains superconducting, while for $R>R_Q$ phase delocalization drives it insulating (Paris et al., 2024).

An equivalent physical picture maps the junction to a quantum particle moving in the periodic potential

$U(\varphi) = -E_J \cos\varphi ,$

with the junction self-capacitance $(\alpha,E_J/E_C)$0 playing the role of inertia and the resistive environment providing dissipation. The competition is between coherent phase dynamics and Ohmic damping. In the superconducting phase, $(\alpha,E_J/E_C)$1 is localized in one minimum of the cosine potential and the junction supports a coherent supercurrent. In the insulating phase, strong quantum fluctuations delocalize $(\alpha,E_J/E_C)$2, suppress coherent supercurrent, and promote phase slips (Kuzmin et al., 2023).

The low-frequency control parameter is often written either as the environmental resistance $(\alpha,E_J/E_C)$3 or as

$(\alpha,E_J/E_C)$4

These parametrizations encode the same threshold: $(\alpha,E_J/E_C)$5 corresponds to $(\alpha,E_J/E_C)$6. In phase–charge dual language, small $(\alpha,E_J/E_C)$7 localizes phase and allows strong charge fluctuations, whereas large $(\alpha,E_J/E_C)$8 localizes charge and produces strong phase fluctuations (Subero et al., 24 Sep 2025).

2. Critical line, boundary field theory, and duality

At weak Josephson coupling, $(\alpha,E_J/E_C)$9, the problem maps to a boundary sine-Gordon model,

$$H = 4E_C(\hat N-\hat N_0)^2 - E_J \cos \hat\varphi + \sum_q \hbar\omega_q a_q^\dagger a_q,$$0

with $$H = 4E_C(\hat N-\hat N_0)^2 - E_J \cos \hat\varphi + \sum_q \hbar\omega_q a_q^\dagger a_q,$$1. In this representation the operator $$H = 4E_C(\hat N-\hat N_0)^2 - E_J \cos \hat\varphi + \sum_q \hbar\omega_q a_q^\dagger a_q,$$2 has scaling dimension $$H = 4E_C(\hat N-\hat N_0)^2 - E_J \cos \hat\varphi + \sum_q \hbar\omega_q a_q^\dagger a_q,$$3, so the Josephson term is relevant for $$H = 4E_C(\hat N-\hat N_0)^2 - E_J \cos \hat\varphi + \sum_q \hbar\omega_q a_q^\dagger a_q,$$4 and irrelevant for $$H = 4E_C(\hat N-\hat N_0)^2 - E_J \cos \hat\varphi + \sum_q \hbar\omega_q a_q^\dagger a_q,$$5, already indicating a transition at $$H = 4E_C(\hat N-\hat N_0)^2 - E_J \cos \hat\varphi + \sum_q \hbar\omega_q a_q^\dagger a_q,$$6 (Paris et al., 2024).

A more refined renormalization-group analysis expands the $$H = 4E_C(\hat N-\hat N_0)^2 - E_J \cos \hat\varphi + \sum_q \hbar\omega_q a_q^\dagger a_q,$$7-function to third order in the Josephson coupling and finds

$$H = 4E_C(\hat N-\hat N_0)^2 - E_J \cos \hat\varphi + \sum_q \hbar\omega_q a_q^\dagger a_q,$$8

At $$H = 4E_C(\hat N-\hat N_0)^2 - E_J \cos \hat\varphi + \sum_q \hbar\omega_q a_q^\dagger a_q,$$9, the $E_C=e^2/(2C)$0-function vanishes through third order in $E_C=e^2/(2C)$1. This nontrivial cancellation strongly supports a fixed line at $E_C=e^2/(2C)$2, rather than an $E_C=e^2/(2C)$3-dependent critical manifold (Paris et al., 2024).

The non-perturbative description of the critical line identifies the low-energy theory at $E_C=e^2/(2C)$4 with a conformally invariant fermionic boundary CFT,

$E_C=e^2/(2C)$5

with $E_C=e^2/(2C)$6 interpolating between Neumann/open and Dirichlet/closed boundary conditions. Because this model is exactly marginal, the mobility satisfies

$E_C=e^2/(2C)$7

so the infrared theory is scale invariant and $E_C=e^2/(2C)$8 labels a continuous line of fixed points. Since $E_C=e^2/(2C)$9 may vary with $E_J$0 while remaining on $E_J$1, the phase boundary is vertical in the $E_J$2 plane. Exact diagonalization of the full shunted-junction Hamiltonian was reported to agree excellently with the predicted critical spectrum even for moderate system sizes (Paris et al., 2024).

The opposite limit, $E_J$3, admits an adiabatic dual description appropriate to the transmon regime. Projecting onto the lowest band yields the dual boundary sine-Gordon model

$E_J$4

with the transmon plasma frequency

$E_J$5

setting the ultraviolet scale. This dual formulation again predicts a transition at $E_J$6, reinforcing the conclusion that the critical line is controlled by the shunt resistance rather than by the microscopic ratio $E_J$7 (Paris et al., 2024).

3. Experimental realizations in engineered and true Ohmic environments

A central experimental realization implemented the dissipative environment not as a lumped resistor but as a long superconducting transmission line. In the Caldeira–Leggett spirit, the resistor was represented by a bath of harmonic modes, here realized physically as standing-wave modes of a long high-impedance line. Each device used two parallel arrays of about $E_J$8 Josephson junctions to form a transmission line with wave impedance $E_J$9, terminated by a small flux-tunable Josephson junction. The junction was implemented as a symmetric SQUID, allowing $\hat\varphi$0 to be tuned by magnetic flux $\hat\varphi$1; at half-integer flux quanta, $\hat\varphi$2, effectively removing the junction and providing a baseline for the line modes (Kuzmin et al., 2023).

This architecture enabled an “inside the resistor” perspective: instead of inferring dissipation only from low-frequency transport, one could observe how the junction scatters environmental photons of the line. The reported transition occurred for $\hat\varphi$3, with a measured critical point

$\hat\varphi$4

The authors emphasized that the phase boundary is essentially independent of junction details such as $\hat\varphi$5 and $\hat\varphi$6 over a wide range, and is set by the environment (Kuzmin et al., 2023).

A later transport experiment used a true resistive environment formed by an on-chip metallic Cr resistor placed only a few micrometers from the junction. Two device classes were studied: a single Al/AlOx/Al Josephson junction and a SQUID with flux-tunable $\hat\varphi$7. In this setting the resistor provided broadband, nearly Ohmic dissipation at low frequency, charge-relaxation channels for tunneling Cooper pairs, and a controlled way to tune $\hat\varphi$8 across $\hat\varphi$9. The reported low-frequency $\hat N$0-$\hat N$1 characteristics showed that $\hat N$2 develops a zero-bias peak for $\hat N$3 and a zero-bias dip for $\hat N$4, with the crossover near

$\hat N$5

In the phase diagram plotted versus $\hat N$6 and $\hat N$7, the transition line occurred near $\hat N$8 and was described as largely insensitive to $\hat N$9 within experimental resolution (Subero et al., 24 Sep 2025).

The SQUID measurements further tested the dependence on Josephson coupling. For $R<R_Q$0, a conductance dip at zero bias persisted across flux values, whereas for $R<R_Q$1 and $R<R_Q$2, a zero-bias peak appeared. This was presented as support for the claim that the decisive control parameter is the environmental resistance relative to $R<R_Q$3, rather than the bare Josephson coupling alone (Subero et al., 24 Sep 2025).

4. Spectroscopic, transport, and scattering signatures

The sharp transition is defined in the zero-frequency, zero-temperature limit. At low frequency, the junction’s elastic response changes discontinuously across the phase boundary. On the superconducting side, $R<R_Q$4, the junction terminates the line like a short circuit and the phase shift is

$R<R_Q$5

On the insulating side, $R<R_Q$6, it acts like an open circuit and

$R<R_Q$7

Away from zero frequency, the finite-frequency response distinguishes inductive and capacitive behavior through the sign of the phase shift: $R<R_Q$8 corresponds to an inductive response and $R<R_Q$9 to a capacitive response (Kuzmin et al., 2023).

Microwave spectroscopy formulates these observables in terms of the reflection coefficient

$R>R_Q$0

and the admittance

$R>R_Q$1

For a finite resonator length $R>R_Q$2, the standing-wave condition is

$R>R_Q$3

and the resonance shift and internal quality factor are given by

$R>R_Q$4

with

$R>R_Q$5

A recurrent misconception is that finite-frequency spectroscopy observes the transition itself. The finite-frequency theory instead states that the Schmid transition is sharply defined only in the dc, $R>R_Q$6 limit; at finite $R>R_Q$7, one observes a crossover in admittance and reflection phase shift (Houzet et al., 2023).

The inelastic channel is especially characteristic. Environmental photons can be down-converted by the nonlinear junction, so one incoming photon scatters into several lower-frequency photons. For frequencies above the renormalized Josephson scale $R>R_Q$8, the inelastic scattering rate scales as

$R>R_Q$9

The same analysis states that $U(\varphi) = -E_J \cos\varphi ,$0 as $U(\varphi) = -E_J \cos\varphi ,$1 and $U(\varphi) = -E_J \cos\varphi ,$2 for $U(\varphi) = -E_J \cos\varphi ,$3. At the critical point $U(\varphi) = -E_J \cos\varphi ,$4, the exponent vanishes and the inelastic down-conversion probability becomes frequency-independent, which was proposed and observed as a marker of quantum-critical behavior (Kuzmin et al., 2023).

Transport measurements provide a complementary scaling signature. At $U(\varphi) = -E_J \cos\varphi ,$5 and very small bias, the current is expected to obey

$U(\varphi) = -E_J \cos\varphi ,$6

The critical condition is

$U(\varphi) = -E_J \cos\varphi ,$7

Reported extracted exponents followed this trend and were used to support the identification of the low-bias crossover with the dissipation-driven quantum transition (Subero et al., 24 Sep 2025).

5. Critical behavior, ideal resistance, and universal scaling

At the phase boundary the junction is neither a pure short nor a pure open circuit. The reported interpretation is that inductive and capacitive responses both diverge because of strong quantum fluctuations, and the junction appears as an ideal resistance to the line modes. In this sense, “junction behaves as ideal resistance” means that it no longer supports coherent superconducting phase locking, no longer acts as a static capacitor or inductor, and instead becomes a scale-invariant dissipative scatterer for photons. For $U(\varphi) = -E_J \cos\varphi ,$8, the effective resistance at the critical point was noted to be approximately

$U(\varphi) = -E_J \cos\varphi ,$9

highlighting the role of nonlinear quantum renormalization in the critical response (Kuzmin et al., 2023).

Finite-frequency theory organizes this behavior in terms of crossover scales and universal scaling functions. In the transmon-like regime $(\alpha,E_J/E_C)$00, the phase shift takes the form

$(\alpha,E_J/E_C)$01

while in the charge-qubit regime $(\alpha,E_J/E_C)$02,

$(\alpha,E_J/E_C)$03

The two descriptions are related by duality,

$(\alpha,E_J/E_C)$04

These scaling functions encode the finite-frequency crossover between inductive and capacitive response, rather than a sharp singularity (Houzet et al., 2023).

Near the transition, inelastic effects become central. The imaginary part of the complex phase shift determines inelastic scattering, and the finite-frequency analysis states that it becomes significant near the crossover scale. At $(\alpha,E_J/E_C)$05, scattering is fully inelastic in the scaling limit, while a small deviation from the critical point yields

$(\alpha,E_J/E_C)$06

This formulation connects the scale-invariant critical line to experimentally accessible microwave phase shifts and internal losses (Houzet et al., 2023).

6. Controversies, alternative formulations, and scope

Despite recent experimental support for a universal threshold at $(\alpha,E_J/E_C)$07, the Schmid–Bulgadaev transition remains theoretically contested at the level of modeling assumptions. One prominent critique argues that the standard Caldeira–Leggett description is only exact when one applies approximations that decompactify the superconducting phase. If phase compactness is retained, the renormalization-group flow changes and the phase diagram depends on four parameters rather than two: $(\alpha,E_J/E_C)$08, $(\alpha,E_J/E_C)$09, the capacitive coupling $(\alpha,E_J/E_C)$10, and the electric/contact coupling $(\alpha,E_J/E_C)$11. In this treatment, an SB-like transition survives in the transmon regime, but the critical parameter is controlled exclusively by the capacitive coupling, not by the full measured resistance. In the Cooper-pair-box regime, the model maps to an anisotropic Kondo problem, and a pseudoferromagnetic phase is argued not to be allowed for regular electrostatic interactions (Kashuba et al., 2023).

This compact-phase critique directly challenges the conventional statement that the total low-frequency resistance alone determines the transition. The same analysis argues that for normal-metal and Josephson-junction-array resistors, the Caldeira–Leggett bath is not generally an exact microscopic description once charge quantization and phase winding are kept explicit. A plausible implication is that the universality of the experimentally inferred threshold and the exactness of the underlying dissipative field theory are separable questions (Kashuba et al., 2023).

A separate controversy concerns the relation between the Schmid–Bulgadaev transition in Josephson circuits and the spin-boson model. One line of work on the spin-boson Hamiltonian

$(\alpha,E_J/E_C)$12

argues that, for any nonzero tunneling amplitude $(\alpha,E_J/E_C)$13, the exact ground state has definite parity and lies below the lowest possible opposite-parity degenerate energy, so that a genuine parity-breaking ground-state quantum phase transition does not occur. In that interpretation, the transition reported in numerical studies is a finite-truncation artifact caused by symmetry breaking in a truncated bosonic basis (Liu et al., 2013).

That negative result is not a statement about the Josephson-junction experiments themselves, which explicitly probe a resistively shunted junction or transmission-line realization. It does, however, show that not every dissipative impurity formulation is taken to be interchangeable in the literature. The modern discussion of the Schmid–Bulgadaev transition therefore includes both strong evidence for a universal critical resistance in Josephson circuits and continuing disagreement over the exact microscopic conditions under which the standard dissipative field theory is valid (Kuzmin et al., 2023, Subero et al., 24 Sep 2025, Liu et al., 2013).