Microscopic origin of 1/f flux noise in SQUIDs and qubits
In plain words
Superconducting loops show slow magnetic-flux fluctuations of nearly the same size whatever the loop size or material. Unpaired electron spins on surfaces are blamed, but what they are and how they produce a 1/f spectrum is not settled.
Precise statement
Low-frequency flux noise S_Phi(1 Hz)^(1/2) ~ 1-10 micro-Phi_0/Hz^(1/2) ($\Phi_0=hc/2e$), weakly dependent on loop area, is attributed to surface spins of density $\sim 5e13\,\mathrm{cm}^{-2}$ (Koch, DiVincenzo and Clarke 2007). Identify the spin species (adsorbed O2, hydrogen, dangling bonds) and the dynamics producing $1/f$ (spin clusters, spin-glass dynamics, spin diffusion), consistent with the measured temperature dependence and the correlation between flux noise and inductance noise.
What would settle it
Surface treatments that remove a candidate species and reduce flux noise by the predicted amount, together with a spin model reproducing spectrum, temperature dependence and cross-correlations.
Status in the literature
Experiments since 2016 link part of the noise to adsorbed molecular oxygen; the spin dynamics producing 1/f remain disputed.