Andreev bound states are fermionic states localized in weak links between superconductors which can be occupied with spinful quasiparticles. Microwave experiments using superconducting circuits with InAs/Al nanowire Josephson junctions have recently enabled probing and coherent manipulation of Andreev states but have remained limited to zero or small magnetic fields. Here we use a flux-tunable superconducting circuit compatible in magnetic fields up to 1T to perform spectroscopy of spin-polarized Andreev states up to ∼ 250 mT, beyond which the spectrum becomes gapless. We identify singlet and triplet states of two quasiparticles occupying different Andreev states through their dispersion in magnetic field. These states are split by exchange interaction and couple via spinorbit coupling, analogously to two-electron states in quantum dots. We also show that the magnetic field allows to drive a direct spin-flip transition of a single quasiparticle trapped in the junction. Finally, we measure a gate- and field-dependent anomalous phase shift of the Andreev spectrum, of magnitude up to ∼ 0.7π. Our observations demonstrate new ways to manipulate Andreev states in a magnetic field and reveal spin-polarized triplet states that carry supercurrent.
README CONTENTS (also included in the .ZIP)
---------------------------------------------------------------------------
# Data repository for: Microwave spectroscopy of interacting Andreev spins
Manuscript located at https://arxiv.org/abs/2208.11198
Authors: J. J. Wesdorp, F. J. Matute-Caňadas, A. Vaartjes, L. Grünhaupt, T. Laeven, S. Roelofs, L. J. Splitthoff, M. Pita-Vidal, A. Bargerbos, D. J. van Woerkom, P. Krogstrup, L. P. Kouwenhoven, C. K. Andersen, A. Levy Yeyati, B. van Heck, G. de Lange
This is a self-contained repository with the code that analyses the raw data (exported to .nc format).
Required packages are:
* matplotlib
* xarray
* h5py
* h5netcdf
* lmfit
* scipy
* numpy
* jupyter-notebook
* jupyter_contrib_nbextensions (recommended to see the table of content per notebook)
* imageio (for gif generation)
* qcodes (only for the notebook qcodes to xarray conversion, the rest of the code is qcodes independent).
*An environment.yaml file is provided for conda based installations. The exact versions are denoted there.*
Folder structure:
- ./ : The top level folder contains the analysis notebooks that were used to process the data and fit to the minimal theoretical model. The Figures.ipynb contains the figure generation code and can be run standalone.
- data : All data files used to generate the figures in the paper. The data is stored in netcdf format with h5netcdf as engine for loading with xarray.
- data/processing is where intermediate analysis results are stored
- data/processing/phi_0 analysis tool/verification_gifs : The GIF are stored here that show the phi_0 two tone data with traced minima/maxima for verification.
- img : The raw output (separate panels) of the figure generation code is stored in the img folder
- lib: custom code used for analysis and plotting of the data
- kwant_simulation : code used for simulating and generating the figures the non-interacting model used for figure S19-S22.
- kwant_simulation/data : simulation output of the kwant simulations.
One dataset (20201109_1372_parity_selection.nc) for figure S11 is added as a separate file and should be copied to the data folder because of its large file size. This allows for a direct preview of the rest of the repository in the 4TU repository.