October 9, 2019 Dataset Open Access

Three-dimensional subnanoscale imaging of unit cell doubling due to octahedral tilting and cation modulation in strained perovskite thin films

Nord, Magnus; Ross, Andrew; McGrouther, Damien; Barthel, Juri; Moreau, Magnus; Hallsteinsen, Ingrid; Tybell, Thomas; MacLaren, Ian

Transmission electron microscopy data used in the journal publication "Three-dimensional subnanoscale imaging of unit cell doubling due to octahedraltilting and cation modulation in strained perovskite thin films"

Data files

There are two data types:

• Scanning TEM (STEM) diffraction patterns acquired with a Medipix3 detector (Merlin): m004_LSMO_LFO_STO_medipix.hdf5
  • Acquired on a probe corrected Jeol ARM200CF
  • Acceleration voltage: 200 kV
  • Convergence semi-angle: 20.4 mrad (calibrated using the SrTiO₃ substrate HOLZ ring)
  • Detector calibration: 1.357 mrad per pixel (calibrated using the SrTiO₃ substrate HOLZ ring)
• Atomic resolution STEM data, both annular dark field (ADF) and annular bright field (ABF), which were acquired simultaneously: s007_ADF.hdf5, s007_ABF.hdf5

The data can be loaded in python using h5py.

For the Medipix3 data:

import h5py
f = h5py.File('m004_LSMO_LFO_STO_medipix.hdf5', mode='r')
data = f['fpd_expt/fpd_data/data']
data_subset = data[0:16, 0:16, :, :]

For the STEM-ADF or STEM-ABF data:

import h5py
f = h5py.File('s007_ADF.hdf5', mode='r')
data = f['Experiments/__unnamed__/data']

Exploring the Medipix3 dataset lazily, i.e. without loading the whole dataset into memory at the same time. Using pixStem:

import pixstem.api as ps
s = ps.load_ps_signal("003_stripe1.hdf5", lazy=True)
s.plot()

Loading the STEM-ADF or STEM-ABF data using HyperSpy, which automatically loads the probe scaling:

import hyperspy.api as hs
s = hs.load("s007_ADF.hdf5")
s.plot()

Processing files

All the TEM data has been processed using python scripts, which is named based on the type of processing:

• d00N_...: Medipix3 data processing
• a00N_...: Atomic resolution STEM-ADF and STEM-ABF processing using Atomap

The scripts generate intermediate files, which are saved in folders with the same prefix as the scripts. So the d001_... script makes a folder named d001_... . These intermediate files are included here as zip-files, since Zenodo doesn't support folder structures.

The python libraries required to run the scripts are listed in requirements.txt. Newer versions of the libraries will most likely also work.

To setup the python environment with the required libraries, and run all the scripts:

pip3 install -r requirements.txt
python3 run_all_scripts.py