All the data was acquired in an FEI Sirion 200 field-emission scanning electron microscope at the University of Strathclyde. The file names correspond to figure numbers in the review article by G. Naresh-Kumar et al. (2016) published in the journal Materials Science in Semiconductor Processing. The images have been cropped in the paper, but full uncropped data is given here. Different cropped areas of Fig_4a.tif and Fig_4b.tif have also been published in G. Naresh-Kumar et al. (2014) Microscopy and Microanalysis 20 55-60, where they appear in Figure 3.

ECCI data:

All the ECC images were acquired with an electron beam spot of ≈ 4 nm, a beam current of  ≈ 2.5 nA and a beam divergence of ≈ 4 mrad. The dimensions may be determined from the scale bar, which can applied in both vertical and horizontal directions since the images have been tilt-corrected. Three of the ECC images have no scale bar shown; these images have the following dimensions:
Fig_1c_2b.tif magnification is ×35k, scan area 3.6 × 2.7 µm
Fig_1d.tif magnification is ×25k, scan area is 4.95 × 3.7 µm
Fig_2c.tif magnification is ×20k, scan area is 6.3 × 4.75 µm

CL data:

The CL data was acquired with the following experimental conditions:
5 keV beam energy
≈ 6 nA beam current
≈ 17 nm sprobe diameter
400 × 340 spatial pixels
40 nm per pixel step size
50 ms acquisition time per spectrum

The CL intensity values given are the integrated intensity of the 3.4 eV GaN near-band-edge CL peak in units of CCD counts, determined by fitting a Voigt peak profile to each spectrum in the CL hyperspectral image. The TIFF file shows the data truncated to an 8-bit integer format; the original data is given as floating-point values in the ASCII text file.