The file "figures.hdf5" contains all numbers plotted in the figures of the main manuscript. Each entry in the first hierarchy level of the file, e.g. "Figure 2a", corresponds to one subfigure. Each entry in the second hierarchy level (if existent) corresponds to one of the curve/subset of the the respective figure, e.g. if curves for multiple fluences are plotted. The innermost hierarchy level contains the data arrays. The dataset name corresponds to the axis label and units. The file "analyzed.hdf5" contains frequencies, quality factors, stress and Young's modulus fit results for each measured nanostring device on each measured sample. The first hierarchy level represents the sample (A or B). The second hierarchy level represents the accumulated implantation fluence that the sample has seen before the respective measurements. The third hierarchy level represents the write field, i.e. string array index, on the chip (0-3) and the fourth hierarchy level represents the string length within the write field. Each length exists exactly once in each write field. The innermost hierarchy level contains arrays of mode number, frequency, quality factor and "raw data indices" (see next paragraph) representing each measured resonance. The fields Young's modulus and pre-stress are scalars containing the respective fit result and its uncertainty. The file "raw.hdf5" finally contains all raw spectra. The first hierarchy level corresponds to unique indices of the respective measurement. This is intended as a look up table for the "raw data indices" of the "analyzed.hdf5" file. Using the index found in the "analyzed.hdf5", one can obtain the raw frequency sweep data and metadata. The file "srim-VACANCY.txt" is the vacancy output file of the SRIM simulation discussed in the main manuscript.