Exposure of a sample to an x-ray beam can cause damage to the sample. This is called radiation damage, or beam damage, or sample damage. It is important to be aware of this possibility, in order to avoid data that is contaminated by damage artifacts.

The overall mechanism of beam damage is complex, stochastic, and likely involves multiple pathways. It is thus not entirely understood. Damage may occur through absorption, with secondary bond-breaking, radical formation, and/or heating causing local damage. It is also possible that damage occurs through generation of secondary photons or photo-electrons.

To a first approximation, damage is expected to only occur within the beam footprint. On the other hand, it may be that x-rays become scattered far from the beam footprint, leading to damage in other areas. With respect to x-ray energy, there is a tradeoff: higher-energy x-rays are more damaging per photon, yet they are generally more penetrating (they interact weakly, and are not strongly absorbed). As such, lower-energy x-rays are likely to cause more sample damage. X-ray energies near an absorption edge would be expected to be especially dangerous (c.f. resonant scattering).

## Practical

In practice, one can attempt to assess sample damage by measuring a spot repeatedly, and observing what exposure timescale is required to begin altering the scattering pattern. For many BioSAXS and soft-matter samples, one can measure the average particle size using a Guinier analysis; beam damage causes aggregation and a measurable increase in $\scriptstyle R_g$.

For a 'typical' synchrotron beam (a ~100 micron beam with ~1012 photons at ~13 keV), hard-matter systems are very robust against beam damage (long exposures no problem), while soft-matter materials will exhibit damage on a timescale of seconds to a few minutes. Higher-flux sources of course generate damage more quickly (sub-second). Ambient oxygen and/or water may accelerate damage (thus, measurements under vacuum may mitigate some kinds of beam damage). Samples containing solvent (residual solvent, swollen with vapor, immersed in bulk solvent) are typically much more sensitive (and will exhibit damage after 1-30 s exposure). Cryo-cooling of the sample (e.g. with liquid nitrogen) can be used to minimize beam damage.