Respect the synchrotron beam strength: how to model it, measure it and mitigate it for various scientific fields

Extremely bright synchrotron radiation sources give extremely strong intensities at the sample. Lawrence Bright et al. (2021) [J. Synchrotron Rad. (2021), 28, 1377–1385] dive into the details for materials science. I offer a Commentary including a historical context.

publication appeared (Kuzay et al., 2001) and they even gave me a mention in their Contents' page Synopsis, although not quite how I would appreciate it: 'Improved thermal models that include convection are developed which replace Helliwell's adiabatic approximation. Temperature rises of 6 K are calculated for a cryocooled 100 micron thick crystal and of 18 K for a room-temperature air-cooled 1 mm thick crystal for an 8 keV 10 13 photons s À1 mm À2 beam. The importance of internal heat conduction within the crystal is also carefully examined.' Apparently I was to be replaced rather than built upon. I presented the different aspects of the Helliwell & Fourme (1983) ESRP internal report in the journals' literature firstly in Helliwell (1984). The beam heating calculations had of course provided a basic 'worst case' adiabatic model. Kuzay et al. (2001) initiated the modelling of heat transfer from the sample, what I called an isothermal (i.e. steady state) model. We had of course immediately proposed an experimental solution, as mentioned above (Fig. 2). Further modelling studies, and measurements, followed [such as Snell et al. (2007)]. All these, including Kuzay et al. (2001), restricted themselves to modelling the case of a cryocooled protein crystal sample. An important exception was the extensive studies by Cherezov et al. (2002) who were constrained in their structural studies of lipid membranes and mesophases to room temperature.
Lawrence Bright et al. (2021) now lead the way for considering materials science and their samples' beam heating. They provide as context for their initiative a full citing of the macromolecular crystallography beam heating papers, to my knowledge [except the Cherezov et al. (2002) studies] and a selection of the radiation damage papers. This latter is a much bigger literature thanks to the major efforts of the International Symposia on Radiation Damage in Macromolecular Crystallography [see, for example, Garman & Weik (2015)]. As part of the proposed ESRP and proposed New Rings at the Stanford Synchrotron Radiation Laboratory initiative led by Professors Keith Hodgson and Herman Winick, we also made evaluations of overall and specific (disulfide bond breaking) radiation damage (Hedman et al., 1985;Helliwell, 1988).
The context for Lawrence Bright et al. (2021) is that the ESRF's Extremely Bright Source (EBS) upgrade has been completed and even greater spectral source brightnesses duly realized at the turn of the year. They certainly respect their new ESRF EBS synchrotron beam strength. They offer their own ways of how to model it and measure it with different test samples relevant to materials science. They also offer ways to mitigate it. More than that, they suggest ways for experimenters to validate their measurements, learning from those experimental situations where experimenters should be concerned. As for the macromolecular crystallographers, the paper of Halle (2004) on cryo structural artefacts is encouraging us to more firmly establish when those artefacts might be serious. Our proposed sample mount (Fig. 2) may yet come into fashion. Also the helium gas would be in an enclosure allowing control of the crystal's temperature to be at 37 C, i.e. physiological for mammals. That temperature would not allow much headroom as the protein melting temperature for mammalian proteins is around 60 C. The alternative solution to Fig. 2, imported from the XFEL facilities, of serial crystallography of course implicitly assumes that every sample is identical. In a recent study I was thwarted in my detailed analyses because I merged the cryodiffraction data of two apparently identical crystals; I only made headway when I analysed them separately (Govada et al., 2021). Meanwhile, cryocrystallography continues to prove very effective in synchrotron facility measurement pipelines and with more and more remote working, as well as mitigating beam heating. These cryo crystal structures, the vast majority in the Protein Data Bank, have also dominated the training set though for prediction of 3D structures using deep learning from amino acid sequences (Helliwell, 2020).
Finally, I note that in Lawrence Bright et al. (2021) the adiabatic approximation has enjoyed a revival with six detailed mentions, and with much more besides. Suffice to say, Lawrence Bright et al. (2021) is I think an important paper.

Figure 1
A piece of lead, aligned horizontally, at the SRS 9.6 melts due to the white beam.

Figure 2
A strategy to control crystal sample heating was offered for roomtemperature X-ray diffraction measurements at the exceptionally large ESRF X-ray intensities that were planned (Helliwell & Fourme, 1983).