Single-crystal X-ray crystallography provides the 3D structure of crystalline materials and has long been considered the benchmark technique to determine chemical structure to better than atomic resolution. In recent years, dramatic increases in the X-ray brilliance available at X-ray Free Electron Laser (XFEL) and upgraded synchrotron facilities around the globe are opening up exciting possibilities to study crystals that were previously too small or weakly diffracting, and to conduct very challenging in situ experiments.^1-4 However, these ultrabright light sources often inflict unprecedented levels of X-ray damage on individual crystals, particularly at XFELs where a single X-ray pulse can entirely destroy a crystal.^5,6 In order to fully-utilize the benefits of these new and upgraded light sources, it has become necessary to innovate crystallographic data collection procedures. Multi-crystal methods such as serial crystallography have emerged as a route to outrun decay processes, in which an individual data image is collected from many thousands of microcrystals at different orientations, before the data from all crystals is later combined to generate the full 3D structure.

Serial crystallography was initially developed at XFELs and pioneered by the macromolecular crystallography community.⁷ This is due, in part, to the fact that macromolecular crystals tend to display large unit cells (unit cell = the smallest portion of the crystal that is repeated through translation in 3D to build up the periodic crystal structure) that, in turn, provide a greater number of diffraction reflections on the single data image collected from each individual crystal. This is beneficial for data processing, as the more diffraction reflections that are present then the easier it is to determine the unit cell parameters of that crystal and, consequently, determine its orientation in the X-ray beam. This information is crucial to enable the combination of all the images from different crystals into one, coherent dataset for final structure determination. Conversely, chemical samples such as organic, organometallic or hard inorganic crystal systems more commonly display much smaller unit cells and thus, to-date, present a significant challenge for study via serial crystallography approaches.

In our paper, we present a small-rotative fixed-target synchrotron serial crystallography (SR-FT-SSX for short!) methodology that can enable routine crystal structure determination of chemical samples with small unit cells, which is benchmarked against an archetypal crystal system sodium nitroprusside dihydrate (SNP.2H₂O, Na₂[Fe(CN)₅NO]·2H₂O).^8,9 We adopt a fixed-target method, previously utilized in macromolecular SSX approaches,^10,11 where microcrystals of SNP.2H₂O are dispersed onto a silicon nitride grid and mounted at the data collection position, allowing us to deliver hundreds of crystals into the synchrotron X-ray beam (Figure 1).

Figure 1 a) schematic description of components in a fixed-target grid, b) SEM image of a section of wells containing a single crystal (x270 zoom), c) an image of the fixed-target grid mounted at the data collection position.

Using a specifically-designed XYZ mechanical stage to control the motions of the fixed-target grid, we are able to scan through the grid and, critically, at each well position we implement a small (ca. 5°) rotation of the grid that allows us to collect a short series of several diffraction images on each crystal. This method produces a series of partial datasets, one from every well position, that are more similar to a traditional (though incomplete) single-crystal X-ray dataset and can be processed using routine approaches. This not only increases the number of diffraction reflections that can be used to determine the orientation of each crystal, aiding the combination of datasets into the final complete crystal structure, but it also dramatically improves the data quality as we have a better understanding of the shape of the diffraction peaks (their “profiles”) in 3D due to our small rotative methodology.

In our article we present an iterative workflow to permit the study of microcrystal batches of both known and unknown materials by SR-FT-SSX, which is adaptable to other set-ups. It is our hope that this publication will facilitate further studies of small molecule crystals at other facilities, helping to make multi-crystal methods more accessible for a range of different crystal systems.

To read the full story, find our article in Nature Communications Chemistry:

Small-rotative fixed-target serial synchrotron crystallography (SR-FT-SSX) for molecular crystals