You are absolutely correct in mentioning that the capillaries are 100 microns by 150 microns. If the capillaries become much larger, convective phenomena begin to dominate, which we aim to avoid.

In general crystals grown within the Crystal Former channels are up to 100 microns in size. Fortunately such crystals are large enough to give good diffraction data at most beam lines that are not optimized for microcrystals (such as the X6A at Brookhaven). In fact Vivian Stojanoff, Jean Jakoncic and myself were able to collect full datasets from individual crystals obtained in the initial screening of some of the model proteins we studied. The resolution of the refined structures ranged from 1.3 – 1.5 Angstrom.