Predicting the effect of chemical factors on the pH of crystallisation trials

In macromolecular crystallisation, success is often dependent on the pH of the experiment. However, little is known about the pH of reagents used and it is generally assumed that the pH of the experiment will closely match that of any buffering chemical in the solution. We use a large data set of experimentally measured solution pH values to show that this assumption can be very wrong and generate a model which can be used to successfully predict the overall solution pH of a crystallisation experiment. Further, we investigate the time dependence of the pH of some polyethylene glycol polymers widely used in protein crystallisation under different storage conditions.