Charge-flipping method

The phase problem in crystallography is on its way to being solved by completely ex novo procedures like the charge flipping method. Fast computers sweep away the problem by carrying out hundreds of thousands of stmcture factor calculations in a few hours. 

There are nonzero chances that the charge flipping algorithm may make direct methods obsolete. Performing as it is, the algorithm is obscure in its fundamentals. No equations, formulas, or proofs are proposed or derived. In the authors own words [14] we admit our own lack of understanding of it beyond the level of intuition. This is a patent case in which the human brain is almost made superfluous by fast computing. Indeed, this facet of scientific activity has few precedents. 

The traditional approach for structure solution follows a close analogy to the analysis of single-crystal XRD data, in that the intensities 1(H) of individual reflections are extracted directly from the powder XRD pattern and are then used in the types of structure solution calculation (e.g. direct methods, Patterson methods or the recently developed charge-flipping methodology [32-34]) that are used for single-crystal XRD data. As discussed above, however, peak overlap in the powder XRD pattern can limit the reliability of the extracted intensities, and uncertainties in the intensities can lead to difficulties in subsequent attempts to solve the structure. As noted above, such problems may be particularly severe in cases of large unit cells and low symmetry, as encountered for most molecular solids. In spite of these intrinsic difficulties, however, there have been several reported successes in the application of traditional techniques for structure solution of molecular solids from powder XRD data. 

---