research papers
This paper discusses the concept that crystallinity is not an essential requirement for applying the techniques of X-ray crystal structure analysis. Assuming this to be true, the removal of crystallinity as a prerequisite for the techniques would allow the imaging of structures well beyond the present range of sizes accessible to X-ray crystallography. An example of an imageable structure could be a single small biological cell, containing perhaps 1013-1014 Da. The proposed concept differs from the usual diffraction method of studying noncrystalline structure, i.e. small-angle scattering, in carrying out the diffraction experiment and subsequent processing as if the structure being studied were in fact the asymmetric unit of a crystal: i.e. orienting the structure in all directions in the X-ray beam needed to explore its Fourier transform (F transform), and phasing and inverting the transform to obtain the electron-density image of the structure. The one actual difference from the crystal case is that the F transform is faint and also continuous, rather than displaying discrete intense Bragg spots. As a result, to get a readable pattern, the structure must be exposed to high levels of radiation. This last fact creates the principal limitation of the technique. With single air-dried biological cells at room temperature as diffracting specimens and soft X-rays in the wavelength range 18-32 Å, patterns to date have not been observed beyond resolutions of 140-300 Å before radiation damage has become evident. At this resolution, the technique nevertheless would lie on the same curve of resolution vs specimen size as do the existing major imaging techniques of X-ray crystallography, electron microscopy and light microscopy, falling directly between the latter two. Thus, X-ray diffractive imaging is not destroyed by the withdrawal of crystallinity but instead is shifted to a new size range of structures, which have hitherto been somewhat inaccessible to imaging. A method for phasing the diffraction pattern, based on the ability to sample the pattern more finely than in the case of the crystalline specimen, is giving good results in preliminary testing. The principal need at present is for better instrumentation for collecting the diffraction data, including the additional motions needed for collecting data in three dimensions.