Seeing Atoms Through Diffraction
X-ray crystallography determines the three-dimensional arrangement of atoms within a crystal by analyzing the pattern of X-rays diffracted from its regularly spaced lattice planes. Since Max von Laue first demonstrated X-ray diffraction in 1912, the technique has revealed the structures of over one million crystals, from simple salts to complex proteins. It remains the most definitive method for determining molecular structure at atomic resolution.
The impact of X-ray crystallography on science is immeasurable. It gave us the double helix of DNA (Watson and Crick, 1953, using Rosalind Franklin's diffraction data), the structure of hemoglobin (Perutz, 1959), the mechanism of ribosomal protein synthesis (Ramakrishnan, Steitz, and Yonath, 2009 Nobel Prize), and the atomic details of countless drug-receptor interactions that drive modern pharmaceutical development.
The Physics of Diffraction
When a beam of monochromatic X-rays strikes a crystal, the electrons in each atom scatter the X-rays in all directions. Because atoms in a crystal are arranged in a periodic lattice, scattered waves from parallel planes of atoms interfere constructively only at specific angles. This condition is described by Bragg's law: n x lambda = 2d x sin(theta), where lambda is the X-ray wavelength, d is the spacing between lattice planes, theta is the angle of incidence, and n is an integer.
Typical X-ray wavelengths (0.5-2.5 angstroms) are comparable to interatomic distances, which is precisely why X-rays are useful for probing crystal structures. Copper K-alpha radiation (1.5418 angstroms) and molybdenum K-alpha radiation (0.7107 angstroms) are the most commonly used laboratory sources. Synchrotron radiation facilities produce extremely intense, tunable X-ray beams that have revolutionized structural biology.
From Diffraction Pattern to Structure
Determining a crystal structure from diffraction data involves several steps. First, the crystal is mounted and rotated in the X-ray beam while a detector (area detector or image plate) records the positions and intensities of thousands of diffracted spots, called reflections. The positions of the reflections reveal the dimensions and symmetry of the unit cell -- the smallest repeating unit of the crystal lattice.
The intensities of the reflections are proportional to the square of the structure factor, which encodes the positions of all atoms in the unit cell. However, the detector records only intensities, not the phases of the diffracted waves. This phase problem is the central computational challenge of crystallography. For small molecules, phases can be determined by direct methods -- mathematical algorithms that exploit statistical relationships among intensities. For proteins, phases are often obtained by molecular replacement (using a known related structure as a starting model), isomorphous replacement (introducing heavy atoms like mercury or selenium), or anomalous scattering (exploiting wavelength-dependent absorption by certain atoms).
Once phases are obtained, an electron density map is calculated by Fourier synthesis. Atomic positions are fitted into this map, and the model is refined by adjusting coordinates, thermal parameters, and occupancies to minimize the difference between observed and calculated intensities. The quality of the final structure is assessed by the R-factor, which measures this agreement; values below 5% are typical for well-determined small-molecule structures, while protein structures typically achieve R-factors of 15-25%.
Crystal Growth
The most challenging aspect of crystallography is often growing suitable crystals. Crystals must be single (not polycrystalline), sufficiently large (typically at least 0.1 mm in each dimension for laboratory sources, though synchrotron microcrystallography can work with crystals as small as 5 micrometers), and well-ordered internally.
For small molecules, crystals are grown by slow evaporation, vapor diffusion, or cooling of saturated solutions. For proteins, the process is far more difficult. Protein crystals are grown by slowly increasing the concentration of a precipitant (ammonium sulfate, polyethylene glycol, or organic solvents) around the protein in solution. Because proteins are large, flexible, and irregularly shaped, growing diffraction-quality crystals can take weeks or months of optimization, and many proteins resist crystallization entirely.
Protein Crystallography
The determination of protein crystal structures has transformed biology. As of 2025, the Protein Data Bank (PDB) contains over 220,000 experimentally determined structures, the majority solved by X-ray crystallography. These structures reveal how enzymes catalyze reactions, how antibodies recognize antigens, how ion channels select specific ions, and how drug molecules bind to their targets.
Structure-based drug design uses crystallographic data to design molecules that fit precisely into the active site of a target protein. The development of HIV protease inhibitors (saquinavir, ritonavir, and others) in the 1990s was one of the first great successes of this approach and helped transform HIV/AIDS from a death sentence into a manageable chronic condition.
Limitations and Complementary Methods
X-ray crystallography provides a static, time-averaged picture. It cannot easily capture conformational dynamics, transient intermediates, or the behavior of intrinsically disordered regions. Cryo-electron microscopy (cryo-EM) has emerged as a powerful complement, determining structures of large complexes and membrane proteins that resist crystallization. Nevertheless, for atomic-resolution structures of small molecules and well-ordered proteins, X-ray crystallography remains unmatched.