Crystallography is the science of determining how atoms are arranged inside solid materials. By directing beams of radiation at a sample and analyzing the patterns that bounce back, researchers can build a three-dimensional map of its atomic structure. This technique has shaped modern medicine, electronics, and chemistry, revealing the structures of everything from DNA to the proteins targeted by new drugs.
How Diffraction Reveals Atomic Structure
The core principle behind crystallography is diffraction. When a beam of X-rays hits a crystal, the regularly spaced atoms scatter the incoming waves. In most directions, those scattered waves cancel each other out. But at specific angles, they line up and reinforce one another, producing bright spots on a detector. The positions and intensities of those spots encode the distances between atoms inside the crystal.
This relationship is captured by Bragg’s Law, a simple equation worked out in 1913: 2d sin(θ) = nλ. In plain terms, it says that when the path difference between X-rays bouncing off neighboring layers of atoms equals a whole number of wavelengths, the reflected waves combine into a strong signal. By measuring the angles at which these strong signals appear, researchers can calculate the spacing between atomic layers and, with enough data points, reconstruct the full three-dimensional arrangement of atoms.
Three Main Types of Diffraction
X-ray diffraction is by far the most common approach and the primary source of three-dimensional structural data worldwide. It works well for locating heavier atoms like carbon, nitrogen, and oxygen, and the results feed into massive databases: the Cambridge Structural Database for small molecules and the Protein Data Bank for biological structures. The limitation is that X-rays scatter off electron clouds, so the lightest atom, hydrogen, is hard to pin down precisely.
Neutron diffraction fills that gap. Neutrons scatter off atomic nuclei rather than electrons, making them far better at locating hydrogen atoms and distinguishing between elements that look similar to X-rays. The trade-off is practical: you need much larger crystals and access to a neutron source like a research reactor or spallation facility. Only about 0.3% of the 1.5 million crystal structures determined so far have used neutron diffraction.
Electron diffraction works on samples too small for conventional X-ray analysis. Because electrons interact with matter much more strongly than X-rays do, this technique can extract structural information from nanoscale crystals, thin films, and surfaces.
Growing Crystals for Analysis
Before any diffraction experiment can happen, researchers need a well-ordered crystal. For proteins, this is often the hardest step. The two main techniques are vapor diffusion and batch crystallization.
In vapor diffusion, a small drop containing a mix of protein and a chemical precipitant is sealed in a chamber alongside a reservoir of pure precipitant. Water slowly evaporates from the drop until the protein concentration rises enough to trigger crystal formation. Researchers set this up as either a “hanging drop” suspended from a cover slip or a “sitting drop” resting on a small shelf. In batch crystallization, the protein is mixed directly with precipitant under a layer of oil, which controls how fast water escapes. Different oil mixtures produce different evaporation rates, giving researchers a way to fine-tune the process. Getting conditions right often requires screening hundreds of combinations of temperature, pH, and precipitant concentration.
Synchrotrons and Modern X-Ray Sources
Early crystallography relied on laboratory X-ray tubes, which bombarded metal targets with electrons to generate radiation. These sources were adequate for small molecules but painfully slow for large biological structures, sometimes requiring weeks to collect a single dataset.
Synchrotrons changed the field entirely starting in the 1970s. These particle accelerators whip electrons around a ring at nearly the speed of light, generating X-ray beams millions of times more intense than laboratory tubes. Third-generation synchrotron rings produce beams 20 million times brighter than the rotating anode sources they replaced. That intensity means researchers can study much smaller crystals, collect data in minutes instead of days, and resolve structures at atomic detail (below 1.2 angstroms).
Synchrotrons also let researchers tune the X-ray wavelength to the precise value needed for a given experiment, something impossible with a fixed laboratory source. And because synchrotron radiation arrives in ultrashort pulses lasting picoseconds, it can capture snapshots of chemical reactions as they happen inside a crystal. Serial femtosecond crystallography, performed at X-ray free-electron lasers, pushes this even further, filming molecular processes in real time.
Drug Design and Medical Impact
One of crystallography’s most consequential applications is structure-based drug design. By solving the three-dimensional structure of a disease-related protein, researchers can see the exact shape of the pocket where a drug molecule needs to bind. They can then design or screen compounds that fit that pocket like a key in a lock.
Drug candidates are introduced to protein crystals either by growing crystals with the drug already present (co-crystallization) or by soaking the drug into preformed crystals. Room-temperature crystallography has proven especially revealing here. In one recent case, researchers studying a cancer-related enzyme found that collecting data at room temperature, rather than the standard frozen conditions, revealed a previously hidden change in how a drug molecule sat in its binding pocket. That conformational shift explained why the drug was less potent than competitors in the same chemical class, something that traditional frozen-crystal methods had completely missed. Room-temperature experiments can also uncover allosteric sites, secondary pockets on a protein that could be targeted by entirely new drugs.
The technique’s medical legacy stretches back decades. Dorothy Hodgkin used crystallography to solve the structure of penicillin in the early 1940s, revealing the unexpected four-membered ring at its core and opening the door to semisynthetic antibiotics. She spent 35 years working on insulin before publishing its structure in 1969. And the double-helix structure of DNA, announced by Watson and Crick in 1953, relied on X-ray diffraction photographs taken by Rosalind Franklin and Maurice Wilkins at King’s College London.
Applications in Materials Science
Crystallography is equally central to developing the materials that power modern technology. Semiconductor manufacturing depends on growing crystals with precisely controlled atomic arrangements, because even tiny defects in a silicon or gallium arsenide crystal can degrade the performance of a chip or sensor. Crystallographic analysis measures defect density, compositional uniformity, and lattice quality, all of which determine whether a batch of semiconductor material will function reliably in devices like integrated circuits, radiation detectors, or infrared sensors.
Research into growing semiconductor crystals in microgravity has highlighted just how sensitive crystal quality is to growth conditions. Crystals produced aboard the International Space Station show fewer impurity striations, lower defect densities, smoother surfaces, and more uniform composition than their Earth-grown counterparts. In one set of experiments, transistors fabricated from microgravity-grown crystals outperformed the best analog circuits made from terrestrial wafers. These findings underscore how crystallographic measurements guide the optimization of real-world manufacturing processes.
AI Prediction and Experimental Crystallography
The arrival of AlphaFold, an AI system that predicts protein structures from their amino acid sequences, has reshaped how crystallographers work without replacing what they do. AlphaFold predictions now routinely serve as starting models for solving crystal structures, a step called molecular replacement that used to require painstaking experimental groundwork. In one case involving a large bacterial protein complex, researchers hit a wall because the complex was too large and too flexible to solve by standard methods. An AlphaFold prediction filled in the missing information and allowed them to complete the structure.
Where AI falls short is in capturing how proteins actually move. AlphaFold produces a single static prediction, while real proteins flex, shift, and adopt multiple shapes depending on their environment. Understanding those dynamics, which drug binding sites open and close, which regions are rigid versus floppy, still requires experimental techniques like crystallography, especially the newer room-temperature and time-resolved methods that can track motion at the atomic level.
Beyond Traditional Crystals
The field has expanded well beyond the study of perfectly ordered crystals. The formal definition of a crystal was revised to accommodate quasicrystals, materials with ordered but non-repeating atomic patterns that were once thought impossible. Pair distribution function analysis, a technique that extracts structural information from less-ordered materials, now lets researchers study nanoparticles, glasses, and even liquids. Surface crystallography examines the one- and two-dimensional structures that govern the behavior of nanomaterials and thin films. Crystallography, in its modern form, is less a single technique than a family of methods united by a common goal: mapping the arrangement of atoms in matter, whether that matter is a flawless gemstone or a disordered glass.