← all systems · see it in The System
Crystallography
Crystallography is the branch of science devoted to the study of molecular and crystalline structure and properties. Its methods rest mainly on the diffraction patterns of a sample struck by a beam (X-rays most often, also electrons or neutrons): the angles and intensities of the diffracted beams are measured, phases are found, and a three-dimensional picture of the density of electrons (or nuclei) and the positions of the atoms is built, refined and deposited for others to use.
Properties published scienceour proposal
Sourced fields, and our own readings.
| aliases | Crystallography (the practice) |
| born at | 1669 (Steno) or 1912 (Laue); contested (C-1526) |
| born precision | years |
| boundary | Inside: the practice of reading how atoms are arranged in solids (and in ordered aggregates of large molecules) from the way a beam is diffracted by them: growing and mounting crystals, the beams (X-rays, electrons, neutrons) and their sources, detectors, indexing and scaling, the phase problem and its solutions, density maps, model building, refinement and validation, the symmetry theory (lattices, point and space groups, reciprocal space) and notation, the reference tables and databases, the people, unions and journals, and the uses (bonding, minerals and alloys, drugs, biological molecules, training structure predictors). Outside: the crystals, atoms and molecules themselves (their own walks), single-particle cryo-EM and solution NMR (rival structure methods, booked as partners and a competition), and what users decide with a structure beyond reading it. Whether crystallography takes in all structure determination is contested (C-1525); what counts as a crystal changed in 1992 (C-1243). |
| nested level | A branch of science shared by physics, chemistry, mineralogy, materials science and biology, carried by people, instruments and large facilities (synchrotrons, free-electron lasers, neutron sources); it works on the crystal's lattice through the electromagnetic field (X-rays, electrons) and the nuclear force (neutrons), and its output (deposited structures) feeds chemistry, biology, drug design and, since 2021, protein structure prediction. |
| scale | social |
Measures (47) published science
| first X-ray diffraction experiment | 1912 year |
| first electron diffraction | 1927 year |
| first macromolecular crystal structure (myoglobin) | 1958 year |
| C-C single bond length in diamond | about 1.52 Å |
| space groups (all) | 230 count |
| space groups allowed for chiral protein crystals | 65 count |
| typical crystal size for single-crystal X-ray work | larger than 0.1 mm in all dimensions |
| crystal centring in the beam | ~25 µm |
| rotation range for a full data set | slightly more than 180 degrees |
| robotic crystallization drop volume | order of 100 nL (10-fold less protein than ~1 µL by hand) |
| direct methods: size and resolution limits | <1000 non-H atoms; better than 1.4 Å atoms; Å |
| R-free rule of thumb | about resolution/10 (0.2 at 2 Å) dimensionless |
| Cambridge Structural Database size (June 2019) | over 1,000,000 structures |
| X-ray crystal structures of biological molecules | 190,000 structures |
| share of PDB deposits by X-ray crystallography, 1971-2014 | 88.8 % |
| crystal structures deposited in the PDB per year (last decade) | about 8000 per year |
| crystal and cryo-EM PDB structures released 2015-2025 | 97,200 crystal; 30,139 cryo-EM structures |
| structures modelling water | 91.8% of crystal structures; 9.7% of cryo-EM structures |
| fraction of interacting photons that diffract (12.4 keV, 100 µm crystal) | 8% (0.15% of incident); 84% absorbed photoelectrically; 2% of incident interact at all |
| Henderson dose limit | 20 MGy |
| measured D1/2 at 100 K (ferritin) | 43 MGy |
| recommended maximum dose at 100 K | 30 MGy |
| dose tolerance gain at 100 K over room temperature | about 70 fold |
| share of protein structures determined cold | over 90 % |
| safe dose at room temperature | 0.2 MGy |
| XFEL peak brilliance over third-generation synchrotrons | nine orders of magnitude ratio |
| LCLS pulse rate | 120 (30 at SACLA) Hz |
| time resolution: synchrotron Laue vs FEL | ~100 ps vs femtoseconds |
| liquid jet speed and flow (SFX) | 10 m/s; 10-30 µl/min m/s; µl/min |
| neutron coherent scattering length, 1H | -3.74 10^-13 cm |
| neutron coherent scattering length, 2H (D) | 6.67 10^-13 cm |
| neutron data collection time | ~5-20 days |
| neutron crystal volume needed (average) | >0.5 mm^3 |
| PDB entries from neutron diffraction | ~100 entries |
| MicroED crystal thickness | 10-400 nm |
| MicroED protein and peptide PDB entries (Feb 2024) | 138 entries |
| electron-diffraction structures in the CSD | more than 740 structures |
| MicroED protein total dose | less than 1 e-/Å^2 |
| predicted structures in AlphaFold DB (Sept 2023) | over 214 million structures |
| PSI structures deposited / crystal share | 5097 / 4598 (90%) structures |
| PSI crystal structures phased by MAD / SAD | 1353 (29%) / 2168 (47%) structures |
| de novo crystal structures through 2012 phased by MAD or SAD | 77 % |
| MR declared / curated, 2013 PDB depositions | 67 / 44 % |
| year SAD overtook MAD in de novo phasing | 2006 year |
| structures possible only with a stabilising ligand (SGC) | about 10 % |
| IUCr redefinition of crystal | 1992 year |
| characteristic clock: one rotation data set's coverage | one image per step through slightly more than 180 degrees images per data set (time not stated in sources read) |
Open contentions (5) open contention
Where sources or readings disagree. Both stay on the map.
- C-413 Where do thoughts come from?
booked: No theory is proven, so the map lists the main ones side by side and doesn't pick one. It shows where thinking is handled in the brain, not where thoughts start.
the other reading: the positions listed in C-413 - C-736 When a system first throws away its excess energy and later learns to store it for reuse, is that one process that grew or two separate ones?
booked: The map keeps both readings side by side and rules neither. A few examples fit the 'one process that grew' reading, but it hasn't been tested.
the other reading: the toll is only a loss - C-1243 What counts as a crystal?
booked: The map uses the definition of a periodic arrangement of atoms, and keeps the crystallographers' broader definition, which includes quasicrystals, alongside.
the other reading: a crystal as a periodic arrangement - C-1525 What counts as crystallography?
booked: The map books it as the practice of reading structure from diffraction, and keeps the wider reading of all study of molecular structure.
the other reading: crystallography as all study of molecular and crystalline structure (the encyclopedia's broad definition, which includes imaging) - C-1526 When did crystallography begin?
booked: The map keeps both starts: Steno measuring crystal faces in 1669, and Laue's diffraction in 1912.
the other reading: start at Kepler 1611 or Steno 1669 (crystal study as such); or at 1912 only (modern crystallography)
Partners and parts published science
Built from / builds (12)
- Human (Homo sapiens)
- Electron microscope
- Free-electron laser
- X-ray crystallography
- Synchrotron light source
- Goniometer
- X-ray area detector (film, CCD, pixel detector)
- Neutron source (reactor or spallation)
- Protein Data Bank
- Cambridge Structural Database
- International Tables for Crystallography
- X-ray crystallography against cryo-EM for new macromolecular structures
Sits inside / holds (3)
- Science (as a human social and cultural system)
- Chemistry (the field: the practice of chemists and the structure of substances and their reactions it works on)
- Condensed matter physics
Trades with (valence) (19)
- Neutron via sys:residual-strong-force · Crystallography gain
- Crystal via sys:electromagnetic-field · Crystallography gain
- Electron via sys:electromagnetic-field · Crystallography gain
- DNA and genes via sys:electromagnetic-field · Crystallography gain
- Protein via sys:hydrogen-bond-network · Crystallography gain
- Human (Homo sapiens) via sys:sensory-environment · Crystallography gain
- X-rays via sys:electromagnetic-field · Crystallography gain
- Mineral via sys:electromagnetic-field · Crystallography gain
- Spectroscopy via sys:electromagnetic-field · Crystallography gain
- Chemistry (the field: the practice of chemists and the structure of substances and their reactions it works on) via sys:electromagnetic-field · Crystallography gain
- Quasicrystal via sys:electromagnetic-field · Crystallography gain
- Free-electron laser via sys:electromagnetic-field · Crystallography gain
- Synchrotron light source via sys:electromagnetic-field · Crystallography gain
- Fourier analysis (the practice of splitting signals into waves and rebuilding them, and the structure it works on) via sys:mental-representation · Crystallography gain
- Protein Data Bank via sys:written-record · Crystallography gain
- Cambridge Structural Database via sys:written-record · Crystallography gain
- Cryo-electron microscopy (single particle) via sys:written-record · Crystallography mixed
- AlphaFold (structure prediction) via sys:written-record · Crystallography gain
- Materials science (the field) via sys:electromagnetic-field · Crystallography gain
expressed by (21)
- Generation/Synthesis
- Scaffolding/Enablement (Meta-Tooling)
- Templating / Standardization
- Interoperability/Transduction (Bridging)
- Amplification/Scale-Up (Manufacture & Distribution)
- Symbiosis/Co-Creation (Collaborative Specialization)
- Sensing/Scanning
- Search/Foraging
- Mapping/Modeling
- Uncertainty Reduction/Learning
- Active Experimentation / Probing (Optimal Experimental Design)
- Multi-Scale Data Fusion / Triangulation
- Trail-Marking / Externalized Memory (Stigmergy)
- Surprise / Change-Point Detection
- Frontier Outposts / Sensing Extension
- Hypothesis Portfolio / Multiple Hypothesis Tracking
- Social Learning / Public Information Use
- Technique-Based Identification (a.k.a. Technique-Based Detection)
- Validation of Effectiveness
- Redundancy & Diversity
- Quality Control & Error Correction
floor filled by (20)
- Neutron
- Electron
- Human (Homo sapiens)
- X-rays
- Chemistry (the field: the practice of chemists and the structure of substances and their reactions it works on)
- Crystallization (nucleation and crystal growth)
- Crystal
- Computer
- Free-electron laser
- Sectors of the chemical industry (oil refining, bulk, fine, pharmaceutical)
- Synchrotron light source
- Fourier analysis (the practice of splitting signals into waves and rebuilding them, and the structure it works on)
- Goniometer
- X-ray area detector (film, CCD, pixel detector)
- Neutron source (reactor or spallation)
- Protein Data Bank
- Cambridge Structural Database
- AlphaFold (structure prediction)
- International Union of Crystallography
- Materials science (the field)
represents (1)
Kinds (8)
- X-ray crystallography
- Neutron crystallography
- Electron crystallography (electron diffraction, MicroED)
- Powder diffraction
- Serial femtosecond crystallography
- Macromolecular crystallography
- Goniometric crystallography (before X-rays)
- Small-molecule crystallography
Notes published science
- When X-rays hit a typical protein crystal, only about 2 in 100 photons interact at all, and of those only about 8 in 100 bounce off in the useful directions that make the picture; about 84 in 100 are absorbed and slowly cook the crystal. Every protein structure is drawn from the rare photons, paid for by the common ones (src:garman-2010).
- To neutrons, ordinary hydrogen and heavy hydrogen look like opposites: hydrogen scatters with a negative strength (-3.74) and shows as a dip in the map, deuterium with a positive one (+6.67) and shows as a peak. So crystallographers soak or grow their crystals in heavy water to make hydrogen atoms appear (src:oksanen-2017).
- Crystallographers supplied 88.8% of everything in the Protein Data Bank from 1971 to 2014; that record trained AlphaFold, whose predicted models are now the first thing crystallographers reach for to solve new crystals. The practice built the tool that may retire its own hardest step (src:wlodawer-2026; src:hendrickson-2023).
Layers
The ways it takes in its world, where they meet, and how it acts back, each from L1 (where it meets the world) up.
crystallography: a beam meets a sample, the diffraction is recorded, phased into a density map, built into an atomic model and used (reader's view)
- L1: The beam meets the sample: X-rays (from tubes, synchrotrons, free-electron lasers), electrons or neutrons, each meeting a different part of the matter (electron density, total charge, nuclei)
- L2: The sample: a crystal grown (often the chief barrier), mounted on a goniometer, usually flash-cooled; its lattice planes diffract the beam (Bragg's law)
- L3: Recording: reflections on film, CCD or pixel detectors while the crystal turns through about 180 degrees; first images give the resolution limit and the crystal's pathologies
- L4: The diffraction data: indexed (unit cell, space group), integrated, merged and scaled into one set, its quality judged by symmetry-equivalent reflections
- L5: Phases and the density map: the phase problem met by direct methods, isomorphous replacement, anomalous diffraction or molecular replacement; Fourier synthesis gives the electron (or nuclear) density
- L6: The atomic model: built into the density, refined against the data (R, R-free), checked for bias, disorder and chemistry
- L7: Use and record: structures deposited for the international community and used to explain bonding, minerals and living molecules, to design drugs, and to train structure predictors
the boardroom (shared floors)
- L4: The diffraction data: reflections (or face angles) indexed to a unit cell and space group, integrated, merged and scaled onto one intensity scale; quality judged by symmetry-equivalent reflections
- L5: Phases and the density map: direct methods, isomorphous replacement, anomalous diffraction or molecular replacement give phases; Fourier synthesis turns amplitudes and phases into electron (or nuclear) density
- L6: The model: atoms built into density, refined against the data (R and R-free on held-out reflections), checked for model bias, disorder and chemistry
- L7: Use and record: structures deposited and used to explain bonding and minerals, to see how drugs meet their targets, to guide new experiments, and as training data for predictors
way in: Chemical knowledge (composition, sequence, bonding rules, ideal geometry)
- L1: Complementary chemical information combined with the diffraction data
- L2: Rules eliminating chemically unlikely structures and giving atoms' relative sizes (Goldschmidt, Pauling)
- L3: Idealised geometry used as restraints in refinement
- L4: Constraints on the density itself: atoms as separate peaks (direct methods' phase relationships) and flat solvent (density modification)
way in: Crystal faces seen and measured (goniometry, the first channel)
- L1: Light off crystal faces, long admired for regularity and symmetry
- L2: The goniometer measures angles between faces; the angles are constant for each kind of crystal (Steno 1669)
- L3: Faces plotted on a stereographic net, each labelled by its Miller index; the plot gives the symmetry
way in: Electron diffraction (MicroED, surfaces)
- L1: Electrons, charged and focusable, meet the total charge of nuclei and electrons
- L2: Crystals 10-400 nm thick (a billionth the volume of usual X-ray crystals); strong, sometimes multiple (dynamical) scattering
- L3: The crystal rotated continuously, patterns recorded as a movie on a fast camera
way in: Known and predicted structures (PDB, CSD, AlphaFold DB)
- L1: Deposited structures in public databases (PDB, CSD, ICSD) and, since 2021, predicted models
- L2: A related structure chosen as search model; AlphaFold models became the search models of choice
- L3: The search model placed in the new cell by rotation and translation searches (rotation function, Phaser's maximum likelihood)
- L4: Phases computed from the placed model for the new structure's Fourier synthesis
way in: Neutron diffraction
- L1: Neutrons from a handful of instruments worldwide; flux-limited
- L2: Scattering by the nuclei, independent of atomic number: hydrogen (negative length, -3.74) shows as troughs, deuterium (+6.67) as peaks, so crystals are soaked or grown in heavy water
- L3: Long collections (about 5 to 20 days) from large crystals (above 0.5 mm3 on average)
way in: Powder diffraction (many crystals at once)
- L1: X-rays or neutrons on a polycrystalline sample; most materials are aggregates of small crystals
- L2: Each phase has its own arrangement of atoms and its own pattern
- L3: The diffractogram compared with simulated patterns (Rietveld refinement)
way in: Serial femtosecond snapshots (X-ray free-electron lasers)
- L1: Femtosecond pulses with peak brilliance nine orders of magnitude above third-generation synchrotrons
- L2: Diffraction before destruction: the pulse passes before significant damage begins; each microcrystal, jetted in at random orientation, is destroyed by its one exposure
- L3: Hits identified and their still patterns merged by specialised programs (CrystFEL, cctbx.xfel, nXDS)
way in: Thought: ideas of what the pattern means
- L1: An idea arrives: crystals as gratings for X-rays came up in a conversation in a Munich park; where a new idea comes from is contested
- L4: Alternative models held and compared with measured intensities (trial and error)
- L5: The problem named: Patterson set phase determination as the problem of X-ray analysis
way in: X-ray diffraction (single crystal)
- L1: X-rays meet the crystal's electrons; for a 100 µm protein crystal at 12.4 keV only about 2% of incident photons interact at all
- L2: Scattering by the electron density: the elastically scattered photons (about 8% of those that interact) leave in Bragg directions; the rest are absorbed and damage the crystal
- L3: The reflections recorded image by image as the crystal is rotated
way out: Deposit: coordinates and data into public databases
- L1: The entry in the record, open to all (PDB, CSD)
- L6: Coordinates with validation reports (Ramachandran, clashscore, rotamer outliers)
- L7: Deciding to share: almost all new structures go to public databases; many commercial ones do not
way out: Design: changing molecules on what the structure shows (drugs, materials)
- L1: New molecules and materials made with tailored properties
- L7: Seeing how a drug meets its target and what change might improve it
way out: Dose: the beam's energy left in the crystal (damage), and the plans that limit it
- L1: Energy absorbed by the crystal: photoelectric absorption and Compton scattering ionise atoms and break the crystal down
- L2: Cooling the crystal (100 K raises tolerance about 70-fold) and choosing beam and exposure
- L4: A dose budget: about 30 MGy at 100 K, 0.2 MGy at room temperature; data merged from several crystals when one fades
way out: Explain: structures that changed what chemistry, mineralogy and biology hold true
- L1: Papers, reference tables and prizes: the International Tables; Nobel Prizes from 1914 (Laue) and 1915 (the Braggs) on
- L7: Explaining: ionic compounds, carbon's tetrahedron, benzene's resonance, the silicates, DNA and proteins
way out: Remake the sample: new conditions, changed molecules, heavy atoms, selenium, deuterium
- L1: New crystals grown by screening hundreds or thousands of conditions
- L2: Readable atoms put in ahead of time: selenomethionine in every methionine, heavy-atom soaks, deuteration
- L6: A failed or ambiguous result sends the crystallographer back to the sample
Sources (13) published science
Garman E.F. (2010) Radiation damage in macromolecular crystallography: what is it and why should we care? Acta Cryst D 66:339. doi:10.1107/S0907444910008656 (PMC2852297) (behind its links)
Hendrickson 2023. Facing the phase problem. IUCrJ. doi:10.1107/s2052252523006449 (CC BY) (behind its links)
Schlichting I. (2015) Serial femtosecond crystallography: the first five years. IUCrJ 2:246. doi:10.1107/S205225251402702X (PMC4392417, CC BY) (behind its links)
Wlodawer A. et al. (2026) Comparative assessment of crystallographic and cryo-EM models in the Protein Data Bank. FEBS Open Bio. doi:10.1002/2211-5463.70326 (PMC13508841) (behind its links)
Unge J., Nannenga B.L., Oliver A.G., Gonen T. (2025) Standards for MicroED. Acta Cryst C. doi:10.1107/S2053229625004875 (PMC12231505) (behind its links)