{
  "id": "crystal-structure",
  "title": "Crystal Structure",
  "lang": "en",
  "source_lang": "en",
  "summary": "Crystal structure describes the ordered arrangement of atoms, ions, or molecules in a crystalline material and connects its symmetry and bonding to measurable properties.",
  "category": "technology",
  "subcategory": "materials",
  "type": "concept",
  "aliases": [
    "Crystalline structure",
    "crystallographic structure"
  ],
  "generated_at": "2026-11-03",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/crystal-structure/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/crystal-structure/",
      "json": "https://aiwiki.page/zh/crystal-structure.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "A crystal structure is the spatial arrangement of atoms, ions, or molecules in a crystal. In an ordinary periodic crystal, this arrangement is described by a repeating atomic motif and a three-dimensional lattice of translations. The structure specifies not only the chemical composition but also the positions, occupancies, and symmetry relationships of the constituent particles. Materials with the same composition can have different crystal structures and consequently very different properties, as illustrated by diamond and graphite.\n\nCrystal structure provides a common framework for mineralogy, chemistry, solid-state physics, and materials engineering. It explains diffraction patterns, constrains physical properties, and helps account for mechanical strength, electrical transport, optical behavior, and chemical reactivity. Although periodic structures are the central case, crystallography also encompasses ordered structures without ordinary three-dimensional periodicity. Real crystals additionally contain thermal motion, defects, surfaces, and domains, so a complete structural description distinguishes an ideal repeating arrangement from its local and time-dependent departures."
    },
    {
      "heading": "Lattices, motifs, and unit cells",
      "text": "The essential distinction in a periodic crystal is between its lattice and its motif, also called a basis. A lattice is a mathematical set of equivalent points generated by translations; it contains no chemical identity by itself. The motif is the atom or group of atoms associated with each lattice point. Repeating the motif at all lattice points produces the crystal structure. Different motifs can occupy the same kind of lattice, and the same chemical constituents can form structures based on different lattices.\n\nThree linearly independent translation vectors, \\(\\mathbf a\\), \\(\\mathbf b\\), and \\(\\mathbf c\\), generate the lattice:\n\n\\[\\mathbf R=n_1\\mathbf a+n_2\\mathbf b+n_3\\mathbf c, \\qquad n_1,n_2,n_3\\in\\mathbb Z.\\]\n\nA unit cell is a region whose translations reproduce the structure without gaps or overlaps. Its dimensions are conventionally expressed by three lengths, \\(a\\), \\(b\\), and \\(c\\), and three angles: \\(\\alpha\\) between \\(\\mathbf b\\) and \\(\\mathbf c\\), \\(\\beta\\) between \\(\\mathbf a\\) and \\(\\mathbf c\\), and \\(\\gamma\\) between \\(\\mathbf a\\) and \\(\\mathbf b\\). These six quantities are the lattice parameters. Their values depend on conditions such as temperature, pressure, and composition.\n\nAn atomic position is usually given in fractional coordinates, so that\n\n\\[\\mathbf r_j=x_j\\mathbf a+y_j\\mathbf b+z_j\\mathbf c.\\]\n\nCoordinates differing by integers describe equivalent positions in different cells. Fractional coordinates are dimensionless and remain convenient when the cell expands, contracts, or changes shape. A structural model also records the chemical species at each site, its occupancy, and parameters describing atomic displacement. Occupancy may be less than one because a site is partly vacant or shared statistically among different elements.\n\nA primitive cell contains one lattice point after accounting for points shared with neighboring cells. A conventional cell may contain several lattice points because it is chosen to display symmetry more clearly. For example, the conventional face-centered cubic cell contains four lattice points: eight corners contribute one point in total, and six face centers contribute three. This does not mean that every face-centered cubic crystal contains only four atoms per conventional cell; the number also depends on the motif.\n\nA structure therefore cannot be identified from cell dimensions alone. Two materials may have similar lattice parameters but different atomic arrangements, while alternative cell choices may describe the same structure. Reliable comparison requires the lattice, chemical contents, atomic coordinates, and symmetry together. Crystallographic coordinate transformations make it possible to recognize equivalence between descriptions that initially appear different."
    },
    {
      "heading": "Symmetry and crystallographic classification",
      "text": "Crystal symmetry consists of operations that leave the ideal structure unchanged. These include translations, rotations, reflections, inversion, and combinations of these operations. A point group describes the symmetry operations that leave at least one point fixed. A space group incorporates the full translational symmetry of the structure, including screw rotations and glide reflections, which combine a rotation or reflection with a fractional translation.\n\nFor an ordinary three-dimensional periodic lattice, the crystallographic restriction permits rotational symmetries of orders 1, 2, 3, 4, and 6. Fivefold rotational symmetry cannot be the exact rotational symmetry of such a lattice. This restriction follows from the requirement that rotations map a discrete translational lattice onto itself; it does not forbid local fivefold atomic clusters or the nonperiodic order found in quasicrystals.\n\nThere are 32 crystallographic point-group types, 14 Bravais lattice types, and 230 space-group types in three dimensions under the standard crystallographic classification. The seven crystal systems are triclinic, monoclinic, orthorhombic, tetragonal, trigonal, hexagonal, and cubic. These systems classify symmetry, rather than simply grouping crystals by external shape. The trigonal and hexagonal systems require particular care: they are distinct crystal systems, although both belong to the hexagonal crystal family, and trigonal structures may be described using rhombohedral or hexagonal settings.\n\nSpace groups constrain the allowed atomic positions. Symmetry-related positions form sets conventionally designated as Wyckoff positions, each with a multiplicity and a site symmetry. Instead of listing every atom in a unit cell independently, a structural description can give an asymmetric unit and generate the rest by symmetry. An atom on a high-symmetry position has fewer independent positional parameters than an atom in a general position. This greatly reduces the number of quantities needed to describe many structures.\n\nSymmetry also constrains physical behavior. A crystal with inversion symmetry cannot exhibit ordinary bulk piezoelectricity, because inversion reverses polarization while leaving mechanical strain unchanged. Cubic symmetry forces several second-rank property tensors to have isotropic form, although a cubic crystal can still be elastically anisotropic. Chirality presents another distinction: a chiral crystal structure cannot contain inversion, mirror, glide-reflection, or other orientation-reversing symmetry operations.\n\nObserved symmetry is not always immediately apparent. Twinning can combine differently oriented domains into a specimen with misleading apparent symmetry, and nearly equal cell parameters can suggest a symmetry higher than the actual one. Structural classification consequently rests on diffraction evidence and atomic arrangement, not on the visual outline of a crystal or a single set of geometric measurements."
    },
    {
      "heading": "Bonding, coordination, and structural stability",
      "text": "A crystal structure reflects a balance among attractive and repulsive interactions, electronic organization, and thermodynamic conditions. Ionic, covalent, metallic, and intermolecular bonding are useful limiting descriptions, but many solids combine several bonding characters. The arrangement that forms is not necessarily the densest possible packing of atoms. Directional bonding, charge balance, electronic energy, and the sizes and shapes of constituent species can favor much more open structures.\n\nThe coordination number of an atom counts neighboring atoms according to a specified bonding or distance criterion. In sodium chloride, each sodium ion has six nearest chloride neighbors, and each chloride ion has six nearest sodium neighbors. In diamond, each carbon atom forms four approximately tetrahedrally directed covalent bonds. In a close-packed elemental metal, an atom has twelve nearest neighbors. These numbers describe different geometries and do not alone determine the complete structure.\n\nCoordination polyhedra provide a useful intermediate scale between individual bonds and the whole crystal. Tetrahedra, octahedra, and other polyhedra may share corners, edges, or faces. In silicate minerals, silicon commonly occupies oxygen tetrahedra, and the way these tetrahedra connect distinguishes isolated groups, chains, sheets, and frameworks. The topology of this connection can remain recognizable even when bond lengths and angles change substantially through chemical substitution or compression.\n\nElectrostatic models help explain ionic crystals, but simple size rules are approximate. Ionic radii depend on oxidation state and coordination environment, while polarization and covalent contributions can alter structural preferences. Likewise, metallic structures depend not only on efficient packing but also on the number and distribution of valence electrons. A quantitative treatment often requires quantum mechanics, because electronic states and bonding energies cannot generally be reduced to fixed spheres interacting through a single universal force.\n\nStability is governed by thermodynamics. At fixed temperature and pressure, the equilibrium phase minimizes the relevant Gibbs free energy,\n\n\\[G=U+pV-TS,\\]\n\nwhere \\(U\\) is internal energy, \\(pV\\) is the pressure–volume contribution, and \\(TS\\) represents the temperature-weighted entropy. Entropy includes contributions from vibrations, configurational disorder, and, where relevant, electronic or magnetic degrees of freedom. Consequently, the lowest-energy arrangement at absolute zero need not be the stable structure at elevated temperature.\n\nKinetics can preserve a structure even when another has lower free energy. Diamond persists under ordinary conditions because conversion to graphite requires extensive rearrangement of strong bonds. Such metastability is central to materials processing: the observed crystal structure depends not only on composition and current conditions but also on growth history, cooling rate, and the barriers separating alternative arrangements."
    },
    {
      "heading": "Common structure types and atomic packing",
      "text": "A relatively small collection of structural prototypes describes many elemental and inorganic crystals. A prototype specifies an arrangement that can be adopted by different chemical substances, sometimes with distortions or substitutions. Recognizing these relationships makes it possible to compare materials without treating every compound as geometrically unrelated.\n\nFace-centered cubic, body-centered cubic, and hexagonal close-packed arrangements are especially important in metallurgy. In the ideal hard-sphere model, face-centered cubic and hexagonal close-packed structures both have coordination number 12 and packing fraction\n\n\\[\\eta=\\frac{\\pi}{3\\sqrt{2}}\\approx0.7405.\\]\n\nThey differ in the stacking of close-packed layers: face-centered cubic follows an \\(ABCABC\\) sequence, whereas hexagonal close packing follows \\(ABAB\\). Body-centered cubic has eight nearest neighbors and a hard-sphere packing fraction of approximately 0.6802. Copper is face-centered cubic under ordinary conditions, magnesium is hexagonal close-packed, and room-temperature alpha iron is body-centered cubic. These geometries influence available slip systems, but packing fraction alone does not determine strength or ductility.\n\nBinary compounds illustrate how sublattices and interstitial positions generate distinct structures. The rock-salt structure, adopted by sodium chloride and magnesium oxide, can be described as two interpenetrating face-centered cubic sublattices with octahedral coordination. The cesium-chloride structure has eightfold coordination and a primitive cubic Bravais lattice with a two-species motif; despite its corner-and-center appearance, it is not a body-centered cubic Bravais lattice. In the fluorite structure of calcium fluoride, calcium forms a face-centered cubic sublattice and fluoride occupies all tetrahedral sites.\n\nDiamond and zinc blende demonstrate the importance of chemical identity within related geometries. Diamond consists of two offset face-centered cubic sublattices occupied by the same element, giving tetrahedral coordination. In zinc blende, the two sublattices contain different elements, as in one form of zinc sulfide. Wurtzite provides a related tetrahedrally coordinated arrangement with a different stacking sequence. The wurtzite structure is technologically important for gallium nitride, while diamond-type silicon underpins much of semiconductor technology.\n\nThe perovskite family is commonly represented by the formula \\(ABX_3\\). In its ideal cubic form, the \\(B\\) cation occupies the center of an octahedron of \\(X\\) anions, and these octahedra share corners. Tilting of octahedra, displacement of cations, and chemical ordering generate many lower-symmetry variants. Related structural principles occur in oxide ceramics and the halide absorbers used in perovskite solar cells.\n\nNot all crystals fit compact inorganic prototypes. Molecular crystals organize discrete molecules through hydrogen bonding, dispersion, and other interactions. Framework materials such as zeolites and metal–organic frameworks contain extended networks surrounding pores. In these materials, connectivity and accessible void space may be more consequential than high packing density."
    },
    {
      "heading": "Defects, disorder, and departures from periodicity",
      "text": "An ideal crystal repeats indefinitely, but real crystals do not. Their structures contain defects over several length scales, and even a chemically pure specimen exhibits thermal atomic motion. A useful distinction separates the average structure, which describes the periodic arrangement inferred from Bragg diffraction, from local structure, which includes deviations and correlations that may be averaged out.\n\nPoint defects include vacancies, interstitial atoms, and substitutional atoms. Their concentrations depend on formation energies, entropy, chemical environment, and charge neutrality. In an ionic crystal, a vacancy may carry an effective charge relative to the perfect lattice and therefore require compensation by other defects or changes in electronic charge. A Schottky defect involves a charge-balanced set of vacancies, whereas a Frenkel defect consists of a vacancy and an interstitial formed when an ion leaves its normal site. Defect chemistry is particularly important in oxides, where oxygen vacancies can change ionic transport and oxidation states.\n\nA dislocation is a line defect associated with a mismatch in lattice registry, characterized by a Burgers vector. Edge and screw dislocations are ideal limiting forms; many real dislocations have mixed character. Their motion allows a crystal to deform plastically without requiring an entire atomic plane to move simultaneously. Interactions among dislocations, solute atoms, and precipitates explain many strengthening mechanisms in steel and other alloys.\n\nPlanar defects include grain boundaries, stacking faults, and twin boundaries. A grain boundary separates regions with different crystallographic orientations. A stacking fault interrupts a regular layer sequence, particularly clearly in close-packed structures. A twin relates neighboring domains by a specific crystallographic operation. Grain size, boundary character, and defect density belong to microstructure, which must be distinguished from the crystal structure within each grain. Materials with the same atomic structure can therefore have very different engineering properties.\n\nDisorder need not destroy crystallinity. Two elements may share one crystallographic site statistically, or molecules may occupy several orientations. Site occupancy describes an average population; it does not mean that an individual atom is fractionally present at an instant. Correlations between neighboring occupations may create short-range order that is invisible to a simple average model but detectable in diffuse scattering or local measurements.\n\nQuasicrystals possess long-range order and sharp diffraction peaks without an ordinary three-dimensional translation lattice. Incommensurately modulated crystals similarly require more than a simple repeating three-dimensional cell for an exact description. Higher-dimensional crystallographic methods describe such order systematically. These cases established that crystallinity is broader than strict periodicity, while preserving the importance of well-defined diffraction order as an experimental characteristic."
    },
    {
      "heading": "Diffraction and structure determination",
      "text": "X-ray crystallography is a principal method for determining crystal structures because X-ray wavelengths are comparable to interatomic distances. X-rays scatter primarily from electron density, and waves scattered by different parts of a crystal interfere. Constructive interference occurs at particular scattering vectors related to the reciprocal lattice, a mathematical lattice whose vectors encode the periodicities of the real-space crystal.\n\nFor a family of lattice planes with spacing \\(d\\), the familiar diffraction condition is Bragg’s law:\n\n\\[2d\\sin\\theta=n\\lambda,\\]\n\nwhere \\(\\theta\\) is the Bragg angle, \\(\\lambda\\) the wavelength, and \\(n\\) an integer. Diffraction peak positions constrain the unit-cell dimensions. Their intensities contain information about the atomic arrangement within the cell. In a simplified independent-atom description, the structure factor is\n\n\\[F_{hkl}=\\sum_j o_j f_j \\exp\\!\\left[2\\pi i(hx_j+ky_j+lz_j)\\right] \\exp(-W_j),\\]\n\nwhere \\(o_j\\) is occupancy, \\(f_j\\) the atomic scattering factor, and \\(W_j\\) a displacement-dependent term. Corrected intensities are proportional to \\(|F_{hkl}|^2\\). Symmetry-related interference can produce systematic absences, helping identify possible space groups.\n\nA central difficulty is that ordinary intensity measurements do not directly provide the phases of the structure factors. This phase problem prevents straightforward reconstruction of electron density from measured intensities alone. Direct methods, Patterson methods, anomalous scattering, and molecular replacement provide different routes to obtaining phase information or an initial model. Once a plausible structure is available, refinement adjusts atomic parameters to improve agreement between calculated and observed data.\n\nSingle-crystal diffraction retains three-dimensional information about individual reflections. Powder diffraction combines contributions from many randomly oriented crystallites into a one-dimensional pattern, so reflections may overlap. It is widely used for identifying phases and determining lattice parameters. Rietveld refinement fits an entire powder pattern using a structural and instrumental model, allowing atomic and microstructural information to be extracted when the data support it.\n\nNeutrons provide complementary contrast because they scatter from nuclei and magnetic moments rather than principally from electron density. They can distinguish certain neighboring elements or isotopes that are difficult to separate with X-rays and are valuable for locating hydrogen in suitable samples. Electrons interact strongly with matter, allowing diffraction from very small crystals and local analysis in an electron microscope, although multiple scattering can complicate interpretation.\n\nNo diffraction model is self-validating. Absorption, preferred orientation, twinning, disorder, limited resolution, and parameter correlations can affect a refinement. Structural reliability therefore depends on chemical plausibility, uncertainty estimates, residual density, and agreement with complementary evidence—not merely on a low numerical fitting residual."
    },
    {
      "heading": "Structural change, polymorphism, and phase transitions",
      "text": "Polymorphism is the ability of a substance to form more than one crystal structure. For elements, the related term allotropy is commonly used. Carbon’s diamond and graphite forms are a prominent example: diamond has a three-dimensional tetrahedral network, whereas graphite contains strongly bonded sheets with much weaker interactions between them. Identical elemental composition therefore does not imply identical density, hardness, electrical behavior, or stability.\n\nA phase transition occurs when changing conditions favor a different state or arrangement. Structural transitions may involve small collective displacements, extensive bond reconstruction, ordering of previously disordered sites, or combinations of these processes. A displacive transition can preserve much of the original connectivity, while a reconstructive transition requires substantial changes in bonding and often encounters a large kinetic barrier.\n\nBarium titanate illustrates a temperature-dependent sequence of related structures. On heating, it passes through rhombohedral, orthorhombic, and tetragonal ferroelectric phases before becoming cubic and paraelectric at approximately \\(120\\,^\\circ\\mathrm C\\), with transition temperatures influenced by composition and sample conditions. Relative displacements of ions create polarization in the ferroelectric phases. The average cubic structure above the transition does not by itself exclude instantaneous local distortions, demonstrating the distinction between average symmetry and local dynamics.\n\nPressure changes the relative importance of volume in free energy and can stabilize denser arrangements. Silica, for example, occurs in structures with different silicon coordination: quartz has tetrahedrally coordinated silicon, whereas the high-pressure phase stishovite has sixfold coordination in a rutile-type structure. Such transformations are important in deep-Earth mineralogy and in interpreting minerals produced by impact events. The transition pathway and the persistence of recovered phases depend strongly on kinetic barriers.\n\nIron provides a practical example of structural change in processing. At atmospheric pressure, pure iron changes from body-centered cubic alpha iron to face-centered cubic gamma iron at about \\(912\\,^\\circ\\mathrm C\\), and returns to a body-centered cubic delta phase at about \\(1{,}394\\,^\\circ\\mathrm C\\) before melting. Carbon and other alloying elements alter phase stability. Rapid cooling of austenite in suitable steels can produce martensite through a diffusionless transformation, generating a distorted structure and a characteristic microstructure.\n\nStructural transitions are studied through diffraction, calorimetry, spectroscopy, and measurements of mechanical, electrical, or magnetic response. Hysteresis, phase coexistence, and abrupt changes in cell volume can reveal first-order behavior, while soft lattice vibrations and evolving symmetry help characterize other transitions. Because stress, defects, grain size, and transformation rate affect the observed behavior, a measured transition temperature belongs to a specified material and experimental condition rather than to structural geometry alone."
    },
    {
      "heading": "Connections to material properties and technology",
      "text": "Crystal structure influences properties by determining both the symmetry of a response and the microscopic pathways through which it occurs. Symmetry can establish that a particular response is forbidden, but it rarely determines its magnitude. Bond stiffness, atomic masses, electronic states, defects, and operating conditions are also required. Structural similarity therefore provides a useful guide, not a guarantee of identical behavior.\n\nElectronic properties emerge from electrons moving in the crystal’s spatially varying potential. Electronic band structure depends on atomic orbital energies, bonding, symmetry, and periodicity. Diamond and graphite again show the consequences: their different carbon bonding networks produce very different electronic behavior. In silicon, substitutional dopants modify carrier populations while largely preserving the host lattice. Strain and changes of structural phase can further shift electronic energies, which is important in semiconductor design.\n\nMechanical response depends on both bonding and crystallographic geometry. Elastic deformation changes interatomic distances and angles reversibly, whereas plastic deformation often involves dislocation motion along specific slip systems. Layered crystals may cleave readily between sheets even when bonding within each sheet is strong. Polycrystalline components add grain-boundary and texture effects: a preferred distribution of grain orientations can make a manufactured sheet or wire directionally dependent even when its chemical composition is uniform.\n\nThermal behavior is closely linked to collective atomic vibrations, called phonons. Their frequencies and propagation depend on atomic masses and force constants. Phonon scattering by defects, boundaries, other phonons, and electrons influences thermal conductivity. Thermal expansion arises from anharmonicity in atomic interactions, and some framework structures display negative thermal expansion over particular temperature ranges because transverse motions and coordinated polyhedral rotations reduce their overall dimensions.\n\nCrystal symmetry is especially important in optical and electromechanical applications. Quartz is piezoelectric, producing an electrical response to applied mechanical stress. Ferroelectric ceramics can be electrically poled to align domains and generate a useful macroscopic response. Structural distortion and domain motion contribute to their operation in transducers and sensors. Optical birefringence, nonlinear optical effects, and selection rules for vibrational spectra likewise depend on symmetry and bonding.\n\nIon transport illustrates the importance of pathways rather than merely composition. Electrode structures in a lithium-ion battery must accommodate lithium insertion and removal, often through connected sites whose accessibility changes during charging. In a solid electrolyte, mobile ions move through a lattice whose site energies and bottlenecks determine migration barriers. Porous frameworks offer a related geometrical function: the dimensions and connectivity of channels govern molecular access and separation. Engineering such materials requires coordination of crystal structure with defects, particle morphology, interfaces, and processing history."
    },
    {
      "heading": "Historical development and computational crystallography",
      "text": "The concept of internal crystalline order developed before atoms could be located experimentally. In 1669, Nicolas Steno described the constancy of corresponding interfacial angles in quartz crystals. During the late eighteenth century, René Just Haüy related crystal forms and cleavage to the repeated arrangement of elementary units. These ideas explained regular external geometry through internal organization, although the proposed units were not yet experimentally established atomic structures.\n\nNineteenth-century work made this geometry systematic. Auguste Bravais classified the 14 three-dimensional lattice types in 1850. Later, Evgraf Fedorov and Arthur Schoenflies independently developed classifications of the 230 space-group types, with William Barlow also contributing to their development. This mathematical framework preceded direct experimental knowledge of atomic positions and supplied the language needed to interpret later diffraction measurements.\n\nThe decisive experimental advance came in 1912, when Max von Laue proposed that a crystal could diffract X-rays, and Walter Friedrich and Paul Knipping demonstrated the effect. William Henry Bragg and William Lawrence Bragg subsequently developed methods for analyzing the patterns; Lawrence Bragg formulated the relation now known as Bragg’s law. Early structure determinations, including those of alkali halides, showed that diffraction could reveal atomic arrangements rather than merely crystal shape.\n\nDuring the twentieth century, structure analysis expanded from simple inorganic compounds to complex molecules and biological macromolecules. Improved phase-determination methods and electronic computing made increasingly large structures accessible. Work by Dorothy Crowfoot Hodgkin on compounds including vitamin B12 and insulin demonstrated the chemical reach of crystallography, while protein crystallography linked molecular architecture to biological function. Synchrotron sources, neutron facilities, and improved detectors later increased sensitivity, speed, and access to extreme conditions.\n\nThe discovery by Dan Shechtman in 1982 of diffraction with noncrystallographic rotational symmetry, reported in 1984, challenged the identification of crystalline order with ordinary periodicity. Subsequent quasicrystal research established a broader structural framework for aperiodic order. The distinction between sharp long-range diffraction order and a repeating three-dimensional unit cell became fundamental to modern crystallographic terminology.\n\nComputational work now complements experimental determination. Density functional theory can estimate electronic energies and optimize atomic positions, while molecular dynamics follows atomic motion under specified interaction models. Structure-prediction methods search possible arrangements using symmetry, chemical constraints, and mathematical optimization; increasingly, machine learning supplies rapid models of energies and forces. Predicted low-energy structures still require careful treatment of temperature, pressure, disorder, and competing phases. Structural databases and standardized crystallographic information files allow measured and calculated models to be compared, reused, and checked for consistency."
    }
  ],
  "links": [
    "atom",
    "ion",
    "molecule",
    "crystal",
    "chemistry",
    "physics",
    "unit-cell",
    "temperature",
    "pressure",
    "space-group",
    "bravais-lattice",
    "coordination-number",
    "quantum-mechanics",
    "thermodynamics",
    "metallurgy",
    "semiconductor",
    "perovskite-solar-cell",
    "zeolite",
    "metal-organic-framework",
    "quasicrystal",
    "x-ray-crystallography",
    "reciprocal-lattice",
    "braggs-law",
    "rietveld-refinement",
    "electron-microscope",
    "phase-transition",
    "spectroscopy",
    "electronic-band-structure",
    "phonon",
    "sensor",
    "lithium-ion-battery",
    "solid-electrolyte",
    "density-functional-theory",
    "mathematical-optimization",
    "machine-learning"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}