Crystals & Lattices
Order in Solid Matter
Why Atoms Line Up
A handful of salt grains and a single grain of sand both look opaque, hard, and ordinary. Look at either one with an electron microscope and you see something that is not ordinary at all. Trillions of atoms are stacked in repeating, geometric patterns that extend with almost machine precision in every direction. Each atom sits in a position chosen by physics, not by chance. A crystal is what matter does when it has time to relax into its lowest-energy arrangement.
The reason crystals form is straightforward. Atoms attract each other at long range and repel at very short range. The attractive part comes from sharing electrons, ionic charge, or weaker dipole effects, depending on the material. The repulsive part comes from quantum exclusion: electron clouds cannot overlap freely without paying a steep energy cost. Between those two forces sits a single distance where the energy is minimum. When billions of atoms try to settle into that minimum at once, the only solution that satisfies every neighbor at the right distance is a regular pattern. Disorder costs energy. Order, given enough time, wins by default.
Time matters because atoms have to rearrange to find their preferred sites. Cool a melt slowly and the atoms have plenty of opportunities to wiggle into place; you get a clean crystal. Quench it instantaneously and they freeze where they are; you get a glass. Window glass is silica that never had time to crystallize. The atoms inside it are still trying to find their preferred sites, just on geological timescales. Old cathedral panes are not actually thicker at the bottom from flowing, but the underlying impulse to relax toward order is real.
The Alphabet of Lattices
In three dimensions, there are exactly fourteen ways for points to repeat themselves with translational symmetry. The French crystallographer Auguste Bravais worked these out in 1850, and the count has not budged since. Every crystal in the universe, no matter how complicated its chemistry, sits on one of those fourteen Bravais lattices. The lattice is the skeleton; the atoms decorating it are called the basis. A diamond and a salt crystal can share the same underlying cubic skeleton while looking and behaving completely differently because what hangs on each lattice point is different.
The most common lattices in nature are also the most efficient at packing spheres. Face-centered cubic (FCC) and hexagonal close-packed (HCP) both fill 74 percent of available space, the maximum possible for identical spheres. Copper, gold, silver, aluminum, lead, and most other ductile metals are FCC. Magnesium, zinc, titanium, and cobalt are HCP. Body-centered cubic (BCC) packs slightly less efficiently at 68 percent and tends to be stiffer and less ductile; iron at room temperature, chromium, and tungsten are BCC. Whether a metal is easy to bend or stubborn to shape often comes down to which Bravais lattice its atoms chose.
The unit cell is the smallest piece of the lattice that, copied and stacked in three directions, reproduces the whole crystal. For salt that cell is a cube about half a nanometer on a side, containing four sodium and four chloride ions. For diamond it is a slightly larger cubic cell containing eight carbon atoms in a tetrahedral arrangement. Once you know the unit cell and the basis, you know the entire crystal. A grain of table salt the size of a poppy seed contains roughly ten quintillion copies of the same eight-ion cube, all aligned, all identical to within a few atomic diameters across the whole grain.
Symmetry Becomes Physics
A crystal has more symmetries than a simple lattice. On top of translation it can have rotation axes, mirror planes, inversion centers, and combinations of them called glide and screw axes. Group these together and you get exactly 230 distinct three-dimensional space groups. Every solid in the universe, from quartz to graphene to high-temperature superconductors, falls into one of those 230 categories. The classification is finite and complete in the same way the list of regular polyhedra is finite and complete.
That symmetry is not just bookkeeping. It controls what the crystal is allowed to do physically. A material can only show piezoelectricity (generate voltage when squeezed) if its space group lacks an inversion center. Quartz qualifies; salt does not, which is why your watch ticks on quartz and not on rock salt. Optical activity, ferroelectricity, the existence of certain electronic band crossings – all are dictated by which symmetries the lattice contains and which it forbids. Group theory is not an aesthetic exercise here. It is a list of which physical phenomena the crystal is permitted to display.
There is a second, equally useful way to think about a crystal. Instead of charting where each atom sits, you can chart which directions and spacings the lattice supports. That second map – one big lattice for the atoms, a paired map for the directions waves can travel through them – is what most of modern solid-state physics is actually drawn on. Electrons gliding through the crystal, vibration waves carrying heat, and X-rays bouncing off the planes all live more naturally on the directions-and-spacings map than on the atoms-in-space one. Switching between the two views is the daily work of the field, and it is why most diagrams in a solid-state textbook look so different from the physical lattice they describe.
How We Know Crystals Are Real
Crystals are too small to see directly with visible light. The waves of visible light are about a thousand times longer than the spacing between atoms in a crystal, which is far too coarse to pick out individual atomic planes – like trying to measure a hair with a yardstick. To get the resolution you need much shorter waves, and those land in the X-ray range. In 1912, Max von Laue realized that crystals would naturally bend X-rays in the same way a finely ruled diffraction grating bends visible light. William Henry Bragg and his son Lawrence then worked out the geometry behind that bending and turned it into a precise measurement tool.
The geometry behind it is intuitive. X-rays bounce off one plane of atoms and also off the next plane down. The two reflected waves only reinforce each other if the second one travels an extra distance equal to a whole number of its own wavelengths – otherwise they cancel out. That reinforcement happens at very specific angles, and those angles depend on the gap between the planes. Measure the angles where reflections appear, run the numbers, and you have the spacings. Photograph many reflections at once and you can reconstruct the entire three-dimensional structure. Almost everything modern science knows about the atomic arrangement of solids – from steel to silicon to DNA – was learned this way.
DNA is the famous case. Rosalind Franklin's Photograph 51, taken in 1952, was an X-ray diffraction pattern of crystallized DNA fibers. The cross-shaped pattern of spots is the unmistakable signature of a helix. Watson and Crick used Franklin's data, alongside their own modeling, to deduce the double-helix structure the next year. The same technique scaled up across decades has now solved the structures of hundreds of thousands of proteins, every commercially important pharmaceutical target, and most of the ingredients of cellular machinery. Crystallography is one of the most consequential measurement techniques ever invented.
The Lattice Is Always Singing
A perfect crystal at zero temperature would be silent. Every atom would sit at its lattice site, motionless. Heat the crystal at all and the atoms begin to vibrate around those sites, pulling on their neighbors as they go. Those coupled vibrations propagate through the lattice as waves. Quantize the waves and you get phonons – the discrete energy packets of lattice motion. Phonons are to a crystal what photons are to the electromagnetic field: the smallest indivisible quantity of vibration energy.
Almost every thermal property of a solid comes from phonons. Heat capacity, thermal expansion, thermal conductivity, the speed of sound, and the temperature at which a metal becomes brittle are all set by how easily phonons are created and how freely they propagate. They also mediate the attractive interaction between electrons that produces conventional superconductivity – lattice vibrations are what hold Cooper pairs together below the critical temperature. The lattice that looks static in a textbook diagram is, at any temperature above absolute zero, humming with billions of overlapping vibrational modes.
Why Steel Bends
A perfect crystal is a theoretical fiction. Real crystals are riddled with imperfections, and counterintuitively, those imperfections are what make solids useful. The simplest defects are point defects: a missing atom (vacancy), an extra atom shoved into a gap between sites (interstitial), or a foreign atom substituted for a native one (substitutional). The diffusion that lets carbon atoms move through iron to make steel is essentially vacancies and interstitials walking through the lattice. Without point defects, atoms could not move through a solid at all.
The defects that govern mechanical strength are one dimension up. A dislocation is a line defect – an extra half-plane of atoms inserted into the lattice, ending in a line. Apply stress to a metal, and dislocations glide through the crystal. Each glide step shifts atoms by one lattice spacing along the slip plane, and the cumulative effect is plastic deformation. The reason copper bends easily is that its FCC lattice has many slip planes and dislocations move freely. The reason ceramics shatter is that their lattices have very few easy slip directions, so dislocations cannot relieve stress and the crystal cracks instead.
Engineering strong materials usually means controlling dislocations. Cold-worked metal is harder than annealed metal because deformation creates so many tangled dislocations that they get in each other's way. Steel is hardened by precipitating tiny carbide particles that pin dislocations in place. The Hall-Petch effect is the observation that smaller grains, with more grain boundaries blocking dislocation motion, give stronger materials. Modern alloy design is, at heart, an exercise in arranging obstacles to dislocation glide. The strongest steel and the most ductile aluminum differ not in their atoms but in how their defects are arranged.
Diamond, Silicon, Salt
A few specific crystal structures have outsized importance. The diamond structure – carbon atoms each bonded to four neighbors at the corners of a tetrahedron – gives diamond its extreme hardness, its high thermal conductivity (better than copper), and its wide electronic band gap. Silicon, sitting directly below carbon in the periodic table, adopts the same crystal structure. Every microprocessor on Earth is patterned into silicon crystal that has been grown to a purity and perfection that has no parallel in any other industrial material. The Czochralski process, which pulls 300-millimeter ingots from molten silicon, routinely produces single crystals with fewer than one defect per billion atoms. The whole digital economy rests on that crystal.
Salt is the simpler example. Sodium chloride forms a face-centered cubic lattice with sodium and chloride ions alternating on the corners and faces. Cleave a salt crystal and it splits along flat faces because the ionic bonds are weakest along certain crystallographic planes. The cubic shape of a table-salt grain, visible to the naked eye, is the unit cell projected up by twenty quintillion repetitions. Almost every macroscopic property of an ionic solid – cleavage, brittleness, optical transparency in the visible band, dissolution behavior in water – flows from this geometric arrangement of charges.
Some crystals have practical magic. Quartz oscillates at extremely stable frequencies when squeezed, which is why it keeps your watch on time. Lithium niobate's nonlinear optical response makes it the workhorse of laser frequency conversion. Yttrium iron garnet (YIG) carries spin waves with vanishing damping and underpins much of modern microwave electronics. Each one is a particular crystal structure performing a particular trick that is impossible without the underlying symmetry. Picking the right crystal is half of solid state engineering.
Glasses, Quasicrystals, and the Spectrum of Order
Not every solid is crystalline. Glass is the textbook example of an amorphous solid – the silicon and oxygen atoms in window glass have well-defined local structure (each silicon bonded to four oxygens, each oxygen bridging two silicons) but no long-range periodicity. Try to draw a unit cell and you cannot; the structure never repeats. Most plastics, gels, soft tissues, and many ceramics are similarly amorphous. They behave as solids on human timescales because diffusion is too slow to let the atoms find a true crystalline arrangement, but they have no underlying lattice.
For most of the twentieth century, the dichotomy was simple: crystalline (periodic, can be diffracted) or amorphous (no order, broad diffuse scattering). Then in 1982 Dan Shechtman observed sharp diffraction spots from a rapidly cooled aluminum-manganese alloy – the hallmark of a crystal – arranged in a pattern with five-fold symmetry. Five-fold symmetry is geometrically forbidden in a periodic lattice; you cannot tile the plane with regular pentagons. The community initially refused to believe him. The alloy turned out to be a quasicrystal: a structure with long-range order but no periodicity, mathematically related to the aperiodic Penrose tiling. Shechtman won the 2011 Nobel Prize in Chemistry. The lesson was that the space of possible orderings is larger than crystallography had assumed for a hundred years.
Liquid crystals occupy yet another point on the spectrum. Their molecules have orientational order – they all roughly point the same way – without positional order. Apply an electric field and the molecules reorient, changing how light passes through. Every laptop screen and most televisions are built on this controlled partial order. Order is not binary. It comes in many degrees, and physics has had to keep enlarging its catalog of what counts as a solid.
The Bigger Picture
Crystals are the simplest way the universe demonstrates emergent order. Atoms by themselves are nearly identical; a sodium atom in salt is the same as a sodium atom in interstellar gas. What changes is their arrangement, and from arrangement alone come all the properties we recognize as the material world – hardness, color, conductivity, magnetism, transparency, the way a wineglass rings when struck. Solid state physics is the long working out of the principle that arrangement matters as much as composition.
Almost every modern technology rides on that principle. The transistors in your phone are dopants in silicon crystal. The display in front of you is liquid crystal modulated by an electric field. The lithium battery powering it is layered intercalation crystals shuttling ions between electrodes. The permanent magnets in its haptic motor are hexagonal-ferrite or rare-earth crystals with carefully chosen anisotropy. None of those technologies would exist without the century of careful work that turned where the atoms sit into a designable engineering parameter.
