A wave meets an atom
The last rung ended with a bold promise: a diffraction pattern is a direct photograph of the reciprocal lattice, the crystal's shadow. But a promise is not an explanation. WHY does a crystal cast a shadow made of sharp spots at all, and why do the spots land exactly where the reciprocal points sit? This whole rung is the answer, and it starts at the smallest possible scale — one wave meeting one atom. Get that single encounter right and the crowd behaviour follows.
An X-ray is a travelling oscillation of the electric field. When it sweeps over an atom, that field grabs the atom's electrons and shakes them back and forth at the wave's own rhythm. An accelerating charge radiates, so each shaken electron becomes a tiny transmitter, broadcasting its own spherical wavelet of exactly the same wavelength as the wave that drove it. That re-broadcasting is scattering, and because the outgoing wavelength matches the incoming one it is called elastic (or coherent) scattering — the only kind that can build a sharp pattern. Picture a buoy dropped into ocean swell: it bobs at the swell's rhythm and throws out its own set of circular ripples. The atom's electrons, jostled by the passing X-ray, do the same with light.
Why the wavelength must match the atoms
To read a ruler you need marks finer than the thing you are measuring; to feel out spacings of one to three angstrom you need a wave whose wavelength is roughly that size. There is a hard geometric reason, not just a rule of thumb. As the next guides will make exact, diffraction needs the path difference between neighbours to reach a whole wavelength, which forces lambda to be no larger than about twice the spacing (lambda ≤ 2d, since a sine cannot exceed 1). A wave far longer than the atomic grid simply cannot register it. This is the wavelength requirement, and it is why ordinary light fails: visible light is around 5000 angstrom, thousands of times too coarse, sailing over the atoms like ocean swell over gravel.
This is exactly where X-rays earn their place. The everyday laboratory line, copper K-alpha, has lambda = 1.54 angstrom — a near-perfect match to the distance between atoms. That line is characteristic radiation: knock an inner electron out of a copper target inside an X-ray source and the atom snaps back with a photon of one sharp, element-specific wavelength. Put a real number through it. For a plane spacing d = 2 angstrom, the first diffracted beam appears at theta = arcsin(1.54 / (2 times 2)) = arcsin(0.385) = 22.6 degrees — a comfortable, easily measured angle. Had lambda been 5 angstrom, the required sine would exceed 1 and no beam could form at all.
Two honest comparisons round this out. Thermal neutrons sit around 1 to 2 angstrom, squarely in range, and they see light atoms that X-rays barely notice. Electrons are the odd one out: at 100 to 300 keV their wavelength is a few picometres (about 0.02 to 0.04 angstrom), far SHORTER than an atom, yet they diffract beautifully — a short wavelength is no obstacle, only a long one is. The catch is that electrons interact with matter thousands of times more strongly than X-rays, so a transmission electron microscope needs ultrathin samples and its beams scatter again and again (dynamical scattering), muddying the simple intensity rules we are about to build. Match-the-spacing is the clean story for X-rays and neutrons; electrons cheat with tiny wavelengths and pay for it in sample prep.
One atom's brightness fades with angle: the form factor
Even a single atom does not scatter equally in all directions. The bookkeeping quantity is the atomic scattering factor f, also called the form factor: it is how strongly the atom scatters, measured in units of what one lone electron would scatter. Straight ahead — into the forward direction, zero scattering angle — every electron in the cloud radiates perfectly in step, so their amplitudes simply add and f = Z, the atomic number (the number of electrons). Copper, Z = 29, forward-scatters like 29 electrons; oxygen like 8; hydrogen, with its single electron, like a feeble 1.
Now swing to a wider angle and the number drops. The electron cloud is not a point — it is smeared over roughly 0.1 to 0.2 nm, a size comparable to the wavelength itself. Wavelets leaving the near side and the far side of that cloud travel slightly different paths, so at anything but forward scattering they slip out of step and partly cancel; f falls below Z, and keeps falling the wider you go. The variable f actually depends on is sin(theta)/lambda, which is just a compact stand-in for the scattering vector — the change in the wave's direction of travel. Bigger turn, more cancellation across the cloud, dimmer atom.
ATOMIC SCATTERING FACTOR f (X-rays; amplitude in units of one electron)
sin(theta)/lambda f(O) f(Fe) f(Cu)
(per angstrom) Z = 8 Z = 26 Z = 29
-----------------------------------------------------------------
0.0 (forward) 8.0 26 29 all electrons in step
0.2 5.6 22 24
0.4 3.2 17 19 cloud starts to cancel
0.6 2.1 13 15
0.8 1.5 11 12
-----------------------------------------------------------------
f = Z straight ahead, then FADES as wavelets from opposite sides
of the finite electron cloud fall out of step at wider angles.
(values approximate / representative, not exact table entries)One honest consequence. For X-rays f roughly tracks Z, so heavy atoms shout and light ones whisper — hydrogen is almost invisible, which is why locating hydrogens by X-rays is notoriously hard. Neutrons flip this: the neutron scattering length does not climb with Z at all, so hydrogen and deuterium scatter strongly and even differ from one isotope to the next. That is precisely why neutron diffraction is the tool of choice for pinning down light atoms and for telling apart neighbours on the periodic table that X-rays see as near-identical.
Interference: how a crowd of atoms picks directions
One atom scatters weakly and almost everywhere. The magic is the crowd. In a crystal, identical scatterers sit on a periodic grid, and their spherical wavelets overlap and add. In nearly every direction the crests and troughs arrive at random offsets and average to nothing — destructive interference, and the scattered light quietly dies. But in a handful of special directions the path difference between one atom and its neighbour is exactly a whole number of wavelengths; then every wavelet arrives crest-on-crest, the amplitudes pile up, and a bright beam flashes out — constructive interference. This is the echo from a row of evenly spaced cliff walls that only booms back at particular angles, the very picture Bragg's law will make exact in the next guide.
- Send a plane wave into the ordered array of atoms.
- Each atom re-radiates a spherical wavelet of the SAME wavelength (that is the scattering from the first section).
- Pick an outgoing direction and find the path difference between the wavelets from two neighbouring atoms.
- If that path difference is a whole number of wavelengths, the wavelets add in phase — a bright beam.
- For any other value they arrive out of step and cancel — darkness.
- So a periodic array turns one smooth incoming wave into just a few sharp outgoing beams — and the more atoms take part, the knife-sharper those surviving beams become.
Steps 3 and 4 — counting the path difference — can be done in two equivalent dialects, and each gets its own guide ahead. Bragg's law (n lambda = 2 d sin theta) counts the path difference between successive PLANES in real space: the familiar reflecting-planes picture. The Laue conditions do the identical counting in reciprocal space, and their geometric embodiment is the Ewald sphere — a reflection fires exactly when a reciprocal point touches that sphere. Same physics, two languages: the real crystal and the shadow world you built last rung. Neither is more true than the other; you reach for whichever makes a given problem simpler.
From amplitudes to what the detector sees
Interference tells you WHERE the beams appear. A second question is HOW BRIGHT each one is, and that is where scattering and interference marry. Add up the wavelets from every atom inside ONE unit cell — each carrying its own form factor f (how loud that atom is) and its own phase (set by where the atom sits in the cell) — and the complex total is the structure factor F_hkl. It is the single number that fixes a reflection's amplitude and phase, and it is built from the lattice-plus-motif recipe: the lattice says where the cells repeat, the motif inside the cell decides how the wavelets add.
But here is the hitch that shapes the rest of your journey. A detector counts photons — it measures the diffracted intensity, not the amplitude, and intensity is proportional to |F|^2, the amplitude squared. Squaring keeps the size of F but discards its phase. That vanished phase is the famous phase problem: to reconstruct where the atoms sit you need the full complex F, yet the experiment only ever hands you |F|. Recovering the missing phase is the central puzzle of structure solution, and no amount of measuring more intensities gives it back directly.
Close on two honesties, because the tidy story has fine print. First, everything above assumed each wavelet scatters just once on its way through the crystal — the kinematical picture. When the interaction is strong, above all for electrons in a TEM, a beam scatters, re-scatters, and re-scatters again, and the clean intensity-proportional-to-|F|^2 rule breaks down; that is dynamical scattering, and it is the price electrons pay for their strength. Second, 'scatter' is not the same as 'diffract sharply'. An amorphous glass still scatters every X-ray that hits it — it just gives broad, blurry halos instead of spots, because it has sharp short-range order but no long-range periodicity. Sharp peaks are the signature of a repeating lattice, not merely of atoms being present. Next guide turns the crowd's special angles into a single exact equation: Bragg's law.