peak position
Peak position is simply where along the 2-theta axis a diffraction peak sits. Of the three things a diffractogram tells you, position is the one that reports on the size and shape of the box the atoms live in — the unit cell — and nothing about what is inside it. Move the box's walls and the peaks slide; rearrange the furniture inside and the peaks stay put (only their heights change).
The link is Bragg's law, lambda = 2 d sin(theta), which turns each measured angle into a plane spacing d, and then the plane-spacing equation ties d to the lattice parameters. For a cubic crystal that equation is 1/d^2 = (h^2 + k^2 + l^2)/a^2. Worked example: with Cu K-alpha (lambda = 1.5406 angstrom), a peak measured at 2-theta = 44.5 degrees means theta = 22.25 degrees, so d = lambda/(2 sin theta) = 1.5406/(2 x 0.3785) = 2.035 angstrom. Do this for every peak, work out which (hkl) each belongs to (indexing), and the whole set of positions pins down the lattice parameters.
Precise positions buy precise lattice parameters, and those in turn reveal thermal expansion (peaks shift as the cell grows with temperature), alloy composition through Vegard's law (the cell size tracks how much of each element dissolves), and residual stress (strain shifts peaks slightly). Be honest about a classic trap: a mis-set sample height or a zero-angle error shifts every peak the same way and masquerades as a wrong lattice parameter — which is why careful workers add an internal standard of known spacing to correct positions.
For aluminium (FCC, a = 4.05 angstrom) the (111) peak sits at 2-theta = 38.5 degrees with Cu K-alpha; heat the sample and the peak creeps to lower angle as the cell expands and d grows.
Position is pure geometry — it fixes the unit cell, not the atoms inside it.
Peak positions give only the cell (geometry), never the motif. And a sample-displacement error shifts positions systematically, faking a change in lattice parameter — always check against a standard.