# How do people check the type of defect in a crystal?

There are many types of defects like

• Vacancy
• Interstitial
• Schottky
• Frenkel
• Non-Stoichiometric
• Alien Impurities

But given a sample, how do scientists know which kind of defect is present in it [practically]?

As you said there are many different kind of defects in crystals. There are basically two different methods to find the Crystallographic defect.

1. Mathematical classification methods
2. Computer simulation methods

Mathematical classification methods:

A successful mathematical classification method for physical lattice defects, which works not only with the theory of dislocations and other defects in crystals but also, e.g., for disclinations in liquid crystals and for excitations in superfluid $\ce{^{3}He}$, is the topological homotopy theory.

Computer simulation methods:

Density-functional theory, classical molecular dynamics and kinetic Monte Carlo simulations are widely used to study the properties of defects in solids with computer simulations. Simulating jamming of hard spheres of different sizes and/or in containers with non-commeasurable sizes using the Lubachevsky-Stillinger algorithm can be an effective technique for demonstrating some types of crystallographic defects.

Source: Wikipedia

Given amount of atoms in typical macroscopic crystal, there is no need to check if any type of defect is present - all of them are presence because of sheer numbers and the law of big numbers. For estimating rough amounts different spectroscopic methods may be used. However, it is an indirect technique, which requires somehow obtained idea of the spectrum of the defect in question. Unless a very rough idea is enough, like detecting impurities using different nuclear spectroscopy methods, the only universal way is usage of quantum chemistry methods, and they are often frustratingly inaccurate when spectrum is a property of interest.

For nanosized structures, surface,near-surface and surface-influenced defects may be detected using hi-definition microscopy techniques, like scanning tunneling microscopy, which may provide subatomic resolution. However, in this case a different problem arises. Humanity usually does not need specific knowledge about some particular crystal, more a typical picture for some kind of systems is of interest. In such a case the only way is to make a lot of observations and somehow reduce it to typical values, and such a procedure is usually debatable, as samples for atomic-resolved spectroscopy must be prepared in very specific ways (see method descriptions for details), and so are very different from what should be present in conditions of interest.