I am often asked how one should choose the wavelength of the laser for LIBS analysis. This is particularly important for applications involving nanosecond lasers, where the laser-material interaction is pronounced. For femtosecond lasers, the laser pulse is so fast that there is little material interaction (few or no lattice vibrations during the pulse), and there’s little plasma interaction (the plasma forms after the pulse). For this reason, these comments are primarily directed toward nanosecond laser selection for LIBS.
There are two primary considerations. The first is absorption of the laser by the (presumed solid) material being ablated . UV lasers (e.g. 266 nm, 213 nm) are absorbed more fully, particularly by transparent samples. So if your sample is, say, glass, consider a UV laser.
The downside to the UV laser, besides typically having a lower pulse energy (due to conversion) for the same price as a visible or IR equivalent, is that it also is not as well-absorbed in the plasma. Plasma absorption of light is a function of many things, a prime factor is electron density. As electron density increases, the wavelength of light absorbed decreases. This means that at early stages of plasma formation, when the electron density is low, a visible or IR laser will “pump” the plasma better than a UV laser.
Summary: better ablation –> UV laser; more intense plasma –> visible or IR laser. There is no right answer. For more discussion, see my article in Spectroscopy linked here.