Before-and-after pictures. LA-ICP-MS experience.

   Before                                                                         After

Remains of amphibole in the ablation crater (center, right image) demonstrate relatively low birefringence color. The pictures were taken using different settings of camera and microscope, XPL.

Closeup of the crater in the same spot. Reflected light.

Imagine you want to determine the concentrations (even below 1ppm=1⋅10^-4%) of  trace elements, the isotope ratios not in a whole rock but in a 30 µm spot of a mineral grain. Laser ablation inductively coupled plasma mass spectroscopy (LA-ICP-MS) makes it possible! For the most part, the method is identical to any kind of ICP-MS. However, the sample introduction happens in situ. The sample is exposed high energy monochromatic radiation - a laser beam (~30 µm in diameter). Subsequently, a part of the sample instantaneously vaporizes and some part of the produced vapor becomes ionized (the trick is to miss melting of the sample). After this happened, carrier gas transports the vapor to inductively coupled plasma (ICP) where all the substance breaks down into charged particles - ions. The ions undergo magnetic field separation (in ICP-SFMS) and detection. 

Any kind of solid samples can be studied, or at least attempted to be studied. The ablated surface should be flat just for aiming purposes, polishing is not required. Ablation is a destructive technique, however diameter of a laser beam is usually around 30-50 µm. Besides the parameters of laser (such as ablation time), the depth of ablation heavily depends on the ability of material to absorb laser radiation. Normally, ablation depth exceeds 30 µm thickness of a regular section and in some cases could be up to 50 µm. In this regard, slightly thicker section was used.

On the picture above we see that the laser made a crater. Since the thickness decreased. the lower birefringence (grey 1st order) is observed in XPL. Assuming that the initial thickness was about 45 µm the birefringence value is equal 0.012 (the beginning of the second order). Using Michel-Lévy nomogramm (which I finally translated and made downloadable), I assume thickness went down to roughly 17 µm. Thus, the ablation depth is 45-17= 28 µm.

The resulted spectrum (count per seconds versus time) looks like this:

Note that the intensities measured for different elements increase and decrease in the same way. The patterns are identical.  That means all the elements were ablated in the same way, without any fractionation.

Since I produced the thin section myself, it is not perfectly flat, so the edges of the section are slightly thinner.

Before. Large crystals of alkali amphibole located on the edge of the section. The thickness is somewhat 30 um. 

The instrument (over)made its job. It produced an empty hole though the crystal. Apparently, the thickness on the edge of the section is around or even less than 30 µm.

After. The laser beam made a hole through the crystal. Probably some of epoxy was ablated as well.

The resulted spectrum:

The spectrum yields non-standard behavior of Zr (and some other elements that are not included in the graph). It is hard to tell what exactly caused this elemental fractionation that but the data cannot be used. Possibly, the ablation of epoxy that underlies the crystal, could affect the result.