Frozen-hydrated samples are the gold standard when doing elemental analysis as there is almost no chance of redistribution or loss of elements – as the sample has simply been rapid frozen, with no or very minimal sublimation. In biology many of the constituent elements, particularly the light elements, diffuse easily and by any other sample preparation there will inevitably be some movement or loss of elements from the sample.

Chlorine, sulphur and oxygen elemental maps of a cryo-planed leaf of Melaleuca systena. The O map is used to show the plant structure. Images courtesy of Caio Guilherme-Pereira, University of Western Australia.

Energy dispersive X-ray microanalysis (EDS) is capable of both qualitative analysis, to determine the presence or absence of elements, and fully quantitative analysis, to determine element concentrations. Element concentration in frozen-hydrated samples is given in terms of mmol kg^-1 wet weight, which is much more biologically meaningful than concentration in terms of dry mass, which is obtained by other common analytical methods.

Cryo-fracturing of the sample is sufficient for qualitative results, but a near-perfect flat sample surface is needed for reliable quantitative analysis. This is because topography dramatically affects X-ray detection, especially of light elements. Cryo-planing, as previously discussed, produces a highly polished, flat sample surface well suited to quantitative analysis but these don’t show any structural information. As carbon and oxygen are found in differing concentrations in different biological structures, maps of the distribution of these elements can be particularly useful for revealing the underlying structure. However, with great care of temperature and time control during sublimation, surface features can also be exposed in order to recognize the structures being analysed while having little effect on elemental concentrations.

Carbon and oxygen elemental maps of a cryo-planed leaf of Grevillea thelemanniana showing the underlying cellular structure that is not readily obvious in the secondary electron image. Images courtesy of Peta Clode, University of Western Australia.

To prevent charging without compromising analytical results, the frozen samples should ideally be coated with a material that does not have X-ray peaks that overlap with the elements of interest. Common examples include carbon, aluminum, or chromium. In biology, an accelerating voltage of 15 kV is sufficient to excite most X-rays of interest (including Mn, Fe, Cu) and offers approximately 2µm analytical resolution.