During June 1999, a query was posted to Histonet asking for demonstration methods for Arsenic in tissues. Following is a summary:
Roy Ellis (1) and Lynette Thibodeau (2) suggest Castel's Method:
Deparaffinized sections show green granules of Scheele's green (CuHAsO3) which, though insoluble in water, is dissolved by acids and by ammonium hydroxide. By substituting copper acetate for the sulfate, the green granular paris green or cupric acetoarsenite is produced. Its solubilities are similar (Castel's method, Bull.Histol.Appliq. V13:106, 1936). A light safranin counterstain gives good contrast. Method courtesy of Lillie, 3rd edition, Histopathologic technic and practical histochemistry, page 445.
John Kiernan (3) writes that arsenic compounds react with hydrogen sulphide to give insoluble As2S3. This is yellow, and unlikely to be visible, but could probably be amplified with a physical developer ("autometallography" or Timm's sulphide-silver method). However, this procedure demonstrates pretty well every metal that has an insoluble sulphide, so it would have no specificity for Arsenic.
John also notes that there is a Japanese investigator, Y. Sumi, who has developed histochemical methods for inorganic substances based on combinations of chromogenic reagents with mixtures of "masking" agents that block the reactivity of elements other than the one you want to stain. His best known methods are for Cd and Hg, but he may have done something for As.
Philip Oshel (4) suggests that if the arsenic is expected to be in nontrace amounts, you can detect, and maybe semi-quatify its presence with energy-dispersive x-ray spectroscopy (EDX or EDS) in a SEM or TEM. SEM might be better, as you could use "bulk" specimens--the small bits of tissue you prepare for sectioning. If there's enough arsenic, you could use paraffin-embedded sections in the SEM, but be aware that you could melt the paraffin. This can cause enough contamination in the column to antagonize the person in charge of the 'scope.
For TEM, you'd have to use thin sections, but this would give better localization, if that's important. Prepare the specimen according to routine procedures for your tissues for either SEM or TEM.