Extended Data Fig. 5: Experiments conducted for exploring possible chemical mechanisms underlying staining contrast generation.
(a,b) Membrane contrast after adding an intermediate CaC washing step between OsO4 and FeCN. Imaged in SEM in high vacuum, electron dose 31 e−/nm2. (c) Membrane contrast after staining with potassium osmate (vi) (note yielded precipitation band). (d) UV-vis spectrum of potassium osmate (vi) in sodium cacodylate buffer over 24 h. (e,f) 2 mm samples stained with ‘reduced osmium’ protocol for 3 h vs. 24 h. (g) Membrane contrast after adding CaCl2 into OsO4 before FeCN incubation. (h) Membrane contrast after OsO4 incubation at 4 °C before FeCN incubation. SEM acquisition for panels c,e,f,g,h was performed in low vacuum at 30 Pa, maximum electron dose of 70 e−/nm2. (i-m) Possible chemical logic underlying staining results (i) Comparison of bright-field color appearance of samples stained in OsO4 for 2 h, 24 h, 24 h (diluted to evaluate color), and stained in OsO4(24 h)->CaC->FeCN (intermediate CaC wash to remove OsO4), top row. For comparison, pure staining solutions are shown below: Os(vi) and in vitro reaction of Os(vi) with FeCN (bottom). Note that sample solution after 24 h OsO4 incubation appears similarly colored as Os(vi) solution (pink); and sample solution after Os-Cac-FeCN staining appears similarly colored as Os(vi)+FeCN reaction in vitro (blue−green). (j). (left) Raman spectrum measurements of the product of Os(vi) + FeCN showing a shifted Os(vi) peak; (middle) in the wavelength range corresponding to FeCN(ii) and FeCN(iii), there was only signal consistent with FeCN(ii) peaks, but not FeCN(iii), suggesting the reaction between Os(vi) and FeCN was not a redox reaction. (k). The same Os(vi) signal was observed in brain staining solution after FeCN staining and in-vitro Os(vi) + FeCN reaction. (l) Sketch summary of the possible chemical reactions (see in Supplementary Results). (m) Summary of Os(vi) coordination chemistry potentially relevant for staining mechanisms.
