It seems advisable to avoid phosphoric [1] and hydrofluoric acids, to exclude possible etching of the silicon nitrate (i.e. silicon(4+) tetranitrate, H8N4O12Si) without the need for protection of the substrate with a 'masking' material. The following alkaline etchants seem particularly promising [2,3]:
a) Add 10 cm3 KOH 40% w/w aq. to 25 cm3 6% (20 vol.) H2O2, where NaOH may substitute for KOH at the same molar amount and the alkaline component may eventually be diluted (from 1:1 to 1:3). Immerse the specimen for 30-60s at 70-80 ºC (recommended for metallography).
b) Use a modified Murakami's etchant: dissolve K3Fe(CN)6 (30% w/w) and NaOH (10% w/w) in water, where KOH may possibly substitute for NaOH at the same molar amount and the alkaline hydroxide can be more concentrated. Note that the original Murakami's etchant is obtained by dissolving K3Fe(CN)6 and KOH in water (both 10% w/w) [e.g. 4].
Safety note: do not use potassium ferricyanide in strongly acid solutions.
References:
[1] K. Nakamura, “Lithography,” in C.Y Chang and S.M. Sze (Eds.),“ULSI Technology", McGraw-Hill, 1996, pp. 270-328.
[2] George F. Vander Voort, "Metallography: Principles and Practice", Material Science and Engineering Series, ASM International, 1999, pp. 692-696.
[3] L. M. Gammon, R. D. Briggs, J. M. Packard, K. W. Batson, R. Boyer, C. W. Domby, "Metallography and Microstructures of Titanium and Its Alloys", in: G. F. V. Voort (Ed.), ASM Handbook, Vol. 9: Metallography and Microstructures, 2004, pp. 899-917 and 543-544 (colour micrographs).
[4] J. B. Fox, D. M. Cheney, A. A. Bauer, R. F. Dickerson, "Studies for the preparation of zirconium-clad uranium-10 w/o molybdenum fuel pins", Report No. BMI-APDA-644, Atomic Power Development Associates / Battelle Memorial Institute, 1958, p. 6. Available at: http://www.osti.gov/bridge/purl.cover.jsp?purl=/4181083-uHtBDq/4181083.pdf