unfortunately you haven't specified which "EM" you are processing tissue for and use for examination: TEM or SEM?
OsO4 one of the eldest (chemical) fixatives used in Histology, as well as transmission and Scanning Electron Microscopy, indispensable perhaps in (in terms of „classical processing“ unproblematic) lipid fixation….etc., etc. The STORAGE, HANDLING of OsO4 CRYSTALs, the MAKING/mixing and proper / SAFE USE and DISPOSAL of „hazardous, toxic and volatile OsO4 solution“ in most academic / scientific institutions consequently needs S.O.P.‘s to be followed strictly.
Furthermore: The assumption/your statement " OsO4 too expensive, toxic and volatile" might be correct in some respect.... but for scientific reasons I have to doubt seriously about the consequence one could deduce: "OsO4 therefore should/could be omitted as a(secondary/tertiary) fixative in EM...of bacterial cell".
First of all: "Electron Microscopy" always has been told and assigned to really be not a cheap but expensive (despite traditionally called/designated to be an "ancillary") technique. So I wonder about why we do it still (at least at special „Centers of excellence“?)
There are many other circumstances making EM (TEM) expensive rather than explicitely "use of OsO4" . I could tell more about that (admitting that I have worked in and was head of a diagnostic TEM-Lab for nearly 35 years and always worked „hands-on“ at the bench and used many other hazardous chemicals you should worry about too).
Regarding your question "is there another substance...." from recent articles in/on modern TEM (digital) imaging I learnt that ultimate use of (toxic) heavy metals (i.e. OsO4, UO2Ac, etc.) for gaining "contrast" of tissue components in ultrathin sections isn't implicitly necessary in the future. But generally it will depend on the TEM „hardware“ you can or are allowed to use. If you haven't access to modern TEM-technology (and most recently developed digital imaging systems) and need to use TEM of vintage e. g. 2000-2005 or older you either have to use OsO4 or you may try other methodological applications (i. e. combined = mixed - fixatives which consist of chemical additives to "classical" fixatives [e.g. using KMnO4-fixative, using "mordanting substances" like Tannic acid, eventually combined with K-ferri-/or ferrocyanide, or even using PPD=paraphenylenediamine as retaining/retention agent] and / or thinking about changing your "classical" staining [e.g. UO2Ac-Pb-citrate] gaining contrast in] your ultrathin sections. There might be solutions without using OsO4 (and there have been some threads on ResearchGate regarding such matter*), nevertheless it will depend on the task of your studies whether you can totally omit OsO4 postfixation or not .
*) e.g. https://www.researchgate.net/post/Is_there_any_OsO4_osmium_tetroxide_alternative_for_fixation_in_immunogold_labeling_for_TEM ;
in the pioneer age of the electron microscopy one of the often used fixatives was the kalium permanganate. If it was single used, the results were not excellent, but I do not know if somebody used it as the post fixative after aldehyde fixation. Try it.
unfortunately you haven't specified which "EM" you are processing tissue for and use for examination: TEM or SEM?
OsO4 one of the eldest (chemical) fixatives used in Histology, as well as transmission and Scanning Electron Microscopy, indispensable perhaps in (in terms of „classical processing“ unproblematic) lipid fixation….etc., etc. The STORAGE, HANDLING of OsO4 CRYSTALs, the MAKING/mixing and proper / SAFE USE and DISPOSAL of „hazardous, toxic and volatile OsO4 solution“ in most academic / scientific institutions consequently needs S.O.P.‘s to be followed strictly.
Furthermore: The assumption/your statement " OsO4 too expensive, toxic and volatile" might be correct in some respect.... but for scientific reasons I have to doubt seriously about the consequence one could deduce: "OsO4 therefore should/could be omitted as a(secondary/tertiary) fixative in EM...of bacterial cell".
First of all: "Electron Microscopy" always has been told and assigned to really be not a cheap but expensive (despite traditionally called/designated to be an "ancillary") technique. So I wonder about why we do it still (at least at special „Centers of excellence“?)
There are many other circumstances making EM (TEM) expensive rather than explicitely "use of OsO4" . I could tell more about that (admitting that I have worked in and was head of a diagnostic TEM-Lab for nearly 35 years and always worked „hands-on“ at the bench and used many other hazardous chemicals you should worry about too).
Regarding your question "is there another substance...." from recent articles in/on modern TEM (digital) imaging I learnt that ultimate use of (toxic) heavy metals (i.e. OsO4, UO2Ac, etc.) for gaining "contrast" of tissue components in ultrathin sections isn't implicitly necessary in the future. But generally it will depend on the TEM „hardware“ you can or are allowed to use. If you haven't access to modern TEM-technology (and most recently developed digital imaging systems) and need to use TEM of vintage e. g. 2000-2005 or older you either have to use OsO4 or you may try other methodological applications (i. e. combined = mixed - fixatives which consist of chemical additives to "classical" fixatives [e.g. using KMnO4-fixative, using "mordanting substances" like Tannic acid, eventually combined with K-ferri-/or ferrocyanide, or even using PPD=paraphenylenediamine as retaining/retention agent] and / or thinking about changing your "classical" staining [e.g. UO2Ac-Pb-citrate] gaining contrast in] your ultrathin sections. There might be solutions without using OsO4 (and there have been some threads on ResearchGate regarding such matter*), nevertheless it will depend on the task of your studies whether you can totally omit OsO4 postfixation or not .
*) e.g. https://www.researchgate.net/post/Is_there_any_OsO4_osmium_tetroxide_alternative_for_fixation_in_immunogold_labeling_for_TEM ;