I also suggest (if possible) to combine with XRD analysis that will provide mineralogical composition of your samples and help you to discriminate between potential minerals. Regards

Dear Mr. Higgs,

you do not reveal very much about your unknown mineral. Therefore it is hardly possible to give you anything more than a rough and general piece of advice. Based upon the atomic % you might get an idea of the mineral. The second pillar upon which you can base your "mineralogical constraints" is the morphology or crystal habit in case of a euhedral or at least subhedral mineral. Both parameters you should evaluate in context with additional information that you might have collected such as the environment of deposition, the diagenetic overprinting or epigenetic alteration.

Best regards

H.G.Dill

Hello,

if you have the chemical composition of your minerals and you want to know what kind of mineral you have there are some useful webpages which show you the ideal composition of a mineral.

http://webmineral.com/chemical.shtml

https://www.mineralienatlas.de/lexikon/index.php/MineralFindElement

http://www.mindat.org/chemsearch.php

There is probably more than one result for your chemical composition. Then you have to make a decision based on the morphology, facies conditions and paragenesis (as Mr. Dill stated).

Best Regards

I also suggest (if possible) to combine with XRD analysis that will provide mineralogical composition of your samples and help you to discriminate between potential minerals. Regards

Hi Caldin,  The above replys are all correct.  I have done many SEM/EDS probable identifications of minerals.  If I suspect a certain mineral or mineral group, I would take a small sample of a known mineral and obtain a spectra and compare the spectra.  It is not accurate but gives one a general indication of what your uk might be.  PXRD is the best to identify the mineral.  Yes, physical properties, such as color, crystal shape are important in the final analysis.

Gail

You need an XRD analysis to identify the Mineral in the sediment sample.

With respect to EDX elemental analysis and mineral identification, the most critical element may be oxygen and how it was analyzed because its energy location is crowded with interfering elements. E.g., was O actually measured relative to a reference standard, or was it analyzed via some standardless method, or was oxygen not analyzed but assumed to be stoichiometric? Do you know (suspect?) if the mineral is hydrated because non-analyzed hydrogen will further confuse the oxygen stoichiometry? It would be helpful to us if you could include a graphic of the x-ray spectrum. 

Please remember that SEM/EDX does not provide mineralogy - it is simply a technique that gives us element compositional information from which we INFER the mineralogy (based on our knowledge of the sample plus other information such as shape, optical properties etc.). Using a SEM/EDX system alone can generate erroneous results particularly in assemblages where there are polymorphs (e.g. andalusite/kyanite) where the distinction on chemistry alone is impossible, or in situations where there are complex solid solutions. You need some sort of direct MINERALOGICAL technique such as XRD which, based on crystal chemistry, will give you a more definitive answer. I note however, that if you are trying to identify a small rare grain, techniques such as XRD are also problematic as they usually have a resolution of ~1wt%. In such situations you could try EBSD on the polished sample. If you don't have access to such equipment you will then have to rely on the SEM/EDX information and make some sort of educated guess as to what the mineral answers.

Mark  

Hi Kenneth, I agree that an old Debye-Scherrer or Gandolfi powder camera would be perfectly suitable but as you say such equipment is rare nowadays (although here at CSIRO we have two but no-one is much inclined to do this sort of work anymore except for a couple of our retired honorary who are world experts). Also, depending on how common, how large etc the mineral grain of interest is, it usually requires quite some expertise to remove the grain from the sample block or slide. I figured EBSD would be simpler and quicker (provided you have access to a suitable SEM equipped with such a detector).

Mark