# Can chemical composition be determined from elemental composition?

I have the energy dispersive X-ray spectroscopy results from a sample of salt that I produced by collecting and processing ocean water by filtering, boiling, evaporating, and additional steps to try to reduce calcium and magnesium content in the final salt crystals:

Atomic %
25.70 C carbon
20.68 O oxygen
20.30 Na sodium
2.71 Mg magnesium
14.25 Al aluminum
1.13 S sulfur
14.46 Cl chlorine
0.77 Ca calcium

I know that I can discount the aluminum content because the mount within the scanning electron microscope is made of aluminum and that's most likely what is being detected, as well as some of the carbon content because the carbon tape holding the salt sample to the mount is contributing to that detected amount.

If there were only one compound present, then I imagine the ratios in the atomic percentages / elemental composition would give us a pretty strong hint at what that compound would be, but evaporating ocean water leaves multiple compounds such as calcium carbonate $$\ce{CaCO3}$$, calcium sulfate aka gypsum $$\ce{CaSO4}$$, sodium chloride $$\ce{NaCl}$$, magnesium chloride $$\ce{MgCl2}$$, and potassium chloride $$\ce{KCl}$$ (https://manoa.hawaii.edu/exploringourfluidearth/chemical/chemistry-and-seawater/salty-sea/weird-science-types-salts-seawater).

Can I know with confidence that these same compounds make up my produced salt? Can I know with confidence that the salt doesn't contain other combinations of the ions present in ocean water, such as sodium carbonate $$\ce{Na2CO3}$$ or calcium chloride $$\ce{CaCl2}$$ or magnesium hydroxide $$\ce{Mg(OH)2}$$?

• Rather than speculating about the carbon and aluminum, you could do a control run with distilled water and subtract.
– user6999
Aug 9 at 13:36
• @BenCrowell Distilled water in scanning electron microscopy? This is not possible. Aug 9 at 14:05
• @M.Farooq I suspect Ben means process distilled water in exactly the same way the seawater was processed (i.e. evaporate it all off), and analyse it. That may be tricky if the processing resulted in a decent amount of mixed salt powder, but starting with a solution of (a known concentration comparable to the seawater) of pure NaCl in distilled water would also provide a control sample - as you would know, now I've seen your answer Aug 9 at 15:06
• What were the additional steps to try to reduce calcium and magnesium content? How could they affect other ions? Aug 9 at 15:10
• @KevinSu, In this way you did not reduce calcium or magnesium content. You just removed insoluble stuff from sea water. Aug 9 at 17:53

The short and sad answer is "no"- you cannot determine the compounds and their formulae just from energy dispersive spectrometry. You have identified the problem correctly as you are not dealing with a single compound but plently of mixtures. There are other deeper issues with EDS. First you should remember that no analytical technique is absolute with two or three exceptions, i.e., the analytical signal is useless in the absence of standards. So you do not have standards data, i.e., signals from known materials.

The next problem is that it your choice that you include or remove a certain element. If you remove carbon, all the atomic % will change. If you remove Al, it will cause further changes in numbers. This adds a lot of uncertainity and of course errors. Is the carbon from the tape or from carbonate? See these types of uncertainities exist.

Third problem is that did you gold coat your sample or not? You must have removed gold peaks too. Insulators are not that great in SEM. Most of it is addressed here Is scanning electron microscopy/energy dispersive X-ray spectrometry (SEM/EDS) quantitative?.

Since seawater is a solution, you cannot say what compounds were originally dissolved in it. You can talk about what ions were present and then you must account for charge balance to guess what were the starting compounds. If you were actually given the task to determine what salts and compounds were actually present, you will need to do a lot of wet chemistry and elemental analysis. The good news is that you do not have to re-invent the wheel. A lot of people have reported and studied seawater composition.

Coming to your isolated and purified salt. I feel the charge balance does not make sense in your data. You need to have a similar number of Cl atoms as sodium. If possible, do a pure NaCl EDS and see what you get.

The proper analysis of this salt will require quantitative ion chromatography for anions. Also then you need elemental data from ICP-OES to see which metals are present and in what ratio (Li, Na, K, Ca, Mg etc.), only then you can estimate what compounds you have in your isolated salt.

• Why wouldn't the charge balance make sense? There's likely a whole lot of carbonate anions in there (perhaps added in the "additional steps to try to reduce calcium and magnesium content"). If all the oxygen came from carbonates, it would account for over 13% (as monovalent), which would be plenty to balance all the cations. Of course some of that oxygen is water, but there's still a lot of room there. Aug 9 at 14:56