# What substances in ash may create a green colour and are also soluble in water?

I have an ash that is high in $\ce{K2O}$ ($\gt 30 \text{%}$), $\ce{SiO2}$ ($\gt 10 \text{%}$) and $\ce{CaO}$ ($\gt 10 \text{%}$), with $\ce{P2O5}$ ($5 \text{%}$) and $\ce{MgO}$ ($3 \text{%}$) also present. When the ash is heated to above $700^\circ \pu{C}$ for $16$ hours for decarburizaton, a color change is observed, from beige to a shade of turquoise, which at higher temperatures ($1300^\circ \pu{C}$ for $10$ mins) becomes a deep shade of green during cooling (the green substance appears to be condensing through the sample).

Whatever substance that is causing this green color appears to be inhibiting melting within the sample (melting temperature should be around $1200^\circ \pu{C}$, according to ash fusion testing), and is soluble in water. I've found through XRD that the ash has significant peaks that correspond to $\ce{KCl}$ in the form of sylvine, but I'm not sure that there is enough chlorine to cause such a deep color change, or that it could cause the color change at $700^\circ \pu{C}$.

So, my question is, does it seem possible that $\ce{KCl}$ is what is causing this color change, and are there other possible candidates that I should test for? (For instance, I've read that potassium manganate will dissolve in water, resulting in a green color too, but I'm not sure how feasible that is).

• A little impurity can go a long way in the color world. Replace 1% of aluminum atoms in $\ce {Al2O3}$ with chromium and you get ... en.m.wikipedia.org/wiki/Ruby. – Oscar Lanzi May 25 '17 at 10:11
• Yeah, that's part of my problem: the ash I'm looking at has so many components! At first I thought it might have been a phase change to Amazonite due to some lead impurities, but after seeing the colour change appear to be from condensing I'm not so sure any more. Any ideas that might narrow my options a bit are appreciated! – Lee Roberts May 25 '17 at 10:25
• Is it $\ce{FeSO4}?$ Wikipedia says: All the iron(II) sulfates dissolve in water to give the same aquo complex $\ce{[Fe(H2O)6]^2+}$. – Berry Holmes May 25 '17 at 12:29
• I haven't got a sulphate analysis of the ash yet, but oxide analysis only accounts for some 65% of the ash, and carbonates only 3%, so that's definitely something I'd like to look into. Thanks! – Lee Roberts May 25 '17 at 12:34
• We've just finished some hot-stage XRD scans, and found that we created a pretty thick blanket of KCl over the top of the sample during cooling. I'm starting to think that this KCl coming into contact with moist air while the sample is still hot is what's causing the colour change. There are other peaks to look at though, so I'll have to see how iron and nickel fit if at all. Thanks for the help! – Lee Roberts May 25 '17 at 15:23