I am looking for a liquid which is not miscible (or very close) with water or vegetable oil. Ideally, I want to have a container which shows three layers (the three liquids) with clear boundary layers.

I was thinking that I should find a non-polar compound of very high molecular weight - such that difference in the density of that compound and oil is enough to draw it down out of the oil. However, I haven't found it yet.

I would prefer a compound which is non-toxic and non-flammable if possible.

It feels like such a simple question, I must be missing something.

  • 7
    $\begingroup$ One thing you might want to add to your question is how much stability you want/need - for example, if you're careful, you can get quite a lot of layers set up, but if you plan to shake it, most of these will combine. $\endgroup$
    – Aesin
    Commented Jun 13, 2012 at 20:07
  • $\begingroup$ Remember, miscibility isn't that much to do with density as it has to do with polar/nonpolar. If you get a third (polar) layer above the oil, on shaking it'll mot probably dissolve in water. $\endgroup$ Commented Jun 14, 2012 at 1:18
  • $\begingroup$ @Aesin: You can make that into a nice answer, especially if you explain how miscibility works and why shaking it kills the system. $\endgroup$ Commented Jun 14, 2012 at 7:39
  • $\begingroup$ Sorry, I should have been more specific. There will definitely be moderate mixing in the system. Using metal is possible, but I would like there to be a fairly significant amount of stuff (aka 5 - 10 gallons worth) because I'd like it to be a easy to see visual display, so cost may be an issue there. $\endgroup$ Commented Jun 14, 2012 at 12:25
  • $\begingroup$ David E. H. Jones demonstrated a jar with seven or eight liquid phases in the 70ties. (of course shaking-proof!) $\endgroup$
    – Georg
    Commented Jun 28, 2012 at 20:29

2 Answers 2


Fluorocarbons are compounds that are hydrophobe and lipophobe. It's a special property of perfluorinated compounds: They are non-polar and thus hydrophobic but in addition as the high electronegativity of fluorine reduces the polarizability of the atom, fluorocarbons are only weakly susceptible to the fleeting dipoles that form the basis of the London dispersion force. As a result, fluorocarbons have low intermolecular attractive forces and are lipophobic. I'm not sure about their toxicity, but this seems to be an issue, e.g. here.

  • $\begingroup$ Good suggestion. Many are highly toxic and not desirable, but some of them are used in medical applications and would be acceptable. $\endgroup$ Commented Jun 14, 2012 at 16:38
  • $\begingroup$ @Greg: you'll probably be interested in those perfluorocarbons used as "blood substitutes", due to their capacity for retaining oxygen. They are nontoxic, AFAICT. $\endgroup$
    – user95
    Commented Jul 7, 2012 at 17:12

How about Mercury and Gallium? Mercury is rather toxic, but you can buy Gallium online and it's a liquid at 85 degrees Fahrenheit - so it might work if you left it in some sunlight when you wanted the Gallium to be liquid.

If you don't plan on opening it often or need to worry about it breaking, I'd try Mercury.

  • 5
    $\begingroup$ Elemental mercury is liposoluble. $\endgroup$
    – F'x
    Commented Jun 14, 2012 at 14:59
  • 2
    $\begingroup$ Interesting - though it would still form layers if you added the Mercury, then water, then oil. Just don't shake it around too much I guess. $\endgroup$
    – Ehryk
    Commented Jun 14, 2012 at 18:46
  • $\begingroup$ see Aesin's link in his comment below the question for adding many layers if you don't shake things. Alternate layers of water and oil do the trick, too, if you pour them slowly and do not shake. $\endgroup$
    – F'x
    Commented Jun 14, 2012 at 18:58
  • $\begingroup$ Would the mercury/gallium not pull out of the oil over time due to their weight, and eventually resettle? Or would the mercury be bound into the oil? If that's the case, you could just add so much mercury that no more can dissolve in the oil. $\endgroup$
    – Ehryk
    Commented Jun 14, 2012 at 22:09

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.