Or can there be a homogeneous mixture that is not a solution?

  • 6
    $\begingroup$ I think this largely rests on what one means by "homogeneity". It's a concept which changes depending on the scale. The entire Universe is homogeneous on the scales of billions of light years, but below that it is not. A steel plate might seem homogeneous at the metre scale, but view it with an electron microscope and all sorts of aggregates and domains become visible. The limits of homogeneity are somewhat loosely defined. $\endgroup$ – Nicolau Saker Neto Jun 23 '15 at 17:08

Homogeneous and heterogeneous classifications are scale-dependent. What we mean when we say a mixture is homogeneous is that there are no visible phase differences on the scale of interest.

For example, milk (when homogenized) and lotions are both examples of colloidal suspensions. There are no phase differences at the human scale, but neither one is a solution. On a more microscopic scale, you could call the suspended micelles phase boundaries, and in that case might not call the mixture homogeneous.

"Solution" always implies mixing on the molecular level (single-phase).

To summarize:

Solutions are always homogeneous mixtures, but homogeneous mixtures are not always solutions.


According to IUPAC gold book:

A solution is a liquid or solid phase containing more than one substance...

So, the concept "solution" is reserved to solid and liquid phases. In gas phase: Air is an example of homogeneous mixture of nitrogen, oxygen, carbon dioxide and other gases present in air. It's obvious that air is not a solution.


For the liquid phase, I think it is probably true that all homogeneous mixtures are solutions.

But the situation is more complicated for solid phases. In a diamond crystal, if the $\ce{^{13}C}$ atoms are homogeneously distributed in the crystal, could you really say that the diamond is a "solution" of $\ce{^{13}C}$ in a $\ce{^{12}C}$ "solvent"? I don't think so. $n$-doped or $p$-doped silicon is another example. In contrast, other homogeneous solids, like brass, are correctly regarded as solid solutions.

No one usually regards gas phases as "solutions", so mixtures of gases are probably not solutions.

  • $\begingroup$ Your comment suggests another part of the definition of a solution: the components of a solution cannot be covalently bonded together. $\endgroup$ – iad22agp Jun 23 '15 at 17:33
  • $\begingroup$ @iad22agp - But now you have ruled out things like Si-Ge that forms a continuous solid solution in the diamond cubic structure. It is homogeneous, and is definitely a solution (and a near-ideal solution as well). So I do not believe that the type of bonding has anything to do with the question. $\endgroup$ – Jon Custer Jun 23 '15 at 18:12
  • $\begingroup$ Well, metal bonding is not exactly covalent. I think the key here may be that, when liquefied, a true solution's components able to freely diffuse within the bulk. $\endgroup$ – iad22agp Jun 23 '15 at 19:00
  • 1
    $\begingroup$ Si-Ge is not metallic - it is a covalently bonded semiconductor. And, diffusion in the liquid may be quite different from diffusion in the solid, since the bonding environment may be very different. Again, Si is a metallic liquid, but not a metallic solid. $\endgroup$ – Jon Custer Jun 23 '15 at 19:30

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