Is there a way to generalize the solubility of any organic compound based on their functional groups and the length of their carbon chains? I know that beyond 3 carbons, alcohols are insoluble, and beyond 4 carbons, carboxylic acids are insoluble. But what if the alcohol had 2 hydroxyl groups? What if the carboxylic acid had 2 carboxyl groups? How many carbons would they each need for the compound to become insoluble in water again? Are there any quantitative guidelines for determining this?

And what about the solubility of esters and haloalkanes? How many carbons does an ester and haloalkane need to have for them to be insoluble? Again, what if there are multiple ester or halide groups?

Lastly, how do you determine solubility when you’ve got a mixture of functional groups? Take aspirin for example, whose structure contains a benzene ring of 6 carbons, a carboxyl group attached to one corner, an ester group attached to another corner, and one methyl group attached to the ester. How would you systematically determine if this compound was soluble in water or not?

enter image description here

(Image credit: http://www.rxlist.com/images/rxlist/aggrenox2.gif)

  • 1
    $\begingroup$ It's much too broad. And there's not much point - in practice experimental data are used. More polar groups can offset non-polar, but they can interact with each other and steric effects can be important. $\endgroup$
    – Mithoron
    Commented Oct 8, 2015 at 15:49
  • $\begingroup$ There are some situations where you need a way to estimate a chemical's water solubility--for example when EPA reviews a new chemical premanufacture notice. The submitter may not provide data on solubility and the substance itself is not available. So you use software tools to develop an estimate. Some of the tools use a group contribution approach as you suggest in the question. $\endgroup$
    – iad22agp
    Commented Oct 8, 2015 at 17:52

1 Answer 1


To determine the solubility of some compound, a chemist would systematically throw it into water and see what happens. You may draw some correlations like "plus six for two hydroxyls, minus five for five carbons, should be soluble"; they are somewhat useful, but mostly qualitatively.

There is a whole hierarchy of computational methods for prediction of solubility, but these are highly complicated.

BTW, haloalkanes are almost as insoluble in water as alkanes themselves.


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