Solubility from Molecular Structure
Date: Winter 2009-2010
Hi... I am Karina and I am wondering how might you know
whether a substance would dissolve in water without doing an actual
experiment? I know that it has to be soluble and polar but how do
you know if they are soluble and polar by just looking at them?
If you know the name of the substance (or its constituents) you can look
up the formula, structure and solubilities in a current edition of the
"Handbook of Chemistry and Physics." If you do not have access to a handbook
in your class room, check out the library. Otherwise, Google the substance.
We can get a reasonable estimate on the polarity of a molecule by
using what we know about likely shapes of molecules. Using such
estimates and comparing these estimates to actual solubility data
from experiments, we find that there is a trend with regard to the
solubility of substances into water. We find that the more polar a
molecule is, the more likely it would be soluble in water.
For example, a molecule like NH3 is known to have N in the middle
with 3 H's attached to it. We could arrange the H's so that they
form a triangle around the N, but we also know that N has what is
known as a "lone-pair" that is also attached, like the H, to the N.
So then the triangle of H's have to bend downward in order for there
to be space to put the lone-pair. So the shape we get is like a
pyramid with the N at the top and the 3 H's forming the base of the
pyramid. This process is a combination of knowing Lewis Structures
and electron domain (or VSEPR) rules.
At this point we apply what we know about bond polarity. We know
that N is more electronegative than H, that is, if N were attached
to H the N would pull electrons toward itself away from the H. So
NH3 ends up with a lot more negative electrons near the N and the
H's would be more positive because of the lack of electrons. This
process is known as bond polarity and molecular polarity analysis.
Using all of these then, we know that NH3 is polar. - and we can
expect that it would be soluble in an equally polar solvent like water.
Greg (Roberto Gregorius)
If you are looking at a purely covalent molecule then you should
consider two main factors: the molecular weight of the compound and
the number of hydrogen bonding sites it has. These are the
principles used by the pharmaceutical industry to determine the
likelihood of a potential drug being soluble in the blood. Very
large molecules tend to be insoluble due to strong intermolecular
forces between solute molecules. Large numbers of H-bonding sites
tend to increase the intermolecular forces between the solute
particles and the solvent, so making dissolving more enthalpically
favored. Most small molecular compounds with any sort of polarity or
H-bonding capability tend to be soluble in water, to some extent.
A couple of other points to bear in mind:
-simply having an overall molecular dipole (being "polar") will not
always give an appreciable solubility in water, for example
dichloromethane CH2Cl2 ("polar") has a low solubility in water
despite its polarity. H-bonding is a much stronger intermolecular
force and has a larger affect on solubility than dipole-dipole interactions.
-all dissolving is a balancing act between the strength of the
forces that must be broken (solute-solute and solvent-solvent) vs
those made (solute-solvent) along with the change in entropy of the
dissolving process at a certain temperature. This tricky balancing
act makes it easier to explain solubility that you observe, rather
than predict solubility from a molecular structure!!
Predicting solubility based solely on molecular structure is tricky. First,
you need to define "solubility". Do you mean soluble at all concentrations?
Or what other criterion. Common table salt dissolves in water but not at all
concentrations. On the other hand, ethanol is miscible with water in all
proportions. Predicting solubility is a tricky business.
Click here to return to the Chemistry Archives
Update: June 2012