Precipitation (from Belter et al):

Precipitation is commonly carried out either by addition of a solvent or salt; or by manipulation of the solution pH or
temperature, commonly the latter.

Precipitation with a solvent:

The solvent is miscible with the feed solution, but not with the solute (for which it is effectively a nonsolvent), so
that the latter precipitates. At equilibrium, the chemical potential of the j-th species (the solute) in the precipitate
phase and solution phase are equal, where yj is the solute concentration in solution.
j ( ppt ) j ( solution) 0j ( solution) RT ln y j

On addition of the solvent/nonsolvent, the standard state chemical potential is increased, consequently the solution
becomes supersaturated and precipitation results till a new saturated concentration is attained.

For molecules of moderate molecular weight, e.g. antibiotics, the feed solvent is commonly ethanol, acetone or t-
butanol, whereas the nonsolvent is water. For proteins, however, the feed is commonly an aqueous solution at the
isoelectric pH, where its solubility is minimum; and the nonsolvent is a water-miscible solvent like acetone or
ethanol. A water-soluble polymer like polyethylene glycol (PEG) is also frequently used as a nonsolvent, provided
precipitation is not associated with excessive viscosity. Caution should always be exercised before using a solvent as
the risk of protein denaturation is omnipresent.

Temperature induced precipitation:

While cooling can precipitate moderate molecular weights like antibiotics, for proteins temperature increase is used
as a tool to selectively denature and thereby precipitate a protein, particularly when salting out has not been found to
be effective. Understandably, selective denaturation as a tool is inherently risky, as protein properties like enzymatic
activity may be affected significantly.

Denaturation is frequently modeled by first-order decay kinetics:

dC A
kC A
dt
or its equivalent integrated form - an exponential decay:
CA
e k At
C A0

with temperature dependence of the rate constant kA describable by the Arrhenius law:
k A k A0 e EA / RT

Where CA is the dissolved protein concentration at any time t and CA0 is the value at t=0.
EA, the activation energy of denaturation, which can vary significantly from one protein to another, can be
manipulated by a substantial amount by only slight change in temperature. Again, the pre-exponential factor k A0 also
varies from protein to protein. Therefore, the rate constant will vary with temperature and also differ considerably
CA CB
C A0 CB0
from one protein to another. As a consequence, for a solution of two proteins A and B, the ratios and may
remain identical (and very close to 1) at a certain temperature, but may be found to differ considerably if the
temperature is increased sufficiently thereby affording separation by precipitation.