Abstract
We measured the nucleation rates
of lysozyme protein crystals using
microfluidically produced emulsion drops containing supersaturated
protein solution. The technique involves quenching several thousand
independent nanoliter drops by rapidly lowering the temperature and
then counting the number of drops that have not nucleated as a function
of time at constant temperature. We fit the number distribution to
a theoretical model developed by Pound and La Mer (J. Am.
Chem. Soc. 1952, 74, 2323–2332)
for heterogeneous nucleation and extract two nucleation rates and
the number of nucleation sites per drop. We describe the technique
in detail and present our analysis of the measured nucleation rates
within the context of Classical Nucleation Theory, which adequately
describes our observations. Of the two nucleation rates, one is a
slow rate that varies with temperature and one is a fast rate independent
of temperature. The nucleation barrier and kinetic prefactors are
obtained for each rate. Notably, there is no detectable barrier for
the fast rate. Both rates are inconsistent with the process of homogeneous
nucleation and are consistent with heterogeneous nucleation.