Research activities

During the past six years, we made significant progress in the numerical
simulation of crystal nucleation. In particular, we developed a technique
to compute the absolute rate of crystal nucleation of 'hard' colloids.
This calculation is significant because it is the first example of a truly
parameter-free prediction of a nucleation rate under realistic conditions
of supersturation. We subsequently extended the method to other
systems, such as simple ionic crystals (NaCl). At the moment, there is
still a great scarcity of reliable experimental data for comparison.

We performed extensive studies on the competition between
crystallization and liquid-liquid demixing in polymer systems. Our
simulations (of a lattice model) showed that the kinetics of crystallization
is strongly influenced by the proximity of the liquid-liquid spinodal.
Moreover, we studied the effect of polymer pre-alignment of the resulting
crystal morphology. To our surprise, we found that a single aligned
polymer is enough to induce the formation of the so-called 'shish-kebab'
morphology observed in many experiments.

The easy accessibility of 'made-to-order' DNA sequences is changing
materials science. We performed a theoretical analysis of the phase
behaviour of colloidal particles coated with single-stranded DNA that can
be linked by complementary single-stranded DNA molecules. Our analysis
showed that the 'freezing' that is observed in experiments is, in fact,
a liquid-liquid demixing followed by structural arrest. This work has
resulted in one patent.

As microfluidic devices become nanofluidic devices, the problem of fluid
transport on a chip becomes acute: the viscous drag in narrow channels
is enormous and hence pumping may not be the best way to transport
liquids. An alternative is to use electric fields to drive the flow of
electrolyte solutions. We developed a Lattice-Boltzmann method to model
such flows for arbitrary geometries. With this method, we are able to
predict electrokinetic phenomena under experimentally relevant
conditions.

Even though computing power continues to grow according to Moore's
law, there are many problems that cannot be addressed without dramatic
improvements in simulation algorithms. We proposed a method that
allows us to change the strategy of Monte Carlo simulations, in such a
way that information about rejected trial moves is not discarded. This
scheme is particularly powerful in parallel simulations (where it can lead
to exponential speed-up). We have applied this new method with success
to the calculation of the folding pathway of simple model proteins.

Future directions
The focal point of the Computational Physics group's research will shift
towards coarse-grained simulations of organization and transport of
biomolecules and materials containing bio-molecular building blocks.

One key line of research has been and is the study of novel DNA-coated
colloids. During the past year, we obtained funding for a joint
experimental-numerical project to study such materials. In the longer
term, we aim to couple this with the study of active (i.e. energyconsuming)
assembly of complex materials. This research is linked to
parallel activities within the AMOLF nanophotonics programme, and at
the University of Amsterdam, the Free University in Amsterdam and
Leiden University.

We aim to gain a better understanding of the role of charge on the
crystallization of proteins and (nano)colloidal salts. Recent experiments
show that these systems can form surprisingly complex binary crystal
structures, some of which have no simple molecular counterpart. We aim
to study the factors that control the kinetics of the formation of complex
binary structures and to explore the limits of 'spontaneous' self-assembly
in multicomponent systems of charged (nano)colloids.
A third line of research is the study of substrate-induced conformational
changes in simple models for proteins. We hope this will lead to an
understanding of how to 'design' artificial binding sites with ultrahigh
specificity.