Colloidal self-assembly

The spontaneous assembly of simple components into well-defined structures is central to a number of chemical and biological systems: sugar molecules crystallize to form rock candy, dozens of proteins assemble to form the capsid shell of many viruses, and long, linear sequences of nucleotides or amino acids fold into specific three-dimensional arrangements, which are key to how they function. Are there general principles shared between these seemingly disparate settings that enable them to order, assemble, or fold in a robust and efficient manner? Our group seeks to answer this and related questions using a bottom-up approach, in which we apply modern optical techniques to characterize the behavior of model experimental systems combining colloids and DNA nanostructures.

One of the experimental systems that we use are DNA-coated colloids: micrometer-scale spherical particles, which are coated with tens of thousands of short, single-stranded molecules of DNA. Here the colloidal particles play the role of the building blocks. They are large enough that we can see them with an optical microscope, yet small enough that they can move around spontaneously and interact with one another via thermal motion. The DNA encodes the interactions between the particles and “tells” them how to self-assemble. For instance, particles coated with complementary sequences—in which the As pair with the Ts and Cs pair with the Gs—will be linked together, whereas particles bearing non-complementary sequences will not interact. In this way we can program complex sets of interactions between a variety of particle types and then study the behavior and structures that emerge.

We have three primary directions within this research area:

  • Quantifying dynamics of assembly. What are the pathways by which a disordered fluid of colloids spontaneously assembles into an ordered crystal? And how do these pathways depend on the details of the structure that’s formed?

  • Creating new interactions. How many unique components would we need to build whatever structure we want? And how could we program the interactions between these components to favor the formation of our chosen target?

  • Building new building blocks. What new kinds of structures could we assemble if he had more complex building blocks, like anisotropic colloids with specific and valence-limited interactions? How do we build these kinds of particles in practice?

Relevant publications:

Using DNA to program the self-assembly of colloidal nanoparticles and microparticles, Nature Reviews Materials, 1 (2016) 16008
Programming colloidal phase transitions with DNA strand displacement, Science, 347 (2015) 639-642


Quantifying dynamic pathways to assembly

Self-assembly describes a dynamic process by which a disordered set of components assembles into an ordered structure. But how does the ordered phase first emerge from the disordered phase? And how does it grow to become the final structure?

To answer these questions, we use a technique called “microfluidics” to create hundreds of identical water droplets, which are each filled with DNA-coated colloids. When observe each drop as its colloidal suspension transitions from the initial disordered phase to the final ordered one, gathering data for hundreds of droplets simultaneously using an optical microscope and a digital video camera. We then extract quantitative information from the videos using image analysis, which can be parsed in various ways. For instance, we can quantify rates of nucleation by doing a statistical analysis of the fraction of droplets that crystallize over time and we can measure rates of growth by monitoring the size of the crystal in each droplet over time. A key feature of this approach is that it does not require single-particle tracking, which enables us to explore the behavior of systems with components that are smaller than the diffraction limit.

We also perform experiments investigating self-assembly of the same colloids confined to two-dimensions, which enables us to track every individual particle throughout the entire assembly process. Here we can answer new types of questions: What is the size of the critical nucleus? Is the ordered phase that nucleates first the same as the final assembled structure? Or does nucleation and growth proceed through a series of intermediates?

These experiments are complemented by a suite of computer simulations, which we perform together in collaboration with Mike Hagan’s group. Here we use a combination of Monte Carlo and Molecular Dynamics simulations to

Programming new colloidal interactions

We are also developing new ways to prescribe interactions between colloidal particles. Traditional approaches to controlling colloidal interactions using DNA rely on direct binding of DNA molecules grafted to the surface of colloidal particles. However, these approaches are typically limited to prescribing a small number of interactions between just a few species.

With the ultimate goal of prescribing assembly of complex, user-prescribed structures, we are developing new approaches to programming interactions between colloids, in which single-stranded DNA oligomers dissolved in solution “link” colloidal particles together. In this new paradigm, each particle type is coated in a unique DNA sequence. These sequences encode the addresses of the building blocks. Then we design a set of linker DNA sequences, which have two domains complementary to the two particles being assembled. Our exploration into this alternative method indicates that linker-mediated binding of ssDNA-coated colloids can prescribe multiple orthogonal binding interactions to the same particle, increasing the addressability needed for asymmetric structures. Furthermore, by putting the binding instructions into the surrounding solution, we can tune the strengths of these interactions and the same grafted colloids can be used in a variety of structures without having to be re-synthesized.

Relevant publications:

Linker-mediated phase behavior of DNA-coated colloids, arXiv, (2019) 1902.08883
Using DNA strand displacement to control interactions in DNA-grafted colloids, Soft Matter, 14 (2018) 969-984
Programming colloidal phase transitions with DNA strand displacement, Science, 347 (2015) 639-642

Building new model colloids

Nature, in its mastery of self-assembly, is able to limit the size of a given structure: amoebas don’t make it past a centimeter in diameter, and humans don’t keep getting taller and taller into old age. At a wide range of length scales, life seems to know when to say “Enough!” In collaboration with theorists at University of Massachusetts Amherst, we aim to create self-assembling structures of limited size using geometrical frustration. Geometric frustration occurs when there is attraction between the components, but the physical shape of these components means that they have to deform to fit together. Each piece added to the assembling structure must deform more and more until the energy cost of deformation becomes too high, limiting the size of the assembled structure.