Everyone knows how to assemble things: Just grasp the parts and put them together. Self-assembly, though, doesn’t work at all like this, and as a consequence, it presents special challenges. Despite its inherent difficulties and limitations, self-assembly is the leading means for implementing atomically precise nanotechnologies today, and I expect it will lead for years to come. Self-assembly is a powerful method, and powerful enough to provide a path to nanotechnologies that are yet more powerful. Improving methods for making complex structures by self-assembly is an enormously important area of research.
I’ve already discussed the near-term promise of nanotechnologies based on composite nanosystems that use biomolecules to organize non-biological components. Here, I’d like to outline some fundamental principles will shape the role of self-assembly as nanotechnology advances. (I will say more soon on E-drexler.com.)
Advantages of self-assembly
Self-assembly has a fundamental advantage over mechanically directed assembly: It requires no machinery to move and orient components, letting random, Brownian motion do the job instead. Selective binding between uniquely matching surfaces compensates for the randomness of the motions that bring components together.
Molecular synthesis methods and self-assembly can be used to produce atomically precise nanosystems by the billions, and even by the ton, thereby establishing a technology base with wide-ranging applications that can drive development forward.
The architecture of biomolecular fabrication is based on the use of programmable machines to produce the complex parts necessary for self-assembly of complex systems. The same fundamental architecture can be extended to use artificial biomolecular machines (and then non-biomolecular machines), resulting in products made of better and more diverse engineering materials.
Disadvantages of self-assembly
The most fundamental disadvantage of pure self-assembly is that for every product, the structure of the parts must encode the structure of the whole. This requires that components be more complex, which tends to make design and fabrication more difficult. Another consequence is that a self-assembled product will be partitioned by complex internal interfaces that have no operational function. Unless they are strengthened after assembly, these interfaces will weak. These are major constraints.
Mechanically directed assembly avoids these constraints. Because components need not encode the structure of a product, they can be simple and standardized, and they can be chosen for their functional properties with less concern for how they are put together. This will enable more straightforward design and fabrication, but one must make the necessary machinery — and I expect that this will be accomplished by means of self-assembly.
I’m working on a page for E-drexler.com that will discuss this topic in more depth. In connection with this, I will describe a spectrum of fabrication methods that spans the range from pure self-assembly, through very soft machines, and onward to rigid structures, small building blocks, and advanced mechanosynthesis. More about this later.