Chemical Computing

Chemical computing is the use of chemical reactions — typically in aqueous solution — to perform information-processing operations. Unlike conventional silicon-based computation, where information is encoded in voltage states and manipulated by Boolean logic gates, chemical computing encodes information in the concentrations of molecular species and manipulates it through coupled reaction kinetics. A concentration above a threshold represents a logical 1; below it, a 0. The collision of travelling waves in a BZ medium implements AND gates; the phase of an oscillating reaction implements NOT. The substrate is slow, noisy, and irreversible, but it offers properties silicon cannot match: massive parallelism, fault tolerance through chemical diffusion, and operation in environments hostile to electronics.

The theoretical foundations of chemical computing were laid by chemical reaction network theory, which classifies reaction networks by their dynamical capacity — whether they can compute, oscillate, or reach stable equilibrium. Recent work has moved from proof-of-concept to functional systems: DNA strand-displacement circuits that implement neural networks, enzyme cascades that solve optimization problems, and synthetic gene circuits that count events inside living cells. The boundary between chemical computing and synthetic biology is increasingly arbitrary; both use molecular kinetics as a programming substrate.

The deeper significance of chemical computing is epistemological. It demonstrates that computation is not a property of engineered artifacts but a natural behavior of sufficiently complex chemical dynamics. A Turing machine is a special case; a reaction-diffusion system is the general form.