The Interface Between Mathematical Chaos and Experimental Chemistry

Abstract

If the mechanism of a chemical reaction is assumed, presently available computer technology is adequate to model the dynamic behavior of a real system by means of deterministic equations. Our own computations have not yet discovered significant ranges of parameter space within which such behavior is chaotic. If experimental measurements exhibit chaotic responses under carefully controlled conditions, those responses need not reflect intrinsic chaos in the chemical mechanism unless sensitivity to uncontrolled fluctuations can be ruled out. The measuring apparatus itself is usually sensitive enough that it need not be a cause of concern, and the quantum-mechanical uncertainty principle is not important in macroscopic systems. However, fluctuations in experimentally controlled parameters and in statistical-mechanical populations of species may both be amplified many orders of magnitude if they occur when the system is making a transition between manifolds. It appears very difficult to rule out the possibility that such fluctuations are responsible for any specific example of experimentally observed chaotic behavior.