The theory of sample zone dispersion is well known for most cases of practical interest in flow injection analysis. This paper offers a theoretical analysis which allows for the optimal design of single-line flow systems. For various reactor types, a detailed analysis is provided in terms of physical constants, design parameters and constraints.
It is shown that, within practical constraints and using a pressure drop of less than 1 bar, it is possible to operate flow systems at 100 samples per hour, with a residence time of 100 s and a reagent consumption of 8 rl for each determination. Further miniaturization of flow systems must rely on smaller detector volumes than those currently available, a situation not unlike that in liquid chromatography.
This paper is concerned with the performance of systems for flow injection analysis (f.i.a.). Three types of parameters of the flow system can be distinguished: physical constants which are characteristic of dilute aqueous solutions, like the molecular diffusion coefficient D, and the viscosity; design parameters like the helix diameter in a coiled reactor or the bed diameter in a packed-bed reactor; and performance-limiting parameters (constraints) such as the maximum pressure drop over the reactor, the smallest acceptable tube diameter or the smallest detector volume that is allowed for. Equations for the optimal design of flow injection systems are presented, which show clearly when a constraint parameter actually limits the performance. The values of the constraint parameters are obtained from f .i.a. practice.
Although R&iEka and Hansen [l] have pointed out correctly the differences between f.i.a. and high-performance liquid chromatography (h.p.l.c.), the present treatment is based on optimization studies in h.p.1.c. by Knox et al. [ 2,3] and Guiochon [ 41. Throughout this work, it is assumed that the elution (response) curve of the flow reactor (column) is Gaussian.