The azo coupling of I-naphthol with diazotized sulfanilic acid has been studied in detail focusing on the practical use of this reaction as a micromixing test reaction, as developed by Bourne and coworkers. The reaction is a fast, competitive, consecutive reaction whose final product distribution is affected greatly by mixing. Problems that occur in the isolation of the pure-dye products and quantification of the product distribution are addressed. Previously unreported in formation is given about the structure and properties of one of the products as well as the existence of an additional unknown product. The reaction was used to characterize the spatial heterogeneity of micromixing in a 14-L stirred-tank fermenter. Results show large differences in the product distribution dependent on the depth and radial position of the feed pipe in the tank.
introductionIt has long been recognized (Danckwerts, 1958;Zwietering, 1959) that for nonfirst-order reaction kinetics the residence time distribution of a reactor is insufficient information to predict conversion. In such cases, it is necessary to consider the degree and intensity of segregation, defined by Danckwerts, which the molecules experience while undergoing chemical reaction. Ultimately, molecules are mixed by small-scale turbulent eddies whose size and velocity depend on the local rate of energy input. Consideration of mixing at the scale of the smallest turbulent eddies is usually termed micromixing. Micromixing becomes important particularly in the case of multiple simultaneous reactions where it can affect the selectivity of a desired product, Some general requirements for a reaction to display mixing sensitivity have been outlined by Bourne et al. (1977) and David and Villermaux (1987).Mixing effects are also seen in bioreactors, in which microorganisms aregrown (Dunlop and Ye, 1990; Tomaet al., 1991). For the purpose of oxygen transfer, these reactors are often highly agitated, which may lead to the assumption of "perfect" mixing in the design of bioreactors. However, changes in micro-organism productivity as a function of substrate injection Correspondence concerning this article should be addressed to E. H. Dunlop location (Hansford and Humphrey, 1966), stirrer speed (Torna et al., 1991), or grid generated turbulence (Fowler and Dunlop, 1989) suggest that the mixing is often less than perfect. Since micro-organisms are of the same length scale or smaller than the smallest turbulent eddies in a fermenter (Dunlop and Ye, 1990), micromixing concepts may be able to explain these effects. A micromixing test reaction, described below, is being used in the present study to measure micromixing in biochemical reactor systems.Although much work has been done to model the process of micromixing (see, for example, Baldyga and Rohani, 1987;David and Villermaux, 1987;Patterson, 1981), there still exists no way to predict its effects for an arbitrary reactor design and reaction kinetics. Micromixing effects, however, can be measured through the use of reactive tracers...