Chlorite-iodide-malonic acid reaction Turing patterns

De Kepper, P., V. Castets, E. Dulos J. Boissonnade. 1991. Turing-type chemical patterns in the chlorite-iodide-malonic acid reaction. Physica 49D 161-69. 

Another set of pattern formation phenomena involve stationary, or Turing patterns (77), which arise in systems where an inhibitor species diffuses much more rapidly than an activator species. These patterns, which are often invoked as a mechanism for biological pattern formation, were first found experimentally in the chlorite-iodide-malonic acid reaction (72). Examples of typical spot and stripe patterns appear in Figure 3. Recently, experiments in reverse microemulsions have given rise not only to the waves and patterns described above, but to a variety of novel behaviors, including standing waves and inwardly moving spirals, as well (75). 

Figure 3. Turing patterns in the chlorite-iodide-malonic acid reaction. Dark areas show high concentrations of starch-triiodide complex. Each frame is approximately 1,3 mm square. Images courtesy of Patrick De Kepper,...

Perraud, J, J. Agladze, K. Dulos, E, De Kepper, P. 1992, Stationary Turing Patterns versus Time-Dependent Structures in the Chlorite-Iodide Malonic Acid Reaction, Physica A 188, 1-16. 

The chemical system used for our study is a chlorite-iodide-malonic acid (CIMA) reaction in an acidic (sulfuric acid) aqueous solution. The CIMA reaction exhibits a rich variety of phenomena oscillations in a batch reactor or in a CSTR [26], transient target waves in a closed Petri dish [26], bistability in a CSTR [26, 27], front structures in a Couette reactor [27-30], and Turing patterns in open gel reactors [7-10]. In our two-side-fed reactor. Figure lb, components of the reaction are distributed in the two compartments in such a way that neither compartment is separately reactive. Chlorite is only in compartment A , and malonic acid is only in compartment B thus there are opposing chemical concentration gradients in the direction normal to the plane of the gel. The other chemical species are contained in equal amounts in both reservoirs, except for sulfuric acid, which is more concentrated in compartment B than in compartment A. Note that chlorite and iodide in compartment A are at a low acid concentration they would react rapidly at high acid conditions. 

Only the BZ reaction has played a more central role in the development of nonlinear chemical dynamics than the chlorite-iodide reaction (De Kepper et al., 1990). This latter system displays oscillations, bistability, stirring and mixing effects, and spatial pattern formation. With the addition of malonic acid, it provides the reaction system used in the first experimental demonstration of Turing patterns (Chapter 14). Efforts were made in the late 1980s to model the reaction (Epstein and Kustin, 1985 Citri and Epstein, 1987 Rabai and Beck, 1987), but each of these attempts focused on a different subset of the experimental data, and none was totally successful. Since each model contains a different set of reactions fitted to a different set of data, individual rate constants vary widely among the different models. For example, the rate constant for the reaction between HOCl and HOI has been given as zero (Citri and Epstein, 1987), 2 x10 s (Rabai... 

Historically, it was the CIMA reaction in which Turing patterns were first found. Under the conditions of these experiments, however, our analysis suggests that, after a relatively brief initial period, it is really the CDIMA reaction that governs the formation of the patterns. Even when the input feeds consist of chlorite and iodide, chlorine dioxide and iodine soon build up within the gel and play the role of reactants whose concentrations vary relatively slowly compared with those of C102 and I . We have therefore found it more practical to work with the CDIMA system, using chlorine dioxide and iodine along with malonic acid as the input species, since in this way the relevant parameters can more easily be measured and controlled. Working with the CDIMA system also leads us naturally toward a simpler version of the model described by eqs. (14.22)-( 14.24). 

---