Supersaturations, seeded runs

Seeded runs Table I at several initial supersaturations ... 

A complete phase diagram of the system caffeine, glutaric add and acetonitrile was constructed from 10 to 35 °C (see Figure 9.8(a)). Batch runs were carried out in a thermostatted stirred vessel. A saturated solution with respect to the caffeine/glutaric acid co-crystal solid phase was prepared at 35 °C. The solution was cooled down to 34 °C to be slightly supersaturated. Seeding with cocrystals was then performed in order to avoid uncontrolled primary nucleation. 

The possibility that a separate BaC03 phase precipitated in the speed runs (at levels undetectable by x-ray diffraction) also must be noted. Since calcite is approximately three times as soluble as witherite (BaC03), the latter can precipitate from a solution at equilibrium with calcite only if the molar ratio of Ba2 to Ca2 exceeds 1 3. Starting compositions for the speed runs had Ba2 /Ca2 ratios from 3.6 x 10-if to 3.9 x 10 3 and final compositions from 2.6 x 10 3 to 2.3 x 1(T2. To exceed the solubility of witherite during these runs, in which the rapid decarboxylation supersaturates the solution relative to calcite, minimum supersaturations ranging from 14 to 130 with respect to calcite would have been required. Such supersaturations would appear difficult to achieve by decarboxylation, especially in the presence of calcite seed. Viewed in another fashion, the C032 saturation concentration of 0.11 g calcite per 200 ml solution at the start of a run is 9.1 x 10 7M. After decarboxylation to atmospheric levels of C02 and precipitation of calcite... 

After the addition of the seed crystals the relative solution concentration was periodically measured. In Figure 3 the changes of L-threonine concentration in the solution were compared for four experimental runs with different initial supersaturations. Here the relative concentration is defined as the ratio of L-threonine to the total enantiomers. For the case of lower initial supersaturations... 

Effect of Temperature. To study the effect of temperature, the nucleation runs were conducted by keeping all the other variables constant i.e. seed density = 200 g/l, impeller speed at 350 rpm and solution concentration adjusted such that the initial absolute supersaturation (AG C ) r 0.9 at the temperatures studied. The results are shown in Figures 5(a) and 5(b). 

It may happen that the hot solution remains supersaturated, and crystallization does not take place, until the solution is cooled. The checkers obtained a poor yield when this occurred. They, therefore, seeded other runs during the boiling process in order to bring about crystallization. Posner1 had to concentrate the solution to about half its volume in order to bring about crystallization. 

For an isothermal run, the growth constant K may be evaluated by determining the initial and final weights of the seeds, the number of crystals, and the variation of supersaturation with time. Values of the growth constant obtained at several different temperatures may be used with Eq. (26) to predict nonisothermal operation. Palermo s work is in agreement with McCabe s (Ml) earlier work, for it is in essence an analysis limited to a single crystal size, rather than a distribution of sizes. 

These interacting process parameters are illustrated in Fig. 5-3, where supersaturation is plotted against an average solution concentration that would be experienced during addition over the indicated time interval (each point represents an average—one point per run, not a sequence of points in one mn). The amount of seed is shown as a parameter in allowing an increased addition rate. This concept is also valid for antisoivent addition time and reactive reagent addition time. 

As with any crystallization process, reactive crystallization will, in general, produce fine particles unless the entire operation is run within the metastable region. This condition can be realized by provision of heavy seeding and by slow addition to control supersaturation at a low level. Adequate mixing is necessai-y, but shear damage must be avoided by selection of the correct impeller speed and type. 

A given supersaturation thus corresponds to a crystal seed of length in labile equilibrium. Aggregation of molecules to form clusters (ca. 20-100 molecules) with the critical crystallite size is only possible in the case of a local deviation in the supersaturation of the solution. Figure 2.3.5-3 shows a schematic plot of the supersolubility curves, which lie above the saturation curve and run approximately parallel to it. They indicate to which degree of supersaturation spontaneous crystallization is not observed in a technically acceptable time (20-30 min).This experimentally determined time is an important quantity in the design of industrial crystallizers. 

For the seeded MSMPR experiment, the rate of de-supersaturation was also veiy fast for the first 15 minutes of the ciystallisation run, as can be seen in Figure 5. The rapid decrease in Ca also indicated that there was no induction time. After the first residence time (15 minutes), the reduction was gradually slow and eventually the curve levelled off. This is obviously due to nucleation and growth of calcium sulphate crystals at the expense of Ca. Finally, the concentration of Ca in the crystallising solution was steady after seven residence times, i.e. after 7x15 minutes or after 105 minutes. 

---