Slurry seeding

As mentioned previously, scale-up of crystallization processes from the laboratory is far from straightforward. Various parameters need to be maintained to be as close to those used in the laboratory as possible in order to reproduce the results from the laboratory. For scale-up, supersolubility, agitation (and its effect on secondary nucleation throughout the vessel), fraction of solids in the slurry, seed number and sizes, contact time between growing crystals and liquid all need to be maintained. 

Seeding with dry powder seed through a vessel head nozzle, while widely practiced in the past, is now limited because of safety and exposure considerations. Slurry seed additions by pump or from seed tanks are preferred and are superior to powder addition for the reasons discussed above. The advantages of retaining the seed in the system, as utilized in continuous operation or in heel recycling are also indicated above. 

Results from the factorial study indicate that the primary input variables affecting filtration resistance were seed type and mixing intensity. When slurry seed was used, insufficient slurry seed led to... 

The process for the final step in a multistep synthesis produces a sodium salt of free phenol. Figure 10-9 shows the flowsheet of the crystallization process. Two streams—stream A containing free phenol dissolved in n-methyl pyrolidone (NMP) and stream B containing ethanolic sodium ethoxide solution—ai e charged simultaneously to a slurry seed bed of sodium-salt of phenol. The addition time is several hours. At the end of each batch, a portion of the batch slurry is wet-milled and used as seed for the next batch. The amount of seed varies from 5% to 20% of the size of the batch. 

The softened seawater is fed with dry or slaked lime (dolime) to a reactor. After precipitation in the reactor, a flocculating agent is added and the slurry is pumped to a thickener where the precipitate settles. The spent seawater overflows the thickener and is returned to the sea. A portion of the thickener underflow is recirculated to the reactor to seed crystal growth and improve settling and filtering characteristics of the precipitate. The remainder of the thickener underflow is pumped to a countercurrent washing system. In this system the slurry is washed with freshwater to remove the soluble salts. The washed slurry is vacuum-filtered to produce a filter cake that contains about 50% Mg(OH)2. Typical dimensions for equipment used in the seawater process may be found in the Hterature (75). 

Precipitation can be continuous or batch. Modem plants use the continuous system where 10 to 14 flat-bottom, internally agitated tanks, approximately 30 m in height and 10—12 m in diameter, are placed in series so that flow of the Hquor—seed slurry moves by gravity through launders connecting the tank tops. 

Cla.ssifica.tion. Slurry leaving precipitation is classified into a coarse and one or more fine fractions, usually by elutriation in hydroclassifiers. Cyclones and combinations of hydroclassifiers and cyclones are gaining popularity. In smelting grade alurnina plants, the coarse fraction, called primary product, is sent to calcination the fine fractions, called secondary and tertiary seed, are recycled to be grown to product size. 

SO that the batch may be dumped into a tank for processing by successive steps, which normally include centrifugation, filtration, and/or drying. In some cases, a small heal of slurry is left in the batch ciys-tallizer to act as seed for the next batch. 

As seen in Chapter 7, the operation of bateh erystallizers is inherently unsteady-state. Transient values oeeur of the major operating variables sueh as slurry density, supersaturation, temperature and mean partiele size. Methods of operational eontrol sueh as by use of seeding and temperature programming were also eonsidered in detail. 

Then, as described in U.S. Patent 3,158,648, the optical isomers may be resolved as follows. 37 g of racemic a-methYl-3,4-dihYdroxYphenylalanine are slurried at 35°C in 100 cc of 1.0 N hydrochloric acid. The excess solids are filtered leaving a saturated solution containing 34.6 g of racemic amino acid of which about 61% is present as the hydrochloride. The solution Is then seeded at 35°C with 7 g of hydrated L-o -methYl-3,4-dihYdroxYphenYlalanine (6.2 g of anhydrous material). The mixture is then cooled to 20°C in 30 minutes and aged one hour at 20°C. The separated material Is isolated by filtration, washed twice with 10 cc of cold water and dried in vacuo. The yield of product is 14.1 g of L-a-methYl-3,4-di-hydroxyphenylalanine in the form of a sesquihydrate of 100% purity as determined by the rotation of the copper complex. 

Four samples of faujasite were synthesized at Si/Al ratios of 2.61, 2.80, 2.97 and 3.03 using published methods from seeded slurries (8-9) and using proprietary methods. One additional sample of Si/Al ratio 2.58 was purchased from Union Carbide. The samples were characterized by X-ray powder diffraction, by surface area measurements, and by wet chemical analysis. The results of these measurements are contained in Table I. 

During the germination trials 400-400 pieces of tomato seeds were applied. As nutrient liquid raw slurry, purified liquid phase diluted with water in ratio of 1 5 and tap water as control has been used. On the sixth day of treatment 34% of the total seed germinated on the filterpaper moistured with raw slurry, 95% on the filterpaper moistured with purified liquid phase, while in the control sample with tap water 68 per cent. 

Figure 15—3 lays out the four-step process, starting with germinating the seed from which everything sprouts. Triethyl aluminum is created from aluminum, hydrogen, and ethylene in step one, which itself has several parts. Powdered aluminum in a toluene slurry is fir-st converted to diethyl aluminum hydride, HA1(C2H5)2, at 212—300 F and 1500 psi. This product is then fed to a tubular reactor with ethylene at 212 F and 300 psi to produce triethyl aluminum. Yields are about 90%. 

The seed crystals of D-fructose were obtained by ball milling crystals produced by spontaneous nucleatlon from an aqueous ethanolIc solution of fructose and allowing them to stand at room temperature in slightly supersaturated ethanolIc solution until the desired crystal size (20-40 microns) was achieved. They were then stored at 30 C in saturated, anhydrous ethanol to prevent further crystal growth. A portion of this slurry was added to the crystallizer as the seeds. 

When the crystal purity is plotted against the total crystal mass in the slurry calculated from the mass balances, the purity decrease seems to start at some constant value of the crystal mass as shown in Figure 5. As mentioned earlier in the text, there are possibilities of existence of the D-enantiomer as small particles on the surface of the seed crystals. If we assume that the breeding of the D-enantiomer starts only when that enantiomer has grown to a certain size, the amount of the L-enantiomer crystals must have also increased to a certain value, the latter being proportional to the former. The crystallization kinetics of the both enantiomers are believed to be the same, the relative amounts of crystals of the both enantiomers must therefore be constant before nucleation of the D-enantiomer starts. 

Figure 2. Change in average crystal size with time for ice ciystal seeds grown in 6% lactose under semi-batch conditions with removal of crystal slurry.

Once temperature equilibrium was established, 600 g of alumina seeds were added to the crystallizer to make up a slurry of solids density 200 g/1. Samples for size distribution analysis and solution concentration determinations were t2dcen periodically for some 24 hours. 

---