O.S.L. Bruinsma

Calcium phosphate precipitation in a fluidized bed in relation to process conditions: A black box approach

The precipitation features of calcium phosphate in a fluidized bed reactor in the concentration range between 5 and 100 mg P 1−1 were studied, and the conditions for optimum phosphate removal efficiency were established. The supply of calcium ions should be such that a Ca/P molar ratio of 3 at the inlet of the reactor is achieved. If the water to be treated does not contain magnesium or carbonate ions, the supply of base should suffice to promote a conversion of 50–65% of the incoming phosphate to the solid phase. In the presence of carbonate and magnesium ions, the base supply should provide a conversion of 80–95%. Magnesium and carbonate ions did not have a detrimental effect on the phosphate removal efficiency for inlet concentrations of up to 4.8 × 10−3 (Mg/P < 2 mol mol−1) and 1.8 × 10−3 kmol m−3, respectively. The feasibility of a process based on the precipitation of magnesium phosphate instead of calcium phosphate was demonstrated for waters with a low calcium content (Ca/P < 0.8 mol mol−1). Finally a method is presented to select process conditions where co-precipitation of unwanted phases can be avoided.

Influence of mixing on the product quality in precipitation

Precipitation of barium sulfate has been studied in a rectangular flat reactor with jet mixing, both experimentally and with Computational Fluid Dynamics (CFD) modeling. The precipitation kinetics and mass balances were implemented in the CFD software (Fluent® version 4.25). The influence of the hydrodynamics on the quality of the precipitate has been studied by changing the inlet velocity ratio while keeping the residence time and species concentration in the reactor constant. The CFD model was able to predict the influence of the inlet velocity ratio on the area mean particle size, the coefficient of variation of the particle size distribution, and the degree of conversion. Flow visualization experiments show that the simulated flow pattern was in good agreement with the experiments. A non-symmetric flow pattern was found, which was in agreement with the study of Murai e.a., 1989.

Phosphate removal in a fluidized bed—II. Process optimization

The aggregation of fine primarily formed calcium phosphate particles with sand grains in a fluidized bed for phosphate removal was studied experimentally by means of a set-up which isolated aggregation from other processes during calcium phosphate precipitation, as well as through experiments under normal operation of the fluidized bed. The net aggregation process was described by means of a mathematical model which takes into account two competing mechanisms: orthokinetic aggregation and breakage. The net aggregation process was found to account for ∼ 60% of the phosphate removed by the fluidized bed. It was found that the orthokinetic aggregation can be improved by spreading the supersaturation more evenly throughout the reactor, and breakage can be diminished by a low energy dissipation rate in the bed. Optimization of the phosphate removal efficiency was therefore achieved by selecting sand grains of small sizes (0.1–0.3 mm) and a low superficial velocity (7·10−3 m/s), and by spreading the addition of the NaOH solution (reactant) over two dosage points. Under these conditions the phosphate removal efficiency was ∼ 80%.

Phosphate removal in a fluidized bed—I. Identification of physical processes

The main processes concerning the precipitation of calcium phosphate in a fluidized bed were studied. Primary nucleation and molecular growth occur at the bottom of the bed, where the incoming phosphate is mixed with the reactants, producing fine particles which partly aggregate with the grains of the bed and further leave the reactor with the liquid effluent. This aggregation seems to be the most important process for phosphate removal, while molecular growth of the grains seems to be of secondary importance.

Dry-spraying of ascorbic acid or acetaminophen solutions with supercritical carbon dioxide

Carbon dioxide is a very poor solvent for many organic compounds, which makes it a good anti-solvent. When a solution is sprayed into carbon dioxide vapour the anti-solvent reduces the solubility within several tens of milliseconds and the solute precipitates. Two distinct regions can be identified, below and above the mixture critical pressure. Below this critical pressure the yield remains relatively low and the process is not well controlled. Above the critical pressure small crystals are obtained of about 2 μm with a yield of 90%.

Nucleation and growth of fine crystals from supercritical carbon dioxide

Abstract Fine benzoic acid and phenanthrene crystals were produced by expansion of a supercritical carbon dioxide solution through a small capillary nozzle. The produced crystals had a size of 2 to 10 μm. Similar trends for the number average particle size were observed for both compounds. A higher pre-expansion pressure leads to smaller particles at the same entrance mole fraction. At a given pre-expansion pressure, the average particle size first increases at higher entrance mole fractions until a maximal value is reached, and then the particle size starts to lower again. Models are presented for the pressure and temperature profiles in the nozzle, which allow calculation of the supersaturation profiles along the nozzle to be carried out.

The furification process in hydraulic packed-bed wash columns

Abstract Hydraulic packed-bed wash columns are continuous solid-liquid separators, which can be applied as the final step in fractional suspension crystallization processes aimed at ultrapurification of organic compounds. In this paper an experimental study and a mathematical model of the purification process in packed-bed wash columns are presented. The developed model can be used to calculate concentration, temperature and porosity profiles along the column axis and to predict the effect of operating conditions on the product purity.

Three-zone approach for precipitation of barium sulphate

Abstract The three-zone approach, introduced by Gosele and Kind [1] [Chem. Ing. Tech. 63 (1991) 59], has been used to investigate BaSO 4 precipitation. Experimental relations for the growth and nucleation of BaSO 4 are found that show good agreement with those of Angerhofer [2] [Thesis, TU Munchen, 1994] and Van der Leeden [3] [Thesis, TU Delft, 1991]. The parameters studied in the three-zone approach are the feed concentration, circulation time and volume of the zones. Preliminary results show that the model used correctly predicts the influence of the above-mentioned parameters on the mean particle size and the conversion. Experiments conducted at various mixing conditions in the 3 zones show that the influence of the parameters becomes stronger when the mixing in the 3 zones becomes less intensive. This influence cannot be explained by the present model, which assumes that the zones are ideally mixed.

Recrystallization of calcium sulfate in phosphoric acid solutions; batchwise operation

A model is presented for the solvent mediated batchwise recrystallization of hemihydrate into dehydrate. The model takes into account the rates of dissolution of hemihydrate and of growth of gypsum. These rates result from the driving forces (under- and supersaturation) that are influenced by temperature and acid concentrations, and from the rate constants, which are mainly affected by impurities. Secondary nucleation can be neglected with respect to the observed recrystallization rates. The presented model can also be applied more generally.

INFLUENCE OF HYDRODYNAMICS ON PRECIPITATION: A COMPUTATIONAL STUDY

A method is proposed to describe the influence of hydrodynamic conditions or. precipitation processes. It is based on spatially distributed moment equations, on relations for the kinetics of nucleation and growth of the precipitating compound and on the solution of the flow field as calculated by the commercially available software package ‘Fluent’. Local quantities are calculated, such as liquid velocities, local supersaturation and local moments of the particle size distribution. The method was applied to simulate a double-jet continuous reactor for the precipitation of calcium phosphate. The influence of the location of feed points on the conversion, product average size and coefficient of variation of the product size distribution was shown. The conversion varied within the range of 0.57 to 0.83, the average size changed by a factor 1.9 and the coefficient of variation by a factor 1.5. It was shown that only a small fraction of the reactor volume is effectively used for nucleation and growth. The resu...