Refine sodium bicarbonate (NaHCO3) is produced industrially in bubble columns according to the reaction: Na2CO3 + CO2 + H2O « 2 NaHCO3
The reaction involves three phases and thus transfers between gas and liquid, and liquid and solid. The supersaturation is created by the transfer of CO2 in the liquid. The presence of crystals in suspension in the liquid phase modifies the bubble behavior and therefore the gas-liquid transfer. The crystal size distribution, the purity and the morphology of the product depend on a wide variety of phenomena and conditions. To elucidate the key parameters, we are developing an in-depth study of the mechanisms controlling the nucleation and growth (liquid-solid transfer) in conditions close to industrial ones.
To be able to study separately the influence of supersaturation, crystallization temperature, sodium carbonate concentration, magma density, agitation and impurities on nucleation and growth, we carried out experiments in a simplified set-up, a 4L - MSMPR (mixed solution mixed product removal) reactor. This type of reactor was chosen for two reasons: first the lowest part of an industrial bubble column can be considered as a MSMPR reactor, second models are available for extracting growth rate and nucleation rate for such a reactor.
The influence of supersaturation level, crystallization temperature, and sodium carbonate concentration on the nucleation and growth kinetics will be presented here.
In all experiments, the supersaturation is obtained by cooling a solution saturated at higher temperature, the residence time is kept constant at 26 min and the product is collected when a steady-state is reached. The sodium carbonate is introduced in the NaHCO3 solution at a continuous rate just before it is fed in the reactor.
The composition of the solution is followed during the crystallization process by titration and densimetry (method developed by Zhu). ICP analyses give us the purity of the solution and crystals. The crystal size distribution (CSD) is obtained by sieving, and the crystal morphology is characterized by scanning electron microscopy (SEM) and optical microscopy.
A new method of data analysis using a fitting square method was developed specially to decrease the error on the data analysis. The growth rate seems to be affected mainly by the crystallization temperature and is barely modified by the two other parameters in the range we investigated in this study. At the contrary, the nucleation rate is strongly affected by the three parameters. Furthermore, an ANOVA analysis of the nucleation rate results shows that parameters are coupled.