Simulation And Optimization Of Crystallization Process Based On Multidimensional Crystal Growth Kinetics Measurement And Morphological Population Balance Model

As a unit operation that produces high value-added products with low energy consumption,crystallization found wide application in pharmaceutical,chemical and fine chemical,material,food,energy and military industries.In defining product quality of crystals,crystal size distribution(CSD)is a key indicator where the size of a crystal is often defined as the volume equivalent diameter of a sphere,and this approach obviously ignores important morphological information,which could have a major impact on the fluidity,stacking density,dissolution rate,stability and other important properties of the crystal products,as well as the efficiency of downstream processing of crystals infiltration,drying,and storage.With the rapid development of science and technology and the increasing demand of product quality improvement,there is no doubt that more strict quality control of crystallization processes is needed.However,it is difficult to meet the product quality requirements by merely obtaining crystal products with satisfactory CSD.In fact,the stringent requirements on crystal morphology and morphology distribution(CShD)of crystal products have set higher standards for the optimization and control of industrial crystallization.Motivated by the above observations,this work was aimed at developing methods for modelling and optimization of CShD in crystallization processes.To achieve the ultimate goal research gaps were identified and filled,and a systematic approach proposed and validated by reference to crystallization of sodium nitrate in aqueous solution.A new device for measurement of face specific growth kinetics was designed and built,with which the face specific growth kinetics of sodium nitrate crystals were measured.Based on this,a new morphological population balance model(MPBM)is established using modeling sodium nitrate as studied compound.In this novel MPB model,the shape of a crystal is quantitatively defined as the distances of all its independent faces to the geometric centre.Furthermore,the concept of multi-dimensional size distributions,which we call shape distribution for the crystal population in the crystallizer,is proposed and set as the optimization target.This represents a major development in optimization of crystallization processes compared with the traditional CSD-targeted practice.With these novel model and definition of target CShD,the crystallization process of sodium nitrate solution was optimized in this thesis.The research work presented here can be summarized as follows:Determination of the thermodynamic properties of sodium nitrate: the morphology of sodium nitrate crystals at different crystallization conditions and their formation was examined and explained.On the bases of solubility data taken from literature,the metastable zone width(MSZW)of sodium nitrate solutions in a 1L crystallizer under different experimental conditions were measured and fitted to different theoretical models and got satisfactory results.These models can also be regarded as approaches for MSZW prediction for sodium nitrate aqueous solution.Determination of the kinetic properties of sodium nitrate: face specific growth rates of three independent crystal faces of sodium nitrate single crystals in the temperature range of 298.15 – 318.15K(with 5K interval)and in the supersaturation range of 0.01 – 0.11 was investigated with the help of the newly developed stereo imaging device for measuring crystal face specific growth kinetics.The growth process was separated into two stages: the initial and the main one.The growth rate at the former stage was generally lager than in the latter stage.The face specific growth rate vs.relative supersaturation(G-?)isotherm can be divided into three parts: a scattering part at medium supersaturations,and a head and tail parts where growth rates increased with supersaturation.The presence of three distinct parts in G-? isotherm was explained by a theory of co-existence and competition of different growth mechanisms,which was validated by experimental data and analogous literature information.All the face specific growth rates data were correlated to solution supersaturation and temperature according to this theory.The presence of growth rate dispersion(GRD)was also explained by this theory.The device for measurement of face specific growth kinetics of crystals growing in a solution consists of three major components,a stereovision imaging system that captures in real-time the size change in each face direction of the crystal,a ATR-FTIR probe measuring solution concentration and supersaturation,and a small purpose-designed crystallizer that allows a crystal fixed in the solution and adequate solution mixing so the concentration measured can reflect the solution concentration near the surfaces of the crystal.This system can be applied to the measurement of face specific growth kinetics of different crystals in various solution system;in this thesis it was applied to study the crystallization of sodium nitrate(NaNO3)and potassium dihydrogen phosphate(KH2PO4,KDP).In order to measure solute concentration during the experiments for the estimation of the face specific growth kinetics,a prediction model for sodium nitrate solutions was established by using Attenuated Total Reflection Fourier Transform Infrared Spectroscopy(ATR-FTIR)and partial least squares(PLS)regression.Additionally,in this thesis seeded cooling crystallization of sodium nitrate in aqueous solutions was simulated and optimized with the developed MPBM,using the measured face specific growth kinetics.The influences of several important parameters such as seed loading,cooling rates,seed crystal size and initial supersaturation on crystallization process and CShD of final crystal products were examined with the MPBM.The proposed morphological population balance model for sodium nitrate crystallization was also applied to investigate and control the growth of individual faces with the aim of obtaining a desired CShD.Process optimization techniques were tested with the MPB model for optimizing the product CShD.Optimal temperature and supersaturation profiles leading to the desired CShD were estimated.Since imposing the optimal temperature or supersaturation trajectory can be easily implemented in a crystallization process by manipulating the temperature in the crystallizer jacket,the proposed methodology provides a feasible closed-loop mechanism for sodium nitrate crystal shape tailoring and control.