Continuous Mixed Suspension Mixed Product Removal Research Articles

In-situ wet milling for particle size control in continuous crystallization: Expanding the attainable region

Continuous mixed-suspension mixed-product removal crystallizer cascades are robust and proven options for continuous manufacturing. However, when applied to a late-stage drug substance the desired particle size could not be achieved due to rapid crystal growth and slow secondary nucleation kinetics. This necessitated a batch terminal milling operation to achieve particle size control (Johnson et al., 2021). To overcome this and to intensify the process, an in-situ high-shear wet mill was integrated into the continuous equipment train to achieve the target physical properties of the drug substance that met the quality and manufacturability criteria needed for the drug product. This article describes the workflow used to optimize the equipment set and operational parameters. Incorporation of a high-shear wet mill operated intermittently in the continuous drug substance process greatly expanded the attainable particle size region, improved product quality, reduced cycle time, and simplified process development and scale-up compared to the process of batch milling after continuous crystallization. Small scale batch milling experiments were used to characterize milling performance at lab and production scales and inform operational space. Several lab-scale continuous runs were then used to define the process operating space and validate the process parameters necessary to meet the particle specifications of the product. Scale-up of the equipment set for in-situ milling was also demonstrated. This approach was tested at lab-scale on representative material operating for 70 h while changing the particle size of the product in real time. This illustrated the potential of intermittent milling for model-free particle size control.

Optimal start-up strategies of a combined cooling and antisolvent multistage continuous crystallization process

A model-based optimization methodology was developed to systematically address all possible optimal start-up scenarios of a multistage combined cooling and antisolvent continuous MSMPR (mixed-suspension, mixed-product removal) crystallizer. The crystallization of aspirin (acetylsalicylic acid, ASA) in ethanol (solvent) and water (antisolvent) was used as a case study. A steady state optimization was firstly conducted to optimize the crystallizer design and operating conditions and set-up the reference trajectories. The discretised profiles of the jacket temperatures, antisolvent flowrates and seeding policies, which includes the seed flowrates, loading and mean size, were used as decision vectors for the dynamic optimization problem. Scenarios with single and combined decision vectors were addressed and analysed in start-up situations involving prefilled or empty vessels. The impact of simplified optimization and operating procedures or refined discretization schemes were also addressed. It was shown that several options may provide comparable performance and the start-up time can be improved by nearly 80% in the case of combined decision vectors involving all possible decision vectors.

Continuous Protein Crystallization in Mixed-Suspension Mixed-Product-Removal Crystallizers

The metastable needlelike form of lysozyme can be obtained from continuous crystallization in mixed-suspension mixed-product-removal (MSMPR) crystallizers for conditions at which the stable tetragonal form is obtained from batch crystallization. A 3D-printed airlift crystallizer and stirred tank crystallizer are used, which both approximate an MSMPR crystallizer well. The influence of residence time, precipitant concentration, and agitation mechanism on the attainable solid-state form is presented. The needlelike crystals appear after nucleation of the tetragonal form. An earlier transition from the tetragonal form to the needlelike form is observed in an airlift crystallizer compared to a conventional stirred crystallizer, which is possibly caused by reduced attrition and the presence of a gas–liquid interface in the airlift crystallizer. Furthermore, a long residence time and high precipitant concentration favor the formation of the needlelike crystals. This work demonstrates how the mode of operation of well-mixed crystallizers affects the attainable solid-state form of a model protein. The results generally show the importance of obtaining a quantitative understanding of the dynamic interplay between stochastic primary nucleation processes and deterministic growth and secondary nucleation processes so that future continuous MSMPR crystallizers can be designed and operated robustly to produce protein crystals of a desired solid-state form.

Struvite nucleation and crystal growth kinetics from cattle liquid manure

Cattle liquid manure (CLM) in future may become excellent raw material for the solid mineral fertilizers production. Phosphates(V) recovery from CLM can be done, e.g. through reaction crystallization of sparingly soluble salts: magnesium and ammonium phosphate(V) hexahydrate MgNH4PO4·6H2O (MAP, struvite) or/and calcium phosphates(V). The original kinetic calculations based on experimental test results concerning continuous reaction crystallization of sparingly soluble phosphates(V) from cattle liquid manure in a continuous laboratory DT MSMPR (Draft Tube, Mixed Suspension Mixed Product Removal) crystallizer with internal circulation of suspension driven by propeller stirrer are presented. Two kinetic models of continuous mass crystallization in MSMPR crystallizer: SIG (Size Independent Growth) and RE SDG (Rojkowski Exponential Size Dependent Growth) were adopted. Advantageous effects of: low pH (8.5), elongated mean residence time of suspension in a crystallizer (3600 s) and magnesium ions excess in relation to phosphate(V) ions concentration (1.2:1) on the process results were observed. Under the considered process conditions linear crystal growth rates of struvite were: G from 4.11 × 10−3 to 1.01 × 10−2 µm s−1 and G∞ from 5.30 × 10−3 to 1.02 × 10−2 µm s−1.

Technoeconomic Optimization of Continuous Crystallization for Three Active Pharmaceutical Ingredients: Cyclosporine, Paracetamol, and Aliskiren

Mixed suspension, mixed product removal (MSMPR) crystallizers are widely implemented for the continuous crystallization of active pharmaceutical ingredients (APIs), allowing enhanced efficiency, flexibility, and product quality compared to currently dominant batch crystallizer designs. Establishing cost-effective continuous crystallization process configurations for societally and economically important APIs is essential to ensure the successful implementation of end-to-end continuous pharmaceutical manufacturing (CPM) campaigns. Process modeling and optimization allow rapid, systematic comparative technoeconomic evaluations. This paper pursues total cost minimization of different crystallizer configurations of three APIs—cyclosporine, paracetamol, and aliskiren hemifumarate—whose continuous MSMPR crystallization has been experimentally demonstrated. Nonlinear optimization for total cost minimization is implemented for one to three crystallizers for different plant API capacities with crystallizer tempera...

Design and optimization of production parameters for boric acid crystals with the crystallization process in an MSMPR crystallizer using FBRM® and PVM® technologies

In crystallization studies, newly developed technologies, such as Focused Beam Reflectance Measurement (FBRM) and Particle Vision and Measurement (PVM) are applied for determining on-line monitoring of a representation of the Chord Length Distribution (CLD) and observe the photographs of crystals respectively; moreover recently they are widely used. Properly installed, the FBRM ensures on-line determination of the CLD, which is statistically associated to the Crystal Size Distribution (CSD). In industrial crystallization, CSD and mean crystal size as well as external habit and internal structure are important characteristics for further use of the crystals. In this paper, the effect of residence time, stirring speed, feeding rate, supersaturation level and the polyelectrolytes such as anionic polyacrylamide (APAM) and non-ionic polyacrylamide (NPAM) on the CLD as well as the shape of boric acid crystals were investigated by using the FBRM G600 and the PVM V819 probes respectively in an MSMPR (Mixed Suspension Mixed Product Removal) crystallizer. The CSD and kinetic data were determined experimentally using continuous MSMPR crystallizer running at steady state. The population density of nuclei, the nucleation rate and the growth rate were determined from the experimental population balance distribution when the steady state was reached.

Comparison of the Method of Classes and the Quadrature of Moment for the Modelling of Neodymium Oxalate Precipitation

Oxalic precipitation is generally used in the nuclear industry to deal with radioactive waste and recover the actinides from a multicomponent solution. To facilitate the development of experimental methods and data acquisitions, actinides are often simulated using lanthanides, gaining experience more easily.The purpose of this article is to compare the results achieved by two methods for solving the population balance during neodymium oxalate precipitation in a continuous MSMPR (Mixed Suspension Mixed Product Removal). The method of classes, also called discretized population balance, used in this study is based on the method of Litster. Whereas, the Quadrature Method of Moment (QMOM) is written in terms of the transport equations of the moments of the number density function. All the integrals are solved through a quadrature approximation thanks to the product-difference algorithm or the Chebyshev algorithm. Primary nucleation, crystal growth and agglomeration are taken into account. Agglomeration phenomena have been found to be represented by a loose agglomerates model. Thermodynamic effects are modeled by activity coefficients which are calculated using the Bromley model. The sizes of particles predicted by the two methods are in good agreement with experimental measurements.

Pharmaceutical crystallisation processes from batch to continuous operation using MSMPR stages: Modelling, design, and control

In pharmaceuticals manufacturing, the conversion of conventional batch crystallisations to continuous mode has the potential for intensified, compact operation and more consistent production via quality-by-design. A pragmatic conversion approach is to utilise existing stirred tank batch crystallisers as continuous mixed-suspension mixed-product removal (MSMPR) stages. In this study, a rigorous and general mathematical model is developed for a pharmaceutical crystallisation process under continuous MSMPR operation. In the proposed changeover from batch to continuous operation, concentration control (C-control), which has been well accepted in batch crystallisation operation, is further extended to facilitate the convenient design of the steady-state operating point of a continuous MSMPR crystalliser; an objective is to ensure that the start-up procedures and on-line control conditions fall within the design-space of the original batch operation. Both single-stage and cascaded two-stage MSMPR crystallisers were investigated and compared to the conventional batch operation. It was observed that despite the production of a smaller number-based mean crystal size, the proposed continuous MSMPR operation achieved higher production capacity with shorter mean residence time and comparable product yield to the batch operation. Lastly, the robustness of C-control strategy against uncertainties in crystallisation kinetics was also demonstrated for the proposed continuous MSMPR operation.

Obtaining Monodisperse Melamine Phosphate Grains by a Continuous Reaction Crystallization Process

A continuous MSMPR (mixed-suspension mixed-product removal) crystallizer was used in the process of obtaining melamine phosphate in a precipitation reaction. The effects of the mean residence time of the suspension in the crystallizer and of the concentration of the melamine suspension in water on the efficiency of the process and the quality of the product obtained were studied. Kinetic parameters of the reaction crystallization process under study were estimated. The grain size distribution of the crystals and the specific surface area of the grains were determined. Morphology was defined by means of a shape factor and coefficients of inhomogeneity and variation. Well-formed, homogeneous crystals of similar specific surface areas and practically identical small pore volumes were obtained. This work shows that, within the studied range of solid-phase concentrations and mean residence times, it is possible to obtain a product of monodisperse grains and predicted grain size by controlling the concentration...

Effect of calcium based additives on the sodium bicarbonate crystallization in a MSMPR reactor

The objective of this work was to study the impact of calcium ions addition on the crystallization of sodium bicarbonate during the reaction of sodium carbonate with carbon dioxide in aqueous solutions. A continuous MSMPR (Mixed Suspension Mixed Product Removal) reactor for the production of sodium bicarbonate was developed, built and tested at pilot scale. This experimental set-up was used to study the impact of process parameters and the effect of calcium based additives on the quality of the NaHCO3 crystals obtained. The nucleation and crystal growth rates were determined from the experimental population density distribution when the steady state was reached. Experiments have firstly shown that the addition of calcium ions greatly reduced the nucleation rate. In contrast, the addition of calcium ions had little impact on crystal growth rate in the temperature range between 70°C and 80°C.

Phosphates (V) recovery from phosphorus mineral fertilizers industry wastewater by continuous struvite reaction crystallization process

Continuous DT MSMPR (Draft Tube Mixed Suspension Mixed Product Removal) crystallizer was provided with typical wastewater from phosphorus mineral fertilizers industry (pH < 4, 0.445 mass % of PO43−, inorganic impurities presence), dissolved substrates (magnesium and ammonium chlorides) and solution alkalising the environment of struvite MgNH4PO4·6H2O reaction crystallization process. Research ran in constant temperature 298 K assuming stoichiometric proportions of substrates or 20% excess of magnesium ions. Influence of pH (8.5–10) and mean residence time (900–3600 s) on product size distribution, its chemical composition, crystals shape, size–homogeneity and process kinetics was identified. Crystals of mean size ca. 25–37 μm and homogeneity CV 70–83% were produced. The largest crystals, of acceptable homogeneity, were produced using 20% excess of magnesium ions, pH 9 and mean residence time 3600 s. Under these conditions nucleation rate did not exceed 9 × 107 1/(s m3) according to SIG (Size Independent Growth) MSMPR kinetic model. Linear crystal growth rate was 4.27 × 10−9 m/s. Excess of magnesium ions influenced struvite reaction crystallization process yield advantageously. Concentration of phosphate(V) ions decreased from 0.445 to 9.2 × 10−4 mass %. This can be regarded as a very good process result. In product crystals, besides main component – struvite, all impurities from wastewater were detected analytically.

A normal vector approach for integrated process and control design with uncertain model parameters and disturbances

In this work the normal vector method is extended to the simultaneous treatment of parametric uncertainty and disturbances. This method ensures that desired dynamic properties hold despite parametric uncertainty by maintaining a minimal distance between the operating point and so-called critical manifolds where the process behavior changes qualitatively. Here, unknown exogeneous disturbances and uncertain model and process parameters are considered simultaneously. To address this simultaneous problem formulation, the augmented systems developed for only parameterized disturbances in previous works have to be modified and extended. A generalized formulation of the robust optimization problem results, which includes normal vector constraints on critical manifolds of steady states and of bounds on the state transient. The numerical methods are further developed to prepare for the treatment of high-dimensional problems. Illustrative case studies considering the design of a continuous mixed-suspension mixed-product removal crystallization process and the Tennessee Eastman process are presented.

Size-dependent nucleation and growth kinetics model for potassium chloride—Application in Qarhan salt lake

The nucleation process and crystallization kinetics of KCl were studied in order to investigate the problems in KCl production in Qarhan salt lake. The correlation between particle size and KCl crystal growth rate was studied in a continuous mixed-suspension mixed-product removal (MSMPR) crystallizer. Classical theory of primary nucleation was used to study the homogeneous and heterogeneous nucleation mechanism of KCl crystallization. Several size-independent growth models and size-dependent growth models, such as Bransom, C-R, MJ2, ASL, and MJ3, were examined by the population balance theory. Results showed that MJ3 model closely fitted the experimental data, the mean relative deviation was 0.94%, the crystal growth rate G=2068{1−exp[−1.081×10−5( L+961.882)]}.

Integrated design and control for robust performance: Application to an MSMPR crystallizer

The interplay of different types of performance constraints in an integrated design and control problem is studied by means of a case study. This integrated problem is based on a recent method for robust process design [M. Mönnigmann, W. Marquardt, Normal vectors on manifolds of critical points for parametric robustness of equilibrium solutions of ODE systems, J. Nonlinear Sci. 12 (2002) 85–112; M. Mönnigmann, W. Marquardt, Steady state process optimization with guaranteed robust stability and feasibility, AIChE J. 49 (12) (2003) 3110–3126; M. Mönnigmann, W. Marquardt, Steady state process optimization with guaranteed robust stability and flexibility: application to HDA reaction section, Ind. Eng. Chem. Res. 44 (2005) 2737–2753; W. Marquardt, M. Mönnigmann, Constructive nonlinear dynamics in process systems engineering, Comput. Chem. Eng. 29 (2005) 1265–1275; M. Mönnigmann, Constructive nonlinear dynamics methods for the design of chemical engineering processes, Ph.D. Thesis, RWTH Aachen University, 2003]. The design is found by means of a steady-state optimization problem accounting for process economics and performance requirements. In particular, the latter are represented by constraints which guarantee a user-specified performance of the design in spite of parametric uncertainties. For the first time, two types of dynamic performance constraints are used simultaneously within the adopted framework. These are constraints on time-domain performance indicators as well as on the asymptotic dynamic process behavior. Furthermore, the effect of uncertainty in both, design and model parameters, is accounted for. A key strength of the suggested framework is the direct quantification of the trade-offs between economics and dynamic performance requirements for a selection of uncertainty scenarios. A series of different integrated design and control problems are formulated and solved for a continuous mixed-suspension mixed-product removal (MSMPR) crystallizer. The process exhibits a complex nonlinear behavior and represents a challenging example. The results of the case study allow an in depth understanding of the interactions of design and control for the underlying process.

Stability and dynamics of isothermal CMSMPR crystallizers

Detailed analysis of stability and bifurcation properties and of the dynamic behaviour of isothermal continuous mixed suspension, mixed product removal (CMSMPR) crystallizers is presented using the moment equation model. It is shown that isothermal CMSMPR crystallizers may exhibit not only limit cycle oscillations but, in the case of magma-dependent nucleation, also saddle point instabilities. Applying the Mikhailov stability criterion, two equivalent sets of necessary and sufficient conditions are derived for checking the stability of steady states. Stability maps are presented in the planes of different pairs of system parameters. It is shown that saddle node and boundary bifurcation, as well as bifurcation at infinity may arise in the case of magma-dependent secondary nucleation. Oscillations of the population density and distribution functions are also presented and analysed, approximating the size distribution by gamma distribution. In dynamic states, the crystal size distribution may differ significantly from the exponential one, while the mean values of oscillations of supersaturation are lower than the corresponding steady state values. A causal loop diagram between the variables reveals that instabilities are created by the interactions of the autoinhibition generated negative feedback and a positive feedback between the four leading moments of crystal size.

Evolution of crystal size distributions in a CMSMPR—the effect of temperature, flow rate, and mixing speed with time

A continuous mixed-suspension, mixed-product removal (CMSMPR) crystallizer with automatic computer control and data acquisition was constructed. Operating this CMSMPR at rpm’s greater than 100 rpm assures well-mixed conditions. The crystallization of NaCl, an impurity in a concentrated CaCl2 solution, was performed by cooling under well-mixed conditions. Atomic absorption spectrometry (AAS) and inductively coupled plasma-atomic emission spectrometry (ICP-AES) have been used to measure solution concentrations. Scanning electron microscopy (SEM) with energy dispersive spectroscopy (EDS), light microscopy and particle size analyzer with advanced laser light scattering have been used to characterize the crystals produced by the CMSMPR crystallizer. Octahedral NaCl crystals are observed. Steady state crystal size distribution (CSD) is observed at nine to ten times the mean residence time for the crystallizer. The CSD has been analyzed to look at the details of nucleation, growth, and aggregation rate as a function of crystallization temperature, rpm, and feed flow rate with the application of a Hounslow’s constant aggregation rate equation.

Distribution of dye into KDP crystals in a continuous MSMPR crystallizer

Colored potassium dihydrogen phosphate (KDP) crystals from a dye (sky blue) solution were produced in a mixed suspension mixed product removal (MSMPR) crystallizer. The crystal size distribution (CSD) of the KDP crystals was correlated by a size-dependent growth rate model. The growth rate increased with crystal size, but it decreased as the residence time increased. Two different concentrations of dye in the solution were examined for MSMPR crystallization of KDP crystals. The colored crystals from the solution with low dye-concentration generated regular-shaped KDP crystals. The dye distribution coefficient into the regular-shaped KDP crystals was less than unity. However, the colored crystals from the solution with high dye-concentration formed the special-shaped KDP crystals having no (1 0 0) sector. The dye distribution coefficient into the special-shaped KDP crystals from the dye solution was larger than unity. The dye distribution coefficient as a function of crystal size was presented, and then the relationship between the aspect ratio of the morphology of KDP crystals and dye distribution coefficient was discussed. As a result, a distribution model of dye impurity in KDP crystals for an MSMPR crystallizer could be first proposed in this work.

Inclusion of mother liquor inside KDP crystals in a continuous MSMPR crystallizer

The crystal size distribution (CSD) of potassium dihydrogen phosphate (KDP) and inclusion ratio of mother liquor have been measured in a mixed suspension mixed product removal (MSMPR) crystallizer. A small amount of sodium dihydrogen phosphate (SDP) was added as an impurity tracer in order to measure the inclusion ratio of mother liquor inside the KDP crystals. The effects of residence time on the CSD and inclusion of mother liquor were investigated. The crystal growth rate and nucleation rate were determined from the population density balance theory by assuming a size-dependent function of the growth rate, because the size-dependent growth rate of KDP crystals was observed under a microscope. MSMPR operation with a shorter residence time produced larger KDP crystals, and showed more size-dependency of the growth rate. The inclusion ratio was depicted as a function of crystal size, growth rate, and agglomeration. The degree of agglomeration was effectively correlated with the inclusion ratio of the KDP crystals that had a large hardness property.

Equation of purity for crystals produced from a cmsmpr crystallizer

Literature data of the amount of liquid contained per crystal w L ( L ) measured for various chemicals are correlated only with the crystal size L as: w L ( L ) L 4 . Using the empirical equation, an equation of purity P for crystals produced from a continuous mixed-suspension, mixed-product-removal (CMSMPR) crystallizer at steady state is successfully derived as: P = 100 \ imes \\left( 1 - {4k \\cdot G \\cdot \ au \\over \\rho_{\\rm C} \\cdot k_{\\rm v} + 4k \\cdot G \\cdot \ au} \\right).

Neuro‐Fuzzy Control of a Continuous Cooled MSMPR Crystallizer

AbstractThis work deals with the design and application of a neuro‐fuzzy controller of the magma density of the fine crystals for the stabilisation of the crystal size distribution (CSD) of the product from an MSMPR (Mixed Suspension, Mixed Product Removal) continuous crystallizer. The cooling crystallization of potassium sulphate from aqueous solutions in a pilot‐scale stirred vessel was investigated. A control strategy, based on the elimination of fines by using a combined sedimentation‐dissolution device, is presented. The control scheme was successfully applied to a pilot‐scale draft‐tube (DT) crystallizer and yielded promising results. The proposed process control system allowed the reduction of the fines fraction by more than 30 %, while maintaining operation stability and short transient responses.