The work aims to develop a process for freezing desalination of seawater on cold walls. The experiments were performed with water/NaCl solutions of different concentrations and with samples of seawater. The pilot crystallizer consists of a cooled tube immersed in a cylindrical double jacketed tank containing water to be treated. The complete process of desalination is conducted in two steps: the freezing step, leading to the crystallization of the ice layer and the sweating step, which consists of purifying the ice layer in depth by melting the impure zones. The systematic study of the influence of operating parameters has highlighted the important role of the cooling ramp and salinity of the solution on the purity of the ice produced. In the absence of stirring, the temperature of the double jacket also has a noticeable effect on the purity of the ice due to temperature gradients, and consequently, convection currents that may result in the solution. The results of this work show the feasibility of the technique and give a good indication of operating conditions that can be used to produce drinking water.

An original process of actinides coprecipitation based on pulsed flow column is studied. The novelty of this process lies in the confinement of the aqueous reagents in separated droplets, dispersed in an inert organic phase (W/O emulsion). Precipitation occurs inside drops when they coalesce. Besides the implementation of well-known technologies of the nuclear industry, this precipitation in emulsion process is particularly convenient for the control of supersaturation, and ensures the sticky precipitates’ confinement within drops, thereby limiting the fouling risk and its adverse consequences on productivity and safety.A thorough understanding of the precipitation mechanisms and their interactions with the hydrodynamic conditions prevailing around and inside the drops is essential for the process optimization. In this context, numerical simulations were conducted, accompanying experiments, to study the process sensitivity. Different levels were considered in the modeling task, going from the emulsion behavior inside the column, to the reagents mixing and precipitation within the drops.Regarding the drop scale, on which we focus in this paper, preliminary static and dynamic observations revealed a stage of mixing of the reagents, followed by a progressive concentration of particles at the drop center leading to their agglomeration. In the modeling three configurations of the reagents’ mixing were therefore considered relatively to the Hill vortices experimentally noticed. CFD simulations allowed calculating possible mean supersaturation profiles in the drop. Two simplified models were proposed to simulate the precipitation inside the drops. Based on the population balance modeling framework, they consider primary nucleation and growth mechanisms and take into consideration either instantaneous or progressive mixing of the reagents. These simplified models were validated with CFD+PBM simulations. Based on the drop scale simulations, process modeling has been discussed.

This work aims at developing a dynamic layer crystallizer operated batchwise, for freezing desalination of sea water. The experiments were performed with water/NaCl solutions and with samples of sea water from Nice, Rabat and Marseille. The pilot crystallizer consists of a cooled tube immersed in a cylindrical double jacketed tank. The solution is poured into the tank and the crystallization takes place on the external surface of the tube, by applying a cooling ramp in the tube. The solution is agitated by air bubbling. The whole process involves the freezing step, leading to the crystallization of the ice layer and the sweating step, which consists of purifying in depth the ice layer by melting the impure zones. A parametric study on the effect of the operating parameters has allowed quantifying the role of the different key parameters of the freezing and sweating steps. Three experiments allowed reaching salinities lower than 0.5g/kg, satisfying the standards of drinking water. The duration of the whole process dropped to only 8h (5h for freezing and 3h for sweating), with a yield of sweating equal to about 50%, provided severe conditions were applied for sweating. Higher yields required longer times. Overall, the results show the feasibility of the technique.

Two in situ process analytical technologies were used to investigate a cocrystallization process: a focused beam reflectance measurement probe and a video probe. The cocrystallizations were conducted batch-wise in a stirred thermostatted vessel with carbamazepine and nicotinamide. The experiments were started in three domains of the phase solubility diagram. When carbamazepine crystals were also formed, a solution mediated phase transition from carbamazepine crystals towards cocrystals was initiated by the addition of the cocrystallizing agent (nicotinamide crystals). As the carbamazepine crystals and the carbamazepine/nicotinamide cocrystals exhibit prism-like and needle-like habits, respectively, it was possible to discriminate between them with the two sensors. The mechanisms involved in the solution mediated phase transition were investigated.

The crystallization-in-emulsion process allows the production of solid particles exhibiting specific features. Here, the batch crystallization process carried out by cooling a melted oil dispersed as an oil-in-water emulsion was studied. Two experimental set-ups allowing the in situ visualization of the nucleation and growth phenomena occurring in the dispersed liquid phase were developed. Observations in quiescent medium of motionless droplets having a diameter of few tens of micrometers showed that primary nucleation started on the inner surface of the droplets. The fast growth of the crystals consumed all the liquid contained within each droplet and was confined within each droplet by the oil–water interface. Solid polycrystalline particles similar in size to the parent droplets were produced. Dynamic tracking of the transient evolution of the size distributions of the two populations of droplets and solid particles during the cooling process in a stirred vessel was carried out using an in situ optical probe. It was shown that the droplets crystallized very progressively during cooling, starting with the largest droplets and ending with the smaller size droplets since the induction time of primary nucleation was dependent on droplet volume. In dilute conditions (1%wt% of dispersed phase) each droplet was converted into a single solid particle. Secondary nucleation based on inter-droplet collisions was not observed in these conditions.

Phase and habit selection is a very important step in the early stages of pharmaceutical development of new APIs. In this paper, we show how observation, diffraction and thermal analysis are complementary methods of solid habit and phase characterization. At the end of phase screening of an API several habits and phases can be discriminated by microscopy, XRPD or Raman spectroscopy. Using thermal methods here allows us to separate the 12 phases discriminated by XRPD into: anhydrous, monohydrate, organic monosolvate and heterosolvate phases.

Understanding the evolution of liquid–liquid dispersion is a key factor in operation and control of emulsification process. A dynamic tracking of an evolving droplet size distribution (DSD) in a dilute oil-in water (O/W) emulsion has been developed with the implementation of an in situ video probe dipped in a stirred vessel and coupled with an image analysis treatment. The optical probe allows real time recording of 2D images of the droplets. Recent image analysis software developed originally for an application in micromechanics was adapted and applied in delayed time on the video sequences regularly recorded. It automatically provided a measurement of the diameter of several thousands of droplets in the range of 10–100μm. The relative accuracy on the droplet number average diameter was 10%. This measurement technique was then used to investigate at lab scale in a stirred vessel and in warm conditions the influence of several process parameters on the evolution with time of a dispersed melted cosmetic ingredient in a water solution containing a surfactant. The specific power input of stirring was the main parameter acting on the reduction of the mean droplet diameter and of the width of the DSD owing to its action on the droplet break-up mechanism. The surfactant concentration was a parameter of secondary relevance on the DSD probably due to the reduction of the coalescence rate and to a faster stabilization of the O/W interface. The use of a flat blade propeller instead of a Rushton turbine was preferable for the production of a narrower DSD. Finally the time required to reach equilibrium was found higher by a factor of 3–4 than predicted in the literature.

The purpose of this work was to assess the possibility of inducing solution mediated phase transition (SMPT) by manipulating the amount of the cocrystallizing agent. The cocrystal, composed of an active pharmaceutical ingredient (carbamazepine, CBZ) and its cocrystallizing agent (a vitamin—nicotinamide, NCT), was selected as a model compound. Batch experiments were performed in a stirred vessel. The solute concentrations of both CBZ and NCT were monitored using in situ ATR-FTIR spectroscopy. The introduction of NCT in dry form allowed a shift in the phase diagram, leading to an SMPT from CBZ crystals toward cocrystals. The concentration profiles gave information on the phase transition kinetics, i.e., the kinetics of nucleation, growth and dissolution mechanisms of the solid phases involved. Several situations were analyzed. This procedure could also be used to correct a process deviation that led to CBZ crystals instead of cocrystals.

In the process of desalination by indirect freezing, ice formed on the cooled surface contains impurities due to kinetic effects. Sweating is an efficient method by which ice layer is purified under the effect of temperature gradient. The main kinetic parameters influencing sweating of ice are initial concentration of ice, sweating temperature and sweating time. In the present work, the effects of these parameters were studied using an experimental design. A statistical model for ice weight and ice purity was developed. The graphical representation of this model in the space of the variables enabled optimization of the whole desalination process time, which led to ice salinity less than the drinking water standards (0.5g/kg).

This work aims in developing a static layer crystallizer for freezing desalination of sea water. The experiments were performed with a simple system of H2O-NaCl and with samples of sea water from Rabat. The pilot crystallizer consists in a tube cooled by means of a thermostatic bath. The tube is immerged in a cylindrical double jacketed tank cooled by means of a second thermostatic bath. The brine is poured into the tank and the crystallization takes place on the external surface of the tube. The global process is divided into 4 steps: (i) crystallization of the ice layer by controlling the cooling rate in the tube (ii) draining off the concentrated brine (iii) purification of the layer by sweating and (iv) melting of the ice to recover the fresh water. A parametric study of the effect of the operating parameters has allowed us to quantity the role of the different key parameters of the crystallization step. Within the studied domain, the purity of the crystalline layer was mainly affected by the initial salinity of the brine. The growth rate of the layer, controlled by the cooling rate in the tube, had also a significant effect. Experiments performed with Rabat sea water showed that a fresh water of salinity close to the drinking water standards Could be obtained in one stage within 31 h. Desalination operated in two consecutive stages (10 h + 11 h) gave salinity below the standards with a comfortable safety margin. If sufficiently severe operating conditions are applied, sweating is able to purify the interior of the ice layer and to reach the drinking water standards, provided the impurity concentration of the ice produced in the crystallization step is low enough. The mass loss induced by sweating is also high when file impurity concentration is high. These first results are promising and show the feasibility of the process which still requires to be optimized.