Ce travail se situe dans le cadre des recherches du laboratoire sur un procédé d’extraction et desextraction, qui implique trois phases liquides, deux phases aqueuses et une phase organique. La faisabilité d’un tel procédé passe par le choix de la phase organique et par l’étude des équilibres mis en jeu lors du processus d’extraction.Le principal objectif concerne : la récupération et la concentration de phénol contenu dans une solution aqueuse, en utilisant un pertracteur à disques tournants alternatifs en régime batch, semi-batch et continu. Par un mécanisme du transfert de matière entre les phases, le phénol passe de la phase aqueuse d’alimentation vers la phase organique puis de la phase organique vers la phase aqueuse réceptrice.La faisabilité de la méthode est testée en mode fermé. Nous étudions l’influence de certains paramètres : tels que la concentration du phénol, la variation de pH de la phase d’alimentation, la vitesse de rotation des disques et le volume de la phase organique, sur l’évolution du transfert du phénol entre les phases. Une solution aqueuse à pH 2, ayant une concentration en phénol égale à 50, 100 ou 300 mg.L-1, est mise en contact avec de l’huile de colza, qui est elle-même en contact avec une solution aqueuse à pH 13. Les résultats montrent qu’au bout de six heures de fonctionnement, 70% à 99% du phénol contenu initialement dans la solution aqueuse à pH 2, se trouve dans la phase aqueuse à pH 13.Afin de récupérer et de concentrer le phénol dans la phase réceptrice, nous avons réalisé des expériences dans le pertracteur fonctionnant en mode semi-ouvert et en mode ouvert.En système semi-ouvert, l’influence de trois paramètres est étudiée : la concentration initiale de phénol, la vitesse de rotation des disques et le débit de la phase d’alimentation. En système ouvert, seul l’influence du débit à la phase aqueuse est analysée. Enfin, grâce à la modélisation basée sur la théorie du double film, nous avons estimé l’ordre de grandeur des coefficients de transferts de matière partiels pour différentes vitesses de rotation en système fermé. / This work is focused on a pertraction process, coupling extraction and stripping steps in the same apparatus and involving three liquid phases, two aqueous phases and one organic phase. The process feasibility requires the choice of a appropriate organic phase (membrane) and a phase equilibria studies.The main objective was recovery and concentration of phenol contained in dilute aqueous solutions using a rotating discs pertractor in batch, semi-batch and continuous mode. The phenol is transferred from the feed phase (aqueous phase) through the membrane (organic phase) in the receiving phase (aqueous phase).The feasibility of the method is tested in a batch system. We studied the influence of some parameters such as the phenol concentration, the feed pH, the discs rotational speed and the volume of the organic phase. An aqueous solution at pH 2 (feed) with a phenol concentration equal to 50, 100 and 300 mg.L-1, is placed in contact with rapeseed oil (liquid membrane), that is itself in contact with an aqueous solution at pH 13 (receiving phase). The obtained results show that after 6 hours, 70% to 99% of phenol initially contained in the aqueous solution at pH 2, is transferred in the aqueous phase at pH 13.To concentrate phenol in the receiving phase, we performed experiments in semi-batch and continuous mode.In semi-batch system, the influence of three parameters is studied: the feed concentration, the discs rotation speed and the feed flowrate. In open system, (continuous mode), only the influence of the aqueous phases flowrates is analyzed.Finally, using mass transfer model based on a double film theory, we estimate the partial mass transfer coefficients for three phases at different rotation speeds in batch system.

Separation of dissolved metals from aqueous solutions using liquid membrane technology is highly advantageous owing to the degree of separation achieved, efficiency and application potential. In the present investigation four types of liquid membranes - bulk liquid membrane (BLM), emulsion liquid membrane (ELM), electrostatic pseudo liquid membrane (ESPLIM) and unified liquid membrane (ULM) have been extensively studied, for their application in extraction and concentration of dissolved metals from effluents. Experiments were conducted with various metal systems to optimize both system and process conditions and to find out the effect of various parameters on the performance of the process. Different mass transport models were proposed for each type, taking diffusional and kinetic resistances into account. Models were extended for simultaneous extraction systems and were verified by different metal-carrier experiments. Good agreement was found between the concentration profiles obtained from the models and the experimental data, thereby establishing the validity of models for all the four types of liquid membranes. The stirred cell employed in BLM process eliminates emulsification and demulsification processes. It also provides simultaneous contact of the organic liquid membrane phase with aqueous feed and strip phases. Overall rate expressions for extraction and stripping in BLM are based on an assumed kinetic mechanism to explain the process qualitatively. It was found that the magnitude^ of diffusional and kinetic resistances determines the overall mass transfer coefficient. The relative magnitude of mass transfer coefficient, reaction rate constants and equilibrium constants enables to visualize the controlling regime of the process. The problem of low flux rate due to high diffusion resistances, inefficient operation and exorbitant costs encountered in bulk and supported liquid membranes (SLM) are overcome in an ELM. In the ELM process, an emulsion of organic membrane phase and aqueous inner phase, is dispersed in the continuous aqueous feed phase. This gives a highly selective and ultra thin liquid film generating a large mass transfer area for separation. Experimental results on membrane instability and emulsion swelling indicate that volumetric leakage rate depends linearly on the stirring speed and that the nature of surfactant does not have any appreciable effect on emulsion swelling. A general permeation model was developed taking into account the external mass transfer around the emulsion drop, diffusion in the drop, reaction at the aqueous-organic interface, leakage of the internal phase to the external phase due to membrane breakup and emulsion swelling due to osmotic pressure difference. Model equations with appropriate boundary conditions were numerically solved by orthogonal collocation technique for a set of model parameters obtained either from known correlations or from independent experiments. Comparison of the model predictions with experimental data from the batch permeation of chromium and other metals using carrier Alamine 336 or LIX 64N, shows that the model predictions are in very good agreement with the experimental findings. Further this model can be used to simulate the effects of various experimental conditions such as metal and hydrogen ion concentrations, carrier concentration, drop diameters, etc., for similar systems. Studies on ESPLIM were conducted with the aim of demonstrating the effectiveness of this new separation process and to develop a simple transport model for metal permeation. In the ESPLIM process, a high electrical field (3-5 kV A.C.) is used for phase dispersion. This system consists of a rectangular reactor filled with membrane solution divided into extraction and stripping cells by a centrally placed integrated type baffle which also acts as an electrode. Two more electrodes were placed in the extraction and stripping cells, where feed and strip phases are introduced from the top of the reactor. When high electrical field is applied across the electrodes, fine droplets of feed and strip are formed and are dispersed in extraction and stripping cells where simultaneous extraction and stripping occurs. The process can be viewed as simultaneous counter current extraction and stripping. The aqueous drops coalesce in the settlers at the bottom of the reactor and are removed continuously. Steady state mass transport model proposed for ESPLIM system accounts for the vertical counter-current extraction and stripping processes taking place in the extraction and stripping cells, together with the lateral transport process of the metal-complex and carrier across the two cells through the integrated baffle zone. The model equations were solved analytically to obtain concentration profiles as a function of the height of the reactor. The required parameters such as mass transfer coefficients, diffusion coefficients etc. were estimated using different correlations. Model predictions agreed remarkably well with the experimental data under various process conditions. From this investigation, it was found that ESPLIM is a simple, efficient and economical process and can be applied in a variety of situations. Based on a suitable combination of solvent extraction, dispersion and liquid membrane technique, a new type of separation system called " Unified Liquid Membrane " was developed. The ULM unit was designed and fabricated, and experiments were conducted to evaluate its performance. The ULM is basically derived from ESPLIM by changing the reactor, baffle design and dispersion technique. Aqueous feed and strip phases were atomized using compressed air through a fine nozzle and are dispersed on either side of an integrated baffle plate that divides the reactor into extraction and stripping cells. Tapering bottom of the reactor reduces the dead volume of the liquid in the settlers and the baffle plate remarkably reduces the leakage problem as well as the resistance through the baffle. Experiments were conducted using LIX 64N and Alamine 336 as carriers for copper and chromium and / or zinc. Mass transport model proposed considers both chemical and phase equilibria in extraction and stripping cells, vertical and lateral transport of carrier and complex across the extraction and stripping cells through the baffle zone. The model equations were solved using initial conditions at the top of the reactor, and equilibrium data for extraction and stripping cells. Effect of various experimental conditions and process parameters was simulated using this model and the model predictions are found to be in excellent agreement with the experimental data. The ULM system developed in this investigation overcomes the major limitations encountered with the other types of liquid membranes while retaining all the advantages of this technology. The problem of high mass transfer resistance from bulk phase to metal permeation as in the case of BLM was eliminated by good phase dispersion. Additional resistance to mass transport from solid membrane as in the case of SLM was removed by using an integrated baffle which also avoids problems of membrane instability, pore clogging and selectivity. The complex problems of emulsification and demulsification were completely eliminated making the system much simpler and efficient. Very good phase dispersion was obtained by atomization without the need for either stirring the whole system or application of high electrical field in the reactor. The membrane liquid within the integrated baffle elements allows easy transport of different species between extraction and stripping cells while completely preventing the mixing of the two aqueous phases. The problems of leakage, swelling and occlusion were avoided due to very short residence time of the aqueous drops in the reactor. It was found that the new ULM configuration is simple, elegant, highly efficient and superior to the other types of liquid membrane systems.