Rozprawy doktorskie na temat „Ultrafiltration”

Les membranes utilisees sont des membranes minerales. Pour l'ultrafiltration, la fluidisation permet d'ameliorer notablement les debits de permeation sans qu'il y ait pour autant deterioration de la retention en macromolecules. Mise en evidence de conditions optimales de porosite du lit fluidise. L'action des particules fluidisees est moins significative pour l'electroultrafiltration

Ce travail concerne la conception et le controle d'installations destinees a concentrer du lactoserum doux par ultrafiltration sur membranes minerales. Une etude bibliographique met en evidence la necessite d'etablir des lois reliant le transfert de matiere aux conditions operatoires (pression, regime hydrodynamique, concentration, geometrie de membrane). Ces lois sont etablies empiriquement en utilisant divers lactoserums et des solutions de concentrats de proteines. Deux methodologies experimentales (pression constante ou pression variable, a concentration constante) sont envisagees. Enfin, l'etude du fonctionnement d'une boucle de concentration a l'echelle du laboratoire est menee et les resultats compares a un modele integrant les lois de transfert empiriques determinees precedemment. Le bon accord entre les calculs et les resultats experimentaux permet d'utiliser le modele pour analyser l'influence des caracteristiques de l'installation (volume mort, surface de membrane) sur le temps de mise en regime des installations

Molasses is a dark brown by-product of sugar production and refining, rich in sugar, vitamins, and salts. The dark coloured pigments are mainly melanoidins, which are produced through the Maillard reaction between amino acids and sugar. Molasses is widely used as an inexpensive source of sugar in biorefineries to produce a range of products including baker’s yeast. Due to adsorption of melanoidins, Baker’s yeast has a dark colour and needs to be washed in several steps to be acceptable to the market. This washing increases the volume of wastewater, which causes major wastewater treatment issues. Ultrafiltration of molasses has been proposed to minimise the need for downstream washing processes, by removing the coloured compounds, whilst also sterilising the molasses prior to fermentation. The ultrafiltration of molasses poses a number of scientific questions and technical challenges. In particular; what is the fouling mechanism of molasses? What determines the sugar retention in such systems? In addition, what is the optimal membrane for molasses decolourisation? The approach used in this thesis is to systematically investigate the following areas: (1) membrane characterisation; (2) molasses characterisation; (3) investigating the effect of operating conditions on ultrafiltration; (4) investigating modelling of fouling and optimising the cleaning procedure i.e. the interaction of three other areas (5) modelling of cane and beet molasses ultrafiltration to select the optimal membrane. The literature review found that the prior research has generally been limited with regards to the effect of membrane materials on ultrafiltration of molasses or other sugar streams. Previous researchers have typically used a single type of molasses and a small number of membranes. In addition, most of the limited existing literature has focused on the relationship between the manufacturers specified Molecular Weight Cut-Off (MWCO) and colour removal from the sugar stream for a single molasses type. Generally, a relationship was found in each single study but these relationships were inconsistent between studies. Therefore, a comprehensive set of experiments was carried out on ultrafiltration of molasses at different operating conditions (5 and 10 bar), with a range of membranes of different MWCO (2-30 kDa), for polysulfone, polyethersulfone, polyacrylonitrile and polyvinylidenedifluoride, on both cane and beet molasses. Two membranes geometries (flat sheet and spiral wound) were also used in ultrafiltration of cane and beet molasses. As molasses characteristics are crucial factors in the ultrafiltration process, six different types of molasses, three beet and three cane, were characterized based on their chemical and physical properties. These properties can vary considerably depending on the type of molasses (cane or beet), the origin of molasses and the production process. These variations are of importance to the bioprocessing industries, which use molasses as a feedstock. Despite this, few authors have examined characteristics relevant to industrial processing such as the viscosity of different types of molasses and only a few studies have investigated colourant molecular weight distribution. This gap has been addressed in the present thesis. To this end, a comprehensive set of experiments was undertaken to characterise molasses with a view to the industrial application of molasses ultrafiltration in baker’s yeast production plant. It was found that cane molasses contains higher amounts of suspended solids compared to beet molasses (11.3-21.5 g/L for cane molasses compared to 6.6-7.8 g/L for beet molasses), accounting for higher sugar retention and rapid fouling during ultrafiltration which was found to be complete pore blocking. In addition, using Liquid Chromatography–Organic Carbon Detection (LC-OCD) method, it was shown that molecular weight distribution of colour pigments in cane molasses is more skewed toward higher molecular weights. The average molecular weight of melanoidins in cane molasses is 2.5 to 3 times higher than melanoidins in beet molasses. As part of this thesis, the effect of operating conditions on ultrafiltration performance was quantified through systematic laboratory experimentation. Changes in permeate flux and its colour and sugar content in relation to changes in the operating conditions was examined for various molasses-membrane combinations. For both beet and cane molasses ultrafiltration 5 kDa flat sheet membrane made of polyethersulfone yielded the highest colour removal (87% for cane molasses and 70% for beet molasses). The effect of increasing pressure from 5 bar to 10 bar was negligible in terms of colour removal. However, it increases the sugar retention by more than 100% (from 10% to 21 % for cane molasses and from 1% to 12% for beet molasses) suggesting different mechanism for sugar/colour separation and a relationship between sugar retention and fouling. These findings provide an opportunity to optimise ultrafiltration process by operating at low pressures. In ultrafiltration of cane molasses, it was found that less hydrophilic membranes (contact angle ~ 75o) tend to show slower fouling and higher colour removal compared to more hydrophilic membranes (~ 37o). In addition, it was found that membranes of the same nominal size (MWCO: molecular weight cut off), can have different pore sizes (Pore size of Alfa Laval 5 kDa membrane is 10.8 nm compared to pore size of Synder 5 kDa membrane which is 8.8 nm) that will affect the permeability of membranes and hence the flux. The fouling mechanism was modelled for the ultrafiltration of cane and beet molasses, using first and second derivatives of flux. It was found that the fouling mechanism for ultrafiltration of cane molasses is complete pore blocking at both 5 and 10 bar. Complete pore blocking in cane molasses ultrafiltration can be attributed to the presence of high amounts of suspended solids in cane molasses (TSS=16.3 g/L) and their large size (>22 µm). A range of different cleaning protocols were tested on spiral wound membranes in order to maintain the durability and efficiency of membranes and keep the maintenance cost of ultrafiltration minimal. A combination of water washing, followed by caustic washing was successful in removing organic fouling. The highest flux recovery of 65% was achieved using this procedure. Higher flux recovery could be potentially achieved by backwashing. However, both membrane manufacturers recommended against backwashing. More trials on molasses ultrafiltration using ceramic membranes could increase the flux recovery. The concluding work of this thesis involved statistical modelling of molasses ultrafiltration that was carried out on cane and beet molasses in order to find the optimal membrane selection for each cane and beet molasses ultrafiltration. It was found that permeate flux for both cane and beet molasses is positively correlated to average pore size of membranes. However, the determining factors in colour removal in cane and beet molasses ultrafiltration are different. In cane molasses due to presence of hydrophobic compounds, hydrophobicity of membranes (i.e. the contact angle) as well as MWCO plays important roles. Colour removal in cane molasses ultrafiltration was positively correlated with contact angle (r=0.906) and negatively with MWCO (r= -0.88). Whilst in beet molasses ultrafiltration, the relationship was not as strong between colour removal and MWCO (r=-0.444). In beet molasses ultrafiltration, a spurious correlation was found between colour removal and average pore size, which was not mechanistically meaningful. In both cane and beet molasses ultrafiltration, flux was exponentially correlated to the average pore size of membranes. It was found that the cause of this was that the optimal tested membrane in terms of maximum colour removal, minimum sugar retention, and highest flux for both cane and beet molasses is polyethersulfone 5 kDa membrane and this was reflected in the results of the statistical models. Now the optimum membrane based on membrane characteristics (MWCO of 5 kDa and 75o contact angle) can be chosen for plant trials. Modelling the system design for ultrafiltration of cane and beet molasses concluded that a three stage ultrafiltration unit, can obtain a sugar recovery of more than 95%. By using ultrafiltration upfront in molasses based industries, energy consumption cost for the wastewater treatment reduced by 84%. Results of this research highlight the significant merit of using ultrafiltration upfront in molasses based industries to overcome wastewater treatment issues in molasses-based industries.

Desizing wastewater is largely responsible for the chemical oxygen demand (COD) load in the textile industry wastewater. A larger portion of COD comes from degraded starch in desizing wastewater. Removing the starch from the wastewater by an ultrafiltration process may reduce the environmental problem caused by the textile factory. If the treatment is made in such a way that all starch components are removed from the wastewater, the treated water can be reused by the factory. If the starch in the concentrate is stable, it can also be reused as a sizing agent. This will give the factory an economic advantage.

In this thesis we have studied the fouling mechanisms involved in the ultrafiltration of solution with partially degraded starch in order to find the treatibility of such solutions. The work has mainly been directed to uncover how the different fouling mechanisms depend on the operating parameters, and to find the performance of selected membranes. In addition, different models were evaluated for their validity in predicting the performance of the membranes and the data was fitted to the model that give the best prediction and are physically more meaningful. In addition, the starch solution was concentrated, and the flux, concentration, retention and rejection profiles as a function of concentrating time were investigated.

For the study, we used a partially degraded starch solution as a model solution. The solution was prepared in the laboratory by enzymatic degradation of potato starch to different levels. In order to evaluate the reproducibility of the degraded starch, three replicates were prepared. The reproducibility was determined by comparing the molar mass distribution from HPLSEC analysis and the concentration of reducing sugar from a DNS test for the replicates. The analyses show very good reproducibility. Three starch model solutions with three different degradation levels were chosen for our ultrafiltration experiments to investigate the effect of average molar mass of the starch.

For the ultrafiltration of the solution ES625 (from PCI) and MPT-U20 (from KOCH) membranes were used. Both membranes were used in the investigation of the contribution of different fouling mechanisms to the flux decline during ultrafiltration of the solution. In the evaluation of the performance of ultrafiltration of the starch solution, however, only the ES625 membrane was used. According to the manufacturers, both membranes have nearly equal pure water flux and MWCO. But in our test, we observed a higher and different pure water flux for each type of membrane. The ES625 had a lower flux (higher retention) than the MPTU20 membrane.

In the ultrafiltration of partially degraded starch solution the permeate flux declines very fast and, for a low feed concentration, it reaches a steady state in a very short time. The steady state time was observed to increase with concentration, molar mass and transmembrane pressure drop, and to decrease with cross flow velocity. All the three fouling mechanisms (concentration polarization, adsorption and deposition) were responsible for the flux decline. The major observed contributors are, however, adsorption and deposition. Adsorption is largely responsible at low-pressure operation while the deposition fouling effect is dominant at higher pressures, near or beyond the limiting flux.

For the ES625 membrane, the contribution of adsorptive fouling increases with concentration and decreases with molar mass of the starch, temperature and pH at a given transmembrane pressure and cross flow velocity. The effect of the operating parameters on the depositional fouling is in line with literature. It increases with pressure, concentration, molar mass and temperature, and decreases with cross flow velocity. Its dependence on pressure can be expressed by a power function with exponent larger than 1.0. This seems to due to an increase in thickness and compaction of the starch gel/deposit at the membrane surface as the transmembrane pressure drop is increased. The contribution of the concentration polarization is also dependent on concentration, cross flow velocity and pressure. Its relative contribution increases with concentration while it decreases with an increase in cross flow velocity. In the turbulent flow regime the relation between the resistance contributed by concentration polarization increases almost linearly with transmembrane pressure drop. In the laminar flow regime, however, the relative contribution of the resistance due to concentration polarization increases for the lower range of pressure and decreases for the higher range of pressure. Its relative contribution also increases with temperature and decreases with increasing molar mass. But the overall fouling resistance in the ultrafiltration of the starch solution increases with feed concentration, molar mass of the starch and transmembrane pressure drop and decreases with cross flow velocity and temperature.

The trend of the flux loss due to all fouling mechanisms for MPT-U20 membrane is similar to ES625 membrane except for adsorption and concentration polarization with changes in concentration and molar mass. The difference could be a result of the difference in morphological properties between the two membranes and the experimental procedures used in determining flux data that used for calculating the contributions. From the pure water flux and the retention data, the ES625 membrane seemed to have a smaller pore size than the MPT-U20 membrane.

Among the ultrafiltration models, the resistances-in-series model was chosen for its provision to include all the fouling mechanisms into the model. When our permeate flux data was fitted to the model, it gives a good fit. However, the model fails to give realistic estimates of the contribution of the individual fouling mechanisms. In order to improve this problem, the model was modified by introducing osmotic pressure across the membrane in such a way that the effect of concentration polarization is accounted for. This modified model is more physically meaningful and gives a realistic estimate of the contribution the reversible and irreversible fraction of the overall resistance.

In concentrating mode operation, the permeate was continuously withdrawn and hence, the concentration of starch in the feed tank was increased. At an early stage of ultrafiltration, the permeate flux appeared to increase slightly, which seems, a result a shear thinning of the starch solution when the solution was pumped through the system. For the rest of the operation, the flux was decreasing, the retention was increasing and the rejection of the membrane was shifted to a lower molar mass as the solution in the feed tank got more concentrated as expected. The shift of rejection to the lower molar mass region is due to the fouling layer that reduces the accessibility of the pores of the membrane.

Generally, the flux we obtained in ultrafiltration of a partially degraded starch solution with the ES625 tubular membrane is equal or better than the reported values from an existing ultrafiltration plant that has been used in the textile industry to recover a synthetic sizing agent (PVA) from the desizing wastewater. The retention is, however, rather low. Two or more stages of treatment are needed to get all starch components removed from the wastewater and make the treated water reusable (recycled).

In axial electrofiltration, a DC electric field is imposed between a rotating inner cylinder and a stationary outer cylinder giving rise to four mechanisms which act to minimize solute accumulation at the filter surface: turbulence, centrifugal force, electrophoresis and shear stress which removes solute aggregates.
Four dynamic membranes, Zr(IV) oxide, calcium oleate, poly-2-vinylpyridine and cadmium sulfide, were used to filter bovine serum albumin (BSA) in a disodium phosphate solution at pH = 8 and Prussian blue in distilled water. Prussian blue is a particle of 0.01(mu)m diameter with a zeta potential of -41mV while BSA is a macromolecule of 69,000 molecular weight, a Stokes-Einstein radius of 0.0038(mu)m and a zeta potential of -23.3mV at pH = 8. For BSA, the flux declined with time while the rejection increased. Filtrate fluxes increased with rotation rate and electric field and declined with concentration for both feeds. The flux declined beyond N = 2000rpm and was constant above C(,0) = 5.0wt%. For Prussian blue, the rejection was greater than 90% at all levels of E, N and C(,0). For BSA, the rejection increased with rotation rate and declined with concentration. The BSA rejection declined above N = 2000rpm and was constant beyond C(,0) = 0.5wt%.
A mathematical model was derived to predict the time variation of filtrate flux and a rejection model was used to predict the effect of surface concentration on BSA rejection.

Osmotic pressure and diffusion coefficients were studied experimentally and theoretically in order to facilitate fully quantitative descriptions of ultrafiltration processes. These parameters may be incorporated in the solution of the equations for hydrodynamics and mass transfer in the filtration module. Theoretical descriptions are presented to allow calculation of osmotic pressure, taking account of the electrostatic repulsive interactions, London Van Der Waals forces, configurational entropy and additional interactions of unknown source. Approaches of constant zeta potential and charge regulation conditions were investigated for BSA, lactoferrin and silica colloidal dispersions. An experimental technique for measuring osmotic pressure was established and verified by comparison of the measured osmotic pressure with known data using BSA protein dispersions. The measured osmotic pressures of BSA solutions also agreed well with the theoretical predictions. Osmotic pressures of the commercially important protein lactoferrin solutions were also measured. Using the experimental results further refinement of the charge regulation model was made by modifying the expression of the extra attractive force, which is probably caused by interaction between sugar molecules on the protein surfaces. The refined model gives greatly improved prediction of osmotic pressure as well as ultra-filtration rate over a wide range of solution conditions. Osmotic pressures of silica dispersions were also measured and a charge regulation model was developed for the silica colloidal system. This model gives a good description of the experimental results over the conditions studied. The gradient diffusion coefficients of BSA solutions were measured by the modified Taylor's capillary method and compared with light scattering data. An analysis of electrokinetic effects in membrane pores containing electrolytes has also been carried out. The analysis uses a numerical solution of the non-linear Poisson-Boltzmaan equation and allows for the mobilities of anions and cations to be individually specified. It is shown that it is very important to use such an approach to calculate zeta-potentials from the basic electrokinetic data.

The influence of the physicochemical conditions on the permeation rate in dead-end and cross flow ultrafiltration of protein dispersions is investigated. The rigorous mathematical models developed are based on sophisticated descriptions of the protein-protein interactions within the layer close to the membrane surface which is responsible for controlling the permeation rate. The calculations are expressed in terms of an osmotic pressure which accounts for the multiparticle electrostatic interactions, dispersion forces and configurational entropy. Electrostatic interactions are accounted for by a Wigner-Seitz cell approach including a numerical solution of the non-linear Poisson-Boltzmann equation. London-van der Waals forces are calculated using a computationally efficient means of approximating screened, retarded Lifshitz-Hamaker constants. Configurational entropy effects are calculated using an equation of state giving excellent agreement with molecular dynamic data. Electroviscous effects are also taken into account. These descriptions of colloidal interactions are used to develop a priori models, with no adjustable parameters, that allow quantitative prediction of the osmotic pressure and the rate of filtration of protein dispersions as a function of zeta potential (and hence pH), protein size, ionic strength, protein concentration, applied pressure and membrane resistance. The model shows good agreement with literature experimental data for the osmotic pressure of the protein bovine serum albumin (BSA). A further refinement to the models has been the development of a surface charge regulation model from knowledge of the amino acid groups giving rise to the protein charge. For cross flow ultrafiltration, the osmotic pressure is further used in the calculation of the gradient diffusion coefficient from the generalised Stokes-Einstein equation. The filtration models have been tested by conducting dead-end and cross flow ultrafiltration experiments with protein dispersions of BSA and lactoferrin. Good agreement between theory and experiment has been obtained for the dead-end model, but further work is required for the cross flow filtration model.

L'eau residuaire de raffinerie petroliere traitee contient un melange d'hydrocarbures (hc) et de matieres en suspension (mes) resultant principalement d'un traitement biologique. L'ultrafiltration (uf) semblerait etre en mesure de resoudre le probleme de rejets en respectant les normes europeennes. La principale limitation economique resulte de la densite de flux de permeat qui doit etre suffisamment elevee pour minimiser la surface filtrante. Le but de ce travail est l'identification des mecanismes de limitation de transfert de matiere. La premiere etape a ete l'identification d'une membrane potentielle. La membrane m9 carbosep a ete selectionnee car elle a la meilleure permeabilite et elle constitue une barriere totale pour les hc et les mes. Dans l'etude de l'uf d'hc seuls, de mes seules ou de suspensions mixtes, on a mis en evidence une incoherence entre les modeles classiques et l'experience. Cependant, le modele du depot avec retroflux est apparu tres satisfaisant. Enfin, dans le but d'ameliorer le flux de transfert, on a introduit un promoteur de turbulence dans le tube membranaire

L'analyse des potentialites de fractionnement de melanges de macromolecules par ultrafiltration et electro-ultrafiltration est basee sur l'etude des melanges albumine-lactoferrine et albumine-poly(ethylene glycol). L'etude du premier melange montre que la transmission d'une proteine a travers une membrane depend de l'exclusion electrostatique et sterique. Elle met egalement en evidence des conditions de ph et de force ionique pour lesquelles les interactions lactoferrine-lactoferrine et lactoferrine-albumine modifient considerablement les mecanismes qui controlent habituellement la selectivite d'une separation et conduisent a l'inefficacite d'un champ electrique. L'etude du second melange conduit a un paradoxe. Elle montre que la transmission d'un poly(ethylene glycol) est favorisee par son accumulation a la surface de la membrane, mais limitee par l'accumulation de l'albumine. Dans des conditions operatoires optimisees l'albumine est separee d'un poly(ethylene glycol) de masse molaire inferieure a 20 kd. Une selectivite plus elevee necessite la dissociation des transports des deux macromolecules. La difference de charge portee par les solutes est exploitee pour dissocier leurs transports. Un champ electrique applique durant une ultrafiltration permet de limiter l'accumulation d'albumine a la surface de la membrane tout en maintenant une transmission elevee du poly(ethylene glycol). Ainsi, l'electro-ultrafiltration peut fractionner l'albumine du poly(ethyleneglycol) 20 kd dans un mode de concentration ou de diafiltration, alors que ces deux macromolecules ont des tailles apparentes identiques

Proteins and fatty acids often exist in solutions containing biological matter that are treated with membranes. These proteins and fatty acids interact with each other as well as with the membranes thereby affecting the flux. Binding of fatty acids to proteins results in complexes that are much larger than fatty acid molecules. Exploitation of this size difference to remove difficult to separate fatty acids from aqueous solutions by ultrafiltration was investigated in this study. In addition, the fouling of membrane by the protein-fatty acid mixtures containing free dissolved fatty acids was studied using bovine albumin (BSA)-caprylic system. Binding of caprylic acid to native and pasteurized BSA was examined by diafiltering pre equilibrated fatty acid-BSA mixtures. The rate of mass transfer of fatty acid molecules through boundary film surrounding the protein molecules was estimated using a BSA solution as the adsorbent phase in an agitated column. A stirred cell fitted with a polyethersulfone membrane (30 kDa) was used for the diafiltrations. Accumulation of fatty acid in the BSA layers fouled on the membrane was also estimated. Binding studies indicate that a native BSA molecule (at pH 6.8) could bind 7 fatty acid molecules in specific binding cavities while approximately 44 molecules are bound onto the surface. When BSA was pasteurized the specific binding decreased from 7 to 2 indicating unfolding of the molecule. In addition, the total binding capacity decreased from 44 to 24 moles/BSA mole and the rate of mass transfer decreased from 4.5/min to 3.6/min, indicating heat induced aggregation of BSA. At alkaline pH levels fatty acid anion acts as an anionic surfactant stabilizing the molecular conformation of the protein and reducing fouling. When pH was lowered to 3, flux severely declined. Unusually large accumulation of fatty acid in the deposited protein layers (caprylic/BSA ~ 10,000 moles) occurred indicating capillary condensation of undissociated fatty acids in the protein layer. Agitated column studies showed that proteins could be used as an adsorbent to remove hard to separate dissolved fatty acids from aqueous solutions. The separated protein-fatty acid complex may be further processed to manufacture animal feed.

The development of quantitative predictive models is of great significance for the successful application of membrane separation processes in the process industries. Successful models exist for the ultrafiltration of macromolecules and particles with sizes 2 to 5nm and 500nm to 10μm, respectively. None of the approaches could predict colloidal ultrafiltration rates in the intermediate size range of 5 to 500nm. This is due to interparticle interactions, especially electrostatic interactions which play an important role and which have been widely neglected or even overlooked by many membrane scientists. As a result of such interactions, changes in the ionic environment and particle zeta potential give order of magnitude changes in the rate of ultrafiltration. However, there has been no rigorous mathematical model for the ultrafiltration of colloids in this range that takes into account such interparticle interactions. The aim of the present dissertation was to develop and test rigorous mathematical models for membrane ultrafiltration of charged colloidal particles, which include quantitative calculations of particle-particle interactions within filter cakes, which are responsible for controlling permeation rates. This has been realized by two approaches: 1) an extended pairwise summation of interaction energies due to electrostatic or double layer forces and London-van der Waals forces through the DLVO theory and 2) by a multiparticle interaction approach based on the use of a Wigner-Seitz cell model to calculate the grand canonical electrostatic potential energy of a concentrated particle system. Dynamic ultrafiltration models for filter cakes of charged particles have been developed ('From Physics To Filtration') based on disjoining pressure calculations between hexagonal close packed particle layers which take the former interaction calculations into proper account. The models require no adjustable parameters. A comparison with experimental ultrafiltration results revealed that the dynamic model based on pairwise summation of interaction energies was in good agreement with the experimental filtration flux over a limited range of conditions, whereas the cell model predictions were in excellent quantitative agreement over a wide range of ionic and zeta potential conditions.

On etudie l'ultrafiltration de fluides contenant des proteines et plus particulierement du lait et de ses derives. On etudie plus precisement le phenomenes de diffusion d'encrassement et de colmatage de la membrane. On observe le flux limite de solute en fonction de la polarisatiuon, de la concentration et des proprietes rheologiques des solutions proteiques utilisees

Cette etude est consacree a la retrofiltration haute frequence et ses applications economiquement importantes. Dans un premier temps, la microfiltration d'une suspension de bentonite, a travers une membrane ceramique tubulaire de 0,2 m, a valide le montage par l'obtention de flux de l'ordre de 900 l. H - 1. M - 2. L'amortissement interne a ete quantifie par un montage specifique et par l'observation des regimes transitoires. Une etude systematique a permis de determiner l'influence de chaque variable operatoire. La retrofiltration n'empechant pas une decroissance du flux vers un equilibre cinetique, l'influence de la frequence a ete minimisee. Le couple (pression de retrolavage, duree de retrolavage) depend des conditions hydrodynamiques et devra etre suffisant pour renouveler la sous-couche laminaire. L'emploi de membrane a peau externe ne s'est pas avere satisfaisant. Le montage a ete soumis a des suspensions biologiques mais n'a pas permis de depasser 200 l. H - 1. M - 2. Ces gains auraient pu etre obtenus par simple augmentation de la vitesse tangentielle. Afin de prevoir et de permettre le dimensionnement d'unites, une modelisation ou, du moins une formalisation de la retrofiltration a ete proposee. Cette technique permet, avec un nombre restreint d'essais d'evaluer l'efficacite et la frequence optimale et fait le lien entre la reponse a une impulsion et une retrofiltration haute frequence.

The mass of five proteins (Bovine serum albumin (BSA), casein, lysozyme, ovalbumin and pepsin) adsorbed to five different membrane materials (of various hydrophobicities) was quantified using a static system and analysed to establish any trends. Comparing the results from the five membranes it seems that there were no obvious trends between the protein masses adsorbed indicating that it may not be just one aspect of protein structure that is important in the adsorption process. Many investigations have indicated that the protein may undergo a conformational change during the adsorption process. Disulphide bridges contribute readily to the stability of the protein molecule and it was hypothesised that if such a structural change occurred, it would result in the breakage of these covalent bonds. To this end, the free thiol group content of the proteins was quantified before and after adsorption.

The effect of ultrafiltration, diafiltration, and preacidification of milk on the partition of specific milk components between retentate and permeate was studied. Percent retention (for any component Y) was determined as: [ l - (%Y in soln. permeate/%Y in soln. retentate)] X 100 where %Y in soln. = [%Y/(%Y + %H 20)] Simultaneous samples of retentate and permeate were taken at several points during each process. Percent retention of total solids, fat, total protein, rennet clottable nitrogen, lactose, total calcium, ionic calcium, sodium, phosphorous, and riboflavin was determined at each sampling point. Percent retention of β-carotene, vitamin B12, retinol, and zinc was determined at different stages of ultrafiltration only. As UF proceeded, percent retention of total solids, total protein, total calcium, ionic calcium, sodium, phosphorous, and riboflavin increased. Percent retention of lactose was not affected and was 0-4%. Percent retention of fat, rennet clottable nitrogen, zinc, retinal, β-carotene, and vitamin B12 was 99-100%. As diafiltration proceeded, percent retention of total solids, lactose, total calcium, ionic calcium, sodium, phosphorous, and riboflavin increased. Percent retention of these nutrients was increased when compared to ultrafiltration alone. Milks with lower pH values (resulting from preacidification) also had lower percent retention of total calcium and phosphorous than milks with normal pH values. Percent retention of sodium was lower during ultrafiltration and diafiltration of acidified milk when compared to ultrafiltration and diafiltration of normal milk. Percent retention of other nutrients was not affected by acidification. Loss of whey proteins into permeate resulted in a lower recovery of total protein after diafiltration than ultrafiltration alone and ultrafiltration of acidified milk. These proteins were determined to be α-lactogobulin and β-lactogobulin.

Membrane technology has acquired significant importance in a variety of applications such as water treatment, health sector and food industry. Nevertheless there is a growing demand of more functional and stable membranes. This thesis is focused on improving solvent stability of polysulfone based membranes without compromising the flux properties. Polysulfone (PSF) is a widely used membrane material for ultrafiltration process because PSF can be easily fabricated into highly porous structure via non–solvent induced phase separation (NIPS). Such a structure is imperative for ultrafiltration process as efficiency of filtration largely depends upon the size of pores and overall porosity of membrane. Generally membranes made of pristine PSF have good chemical and mechanical stability but get dissolved in many of organic solvents. Though previous attempts somehow attained solvent stability but membrane structure was deteriorated and so does the transport property. In order to increase the solvent stability, an approach based on UV induced acrylic functionalization of PSF is described in this thesis. A two-steps method of membrane fabrication, involving NIPS and UV curing, was established. At first the acrylic functionality was introduced on the backbone of polysulfone chain through synthesis of methacrylated polysulfone (PSF–DM) macro-monomer. The thin films of PSF-DM containing phenylbis (2,4,6-trimethylbenzoyl) phosphine oxide (BAPO) photoinitiator cast on the glass plates were subjected to UV curing and NIPS processes to prepare the membranes. The reverse process i.e. NIPS followed by UV curing was also applied. The developed membranes were thoroughly characterized and the flux properties were evaluated in a dead–end filtration apparatus. The solvent stability was good in DMSO and acetone. Regardless of sequence of operations the average flux values were in the range of tight ultrafiltration. The sequence UV–NIPS resulted into more uniform and structurally stable membranes. The next approach was based on incorporating acrylic functionality in a way that a diacrylate monomer was added into pristine PSF solution in DMF and the membranes were fabricated following the previously established method i.e. UV curing followed by NIPS. For this purpose two structurally different di-acrylates, Bisphenol A ethoxylate diacrylate (BEDA) and Poly (ethylene glycol) diacrylate (PEGDA) were incorporated and in order to study the effect of concentration, different amounts of acrylic resins were investigated. The acrylic double bond conversions were studied through FTIR and ATR–FTIR. The viscoelastic properties were analyzed via dynamic mechanical thermal analysis (DMTA) and morphological properties were studied through field emission scanning electron microscopy (FESEM). The solvent stability was analyzed by immersing the membranes cutouts in a number of solvents for 120 hrs at room temperature. The flux properties were evaluated in a pressurized stirred cell apparatus and the rejection of 27 nm polystyrene particles was also collected. Almost all the formulations came up with excellent solvent stability in DMF, DMSO, THF, Acetone, Toluene and Ethyl Acetate. The 10 % BEDA functionalized PSF membrane exhibited the best flux properties, i.e. very close to that of pristine PSF, along with good rejection (>90 %) of 27 nm polystyrene particles. The UV cured acrylic functionalized PSF membranes were further applied to specific contaminant removal. The previously developed best UF membranes were coated with polydopamine (PDA) in order to combine the rejection of ultrafiltration membrane with the adsorption of contaminant by PDA layer. As BEDA cross-linked membranes were developed through UV curing followed by NIPS so the strategy of PDA coating was coupled with NIPS either as a separate action next to NIPS or in a more facile and advantageous one–step method where NIPS and PDA coating occurred at the same time. In order to study the increased adsorptive behavior of PDA coated membranes, methylene blue (MB) was taken as a model contaminant. The removal of MB was investigated both in batch and continuous filtration. The Zwitterionic behavior of PDA was effectively exploited and the coated membranes released MB in acidic condition thus regenerated for next adsorption. The cyclic stability of several adsorption and desorption cycles was studied. Above stated all three works discussed the development of UV cured membranes through NIPS In recent years, electrospun nano-fibrous membranes (ENMs’) have been employed as successful replacement of phase separation membranes because of high porosity thus much lower mass transfer resistance. We designed a reactive electrospinning setup where electrospinning was coupled with online UV irradiation. The same BEDA functionalized PSF formulation was taken as feed solution. All the parameters of electospinning and UV curing were thoroughly investigated, optimized and correlated so the optimum curing of acrylic monomer and production of nano fibers take place simultaneously. The electrospun mats were also prepared via offline curing method and compared with the counterparts. All the UV cured mats were fully characterized and solvent stability was assessed in the same way as reported before.

Le travail de thèse présenté dans ce manuscrit a pour objectif de mettre en place un nouveau procédé d’élaboration de membranes à base de chitosane pour le traitement d’effluents acides contenants des ions métallique. Soluble en milieu aqueux acide, le chitosane présente la propriété de gélifier lorsque le pH devient basique, ce qui permet d’envisager l’élaboration de membrane sans solvant organique contrairement aux polymères synthétiques classiques. Par ailleurs, ce travail de thèse s’est intéressé à un procédé original de gélification par voie enzymatique, dans lequel la gélification in-situ de la solution de chitosane permet une structuration contrôlée de la membrane contrairement aux procédés classiques qui donnent lieu à un front de gélification. Une étude des cinétiques de gélification en fonction des paramètres d’élaboration (température et concentration en urée) a mis en évidence que seule la température est significativement influente sur le temps de gélification dès lors que la concentration en urée n’est pas limitante. Un modèle a été mis en place pour décrire la gélification enzymatique du chitosane afin de comprendre les mécanismes des cinétiques réactionnelles et de transferts lors de la formation du gel. Des membranes de chitosane ont ainsi été élaborées par le procédé par voie enzymatique, la porosité de ces membranes ayant été générée avec un agent porogène (PEG 6000) et une réaction d’acétylation du chitosane ayant permis d’obtenir des membranes insolubles en milieu acide. Les membranes obtenues ont été caractérisées en termes de morphologie et de propriétés fonctionnelles (filtration, sorption du Cu(II) comme élément métallique modèle)
The Ph-D work presented in this manuscript aims to develop a new process for elaborate chitosan membranes for treatment of acidic media containing metal ions. Soluble In acidic aqueous media, gelation occurs when the pH becomes basic, allowing elaboration of membrane without the use of organic solvents unlike classical synthetic polymers. Moreover, this Ph-D work has focused on an original process enzymatic gelation which the in-situ gelation of chitosan solution allows a controlled structuration of the membrane unlike conventional processes which leads to a front gelling.A study of gelation time as a function of the elaboration parameters (temperature and urea concentration) highlighted that only the temperature is a main parameter on gelation time since the urea concentration is not limiting. A model was developed to describe the chitosan enzymatic gelation in order to understand mechanisms of reaction kinetics and transfers during the gel formation.Chitosan membranes have been prepared by enzymatic process, the porosity of such membranes have been generated with a blowing agent (PEG 6000) and an acetylation of chitosan having yielded insoluble membranes in acid medium. The resulting membranes were characterized by their morphology and functional properties (filtration, sorption of Cu (II) as model metal element)

Le procede d'ultrafiltration est limite dans son utilisation par ses caracteristiques propres. Ainsi il est impossible de pretendre a un transport selectif d'ions de petites tailles. D'autre part, lors d'une filtration d'un fluide contenant des macrosolutes, il se forme un encrassement fortement penalisant au niveau des flux de permeat. Le double objectif presente dans ce memoire, repond a ces limitations en utilisant le procede d'electro-ultrafiltration. A partir du pilote que nous avons realise, nous mettons en evidence qu'il est possible d'obtenir une epuration ou une concentration de solution contenant des nitrates, ainsi que de proceder au decolmatage de membranes minerales en zircone possedant un support macroporeux en carbone. Dans une premiere partie, le procede d'electro-ultrafiltration est presente. Sa particularite est d'utiliser le support de la membrane comme electrode externe. Une tige de carbone ou de platine placee coaxialement a la membrane tubulaire permet d'obtenir dans la solution des champ electriques eleves sous de faibles differences de potentiel. Le montage concu est utilise sur des solutions de nitrates permettant d'observer les effets de differents parametres : contre ion associe au nitrate, seuil de coupure des membranes, difference de pression transmembranaire, vitesse d'ecoulement de la charge. Il est mis en evidence la possibilite de concentrer des solutions en ions ou de les epurer suivant la polarite de la membrane. Un bilan des especes presentes dans le permeat permet de relier la concentration en ions nitrates a la valeur mesurable du ph accessible par mesure. Enfin, la troisieme partie concerne la possibilite de decolmater des membranes encrassees par une filtration de lait. La generation de gaz, par electrolyse, dans le support macroporeux de la membrane chasse les molecules limitant le flux du permeat. L'operation dure quelques minutes et permet un nettoyage aussi efficace qu'un cycle chimique industriel de plusieurs heures. D'autre part l'effet d'un champ electrique applique durant la phase de filtration permet d'augmenter de facon sensible le debit du permeat, le taux d'azote total variant peu.

Cette étude traite du colmatage des membranes basse pression par les matières organiques des eaux naturelles. Le premier objectif a été d'évaluer la capacité d'unités pilotes à apprécier le potentiel colmatant. Ces unités sont alimentées par de l'eau renouvelée quotidiennement ce qui permet de se rapprocher des conditions d'exploitation sur usine. Les profils de perte de flux enregistrés sur plusieurs campagnes d'essais sur site ont permis de valider l'intérêt de notre outil expérimental. Le second objectif portait sur la caractérisation du colmatage de nature organique. Une grande diversité de techniques analytiques a été développée. Quel que soit la membrane, le matériel colmatant englobe des substances humiques, des substances protéiques, des sucres aminés (parois microbiennes) et des polysaccharides. Les résidus microbiens composantes ubiquistes des eaux de surface et structures résultant d'un possible développement bactérien à la surface des membranes semblent constituer la source prépondérante du matériel colmatant
This project relates to NOM fouling of low pressure membranes. The first objective was to test the capacity of small pilot units to evaluate the fouling properties of natural and treated waters. Pilot testing was conducted on site (drinking water treatment plants). Flux decline recorded during several campaigns with different water qualities confirmed the efficacy of our pilot units to identify fouling problems based on a short time experiment. The second objective was to characterize the organic foulants using a large set of analytical tools. Our results showed the need to use both membrane surface autopsy and water quality characterization to identify membrane foulants. NOM foulant was found to be a mixture of proteins, polysaccharides, polyhydroxyaromatics and aminosugars. Residus of microorganisms (membrane cell walls, exopolymers) seem to play a major role in membrane fouling, phenomenon enhance in the humic type NOM

Ziel der Arbeit ist die Beschreibung des Membranverhaltens bzw. der Membraneigenschaften bei der Regeneration von Extraktionsmitteln der Metallextraktion durch Trennung von beladenen und unbeladenen Extraktionsmitteln mittels Ultrafiltration. Dazu wurden die Trennleistungen als Funktion von Druck, Temperatur und Solventzusammensetzung für verschiedene Membranmaterialien und -typen untersucht. Anhand von Sorptionsversuchen konnte die Löslichkeit der Komponenten der organischen Phase in den verschiedenen Polymermatrizen von Membranen aus regenerierter Cellulose bzw. Polyethersulfon und die adsorptive Gelschichtbildung als Funktion der Solventzusammensetzung, der Temperatur und des Membranmaterials aufgezeigt werden. Der Einfluss der Sorption, des Quellens und Kompaktierens auf das Materialverhalten konnte mittels einer Hochdruckmikrowaage und eines Laserscanner-Messstandes ermittelt werden. Die durch Veresterung und Wasserstoffbrückenbindung entstehende Gelschicht auf der Membranoberfläche, die letztendlich für die Rückgewinnung von unbeladenem Extraktionsmittel ausschlaggebend ist, kann mit Hilfe der Röntgenphotoelektronenspektroskopie, Rasterelektronen- und Rasterkraftmikroskopie nachgewiesen werden.

The project was concerned with studying the capability of a micellar-enhanced ultrafiltration system (MEUF) to remove platinum group metal ions namely Pt (lV) and Pd (ll) chloro anions from aqueous industrial waste effluents. South Africa has the world’s largest reserves of platinum group metals (PGMs) and other valuable metals such as manganese, chrome ores, titanium minerals etc. which are required for new automotive and other technologies, including fuel cells, catalytic converters and lighter components. The consistent loss with the industrial waste stream and the toxicological effects of these precious metals led to the need to develop new and effective methods to recover them from industrial waste effluents. With such a wide variety of fields where these PGMs are used and the failure of the traditional techniques namely sedimentation, fermentation etc. to effectively reduce or recover these highly toxic and precious metal ions prior to discharging industrial waste effluents, it is necessary to explore other techniques such as membrane technology that can be used to recover these valuable species from industrial waste streams. The present study involved the use of a cationic surfactant, viz cetylpyridinium chloride, which was introduced into an aqueous solution containing palladium and platinum metal anions. The surfactant forms charged micelles above a certain critical concentration value. The metal anions adsorb onto the available oppositely charged sites on the micelle surfaces and are then able to be retained by a suitable membrane. Hollow fibre ultrafiltration membranes with the MWCO of +/- 10 kD and +/-30nm pore size were used as a filter component in this study. For this MEUF system to be effective, it was vital that the anionic metal ion species adsorbed sufficiently onto the available oppositely charged sites of the micelles and that the micelles were retained efficiently by the membrane. Results obtained during the investigation made it possible to make certain predictions about the micellisation process. It was also found that, it was not only the metal ion: surfactant (M:S) ratio that was critical, but the presence of other electrolytes in the aqueous stream proved to have a huge impact on the capability of the MEUF system. Findings of this research study showed that the MEUF system using cetylpyridinium chloride (CPC) can be used to recover or retain Pt (lV) and Pd (ll) anions from industrial waste effluents. It was also found that PtCl6 2-, due to its greater adsorption capabilities onto the micelle surface than PdCl4 2- or PdCl3(H2O)-, was preferentially retained in neutral medium. This may be exploited as a possible means of separating the two metal ions. The developed system offers the following advantages over some traditional and current methods: simplified unit operation line flow process, smaller amounts of chemical usage and no solid toxic sludge to be disposed of. Applications of this work could be of vital importance in catalytic converter recycling, especially in Port Elizabeth where extensive automobile parts manufacturing occurs.

Concerns over possible waterborne disease forced drinking water supply companies in England and Wales to adopt microfiltration and ultrafiltration technologies rapidly. MF and UF membrane plants are designed to produce water of a consistent quality regardless of throughput and fluctuations in the feedwater quality. To operate well they need to maintain flux and balance the rate of fouling, and chemical cleaning performance is critical to this. Giant steps have been taken into characterizing the foulants scientifically in the last few years while cleaning is reactive and ad hoc. This thesis explores the basis for a corresponding cleaning science for the technology to develop quantitively. Cleaning performance was defined in terms of a response to combinations of explanatory variables in a materials limited cleaning envelope. The study focused on applying variations of cleanant concentration, applied temperature and soak times to a variety of membranes fouled with different waters and regimes. An experimental design was developed and applied consistently to a number of different sampled sites; allowing an optimised recovery from the polynomial expressions for each treatment, through factorial analysis of the data. The size and variety of the data set analysed allowed comparison and quantification of the different deviations from optimal cleaning response. This effect was seen to vary temporally and with operating regime and the methods usefulness as a practical tool in the membrane plant lifecycle was considered. Cost evaluation of the variation in cleaning response showed that sub-optimal cleaning costs and energy use may be significant and the thesis also illustrated how module geometry affects initial cake deposition and thus cleanability. By demonstrating the potential for cleaning factor analysis, the potential for a combined heuristic and predictive cleaning control science is possible, but will need new strategies to manage technology change.

Lime softening, chlorination, clarification and filtration have been long recognized treatment processes for beverage water specifically the carbonated soft drink (CSD) because it provides consistent water quality required for bottling plants, however these processes are becoming uneconomical and causes more problems than the benefits they offer. These processes require very large foot print, occupy large plant volume, and generate large volume of sludge which causes disposal problems. Chlorination produces trihalomethanes (THMs) and other by-products which are detrimental to health and imparts tastes to the final products. Using the newly developed submerged spiral wound ultrafiltration membranes in conjunction with lime softening may replace the conventional lime softening, clarification and filtration processes. This research was conducted to demonstrate the feasibility of integrating immersed ultrafiltration (UF) membrane with lime softening. The objectives of this research was to achieve the water quality required by the CSD bottlers; determine the relationships of operating parameters such as pH and membrane flux with trans-membrane pressure (TMP), and membrane permeability; determine the optimum dosage of lime; evaluate the operating parameters as basis for the design and construction of the full scale plant; and predict the membrane cleaning intervals. A pilot unit consisting of lime reactor and UF system was designed and built for this research. The pilot unit was operated at various pH ranging from 7.3 to 11.2 and at membrane flux rates of 15, 30 and 45 gfd. The pilot unit was also operated at the CSD bottler’s operating conditions which is pH 9.8 at flux of 30 gfd. The pilot unit operated for a total of 1800 hours. The raw water source was from city water supply. The filtrate from the pilot unit achieved alkalinity reduction to 20 to 30 mg/L preferred by CSD bottlers, with lime dosage close to the calculated value. The filtrate turbidity during the test was consistently within 0.4 to 0.5 NTU. The TMP values obtained during the test ranges from 0.1 to 2.5 psi, while the permeability values ranges from 18.19 to 29.6 gfd/psi. The increase in flux results to corresponding increase in TMP, and increase in operating pH, increases the rate of TMP. Permeability decreases with increasing operating pH. The TOC reduction ranges from 2.6 % to 15.8% with increasing operating pH. No scaling of the UF membranes was observed during the test. Thirty days UF membrane cleaning interval was predicted. The results from this research can use as the basis of designing and operating a full scale Lime Softening UF Treatment Plant.

Le problème abordé dans ce travail est celui de la mise au point d'un protocole raisonné de détermination de l'abattement du nombre de virus par une membrane de filtration, dans des conditions proches de celles que l'on rencontre dans le domaine du traitement de l'eau. Pour cela, ont été recherchées parmi les conditions opératoires, celles qui conduisent au transfert le plus important, de manière à révéler d'éventuelles faiblesses dans les matériaux testés. Au cours de cette étude, les phages MS2 et Qbeta ont été choisis comme indicateurs. Leurs comportements en suspension dans divers solvants, en statique aux interfaces puis en dynamique en cours de filtration ont été étudiés. Le protocole mis au point a été ensuite testé à l'échelle du laboratoire sur diverses membranes intègres, sur des membranes présentant un défaut (trou) et enfin à l'échelle pilote sur des installations mises à notre disposition par deux industriels. Les méthodes de détection et de quantification employées et comparées sont le dénombrement après mise en culture et la RT-PCR quantitative avec et sans extraction. Ces méthodes permettent au sein d'un prélèvement l'identification des différentes formes sous lesquelles les particules virales se trouvent : infectieuse, désactivée et détériorée
Problem addressed in this work is the development of a rational protocol for determining the reduction of viruses by membrane filtration in conditions close to those encountered in water treatment. To do this, were researched among the operating conditions, those leading to the largest transfer, in order to reveal possible weaknesses in the materials tested. In this study, bacteriophages MS2 and Qbeta were chosen as indicators. Their behaviors in suspension in various solvents, in static during contact with interfaces and in dynamic during filtration were studied. The protocol developped was then tested in various laboratory scales using virgin membrane and through membranes with one defect (hole). Finally the validation of the protocol has been conducted on pilot scale facilities supplied by two manufacturers. The detection and quantification methods used and compared are the cell culture (PFU method) and RT-PCR with and without extraction. These methods allow in a sample the determination of the various forms in which viral particles are likely to be present: infectious, disactivated and broken

The interest in the design of portable and wearable medical devices is related to both the relevant clinical and social benefits for patients and the potential economic savings for national health services. Biomedical technologies are improving at a very fast rate and represent an extraordinary means to develop innovative portable and wearable devices which can help people live in a prosperous way, in particular reducing sorrow in case of disease. This leads to a widespread effort to develop devices which can execute at home therapies that are usually performed in hospitals. This thesis presents a new wearable and portable device for extracorporeal blood ultrafiltration, named WUF (Wearable UltraFiltration device), able to remove excess fluids from fluid overload patients with chronic kidney disease and/or congestive heart failure. The design requirements that a modern wearable device for extracorporeal ultrafiltration must meet have been identified thanks to a thorough literature review on previous similar proposals followed by an extensive risk analysis. The design of the WUF prototype has faced several difficulties, ranging from the identification or conceivement of safe and reliable components to the design of a compact and neat layout. For most components it was possible to identify commercial (off-the-shelf) products meeting the requirements, nonetheless for some others, specific investigations, studies and developments were needed and led to the design of customized solutions or the formulation of original approaches. The design of an effective, efficient, safe and reliable control architecture, based on two microcontrollers and one microcomputer, the implementation of the control logic and of a graphical user interface have been carried out too being essential features of such a mechatronic device. A backpack/trolley design has been chosen as the layout for the device, since such a solution guarantees the best tradeoff between miniaturization and ergonomics. The design introduces an original positioning of the vast majority of components in three independent planar panels: one for disposable components, one for non-disposable devices and one for electronic boards and controllers. This arrangement of components can drastically simplify and speed up the in-hospital operations needed before and after a therapy with the WUF.
L'interesse per la progettazione di dispositivi medici portatili e indossabili è legato sia ai vantaggi clinici e sociali rilevanti per i pazienti, sia ai potenziali risparmi economici per i servizi sanitari nazionali. Le tecnologie biomediche stanno migliorando a un ritmo molto rapido e rappresentano un mezzo straordinario per sviluppare dispositivi portatili e indossabili innovativi che possano garantendo ai pazienti una migliore qualità di vita. Ciò implica uno sforzo per sviluppare dispositivi che possono eseguire terapie domiciliari che vengono solitamente eseguite negli ospedali. Questa tesi presenta un nuovo dispositivo indossabile e portatile per l'ultrafiltrazione extracorporea del sangue, denominato WUF (Wearable UltraFiltration device), in grado di rimuovere i liquidi in eccesso da pazienti con sovraccarico di liquidi con malattia renale cronica e / o insufficienza cardiaca congestizia. I requisiti di progettazione che un moderno dispositivo indossabile per l'ultrafiltrazione extracorporeo deve soddisfare sono stati identificati grazie a un'accurata revisione della letteratura, basata su simili proposte precedenti, e da un'estesa analisi dei rischi. Il design del prototipo WUF ha incontrato diverse difficoltà, a partire dall'identificazione o concezione di componenti sicuri e affidabili, alla progettazione di un layout compatto. Per la maggior parte dei componenti è stato possibile identificare prodotti commerciali che hanno soddisfatto i requisiti, tuttavia per altri sono stati necessari studi specifici che hanno portato alla progettazione di soluzioni customizzate o alla formulazione di approcci originali. La progettazione di un'architettura di controllo efficace, efficiente, sicura e affidabile, basata su due microcontrollori e un microcomputer, l'implementazione della logica di controllo e di un'interfaccia grafica per l’utente sono state altresì sviluppate, essendo caratteristiche essenziali di un dispositivo meccatronico. Un design a zaino/trolley è stato scelto come layout per il dispositivo, poiché tale soluzione garantisce il miglior compromesso tra miniaturizzazione ed ergonomia. L’innovativo design introduce un posizionamento originale della maggior parte dei componenti in tre pannelli planari indipendenti: uno per componenti monouso, uno per dispositivi non monouso e uno per schede elettroniche e microcontrollori. Questa disposizione dei componenti può drasticamente semplificare e accelerare le operazioni in ospedale necessarie prima e dopo la terapia con il dispositivo WUF.

Ce memoire est une contribution a l'etude d'une nouvelle technique de lutte contre le colmatage en filtration tangentielle. Le principe repose sur l'utilisation d'ecoulements secondaires, appeles vortex de dean, qui apparaissent dans un module helicoidal ou des fibres creuses sont enroulees en helice. Dans une premiere partie, l'analyse de l'evolution du colmatage des membranes lors de l'ultrafiltration de suspensions de levures de boulangerie fait apparaitre l'importance des differentes fractions constituant la suspension. Le colmatage irreversible (adsorption et formation d'un gel) est attribue aux composes extracellulaires presents dans le milieu alors que le colmatage reversible est essentiellement du a la formation d'un gateau de levures. En comparant les flux limites obtenus dans un module droit et helicoidal, a meme depense energetique specifique, des gains compris entre 2 et 5 ont ete mesures. Les gains les plus eleves sont observes en regime laminaire et augmentent avec la contrainte de cisaillement a la paroi et avec la concentration. Le flux limite de permeat est relie par une loi semi-empirique a la contrainte moyenne de cisaillement a la paroi. La correlation ainsi etablie est verifiee pour la filtration tangentielle de differents fluides simples (bentonite, dextrane) ou complexes (levures, biere) quelle soit la geometrie du module. Les parametres lies a la conception du module sont tous rassembles dans la valeur de la contrainte parietale : sa determination est primordiale et une methode electrochimique fiable a ete developpee pour la quantifier localement. L'etude de la filtration de biere de garde confirme, sur un fluide industriel, l'interet potentiel des ecoulements de dean sur l'augmentation des flux de permeat et la transmission de composes a l'origine de la qualite de la biere.

In Micellar-Enhanced Ultrafiltration (MEUF), a surfactant is added to a wastewater stream containing an organic solute. Most of the solute is solubilized in micelles, which are aggregates of surfactant molecules, formed in the feed solution. Separation is achieved through retention of the micelles by an ultrafiltration membrane. Since much of the solute is within the micelles, a membrane with a molecular weight cut off (MWCO) much larger than the molecular weight of the solute is used. Although some unsolubilized solute and surfactant monomers pass through the membrane, the use of a surfactant with a low critical micelle concentration (CMC) ensures little contamination of the permeate by surfactant monomers.
Solubilization tests and MEUF experiments were performed with an aqueous feed solution of benzoic acid (molecular weight 122) and Rhodameen T12/90 (molecular weight 798), a twin-head, cationic tertiary amine surfactant. This surfactant has an 18 carbon hydrophobic tail and two hydrophilic polyoxyethylene heads. The equilibrium solubilization of benzoic acid by surfactant solutions was measured by semiquilibrium dialysis. The maximum solubilization was nearly one mole of benzoic acid per mole of Rhodameen.
MEUF was carried out in hollow fiber polysulfone membrane units with MWCO of 5,000 and 30,000. The permeate flux and overall rejection of benzoic acid and surfactant were measured for different initial feed compositions and initial surfactant concentrations under varying transmembrane pressures (TMP). The flux increased with TMP and decreased with Rhodameen concentration, but was not affected by the benzoic acid concentration. The highest rejections of benzoic acid (90%) and Rhodameen (95%) were obtained when the molar ratio of surfactant to solute in the feed was 1.2.

The influence of physiochemical conditions on the fouling and fractionation of charged solutes, particularly proteins, during ultrafiltration has been investigated in this study. There exists great scope for improvement of ultrafiltration selectivity and efficiency through manipulation of pH, ionic strength and pressure. With a knowledge of the inherent physiochemical properties of the (bio)colloids and membranes one can apply colloid science insights to model interactions taking place a the membrane surface. Charged solutes are driven towards the membrane pores by hydrodynamic forces. As they approach the pore entrance they usually experience repulsive electrostatic forces. An important concept to be discussed is that of the 'critical pressure', the applied pressure at which the hydrodynamic forces and repulsive electrostatic forces are exactly balanced. This is calculated using finite element techniques. If pressures below the critical pressure are applied the solute will not come into intimate contact with the membrane. Solutes will be retained by the membrane at pressures below the critical pressure, and at pressures above the critical pressure they may be transmitted through the membrane. Experiments with dead-end ultrafiltration of colloidal silica through fully retentive membranes have tested the model and allowed study of electrostatic effects in ultrafiltration in detail. Filtration and fractionation experiments have also been conducted with the proteins bovine serum albumin (BSA), hemoglobin and lactoferrin using polycarbonate ultrafiltration membranes. These proteins have a similar size but different physiochemical properties, which were measured experimentally. Results for single protein transmission and mixed protein fractionation show good agreement with the model predictions. The work has tested an ab inititio based model to predict the rejection and non-fouling conditions in which to run a membrane separation. It has highlighted the additional important effects of protein-protein interactions in the concentration polarisation layer, with which the model's predictive capabilities could be furthered.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Chemical Engineering, 1994.
Includes bibliographical references (leaves 244-254).
by Maria Cláudia Ferreira Drumond de Sousa.
Ph.D.

Thesis (MScEng (Process Engineering))--University of Stellenbosch, 2009.
ENGLISH ABSTRACT: Membrane fouling is universally considered to be one of the most critical problems in the wider application of membranes in filtration separation. Fouling is caused by the deposition of particles not only on the surface of the membrane, but also inside the membrane pores, which reduces permeate flux and leads to a reduction of the efficiency and the longevity of the membrane. The backpulsing cleaning method can be used to remove deposited foulants from the surface of the membrane, without having to shut down the plant. Ultrasonic time-domain reflectometry (UTDR) is a nondestructive technique, used to detect and measure the growth of fouling layer on the membrane surface during microfiltration and ultrafiltration processes. In this study flat-sheet microfiltration (MF) and ultrafiltration (UF) membranes were fouled during a cross-flow filtration processes using dextrin, yeast or alumina (feed pressure 100 kPa and feed flow rate 0.45 liter/minute), in a flat cell. Infrasound frequency backpulsing, in the permeate space, was used to clean the membranes. Backpulsing was carried out using the permeate water or soap solutions. The peak pressure amplitude of the pulses used to clean the membranes was 140 kPa, the pulsing was applied at a frequency of 6.7 Hz. The main objectives of this research were: (1) to obtain a fundamental understandimg of how foulants deposit on membrane surfaces and how the foulant deposits can be removed using the backpulsing cleaning technique during MF and UF, (2) to use the ultrasonic measurement technique for monitoring the growth and removal of the fouling layer on the membrane surface and (3) Use scanning electron microscopy (SEM) as a direct measurement technique to analyze the structure the foulant deposits on membrane surfaces before and after cleaning. Results showed that a flux value of between 55% and 98% of the clean water flux value can be achieved by backpulsing cleaning. UTDR was successfully applied to monitor membrane cleaning and provide information about the growth and removal of fouling layers on the membrane surface.
AFRIKAANSE OPSOMMING: Membraanaanvuiling is wêreldwyd bekend as een van die mees kritieke probleme wat die wyer aanwending van membrane vir skeidingsprosesse benadeel. Aanvuiling word veroorsaak deur die deponering van partikels, nie net op die oppervlak van die membraan nie, maar ook binne-in die membraanporieë, wat die volgende tot gevolg het: 'n afname in vloed deur die membraan, 'n afname in die effektiwiteit van die membraan, en 'n korter membraanleeftyd. Die teenpulsskoonmaakmetode kan gebruik word om die aanvuilingslaag vanaf die membranoppervlakte te verwyder sonder dat dit nodig is om die membraantoetsapparaat af te skakel. Ultrasoniese-tydsgebied-weerkaatsing (UTW) is 'n nie-vernietigende tegniek wat gebruik kan word om die groei van 'n aanvuilingslaag op 'n membraanoppervlakte tydens mikrofiltrasie (MF) of ultrafiltrasie (UF) te identifiseer en te meet. In hierdie studie is plat-vel MF en UF membrane bevuil gedurende 'n kruisvloeifiltrasieproses deur gebruik to maak van dekstraan, gis of alumina, in 'n plat sel. Infraklank-frekwensieteenpols, in die permeaatgebied, is gebruik om die membrane skoon te maak. Hiervoor is die proseswater of 'n seepoplossing gebruik. Die maksimum drukamplitude van die pulse wat gebruik is was 140 kPa, en die puls was aangewend teen 'n frekwensie van 6.7 Hz. Die hoofdoelwite van hierdie studie was die volgende: (1) om inligting in te win oor hoe aanvuilingsmateriale op membraanoppervlaktes gedeponeer word tydens MF en UF en hoe hulle verwyder kan word deur gebruik te maak van die teenpulsskoonmaaktegniek; (2) om van die teenpulsskoonmaaktegniek gebruik te maak om die groei van die bevuilingslaag asook die verwydering daarvan op die membraanoppervlakte te monitor; en (3) om van skandeerelektronmikroskopie (SEM) as 'n direkte analitiesetegniekgebruik te maak om die struktuur van die aanvuilingsmateriaal voor en na die die skoonmaakproses te analiseer. Deur gebruik te maak van teenpulsskoonmaak kon die membraanvloed tot tussen 55–98% van die oorspronklike suiwerwatervloed verbeter word. Sodoende is ultrasoniese-tydsgebiedweerkaatsing suksesvol gebruik om die skoonmaak van membrane te monitor asook om inligting in te win i.v.m. die groei en verwydering van die aanvuilingslae op die membraanoppervlaktes.