Influence of inclination on gas-sparged cross-flow ultrafiltration through an inorganic tubular membrane

The effect of membrane inclination on the flux of single-phase or gas–liquid two-phase ultrafiltration in a tubular membrane has been investigated. Experimental result shows that membrane inclination has a significant enhancement on the flux of two-phase ultrafiltration operated at slug flow pattern. As the angle of inclination from the horizontal increases, the flux increases, reaches a maximum, and then decreases. The flux may be enhanced more than 1.5 when the membrane is inclined from 0 to 50°. The flux enhancement due to membrane inclination increases with increasing the gas velocity, the feed concentration, and the transmembrane pressure, while it decreases with increasing the liquid velocity. The optimal inclination angle of the membrane in a slug-flow ultrafiltration is close to 50°. An equation for determining the optimal inclination angle was also proposed in this work.     

How slug flow can enhance the ultrafiltration flux in mineral tubular membranes

The study deals with the use of a gas-liquid two-phase flow to reduce tubular mineral membrane fouling by injecting air directly into the feed stream. The injected air is supposed to create complex hydrodynamic conditions inside the ultrafiltration module which destabilize the concentration layer over the membrane surface. The experimental study was carried out by filtering suspensions (bentonite and yeast) through an ultrafiltration tubular mineral membrane. A range of transmembrane pressures and various liquid and gas flow-rates were tested. Results related to the permeate flux showed an enhancement by a factor of 3, with a slug flow-structure for the two kinds of suspension (200% of flux increase). Furthermore, the applicability of such an unsteady technique was examined with a view to reduce energy consumption.     

Effect of colloids on salt transport in crossflow nanofiltration

The role of colloid deposition on the performance of a salt-rejecting NF membrane was evaluated by modeling salt transport using a two-layer transport model, which quantified the relative contributions of advection and diffusion in the cake and the membrane layers, and the effects of flux on the membrane sieving coefficient. The model was able to accurately describe how the measured permeate concentration, rejection, osmotic pressure, and flux decline varied with time. The two-layer model confirmed that the Peclet number in the cake layer was about an order of magnitude higher than that in the membrane layer, leading to significant concentration polarization at the membrane surface, as shown by others. However, the cake layer also increased overall resistance, which resulted in flux decline during constant pressure operation. Flux decline caused an increase in the actual sieving coefficient, leading to higher solute flux, lower observed rejection, and thus lower the bulk concentration. These coupled phenomena tended to mitigate the increase in concentration polarization caused by the cake. Therefore, as predicted by the model and verified by experiment, the osmotic pressure does not increase monotonically as the cake grows, and in fact can decrease when the cake layer is thick and the flux decline is significant. In our experimental system, the pressure drop across the cake layer, which was proportional to the cake thickness, was significant under the conditions studied. The effects of cake-enhanced osmotic pressure analyzed here are lower than those observed in previous studies, possibly because the transport model employed explicitly accounts for the effect of flux decline due to cake growth on the membrane sieving coefficient, and possibly because we used a somewhat different methodology to estimate cake porosity.     

Effect of operating variables on the flux and selectivity in sweep gas membrane distillation for dilute aqueous isopropanol

Sweep gas membrane distillation was examined as a possible technique for isopropanol (IPA)–water separation using PTFE hollow fiber membrane module. The composition and flux of the permeate were monitored when feed concentration, operating temperature and flow rate were varied. The upper feed concentration tested was 10 wt.% IPA. Within the feed temperature range of about 20–50°C, IPA selectivity of 10–25 was achieved. Since the concentration near the surface on the membrane increased by the selective adsorption of IPA on the hydrophobic membrane, the selectivity increases. The permeate flux and IPA selectivity increase as feed temperature increase. The flux and selectivity increase at higher flow rates is mainly due to the reduced effects of concentration and temperature polarization. The effect of salt addition to the feed mixture was also examined.     

Hydrodynamics and mixing in three-phase fluidized bed bioreactors

The effects of solid particles on hydrodynamics and mixing in a three-phase fluidized bed bioreactor were discussed. The gas holdup, bubble size, and liquid-phase axial dispersion coefficient were measured in a 0.25-m id bubble column bioreactors containing low-density particles. The presence of low-density solid particles slightly increased gas holdup. The decrease in average bubble diameter with solid presence was found. For the three-phase system, the liquid-phase axial dispersion coefficients were higher than for the two-phase system. We extended a model for a gas holdup developed for a gas-liquid two-phase bubble column bioreactor to a gas-liquid-solid three-phase fluidized bed bioreactor. Using the present data and available data in the literature, the predictions of the proposed model were examined. The proposed model agreed with a wide range of the experimental data. A theoretical correlation for liquid-phase axial dispersion coefficient was developed using Kolmogoroff's theory of isotropic turbulence. Reasonable agreement was obtained between the predicted and experimental values of axial dispersion coefficient.     

Quantitative measurements of the concentration polarisation layer thickness in membrane filtration of oil-water emulsions using NMR micro-imaging

In this paper we report measurements of the thickness of the concentration polarisation layers formed during crossflow membrane filtration of an oil-water emulsion. The formation and development of the oil polarisation layers was visualised non-invasively using NMR chemical shift selective micro-imaging. A series of images was acquired during the transient state of the filtration, (i.e. while the polarisation layer was forming and the flux of filtrate was changing), prior to the establishment of steady state conditions. An estimate of the specific resistance of the concentration polarisation layers was then obtained by determining the average oil layer thicknesses and concentration at a given time from the resulting images and measuring the corresponding (length averaged) flux of filtrate gravimetrically. After the establishment of steady state conditions, the dependence of the steady state filtrate flux on crossflow Reynolds number was found to be consistent with Brownian diffusion being the main mechanism controlling the build-up of the oil polarisation layers, at least under our range of operating conditions.     

Modelling the clarification of lactic acid fermentation broths by cross-flow microfiltration

This work is focused on modelling microfiltration for clarifying fermentation broths for the production of lactic acid. The hydraulic resistance-in-series model was used with membrane resistance, bacterial cell cake resistance, adsorption resistance and solute concentration polarisation resistance. Most of the model parameters were determined from independent experiments. This model was applied for microfiltrations operated either under constant transmembrane pressure or under constant permeate flux. Resistances due to adsorption and to solute concentration polarisation dominated. Bacterial cake resistance was found to be very low or equal to zero when microfiltration was below the critical flux.     

How slug flow can improve ultrafiltration flux in organic hollow fibres

The study deals with the use of a gas-liquid two-phase flow to reduce particle membrane fouling in organic hollow fibres by injecting air directly into the feed stream. A theoretical approach of slug flow in fibres demonstrates that the slugs created inside the fibres induce high wall shear stresses. Moreover, the membrane surface is alternately submitted to positive and negative shear stresses. This succession of stresses is expected to prevent filtered particles from settling on the membrane surface and then enhance the ultrafiltration mass transfer. Experiments were carried out with clay suspensions in hollow fibre membrane. A range of various air velocities and particle concentrations was examined and the effect of a steady gas flow was compared to that of an intermittent one. As expected, the injecting air process leads to an increase of the permeate flux by up to 110% for U_g=1 m s⁻¹ (flux multiplied by 2.1), for all the various concentrations studied. Furthermore, even at a low air velocity a significant enhancement can be achieved (+60% for U_g=0.1 m s⁻¹, flux multiplied by 1.6). An intermittent gas flow seems to be less effective than a steady one in similar experimental conditions.     

Concentrating the extract of traditional Chinese medicine by direct contact membrane distillation

This study applies direct contact membrane distillation (DCMD) to concentrating the extract of traditional Chinese medicine (TCM). The trans-membrane flux under various operation conditions was measured in real-time during concentration process. By decoupling the factors affecting the trans-membrane flux decline, it was found that the observed flux decline throughout the process could be attributed to the membrane fouling, the reduction of water vapor pressure and the increase of transport resistance at feed side. Analysis of the combined factors was given to show in detail the mechanism of flux decline. Factors that may affect the flux level, such as feed velocity, feed temperature and pretreatment were experimentally examined. Gas bubbling or sparging was introduced into DCMD system for reducing membrane fouling, and it was found that both gas–liquid two-phase flow at the feed side and gas back-washing within membrane module are effective ways to control membrane fouling.     

The effect of the physical properties of the liquid phase on the gas-liquid mass transfer coefficient in two- and three-phase agitated systems

The results of studies concerning two- and three-phase systems in an agitated vessel are presented. The aim of our research was to investigate the effect of the physical properties of the liquid phase on the value of the volumetric gas-liquid mass transfer coefficient in mechanically agitated gas-liquid and gas-solid-liquid systems. Our experimental studies were conducted in a vessel with an internal diameter of 0.288 m. The flat bottom vessel, equipped with four baffles, was filled with liquid up to a height equal to the inner diameter. The liquid volume was 0.02 m³. Three high-speed impellers of a diameter equal to 0.33 of the vessel diameter were used: Rushton turbine (RT), Smith turbine (CD 6), or A 315 impeller. The measurements were carried out in coalescing and non-coalescing systems. Distilled water and aqueous solutions of an electrolyte (sodium chloride) of two different concentrations were used as the liquid phase. The gas phase was air. In the three-phase system, particles of sea sand were used as solid phase. The measurements were conducted at five different gas-flow rates and three particle loadings. Volumetric gas-liquid mass transfer coefficients were measured using the dynamic method. The presence and concentration of an electrolyte strongly affected the value of the gas-liquid mass transfer coefficient in both two- and three-phase systems. For all agitators used, significantly higher k _l a coefficient values were obtained in the 0.4 kmol m⁻³ and 0.8 kmol m⁻³ aqueous NaCl solutions compared with the data for a coalescing system (with distilled water as the liquid phase). The k _l a coefficient did not exhibit a linear relationship with the electrolyte concentration. An increase in the sodium chloride concentration from 0.4 kmol m⁻³ to 0.8 kmol m⁻³ caused a considerable decrease in the volumetric mass transfer coefficient in both the two-phase and three-phase systems. It was concluded that the mass transfer processes improved at a certain concentration of ions; however, above this concentration no further increase in k _l a could be achieved.     

Nature of flow on sweeping gas membrane distillation

The process of sweeping gas membrane distillation (SGMD), with the liquid feed and the sweeping gas counterflowing in a plate and frame membrane module, has been studied. A theoretical model, which was presented in a previous paper and permitted to obtain the temperature profiles inside the fluid phases, has been developed in order to analyse the physical nature of the transmembrane water flux. Two porous hydrophobic membranes have been studied in different experimental conditions. The influence of some relevant parameters, such as the inlet and outlet temperatures or the circulation velocities of the fluids, has been studied. The experimental results have been analysed according to the model and the conclusion is that the water transport takes place, apparently, via a combined Knudsen and molecular diffusive flow mechanism. From the temperature profiles, a local temperature polarisation coefficient may be defined. From this local value, an overall one for the whole system is then defined. The new theoretical predictions have been applied to the obtained results and the accordance may be considered good.     

Free-solvent model shows osmotic pressure is the dominant factor in limiting flux during protein ultrafiltration

In protein ultrafiltration (UF), the limiting flux phenomenon has been generally considered a consequence of the presence of membrane fouling or the perceived formation of a cake/gel layer that develops at high operating pressures. Subsequently, numerous theoretical models on gel/cake physics have been made to address how these factors can result in limiting flux. In a paradigm shift, the present article reestablishes the significance of osmotic pressure by examining its contribution to limiting flux in the framework of the recently developed free solvent osmotic pressure model. The resulting free-solvent-based flux model (FSB) uses the Kedem–Katchalsky model, film theory and the free solvent representation for osmotic pressure in its development. Single protein tangential-flow diafiltration experiments (30 kDa MWCO CRC membranes) were also conducted using ovalbumin (OVA, 45 kDa), bovine serum albumin (BSA, 69 kDa), and immuno-gamma globulin (IgG, 155 kDa) in moderate NaCl buffered solutions at pH 4.5, 5.4, 7 and 7.4. The membrane was preconditioned to minimize membrane fouling development during the experimental procedure. The pressure was randomly selected and flux and sieving were determined. The experimental results clearly demonstrated that the limiting flux phenomenon is not dominated by membrane fouling and the FSB model theoretically illustrates that osmotic pressure is the primary factor in limiting flux during UF. The FSB model provides excellent agreement with the experimental results while producing realistic protein wall concentrations. In addition, the pH dependence of the limiting flux is shown to correlate to the pH dependency of the specific protein diffusion coefficient.     

Air gap membrane distillation of sucrose aqueous solutions

In this paper results obtained with air gap membrane distillation (AGMD) using sucrose aqueous solutions are shown. The role of the relevant process parameters has been investigated experimentally (the flow rate through the cell, the feed initial concentration, the type of membrane, the air gap thickness, etc.). Equations have been proposed to estimate the intermediate temperatures for the air gap configuration. The fluxes given by different gas stagnant film diffusion models showed good agreement with the experimental results over the entire range of temperatures studied. Also a model which accounts for the thermal diffusion phenomenon was used. From the fits of the experimental flux data to the theoretical equations, the diffusion coefficient of the water vapour–air mixture, D_AB, and the thermal diffusion coefficient, K_T (only in the last case), were obtained and the results were analysed. For the D_AB coefficient higher values than the tabulated ones have been obtained, although of the same order of magnitude, and still higher when the thermal diffusion is considered.     

Sieving variations due to the choice in pore size distribution model

The effect of the choice of the standard probabilistic model to describe the pore size distribution was theoretically studied on predicting membrane performance parameters, area average water flux and area average membrane sieving coefficient. Preliminary discrete pore size distributions were generated from rejection profiles of dextran and PEG for 10,000, 30,000 and 100,000 molecular weight cutoff (MWCO) polysulfone and cellulose acetate membranes. The standard probability distribution functions (PDF), gamma, lognormal, normal, Weibel and Rayleigh were used to fit the resulting pore size distribution data. It was observed that the area averaged sieving coefficients are sensitive to the choice of the PDF. These results implied that an uncertainty in the choice of distribution in describing the membrane morphology could lead to a propagated uncertainty in predicting overall membrane performance.     

Ideal limiting fluxes in ultrafiltration: comparison of various theoretical relationships

The ultrafiltration of macromolecules is characterised by a limiting flux at high transmembrane pressures. There is also some evidence that at high pressures and low crossflow velocities the flux decreases slightly with increasing pressure. It is confirmed from a theoretical viewpoint that this can only be caused by a decrease in the average mass-transfer coefficient due to concentration increases in the boundary layer. At the practical level, we propose an expression which, for a given system, enables the ideal flux to be estimated a priori as a function of the transmembrane pressure. The ideal flux is defined as that flux which would occur in the absence of fouling and gelation. The model includes the influence of both osmotic pressure and the variation in viscosity due to concentration polarisation. Thus for predictive purposes knowledge of osmotic pressure and viscosity as a function of concentration is required. The only membrane parameter that has to be experimentally determined is the membrane permeability. In the absence of adsorption (which is the ideal case) this is the permeability to the pure solvent. The model has been tested against Jonsson's data for the ultrafiltration of dextran solutions. The results are most encouraging.     

Microfiltration of amphoteric surfactant using ceramic membranes

The microfiltration of commercially available amphoteric surfactant using ceramic membranes has been investigated. Various combinations operating conditions such as pH, electrolyte and surfactant concentrations were employed. Zeta potential and adsorption isotherms were obtained for the components of membrane surfactant system as functions of pH using surfactant or indifferent electrolyte (KCl). The shift in the membrane isoelectric point induced by the surfactant is linked to the carboxylic groups present on the surfactant which are believed to play a dominant role in the net surface charge of the membrane. A minimum in the permeate flux was found at the pH corresponding to the isoelectric point of the zwitterionic surfactant. This behaviour is ascribed to the interactions occurring between the surfactant–surfactant molecules and the surfactant–membrane. The higher fluxes obtained at low pH as compared to high pH arise from different fouling mechanisms and ionic strengths. Lower fluxes were found when inorganic electrolytes were used in conjunction with surfactant. However, as the valency of the salt increases, flux behaviour of the zwitterionic surfactant (close to isoelectric point) does not vary whilst the cationic and anionic state of the surfactants are much more affected. Interactions between surfactant molecules as a result of the charge screening effects by the larger valence ions are encouraged. The permeate flux declines with an increasing surfactant concentration even though some concentrations fall under the critical micelle concentration (c.m.c.). This is attributed to concentration polarisation in which the accumulated surfactant concentration at the membrane surface could form a stable viscous phase which is resistant to permeate flow in the secondary layer next to the membrane surface. This paper demonstrates the role interactions such as surfactant–surfactant and surfactant–membrane play in influencing the filterability of surfactant solutions using ceramic membranes.     

A liquid porosimetry technique for correlating intrinsic protein sieving: Applications in ultrafiltration processes

The understanding of variation in sieving properties of membranes is of great importance for the successful development of ultrafiltration applications. A liquid porosimetry technique is presented to quantify the sieving variation among several polyethersulfone ultrafiltration membranes. Observed sieving coefficients were measured with proper precautions taken to control and minimize fouling. These data were translated to intrinsic sieving coefficients using a stagnant film model. The intrinsic membrane sieving coefficient correlated well with the liquid porosimetry data. This liquid porosimetry technique can distinguish between membranes of different molecular weight cut-off and is sensitive enough to capture slight changes in the sieving coefficient of variants of the same cut-off membrane. This technique has several attractive features: it is non-destructive, independent of the module configuration and relatively simple to perform. Two potential applications of this technique are also examined: (1) quantification of the effect of membrane variation on high performance tangential flow filtration (HPTFF) for protein separations and (2) development of a membrane integrity test to ensure batch-to-batch consistency. This technique has the potential for use in membrane quality control, membrane selection, and validation of industrial ultrafiltration processes.     

Coupled pervaporation/mass spectrometry for investigating membrane mass transport phenomena

The integration of organophilic pervaporation into processes of varying feed concentration, such as bioconversions, chemical reactions, or analytical sample preparation, requires not only the understanding of mass transport phenomena across the membrane under steady-state conditions, but also the insight into the transient response of the pervaporation membrane to changes as they occur in practice. For this purpose, a laboratory-scale pervaporation unit was coupled to a mass spectrometer for on-line permeate analysis, maintaining the overall pervaporation operating conditions controllable independently, and without introducing any inert gases for sample transfer.The experimental set-up was employed for investigating the transport of aroma compounds across a POMS–PEI composite membrane, focusing in particular on the so-called “membrane conditioning”; the possible synergetic effect of ethanol on the flux of one model aroma compound, ethyl hexanoate; the application of the system proposed to the rapid screening of the effect of the hydrodynamic upstream conditions on the degree of concentration polarisation.The method proposed proved to be robust and flexible, not only allowing insights into transient mass transport phenomena otherwise not attainable, but also reducing experimental workload significantly when characterising the effect of varying operating conditions on the pervaporation performance.     

Characterization of the pore area distribution in porous membranes using transport measurements

An approach originally proposed by Mason and coworkers has been applied to model porous membranes to show that transport measurements with small and large solutes can be used to distinguish between porous membranes with the same average pore size but different pore size distributions. In addtion, it is shown that such measurements can be used to account for membrane heteroporosity when predicting the sieving characteristics of a membrane. This is done by applying moment theory to results from flux measurements for a small solute at Pe ≈ 1 or to results from measurements of the reflection coefficient for a large solute at infinite Pe. No a priori assumptions about the nature of the distribution of pore areas are necessary.In this paper, the results from calculations performed with three different model membranes with log-normal pore size distribution are reported. These results show that one can begin to distinguish between membranes by measuring the hydraulic and diffusive permeability and performing at least one additional flux measurement — with either a small, non-hindered solute at Pe ≈ 1 or a large solute at infinite Pe. Results also show that a fairly narrow window can be placed on the sieving curve for a heteroporous membrane without performing any sieving measurements. This is an interesting and encouraging result because it means that many of the problems that arise from measuring and interpreting pore size distributions using more traditional techniques can be avoided by using small solute flux measurements to predict the separation characteristics of many porous membranes.     

Factors affecting flux and ethanol separation performance in vacuum membrane distillation (VMD)

Membrane distillation (MD) has a great potential as a concentration process for temperature labile liquids such as fruit juices, etc. Besides water, also aroma compounds will permeate through the membrane depending on their volatility and how the MD process is operated on the permeate side.In this paper, an experimental and theoretical investigation of the influence of concentration polarisation and temperature polarisation on the flux and selectivity of binary aqueous mixtures of ethanol is presented for vacuum membrane distillation (VMD) processes. Experimental results include changes of the following parameters: nature of solutions, membrane material and pore size, feed temperature, recirculation flow rate. One method was proposed in order to evaluate the concentration polarisation effects from the fit of the experimental data. General models taking into account Knudsen and viscous flows were proposed, but viscous contribution resulted to be negligible under our operating conditions. Therefore, theoretical fluxes were estimated using Knudsen model and a good agreement between them and the experimental ones was found.