Ceramic Membrane Fouling Research Articles

Whey protein fouling of large pore-size ceramic microfiltration membranes at small cross-flow velocity

Microfiltration of whey protein solutions by tubular ceramic membranes, under constant cross-flow and trans-membrane pressure, with periodic backwashing, is investigated using a fully instrumented pilot unit. Relatively large nominal membrane pore size (0.8 μm) insures very high protein transmission, which is desirable in applications such as microbial load reduction. In the first of a sequence of three filtration-backwashing cycles, irreversible and reversible fouling are identified, over the tested pressure range of 5–17.5 psi. Early in the first cycle, especially at the higher pressures, a pore constriction/blocking mechanism appears to be responsible for the irreversible fouling. In the other two cycles only the reversible fouling is significant, possibly due to some kind of protein layer formation on the membrane surface. The permeate flux level tends to increase by increasing trans-membrane pressure up to a near-optimum value of 10 psi, beyond which pressure has a negative effect. This interesting trend is attributed to the interplay of cross-flow velocity, which tends to reduce fouling by promoting re-suspension and breakage of colloidal protein agglomerates, with the trans-membrane pressure (and related flux) which leads to protein layer formation on the membrane and may impose compressive stresses, thereby increasing its resistance to permeation.

Fouling and regeneration of ceramic membranes used in recovering titanium silicalite-1 catalysts

Ceramic ultrafiltration (UF) membranes were used to separate titanium silicalite-1 (TS-1) catalysts from slurry in catalytic ammoximation of cyclohexanone to the oxime. Both the fouling mechanisms and the regeneration of fouled membranes were studied. The fouling was mainly caused by TS-1 particles deposition, silica additive adsorption and iron precipitation. The interaction between the two solutes in the system, silica additive and TS-1 catalysts, played a significant role in the formation of dense cake layers at the membrane surface, leading to a greater flux decline during ultrafiltration of TS-1 particles. Estimation of hydrodynamic forces acting on a single particle shows crossflow velocity (CFV) has an important effect on the deposition of TS-1 particles. However, after particles have deposited, increasing CFV will not resuspend them due to the strong and dense cake layer formation. The cleaning of a fouled membrane can be achieved by sequential use of strongly basic and acidic solutions. In addition, a particle-cleaning method was applied. Micro-sized alumina particles were used for their scouring effect to remove the cake layer from the membrane surface effectively. The particle cleaning combined with acid cleaning was effective to fully restore membrane flux.

Whey protein fouling of microfiltration ceramic membranes—Pressure effects

Permeate flux decline in dead-end microfiltration of whey protein isolate solutions is studied, using disc-type ceramic membranes of nominal pore size 0.8 μm. The tests involve five successive filtration cycles, under fixed filtration pressure, with intermediate backwashing. Flux decline analysis and membrane resistance are employed to determine fouling mechanisms, with respect to applied pressure, in the range 2.5–10 psi. Permeate and retentate concentration data complement the permeation rate measurements. Both irreversible and reversible fouling is identified for the backwashing mode employed. Data interpretation suggests that more rapid fouling, as well as significant surface-layer “compaction” effects may occur at the higher pressures employed. There is evidence that irreversible fouling effectively develops during the first cycle of the tests (of 30 min duration) and apparently does not significantly increase later on; however, reversible fouling occurs through the entire filtration series, being more intense during the first minutes of each cycle. The effect of protein aggregates is investigated with the aid of DLS measurements and filtration tests with prefiltered solutions; it appears that whey protein aggregates present in the solution, are responsible (almost entirely) for the observed membrane fouling. An effective membrane cleaning procedure is proposed. Finally, the possible advantage of operating large-pore size MF (for the process considered) at rather low pressure is suggested.

Ultrasonic control of ceramic membrane fouling: Effect of particle characteristics

In this study, the effect of particle characteristics on the ultrasonic control of membrane fouling was investigated. Ultrasound at 20 kHz was applied to a cross-flow filtration system with γ-alumina membranes in the presence of colloidal silica particles. Experimental results indicated that particle concentration affected the ability of ultrasound to control membrane fouling, with less effective control of fouling at higher particle concentrations. Measurements of sound wave intensity and images of the cavitation region indicated that particles induced additional cavitation bubbles near the ultrasonic source, which resulted in less turbulence reaching the membrane surface and subsequently less effective control of fouling. When silica particles were modified to be hydrophobic, greater inducement of cavitation bubbles near the ultrasonic source occurred for a fixed concentration, also resulting in less effective control of fouling. Particle size influenced the cleaning ability of ultrasound, with better permeate recovery observed with larger particles. Particle size did not affect sound wave intensity, suggesting that the more effective control of fouling by large particles was due to greater lift and cross-flow drag forces on larger particles compared to smaller particles.

Ultrasonic control of ceramic membrane fouling caused by natural organic matter and silica particles

The ultrasonic control of membrane fouling caused by natural organic matter (NOM) and silica particles was investigated under various solution conditions. Twenty kilohertz ultrasound was applied to a cross-flow ultrafiltration system with γ-alumina membranes. Experimental results indicated that solution chemistry affected the ability of ultrasound to control membrane fouling. More effective control of membrane fouling occurred at high pH, low ionic strength, and in the absence of divalent cations, consistent with the expectation of stronger electrostatic repulsion between foulant–foulant and foulant–membrane components. Fourier transform infrared spectra of the NOM foulant showed the presence of aliphatic carbon and carboxylate functional groups, both of which decreased in magnitude with increasing pH, in the absence of Ca 2+, or the presence of ultrasound. A decrease in the NOM rejection rate during ultrasound was observed. This likely occurred due to decreased steric exclusion of NOM through the fouled membrane, as scanning electron microscope images indicated that ultrasound partially removed the foulant layer from the membrane surface allowing larger NOM molecules to pass through the membrane. No damage to the membrane by ultrasound was observed.

Ultrasonic control of ceramic membrane fouling by particles: Effect of ultrasonic factors

Ultrasound at 20 kHz was applied to a cross-flow ultrafiltration system with γ-alumina membranes in the presence of colloidal silica particles to systematically investigate how ultrasonic factors affect membrane cleaning. Based on imaging of the ultrasonic cavitation region, optimal cleaning occurred when the membrane was outside but close to the cavitation region. Increasing the filtration pressure increased the compressive forces driving cavitation collapse and resulted in fewer cavitation bubbles absorbing and scattering sound waves and increasing sound wave penetration. However, an increased filtration pressure also resulted in greater permeation drag, and subsequently less improvement in permeate flux compared to low filtration pressure. Finally, pulsed ultrasound with short pulse intervals resulted in permeate flux improvement close to that of continuous sonication.

Hydraulic resistance in microfiltration of titanium white waste acid through ceramic membranes

Effects of operational parameters on fouling of ceramic membranes during microfiltration of titanium white acid are studied. Based on resistance-in-series model, the total filtration resistance (Rt) is divided into four parts: polarization resistance (Rp), cake resistance (Rc), internal plugging resistance (Ri) and membrane's own resistance (Rm). Their values and the relative percentages as a function of the operating parameters is determined. The mass of cake (Mc) is measured and average specific resistance of cake (αc) is also calculated for the further analysis of the change of Rc. The study will provide theoretical foundations for taking necessary actions to avoid or reduce fouling.

Fouling and regeneration of ceramic microfiltration membranes in processing acid wastewater containing fine TiO 2 particles

Both the fouling mechanisms and the regeneration of two asymmetric microporous Al 2O 3 MF membranes for the recovery of fine TiO 2 particles in acid wastestreams were investigated. The fouling was mainly caused by particle deposition, ion adsorption, and gel polarization. The foulants include TiO 2 particles and sulfates. The effect of operation conditions including membrane pore size, transmembrane pressure, and cross-flow velocity (CFV) on the fouling were also investigated, and the appropriate conditions were obtained. Membrane regeneration was achieved by backpulsing and chemical cleaning. Backpulsing is an effective method for reducing fouling extent of 1.0 μm pore-size membrane. The optimum backpulsing conditions are also determined. The effective cleaning agent is 0.5 mol/l oxalic acid solution.

Fouling and rejection behavior of ceramic and polymer-modified ceramic membranes for ultrafiltration of oil-in-water emulsions and microemulsions

The effectiveness of poly(vinylpyrrolidone) (PVP)-modification of a zirconia-based ultrafiltration membrane was investigated for the treatment of oil-in-water (o/w) emulsions. Fouling, hydraulic permeability, flux decline, and solute rejection for modified and native membranes were evaluated using a diagnostic o/w emulsion as well as an emulsion prepared using a commercial cutting oil. The native membrane was irreversibly fouled by both the o/w microemulsion and the commercial cutting oil emulsion. In contrast, irreversible fouling was not observed for the PVP-modified membrane. The main fouling agent for the diagnostic o/w microemulsion was identified as the anionic surfactant octanoate, which lead to an irreversible decline of initial membrane hydraulic permeability of up to 20%. Adsorption of surfactant species onto the native membrane surface were attributed to charge screening due to high ionic strength environment and presence of hydroxyl groups on the zirconia surface. Analysis of permeate flux decline and membrane resistance behavior suggests that membrane fouling was mainly due to a solute cake (gel) layer buildup at the membrane surface. However, the possibility of some internal pore plugging, due to deposition of small surfactant molecules, micelles or oil droplets, during the initial few minutes of filtration runs, could not be ruled out. Improved oil rejection (two-fold for the microemulsion and over 20% for the cutting oil emulsion) with the modified membrane compared to the native membrane was attributed to repair (or narrowing) of defects (or ‘pin-holes’) of the native membrane upon polymer grafting.

Protein Adsorption onto Zirconia Modified with Terminally Grafted Polyvinylpyrrolidone

The potential effectiveness of poly(vinyl pyrrolidone) (PVP) as a zirconia surface modifier for protein adsorption reduction was investigated using lysozyme (LYS). The relatively small size of LYS (45 × 30 × 30 Å) allowed for testing the adequacy of the graft polymerization method for producing a dense surface chain coverage to exclude LYS from direct interaction with the zirconia surface. The study demonstrated that a PVP brush layer is capable of reducing lysozyme adsorption. Overall, the maximum adsorption capacity decreased (by up to about 76%) due to surface modification with increasing polymer/silane surface coverage ratio (mol/mol). Adsorption reduction, due to protein exclusion from the surface by the tethered polymer layer, increased significantly when the distance between surface chains was less than the large axis of LYS (i.e., 45 Å). The present results are encouraging and suggest further consideration of polymer-modified ceramic surfaces for reducing fouling of ceramic membranes during protein ultrafiltration and producing ceramic biocompatible surfaces for biomedical applications.

Fouling of ceramic membranes by albumins under dynamic filtration conditions

The fouling of ceramic ultrafiltration membranes by bovine serum albumin (BSA) and human serum albumin (HSA) has been followed by means of small angle neutron scattering (SANS) and simultaneous monitoring of the decline of permeate flux with time under cross-flow conditions. SANS is primarily sensitive to the gradual buildup of the protein layer deposited inside the membrane pores while the permeate flux is a measure of the overall blockage of the membrane. The combined measurement allows the determination of the location of protein fouling. The effect of solution pH on the fouling by the two albumins was characterised at a fixed protein concentration of 1 g dm −3 on alumina membranes with nominal pore diameter of 200 Å. The filtration results show that the pattern of fouling is similar for both BSA and HSA, that there is a rapid permeate flux decline at pH 3 and 5, and a slower decline at pH 7. The simultaneous SANS experiment shows that at pH 5 there is strong deposition of protein layer inside the membrane pores, suggesting that at this pH blockage inside the membrane pores is primarily responsible for fouling. However, at pH 3, SANS showed that amount of BSA and HSA deposition inside the pores was at its smallest, suggesting that, in this case, the permeate flux decline is caused by the blockage of the front surface of the membrane. At pH 7 the extent of protein deposition inside the membrane is intermediate between the values for the other two pH, suggesting that the blockage at this pH mainly arises from the slow deposition of protein inside the membrane pores. Similar pH effects were observed on alumina membranes with a nominal pore diameter of 2000 Å. Although there is a general similarity in the fouling pattern of the two proteins, HSA causes less blockage with time at the pH conditions away from the isoelectric point of the albumins. The effect of cross-flow velocity on the fouling was also examined. It was found that permeate flux dropped faster at an elevated cross-flow velocity, although the extent of protein deposition inside the membrane pores was found to be similar, indicating that blockage at the front, outer surface of the membrane is intensified by a higher flow velocity.

Identification of the location of protein fouling on ceramic membranes under dynamic filtration conditions

The fouling of ceramic ultrafiltration membranes by bovine serum albumin (BSA) under dynamic filtration conditions has been studied by small angle neutron scattering (SANS) combined with the monitoring of the decline of permeate flux with time under the crossflow condition. Although the dropping of permeate flux is indicative of the blockage of the membrane pores it offers no information about the location of the protein deposition. In comparison, neutron scattering is sensitive to the structural composition of the membrane pores and the measurement of the scattering intensity with time offers information about the gradual buildup of protein deposit inside the membrane pores. The combined experiment of the SANS and permeate flux measurement thus allows a reliable determination of the location of protein fouling. BSA solution was pumped through a thin flat sheet of commercial alumina membrane in a quartz cell specially designed to allow a simultaneous experiment of permeate flux recording and small angle neutron scattering. The fouling experiment was first made at pH 5, close to the isoelectric point (IP) of 4.8 for BSA at the BSA concentration ranged from 0.1 to 1 g dm −3 . The permeate flux was found to decrease with BSA concentration and the trend of decrease is entirely consistent with the thickening of the protein layer adsorbed on the surface of the membrane pores, suggesting that under this condition blockage inside the membrane pores is primarily responsible for fouling. Subsequent measurements were made at different pH values away from the BSA isoelectric point and at a fixed BSA concentration of 1 g dm −3 . At pH 3 the permeate flux showed a fastest decaying with time while SANS showed a least amount of BSA deposition inside the membrane pores. This observation clearly indicates that it is the front surface of the membrane that is blocked. At pH 7 the permeate flux showed a slowest decaying with time but the extent of protein deposition inside the membrane pores is intermediate, suggesting that at least at the early stage of the filtration, front blockage is insignificant and fouling mainly arises from the slow deposition of protein inside membrane pores.

Application of Small Angle Neutron Scattering to the in Situ Study of Protein Fouling on Ceramic Membranes

The fouling of ceramic ultrafiltration membranes by protein under dynamic filtration conditions has been investigated by small angle neutron scattering (SANS). Bovine serum albumin (BSA) solution was pumped through a thin commercial alumina membrane in a quartz cell specially designed to allow simultaneous SANS. SANS from clean membranes immersed in D2O obeyed Porod's law and showed that the pores in these ultrafiltration membranes more closely resemble fractal or random pore structures than well-oriented, monodisperse arrays of cylindrical pores. During protein filtration, the SANS patterns showed a buildup with time of protein on all parts of this internal surface. The effects of time and pH indicate that the performance with respect to filtration is closely linked to the buildup of the adsorbed layer of protein.

A model of ceramic membrane fouling during heavy oil ultrafiltration

AbstractFouling of asymmetric, single tube ceramic membranes during ultrafiltration of Cold Lake heavy oil, is described by a pore restriction fouling model that predicts reduced permeate flux but increased asphaltene rejection with increased time‐on‐stream. Application of the model to experimental data showed that a 2nd‐order rate of asphaltene adsorption/deposition within the membrane pores, described the data better than a 1st‐order model. The asphaltene molecular radius, estimated as a parameter of the model, was in the range 3.8 ‐ 14.2 nm, depending on the membrane pore diameter. These values are in good agreement with previously published estimates of asphaltene particle size, determined using different experimental techniques.

Fouling of ceramic membranes during microfiltration of model solutions

A model study of the microfiltration of fatty acids (C2 to C8) dissolved in a synthetic ultrafiltrate was performed. Ceramic membranes of 0.1 and 0.2 μm were used. Results showed that minerals from the ultrafiltrate were involved in membrane fouling, while fatty acids were poorly rejected.