A proposta deste trabalho foi a concepção de um modelo para o dimensionamento otimizado de sistemas para tratamento de água baseados na tecnologia de Ultrafiltração com membranas de fibra-oca pressurizadas. O modelo relaciona o comportamento das membranas com a qualidade da água bruta através de resultados de ensaios em unidade piloto ou de bancada e equações de bloqueio de poros. A validação desta relação foi realizada através da análise dos dados de uma planta piloto operada em dois períodos distintos em regimes de fluxo variando de 60 a 70L/m².h e alimentada com água bruta proveniente de um manancial de superfície. Os resultados apontaram para a predominância do mecanismo de obstrução através da formação de torta e indicaram uma boa aderência das equações do modelo físico à realidade observada. Variáveis econômicas foram incorporadas ao modelo de forma a permitir a otimização através da busca do mínimo custo total de propriedade. Uma análise de sensibilidade demonstrou que os parâmetros de projeto mais impactantes no custo total, quantidade de membranas, duração do ciclo de filtração e duração do ciclo de contralavagem, podem variar seu peso em diferentes regiões do mundo influenciando o dimensionamento. Outras variáveis de projeto como a vazão de ar de limpeza demonstraram ser pouco significantes para a redução do custo total. A otimização do modelo foi realizada através de um algoritmo de busca numérica para as variáveis de duração do ciclo de filtração e quantidade de membranas, os resultados a partir das informações colhidas na planta piloto levaram a um projeto arrojado, porém dentro das recomendações gerais dos fabricantes de membranas. Como conclusão é possível afirmar que o modelo de dimensionamento do projeto é capaz de reduzir os custos totais de uma estação de tratamento de água baseada na tecnologia de ultrafiltração além de demonstrar potencial para a otimização dinâmica de plantas já instaladas. / The goal of this work was the development of an optimal design model for water treatment plants based in the pressurized hollow-fiber ultrafiltration membrane technology. The model uses operational data from pilot plants or bench scale units as input and pore blocking equations to predict the behavior of the membranes. The correlation between the fouling model adopted and the pilot plant results was evaluated using a pilot plant operated in two different time frames with flux rates from 60L/m²h to 70L/m²h and fed with raw water from a lake. The results from this validation have shown that the major fouling mechanism is the cake filtration and the theoretical curves had a good fitting with the operational data. To allow the cost optimization, economic variables were added to the model. A sensitivity analysis demonstrated that the most significant design parameters on the overall cost were the membrane area, the duration of the filtration cycle and the duration of the backwash cycle. According to the analysis, since the costs of the commodities and membranes are different from a region to the other, the optimal system design will also be different. Other design parameters like the membrane aeration rate have shown almost no impact on the total operating cost. The model optimization for the membrane quantity and the filtration cycle duration was based on bi-dimensional discrete numeric algorithm. The results from the optimization using the pilot plant data were compatible with the typical ranges and limits proposed by the membrane manufacturers. The design model proposed was able to reduce the total costs of new plants and demonstrated a good potential for the dynamic optimization of existing plants.

Chapter 1 briefly introduces the theoretical background of how applications and lists some of resent research progresses in this area. Polymer with various configurations and conformations pass through nanopores; including polymer linear chains, stars polymer, branched polymers, polymer micelles are introduced. Among them, the de Gennes and Brochard-Wyart's predictions of polymer linear and star chains passing through nanopores are emphasized, in which they predicted that qc of linear chain is qc ≃ kBT/(3pieta), where kB, T and eta are the Boltzmann constant, the absolutely temperature, and the viscosity of solvent, respectively, independent of both the chain length and the pore size; and for star chains passing through nanopores, there exist a optimal entering arm numbers, namely, the star chains passing through nanopores. / Chapter 2 details basic theory of static and dynamic laser light scattering (LLS), including its instrumentation and our ultrafiltration setup. / Chapter 3 briefly introduces the sample preparation, including the history and mechanism of anionic living polymerization, as well as how we used a novel home-made set-up to prepare linear polystyrene with different chain lengths and star polystyrene with various arm numbers and lengths. / Chapter 4 summarizes our measured critical flow rates (qc) of linear polymer chains with different lengths for nanopores with different sizes, since the flow rate is directly related to the hydrodynamic force, we have developed a sensitive method (down to tens fN) to directly assess how much the hydrodynamic force (Fh) is required to overcome the weak entropy elasticity and stretch individual coiled chains in solution. Our method is completely different from the using existing optical tweezers or AFM, because they measure the relatively stronger enthalpy elasticity. Our results confirm that qc is indeed independent of the chain length, but decreases as the pore size increases. The value of qc is ∼10--200 times smaller than kBT/(3pieta). Such a discrepancy has been attributed to the rough assumption made by de Gennes and his coworkers; namely, each chain segment "blob" confined inside the pore is not a hard sphere so that the effective length along the flow direction is much longer than the pore diameter. Finally, using the solution temperature, we varied the chain conformation, our result shows that q c has a minimum which is near, but not exactly located at the theta temperature, might leading to a better way to determine the true ideal state of a polymer solution, at which all viral coefficients, not only the second vanish. / Chapter 5 uses polymer solutions made of different mixtures of linear and star chains, we have demonstrated that flushing these solution mixtures through a nanopore with a properly chosen flow rate can effectively and cleanly separate linear and star chains no matter whether linear chains are larger or smaller than star chains. / Chapter 6 further investigates how star-like polystyrene pass through a given nanopore under the flow field. Star polystyrene chains with different arm lengths (LA) and numbers (f) passing through a nanopore (20 nm) under an elongational flow field was investigated in terms of the flow-rate dependent relative retention ((C0 - C)/C0), where C 0 and C are the polymer concentrations before and after the ultrafiltration. Our results reveal that for a given arm length (LA), the critical flow rate (qc,star), below which star chains are blocked, dramatically increases with the total arm numbers (f); but for a given f, is nearly independent on LA, contradictory to the previous prediction made by de Gennes and Brochard-Wyart. We have revised their theory in the region fin < fout and also accounted for the effective length of each blob, where fin and fout are the numbers of arms inside and outside the pore, respectively. In the revision, we show that qc,star is indeed independent of LA but related to f and f in in two different ways, depending on whether fin ≤ f/2 or ≥ f/2. A comparison of our experimental and calculated results reveals that most of star chains pass through the nanopores with fin ∼ f/2. Further study of the temperature dependent (C0 - C)/C 0 of polystyrene in cyclohexane reveals that there exists a minimum of qc,star at ∼38 °C, close to its theta temperature (-34.5 °C). / This Ph. D. thesis presents our study on the ultrafiltration of polymers with different configurations and conformations; namly, theoretically, the passing of polymer chains through a nanopore under an elongational flow filed has been studied for years, but experimental studies are rare because of two following reasons: (1) lacks a precise method to investigate how individual single polymer chain pass through a nanopore; (2) it is difficult, if not impossible, to obtain a set of polymer samples with a narrow molar mass distribution and a uniform structures; except for linear chains. The central question in this study is to find the critical (minimum) flow rate (qc) for each kind of chains, at which the chains can pass through a given nanopore. A comparison of the measured and calculated qc leads to a better understanding how different chains are deformed, stretched and pulled through a nanopore. We have developed a novel method of combinating static and dynamic laser light scattering (LLS) to precisely measure the relative retention concentration ((C0 - C)/C0). / Ge, Hui. / Adviser: Chi Wu. / Source: Dissertation Abstracts International, Volume: 73-01, Section: B, page: . / Thesis (Ph.D.)--Chinese University of Hong Kong, 2010. / Includes bibliographical references. / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Electronic reproduction. [Ann Arbor, MI] : ProQuest Information and Learning, [201-] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Abstract also in Chinese.

A proposta deste trabalho foi a concepção de um modelo para o dimensionamento otimizado de sistemas para tratamento de água baseados na tecnologia de Ultrafiltração com membranas de fibra-oca pressurizadas. O modelo relaciona o comportamento das membranas com a qualidade da água bruta através de resultados de ensaios em unidade piloto ou de bancada e equações de bloqueio de poros. A validação desta relação foi realizada através da análise dos dados de uma planta piloto operada em dois períodos distintos em regimes de fluxo variando de 60 a 70L/m².h e alimentada com água bruta proveniente de um manancial de superfície. Os resultados apontaram para a predominância do mecanismo de obstrução através da formação de torta e indicaram uma boa aderência das equações do modelo físico à realidade observada. Variáveis econômicas foram incorporadas ao modelo de forma a permitir a otimização através da busca do mínimo custo total de propriedade. Uma análise de sensibilidade demonstrou que os parâmetros de projeto mais impactantes no custo total, quantidade de membranas, duração do ciclo de filtração e duração do ciclo de contralavagem, podem variar seu peso em diferentes regiões do mundo influenciando o dimensionamento. Outras variáveis de projeto como a vazão de ar de limpeza demonstraram ser pouco significantes para a redução do custo total. A otimização do modelo foi realizada através de um algoritmo de busca numérica para as variáveis de duração do ciclo de filtração e quantidade de membranas, os resultados a partir das informações colhidas na planta piloto levaram a um projeto arrojado, porém dentro das recomendações gerais dos fabricantes de membranas. Como conclusão é possível afirmar que o modelo de dimensionamento do projeto é capaz de reduzir os custos totais de uma estação de tratamento de água baseada na tecnologia de ultrafiltração além de demonstrar potencial para a otimização dinâmica de plantas já instaladas. / The goal of this work was the development of an optimal design model for water treatment plants based in the pressurized hollow-fiber ultrafiltration membrane technology. The model uses operational data from pilot plants or bench scale units as input and pore blocking equations to predict the behavior of the membranes. The correlation between the fouling model adopted and the pilot plant results was evaluated using a pilot plant operated in two different time frames with flux rates from 60L/m²h to 70L/m²h and fed with raw water from a lake. The results from this validation have shown that the major fouling mechanism is the cake filtration and the theoretical curves had a good fitting with the operational data. To allow the cost optimization, economic variables were added to the model. A sensitivity analysis demonstrated that the most significant design parameters on the overall cost were the membrane area, the duration of the filtration cycle and the duration of the backwash cycle. According to the analysis, since the costs of the commodities and membranes are different from a region to the other, the optimal system design will also be different. Other design parameters like the membrane aeration rate have shown almost no impact on the total operating cost. The model optimization for the membrane quantity and the filtration cycle duration was based on bi-dimensional discrete numeric algorithm. The results from the optimization using the pilot plant data were compatible with the typical ranges and limits proposed by the membrane manufacturers. The design model proposed was able to reduce the total costs of new plants and demonstrated a good potential for the dynamic optimization of existing plants.

The study involves two different approaches to increasing cottage cheese yields by the inclusion of whey proteins in the curd. They were the use of high heat treated skim milk and ultrafiltered skim milk retentates for cottage cheese making. Increasing the pasteurization temperature of skim milk from 61.8 to 79.4 C for 30 minutes resulted in 15.6% increase in cultured cottage cheese yields. However, the high heat cheese exhibited variable quality and was generally inferior to the control. Cheese making from high heat skim milk was done by cutting at the A.C. end point and using 30 ml of rennet per 454 kg of skim milk. There appeared to be no relationship between the rate of heating to 79.4 C and the quality of cottage cheese produced. Skim milk was concentrated by ultrafiltration to 12, 16, and 20% solids. The 20% retentate was produced by concentrating skim milk 5:1 and then subjecting it to diafiltration with an equivalent volume of deionized water. The acidification of the retentates was accomplished by metering concentrated hydrochloric acid into the vortex of the centrifugal Culture growth in 20% retentates was inhibited below and soluble phosphate seemed to be an important factor in this inhibition. The yield trials indicated that the increase in yields over cultured cottage cheese made from skim milk was 12.4, 15.3, 5.6 and 1.6% for 16 % cultured, 16 % direct acid, 20% cultured and 20% direct acid cottage cheeses made from retentates. The lower yield increases from 20% retentates was because of the shattering of the curd while cooking, and the diffusion of the whey proteins. While the quality of cultured cottage produced by ultrafiltration was as good as the cultured cottage cheese from skim milk, the direct acid cottage cheeses were much inferior. Considering both high heat treatment of skim milk and ultrafiltration for increasing cottage cheese yields, ultrafiltration seems to be the better method owing to the much better quality of cheese produced. To obtain maximum benefits from ultrafiltration, problems of lactic culture growth in retentates must be solved and non conventional cooking methods have to be developed for handling curds from high solids retentates.

Membrane processes have found increasing industrial applications worldwide. For membrane processes to deliver their desired performances and mitigate the effect of disturbances, automatic controllers must be installed. Before the installation of controllers, operability analysis is a crucial step to evaluate how well the processes can be controlled, and to determine how process design can be improved for better control. However, existing applications of operability analysis in membrane processes are limited. This thesis extends the application of operability analysis to a multiple-stage membrane process, exemplified by a detailed case study of a 12-stage industrial whey ultrafiltration (UF) process. Process dynamic models are determined to describe the transient behaviour of process performance caused by disturbances and long-term fouling. Steady-state nonlinear operability analysis is conducted to identify inherent limitations of the process. Using the process dynamic models, dynamic operability analysis is performed to determine the effects of dynamic behaviour on process and controller design. Steady-state operability analysis shows that the whey UF process is not able to mitigate the effects of high concentrations of true protein in the fresh whey feed. The ability of the process to mitigate the effects of disturbances is also adversely affected by long-term membrane fouling. Mid-run washing is therefore necessary to restore control performance after long periods of operation. Besides demonstrating the adverse effects of long-term membrane fouling on operability, dynamic operability analysis identifies the manipulated variables that can deliver the best control performance. It also indicates that control performance can be improved by installing equipment (e.g. buffer tanks) upstream of the process. Dynamic operability analysis shows that recycling of the retentate stream has a profound effect on the plant-wide dynamics and reduces significantly the achievable speed of process response under automatic control. However, retentate recycling is essential during operation to minimize membrane fouling. Although reducing the number of stages in the whey UF process can improve the achievable speed of process response under automatic control, process performance will fluctuate significantly from its desired level. A trade-off therefore exists between process performance and control performance that should be addressed during process and controller design.

Membrane processes are increasingly used in water treatment. Experiments were performed using stirred cell equipment, polymeric membranes and synthetic surface water containing natural organics, inorganic colloids and their aggregates, and cations. All processes could remove a significant amount of natural organics. Pretreatment with ferric chloride was required to achieve significant organic removal with MF and high MWCO UF. Additionally, fouling mechanisms for the three processes were investigated. Crucial parameters were aggregate characteristics (fractal structure, stability, organic-colloid interactions), solubility of organics and calcium, and hydrodynamics. In MF, fouling by pore plugging was most severe. Variations in solution chemistry changed the aggregation state of the colloids and/or natural organic matter and dramatically affected rejection and fouling behaviour. UF membrane fouling was mainly influenced by pore adsorption and could improve natural organics rejection significantly. Coagulant addition shifted fouling mechanism from pore adsorption to cake formation. Aggregate structure was most significant for flux decline. In NF, rejection of natural organics involved both size and charge exclusion. Fouling was caused by precipitation of a calcium-organic complex. Fouling could be avoided by pretreatment with metal salt coagulants. Thorough chemical characterisation of the organics used demonstrated that only size and aromaticity can be related to fouling. The study is concluded with a process comparison based on a water quality parameter and a cost comparison. Treatment cost of microfiltration with chemical pretreatment was similar to that of nanofiltration at a comparable natural organics rejection.

Membrane filtration of protein solutions is influenced by a wide range of processing and physicochemical conditions. Monitoring and optimizing membrane filtration may have advantages for achieving, in a cost effective manner, improved bioproduct purification and membrane performance which is relevant to pharmaceutical and biochemical applications. The motivation of this work was to examine the feasibility of applying two-dimensional fluorescence spectroscopy in conjunction with chemometric techniques for monitoring and possibly optimizing the performance of membrane processes. Preliminary work focused on assessing the use of multivariate calibration tools in conjunction with the sensitivity of intrinsic protein fluorescence towards changes in environmental conditions was to predict protein concentration and aggregation behavior. A model protein, β-lactoglobulin (β-LG), was used as a first simple case scenario. Results showed very good agreement between the fluorescence based predictions and measurements obtained by HPLC and gravimetric analysis regardless of the conditions. PLS analysis of excitation-emission matrices revealed unique spectral fingerprints that are most likely associated with the heat-induced denaturation and aggregation. Standard Normal Variate, a signal preprocessing and filtering tool, was shown to have a significant effect on enhancing the predictive accuracy and robustness of the PLS model as it reduced the effect of instrumental noise. The methodology was then extended to a two-component protein system consisting of α- lactlalbumin (α-LA) and β-lactoglobulin (β-LG). The process of thermal induced aggregation of β-LG and α-LA protein in mixtures, which involves the disappearance of native-like proteins, was studied under various treatment conditions including different temperatures, pH, total initial protein concentration and proportions of α-LA and β-LG. A Partial Least Squares (PLS) regression algorithm was used to correlate the concentrations of α-LA and β-LG to the fluorescence spectra obtained for mixtures.The results illustrated that multivariate models could effectively deconvolute multiwavelength fluorescence spectra collected for the protein mixtures and thereby provide a fairly accurate quantification of respective native-like α-LA and β-LG despite the significant overlap between their emission profiles. It was also demonstrated that a PLS model could be used as a black-box prediction tool for estimating protein aggregation when combined with simple mass balances. Ultrafiltration experiments of the whey protein isolate solutions were carried out in dead-end filtration mode and fluorescence measurements of permeate and retentate solutions were acquired in synchronous scanning mode using a fiber optic probe. By implementing a dilution strategy for the retentate side, the fluorescence based PLS model encompassed a low protein concentration range where fluorescence was not expected to be significantly influenced by concentration-dependent interferences. It was also demonstrated that synchronous spectra can provide good predictions and consequently the use of the full spectrum may not be necessary for monitoring with corresponding savings in acquisition time. Membrane performance variables that are difficult to measure, such as individual protein transmission and membrane selectivity could be estimated directly from fluorescence-based predictions of protein concentrations in the retentate and permeate streams. Multiwavelength light scattering spectra, acquired using the fiber optic probe, were shown to be a useful indicator for protein self-association behavior, which is known to influence the membrane filtration. High fouling potential were observed for protein solutions that exhibited significant Rayleigh scattering. A predictive PLS model for estimating protein aggregation from Rayleigh scattering measurements was developed and it was tested by using molecular weight experimental values obtained from the literature. Although this comparison was only partial due to the limited amount of molecular weight data available, the findings verified the possibility of estimating the aggregate size from multiwavelength Rayleigh scattering spectra acquired using a conventional spectrofluorometer. Thus, the results implied that both intrinsic fluorescence and light scattering multiwavelength measurements could provide complementary information about the filtration process.

Membrane filtration of protein solutions is influenced by a wide range of processing and physicochemical conditions. Monitoring and optimizing membrane filtration may have advantages for achieving, in a cost effective manner, improved bioproduct purification and membrane performance which is relevant to pharmaceutical and biochemical applications. The motivation of this work was to examine the feasibility of applying two-dimensional fluorescence spectroscopy in conjunction with chemometric techniques for monitoring and possibly optimizing the performance of membrane processes. Preliminary work focused on assessing the use of multivariate calibration tools in conjunction with the sensitivity of intrinsic protein fluorescence towards changes in environmental conditions was to predict protein concentration and aggregation behavior. A model protein, β-lactoglobulin (β-LG), was used as a first simple case scenario. Results showed very good agreement between the fluorescence based predictions and measurements obtained by HPLC and gravimetric analysis regardless of the conditions. PLS analysis of excitation-emission matrices revealed unique spectral fingerprints that are most likely associated with the heat-induced denaturation and aggregation. Standard Normal Variate, a signal preprocessing and filtering tool, was shown to have a significant effect on enhancing the predictive accuracy and robustness of the PLS model as it reduced the effect of instrumental noise. The methodology was then extended to a two-component protein system consisting of α- lactlalbumin (α-LA) and β-lactoglobulin (β-LG). The process of thermal induced aggregation of β-LG and α-LA protein in mixtures, which involves the disappearance of native-like proteins, was studied under various treatment conditions including different temperatures, pH, total initial protein concentration and proportions of α-LA and β-LG. A Partial Least Squares (PLS) regression algorithm was used to correlate the concentrations of α-LA and β-LG to the fluorescence spectra obtained for mixtures.The results illustrated that multivariate models could effectively deconvolute multiwavelength fluorescence spectra collected for the protein mixtures and thereby provide a fairly accurate quantification of respective native-like α-LA and β-LG despite the significant overlap between their emission profiles. It was also demonstrated that a PLS model could be used as a black-box prediction tool for estimating protein aggregation when combined with simple mass balances. Ultrafiltration experiments of the whey protein isolate solutions were carried out in dead-end filtration mode and fluorescence measurements of permeate and retentate solutions were acquired in synchronous scanning mode using a fiber optic probe. By implementing a dilution strategy for the retentate side, the fluorescence based PLS model encompassed a low protein concentration range where fluorescence was not expected to be significantly influenced by concentration-dependent interferences. It was also demonstrated that synchronous spectra can provide good predictions and consequently the use of the full spectrum may not be necessary for monitoring with corresponding savings in acquisition time. Membrane performance variables that are difficult to measure, such as individual protein transmission and membrane selectivity could be estimated directly from fluorescence-based predictions of protein concentrations in the retentate and permeate streams. Multiwavelength light scattering spectra, acquired using the fiber optic probe, were shown to be a useful indicator for protein self-association behavior, which is known to influence the membrane filtration. High fouling potential were observed for protein solutions that exhibited significant Rayleigh scattering. A predictive PLS model for estimating protein aggregation from Rayleigh scattering measurements was developed and it was tested by using molecular weight experimental values obtained from the literature. Although this comparison was only partial due to the limited amount of molecular weight data available, the findings verified the possibility of estimating the aggregate size from multiwavelength Rayleigh scattering spectra acquired using a conventional spectrofluorometer. Thus, the results implied that both intrinsic fluorescence and light scattering multiwavelength measurements could provide complementary information about the filtration process.

Oxidation plays an important role in the basic processes of life, such as the production of energy and phagocytosis employed by the immune system. However, when an imbalance between oxidants and antioxidants exists in vivo, oxidation can become uncontrolled and result in diseases such as arthritis, cancer, artherosclerosis, and Alzheimer’s Disease. Dietary antioxidants including polyphenolic compounds, proteins, and peptides have been identified as being physiologically functional foods capable of contributing to the restoration of this oxidant-antioxidant balance. The objective of this study was to explore the production of antioxidant soy peptides from a commercially available soy protein isolate (SPI) by enzymatic hydrolysis in a process similar to that occurring in the human digestive tract. In this study Archer-Daniels Midland SPI PRO-FAM 974 was used as a raw material for the production of antioxidant soy peptides. The digestion consisted of enzymatic digestion of the SPI (3.12 wt %) with pepsin (37ºC, pH 1.5) and/or pancreatin (40ºC, pH 7.8) either individually or sequentially. The enzyme concentration and digestion time for each enzyme was optimized using a 2^4 factorial experimental design to produce the greatest concentration of peptides quantified in PheGly equivalents by the OPA assay. A maximum peptide concentration of approximately 65 mM PheGly equivalents was achieved in the follow-up digests resulting from this factorial design model, using pepsin (0.15 g/L, 15 minutes) and pancreatin (4.5 g/L, 120 minutes) sequentially to digest the SPI. Fractionation of the peptides by sequential dead-end membrane ultrafiltration with molecular weight cut-offs (MWCO) of 3 kDa and 1 kDa was performed to produce peptide fractions with increased antioxidant capacity. The permeate flux as a function of time was fit to empirical models, revealing that the membrane fouling resulting in the permeate flux decline is largely reversible and most likely the result of cake filtration. Antioxidant capacity was quantified by the DPPH, FCR, and ORAC assays to determine the electron-donating and proton-donating capacities of the soy peptides. The electron-donating DPPH assay was not suitable to quantify the antioxidant capacity of the soy peptides due to poor peptide solubility in the assay media and sensitivity. The electron-donating FCR assay and the proton-donating ORAC assay were used to distinguish between the ultrafiltration and digestion conditions employed to produce the soy peptides and the antioxidant capacity was quantified in equivalence to the standard antioxidant Trolox. The soy peptide fraction with the greatest antioxidant capacity was produced by enzymatic digestion with pancreatin (4.5 g/L, 120 minutes) alone and had a molecular weight cut-off of between 3 kDa and 1 kDa. This fraction had an equivalent antioxidant capacity of approximately 190 mg Trolox/g sample in the ORAC assay and approximately 180 mg Trolox/g sample in the FCR assay. A preliminary linear model for the optimum digestion and ultrafiltration conditions for the production of antioxidant peptides with the greatest ORAC antioxidant capacity was also developed. The model includes a positive pancreatin digestion time term and a negative pepsin digestion time term. No ultrafiltration terms were found to be significant in this preliminary model, but a large constant term persisted. In conclusion, the enzymatic digestion of commercially available SPI with pancreatin and fractionated by ultrafiltration successfully produced a soy peptide fraction with increased antioxidant capacity.

Application of low pressure membranes in drinking water treatment, including both microfiltration (MF) and ultrafiltration (UF), have witnessed a rapid increase in the past decades. Low pressure membranes are considered a good technology in retrofitting existing conventional drinking water treatment plants or in newly constructed plants to meet the stringent regulations for drinking water treatment that aim at preventing health risks of waterborne diseases. Enteric viruses are one of the major types of waterborne pathogens, and they can be commonly found and are persistent in the environment. Both the United States and Canada require a 99.99% (4-log) removal of viruses during the drinking water treatment train. Unlike MF membranes, UF membranes have a very good potential for removing enteric viruses from the water due to their smaller pores comparable to the size of viruses. Drinking water regulations/guidelines in both the United States and Canada do not grant UF membranes any removal credit for viruses by default; however they have the provision that, in certain cases, virus removal credit may be granted based on pilot scale challenge testing. A better understanding of the interaction between the UF membranes and virus rejection can help to establish a removal credit for UF membranes. An essential part of this will be the effect of the membrane operation on the rejection of viruses to determine if UF membranes can offer a consistent removal of viruses. Membrane fouling is one of the major problems in membrane operation and it can affect the rejection characteristics of the membrane and improve its performance. The aim of this study was to investigate the removal of virus surrogates (MS2 and φX174 bacteriophage) using a commercial UF membrane under different conditions, to obtain information about the removal mechanisms of viruses. The experimental filtration unit was designed to have similar conditions like the full scale membrane treatment plants. The UF membrane used in this study provided very good removal of both MS2 and φX174 bacteriophage. The obtained results were consistent and in agreement with the expected removals based on the membrane characterization results and types of virus surrogate. As part of this work, a detailed study to improve methods for characterizing the pore size distribution of membranes was conducted. In the second part of the study, two different types of surface waters were used to study the effect of membrane fouling on virus removal. It was found that mainly hydraulically irreversible fouling could significantly improve the virus removal by UF membranes. Different cleaning regimes that are used in treatment plants had varying effects on virus removal. After maintenance cleaning, virus removal remained higher than that of clean membranes, and only chemical cleaning was effective for completely removing membrane foulants and returning virus removal back to base levels. Advanced analytical techniques were used to define the nature of the fouling layer on the membrane surface and how the foulants affected the rejection of viruses. Finally, our study showed that UF membranes are a robust treatment technology for removing different types of enteric virus surrogates from water under different operational conditions. Close monitoring of the UF unit performance and direct integrity testing can possibly detect membrane problems that can affect the rejection of viruses. Based on the virus physical characteristics and a detailed study of the membrane surface characteristics, especially the pore size distribution of the membrane, the removal of the specific virus can be closely estimated.