Rejection Of Organic Solutes Research Articles

Molecular fingerprint-aided prediction of organic solute rejection in reverse osmosis and nanofiltration

Reverse osmosis and nanofiltration are used to purify feedwaters that contain a range of harmful organic solutes. The rejection of many of these solutes is poorly understood due to our limited ability to experimentally measure removal of any given compound. In this work, we present a machine learning approach that predicts organic solute rejection using molecular fingerprints that encode chemical structure features, such as functional groups and rings, into simple binary vectors. We trained machine learning models on a database of 1906 membrane rejection measurements including 228 organic compounds and 39 types of reverse osmosis and nanofiltration membranes. Three types of molecular fingerprint models (structural key, circular, and path based) were compared, and we observed that the Molecular Access System (MACCS) structural key had high performance (coefficient of determination of 0.87 with the testing set), fast calculation time due to its short bit-length, and easy interpretability. In addition to evaluating prediction performance, Shapley Additive Explanations (SHAP) analysis was implemented to gain a better molecular-scale understanding of membrane rejection, identifying molecular substructures that are important in determining their rejection. Overall, this work presents a method to predict the rejection of compounds that uses readily available molecular structure information and improves our ability to understand rejection mechanisms.

Leaching of dissolved organic carbon from commercial reverse osmosis membranes exacerbates interferences in organic solute rejection during bench-scale filtration

Materials in contact with purified (e.g., Drinking water, ultrapure-water) water are sources of dissolved organic carbon which can increase the regrowth potential. Reverse osmosis (RO) membrane materials are composed of flexible polymers that are also in contact with treated water, but their contribution to DOC and microbial regrowth in bench-scale filtration studies is not reported. Bench-scale filtration studies use less volumetric throughput compared to pilot and full-scale filtration systems, which likely exacerbates the influence of leachates on for DOC measurement in permeates. This limits our ability to determine if standard membrane preparation procedures ensure sufficient removal of leachable organics. Migration of DOC from 8 different reverse osmosis membrane filters was assessed by two different experimental approaches; (1) Immersion assay—membranes were soaked in a pasteurized, isotonic water solution with a low and constant concentration of DOC. (2) Volume throughput (VT) test—Permeate was collected at defined VT steps to determine the impact of VT on the washout of DOC during system start-up. Test water from the immersion assay and the filtration test were analyzed with respect to (i) DOC (mg/L), (ii) microbial regrowth potential (% increase in bacterial activity), (iii) DOC fractional composition via LC-OCD. The presented study identified variations in the leachability of DOC between commercial RO membranes. Leaching of DOC from RO membranes varied by supplier and membrane composition. Alfa Laval RO98 exhibited the highest release of DOC compared to the other membranes tested during the immersion assay, suggested by an increase in test water DOC from 0.3 mg/L to 7.43 ± 1.54 mg/L, (n = 3) within 48 h. Aquaporin biomimetic-based membranes had the lowest release of DOC after 48 h of immersion (0.702 ± 0.17 mg DOC/L, n = 3). Higher concentrations of leached DOC did not lead to a higher regrowth potential on the released solute. Results of the filtration test demonstrate that DOC leaching from the membrane can exceed the point of flux-stabilisation, which is the common endpoint of membrane preparation. In all membranes tested, the chemical composition of DOC was dominated by low molecular weight acids, indicative of monomers used during membrane interfacial polymerization. The implication of this study suggests that on bench-scale filtration less volumetric throughput compared to pilot-full scale filtration exacerbates the influence of leachates on for example DOC measurement in permeates; promoting interferences in DOC permeate measurement and corresponding rejection coefficient. Relevance for full-scale filtration systems is limited due to higher volume throughputs applied during system start-up and commissioning.

How permeable could a reverse osmosis membrane be if it was specifically developed for uncharged organic solute rejection?

AbstractHere, we present both experimental data and modeling results of the removal of four different polar organic solutes by six different commercial nanofiltration (NF) and reverse osmosis (RO) membranes in a natural groundwater matrix. These results are significant for two reasons. First, three of the polar organic solutes are contaminants commonly found in both drinking water aquifers and municipal wastewaters and, hence, are particularly relevant to the protection of public health in drinking water production and potable reuse applications. Second, our membrane transport model can be used (1) to understand the fundamental mechanism governing uncharged, polar organic solute rejection and (2) to make forward predictions about how one might alter NF/RO membrane chemistry and structure to achieve a desired level of rejection. Perhaps a disappointing conclusion is that, for uncharged, polar organic solutes, the only obvious means of improving rejection is to reduce pore size, increase barrier layer thickness, and/or decrease barrier layer porosity, all of which reduce water permeability and energy efficiency. However disappointing this conclusion may be, we believe the fundamental insights that can be derived from this work are significant and important for the drinking water and water reuse scientific communities.

Application of a lyotropic liquid crystal nanofiltration membrane for hydraulic fracturing flowback water: Selectivity and implications for treatment

A thin-film composite, bicontinuous cubic lyotropic liquid crystal polymer (TFC QI) membrane with uniform-size, ionic nanopores was studied for the treatment of hydraulic fracturing flowback water. The TFC QI membrane performance was compared to those of a commercial nanofiltration (NF) membrane (NF270) and a commercial reverse osmosis (RO) membrane (SW30HR) for the filtration of flowback water from the Denver-Julesburg Basin. The permeability, salt rejection, and organic solute rejection for each membrane was evaluated. The results illustrate that the TFC QI membrane maintained its performance to a similar degree as the commercial NF and RO membranes while demonstrating a unique selectivity not observed in the commercial membranes. Specifically, the TFC QI membrane rejected 75% of the salt while recovering 9.6% of the dissolved organic carbon (DOC) and 50% of the water. Of particular interest was the recovery of labile DOC, which was assessed through biodegradation experiments. Analysis following biodegradation of the TFC QI membrane permeate demonstrates the membrane's ability to recover labile DOC in a reduced-saline permeate. Improved recovery of labile DOC (increased to 22%) was demonstrated by reducing the pH of the flowback water. Therefore, the selectivity of the TFC QI membrane provides an opportunity to recover resources from hydraulic fracturing flowback.

Application of quantitative structure–property relationships (QSPRs) to predict the rejection of organic solutes by nanofiltration

Recent interest in quantitative structure property/activity relationship (QSPR/QSAR) models to predict the removal of organic contaminants by membrane processes has highlighted the need to develop models applicable to different operating conditions, such as flux. In this study, two types of QSPR models were developed to predict removal of nonionic organic compounds by a nanofiltration membrane (NF270) including: (1) QSPR models with flux as an independent (or predictor) variable; and (2) QSPR models to predict fitting parameters of a fundamental model (i.e., Spiegler–Kedem model). Rejection data for 67 nonionic organic compounds and an NF membrane at five solvent fluxes (10–120Lm−2h−1) were generated and used for model development and validation. In order to select the best statistical method to develop QSPRs, several models were developed using multiple linear regression (MLR), partial least squares (PLS), and artificial neural network (ANN) with molecular descriptors selected by different methods (i.e., filtering, hybrid, and wrapper methods). The most effective linear and nonlinear models were developed using PLS and ANN with molecular descriptors selected by variable importance plot and feed forward selection. The rejection of nonionic compounds with minimal solute-membrane affinity could be described by permeate flux, and a compound’s molecular depth and diffusion coefficient. Fitting parameters of Spiegler–Kedem model (reflection coefficient and permeability coefficient) could be best described by size parameters and diffusivity. These results indicate that the rejection of the majority of the nonionic organic compounds evaluated was mainly due to size exclusion. Generally, both QSPR methods developed during this study were found to be effective methods for predicting solute rejection by the nanofiltration membrane, and nonlinear models provided better fits (based on statistical parameters) of both training and validation rejection datasets.

Maximal pore size in UF membranes

The ultrafiltration membrane rejection capability is most often characterized by molecular weight cut-off (MWCO). The value is found by rejection of organic solutes and the evaluation of particle retention requires a conversion of either MWCO to pore size or particle diameter to molecular weight. The conversion affects the accuracy of reported values and results in a gap between reported and measured retentions.We suggest a novel, simple and effective pore size test based on synthesis and membrane transfer of rigid nanoparticles. Gold and silver 3–50nm monodispersions had delivered a comprehensive pore size distribution including d100, a pore diameter for which a membrane has a 100% retention capability. The maximal pore size in UF membrane structure can hardly be detected with other methods although it is much needed for precise separation analysis.The d100 values in tested UF membranes vary between 40nm and 50nm depending on the membrane material. The polymer membranes are more flexible than the ceramics and their d100 is usually much higher than MWCO. The d100 increases with high transmembrane pressure or after oxidative chemical cleaning. For some membranes the d100 values can be correlated with d90 but not with d50.

Water filtration performance of a lyotropic liquid crystal polymer membrane with uniform, sub-1-nm pores

The water filtration performance of a supported, nanostructured, lyotropic (i.e., surfactant) liquid crystal (LLC) polymer membrane containing uniform, sub-1-nm pores was examined in more detail using more extensive aqueous filtration tests and a more structurally representative nanopore model. This nanoporous membrane material is based on the radical cross-linking of an ionic, LLC monomer that forms a type I bicontinuous cubic (Q I) phase with a 3D interconnected, annulus-like, water pore network. Previously, initial water transport studies using a limited number of solutes demonstrated initial proof-of-concept that this LLC polymer membrane can reject small molecules based on molecular size discrimination. In this study, more comprehensive transport experiments using aqueous solutions containing a wider range of different size neutral organic molecules and salt ions were undertaken. The results of these experiments showed that the LLC polymer membrane has a rejection performance in between that of a commercial, high-performance reverse osmosis (RO) membrane (Dow SW30HR) and that of a conventional porous nanofiltration (NF) membrane (Dow NF-270) with non-uniform pore sizes. Specifically, the LLC membrane was found to reject inorganic salt ions much better than a NF membrane, and on a level comparable to a RO membrane. It does not reject small neutral organic solutes as well as a RO membrane, but it was found to reject organic solutes better than a conventional NF membrane. Also, the organic solute rejection behavior was fitted using two different Donnan-steric pore models (DSPM) with different pore shapes (i.e., cylindrical and straight slit). It was determined that the Q I-phase LLC membrane has an effective pore radius between 0.29 nm and 0.45 nm. This modeling also suggested that the LLC nanopores are indeed uniform in size. Initial chemical stability and fouling experiments suggest this nanoporous LLC polymer membrane is also very resistant to chlorine degradation and protein fouling. This combination of properties makes this nanoporous LLC polymer a very promising membrane material for water desalination and other NF processes.

Artificial neural network models based on QSAR for predicting rejection of neutral organic compounds by polyamide nanofiltration and reverse osmosis membranes

Prediction of the rejection of neutral organic compounds by polyamide nanofiltration (NF) and reverse osmosis (RO) membranes was tested in this study, using artificial neural network (ANN) models. The ANN models were based on a quantitative structure–activity relationship (QSAR) equation that defined an appropriate set of solute and membrane variables able to represent and describe rejection. The QSAR equation was defined using principal component analysis and multiple linear regression. The QSAR model combines interactions between solute properties, membrane characteristics and operating conditions. Solute size descriptors (molecular length and equivalent width) and the hydrophobicity descriptor represented by log K ow, were the main physicochemical properties of the neutral organic compounds able to describe rejection. Regarding the membrane characteristics, it was found that salt rejection (in combination with the aforementioned solute properties) can be practically applied to predict organic solute rejection. An experimental database produced by the authors in combination with data collected from literature comprising a total of 161 rejections of 50 organic neutral compounds (including pharmaceuticals, endocrine disrupting compounds, pesticides, alcohols, phenols and solvents) by six NF and nine RO membranes was used to produce the ANN models. Sixty percent of the data were used to generate the model, and validation was performed with 20% of the total data. Independent rejection predictions were calculated for the remaining 20% of the data. For the most promising ANN models, the independent predicted rejection values were compared to measured rejections and good correlations were found ( R 2 = 0.97) and predicted rejections presented and standard deviation of error of ca. 5%.

Influence of Solute−Membrane Affinity on Rejection of Uncharged Organic Solutes by Nanofiltration Membranes

A simple, analytical method for predicting transport of uncharged organic solutes through nanofiltration (NF) and reverse osmosis (RO) membranes is presented in this paper. The method requires characterization of key solute and membrane parameters-namely, solute size, membrane pore size, and solute-membrane affinity. All three parameters can be experimentally determined from relatively simple permeation tests and contact angle analyses. The parameters are fed into an analytical model of solute transport, which accounts for hindered convection and diffusion of solutes in the membrane pores, as well as the combined effects of steric exclusion and solute-membrane affinity on solute partitioning from the feed solution into the membrane pores. Overall model predictions for organic solute rejection agreed well with experimental data for three different solutes and two different polymeric NF membranes. Further, the model demonstrates the dramatic influence of solute-membrane affinity on organic rejection by NF and RO membranes. Solute transport predictions made assuming only steric exclusion significantly overestimated rejections for solutes with strong affinity for membrane polymers and similarly underestimated rejections for solutes that were strongly repelled by membrane polymers.

The role of electrostatic interactions on the rejection of organic solutes in aqueous solutions with nanofiltration

The effects of electrostatic interactions on the rejection of organic solutes with nanofiltration membranes were investigated. For two different membranes, the rejection of selected organic acids, positively and negatively charged pharmaceuticals and neutral pharmaceuticals was investigated at different feed water chemistries (different ionic strengths and pH conditions, with and without the presence of NOM and divalent cations). It was concluded that for negatively charged membranes, electrostatic repulsion leads to an increase of the rejection of negatively charged solutes and electrostatic attraction leads to a decrease of the rejection of positively charged solutes, compared to neutral solutes. Neutral and positively charged solutes engage in hydrophobic interactions with negatively charged membranes, whereas negatively charged solutes do not engage in hydrophobic interactions since they cannot approach the membrane surface. This provides proof for the theory of an increased concentration of positively charged organic solutes and a decreased concentration of negatively charged organic solutes at the membrane surface compared to the bulk fluid. This concept may be denoted as “charge concentration polarisation”. The concept was further used as a modelling tool to predict the effects of electrostatic interactions on the rejection of trace organic solutes.

Thermodynamic Factors in Partitioning and Rejection of Organic Compounds by Polyamide Composite Membranes

The paper analyzes the mechanism of partitioning and rejection of organic solutes by polyamide membranes for reverse osmosis and nanofiltration. The partitioning of homologous series of alcohols and polyols, in which polarity changes with size in opposite ways, was measured using attenuated total reflection IR spectroscopy. The results show that the partitioning of polyols monotonously decreases with size, whereas for alcohols it is not monotonous and slightly decreases for small C1-C3 alcohols followed by a sharp increase for larger alcohols. These results may be explained by assuming a heterogeneous structure of polyamide comprising a hydrophobic polyamide matrix and a polar internal aqueous phase. The partitioning data could consistently explain the results of rejection in standard filtration experiments. They clearly demonstrate that high/low partitioning may play a significant role in achieving a low/high rejection of organics. In particular, this points to the need to account for the partitioning effect while using molecular probes such as polyols or sugars for estimating the effective "pore" size or molecular weight cutoff of a membrane and for choosing/developing organic-rejecting membranes.

A comparison of nanofiltration with aqueous and organic solvents

This paper reports rejection of organic solutes with molecular weight (MW) in a range of 180–600 in aqueous and organic solvents through several commercial solvent resistant nanofiltration membranes (SRNF). The organic solutes investigated include charged and neutral molecules and the selected solvents are water, water–methanol mixture, methanol, ethanol, acetone and ethyl acetate. The higher rejection in water than in organic solvents was observed for Desal-DK, MPF-44 and STARMEM™ series membranes. However, MPF-50 showed “uncommon” rejection characteristics of higher rejection in organic solvents (methanol, ethanol and acetone) than in water for neutral molecules compared to charged molecules. The rejection was analysed in accordance with the charge effect and hydration/solvation mechanism.

High Flux Filtration Medium Based on Nanofibrous Substrate with Hydrophilic Nanocomposite Coating

A novel high flux filtration medium, consisting of a three-tier composite structure, i.e., a nonporous hydrophilic nanocomposite coating top layer, an electrospun nanofibrous substrate midlayer, and a conventional nonwoven microfibrous support, was demonstrated for oil/water emulsion separations for the first time. The nanofibrous substrate was prepared by electrospinning of poly(vinyl alcohol) (PVA) followed by chemical cross-linking with glutaraldehyde (GA) in acetone. The resulting cross-linked PVA substrates showed excellent water resistance and good mechanical properties. The top coating was based on a nanocomposite layer containing hydrophilic polyether-b-polyamide copolymer or a cross-linked PVA hydrogel incorporated with surface-oxidized multiwalled carbon nanotubes (MWNTs). Scanning electron microscopy (SEM) examinations indicated that the nanocomposite layer was nonporous within the instrumental resolution and MWNTs were well dispersed in the polymer matrix. Oil/ water emulsion tests showed that this unique type of filtration media exhibited a high flux rate (up to 330 L/m2-h at the feed pressure of 100 psi) and an excellent total organic solute rejection rate (99.8%) without appreciable fouling. The increase in the concentration of surface-oxidized MWNT in the coating layer generally improves the flux rate, which can be attributed to the generation of more effective hydrophilic nanochannels for water passage in the composite membranes.

Rejection mechanisms of plastic additives and natural hormones in drinking water treated by nanofiltration

Rejection mechanism of nanofiltration (NF) membrane for organic solutes (including alkyl phenol, bisphenol A, phthalate acid esters, phosphoric acid esters, other plastic additives, and natural hormones like 17β estradiol, etc) was examined. The experiments indicated almost complete removal of these solutes by the tight NF membranes, while the rejection obtained by the loose membranes was dependent on solute species. Compounds with flat-shaped structure like 2-H-benzothiazol and dimethyl phthalate were more permeable than other target solutes within the same molecular size range. Of the four methods of size determination used, there was no significant difference in the accuracy of prediction of rejection from the solute sizes. The size configurations of molecules have been identified to greatly influence the permeation or retention, yet the size parameter was thought to be incomplete as a single parameter in describing rejection of organic solutes. The logarithm of octanol-water partitioning coefficients of solutes appears not to be an effective parameter in modelling rejections as its contribution was only slight or minor. We obtained better correlations with structural configuration of molecules (by molecular width and length) and polarity of the molecules.

Nanofiltration membranes based on poly(vinyl alcohol) and ionic polymers

Nanofiltration membranes based on poly(vinyl alcohol) (PVA) and ionic polymers, such as sodium alginate (SA) and chitosan, were prepared by casting the respective polymer solutions. The membranes prepared from PVA or PVA–ionic polymer blend were crosslinked in a isopropanol solution using glutaraldehyde as a crosslinking agent. The membranes were characterized with Fourier transform infrared spectroscopy and X-ray diffractometry and swelling test. The membranes crosslinked through the acetal linkage formation between the OH groups of PVA and the ionomer and glutaraldehyde appeared to be semicrystalline. To study the permeation properties, the membranes were tested with various feed solutions [sodium sulfate, sodium chloride, poly(ethylene glycol) with 600 g/mol of molecular weight (PEG 600), and isopropyl alcohol]. For example, the permeance and the solute rejection of the 1000 ppm sodium sulfate at 600 psi of upstream pressure through the PVA membrane were 0.55 m3/m2 day and over 99%, respectively. The effects of the ionomers on the permeation properties of the PVA membranes were studied using the PVA–SA and PVA–chitosan blend membranes. The addition of small amount of ionic polymers (5 wt %) made the PVA membranes more effective for the organic solute rejection without decrease in their fluxes. The rejection ratios of the PEG 600 and isopropanol were increased substantially. © 1999 John Wiley & Sons, Inc. J Appl Polym Sci 72: 1755–1762, 1999

逆浸透膜による低分子量有機物水溶液の分離特性の研究 その１ ＮＳ‐１００および酢酸セルロース非対称膜

Reverse osmotic rejections of some organic acids (acetic acid, propionic acid, butyric acid, valeric acid, succinic acid) and aldehydes (acetaldehyde, propionaldehyde, butyraldehyde) in aqueous solution have been studied using two composite membranes (NS-100 (P3500), NS-100 (P1700)) and cellulose acetate membrane. Separation of solute and permeation through the membrane were measured under the following condition : concentration of solute in most cases is in the range of 50-500 ppm. operating pressure 0.98-2.94 MPa. An analysis of data with Spiegler-Kedem's transport model was carried out to obtain membrane constants such as reflection coefficient σ, solute and hydraulic permeabilities ω and Lp.The result showed that NS-100 membrane gave the superior solute rejection and superior permeation flux. Its rejection was in the range of 40-90%. CA membrane gave the higher salt rejection, but had lower organic solute rejection and permeation flux. In the case of the concentration of solution bellow the 1000 ppm, the rejection does not depend on concentration. But beyond the 1000 ppm, the rejection decreases with the increase of concentration. For organic compounds having the same functional group, the rejection of a solute increases with the increase of molecular weight. Using NS-100 membrane, for organic solute of acids and aldehydes having the same carbon atom numbers, the rejection for acids is lower. This might be attributed to the electrostatic attraction force acting between the solute molecular and skin layer of the NS-100 membrane.

Application of regular solution theory to organic solute rejection by reverse osmosis

Regular solution theory was applied to the prediction of reverse osmosis rejection of organic solute from aqueous solution, assuming that the rejection is primarily determined by the distribution of solute between the membrane and the aqueous solution. It indicates that this treatment can serve as a basis for predicting solute rejection in reverse osmosis if the membrane parameters are estimated reasonably.

Factors influencing reverse osmosis rejection of organic solutes from aqueous solution

Factors influencing reverse osmosis rejection of organic solutes from aqueous solution