scispace - formally typeset
Search or ask a question
Journal ArticleDOI

A method for the simultaneous determination of transport and structural parameters of forward osmosis membranes

01 Oct 2013-Journal of Membrane Science (Elsevier)-Vol. 444, pp 523-538
TL;DR: In this paper, a simple and rapid methodology to characterize the water and solute permeability coefficients (A and B, respectively) and structural parameter (S) of forward osmosis (FO) membranes is presented.
About: This article is published in Journal of Membrane Science.The article was published on 2013-10-01 and is currently open access. It has received 394 citations till now. The article focuses on the topics: Forward osmosis & Membrane.

Summary (5 min read)

1. Introduction

  • Forward osmosis (FO) utilizes the osmotic pressure difference developed across a semi-permeable membrane separating two solutions of different concentrations to drive the permeation of water [1].
  • In particular, thin-film composite (TFC) ll rights reserved.
  • A convenient and consistent methodology to characterize FO membranes is of critical importance to advance this technology onto its mature phase, facilitating the sharing of data, their interpretation, and comparison.
  • These three parameters are univocal and can be used with the respective governing equations to accurately predict the water and salt flux performance of a membrane sample in any laboratory-scale FO system.
  • These protocols are cumbersome and laborious, requiring multiple experiments in different experimental setups.

2. A single FO experiment to characterize osmotic membranes

  • A single and facile FO experiment is proposed to characterize the intrinsic transport parameters, A and B, and the structural parameter, S, of an FO membrane by measuring the water and reverse solute flux across the membrane under different draw solution concentrations.
  • In their study, the experiments were carried out in four stages.
  • At the end of the first stage, a known volume of concentrated draw ities are schematically represented as lines across the time scale for each of the four ntration (red), cF, are represented as single lines in the top plot.
  • The osmotic pressure and salt concentration difference across the membrane increased and, as a result, both the FO water flux and the reverse solute flux augmented to reach values Jw,2 and Js,2, respectively, in stage 2.
  • Values of water flux and reverse solute flux were experimentally measured at every stage.

3.1. Water and salt flux governing equations in FO

  • The mass transport across a membrane in FO can be expressed in terms of the membrane characteristic properties, the hydrodynamics in the membrane flow cell, and experimentally accessible parameters: the bulk solute concentration of the draw solution, cD, feed solution concentration, cF, and the corresponding osmotic pressures, πD and πF.
  • The water permeability coefficient, A, and salt permeability coefficient, B, are intrinsic properties of the membrane active layer.
  • The support layer structural parameter, S, is defined as tsτ=ε, with ts being the thickness of the support layer, τ its tortuosity, and ε its porosity.
  • In these equations, the terms expðJw=kÞ and expð−JwS=DÞ account for concentrative external concentration polarization (ECP) and dilutive internal concentration polarization (ICP), respectively.

3.2. Calculating A, B, and S numerically by minimization of a global error

  • The authors have developed an algorithm to calculate the membrane parameters, A, B, and S, from experimental water and salt flux data.
  • The FO transport Eqs. (1) and (2) were fitted to the experimental fluxes by least-squares non-linear fitting, using A, B, and S as regression parameters, and cD (πD) and cF (πF) as independent experimental variables.
  • The input parameters included the average draw and feed solution concentrations in each stage, system temperature, salt diffusion coefficient in the bulk solution, measured water and salt fluxes, and the initial guesses for the parameters to be calculated.
  • Details on the theory underlying each of the numerical methods may be found in standard numerical methods textbooks [38].

3.3. Experimental and modeling assumptions

  • That is, the membrane has a dense selective layer that is able to maintain virtually the entire osmotic pressure difference across it (i.e., s approaches unity) [39].
  • ICP is partially mitigated by transport of solute from the draw solution into the membrane support layer.
  • It is important to note that, while the draw solute diffusivity is dependent on the local salt concentration—and therefore differs across the membrane support layer—the mass transport model from which Eqs. (1) and (2) were derived assumes a constant D within the membrane support layer.
  • This relation was applicable for the procedure outlined here because the solute concentration at the support/ active layer interface was significantly lower than that of the bulk solutions, due to the effects of ICP.
  • From this assumption follows a linear relationship between the experimental value of Jw/Js and the ratio A/B, namely ðJw=JsÞ ¼ ðA=BÞυRgT [40].

4.1. FO membranes

  • Both hand-cast TFC and commercial membranes were characterized.
  • To begin casting the membrane, a commercial polyester non-woven fabric (PET, Grade 3249, Ahlstrom, Helsinki, Finland) was taped on a glass plate and was wet with 1-methyl-2pyrrolidinone (NMP, anhydrous, 99.5%, Sigma-Aldrich).
  • Additionally, prototype thin-film composite FO membranes were obtained from Oasys Water (Oasys Water Inc., Boston, MA).
  • All membranes were thoroughly wet prior to the experiments by immersing in 25% isopropanol solutions for 30 min.

4.2. FO setup and experimental conditions

  • FO water fluxes and reverse solute fluxes were determined in an experimental cross-flow FO system described in their previous studies [22,35].
  • All characterization tests were conducted with the membrane in FO configuration, i.e., porous support layer facing the draw solution and active layer facing the feed solution.
  • For FO characterization tests, a stock 5 M NaCl solution was prepared with sodium chloride (NaCl) from J.T. Baker (Phillipsburg, NJ), by dissolving the appropriate amount of NaCl in DI water (Milli-Q, Millipore, Billerica, MA).
  • An appropriate volume of the NaCl stock solution was added to the draw solution to obtain the desired concentration and initiate the first stage.
  • After attainment of steady state, the water flux, JEXPw;i , was determined by monitoring the rate of change in weight of the draw solution, and the solute concentration in the feed was measured at 1 min intervals with a calibrated conductivity meter (Oakton Instruments, Vernon Hills, IL).

4.3. Experimental results

  • The complete set of water and salt fluxes measured during the experiments, together with the draw and feed solution concentrations and osmotic pressures.
  • Fig. 2 presents the experimental results plotted as the Jw/Js values against the bulk osmotic pressure difference between the feed and the draw side.
  • The latter was calculated from NaCl concentrations using the van’t Hoff equation.
  • The percent figure beside each data series represents the coefficient of variation (CV) calculated between the Jw/Js values of the different stages in each experiment.
  • Different draw solute concentrations were employed for the various membrane types, due to their dissimilar performance.

4.4. Calculated parameters

  • Table 1 summarizes the transport parameters and the related coefficients of determination, calculated by the model for each experiment.
  • The Matlabbased algorithm produced practically identical results.
  • The values of the measured and calculated water and salt fluxes are compared on a linear–linear plot in Fig. 3 (solid symbols).
  • The equivalence among the two separate algorithms, the robustness of the calculation with respect to the initial parameter estimates, and the high coefficients of determination suggest that the methodology is reliable and the calculation codes well posed.

5.1. Comparison to current approach

  • It is important to make a comparison between the parameters calculated using their method and those obtained by the current approach, whereby membrane active layer properties are measured in an RO experiment and the structural parameter is calculated using additional FO measurements.
  • The values of SRO+FO presented in Table 2 are the average results obtained by applying ARO and BRO in the four stages of each respective FO experiment.
  • Agreement between the active layer properties determined by the two methods was generally poor.
  • The average discrepancy is approximately 46%, with the largest deviation observed in the membrane salt permeabilities.
  • Overall, membrane parameters derived using the RO-FO method gave poorer prediction of the water and salt fluxes, compared to the FO-only approach proposed here (solid symbols).

5.2. The observed discrepancy requires further analysis

  • A noteworthy trend emerged upon examining the calculated A/B ratios, a parameter quantifying the permeability–selectivity of the membrane active layer (Table 2).
  • Other work has also reported disagreement between recorded values of the A/B ratio (or the equivalent Jw/Js ratio) obtained in FO and in RO experiments [45,46].
  • The difference in membrane performance reported in past studies suggests that the driving force has certain influence on mass transport across the membrane active layer, and plausibly explains the discrepancies observed here.
  • This is only one hypothesis and further investigations on the phenomenon are necessary.

5.3. Influence of the number of stages

  • The least-squares non-linear regression method presented here utilizes an over-determined system of equations to calculate the membrane properties in FO.
  • The calculated membrane properties of the simulated characterization experiments were normalized (i.e., divided by the true value) and presented in Fig. 5(A–C) as a function of R2 for both water and reverse salt flux (blue cross and red plus symbols, respectively).
  • First, the symbols gradually spread to the left (i.e., lower coefficients of determination) to indicate relatively poorer fits and, second, the data points cluster towards the normalized value of one (horizontal dashed line).
  • Thus, setting a stricter requirement for the coefficients of determination, an indicator of water and reverse salt flux measurement quality, eliminates the weaker data and minimizes the likelihood of the calculated properties deviating exceedingly from the actual values.
  • As the number of stages increases, the calculated membrane parameters fall closer to the true values for the same coefficient of determination benchmark, but the fraction of the sample population fulfilling the criteria is also reduced.

5.4. Influence of the consistency in experimental Jw/Js throughout the stages

  • The reverse flux selectivity, the ratio of water flux to reverse salt flux, is a constant factor that is solely dependent on the intrinsic membrane active layer characteristics, A and B, for the solution concentrations used in this study (Section 3.3).
  • The calculated membrane parameters are normalized by the true values and plotted in Fig. 6 (A) as a function of the coefficient of variation of Jw/Js (CV, defined as the standard deviation divided by the arithmetic mean) for the analysis performed in Section 5.3.
  • Only the results for characterizations carried out in four stages are presented.
  • This is due to the higher data quality expected at lower CVs.
  • Therefore, the CV of Jw/Js can be employed as a prerequisite for screening data reliability—conspicuous fluctuations would suggest large experimental errors beyond the tolerance of the algorithm.

5.5. Outlook and recommendations

  • Analysis of the robustness and sensitivity of the proposed methodology suggests that a first requirement for its implementation is a consistent experimental value of Jw/Js between the various stages of each FO experiment.
  • Results the references to color in this figure legend, the reader is referred to the web version analysis.
  • A second control for the reliability of the results can be identified in the coefficients of determination for both water and salt flux.
  • When one or both fall below 0.9, results should be taken cautiously.
  • Meeting the dual conditions of CV of Jw/Js lesser than 10% and both Rw2 and Rs2 greater 0.95 will likely yield calculated parameters that are off by 10% or lower from their true values.

6. Concluding remarks

  • The complete equations governing mass transport in FO are presented and used in combination with experimental measurements to determine the target parameters by an error minimization algorithm.
  • Thorough analysis indicates that the deviation of the calculated parameters from their true values is likely to be less than 3.5%, further reinforcing the robustness of the protocol.
  • For all TFC membranes, the A/B ratio measured by the FO methodology presented here was considerably lower (consistently between 30 and 50%) than that determined by RO experiments.
  • This observation casts doubt on the theoretical notion that membrane parameters are conserved in pressure- and osmotically-driven configurations.
  • Regardless of the exact reasons for this disagreement, the difference in calculated membrane parameters and the greater prediction accuracy of the methodology presented here indicate that FO membranes should be characterized only by means of FO experiments.

Acknowledgments

  • The authors acknowledge the support received from the National Science Foundation under Award Number CBET 1232619.
  • The authors also acknowledge the Graduate Fellowship (to N.Y.Y.) made by the Environment and Water Industrial Development Council of Singapore and the National Science Foundation Graduate Research Fellowship awarded to A.P.S.

Did you find this useful? Give us your feedback

Figures (8)
Citations
More filters
Journal ArticleDOI
TL;DR: In this paper, the energy efficiency of the forward osmosis (FO) process is analyzed and the potential use of low-cost energy sources is highlighted, emphasizing the importance of the structural parameter, reverse solute flux selectivity, and the constraints imposed by the permeability selectivity tradeoff.

686 citations


Cites background or methods from "A method for the simultaneous deter..."

  • ...As discussed in Section 3, TFC membranes are the current state-of-the-art for FO because they exhibit a higher water permeability and salt rejection than CTAmembranes [5]....

    [...]

  • ...RSFS represents the volume of water produced per mass of draw solute lost, and it follows a simple, experimentally-verified relationship based on the van't Hoff equation [9,10]:...

    [...]

  • ...These studies on FO involved membrane development [5–8], mass transfer analysis [4,9], membrane characterization [10,11], fouling phenomena [12–16], and introduction and characterization of new draw solutions [17–20]....

    [...]

  • ...However, with the recent commercialization of TFC-FO membranes, which havemuch higher RSFS than CTAmembranes [10,60], reverse draw solute flux for small inorganic solutes has been greatly reduced....

    [...]

  • ...The water flux model is based on the film theory [10]:...

    [...]

Journal ArticleDOI
TL;DR: In this article, pressure-retarded osmosis (PRO) is investigated as a method to extract energy from salinity gradients, and the authors define the maximum energy that can be obtained from the process, quantify losses and energetic costs that will reduce the net extractable energy, and explain how membrane modules can be improved.
Abstract: The enormous potential of harvesting energy from salinity gradients has been discussed for decades, and pressure-retarded osmosis (PRO) is being increasingly investigated as a method to extract this energy. Despite advancements in membranes and system components, questions still remain regarding the overall viability of the PRO process. Here, we review PRO focusing on the net energy extractable and the ultimate feasibility of the most widely explored configurations. We define the maximum energy that can be obtained from the process, quantify losses and energetic costs that will reduce the net extractable energy, and explain how membrane modules can be improved. We then explore the potential of three configurations of PRO: systems designed to control mixing where rivers meet the sea, power plants that utilize the high concentration gradients available from hypersaline solutions, and PRO systems incorporated into reverse osmosis desalination plants to reduce electricity requirements. We conclude by considering the overall outlook of the process and identifying the most pressing challenges for future research.

297 citations

Journal ArticleDOI
TL;DR: In this article, the authors review mechanisms and models of solute transport relevant to nanofiltration (NF), reverse osmosis (RO), and forward Osmosis(FO) membrane separation processes.

262 citations

Journal ArticleDOI
TL;DR: Osmotic membrane bioreactor (OMBR) is an emerging technology integrating a forward osmosis (FO) process into a membrane Bioreactor as discussed by the authors, which has been gaining increasing popularity in wastewater treatment and reclamation due to its excellent product water quality, low fouling tendency and high fouling reversibility over conventional MBRs.

229 citations

Journal ArticleDOI
TL;DR: In this article, a systematic investigation on the influence of support layer pore size on the osmotic performance of thin film composite membranes is conducted for the first time, where TFC membranes were made by interfacial polymerization to form a polyamide selective layer on top of commercially available nylon 6,6 microfiltration membranes with similar physical and chemical properties but different pore sizes.

228 citations

References
More filters
Journal ArticleDOI
05 Aug 2011-Science
TL;DR: The possible reductions in energy demand by state-of-the-art seawater Desalination technologies, the potential role of advanced materials and innovative technologies in improving performance, and the sustainability of desalination as a technological solution to global water shortages are reviewed.
Abstract: In recent years, numerous large-scale seawater desalination plants have been built in water-stressed countries to augment available water resources, and construction of new desalination plants is expected to increase in the near future. Despite major advancements in desalination technologies, seawater desalination is still more energy intensive compared to conventional technologies for the treatment of fresh water. There are also concerns about the potential environmental impacts of large-scale seawater desalination plants. Here, we review the possible reductions in energy demand by state-of-the-art seawater desalination technologies, the potential role of advanced materials and innovative technologies in improving performance, and the sustainability of desalination as a technological solution to global water shortages.

4,840 citations

Book
16 Dec 1999
TL;DR: Overview of membrane science and technology membrane transport theory membrane and modules concentration polarization reverse osmosis ultrafiltration microfiltration gas separation pervaporation ion exchange membrane processes - electrodialysis carrier facilitated transport medical applications of membranes other membranes processed.
Abstract: Overview of membrane science and technology membrane transport theory membrane and modules concentration polarization reverse osmosis ultrafiltration microfiltration gas separation pervaporation ion exchange membrane processes - electrodialysis carrier facilitated transport medical applications of membranes other membranes processed.

3,680 citations


"A method for the simultaneous deter..." refers methods in this paper

  • ...The water flux, Jw, across the active layer is given by [42] Jw 1⁄4 AðΔπmÞ ðA1Þ...

    [...]

  • ...The whole composite was immediately immersed in a precipitation bath containing 3 wt% DMF in deionized (DI) water at room temperature to initiate non-solvent induced phase separation [41,42]....

    [...]

Book
01 Jan 1966
TL;DR: Reference Record created on 2005-11-18, modified on 2016-08-08 as discussed by the authors, created on 2011-11 -18, created on 2006-11 18 and modified on 2008-08 -08,08
Abstract: Keywords: analyse ; methodes : numeriques ; equations : lineaires ; calcul : integral ; equations : differentielles Reference Record created on 2005-11-18, modified on 2016-08-08

2,377 citations

Journal ArticleDOI
TL;DR: In this paper, the state-of-the-art of the physical principles and applications of forward osmosis as well as their strengths and limitations are presented, along with a review of the current state of the art.

2,235 citations

Journal ArticleDOI
TL;DR: In this article, a review of new experimental and theoretical physical research related to the formation of polymeric membranes by phase separation of a polymer solution, and to the morphology of these membranes is presented.

1,246 citations


"A method for the simultaneous deter..." refers methods in this paper

  • ...The whole composite was immediately immersed in a precipitation bath containing 3 wt% DMF in deionized (DI) water at room temperature to initiate non-solvent induced phase separation [41,42]....

    [...]

Frequently Asked Questions (18)
Q1. What contributions have the authors mentioned in the paper "A method for the simultaneous determination of transport and structural parameters of forward osmosis membranes" ?

The authors present a simple and rapid methodology to characterize the water and solute permeability coefficients ( A and B, respectively ) and structural parameter ( S ) of forward osmosis ( FO ) membranes. The authors compare the membrane properties obtained with their FO-based methodology with those derived from existing protocols based on an RO experiment followed by an FO experiment. Additionally, the existing and proposed approaches yield consistently dissimilar results for some of the analyzed membranes, indicating a discrepancy that might be attributed to the different driving forces utilized in RO and in FO that should be further investigated. 

Water permeating across the active layer dilutes the draw solution in the support layer, resulting in dilutive ICP, the effect of which is to decrease the net osmotic driving force. 

To simulate uncertainty in experimental measurements, errors ranging from 0 to 15% were deliberately introduced to the water and reverse salt flux readings. 

As the CV of Jw/Js approaches zero, the normalized parameters converge towards unity (horizontal dashed line), signifying the membrane properties are more accurately predicted. 

Due to both the permeation of water and the salt leakage, the concentration of thedraw solution was diluted throughout the experiment. 

The least-squares non-linear regression method presented here utilizes an over-determined system of equations to calculate the membrane properties in FO. 

The membrane characteristic parameters can be determined numerically by solving a system of equations if all the other variables, water and salt flux, feed channel mass transfer coefficient, salt diffusivity, and concentrations or osmotic pressures of the solutions, are known. 

The support membrane remained in the precipitation bath for 10 min before being transferred to a DI water bath for storage until polyamide (PA) formation. 

Constancy of the reverse flux selectivity between the characterization stages gives not only better prediction of membrane parameters, but also enhances the fit of the calculated fluxes to the experimental readings. 

A convenient and consistent methodology to characterize FO membranes is of critical importance to advance this technology onto its mature phase, facilitating the sharing of data, their interpretation, and comparison. 

The osmotic pressure and salt concentration difference across the membrane increased and, as a result, both the FO water flux and the reverse solute flux augmented to reach values Jw,2 and Js,2, respectively, in stage 2. 

The simulated fluxes (offset by random errors), along with the corresponding draw and feed solution osmotic pressures and concentrations, were input into the Excel algorithm described in Section 3 to determine the membrane parameters A, B, and S that best fit the simulated Jw and Js “measurements”. 

In addition, the approximation was justified by the hydrodynamic conditions maintained at the stirred boundary layer on the feed side, where typical values of the feed solution mass transfer coefficient, k, far exceed the permeating water flux. 

Although the sodium chloride draw solution deviates from ideal behavior at high concentrations, this relation was applicable for the procedure outlined here because the solute concentration at the support/ active layer interface was significantly lower than that of the bulk solutions, due to the effects of ICP. 

An examination of the figure shows that attaining good CV and R2 values concurrently is essential to minimize errors in the determined parameters. 

The whole composite was immediately immersed in a precipitation bath containing 3 wt% DMF in deionized (DI) water at room temperature to initiate non-solvent induced phase separation [41,42]. 

The authors recommend a CV within 10% to confidently continue theanalyzed by the simulations described in Appendix D. (A) Simulated parameters, A, of variation for Jw/Js values of the different stages in each experiment (a value of f the average error of each simulated parameter, A, B, and S (blue, red, and green, nd salt (orange) fluxes, for chosen ranges of coefficient of variation obtained during ariation requirements is also reported on top of each set of histograms. 

The deviation of the calculated values was especially pronounced for the reverse salt fluxes, where the higher reverse flux selectivity (A/B) determined for TFC membranes using the RO-FO method (Table 2) resulted in predicted salt fluxes that are significantly lower than the experimental FO values (Fig. 3(E), (G), and (H)).