Can the variance in membrane performance influence the design of organic solvent nanofiltration processes?

doi:10.1016/j.memsci.2018.12.077

Journal of Membrane Science

Volume 575, 1 April 2019, Pages 217-228

https://doi.org/10.1016/j.memsci.2018.12.077 Get rights and content

Highlights

•
Standard experimental procedure to characterize OSN membranes.
•
High-quality rejection and flux data of membrane-solvent-solute systems.
•
Best case approximations of comparability from experimental data.
•
Statistical analysis and evaluation of results from five research groups.
•
Effect of uncertainties of flux and retention measurement on process design.

Abstract

The development of organic solvent nanofiltration (OSN) membranes has been a continuous effort during the past decade. Several groups generated and published experimental results and simulations with either tailor-made membranes or industrial products. The published data space is diverse, with a single publication often focusing on a certain membrane only, a group of specific solutes only, on the effect of solvent only. A comprehensive comparison of all of these is still lacking. An analysis on the reliability of quantified transport parameters is missing, in particular when different analysis methods and different membrane systems are used. The technology is in its incubation phase with a need for reliable data and measurement standards to ground the process design on reliable membrane transport properties. This study uses an unprecedented extensive standard experimental procedure to measure separation characteristics of polymer membranes for the separation of organic solutes from organic solvents.

For a variety of different solvents, solutes and membranes, flux and retention measurements are rigorously characterized by round robin tests at different labs using different analytical systems. This extensive collaborative study evaluates for the first time which fluctuations in results occur under comparable conditions of lab-scale experiments at different locations and which influence these have on the design of OSN processes.

Utilizing the variance in the membrane transport data, a process design study and optimization is presented. For the first time ever, we report how variance in membrane transport properties can influence the process design ranging from a simple single stage system for values for the upper transport limits, to a two stage system with a permeate recirculation for mean values, to a complex three stage system for the lower limits.

Graphical abstract

Introduction

Taking membrane research from its initial phase of concept development, material and membrane design into process development requires substantial resources and time. This is not special but true for many innovations starting with an innovation trigger. This is also known as the Gartner Hype Cycle [1]. Judgment of the quality of the innovation trigger occurs often through reproducibility and applications studies and generally also creates - after a peak of inflated expectations - a trough of disillusion. In membrane research, the innovation trigger is assumed to occur through the availability of new materials and membranes made thereof. Yet, it is often not well known how statistically significant the mass transport properties of new membrane materials and new membrane products are. In a recent meta-study for example [2], published in this journal with more than 3000 references, the authors pose the fundamental questions (a) whether there is an average proton conductivity for Nafion membranes, (b) how the proton conductivity due to Nafion modification has evolved over past 10 years, (c) which additives really contribute to a conductivity increase, (d) how temperature or humidity affect conductivity. While such questions are not scientifically inspiring and do not initiate a new hype cycle, such questions are important in order to understand and judge the significance of a new material or membrane type. Here we pose similar questions with respect to the statistical significance and comparability of characterization methods to quantify mass transport properties for membranes used in organic solvent nanofiltration. As opposed to the meta-study, we have chosen to engage different labs and research groups with comparable analytical infrastructure, to use a round robin test evaluating the results of a developed standard experimental procedure. This leads to conclusions on standard deviations in flux and retention values in a complex parameter space for different membrane materials in various solvent/solute systems.

Section snippets

Background

Organic solvent nanofiltration (OSN) became a new member of the family of pressure-driven membrane processes by the discovery of solvent resistant membranes. Up to now, a great number of various applications were developed since the production and reactions in organic solvents comprises a wide range [3], [4]. Although OSN is still in the focus of industry as an energy-efficient separation process, there are only a few established plants [5], [6], [7]. However, the progress in the investigation

Materials and categorization

An appropriate categorization usable for test systems has to be as simple as possible and as accurate as necessary. Following this maxim, we decided to categorize the solvents in protic and non-protic solvents. In protic solvents, we extended the category with two degrees of polarity and for non-protic solvents with three degrees of polarity. Therefore, the normalized polarity index of Reichardt [32] supplies the values. The same as the polarity, in this way a wide range of the

Statistical analysis and evaluation

While it is often stated that comparable results have been obtained in different studies, these statements are qualitative and therefore subjective. To avoid such subjectivity and allow for a meaningful quantitative reference, we considered a statistical evaluation of the results. Thus, we defined a degree of congruence. The difference between a random sample and a reference or confidence intervals compared to each other are typically used. Confidence intervals should give the precision of an

Results and discussion

Considering the spent effort devoted to the standardized experimental procedure and the validation of the analytic methods, it can be concluded that the obtained results of the round robin test are a best case representative for the diversity of measurements in diverse laboratories. Differences are quite normal since locations, set-ups, measurement devices, analytic methods, time of measurements and experimenter vary from each other. The contributors endeavored to operate as equal as possible

Conclusion

The data quality of results from membrane characterizations has never been stated as a basis for the OSN users. This uncertainty hinders the progress in finding one or more reasonable standard systems for OSN membrane characterization. Such standardized procedure can simplify the assessment of membrane performance and their applicability in real separation processes. As a result, the process design of a OSN plant in industrial scale would be less difficult and troublesome. In this study, we

Acknowledgment

The authors wish to acknowledge the German Federal Ministry for Economic Affairs and Energy (BMWi) for financial support via the project “Energieeffiziente Stofftrennung in der chemischen und pharmazeutischen Industrie durch Membranverfahren - ESIMEM” (03ET1279E).

References (46)

L. Liu et al.
An overview of the proton conductivity of nafion membranes through a statistical analysis
J. Membr. Sci.
(2016)
L. Peeva et al.
2.05 - nanofiltration operations in nonaqueous systems
X. Yang et al.
Experimental observations of nanofiltration with organic solvents
J. Membr. Sci.
(2001)
H.J. Zwijnenberg et al.
Important factors influencing molecular weight cut-off determination of membranes in organic solvents
J. Membr. Sci.
(2012)
S. Postel et al.
Solvent dependent solute solubility governs retention in silicone based organic solvent nanofiltration
J. Membr. Sci.
(2016)
P. Schmidt et al.
Characterisation of organic solvent nanofiltration membranes in multi-component mixtures: membrane rejection maps and membrane selectivity maps for conceptual process design
J. Membr. Sci.
(2013)
S.R. Hosseinabadi et al.
Solvent-membrane-solute interactions in organic solvent nanofiltration (osn) for grignard functionalised ceramic membranes: explanation via spiegler-kedem theory
J. Membr. Sci.
(2016)
D. Stamatialis et al.
Observations on the permeation performance of solvent resistant nanofiltration membranes
J. Membr. Sci.
(2006)
S. Postel et al.
On negative retentions in organic solvent nanofiltration
J. Membr. Sci.
(2013)
C.J. Davey et al.
Molecular weight cut-off determination of organic solvent nanofiltration membranes using poly(propylene glycol)
J. Membr. Sci.
(2017)

J. da Silva Burgal et al.

Organic solvent resistant poly(ether-ether-ketone) nanofiltration membranes

J. Membr. Sci.

(2015)

Y.H. See Toh et al.

In search of a standard method for the characterisation of organic solvent nanofiltration membranes

J. Membr. Sci.

(2007)

M. Bastin et al.

Solvent resistant nanofiltration for acetonitrile based feeds: a membrane screening

J. Membr. Sci.

(2017)

P. Schmidt et al.

Characterisation of organic solvent nanofiltration membranes in multi-component mixtures: process design workflow for utilising targeted solvent modifications

Chem. Eng. Sci.

(2014)

B. Shi et al.

Will ultra-high permeance membranes lead to ultra-efficient processes? Challenges for molecular separations in liquid systems

J. Membr. Sci.

(2017)

B. Shi et al.

Performance of spiral-wound membrane modules in organic solvent nanofiltration - fluid dynamics and mass transfer characteristics

J. Membr. Sci.

(2015)

G. Wypych

{PMMA polymethylmethacrylate}

M. Skiborowski et al.

Model-based structural optimization of seawater desalination plants

Desalination

(2012)

M. Scholz et al.

Structural optimization of membrane-based biogas upgrading processes

J. Membr. Sci.

(2015)

B. Ohs et al.

Optimization of membrane based nitrogen removal from natural gas

J. Membr. Sci.

(2016)

Y. Thiermeyer et al.

Solvent dependent membrane-solute sensitivity of osn membranes

J. Membr. Sci.

(2018)

A. Linden et al.

Understanding gartner’s hype cycles, Strategic Analysis Report No. R-20-1971

(2003)

P. Marchetti et al.

Molecular separation with organic solvent nanofiltration: a critical review

Chem. Rev.

(2014)

Cited by (31)

Solvent and thermally stable polymeric membranes for liquid molecular separations: Recent advances, challenges, and perspectives
2023, Journal of Membrane Science
The transition to a sustainable economy requires a more effective and less energy-intensive industry. Membrane technology could augment or partially substitute classical molecular separation processes such as distillation to reduce energy, carbon, and water intensity. Organic solvent nanofiltration and reverse osmosis (OSN and OSRO) can positively impact the petrochemical, pharmaceutical, food, and fine chemical industries, among others, if broadly implemented. While hybrid and inorganic materials have the potential for game-changing performance, polymeric membranes provide key advantages in scalability and processability. Improved materials able to operate in challenging conditions, including combinations of organic solvents, high temperatures, extreme pHs, and oxidative environments are crucial. This is a comprehensive review of the state-of-the-art of polymeric membranes for use in OSN and OSRO, including a critical analysis of current academic approaches and potential polymer systems capable of enabling high-temperature liquid phase membrane separations. The challenges and prospects of OSN and OSRO membranes are discussed in the final section.
Synergizing machine learning, molecular simulation and experiment to develop polymer membranes for solvent recovery
2023, Journal of Membrane Science
Organic solvent nanofiltration (OSN) is a robust membrane technology for solvent recovery and molecular separation in harsh conditions. However, the current OSN membranes are largely produced through trial-and-error methods. In this study, machine learning (ML), molecular simulation (MS) and experiment are synergized for the development of OSN membranes. Using three different learning strategies, ML models are first constructed to identify critical gross properties (i.e., solvent viscosity, membrane thickness and water contact angle) and establish a phenomenological relationship for permeability prediction. Subsequently, ML models based on molecular representation via concatenated fragments are developed to predict methanol permeabilities in three polymer of intrinsic microporosity (PIM) membranes (PIM-A1, CX-PIM-A1 and PIM-8). The methanol permeability predicted in PIM-A1 is the highest among the three and also higher than that in archetypal PIM-1. Next, MS is conducted to provide microscopic insights into swelling behavior and methanol permeation in the three PIM membranes. Finally, the PIM-A1 membrane is experimentally fabricated and found to exhibit nearly complete solute rejection and methanol permeability of 2.33 × 10⁻⁶ L·m/m²·h·bar, which validates the ML prediction. This study demonstrates that the synergy of ML, MS and experiment can fundamentally elucidate and quantitatively predict solvent permeation in polymer membranes, and the holistic approach may advance the development of new membranes for solvent recovery and other important separation processes.
Solutes in solvent resistant and solvent tolerant nanofiltration: How molecular interactions impact membrane rejection
2023, Journal of Membrane Science
Solvent resistant nanofiltration (SRNF) and solvent tolerant nanofiltration (STNF) are energy-efficient membrane-based technologies to purify solvents or water-solvent mixtures, respectively, and/or to recover solutes from these streams. SRNF and STNF are rapidly gaining importance as they contribute to process intensification and to a circular economy. Whilst solute-solvent-membrane interactions determine the membrane performance, they are often not sufficiently taken into account, which possibly leads to misinterpretation of the obtained data. This review focusses specifically on the environment-dependent behavior and properties of commonly used solutes (dyes, oligomers, n-alkanes, esters, triglycerides, sugars and inorganic salts) through a theoretical physicochemical lens that focusses on molecular interactions. It reveals that the assumed fixed solute properties do not necessarily coincide with those in the chosen solvent medium. Indeed, solutes can adopt different configurations, charges or sizes in different (hydro-)organic media and this can significantly impact the membrane performance. By highlighting these complex solute-solvent interactions from a theoretical point of view, this review aims at more correctly interpreting, describing and representing membrane performance data, and to facilitate the selection of suitable solutes for a specific separation. Furthermore, the most adequate characterization techniques per solute class are discussed and some solutes that are less environment-dependent are proposed as more appropriate alternatives. In addition to achieving a better understanding of the fundamental solute-solute and solute-solvent interactions, the outlined insights aim at assisting membrane selection by end-users and at promoting improved data reliability.
From academia to industry: Success criteria for upscaling nanofiltration membranes for water and solvent applications
2023, Journal of Membrane Science
Nanofiltration (NF) is an established membrane technology to purify aqueous streams. In addition, there is a strong driving force to recycle and recover used solvents and dissolved compounds in industry. Solvent resistant nanofiltration (SRNF) and solvent tolerant nanofiltration (STNF), two young but highly promising membrane-based separation technologies, represent an economical and environment-friendly solution to this need. However, there exists a significant gap between academic research and industrial implementation of these technologies. This critical review addresses the challenges that need to be overcome to facilitate industrial valorization of academic breakthroughs by disclosing the so-called evaluator chart. This chart comprises industrial operating conditions, long-term performance, capital expenditures (CAPEX) and operational expenditures (OPEX) analysis, membrane module manufacturing, toxicity, and waste management. It is applied here to the state-of-the-art NF and SRNF membranes developed in academia to assess their industrial potential. To further direct future research towards industrial scale, invention trends in SRNF in both academia and industry from 2015 to 2020 are highlighted.
Sustainable organic solvent nanofiltration membranes
2023, Green Membrane Technologies towards Environmental Sustainability
Various green alternatives and membrane manufacturing strategies have been introduced recently to make organic solvent nanofiltration technology more sustainable. These include the use of less toxic solvents to dissolve the polymer and polymerization process, the use of degradable and renewable polymer sources, and the reduction of solvent consumption. Since the reliability and longevity of the membrane contribute to the “greenness” of the technique, the longer a membrane can function, the greener it is in comparison to other separation techniques. All aspects of this chapter, including the monomer source, membrane fabrication, and solvent recyclability, were in accordance with sustainability and circular economy principles.
Explainable machine learning for unraveling solvent effects in polyimide organic solvent nanofiltration membranes
2023, Advanced Membranes
Understanding the effects of solvents on organic solvent nanofiltration currently depends on results obtained from small datasets, which slows down the industrial implementation of this technology. We present an in-depth study to identify and unify the effects of solvent parameters on solute rejection. For this purpose, we measured the rejection of 407 solutes in 11 common and green solvents using a polyimide membrane in a medium-throughput cross-flow nanofiltration system. Based on the large dataset, we experimentally verify that permeance and electronic effects of the solvent structure (Hildebrand parameters, electrotopological descriptors, and LogP) have strong impact on the average solute rejection. We furthermore identify the most important solvent parameters affecting solute rejection. Our dataset was used to build and test a graph neural network to predict the rejection of solutes. The results were rigorously tested against both internal and literature data, and demonstrated good generalization and robustness. Our model showed 0.124 (86.4% R²) and 0.123 (71.4 R²) root mean squared error for the internal and literature test sets, respectively. Explainable artificial intelligence helps understand and visualize the underlying effects of atoms and functional groups altering the rejection.

View all citing articles on Scopus

View full text