ACS Publications. Most Trusted. Most Cited. Most Read
My Activity
CONTENT TYPES
RETURN TO ISSUEPREVResearch ArticleNEXT

Inverse Design of Pore Wall Chemistry To Control Solute Transport and Selectivity

Cite this: ACS Cent. Sci. 2022, 8, 12, 1609–1617
Publication Date (Web):November 30, 2022
https://doi.org/10.1021/acscentsci.2c01011

Copyright © 2022 The Authors. Published by American Chemical Society. This publication is licensed under

CC-BY 4.0.
  • Open Access

Article Views

2402

Altmetric

-

Citations

LEARN ABOUT THESE METRICS
PDF (6 MB)
Supporting Info (1)»

Abstract

Next-generation membranes for purification and reuse of highly contaminated water require materials with precisely tuned functionality to address key challenges, including the removal of small, charge-neutral solutes. Bioinspired multifunctional membrane surfaces enhance transport properties, but the combinatorically large chemical space is difficult to navigate through trial and error. Here, we demonstrate a computational inverse design approach to efficiently identify promising materials and elucidate design rules. We develop a combined evolutionary optimization, machine learning, and molecular simulation workflow to spatially design chemical functional group patterning in a model nanopore that enhances transport of water relative to solutes. The genetic optimization discovers nonintuitive functionalization strategies that hinder the transport of solutes through the pore, simply by patterning hydrophobic methyl and hydrophilic hydroxyl functional groups. Examining these patterns, we demonstrate that they exploit an unexpected diffusive solute hopping mechanism. This inverse design procedure and the identification of novel molecular mechanisms for pore chemical heterogeneity to impact solute selectivity demonstrate new routes to the design of membrane materials with novel functionalities. More broadly, this work illustrates how chemical design is a powerful strategy to modulate water-mediated surface–solute interactions in complex, soft material systems that are relevant to diverse technologies.

This publication is licensed under

CC-BY 4.0.
  • cc licence
  • by licence

Synopsis

Water scarcity is a grand challenge. Here, a computational inverse design approach discovers new, chemically patterned materials strategies that optimize the separation of contaminants from water.

Introduction

ARTICLE SECTIONS
Jump To

The in silico inverse design of materials has long been a goal of computational modeling, promising accelerated materials discovery combined with reduced experimental cost. While system size and complexity have challenged earlier efforts, modern hardware and algorithms have given rise to many recent successes in the optimization of chemical functionalities. These efforts span a wide range of materials, from the design of small molecules, (1,2) molecular complexes, (3) and functionalized interfaces (4,5) to self-assembling colloidal (6) and polymeric systems. (7) In some cases, existing structure–property relationships or learned empirical models simplify navigation of design space. Often, however, such models are inadequate or nonexistent, and molecular simulations are required to predict material properties, greatly increasing the computational expense of these design efforts. Soft material systems involving solution phases pose particularly difficult design problems that are due to often significant chemical heterogeneity, a concert of different interactions (excluded volume, electrostatic, van der Waals), and the emergence of complex driving forces involving entropies, such as hydrophobic interactions, that necessitate atomistic-level fidelity and long simulation time scales. Such systems often involve very large design spaces that are not easily enumerable or navigable as they lack obvious order parameters along which optimization can proceed, presenting additional challenges. There is therefore a pressing need for soft materials optimization approaches that make efficient use of generated simulation data, such as by incorporating on-the-fly learned models. (8)
A particularly important, complex class of materials that encompasses these challenges and is a major opportunity for computational soft materials design is water purification membranes. In this work, we demonstrate an inverse design approach to design the chemical functionality of a nanopore wall to optimize solute rejection in water separation membranes. We establish an optimization workflow coupling molecular simulations and machine-learned surrogate models and demonstrate its capabilities for complex soft material problems. We furthermore demonstrate how algorithm-generated materials can educe novel and nonintuitive design rules to guide synthetic efforts and advance the fundamental understanding of the relationship between surface chemistry and the transport properties of water and solutes.
Water scarcity will become an increasingly pressing issue in the coming decades. (9) Energy security is intimately linked to the sustainability of current water resources, (10) and thus, efficient water treatment processes are crucial for employing alternative lower quality water sources. The maturity of membrane technologies for low-energy, reliable treatment of seawater through reverse osmosis has led to their widespread adoption for desalination (11,12) and utility for water reuse and other advanced water treatment systems. (13) However, membranes are not yet capable of economically treating highly impaired waters, such as oilfield-produced water, currently an enormous wastewater stream on the order of 2 billion gallons per day. (14) The development of membrane separation processes providing energy-efficient routes to converting such highly contaminated sources into “fit-for-purpose” water (15) (e.g., for irrigation, streamflow augmentation, power) as well as recovering valuable resources from these waste streams remains a grand goal (16−19) but is hindered by key challenges associated with the concentration and broad chemical variety of contaminants. For instance, the separation of small, charge-neutral solutes is particularly difficult for current membranes that leverage charge and size separation mechanisms. (16) One exemplary solute is boric acid. While the World Health Organization guidelines recommend drinking water concentrations of less than 2.4 mg/L of boron, (20) water reuse for agricultural irrigation of most crops requires even lower concentrations (<1 mg/L). (21,22)
Biology provides inspirational design considerations. Transmembrane proteins, such as aquaporin (23) and potassium ion channels, (24) achieve impressive permeability and selectivity through nanoscale confinement and precise arrangement of protein side-chain functional groups within the pores, shaping the chemical and topological landscape through which water and solutes transport. A number of efforts have made progress along this direction in synthetic systems. (25) Nanochannels such as carbon nanotubes can exhibit unique (26,27) and ultrafast transport of water, (28,29) where the strength of the wall–water interaction modulates water transport. (30) Functionalizing the pore also affects water adsorption (31) and flux (32) and modulates solute rejection, (33−37) depending on the functionalization chemistry. Previous studies have also demonstrated that the location of the functional groups affects the water behavior (31,32) and solute transport. (38) However, a critical knowledge gap is understanding the potential impact of patterning pore surfaces with more than one type of functional group on engineering solute rejection and selectivity for the large class of neutral small-molecule solutes. The presence of more than one chemistry at a surface exponentially expands the design space. Indeed, previous computational work has shown that both the number and the arrangement of chemical functional groups at an extended surface can significantly tune the surface affinity of model solutes (5) as well as the local diffusivity of water. (4) Design rules and strategies to identify optimal, heterogeneous chemistries are largely unknown and would have implications beyond water filtration membranes to other chemically heterogeneous surfaces, such as porous adsorbents, catalytic surfaces, and antifouling interfaces, as well as heterogeneous macromolecules such as proteins and sequence-defined polymers.
In this work, as a case study, we focus on rejection of a neutral solute, boric acid, and demonstrate a novel workflow to precisely tune membrane pore wall spatial chemistry to enhance solute rejection. While Edisonian approaches in the membrane community have identified important heuristic membrane material characteristics affecting, e.g., fouling (39,40) and solute uptake, (41−44) such approaches only skim the large design space associated with multifunctional chemistries. Instead, here, we describe a computational inverse design workflow that optimizes the relative transport of water over boric acid, merely by patterning nonpolar methyl and polar hydroxyl groups tethered to a model pore. This proof of concept demonstrates the potential of computational materials discovery in uncovering the existence of nonintuitive, heterogeneous pore wall chemistries that enhance solute rejection. In particular, the optimization workflow suggests that pores with rings of hydroxyl groups pose selective barriers to boric acid transport over water, suggesting a new class of materials to explore for water separations. On the basis of these computationally designed pore surfaces, we identify a mechanism behind the enhancement in boric acid rejection captured by a simple, predictive analytical model. We then demonstrate that the optimal pore design gives enhanced selectivity for a range of other solutes that vary with respect to polarity, hydrogen-bonding capability, and surface affinity, suggesting the generality of the proposed mechanism.

Results and Discussion

ARTICLE SECTIONS
Jump To

Genetic Algorithm Optimization

Our basic question is: to what extent can surface chemical patterning be leveraged to optimize the relative transport of water versus neutral solutes in a pore? Here, we use boric acid as a model solute, which is difficult to separate due to its small size and charge neutrality at neutral pH, and we consider nanopores of varied hydrophobic/hydrophilic content, tuned by patterning the interior of a carbon nanotube (CNT) with hydroxyl and methyl functional groups (Figure 1a). To compute transport properties of water and solutes inside the pore, we first simulate a “finite” pore attached to a water reservoir to equilibrate the density of water inside the pore and then simulate an “infinite” pore with the same density (Figure 1b) from which we compute water and boric acid transport metrics, such as equilibrium diffusivity and nonequilibrium, pressure-driven flux (details in Methods). To assess the potential of chemical patterning to enhance pore transport of water relative to boric acid, we use a bioinspired evolutionary or genetic algorithm (Figure 1c) that identifies arrangements of methyl and hydroxyl functional groups on the pore wall that optimize the difference between the computed water and boric acid fluxes. To accelerate the optimization, we employ a machine-learned surrogate objective function trained on prior explicitly simulated patterns to predict the flux difference. The surrogate function is periodically retrained with additional batches of explicit molecular dynamics (MD) calculations. The surrogate model allows exploration of orders of magnitude more generations than is possible with MD alone due to computational expense. Because we use a linear model in the feature space, the surrogate model also serves to average out statistical noise from the MD-computed flux, giving the genetic algorithm robustness to uncertainty (Figure S14).

Figure 1

Figure 1. (a) Model nanopore system allows for precise patterning of nonpolar methyl groups and polar hydroxyl groups along the CNT backbone. For ease of visualization, we present the flattened pore patterns with blue triangles representing methyl groups and red circles representing hydroxyl groups. (b) We first simulate the finite pore to compute the density of water, which we then set in the infinite pore to compute water and boric acid transport properties. (c) A genetic algorithm incorporating MD simulations and a learned surrogate function allows for automated identification of pore wall chemical patterns optimizing pore transport properties.

This optimization approach detects novel functionalization strategies: Figure 2 shows the trajectory of the flux difference versus the number of generations for genetic algorithm runs fixing the hydroxyl group fraction at 0.25 (32 hydroxyl groups) (Figure 2a) and allowing the hydroxyl fraction to vary (Figure 2b). Both trajectories span a large range of flux differences and thus demonstrate that chemical patterning can be tuned to enhance solute rejection. The optimal pattern found at a fixed hydroxyl fraction contains two nearly perfect rings of hydroxyl groups (shown in the 2D schematic as rows), while the variable hydroxyl case with a much larger search space identifies an optimal pattern that contains a single ring of hydroxyl groups. The optima suggest that rings of hydroxyl groups inside the nanopore particularly inhibit boric acid transport with respect to water transport, a nonobvious result. Patterns that include these features are not identified in the first set of simulated generations, showing that the surrogate function is key to exploring a sufficiently large number of generations to identify these optimal motifs.

Figure 2

Figure 2. Genetic algorithm trajectories maximizing the difference between water flux and boric acid flux for (a) fixed hydroxyl group fraction and (b) varying hydroxyl group fraction show optimization to surfaces with rows of hydroxyl groups. Each marker corresponds to a pattern discovered by the genetic algorithm. Simulated generations are shown as green stars and surrogate model-driven generations as black circles. Inset patterns show the optimal patterns found after (1) the first set of simulated generations, (2) the first set of model-driven generations, (3) the second set of simulated generations, and (4) the second set of model-driven generations. The red line shows the maximum value of the objective function at each point in the optimization.

The surrogate model is key to enabling high-throughput optimization. We find that it only needs to maintain accuracy in predicting the flux difference of patterns similar to those currently explored by the genetic algorithm (“local” in pattern space) to enhance performance, with periodic retraining through new MD simulations ensuring its ongoing relevance. Figure S15 shows that the linear model is indeed able to fit the “training” data (a single set of explicit MD simulations) well, suggesting good local accuracy. The model also predicts the flux difference of patterns outside of the training set qualitatively well, giving accurate directional trends, which is the key requirement to advance the genetic algorithm. The quantitative accuracy is also reasonable given the statistical uncertainty associated with the MD simulations (Figure S15). Beyond the linear model, we find that neural-network-based surrogates do not perform significantly better (Figure S16) and furthermore require more extensive hyperparameter selection while sacrificing interpretability. We instead take advantage of the interpretability of simpler models (specifically, LASSO and random forest regression) to compute feature importance metrics demonstrating that spatial correlations of hydroxyl groups along the pore axis and circumference, and especially rows of hydroxyl groups, are the most important features for the regression (Table S4), reinforcing the idea that these motifs are significant.

Behavior of Rationally Designed Patterns Suggests Mechanism for Boric Acid Rejection

Inspired by the genetic algorithm results, a rational exploration of the effect of pore patterning on water and solute transport reveals a rich design landscape. Figure 3 visualizes this design space, by showing the water flux (Figure 3a) and boric acid flux (Figure 3b) as a function of hydroxyl fraction in regularly patterned pores (Table S5), chosen as idealized versions of the motifs identified from the genetic algorithm-discovered patterns (Figure S17). Figure 3a clearly illustrates that the pore wall chemistry significantly modulates water transport: increasing the number of hydroxyl groups generally reduces water flux, with the fully methylated and fully hydroxylated pores capturing the extreme cases with an approximately 4-fold difference in flux. However, even at a fixed fraction of hydroxyl groups, the specific pattern of the groups also significantly affects transport: for instance, patterns with a hydroxyl group fraction of 0.25 span a 2-fold change in flux difference. In contrast, previous work that “patterned” a CNT with chemically unspecific LJ sites of varying wall–water interaction strengths revealed that functional group arrangement did not affect water flux. (45) The present study thus suggests that functionalization with chemically specific moieties involving, e.g., hydrogen-bonding interactions is required to observe pattern-dependent trends, consistent with previous work showing the dependence of water diffusivity on surface patterning. (4)

Figure 3

Figure 3. Pattern of methyl and hydroxyl groups has a significant effect on the flux of (a) water and (b) boric acid in nonequilibrium simulations and on (c) the diffusivity of boric acid in equilibrium simulations. For each panel, orange triangles represent patterns with a “single patch” of hydroxyl groups, spanning the pore axis, while blue circles represent “fully dispersed” patterns (examples of both are in the inset of a). Green diamonds represent other patterns where the hydroxyl groups form smaller patches (“patchy”). Two examples of such patterns are shown in the inset of b. These patterns also show lower boric acid flux compared to the fully dispersed patterns. The inset also includes a schematic of the periodically varying free energy landscape used in the analytical model to predict the reduction in boric acid diffusivity in the ringed pores. In c, green diamonds are only shown for patterns with rows of hydroxyl groups. The dashed line shows the analytical prediction described in the text based on a model of the periodic free energy landscape depicted in the inset of b.

In particular, the patterns with the hydroxyl groups in a single patch spanning the pore axis (orange triangles) lead to maximal water flux out of all patterns examined, while those that disperse the hydroxyl groups (blue circles) produce minimal flux; intermediate behaviors are seen from patterns with hydroxyl groups arranged in multiple, smaller patches (green diamonds). The diffusivity of water in these pores from equilibrium simulations recapitulates the same nonequilibrium flux trends (Figure S18). These behaviors are consistent with earlier observations at extended self-assembled monolayer surfaces in which both increasing the number of hydroxyl groups and dispersing them more evenly across the surface lead to a reduction in surface water diffusivity. (4) More broadly, the pattern-dependent trends follow similar findings for water near surfaces where increased dispersion of polar sites leads to reduced surface hydrophobicity, which manifests as reduced water density fluctuations, (46,47) reduced tendency to dewet, (48−50) and reduced solute affinity. (5)
The design space for boric acid in Figure 3b shows that the pore wall chemistry generally affects its transport similarly to water; however, the patterns that minimize boric acid transport for a given hydroxyl fraction are not fully dispersed but instead contain rows of hydroxyl groups (Figure 3b, inset). These motifs use rings of hydroxyl groups to reduce boric acid flux relative to water (Figure S19), rationalizing the optimal surfaces located by the genetic algorithm. For the ringed pores (Figure 3b, inset), boric acid flux is less than half that of water, a reduction that is sufficient to reduce boric acid concentration in seawater (4.5 mg/L) (21) to the target concentration for drinking water (2.4 mg/L). (20) We note, though, that treating groundwater, which can contain higher concentrations of boric acid, or achieving a lower target concentration for irrigation water would require recycling.
To understand the origin of this unique behavior, we hypothesize that two mechanisms are important: (1) boric acid strongly partitions to the hydrophobic, methylated regions of the pore wall, and (2) it is then forced to “hop” over the hydroxyl rings with an energy barrier determined by rearrangements to the hydrogen-bonding network. Boric acid, while highly polar, shows affinity for methylated wall sections of the pore due to its small size; earlier simulation results demonstrated that water drives boric acid to methylated surfaces as a way to minimize the penalty associated with water restructuring around the solute in bulk solution distant from the interface. (5) To assess this hypothesis, we compared measurements of the diffusivity from equilibrium simulations to predictions of an analytical model from ref (51) that describes the 1D diffusive transport of a particle (here, boric acid) in a periodic energy landscape (Figure 3b, inset). We estimate the height of the free energy barrier from moving between methylated sections (across hydroxylated sections) to be 2.4kBT, based on the partition coefficients of boric acid into fully methylated versus fully hydroxylated pores. The periodicity of the landscape is then set by the geometry of the pattern, i.e., the number of methylated and hydroxylated rows (details in section S5D, Supporting Information). We combine this prediction with a simple exponential decay that describes the background reduction in diffusivity due to the surface hydroxyl group fraction (section S5D, Supporting Information). The resulting model then predicts the boric acid diffusivity (Figure 3c, dashed line) in agreement with equilibrium simulations for pores with various arrangements of rows of hydroxyl groups. This mechanism is consistent with earlier work on the reduction in water dynamics in rough versus smooth pores, concluding that the primary mechanism is the creation of a rough energy landscape. (52)
At higher fractions of hydroxyl groups (0.50 and 0.75), not all patterns with rows of hydroxyl groups agree with the analytical prediction. We hypothesize that at least two consecutive rows of methyl groups are required to produce an effective hydrophobic surface environment, as previous work has shown that polar groups affect the local water behavior near adjacent methylated regions (48) such that isolated methyl groups have little effect on the local water behavior. (47) Consistent with this hypothesis, the patterns showing the largest deviations from the analytical model each contain isolated single rows of methyl groups. At lower hydroxyl group fractions, the agreement between the analytical model and the MD diffusivities is consistent with the proposed hopping mechanism, reinforcing this potential route to inhibiting solute transport while maintaining water transport. These observations underscore the relevance of nonadditive, context-dependent interactions to the design of hydrated systems. Here, the arrangement of hydroxyl groups affects solute transport, independent of the average density of the hydroxyl groups and even beyond basic characterization of the pattern (i.e., degree of clustering), thus demonstrating behaviors that are highly nonlinear on a nanoscopic scale.
Transport through the pore is not the sole factor affecting membrane performance. Water and solutes must first partition into it: both the number and the arrangement of hydroxyl groups affects water and boric acid partition coefficients (Figure S21). While patterns with more hydroxyl groups or dispersed hydroxyl groups lead to enhanced water partitioning and reduced boric acid partitioning and flux, they also produce reduced water flux, suggesting a trade-off between permeability and selectivity, a commonly observed relationship for other membrane materials that captures how improvements to solute rejection tend to be at the expense of reducing throughput. (41) This trade-off is a key challenge for membrane design efforts. To assess this behavior, we compute the permeability from water’s partition coefficient (tendency to go into the pore) multiplied by its in-pore flux (transport once inside the pore). We compute the selectivity from the ratio of the permeability of water to that of boric acid. The calculated permeability and selectivity values generally reinforce the trade-off: for instance, water permeability increases from the fully hydroxylated pore (0.7034 mol-nm/cm2-s-MPa) to the pore with an equal number of hydroxyl and methyl groups arranged in single, axial patches (Figure 1a) (1.3004 mol-nm/cm2-s-MPa) and again to the fully methylated pore (2.8668 mol-nm/cm2-s-MPa). However, selectivity for water over boric acid decreases along that same series (571 × 104, 9.32 × 104, and 5.83 × 104). Beyond these specific patterns, randomly generated pores also demonstrate this trade-off (Figure S22). However, the optimal genetic optimization patterns display improved permeability–selectivity behavior compared to randomly patterned pores (Figure S22), suggesting that optimization of the flux difference may produce pattern motifs that address the permeability–selectivity trade-off.
To further demonstrate the generality of these pattern motifs, we choose a diverse set of other neutral solutes (phenol, benzene, isopropanol, ammonia, arsenous acid) to simulate with several exemplar patterns. These solutes possess a wide range of affinities to methylated and hydroxylated self-assembled monolayers (measured in ref (5)) as well as a diverse set of surface hydrogen-bonding behaviors, as quantified by the average number of hydrogen bonds formed with the fully hydroxylated pore (Table S6). Out of these solutes, phenol, benzene, and isopropanol similarly demonstrate enhanced water flux relative to boric acid flux in pores with rings of hydroxyl groups (Figure 4), suggesting that the proposed mechanism may be broadly generalized to enhance the separation of other solutes from water. Consistent with the proposed diffusive hopping mechanism, there is a correlation between the enhancement in solute rejection in ringed pores and the difference in affinity for methylated versus hydroxylated SAMs (Table S6), as larger differences in affinity raise the free energy barrier to move between ring-separated methylated sections. For instance, ammonia, which has the lowest difference in affinities for the methylated and hydroxylated surfaces, has the lowest flux relative to water in the fully methylated pore. Measures of affinity to homogeneous functionalized surfaces are thus predictive of this effect of patterned functionalizations on solute rejection.

Figure 4

Figure 4. Other solutes exhibit similar trends as boric acid in a selection of rationally patterned pores. Benzene, phenol, and isopropanol show maximal water flux relative to boric acid flux in pores with rows of hydroxyl groups, while for ammonia and arsenous acid, the optimal pattern is the fully methylated one. For the simulations with solutes besides boric acid, we use the same workflow as described for boric acid except that the production simulations in the infinite pore are run for 1 μs. The reported uncertainties are the standard error of the mean from three independent simulations.

Conclusions

ARTICLE SECTIONS
Jump To

The present simulations demonstrate that pore wall patterning of nonpolar and polar functional groups significantly affects the partitioning and transport of water and solutes and offers new strategies to design pore chemistries. We show that inverse design algorithms automatically identify unique, nonintuitive chemical patterning strategies and associated mechanisms that decouple and thus optimize the flux of water relative to solutes of interest. Here, for example, we find that pores with higher rates of boric acid rejection contain rings of hydroxyl groups. The underlying mechanism identified by the evolutionary optimization is unexpected: boric acid transport in optimal pores is well described by a periodic barrier-hopping process, whereby it adsorbs at the methylated regions of the pore wall and then must hop over hydrogen-bonding hydroxyl rings. These results suggest that leveraging functional groups to create a fluctuating free energy landscape for a solute in a pore may be an important mechanism for enhancing solute rejection at little cost to water permeability. More generally, this approach illustrates the use of computational inverse design algorithms to design membranes and other interfacial systems to identify unique, nonobvious solutions that can then suggest theoretical models for guiding design strategies.
Our specific results imply that as synthetic advances make multifunctional membrane surfaces more accessible, inverse design procedures will be key to exploring the larger design space. While direct functionalization of membrane surfaces with precisely patterned methyl and hydroxyl groups may not be currently feasible, lithographic techniques with block copolymer assembly as well as covalent or noncovalent functionalization of nanochannels with biomolecules or polymers (36,53,54) may be promising routes to introducing multiple chemistries at a pore surface. The actual design space therefore potentially includes a large set of possible functionalities, much larger than the two functionalities examined here. Beyond multifunctional pore walls, other possible extensions of this pore design problem include (1) varying the pore radius, as the difference in solute and water affinity for the pore wall may offer a means to enhance water flux in the bulk region of the pore while still relying on the pore wall to enhance solute rejection, and (2) varying the topology or “roughness” of the pore surface, which has been shown to modulate local water behavior. These expanded design spaces make the development of automated design algorithms even more important. Beyond membranes, workflows for the precise and efficient in silico design of chemical patterns would be useful for identifying optimal antifouling surfaces and directing biomolecular folding and assembly, among other applications.

Methods

ARTICLE SECTIONS
Jump To

Simulation Models

To model an ideal nanopore with tunable pore wall chemistry, we simulate a (16,0) zigzag single-wall carbon nanotube (CNT) with nonpolar methyl and polar hydroxyl functional groups directly tethered to the interior of the pore. At smaller radii, the diffusivity of water inside the pore decreases rapidly due to strong confinement effects (Figure S1), consistent with experimental work on water in porous silica showing an enhancement in water network structures with decreasing pore size. (55,56) We thus choose a slightly larger radius of 1.25 nm, for which water reaches its bulk density in the middle of the pore (Figure S2), in order to focus on the effects of surface chemistry as opposed to confinement, since confinement can significantly alter water and solute behavior in nanoporous environments. (57−59) The functional groups are placed every 4 carbons (Figure S3) to achieve a surface density of 0.21 nm2/group, comparable to the surface density for realizable functionalized surfaces, such as self-assembled monolayers. (60) We obtain nonbonded force field parameters for the CNT and functional groups from ref (61). Section S1C, Supporting Information, describes the bonded parameters in more detail. Parameters for water are from the TIP3P water model, (62) and those for boric acid are based on the classical model parametrized in ref (63) (section S1CII, Supporting Information). Section S5E, Supporting Information, describes the parameters for the other solutes examined.
To compute the transport properties of water and boric acid through the pore, we simulate an “infinitely” long nanopore (Figure 1b) with 16 rows of carbon atoms, yielding a pore of length 3.4 nm with 128 functional groups, allowing a diverse set of patterns. To compute water and solute densities inside the pore, we first simulate a “finite” pore in between two water reservoirs bounded by graphene sheets (Figure 1b, details in section S1, Supporting Information). We model all systems using the OpenMM simulation engine. (64) Electrostatic interactions are computed using particle-mesh Ewald. Lennard–Jones (LJ) nonbonded interactions are cut off at 10 Å. Hydrogen bonds are constrained with SHAKE, (65) and water is kept rigid with SETTLE. (66) We perform an energy minimization before all simulations. We use Langevin dynamics with a time step of 2 fs, a temperature of 298 K, and a friction coefficient of 1 ps–1. For NPT simulations, we use an anisotropic Monte Carlo (MC) barostat that only adjusts the box size in the axial direction, with 200 fs between MC moves to set the pressure to 1 bar. All carbon atoms in the nanotube and graphene sheets are fixed during the simulation. More details, including the lengths of the equilibration and production stages for each simulation, are given in section S1D, Supporting Information.

Finite Nanopore Simulation and Partition Coefficient Calculation

For all simulations with the finite pore, we restrain the boric acid molecule so that it stays within the nanopore (section S1E, Supporting Information). To determine the partition coefficient of boric acid, we first compute its free energy of solvation inside the pore via an expanded ensemble simulation where the boric acid LJ ε parameters and partial charges are slowly turned on in several distinct states. The procedure is similar to that in ref (5) except that an additional restraint keeps boric acid inside the pore (section S1E, Supporting Information). The free energy of solvating boric acid, ΔG, is the difference between the states in which the boric acid molecule is fully on and off, which we compute using MBAR. (67,68) We then compute the partition coefficient, K, from K = K0 exp[−(ΔG – ΔG0)], where ΔG0 = −3.4kBT and K0 = 8.0 are reference values computed from an unbiased simulation of the fully methylated pore (section S1F, Supporting Information).

Infinite Nanopore Simulation and Transport Metric Calculations

To probe the diffusivity of water and boric acid, we compute the axial mean squared displacement for the water oxygens and boron atom and extract the diffusivities from the slope of the MSDs. To compute the nonequilibrium flux of water and boric acid under an applied pressure gradient, we add a constant acceleration to each water and boric acid atom such that the pressure gradient is 12 MPa/nm. In this regime, water flux remains linear with pressure gradient (Figure S9), allowing extrapolation to lower, more realistic gradients. We run a simulation with the applied pressure gradient and compute the total net displacements of the water oxygens and boron atoms in the axial direction. The in-pore flux is the displacement divided by the length of the simulation, the volume of the pore, and the number of molecules. This “flux” calculation measures the rate at which individual water or solute molecules move through the pore and does not have a direct contribution from their densities in the pore, which is convenient for separating the effects of pore wall chemistry on partitioning into the pore and transporting through the pore. To obtain metrics that combine both effects, we compute the permeability by multiplying the partition coefficient by the computed in-pore flux, pore volume, and bulk water density and dividing by the pressure gradient. We compute the selectivity from the ratio of the product of the partition coefficient and in-pore flux for water and boric acid and then divide by the approximate mole fraction of boric acid in seawater, (21) approximately 1.4 × 10–6.

Genetic Algorithm Structure

The genetic algorithm, depicted in Figure 1c and in more detail in Figure S10, follows the general procedure established in ref (21) but is used here to find the arrangement of 128 hydroxyl and methyl groups that optimizes the difference between water and boric acid flux. We perform two genetic algorithm runs, one in which the number of surface hydroxyl groups is fixed at 32 (“fixed fraction”) and one in which it is free to vary (“varying fraction”). Both runs begin with 5 generations of randomly generated patterns. For the fixed fraction run, we generate initial patterns by randomly choosing 32 sites for the hydroxyl groups. For the varying fraction run, the number of hydroxyls is randomly chosen at a given hydroxyl fraction spanning 0–1. We then quench these initial randomly generated patterns to minimize the pattern entropy, creating a more feature-diverse set of motifs (Figure S13) to initialize the algorithm, improving its convergence. Here, the pattern entropy is Spat=i=1512PilogPi, where Pi gives the probability (fraction) of 3 × 3 subsections with a given pattern among all 512 distinct 3 × 3 motifs (see section S2A, Supporting Information, for details).
In the genetic optimization runs, the initial “random” generations are followed by a subsequent set of maximization generations where the flux difference is evaluated from molecular dynamics (MD) simulations, as described above. Because the simulation–evaluated flux differences are computationally expensive, we fit a surrogate model to the initial simulated generations to predict the flux difference based on a linear combination of 22 pattern features, chosen to elucidate various spatial correlations in the arrangement of the functional groups (section S3A, Supporting Information). We then rapidly iterate over many generations using the surrogate model, followed by additional generations with explicit MD evaluation of the best candidates, and finally additional surrogate-model-evaluated generations with a retrained surrogate model. Table S3 gives the number of generations in each stage. Because the varying hydroxyl group fraction case explores a much larger pattern space, we increase the number of generations compared to the fixed fraction run.

Supporting Information

ARTICLE SECTIONS
Jump To

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acscentsci.2c01011.

  • Details regarding the simulation models and property calculations; genetic algorithm workflow; surrogate function; spectral clustering of the genetic algorithm-identified patterns; types of rationally designed patterns explored; other quantities computed for the rationally designed patterns including the partition coefficients of water and boric acid; analytical model describing boric acid diffusivity; details regarding the simulation of other solutes in the nanopore; permeability–selectivity behavior of random and genetic-algorithm-discovered patterns (PDF)

Terms & Conditions

Most electronic Supporting Information files are available without a subscription to ACS Web Editions. Such files may be downloaded by article for research use (if there is a public use license linked to the relevant article, that license may permit other uses). Permission may be obtained from ACS for other uses through requests via the RightsLink permission system: http://pubs.acs.org/page/copyright/permissions.html.

Author Information

ARTICLE SECTIONS
Jump To

  • Corresponding Author
  • Authors
    • Sally Jiao - Department of Chemical Engineering, University of California, Santa Barbara, California93106, United StatesOrcidhttps://orcid.org/0000-0003-1729-0414
    • Lynn E. Katz - Department of Civil, Architectural and Environmental Engineering, University of Texas at Austin, Austin, Texas78712, United States
  • Notes
    The authors declare no competing financial interest.

Acknowledgments

ARTICLE SECTIONS
Jump To

This work was supported by the Center for Materials for Water and Energy Systems (M-WET), an Energy Frontier Research Center funded by the U.S. Department of Energy, Office of Science, Basic Energy Sciences under Award DE-SC0019272. Use was made of computational facilities purchased with funds from the National Science Foundation (OAC-1925717) and administered by the Center for Scientific Computing (CSC). The CSC is supported by the California NanoSystems Institute and the Materials Research Science and Engineering Center (MRSEC; NSF DMR 1720256) at UC Santa Barbara. S.J. acknowledges support by the National Science Foundation Graduate Research Fellowship (DGE 1650114). We thank M. P. Howard for helpful discussions concerning nanopore transport. We also thank M. R. Landsman for insight regarding boric acid solutes. Finally, we thank J. B. Rawlings for useful suggestions related to developing the optimization workflow.

References

ARTICLE SECTIONS
Jump To

This article references 68 other publications.

  1. 1
    Austin, N. D.; Sahinidis, N. V.; Trahan, D. W. Computer-Aided Molecular Design: An Introduction and Review of Tools, Applications, and Solution Techniques.. Chem. Eng. Res. Des. 2016, 116, 226,  DOI: 10.1016/j.cherd.2016.10.014
  2. 2
    Mohr, B.; Shmilovich, K.; Kleinwächter, I. S.; Schneider, D.; Ferguson, A. L.; Bereau, T. Data-Driven Discovery of Cardiolipin-Selective Small Molecules by Computational Active Learning.. Chem. Sci. 2022, 13 (16), 44984511,  DOI: 10.1039/D2SC00116K
  3. 3
    Janet, J. P.; Ramesh, S.; Duan, C.; Kulik, H. J. Accurate Multiobjective Design in a Space of Millions of Transition Metal Complexes with Neural-Network-Driven Efficient Global Optimization.. ACS Cent. Sci. 2020, 6 (4), 513524,  DOI: 10.1021/acscentsci.0c00026
  4. 4
    Monroe, J. I.; Shell, M. S. Computational Discovery of Chemically Patterned Surfaces That Effect Unique Hydration Water Dynamics.. Proc. Natl. Acad. Sci. U. S. A. 2018, 115 (32), 80938098,  DOI: 10.1073/pnas.1807208115
  5. 5
    Monroe, J. I.; Jiao, S.; Davis, R. J.; Robinson Brown, D.; Katz, L. E.; Shell, M. S. Affinity of Small-Molecule Solutes to Hydrophobic, Hydrophilic, and Chemically Patterned Interfaces in Aqueous Solution.. Proc. Natl. Acad. Sci. U. S. A. 2021, 118 (1), e2020205118,  DOI: 10.1073/pnas.2020205118
  6. 6
    Ma, Y.; Aulicino, J. C.; Ferguson, A. L. Inverse Design of Self-Assembling Diamond Photonic Lattices from Anisotropic Colloidal Clusters.. J. Phys. Chem. B 2021, 125 (9), 23982410,  DOI: 10.1021/acs.jpcb.0c08723
  7. 7
    Meenakshisundaram, V.; Hung, J.-H.; Patra, T. K.; Simmons, D. S. Designing Sequence-Specific Copolymer Compatibilizers Using a Molecular-Dynamics-Simulation-Based Genetic Algorithm.. Macromolecules 2017, 50 (3), 11551166,  DOI: 10.1021/acs.macromol.6b01747
  8. 8
    Sherman, Z. M.; Howard, M. P.; Lindquist, B. A.; Jadrich, R. B.; Truskett, T. M. Inverse Methods for Design of Soft Materials.. J. Chem. Phys. 2020, 152 (14), 140902,  DOI: 10.1063/1.5145177
  9. 9
    Rijsberman, F. R. Water Scarcity: Fact or Fiction?. Agricultural Water Management 2006, 80 (1–3), 522,  DOI: 10.1016/j.agwat.2005.07.001
  10. 10
    Hamiche, A. M.; Stambouli, A. B.; Flazi, S. A Review of the Water-Energy Nexus.. Renew. Sust. Energy Rev. 2016, 65, 319331,  DOI: 10.1016/j.rser.2016.07.020
  11. 11
    Jones, E.; Qadir, M.; van Vliet, M. T. H.; Smakhtin, V.; Kang, S. The State of Desalination and Brine Production: A Global Outlook.. Sci. Total Environ. 2019, 657, 13431356,  DOI: 10.1016/j.scitotenv.2018.12.076
  12. 12
    Peñate, B.; García-Rodríguez, L. Current Trends and Future Prospects in the Design of Seawater Reverse Osmosis Desalination Technology.. Desalination 2012, 284, 18,  DOI: 10.1016/j.desal.2011.09.010
  13. 13
    Scholes, R. C.; Stiegler, A. N.; Anderson, C. M.; Sedlak, D. L. Enabling Water Reuse by Treatment of Reverse Osmosis Concentrate: The Promise of Constructed Wetlands.. ACS Environ. Au 2021, 1 (1), 717,  DOI: 10.1021/acsenvironau.1c00013
  14. 14
    Clark, C. E.; Veil, J. A. Produced Water Volumes and Management Practices in the United States; ANL/EVS/R-09-1; Argonne National Laboratory, U.S. Department of Energy, 2009; 1007397.  DOI: 10.2172/1007397 .
  15. 15
    Landsman, M. R.; Sujanani, R.; Brodfuehrer, S. H.; Cooper, C. M.; Darr, A. G.; Davis, R. J.; Kim, K.; Kum, S.; Nalley, L. K.; Nomaan, S. M. Water Treatment: Are Membranes the Panacea?. Annu. Rev. Chem. Biomol. Eng. 2020, 11 (1), 559585,  DOI: 10.1146/annurev-chembioeng-111919-091940
  16. 16
    Sujanani, R.; Landsman, M. R.; Jiao, S.; Moon, J. D.; Shell, M. S.; Lawler, D. F.; Katz, L. E.; Freeman, B. D. Designing Solute-Tailored Selectivity in Membranes: Perspectives for Water Reuse and Resource Recovery.. ACS Macro Lett. 2020, 9 (11), 17091717,  DOI: 10.1021/acsmacrolett.0c00710
  17. 17
    Cooper, C. M.; McCall, J.; Stokes, S. C.; McKay, C.; Bentley, M. J.; Rosenblum, J. S.; Blewett, T. A.; Huang, Z.; Miara, A.; Talmadge, M. Oil and Gas Produced Water Reuse: Opportunities, Treatment Needs, and Challenges.. ACS EST Engg. 2022, 2 (3), 347366,  DOI: 10.1021/acsestengg.1c00248
  18. 18
    Sabie, R. P.; Pillsbury, L.; Xu, P. Spatiotemporal Analysis of Produced Water Demand for Fit-For-Purpose Reuse─A Permian Basin, New Mexico Case Study.. Water 2022, 14 (11), 1735,  DOI: 10.3390/w14111735
  19. 19
    Guest, J. S.; Skerlos, S. J.; Barnard, J. L.; Beck, M. B.; Daigger, G. T.; Hilger, H.; Jackson, S. J.; Karvazy, K.; Kelly, L.; Macpherson, L. A New Planning and Design Paradigm to Achieve Sustainable Resource Recovery from Wastewater.. Environ. Sci. Technol. 2009, 43 (16), 61266130,  DOI: 10.1021/es9010515
  20. 20
    World Health Organization. A Global Overview of National Regulations and Standards for Drinking-Water Quality, 2nd ed.; World Health Organization: Geneva, 2021.
  21. 21
    Hilal, N.; Kim, G. J.; Somerfield, C. Boron Removal from Saline Water: A Comprehensive Review.. Desalination 2011, 273 (1), 2335,  DOI: 10.1016/j.desal.2010.05.012
  22. 22
    Zaman, M.; Shahid, S. A.; Heng, L. Guideline for Salinity Assessment, Mitigation and Adaptation Using Nuclear and Related Techniques; Springer International Publishing: Cham, 2018.  DOI: 10.1007/978-3-319-96190-3 .
  23. 23
    Agre, P. Aquaporin Water Channels (Nobel Lecture).. Angew. Chem. Int. 2004, 43 (33), 42784290,  DOI: 10.1002/anie.200460804
  24. 24
    Latorre, R.; Miller, C. Conduction and Selectivity in Potassium Channels.. J. Membr. Biol. 1983, 71 (1–2), 1130,  DOI: 10.1007/BF01870671
  25. 25
    Werber, J. R.; Osuji, C. O.; Elimelech, M. Materials for Next-Generation Desalination and Water Purification Membranes.. Nat. Rev. Mater. 2016, 1 (5), 16018,  DOI: 10.1038/natrevmats.2016.18
  26. 26
    Hummer, G.; Rasaiah, J. C.; Noworyta, J. P. Water Conduction through the Hydrophobic Channel of a Carbon Nanotube.. Nature 2001, 414 (6860), 188190,  DOI: 10.1038/35102535
  27. 27
    Kalra, A.; Garde, S.; Hummer, G. Osmotic Water Transport through Carbon Nanotube Membranes.. Proc. Natl. Acad. Sci. U. S. A. 2003, 100 (18), 1017510180,  DOI: 10.1073/pnas.1633354100
  28. 28
    Ultrafast Transport in Nanotubes and Nanosheets. In Nanotubes and Nanosheets; Chen, Y. (Ian), Ed.; CRC Press, 2015; pp 290323.  DOI: 10.1201/b18073-15 .
  29. 29
    Tunuguntla, R. H.; Henley, R. Y.; Yao, Y.-C.; Pham, T. A.; Wanunu, M.; Noy, A. Enhanced Water Permeability and Tunable Ion Selectivity in Subnanometer Carbon Nanotube Porins.. Science 2017, 357 (6353), 792796,  DOI: 10.1126/science.aan2438
  30. 30
    Köhler, M. H.; Bordin, J. R. Surface, Density, and Temperature Effects on the Water Diffusion and Structure Inside Narrow Nanotubes.. J. Phys. Chem. C 2018, 122 (12), 66846690,  DOI: 10.1021/acs.jpcc.8b00112
  31. 31
    Striolo, A.; Chialvo, A. A.; Cummings, P. T.; Gubbins, K. E. Simulated Water Adsorption in Chemically Heterogeneous Carbon Nanotubes.. J. Chem. Phys. 2006, 124 (7), 074710,  DOI: 10.1063/1.2171349
  32. 32
    Majumder, M.; Chopra, N.; Hinds, B. J. Mass Transport through Carbon Nanotube Membranes in Three Different Regimes: Ionic Diffusion and Gas and Liquid Flow.. ACS Nano 2011, 5 (5), 38673877,  DOI: 10.1021/nn200222g
  33. 33
    Fornasiero, F.; In, J. B.; Kim, S.; Park, H. G.; Wang, Y.; Grigoropoulos, C. P.; Noy, A.; Bakajin, O. PH-Tunable Ion Selectivity in Carbon Nanotube Pores.. Langmuir 2010, 26 (18), 1484814853,  DOI: 10.1021/la101943h
  34. 34
    Corry, B. Water and Ion Transport through Functionalised Carbon Nanotubes: Implications for Desalination Technology.. Energy Environ. Sci. 2011, 4 (3), 751,  DOI: 10.1039/c0ee00481b
  35. 35
    Sadeghi, I.; Asatekin, A. Membranes with Functionalized Nanopores for Aromaticity-Based Separation of Small Molecules.. ACS Appl. Mater. Interfaces 2019, 11 (13), 1285412862,  DOI: 10.1021/acsami.9b00090
  36. 36
    Thomas, M.; Corry, B. Modifying Water Flow, Ion Selectivity, and Salt Rejection in Carbon Nanotubes via Surface Adsorption.. J. Phys. Chem. C 2020, 124 (6), 38203826,  DOI: 10.1021/acs.jpcc.9b08485
  37. 37
    Lyu, Q.; Kang, D.-Y.; Hu, S.; Lin, L.-C. Exploiting Interior Surface Functionalization in Reverse Osmosis Desalination Membranes to Mitigate Permeability–Selectivity Trade-off: Molecular Simulations of Nanotube-Based Membranes.. Desalination 2020, 491, 114537,  DOI: 10.1016/j.desal.2020.114537
  38. 38
    Gong, X.; Li, J.; Xu, K.; Wang, J.; Yang, H. A Controllable Molecular Sieve for Na + and K + Ions.. J. Am. Chem. Soc. 2010, 132 (6), 18731877,  DOI: 10.1021/ja905753p
  39. 39
    Miller, D. J.; Dreyer, D. R.; Bielawski, C. W.; Paul, D. R.; Freeman, B. D. Surface Modification of Water Purification Membranes.. Angew. Chem. Int. 2017, 56 (17), 46624711,  DOI: 10.1002/anie.201601509
  40. 40
    Zhao, X.; Zhang, R.; Liu, Y.; He, M.; Su, Y.; Gao, C.; Jiang, Z. Antifouling Membrane Surface Construction: Chemistry Plays a Critical Role.. J. Membr. Sci. 2018, 551, 145171,  DOI: 10.1016/j.memsci.2018.01.039
  41. 41
    Park, H. B.; Kamcev, J.; Robeson, L. M.; Elimelech, M.; Freeman, B. D. Maximizing the Right Stuff: The Trade-off between Membrane Permeability and Selectivity.. Science 2017, 356 (6343), eaab0530,  DOI: 10.1126/science.aab0530
  42. 42
    Meier, I. K.; Langsam, M.; Klotz, H. C. Selectivity Enhancement via Photooxidative Surface Modification of Polyimide Air Separation Membranes.. J. Membr. Sci. 1994, 94 (1), 195212,  DOI: 10.1016/0376-7388(93)E0174-I
  43. 43
    Peng, L. E.; Yao, Z.; Liu, X.; Deng, B.; Guo, H.; Tang, C. Y. Tailoring Polyamide Rejection Layer with Aqueous Carbonate Chemistry for Enhanced Membrane Separation: Mechanistic Insights, Chemistry-Structure-Property Relationship, and Environmental Implications.. Environ. Sci. Technol. 2019, 53 (16), 97649770,  DOI: 10.1021/acs.est.9b03210
  44. 44
    Guo, H.; Deng, Y.; Tao, Z.; Yao, Z.; Wang, J.; Lin, C.; Zhang, T.; Zhu, B.; Tang, C. Y. Does Hydrophilic Polydopamine Coating Enhance Membrane Rejection of Hydrophobic Endocrine-Disrupting Compounds?. Environ. Sci. Technol. Lett. 2016, 3 (9), 332338,  DOI: 10.1021/acs.estlett.6b00263
  45. 45
    Moskowitz, I.; Snyder, M. A.; Mittal, J. Water Transport through Functionalized Nanotubes with Tunable Hydrophobicity.. J. Chem. Phys. 2014, 141 (18), 18C532,  DOI: 10.1063/1.4897974
  46. 46
    Xi, E.; Venkateshwaran, V.; Li, L.; Rego, N.; Patel, A. J.; Garde, S. Hydrophobicity of Proteins and Nanostructured Solutes Is Governed by Topographical and Chemical Context.. Proc. Natl. Acad. Sci. U. S. A. 2017, 114 (51), 1334513350,  DOI: 10.1073/pnas.1700092114
  47. 47
    Acharya, H.; Vembanur, S.; Jamadagni, S. N.; Garde, S. Mapping Hydrophobicity at the Nanoscale: Applications to Heterogeneous Surfaces and Proteins.. Faraday Discuss. 2010, 146, 353,  DOI: 10.1039/b927019a
  48. 48
    Giovambattista, N.; Debenedetti, P. G.; Rossky, P. J. Hydration Behavior under Confinement by Nanoscale Surfaces with Patterned Hydrophobicity and Hydrophilicity.. J. Phys. Chem. C 2007, 111 (3), 13231332,  DOI: 10.1021/jp065419b
  49. 49
    Hua, L.; Zangi, R.; Berne, B. J. Hydrophobic Interactions and Dewetting between Plates with Hydrophobic and Hydrophilic Domains.. J. Phys. Chem. C 2009, 113 (13), 52445253,  DOI: 10.1021/jp8088758
  50. 50
    Willard, A. P.; Chandler, D. Coarse-Grained Modeling of the Interface between Water and Heterogeneous Surfaces.. Faraday Discuss. 2009, 141, 209220,  DOI: 10.1039/B805786A
  51. 51
    Reimann, P.; Van den Broeck, C.; Linke, H.; Hänggi, P.; Rubi, J. M.; Pérez-Madrid, A. Giant Acceleration of Free Diffusion by Use of Tilted Periodic Potentials.. Phys. Rev. Lett. 2001, 87 (1), 010602,  DOI: 10.1103/PhysRevLett.87.010602
  52. 52
    Harrach, M. F.; Klameth, F.; Drossel, B.; Vogel, M. Effect of the Hydroaffinity and Topology of Pore Walls on the Structure and Dynamics of Confined Water.. J. Chem. Phys. 2015, 142 (3), 034703,  DOI: 10.1063/1.4905557
  53. 53
    Ding, D.; Gao, P.; Ma, Q.; Wang, D.; Xia, F. Biomolecule-Functionalized Solid-State Ion Nanochannels/Nanopores: Features and Techniques.. Small 2019, 15 (32), 1804878,  DOI: 10.1002/smll.201804878
  54. 54
    Tang, C.; Lennon, E. M.; Fredrickson, G. H.; Kramer, E. J.; Hawker, C. J. Evolution of Block Copolymer Lithography to Highly Ordered Square Arrays.. Science 2008, 322 (5900), 429432,  DOI: 10.1126/science.1162950
  55. 55
    Le Caër, S.; Pin, S.; Esnouf, S.; Raffy, Q.; Renault, J. Ph.; Brubach, J.-B.; Creff, G.; Roy, P. A Trapped Water Network in Nanoporous Material: The Role of Interfaces.. Phys. Chem. Chem. Phys. 2011, 13 (39), 17658,  DOI: 10.1039/c1cp21980d
  56. 56
    Knight, A. W.; Kalugin, N. G.; Coker, E.; Ilgen, A. G. Water Properties under Nano-Scale Confinement.. Sci. Rep. 2019, 9 (1), 8246,  DOI: 10.1038/s41598-019-44651-z
  57. 57
    Rasaiah, J. C.; Garde, S.; Hummer, G. Water in Nonpolar Confinement: From Nanotubes to Proteins and Beyond.. Annu. Rev. Phys. Chem. 2008, 59 (1), 713740,  DOI: 10.1146/annurev.physchem.59.032607.093815
  58. 58
    Chakraborty, S.; Kumar, H.; Dasgupta, C.; Maiti, P. K. Confined Water: Structure, Dynamics, and Thermodynamics.. Acc. Chem. Res. 2017, 50 (9), 21392146,  DOI: 10.1021/acs.accounts.6b00617
  59. 59
    Zhang, S.; Hedtke, T.; Zhou, X.; Elimelech, M.; Kim, J.-H. Environmental Applications of Engineered Materials with Nanoconfinement.. ACS EST Engg. 2021, 1 (4), 706724,  DOI: 10.1021/acsestengg.1c00007
  60. 60
    Love, J. C.; Estroff, L. A.; Kriebel, J. K.; Nuzzo, R. G.; Whitesides, G. M. Self-Assembled Monolayers of Thiolates on Metals as a Form of Nanotechnology.. Chem. Rev. 2005, 105 (4), 11031170,  DOI: 10.1021/cr0300789
  61. 61
    Wang, S.; Lu, L.; Lu, X.; Cao, W.; Zhu, Y. Adsorption of Binary CO 2 /CH 4 Mixtures Using Carbon Nanotubes: Effects of Confinement and Surface Functionalization.. Sep. Sci. Technol. 2016, 51 (7), 10791092,  DOI: 10.1080/01496395.2016.1150296
  62. 62
    Jorgensen, W. L.; Chandrasekhar, J.; Madura, J. D.; Impey, R. W.; Klein, M. L. Comparison of Simple Potential Functions for Simulating Liquid Water.. J. Chem. Phys. 1983, 79 (2), 926935,  DOI: 10.1063/1.445869
  63. 63
    Risplendi, F.; Raffone, F.; Lin, L.-C.; Grossman, J. C.; Cicero, G. Fundamental Insights on Hydration Environment of Boric Acid and Its Role in Separation from Saline Water.. J. Phys. Chem. C 2020, 124 (2), 14381445,  DOI: 10.1021/acs.jpcc.9b10065
  64. 64
    Eastman, P.; Swails, J.; Chodera, J. D.; McGibbon, R. T.; Zhao, Y.; Beauchamp, K. A.; Wang, L.-P.; Simmonett, A. C.; Harrigan, M. P.; Stern, C. D. OpenMM 7: Rapid Development of High Performance Algorithms for Molecular Dynamics.. PLoS Comput. Biol. 2017, 13 (7), e1005659,  DOI: 10.1371/journal.pcbi.1005659
  65. 65
    Andersen, H. C. Rattle: A “Velocity” Version of the Shake Algorithm for Molecular Dynamics Calculations.. J. Comput. Phys. 1983, 52 (1), 2434,  DOI: 10.1016/0021-9991(83)90014-1
  66. 66
    Miyamoto, S.; Kollman, P. A. Settle: An Analytical Version of the SHAKE and RATTLE Algorithm for Rigid Water Models.. J. Comput. Chem. 1992, 13 (8), 952962,  DOI: 10.1002/jcc.540130805
  67. 67
    Shirts, M. R.; Chodera, J. D. Statistically Optimal Analysis of Samples from Multiple Equilibrium States.. J. Chem. Phys. 2008, 129 (12), 124105,  DOI: 10.1063/1.2978177
  68. 68
    Chodera, J. D.; Shirts, M. R. Replica Exchange and Expanded Ensemble Simulations as Gibbs Sampling: Simple Improvements for Enhanced Mixing.. J. Chem. Phys. 2011, 135 (19), 194110,  DOI: 10.1063/1.3660669

Cited By

ARTICLE SECTIONS
Jump To

This article is cited by 4 publications.

  1. Sally Jiao, Dennis C. Robinson Brown, M. Scott Shell. Relationships between Water’s Structure and Solute Affinity at Polypeptoid Brush Surfaces. Langmuir 2024, 40 (1) , 761-771. https://doi.org/10.1021/acs.langmuir.3c02971
  2. Alexandra Khlyustova, Yifan Cheng, Rong Yang. Membrane Pore Size Distribution by Design via Kinetic Engineering Using Initiated Chemical Vapor Deposition. Macromolecules 2023, 56 (16) , 6492-6500. https://doi.org/10.1021/acs.macromol.3c01078
  3. Hang Zhang, Sankaran Sundaresan, Michael A. Webb. Molecular Dynamics Investigation of Nanoscale Hydrophobicity of Polymer Surfaces: What Makes Water Wet?. The Journal of Physical Chemistry B 2023, 127 (22) , 5115-5127. https://doi.org/10.1021/acs.jpcb.3c00616
  4. Audra J. DeStefano, My Nguyen, Glenn H. Fredrickson, Songi Han, Rachel A. Segalman. Design of Soft Material Surfaces with Rationally Tuned Water Diffusivity. ACS Central Science 2023, 9 (5) , 1019-1024. https://doi.org/10.1021/acscentsci.3c00208
  • Abstract

    Figure 1

    Figure 1. (a) Model nanopore system allows for precise patterning of nonpolar methyl groups and polar hydroxyl groups along the CNT backbone. For ease of visualization, we present the flattened pore patterns with blue triangles representing methyl groups and red circles representing hydroxyl groups. (b) We first simulate the finite pore to compute the density of water, which we then set in the infinite pore to compute water and boric acid transport properties. (c) A genetic algorithm incorporating MD simulations and a learned surrogate function allows for automated identification of pore wall chemical patterns optimizing pore transport properties.

    Figure 2

    Figure 2. Genetic algorithm trajectories maximizing the difference between water flux and boric acid flux for (a) fixed hydroxyl group fraction and (b) varying hydroxyl group fraction show optimization to surfaces with rows of hydroxyl groups. Each marker corresponds to a pattern discovered by the genetic algorithm. Simulated generations are shown as green stars and surrogate model-driven generations as black circles. Inset patterns show the optimal patterns found after (1) the first set of simulated generations, (2) the first set of model-driven generations, (3) the second set of simulated generations, and (4) the second set of model-driven generations. The red line shows the maximum value of the objective function at each point in the optimization.

    Figure 3

    Figure 3. Pattern of methyl and hydroxyl groups has a significant effect on the flux of (a) water and (b) boric acid in nonequilibrium simulations and on (c) the diffusivity of boric acid in equilibrium simulations. For each panel, orange triangles represent patterns with a “single patch” of hydroxyl groups, spanning the pore axis, while blue circles represent “fully dispersed” patterns (examples of both are in the inset of a). Green diamonds represent other patterns where the hydroxyl groups form smaller patches (“patchy”). Two examples of such patterns are shown in the inset of b. These patterns also show lower boric acid flux compared to the fully dispersed patterns. The inset also includes a schematic of the periodically varying free energy landscape used in the analytical model to predict the reduction in boric acid diffusivity in the ringed pores. In c, green diamonds are only shown for patterns with rows of hydroxyl groups. The dashed line shows the analytical prediction described in the text based on a model of the periodic free energy landscape depicted in the inset of b.

    Figure 4

    Figure 4. Other solutes exhibit similar trends as boric acid in a selection of rationally patterned pores. Benzene, phenol, and isopropanol show maximal water flux relative to boric acid flux in pores with rows of hydroxyl groups, while for ammonia and arsenous acid, the optimal pattern is the fully methylated one. For the simulations with solutes besides boric acid, we use the same workflow as described for boric acid except that the production simulations in the infinite pore are run for 1 μs. The reported uncertainties are the standard error of the mean from three independent simulations.

  • References

    ARTICLE SECTIONS
    Jump To

    This article references 68 other publications.

    1. 1
      Austin, N. D.; Sahinidis, N. V.; Trahan, D. W. Computer-Aided Molecular Design: An Introduction and Review of Tools, Applications, and Solution Techniques.. Chem. Eng. Res. Des. 2016, 116, 226,  DOI: 10.1016/j.cherd.2016.10.014
    2. 2
      Mohr, B.; Shmilovich, K.; Kleinwächter, I. S.; Schneider, D.; Ferguson, A. L.; Bereau, T. Data-Driven Discovery of Cardiolipin-Selective Small Molecules by Computational Active Learning.. Chem. Sci. 2022, 13 (16), 44984511,  DOI: 10.1039/D2SC00116K
    3. 3
      Janet, J. P.; Ramesh, S.; Duan, C.; Kulik, H. J. Accurate Multiobjective Design in a Space of Millions of Transition Metal Complexes with Neural-Network-Driven Efficient Global Optimization.. ACS Cent. Sci. 2020, 6 (4), 513524,  DOI: 10.1021/acscentsci.0c00026
    4. 4
      Monroe, J. I.; Shell, M. S. Computational Discovery of Chemically Patterned Surfaces That Effect Unique Hydration Water Dynamics.. Proc. Natl. Acad. Sci. U. S. A. 2018, 115 (32), 80938098,  DOI: 10.1073/pnas.1807208115
    5. 5
      Monroe, J. I.; Jiao, S.; Davis, R. J.; Robinson Brown, D.; Katz, L. E.; Shell, M. S. Affinity of Small-Molecule Solutes to Hydrophobic, Hydrophilic, and Chemically Patterned Interfaces in Aqueous Solution.. Proc. Natl. Acad. Sci. U. S. A. 2021, 118 (1), e2020205118,  DOI: 10.1073/pnas.2020205118
    6. 6
      Ma, Y.; Aulicino, J. C.; Ferguson, A. L. Inverse Design of Self-Assembling Diamond Photonic Lattices from Anisotropic Colloidal Clusters.. J. Phys. Chem. B 2021, 125 (9), 23982410,  DOI: 10.1021/acs.jpcb.0c08723
    7. 7
      Meenakshisundaram, V.; Hung, J.-H.; Patra, T. K.; Simmons, D. S. Designing Sequence-Specific Copolymer Compatibilizers Using a Molecular-Dynamics-Simulation-Based Genetic Algorithm.. Macromolecules 2017, 50 (3), 11551166,  DOI: 10.1021/acs.macromol.6b01747
    8. 8
      Sherman, Z. M.; Howard, M. P.; Lindquist, B. A.; Jadrich, R. B.; Truskett, T. M. Inverse Methods for Design of Soft Materials.. J. Chem. Phys. 2020, 152 (14), 140902,  DOI: 10.1063/1.5145177
    9. 9
      Rijsberman, F. R. Water Scarcity: Fact or Fiction?. Agricultural Water Management 2006, 80 (1–3), 522,  DOI: 10.1016/j.agwat.2005.07.001
    10. 10
      Hamiche, A. M.; Stambouli, A. B.; Flazi, S. A Review of the Water-Energy Nexus.. Renew. Sust. Energy Rev. 2016, 65, 319331,  DOI: 10.1016/j.rser.2016.07.020
    11. 11
      Jones, E.; Qadir, M.; van Vliet, M. T. H.; Smakhtin, V.; Kang, S. The State of Desalination and Brine Production: A Global Outlook.. Sci. Total Environ. 2019, 657, 13431356,  DOI: 10.1016/j.scitotenv.2018.12.076
    12. 12
      Peñate, B.; García-Rodríguez, L. Current Trends and Future Prospects in the Design of Seawater Reverse Osmosis Desalination Technology.. Desalination 2012, 284, 18,  DOI: 10.1016/j.desal.2011.09.010
    13. 13
      Scholes, R. C.; Stiegler, A. N.; Anderson, C. M.; Sedlak, D. L. Enabling Water Reuse by Treatment of Reverse Osmosis Concentrate: The Promise of Constructed Wetlands.. ACS Environ. Au 2021, 1 (1), 717,  DOI: 10.1021/acsenvironau.1c00013
    14. 14
      Clark, C. E.; Veil, J. A. Produced Water Volumes and Management Practices in the United States; ANL/EVS/R-09-1; Argonne National Laboratory, U.S. Department of Energy, 2009; 1007397.  DOI: 10.2172/1007397 .
    15. 15
      Landsman, M. R.; Sujanani, R.; Brodfuehrer, S. H.; Cooper, C. M.; Darr, A. G.; Davis, R. J.; Kim, K.; Kum, S.; Nalley, L. K.; Nomaan, S. M. Water Treatment: Are Membranes the Panacea?. Annu. Rev. Chem. Biomol. Eng. 2020, 11 (1), 559585,  DOI: 10.1146/annurev-chembioeng-111919-091940
    16. 16
      Sujanani, R.; Landsman, M. R.; Jiao, S.; Moon, J. D.; Shell, M. S.; Lawler, D. F.; Katz, L. E.; Freeman, B. D. Designing Solute-Tailored Selectivity in Membranes: Perspectives for Water Reuse and Resource Recovery.. ACS Macro Lett. 2020, 9 (11), 17091717,  DOI: 10.1021/acsmacrolett.0c00710
    17. 17
      Cooper, C. M.; McCall, J.; Stokes, S. C.; McKay, C.; Bentley, M. J.; Rosenblum, J. S.; Blewett, T. A.; Huang, Z.; Miara, A.; Talmadge, M. Oil and Gas Produced Water Reuse: Opportunities, Treatment Needs, and Challenges.. ACS EST Engg. 2022, 2 (3), 347366,  DOI: 10.1021/acsestengg.1c00248
    18. 18
      Sabie, R. P.; Pillsbury, L.; Xu, P. Spatiotemporal Analysis of Produced Water Demand for Fit-For-Purpose Reuse─A Permian Basin, New Mexico Case Study.. Water 2022, 14 (11), 1735,  DOI: 10.3390/w14111735
    19. 19
      Guest, J. S.; Skerlos, S. J.; Barnard, J. L.; Beck, M. B.; Daigger, G. T.; Hilger, H.; Jackson, S. J.; Karvazy, K.; Kelly, L.; Macpherson, L. A New Planning and Design Paradigm to Achieve Sustainable Resource Recovery from Wastewater.. Environ. Sci. Technol. 2009, 43 (16), 61266130,  DOI: 10.1021/es9010515
    20. 20
      World Health Organization. A Global Overview of National Regulations and Standards for Drinking-Water Quality, 2nd ed.; World Health Organization: Geneva, 2021.
    21. 21
      Hilal, N.; Kim, G. J.; Somerfield, C. Boron Removal from Saline Water: A Comprehensive Review.. Desalination 2011, 273 (1), 2335,  DOI: 10.1016/j.desal.2010.05.012
    22. 22
      Zaman, M.; Shahid, S. A.; Heng, L. Guideline for Salinity Assessment, Mitigation and Adaptation Using Nuclear and Related Techniques; Springer International Publishing: Cham, 2018.  DOI: 10.1007/978-3-319-96190-3 .
    23. 23
      Agre, P. Aquaporin Water Channels (Nobel Lecture).. Angew. Chem. Int. 2004, 43 (33), 42784290,  DOI: 10.1002/anie.200460804
    24. 24
      Latorre, R.; Miller, C. Conduction and Selectivity in Potassium Channels.. J. Membr. Biol. 1983, 71 (1–2), 1130,  DOI: 10.1007/BF01870671
    25. 25
      Werber, J. R.; Osuji, C. O.; Elimelech, M. Materials for Next-Generation Desalination and Water Purification Membranes.. Nat. Rev. Mater. 2016, 1 (5), 16018,  DOI: 10.1038/natrevmats.2016.18
    26. 26
      Hummer, G.; Rasaiah, J. C.; Noworyta, J. P. Water Conduction through the Hydrophobic Channel of a Carbon Nanotube.. Nature 2001, 414 (6860), 188190,  DOI: 10.1038/35102535
    27. 27
      Kalra, A.; Garde, S.; Hummer, G. Osmotic Water Transport through Carbon Nanotube Membranes.. Proc. Natl. Acad. Sci. U. S. A. 2003, 100 (18), 1017510180,  DOI: 10.1073/pnas.1633354100
    28. 28
      Ultrafast Transport in Nanotubes and Nanosheets. In Nanotubes and Nanosheets; Chen, Y. (Ian), Ed.; CRC Press, 2015; pp 290323.  DOI: 10.1201/b18073-15 .
    29. 29
      Tunuguntla, R. H.; Henley, R. Y.; Yao, Y.-C.; Pham, T. A.; Wanunu, M.; Noy, A. Enhanced Water Permeability and Tunable Ion Selectivity in Subnanometer Carbon Nanotube Porins.. Science 2017, 357 (6353), 792796,  DOI: 10.1126/science.aan2438
    30. 30
      Köhler, M. H.; Bordin, J. R. Surface, Density, and Temperature Effects on the Water Diffusion and Structure Inside Narrow Nanotubes.. J. Phys. Chem. C 2018, 122 (12), 66846690,  DOI: 10.1021/acs.jpcc.8b00112
    31. 31
      Striolo, A.; Chialvo, A. A.; Cummings, P. T.; Gubbins, K. E. Simulated Water Adsorption in Chemically Heterogeneous Carbon Nanotubes.. J. Chem. Phys. 2006, 124 (7), 074710,  DOI: 10.1063/1.2171349
    32. 32
      Majumder, M.; Chopra, N.; Hinds, B. J. Mass Transport through Carbon Nanotube Membranes in Three Different Regimes: Ionic Diffusion and Gas and Liquid Flow.. ACS Nano 2011, 5 (5), 38673877,  DOI: 10.1021/nn200222g
    33. 33
      Fornasiero, F.; In, J. B.; Kim, S.; Park, H. G.; Wang, Y.; Grigoropoulos, C. P.; Noy, A.; Bakajin, O. PH-Tunable Ion Selectivity in Carbon Nanotube Pores.. Langmuir 2010, 26 (18), 1484814853,  DOI: 10.1021/la101943h
    34. 34
      Corry, B. Water and Ion Transport through Functionalised Carbon Nanotubes: Implications for Desalination Technology.. Energy Environ. Sci. 2011, 4 (3), 751,  DOI: 10.1039/c0ee00481b
    35. 35
      Sadeghi, I.; Asatekin, A. Membranes with Functionalized Nanopores for Aromaticity-Based Separation of Small Molecules.. ACS Appl. Mater. Interfaces 2019, 11 (13), 1285412862,  DOI: 10.1021/acsami.9b00090
    36. 36
      Thomas, M.; Corry, B. Modifying Water Flow, Ion Selectivity, and Salt Rejection in Carbon Nanotubes via Surface Adsorption.. J. Phys. Chem. C 2020, 124 (6), 38203826,  DOI: 10.1021/acs.jpcc.9b08485
    37. 37
      Lyu, Q.; Kang, D.-Y.; Hu, S.; Lin, L.-C. Exploiting Interior Surface Functionalization in Reverse Osmosis Desalination Membranes to Mitigate Permeability–Selectivity Trade-off: Molecular Simulations of Nanotube-Based Membranes.. Desalination 2020, 491, 114537,  DOI: 10.1016/j.desal.2020.114537
    38. 38
      Gong, X.; Li, J.; Xu, K.; Wang, J.; Yang, H. A Controllable Molecular Sieve for Na + and K + Ions.. J. Am. Chem. Soc. 2010, 132 (6), 18731877,  DOI: 10.1021/ja905753p
    39. 39
      Miller, D. J.; Dreyer, D. R.; Bielawski, C. W.; Paul, D. R.; Freeman, B. D. Surface Modification of Water Purification Membranes.. Angew. Chem. Int. 2017, 56 (17), 46624711,  DOI: 10.1002/anie.201601509
    40. 40
      Zhao, X.; Zhang, R.; Liu, Y.; He, M.; Su, Y.; Gao, C.; Jiang, Z. Antifouling Membrane Surface Construction: Chemistry Plays a Critical Role.. J. Membr. Sci. 2018, 551, 145171,  DOI: 10.1016/j.memsci.2018.01.039
    41. 41
      Park, H. B.; Kamcev, J.; Robeson, L. M.; Elimelech, M.; Freeman, B. D. Maximizing the Right Stuff: The Trade-off between Membrane Permeability and Selectivity.. Science 2017, 356 (6343), eaab0530,  DOI: 10.1126/science.aab0530
    42. 42
      Meier, I. K.; Langsam, M.; Klotz, H. C. Selectivity Enhancement via Photooxidative Surface Modification of Polyimide Air Separation Membranes.. J. Membr. Sci. 1994, 94 (1), 195212,  DOI: 10.1016/0376-7388(93)E0174-I
    43. 43
      Peng, L. E.; Yao, Z.; Liu, X.; Deng, B.; Guo, H.; Tang, C. Y. Tailoring Polyamide Rejection Layer with Aqueous Carbonate Chemistry for Enhanced Membrane Separation: Mechanistic Insights, Chemistry-Structure-Property Relationship, and Environmental Implications.. Environ. Sci. Technol. 2019, 53 (16), 97649770,  DOI: 10.1021/acs.est.9b03210
    44. 44
      Guo, H.; Deng, Y.; Tao, Z.; Yao, Z.; Wang, J.; Lin, C.; Zhang, T.; Zhu, B.; Tang, C. Y. Does Hydrophilic Polydopamine Coating Enhance Membrane Rejection of Hydrophobic Endocrine-Disrupting Compounds?. Environ. Sci. Technol. Lett. 2016, 3 (9), 332338,  DOI: 10.1021/acs.estlett.6b00263
    45. 45
      Moskowitz, I.; Snyder, M. A.; Mittal, J. Water Transport through Functionalized Nanotubes with Tunable Hydrophobicity.. J. Chem. Phys. 2014, 141 (18), 18C532,  DOI: 10.1063/1.4897974
    46. 46
      Xi, E.; Venkateshwaran, V.; Li, L.; Rego, N.; Patel, A. J.; Garde, S. Hydrophobicity of Proteins and Nanostructured Solutes Is Governed by Topographical and Chemical Context.. Proc. Natl. Acad. Sci. U. S. A. 2017, 114 (51), 1334513350,  DOI: 10.1073/pnas.1700092114
    47. 47
      Acharya, H.; Vembanur, S.; Jamadagni, S. N.; Garde, S. Mapping Hydrophobicity at the Nanoscale: Applications to Heterogeneous Surfaces and Proteins.. Faraday Discuss. 2010, 146, 353,  DOI: 10.1039/b927019a
    48. 48
      Giovambattista, N.; Debenedetti, P. G.; Rossky, P. J. Hydration Behavior under Confinement by Nanoscale Surfaces with Patterned Hydrophobicity and Hydrophilicity.. J. Phys. Chem. C 2007, 111 (3), 13231332,  DOI: 10.1021/jp065419b
    49. 49
      Hua, L.; Zangi, R.; Berne, B. J. Hydrophobic Interactions and Dewetting between Plates with Hydrophobic and Hydrophilic Domains.. J. Phys. Chem. C 2009, 113 (13), 52445253,  DOI: 10.1021/jp8088758
    50. 50
      Willard, A. P.; Chandler, D. Coarse-Grained Modeling of the Interface between Water and Heterogeneous Surfaces.. Faraday Discuss. 2009, 141, 209220,  DOI: 10.1039/B805786A
    51. 51
      Reimann, P.; Van den Broeck, C.; Linke, H.; Hänggi, P.; Rubi, J. M.; Pérez-Madrid, A. Giant Acceleration of Free Diffusion by Use of Tilted Periodic Potentials.. Phys. Rev. Lett. 2001, 87 (1), 010602,  DOI: 10.1103/PhysRevLett.87.010602
    52. 52
      Harrach, M. F.; Klameth, F.; Drossel, B.; Vogel, M. Effect of the Hydroaffinity and Topology of Pore Walls on the Structure and Dynamics of Confined Water.. J. Chem. Phys. 2015, 142 (3), 034703,  DOI: 10.1063/1.4905557
    53. 53
      Ding, D.; Gao, P.; Ma, Q.; Wang, D.; Xia, F. Biomolecule-Functionalized Solid-State Ion Nanochannels/Nanopores: Features and Techniques.. Small 2019, 15 (32), 1804878,  DOI: 10.1002/smll.201804878
    54. 54
      Tang, C.; Lennon, E. M.; Fredrickson, G. H.; Kramer, E. J.; Hawker, C. J. Evolution of Block Copolymer Lithography to Highly Ordered Square Arrays.. Science 2008, 322 (5900), 429432,  DOI: 10.1126/science.1162950
    55. 55
      Le Caër, S.; Pin, S.; Esnouf, S.; Raffy, Q.; Renault, J. Ph.; Brubach, J.-B.; Creff, G.; Roy, P. A Trapped Water Network in Nanoporous Material: The Role of Interfaces.. Phys. Chem. Chem. Phys. 2011, 13 (39), 17658,  DOI: 10.1039/c1cp21980d
    56. 56
      Knight, A. W.; Kalugin, N. G.; Coker, E.; Ilgen, A. G. Water Properties under Nano-Scale Confinement.. Sci. Rep. 2019, 9 (1), 8246,  DOI: 10.1038/s41598-019-44651-z
    57. 57
      Rasaiah, J. C.; Garde, S.; Hummer, G. Water in Nonpolar Confinement: From Nanotubes to Proteins and Beyond.. Annu. Rev. Phys. Chem. 2008, 59 (1), 713740,  DOI: 10.1146/annurev.physchem.59.032607.093815
    58. 58
      Chakraborty, S.; Kumar, H.; Dasgupta, C.; Maiti, P. K. Confined Water: Structure, Dynamics, and Thermodynamics.. Acc. Chem. Res. 2017, 50 (9), 21392146,  DOI: 10.1021/acs.accounts.6b00617
    59. 59
      Zhang, S.; Hedtke, T.; Zhou, X.; Elimelech, M.; Kim, J.-H. Environmental Applications of Engineered Materials with Nanoconfinement.. ACS EST Engg. 2021, 1 (4), 706724,  DOI: 10.1021/acsestengg.1c00007
    60. 60
      Love, J. C.; Estroff, L. A.; Kriebel, J. K.; Nuzzo, R. G.; Whitesides, G. M. Self-Assembled Monolayers of Thiolates on Metals as a Form of Nanotechnology.. Chem. Rev. 2005, 105 (4), 11031170,  DOI: 10.1021/cr0300789
    61. 61
      Wang, S.; Lu, L.; Lu, X.; Cao, W.; Zhu, Y. Adsorption of Binary CO 2 /CH 4 Mixtures Using Carbon Nanotubes: Effects of Confinement and Surface Functionalization.. Sep. Sci. Technol. 2016, 51 (7), 10791092,  DOI: 10.1080/01496395.2016.1150296
    62. 62
      Jorgensen, W. L.; Chandrasekhar, J.; Madura, J. D.; Impey, R. W.; Klein, M. L. Comparison of Simple Potential Functions for Simulating Liquid Water.. J. Chem. Phys. 1983, 79 (2), 926935,  DOI: 10.1063/1.445869
    63. 63
      Risplendi, F.; Raffone, F.; Lin, L.-C.; Grossman, J. C.; Cicero, G. Fundamental Insights on Hydration Environment of Boric Acid and Its Role in Separation from Saline Water.. J. Phys. Chem. C 2020, 124 (2), 14381445,  DOI: 10.1021/acs.jpcc.9b10065
    64. 64
      Eastman, P.; Swails, J.; Chodera, J. D.; McGibbon, R. T.; Zhao, Y.; Beauchamp, K. A.; Wang, L.-P.; Simmonett, A. C.; Harrigan, M. P.; Stern, C. D. OpenMM 7: Rapid Development of High Performance Algorithms for Molecular Dynamics.. PLoS Comput. Biol. 2017, 13 (7), e1005659,  DOI: 10.1371/journal.pcbi.1005659
    65. 65
      Andersen, H. C. Rattle: A “Velocity” Version of the Shake Algorithm for Molecular Dynamics Calculations.. J. Comput. Phys. 1983, 52 (1), 2434,  DOI: 10.1016/0021-9991(83)90014-1
    66. 66
      Miyamoto, S.; Kollman, P. A. Settle: An Analytical Version of the SHAKE and RATTLE Algorithm for Rigid Water Models.. J. Comput. Chem. 1992, 13 (8), 952962,  DOI: 10.1002/jcc.540130805
    67. 67
      Shirts, M. R.; Chodera, J. D. Statistically Optimal Analysis of Samples from Multiple Equilibrium States.. J. Chem. Phys. 2008, 129 (12), 124105,  DOI: 10.1063/1.2978177
    68. 68
      Chodera, J. D.; Shirts, M. R. Replica Exchange and Expanded Ensemble Simulations as Gibbs Sampling: Simple Improvements for Enhanced Mixing.. J. Chem. Phys. 2011, 135 (19), 194110,  DOI: 10.1063/1.3660669
  • Supporting Information

    Supporting Information

    ARTICLE SECTIONS
    Jump To

    The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acscentsci.2c01011.

    • Details regarding the simulation models and property calculations; genetic algorithm workflow; surrogate function; spectral clustering of the genetic algorithm-identified patterns; types of rationally designed patterns explored; other quantities computed for the rationally designed patterns including the partition coefficients of water and boric acid; analytical model describing boric acid diffusivity; details regarding the simulation of other solutes in the nanopore; permeability–selectivity behavior of random and genetic-algorithm-discovered patterns (PDF)


    Terms & Conditions

    Most electronic Supporting Information files are available without a subscription to ACS Web Editions. Such files may be downloaded by article for research use (if there is a public use license linked to the relevant article, that license may permit other uses). Permission may be obtained from ACS for other uses through requests via the RightsLink permission system: http://pubs.acs.org/page/copyright/permissions.html.

Pair your accounts.

Export articles to Mendeley

Get article recommendations from ACS based on references in your Mendeley library.

Pair your accounts.

Export articles to Mendeley

Get article recommendations from ACS based on references in your Mendeley library.

You’ve supercharged your research process with ACS and Mendeley!

STEP 1:
Click to create an ACS ID

Please note: If you switch to a different device, you may be asked to login again with only your ACS ID.

Please note: If you switch to a different device, you may be asked to login again with only your ACS ID.

Please note: If you switch to a different device, you may be asked to login again with only your ACS ID.

MENDELEY PAIRING EXPIRED
Your Mendeley pairing has expired. Please reconnect