Fast Ions Transportation in Nanochannel with ATPase‐Like Structure

Ion transport plays an important role in various biological processes because of the ability of ions to move rapidly in biological ion channel‐confined spaces. For example, rapid proton transport in ATPases is attributed to confined channel spaces and conjugated sites. According to molecular dynamics simulations, the confined spaces and conjugated sites in nanochannels can enhance ion transport. Herein, it is demonstrated that the ATPase‐like structures of sulfonic acid‐modified covalent organic framework nanochannels, which promote the formation of highly ordered and continuous water molecular chains and confined spaces, can support ion (H+, Li+, Na+, and K+) transport rates that are an order of magnitude higher than those of bulk water. The ion transport rates in the nanochannel are superior to those in other artificial channels. Moreover, the selectivity of cations in the nanochannel is evaluated using the diffusion potential with a concentration gradient. The simulations and experimental results demonstrate that confined spaces and conjugated sites are crucial for efficient ion transport in nanochannels modified by sulfonic acid groups as cation conductor materials.


Introduction
Among the numerous ion transport across membranes, fast ion transportation through biological ion channels is crucial for various life processes such as ATP synthesis, [1][2][3] physiological environment pH stabilization, [4] and charge homeostasis. [5]he extensively studied biological proton channels include gramicidin A (GramA), [6] M2 proton channels, [7] torque flagellum of MotA, [8,9] and the F0 region of ATPase, [2,10] which are involved in various life processes.Although the structures of these biological channels differ, the confinement effect and binding sites of the nano/ sub-nanochannels contribute to an orderly arrangement of ions, facilitating fast ion transportation. [11][14] Based on this result, we hypothesize that materials with similar structures to ATPase may result in fast ion transport (Figure 1b).The ordered arrangement of ions in a confined space facilitates fast ion transport, whereas the binding sites facilitate selective ion transport.
[30][31][32][33][34][35][36][37] However, artificial ion channels do not possess continuous, confined spaces.[51] However, the synthesis of MOF requires the use of metal ions.Covalent organic frameworks (COFs) are self-supporting and provide a confined space with an abundance of functional groups. [52,53]Because of these characteristics, we considered that COFs met the design criteria to achieve ion selectivity and rapid transportation.
To investigate the ion transport properties of ATPase-like structural channels, we focused on cations (H þ , Li þ , Na þ , and K þ ) transport through continuous confined channels of COFs.We used molecular dynamics (MD) simulations to predict the ion transport properties of three structurally similar channels with different functional groups and found that all channels exhibited both fast ion transport and strong cation selectivity when the functional group was sulfonic acid.
Based on the simulation results, we prepared self-supported NUS-9 membranes containing highly ordered one-dimensional channels with a radius of 7 Å radius to investigate proton and metal ions transport properties of the channel.The channel exhibited a higher cations transport number and mobility than the bulk.In addition, the MD simulation and experimental results showed that metal ion ordering was enhanced in the NUS-9 channels, facilitating rapid ion transport.Therefore, the unique ATPase-like structure of the NUS-9 channel demonstrated fast ion transport, and the bionic structure prompted further research in the fields of proton transport, biomimetics, and energy conversion.

Results and Discussion
To predict ion transport in nanochannels, MD simulations were performed using the large-scale atomic/molecular massively parallel simulator (LAMMPS) program.We constructed COFs models with different functional groups consisting of 10 layers of construction units, which were sliced from the unit cells.Channel-1, 2, and 3 represent the channels with functional groups R = -H, -COOH, and -SO 3 H, respectively (Figure 2).A KCl solution was used to simulate the properties of ions in channels.As shown in Figure 2a-c, the speed of the ions in the channels was higher than that in the bulk phase, demonstrating that the confined spaces in the channels facilitated ion transport.Subsequently, the ion density was analyzed to study the ion selectivity of the nanochannels, as shown in Figure 2d-i.The density distribution of K þ in the three channels is shown in Figure 2d-f.Channel-3 exhibited the highest K þ concentration.Figure 2g-i shows the density distribution of Cl À in the three channels, with Channel-3 exhibiting the lowest Cl À concentration.The largest disparity in the distribution of the anion and cation concentrations in Channel-3 confirmed its high cation selectivity.Based on these MD simulations, we predicted that Channel-3 was highly cation-selective and facilitated rapid transport.
To investigate ion transport through NUS-9, a suitable synthesis strategy for COFs is crucial for obtaining a self-supporting membrane.As shown in Figure 3a, two monomers, 1,3,5tricarboxyresorcinol (Tp) and 1,4-phenylenediamine-2-sulfonic acid (Pa-SO 3 H), were dissolved in different solvents, and the mixture was heated to obtain crystallized NUS-9 membranes (Figure 3b).A smooth surface was observed on the pristine NUS-9 nanochannel membrane using scanning electron microscopy (SEM) (Figure S1a, Supporting Information), which was attributed to NUS-9 crystallization.Moreover, the thickness of the dense channel membrane was %8 μm (Figure S1b, Supporting Information).The simulated NUS-9 was a quasi-hexagonal loop structure containing three sulfonic acid groups with a 1.4 nm aperture size. [52]The high crystallinity was confirmed using transmission electron microscopy (TEM) (Figure 3c,d).Furthermore, X-ray diffraction (XRD) patterns were obtained to investigate the crystallinity of the prepared NUS-9 channel membrane (Figure 3e).The NUS-9 channel membrane showed diffraction peaks at %4.7°, 8.9°, and 28°, which corresponded to the (100), (110), and (001) crystal planes, respectively. [53]We have added the measurement of the pore size of membrane, as shown in Figure S2, Supporting Information, the pore width centered at 13.6 Å, which was close to the theoretical pore size of 14 Å measured from the simulated crystal structure. [54]The Fourier transform infrared (FTIR) spectra peaks observed at 1,070 and 1,409 cm À1 along with symmetric and asymmetric O = S = O stretching bands, confirming the existence of sulfonic acid groups, and the typical stretching bands at 1,236 and 1,580 cm À1 assignable to C = C and C-N stretching bands (Figure S3, Supporting Information). [55]he ion transport properties of the NUS-9 channel were evaluated using I-V curves at various concentrations of HCl solution (Figure 4a).The ionic current measurements in the transmembrane revealed that the nanochannels transported H þ ions.As shown in Figure 4b, when the H þ concentration was lower than 0.001 M, the ion conductance decreased nonlinearly with the reduction in electrolyte concentration.This demonstrated that the proton transport was controlled by the surface charge of the sulfonic acid groups in the channels.To compare the transport of protons and other alkali metal ions (Li þ , Na þ , and K þ ) through the NUS-9 nanochannels, we plotted their I-V curves under the same conditions (Figure S4, Supporting Information) and obtained their ionic conductance.As shown in Figure 4c, a linear relationship between the ionic conductance and concentration was observed, indicating that the surface charge of the channel controlled ion transportation.Moreover, the ionic conductance of the protons was the highest at the same concentration.This result was more pronounced at low concentrations and was attributed to the selectivity of the electric double layer to counterions.
Eliminating the effect of the applied electric field on ion transport in the NUS-9 channel and considering that the ion transport through transmembrane is typically driven by energy consumption or concentration gradients in organisms, we created concentration gradients on both sides of the membrane to study the ion transport through the transmembrane under the same gradient conditions (Figure S5, Supporting Information).The transfer number of different ions was calculated by using the Nernst equation.As shown in Figure 4d, the diffusion potential (E diff ) and osmotic current (I os ) were obtained from I-V curves with a concentration gradient, and a typical I-V curve was is composed of the redox potential (E redox ) of the electrodes and the diffusion potential (E diff ) of the channels.The short-circuit current (I SC ) and V OC were calculated from the intersection points of the curve and coordinate axis, respectively.E diff was calculated by subtracting the contribution of E redox to the electrodes from the original total voltage (V OC ).The I-V curves of electrolytes (HCl, LiCl, NaCl, and KCl) with concentration gradients ranging from 10 to 10 5 (low concentration was always maintained at 10 À6 M) are shown in Figure 4e and S6, Supporting Information.As shown in Figure 4f, both V OC and I SC increased with increasing concentration gradients (obtained from I-V curves in Figure 4e and S6, Supporting Information).The ionic current of HCl was higher at the same concentration gradient, indicating that the proton transfer is faster than that of the other three ions (Figure 4g).Combined with the results shown in Figure 4c, these results demonstrated that protons were transported quickly through the NUS-9 channels.The cation transfer numbers (t þ ) were calculated using E diff (Figure 4h); the channels exhibited cationic selectivity.The t þ value of the proton was as high as 0.96, which was higher than those of the other three cations with different concentration gradients.This demonstrates the remarkable proton selectivity of the channel.The Grotthuss mechanism was proposed to study proton transport, and the reported activation energies obtained using AC impedance data for the Grotthuss mechanism were in the range of 9.65-38.59[58][59] These values were obtained by measuring the ions transport kinetics as a function of temperature in the range of 30-70 °C. Figure 4i shows the activation energies of ions obtained from the linear Arrhenius plots of ions (H þ , Li þ , Na þ , and K þ ) (Figure S7, Supporting Information).The activation energy of H þ was 13.63 kJ mol À1 , which was lower than 38.59 kJ mol À1 and higher than 9.65 kJ mol À1 , indicating that Grotthuss transportation occurred in the channels of NUS-9.Owing to the unique proton hop-turn mechanism, the orderliness of the transport path is a significant factor in the rapid transport of protons formed in the confined space of the nanochannel.The proton transfer was faster than that of the other three ions because of its lower activation energy than that of the other ions.
A numerical simulation was conducted to investigate the transport of different ions in the nanochannel and the activation energy of the ionic transmembrane.A model of NUS-9 channels was constructed based on the Poisson-Nernst-Planck (PNP) and Einstein-Stokes (ES) equations (Figure S6, Supporting Information).Figure 5a shows the I-V curves of the ions (H þ , Li þ , Na þ , and K þ ) at 0.1 mM of electrolyte concentration.Consistent with the experimental results, the protons exhibited the highest ionic current, indicating that the proton transmission was faster.Subsequently, the ionic current of the transmembrane was simulated at different temperatures (30-70 °C) by adjusting relevant parameters to estimate the activation energy of ion transport (Figure 5b,c and S9, Supporting Information).The ion current gradually increased with increasing temperature and was used to calculate the simulated activation energies.As shown in Figure 5d,e, Arrhenius plots were obtained by plotting transport kinetics as a function of temperature, which revealed a lower energy barrier for proton transport.The ion transport driven by the concentration gradient was obtained by setting the 10-fold gradient on both sides of the channel (Figure 5f ).The proton transport in the channel remained the highest, demonstrating the low energy barrier of proton transport.The experimental and simulated results indicated that the proton transport performance can be attributed to the low energy barrier of the transmembrane nanochannels.
To evaluate ion transport, the mobility of the ions in the NUS-9 channel was measured.Figure 6a shows the I-V curves of the NUS-9 channels with different electrolytes (HCl, LiCl, NaCl, and KCl) at a concentration of 10 À4 M, which were used to calculate the ionic mobility.Evidently, the ionic current of the protons was the highest at the same voltage.As shown in Figure 6b and Table S1, Supporting Information, NUS-9 channels exhibited remarkable proton mobility of 17.82 Â 10 À7 m 2 V À1 s, outperforming bulk (3.24 Â 10 À7 m 2 V À1 s), and other previous artificial proton channels [11,50,51] such as the 0.8 nm CNT and MIL-53-COOH (11.3 Â 10 À7 m 2 V À1 s) with ultrafast proton conduction.Compared with other ultrafast proton transfer channels, NUS-9 channels possess both confined spaces and an F1-like structure, which facilitate rapid proton transport.As shown in Figure 6c, the mobilities of Li þ , Na þ , and K þ ions were 3.33, 3.91, and 5.45 Â 10 À7 m 2 V À1 s, respectively, which were larger than those in the bulk phase.To explain this phenomenon, MD simulations were performed, and the ion distributions inside and outside the channel were studied.A simulation model was constructed to visualize the water conformation in a 1.4 nmsized confined NUS-9 channel.The polar sites in the basic molecular structure of NUS-9 included hydroxyl oxygen (O1) and sulfonyl oxygen (O2) (Figure 6d), which facilitated the formation of hydrogen bonds between the water molecules and sulfonic acid groups inside the channels.Figure 6e,f shows the simulation results of water molecular chains confined in NUS-9 channels.The water configuration was observed from the axial (Figure 6e) and radial (Figure 6f ) directions, and synergistic and highly ordered hydrogen bond networks were found between the water molecule chains in the COF channels and the NUS-9 wall, providing an efficient pathway for proton conduction.The sulfonic acid groups inside the channel formed hydrogen bonds with the confined water molecules, resulting in three highly ordered and continuous hydrogen-bonded chains in the channel for proton transfer.As shown in Figure 6g, the ion distribution inside the channel was more ordered than that in the bulk phase, explaining the fast transport of metal ions.Among the four ions, the protons exhibited the highest mobility, revealing that the Grotthuss hop-turn mechanism was crucial for proton transport in the NUS-9 channels.This indicated that the highly ordered and successive water molecule chains facilitated proton transmission.Because metal ion (Li þ , Na þ , and K þ ) transport differed from the Grotthuss hop-turn mechanism of protons, the three chains of ordered water molecules confined in the NUS-9 channels had no favorable effect on metal ion transport.Therefore, the proton mobility was higher than that of the other three ions.

Conclusion
MD simulations revealed that highly ordered and continuous chains of water molecules were formed inside the F1-like structure of the NUS-9 channel, resulting in rapid proton transport via the Grotthuss mechanism.Our results demonstrated that the confined spaces and binding sites in the channels promote the formation of ordered and continuous water molecule chains and cation selectivity, improving the mobility of protons and alkali metal ions (Li þ , Na þ , and K þ ) via sequential transportation.The mobility and cation transport numbers for the NUS-9 nanochannels confirmed our hypothesis of enhanced cation selectivity and transport rates in the channel and showed that the ion mobility exceeded that of the bulk.Moreover, a COF membrane (NUS-9) exhibited rapid proton transport driven by both an electric field and a concentration gradient.The fastest proton transport among the four ions (H þ , Li þ , Na þ , and K þ ) was attributed to the lowest proton transport activation energy, as confirmed by consistent experimental and simulation results.The confined spaces and binding sites of the channels caused ions to order, accelerating the transport and selectivity of protons and metal ions.We believe that this work will prompt further research into NUS-9 nanochannels for applications in fast ion transport, osmotic power harvesting, and proton pumping.
Fabrication of NUS-9 Membrane: Tp (0.1 mM) and Pa-SO 3 H (0.15 mM) were sonicated for 30 min in 2 mL NMP and 2 mL DMSO in glass bottles to ensure complete dissolution, respectively.The two solutions were mixed and continued to be sonicated for 30 min with 298 K to mix well.The mixture was spread evenly onto a clean glass surface using a homogenizer and heated at 80 °C in the oven for 5 days to complete the reaction and crystallization of the two monomers.The NUS-9 membrane was obtained by immersing the glass into water.The membrane was characterized and tested after drying at room temperature.
Characterizations: The surface and cross-sectional morphologies of the membrane were observed by scanning electron microscope (SEM, FEI Quanta FEG 250, USA).The X-ray diffraction (XRD) patterns were obtained by using Ultima IV XRD meter (Rigaku Corporation, Japan).
Electrical Measurements: The I-V curves of NUS-9 channels were measured using 6487 picoammeter (Keithley Instruments, Cleveland, OH).The membrane was clutched between two electrolytic cells, the two Ag/AgCl electrodes were used to test.
Numerical Simulation: The ionic transport behaviors of NUS-9 nanochannels were calculated with 2D PNP equations by using the commercial package COMSOL Multiphysics software (version 5.6) under steady state.The detailed model parameters and theoretical analysis were performed in Section S5 (Supporting Information).
Molecular Dynamics (MD) Simulation: The simulation system was composed of two water reservoirs and the COF channel, which was sliced from the unit cells.The detailed model parameters and theoretical analysis were performed in Section S8 (Supporting Information).

Figure 1 .
Figure 1.a) Schematic diagram of F-ATPase and b) ion transport through the NUS-9 channel.The three functional groups (R) in the nanochannel are similar to the three β binding sites of the F-ATPase.

Figure 2 .
Figure 2. MD simulation results for the NUS-9 channels: a-c) Ion transport speeds in different channels with different functional groups; d-f ) Density distribution of Kþ ions in three channels; g-i) Density distribution of Cl-ions in three channels.

Figure 3 .
Figure 3. a) Fabrication process of NUS-9 channel membrane; Characterization of NUS-9 membrane nanochannels; b) Optical photograph of NUS-9 membrane; c,d) TEM images of NUS-9 membrane; e) XRD patterns of COF membrane and simulated COF crystal.

Figure 4 .
Figure 4. Measurement of ion transport in NUS-9 nanochannels: a) I-V curves of NUS-9 nanochannels at different concentrations of HCl solution; b) Ion conductance of NUS-9 nanochannels calculated from I-V curves; c) Plots of ion conductance (H þ , Li þ , Na þ , and K þ ) of NUS-9 nanochannels versus concentration; d) I-V curves under a 100-fold HCl gradient (high concentration c H = 10 À4 M/low concentration c L = 10 À6 M).The insert diagram was the equivalent circuit of the system depicting the relationship between different components.The short-circuit current (I SC ) and open-circuit voltage (V OC ) were obtained from the intersection points of the curve and the coordinate axis, respectively.E diff was evaluated by subtracting the contribution of the E redox on electrodes from the original total voltage (V OC ).e) I-V curves of the NUS-9 membrane in HCl solution at different concentration gradients.f ) V OC and I SC values of the NUS-9 membrane in HCl solution at different concentration gradients; g) I SC values of different ions at the same concentration gradient; h) The cation transfer numbers (t þ ) of ions; i) Activation energy for ion transport calculated from the slope of the fitted Arrhenius plots (Figure S7, Supporting Information).

Figure 5 .
Figure 5. Numerical simulation analysis of ion transport: a) Simulated I-V curves of different ions for the NUS-9 nanochannel membrane; b) Simulated current versus temperature (I-T ) curve of proton; c) Calculated cation distribution based on the NUS-9 model at different temperatures; d) Simulated Arrhenius plots of different ions (H þ , Li þ , Na þ , and K þ ) for the NUS-9 membrane nanochannel, which are obtained from the I-T curves (Figure 3b and S6, Supporting Information); e) Simulated activation energies of different ions for transport through the NUS-9 channel model, which are obtained from the slope of the fitted line.f ) Simulated I-T curves of ions under a 10-fold concentration gradient.

Figure 6 .
Figure 6.a) I-V curves of the NUS-9 channel in different electrolytes of 0.0001 M concentration; b) Proton mobility in various artificial proton channels and bulk water; c) Ionic mobility in NUS-9 nanochannels and the corresponding bulk solutions; d) Basic molecular structure of NUS-9 with polar sites being hydroxyl oxygen (O1) and sulfonyl oxygen (O2).The water configuration in the NUS-9 channel model was observed from e) axial and f ) radial directions, and the synergistic and highly ordered hydrogen bond networks were found between water molecule chains in COF channels and the NUS-9 wall.g) Metal ion distribution (Li þ , Na þ , and K þ ) in the NUS-9 channel and bulk phase.