Changing Water Affinity from Hydrophobic to Hydrophilic in

Article pubs.acs.org/Langmuir

Changing Water Affinity from Hydrophobic to Hydrophilic in Hydrophobic Channels Tomonori Ohba,*,† Shotaro Yamamoto,† Tetsuya Kodaira,‡ and Kenji Hata‡ †

Graduate School of Science, Chiba University, 1-33 Yayoi, Inage, Chiba 263-8522, Japan National Institute of Advanced Industrial Science and Technology (AIST), 1-1-1 Higashi, Tsukuba, Ibaraki 305-8565, Japan

‡

S Supporting Information *

ABSTRACT: The behavior of water at hydrophobic interfaces can play a significant role in determining chemical reaction outcomes and physical properties. Carbon nanotubes and aluminophosphate materials have one-dimensional hydrophobic channels, which are entirely surrounded by hydrophobic interfaces. Unique water behavior was observed in such hydrophobic channels. In this article, changes in the water affinity in one-dimensional hydrophobic channels were assessed using water vapor adsorption isotherms at 303 K and grand canonical Monte Carlo simulations. Hydrophobic behavior of water adsorbed in channels wider than 3 nm was observed for both adsorption and desorption processes, owing to the hydrophobic environment. However, water showed hydrophilic properties in both adsorption and desorption processes in channels narrower than 1 nm. In intermediate-sized channels, the hydrophobic properties of water during the adsorption process were seen to transition to hydrophilic behavior during the desorption process. Hydrophilic properties in the narrow channels for both adsorption and desorption processes are a result of the relatively strong water−channel interactions (10−15 kJ mol−1). In the 2−3 nm channels, the water−channel interaction energy of 4−5 kJ mol−1 was comparable to the thermal translational energy. The cohesive water interaction was approximately 35 kJ mol−1, which was larger than the others. Thus, the water affinity change in the 2−3 nm channels for the adsorption and desorption processes was attributed to weak water−channel interactions and strong cohesive interactions. These results are inherently important to control the properties of water in hydrophobic environments.

■

INTRODUCTION Detailed understanding of water−solid interfaces is of significant importance in many chemical reactions and biological systems. Particularly, the role of hydrophobic solid nanopores has been a focus for batteries, electric double-layer capacitors, and biological membranes.1−8 Unique behavior of water in hydrophobic nanopores is induced by hydrogen bonding and controls various activities. Graphite typically presents hydrophobic interfaces, and thus graphitic carbon nanopores are hydrophobic. Water vapor adsorption isotherms are influenced significantly by the hydrophobicity/hydrophilicity of carbon nanopores in activated carbons.9−11 Water vapor was seen to rarely adsorb in hydrophobic carbon nanopores at low pressures, while hydrophilic carbon nanopores adsorbed water vapor at low pressure.10,12−15 In the water adsorption process, water condensation was also observed in hydrophobic zeolites.16,17 Grand canonical Monte Carlo (GCMC) simulations revealed that water vapor could be adsorbed even in purely hydrophobic nanopores.18,19 Water vapor adsorption isotherms sometimes demonstrate an adsorption hysteresis loop by having differently shaped adsorption and desorption isotherms.11,18−24 Adsorption hysteresis is caused by kinetically forbidden structural transformations of adsorbed water.20 Thus, the adsorbed structures © 2015 American Chemical Society

of water in the adsorption and desorption processes were different, inducing adsorption hysteresis; water clusters were formed in the adsorption process, and a monolayer-like structure was observed in the desorption process.25 Cluster formation during adsorption is a key for water vapor adsorption in a hydrophobic environment because water can be stabilized by cluster formation.26,27 Thus, water condensation in hydrophobic nanopores is observed to proceed through a cluster growth mechanism.28 Carbon nanotubes (CNTs) and aluminophosphate (AlPO4) materials are also considered hydrophobic and have onedimensional channels, which could be useful models of biological channels.29−33 Unusual water structures in CNTs have been reported.34−39 Wang and co-workers indicated the temperature dependence of hydrophobic and hydrophilic properties of water in CNTs induced by water structure change.40 An anomalous ice phase was observed in CNTs at high pressure and even at ambient pressure in molecular dynamics simulations.34,35 We also experimentally observed such ice-like clusters in CNTs under ambient conditions.37 Received: November 20, 2014 Revised: December 24, 2014 Published: January 14, 2015 1058

DOI: 10.1021/la504522x Langmuir 2015, 31, 1058−1063

Article

Langmuir

Figure 1. Structural characterization of carbon nanotubes. (a) N2 adsorption isotherms at 77 K, (b) Raman scattering from the G- and D-bands of carbon nanotubes, and (c) C 1s peaks and (d) O 1s peaks of XPS spectra of carbon nanotubes. Blue, red, green, and black curves represent the 1, 2, 3, and 5 nm carbon nanotubes, respectively. 0.73 and 0.36 nm in diameter, respectively, based on their crystal structures.32 Here 1, 2, 3, and 5 nm CNTs were single-, single-, double-, and multiwalled CNTs, respectively. All CNTs were heated at 673 K for 1 h in an oxygen atmosphere to remove the CNT end-caps. Adsorption isotherms of N2 at 77 K and water vapor at 303 K were measured using a volumetric adsorption apparatus (Autosorb-1; Quantachrome Instruments Co., Boynton Beach, FL) and an inhouse-built gravimetric water vapor adsorption apparatus, respectively. CNT samples were evacuated at 423 K, below 10 mPa for more than 2 h prior to adsorption measurements. Channel diameters were also evaluated from αS analysis of the N2 adsorption isotherms. Raman spectra were measured using an Nd:YAG laser at a power of 0.1 mW (NRS-3000; JASCO Co., Tokyo, Japan). The surface oxygen groups of CNTs were assessed by X-ray photoelectron microscopy (XPS) every 0.1 keV using Mg Kα radiation at 10 kV and 10 mA (JPS-9010MX, JEOL Co., Tokyo, Japan). GCMC simulations of water in CNTs were performed using the Metropolis sampling scheme to evaluate the amount of water adsorbed in the CNTs. An armchair-type CNT of 4.8 nm length was positioned in the center of a 10 × 10 × 10 nm3 unit cell. A 3 × 3 × 3 nm3 unit cell was used for bulk liquid water. Periodic boundary conditions were applied in all three dimensions. Each calculation cycle was 1 × 107 steps. The intermolecular interaction potentials of water molecules and carbon atoms were calculated using the Lennard-Jones and Coulomb potential models. The potential parameters in the calculations were as follows: εH2O/kB = 80.5 K, σH2O = 0.312 nm, qH2O/e = ±0.241 C, εC/kB = 28.0 K, and σC = 0.340 nm. Quadrupole interaction of a CNT was neglected in the calculations. The Ewald summation technique was simply used in the three Cartesian directions in the unit cells to ensure accurate calculation of the Coulomb interaction between partial charges. Intermolecular interactions of water adsorbed in an internal channel of a CNT were calculated using the above potential models. Details of the simulation procedure have been reported in our previous papers.25,38

Weak hydrogen bonding might be assessed by strong orientation effects of water assemblies on hydrophobic interfaces.41 However, ice-like clusters by strong hydrogen bonding were observed in the 2−3 nm CNTs.38 This anomalous structure formation is a result of unusual hydrogen bonds in CNTs.39 Structural properties of water adsorbed in CNTs affected the water dynamics as well. Holt and co-workers reported fast water transport through CNTs with narrower diameters than 2 nm.42 In such CNTs, adsorbed water arranged in a single file.43 Fast water transport is caused by free hydrogen bonds in these narrow CNTs.44,45 Thus, a hydrophobic environment in CNTs is essential for fast water transport as well as unusual structure formation. Hydrophobicity in such one-dimensional channels is key to observe these unique physical properties of water, or in other words, the affinity of water in one-dimensional channels is crucial for observing such unique behaviors. An adsorption isotherm of water vapor is a measure to assess hydrophobic/ hydrophilic nature in one-dimensional channels. In this study, hydrophobicity/hydrophilicity in one-dimensional channels was evaluated by water vapor adsorption isotherms in various CNTs and AlPO4, and the water vapor adsorption mechanism in onedimensional channels was evaluated using GCMC simulations.

■

EXPERIMENTAL SECTION

Four CNTs were prepared using high-pressure CO conversion and chemical vapor deposition (CNTs were provided by Unidym Inc., Menio Park, CA; Hata group, AIST; NanoLab Inc., Waltham, MA).46 AlPO4-5 and AlPO4-C provided by Kodaira were also used for supplying cylindrical nanoporous materials.33 The CNTs were named as 1, 2, 3, and 5 nm CNTs, according to the CNT diameters, which were obtained from transmission electron microscopy, reported in our previous paper.38 AlPO4-5 and AlPO4-C have cylindrical channels of 1059

DOI: 10.1021/la504522x Langmuir 2015, 31, 1058−1063

Article

Langmuir

■

RESULTS AND DISCUSSION

The channel structures of CNTs were evaluated by the N2 adsorption isotherms at 77 K in Figure 1a. Despite having the largest channels, the adsorption amount in the 5 nm CNT was smaller than the others because the 5 nm CNT comprised multiple carbon walls. In the same way, the 3 nm CNT composed of double carbon walls had a relatively small amount of adsorbed N2. Thus, the maximal adsorption amount was seen in the single-walled 2 nm CNT. However, only a small adsorbed amount for the single-walled 1 nm CNT was observed because of its extremely narrow channels. The channel volumes, as evaluated from the αS analysis, of the 1, 2, 3, and 5 nm CNTs were 0.168, 0.811, 0.713, and 0.287 mL g−1, respectively. The channel diameters for the same CNT species were 1.4, 2.4, 2.9, and 4.6 nm, by assuming cylindrical channel structure. The channel diameters agree well with those found in the preceding study obtained from the transmission electron microscopy images.38 The G- and D-bands of CNTs in Raman scatterings, as shown in Figure 1b, originate from ordered and less ordered carbons, respectively. The G/D ratios of the 1, 2, 3, and 5 nm CNTs were 8.4, 4.7, 8.1, and 1.0, respectively. The 1 and 3 nm CNTs thus have highly ordered structure, whereas the 5 nm CNT was composed of somewhat less ordered carbon atoms. The G-bands of the CNTs were observed around 1590 cm−1, but also at 1540 cm−1 for the 1 nm CNT and 1620 cm−1 for the 5 nm CNT. The shoulder peaks indicated that the 1 and 5 nm CNTs had metallic and semiconducting CNT features, respectively.47 Surface oxygen groups on the CNTs were characterized by XPS C 1s and O 1s peaks. The O/C ratios of the 1, 2, 3, and 5 nm CNTs were 1.3, 0.5, 1.2, and 1.2%, respectively; thus, channel surfaces of CNTs have a highly hydrophobic nature. AlPO4 materials also have a hydrophobic nature, estimated from water desorption temperature reported in the previous paper.33 Water vapor adsorption isotherms on CNTs and AlPO4 are shown in Figure S1. Here, P and P0 are the water vapor pressure and saturated water vapor pressure. Filling factors were obtained by dividing the adsorption amounts by the channel volumes, as shown in Figure 2. Channel volumes of CNTs were evaluated from the above αS analysis of the N2 adsorption isotherms, while those of AlPO4 were calculated from the crystal structures (0.29 mL g−1 for the AlPO4-5 and 0.24 mL g−1 for the AlPO4-C). Water vapor was rarely adsorbed in CNTs, except for the 1 nm CNT, below P/P0 = 0.5, as was expected from the hydrophobic nature of the CNTs. Significant adsorption of water vapor was seen in the relatively high pressure range between 0.6 and 1.0. Water vapor adsorption in the 1 nm CNT was gradually observed in the low pressure regime. Water vapor adsorption and desorption isotherms for the 1, 2, and 3 nm CNTs indicated adsorption hysteresis associated with concave and convex curves. However, for the 5 nm CNT, no obvious adsorption hysteresis was observed. Desorption curves were apparently dependent on the CNT diameter: from concave to convex curves according to decreasing CNT diameter for the 2−5 nm CNTs. Water vapor adsorbed in the 1 nm CNT gradually desorbed, and the desorption curve approached the adsorption curve below P/P0 = 0.4, which was the lowest decay pressure of the CNTs. Water vapor quickly adsorbed in AlPO4-5 at P/P0 = 0.3 and desorbed at P/P0 = 0.2. Water adsorbed in AlPO4-5 was less packed in the channels because the maximum filling factor in AlPO4-5 is 0.7a result of significant structural restriction in the channels.

Figure 2. Water vapor adsorption isotherms of CNTs (a) and AlPO4 (b) at 303 K. Solid and open symbols represent adsorption and desorption curves, respectively. (a) ●, 1 nm CNT; ■, 2 nm CNT; ◆, 3 nm CNT; ▲, 5 nm CNT. (b) ●, AlPO4-5 (diameter 0.73 nm); ■, AlPO4-C (diameter = 0.36 nm).

Conversely, water vapor adsorption in AlPO4-C started below P/P0 = 0.05, and water vapor was fully adsorbed. Furthermore, water vapor was hardly desorbed by evacuation (until P/P0 = 0.006). Thus, water vapor was strongly captured in AlPO4-C, despite its hydrophobic nature. Adsorption and desorption pressures at the filling factor of 0.5 were obtained from the adsorption isotherms, as shown in Figure 3a. Adsorption uptakes were near unity at P/P0 = 0.8 for channel diameters larger than 2 nm. However, the adsorption pressure significantly decreased with decreasing channel diameter less than 1 nm. Desorption pressures gradually decreased with decreasing channel diameter, and the desorption pressure curve was close to the adsorption pressure curve. Thus, obvious adsorption hysteresis was diminished in channels narrower than 1 nm. Adsorption and desorption pressures were strongly related to the water affinity in the hydrophobic channels. That is, a high water affinity in adsorption processes was observed in the narrower channels, even in a hydrophobic environment, and the water affinity in desorption processes was relatively higher than that in adsorption processes. We attempted to employ the modified Freundlich equation to determine the water affinity to hydrophobic channels, as follows: n = K f ((P − P′)/P0)1/ m

(1)

where n, Kf, m, and P′ are the adsorption amount, coefficient, Freundlich exponent, and pressure shift The pressure shift coefficient was determined pressures of water vapor adsorption uptake and 1060

Freundlich coefficient. from the desorption

DOI: 10.1021/la504522x Langmuir 2015, 31, 1058−1063

Article

Langmuir

The Kf in an adsorption process was 0.2−0.3 in the channels larger than 1 nm but significantly increased in the subnanometer channels. Conversely, the Kf in the desorption process was larger than 0.8 except in the 5 nm channels. Therefore, water vapor adsorption affinity was particularly high in the subnanometer channels due to channel size and curvature effects, even in the hydrophobic environment, and the affinity in a desorption process was significant, except in the 5 nm channels. Thus, water interaction in hydrophobic channels was changed from hydrophobic to hydrophilic by adsorption in narrower channels as well as the process change from adsorption to desorption. GCMC simulations of water vapor adsorption in CNTs were performed to investigate why the water vapor affinity changed according to channel diameter as well as examine the processes of adsorption and desorption. Simulated adsorption uptake of water vapor was also shifted to higher pressures with increasing channel diameter, as shown in Figure 4a. The tendency agrees with the experimental adsorption isotherms in Figure 2, although not enough water could be adsorbed in the 3 nm channel below P/P0 = 1.0 in this simulation. The adsorption uptakes were attributed to stabilization of water adsorbed in channels, as shown in Figure 4b. The whole stabilization and water intermolecular interaction energies are separately plotted in Figure 4b. The intermolecular interaction energy in bulk liquid water was 42 kJ mol−1, roughly corresponding with the literature value of 44 kJ mol−1.48 Water−channel interaction energies were nearly 10−15 kJ mol−1 in the 0.8 and 2 nm channels, whereas water−channel interactions were 3−5 kJ mol−1 in the 2 and 3 nm channels. Thus, adsorption uptakes from low pressures in narrow channels are derived from the high stabilization energy of 10−15 kJ mol−1 by the water− channel interaction. Intermolecular interaction energies of the adsorbed water were approximately 22, 27, 37, and 27 kJ mol−1 in the 0.8, 1, 2, and 3 nm channels, respectively. The water− channel interaction energies in the 0.8 and 1.0 nm channels could be more comparable to the water intermolecular interactions than those in the larger channels, and the overall stabilization energies were approximately 38 kJ mol−1. Those stabilization energies lead to the hydrophilic properties of water adsorbed in such narrow channels in both the adsorption and desorption processes. The stabilization energy by the water−channel interaction in the 2 and 3 nm channels was comparable to the thermal energy

Figure 3. Adsorption and desorption threshold pressures (a) and modified Freundlich coefficient (b) as a function of channel diameter. Solid and open symbols (and solid and dashed curves) represent threshold pressures in adsorption and desorption processes, respectively. Blue and red symbols are for CNTs and AlPO4, respectively.

release near a filling factor of 0.1, as shown in Figure 3a. The modified Freundlich equation is the same as Freundlich equation when P′ = 0. Fitting curves of the adsorption and desorption isotherms are shown in Figure S1. The Freundlich coefficient is a measure of the adsorption affinity between adsorbate and adsorbent; strong adsorption affinities are indicated when Kf > 1 and weak adsorption affinity when Kf < 1. Therefore, the water vapor adsorption affinity was clearly assessed from Kf of the modified Freundlich equation in adsorption and desorption processes, as shown in Figure 3b.

Figure 4. Simulated water adsorption isotherms at 303 K (a) and water stabilities (b). Channel sizes were 0.8 (●), 1.0 (■), 2.0 (▲), and 3.0 nm (◆). Solid and dashed curves (filled and open symbols) represent the overall stabilization energy (including water intermolecular interactions and water−channel interaction energies) and the stabilization energy by water intermolecular interaction energies, respectively. 1061

DOI: 10.1021/la504522x Langmuir 2015, 31, 1058−1063

Article

Langmuir (3/2)kBT of 3.8 kJ mol−1. Thus, water in wider channels could not be adsorbed in a single form, that is, only by water−channel interaction, and thus, water vapor was rarely adsorbed in the low-pressure regions. However, in high-pressure regions, water molecules have a chance to aggregate with each other, even in wider channels. Water could gain enough stabilization energy, mainly by intermolecular interactions of water in the 2 nm channel, through aggregation to be equivalent to the water condensation energy. The stabilization energy by intermolecular interaction of water in the 2 nm channel was over 40 kJ mol−1 after water condensation in this channel, and thus, the high water stability prevents water desorption in the highpressure regime. The stability change of water in the adsorption and desorption processes leads to hydrophobic and hydrophilic properties of water in the 2 nm channel, respectively. Hydrophobic properties of water in the 3 nm channel for both adsorption and desorption processes were attributed to a low stabilization energy of 30 kJ mol−1, even at P/P0 = 1.0.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This research was supported by a JSPS KAKENHI Grant (No. 26706001) and a Research Fellowship from the Futaba Electronics Memorial Foundation.

■

(1) Agre, P.; Brown, D.; Nielsen, S. Aquaporin water channels: unanswered questions and unresolved controversies. Curr. Opin. Cell. Biol. 1995, 7 (4), 472−483. (2) Agre, P.; Kozono, D. Aquaporin water channels: molecular mechanisms for human diseases. FEBS Lett. 2003, 555 (1), 72−78. (3) MacKinnon, R. Potassium channels and the atomic basis of selective ion conduction (Nobel Lecture). Angew. Chem., Int. Ed. 2004, 43 (33), 4265−4277. (4) Simon, P.; Gogotsi, Y. Capacitive energy storage in nanostructured carbon-electrolyte systems. Acc. Chem. Res. 2012, 46 (5), 1094−1103. (5) Chmiola, J.; Yushin, G.; Gogotsi, Y.; Portet, C.; Simon, P.; Taberna, P. L. Anomalous increase in carbon capacitance at pore sizes less than 1 nm. Science 2006, 313 (5794), 1760−1763. (6) Simon, P.; Gogotsi, Y. Materials for electrochemical capacitors. Nat. Mater. 2008, 7 (11), 845−854. (7) Wu, C.; Zhu, X.; Ye, L.; Ouyang, C.; Hu, S.; Lei, L.; Xie, Y. Necklace-like hollow carbon nanospheres from the pentagon-including reactants: synthesis and electrochemical properties. Inorg. Chem. 2006, 45 (21), 8543−8550. (8) Huang, J.; Sumpter, B. G.; Meunier, V. A universal model for nanoporous carbon supercapacitors applicable to diverse pore regimes, carbon materials, and electrolytes. Chem.Eur. J. 2008, 14 (22), 6614−6626. (9) McBain, J. W.; Porter, J. L.; Sessions, R. F. The nature of the sorption of water by charcoal. J. Am. Chem. Soc. 1933, 55 (6), 2294− 2304. (10) Pierce, C.; Smith, R. N.; Wiley, J. W.; Cordes, H. Adsorption of water by carbon 1. J. Am. Chem. Soc. 1951, 73 (10), 4551−4557. (11) Wongkoblap, A.; Do, D. D. Adsorption of water in finite length carbon slit pore: comparison between computer simulation and experiment. J. Phys. Chem. B 2007, 111 (50), 13949−13956. (12) Lee, W. H.; Reucroft, P. J. Vapor adsorption on coal- and woodbased chemically activated carbons. Carbon 1999, 37 (1), 7−14. (13) Müller, E. A.; Gubbins, K. E. Molecular simulation study of hydrophilic and hydrophobic behavior of activated carbon surfaces. Carbon 1998, 36 (10), 1433−1438. (14) McCallum, C. L.; Bandosz, T. J.; McGrother, S. C.; Müller, E. A.; Gubbins, K. E. A molecular model for adsorption of water on activated carbon: comparison of simulation and experiment. Langmuir 1999, 15 (2), 533−544. (15) Müller, E. A.; Rull, L. F.; Vega, L. F.; Gubbins, K. E. Adsorption of water on activated carbons: A molecular simulation study. J. Phys. Chem. 1996, 100 (4), 1189−1196. (16) Desbiens, N.; Demachy, I.; Fuchs, A. H.; Kirsch-Rodeschini, H.; Soulard, M.; Patarin, J. Water condensation in hydrophobic nanopores. Angew. Chem., Int. Ed. 2005, 44 (33), 5310−5313. (17) Cailliez, F.; Trzpit, M.; Soulard, M.; Demachy, I.; Boutin, A.; Patarin, J.; Fuchs, A. H. Thermodynamics of water intrusion in nanoporous hydrophobic solids. Phys. Chem. Chem. Phys. 2008, 10 (32), 4817−4826. (18) Liu, J. C.; Monson, P. A. Does water condense in carbon pores? Langmuir 2005, 21 (22), 10219−10225. (19) Striolo, A.; Gubbins, K. E.; Gruszkiewicz, M. S.; Cole, D. R.; Simonson, J. M.; Chialvo, A. A.; Cummings, P. T.; Burchell, T. D.; More, K. L. Effect of temperature on the adsorption of water in porous carbons. Langmuir 2005, 21 (21), 9457−9467.

■

CONCLUSION In this study, water affinity changes in hydrophobic channels of 0.4−5 nm were characterized by water vapor adsorption isotherms and molecular simulations. Water affinity apparently depends on the channel size and the specific processes of adsorption and desorption. Hydrophilic behavior was observed in the adsorption and desorption processes for water in channels with diameters narrower than 2 nm. Conversely, in channels wider than 3 nm, the water vapor adsorption and desorption were consistent with hydrophobic behavior. Water vapor in intermediate sized channels of 2−3 nm displayed hydrophobic adsorption and hydrophilic desorption behavior. The dependence of water affinity on the channel size results from specific water and water−channel intermolecular interactions. An extremely narrow hydrophobic channel provides relatively strong stabilization energy for water, despite the somewhat restricted water−water intermolecular interaction. Water vapor adsorption is promoted by this strong stabilization, and thus, hydrophilic properties were observed despite being in a hydrophobic channel. In the medium-sized channels, weak water−channel interactions and strong water−water intermolecular interactions exist, inducing hydrophobic and hydrophilic properties in the adsorption and desorption processes, respectively. Here, the water−channel interaction is dominant in the adsorption process because water assemblies were rarely formed, and the contribution of water cohesive interactions became weak. However, both cohesive water and water− channel interactions were influential in the desorption process. Therefore, the water−channel interaction controls the hydrophilic/hydrophobic properties in the initial adsorption process, while stability contributions from both intermolecular interactions between water and water−channel controls the hydrophilic/hydrophobic properties in the desorption process.

■

ASSOCIATED CONTENT

S Supporting Information *

Water vapor adsorption isotherms of CNTs and AlPO4 at 303 K. This material is available free of charge via the Internet at http://pubs.acs.org.

■

REFERENCES

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected] (T.O.). 1062

DOI: 10.1021/la504522x Langmuir 2015, 31, 1058−1063

Article

Langmuir

through sub-2-nanometer carbon nanotubes. Science 2006, 312 (5776), 1034−1037. (43) Berezhkovskii, A.; Hummer, G. Single-file transport of water molecules through a carbon nanotube. Phys. Rev. Lett. 2002, 89 (6), 064503. (44) Joseph, S.; Aluru, N. R. Why are carbon nanotubes fast transporters of water? Nano Lett. 2008, 8 (2), 452−458. (45) Ohba, T.; Kaneko, K.; Endo, M.; Hata, K.; Kanoh, H. Rapid water transportation through narrow one-dimensional channels by restricted hydrogen bonds. Langmuir 2013, 29 (4), 1077−1082. (46) Hata, K.; Futaba, D. N.; Mizuno, K.; Namai, T.; Yumura, M.; Iijima, S. Water-assisted highly efficient synthesis of impurity-free single-walled carbon nanotubes. Science 2004, 306 (5700), 1362− 1364. (47) Dresselhaus, M. S.; Dresselhaus, G.; Saito, R.; Jorio, A. Raman spectroscopy of carbon nanotubes. Phys. Rep. 2005, 409 (2), 47−99. (48) Lide, D. R. CRC Handbook of Chemistry and Physics, 84th ed.; Taylor & Francis: London, 2003.

(20) Ohba, T.; Kaneko, K. Kinetically forbidden transformations of water molecular assemblies in hydrophobic micropores. Langmuir 2011, 27 (12), 7609−7613. (21) Rudisill, E. N.; Hacskaylo, J. J.; LeVan, M. D. Coadsorption of hydrocarbons and water on BPL activated carbon. Ind. Eng. Chem. Res. 1992, 31 (4), 1122−1130. (22) Mann, R.; Yousef, H. N. S.; Friday, D. K.; Mahle, J. J. Interpretation of water isotherm hysteresis for an activated charcoal using stochastic pore networks. Adsorption 1995, 1 (3), 253−264. (23) Inagaki, S.; Fukushima, Y. Adsorption of water vapor and hydrophobicity of ordered mesoporous silica, FSM-16. Microporous Mesoporous Mater. 1998, 21 (4-6), 667−672. (24) Liu, J. C.; Monson, P. A. Monte Carlo simulation study of water adsorption in activated carbon. Ind. Eng. Chem. Res. 2006, 45 (16), 5649−5656. (25) Ohba, T.; Kaneko, K. Cluster-associated filling of water molecules in slit-shaped graphitic nanopores. Mol. Phys. 2007, 105 (2-3), 139−145. (26) Ohba, T.; Kanoh, H.; Kaneko, K. Affinity transformation from hydrophilicity to hydrophobicity of water molecules on the basis of adsorption of water in graphitic nanopores. J. Am. Chem. Soc. 2004, 126 (5), 1560−1562. (27) Ohba, T.; Kanoh, H.; Kaneko, K. Cluster-growth-induced water adsorption in hydrophobic carbon nanopores. J. Phys. Chem. B 2004, 108 (39), 14964−14969. (28) Ohba, T.; Kanoh, H.; Kaneko, K. Water cluster growth in hydrophobic solid nanospaces. Chem.Eur. J. 2005, 11 (17), 4890− 4894. (29) Sansom, M. S.; Biggin, P. C. Water at the nanoscale. Nature 2001, 414 (6860), 156−159. (30) Capener, C. E.; Sansom, M. S. P. Molecular dynamics simulations of a K channel model: Sensitivity to changes in ions, waters, and membrane environment. J. Phys. Chem. B 2002, 106 (17), 4543−4551. (31) Hu, J.; Zhai, J.; Wu, F.; Tang, Z. Molecular dynamics study of the structures and dynamics of the iodine molecules confined in AlPO4-11 crystals. J. Phys. Chem. B 2010, 114 (49), 16481−16486. (32) Floquet, N.; Coulomb, J.; Dufau, N.; Andre, G. Structure and dynamics of confined water in AlPO4-5 zeolite. J. Phys. Chem. B 2004, 108 (35), 13107−13115. (33) Kodaira, T.; Ikeda, T. The selective adsorption of tellurium in the aluminosilicate regions of AFI- and MOR-type microporous crystals. Dalton Trans. 2014, 43 (37), 13979−13987. (34) Koga, K.; Gao, G.; Tanaka, H.; Zeng, X. C. Formation of ordered ice nanotubes inside carbon nanotubes. Nature 2001, 412 (6849), 802−805. (35) Mashl, R. J.; Joseph, S.; Aluru, N. R.; Jakobsson, E. Anomalously immobilized water: A new water phase induced by confinement in nanotubes. Nano Lett. 2003, 3 (5), 589−592. (36) Ohba, T.; Taira, S.; Hata, K.; Kaneko, K.; Kanoh, H. Predominant nanoice growth in single-walled carbon nanotubes by water-vapor loading. RSC Adv. 2012, 2 (9), 3634−3637. (37) Ohba, T.; Taira, S.; Hata, K.; Kanoh, H. Mechanism of sequential water transportation by water loading and release in singlewalled carbon nanotubes. J. Phys. Chem. Lett. 2013, 4 (7), 1211−1215. (38) Ohba, T. Size-dependent water structures in carbon nanotubes. Angew. Chem., Int. Ed. 2014, 126 (31), 8170−8174. (39) Byl, O.; Liu, J. C.; Wang, Y.; Yim, W. L.; Johnson, J. K.; Yates, J. T., Jr. Unusual hydrogen bonding in water-filled carbon nanotubes. J. Am. Chem. Soc. 2006, 128 (37), 12090−12097. (40) Wang, H. J.; Xi, X. K.; Kleinhammes, A.; Wu, Y. Temperatureinduced hydrophobic-hydrophilic transition observed by water adsorption. Science 2008, 322 (5898), 80−83. (41) Scatena, L. F.; Brown, M. G.; Richmond, G. L. Water at hydrophobic surfaces: weak hydrogen bonding and strong orientation effects. Science 2001, 292 (5518), 908−912. (42) Holt, J. K.; Park, H. G.; Wang, Y.; Stadermann, M.; Artyukhin, A. B.; Grigoropoulos, C. P.; Noy, A.; Bakajin, O. Fast mass transport 1063

DOI: 10.1021/la504522x Langmuir 2015, 31, 1058−1063