First Principles Study of Water Nanotubes Captured Inside Carbon

Subscriber access provided by ST FRANCIS XAVIER UNIV

New Concepts at the Interface: Novel Viewpoints and Interpretations, Theory and Computations

First Principles Study of Water Nanotubes Captured Inside Carbon/Boron Nitride Nanotubes Farzaneh Shayeganfar, Javad Beheshtian, and Rouzbeh Shahsavari Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.8b00856 • Publication Date (Web): 23 Aug 2018 Downloaded from http://pubs.acs.org on August 24, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

First Principles Study of Water Nanotubes Captured Inside Carbon/Boron Nitride Nanotubes Farzaneh Shayeganfar,1, 2 Javad Beheshtian*,3 and Rouzbeh Shahsavari*1 1

Department of Civil and Environmental Engineering, Rice University, Houston, TX 77005 2

3

Department of Energy Engineering and Physics, Amirkabir University, 14588, Tehran

Department of Chemistry, Shahid Rajaee Teacher Training University, 16875-163, Tehran

Abstract Water confined to nanopores such as carbon nanotubes (CNTs) exhibit different states, enabling the study of solid-like water nanotubes (WNTs) and the potential application of their properties due to confined effects. Herein, we report the interfacial interaction and particular stabilized boundaries of confined WNTs within CNTs and boron nitride nanotubes (BNNTs) using first principles calculations. We demonstrate that the intermolecular potential of nanotube walls exert diameter-dependent additive or subtractive van der Waals (vdW) pressure on the WNTs, altering the phase boundaries. Our results reveal that the most stable WNT@CNT is associated with a CNT diameter of 10.5 Å. By correlating the stability of WNTs with interfacial properties such as the vdW pressure and vibrational phonon modes of confined WNTs, we decode and compare various synergies in water interaction and stabilized states within the CNTs and BNNTs, including interfacial properties of WNT@BNNTs that are more significant than those of WNT@CNTs. Our results suggest that the transition of a water tube to an ice tube is strongly dependent on the diameter of the confining CNT or BNNT, providing new insights on leveraging the interfacial interaction mechanism of confined WNTs and their potential application for fabricating nanochannels and nanocapacitors.

ACS Paragon Plus 1 Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1.

INTRODUCTION

The unique properties of water, a ubiquitous medium and exible hydrogen network, strongly depend on intermolecular interactions.1 The dipolar and hydrogen bonding interactions due to polar OH bonds are responsible for the phase behavior of water and also play a great role in supramolecular processes.1 Strong confinement of water leads to a modified hydrogen network, exhibiting unusual dynamic and structural properties.2 Scientists have proposed some ways to better understand interactions in water when it is confined to nanoscopic spaces.3,4 For instance, Kurotobi and Murata5 recently captured a single molecule of water6 inside a cage of buckminsterfullerene.5 Moreover, Thilgen1 studied a single water molecule trapped inside hydrophobic C60. Shultz et al.7 elucidated local water structure by investigating local- and longrange responses of the hydrogen-bond network that exists in water with the C60 surface. Intermolecular coupling and cooperative motion of water are indicative of hydrogen bond potential.8-10 Recent observation of a unique phase transition in water when confined to isolated CNTs provides evidence for ice nanotubes.11 However, interpretation of the experimental data requires quantum computational modeling. When water is confined to nanometer-sized channels and pores, such as graphene layers and CNTs, the intermolecular energy of the CNT wall exerts vdW pressure on square-tube water, which is estimated to be in the gigapascal (GPa) range.12 Some computational techniques such as ab intio methods based on density functional theory (DFT)13 and quantum Monte Carlo (QMC) have recently been used to investigate experimental results of two-dimensional (2D) ice.12,14 For instance by DFT Chen et al.13,15 found that at high pressure, square ice exhibited the lowest enthalpy phase, supporting the experimental claim. Moreover, at lower pressures, a "pentagonal ice" phase had the lowest enthalpy, while a "hexagonal ice” phase dominated a previous phase at ambient pressure.13


Page 2 of 33

Page 3 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Using molecular dynamic simulation (MD), Chakraborty et al.2 studied the diffusion(?) of single water molecules confined to narrow CNTs, which exhibited solid-like ordering due to strong hydrogen bonding between nearest-neighbor molecules.2 Moreover, Sahu and Musharaf Ali16 investigated the behavior of both polar and nonpolar fluid(?) molecules within a CNT at the molecular level, reflecting the transport behavior of fluid molecules.16 In this paper, we employed first-principles calculations based on density functional theory (DFT) to probe the electronic properties of WNT and to determine whether WNT stability, when confined inside CNTs and BNNTs, was diameter-dependent over a vast diameter range of 8-12 Å. Ab initio techniques revealed that vdW pressure and phonon dispersion was the key feature for phase behavior of a WNT inside either a CNT or a BNNT. It is worth noting that vdW interactions affected the interfacial behavior of the confined WNT, i.e., encapsulated in CNT(BNNT) near the vdW boundary (hence the name vdW pressure). Moreover, the WNT had stronger interactions inside BNNTs than inside CNTs, likely because of the polarizable structure of BN. Studies of local WNT interactions within nanotubes suggest that a picture of the state diagram for a WNT can be deduced from the vibrational spectrum of WNT. The interaction properties of water and carbon nanomaterials introduce practical applications in water desalination and treatment and in hydrogen storage.17

2.

COMPUTATIONAL METHODS In order to study the structural and interfacial properties of WNT@CNTs and WNT@BNNTs,

we carried out ab initio computation based on a density functional theory (DFT) technique implemented in the SIESTA package.18 The initial geometry optimization of DFT calculations was acquired using periodic boundary conditions (PBC) within the generalized gradient approximation (PBE+D). The norm-conserving approach was used for Trouillier-Martins pseudopotentials and double-polarized atomic basis sets. ACS Paragon Plus 3 Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For the k-point sampling, a 12 ! 1 ! 1 Monkhorst-Pack grid was used for all configurations. Concerning the role of vdW force, we compared the DFT results based on PBE+D and vdW function (vdW-DF),19 as described by Roman-Perez and Soler.20 Geometry and structure were considered optimized when the force and total energy were less than 0.01 eV/Å and 10 eV, respectively. Chen et al.12,13 reported the formation of double-layer ice in a nano-confinement array, which is in agreement with monolayer and multilayer square ice observed by transmission electron microscopy experiments.14 Chen’s group carried out computations on randomly oriented water molecule structures, starting with pentagonal, hexagonal, square, and hexagonal-closepacked (HCP), in which the structures were non-confined (ambient) and confined (high pressure).12 The phase diagram indicates that double-layer ice favors square-tube (interlayer hydrogen bonding) as the stable structure in the GPa regime.12 Moreover, Kumar et al.21 investigated the structural stability of one-dimensional (1D) ice nanotubes (INTs) using DFT. They considered four different INTs, namely (4,0)-INT (square-tube), (5,0)-INT (pentagonal), (6,0)-INT (hexagonal), and (8,0)-INT (octagonal), with different lengths.21 They found that increasing the size of the ice tube enhanced the stabilization energy (SE). The calculated SE/H-bond was greater for (4,0)-INT64 [5.8 kcal/mol] than for (5,0)-INT65 and for (6,0)-INT66 [5.5 kcal/mol], indicating (4,0)-INT64 or square-tube water is the most stable structure. Among the hexagonal, pentagonal, square-tube and HCP structures, a square-tube structure creates interlayer hydrogen bonding (no lateral hydrogen bonding between tubes) in an AA stacking arrangement, as reported by Chen et al.12 Our goal was to investigate interfacial interactions and in particular the stability state of water within CNT/BNNTs and compare with results reported by Agrawal et al.11 The intermolecular energy of WNT inside CNTs (BNNTs), Einter, is defined by:


Page 4 of 33

Page 5 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Einter = E(WNT@CNT) - [ E(CNT) + E(WNT) ]


(1)

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

in which E(WNT@CNT) is the total energy for an optimized water nanotube layer within a CNT, E(CNT) is the total energy for the CNT, and E(WNT) is the total energy of isolated double-layer water in a supercell.

The vibrational frequencies of WNT@CNT/BNNT were calculated in a linear response framework with regard to replacing phonon frequency modes in the real space, indicating structure stability. Dynamical matrices with eight wave (q) vectors were computed in the irreducible wedge of the Brillouin zone (BZ), i.e., on an 8 ! 8 ! 8 grid.22,23

3.

RESULTS AND DISCUSSION Experimental observation that a unique phase transition of water occurs when confined to

isolated CNTs provides evidence for ice nanotubes, predicting to be extremely sensitive to the diameter of the CNT confining the water.11 Agrawal et al.11 studied the phase boundaries of water when confined to several isolated carbon nanotubes of different diameters using Raman spectroscopy. Their findings show that the single-walled carbon nanotubes with diameter 1.05 and 1.06 nm have a reversible freezing transition to 105-151 oC and 87-117 oC, respectively. However, interpretation of the experimental data requires quantum computation.

Our goal in this paper was to correlate the relative stability and vdW pressure of confined WNTs and compare with previous experimental observations. In this regard, we started with square tube water, which is the most stable structure in the GPa regime.12,24 The CNTs investigated in this study were zigzag by their chiral indices (n,0) for the following reasons.


Page 6 of 33

Page 7 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

First, Liu et al.25 reported the effect of CNT topology on the diffusion dynamics of water confined in zigzag and armchair CNTs. They found that the diffusion of water molecules was slower in zigzag CNTs than in armchair CNTs, consequently causing the water to behave in a solid-like phase.25 Second, Liu et al.26 investigated the transport properties (diffusivity, thermal conductivity and shear viscosity) and molecular distribution of water molecules confined in CNTs. The pore diameter, the potential of the confining wall, and the helicity of CNTs significantly affected the state of water. The structure of water, ordered layer of water, and solidlike behavior of water in CNTs was dominate in zigzag CNTs.26,27 Third, we performed periodic boundary calculation (PBC) for in nite(?) WNTs and CNTs, in which the periodic bond length of the WNT matches well with that of the zigzag CNT, as long as the armchair chirality did not commensurate with in nite(?) WNT geometry. Square-tube water or WNT not only has the proper geometry relevant to CNT, but also has a larger number of interlayer hydrogen bonding; therefore, WNT o ers(?) the perfect starting point. WNT as an exclusive water structure is an insulator, indicating a bandgap of Eg ~ 6 eV with a 1.72 Å hydrogen bond length. In the supporting information (SI), we carried out firstprinciples calculation on armchair CNTs, comparing with zigzag CNTs. Since no covalent bonding between WNTs and CNTs is established for a physisorbed state, we can expect only minor differences between armchair and zigzag CNTs.

3.1.

WNT@CNT

We first examined WNTs located inside different CNTs by using ab initio computations. We focused on the interfacial properties of WNTs trapped inside CNTs of different diameters, starting with a detailed description of the optimized structures for the square-tube WNT@CNT (Figure 1). ACS Paragon Plus 7 Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 33

For the intermolecular energy of WNT@CNTs, we used supercells in which CNT(10,0) contained 160 C atoms; CNT(11,0) 176 C atoms; CNT(12,0) 192 C atoms; CNT(13,0) 208 C atoms; CNT(14,0) 224 C atoms; and CNT(15,0) 240 C atoms. Then 24 water molecules (72 atoms) were inserted inside each CNT. Table 1 shows the DFT results in terms of the intermolecular energy (Einter), the optimized CNT diameter (d), the hydrogen bond length, the net Mulliken charge on water, and the intermediate pressure (P!

!!! !!!!

!

, in which Fi,j is the force

on the ith atom of adsorbate due to the jth atom of the substrate and A is the substrate area) exerted on the WNT by the CNT. In regards to evaluating the adsorption energy and the relative stability of the different CNTs, we performed vdW-DF by considering dispersion into the DFT calculations (Table 1,Table 2). It is also well known that this correction will mainly affect the total energy (increased total energy and bandgap) but that the electronic structure of the system will remain essentially intact. In addition, a previous comparison between LDA and GW approaches28 on graphene clearly showed that the band structures were similar except for an energy shift in GW with respect to LDA.

The trends presented in Table 1 are straightforward: (i) CNT(13,0) is the most stable structure, followed by CNT(12,0), (14,0), (15,0) and (11,0), whereas CNT(10,0) is unstable with positive energy adsorption, (ii) the hydrogen bond length increases from CNT(10,0) to (15,0), (iii) intermediate pressure peaks on the WNT by unstable CNT(10,0) (P!

!!! !!!!

!

, in which Fi,j is the

force on the ith atom due to the jth atom in the WNT and A is the WNT area), whereas it decreases for the most stable WNT@CNT(13,0), and (iv) the net charge is more important for the unstable case, i.e., CNT(10,0), and for the metastable case, i.e., CNT(11,0). ACS Paragon Plus 8 Environment

Page 9 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

To summarize the data presented in Table 1, the most stable WNT is CNT(13,0), in which the vdW pressure on the water tube is less important. Furthermore, such a vdW pressure (0.28 GPa) is consistent with improving intramolecular electrostatic attraction between O and H atoms to break hydrogen bonds in CNT(10,0). We validated our computational approach by directly comparing the calculated values for CNT(13,0) to an accurate experimental measurement, specifically a phase transition of water tube to ice NT in a CNT with a diameter of 10.5 Å reported by Agragwal et al.11 The WNTs’ net charge in Table 1 are negative, yielding to positively charged or p-doped CNTs. This result is in agreement with the electron withdrawing character of oxygen atoms, which usually leave CNTs positively charged. The magnitude of the vdW pressure for different CNTs along with the optimized CNT diameter support a more comprehensive description of WNT@CNT stability.

In brief words, the high vdW pressure breaks WNT hydrogen bonds (1.45 Å) and facilitates an overlap (donation) between oxygen atoms and CNTs, making the WNT more polarized. However, the inverse mechanism, in which the vdW pressure is smaller for CNT(12,0) and (13,0) than for unstable (11,0) and metastable (10,0), facilitates the intramolecular energy of the WNT to create hydrogen bonds (1.75 Å) and exhibit a stabilized solid-like state. Consequently, this mechanism would also be consistent for larger CNTs such as (14,0) and (15,0). We analyzed the phonon dispersion as well as electronic structure properties of a water tube within CNTs in the following section.

3.1.1. Electronic properties ACS Paragon Plus 9 Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

We calculated a small charge transfer from CNTs to WNTs, confirming that no covalent bonding exists between the two species. In order to clarify the influence(?) of WNTs on the electronic structure of CNTs, we present total density of states (DOS) of WNT@CNT compared with the DOS of a pristine CNT (Figure 2). First, the direct comparison of DOS for WNT@CNT with that for a pristine CNT shows that CNT(10,0) and (11,0) are more affected by the water tube when it is around the Fermi level than other CNTs. Intermolecular interactions of the water tube perturbed the electronic states of CNT(10,0) and (11,0); for example, electron-withdrawing oxygen atoms of water result in transfer of some electrons from the valence band of the CNT to the water tube. Hence, the DOS of the WNT@CNT system is slightly shifted below the Fermi energy of an isolated CNT as a reference. This Fermi energy shifting (Figure 1, panels d and e) proves p-doping of the CNT, confirming charge transfer from the CNT to the WNT by values shown in Table 1. The electronic properties of WNT@CNT(12,0), (13,0), (14,0), and (15,0) are also consistent with the results obtained for the intermolecular energy reported in Table 1, revealing that the formation of highly directional hydrogen bonds makes the WNT the most stabilized state.

To check and compare electronic properties of our model, we calculated band structures of the six systems with respect to a pure CNT, shown in Figure 3. A small but systematic bandgap closing in the CNT is observed for semiconductor CNTs such as CNT(10,0), (11,0), (13,0), and (14,0) and bandgap opening is observed in metallic CNTs such as CNT(12,0) and (15,0) (Figure 3). The latter results suggest that such an intermolecular interaction between water and the CNT improves the overlap and mixing between O (water) and C atoms (CNT) states, to nally(?) introduce a bandgap in the CNT through a local electronic perturbation by O atoms.29-31


Page 10 of 33

Page 11 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

To calculate phonon dispersion at points of high symmetry in the Brillouin zone, we performed lattice dynamics to consider phonon frequency modes. Phonon dispersion plots of WNTs within CNTs are depicted in Figure 4, where an isolated WNT is plotted for comparison. The calculated phonon of an isolated WNT is in good agreement with the experimental phonon band dispersion reported by Lyo and his co-workers.32 Figure 4 shows variation of the vibrational phonon modes for WNTs inside CNTs, giving additional details on the intermolecular interaction. Concerning stability, existence of negative phonon vibrations for both CNT(10,0) - the most unstable system - and CNT(11,0) for the metastable system support our DFT results presented in Table 1. Second, we observed that splitting occurs for low frequency (long wavelengths, acoustic modes) for almost all WNTs inside the CNTs. WNT can be considered a polar structure (see Figure S2), which affects its interfacial properties inside a CNT, creating an interior electrical field and broken symmetry. The long range of Coulomb interactions between the water tube (a square bonded polar system) with the CNT causes splitting of the long wavelength. The origin of the splitting is the interior electrical field created by the natural polarity existing in the WNT, in line with previous reports for ice systems.33

Third, there are some bandgaps or stop bands in the frequency spectra due to the non-unity mass ratio of water tubes.33 The dispersion of bands is strongly influenced by the intermolecular interaction of WNTs trapped inside CNTs, while the stop bands (bandgap) are being shifted. The phonon group velocity related to the slope of dispersion curves in the phonon spectrum gives information about atomic mobility. For instance, flat bands lead to slower group velocity and atomic mobility, while dispersed bands cause higher atomic mobility (orientational mobility). ACS Paragon Plus 11 Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

These plots inform the fact that dispersed bands lead to higher atomic mobility (orientational mobility), while flat bands indicate more localized carriers.30 Analysis of these phonon spectrum shows that the lower frequency or acoustic phonons for all WNT@CNTs include more dispersed bands or higher atomic mobility than what appears in the phonon spectrum for isolated WNTs. This finding suggests that confined WNTs within CNTs can modulate the electronic properties of the CNT due to the intermolecular interaction between them, improving the correlation between electrons and confirming our DFT results about bandgap closing in the system. However, larger frequency or optical phonons belong to both isolated WNTs and WNT@CNTs, considering the low phonon group velocity or atomic mobility. Therefore, these phonon bands support the notion that isolated and confined WNTs exhibit localized, slow electron mobility, or in other words behave solid-like, in close agreement with previous observations.2,21,24

Figure 5 is a schematic of the state diagram vs. intermolecular energy of WNT@CNTs and the interfacial pressure for CNTs of different diameters. The blue diamonds plot the intermolecular energy of different systems, showing WNT@CNT(10,0) is unstable, WNT@CNT(11,0) is metastable and WNT@CNT(13,0) is the most stable structure. In this regard, black triangles plot the vdW pressure exerted on a WNT in different CNTs, showing how the pressure varies with CNT diamater. The unstable system i.e, WNT@CNT(10,0) suffers from high pressure (~ 0.3 GPa), while for the most stable system i.e, WNT@CNT(13,0), the pressure is decreased to ~ 0.1 GPa.

3.2.

WNT@BNNT

Encouraged by our results described above and in order to evaluate the influence of a WNT confined within a BNNT, we constructed a system with square-tube water (WNT) inside a ACS Paragon Plus 12 Environment

Page 12 of 33

Page 13 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

BNNT. BNNTs are promising candidates for water desalination34 and biological nanochannels due to their wettability character.35 Hence, we compared interfacial properties of WNTs confined in CNTs or in BNNTs, which are of extreme interest from both scientific and technological standpoints. To compare results from BNNTs with those from CNTs, we chose the same supercell used to study CNTs for BNNTs. For the intermolecular energy of WNT@BNNTs, we used supercells in which the BNNT(10,0) contained 160 B,N atoms; BNNT(11,0) 176 B,N atoms; BNNT(12,0) 192 B,N atoms; CNT(13,0) 208 B,N atoms; BNNT(14,0) 224 B,N atoms; and BNNT(15,0) 240 B,N atoms, in which 24 water molecules (72 atoms) had been inserted inside each BNNT (Figure 6). The first interesting properties to compare between WNT@CNTs and WNT@BNNTs remain the intermolecular energy (Einter) and vdW pressure, as Table 2 reports the DFT structural parameters for WNT@BNNTs. The same trend observed for CNTs existed for BNNTs, as shown in Table 2. However, Einter and vdW pressure were significantly greater for WNT@BNNTs than for WNT@CNTs. One plausible explanation for different Einter and vdW in WNT@BNNTs takes into account that BNNTs have a different polarization character compared to CNTs because of B and N atoms. The partly ionic structure of BN reduces electrical conductivity, whereas the intermolecular interaction increases. The reduced electron-delocalization in BN causes a large bandgap relative to graphene. The resulting DOS and electronic band structure shown in Figures 7 and 8 support this suggestion. As listed in Table 2, WNT@BNNT(13,0) shows higher intermolecular energy and less vdW pressure, resulting in the most stable structure. This result supports the fact that BNNT(13,0) interacts weakly with the water tube, then converts it to an ice NT network.

3.2.1. Electronic properties ACS Paragon Plus 13 Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Finally, a series of DOS for WNT@BNNTs and pure BNNT was calculated and plotted in Figures 6, panel d and 7. A careful analysis of Figure 7 reveals that the electronic states of BNNTs are less affected than those of CNTs when the water tube is around the Fermi level. This effect can be effectively interpreted as more localized carriers or weaker interactions form between adjacent orbitals, originating from the insulator character of BNNTs. The inverse scenario works for WNT@BNNTs in regards to WNT@CNTs. As mentioned, the intermolecular interaction of water with CNTs affects the electronic states of the CNT due to higher carrier mobility, while the DOS of BNNTs remains intact. BNNT has altered water frequency phonon vibrations (Figure 9), confirming localized and strong interaction on water tubes.

To complete the first part of this analysis of electronic properties, the band structures of the WNT@BNNTs were compared with that of an isolated BNNT, as shown in Figure 8. Consistent with previous DOS diagrams in Figure 7, the Fermi level and dispersion of the bands more or less remain the same for the isolated BNNT and for the WNT@BNNTs.

Phonon dispersion plots of the WNT@BNNTs are depicted in Figure 9; isolated water is plotted for comparison. A careful analysis of Figure 9, which shows variation of the vibrational phonon modes for WNTs within BNNTs, gives additional details on the intermolecular interaction. From these plots, a number of general trends can be discerned. First, the number of flat bands increases in almost all WNT@BNNT systems, even for small frequencies to an extent. As previously mentioned for WNT@CNT systems, flat bands indicate more localized carriers, which cause strong intermolecular interaction between the water tube and BNNTs. These results are consistent with DFT intermolecular energies presented in Table 2. ACS Paragon Plus 14 Environment

Page 14 of 33

Page 15 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Second, the distribution of stop bands (no phonon wave propagation) differs with that of the isolated WNT, in which the stop bands are shifted up and become wider and denser for larger tubes. The phonon group velocity of flat bands related to the slope of the flat dispersion curves becomes smaller or zero when the branches are completely at something missing here, leading to localized intermolecular interactions between the WNT and the BNNT. These flat bands originate from the insulator feature of BNNTs and dipole-dipole interactions between the water tube as a polar system and the BNNT.

Figure 10 is a schematic of the state diagram vs. the intermolecular energy of the WNT@BNNTs and interfacial pressure for BNNTs of different diameters. The blue diamonds plot the intermolecular energy of different systems, showing WNT@BNNT(10,0) is unstable, WNT@BNNT(11,0) is metastable, and WNT@CNT(13,0) is the most stable structure. In this regard, black triangles give the vdW pressure exerted on the WNT in different BNNTs, showing how the pressure varies with the BNNT diameter. The unstable system i.e, WNT@BNNT(10,0) suffers from high pressure (~ 0.5 GPa) while for the most stable system i.e, WNT@BNNT(13,0) the pressure is decreased to (~ 0.1 GPa).

4.

CONCLUSION

In summary, our comprehensive theoretical computations revealed that considerable intermolecular interactions and vdW pressure exist between a water tube and a CNT/BNNT, and demonstrated that the phase behavior of the water tube captured within a CNT or BNNT is diameter-dependent. The state diagram of this system shows that decreasing vdW pressure helps to stabilize a square water NT, converting it to an ice NT. The vdW pressure is 0.3 (0.5) GPa for ACS Paragon Plus 15 Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

the most unstable system i.e., CNT/BNNT(10,0), which decreases to 0.09 (0.12) GPa for the most stable system i.e., CNT/BNNT(13,0). The interfacial properties of WNT@BNNT are stronger than those of CNTs.

The phonon calculation of different WNT@CNTs(BNNTs) compared with an isolated square-tube water confirmed that (10,0) CNT/BNNT is unstable, while (11,0) is metastable with negative phonon bands. However, the remaining structures are stable and CNT/BNNT(13,0) is the most stable system with a diameter of 10.5 Å, in line with experimental observation.11 Our results suggest that the water tube transition to ice NT is strongly diameter-dependent. This work suggests that the intermolecular interaction between a WNT and a hydrophobic CNT/BNNT could lead to design and fabrication of nanometer-scale capillaries.

Supporting Information Available The supporting information (SI.pdf) file is available free of charge.

Acknowledgements The supercomputer machines utilized in this work were supported in part by the NIH award NCRR S10RR02950 and an IBM Shared Rice University Research (SUR) grant.

Competing Financial Interest The authors declare no competing financial interest.


Page 16 of 33

Page 17 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

References: [1] Thilgen, C. A Single Water Molecule Trapped Inside Hydrophobic C60. Angew. Chem. Int. Ed. 2012, 51, 587-589.

[2] Chakraborty, S.; Kumar, H.; Dasgupta, C. and Maiti, P. K. Confined Water: Structure, Dynamics, and Thermodynamics. Acc. Chem. Res. 2017, 50 (9), 2139-2146.

[3] Ludwig, R. Water: From Clusters to the Bulk. Angew. Chem. Int. Ed. 2001, 40, 1808-1827.

[4] Yoshizawa, M.; Kusukawa, T.; Kawano, M.; Ohhara, T.; Tanaka, I.; Kurihara, K.; Niimura, N.; Fujita, M. Endohedral Clusterization of Ten Water Molecules into a Molecular Ice within the Hydrophobic Pocket of a Self-Assembled Cage. J. Am. Chem. Soc. 2005, 127, 2798-2799.

[5] Kurotobi, K.; Murata, Y. A Single Molecule of Water Encapsulated in Fullerene C60. Science 2011, 333, 613-616.

[6] Noh, T. H.; Heo, E.; Park, K. H.; Jung, O.-S. Motion of an Isolated Water Molecule within a Flexible Coordination Cage: Structural Properties and Catalytic Effects of Ionic Palladium(II) Complexes. J. Am. Chem. Soc. 2011, 133, 1236-1239.

[7] Shultz, M. J.; Vu, T.H.; Meyer, B. and Bisson, P. Water: A Responsive Small Molecule. ACS Paragon Plus 17 Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Acc. Chem. Res. 2012, 45, 1522.

[8] Takahashi, K.; Kramer, Z. C.; Vaida, V.; Skodje, R. T. Vibrational Overtone Induced Elimi-

nation Reactions within Hydrogen-Bonded Molecular Clusters: the Dynamics of Water Catalyzed Reactions in CH2F OH3(H2O)n. Phys. Chem. Chem. Phys. 2007, 9, 38643871.

[9] Vaida, V.; Kjaergaard, H. G.; Hintze, P. E.; Donaldson, D. J. Donaldson, D. J. Photolysis of Sulfuric Acid Vapor by Visible Solar Radiation.Science 2003, 299, 15661568.


Page 18 of 33

Page 19 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

[10] Sakhavand, N.; Muthuramalingam, P. and Shahsavari, R. Toughness Governs the Rupture of the Interfacial H-Bond Assemblies at a Critical Length Scale in Hybrid Materials. Langmuir 2013, 29 (25), 8154-8163.

[11] Agrawal, K. V.; Shimizu, S.; Drahushuk, L. W.; Kilcoyne, D. and Strano, M. S. Observation of Extreme Phase Transition Temperatures of Water Confined Inside Isolated Carbon Nanotubes. Nat. Nanotech 2017, 12, 267-273.

[12] Chen, J.; Zen, A.; Brandenburg, J. G.; Alf`e, D. and Michaelides, A. Evidence for Stable Square Ice from Quantum Monte Carlo, Phys. Rev. B 2016 94, 220102(R).

[13] Chen, J.; Schusteritsch, G.; Pickard, C. J.; Salzmann, C. G. and Michaelides, A. DoubleLayer Ice from First Principles. Phys. Rev. B 2017, 95, 094121.

[14] Algara-Siller, G.; Lehtinen, O.; Wang, F.C.; Nair, R. R.; Kaiser, U.; Wu, H. A.; Geim, A. K. and Grigorieva, I. V. Square Ice in Graphene Nanocapillaries. Nature 2015 519, 443.

[15] Shephard, J. J.; Ling, S.; Sosso, G. C.; Michaelides, A.; Slater, B. and Salzmann, C. G. Is High-Density Amorphous Ice Simply a Derailed State along the 2 Ice I to Ice IV Pathway? J. Phys. Chem. Lett. 2017, 8 (7), 1645-1650.

[16] Sahu, P. and Musharaf Ali, SK. Curious Characteristics of Polar and Nonpolar Molecules Confined within Carbon Nanotubes (CNT) of Varied Diameter: Insights from Molecular Dynamics Simulation. J. Chem. Eng. Data 2017, 62 (8), 2307-2315.

[17] Al-Hamdani, Y. S.; Alfe, D. and Michaelides, A. How Strongly Do Hydrogen and Water Molecules Stick to Carbon Nanomaterials? J. Chem. Phys 2017, 146, 094701.

[18] Soler, J. M.; Artacho, E.; Gale, J.D.; Garc a, A.; Junquera, J.; Ordejon, P.; Sanchez-Portal, D. The SIESTA Method for Ab Initio Order-N Materials Simulation. J. Phys.: Condens. Matter 2002, 14, 2745-2779.


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

[19] Dion, M.; Rydberg, H.; Schroder, E.; Langreth, D. C.; Lundqvist, B. I. Van der Waals Density Functional for General Geometries. Phys. Rev. Lett. 2004, 92, 246401.

[20] Roman-Perez, G.; Soler, J. M. E cient Implementation of a Van der Waals Density Functional: Application to Double-Wall Carbon Nanotubes. Phys. Rev. Lett. 2009, 103, 096102.

[21] Mahesh Kumar, R.; Elango, M.; Parthasarathi, R. and Subramanian, V. Density Functional Theory Studies on Ice Nanotubes. J. Phys. Chem. A 2011, 115, 12841-12851.

[22] Shayeganfar, F.; Vasu, K. S.; Nair, R. R.; Peeters, F. M. and Neek-Amal, M. Monolayer Alkali and Transition-Metal Monoxides: MgO, CaO, MnO and NiO.Phys. Rev. B 2017, 95, 144109.

[23] Mirabbaszadeh, K.; Torkashvand, Z.; Shayeganfar, F. Electro/Mechanical Mutable Properties of Black Phosphorene by Electric Field and Strain Engineering. Materials Research Express 2018, 5 (6), 066307.

[24] Jay Mashl, R.; Joseph, S.; Aluru, N. R. and Jakobsson, E. Anomalously Immobilized Water: A New Water Phase Induced by Confinement in Nanotubes. Nano Lett., 2003, 3 (5), 589-592.

[25] Liu, Y-C.; Shen, J-W.; Gubbins, K. E.; Moore, J. D.; Wu, T. and Wang. Q. Di usion Dynamics of Water Controlled by Topology of Potential Energy Surface Inside Carbon Nanotubes. Phys. Rev. B 2008, 77, 125438.

[26] Liu, Y.; Wang, Q.; Zhang, L. and Wu, T. Dynamics and Density Pro le of Water in Nanotubes as One-Dimensional Fluid. Langmuir 2005, 21, 12025-12030.

[27] Hernandez-Rojas, J.; Calvo, F.; Breton, J. and Gomez Llorente, J. M. Confinement Effects ACS Paragon Plus 20 Environment

Page 20 of 33

Page 21 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

on Water Clusters Inside Carbon Nanotubes. J. Phys. Chem. C 2012, 116, 17019-17028.

[28] Kharche, N.; Nayak, S. K. Quasiparticle band gap engineering of graphene and graphone on hexagonal boron nitride substrate. Nano Lett. 2011, 11, 5274-5278.

[29] Shayeganfar, F. and Rochefort, A. Electronic Properties of Self-Assembled Trimesic Acid Monolayer on Graphene. Langmuir 2014, 30, 9707-9716.

[30] Shayeganfar, F. and Shahsavari, R. Electronic and Pseudomagnetic Properties of Hybrid Carbon/Boron-Nitride Nanomaterials via Ab-Initio Calculations and Elasticity Theory. Carbon 2016, 99, 523-532.

[31] Shayeganfar, F.; Beheshtiyan, J.; and Shahsavari, R. Electro- and Opto-Mutable Properties of MgO Nanoclusters Adsorbed on Mono- and Double-Layer Graphene. Nanoscale 2017, 9 (12), 4205-4218.

[32] Renker, B. Physics and Chemistry of Ice (University of Toronto Press, Toronto, p. 82,1973).

[33] McGaughey, A. J. H.; Husseinz, M. I.; Landry, E. S.; Kaviany, M. and Hulbert, G. M. Phonon Band Structure and Thermal Transport Correlation in a Layered Diatomic Crystal. Phys. Rev. B 2006, 74, 104304.

[34] Hilder, T. A.; Gordon, D.; Chung, S. H. Boron Nitride Nanotubes Selectively Permeable to Cations or Anions.Small 2009, 5, 2183-2190.

[35] Curtis, A. S. G.; Forrester, J. V.; Mcinnes, C.; Lawrie, F. Adhesion of Cells to Polystyrene Surfaces.Eur. J. Cell Biol. 1983, 97, 1500-1506.


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 33

TABLE 1. Structural analysis of optimized confined WNT@CNTs. The numbers inside the parentheses were calculated by vdW-DF. W NT @CNT E inter(eV) Opt diameter (Å)

(10,0)

(11,0)

5.63(3.64) -1.19(-4.45)

(12,0)

(13,0)

(14,0)

(15,0)

-2.93(-6.05) -5.83(-12.02) -2.01(-5.61) -1.68(-5.38)

8

8.9

9.5

10.5

11

12.2

1.45 -0.28

1.6 -0.35

1.67 -0.14

1.75 -0.23

1.78 -0.18

1.82 -0.07

Pressure (GPa)

0.13(0.28)

0.11(0.23)

0.09(0.20)

0.06(0.09)

0.07(0.11)

0.05(0.08)

Bandgap (eV)

0.61

1.02

0.1

0.53

0.74

0.09

Pristine gap (eV)

0.82

1.11

0

0.53

0.74

0

Hydrogen bond (Å) WNT net charge (|e|)

TABLE 2. Structural analysis of optimized confined WNT@BNNTs. The numbers inside parentheses were calculated by vdW-DF. W NT @BNNT

(10,0)

(11,0)

(12,0)

(13,0)

(14,0)

(15,0)

E inter (eV)

11.28(8.32)

0.68(-3.7)

-3.83(-9.36)

-4.6(-10.48)

-3.38(-8.77)

-2.13(-4.8)

Opt diameter (Å)

8.26

8.76

9.5

10.52

11.21

12.2

Hydrogen bond (Å)

1.42

1.48

1.62

1.72

1.75

1.75

Elecrical dipole (px)

0.09

0.18

0.08

0.04

0.2

0.42

TMA net charge (|e|) Pressure (GPa) Bandgap (eV)

-0.1 -0.07 -0.11 -0.15 -0.09 -0.08 0.351(0.511) 0.243(0.396) 0.136(0.201) 0.081(0.125) 0.101(0.184) 0.077(0.11) 3.94

4.01

22 ACS Paragon Plus Environment

4.04

4.21

4.25

4.3

Page 23 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Page 24 of 33

Page 25 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Page 26 of 33

Page 27 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Page 28 of 33

Page 29 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Page 30 of 33

Page 31 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Page 32 of 33

Page 33 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir


First Principles Study of Water Nanotubes Captured Inside Carbon

Recommend Documents