Itsuo Hanasaki

Flow structure of water in carbon nanotubes: Poiseuille type or plug-like?

We have conducted molecular dynamics simulations of water flow in carbon nanotubes (CNTs) for (6,6) to (20,20) CNTs at a streaming velocity of 100 ms. The fluidized piston model (FPM) and the ice piston model (IPM) are employed to drive flow through the CNTs. The results show that the single-file water flow inside (6,6) CNT has a convex upward streaming velocity profile, whereas the velocity profiles in (10,10) to (20,20) CNTs are flat except near the tube wall. The flow structure of cylindrical water in the (8,8) CNT is intermediate between that for the (6,6) CNT and the larger CNTs. The flow parameters are found not to exhibit any dependence on streaming velocity at up to 300 ms in the (12,12) CNT. The hydrogen bond lifetimes of water flowing in CNTs tend to be longer than for the corresponding equilibrium states, and nonzero flow does not reduce the microscopic structure or structural robustness (hydrogen bond lifetime). Although the atomic density profile varies with tube diameter, reflecting the change in static microscopic structure of flow from single file to cylindrical, tube diameter does not induce a clear transition in streaming velocity, temperature, or hydrogen bond lifetime over this diameter range. The results suggest that water flow in CNTs of this size is more pluglike than Poiseuille type, although the flow structure does not strictly accord with either definition.

Hydrogen bond dynamics and microscopic structure of confined water inside carbon nanotubes

We have investigated the density and temperature dependences of microscopic structure and hydrogen bond dynamics of water inside carbon nanotubes (CNTs) using molecular dynamics simulation. The CNTs are treated as rigid, and smoothly truncated extended simple point charge water model is adopted. The results show that as the overall density increases, the atomic density profiles of water inside CNTs become sharper, the peaks shift closer to the wall, and a new peak of hydrogen atomic density appears between the first (outermost) and second layer. The intermittent hydrogen bond correlation function C(HB)(t) of water inside CNTs decays slower than that of bulk water, and the rate of decay decreases as the tube diameter decreases. C(HB)(t) clearly decays more slowly for the first layer of water than for other regions inside CNTs. The C(HB)(t) of the interlayer hydrogen bonds decays faster than those of the other regions and even faster than that of the bulk water. On the other hand, the hydrogen bond lifetimes of the first layer are shorter than those of the inner layer(s). Interlayer hydrogen bond lifetimes are clearly shorter than those of the constituent layers. As a whole, the hydrogen bond lifetimes of water inside CNTs are shorter than those of bulk water, while the relaxation of C(HB)(t) is slower for the confined water than for bulk water. In other words, hydrogen bonds of water inside CNTs break more easily than those of bulk water, but the water molecules remain in each others vicinity and can easily reform the bonds.

Water flow through carbon nanotube junctions as molecular convergent nozzles

Molecular dynamics (MD) simulations are conducted for water flow through carbon nanotube (CNT) junctions as molecular nozzles. The fluidized piston model (FPM) is employed to drive the inlet flow at streaming velocities of 25 and 50 m s−1. Water flow through the CNT junctions is found to undergo an increase in streaming velocity, a decrease in pressure, and an increase in temperature. Although the difference of the upstream velocities does not generally lead to an appreciable density difference in the downstream CNT, the higher streaming velocity causes the upstream density to increase. The streaming velocity remains almost constant in the upstream CNT, but increases dramatically in the junction region. The ratio of downstream to upstream streaming velocities increases with the ratio of upstream to downstream cross section. A higher inlet velocity results in larger acceleration, which is generally more noticeable at larger cross-sectional ratios, and less prominent in junctions with smaller cross-sectional ratios. The cross-sectional ratio calculated from the internal radii of the CNTs based on the oxygen atomic density profile of water is closer to the ratio of downstream to upstream streaming velocities than the cross-sectional ratio calculated from the radii given by the carbon atomic centres.

Structure and stability of water chain in a carbon nanotube

Water molecules form a single-file chain structure in a (6, 6) carbon nanotube (CNT), and this stability is different from that of water molecules confined in CNTs with larger diameters, let alone the bulk. Using the molecular dynamics (MD) method and quantum mechanical (QM) calculations, we investigate the characteristics in the context of density dependence of the collective structure and hydrogen bond behavior. The results obtained from MD show that high water density leads to substantially longer hydrogen bond lifetimes. On the other hand, the hydrogen bond lifetime does not noticeably decrease with decreasing density but remains roughly the same when the density is lower than a certain critical value. The mean molecular orientation angle of the water molecule, defined by the angle that comprises the water dipole moment and the CNT axis, is smaller for higher densities, and asymptotically approaches 33° on the low density side. Such an asymptotic nature of the structure and stability stems from non-uniform distribution of water molecules. The mean orientation angle obtained from QM calculations using density functional theory coincides with the MD result. QM analysis also suggests that the charge distribution of water in the CNT originates from the molecular configuration due to spatial confinement rather than strong electronic interaction between water and the CNT.

Molecular dynamics study of Ar flow and He flow inside carbon nanotube junction as a molecular nozzle and diffuser

Abstract A carbon nanotube junction consists of two connected nanotubes with different diameters. It has been extensively investigated as a molecular electronic device since carbon nanotubes can be metallic and semiconductive, depending on their structure. However, a carbon nanotube junction can also be viewed as a nanoscale nozzle andv diffuser. Here, we focus on the nanotube junction from the perspective of an intersection between machine, material and device. We have conducted a molecular dynamics simulation of the molecular flow inside a modeled (12,12)–(8,8) nanotube junction. A strong gravitational field and a periodic boundary condition are applied in the flow direction. We investigated dense-Ar flows and dense-He flows while controlling the temperature of the nanotube junction. The results show that Ar atoms tend to be near to the wall and the density of the Ar is higher in the wide (12,12) nanotube than in the narrow (8,8) nanotube, while it is lower in the wide tube when no flow occurs. The streaming velocities of both the Ar and the He are higher in the narrow nanotube than in the wide nanotube, but the velocity of the Ar is higher than the velocity of the He and the temperature of the flowing Ar is higher than the temperature of the He when the same magnitude of gravitational field is applied.

Single-molecule measurements and dynamical simulations of protein molecules near silicon substrates

Interactions between protein molecules and inorganic substrates were studied both experimentally and numerically to obtain fundamental insight into the assembly of biomacromolecules for engineering applications. We experimentally traced individual fluorescent-labelled lysozyme (F-lysozyme) molecules, diffusing in the vicinity of interfaces between a protein solution and oxidized Si(1?0?0) and glass plates. The results indicate that diffusion coefficients of F-lysozyme molecules on both substrates are more than three orders of magnitude smaller than those in a bulk solution. The molecular dynamics simulations reveal a drastically diminished diffusion coefficient of lysozyme on the substrates of pure Si(1?1?1) and oxidized Si(1?0?0) with a hydroxy-terminated surface compared with that in bulk solution due to molecular adsorption behaviour on the substrate, which is in good agreement with experimental results. Furthermore, full atomistic description of the behaviour provides detailed information of deformation due to the adsorption process. Lysozyme on pure Si(1?1?1) undergoes substantial deformation whereas that on oxidized Si(1?0?0) does not, which indicates the importance of substrate surface condition to preserve the structure, i.e. functionality of adsorbed biomolecules.

Fluidized piston model for molecular dynamics simulations of hydrodynamic flow

We have developed a new method, the fluidized piston model (FPM), to drive dense fluid flow in molecular dynamics (MD) simulations. Fluid in the upstream region acts as a fluidized piston (FP) that continuously presses and supplies the downstream fluid. The FPM can treat dense and polyatomic fluid flow driven by a pressure gradient efficiently with the prescribed inlet streaming velocity at high speed, which is usually used for hydrodynamic MD studies. Furthermore, the temperature inside the sampling region is kept constant without direct control. We apply the FPM to the 100 m s−1 water flow inside (20, 20), (12, 12), (10, 10), (8, 8) and (6, 6) carbon nanotubes (CNTs). The results show that the appropriate density in the sampling region is adaptively obtained during the simulation and the desired streaming velocity is obtained except inside the (6,6) CNT. The deviation of the obtained streaming velocity from the desired value inside the (6,6) CNT is related to the current technical limit of the temperature control method applied in the upstream FP region and the highly discrete nature of the single-file water flow.

The antigen–antibody unbinding process through steered molecular dynamics of a complex of an Fv fragment and lysozyme

We have investigated the antigen–antibody unbinding process using steered molecular dynamics (SMD) simulations. We focus on a complex system consisting of an Fv fragment of an antibody molecule and a lysozyme as an antigen molecule. The Fv fragment consists of a VL and VH chain. The results show that the VH chain is unbound earlier than the VL chain, which is confirmed by the ensemble average of the distance profile obtained from 40 unbinding trajectories. The use of lysozyme as an antigen molecule instead of a small hapten molecule reveals the fact that the induced fit, estimated by the deformation accompanying the unbinding process, is more noticeable for the antigen molecule than for the antibody molecule. The SMD also reveals the non-Gaussian distribution of maximum force necessary for the unbinding process.

Coarse-grained picture of Brownian motion in water: Role of size and interaction distance range on the nature of randomness

The Brownian motion of a particle in a fluid is often described by the linear Langevin equation, in which it is assumed that the mass of the particle is sufficiently large compared to the surrounding fluid molecules. This assumption leads to a diffusion coefficient that is independent of the particle mass. The Stokes-Einstein equation indicates that the diffusion coefficient depends solely on the particle size, but the concept of size can be ambiguous when close to the molecular scale. We first examine the Brownian motion of simple model particles based on short-range interactions in water by the molecular dynamics method and show that the diffusion coefficient can vary with mass when this mass is comparable to that of the solvent molecules, and that this effect is evident when the solute particle size is sufficiently small. We then examine the properties of a water molecule considered as a solute in the bulk solvent consisting of the remainder of the water. A comparison with simple solute models is used to clarify the role of force fields. The long-range Coulomb interaction between water molecules is found to lead to a Gaussian force distribution in spite of a mass ratio and nominal size ratio of unity, such that solutes with short-range interactions exhibit non-Gaussian force distribution. Thus, the range of the interaction distance determines the effective size even if it does not represent the volume excluded by the repulsive force field.

Suppressing the coffee-ring effect of colloidal droplets by dispersed cellulose nanofibers

Graphical Abstract We report that the addition of a small amount of cellulose nanofibers (CNFs) into an aqueous dispersion of colloidal particles suppresses the coffee-ring effect when the dispersion dries on a solid substrate, as revealed by the computational analysis of experimental time-series images and by particle image velocimetry. The addition of CNFs is much more effective than the increase of colloidal particle concentration at the same weight percentage; it is also more environment friendly than the use of typical molecular surfactants. This finding is promising for the fabrication of metamaterials from colloidal dispersions and for ink printing in electronics, where CNFs can also serve as a substrate for flexible devices.