Water motions and orientations in nanotubes of various dimensions. (Simulations done by Jay Mashl, University of Illinois/NCSA/Beckman Institute Computational Biology/Nanoscience Group)

System Setup Carbon nanotube (fixed) 8 sizes of nanotubes ranging 5.4  -  16.3 Å dia., armchair (5,5) -  (12,12). Length ~40 Å. Bilayer mimetic (hcp CH 2 's, fixed) SPC/E water (T = 300 K) Electrostatics: PME Nose-Hoover coupling Pressure piston (P z = 1 bar) 2q H = -q O = e Runs of ~2 ns each using GROMACS ( See www. gromacs.org) Simulations done on NCSA IA32 and IA64 Linux superclusters

Relative Diffusion coefficients Water in Nanotube vs. bulk(=1) D z (tube) / D z (bulk) Nanotube diameter (Å) "Critical" slowing

Diffusion of Water: Two Dimensions Time lag (ps) Mean-square displacement (Å 2 )

Water Dipole Autocorrelation Time lag (ps) / p 2

Water Velocity Autocorrelation (nm / ps) 2 Time lag (ps)

Snapshots of Water Configurations (6,6) (12,12) (9,9) Critical size for order Bulk water properties not yet achieved T = 300 K (water), fixed tube & slab 2-D hydrogen bonding 1-D hydrogen bonding Single file

Summary Water motions and orientations are qualitatively modified by confinement in nanotubes. Depressed transverse diffusion Modified dipole autocorrelations Collective oscillations in narrow tubes Point for discussion and future work: Could anomolous solvent behavior at critical dimensions be modulated by an external stimulus (change in electric field for example) to serve as basis for switching behavior? Anomalous behavior (extreme ordering and reduced mobility) is seen at critical nanotube dimensions