As far as I remember the crystallization or  quasi-polymerization of water molecules were  observed in ultra microscopic capillary tube few decade ago. The idea was the short range capillary forces acting between the tube walls and water molecules. But it was nothing to do with the structure of the bulk ice.  I don' t fully agree with the comment made by Dr. Pant since the tubes were involved was quartz!!!  Of course hydrophobic forces may enhance this chain type molecular ordering.

Dr. Tarik glad to see your answer, but i was a bit interested to know does hydrophobic forces also play some roles in such ice formations or is it just the capillary forces you mentioned ? What size are you pointing to by saying ultra microscopic capillary because for even large CNT ordering solid like is seen at the interface (inside) while bulk water is seen at the center of the channel.

What I was talking about  it was   the one dimensional ordered structure of water molecules  along the capillary tube (like wiskers having helical screw dislocation like structure)  not the formation crystalline lattice having translational  symmetry. The size was few fraction of a micrometer similar to the fluid carrying channels in plants and other living bodies . Your comment shows that the short range interaction  taking place at the interface is primary factor for the observed ordering phenomena, where carbon tube surface acts as a temp plate for the ordering!! Far distance away from the interface,  the gravitational forces, and molecular Brownian  motions may be play negative role for the formation ordered structure?  

Yes sir, it seems like that only. At the interface enhanced ordering is seen and on increasing the diameter of the CNT actually at the interface you can see a complete ordered nanotube made up of water. While if you go to the center water is back to its original form i.e. disordered !!

We should not forget there is always stuck boundary layer next to the interface where no convection takes place but if you go further away there is always micro convection or vortex motion driven by the local temperature fluctuations occurs in the direction of earth gravitation force due to the anomalous temperature dependence of water specific density. Macroscopic scale anomalous behavior of  water may be observed by the existence of the sea  water circulations between the equator and the polar  regions:  the warm current  goes deep under and cold current  just below the ocean's  surface.

I do believe that those vortex motions  or micro circulations  away from the interface dominates the whole ordering process, and results instability. I think it would be very interesting to do this experiment in absence of gravitation in space?

The more puzzling question for me is the opposite - could this ordering near the interface lead to a suppression of ice formation in the bulk, which would have large practical implications?

Dear Dr. Sobisch; if you look at the problem  from the thermodynamics point of view  the crystallization of water is exzothermic reaction. Therefore,    in order to the nucleus of the crystallite   to grow  the  rejected  heat of crystallization should be taken away from the interface between the solid  phase and liquid phase. This rejected heat will be taken out from the walls of the carbon tube to the surrounding environment. Carbon nano tube walls as well as the ordered or crystallize water layer attached to  the tube walls act as a thermal  barrier  for the heat transfer, which is   proportional with the thickness . Similarly,  the crystallization of water in the bulk involves homogeneous nucleation, which requires very much supercooling than the heterogenous nucleation takes place at the tube walls. Even for the standard mold casting problems,  the solidification of the liquid phase  at the central region by homogenous nucleation needs few hundred degrees of supercooling for cooper alloys having a Tm in the range of 1050 C.

I am looking very emphatic for  Dr. Sobisch  comments or guess on the negative role of ordered phase of water formed at the carbon nano tube walls. I also think that ordered water molecular structure might have highly anisotropic heat conductivity similar to the Pyrolytic carbon such as:  very low heat conduction normal to the sheets, which are made by water molecules by non-covalent bonds ( an electron is not localized by  resonating between to moleculus-) and otherwise the sheets are put together by van der Walls forces., and then there is a  high heat conductivity along the direction of tube axis but not in the radial direction.!!! 

Water is adsorbed as clusters. You should take a look at water adsorption modelling inside nanospaces. Dr. Ohba is an expert on this field, for example:

Initial filling mechanism of predominant water adsorption on hydrophobic slit-shaped carbon nanopores

or

Cluster-Growth-Induced Water Adsorption in Hydrophobic Carbon Nanopores

or

Structural mechanism of water adsorption in hydrophobic micropores from in situ small angle X-ray scattering

http://iopscience.iop.org/1742-6596/177/1/012001/pdf/1742-6596_177_1_012001.pdf

http://pubs.acs.org/doi/abs/10.1021/jp048323v

Hi Shashank,

must admit I've got not so much experience with nanotube ice, but heavily dealt with nanotube water for sure (however sound the concept of water and ice could be at such scale).

As to the latter, strong confining conditions induced by both nanopore size and hydrophobicity produce a peculiar, highly-ordered molecular configuration (http://www.physics.iisc.ernet.in/~maiti/download/maiti_biswa_ACS_nano.pdf) characterized by heavy modifications in the orientational relaxation of the system (http://arxiv.org/abs/1108.3908). Thus, yes, the experimentally undeniable polar wetting of a nonpolar channel is generated by the hydrophobic constraints as well - and the counterintuitive wetting should be preserved, among other mechanisms, via rotational-entropy gain, which would balance the lack in bond energy.

Dear Alexandra,

Even i am new to this area and it seems like the ordering (ice like structure) has nothing to do with the confinement, hydrogen bonding. Its just  the potential in my case (LJ) playing role. I think i got some answers for this type of ordered coarse grained water nanotube being formed inside the nanotube itself. 

I am just curious to know if you have dealt with the dynamics of water inside CNT particularly diffusion calculations ? If yes i will be really glad if we can discuss diffusion calculations inside CNT.

Thanks

Shashank

Dear Dr. Pant, what we are talking  about the crystalline like transformation of water molecules  at immediate vicinity of  carbon nanotubes ( or fine capillary tubes made by ordinary glass or quartz)   or nano-size cavities,  well above the freezing point of water under the atmospheric pressure.  First of all; LJ doesn't play any role when one comes to water molecules whether they are in liquid  or crystalline forms.  Where,  what is called hydrogenic (or hydroponic) bond  plays  the major role, which has electric dipole in character, and it can only interact  with the electric dipole, and the interaction force varies with the distance as  1/r4. This force may be given by   M'.del [M.deldel 1/r]  , which has  much longer range then  the 6-12  LJ   potential that results   r-7 and  r-13 interaction forces. Where M', M      are  electrostatic dipole moments  associated to the wall or  substrate  and the   water molecules, respectively.  Also there is a non-vanishing torque acting, which varies as  1/r3  that generates long range rotation!  Regards.

Dear Dr. Tarik

Glad to hear from you. Well i should tell you in brief about my system. Its CNT and instead of water i have used a coarse grained water model (which i published last year in JCP). In this model i have not considered electrostatic terms still it is able to express water and water-like anomalies (translational order, excess entropy, diffusion entropy and density anomaly). So further i am using this model to study it inside CNT where CNT is modeled by LJ potential as proposed long back by Hummer et al 2001 Nature. So by potential i meant that long ordering is present around the CNT (surface) is because there coarse grained water can experience attractive interactions by CNT so they prefer to be ordered perfectly around the CNT. This type of ordering is present outside and inside the CNT nanopore.

Thanks

Shashank

Dear Dr. Pant; Firstly, Please,   I would like to see your   JCP  paper, Because  I am especially interesting in knowing  how did you formulate the diffusional entropies  associated with water in liquid and crystalline forms, which require knowledge on the variations of the Gibbs free energy with respect to temperature under constant pressure or the Helmholtz free energy change with respect to temperature at constant volume. Up to now I haven't seen single theoretician dealing with the calculation of the Gibbs free energy and its change with temperature under the constant  positive or negative hydrostatic stress. As I remember the whole statistical thermodynamics as well as the mechanics rely on the Helmholtz free energy as clearly stated in the classical book by Professors Fowler and Guggenheim.

Your selection of LJ potential is arbitrary like the selections of all those ad hoc potentials themselves.  

Hi again,

yes, I did (diffusion measurements, I mean) - even though investigation is still in progress.

This discussion is interesting indeed.  I did not catch what you mean for attractive interactions by the CNT, which is a prototype of nonpolar pore. In my mind (I'm an experimentalist, not a theoretician) the polar wetting of  a small-diameter CNT  is an 'entropic trick'  triggered by ultra-confinement rather than the result of an attractive wall potential.

Also, the behavior of simulated water (as well as other HB molecular systems) is hopelessly model-dependent, and even refined electrostatic configurations are not able to fully reproduce many important experimental trends (such as neutron scattering data for zero-point proton kinetic energies) in the bulk fluid (http://www.ncbi.nlm.nih.gov/pubmed/22010722), not to mention the low-dimensional one.  A non-electrostatic model is hardly comparable to real water.

Finally, how sure can we be that traditional water anomalies get preserved (rather than shifted, or even suppressed) under strong confinement? No safe link between bulk- and confined-liquid thermodynamics has been retrieved yet.

Hi again , 

first of all i am a theoretician. You should look at a very recent theoretical finding published in Angewchem which talks about the water inside the CNTs published in 2014. Well i am afraid to differ from your confinement comment. I think the ordered water has nothing to do with the confinement issue because i have done some work where i found that even for large diameters of CNTs there is ordered but only at the interface and bulk water at the center of the CNT. So, i don't think the orderness is due to confinement.

Shashank

Hi,

your result is completely consistent with experiments. Yet, water at interfaces is actually 2D confined. Many nice experiments can be performed to highlight that  - for example, by 1D NMR, which signals a several-ppm chemical shift of interface water with respect to the the main bulk line, or 2D diffusion NMR, which reports peculiar long-time-limit coefficients further depending on tortuosity of the medium.

Gonna read your article soon - that's really a nice topic.

Sorry, several tenths of ppm.   :)