Clarifying the microscopic structure and movement of water
While being so near to us, water is the most puzzling substance in our life. After 40 years of research works on the mysterious movement of water with the help of computer simulation and the advance of analysis skill, have we actually found a new story that can tell us more about water?
It has been four decades since A. Rahman and F. Stillinger run molecular dynamics simulation for the first time , and the first subject they chose for that simulation was none other but water. We all know that water is the most essential liquid for life. On the other hand, despite of having a very simple molecule structure, water does not share many properties with other liquid. Within this four decades, a lot of discoveries on water have stunned us, and here we would like to introduce some of properties that have been discovered by computer simulation.
I. Hydrogen-bonding network, cooperative motion, and anomaly at low-temperature
A hypothesis that says that water molecules in liquid phase form a three-dimensional network through hydrogen-bonding was proposed about 60 years ago . But it was computer simulation that can confidently explain the structure and the dynamics of that network. Even at present, it is very hard to observe in detail the hydrogen bond structure through experimental work.
In ice phase, a water molecule forms hydrogen-bond with four neighboring water molecules. Even in liquid phase, at ambient temperature, a water molecule forms in average four weak hydrogen-bond with the surrounding water molecules. This network dynamically change with each hydrogen bond has a lifetime of no more than several pico seconds (equal to one per trillion second). When an exothermic chemical reaction occurs in liquid water, the released heat will be rapidly dissipated through the dynamic change of hydrogen-bond network, thus reversed reaction become less likely and a chemical reaction in aqueous solution is generally fast. This is also an important characteristic that enable biopolymer to exhibit its unique properties inside water .
At low temperature, as the lifetime of hydrogen-bond become longer, the tetrahedron molecular arrangement become preferred and the network structure become geometrically restricted. The uniquely ordered molecular arrangement of supercooled water, the various anomalies of water at low temperature (such as swelling at temperature lower than 4oC), are all considered to be caused by that restriction . Later, Mishima and co-workers reported the existence of two types of amorphous structure of ice: high-density amorphous ice under high pressure and low-density amorphous ice under ambient pressure . The best part is, simulation study also can shows two types of liquid state that correspond to two types of amorphous state by evading the crystallization process and put water in supercooled state . In this way, the macroscopic observation of the existence of two types of liquid phase (amorphous phase) is likely to be true. But the principle questions from the microscopic view, such as what is the difference in molecular arrangement between those two, what kind of structure exists between the interface between those two, why does the surface energy is high, are not yet clarified. Liquid-liquid transition was also observed for particular substances other than water, however none of them have yet sufficient explanation from molecular scale.
II. Solid water: three, two, and one dimensional
Water become ice upon cooling and ice become water upon melting. Even for such long-time and widely known phenomena, it is very hard for the experimental work to grasp the origin of such phenomena. In simulation, we make an ice nucleus from a fluctuated system, then while observing the change in structure, we can grasp in detail the process of phase transition. Even though phase transition is a macroscopic phenomena, the origin and the phenomena at the interface that connect those two phases are reflected by its individual particle. If we look closely on the crystallization and the melting process of water, the previously mentioned liquid-liquid transition and the geometry characteristic of hydrogen bond-network are tightly involved at the beginning of each process . From now on, together with the rapid development of computer speed, we can expect to discover various types of phase-transition of diverse kinds of substance.
When water is trapped between narrow slits or inside nanopores, it creates a unique network structure in order to maintain the nature of four-hydrogen bonded system. For example, simulation work predicts that water formed an ordered prism or spiral structure when it is confined inside carbon nanotube, which then successfully proved by experimental work . On the other hand, when aqueous solution is loaded inside carbon nanotube, the degree of ion intrusion is dramatically changed. Here we learned that there are still a lot of unique and undiscovered properties of water when it is confined in narrow space, and the most exciting part is to know that water in biological matter is actually in the same condition as confined water in narrow space. We can even ask, does the living actually use that unique behaviors of water inside narrow space.
III. Proton transfer reaction
Unlike other ions, hydrogen ion reacts with water molecule and rapidly transferred inside water. Solvent usually does not directly participate in chemical reaction inside a solution, but since water is reacting, the rate of proton transfer is directly affected by the arrangement of water molecules. In particular, biological substance use this fact for its benefit. A lot of biological substances was found to send signal and energy through hydrogen ion by preparing a proton wire inside protein to control the orientation of molecule. It turns out that water is not simply act as a solvent for biological substance, rather it is an integrated part of biological system from which a specific function is born. From here, if the research are making a big progress, we are going to find more cases where the living things utilize water in a more sophisticated way.
IV. Ice under high pressure
The form of ice that exists and stable under pressure about 2 GPa or higher is ice VII. It has two stacked, but not interconnected, network of diamond structure, with the oxygen atoms are arranged according to body-centered cubic lattice. With the increase of pressure, the distance between oxygen atoms gradually shorten, at 100 GPa two oxygen atoms finally sit next to each other. At high temperature and high pressure, while oxygen atom stays in its position in lattice, proton can freely move and the ice become a super ion conductor. On the other side, ice VII is a molecular crystal under pressure of several GPa, and it undergoes melting process accompanied by an increase in volume if heated. According to experimental result, this melting curve is largely uneven, but molecular dynamics simulation showed that there lies a phase between crystal and liquid phase, and that phase is called plastic phase of ice . In this plastic ice, water molecules are fixed on their position according to body-centered cubic lattice, but they are mostly free to rotate. With the molecular structure is similar to hydrogen sulfide, plastic phase has already been discovered by experimental work, thus the observation of plastic ice from experimental work can be expected. Water and ice under high pressure are largely exist in outer planet and satellite, and they are considered to play important role in affecting weather and soil condition.
V. Closing remarks
Computer simulation enable us to observe the motion of a large number of molecules in detail. But, we feel that we still lack of words that can fluently explain all things that happen inside the simulation. Even though we can explain the position and movement of surrounding water individually, but like the case of proton transfer reaction inside water, we are still poor in vocabulary to be able to describe the phenomena of cooperative movement of several water molecules. Like wine tasting, we cannot grasp and describe what we taste if there is no word that can describe it. This will certainly become a serious problem as computer keep advancing and simulation-scale continue to expand.
But there is no need to be pessimistic here, because even the simulation scale grow larger, if the unknown thing does not grow, then there is no need to have large scale computation. The next generation super computer is a machine that is designed to handle extremely complex phenomena at large scale, and we are now looking to feel and experience the result that born from that super computer.
 A. Rahman, F. H. Stillinger, J. Chem. Phys., 55, 3336 (1971).
 A. Pople, Proc. R. Soc. A, 205, 163 (1951).
 I. Ohmine, J. Chem. Phys., 85, 3342 (1986).
 F. H. Stillinger, Science, 209, 451 (1980).
 O. Mishima, L. D. Calvert, E. Whalley, Nature, 310, 393 (1984).
 D. Paschek, Phys. Rev. Lett., 94, 217802 (2005).
 M. Matsumoto, S. Saito, I. Ohmine, Nature, 416, 409 (2002).
 K. Koga, G. T. Gao, H. Tanaka, X. C. Zeng, Nature, 412, 802 (2001).
 Y. Maniwa, H. Kataura, M. Abe, A. Udaka, S. Suzuki, Y. Achiba, H. Kira, K. Matsuda, H. Kadowaki, Y. Okabe, Chem. Phys. Lett., 401, 534 (2005).
 Y. Takii, K. Koga, H. Tanaka, J. Chem. Phys., 128, 204501 (2008).