1
Water Runs Deep

Nicholas E. Ernst and Bruce C. Gibb

Tulane University, Department of Chemistry, 2015 Percival Stern Hall, 6400 Freret Street, New Orleans, LA, 70118, USA

1.1 The Control of Water

There is current consensus among archeology scholars that there are five cradles of civilization, geographical locations where the first civilizations emerged thousands of years ago. These five areas - the Fertile Crescent, the Indus River, the Yellow River, the Central Andes, and Mesoamerica - each independently gave rise to a new level of human existence [1-3]. What do these sites have in common? They all shared a combination of favorable geography and the opportunity to control water.

Because of its central role in life on Earth, the idea of controlling water manifests in many aspects of human endeavor. In agriculture, the key is irrigation - how to bring water to the fertile land and crops. "Controlling" the oceans and rivers requires harbors, moorings, and boats to open up trading routes or the possibility of fishing, while controlling sewage has been key to the success of cities. Controlling water also allowed energy production in the form of watermills and, in more modern times, dams, tidal barrages, and wave energy convertors. Unfortunately, the control of water can also be used to justify war. Conflicts between nations, states, or groups, for example, the violence in the war in Sudanese Darfur, can be partially attributed to water [4, 5].

For many different reasons, the control of water at the molecular and nanoscale levels is also key. Many scientific disciplines work toward being able to understand and control water purification for waste treatment and desalination [6, 7], water flow for microfluidics and ultra-hydrophobic materials [8], and solvation to in turn control dissolved matter. At this level, the control of water is free of famine and conflict. However, it is plagued by beguiling complexity; the solvent of life, this "matrix of the world and of all its creatures" (Paracelsus, 1493-1541 CE) is a truly complex substance possessing endless emergent phenomena [9, 10]. This means that it is easy for pathological science [11] to make wild claims about the memory properties of water [12, 13] or the existence of polywater [14, 15]. Although, at the time, the latter stoked a public fear reminiscent of Vonnegut's Cat's Cradle [16], polywater has passed into history. Alas the same cannot be said for memory water and infinitely diluted homeopathic remedies. And lest it be thought that there are no new controversies around water, consider also the relatively recent idea that water is capable of storing charge [17-19].

As the complexity of water means that it is exceedingly difficult to truly understand how water interacts with itself and how it interacts with solutes [20], the primary aim of this chapter is to outline what we do and do not know about supramolecular interactions in aqueous solution. Correspondingly, a secondary aim is to minimize situations wherein an observation might be explained away by the waving of hands. For as has been noted previously, "When confronted with unexpected experimental results water structure has commonly been used deus ex machina for explaining the observations." [21] We can and should do better than this.

1.2 The Shape of Water

At the heart of water and aqueous solutions lies a contentious question that goes back millennia - what is the structure of liquid water? The Greek philosopher Empedocles (490-430 BCE), like many thinkers around the world, considered water to be one of the four classical elements: earth, water, air, and fire. More helpfully, Democritus (460-370 BCE) proposed that all matter was made up of small indivisible particles called atoms (atomism). Around the same time, Plato (428-348 BCE) was being less helpful but slightly prophetic. He merged geometry with the four elements of the universe to propose that water was one of five (Platonic) solids: tetrahedron, cube, octahedron, dodecahedron, and icosahedron. Specifically, Plato proposed that water is an icosahedron, which explained its ability to flow across (cubic) the Earth.

It was not until 1781 that Henry Cavendish showed water to consist of hydrogen (inflammable air) and oxygen (dephlogisticated air) by burning a mixture of the two [22]. This discovery was built further upon by Amadeo Avogadro, who was the first to clearly differentiate between atoms and molecules, and that the latter must be represented by an empirical formula. He also hypothesized that equal volumes of gas at equal pressure contain the same amounts of molecules regardless of the nature of the gas, and from this was able to repeat Cavendish's experiment to show that water was two parts hydrogen and one part oxygen [23].

Moving into the late 1800s and the early 1900s, a combination of Röntgen's discovery of X-rays, Einstein's concept of a photon, and Compton's confirmation of X-ray scattering by electrons ultimately led to the development of X-ray crystallography. This led Bragg and others to determine the structure of ice. In the solid state, water possesses tetrahedral geometry [24] in which oxygen atoms are 2.75?Å from each other. There was, after all, an element of truth to what Plato had thought: water was not an icosahedron, but four water molecules did make a tetrahedron. That same decade saw the nucleation of the idea of hydrogen bonding (H-bonding). In 1920 Latimer and Rodebush proposed that a "free pair of electrons might be able to exert sufficient force on a hydrogen held by a pair of electrons on another water molecule to bind the two molecules together." However, it was not until 1939 and the publication of Pauling's The Nature of the Chemical Bond that the H-bond began to meet with widespread acceptance. Hexagonal ice, the form of all natural snow and ice on Earth, has the perfect tetrahedral structure with bond-bond, bond-lone pair, and lone pair-lone pair angles of 109.47° [25]. The reason for the high boiling point, large heat of vaporization, high heat capacity, and high surface tension of water was evident.

Partially through the development of X-ray crystallography, liquids began to be thought of as either dense ordered gases or imperfect disordered crystals, and with its H-bonding network, water was a prime candidate for this line of thinking. For example, an influential model by Bernal and Fowler treated water as a point charge with tetrahedral geometry whose structure was akin to that of disordered quartz [26]. As we shall shortly see (vide infra), this idea of water possessing considerable structure quickly moved into adjacent areas of research.

The Bernal and Fowler paper influenced thinking through the 1950s. Thus, Erying [27] proposed a liquid state structure of water consisting of crystalline close packing threaded with many dislocations. According to this model, molecules escaping the close packing could wander almost gas-like between the clusters. Similarly, Pauling proposed a disordered crystal based on the clathrate structure, while Bernal proposed a model of a random H-bonded network where water molecules gave rise to connected four- and seven-membered rings.

The last half-century or so has seen a dipartite approach to studying water. On the one hand, there have been the continued improvements in established spectroscopic techniques as well as the development of new spectroscopies [28]. The suite now available is extensive and includes Raman/Raman-multivariate curve resolution (Raman-MCR), IR, 2D-IR, sum-frequency generation techniques, and multiple X-ray and neutron approaches such as scattering, small-angle scattering, X-ray absorption spectroscopy (), and X-ray emission spectroscopy (). Complementing this approach has been the development of computational chemistry. Thus, the combination of increasingly powerful computers, more accurate water models (e.g. TIP4P-Ew) [29], and new strategies has added greatly to our understanding of water and solvation at the molecular level [30].

We now know that the mean H─O bond length in liquid water is 0.97?Å, the mean HOH bond angle 106°, and the mean negative charge on the O atom ~70% of an elementary charge, with each positively charged H atom sharing the neutralizing charge. We also know that although water makes four H-bonds in the solid state, as a liquid it forms ~3.6 H-bonds on average. However, this number varies according to the specific analytical technique used. Furthermore, there is still some debate regarding the symmetry of the H-bonds in liquid water. In the gas phase, water forms minimum energy clusters [31-33]; however, it is clear from simulations that these structures do not persist in bulk solution at ambient temperature. Rather, the current consensus is that water possesses primarily tetrahedral structure [34, 35], although this is not universally agreed upon. Thus to pick one controversial example, XAS suggests that ~80% of water molecules have one strong and one weak...