We have recently described a three-dimensional self-assembling model (SAM) for water (1). Each water molecule is modeled by a sphere containing four tetrahedrally placed magnets, two with their north poles facing out (representing the hydrogens of water) and two with their south poles facing out (representing the electron pairs of the oxygen).

In these models the north—south attraction of the magnets represent the dipolar and hydrogen bonding interactions that are present in water. While hydrogen bonding forces are generally stronger than dipolar forces, the two share a dependence on distance and directionality. Therefore, both can be modeled by the magnetic interactions of the SAM models. In fact, the mathematical equations for dipolar forces are similar to the equations for the interactions between magnets. Furthermore, the interactions between magnets and magnetizable objects are analogous to the interactions of a dipole and a polarizable body, and are described by the same equations as are induction forces.

These models tend to spontaneously assemble themselves into the three dimensional polyhedra that are characteristic of the structure of ice, that is, each molecule is attached to four others at the tetrahedral angles so that large void spaces are created. If pressure is applied to these structures they are easily compressed and their ice-like lattice is destroyed. If they are mixed with similar sized spheres that do not contain magnets, and then agitated, they spontaneously separate into separate regions.

Clark’s letter suggests that we have equated magnetic interactions with the chemical bond. We did not. We are simply making an analogy between dipolar magnetic interactions and the nonbonded interactions among water molecules. In fact, we do not even mention covalent or ionic chemical bonds.

These models do not imply anything about either the nature of the chemical bond or the thermodynamics of intermolecular interactions. The interactions among the SAMs are analogous to those among water molecules in their strength, directionality, and ability to break and reform readily. As such, they can demonstrate some but not all of the properties of water. In spite of the fact that they cannot model entropy, the SAMs can be used to physically mimic the formation of ice and its ability to float on water. They are also useful in demonstrating the phase separation of nonpolar liquids and water.

It is important to note that the quotation in the preceding article is not from our work and it does not in any way reflect our view on the subject. Furthermore, it is no less appropriate to model a bond by magnetic attraction than by the plastic or wooden dowels used in most ball-and-stick models. Finally, since the references cited by Clark do not demonstrate tetrahedral models or models of water, they were not cited in our article (1). The closest magnetic demonstrations of water–water interactions were proposed by Davies (2) and Pravia and Maynard (3), who both used magnets for two-dimensional overhead projections.

We appreciate Clark’s willingness to point out what he feels might be confusing to readers as well as the opportunity to clear up any misconceptions that our model is meant to be anything other than a model.

Literature Cited

1. Johnson, J. L. H.; Yalkowsky, S. H. J. Chem. Educ. 2002, 79, 1088–1091.
2. Davies, W. G. J. Chem. Educ. 1991, 68, 245.
3. Pravia, K.; Maynard, D. F. J. Chem. Educ. 1996, 73, 497.