Ionic defects in ice

Scheiner S, Nagle JF (1983) Ab initio molecular orbital estimates of charge partitioning between Bjerrum and ionic defects in ice. J Phys Chem 87 4267 1272... 

The results of the Devlin group raise several issues. Recall that the ice-Ih/XI phase transition at 72A is catalyzed by hydroxide."" However, it appears that no ionic defects are actively diffusing at 72A. Furthermore, even though hydroxide, not protons, catalyze this phase transition, the results of Devlin s group indicate that protons are initially the most active diffusing species as temperature is raised. Hence, the results of the Devlin group call out for a better understanding of the behavior of ionic defects in ice, and in particular, the mechanism by which hydroxide catalyzes the Ih/XI phase transition. 

M. Benoit, D. Marx, and M. Parrinello (1999) The role of quantum effects and ionic defects in high-density ice. Solid State Ionics 125, p. 23... 

The nature of the charge carriers is not yet clear and there could be at least two possibilities ionic impurities or H-bond network defects (so-called orientation and ionic defects) similar to those considered in the conduction and relaxation of ice [242,243]. Conductivity behavior due to ionic impurity is expected to increase linearly with the impurity concentration per unit volume, and the impurities would be proportional to water content. However, the normalized conductivity did not show such a linear behavior (Fig. 39), although some tendency of increase with water content can be observed. This increase of the normalized conductivity does not immediately contradict the concept of defect-conductivity, because the number of defects can also be increased by increasing water content. At the very least, diffusion of ions should also be accompanied by breaking or changing of H-bond networks around the ion molecule. Such rearrangements of H-bond networks around ions can affect the EW and may influence the main relaxation process on a cluster level. Thus, the existence of universality is most likely due to the presence of an H-bond network and its defect structure. 

One of the most difficult and intriguing aspects of ice physics is the behavior of defects. Although defects have little effect on the statics of phase transitions, they are the key to understanding the dynamics and mechanism of phase transitions. For example, ice-Ih must be doped with hydroxide to catalyze the transition to ice-XI. The mechanism by which hydroxide (OH ) catalyzes the ice-Ih/XI transition is not understood. In fact, it is not even clear that hydroxide defects have significant mobility in this temperature range near 70/v. The structure and transport properties of defects is relevant to problems in environmental and atmospheric science " and glaciology. In this work, we introduce techniques applicable to the study of ionic defects, and OH , and neutral defects, such as the OH radical, in ice. 

To date, for pure water ice phases only second order invariants generated by projection on a small number of nearby bond pairs were needed. For example, for ice-Ih three second order invariant functions provided an accurate parameterization of the energy. We used those same three invariant functions with identical a coefficients to describe the pure water portion of the system with an L- and ionic defect. We incorporated 6 additional invariants of the form given in Eq. (2) involving a 6- and closeby c-variable. On physical grounds, we expect charge-dipole interactions to be important in the presence of ionic defects. The... 

The dielectric relaxation process of ice can be understood in terms of proton behavior namely, the concentration and movement of Bjerrum defects (L- and D-defect) and ionic defects (HaO and OH ), which are thermally created in the ice lattice. We know that ice samples highly doped with HE or HCl show a dielectric dispersion with a short relaxation time r and low activation energy of The decreases in the relaxation time and... 

Seidenstickers model (138) accounts for the d.c. and a.c. conductivity measurements on NHs-doped ice in terms of hybrid levels, that is, levels where both Bjerrum (D) and ionic defects are present at the same nitrogen site. If, further, the D defects associated with ammonia are only slightly dissociated, the model predicts that both a.c. and d.c. conduc-... 

This information allows us to make tentative estimates of the concentration of orientational defects in pure ice, using an equation like (7.1), and of their mobility. It is clear from the energies involved that they should be much more numerous in pure ice than are the ion states. The energy barrier to proton motion is comparable in height to that for ion states but twice as wide, so that it is possible, and indeed turns out to be the case, that the anomalously high mobility of ionic states does not extend to orientational defects. Experimental information, derived from studies of the electrical properties of ice, is summarized for convenience in table 7.3. 

Notwithstanding the fact that motion of ionic defects is connected with the motion of protons, it does not imply a real proton transfer. Actually there occur only successive local displacements of protons along the bonds. Due to the memory of the protonic system, after a certain amount of current has passed, all the bonds appear to be blocked, and further passage of current is impossible. It should be remembered however, that there are also defects of the second type in the system, namely, D and L defects, shown in Fig. 10.2, which represent violations of the first Bernal-Fowler rule. A D defect moving in the same direction as an defect, polarizes the bonds in the opposite direction, that is, it unblocks them as seen in Fig. 10.2. In an analogous way the motions of OH and L defects are related. So, by a combined motion of all the defects (or of only an D pair) a current may pass through ice indefinitely. D and... 

Electrical measurements of ice are diflBcult to interpret because of polarization effects, surface conductivity, injection of defects and/or impurity atoms from sandwich electrodes, diffusion effects, differential ion incorporation, and concentration gradients due to nonsteady state impurity distribution. Theories formulated for pure ice and for ice doped with HF (KF and CsF) in terms of ion states and valence defects, qualitatively account for experimental data, although the problem of the majority and minority carriers in doped ice, as a function of concentration and temperature, requires further examination. The measurements on ice prepared from ionic solutes other than HF, KF, and CsF are largely unexplained. An alternative approach that treats ice as a protonic semiconductor accounts for results obtained for both the before-named impurities as well as ammonia and ammonium fluoride. 

---