Fig. 9.20. Electric conductivity of water versus pressure. 1–5, Yakushev et al. (2000); 6,

Mitchell and Nellis (1982); 7, Hamann and Linton (1966).

potentials.

Figure 9.20 summarizes the experimental results on the electrical conductivity of water. It can be seen that starting from 30 GPa the conductivity depends weakly on pressure and saturates at approximately 150 ohm−1cm−1. Following Hamann and Linton (1966), this can be attributed to the complete dissociation of water. In electrochemical experiments, galvanic cells having electrodes of various metals and water as the electrolyte were subjected to dynamic compression. The characteristics of the recorded e.m.f. of these cells, along with the fact that the measured conductivity is far from a typical metallic value, indicate that the high electrical conductivity of highly compressed water is of an ionic nature.

9.3.7Dielectrization of metals

The conventional point of view in solid states physics is that the structural phase transitions occur in a solid as the density and pressure increase, thus causing it to form a close–packed phase with maximal coordination number. As a result, insulators become conductors and poor metals improve the metallic properties. Numerous experiments, including those discussed above, seem to support this general picture. However, modern sophisticated quantum–mechanical calculations (Neaton and Ashcroft 1999) predict much reacher and interesting behavior of matter at high pressures. For example, lithium, like other alkali metals, has long been considered as a prototype “simple” metal. At normal conditions, alkali metals have a simple bcc structure, metallic sheen and conductivity. However, theory shows that lithium under pressure transforms into an orthorhombic phase at 50 GPa. At higher pressures near 100 GPa, lithium nuclei form pairs, which

results in the formation of structures similar to condensed phases of molecular hydrogen, with the electron properties being close to the properties of semiconductors with narrow energy gap. Finally, at even higher pressures, lithium reverts to a monoatomic metal.

In experiments carried out in diamond anvil cells to a pressure of 60 GPa (Struzhkin et al. 1999; Mori and Ruo 1999) it was found that the metallic sheen of lithium disappears under compression – it becomes gray and then black (i.e., strongly absorbing) at p = 50 GPa. Recent X–ray di raction studies carried out at pressures up to 50 GPa by Hanfland et al. (2000) have also revealed a sequence of structural transitions. According to these measurements, near 39 GPa lithium transforms from a high–pressure fcc phase through an intermediate rhombohedral phase to a complex bcc phase with 16 atoms per cell. Calculations performed by Hanfland et al. (2000) predicted high stability of this phase up to pressures of about 165 GPa.

In experiments by Fortov et al. (1999b, 2001), direct measurements of electrical resistivity of compressed lithium revealed anomalous behavior of its electrophysical properties. Lithium was compressed in multistep shock experiments up to a pressure of 210 GPa and density of 2.3 g cm−3. The data obtained by Fortov et al. (2001) are shown in Fig. 9.21 in the form of normalized resistivity ρ/ρ293 as a function of density. As one can see, the resistivity increases monotonically with density for all experiments corresponding to a maximum pressure of 100 GPa at initial temperatures of 77 K and 293 K. The data obtained at higher pressures (160 and 212 GPa) yield the same dependence in the investigated density range up to 1.75 g cm−3. At higher densities of 2.0–2.3 g cm−3, the resistivity decreases dramatically. Lithium melts under conditions of dynamic experiment in the first or second shock at pressures below 7.3 GPa and temperatures below 530 K, depending on the intensity of the incident shock wave. The final states of dynamically compressed lithium, according to the results of 1D numerical modeling with real equation of state, correspond to the liquid state at temperatures from 955 to 2833 K. The estimated thermal contribution to the lithium resistivity is about 20–25% of the total value at the maximum density. Therefore, the main reason for the change in lithium resistivity is a decrease in the interatomic distances. One should note that the dependence of resistivity on density is similar both in the solid and liquid states. Another interesting and unusual fact is that, under conditions of dynamic experiments, liquid lithium is a poor conductor up to 160 GPa, whereas at higher pressures the resistivity is decreasing. Presumably this suggests that compressed lithium has an ordered structure which is destroyed at 160 GPa, and then lithium again becomes a “good” metal.

Recently, the existence of new phases in compressed sodium at pressures p > 130 GPa has been predicted by Neaton and Ashcroft (2001), based on the density functional method employed to calculate the atomic and electron structure. Similar to the case of lithium, the new phases are di erent from those expected for “simple” metals – they have low structural symmetry and semimetallic electron properties. The experimental confirmation of a possible transi-