No abstract

Platinum metal was shock compressed to 660 GPa using a two-stage light-gas gun to qualify this material as an ultrahigh-pressure standard for both dynamic and static experiments. The shock velocity data are consistent with most of the previously measured low-pressure data, and an overall linear us−up relationship is found over the range 32–660 GPa. As a part of this work, we have also extended the Hugoniot of the tantalum standard we use to 560 GPa; we have included these data into a new linear fit of the tantalum Hugoniot between 55–560 GPa. We also present the results of a first-principles theoretical treatment of compressed platinum. The fcc phase is predicted to remain stable to beyond 550 GPa. In addition, we have calculated the 300-K pressure-volume isotherm and the Hugoniot. The latter is in excellent agreement with experimental results and qualifies the former to at least 10% accuracy.

Electrical conductivity measurements indicate that fluid hydrogen achieves the minimum conductivity of a metal at 140 GPa, ninefold initial liquid-H 2 density, and 2600 K. Metallization density is defined to be that at which the electronic mobility gap E g is reduced by pressure to E g ϳk B T, at which point E g is filled in by fluid disorder to produce a metallic density of states with a Fermi surface and the minimum conductivity of a metal. High pressures and temperatures were obtained with a two-stage gun, which accelerates an impactor up to 7 km/sec. A strong shock wave is generated on impact with a holder containing liquid hydrogen at 20 K. The impact shock is split into a shock wave reverberating in hydrogen between two stiff Al 2 O 3 anvils. This compression heats hydrogen quasi-isentropically to about twice its melting temperature and lasts ϳ100 ns, sufficiently long to achieve equilibrium and sufficiently short to preclude loss of hydrogen by diffusion and chemical reactions. The measured conductivity increases four orders of magnitude in the range 93 to 140 GPa and is constant at 2000 ͑⍀ cm͒ Ϫ1 from 140 to 180 GPa. This conductivity is that of fluid Cs and Rb undergoing the same transition at 2000 K. This measured value is within a factor of 5 or less of hydrogen conductivities calculated with ͑i͒ minimum conductivity of a metal, ͑ii͒ Ziman model of a liquid metal, and ͑iii͒ tight-binding molecular dynamics. At metallization this fluid is ϳ90 at. % H 2 and 10 at. % H with a Fermi energy of ϳ12 eV. Fluid hydrogen at finite temperature undergoes a Mott transition at D m 1/3 a*ϭ0.30, where D m is the metallization density and a* is the Bohr radius of the molecule. Metallization occurs at a lower pressure in the fluid than predicted for the solid probably because crystalline and orientational phase transitions in the ordered solid do not occur in the fluid and because of many-body and structural effects. Tight-binding molecular dynamics calculations by Lenosky et al. suggest that fluid metallic hydrogen is a novel state of condensed matter. Protons are paired transiently and exchange on a timescale of a few molecular vibrational periods, ϳ10 Ϫ14 sec. Also, the kinetic, vibrational, and rotational energies of the dynamically paired protons are comparable. ͓S0163-1829͑99͒02805-2͔

Hugoniot curves for Al (alloy 11000), Cu (type oxygen-free high-conductivity), and Ta have been measured in the shock pressure range 30–430 GPa (0.3–4.3 Mbar) with a two-stage light-gas gun. Impactor velocities were measured to 0.1% by flash radiography. Shock velocities were measured to 0.5–1.2% with an electronic detection system with subnanosecond time resolution. Our data and those of other workers were fitted to a linear relation between shock and mass velocities. The fractional standard deviations of the data from the fits range from 0.6 to 0.9% for the three metals. Methods of data analysis and error analysis for individual data points and for the least-squares fitting to the data sets are presented. Bands of uncertainty about the fits, arising from experimental uncertainties in the data, are presented and are used to calculate the systematic error introduced by the method of shock-impedance matching. The accuracy of the data and of the fits qualifies these metals as equation-of-state standards for shock-wave experiments.

Temperatures of shock compressed liquid deuterium and hydrogen up to 5200 K were measured at pressures up to 83 GPa (830 kbar). The measurements are in excellent agreement with earlier calculations to about 20 GPa and show evidence for dissociation above 20 GPa. At the highest measured temperatures and pressures current theories break down and a revised theory is proposed.

Dynamic equation-of-state data for liquid H2O and NH3 were measured in the shock pressure range 30–230 GPa (0.3–2.3 Mbar) using a two-stage light-gas gun. Electrical conductivities of water were also measured in the shock pressure range 28–59 GPa (280–590 kbar). The experimental techniques to measure the electrical conductivity in a 50 ns time interval and to cool the target holders to liquid ammonia temperatures (230 K) are described. The H2O data are discussed in terms of the statistical mechanics model of Ree. At temperatures above 3000 K significant molecular ionization occurs.

Dynamic equation-of-state data for D2 and H2 were measured in the pressure range 2–76 GPa (20–760 kbar) using a two-state light-gas gun. Liquid specimens were shocked from initial states near the saturation curve at 20 K. Maximum compression was sixfold over initial liquid density at a calculated temperature of 7000 K for D2. The data is discussed in terms of the theory of Ross et al., which includes an effective intermolecular pair potential, molecular vibration, free molecular rotation, and molecular dissociation.

Hugoniot points of liquid D 2 were measured at shock pressures of 107, 54, and 28 GPa using converging explosively driven systems ͑CSs͒. The two data sets measured with a laser ͑L͒ and pulsed currents ͑PCs͒ differ substantially. Our results are in excellent agreement with the PC data and the error bars of the CS-PC data are less than half those of the L data. The limiting compression obtained from the best fit to the CS-PC data is 4.30± 0.10 at 100 GPa. The CS-PC data are in good agreement with path integral Monte Carlo and density functional theory calculations, which is expected to be the case at even higher shock temperatures and pressures, as well.The single-shock compression curve ͑Hugoniot͒ of deuterium up to 100 GPa ͑1 Mbar͒ pressures has been controversial because limiting shock compression close to sixfold of initial liquid density has been reported using a high-intensity laser ͑L͒ ͑Ref. 1͒ and limiting compression close to fourfold has been reported using large pulsed currents ͑PCs͒. 2-4 That is, as pressure achieved with a single shock increases, so too does temperature, which limits compression at sufficiently high pressures. Examination of the systematics of singleshock compression of diatomic liquids suggests that the PC data are correct. 5 Deuterium in all the shock experiments is in thermal equilibrium because there are more than 10 4 collisions between atoms and/or molecules within the respective time resolutions. Deuterons in these experiments are classical. 6 Thus, there is no a priori reason why fluid deuterium would be expected to behave differently than other low-Z diatomics, as reported in Ref. 1. In order to determine the correct Hugoniot of D 2 , we began experiments on solid 7,8 and liquid samples in 1999. In this paper we report Hugoniot points at 109, 54, and 28 GPa for liquid D 2 samples.Strong shock waves were generated with hemispherical convergence driven by explosives ͑CS͒, the same method we used previously to measure points at 121 and 61 GPa for solid samples. 7,8 Our points at 109 and 121 GPa achieve limiting compression. Our points at 109 and 121 GPa and at 54 and 61 GPa used liquid and solid D 2 samples, respectively, to demonstrate self-consistency and reproducibility. Our point at 28 GPa demonstrates agreement with data measured at lower pressures with a two-stage gas gun ͑GG͒. 9 To minimize uncertainties, our CS method 10,11 requires that a given experiment be repeated several times and the results averaged. We have performed thirteen cryogenic, explosively driven experiments to obtain the three data points for liquid samples reported here. Our method produces data at 100 GPa pressures in the simple materials Al and Cu which are in excellent agreement with data obtained with a two-stage gun and planar explosives. 12,13 Thus, while our method produces relatively few data points, our results are in excellent agreement with data obtained by other techniques of demonstrated accuracy at shock pressures which can be obtained with all three methods.High shock pressures were gener...