UCRL-JC-126769
1. Do any metallic systems have electrical
conductivities similar to those of metallic hydrogen? Yes. The electrical
conductivities of metallic fluid Cs and Rb at 2000 K are identical to those of
fluid hydrogen when all three undergo the same continuous transition from a
semiconducting to metallic fluid.
2. What is different about these shock
experiments compared to previous ones which tried to metallize hydrogen at
static high pressures in a diamond cell? We shock-heated hydrogen to about
3000 K, which produced a fluid. Previous experiments with static megabar
pressures were performed at room temperature or below, which produced solid
hydrogen.
3. Why does metallization occur in the fluid at
a lower pressure than for the solid? Metallization in solid hydrogen is
inhibited by phase transitions in crystal structure and molecular orientation.
Neither exists in the disordered fluid.
4. Does any other element become metallic at a
lower pressure in the fluid than in the solid? Yes. Iodine becomes
metallic at 30 kbar in the fluid and 160 kbar in the solid.
5. What do you mean by a metal? Metallic
fluid hydrogen is achieved when high pressures reduce the energy gap Eg between the filled
valence-electron band and the unfilled conduction-electron band down to Eg~kBT,
where kB is Boltzmann's constant and T is the temperature. When Eg~kBT, thermal
smearing in the disordered fluid fills in the energy gap, a metallic density of
electronic states is achieved, and the electronic system has a Fermi surface
characteristic of a metal.
6. Why not heat hydrogen in a diamond cell to
achieve the same high pressures and temperatures as in the shock experiments?
Because of the high mobility of hydrogen at high temperatures, the diamond cell
is essentially transparent to hydrogen diffusion. The highest recorded
temperature of hydrogen in a diamond cell is 500 K. At higher temperatures the
cell is empty because hydrogen diffuses away.
7. How does shock compression heat hydrogen?
Because shock compression is so fast, it is also adiabatic. That is, heat
produced by compression has insufficient time to be conducted or radiated away.
Also, by reaching final pressure with a shock reverberating in soft hydrogen
between stiff sapphire anvils, the final hydrogen temperature is about 1/10 what
it would be if final pressure were reached in a single shock. In this way the
temperature is about twice the melting temperature and 2% of the Fermi
temperature, just the right temperature to produce melted condensed matter.
8. Why doesn't hydrogen diffuse out of a
shock-compressed sample? The experiment lasts only 100 ns, too short a time
for the hydrogen to diffuse away.
9. Is metallic hydrogen in thermal equilibrium?
Yes. Within the 3-ns time resolution of the diagnostic system, there are ~105 intermolecular
collisions and 4 times as many vibrations. This number of collisions is larger
by 3-4 orders of magnitude than required to achieve thermal equilibrium.
10. Is metallic hydrogen in electrical
equilibrium? Yes. The thickness of the hydrogen layer decreases from the
initial value of 500 mm
down to the compressed value of 50 mm. The calculated flux diffusion time for a layer 50 mm thick with our highest
electrical conductivity of 2000 (W-cm)-1 is < 1 ns, which indicates that the electrical current
reaches its equilibrium flow pattern in <1 ns.
11. Is the experiment affected by hydrodynamic
interfacial instabilities? No. Rayleigh-Taylor and Richtmyer-Meshkov
instabilities did not occur when shock waves traversed the planar interfaces
between sapphire and hydrogen because the initial surfaces of the sapphire
crystals were optically flat (300 A rms surface roughness) and the time duration
of the experiment (100 ns) was too short to allow such small instabilities to
grow during the duration of the measurement.
12. Is the temperature lowered by radiative
cooling? No. The temperature lost by radiation in 100 ns is <1 K out of
3000 K, a negligible amount. Thermal radiation is emitted from about 2 optical
depths (640 A@) at the surface of the metallic hydrogen. The energy radiated at
3000 K was calculated from the Stefan-Boltzmann radiation law and converted to
the radiated temperature using the calculated heat capacity of 2 optical depths
of metallic hydrogen.
13. Is the temperature lowered by thermal
conduction? No. An interfacial layer of metallic hydrogen ~0.5 mm thick is cooled about 200 K in
100 ns; both are negligible compared to the 50-mm total thickness at 3000 K. The heat conducted was calculated
using values of thermal conductivity and diffusivity of metallic hydrogen
calculated with the Wiedemann-Franz law; values for alumina at 1500 K, its
calculated shock temperature, were taken as handbook values.
14. Is the measured conductivity caused by a
metallic interfacial layer formed by high-temperature chemical reactions?
No. Such a layer, if it were to exist, would be too thin after 100 ns and its
electrical conductivity too small to account for the measured signal. For
example, there is insufficient time to form a thin metallic layer of pure Al by
the reduction of Al2O3by metallic hydrogen. In 100 ns the conservatively estimated
diffusion constant of H2into Al2O3is too small for the hydrogen to get sufficiently deep inside
the Al2O3and the diffision constants of O2, H2O, and OH-are far too small for
them to get out of the anvil to allow formation of a metallic layer of Al.
15. Is current carried by ions, rather than by
electrons. No. Current is carried by electrons. The Drude conductivity
depends inversely on the mass of the carrier. The masses of an electron and of
a proton, the lightest possible ion, differ by a factor of 2000, which indicates
that electronic conduction dominates. For example, the electrical
conductivities of ionic alkali-halide fluids are typically about 1 (W-cm)-1, while
conductivities of pure metallic alkali fluids are typically a few 1000 (W-cm)-1. We measured a
metallic hydrogen conductivity of 2000 (W-cm)-1.
16. Is current carried by electrons in an
impurity band of H monomers in a semiconducting H2host, as in the
degenerate doped semiconductor Si(P)? No. The
electronic structures of the H atom and the H2molecule are very
similar, which means the energy gap we observe is that of the H2-H mixture.
Also, fluid H2at 3000 K does not have a high dielectric constant, which means
that H2and
H only interact at short range. In contrast, the electronic structure of
crystalline Si and P are very different. Also, crystalline Si has a high
dielectic constant, which permits dilute concentrations of P to interact at long
range in a semiconducting Si host and form a conducting P impurity band in the
bandgap of Si. Metallic fluid hydrogen is nothing like Si(P).
17. Is the value of the electrical resistivity
of metallic hydrogen consistent with simple expectations? Yes. The
measured value of 500 mW-cm is bracketed by simple models. The formula for the
calculated maximum resistivity of a metal gives a value of 250 mW-cm. A calculation of the
electrical resisivity using the Ziman weak-scattering model for a molecular
liquid metal gives a value of 100 mW-cm. The electrical resistivity calculated in the
strong-scattering free-electron model gives 1700 mW-cm. All these values are within a factor of 5 or less of the
measurement, which indicates the measured value is reasonable.
18. Why has the measured resistivity value of
metallic hydrogen, 500 mW-cm, not been calculated theoretically? Because no
theory exists for this novel state of condensed matter. For example,
tight-binding molecular-dynamics calculations indicate that the energies of
translation, vibration, and rotation of H2are all comparable in fluid metallic hydrogen. Present
theories of liquid metals assume these energies are significantly different from
one another.
19. If metallic hydrogen could be retained metastably on release of pressure, what properties would it have? Metallic hydrogen is speculated to have a number of interesting properties and important applications, if it could be quenched from high pressures to ambient. It is not known how, nor even whether, metallic hydrogen could be quenched to ambient but the potential benefits are enormous if it could be. For this reason it is worth speculating on possible uses of metastable solid metallic hydrogen.
The ten times higher density of metallic DT fuel pellets, relative to molecular solid ones, would increase substantially the energy produced in laser-driven ICF, giving giant lasers, such as LLNL's NOVA and future NIF, an even larger margin for success than expected previously. The higher starting density of the fuel might produce a sufficiently large increase in ICF energy yield such that it might be possible to use only deterium as the fuel. The absence of radioactive tritium might make the ICF energy source much more attractive to commercial energy producers.
Metallic solid hydrogen has been predicted to be a room-temperature superconductor, which would result in substantial energy conservation nationwide.
Metastable metallic hydrogen would have a very high density of stored energy because it would have a density about ten times that of liquid H2 at 1 bar. Thus, the stored energy released by reversion to the diatomic insulating fluid would also be very large and metastable metallic hydrogen would have widespread applications as fuels. If this energy were released relatively slowly or quickly, metallic hydrogen would be either a clean propellant, as gasoline, or an explosive, respectively. The predicted specific impulse of metallic hydrogen is about 5 times that of liquid H2/O2fuel now used to launch rockets into space. This large increase in specific impulse would result in smaller cheaper spacecraft, which would greatly facilitate space travel.
If solid metallic hydrogen has sufficient strength, it might be useful as a light-weight structural material. For example, automobiles made of metallic hydrogen would be ~10 times lighter than current ones made of steel, enhancing fuel efficiency and reducing conventional fuel emissions. The ideal would be to synthesize metallic hydrogen to be either extremely metastable, as diamond, for use as a structural material or readily reactive, as gasoline.
Large quantities of metallic hydrogen might be made by shock recovery methods using large systems of chemical explosives, as DuPont now shock-synthesizes diamond.
Web Weaver of this page:
Kelly A. Barrett-- kab@llnl.gov
UCRL-JC-126769