HYPERCONDUCTIVITY IN CHILLED BERYLLIUM METAL

lt is shown that in the vicinity of 77 K beryl!ium has a superior specific conductance compared with the nominally excellent metallic conductors aluminum and copper. lt is concluded that beryllium should be considered for some conduction applications, despite its well known toxicity problems.

(Received 13 December 1989; accepte<i for publication 7 May 1990) lt is shown that in the vicinity of 77 K beryl!ium has a superior specific conductance compared with the nominally excellent metallic conductors aluminum and copper. lt is concluded that beryllium should be considered for some conduction applications, despite its well known toxicity problems.
Bery11ium has several unusual and even unique properties when ccmpared with other metallic conductors. lt is a good conductor at room temperat.ure (3.328 µ,fl cm in polycrystalline form) compared with Al (2.24 µü, cm) and Cu ( 1.64 µ0 cm). lt has a Poisson ratio of 0.02-0.08 and an elastic moduius of 275-300 GPa, 1 has a high Debye temperature 0 2 (1481 K) , a modest magnetor.esistivity, 3 a high rnelting temperature ( 1575 K), and excellent thermal conductivity. 4 In a sense, Be is an hcp metallic form of diamond, sharing diamond's covalent sp bonding, strength, and phononic properties, but with the added feature of high electrical conductivity.
In this letter we focus on a figure of merit defined as the specific conductance ( the ability of a material to conduct electrical current per kilogram). We show that Be is so superior in this regard compared to other metallic conductors at 77 K that it may have important new hyperconductivity 3 applications, despite its toxicity and expense. 5 Beryllium in the form of airborne fine particulates has a nasty toxicity which became notcrious. This makes fabrication of Bea process requiring careful control, whose safety was embodied in specific laws even before awareness of other possible environmental problems had become so common. It is possible for such reasons that Be has perhaps been neglected and its unusuai properties forgotten. That most met.als have lowered resistivity at low temperatures has been studied since at least the time of Matthiessen and Vogt 6 with the resuit that generally the resistivity of any metal is separable into two component.s: where for reasonably pure materials p 0 is temperature independent and primari!y due to residual im.purities or surface scattering and p;( T) is due to the tempernturedependent occupation of tiuctuating phononic, magnetic, or other dissipative vibronic modes. It is common 7 to plot the normalized resistivity (=p(T)/p(298)j as a function of reduced temperature t( = T /(}) . In this case the electrical resistivity of many metals falls on a single curve: an early triumph for solid-state physics and the Bloch-Grueneisen law. But this common practice precisely con-cea1s the point we wish to emphasis her e. In Note that we plot here the total residual resistivity in all cases, since this is the dissipative resistance seen by any power source. We have used the resistivity at 298 K to normai.ize all data plotted in measurements is ' 5%, with an absolute uncertainty of ±20%. The RBS surface measurements are accurate to ± 5% and the RBS bulk measurements to ± 20%. The total impurity content of the sample was 6150 ppm (atomic fraction) ofwhich 6000 ppm was oxygen and carbon as found in the outer 8000 A of the surface. lt is difficult to extrapoiate such a measurement to determine the bulk oxygen or carbon concentration. Nevertheless 60C-O ppm is a reasonable upper limit. W e have not made a study of which impurities cause the highest resistivity increase in Be, or of whai techniques would remove them, although for further work or applications this is surely important. We conci.ude that the samples used in our investigation were certainly excellent for the time they were grown. However, they may not represent the metallurgical limit of what can be achieved in Be nowadays in terms of either impurity content, microstructure, or RRR.  The noteworthy feature ofFig. 1 is that all three sets of Be data show a substantial leveling off in the resistivity near 77 K. This leveling-off does not happen until neariy 20 K in other good conductors such as Al or Cu. The main reason ts not hard to discern: Be has a substantially higher e by a factor of 3 er 4 than all the others. At and below boiling nitrogen temperature, most phonons of Be are frozen out, exposing the effects of residual impurities on the resistivity.
Based on the results of Fig. l, we were intrigued t.o see how weil Be compared to other materials near 77 K . To make a useful comparison for ccnducting materials, we propose a figure of merit (FOM) which is the specific conductance (mh.os per kilogram). We envisage a task of making a 10 m conClucting bar from 1 kg of material with the maximum conductance. As is clear, such a FOM essentially scales the cross-sectional area A of the bar: the iighter the material, the thicker the cross section and the sma!let the bar's resistance. Fora fiJ<.ed Iength L, thi.s FOM devo!ves into the ratio of the temperature-dependent conductivity to tJ1.at of the density: where a is the conductivity and 6 is the density. ( In Fig. 2  Tbe results we have given here are tailored for an act.ively cooled system. We have not considered more complicated and dynamical system parameters such as the power transfer time, relative cost, refrigeration efficiency, cooling method, load Jevel, and system reliabiiity. Ncvertheless, in space-borne app1ications, Be conductors may have significant benefit compared with Cu or AI. Since the radiant temperature of space is about 80, 14 Be conductors might be useable without any cooling. 15 SimiJarly Be rnay have significant impact on advanced naval, land, or air app!ications where weight savings in power transfer is important. Futhermore, with bulk high-temperature superconductors presently limited to a critical current density of 10 8 A/m 2 , beryllium conductors may be of superior and more immediate utility for many conduction applications near 77 K.
We would like to thank R . Borg, M. A. Hill, H. L. Laquer, W. Stewart, J. Rogers, and F. R. Fickett for helpful discussions on the conductive properties of high-purity met.als. Part ofthis work was performed under the auspices of the United States Department of Energy. 242