Covalent semiconductors are such solids as diamond, silicon, germanium, silicon carbide and silicon-germanium where atoms are linked by covalent bonds. Most of those materials, at least in their bulk form, are well studied and rarely hit the front pages of the top scientific journals in the last decade. However, issue 23 of volume 93 (2004) of a major physics journal Physical Review Letters contained as many as 4 papers on diamond.[1] Those papers were a reaction to a breakthrough discovery of superconductivity in synthetic diamond grown by high-pressure high-temperature (HPHT) method.[2] The discovery had no practical importance but surprised most scientists as superconductivity has not been considered seriously in covalent semiconductors.
Contents |
Superconductivity in diamond was achieved through heavy p-type doping by boron such that the individual doping atoms started interacting and formed an "impurity band". The superconductivity was of type-II with the critical temperature Tc = 4 K and critical magnetic field Hc = 4 T. Later, Tc ~ 11K has been achieved in homoepitaxial CVD films.[3][4]
Regarding the origin of superconductivity in diamond, three alternative theories exist at the moment: conventional BCS theory based on phonon-mediated pairing, correlated impurity band theory[5] and spin-flip-driven pairing of holes weakly localized in the vicinity of the Fermi level.[6] Whereas there is no solid experimental support for either model, recent accurate measurements of isotopic shift of the transition temperature Tc upon boron and carbon isotopic substitutions favor the BCS theory.[7]
It was suggested[2] that "Si and Ge, which also form in the diamond structure, may similarly exhibit superconductivity under the appropriate conditions", and indeed, discoveries of superconductivity in heavily boron doped Si (Si:B)[8] and SiC:B[9] have quickly followed. Similar to diamond, Si:B is type-II superconductor, but it has much smaller values of Tc = 0.4 K and Hc = 0.4 T. Superconductivity in Si:B was achieved by heavy doping (above 8 at.%), realized through a special non-equilibrium technique of gas immersion laser doping.
Superconductivity in SiC was achieved by heavy doping with boron[10] or aluminum.[11] Both the cubic (3C-SiC) and hexagonal (6H-SiC) phases are superconducting and show a very similar Tc of 1.5 K. A crucial difference is however observed for the magnetic field behavior between aluminum and boron doping: SiC:Al is type-II, same as Si:B. On the contrary, SiC:B is type-I. In attempt to explain this difference, it was noted that Si sites are more important than carbon sites for superconductivity in SiC. Whereas boron substitutes carbon in SiC, Al substitutes Si sites. Therefore, Al and B "see" different environment that might explain different properties of SiC:Al and SiC:B.[12]
Superconductivity in carbon nanotubes has been observed experimentally in 2001.[13] Note however a crucial difference between nanotubes and diamond: Although nanotubes contain covalently bonded carbon atoms, they are closer in properties to graphite than diamond, and can be metallic without doping. Meanwhile, undoped diamond is an insulator.
When metal atoms are inserted (intercalated) between the graphite planes, several superconductors are created with the following transition temperatures:[14][15]
Material | CaC6 | Li3Ca2C6 | YbC6 | SrC6 | KC8 | RbC8 | NaC3 | KC3 | LiC3 | NaC2 | LiC2 |
---|---|---|---|---|---|---|---|---|---|---|---|
Tc (K) | 11.5 | 11.15 | 6.5 | 1.65 | 0.14 | 0.025 | 2.3-3.8 | 3.0 | <0.35 | 5.0 | 1.9 |
The priority of many discoveries in science is vigorously disputed (see, e.g., Nobel Prize controversies). Another example, after Sumio Iijima has "discovered" carbon nanotubes in 1991, many scientists have pointed out that carbon nanofibers were actually observed decades earlier. The same could be said about superconductivity in covalent semiconductors. Superconductivity in germanium and silicon-germanium was predicted theoretically as early as in the 1960s.[16][17] Shortly after, superconductivity was experimentally detected in germanium telluride.[18][19] In 1976, superconductivity with Tc = 3.5 K was observed experimentally in germanium implanted with copper ions;[20] it was experimentally demonstrated that amorphization was essential for the superconductivity (in Ge), and the superconductivity was assigned to Ge itself, not copper.