A group of researchers working with Séamus Davis, a member of Berkeley Lab's Materials Sciences Division and a professor of physics at UC Berkeley, have devised an extraordinary probe for examining superconductivity on the nanoscale.
In the first test of the new technique, a special scanning tunneling microscope (STM) in Davis's lab was used to study the influence of individual nickel atoms implanted in a high-temperature (high-Tc) superconductor. The results, published in the June 21, 2001, issue of Nature, shed light on puzzling aspects of these still-mysterious materials.
"One of the perplexing things about superconductivity is that magnetic impurities destroy superconductivity in conventional low-temperature superconductors, whereas high-Tc superconductors may actually depend on some kind of magnetic mechanism," says Davis.
Davis and his colleagues set out to investigate this phenomenon directly at the atomic scale in a superconductor known as BSCCO (pronounced "bisko," for bismuth strontium calcium copper oxide), in order to determine what influence individual impurity atoms have on electronic structure in their immediate neighborhoods.
Superconductivity is the flow of charged particles through a material without resistance, which happens when electrons form so-called Cooper pairs. Cooper pairs form below the superconducting transition temperature (Tc), which is only a few degrees above absolute zero in conventional superconductors, as cold as liquid helium or colder. Phonons, quantized vibrations of the material's crystal lattice, help create regions of positive charge between the two electrons, "holes" which overcome the mutual repulsion of the electrons' negative charges.
Associated with each Cooper pair is another kind of pair, formed by each electron and its accompanying hole. These "quasiparticles" are fictitious representations of real particle systems, including the quantum states by which they are identified, but they make computation managable in a way impossible for complete quantum solutions.
Magnetism enters the picture, Davis explains, because "in Cooper pairs one electron's spin" its quantized angular momentum "points 'up' and the other's points 'down,' which gives rise to oppositely oriented magnetic moments."
If an external magnetic field is applied to nonsuperconducting systems, electrons of similar energy are separated by their spins. This splitting doesn't affect superconductors, however, because "magnetic fields cannot penetrate the surface region of superconductors." On the other hand, Davis says, "if a conventional superconducting material is riddled with magnetic impurities, magnetism can't be kept out the Cooper pairs are split apart in the vicinity of each magnetic atom, which in conventional superconductors destroys their superconductivity."
This is not so for high-Tc superconductors, whose transition temperatures are warmer than liquid nitrogen. "Nickel atoms are magnetic, but nickel impurities have a weak effect on superconductivity in high-Tc superconductors like BSCCO," says Davis. "Oddly, zinc impurities disrupt it, and zinc atoms are nonmagnetic."
Part of the explanation lies in the electronic states characteristic of high-Tc superconductors. All the highest-Tc superconductors found so far are copper oxide ceramics having the crystal structure of the mineral perovskite, with planes of copper and oxygen atoms (where superconductivity is thought to occur) interlayered with planes of other atoms.
Working with the complex perovskite BSCCO what he calls "this horrible structure" Davis and his colleagues substituted several nickel atoms for every thousand copper atoms in the copper oxide planes. With their scanning tunneling microscope, they identified the nickel atoms in a small area of the plane.
The electronic states of Cooper pairs in high-Tc superconductors are markedly different from those in conventional ones: the two electrons revolve around each other much faster and farther apart, as do their associated quasiparticles. These wider orbits are analogous to the higher-energy d orbitals of electrons around an atom, and high-Tc superconductors are often called d-wave superconductors.
Davis and his colleagues set out to test the theory that a magnetic atom should capture the electron parts of the quasiparticles associated with Cooper pairs, leaving the hole parts to orbit rapidly in the d-wave configuration.
Making use of the STM's ability to measure electronic states on the scale of individual atoms, Davis and his colleagues were able to construct images of regions of positive and negative bias in the immediate vicinity of each nickel atom. Positive bias on the sample revealed a cross-shaped distribution of negative charges electrons while negative bias revealed the x-shaped distribution of positive charges holes.
Moreover, the researchers discovered two peaks in energy near each nickel atom, corresponding to the opposing up and down spins of the quasiparticle pairs.
"This shows that a nickel atom retains its overall magnetic moment in the superconducting state and doesn't disturb that state," Davis says. "In fact, it helps maintain the particular magnetic properties of the cuprate perovskite system."
Coupled with the observation that similar doping with zinc impurities destroys superconductivity in these systems each zinc atom destroying superconductivity within a radius of 1.5 nanometers, possibly because zinc atoms form nonmagnetic voids this is good evidence that high-Tc superconductivity depends on uninterrupted magnetic pathways to aid the flow of charge.
For the first time, researchers have a technique for testing their theories of the microscopic organization of high-Tc superconductors on an atom by atom basis.
Davis's newfound ability to measure the quantum spin states of individual atoms opens even larger vistas of possibility, including a potential mechanism for getting information into and out of the would-be superfast quantum computers of the future.