Resumen de Scanning tunneling spectroscopy of superconductors close to a quantum critical point and at magnetic impurities
Immediately after discovering superconductivity in Pb, H. Kamerlingh Onnes tried to wind a Pb wire to make a magnet and obtain very high magnetic fields. He wanted to address an application with his discovery, aiming at building dissipation-free electrical motors. He rapidly saw, however, that superconductivity disappeared as soon as the current in the solenoid created a field of a few tens of Gauss. His disappointment lead him to center efforts in a more fundamental study of superconductivity. But his observation already pointed out a relevant fundamental aspect of superconductivity, developed by Ginzburg 40 years later, namely that superconductivity and magnetism are antagonistic phenomena.
This antagonism provides an interesting playground in modern superconductivity, as there are many situations where it might turn into a positive interaction, often connected with an increase in the critical temperature. In that sense, Kamerlingh Onnes’ disappointment was leading us to high critical temperature superconductivity! In this Ph.D. thesis I have explored this issue using scanning tunneling microscopy (STM). STM provides an atomic view of the superconducting properties. I chose to study BaFe2(As1−xPx)2, a system where superconductivity has maximum Tc exactly at the P concentration where magnetism disappears (at x =0.3), and isolated magnetic impurities in 2H-NbSe2−xSx. In the course of my studies, I also explored the Josephson effect at atomic level and found features that are completely new.
In the iron based pnictide superconductors, I have analyzed a sample exactly at the quantum critical point where magnetism disappears, BaFe2(As0.7P0.3)2, and another sample in the paramagnetic state, BaFe2(As0.56P0.44)2. Contrary to other doped superconductors, here the As by P substitution does not influence significantly the sample quality. I have observed the As/P atomic lattice and for the first time the vortex lattice in both samples. In BaFe2(As0.7P0.3)2, I find a strong tendency to form a well ordered square vortex lattice locked to the atomic lattice, connected to a strong superconducting gap anisotropy. These effects are absent in BaFe2(As0.56P0.44)2, suggesting that they are caused by magnetism. Furthermore, I have measured the band structure using quasiparticle interference, finding a strong superconducting gap anisotropy in BaFe2(As0.7P0.3)2. A careful analysis of the scattering patterns allows us to extract information about the gap opening and the band structure. When analyzing the magnetic field dependence of the vortex core size, I found an increase and a subsequent decrease of the vortex core size at BaFe2(As0.7P0.3)2. Such a behavior is absent in BaFe2(As0.56P0.44)2 and is contrary to the expectation for a usual superconductor, which consists of a magnetic field induced decrease in the vortex core size. The peculiar behavior is related to the properties of isolated vortex cores in BaFe2(As0.7P0.3)2, which are, at low fields, in the so-called quantum limit, where the core level spacing exceeds the thermal energy. This is an interesting consequence of the diverging mass at the quantum critical point.
I furthermore analyzed the transition metal dichalcogenides 2H-NbSe2 and 2H-NbSe1.8S0.2 with Fe magnetic impurities. I show that we can obtain gapless superconductivity with a very small amount of magnetic impurities in 2H-NbSe1.8S0.2. I carefully analyzed scattering patterns of the in-gap states induced by the Fe magnetic impurities. I found oscillatory patterns due to the Fe impurities that are associated to portions of the band structure with an increased density of states. The oscillatory patterns are absent in pure 2H-NbSe2, suggesting that the S concentration reduces the dimensionality, increasing the scattering signal due to a stronger 2D-character. When there are many magnetic impurities, these overlap and lead to gapless superconductivity. Moreover, I studied in detail the interaction between dilute states from Fe impurities and from quantized states inside vortex cores. I found that the electron-hole asymmetric character of Fe impurity states is transferred to the much larger vortex core states through exchange interaction, causing vortex cores to become axially asymmetric in presence of magnetic impurities.
I have also discovered an unexpected low frequency time dependent AC Josephson signal. I have characterized the effect carefully and shown that it is a novel feedback effect acting on the Josephson junction. The feedback provides a significant enhancement of the Josephson coupling observed in the experiment, improving the sensitivity of Scanning Josephson Spectroscopy. It leads to an interesting bistable behavior, in which the junction switches between the usual AC Josephson effect, with a non-zero voltage, and the zero voltage state with a DC Josephson current.
