Spin triplet supercurrents in thin films of ferromagnetic CrO2
Anwar, M.S.
Citation
Anwar, M. S. (2011, October 19). Spin triplet supercurrents in thin films of ferromagnetic CrO2. Casimir PhD Series. Retrieved from https://hdl.handle.net/1887/17955
Version: Corrected Publisher’s Version
License: Licence agreement concerning inclusion of doctoral thesis in the Institutional Repository of the University of Leiden
Downloaded from: https://hdl.handle.net/1887/17955
Note: To cite this publication please use the final published version (if applicable).
Appendix B
Resistance anomaly of CrO 2 based SFS devices
We fabricate CrO2based SFS devices in the lateral geometry using an a-MoGe superconductor deposited on unstructured 100 nm thick CrO2 thin films, as discussed in detail in Chapter 5. In Chapter 5, it is also described that R(T ) data at lower temperatures of our devices have an anomalous behavior of showing an up-jump at Tc. There is no plausible explanation of such anomalous behavior. To understand it bit more, we remark on the following points, (i) the lateral-geometry of the junction; and (ii) the discontinuous resistance change;
(iii) the cleaning procedure of the CrO2.
Junction Geometry, The contact geometry of the full film is very unusual, with the CrO2 having a very low resistivity (5 µΩcm) and contacts a very high one (200 µΩcm). Current therefore would tend to flow through the film along unknown paths. If this leads to an up-jump in R when the contacts go from highly resistive to superconducting, we should observe this for low-resistive metallic films other than CrO2. The R(T ) for a junction fabricated with Au (see Fig. B.2) and with a Co (see Figs. A.3 and A.5) films show a normal down jump at the Tc, which suggests that the the up jump cannot be related with the lateral geometry of the junction. There is another reason it cannot be related with full CrO2film underlying the superconductor because we also measured the same up-jump for a device for which we etched away the extra CrO2and kept it only between and under the superconducting leads.
Discontinuous Resistance Change, There are two mechanisms which can be thought of involving the superconducting transition. One is the fact that Andreev reflection are fully suppressed by the 100% spin-polizaed ferromagnet, which might somehow lead to an increased interface resistance. However, this effect should show a temperture-dependence which reflects the opening of the
107
108 Chapter B. Resistance anomaly of CrO2 based SFS devices superconducitng gap. The jump being discontinuous, does not show this in the R(T ) for a Co (about 50% spin polarization) based junction reveals that the suppression of AR cannot give such an up-jump.
The other mechanism is charge imbalance relaxation. This effect actu- ally does give a jump-like change in resistance at Tc. The geometry to see it is an S/N (or S/F) interface, with one voltage probe measuring the chemi- cal potential on the N-side, and one probe made of an N-contact measuring the quasi-particle (qp) chemical potential on the S-side (an S contact would measure the Cooper pair chemical potential). Due to the diverging change im- balance length, the voltage difference for the qp’s is larger just below Tc than given by S-metal just above the Tc. However, this is not our measurement geometry.
Figure B.1: (a) Resistance as a function of temperature for an uncleaned SFS device fabricated on pretreated TiO2. The inset shows the R(T ) data for a device fabricated with an old CrO2film, where a thick insulating layer of Cr2O3
was present. (b) Differential resistance as a function of applied voltage for an uncleaned SFS device. (c) R(T ) for a cleaned device prepared on a pretreated TiO2 substrate and (d) its differential resistance versus applied voltage.
109 Cleaning, to check the effect of the cleaning, we studied two SFS devices using 100 nm thick CrO2 films deposited on pretreated TiO2 substrates (also mentioned in Chapter 5). The first one was uncleaned (uncleaned device) and the second one was properly cleaned for five minutes (cleaned device). The R(T ) data for an uncleaned device show a down jump followed by a small up jump (see Fig. B.1a). The differential conductance as function of applied voltage shows a clear gap structure around the zero voltage. It is interesting to note the linear increase of the conductance just below the gap which might come from the tunneling behavior of the interface and might provide evidence of the presence of an insulating barrier. On the other hand, the R(T ) data of a cleaned device show an up-jump. For this device, the differential conductance is not showing a clear superconducting gap-like structure. A clean interface might be the reason. It is making it clear that up-jump occurs for cleaned devices when we do not have an insulating barrier at the interface. It is also confirmed with the only down jump for the device fabricated on an old film with rather thick insulating layer of Cr2O3 (see the inset of Fig. B.1a). The devices fabricated with Nb as a superconducotr also show an up-jump in the same way, which reveals that up-jump is not connected with a-MoGe.
From these investigations it can be concluded that the up-jump only occurs for cleaned junctions fabricated with CrO2. A specific source of the up-jump is still not known but might be related to material properties of CrO2, in particular the full spin polarization.
Figure B.2: R(T ) for a device that is fabricated with a Au film rather than a CrO2film.
