Superconductor physics is a subject that is shroudedin mystery. In this ?eld, Mott insulators are a spe-cial kind of material since their electric conductivitydoes not convey the standard band theory model. Es-pecially at very low temperatures, they become insu-lating. To gain more knowledge in this ?eld, we havestudied the physics behind a Mott insulator junctionmade of alternating layers LSCO (La2xSrxCuO4)and NCCO (Nd1?xCexCuO4), capped with an STO(SrTiO3) layer. These layers are grown on an LSAT-crystal substrate ((LaAlO3)0.3(Sr2AlT aO6)0.
7) usingPulsed Laser Deposition (PLD). This can be seen inFigure 1.What we want to ?nd out are the characteristics ofthin-?lm pn-junctions under cold circumstances (<100K), and if there is a link between theory fromthe ?eld of semiconductors and superconducting pn-junctions.
Since LSCO (x=0.09) is positively andNCCO (x=0.1) negatively doped these two can formthe required pn-junctions. In the researched situation,the LSCO is expected to be superconducting while theNCCO is expected to be anti-ferromagnetic.This anti-ferromagnetic property is important in ourresearch, since an anti-ferromagnetic phase makes thematerial an insulator (Coulomb repulsion). When thematerials in question are cooled below their criticaltemperature, this anti-ferromagnetic phase will van-ish as the result of differences appearing in electronspins (creation of cooper pairs), albeit the material be-comes superconducting.
The expected outcome of theresearch is that the material (Mott insulator, normallyinsulating) will behave like a superconducting semi-conductor when brought in an extremely cold thermalequilibrium. The Fermi level in the material will alignsimilar to the level of energy bands in semiconduc-tors. The measured semiconductor band of the super-conducting material will then look similar to ordinarypn-junctions.
Fig. 1: Multilayer LSCO/NCCO structure2 METHODOLOGY2.1 Structuring and DepositionThe goal is to achieve an outline as shown in Fig-ure 2. Here, the black part represents the multilayerstack with around it its measurement leads. This hasbeen achieved using lithography, etching and sput-tering respectively 1, 2, 3. After creating the out-line, pre-prepared LSCO and NCCO targets are ap-plied onto the substrate inside a PLD chamber. STO-capping prevents conduction from the leads into themultilayer stack.
The black square is where a photo-resistive layer has been applied, after which the excessSTO, NCCO and LSCO is removed using an argon-ion beam.Fig. 2: Outline with contact leads2.2 Quality ChecksSince this is the ?rst time such a sample has beengrown, a thorough quality check is needed to ensurethe reliability of the data that is being extracted frommeasurements. Used methods are X-ray diffraction todetermine deposited materials, and X-ray re?ectivityto estimate layer thickness. Material deposition errorsare mostly found between LSCO and NCCO layers,with a maximum error of 4.6%.
Layer thickness hasbeen measured at 48nm, close to the 50nm thicknessbeing expected. Both measurements are consideredsuitable for experiments.3 MEASUREMENTSTwo methods of measuring have been applied. The?rst measurement method is done by placing the sam-ple in a bath cryostat to measure at low temperatures(T = 4.2K).
The second using Physical PropertyMeasurement System (PPMS) to measure in a widertemperature range. Both measurement methods havebeen applied to a single-layer and multilayer sample.The single pn-layer acts like a control group to in-dicate if there is pn-behaviour present, which is thenscaled up to a multilayer measurement.4 RESULTS4.1 Single-layer pn behaviourFigures 3 and 4 show the curves of a single pn-layer, measured 300K and 4.2K.
Already beingnoticed is the nonlinear behaviour around the zerovoltage. This is expected to be a depletion layerdue to the pn-contacts 4. In general, resistance ishigher at 4.2K, possibly by emergence of an anti-ferromagnetic phase.Figure 5 shows the IVT-curves of the single-layersample with the absolute value of the current beingcoloured. The pn-behaviour is clearly visible. The re-sistance increases in the 300K 180K range and below40K again.
4.2 Multilayer SampleInteresting results arise from measurements on themultilayer sample. In Figure 6 you can see thataround 140K the resistance decreases slightly. The(a)(b)Fig. 3: IV-curves for junctions 1-6NCCO layer has a phase transition into the anti-ferromagnetic region under 140K. Another interest-ing measurement comes from junction 11, Figure 7.At around 12K the resistance drops signi?cantly, in-dicating a transition to its superconducting phase 5.It is not clear why it does not become completely su-perconducting, but it is is assumed that this layer be-comes superconducting before the NCCO layers haveanti-ferromagnetic properties.
These results indicate that below 140K, the LSCOdecreases in resistance. When the temperature dropseven further, the NCCO layers enter the high-resistiveanti-ferromagnetic phase. A resistance drop at 20Kthen indicates a superconducting phase.5 CONCLUSIONSWe can conclude that in the single-layer sample, pn-behaviour is clearly present.
Possible improvementslie in the lowering the deviation of optimal dopingduring deposition of the PLD plume. This will im-(a)(b)Fig. 4: IV-curves for junctions 7-12Fig. 5: IVT-curvesFig. 6: IVT-curve junction 7, LSCOFig.
7: IVT-curve junction 11, LSCOprove the sample quality. IV-curves at 300K showohmic and nonlinear behaviour. Low resistance hasbeen measured in the LSCO layer in the 10 ?20Ktemperature range, indicating superconductivity.6 DISCUSSIONSince this is the ?rst time such a multilayer stack hasbeen built, the growth conditions have mostly beenan educated guess. Nonetheless is the sample of ac-ceptable quality that has given off reasonable results.
Also, Mott-insulators is still a ?eld where not a lot ofresearch has been done. That has made it dif?cult topresent ?rm conclusions about physical effects arisingin the data.