Le sujet de ce travail a été une étude détaillée du supraconducteur CuxTiSe2. Des échantillons avec différentes teneurs en cuivre x, ont été étudiés par magnétométrie locale à sonde de Hall. Nous avons tout d'abord construit les profils de champ magnétique dans les échantillons pour déterminer quand les premiers vortex pénètrent dans l'échantillon. La forme en dôme des profils a indiqué que le piègeage est très faible et que la pénétration des vortex est principalement régie par des barrières géométriques, ce qui nous a permis de déterminer le champ critique inférieur, Hc1. Un bon accord entre la théorie et Hc1 (T) a été obtenu en utilisant un modèle à deux gaps, même si l'origine du plus petit de ces gaps ne peut être liée à la structure électronique de CuxTiSe2 . En combinant ces mesures avec des mesures de chaleur spécifique effectuées précédemment sur ces échantillons, nous avons obtenu un comportement non trivial de l'énergie de condensation. En outre, les mesures effectuées pour diverses orientations du champ magnétique ont révélé la présence d'un effet de "lock-in" dans CuxTiSe2, montrant que les vortex restent piégés le long d'une structure en couches même pour les champs magnétiques inclinés. L'orientation des vortices a été déterminée expérimentalement et a pu être bien décrite par un modèle tenant compte de la présence d'une "sur-structure" dans laquelle la supraconductivité est - au moins - partiellement détruite. Cet effet de lock-n a été analysé sur trois échantillons et nous avons montré que la force de piégeage est indépendante du dopage du cuivre. La nature de cette sur-structure est encore inconnue mais pourrait être liée à l'observation récente de parois de domaine des ondes de densité de charge, qui pourraient induire de fortes variations du paramètre d'ordre supraconducteur. / The subject of this work was a detailed study of superconducting CuxTiSe2. Samples with various copper content x, from underdoped to overdoped regions of the phase diagram, were investigated by local Hall probe magnetometry. At first, we constructed the magnetic field profiles in the samples to determine when the first vortices penetrate into the sample. The observed dome-like shape of the profiles indicated that pinning is very small and that the vortex penetration is primarily governed by geometrical barriers, hence enabling us to determine the lower critical field, Hc1. A good agreement between theory and the Hc1(T) data was obtained by using a model with two energy gaps, even though the origin of the smaller energy gap in CuxTiSe2 cannot be related to its electronic structure. Combining those measurements with heat capacity measurements previously performed on these samples, we obtained a non trivial condensation energy. Moreover, measurements performed for various orientations of the magnetic field revealed the presence of a lock-in effect in CuxTiSe2, when superconducting vortices are trapped along a layered structure even for tilted magnetic fields. The orientation of the vortices was experimentally determined and could be well described by a model considering the presence of an additional layered structure in which superconductivity is at least partially suppressed. The lock-in effect was analysed in three different samples and we have shown that the strength of the lock-in is independent of the copper doping, even though this effect can be masked by strong demagnetisation effects in very thin samples. The strength of this lock-in effect indicates that the superconductivity is strongly suppressed on certain layered structure. The nature of this "extrinsic" pinning structure is still unknown but might be related to the recent observation of charge density waves domain walls which may induce strong variations of the superconducting order parameter.

The superconductivity in magnesium diboride (MgB2) was discovered in 2001. As a BCS superconductor, MgB2 has a record-high Tc of 39 K, high Jc of > 107 A/cm2 and no weak link behavior across the grain boundary. All these superior properties endorsed that MgB2 would have great potential in both power applications and electronic devices. In the past 15 years, MgB2 based power cables, microwave devices, and commercial MRI machines emerged and the next frontier are superconducting radio frequency (SRF) cavities. SRF cavities are one of the leading accelerator technologies. In SRF cavities, applied microwave power generates electrical fields that accelerate particle beams. Compared with other accelerator techniques, SRF cavity accelerators feature low loss, high acceleration gradients and the ability to accelerate continuous particle beams. However, current SRF cavities are made from high-purity bulk niobium and work at 2 K in superfluid helium. The construction and operational cost of SRF cavity accelerators are very expensive. The demand for SRF cavity accelerators has been growing rapidly in the past decade. Therefore, a lot of effort has been devoted to the enhancement of the performance and the reduction of cost of SRF cavities. In 2010, an acceleration gradient of over 50 MV/m has been reported for a Nb-based SRF cavity. The magnetic field at the inner surface of such a cavity is ~ 1700 Oe, which is close to the thermodynamic critical field of Nb. Therefore, new materials and technologies are required to raise the acceleration gradient of future SRF cavity accelerators. Among all the proposed approaches, using MgB2 thin films to coat the inner surface of SRF cavities is one of the promising tactics with the potential to raise both the acceleration gradient and the operation temperature of SRF cavity accelerators. In this work, I present my study on MgB2 thin films for their application in SRF cavities. C-epitaxial MgB2 thin films grown on SiC(0001) substrates showed Tc > 41 K and Jc > 107 A/cm2, which is superior to bulk MgB2 samples. Polycrystalline MgB2 thin films grown on metal substrates showed similar Tc and Jc compared with bulk samples, indicating MgB2 is suitable for coating a metal cavity. Large c-pitaxial MgB2 thin films were grown on 2-inch diameter c-sapphire wafers, showing our technique is capable of depositing large area samples. The lower critical field (Hc1) of MgB2 thin films was measured as well as it is know that bulk MgB2 has a small Hc1 and would suffer from vortex penetration at low magnetic fields. The penetrating magnetic vortices would result in loss in an applied RF field. However, due to the geometry barrier, thin film MgB2 would have a higher Hc1 than the bulk material. In my experiments, the Hc1 of MgB2 thin films increased with decreasing film thickness. At 5 K, a 100 nm epitaxial MgB2 thin film showed enhanced Hc1 ~ 1880 Oe, which is higher than Hc1 of Nb at 2 K. This showed that MgB2 coated SRF cavities have the potential to work at higher magnetic fields and higher temperature. Because the magnetic field distribution in the thin film Hc1 measurement is different from the magnetic field in a real SRF cavity, a few Nb ellipsoids were machined and coated with MgB2. The ellipsoid only has a magnetic field outside its surface and can serve as an inverse SRF cavity in the vortex penetration measurement. In the experiments, vortices penetrate into the bulk Nb ellipsoid at a magnetic field 400 Oe lower than the vortex penetration field of MgB2 coated Nb ellipsoids. This result confirmed our prediction that MgB2 coated SRF cavities could work at higher magnetic fields, thus producing higher acceleration gradients. In the last part of this thesis, I discussed how I used the dielectric resonator technique to measure the surface resistance (Rs) and Tc of MgB2 thin films. While the sensitivity of this technique was not high enough to lead to reliable Rs values, it can still serve for the determination of Tc for large area samples that are too bulky for other measurement systems. / Physics

Cette thèse se décompose en deux parties.Dans la première, nous nous intéressons au premier champ critique, Hc1, du titanate de strontium, que nous mesurons à l’aide d’un réseau de microsondes de Hall taillées dans un gaz bidimensionnel. La valeur du premier champ critique nous permet alors d’évaluer la densité superfluide à six différents dopages couvrant l’ensemble du dôme supraconducteur. À bas dopage, nous trouvons que celle-ci correspond à la densité de porteurs dans l’état normal tandis qu’au-delà du dopage optimal, celle-ci chute drastiquement. En plaçant nos résultats dans le contexte de la loi de Homes, nous voyons que cette chute s’explique par l’entrée dans la limite sale. Un fit multibande de Hc1(T), dans ce contexte semble également indiquer que la supraconductivité émerge de la bande la plus basse et est seulement induite dans les deux autres bandes.Dans la seconde partie, nous regardons l’évolution de la densité de porteurs, n, du cuprate Nd-LSCO. Nous mesurons ainsi six échantillons de dopages proches du point critique pseudogap, p*, via trois sondes de transport : effet Hall, résistivité et effet Seebeck. Nous trouvons que n chute de 1+p à p à l’entrée dans la phase pseudogap. En comparant les différentes sondes, nous montrons que cette chute est due à une reconstruction de la surface de Fermi et qu’il existe sûrement des poches d’électrons et de trous juste en dessous de p*. Ceci est en accord, entre autres, avec un scénario antiferromagnétique. Finalement, nous trouvons que la mobilité est inchangée à l’entrée dans la phase pseudogap et que les mesures de transports semblent insensibles à la divergence de la masse effective vue par chaleur spécifique. / This thesis consists of two parts.The first one is about the lower critical field of strontium titanate, measured thanks to an array of Hall micro-probes tailored in a 2D electron gas. The value of the lower critical field allows us to quantify superfluid density at six different dopings spreading all along the superconducting dome. At low doping, we find that it follows the normal state carrier density while it dramatically falls above optimal doping. Analyzing our results in the context of the Homes law, we understand that this drop is due to the entering into the dirty limit. A multiband fit Hc1(T) in this context seems to indicate that superconductivity is born in the lowest band and only induced in the two others.In the second part, we focus on the carrier density, n, of the cuprate Nd-LSCO. We measure six samples with doping close to the pseudogap critical point p*, thanks to three different transport probes : Hall effect, resistivity and Seebeck effect. We find that entering the pseudogap phase induces a drop in n from 1+p to p. The comparison of the different probes shows that this drop is due to a Fermi surface reconstruction and that both holes and electrons pockets may exist just under p*. This observation is consistent with an antiferromagnetic scenario. Finally, we find that mobility is not affected by the pseudogap and that transport measurements seems insensitive to the diverging effective mass as observed by specific heat.