Jorge L. Gonzalez’s research while affiliated with Federal University of Espírito Santo and other places

Nb/Pt and Nb/Cu nanohybrids were sputtered on Si(100) substrate and systematically studied by transmission electron microscopy and magneto-transport measurements. Our experimental findings show that the colossal spin-valve effect measured in a hybrid formed by thick Pt layers deposited on Nb films is absent in equivalent Nb/Cu nanohybrids. In the latter, an ordinary spin-valve effect was experimentally measured and numerically quantified using the superconducting anisotropic phenomenon based on the Ginzburg–Landau model. The unusual enhancement of the spin-valve effect is explained considering the formation of odd-frequency triplet states of Cooper pairs at the Nb/Pt interface induced by the spin–orbit coupling of the Pt component. In a broad perspective, this study strongly evidences the role that the spin–orbit interaction can play for controlling the spin state of Cooper pairs at interfaces of superconductor-based hybrids in the absence of ferromagnetic materials.

Biochar and ashes derived from coffee husk were used as heterogeneous catalysts in the synthesis of biodiesel through a transesterification reaction. The catalysts were obtained from the carbonization, activation, and combustion of raw coffee husks. Their properties were characterized by solid-state 13C and 31P nuclear magnetic resonance (NMR) spectroscopies and X-ray diffraction, among other techniques. The results evidenced the presence of different inorganic compounds (mostly K- and Ca-containing phases) mixed with the turbostratic structure (in the case of the biochar samples). A biodiesel conversion of 66% (evaluated by 1H NMR analysis of the liquid reaction products) was found for the biochar sample prepared at 700 °C; the activated biochar catalyst prepared at the same temperature showed a higher biodiesel conversion (74%), which can be attributed to its superior specific surface area. The best catalytic efficiency (biodiesel conversion of 93%) was observed for the coffee husk ashes, which is consistent with the higher contents of Ca and K salts in the ashes in comparison with the biochar samples. Reuse tests conducted with the ashes samples showed an efficiency reduction after the second cycle (from around 90% in the first two cycles to 8% in the third cycle), due to the partial removal of active phases (mostly K-containing salts) within the reaction medium. The presented results show that coffee husks are a cheap and environmentally viable source for the production of materials of interest in heterogeneous catalysis, with no need for chemical modification to achieve good efficiency.

Velocity models are crucial intermediate products generated in seismic data processing, and the model's accuracy is essential for constructing quality seismic images. Conventional approaches to velocity model building employ a family of inversion methods, among which are ray-based tomography and full-waveform inversion. These methods have been highly optimized throughout the years but are still heavily dependent on continuous human curation of the results, which leads to an overall high time cost, especially in areas with high structural complexity, such as those containing salt tectonics. We investigate a deep learning approach that accurately defines salt geometries for velocity model building. We train our convolutional neural network on synthetic shot gather data, explore a manner of leveraging information through summation of shot data, and demonstrate the influence that the choice of loss function has on the quality and aspect of predicted velocity models. Our Residual U-Net model trained on data containing only randomly shaped salt bodies can estimate geologically complex salt geometries such as those in 2D SEG/EAGE Salt Model slices. Our results show that deeper encoder-decoder models with shortcut connections resolve velocity model structures better than shallower models. Moreover, network models trained with a composite loss function - combining mean absolute error and the Multi-Scale Structural Similarity Index - better delineate the contours of areas with high-velocity contrast and better recover regions with a uniform velocity trend than network models trained with conventional loss functions like the mean squared error. The Residual U-Net and loss functions we employ are not task-specific and can be extended to other deep learning approaches to velocity model building.

Hybrid Nb(20)/Cu(5)/Py(2), Nb(20)/Cu(5)/Co(40) and Nb(20)/Cu(d$_{Cu}$)/Py(2)/Cu(5)/Co(40) heterostructures (values in nanometers and d$_{Cu}$ = 0, 2.5 and 5.0) were fabricated using a confocal DC magnetron sputtering setup. Performing magnetotransport measurements and using the anisotropic Ginzburg–Landau approach, the stray field values in zero-applied field ($B_\mathrm{0Co}$, $B_\mathrm{0Py}$ and $B_\mathrm{0Py+Co}$) were estimated as being $- 9~$mT ($+ 31~$mT) for the hybrid Nb/Cu/Co (Nb/Cu/Py) trilayers and $+ 36~$mT for the hybrid ordinary Nb/Cu(5)/Py/Cu/Co spin-valve heterostructure. The effective fields acting on the superconducting layer in the hybrid non-ordinary Nb/Cu/Py and Nb/Cu/Co spin-valve heterostructures and its dependence with the applied magnetic field were also quantified, showing that the stray fields of thin ferromagnetic layers are the same order of magnetide but with different strengths. For an applied field of 1 T, the spin-valve effect values of $-~230~$mK ($+~300~$mK), previously found for the hybrid Nb/Cu/Co (Nb/Cu/Py) trilayers and $+~140$ mK for the Nb/Cu(5)/Py/Cu/Co heterostructure, had their physical origin better discussed and demonstrated. The 2-nm-thick ferromagnetic Py layer strongly contributes to the effective fields, and the low spin-valve effect of $+~140$ mK for the ordinary Nb/Cu(5)/Py/Cu/Co spin-valve heterostructure would be a consequence of two contributions of opposite signs governed by the Py and Co layers and also due to the proximity effect contribution.