We report a facile preparation approach of MAPbBr3, MAPbCl3, or FAPbBr3 (where MA = CH3NH3 + and FA = CH­(NH2)2 +) perovskite nanowires via sequential synthesis of MAPbI3 and FAPbI3 nanowires with chemically controlled composition and morphologies followed by an exchange of halide anions. The nanowires formation sequence includes intermediate phases such as MAI-PbI2-DMF and FAI-PbI2-DMF (DMF = dimethylformamide) acting as structure directing agent. The 1D shape of the adduct is preserved during the conversion to perovskite. The adducts play the role of key precursors controlling the final product morphology. Systematic investigations of the observed phase transformations and morphology features on multiple length scales revealed the effectiveness of the suggested synthetic route utilizing an original pseudomorph formation mechanism of the 1D structures to produce partly oriented films and textured layers of the nanowires via only a few experimental steps.

No abstract

The stochastic variance-reduced gradient method (SVRG) and its accelerated variant (Katyusha) have attracted enormous attention in the machine learning community in the last few years due to their superior theoretical properties and empirical behaviour on training supervised machine learning models via the empirical risk minimization paradigm. A key structural element in both of these methods is the inclusion of an outer loop at the beginning of which a full pass over the training data is made in order to compute the exact gradient, which is then used in an inner loop to construct a variance-reduced estimator of the gradient using new stochastic gradient information. In this work we design loopless variants of both of these methods. In particular, we remove the outer loop and replace its function by a coin flip performed in each iteration designed to trigger, with a small probability, the computation of the gradient. We prove that the new methods enjoy the same superior theoretical convergence properties as the original methods. For loopless SVRG, the same rate is obtained for a large interval of coin flip probabilities, including the probability 1 /n, where n is the number of functions. This is the first result where a variant of SVRG is shown to converge with the same rate without the need for the algorithm to know the condition number, which is often unknown or hard to estimate correctly. We demonstrate through numerical experiments that the loopless methods can have superior and more robust practical behavior.Preprint. Under review.

No abstract

No abstract

Solution combustion synthesis (SCS) is typically used to produce nanostructured oxides and bulk metallic materials for a variety of application including catalysis. Here, we report in situ, one-step SCS of high surface area (155 m 2 /g) Ni catalysts supported on fumed silica (SiO 2 ). Time-resolved X-ray diffraction is used to investigate the dynamics of phase formation during combustion of nickel nitrate−glycine−ammonium nitrate reactive gels impregnated onto porous SiO 2 . It is shown that highly dispersed nickel nanoparticles (5 nm) formed in the reaction front are followed by their rapid oxidation by air oxygen. To prevent the undesired oxidation process, the synthesis was conducted in an inert atmosphere (argon, helium). It is demonstrated that low concentration oxygen impurity (less than 0.001 wt %) in the inert gas passivates the Ni nanoparticles through the formation of a thin amorphous oxide layer. The thus prepared Ni/SiO 2 supported catalyst possesses high activity during the ethanol decomposition toward hydrogen at low temperatures (200°C) and excellent stability toward deactivation with essentially no change of catalyst activity over 100 h of operation.

No abstract

No abstract