RESUMO
Cobalt ferrite (CoFe2 O4 ) spinel has been found to produce C2 -C4 hydrocarbons in a single-step, ambient-pressure, photocatalytic hydrogenation of CO2 with a rate of 1.1â mmol g-1 h-1 , selectivity of 29.8 % and conversion yield of 12.9 %. On stream the CoFe2 O4 reconstructs to a CoFe-CoFe2 O4 alloy-spinel nanocomposite which facilitates the light-assisted transformation of CO2 to CO and hydrogenation of the CO to C2 -C4 hydrocarbons. Promising results obtained from a laboratory demonstrator bode well for the development of a solar hydrocarbon pilot refinery.
RESUMO
It has long been known that the thermal catalyst Cu/ZnO/Al2O3(CZA) can enable remarkable catalytic performance towards CO2 hydrogenation for the reverse water-gas shift (RWGS) and methanol synthesis reactions. However, owing to the direct competition between these reactions, high pressure and high hydrogen concentration (≥75%) are required to shift the thermodynamic equilibrium towards methanol synthesis. Herein, a new black indium oxide with photothermal catalytic activity is successfully prepared, and it facilitates a tandem synthesis of methanol at a low hydrogen concentration (50%) and ambient pressure by directly using by-product CO as feedstock. The methanol selectivities achieve 33.24% and 49.23% at low and high hydrogen concentrations, respectively.
RESUMO
Despite the great success of deep neural networks, the adversarial attack can cheat some well-trained classifiers by small permutations. In this paper, we propose another type of adversarial attack that can cheat classifiers by significant changes. For example, we can significantly change a face but well-trained neural networks still recognize the adversarial and the original example as the same person. Statistically, the existing adversarial attack increases Type II error and the proposed one aims at Type I error, which are hence named as Type II and Type I adversarial attack, respectively. The two types of attack are equally important but are essentially different, which are intuitively explained and numerically evaluated. To implement the proposed attack, a supervised variation autoencoder is designed and then the classifier is attacked by updating the latent variables using gradient information. Besides, with pre-trained generative models, Type I attack on latent spaces is investigated as well. Experimental results show that our method is practical and effective to generate Type I adversarial examples on large-scale image datasets. Most of these generated examples can pass detectors designed for defending Type II attack and the strengthening strategy is only efficient with a specific type attack, both implying that the underlying reasons for Type I and Type II attack are different.
RESUMO
Solar hydrogen and electricity are promising high energy-density renewable sources. Although photochemistry or photovoltaics are attractive routes, special challenge arises in sunlight conversion efficiency. To improve efficiency, various semiconductor materials have been proposed with selective sunlight absorption. Here, we reported a hybrid system synergizing photo-thermochemical hydrogen and photovoltaics, harvesting full-spectrum sunlight in a cascade manner. A simple suspension of Au-TiO2 in water/methanol serves as a spectrum selector, absorbing ultraviolet-visible and infrared energy for rapid photo-thermochemical hydrogen production. The transmitted visible and near-infrared energy fits the photovoltaic bandgap and retains the high efficiency of a commercial photovoltaic cell under different solar concentration values. The experimental design achieved an overall efficiency of 4.2% under 12 suns solar concentration. Furthermore, the results demonstrated a reduced energy loss in full-spectrum energy conversion into hydrogen and electricity. Such simple integration of photo-thermochemical hydrogen and photovoltaics would create a pathway toward cascading use of sunlight energy.