A current challenge to alkaline polymer electrolyte fuel cells (APEFCs) is the unexpectedly sluggish kinetics of the hydrogen oxidation reaction (HOR). A recently proposed resolution is to enhance the oxophilicity of the catalyst, so as to remove the H ad intermediate through reacting with OH ad , but this approach is questioned by other researchers.Here we report a clear and convincing test on this problem. By using PtRu/C as the HOR catalyst for APEFC, the peak power density is boosted to 1.0 W/cm 2 , in comparison to 0.6 W/cm 2 when using Pt/C in the anode. Such a remarkable improvement, however, can hardly be explained as an oxophilic effect, because, as monitored by CO stripping, reactive hydroxyl species can generate on certain sites of the Pt surface at more negative potentials than on the PtRu surface in KOH solution. Rather, the incorporation of Ru has posed an electronic effect on weakening the Pt-H ad interaction, as revealed by the voltammetric behavior and from density-functional calculations, which thus benefits the oxidative desorption of H ad , the rate determining step of HOR in alkaline media. These findings further our fundamental understanding of the HOR catalysis, and cast a new light on the exploration of better catalysts for APEFC. 5 process to monitor the generation of reactive hydroxyl species, for the anodic current of CO oxidation has to be triggered by reactive hydroxyl species. 31 As demonstrated in Figure 2, in 0.1 M H 2 SO 4 solution, the CO stripping on Pt/C takes on a single sharp peak at 0.85 V, and, upon alloying with Ru, the CO stripping peak is somewhat broadened and moves negatively by 0.3 V, showing that Ru does accelerate the formation of OH ad in acidic environment.However, the CO stripping on Pt/C behaves rather differently in 0.1 M KOH solution: multiple anodic peaks appear and the onset potential shifts to ~0.2 V, a potential even more negative than the onset of CO stripping on PtRu/C (~0.35 V) in either acid or alkaline media.Such a surprising finding indicates that, in alkaline environment, the reactive hydroxyl species, be it OH ad or OH ad − , can generate on certain sites of the Pt surface more favorably than on the PtRu surface; but when alloyed with Ru, the surface reactivity of Pt is suppressed, thereby no reactive hydroxyl species appearing at the potential region negative to 0.35 V.On the basis of the above observations, the promotion effect of Ru on catalyzing the HOR in alkaline media can hardly be explained as an oxophilic effect. The existence of reactive hydroxyl species on either Pt or PtRu surface at potentials negative to 0.2 V also seems unlikely, as revealed by Figure 2. On the other hand, the Ru has posed an obvious effect on weakening the Pt-H ad interaction, as a consequence of the suppressed surface reactivity of Pt. As illustrated in Figure 3a, in KOH solution, the hydrogen underpotential deposition (H-UPD) and subsequent desorption behavior on PtRu/C is clearly different from that on Pt/C: whereas strong H ad peaks are the major signal for Pt/C...

Anion exchange membrane fuel cells (AEMFCs) have been developed as promising energy conversion devices for stationary and mobile applications due to their potentially low cost. To realize high-performance AEMFCs, new polymeric membranes are needed that are highly conductive and chemically stable. Here we report a systematic study of anion exchange membranes (AEMs) with multiple cations per side chain site to demonstrate how this motif can boost both the conductivity and stability of poly­(2,6-dimethyl-1,4-phenylene oxide)-based AEMs. The highest conductivity, up to 99 mS/cm at room temperature, was observed for a triple-cation side chain AEM with 5 or 6 methylene groups between cations. This conductivity was considerably higher than AEM samples based on benzyl­trimethyl­ammonium or benzyl­dimethylhexyl­ammonium groups with only one cation per side chain site. In addition to high conductivity, the multication side chain AEMs showed good alkaline and dimensional stabilities. High retention of ion exchange capacity (IEC) (93% retention) and ionic conductivity (90% retention) were observed for the triple-cation side chain AEMs in degradation testing under 1 M NaOH at 80 °C for 500 h. Based on the high-performance triple-cation side chain AEM, a Pt-catalyzed fuel cell with a peak power density of 364 mW/cm2 was achieved at 60 °C under 100% related humidity.

We propose a novel data augmentation method for labeled sentences called conditional BERT contextual augmentation. Data augmentation methods are often applied to prevent overfitting and improve generalization of deep neural network models. Recently proposed contextual augmentation augments labeled sentences by randomly replacing words with more varied substitutions predicted by language model. BERT demonstrates that a deep bidirectional language model is more powerful than either an unidirectional language model or the shallow concatenation of a forward and backward model. We retrofit BERT to conditional BERT by introducing a new conditional masked language model 1 task. The well trained conditional BERT can be applied to enhance contextual augmentation. Experiments on six various different text classification tasks show that our method can be easily applied to both convolutional or recurrent neural networks classifier to obtain obvious improvement.

The numerical results from a physics-based global magnetohydrodynamic (MHD) model are used to examine the effect of the interplanetary magnetic field (IMF), solar wind dynamic pressure, and dipole tilt angle on the size and shape of the magnetopause. The subsolar magnetopause is identified using the plasma velocity and density, the cusps are identified using the thermal pressure, and the whole shape of the magnetopause is determined with the three-dimensional streamlines traced through the simulation domain. The magnetopause surface obtained from the simulations is fitted with a three-dimensional surface function controlled by ten configuration parameters, which provide a description of the subsolar magnetopause, the cusp geometry, the flaring angle, the azimuthal asymmetry, the north-south asymmetry, and the twisting angle of the magnetopause. Effects of the IMF, solar wind dynamic pressure, and dipole tilt angle on the configuration parameters are analyzed and fitted by relatively simple functions. It is found that the solar wind dynamic pressure mainly affects the magnetopause size; the IMF mainly controls the magnetopause flaring angle, azimuthal asymmetry, and twisting angle; and the dipole tilt angle mainly affects the magnetopause north-south asymmetry and the cusp geometry. The model is validated by comparing with available empirical models and observational results, and it is demonstrated that the new model can describe the magnetopause for typical solar wind conditions.