The main challenges to the commercial viability of polymer electrolyte membrane fuel cells are (i) the high cost associated with using large amounts of Pt in fuel cell cathodes to compensate for the sluggish kinetics of the oxygen reduction reaction, (ii) catalyst degradation, and (iii) carbon-support corrosion. To address these obstacles, our group has focused on robust, carbon-free transition metal nitride materials with low Pt content that exhibit tunable physical and catalytic properties. Here, we report on the high performance of a novel catalyst with low Pt content, prepared by placing several layers of Pt atoms on nanoparticles of titanium nickel binary nitride. For the ORR, the catalyst exhibited a more than 400% and 200% increase in mass activity and specific activity, respectively, compared with the commercial Pt/C catalyst. It also showed excellent stability/durability, experiencing only a slight performance loss after 10,000 potential cycles, while TEM results showed its structure had remained intact. The catalyst's outstanding performance may have resulted from the ultrahigh dispersion of Pt (several atomic layers coated on the nitride nanoparticles), and the excellent stability/durability may have been due to the good stability of nitride and synergetic effects between ultrathin Pt layer and the robust TiNiN support.

Edge caching is an emerging technology for addressing massive content access in mobile networks to support rapidly growing Internet of Things (IoT) services and applications. However, most current optimization-based methods lack a self-adaptive ability in dynamic environments. To tackle these challenges, current learning-based approaches are generally proposed in a centralized way. However, network resources may be overconsumed during the training and data transmission process. To address the complex and dynamic control issues, we propose a FederAted Deep reinforcement learning-based cooperative Edge caching (FADE) framework. FADE enables base stations (BSs) to cooperatively learn a shared predictive model by considering the first-round training parameters of the BSs as the initial input of the local training, and then uploads near-optimal local parameters to the BSs to participate in the next round of global training. Furthermore, we prove the convergence of the proposed FADE, and it achieves the expectation convergence. Trace-driven simulation results show that the proposed FADE framework reduces 92% of performance loss and average 60% system payment over the centralized deep reinforcement learning (DRL) algorithm, achieves only a 4% performance loss of the desirable omniscient oracle algorithm, and obtains 7%, 11% and 9% network performance improvements compared to some existing schemes, i.e., least recently used (LRU), least frequently

A series of ZIF-derived Fe-N codoped carbon materials with a well-defined morphology, high surface area, tunable sizes and porous nanoframe structure was successfully prepared by synthesizing Fe-doped ZIF-8 through the assembly of Zn ions with 2-methylimidazole in the presence of iron(III) acetylacetonate, followed by pyrolysis at a high temperature and in an Ar atmosphere. The prepared optimum catalyst materials exhibited excellent activity for the oxygen reduction reaction (ORR) and outstanding durability in both acidic and alkaline solutions. We found that Fe doping during the ZIF-8 synthesis stage was crucial to achieve the materials' well-defined morphology, tunable size, good particle dispersion, and high performance. XPS revealed that Fe doping greatly enhanced the fractions of graphitic-N and pyridinic-N and decreased the fraction of oxidized-N. We suggest that the porosity and high surface area of the nanoframe structure originated from the metal-organic frameworks, the high dispersion of Fe in the nanoframe, and the enhanced proportions of active N species, all of which were responsible for the materials' significantly enhanced ORR performance.

Ghrelin, an endogenous ligand of the growth hormone secretagogue receptor, has been reported to have beneficial effects on cardiac function. The authors used the Langendorff model of ischemia/reperfusion (I/R) injury in isolated rat heart to determine whether ghrelin exerts direct cardioprotective effects. Also, the capacity of ghrelin to bind to sarcolemmal membrane fractions before and after ischemia and reperfusion was examined. Compared with vehicle administration, administration of ghrelin (100-10,000 pM) during the reperfusion period resulted in improvement in coronary flow, heart rate, left ventricular systolic pressure, and left ventricular end-diastolic pressure. Ghrelin also enhanced the rates of left ventricular contraction and relaxation after ischemia following reperfusion. Administration of ghrelin during reperfusion reduced myocardial release of lactate dehydrogenase and myoglobin, indicating protection against cardiomyocyte injury. In addition, ghrelin attenuated the depletion of myocardial ATP resulting from ischemia and reperfusion. A receptor-binding assay demonstrated that maximum binding capacity of ghrelin to sarcolemmal membranes was significantly increased after ischemia and was further increased after I/R. However, Scatchard analysis showed that the affinity of ghrelin for its receptor was not altered. The authors have concluded that administration of ghrelin during reperfusion protects against myocardial I/R injury. The cardioprotective effects are independent of growth hormone release and likely involve binding to cardiovascular receptors, a process that is upregulated during I/R.