We demonstrate the improved thermoelectric properties of n-type lead-free transformable AgBiSe2 polymorphs by indium doping on silver sites. X-ray diffraction analysis suggests that complete solid solutions are well formed up to [In] = 0.02. Electrical conductivity and Seebeck coefficient behave in a routinely opposite manner due to the dominant role of the carrier concentration adjusted by the localized indium impurity levels, as also suggested by our density functional theory (DFT) calculations. As indium concentration increases, we observe a drastic variation of the thermoelectric transport properties with temperature, in the range of 450 to 580 K. By performing the isothermal electrical measurements, we attribute this interesting behavior to the ongoing α to β phase transformation process. The In 5s lone pair electrons, as indicated from our DFT calculations, increase the anharmonicity of the chemical bonds and enhance the phonon-phonon scattering. This, together with the introduced InAg.. point defects, further brings down the lattice thermal conductivity. The maximum thermoelectric figure of merit ZT is achieved at 773 K and increases from 0.3 for pristine AgBiSe2 to 0.7 for an optimal [In] = 0.015 doping, a more than two times enhancement.

Cu2S compounds are promising thermoelectric (TE) candidate materials with environmentally friendly and earth abundant chemical constituents. A series of phase transitions occur with temperature whereas only the high temperature stabilized cubic structure (α‐Cu2S) exhibits desirable TE properties. In this work, by alloying Cu sites with Mn, Zn, Ga, and Ge, profound influence on β‐ to α‐Cu2S phase transition and thermoelectric transport properties is observed. Both phase transition temperature (Tc) and the enthalpy of phase change (ΔH) decreases with doping; remarkably, for Cu1.95Mn0.03S, Tc reduces by ≈156 K. The Seebeck anomaly near the critical point of phase transition also vanishes. The electrical conductivity is remarkably improved for doped samples due to the largely elevated hole concentration. In comparison with pristine Cu2S, not only is the peak TE power factor substantially enhanced (by ≈272%), but also the average ZT for 500–823 K is highly improved (by ≈145%) due to the successful stabilization of α‐Cu2S at lower temperatures. The present work offers a clue to enlarge the temperature regime of high TE properties, which is practically useful for a variety of polymorphous thermoelectric compounds.

No abstract

Traditional machine learning relies on a centralized data pipeline for model training in various applications; however, data are inherently fragmented. Such a decentralized nature of databases presents the serious challenge for collaboration: sending all decentralized datasets to a central server raises serious privacy concerns. Although there has been a joint effort in tackling such a critical issue by proposing privacy-preserving machine learning frameworks, such as federated learning, most state-of-the-art frameworks are built still in a centralized way, in which a central client is needed for collecting and distributing model information (instead of data itself) from every other client, leading to high communication burden and high vulnerability when there exists a failure at or an attack on the central client. Here we propose a principled decentralized federated learning algorithm (DeceFL), which does not require a central client and relies only on local information transmission between clients and their neighbors, representing a fully decentralized learning framework. It has been further proven that every client reaches the global minimum with zero performance gap and achieves the same convergence rate O(1/T ) (where T is the number of iterations in gradient descent) as centralized federated learning when the loss function is smooth and strongly convex. Finally, the proposed algorithm has been applied to a number of applications to illustrate its effectiveness for both convex and nonconvex loss functions, time-invariant and time-varying topologies, as well as IID and Non-IID of datasets, demonstrating its applicability to a wide range of real-world medical and industrial applications.

No abstract