The manipulation of crystal orientation from the thermodynamic equilibrium states is desired in layered hybrid perovskite films to direct charge transport and enhance the perovskite devices performance. Here we report a templated growth mechanism of layered perovskites from 3D-like perovskites which can be a general design rule to align layered perovskites along the out-of-plane direction in films made by both spin-coating and scalable blading process. The method involves suppressing the nucleation of both layered and 3D perovskites inside the perovskite solution using additional ammonium halide salts, which forces the film formation starts from solution surface. The fast drying of solvent at liquid surface leaves 3D-like perovskites which surprisingly templates the growth of layered perovskites, enabled by the periodic corner-sharing octahedra networks on the surface of 3D-like perovskites. This discovery provides deep insights into the nucleation behavior of octahedra-array-based perovskite materials, representing a general strategy to manipulate the orientation of layered perovskites.

With the recent advances in deep learning, neural network models have obtained state-of-the-art performances for many linguistic tasks in natural language processing. However, this rapid progress also brings enormous challenges. The opaque nature of a neural network model leads to hard-to-debug-systems and difficult-to-interpret mechanisms. Here, we introduce a visualization system that, through a tight yet flexible integration between visualization elements and the underlying model, allows a user to interrogate the model by perturbing the input, internal state, and prediction while observing changes in other parts of the pipeline. We use the natural language inference problem as an example to illustrate how a perturbation-driven paradigm can help domain experts assess the potential limitation of a model, probe its inner states, and interpret and form hypotheses about fundamental model mechanisms such as attention.

No abstract

Gold nanorods (AuNRs) are potentially useful in tumor theranostics, but the poor stability, high toxicity, and rapid removal by the immune system seriously limit their theranostic applications. In our study, we demonstrate the fabrication of Cu(II)-doped polydopamine-coated AuNR (AuNR@CuPDA), which significantly improves the potentials in tumor theranostics. Besides the improvement of physiological stability and biocompatibility, the PDA shell increases the photothermal performance and prolongs the blood circulation time of AuNRs. The half-life of AuNRs during blood circulation increases from 0.7 to 4.5 h after PDA coating, and the injected dose per gram of tumor tissue is 4.6% ID g for AuNR@CuPDA. In addition to computer tomography imaging, the loading of Cu(II) in PDA shell endows AuNR@CuPDA with magnetic resonance imaging function. Cu(II) doped in PDA shell also exhibits chemotherapeutic behavior, and the tumor inhibitor rate is 31.2%. Further combining 808 nm laser-driven photothermal therapy, tumors were completely ablated, and no recurrence was observed. Liver and renal functions tests and histological analysis of major organs confirm that AuNR@CuPDA is in good safety.

No abstract