Nanohybrids based on biomolecular nanostructures and graphene quantum dots (GQDs) have found wide application in the biological and biomedical fields. Herein, the design of a peptide with trifunctional motifs is reported as the precursor building block for constructing a novel multifunctional protein nanofiber (PNF), and further conjugated with highly fluorescent GQDs by noncovalent interactions. The physicochemical properties of these PNF–GQD nanohybrids are thoroughly characterized by a variety of spectroscopic and microscopic techniques, revealing that the GQDs essentially maintain their favorable optical properties in the nanohybrids. A good biocompatibility of the PNF–GQD nanohybrids is found with cell viability assays. With both, a recognition moiety (RGD) and an imaging probe (GQD), these PNF–GQD nanohybrids possess the capability of targeting and imaging tumor cells simultaneously. A potential application of these novel nanohybrids, i.e., fluorescence imaging of HeLa tumor cells, has been investigated by confocal fluorescence microscopy, which shows much enhanced labeling efficiency compared with GQDs only. Moreover, cellular internalization by nontumorous COS‐7 cells was much weaker than by HeLa cells. Our results show that GQD‐decorated PNF nanohybrids have great potential as multifunctional platforms for biomedical applications, particularly, where the capability of sensitive tracking and efficient labeling is appreciated.

Crack-free amorphous photonic crystals, showing high color visibility and low angle dependence, have been fabricated by simple co-assembly.

In this work, hollow polydopamine (PDA) amorphous colloidal structures with liquid-immune and angle-independent structural colors have been successfully fabricated. Owing to the high refractive index contrast derived from hollow PDA, and its light-absorbing nature, the as-prepared colloidal materials demonstrate brilliant structural colors even in liquid environments.

Large amounts of data has made neural machine translation (NMT) a big success in recent years. But it is still a challenge if we train these models on small-scale corpora. In this case, the way of using data appears to be more important. Here, we investigate the effective use of training data for low-resource NMT. In particular, we propose a dynamic curriculum learning (DCL) method to reorder training samples in training. Unlike previous work, we do not use a static scoring function for reordering. Instead, the order of training samples is dynamically determined in two ways -loss decline and model competence. This eases training by highlighting easy samples that the current model has enough competence to learn. We test our DCL method in a Transformerbased system. Experimental results show that DCL outperforms several strong baselines on three low-resource machine translation benchmarks and different sized data of

A high power and efficient 588 nm yellow light is demonstrated through intracavity frequency doubling of an acousto-optic Q-switched self-frequency Raman laser. A 30-mm-length double-end diffusion-bonded Nd:YVO(4) crystal was utilized for efficient self-Raman laser operation by reducing the thermal effects and increasing the interaction length for the stimulated Raman scattering. A 15-mm-length LBO with non-critical phase matching (theta = 90 degrees, phi = 0 degrees) cut was adopted for efficient second-harmonic generation. The focus position of incident pump light and both the repetition rate and the duty cycle of the Q-switch have been optimized. At a repetition rate of 110 kHz and a duty cycle of 5%, the average power of 588 nm light is up to 7.93 W while the incident pump power is 26.5 W, corresponding to an overall diode-yellow conversion efficiency of 30% and a slope efficiency of 43%.