Gated Recurrent Neural Network for Predicting the Plasmonic Colloid Composition from Spectra.

In current research on the synthesis of colloidal nanostructures, the size and morphology of nanoparticles still exhibit certain dispersion and variation from batch to batch. Characterization of size distribution and morphology distribution of nanoparticles often requires techniques such as scanning electron microscopy or transmission electron microscopy, which involve high vacuum environments, are time-consuming, and costly. Experienced researchers can roughly estimate the size and distribution of nanostructure from spectra for a given synthetic route, but the accuracy is often limited. This paper reports the potential of using neural networks to accurately predict the composition of colloidal nanostructures from spectra. We address several fundamental issues in neural network prediction of colloidal composition. We first demonstrate the prediction of the composition of a colloidal binary mixture of gold nanoparticles using a gated recurrent neural network (GRU). The evolution of prediction errors for scattering, absorption, and extinction spectra of nanostructures with sizes ranging from 5 to 120 nm are analyzed. Furthermore, we demonstrate that the neural network model operates robustly under white noise in experimental testing scenarios. Compared to fully connected neural networks, the gated recurrent unit exhibits better testing accuracy in spectral prediction. When confronted with experimental data that deviates from simulation outputs, minor adjustments to the training set can allow the predictions to align closely with the experimental spectra, paving the way for the characterization of complex colloidal compositions with artificial intelligence.

Nanoscale Valley Modulation by Surface Plasmon Interference.

Excitons in two-dimensional (2D) materials have attracted the attention of the community to develop improved photoelectronic devices. Previous reports are based on direct excitation where the out-of-plane illumination projects a uniform single-mode light spot. However, because of the optical diffraction limit, the minimal spot size is a few micrometers, inhibiting the precise manipulation and control of excitons at the nanoscale level. Herein, we introduced the in-plane coherent surface plasmonic interference (SPI) field to excite and modulate excitons remotely. Compared to the out-of-plane light, a uniform in-plane SPI suggests a more compact spatial volume and an abundance of mode selections for a single or an array of device modulation. Our results not only build up a fundamental platform for operating and encoding the exciton states at the nanoscale level but also provide a new avenue toward all-optical integrated valleytronic chips for future quantum computation and information applications.

Centrifugation-induced assembly of dense hotspots based SERS substrate for enhanced Raman scattering and quenched fluorescence.

Hanging (aggregation stuck to the centrifugal tube) in the centrifugation process is always regarded as an unwanted condition. In this work, we develop a centrifugation-induced assembly of dense hotspots surface-enhanced Raman scattering (SERS) substrates from the hanging phenomenon. We discovered interesting sintering-resistant behavior (maintain the sharp nanotip features) of star-like Au nanoparticles after centrifugation-induced assembly, which is in stark contrast with the sintering phenomenon of sphere-like nanoparticles. We also found that one side of centrifugal-induced Au assemblies is two-dimensional (2D, root mean square (rms) roughness down to ∼10 nm), while the other is three-dimensional (3D, rms roughness more than 100 nm). The close-packed feature of the Au assemblies makes them candidates as dense hotspots based SERS substrates. Through systematic investigation of SERS performance of centrifugation-induced assemblies with different morphology (star-like and sphere-like, 2D and 3D), it was found that the 3D side of star-like Au nanoparticles assembly exhibits the highest SERS enhancement together with quenched fluorescence. The star-like SERS substrate also displays high detection uniformity (with 10-7M Rhodamine 6G) and a low detection limit (down to 10-12M Rhodamine 6G).

Electrospinning Ag-TiO2 Nanorod-Loaded Air Treatment Filters and Their Applications in Air Purification.

The efficient treatment of the problem of air pollution is a practical issue related to human health. The development of multi-functional air treatment filters, which can remove various kinds of pollutants, including particulate matter (PM) and organic gases, is a tireless pursuit aiming to address the actual needs of humans. Advanced materials and nano-manufacturing technology have brought about the opportunity to change conventional air filters for practical demands, with the aim of achieving the high-efficiency utilization of photons, a strong catalytic ability, and the synergetic degradation of multi-pollutants. In this work, visible-responding photocatalytic air treatment filters were prepared and combined with a fast and cost-effective electrospinning process. Firstly, we synthesized Ag-loaded TiO2 nanorod composites with a controlled size and number of loaded Ag nanoparticles. Then, multi-functional air treatment filters were designed by loading catalysts on electrospinning nanofibers combined with a programmable brush. We found that such Ag-TiO2 nanorod composite-loaded nanofibers displayed prominent PM filtration (~90%) and the degradation of organic pollutants (above 90%). The superior performance of purification could be demonstrated in two aspects. One was the improvement of the adsorption of pollutants derived from the increase of the specific surface area after the loading of catalysts, and the other was the plasmonic hot carriers, which induced a broadening of the optical absorption in the visible light range, meaning that many more photons were utilized effectively. The designed air treatment filters with synergistic effects for eliminating both PM and organic pollutants have promising potential for the future design and application of novel air treatment devices.

Investigation of simultaneously existed Raman scattering enhancement and inhibiting fluorescence using surface modified gold nanostars as SERS probes.

One of the main challenges for highly sensitive surface-enhanced Raman scattering (SERS) detection is the noise interference of fluorescence signals arising from the analyte molecules. Here we used three types of gold nanostars (GNSs) SERS probes treated by different surface modification methods to reveal the simultaneously existed Raman scattering enhancement and inhibiting fluorescence behaviors during the SERS detection process. As the distance between the metal nanostructures and the analyte molecules can be well controlled by these three surface modification methods, we demonstrated that the fluorescence signals can be either quenched or enhanced during the detection. We found that fluorescence quenching will occur when analyte molecules are closely contacted to the surface of GNSs, leading to a ~100 fold enhancement of the SERS sensitivity. An optimized Raman signal detection limit, as low as the level of 10-11 M, were achieved when Rhodamine 6 G were used as the analyte. The presented fluorescence-free GNSs SERS substrates with plentiful hot spots and controllable surface plasmon resonance wavelengths, fabricated using a cost-effective self-assembling method, can be very competitive candidates for high-sensitive SERS applications.