Wenjing Wu

Complex refractive indices of materials encode fundamental information on light-matter interactions and are critical for the design of advanced photonic and optoelectronic devices. In many emerging materials, refractive indices change under external stimuli such as temperature, electric fields, or strain. Tracking these changes in-operando is critical for active photonic and optoelectronic device design, but remains challenging. Conventional methods such as ellipsometry rely on labor-intensive model fitting and are often impractical for multilayer stacks or in-operando measurements. Optical reflectometry offers a simpler alternative but suffers from ambiguous extraction of refractive index from reflectance spectra and limited applicability under dynamic modulation. Here, we present ReflectoRNN, an artificial intelligence (AI)-powered reflectometry framework based on recurrent neural networks (RNN), for real-time extraction of complex refractive indices in evolving materials. ReflectoRNN extracts refractive indices from reflectance spectra under thermal, electrical, magnetic, or mechanical stimuli. It achieves a median Pearson’s correlation coefficient (PCC) of 0.998 and a relative accuracy score (RAS) of 0.968 on generated datasets. Validation experiments on MoS₂ and WS₂ across diverse substrates, including single-layer and multilayer dielectric stacks, and distributed Bragg reflectors (DBRs), demonstrate high accuracy and physical consistency, with temperature-dependent exciton resonance energy matching Bose–Einstein predictions. ReflectoRNN enables in-operando optical characterization of materials across complex photonic structures and offers a pathway toward automated, real-time monitoring and accelerated materials discovery.

Surface-enhanced Raman spectroscopy (SERS) offers high sensitivity for biomolecular detection, but its performance is often constrained by noise arising from signal non-uniformity across substrates. Here, we introduce a noise-management−oriented design strategy for hybrid SERS substrates composed of gold nanoparticles (AuNP) and two-dimensional (2D) materials (graphene, MoS₂, and WSe₂). Compared with conventional AuNP substrates, the hybrids exhibit markedly improved spectral uniformity and signal-to-noise ratio (SNR), with the AuNP/graphene platform reducing noise by ∼67% and increasing SNR by ∼279%. Full-wave simulations based on Maxwell’s equations corroborated the experimental results and reveal that optical constants of the 2D material and nanoparticle distribution jointly govern noise characteristics. SNR dependence on nanoparticle density distributions, refractive index (n), and extinction coefficient (k) is further established. As a practical demonstration, the AuNP/graphene substrate enabled detection of the receptor binding domain protein at a limit of detection (LOD) of 10⁻⁹ M, representing a ten-fold improvement over the 10⁻⁸ M LOD of AuNP substrates. These results establish AuNP/2D hybrids as effective platforms for noise-managed SERS, offering enhanced sensitivity for biosensing.

Phonons play a central role in fundamental solid-state phenomena, including superconductivity, Raman scattering, and symmetry-breaking phases. Harnessing phonons to control these effects and enable quantum technologies is therefore of great interest. However, most existing phonon control strategies rely on external driving fields or anharmonic interactions, limiting their applicability. Here, we realize multimode ultrastrong light–matter coupling and theoretically show the modulation of phonon emission. This regime is realized by coupling two optical phonon modes in lead halide perovskites to a nanoslot array functioning as a single-mode cavity. The small mode volume of the nanoslots enables high coupling strengths in the phonon-polariton system. We show theoretically that the nanoslot resonator mediates an effective interaction between phonon modes, leading to superthermal phonon bunching in thermal equilibrium between distinct modes. Our findings are well described by a multimodal Hopfield model. This work establishes a pathway for engineering phononic properties for light-harvesting and light-emitting technologies.

Significant progress has been made in growing large-area transition metal dichalcogenide (TMD) monolayers, but most rely on stitching small flakes, which can affect electronic and optical performance. High-quality single-crystal monolayers are needed for device applications. Here, we report a wafer-bonder-assisted transfer (WBAT) method to produce uniform, crack-free, single-crystal TMD monolayers over large areas. This technique integrates gold-tape exfoliation with wafer-scale bonding, enabling flake sizes over 10⁶ times larger than those from traditional Scotch tape exfoliation. Compared to hand-press transfer, WBAT yields higher-quality monolayers with fewer cracks, reduced strain, and over 2-fold improvement in photoluminescence uniformity, as confirmed by Raman and PL mapping. Fabricated field-effect transistor arrays show high mobility and on-off ratio. The WBAT method offers high yield, reproducibility, and compatibility with various 2D materials, heterostructures, and substrates, aligning with standard semiconductor processes. It provides a scalable solution for integrating high-quality TMD monolayers into next-generation electronic and optoelectronic devices.

Deterministic spatial control of material properties is essential for advanced electronic and optoelectronic device technologies. van der Waals (vdW) materials stand out for their high tunability, yet achieving multifunctional on-chip control remains challenging. Here, we focus on α-MoO₃ and site-selectively modulate both its optical emission and conductivity via electron-beam irradiation. In situ cathodoluminescence spectroscopy reveals a cumulative enhancement of blue emission under the beam, exhibiting superlinear increases in intensity with beam current. Irradiated regions show distinct contrasts in Raman and work functions, indicating oxygen defect formation. Moreover, these local structural modifications substantially increase lateral and vertical conductivities, enabling conductive channels with sharp boundaries. Building on this, we demonstrate stable, deterministic optical patterning with subdiffraction spatial resolution, highlighting the potential for submicron devices without chemical processes or traditional lithography. Our results provide a versatile platform for on-chip light sources, conductive interconnects, and customizable optoelectronic elements, expanding the design space of vdW materials.

Single-photon emitters (SPEs) in two-dimensional WSe₂ offer high extraction efficiency and on-chip compatibility, but achieving high purity remains challenging. We present two strategies to suppress classical emission and enhance purity in WSe₂-based SPEs. In monolayer WSe₂, we exploited the presence and absence of valley-spin locking in free and bound excitons, respectively, to achieve purity of 98.3% via polarization control and 99.0% combined with near-resonant excitation. In bilayer WSe₂, we obtained 97.0% purity without polarization filtering, enabled by the indirect band gap and inversion symmetry. These values represent some of the highest as-measured purities reported for 2D TMD SPEs. Our methods do not require complex fabrication or instrumentation and are supported by first-principles calculations of the vacancy state of Se and spin degeneracy. This work offers practical pathways for realizing high-quality single-photon sources for emerging quantum technologies.

Optical spectroscopy, a noninvasive molecular sensing technique, offers valuable insights into material characterization, molecule identification, and biosample analysis. Despite the informativeness of high-dimensional optical spectra, their interpretation remains a challenge. Machine learning methods have gained prominence in spectral analyses, efficiently unveiling analyte compositions. However, these methods still face challenges in interpretability, particularly in generating clear feature importance maps that highlight the spectral features specific to each class of data. These limitations arise from feature noise, model complexity, and the lack of optimization for spectroscopy. In this work, we introduce a machine learning algorithm─logistic regression with peak-sensitive elastic-net regularization (PSE-LR)─tailored for spectral analysis. PSE-LR enables classification and interpretability by producing a peak-sensitive feature importance map, achieving an F1-score of 0.93 and a feature sensitivity of 1.0. Its performance is compared with other methods, including k-nearest neighbors (KNN), elastic-net logistic regression (E-LR), support vector machine (SVM), principal component analysis followed by linear discriminant analysis (PCA-LDA), XGBoost, and neural network (NN). Applying PSE-LR to Raman and photoluminescence (PL) spectra, we detected the receptor-binding domain (RBD) of SARS-CoV-2 spike protein in ultralow concentrations, identified neuroprotective solution (NPS) in brain samples, recognized WS₂ monolayer and WSe₂/WS₂ heterobilayer, analyzed Alzheimer's disease (AD) brains, and suggested potential disease biomarkers. Our findings demonstrate PSE-LR's utility in detecting subtle spectral features and generating interpretable feature importance maps. It is beneficial for the spectral characterization of materials, molecules, and biosamples and applicable to other spectroscopic methods. This work also facilitates the development of nanodevices such as nanosensors and miniaturized spectrometers based on nanomaterials.

Optical spectroscopy is indispensable for research and development in nanoscience and nanotechnology, microelectronics, energy, and advanced manufacturing. Advanced optical spectroscopy tools often require both specifically designed high-end instrumentation and intricate data analysis techniques. Beyond the common analytical tools, deep learning methods are well suited for interpreting high-dimensional and complicated spectroscopy data. They offer great opportunities to extract subtle and deep information about optical properties of materials with simpler optical setups, which would otherwise require sophisticated instrumentation. In this work, we propose a computational approach based on a conventional tabletop optical microscope and a deep learning model called ReflectoNet. Without any prior knowledge about the multilayer substrates, ReflectoNet can predict the complex refractive indices of thin films and 2D materials on top of these nontrivial substrates from experimentally measured optical reflectance spectra with high accuracies. This task was not feasible previously with traditional reflectometry or ellipsometry methods. Fundamental physical principles, such as the Kramers–Kronig relations, are spontaneously learned by the model without any further training. This approach enables in-operando optical characterization of functional materials and 2D materials within complex photonic structures or optoelectronic devices.