Optical cavities play crucial roles in enhanced light–matter interaction, light control, and optical communications, but their dimensions are limited by the material property and operating wavelength. Ultrathin planar cavities are urgently in demand for large-area and integrated optical devices. However, extremely reducing the planar cavity dimension is a critical challenge, especially at telecommunication wavelengths. Herein, we demonstrate a type of ultrathin cavities based on large-area grown Bi₂Te₃ topological insulator (TI) nanofilms, which present distinct optical resonance in the near-infrared region. The result shows that the Bi₂Te₃ TI material presents ultrahigh refractive indices of >6 at telecommunication wavelengths. The cavity thickness can approach 1/20 of the resonance wavelength, superior to those of planar cavities based on conventional Si and Ge high refractive index materials. Moreover, we observed an analog of the electromagnetically induced transparency (EIT) effect at telecommunication wavelengths by depositing the cavity on a photonic crystal. The EIT-like behavior is derived from the destructive interference coupling between the nanocavity resonance and Tamm plasmons. The spectral response depends on the nanocavity thickness, whose adjustment enables the generation of obvious Fano resonance. The experiments agree well with the simulations. This work will open a new door for ultrathin cavities and applications of TI materials in light control and devices.

Based on the transverse-longitudinal mapping of Bessel beams, we propose a simple method to construct a self-similar Bessel-like beam whose transverse profile maintains a stretched form during propagation. Specifically, the propagating-variant width of this beam can be flexibly predesigned. We experimentally demonstrate three types of self-similar Bessel-like beams whose width variations are linear, piecewise, and period functions of propagation distance, respectively. The experimental results match well with the theoretical predictions. We also demonstrate that our approach enables the generation of self-similar higher-order vortex Bessel-like beams.

The conversion-efficiency for second-harmonic (SH) in optical fibers is significantly limited by extremely weak second-order nonlinearity of fused silica, and pulse pump lasers with high peak power are widely employed. Here, we propose a simple strategy to efficiently realize the broadband and continuous wave (CW) pumped SH, by transferring a crystalline GaSe coating onto a microfiber with phase-matching diameter. In the experiment, high efficiency up to 0.08 %W^-1mm^-1 is reached for a C-band pump laser. The high enough efficiency not only guarantees SH at a single frequency pumped by a CW laser, but also multi-frequencies mixing supported by three CW light sources. Moreover, broadband SH spectrum is also achieved under the pump of a superluminescent light-emitting diode source with a 79.3 nm bandwidth. The proposed scheme provides a beneficial method to the enhancement of various nonlinear parameter processes, development of quasi-monochromatic or broadband CW light sources at new wavelength regions.

Graphene and related two-dimensional materials have attracted great research interests due to prominently optical and electrical properties and flexibility in integration with versatile photonic structures. Here, we report an in-fiber photoelectric device by wrapping a few-layer graphene and bonding a pair of electrodes onto a tilted fiber Bragg grating (TFBG) for photoelectric and electric-induced thermo-optic conversions. The transmitted spectrum from this device consists of a dense comb of narrowband resonances that provides an observable window to sense the photocurrent and the electrical injection in the graphene layer. The device has a wavelength-sensitive photoresponse with responsivity up to 11.4 A/W, allowing the spectrum analysis by real-time monitoring of photocurrent evolution. Based on the thermal-optic effect of electrical injection, the graphene layer is energized to produce a global red-shift of the transmission spectrum of the TFBG, with a high sensitivity approaching 2.167×10⁴ nm/A². The in-fiber photoelectric device, therefore as a powerful tool, could be widely available as off-the-shelf product for photodetection, spectrometer and current sensor.

The time-delay problem, which is introduced by the response time of hardware for correction, is a critical and non-ignorable problem of adaptive optics (AO) systems. It will result in significant wavefront correction errors while turbulence changes severely or system responses slowly. Predictive AO is proposed to alleviate the time-delay problem for more accurate and stable corrections in the real time-varying atmosphere. However, the existing prediction approaches either lack the ability to extract non-linear temporal features, or overlook the authenticity of spatial features during prediction, leading to poor robustness in generalization. Here, we propose a mixed graph neural network (MGNN) for spatiotemporal wavefront prediction. The MGNN introduces the Zernike polynomial and takes its inherent covariance matrix as physical constraints. It takes advantage of conventional convolutional layers and graph convolutional layers for temporal feature catch and spatial feature analysis, respectively. In particular, the graph constraints from the covariance matrix and the weight learning of the transformation matrix promote the establishment of a realistic internal spatial pattern from limited data. Furthermore, its prediction accuracy and robustness to varying unknown turbulences, including the generalization from simulation to experiment, are all discussed and verified. In experimental verification, the MGNN trained with simulated data can achieve an approximate effect of that trained with real turbulence. By comparing it with two conventional methods, the demonstrated performance of the proposed method is superior to the conventional AO in terms of root mean square error (RMS). With the prediction of the MGNN, the mean and standard deviation of RMS in the conventional AO are reduced by 54.2% and 58.6% at most, respectively. The stable prediction performance makes it suitable for wavefront predictive correction in astronomical observation, laser communication, and microscopic imaging.