Search for: All records

Optical multilayer thin film structures are widely used in many photonic applica- tions, including filters, absorbers, photovoltaics, display devices. The important part to enable these applications is the inverse design, which seeks to identify a suitable structure that satisfy desired optical responses. Recently, a Foundation model-based OptoGPT is proposed and has shown great potential to solve a wide range of inverse design problems. However, OptoGPT fails to design certain types of optical responses that are important to practical applications. The major rea- son is that the training data is randomly sampled and it is highly probable that these design targets are not selected in training, leading to the out-of-distribution issue. In this work, we propose a self-improving data augmentation technique by leveraging neural networks’ extrapolation ability. Using this method, we show sig- nificant improvement in various application design tasks with minimum fine-tuning. The approach can be potentially generalized to other inverse scientific foundation models. 

High-quality and brilliant structural colors have been successfully produced using solution-based process over the past decade. 

Iridescent color-shift pigments have been used in some industrial applications, e.g., for cosmetics and packaging. To achieve environmental-friendly and lasting color, thin-film interference is used to generate structural color. By maximizing the refractive index (RI) difference between the thin films (i.e., using an ultralow RI film), super-iridescent structural color can be produced. While the lowest refractive index of a naturally occurring solid dielectric is close to 1.37 (i.e., MgF₂), we synthesized highly porous dielectric SiO₂aerogel to achieve ultralow-RI (n ~ 1.06) and demonstrated a high-refractive index/low-refractive index/absorber (HLA) trilayer structural color. The achieved structural color is highly iridescent and capable of tracing a near-closed loop in CIE color space. By tuning the refractive index, thickness, and geometry of the aerogel layer, we control the reflection dip’s shape, therefore producing a wide range of vivid and iridescent colors. 

Optical multilayer thin film structures have been widely used in numerous photonic applications. However, existing in- verse design methods have many drawbacks because they either fail to quickly adapt to different design targets, or are difficult to suit for different types of structures, e.g., designing for different materials at each layer. These methods also cannot accommodate versatile design situations under different angles and polarizations. In addition, how to benefit practical fabrications and manufacturing has not been extensively considered yet. In this work, we introduce OptoGPT (Opto Generative Pretrained Transformer), a decoder-only transformer, to solve all these drawbacks and issues simultaneously. 

Abstract Printed electronics have made remarkable progress in recent years and inkjet printing (IJP) has emerged as one of the leading methods for fabricating printed electronic devices. However, challenges such as nozzle clogging, and strict ink formulation constraints have limited their widespread use. To address this issue, a novel nozzle‐free printing technology is explored, which is enabled by laser‐generated focused ultrasound, as a potential alternative printing modality called Shock‐wave Jet Printing (SJP). Specifically, the performance of SJP‐printed and IJP‐printed bottom‐gated carbon nanotube (CNT) thin film transistors (TFTs) is compared. While IJP required ten print passes to achieve fully functional devices with channel dimensions ranging from tens to hundreds of micrometers, SJP achieved comparable performance with just a single pass. For optimized devices, SJP demonstrated six times higher maximum mobility than IJP‐printed devices. Furthermore, the advantages of nozzle‐free printing are evident, as SJP successfully printed stored and unsonicated inks, delivering moderate electrical performance, whereas IJP suffered from nozzle clogging due to CNT agglomeration. Moreover, SJP can print significantly longer CNTs, spanning the entire range of tube lengths of commercially available CNT ink. The findings from this study contribute to the advancement of nanomaterial printing, ink formulation, and the development of cost‐effective printable electronics. 

Abstract High‐quality‐factor microring resonators are highly desirable in many applications. Fabricating a microring resonator typically requires delicate instruments to ensure a smooth side wall of waveguides and 100‐nm critical feature size in the coupling region. In this work, a new method “damascene soft nanoimprinting lithography” is demonstrated that can create high‐fidelity waveguide by simply backfilling an imprinted cladding template with a high refractive index polymer core. This method can easily realize high Q‐factor polymer microring resonators (e.g., ≈5 × 10⁵around 770 nm wavelength) without the use of any expensive instruments and can be conducted in a normal lab environment. The high Q‐factors can be attributed to the residual layer‐free feature and controllable meniscus cross‐section profile of the filled polymer core. Furthermore, the new method is compatible with different polymers, yields low fabrication defects, enables new functionalities, and allows flexible substrate. These benefits can broaden the applicability of the fabricated microring resonator.