Search for: All records

The prediction of crystal properties plays a crucial role in materials science and applications. Current methods for predicting crystal properties focus on modeling crystal structures using graph neural networks (GNNs). However, accurately modeling the complex interactions between atoms and molecules within a crystal remains a challenge. Surprisingly, predicting crystal properties from crystal text descriptions is understudied, despite the rich information and expressiveness that text data offer. In this paper, we develop and make public a benchmark dataset (TextEdge) that contains crystal text descriptions with their properties. We then propose LLM-Prop, a method that leverages the general-purpose learning capabilities of large language models (LLMs) to predict properties of crystals from their text descriptions. LLM-Prop outperforms the current state-of-the-art GNN-based methods by approximately 8% on predicting band gap, 3% on classifying whether the band gap is direct or indirect, and 65% on predicting unit cell volume, and yields comparable performance on predicting formation energy per atom, energy per atom, and energy above hull. LLM-Prop also outperforms the fine-tuned MatBERT, a domain-specific pre-trained BERT model, despite having 3 times fewer parameters. We further fine-tune the LLM-Prop model directly on CIF files and condensed structure information generated by Robocrystallographer and found that LLM-Prop fine-tuned on text descriptions provides a better performance on average. Our empirical results highlight the importance of having a natural language input to LLMs to accurately predict crystal properties and the current inability of GNNs to capture information pertaining to space group symmetry and Wyckoff sites for accurate crystal property prediction. 

Abstract The sub‐Terahertz and Terahertz bands play a critical role in next‐generation wireless communication and sensing technologies, thanks to the large amount of available bandwidth in this spectral regime. While long‐wavelength (microwave to mm‐Wave) and short‐wavelength (near‐infrared to ultraviolet) devices are well‐established and studied, the sub‐THz to THz regime remains relatively underexplored and underutilized. Traditional approaches used in the aforementioned spectral regions are more difficult to replicate in the THz band, leading to the need for the development of novel devices and structures that can manipulate THz radiation effectively. Herein a novel organic, solid‐state electrochemical device is presented, capable of achieving modulation depths of over 90% from ≈500 nm of a conducting polymer that switches conductivity over a large dynamic range upon application of an electronically controllable external bias. The stability of such devices under long‐term, repeated voltage switching, as well as continuous biasing at a single voltage, is also explored. Switching stabilities and long‐term bias stabilities are achieved over two days for both use cases. Additionally, both depletion mode (always “ON”) and accumulation mode (always “OFF”) operation are demonstrated. These results suggest applications of organic electrochemical THz modulators in large area and flexible implementations. 

Abstract Neuromorphic photonics has become one of the research forefronts in photonics, with its benefits in low‐latency signal processing and potential in significant energy consumption reduction when compared with digital electronics. With artificial intelligence (AI) computing accelerators in high demand, one of the high‐impact research goals is to build scalable neuromorphic photonic integrated circuits which can accelerate the computing of AI models at high energy efficiency. A complete neuromorphic photonic computing system comprises seven stacks: materials, devices, circuits, microarchitecture, system architecture, algorithms, and applications. Here, we consider microring resonator (MRR)‐based network designs toward building scalable silicon integrated photonic neural networks (PNN), and variations of MRR resonance wavelength from the fabrication process and their impact on PNN scalability. Further, post‐fabrication processing using organic photochromic layers over the silicon platform is shown to be effective for trimming MRR resonance wavelength variation, which can significantly reduce energy consumption from the MRR‐based PNN configuration. Post‐fabrication processing with photochromic materials to compensate for the variation in MRR fabrication will allow a scalable silicon system on a chip without sacrificing today's performance metrics, which will be critical for the commercial viability and volume production of large‐scale silicon photonic circuits. 

Abstract Organic small molecules that exhibit second‐scale phosphorescence at room temperature are of interest for potential applications in sensing, anticounterfeiting, and bioimaging. However, such materials systems are uncommon—requiring millisecond to second‐scale triplet lifetimes, efficient intersystem crossing, and slow rates of nonradiative recombination. Here, a simple and scalable approach is demonstrated to activate long‐lived phosphorescence in a wide variety of molecules by suspending them in rigid polymer hosts and annealing them above the polymer's glass transition temperature. This process produces submicron aggregates of the chromophore, which suppresses intramolecular motion that leads to nonradiative recombination and minimizes triplet–triplet annihilation that quenches phosphorescence in larger aggregates. In some cases, evidence of excimer‐mediated intersystem crossing that enhances triplet generation in aggregated chromophores is found. In short, this approach circumvents the current design rules for long‐lived phosphors, which will streamline their discovery and development. 

Crystalline organic semiconductors are known to have improved charge carrier mobility and exciton diffusion length in comparison to their amorphous counterparts. Certain organic molecular thin films can be transitioned from initially prepared amorphous layers to large-scale crystalline films via abrupt thermal annealing. Ideally, these films crystallize as platelets with long-range-ordered domains on the scale of tens to hundreds of microns. However, other organic molecular thin films may instead crystallize as spherulites or resist crystallization entirely. Organic molecules that have the capability of transforming into a platelet morphology feature both high melting point (Tm) and crystallization driving force (ΔGc). In this work, we employed machine learning (ML) to identify candidate organic materials with the potential to crystallize into platelets by estimating the aforementioned thermal properties. Six organic molecules identified by the ML algorithm were experimentally evaluated; three crystallized as platelets, one crystallized as a spherulite, and two resisted thin film crystallization. These results demonstrate a successful application of ML in the scope of predicting thermal properties of organic molecules and reinforce the principles of Tm and ΔGc as metrics that aid in predicting the crystallization behavior of organic thin films. 

Integrating second order nonlinear (χ(2)) optical materials on chip is an ongoing challenge for Si photonics. Noncentrosymmetric molecular crystals have the potential to deliver high χ(2) nonlinearity with good thermal stability, but so far have been limited to growth from solution or the melt, which are both difficult to control and scale up in manufacturing. Here, we show that large (>100 μm) single crystal domains of the nonlinear molecule 2-[3-(4-hydroxystyryl)-5,5-dimethylcyclohex-2-enylidene] malononitrile (OH1) can be grown monolithically on either glass or Si via vacuum evaporation, followed by a short thermal annealing step. The crystallites are tens of nanometer thick and exhibit strong second harmonic generation with their primary χ(2) tensor component lying predominantly in plane. Remarkably, we find that a single domain can grow uninterrupted through nearby channels etched on a Si wafer, which may provide a path to integrate OH1 on Si or Si3N4 waveguides for a broad range of χ(2)-based photonic integrated circuit functionality. 

Abstract The preferential growth of α‐phase formamidinium perovskite (α‐FAPbI₃) at low temperatures can be achieved with the incorporation of chloride‐based additives, with methylammonium chloride (MACl) being the most common example. However, compared to other less‐volatile chloride additives, MACl only remains in the growing perovskite film for a short time before evaporating during annealing, primarily influencing the early stages of film formation. In addition, evaporation of MACl as methylamine (MA⁰) and HCl can introduce a side reaction between MA⁰and formamidinium (FA), undermining the compositional purity and phase stability of α‐FAPbI₃. In this study, it is demonstrated that addition of iodine (I₂) into the FAPbI₃precursor solution containing MACl suppresses the MA‐FA side reaction during annealing. Additionally, MACl evaporation is delayed owing to strong interaction with triiodide. The added I₂facilitates spontaneous growth of α‐FAPbI₃prior to annealing, with an improved bottom morphology due to the formation of fewer byproducts. Perovskite solar cells derived from an I₂‐incorporated solution deliver a champion power conversion efficiency of 25.2% that is attributed to suppressed non‐radiative recombination.