Diverse Parahydrogen-Induced Hyperpolarization Effects in Chemistry: Expanding Beyond Classical Approaches
Parahydrogen-induced hyperpolarization (PHIP) is a powerful tool for enhancing nuclear magnetic resonance (NMR) signals, with diverse effects beyond the classical PASADENA, ALTADENA, hydrogenative PHIP (hPHIP), and signal amplification by reversible exchange (SABRE) approaches. This review examines less common PHIP phenomena, including photo-PHIP, partially negative line (PNL) effects, oneH-PHIP, metal-free PHIP, and chemically relayed polarization transfer, which provide valuable insights into reaction mechanisms and enable new analytical applications.
Groundbreaking Synthesis and Characterization of a Promethium Complex, Unlocking Insights into the Elusive Element
The first experimental characterization of a promethium complex in solution, providing fundamental insights into the chemistry and properties of this rare and enigmatic element.
LlaSMol: Advancing Large Language Models for Chemistry with a Comprehensive Instruction Tuning Dataset
Our developed LLMs can achieve very strong results on a comprehensive set of chemistry tasks, outperforming the most advanced GPT-4 and Claude 3 Opus by a substantial margin, by leveraging the proposed large-scale, comprehensive, and high-quality instruction tuning dataset SMolInstruct.
L+M-24: Building a Dataset for Language+Molecules @ ACL 2024
Natural language plays a crucial role in molecule design, focusing on compositionality, functionality, and abstraction.
Node-Aligned Graph-to-Graph (NAG2G): Elevating Template-Free Deep Learning Approaches in Single-Step Retrosynthesis
Node-Aligned Graph-to-Graph (NAG2G) revolutionizes single-step retrosynthesis prediction with template-free deep learning.
RetroBridge: Modeling Retrosynthesis with Markov Bridges at ICLR 2024
Modeling retrosynthesis with Markov bridges for accurate prediction of precursor molecules.
EL-MLFFs: Ensemble Learning of Machine Learning Force Fields
EL-MLFFs proposes an ensemble learning framework to enhance force prediction accuracy in machine learning force fields.
Kernel-Elastic Autoencoder for Molecular Design: Innovations and Applications
Kernel-Elastic Autoencoder (KAE) revolutionizes molecular design with enhanced generative capabilities, setting new benchmarks in constrained optimizations and molecular docking.
3D Molecules Generative Modeling via Bayesian Flow Networks
Geometric Bayesian Flow Networks (GeoBFN) achieves state-of-the-art 3D molecule generation performance.
Machine Learning Analysis of Atomic Force Microscopy Images for Image Classification and Sample Surface Recognition
Machine learning offers a seamless approach to analyze multidimensional AFM images, enabling classification of sample surfaces with statistical significance.
© 2024 by Linnk AI