Surface-enhanced Raman spectroscopy (SERS) is a vibrational spectroscopy technique with sensitivity down to the single molecule level that provides fine molecular fingerprints, allowing for direct identification of target analytes. Extensive theoretical and experimental research, together with continuous development of nanotechnology, has significantly broadened the scope of SERS and made it a hot research field in chemistry, physics, materials, biomedicine, and so on. However, SERS has not been developed into a routine analytical technique, and continuous efforts have been made to address the problems preventing its real-world application. The present minireview focuses on analyzing current and potential strategies to tackle problems and realize the SERS performance necessary for translation to practical applications.
Oliver T. Unke, Stefan Chmiela, M. Gastegger
et al.
Machine-learned force fields combine the accuracy of ab initio methods with the efficiency of conventional force fields. However, current machine-learned force fields typically ignore electronic degrees of freedom, such as the total charge or spin state, and assume chemical locality, which is problematic when molecules have inconsistent electronic states, or when nonlocal effects play a significant role. This work introduces SpookyNet, a deep neural network for constructing machine-learned force fields with explicit treatment of electronic degrees of freedom and nonlocality, modeled via self-attention in a transformer architecture. Chemically meaningful inductive biases and analytical corrections built into the network architecture allow it to properly model physical limits. SpookyNet improves upon the current state-of-the-art (or achieves similar performance) on popular quantum chemistry data sets. Notably, it is able to generalize across chemical and conformational space and can leverage the learned chemical insights, e.g. by predicting unknown spin states, thus helping to close a further important remaining gap for today’s machine learning models in quantum chemistry. Current machine-learned force fields typically ignore electronic degrees of freedom. SpookyNet is a deep neural network that explicitly treats electronic degrees of freedom, closing an important remaining gap for models in quantum chemistry.
Abstract Deep eutectic solvents (DESs) are currently being used in different sectors, such as electrochemistry, electrodeposition, organic synthesis, nanoparticle preparation, bioactive compound separation, etc. Their use in analytical chemistry has only recently begun to expand. Despite the publication of a sufficient number of DES-based analytical extraction procedures, some details, such as interaction of DES with the sample and target analytes as well as with water are insufficiently explored and theoretically explained. Here we discuss the role of water in DES-based extraction in terms of analytical chemistry, especially for the pre-treatment of solid samples. We believe that this review will benefit those who have linked their research with DESs and will enable them to speed up their work.
Abstract Wearable potentiometric ion sensors (WPISs) have become an exciting analytical platform that combines chemical, material and electronic efforts to supply physiological information during certain human activities. The real possibility of wearing an analytical device with diverse configurations—sweatband, patches, garments—without disturbing the welfare of the carrier has enabled potentiometric ion sensors both as health quality and sport performance controllers. Recent studies show a large involvement of WPISs in the following of critical biomarkers (timely or continuously), such as sodium, potassium, calcium, magnesium, ammonium and chloride, which are present at relatively high concentrations in sweat (∼mM levels). Certainly, the non-invasive nature of WPISs and other significant features, e.g., simplicity and cost-effectiveness, have broadened new horizons in relation to applied analytical chemistry. This has been pointed out in the literature over the last decade with the predominance of two analytical outcomes: (i) the improvement of sport performance as a result of continuous detection of ions in sweat (health status of the individual) while decreasing physiological complications (injuries, muscle cramps, fatigue and dehydration) during practice; and (ii) advancements in clinical diagnostics and preventive medicine as a consequence of the monitoring of the health status of patients suffering from any kind of disorder. Beyond the undeniable importance of the integration of WPISs to satisfy current societal needs, the following crucial questions about misleading and missing analytical features need to be answered: To what extent is WPIS technology a reliable analytical tool for the quantification of ions? Is cross-validation the current bottleneck toward further progress? Which are the fundamental steps involving the ion-selective electrode side that would benefit WPIS outcomes? Why is sweat the main (and almost the only) biological fluid to be monitored by WPISs? What is the best sampling strategy to be incorporated into WPIS devices for on-body monitoring of sweat? Which precision limits should be considered to assure a reliable decision-making process? Accordingly, this review focuses on the progression of WPISs from an analytical perspective—merely our vision of the field—within the period between 2010 and 2018. An updated search using specific keywords (wearable, ion, potentiometry, sensor) provided 43 contributions, which are herein highlighted, with a sustainable acceleration over the last three years. Thus, this review describes the current state of WPIS technology, the construction of wearable all-solid-state potentiometric sensors, critical requirements of potentiometric sensors to be fulfilled in a wearable configuration and key features regarding the ideal implementation of WPISs as reliable messengers of physiological information in real scenarios.
Yan-zuo JIANG, Sheng-sheng WANG, Zhi-heng QIAN
et al.
The proliferation of fentanyl analogs, marked by their structurally diverse and rapidly evolving molecular frameworks, presents formidable and urgent challenges to traditional detection methods. The existing techniques based on gas chromatography-mass spectrometry (GC-MS) are evidently limited by the subjective judgment of operators. The existing solutions all have obvious shortcomings. First, the manual judgment leads to the low efficiency of the entire data processing workflow. Second, the limited coverage and frequent delay of database make it unable to promptly incorporate newly emerging molecules. Third, the entirely single-point detection fails to form effective linkage, resulting in a severe issue of data silos. To overcome these deficiencies, this study specifically introduced an intelligent detection system based on a scalable cloud platform. The system combines three advanced techniques: large-scale artificial intelligence (AI) models, proprietary high-speed spectral matching algorithms, and an innovative federated retrieval framework. This integration enables a fully automated, one-click analytical workflow that fundamentally redefines the detection process. It supports multimodal inputs including raw analytical data, analytical reports, and spectral images captured by smartphones. Once data have been input, the system will automatically perform high-precision information extraction, feature analysis, and unambiguous substance identification without manual intervention. This study also conducted comprehensive and systematic testing in order to verify the overall performance of the developed system. Core parameters, including analytical accuracy, processing speed, sample throughput, and computational resource utilization, were quantitatively evaluated in the testing process. The results confirm that the system is fully capable of meeting the diverse requirements of rapid on-site sample testing. Additionally, a series of performance comparison tests were examined on the cloud platform hosts with varied performance levels. These hosts allow users to select servers tailored to specific needs, thus balancing performance and cost-effectiveness. This newly-designed system also incorporates a federated search algorithm, thereby establishing a cloud-client collaborative computing mode for detection tasks. When the algorithm is activated, the central processor searches the main path library, while distributing retrieval and matching tasks to associated clients, then the clients join in the matching and searching process. These processes achieve rapid coverage of newly emerged types of unknown controlled samples—especially under the circumstance that the main library lacks real-time updated sample data. This node-distributed network retrieval model can avoid sharing sensitive proprietary raw data, while breaking down the data silos under secure conditions. However, the operation of the algorithm will increase the overall system retrieval time, so it is necessary to carefully consider the actual needs and use the algorithm selectively. This intelligent system provides an effective solution for transforming the traditional mode of drug analysis into an intelligent multi-point collaborative detection approach, while presenting immediate and powerful tools for curbing the spread of fentanyl analogs. It can also offer some similar solutions for other scenarios related to drug control. This study effectively contributes to the integration of artificial intelligence with analytical chemistry, demonstrating the advent of an upcoming intelligent and data-driven era of analysis.