ABSTRACT Raman spectroscopy data of p ‐toluamide (PTA) polycrystals in the spectral region of 100–3500 cm −1 under high pressure have been collected up to 14.8 GPa at room temperature. Raman spectra exhibited distinct changes during compression, including the appearance of new Raman bands, disappearance of original Raman bands, and discontinuous changes in Raman shifts. The pressure dependence of the Raman shifts of PTA has been obtained. Detailed Raman spectral analysis suggested that compression induces three structural phase transitions of PTA within the investigated pressure range: phase I to II at 1.8 GPa, phase II to III at 5.5 GPa, and phase III to IV at 13.1 GPa. Moreover, the effect of pressure on the NH stretching vibrational modes has been analyzed based on the hydrogen bonds rearrangements among PTA molecules. This study provides a comprehensive understanding of PTA vibrational mode assignments and valuable insights into the amide structure stability for medicinal and pharmacological applications.
ABSTRACT Raman spectroscopy has emerged as a powerful analytical tool across diverse industrial sectors, owing to its nondestructive nature, high chemical specificity, and ability to provide unique molecular “fingerprints.” In the steelmaking industry, this technique offers a promising route for the rapid and precise characterization of critical materials such as sinter, a porous agglomerate of iron ore fines essential for blast furnace charging. However, practical deployment is often hindered by the scarcity of labeled spectral data. This work investigates the use of a ‐variational autoencoder (‐VAE) for the generation of high‐fidelity synthetic Raman spectra from a limited set of real sinter samples (321 spectra). With and latent codes sampled from a full‐covariance Gaussian fitted to the aggregate posterior, the model produces genuinely novel spectra whose global distributional fidelity matches that of SMOTE (Fréchet spectral distance: 4.55 vs. 4.41), despite SMOTE being constructed by interpolation of existing spectra; the mean intensity correlation with real spectra is 0.930 0.038. These synthetic spectra were used to augment supervised regression models for basicity () prediction. The LightGBM model achieved (95% CI: [0.72, 0.84]) in cross‐validation with original data; no augmentation method produced a statistically significant improvement (Wilcoxon signed‐rank, all ), and on the held‐out test set (), no augmented model exceeded the original‐only baseline (; S‐VAE: 0.809; Interpolação: 0.786; ‐VAE: 0.780; SMOTE: 0.779). The operational value of generative augmentation emerges instead in a downstream classification experiment: classifiers trained with increasing fractions of synthetic spectra sustain higher macro F1 at 100% synthetic fraction (‐VAE 0.66, S‐VAE 0.60) than SMOTE‐augmented classifiers, which degrade to 0.58. These results establish that the value of VAE‐based generation lies in producing distributionally faithful yet novel spectra that remain useful when consumed directly by downstream tasks such as spectral library expansion and classifier training, rather than in regression gains.
Accurate modeling of collisional linewidths is essential for CARS-based thermometry of complex molecules such as ethylene. In a previous study [JQSRT 234, 24 (2019)], constant isolated-line widths and empirical temperature scaling were used to model experimental CARS spectra of pure CH over the range 293-804 K, yielding satisfactory agreement but limited physical consistency. Here, we employ specially computed transition-dependent collisional widths and associated temperature exponents validated on infrared-absorption measurements to ensure physically meaningful trends and reproduce the experimental spectra with an improved accuracy.
ABSTRACT Experimental measurements of fuel and flame temperatures in a methane/air co‐flow diffusion flame are performed using coherent Stokes Raman scattering. A three‐beam hybrid fs/ps configuration is employed such that the excitation bandwidth of the ultrafast pulses includes the rovibrational transitions in the Fermi‐dyad of CO and the mode of CH. Time‐domain spectroscopic models were developed and employed to simultaneously extract the CH and CO temperatures separately from the same CSRS spectra. Axial and radial temperature profiles of CH and CO in the diffusion flame are presented. Methane temperatures ranged from 300 to 900 K while CO temperatures ranged from 500 to 2500 K. At probe volume locations higher above the burner, the fuel and product temperatures were in strong agreement. Measurement locations near the methane injector revealed inhomogeneities between fuel and flame temperatures before significant mixing by diffusion. For the first time, a three‐beam ultrafast CSRS configuration is demonstrated to simultaneously measure differing CH and CO temperatures.
Large-area soil organic carbon (SOC) measurements provide information for advancing soil health, optimizing agricultural productivity, and helping mitigate climate change; however, conventional soil analysis methods remain difficult to scale. Raman spectroscopy offers high-throughput molecular information, but matrix effects and strong fluorescence interference limit its applicability to soil analysis. Here, we combine shifted excitation Raman difference spectroscopy (SERDS) with machine learning and signal processing to overcome these limitations, as demonstrated by the analysis of more than 900 geographically diverse North American soils. This work introduces common-mode rejection (CMR), a physics-motivated preprocessing method that follows the central SERDS principle by removing the broad background shared by paired shifted-excitation measurements while preserving the noncommon component. Unlike conventional methods, such as asymmetric least squares (ALS) background subtraction, CMR avoids dataset-specific manual tuning of the baseline. This method is also robust to multiplicative scaling and additive offset differences between paired spectra. CMR preserves a reproducible, broad excitation-dependent spectral component that contributes materially to prediction. Multivariate regression models trained by CMR-SERDS spectra afforded a 16% gain in coefficient of determination, R , and a 28% reduction in root mean squared error, RMSE, compared with SERDS spectra preprocessed by ALS to yield flat baselines. These results show that parametric background correction algorithms, deployed aggressively to produce cosmetically flattened baselines, remove analytically useful information and establish CMR-SERDS as a robust, interpretable, and transferable framework for high-throughput SOC analysis, and suggest broader relevance to Raman measurements acquired under strong background interference.
In this work, we present NTECARS, an open-source code written in the programming language Julia for the calculation and fitting of single and dual-pump CARS spectra under nonequilibrium conditions. The code enables flexible modeling and fitting of the rovibrational distributions using multitemperature models, models that do not require assumptions about the vibrational distribution or user-defined distribution functions. NTECARS accounts for collisional narrowing effects, and convolutions with arbitrary laser and instrumental profiles, and allows for a wide range of parameters such as the molar fractions, instrumental profiles, and baseline offsets to be obtained through fitting. At present, and are fully implemented, with a modular design that allows for the inclusion of additional species without structural changes. The code is validated under thermal equilibrium conditions using both synthetic and measured spectra. The nonequilibrium capability is demonstrated by time-resolved CARS measurements on a nanosecond-pulsed discharge in a mixture and excellent agreement with time-resolved absorption spectroscopy measurements from a previous work is found.
ABSTRACT Conventional approaches to live‐cell metabolomics face key barriers: Destructive methods like mass spectrometry lose spatiotemporal resolution, while noninvasive techniques are limited by phototoxicity and probe dependence. These challenges hinder the exploration of tumor metabolic reprogramming, neurodegeneration, and metabolism‐targeted therapies. Surface‐enhanced Raman scattering (SERS), with label‐free detection, single‐molecule sensitivity, and subcellular resolution, offers a promising solution. Recent advances show its capacity to monitor metabolite gradients, redox dynamics (NADH/NAD + ), and enzyme spatial distribution, as well as to visualize compartmentalized metabolism through super‐resolution imaging. This review highlights these achievements and introduces an integrated framework of molecular recognition, dynamic tracking, mechanistic insight, and translational application. Finally, we outline future opportunities combining SERS with organ‐on‐chip models and artificial intelligence, underscoring its potential in personalized medicine and in situ metabolic regulation. Overall, SERS‐based metabolomics represents a methodological breakthrough enabling cross‐scale analysis of metabolic flux in health and disease.
ABSTRACT Cosalite (), a layered sulfosalt mineral with an orthorhombic structure, is investigated through a combined experimental and first‐principles theoretical approach to elucidate its vibrational, nonlinear optical, and thermal transport properties. The unambiguous determination of the non‐centrosymmetric space group for the most stable phase of cosalite is performed using DFT calculations, supported by experimental EBSD data. The experimental Raman spectra recorded at liquid nitrogen temperature reveal eight intense modes in the range 50–300 cm, assigned to Bi‐S and Pb‐S stretching vibrations. The direct optical transition is established by calculation of the band structure. The optical absorption spectrum calculated by TDDFT including SOC effects yields a bandgap value of = 1.1 eV, which is substantially similar to the fundamental Shockley‐Queisser limit value, thus opening the possibility of using cosalite as a photovoltaic element. The strong third‐order nonlinear optical response is obtained in cosalite ( m/V) and is attributed to the dynamics of the electronic lone pair localized on Pb and Bi atoms. The lattice thermal conductivity is established by analysis of the heat flux current using classical molecular dynamics, employing a machine‐learning technique for constructing the interatomic potential. The estimated ultralow value gives rise to potential applications of the crystal as a thermoelectric generator.
ABSTRACT Tip‐enhanced Raman spectroscopy (TERS) enables vibrational imaging with subnanometer spatial resolution, offering label‐free chemical insight into biological structures at the single‐molecule level. This review outlines recent developments in TERS instrumentation, including tip fabrication strategies, excitation geometries, and configurations compatible with ambient and liquid‐phase environments. Emphasis is placed on biological applications, where TERS has been used to probe proteins, nucleic acids, lipid assemblies, and viral particles. These studies demonstrate how localized plasmonic enhancement supports surface‐selective detection and enables mapping of molecular heterogeneity across cellular interfaces and membrane domains. Spatial resolution metrics, both lateral and vertical, are discussed in relation to tip–sample coupling, field confinement, and operational conditions. We also examine methodological challenges such as signal instability, tip degradation, thermal effects, and reproducibility, alongside emerging strategies for mitigation. In parallel, we highlight the role of advanced data analysis frameworks, including multivariate chemometric methods and machine learning approaches, which are increasingly essential for interpreting hyperspectral datasets and extracting relevant insights. Together, these advances position TERS as a versatile platform for nanoscale bioanalysis, with growing relevance for mechanistic studies, molecular diagnostics, and real‐time tracking of biochemical processes.
ABSTRACT This work investigates the pressure‐dependent behavior of Raman spectra of methane, carbon dioxide, nitrogen, and selected minor biogas components in the pressure range up to 100 bar. A clear pressure‐induced red shift was observed for the ν₁ and 2ν 4 bands (≈0.02–0.024 cm −1 ·bar −1 ), in agreement with literature data, while the ν 3 band exhibited a significantly larger shift (0.05 cm −1 ·bar −1 ). In contrast, the ν 2 and 2ν 2 bands were nearly pressure‐insensitive. Pressure‐dependent band broadening was observed mainly for fundamental stretching modes, whereas overtone bands remained largely unaffected. For CO 2 (10–50 bar), pressure sensitivity was confirmed not only for the fundamental and overtone Q‐branches but also for hot bands, which demonstrated measurable red shifts and quadratic broadening behavior. Nitrogen exhibited the smallest pressure shift (0.008 cm −1 ·bar −1 ) and showed band narrowing with increasing pressure. Raman spectra of minor but technically critical components, H 2 S and NH₃, are reported here for the first time in this context. Spectral overlap between H 2 S and methane was identified as a limiting factor for qualitative analysis, whereas ammonia bands were clearly distinguishable. Finally, the applicability of the method was verified by measuring a real biogas sample. The results demonstrate that Raman spectroscopy is a promising tool for high‐pressure qualitative analysis and pressure monitoring in biogas systems.
Variability in instrument calibration, both between different devices and over time, remains a significant obstacle to Raman spectroscopy. Even minor shifts in wavenumber can substantially reduce classification accuracy. Traditional spectral models struggle to remain robust under such conditions, particularly due to the one-dimensional (1D) nature of spectral inputs and the absence of pretrained deep learning architectures for Raman spectra. In this study, we systematically investigate the robustness of a previously introduced spider plot-based spectral representation specifically with respect to wavenumber calibration drift in test data. In this approach, each Raman spectrum is converted into a radial spider plot in which the angle of each sector represents the wavenumber, and the color indicates the normalized intensity, enabling the use of powerful image-based deep learning models. We employed EfficientNetB7, a pretrained convolutional neural network (CNN), to classify transformed spectra. Using a dataset of 5420 spectra from six bacterial species, we introduced synthetic wavenumber shifts of +/- 9 cm-1 to simulate realistic calibration drift in the test data. The traditional method, which combined principal component analysis with linear discriminant analysis (PCA-LDA), showed a significant drop in balanced accuracy from 0.90 to 0.77 when shifts were present. By contrast, our spider plot-based approach retained strong performance, achieving balanced accuracy scores of 0.88 without shifts and 0.83 with shifts. Although no spectral augmentation was done in the training of either classification workflow, we attribute this robustness to the rotational consistency of spider plots under spectral shift and the tolerance to image rotation in CNNs, which are pretrained on large, augmented image datasets. Our findings demonstrate the advantage of using image-based methods to mitigate the effects of calibration variability in Raman spectroscopy.
Two‐dimensional (2D) transition metal carbides and nitrides (MXenes) have attracted considerable attention due to their exceptional potential in electronics, electrochemical energy storage and conversion, optics, biomedicine, and sensing. However, their structural skeleton and surface terminations are highly sensitive to temperature, and thermally induced structural evolution can profoundly modify their physicochemical properties and device performance. Consequently, analytical methods capable of accurately identifying and continuously monitoring the structural dynamics of MXenes are urgently required. Within this context, Raman spectroscopy has emerged as a powerful and accessible technique for sensitively probing MXene lattice vibrations and surface chemistry. Using Ti 3 C 2 T x as a model MXene, this study employs in situ and quasi–in situ Raman spectroscopy to systematically elucidate its temperature‐dependent structural evolution during thermal annealing. The Raman response captures a series of sequential transformations triggered by heating, including interlayer water removal, depletion of –OH terminations accompanied by the growth of =O groups, loss of –F terminations, formation of C–C bonds, lattice oxidation, and eventual TiO 2 formation. By tracking the shifts and intensity variations of characteristic Raman modes, distinct structural states across the thermal evolution pathway are resolved, enabling the construction of a comprehensive spectral fingerprint that delineates the entire heat‐driven transformation process. These in situ Raman insights not only deepen the fundamental understanding of MXene thermal stability and structural evolution mechanisms but also provide an essential diagnostic basis and practical guidance for optimizing their performance and stability in electronic devices, electrochemical energy‐storage systems, and sensing applications.
This study investigates a fiber‐enhanced Raman gas sensor operating under pressurized conditions, which employs a 5‐m‐long nodeless hollow‐core anti‐resonant fiber (HC‐ARF). The system utilizes a 25‐μm core‐diameter optical fiber coupled with a charge‐coupled device (CCD) for spatial filtering. Under normal atmospheric pressure conditions and with an exposure time of 300 s, the system successfully detects atmospheric CH 4 and 13 C 16 O 2 . By increasing the gas pressure inside the HC‐ARF, it was observed that the Raman signal intensity was enhanced while the background noise remained unchanged. Under the conditions of an output power of 1.6 W, an integration time of 300 s, and an atmospheric pressure of 10 bar, the detection limits based on a 5‐m‐long HC‐ARF were determined to be 60 ppb for CH 4 , 130 ppb for C 2 H 6 , and 92 ppb for C 2 H 4 , respectively. Simultaneously, as the pressure increased, the gas filling rate of the gas sensor was also enhanced. At an atmospheric pressure of 10 bar, CH 4 filled the entire 5‐m‐long HC‐ARF within 4 s. These achievements have paved the way for developing portable gas sensors with low detection limits and rapid response times, thereby making the monitoring of trace gases feasible.
Raman spectroscopy, a technique rooted in the inelastic scattering of photons, has emerged as an influential tool in various scientific domains, including biology, materials, and cosmetics science. Raman spectroscopy enables real‐time, in situ, noninvasive analysis of cosmetic ingredients on the skin surface and internally. The employment of the aforementioned method in conjunction with the deep penetration capability of confocal Raman microscopy (CRM) enables layer‐by‐layer tracking of penetration behavior from the stratum corneum (SC) to the dermis. It utilizes molecular vibrational fingerprint analysis to achieve cross‐scale (cell to tissue) in vivo evidence for investigating cosmetic efficacy mechanisms. This review focuses on mechanisms and applications of Raman spectroscopy in cosmetics, emphasizing its unique advantages in skin analysis, efficacy evaluation, and quality control. Combining multivariate analysis with Raman spectroscopy data processing enhances its reliability in product evaluation. Overall, Raman spectroscopy holds great promise for furthering understanding of cosmetic science and improving product development and efficacy evaluation.
Soot, coal, graphite, and similar black-carbon materials contain layered sheets of sp2-bonded carbon in regions larger than about 4 nm. Such materials typically exhibit Raman spectra (RS) with a peak near 1580 cm-1 (G band) and, except in the case of graphene or highly ordered graphite, a peak in the range 1300-1350 cm-1 (D band). Here such materials are termed DG-carbon (DGC). To better characterize DGC, the RS can be represented by multiple D and G bands. Comparisons between "best-fit" D and G bands reported by different researchers are often not possible, partly because in many cases, the choices that must be made in selection of background removal procedures and optimization approach are not stated. Also, automated fitting methods are needed for characterizing large numbers of RS of DGC, such as those measured using Raman hyperspectral imagers, including those used in characterizing atmospheric aerosol particles. To help facilitate both quantitative comparisons of results and automated fits to RS data, we provide here a fitting code which uses MINPACK-1 as called by the LMFIT python package. To illustrate the use of the code and to investigate optimization using different numbers of basis functions, 15 DGC spectra and four non-DGC spectra were fitted. We found, for +/- 20 cm-1 variability of initial peak positions, the RS fits with two Lorentzian functions converged to a single optimized fit. However, for five-band fits, (1) the final fit demonstrated notable sensitivity to the initial values of the functions in 11 of 15 cases and (2) constraints on the peak center locations were required to keep peak center positions within valid ranges in some cases. When the RS were weaker or the luminescence was large, the variability in final fits was greater.
We investigated the physicochemical characteristics of commercially available bone graft substitutes of bovine, human, and natural hydroxyapatite origin, aiming to evaluate whether compositional differences detectable by vibrational spectroscopy may influence their suitability for bone regeneration. Raman and Fourier transform infrared (FTIR) spectroscopy techniques were employed to examine the vibrational signatures of phosphate and carbonate groups in the mineral phase, focusing on spectral descriptors associated with crystallinity, carbonate substitution, and lattice ordering. Carbonate-to-phosphate ratios derived from both Raman and FTIR spectra were used for a comparative assessment of the mineral composition of the different graft types. All materials exhibited characteristic phosphate and carbonate bands consistent with hydroxyapatite as the dominant mineral phase. Minor variations were observed in band positions/widths and carbonate-related features, indicating subtle differences in structural order that can be attributed to specific manufacturing processes in the synthesis. However, the carbonate-to-phosphate ratios showed limited variability, and only minor changes in carbonate-related spectral parameters were observed among bovine, human, and natural hydroxyapatite grafts. This study provides a methodological reference for carbonate substitution and crystallinity analysis, clarifying structural trends across commercial bioapatite grafts and supporting future research in this field. Even though the observed compositional differences are small, this study confirms that commercial grafts share a highly similar mineral framework, indicating a broad physicochemical similarity among the evaluated graft materials.
Colorectal cancer (CRC) remains a leading cause of cancer-related mortality, underscoring the need for improved early detection. Single-cell Raman spectroscopy (SCRS) enables the detection of CRC-associated biomolecular alterations and shows promise as a label-free and complementary auxiliary tool for improved CRC screening and characterisation. This study employed confocal Raman spectroscopy and imaging to analyse CRC cells and normal cell lines alongside 14 paired tissue samples. Biomolecules were quantified from the SCRS data to construct a machine learning classifier, and ultra-performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS) was used for metabolomic profiling. Metabolic mapping revealed consistent alterations in lipids, nucleic acids, proteins, amino acids and sugars across CRC and normal cell lines using Raman imaging and UPLC-MS/MS. Lipids were downregulated in CRC, whereas reduced sugar levels correlated with the tumour differentiation potential. The classification model exhibited high efficacy in diagnosing CRC and predicting its T stage and N stage, with AUC values ranging from 0.94 to 1.00. This study revealed the critical roles of lipid and sugar metabolism in CRC development and differentiation, establishing a foundation for understanding the pathogenesis and identifying therapeutic targets. The AI-assisted SCRS analysis showed high accuracy in classifying cancer, indicating its promise as a simple, label-free auxiliary tool for CRC screening and diagnosis.
Syntheses of new organic molecules and the discovery of high-pressure phases of known organic molecules under extreme conditions have always been a special interest in the fields of chemical science, geological sciences, and astrochemistry, owing to their profound significance for both the fundamental and advanced understanding of the origins of life on Earth. Laboratory-based dynamic shock-wave-recovery experiments on organic species have made a tremendous contribution to the identification of the synthetic role of natural shockwaves on the pre-biological processes of space-related organic molecules. In the present study, we consider L-tartaric acid (LTA-P21) as a test sample, which is a potential molecule that may be present in the interstellar medium (ISM). It is subjected to acoustic shock waves with 0, 50, and 100 shocks, respectively, to understand its high-pressure structural science. Significant differences are observed in the X-ray diffraction lines, Raman spectral lines, morphology, and optical transmittance spectra of the LTA under shocked conditions. To identify the formation of new phases, Le Bail refinements were performed, revealing P21-P1 and P1-P1 transitions under 50- and 100-shocked conditions and further validated by the analyses of the lattice Raman modes, surface morphology, and optical bandgap energies. A possible mechanism is proposed based on the thermal conductivity-dependent shock waves-driven superheating approach. By presenting unequivocal explanations for the present work, it is believed that the acoustic shock waves have enough driving force to discover new crystallographic structures in organic species, such that tabletop shock tubes can be effectively considered powerful tools for discovering new organic species and phases.
1,4-Cubanedicarboxylic acid (CDC) is an important cubane derivative that has attracted considerable attention from researchers in functional materials because of its rigid three-dimensional cage structure and the carboxyl groups with diverse coordination capabilities. Although its crystal structure has been reported, systematic vibrational spectroscopic characterization, especially Raman spectroscopic analysis, has not yet been carried out. The lack of spectroscopic reference data limits the identification of CDC in multicomponent systems and hinders the monitoring of possible structural changes under different conditions. In this study, CDC was investigated by X-ray diffraction, Fourier transform infrared (FTIR) spectroscopy, and Raman spectroscopy, supported by density functional perturbation theory (DFPT) calculations. With the aid of the calculated vibrational wavenumbers, the experimental spectra were further validated, and the main phonon modes were assigned in detail. The characteristic infrared and Raman spectral features of CDC were identified. These results provide a reliable vibrational spectroscopic reference for CDC and lay a foundation for the identification and further study of cubane-based materials.
