Journal of Raman Spectroscopy最新文献

The Raman Laser Spectrometer (RLS), part of the Pasteur analytical suite onboard the ExoMars 2028 Rosalind Franklin rover, is designed to perform structural and compositional analyses of powdered subsurface samples on Mars. Its fully autonomous operation within the constraints of the Pasteur Analytical Laboratory-limited by time, energy, and sample availability-requires an efficient balance between scientific performance and operational viability. This study presents a qualitative analysis of RLS operations under mission-representative conditions using the Flight Spare (FS) model, focusing on the impact of key parameters-number of accumulations, autofocus frequency, and analyzed spots per sample-on the system's detection capabilities. Experimental campaigns were conducted using ESA-selected analog samples representative of Oxia Planum geology. Performance was evaluated using both the RLS FS and the ExoMars Simulator. Results show high consistency (90-95%) in mineral detection between systems, confirming the robustness of the RLS FS under representative scenarios. The instrument demonstrated its ability to identify key phases, including oxides, silicates, carbonates, hydrated sulfates, and amorphous carbon, highlighting its relevance to geological and astrobiological investigations. Operational tests confirmed that reducing the number of accumulations or autofocus activations-under appropriate sample conditions-does not compromise spectral quality. These findings support a flexible strategy that adapts operational parameters to the scientific context, optimizing resource use and preserving long-term instrument reliability. The results will contribute to the refinement of nominal activity plans for ExoMars and reinforce the use of the RLS FS as a critical asset for validating future configurations of the flight model.

The food preparation industry relies on accurately measuring the endpoint temperature (EPT) of cooked pork to ensure the safety and quality of these products. Food thermometers are the current gold standard but become ineffectual as soon as the meat begins to cool. Furthermore, single-point measurements fail to capture the spatial variation of EPT in a sample. Raman spectroscopy offers a non-contact, consistent method of measuring EPT after the meat has cooled. We present two Raman classification models for predicting EPT in cooked pork. A principal component analysis-random forest (PCA-RF) model classifies spectra into temperature categories with 87.5% accuracy. A partial least squares (PLS) model predicts the EPT on a continuous scale with a root mean square (RMS) error of 3.8°C. We apply this PLS model to create hyperspectral images of the EPT in cooked tissue cross-sections. These images show that different cooking methods exhibit varying heat penetration, with expected differences in the distance from the hot outer surface over which EPT decays. This technique for imaging the EPT in cooked pork offers possibilities for further studies of the differences among various cooking methods, informing future improvements in both food safety and quality.

This study examines the temperature-dependent Raman spectra of MBE-grown bismuth telluride (Bi₂Te₃) thin films, analyzing Stokes, and anti-Stokes scattering in two polarizations to resolve symmetry-dependent mode strengths. Density functional theory simulations of Stokes spectra identified the fundamental vibrational modes and anharmonic decay. The temperature evolution of phonon wavenumbers and linewidths revealed the role of anharmonicity: the real part of the phonon self-energy governs the wavenumber shift (redshift) of the $�A_{�1�g�}^2$ mode, while its imaginary part drives the linewidth broadening, both arising from cubic and quartic anharmonic processes which is associated to energy and symmetry of the mode. In contrast, the lower wavenumber $�A_{�1�g�}^1$ and $�E_g^2$ modes exhibited weaker coupling to thermal decay channels, reflected in smaller changes to their self-energy components and longer lifetimes. The intensity of $�A_{�1�g�}^2$ mode decreased significantly with temperature due to multi-phonon decay, whereas $�A_{�1�g�}^1$ and $�E_g^2$ intensities remained stable. These results quantify the distinct mediation of wavenumber renormalization and lifetime effects in Bi₂Te₃ by the real and imaginary components of the phonon self-energy.

Machine learning (ML) techniques are valuable for analyzing complex biological SERS spectra, allowing for the detection of minor differences in cell composition. However, several challenges arise in the data analysis process, such as selecting the appropriate preprocessing methods, machine learning algorithms, and validation strategies to avoid under/overfitting and ensure reliable estimates. This study systematically compared various validation strategies and their impact on multiple ML classifiers using four biological datasets of varying complexities, in terms of class overlap, and sample variability.

Therefore, a machine learning workflow was established, incorporating more than 10 classifiers and using nested cross-validation (CV) for hyperparameter tuning and performance estimation. Five CV strategies were compared: Leave-One-Group-Out, stratified K-Fold, unstratified K-Fold, Leave-One-Out, and nested CV.

Our results demonstrate that stratified K-Fold CV yielded performance nearly equivalent to nested CV in terms of accuracy and efficiency but with a reduced computational cost. Leave-One-Group-Out strategy produced lower performance estimates than the other four methods, which may be more representative of real-world performance.

Conclusively, this work shows that simpler CV strategies can effectively replace computationally expensive nested CV in certain cases, while maintaining comparable performance. Nonetheless, careful consideration of overfitting remains crucial when employing these more efficient methods.

Raman spectroscopy is widely employed for quantitative analysis in aqueous solutions, yet it faces a notable challenge: the overlap of water Raman bands, often used as an internal standard, with analyte bands. To overcome this hurdle, we have developed an automated fitting method that encompasses intensity normalization, solvent subtraction, and quantitative analysis. This method utilizes the isolated Raman water spectrum as an internal standard. The effectiveness of this method has been validated through its application to the Raman spectra of Na₂SO₄, D-glucose, and lysozyme from egg white. These results show a strong correlation between normalized peak intensity and concentration. Furthermore, this method exhibits robustness against noise and fluorescence background, maintaining high accuracy even under low–signal-to-noise ratio (SNR) conditions, down to 20 dB. Most importantly, the fundamental principle of this method is versatile and can be applied to various types of quantitative spectral data derived from solutions.

The presence of quinolinic acid (QA) below 100 nM is a normal condition while its increased amount may cause variety of neurodegenerative diseases. The precise detection of QA in trace amounts (nanomolar) is crucial to control its toxic effects. In the present study, the SERS-based detection of QA and its interaction at varying concentrations in human serum is carried out using silver nanoparticle substrates. The analysis of conformational dynamics of QA across different concentrations ranging from 10⁻³ to 10⁻⁹ M has been done. The adsorption mechanism between QA and silver nanoparticles has been studied using DFT, and the detection of QA up to nanomolar concentration is achieved. A significant shift in the SERS spectra of QA is observed between 10⁻⁴ and 10⁻⁵ M concentration, attributed to changes in adsorption geometry with varying pH and conformational change from zwitterionic QA → neutral QA. These findings are supported by UV–visible spectra, pH measurements, and DFT calculations.