Applied Spectroscopy最新文献

Liquid biopsy is revolutionizing cancer management, with circulating tumor cells (CTCs), offering a transformative approach to screening, diagnosis, and treatment monitoring. However, existing CTC isolation methods relying on antigen expression or physical properties lack robustness, are operator-dependent, and suffer from automation challenges, leading to inconsistent and time-intensive analyses. A universal, unbiased methodology for CTC detection across tumor types is critically needed. Here, we present the first proof-of-concept study demonstrating the use of Fourier transform infrared (FT-IR) microspectroscopy to study cytospun blood samples coupled with a random forest (RF) classifier, for the detection of a single CTC in the blood of a lung cancer patient as confirmed via immunohistochemistry. Notably, our method utilizes glass coverslips as substrates, routinely employed in pathology departments, enabling seamless integration with histopathological analyses (e.g., staining, immunohistochemistry). Using FT-IR spectral data from in vitro growing lung cancer cells as a training model, we achieved precise CTC identification based on biochemical composition, specifically within the fingerprint region (1800cm^-1 to 1350 cm^-1). This study introduces FT-IR microspectroscopy as a novel, label-free approach for CTCs detection in liquid biopsies, with the potential to redefine cancer diagnostics. By enhancing precision and accessibility in CTC identification, the clinical implementation of this methodology may represent a significant advancement in personalized oncology, offering a clinically viable tool for real-time cancer monitoring and improved patient stratification.

Understanding the abundance of atomic oxygen in the vicinity of carbon surfaces exposed to high-enthalpy flows is critical to accurate predictions of the gas-surface interaction. A novel approach for obtaining absolute number density measurements of atomic oxygen in high-enthalpy facilities with nanosecond laser pulses is described and demonstrated using photoionization-dominated, two-photon laser-induced fluorescence. In two-photon laser-induced fluorescence measurements, the depopulation of the excited state is typically dominated by electronic quenching, which depends on the temperature, pressure, and gas composition. To account for the electronic quenching rate, the fluorescence lifetime can be measured by temporally resolving the fluorescence. This can prove challenging in high-temperature and/or high-pressure environments where the fluorescence lifetime can be less than a nanosecond. Instead, by increasing the laser intensity until photoionization dominates the depopulation of the excited state, we create a quenching-independent measurement that is proportional to absolute number density. This technique is demonstrated here in the reacting boundary layer of a graphite sample ablating in the 6000 K plume of an inductively coupled plasma torch. The boundary layer possesses a large temperature gradient that varies from about 2000 K near the sample surface to the plume temperature of 6000 K in a span of approximately 2 mm. The photoionization-dominated technique is calibrated by using the freestream oxygen concentration, assuming the torch plume is in local thermodynamic equilibrium. The spatial resolution of the measurements is 50 µm and we are able to measure the number density of atomic oxygen to within about 60 µm of the graphite sample.

Discrete frequency infrared (IR) imaging is an exciting experimental technique that has shown promise in various applications in biomedical science. This technique often involves acquiring IR absorptive images at specific frequencies of interest that enable pathologically relevant chemical contrast. However, certain applications, such as tracking the spatial variations in protein secondary structure of tissue specimens, necessary for the characterization of neurodegenerative diseases, require deeper analysis of spectral data. In such cases, the conventional analytical approach involves band fitting the hyperspectral data to extract the relative populations of different structures through their fitted areas under the curve (AUC). While Gaussian spectral fitting for one spectrum is viable, expanding that to an image with millions of pixels, as often applicable for tissue specimens, becomes a computationally expensive process. Alternatives like principal component analysis (PCA) are less structurally interpretable and incompatible with sparsely sampled data. Furthermore, this detracts from the key advantages of discrete frequency imaging by necessitating the acquisition of more finely sampled spectral data that is optimal for curve fitting, resulting in significantly longer data acquisition times, larger datasets, and additional computational overhead. In this work, we demonstrate that a simple two-step regressive neural network model can be utilized to mitigate these challenges and employ discrete frequency imaging for retrieving the results from band fitting without significant loss of fidelity. Our model reduces the data acquisition time nearly six-fold by requiring only seven wavenumbers to accurately interpolate spectral information at a higher resolution and subsequently using the upscaled spectra to accurately predict the component AUCs, which is more than 3000 times faster than spectral fitting. Our approach thus drastically cuts down the data acquisition and analysis time and predicts key differences in protein structure that can be vital towards broadening potential applications of discrete frequency imaging.

Rocks are an extremely important and indispensable part of the Earth's crust, with wide applications in various fields such as geology, environmental monitoring, and industry. Traditional methods often rely on a single analytical technique or visual inspection, but this may not achieve the accuracy required for thorough classification. Laser-induced breakdown spectroscopy (LIBS) technology mainly provides information on the composition and content of rock elements, while images can provide appearance information such as color and texture. The multilayer perceptron (MLP) and DenseNet121 models were selected for processing preprocessed LIBS and image data, respectively. When using LIBS and images separately for classification, the accuracy rates were 93.63% and 90.90%, respectively. However, after fusing the bimodal data using LIBS and images, we achieved a significant performance improvement of 97.27% in accuracy. This study indicates that advanced neural network models can effectively integrate LIBS and image data and improve the performance of rock classification.

Proteins are biomolecules with characteristic three-dimensional (3D) arrangements that render them different vital functions. In the last 20 years, there has been a growing interest in biopharmaceutical proteins, especially antibodies, due to their therapeutic application^. The functionality of a protein depends on the preservation of its native form, which under certain stressing conditions can undergo changes at different structural levels that cause them to lose their activity.¹ Although mass spectrometry is a powerful technique for primary structure determination, it often fails to give information at higher order levels. Like infrared (IR), Raman spectra are well known to contain bands (especially the amide I from 1625-1725cm^-1) that correlate with secondary structure (SS) content. However, unlike circular dichroism (CD), the most well-established technique for SS analysis, Raman spectroscopy allows a much wider ranges of optical density, making possible the analysis of highly concentrated samples with no prior dilution. Moreover, water is a weak scatterer below 3000 cm^-1, which confers Raman an advantage over IR for the analysis of complex aqueous pharmaceutical samples as the signal from water dominates the amide I region. The most traditional procedure to extract information on SS content is band-fitting. However, in most cases, we found the method to be ambiguous, limited by spectral noise and subjected to the judgment of the analyzer. Self-organizing maps (SOM) is a type of self-learning algorithm that organizes data in a two-dimensional (2D) space based on spectral similarity and class with no bias from the analyzer and very little effect from noise. In this work, a set of protein spectra with known SS content were collected in both solid and aqueous state with back-scatter Raman spectroscopy and used to train a SOM algorithm for SS prediction. The results were compared with those by partial least squares (PLS) regression, band-fitting, and X-ray data in the literature. The prediction errors observed by SOM were comparable to those by PLS and far from those obtained by band-fitting, proving Raman-SOM as viable alternative to the aforementioned methods.

In this study, we have developed a plasmonic hybrid heterostructure integrating two elements: Two-dimensional (2D) reduced graphene oxide-gold nanostars composite (rGO-GNS), and gold nanostars (GNS) substrate. By harnessing the unique plasmonic properties of rGO in chemical enhancement and that of GNS in electromagnetic enhancement, the hybrid heterostructure offers synergistic enhancement effects that enable ultra-low sensitivity and accurate identification and analysis of trace quantities of target substances. It is noteworthy that the high-density hotspots generated by strong plasmonic coupling of rGO-GNS and GNS results in ultra-high surface-enhanced Raman spectroscopy (SERS) enhancement compared to individual substrate either GNS or rGO-GNS substrate. Moreover, the uniformity and reproducibility of the GNS@rGO-GNS substrate were studied by using thiophenol (TP) as a model analyte, which indicates that the SERS sensor exhibited superior signal reproducibility with an RSD value 5% and long-term stability with a minimal signal loss after 30 days. To demonstrate a potential application of our SERS substrate, SERS detection of the pesticide thiram in river water was realized with a limit of detection (LOD) up to 50 pM, showing the potential for new opportunities for efficient chemical and biological sensing applications.

The impact of pulsing a radio frequency (RF) glow discharge lamp on analytical aspects in glow discharge optical emission spectrometry (GD-OES) was investigated. The experiments were done with a LECO GDS950 spectrometer. This instrument has a fixed pulse frequency of 320  Hz and adjustable pulse duty cycles (PDC) 100%-6.4%. In the first part, emission yields (EY) were studied by measuring coated steel samples in compositional depth profiling (CDP) mode. Variations in EY were measured by integrating the intensity of emission lines from several elements through the entire coatings at several PDC settings. The results show generally small EY variations. For improved accuracy, a set of correction constants is suggested. In the second part, the impact on signal-to-background (S/B), signal-to-noise-noise (S/N), and precision was investigated using "high current" pulsing. This means increased pulse power leaving the average power constant at the different PDC settings. The samples were a low alloy steel and a high purity iron blank (background) sample. The results showed significant increase of the S/B and S/N for four out of six spectral lines investigated at increasing pulse power, showing potential for improved detection limits (DL). Furthermore, there was a tendency towards improved precision with higher pulse power. Finally, the effect on depth resolution in CDP was investigated by running a ZnNi coated steel using "high current" pulsing. It was found that the depth resolution was unaffected up to 30% PDC.

In a previous paper, we proposed the use of a set of colored LEGO blocks as "standard" samples for the evaluation of fluorescence avoidance and mitigation schemes in Raman spectroscopy, as well as for use to evaluate the instruments' performance on dark samples. The purpose of this paper is to establish that this set of LEGO blocks does represent a good test case for fluorescence avoidance and mitigation when using handheld Raman spectrometers, and for the ability to record Raman spectra from dark samples. The performance of ten different instruments, operating using different exciting lines (785, 830/852, and 1064 nm), and different data processing schemes, are compared. The combination of a series of colored blocks (white, yellow, red, and blue), and successively darker tone blocks (white, gray, and black) do challenge these instruments, and shed light on the ways that their manufacturers have optimized these instruments in specific areas and for different purposes.

A new method based on near-infrared (NIR) hyperspectral imaging (HSI) has been developed for online polymer film thickness mapping. Traditional online methods, including X-ray, capacitance, and physical gauging (micrometers), can only determine film thickness for a point with each measurement. The NIR-HIS method allows the determination of film thickness for a line based on each image, thus enabling true real-time two-dimensional (2D) mapping of film thickness as the film translates in front of the instrument. A Specim NIR camera, 1000-2500 nm, 384 (spatial) × 288 (spatial) pixels, was used in this study for various low-density polyethylene (LDPE), and high-density polyethylene (HDPE) films. Sample thickness between μm to mm can be mapped based on the myriad NIR absorbance bands with various molar absorptivity. The 2310 nm NIR peak was found to be the most effective feature for determining film thickness over the range of polyethylene film studied in this project: 10∼100 μm. A good correlation was found between the 2310 nm absorbance and the incumbent X-ray thickness scanner results. Interference fringes were found to be a potential source of error for quantitative analysis of thin films, and a classical least squares (CLS) analysis was found to be effective in removing fringes. This method was implemented to map out film thickness in real-time in an industrial blown film process.