Atomic spectrometry update: review of advances in the analysis of metals, chemicals and functional materials
Ayush Agarwal
ab,
Eduardo Bolea-Fernandezc,
Robert Clough*d,
Andy Fisherd,
Bridget Gibsone and
Steve Hilld
aPaul Scherrer Institute, PSI Center for Energy and Environmental Sciences, 5232 Villigen PSI, Switzerland
bÉcole Polytechnique Fédérale de Lausanne (EPFL), School of Architecture, Civil and Environmental Engineering (ENAC IIE GR-LUD), 1015 Lausanne, Switzerland
cDepartment of Analytical Chemistry, Aragon Institute of Engineering Research (I3A), University of Zaragoza, Zaragoza, 50009, Spain
dSchool of Geography, Earth and Environmental Science, University of Plymouth, Plymouth, PL4 8AA, UK. E-mail: rclough@plymouth.ac.uk
eIntertek Sunbury Technology Centre, Shears Way, Sunbury, Middlesex, UK
First published on 9th October 2025
Abstract
This update covers the literature published between approximately June 2024 and April 2025 and is the latest part of a series of annual reviews. It is designed to provide the reader with an overview of the current state of the art with respect to the atomic spectrometric analysis of various metals, polymers, electronic, nano and other materials. Data processing continues to be the major focus for LIBS and TOF-SIMS analyses, mainly to provide reliable analyte quantification data. A variety of machine learning algorithms and statistical approaches have been used for this, often in multiple steps. Although these algorithms have been used for some years, their use is expanding into new areas. Another development is the combination of complementary techniques on the same instrument platform. This enables data from the two techniques to be obtained simultaneously and from the same spot on the sample. The analysis of polymers and nanomaterials continues to develop, with the prominent platforms used being SP-ICP-MS, SP-ICP-TOF-MS and X-ray based techniques. In addition, efforts are now being accelerated to produce nanomaterial CRMs and RMs, the lack of which has hampered truly robust method validation. For electronic materials XPS, GIXRF, GEXRF and TOF-SIMS remain dominant for surface and depth profiling, whilst for bulk composition LIBS, ICP-MS, and XRF remain prominent. Work in this area is also focussing on the development of advanced sample preparation and microextraction approaches that expand the scope of laser-based spectroscopy.
1. Introduction
This latest update adds to that from last year1 and complements the five other annual Atomic Spectrometry Updates, advances in environmental analysis,2 advances in the analysis of clinical and biological materials, foods and beverages,3 advances in atomic spectrometry and related techniques,4 advances in elemental speciation5 and advances in X-ray fluorescence spectrometry and its special applications.6 To mark the 40th anniversary of the ASU, an overview of its history, topic coverage and a call for writers, has also been published this year.7All the ASU reviews adhere to a number of conventions. An italicised phrase close to the beginning of each paragraph is intended to highlight the subject area of that individual paragraph. A list of abbreviations used in this review appears at the end. It is a convention of ASU that information given in the paper being reported on is presented in the past tense whereas the views of the ASU reviewers are presented in the present tense. Topical review papers which cover a broad subject area appear in their own section whilst those that cover a more specific area are to be found within the relevant subsection.
2. Topical reviews
There has been a number of relevant reviews of techniques published this year. Ion beam analysis is a collective term for several techniques, including RBS and PIXE. A review presented by da Silva, (93 references) discussed the use of machine learning in ion beam analysis.8 The analytical techniques are well known as being capable of surface and sub-surface analysis as well as depth-profiling. However, they often suffer the drawback of requiring significant amounts of data processing which requires high computing capacity and slows the overall analysis. The review was split into sections which discussed supervised learning, unsupervised learning and reinforcement learning. There was also a section where the author presented perspectives and future pathways. It was emphasised that these techniques are still in their infancy but offer great potential for enhancing workflows and for undertaking innovative applications. A review by Robinson and Thissen entitled “Selecting the best surface analysis method for your materials/samples” (85 references) considered the question of what is required from the analysis.9 For instance, does the elemental composition need to be determined (techniques such as XPS, EDX, XRF, SIMS), the chemistry (MALDI, FTIR, Raman and SIMS), morphology (SEM, XRD), layer thickness (SIMS, XRR, ellipsometry), etc. The techniques used to assess these properties as well as optical, electrical and mechanical properties were shown in a useful summarising figure. Useful tables were also presented that detailed the capabilities of each technique, e.g. resolution (lateral or depth), whether a technique is capable of mapping, detection limits obtainable and also provided a few extra notes. As well as the quick reference tables, many of the individual techniques were described in individual paragraph of text. Other sections of the text emphasised the other important/defining factors, e.g. cost of instrumentation and of analysis, availability of instrumentation. Also emphasised was the fact that many of the techniques provide complementary information and do not simply give the same information as each other.The large majority of papers describing the use of TOF-SIMS present only qualitative data, even though it is possible to obtain quantitative data. An overview focussing on the best practices for performing quantitative TOF-SIMS analysis was presented by Spool and Finney.10 The paper contained only 15 references but provided a good base for other workers in the area. The paper was split into five sections, including an introduction, experimental, obtaining repeatable results, obtaining replicable and reproducible results and conclusions. The bulk of the paper was in the section entitled obtaining repeatable results. This contained numerous sub-sections, each providing an important factor to take into consideration when attempting to obtain repeatable data. These sections were: appropriate sample, instrument status, primary ion current, flux, and dose; primary ion beam focus and the effective dose; dead time; charge compensation; topography and sample positioning; standards for normalisation and quantification; instrument maintenance and ANOVA, gauge repeatability and reproducibility. The overall conclusion was that establishing repeatability is the main challenge in obtaining quantitative analysis. It was therefore proposed that a standard protocol be prepared for any set of quantitative analyses. The acquisition setup, the analyser and primary ion gun settings and the primary ion current and dose should be consistent. A normalisation scheme may eliminate the effects of small primary ion current and dose changes but cannot overcome the effects of primary ion current on dead time effects and of primary ion dose on sample damage. Control of the primary ion dose and current was said to provide the best results. By obeying these basic rules and a string of others given in the conclusions, the authors stated that quantitative results can readily be obtained.
In recent years, LIBS has become an increasingly popular technique because of its lack of requirement of complex sample preparation procedures, relative ease of use, simultaneous detection, quasi-non-destructive sample analysis, the ability to map surfaces and also to conduct depth-profile studies. The recent advances in chemical composition imaging operation based on LIBS was the subject of a review (158 references) by Zhao et al.11 The fundamental principles, operation types, applications and technical upgrade schemes were all included in the review. The main focus of the review was the types of LIBS imaging operations and proposals for technological upgrades. The existing problems in LIBS imaging were presented and the authors then provided their own insights into its future development directions, including LIBS signal optimisation, imaging resolution improvement and multiple technology integration. Applications in the fields of biomedical research, industry, materials and minerals were presented.
The last review of interest was presented by Churbanov et al. who gave an overview of the methods used to assess the impurity composition of high purity sulfur.12 The review (63 references) covered the identification of impurities, molecular compounds, and heterophase inclusions in high-purity sulfur. The review was not merely confined to atomic spectrometric techniques. In addition to atomic emission and mass spectrometry, also included were techniques such as colorimetry, gravimetry, titrimetry, turbidimetry, conductometry, gas chromatography, infrared spectrometry, chromatography coupled with mass spectrometry and laser ultramicroscopy.
3. Metals
3.1. Ferrous metals
Following on from the trend noted in previous reviews, the analysis of ferrous metals has attracted a lot of attention. The development of methods to improve calibration accuracy and sensitivity of LIBS analyses, and address problems arising from self-absorption, continues to be a popular study area. Other studies on LIBS have investigated the laser configurations used.Several papers on the different types of lasers for use with LIBS have been published. One, which compared spatially resolved single and double pulse (especially orthogonal double pulse) setups for LIBS to examine and enhance LOD values when analysing duplex stainless-steel alloy, was reported by Alkallas et al.13 The Boltzmann plot method and Stark broadening were used with single pulse and orthogonal pre-pulse double pulse configurations. The experimental results showed that there were significant increases in the spectral intensities of the Fe lines in double pulse LIBS with Ar gas compared to those using He gas. Internal standardisation was used to create calibration curves for the spectral lines of Cr, Cu, and Mn in alloy samples for both studied media. Argon was found to outperform He when determining the elemental composition of steel alloys. Furthermore, the LOD values for both single pulse and double pulse LIBS were calculated and reported to be better in the presence of Ar gas compared to those found using He gas. The nanosecond laser is commonly used in LIBS, however during the nanosecond laser ablation process, there is a thermal effect on the target material, making it difficult to further improve the spatial resolution. A study using a picosecond laser to investigate the effects of the diameter of laser focusing spot, laser energy, and laser irradiation interval on the spatial resolution of LIBS has been published.14 The spatial mapping of metallic coatings using LIBS with a spatial resolution of 1 μm was achieved by using laser energy of 0.4 μm J per pulse and irradiation interval of 0.8 μm. The LIBS measurement results were in good agreement with those arising from the use of SEM-EDS. The study showed that by changing the laser ablation conditions, the spatial resolution of the spatial mapping of metallic coatings by using LIBS based on picosecond laser pulses could be reduced to 1 μm or lower. Picosecond LIBS has also been used by Li et al.15 for high-resolution microanalysis of steel samples. The LIBS system utilised a 9 ps pulsed laser operating at 355 nm and 35 Hz repetition. Successful analysis of Al, Cu, and Mn segregation in steels was achieved. The LIBS mappings with a spatial resolution of 2 μm demonstrated agreement with results obtained from EPMA. In addition, by employing an even higher spatial resolution of 1 μm, results could still be acquired through LIBS on a steel area of 100 μm2. A handheld metal detection instrument based on microjoule high repetition frequency LIBS has been described.16 The instrument used a Raspberry Pi as the control core and a laser with a frequency of 10 kHz and a single pulse energy of 100 μJ as the excitation source. A mini-putter was built into the instrument to move the laser, allowing the ablation of the sample surface line area without external auxiliary equipment. The excitation-generated plasma radiation was collected by a simple optical path and transmitted directly to the spectrometer. A back-propagation neural network (BPNN) model was also designed and trained, based on 12 different grades of alloys and used in the feedback process to the Raspberry Pi. This resulted in the rapid classification of the 12 alloys with >95% accuracy on the handheld instrument. A low pulse energy and high repetition rate diode pumped solid-state laser has been utilised for trigger-free LIBS based quantitative analysis of common steel grades.17 Due to the low pulse energy of 2 mJ, the signal was accumulated over a large number of pulses to obtain a good signal quality and was acquired in a free-running mode, i.e. without any triggering. Although the trigger-free signal acquisition lacked a continuum suppression capability, the weak plasma events caused by the low pulse energy did not produce a significant continuum and hence continuum suppression was not an issue. Further, the 2 kHz repetition rate of the laser required the sample to be continuously scanned to avoid signal decay as a result of the formation of excessively deep craters. As scanning could introduce additional signal fluctuations, a combination of sample scanning velocity and pulse energy leading to the least signal fluctuations was used to acquire LIBS spectra for the steel samples. Of the 19 samples, 15 were used to train random forest regression models for the quantitative analysis of various elements, including C, Co, Cr, Cu, Mn, Mo, Ni, and Si. The remaining 4 samples were used for testing and validating the models. The authors noted that when compared to the full spectral range from 185 to 545 nm, a smaller and information-rich spectral range from 185 to 310 nm produced better results. With the selected spectral range, the relative mean error ranged from 1.4% to 16%, with the lower values belonging to the major elements and the higher values to the minor elements. The rmse values were found to range from 0.014 wt% for C to 0.26 wt% for Ni. The R2 values, which indicated the correlation between the predicted and reference concentrations, were found to be >96% for all the elements except Si and Cu.
The impact of Ar flow rate when using 1.5 m long probe gun with LIBS to determine C, P, and S in molten steel was reported by Zhang et al.18 An optimum gas flow rate of 11 L min−1 was determined which gave the most intense spectral line and stable spectrum. Using this flow rate, C, P, and S were then detected and quantitatively determined in 14 standard steel samples. After internal standard calibration, the obtained LOD of the three elements were <0.000%, 0.04%, and 0.015%, with RSDs of 2.3%.1.1% and 1.0% respectively.
Numerous studies utilising chemometric methods to improve the accuracy of the calibration in LIBS have been presented. One such example is where LIBS measurements are hampered by matrix or self-absorption effect, limiting the precision of linear analytical processes. To overcome this Chen et al.19 have proposed a nonlinear process based on the S-transform to incorporate nonlinearity into the data analysis process. The approach integrated a feature selection unit based on the spectral distance variable selection method (SDVS), a nonlinear processing unit based on the S-transform (ST), and a PLS regression model. To demonstrate the improvement in accuracy achieved through nonlinear processing, a comparative analysis involving five models, Raw-PLS, SDVS-PLS, ST-PLS, SDVS-ANN, and SDVS-ST-PLS was conducted. The results revealed a significant improvement in the performance of the SDVS-ST-PLS model, facilitating wider application. A new method combining the sparrow search algorithm optimised kernel extreme learning machine (SSA-KELM) and LIBS has been proposed to analyse and model the elemental content of 12 groups of steel samples, including medium-low alloy steel and low alloy steel.20 A portable LIBS spectrometer was used to collect data from 12 different steel scrap samples, with 28 different locations on the surface of each sample selected for analysis. The k-value check was used to eliminate gross errors, and the remaining data was averaged to obtain 336 groups of spectrum data from 12 sample groups. Then, the obtained spectral data was subjected to baseline correction and normalisation to reduce the baseline fluctuation. Multiple related spectral lines of the target elements were selected as the input features of the model, and the spectral data was divided into training and testing sets. A random sample from each steel type was selected as the model’s testing set, and the remaining data was used as the model’s training set. The sparrow search algorithm was used to optimise the parameters of the KELM model. The final model for Al, C, Cr, Cu, Mn, Ni, Si, Ti, and V had an average correlation coefficient (R2) and rsme of 0.996 and 0.016, respectively, on the validation set. The quantitative analysis performance of the single variable calibration model and the genetic algorithm optimised KELM (GA-KELM) multivariate calibration model was compared, and the results showed that the SSA-KELM model gave significant improvements in all indicators compared to the single variable calibration model and GA-KELM model. Thus, the combination of KELM and Sparrow search algorithms could effectively reduce the interference of multiple factors on the target elements and enhance the performance of the quantitative analysis. Two publications have reported on the use of a random forest regression models for the analysis of steels, where multiple decision trees were combined into a single ensemble. In the first report,21 the rmse of cross-validation criterion was first used to select the spectral range of the spectral variables for random forest model input, to prevent over-fitting of the random forest model when only a few relevant variables are accompanied by many other variables. Second, the out-of-bag error criterion was used to optimise the numbers of decision trees and characteristic variables in the random forest model, which optimises the random forest structure. The availability of a large amount of relevant spectral information, coupled with the remarkable regression capacity of random forest, greatly improved the carbon analytical accuracy. The results showed that the rmse of prediction was 0.034 wt% for the calibration curve method and 0.023 wt% for the random forest method; the reduction afforded by the latter method was 32.4%. The second study, reported by Jin et al.,22 addressed problems due to laser energy fluctuations and spectral interference when using univariate regression analysis in LIBS, and a convolutional neural network (CNN) model for total Fe content prediction in iron ores. Overall, 339 batches of iron ore samples from five countries were obtained and 2034 representative spectra were collected. The performance of variable importance random forest (VI-RF), variable importance back propagation artificial neural network (VI-BP-ANN), and CNN-assisted LIBS in predicting the total Fe content of iron ores was compared. The coefficient of determination, rmse, mean relative error, and modelling time were also included for evaluation. The result showed that variable importance significantly enhanced the quantitative accuracy and reduced modelling time compared to traditional BP-ANN and random forest models. Moreover, the CNN model outperformed manual feature selection methods (VI-BP-ANN and VI-RF), exhibiting the shortest modelling time, highest coefficient of determination, lowest rmse, and mean relative error. The CNN model’s unique characteristics, such as weight sharing and local connection, made it well suited for analysing high-dimensional LIBS data in multivariate regression analysis. The approach demonstrated the effectiveness of machine learning and deep learning approaches to improve the accuracy of LIBS for total Fe content prediction in iron ores.
Potential problems may arise from LIBS when dealing with redundant or irrelevant features in the data. To enhance the accuracy and interpretability of multivariate classification, a study by Lin et al.23 introduced a hybrid feature selection method that combined the filtering characteristics of the select percentile algorithm with the embedded advantages of the elastic net algorithm. Under this framework, the support vector machine (SVM) algorithm was applied for classification, and demonstrated an accuracy, precision, and machine learning evaluation F1 score of 0.9888, 0.9895, and 0.9889 on the test set, respectively. The study also introduced the local interpretable model-agnostic explanations method which allowed for the visualisation of the importance of each variable, thereby enhancing the interpretability and credibility of the model. Overall, the model and methods proposed showed effectiveness in eliminating redundant or irrelevant features and in precise classification.
The classification of 10 kinds of special steels utilising LIBS combined with particle swarm optimisation (PSO) and SVM has been reported in a Chinese language journal.24 The SVM learned and modelled the spectral data to obtain the rapid steel classification model. However, due to the element composition of different special steels being similar and complex, the performance of classification results may be significantly affected by SVM model parameters. Two different methods of PSO and grid search optimisation were used to optimise the model parameters and speed up the training efficiency. Then, the spectral intensity of 17 characteristic lines of 6 major trace elements (Cr, Cu, Mn, Mo, Ti and V) in samples and 17 feature information variables extracted from the LIBS spectrum data with full variables by PCA were chosen as the input to establish the PSO-SVM, PSO-PCA-SVM, PCA-SVM and SVM models for steel classification respectively. The experimental results showed that compared with the SVM model’s optimisation time of 115 s, the shortest optimisation time of PSO-SVM was 11.5 s. The classification accuracy (96.67%) was not significantly inferior to the accuracy of the PCA-SVM model (97.5%), indicating that LIBS combined with the PSO-SVM algorithm could achieve rapid and precise steel classification.
A further study aimed at improving prediction accuracy when using raw full spectrum data model input has been presented by Wu et al.25 Two engineering techniques (minimum absolute shrinkage and selection operator regression LASSO and sequential backward selection) were combined with machine learning for the quantitative analysis of Cr, Ni, and Ti, in 7 stainless steel samples with different composition. Seventy LIBS spectra were obtained, and four different data preprocessing methods were compared: Maximum Minimum Normalisation, Standard Normal Variation, Savitzky Golay Smooth Filtering, and Internal Standard Method. Finally, Savitzky Golay smoothing filtering was chosen for spectral preprocessing. Effective variables were independently selected for different quantisation elements when selecting features using LASSO and SRS algorithms. Then, three different feature combinations, full spectrum, LASSO selection feature, and sequential backward selection feature were used as inputs to the model. The results showed that the model inputs selected by the two feature selection methods showed better prediction accuracy and stability compared to full-spectrum inputs in different machine learning models. Among them, the LASSO-PLS model achieved the best prediction accuracy in the quantitative analysis of Cr, Ni, and Ti, elements, with an average relative error of 3.50%, 2.66%, and 0.93%, and RSD of 4.55%, 5.23%, and 2.04% respectively.
The accurate evaluation of heat-resistant steel deterioration using LIBS is of importance for the safe operation of high-temperature pressure equipment. Understanding how the plasma impacts on matrix properties and utilising the plasma information can lead to achieving more effective detection methods. In a study by Cai et al.,26 the plasma evolution and pulse fluctuations of a typical heat-resistant steel, T91, were studied using plasma images to understand the different evolution stages and characteristics of the plasma. Specimens with different aging grades were employed to investigate the expression form, evolution and identification of matrix information on plasma. Subsequently, the plasma images and the RSD images based on pulse–pulse RSD were employed to build an aging grade evaluation model, for which the best model accuracies were 96.6% and 96.0%, respectively. A model combining these two image features achieved the highest accuracy at 99.8%. The effects of the delay time, region selection, and data coupling strategy on model performance were also investigated. The results indicated that the temporal-spatial characteristics, identification, and stability of plasma information had a significant effect on the performance of the model.
The surface hardness imaging of a low-alloy steel by LIBS has been reported by Retterath et al.27 Conventional tactile hardness testing methods such as Brinell, Rockwell or Vickers rely on direct mechanical contact, which results in significant surface damage and prohibits their application to complex specimen geometries. Furthermore, the sample must have a certain thickness for tactile testing methods to be applied. In this study, the authors investigated the use of LIBS as a fast non-quantitative alternative for visualising surface hardness gradients. To eliminate the influence of different chemical compositions, they manually heat-treated low-alloy steel pieces cut from the same raw material and batch. By partially quenching a sample piece, they were able to obtain a hardness gradient along the longitudinal axis. A positive correlation between the ratio of ionic to atomic line intensities of Fe (Fe II 263.1 nm/Fe I 358.1 nm) and the mechanical hardness of the sample surface was found. By scanning the surface and measuring the line intensity ratios, it was possible to obtain a spatially resolved map directly correlating with the surface hardness distribution. Additionally, the authors reported that the irradiation of laser pulses resulted in significant surface alterations, thereby invalidating subsequent measurements and scans at identical positions. Sdvizhenskii et al. have investigated the impact of the melt temperature of superheated pig iron on the analytical performance of LIBS.28 The LIBS measurements were conducted on molten pig iron samples at temperatures representative of the upper (1550 °C) and lower (1350 °C) limits of the molten iron temperature range in steel production. For the quantitative determination of silicon in molten pig iron, samples were obtained from blast furnace production, and it was observed that higher accuracy and linearity of calibration curves were achieved at higher temperatures, potentially due to improved mixing of the metal volume in the induction furnace crucible.
Several workers have reported on the coupling of plasma acoustic emission signals with LIBS, including two papers by Xiong et al. In the first of these, the focus was on improving the accuracy of machining process identification for low roughness samples.29 Plasma acoustic emission signals (PAESs) with bimodal information were proposed, and the LIBS spectral data and PAES data of nine types of low roughness samples processed using three machining processes; horizontal milling, plain grinding, and vertical milling. The spectrum intensities of the primary element Fe and trace element Mn, as well as the PAES maximum peak, were compared and examined. Using the PCA-SVM machine learning technique, the three recognition impacts of single LIBS data, single PAES data, and LIBS-PAES bimodal data fusion were examined and compared. When compared to single-modal data recognition, bimodal data fusion greatly improved the recognition ability, reflecting the benefits of bimodal data fusion. In the second publication,30 the technique was used to classify and identify eight steel samples. The LIBS spectral data and PAES data of the eight samples were recorded synchronously, and three mid-level data fusion strategies were proposed: additive fusion, splicing fusion, and multiplicative fusion. The authors discussed the results obtained using machine learning algorithms. The average accuracy of classifying a single LIBS spectrum was 72.5%, whereas the average accuracy of classifying a single PAES data was 78.8%. By combining LIBS spectral data and PAES data in the middle layer, the average accuracy of the splicing fusion classification result was 87.5%, and the average accuracy of the multiplication fusion classification result was 86.25%. The study also found that thermal hardness may be an important physical factor affecting the acoustic emission signal of steel plasma. An assessment of the metal grain size of 12Cr1MoV steel by LIBS coupled with acoustic wave information has been reported by Tang et al.31 The 12Cr1MoV steel was selected as the experimental sample because it has different grain size grades. Spectral and acoustic data were recorded during the laser ablation process. Initially, it was revealed that the acoustic energy did not exhibit a significant downward trend with the continuous laser shots, but the acoustic energy fluctuations became more intense. To enhance the capacity to assess the grain size grade of heat-resistant steel, the researchers integrated acoustic data with spectral data. Two data fusion strategies were proposed: first, dimensionality reduction followed by combination, and second, combination followed by dimensionality reduction. Subsequently, two classification models, linear discriminant analysis (LDA) and SVM, were constructed utilising three data types: spectral data, acoustic spectral data, and the aforementioned combined data set. The performance of the model trained on the combined data obtained based on the first strategy was found to be superior to models trained on a single data type (spectral data or acoustic spectral data), achieving a classification accuracy of 92.3%. The second strategy yielded unsatisfactory results due to the significant difference in dimensions between spectral data and acoustic spectral data, however following modification gave a classification accuracy of 98.9%.
In an ambitious study by Riedo et al.,32 fused deposition modelling technology for 3D printing was employed to produce an ion optical system for use in a reflectron, later to be integrated in a space-prototype mass analyser for LA-ionisation MS. For the insulating parts, polylactic acid filament was used as printing material, while the conductive ion optical parts were printed using polylactic acid impregnated with carbon. Measurements were conducted on a stainless-steel sample (AISI 316L, 1.4435) and NIST SRM 661 to validate the performance of the reflectron. The authors found that the system performed ‘nominally’ in terms of mass resolution and detection sensitivity. Research on a bimodal fusion detection method for surface defects of metal additive manufacturing (also known as metal 3D printing) components based on LIBS has been reported by Lin et al.33 The LIBS technology was used to capture spectral information, and a high-speed camera was utilised to record plasma images to extract pertinent details from each laser event. The LIBS spectral scores were obtained via PCA and plasma image features were extracted to generate a bimodal fusion descriptor. This descriptor was employed to enhance the detection capability of three common surface defects in metal additive manufacturing, specifically holes, cracks and bulges. Building on this foundation, a mid-level data fusion technique was employed to integrate the scores of LIBS spectra derived from PCA with seven features extracted from plasma images, resulting in the development of a bimodal fusion approach. Subsequently, the distribution of spectral data, plasma image features and bimodal fusion descriptors was discussed. Finally, three models (random forest, SVM and LDA) were used to evaluate the recognition accuracy of component defects. The results indicated that the LDA model, utilising bimodal fusion descriptors, yielded the most effective classification and the method improved the recognition of different defects of metal additive manufacturing components.
Several groups have reported on studies to investigate the corrosion of surface coatings on steel. An evaluation of UV-fsLA-ICP-TOF-MS for fast multi-elemental mapping of Hastelloy coated carbon steel (Cr–Fe–Ni–Mo–W matrix produced using laser cladding technology) immersed in a liquid corrosive environment for two months has been made.34 The UV-fs laser was employed to reduce the ablation thermal effects, improving the spatial resolution in the elemental analysis of the coating, interface and substrate. The TOF-MS was equipped with ion beam attenuation grids to expand its linear dynamic range, allowing the determination and spatial distribution of major (Co, Fe, Ni), minor (Mn, Si) and trace elements (Cu, In, Zn) to be determined. Quantification was achieved through relative sensitivity factors calculated using matrix-matched external reference materials and making use of measurements carried out by SEM-EDS to determine the concentration of internal standard elements. Qualitative and quantitative elemental maps highlighted diffusion processes at the coating-substrate interface. Moreover, correlated spatial distributions of ion signals from external corrosive elements (e.g., Ba, Bi, Hg, S and Tl) were detected in localised regions within the material and on the external surface. The corrosion of carbon steel SA106 Gr. B in oxalic acid under the anticipated conditions of a pressurised heavy water reactor during chemical decontamination has been reported.35 The carbon steel specimens were exposed to acidic medium for 18 h at 95 °C to evaluate the corrosion activity. Dissolved Fe was determined in 10 and 20 mmol per L oxalic acid solutions and was found to be 276.5 and 401.3 mg L−1, respectively. Corrosion rates were assessed using the weight loss method. The carbon steel specimens dissolved in 20 mmol per L oxalic acid exhibited a corrosion rate of 6.3 mm per year, while that for samples dissolved in 10 mmol per L oxalic acid was 4.6 mm per year. The corrosion products were characterised by EPMA. The intergranular corrosion was observed prior to the formation of corrosion products and oxalate passivation, potentially providing pathways for contaminant diffusion into the bulk substrate. The performance of the cationic surfactant benzyldimethylstearylammonium chloride (BDSAC) on the inhibition of corrosion of cold rolled steel in 0.10 mol per L trichloroacetic acid solution has been studied utilising weight loss, electrochemical techniques, surface characterisations and theoretical calculations.36 It was found that the maximum inhibition efficiency of 50 mg per L BDSAC reached 96% at 20 °C. The performance of inhibition decreased with increasing temperature. The BDSAC acted as a mixed-type inhibitor, which efficiently retarded both cathodic hydrogen evolution and anodic dissolution. The use of SEM, AFM, confocal microscopy and contact angle measurements confirmed that the BDSAC effectively reduced surface roughness and increases hydrophobicity. Measurements by TOF-SIMS and XPS confirmed the adsorption of BDSAC. The BDSAC effectively adsorbs on metal surface, slowing the migration rate of corrosive ions (H3O+ and Cl3CCOO−) in the adsorption film. An investigation into the corrosion of carbon steels subjected to chloride and sulphate in simulated concrete pore solutions of different pH has been reported.37 The study employed electrochemical techniques, TOF-SIMS, XRD and DFT calculations. The results revealed that higher pH strengthened the passivation film and reduced corrosion by limiting corrosive ions adsorption. Notably, sulphate triggered corrosion at higher thresholds, while chloride alone significantly elevated corrosion rates. The presence of both ions showed mitigated corrosion risk due to competitive absorption onto steel interface of sulphate. A two-stage mechanism involving competitive adsorption and catalytic corrosion was proposed to help elucidate the interaction mechanism of chloride and sulphate during the steel corrosion process.
Continuing the topic of corrosion, a study to investigate the formation mechanism of zinc phosphate conversion coatings on pearlitic steel and ferritic iron with an emphasis on the impact of microstructure and substrate corrosion, has been reported by Alinezhadfar et al.38 The deposition parameters, pH (2 and 2.5) and temperature (50 and 70 °C) were investigated. Open circuit potential monitoring was used to identify different stages of the coating formation. Analysis by SEM and XRD showed that higher temperatures accelerated the growth of phosphate crystals, composed of hopeite (Zn3(PO4)2) and phosphophyllite (Zn2Fe(PO4)2) on both substrates. Deposition at pH 2.5 led to bulk-solution precipitation on substrates, while at pH 2, coatings were growing from the substrate surface. Electrochemical impedance and ICP-MS measurements revealed that Fe corroded around 2.5 times slower than steel. Using ED-XRS and SEM, iron phosphate particles were shown to be formed on both substrates. These particles accumulated in higher amounts on steel, while Fe exhibited minimal corrosion product accumulation. Different phosphating stages were then studied using SEM and TOF-SIMS and highlighted a thicker iron phosphate layer on steel at early phosphating stages, compared to iron. Focused ion beam cross-sections of fully phosphated steel showed a porous interlayer, mainly composed of iron-phosphate, at the coating/steel interface. Zinc phosphate crystals were nucleated on this porous layer or formed by near-surface solution precipitation. The Fe substrates did not show this porous interlayer and had lower phosphatability with only near-surface solution precipitation of zinc phosphate crystals. The corrosion resistance of several types of steel (AISI 410, 321, 316L, 904L) has been determined in a liquid Bi–Li (5 mol%) alloy medium at 650 °C.39 Analysis by ED-XRS, SEM, XRF and ICP-OES characterised the steel structure and alloy composition. Three alloys (AISI 321, 316L and 904L) underwent severe corrosion in liquid Bi–Li alloy, and their corrosion rates were found to depend on the Ni content in the material. The AISI 410 steel exhibited the lowest corrosion rate of all the materials investigated, and the authors suggest that this type of steel could be considered as a reasonable structural material for work in liquid Bi–Li alloy environments. The corrosion rates of AISI 410, 321, 316L and 904L steels in liquid Bi–Li alloy at 650 °C were 77, 244, 252 and 280 μm per year, respectively. It was also found that Cr was etched more intensively than Fe from the surface of steel samples. A proof-of-concept paper looking at proposed polymerised structure, including tetrameric polynuclear species, of solid amorphous oxyhydroxide zirconium conversion coatings on cold-rolled steel using TOF-SIMS has been published.40 Tetramers were formed at pH near 4 (and possibly higher), with thickness increasing over extended conversion times. The use of simulated acid rain further demonstrated that optimal coating formation requires a pH of at least 4, coatings were to thin or absent at lower pH values, and a sufficient conversion time for adequate thickness. Tetramer forms were not observed when the coatings were prepared at lower pH or shorter conversion time, proving that the polymerisation step is crucial for obtaining the coatings offering adequate corrosion protection.
The analysis of austenitic and duplex stainless steels has once again received attention this year. A review of the analysis of Fe 2p(3/2) peak and other transition metals in the austenitic stainless-steel literature using XPS has been presented by Hughes et al.41 In Part 1 of the paper, the authors discuss the significant shortcomings of the most widely used approaches, based on the principle of “chemistry fitting,” where single symmetric peaks are used to represent either individual oxidation states or specific compounds. No meaningful conclusions could be drawn from these commonly employed two- or three-component peak fitting approaches; the implication being that a large portion of the literature that relies on this approach is flawed. As a significantly more accurate and reliable alternative to “chemistry fitting,” the authors also discuss “envelope fitting” (using empirical multiplet structures) and examine its limitations when applied to austenitic stainless-steel data. In Part 2 of the review the authors discuss that for other elements such as Cr 2p, the problems associated with using single components to represent oxidation states or compounds are not found to be as severe. It was found that it does not impact binding energy measurements, but does influence relative intensities, which will have a flow-on effect for oxide thickness calculations and obtaining a correct understanding of the surface more broadly. A comparison of the nitriding behaviour for austenitic (316Ti) and super austenitic (904L) stainless steels has been reported.42 In situ XRD was used to compare nitrogen low-energy ion implantation in both steels. While the diffusion and layer growth were very similar, as derived from the decreasing intensity of the substrate reflection, strong variations in the observed lattice expansion, as a function of orientation, the steel alloy, and nitriding temperature, were observed. Nevertheless, a similar nitrogen content was measured using TOF-SIMS. For some conditions, the formation of a double layer with two distinct lattice expansions was observed, especially for steel 904L. Regarding the stability of expanded austenite, 316Ti had already decayed in CrN during nitriding at 500 °C, while no such effect was observed for 904L. Thus, the alloy composition had a strong influence only on the lattice expansion and the stability of expanded austenite, but not the diffusion and nitrogen content. A study designed to investigate the interactions between hydrogen and deformation induced defects in a 316L austenitic stainless steel has also been reported.43 The studied material was pre-strained up to different levels. The resulting microstructures were characterised by optical and electron microscopy and by XRD. Dislocation densities were systematically quantified showing a more than 50-fold increase between the unstrained and the most pre-strained sample. Cathodic charging was employed to introduce deuterium into the material. Using TD spectrometry and SIMS analyses, the deuterium energetic and spatial distributions were evaluated both just after charging and after an additional isothermal aging treatment. Due to the aging treatments, three different energetic states of deuterium in the 316L SS were evidenced: interstitial 2H, low-energy trapped 2H and high-energy trapped 2H. Probing the low-energy trapping proved challenging – probably due to the proximity between the migration enthalpy and the associated detrapping energy, but aging highlighted it. This low-energy contribution, displayed in the TDS results, increased with the pre-strain level and the dislocation density, and was associated with dislocations elastic fields. Contrary to the authors expectations, the high-energy trapping could not be correlated directly to the pre-strain level nor to the dislocation density. Trapping at vacancies was suggested. Fibre laser based LIBS has been used as a non-destructive and in situ rapid detection technique for aging grade estimation of Duplex stainless steel Z3CN20-09M, used for pipeline within the first circuit of pressurised water nuclear power plants.44 The spectral signals, laser induced plasma properties, and laser ablation processes of various Z3CN20-09M specimens with different aging grades were analysed together with their microstructure and metallography. It was found that the spectral signal intensities of the matrix element (Fe) and the alloying element (Cr) increased with increasing aging grade, which was explained by the increase in plasma temperature and ablation mass. Linear relationships were obtained between the signal intensity ratios of Fe I1/Fe I, Cr I1/Cr I, and Fe I/Cr I, to the aging grades, with the coefficient of determination (R2) ranging from 0.94 to 0.99.
The use of ICP-MS analysis has been reported in several publications. A method for the determination of ultra-trace levels of Ti in bearing steel by ICP-MS-MS has been published.45 The study investigated the effects of spectral interference elimination in different collision/reaction modes (no gas, He, H2, and O2 modes). The spectral interference from Mo++ plasma on Ti+ was eliminated using the O2 mass-shift mode. An internal standard was used to correct for instrument fluctuations, and matrix interference was eliminated using the standard addition method. Under optimised conditions, the mass fraction was in the range of 0.0001–0.010%, the linear correlation coefficient was greater than 0.999, and the LOD values of the four isotopes ranged from 0.00001% to 0.00003%. The RSD was less than 5.0% when steel samples with three additive levels were measured six times. The recovery of Ti ranged from 90% to 120%. A novel electrochemical flow cell, to enable in situ surface scratching, in conjunction with ICP-MS for real time monitoring of tribocorrosion and repassivation dynamics.46 Tribochemical processes occur in various environments, where the combination of mechanical stress and chemical reactions can lead to significant changes in material properties, surface composition, roughness and technological performance. A detailed observation of chemical and mechanical tribodegradation, as well as materials repassivation are difficult to assess in situ. In this study, the cell featured an eccentrically positioned rotating sphere that scratched the surface, with ICP-MS analysis of the degraded material. Experiments were conducted on a stainless-steel sample, Fe(82.5)Cr(17.5), and demonstrated excellent repeatability. Due to tribomechanical damaging, repassivation of damaged areas showed considerably faster first order reaction kinetics and higher dissolution rates, compared to electrochemical passivation. During the scratching process, both ions and particles were observed to dissolve. While high Cr-content (≥ 12%) facilitates full repassivation, low Cr-content (<12%) samples were unable to form a stable passive layer under tribocorrosive conditions. The use of ICP-MS in single particle mode for the characterisation of micro and nanostructured materials is now a growing field of research. Morales et al.47 have used this approach to look at anisotropic structures including solid Pt-nanorods and hollowed Fe2O3-nanotubes. The structures were evaluated by SEM, HR-TEM and SP-ICP-MS techniques. Solid Pt-nanorods (191 ± 18 nm in diameter) showed important heterogeneity in their length, ranging from 42 to 72 nm, due to sample preparation difficulties. The analysis by single particle mode ICP-MS confirmed the presence of two different populations of Pt/nanorods at 19 ± 4 fg and 41 ± 5 fg, respectively, yielding a mean value of 23 ± 12 fg Pt per rod and a length range from 38 to 67 nm, in agreement with TEM measurements. In the case of the two different sized double-walled Fe2O3-nanotubes of 900 nm and 1800 nm in length, the single particle mode ICP-MS measurements provided results of 16 ± 10 and 25 ± 4 fg Fe per nanotube, respectively. From this data, the layer thickness of the Fe2O3 nanotube wall was calculated and gave values ranging between 20 ± 6 and 17 ± 4 nm, respectively. This was in good agreement with the TEM estimations (18 ± 4 nm). A variety of techniques have been used to investigate non-metallic inclusions in steel which may have a detrimental effect on the processing, mechanical properties, and corrosion resistance of the finished product. The potential of elemental and isotopic fingerprinting to trace the sources of macroscopic oxide non-metallic inclusions found in vacuum arc remelting treated steel ingots using SEM-EDX, ICP-MS, LA-ICP-MS, and LA-MC-ICP-MS has been investigated by Walkner et al.48 Main and trace element content and 87Sr![[thin space (1/6-em)]](https://www.rsc.org/images/entities/char_2009.gif)
:86Sr isotope ratios were determined in two specimens of macroscopic non-metallic inclusions, as well as in samples of potential source materials. For both specimens, very similar results were obtained, indicating a common mechanism of formation. The inclusions were thought to be exogenous in origin and were primarily composed of Ca–Al oxides. They appeared to have undergone chemical modification during the casting and remelting process. The results indicated that particles from the refractory lining of the casting system most likely formed the macroscopic inclusions, possibly in conjunction with a second, Ca-rich material.
The use of external particle induced gamma ray emission (PIGE), using a proton beam of 3.5 MeV, has been reported for the determination of the mass fractions of Fe and B as well as the isotopic composition of B in synthetically prepared ferroboron (Fe–B) alloys.49 Powder samples were wrapped in a thin Mylar foil and kept in front of the target flange in the atmosphere for irradiation using an external PIGE set up. Boron mass fractions determined using 429 keV gave better precision than those determined using 718 and 2125 keV with 136 keV of Ta as a current normaliser after irradiating the samples only for 10–15 minutes, owing to a high proton-induced thick target gamma ray yield of 429 keV at 3.5 MeV. Mass fractions of Fe, the major constituent of the Fe–B alloy, were also determined simultaneously with B. The advantage claimed for the method are the direct analysis of powder samples with minimal sample preparation providing high sample throughput and faster results with adequate accuracy and precision.
Several groups have reported research related to the distribution of elements in welds in steel. A study to explore the use of a high-temperature arc, generated during tungsten inert gas welding, to fuse a previously electrodeposited Ni/TiO2 coating to the surface of AISI 1020 steel to enhance the surface mechanical properties has been reported by Cooke et al.50 The surface was characterised by SEM equipped with EDS, an optical microscope equipped with LIBS, Vicker’s microhardness number (VHN), and pin-on-plate wear testing. The treated surface exhibited a unique amalgamation of hardening mechanisms, including nanoparticle dispersion strengthening, grain size reduction, and solid solution strengthening. The thickness of the electrodeposited layer appeared to strongly influence the hardness variation across the width of the treated layer. The hardness of the treated layer when the Ni coating contained 30 nm TiO2 particles was found to be 451 VHN, validating a 2.7-fold increase in material hardness compared to the untreated substrate (165 VHN). Similarly, the treated surface exhibited a twofold improvement in wear resistance (9.0 × 102 μm3 s−1), making it substantially more durable in abrasive environments than the untreated surface. Microstructural and EDS analysis revealed a significant reduction in grain size and the presence of high concentrations of Ni and TiO2 within the treated region, providing evidence for the activation of several strengthening mechanisms. The in situ quantitative analysis of elements in X80 pipeline steel welds has been published in Chinese by Han et al.51 This steel is the main material used in long-distance oil and gas transmission pipelines and there is a need for a fast and accurate in situ quantitative analysis method for the distribution of elements along the weld seam. The paper proposed a method for in situ quantitative analysis of Al, Cr, Mn, Nb, and Ni in the welds using LA-ICP-MS. By optimising the laser pulse frequency to 20 Hz, laser energy to 100% (laser output mode Image Aperture), etching aperture of 100 μm, and zero defocus distance, the strength and the stability of mass spectrometry signals were enhanced. Calibration was achieved using standard samples matched with the matrix, and the matrix element 57Fe was used as the internal standard for correction. The method was then used to determine the distribution of element content in two X80 pipeline steel welds with the same welding material but different base material compositions. The correlation coefficient of this method ranged from 0.9927 to 0.9996, with a quantification limit of 0.23 to 2.57 μg g−1. The results showed that Al, Cr, Mn and Nb, with similar contents in the two base metals, exhibited similar dilution at the root of the weld. In comparison, Ni elements with significant differences in content between the two base metals showed significant differences in content within 8.4 mm from the root of the weld. The impact test results showed that the weld toughness with a high Ni element content in the base material was significantly higher. Analysis by SEM of weld roots showed that the increase of Ni element content was conducive to forming a lath bainite structure and the weld impact toughness could be improved by higher Ni content by promoting the low-temperature lath bainite transformation.
3.2. Non-ferrous metals and alloys
The analysis of numerous non-ferrous metals and alloys have been reported during this review period. As with the analysis of ferrous metals, the most common technique used has been LIBS, although other investigators have employed LA-ICP-MS, SIMS and XRF. A wide range of statistical methods have also been used to help interpret data and improve both accuracy and sensitivity.In a study by Li et al.,60 2D structured light generated using a spatial light modulator was used to investigate the 2D laser energy distribution on the sample surface and the impact on the plasma properties and spectral emission. Various focal patterns with different cross-section areas were tested using metal samples so that the effects of the focal spot structure and cross-section area could be separated. The arrow-target and anticone of 200 μm diameter resulted in a line enhancement factor of more than 3 for a pure Al sample and a factor of about 6 for the pure Cu sample over that without beam modulation. A significant improvement in the signal-to-noise ratio and a reduction in the LOD of Al, Cr, Cu, Mn, and Si by a factor of about 6 were also achieved with the bearing steel sample. To better understand the mechanism behind such improvements, time-resolved spectra were measured, from which the electron density and plasma temperature evolution were calculated. The results showed that with anticone and arrow-target patterns, the electron density was increased by more than 50%, while the plasma temperature varied less than 500 K, suggesting an evident ablation enhancement. The fast images of plasma self-emission proved that the laser energy distribution substantially influenced the morphology of the plasma plume. The plasma cores produced by the anticone and arrow-target focal patterns were more compact after the expansion and cooling process, which could reduce the mixing of plasma plumes and ambient air. Overall, substantial spectral signal enhancement and a much lower LOD were achieved.
The use of LIBS for Al alloy analysis has been reported on. The impact of sample temperature on LIBS emission characteristics and calibration-free LIBS quantitative analysis in vacuum and air has been studied.61 Using an Al–Sn–Cu alloy, the researchers analysed spectral lines within a temperature range from 20 to 170 °C, and evaluated the impact on emission line intensities, plasma parameters, expansion imagery, and laser ablation crater morphologies. The results indicated that even at elevated temperatures, the calibration-free LIBS remained quantitative even for minor elements with weak lines. At atmospheric conditions, Sn segregation occurred, yet calibration-free LIBS remained robust. A discriminative learning approach based on LIBS, utilising the Discriminative Restricted Boltzmann Machine (DRBM) has been reported for spectral feature selection and classification of five distinct small-sample Al alloy samples.62 The learned spectral latent distribution from the generative model component of DRBM effectively regularised the discriminative process, thereby overcoming the problem of training overfitting arising from the high-dimensional small-sample limitation. This resulted in a stable and generalisable qualitative analysis model independent of empirical knowledge. The approach achieved a 100% accuracy, surpassing the best-performing traditional machine learning method (PCA-random forest) by 13.3% in accuracy and demonstrating a similar improvement compared to a BPNN with the same structure.
The most commonly used rare earth magnets, NdFeB magnetic materials, have been the subject of several investigations. In the production and processing of these materials, quality control on crude samples is necessary to reduce production costs. Chen et al.66 have designed and developed an on-line LIBS instrument for crude sample detection of magnetic materials. The instrument meets the requirements for product QC in workshop production. The spectra of 7 kinds of NdFeB crude material samples were obtained by the LIBS instrument. To improve the classification accuracy, the algorithm in the instrument supporting software was used. Appropriate stoichiometric methods were selected, and a sliding window minimum removal base method was independently designed to further optimise the classification accuracy. The overall optimisation improved the classification accuracy from 87% to 99%. Zhou et al.67 have used a combination of LIBS and random forest models to investigate the quantitative analysis of four REEs (Nd, Pr, Tb and Dy) in NdFeB alloys. Firstly, the original LIBS spectra were screened by PCA-mahalanobis distance (PCA-MD). The effects of different data processing methods on the screened LIBS spectra were then explored, and the feature variables were extracted from the pre-processed spectral data by the variable importance measurement (VIM). To further verify the prediction performance of the model, the prediction results of the random forest models based on the different methods were compared. Finally, a PCA-VIM-random forest calibration model was established on the basis of the optimised spectra, selected feature variables and parameters. The results showed that the PCA-VIM-random forest model has better prediction performance than the random forest calibration model based on the raw spectra. A study addressing the sorting of solid used NdFeB magnets upstream of a recycling process has been published by Sirven et al.68 Dysprosium was chosen as representative of heavy REEs, and LIBS was selected to perform online quantitative analysis under a residual coating or oxidation layer when present. The studies were carried out using two different LIBS systems. The first one was a commercial laboratory equipment with a short-range working distance of 30 cm. Laser ablation conditions were optimised based on the LIBS signal intensity and stability, on the shot-to-shot signal evolution, and on the quality of the calibration curve. This led to 60 pre-ablation shots and 40 analysis shots. Under these conditions, the authors demonstrated the feasibility of quantifying the Dy content under a depth of 35 to 60 μm with a relative uncertainty of less than 10% at 95% confidence. The second LIBS system was a made in-house and designed to make measurements at 1 m distance, a working range representative of a possible online implementation above a conveyor belt. The ablation efficiency of this system was found to be much too low, but it was considerably increased by using the laser in the free-running mode. In the standard Q-switched mode, the results also showed the feasibility of quantifying Dy with uncertainties of the same order than with the short-range equipment.
Alloys based on Ni, including those recovered from waste steams have been studied. A pretreatment method for use with ICP-MS has been developed using a micro-reaction sealed digestion vessel under micro-pressure.69 The method was used for the determination of As, Bi, Ga, In, Pb, Sb, Sn and Tl in Ni based superalloys. The mass of the samples and the volume of acid used to dissolve the superalloys were significantly reduced. Internal standards were Sc, Re and Rh, calibration standards were matrix matched to the samples and polyatomic ions were negated using He collision mode. The linear correlation coefficients of the calibration curves were above 0.9995, with linear coefficients ranging from 0.2 to 100 ng g−1. The LOD values for the method ranged from 0.0042 to 0.13 μg g−1, and the quantification limits ranged from 0.014 to 0.41 μg g−1. The micro-reaction test results of the various groups were consistent with those of traditional methods, and the RSD (n = 11) was less than 10%. An ultrafast pulsed LA has been utilised to modify the surface of a nickel-based alloy (VDM (R) alloy 699 XA) by changing its surface chemistry and structure.70 The processed surfaces were characterised by a SEM, ED-XRS, mechanical profilometer, Raman microscopy, and X-ray spectroscopy. The surface roughness at the bottom of the ablated cavities was lower than 0.3 μm. The calculated ablated depth per pulse was 2.34 × 10−6 and 7.81 × 10−6 μm per pulse for laser fluences of 1 and 10 J cm−2, respectively. Laser-induced periodic surface structures covered the bottoms of the fully ablated cavities. The use of EDX showed that the elemental distribution of alloy 699 XA was slightly reduced after laser surface ablation. The XPS analysis showed that the amount of Ni and Cr increased as laser fluences increased. It also indicated that surface oxides, such as Cr2O3 and NiCr2O4, were formed. The manufacture of a 97% pure nickel–cobalt alloy by bringing together three major waste streams, Ni-MH batteries, e-waste plastics, and waste glass has been reported.71 A temperature of 1550 °C aided the reduction of nickel-oxide using e-waste plastic as the reductant and sent rare earth elements present in the waste Ni-MH battery as an oxide mixture to the slag phase. Waste glass powder used in this process functioned as the fluxing agent, so additional flux was not required. The reduction mechanism was gas-based, controlled mainly by hydrogen and carbon monoxide gases released. Formation of the nickel alloy and the enrichment of slag with mixture of rare earth oxides was confirmed by XRD and SEM-EDS whilst ICP-OES and LIBS confirmed the high metal content in the alloy, with a purity (98%) close to the composition of nickel super alloy.
The determination of Cr content in Ni base alloys has also received attention. Chromium has multiple roles in nickel-based alloy, in particular increasing the wear resistance of an alloy coating. In addition, Cr can form a dense protective film under the action of high-temperature gas, which significantly improves the alloy’s resistance to high-temperature oxidation and thermal fatigue performance. However, in the case of a high content, Cr easily forms harmful phases with Al, Mo, Ti, and other elements in the alloy and reduces its strength. Inconel 718 is a Ni–Cr–Fe alloy containing 6 other elements, some at trace levels. Two group have reported on the effects of heat treatment (selective laser melting) on the microstructure and mechanical properties of this alloy. The aim of the first study was to investigate the microstructure evolution and mechanical properties of Inconel 718 under different heat treatment conditions.72 A range of analytical techniques were employed including: SEM, electron backscattered diffraction, ED spectroscopy, XRD, TOF-SIM, and TEM. The experimental findings revealed the presence of cellular high-density dislocation substructures in the as-received specimens, with a significant accumulation of Laves phase precipitates at grain boundaries and subgrain boundaries. After the double aging treatment, the cellular substructure persists, with higher concentrations of γ′′ and γ′ strengthened phases compared to an as-received specimen. Conversely, the second sample types, solid solution followed by double aging, underwent almost complete recrystallisation, resulting in the dissolution of brittle Laves phases and a substantial increase in the content of strengthening phase γ′′ and γ′. As a consequence of the precipitation of the γ′′ and γ′ strengthened phase and the modification of the microstructure, the material exhibited enhanced strength and hardness, albeit at the expense of reduced plasticity. Overall, the relationship between heat treatment processes and precipitation behaviour indicated that the double aging heat treatment yielded favourable mechanical properties with a balance between strength and plasticity. In the second study, by Beyhaghi et al.,73 the detailed structure and time evolution for forming the passivation layer during oxidation of selective laser melted (SLM-IN718) and commercially cast (Comm-IN718) Inconel 718 at 650 °C, was presented. This layer was shown to have a complex 3D structure, with a top layer of Fe–Ni chromite spinels (NiFexCr(2 − x)O4) and a sublayer of Cr2O3. Data obtained by SIMS revealed the depth hierarchy of the passivation layer profile, with the chromite layer covering a thicker Cr2O3 layer forming above the niobium-rich intermetallic layer above the alloy. Since passivation layers are critical to the endurance of superalloys during extreme thermal cycling conditions, the authors suggest that determining the complex stoichiometry and hierarchy of the layer can guide the development of durable superalloys with improved resistance to extreme thermal cycling conditions. The determination of Cr content in Ni base alloys using a matrix matching calibration method with ICP-OES has been reported in a Chinese language paper.74 Nickel base alloy samples were digested by microwave digestion using HF, HCl, and HNO3. Given the potential interference from high concentration elements such as Fe, Mo and Ni when determining Cr, the 266.602 nm analytical spectral line was selected to give fewer interferences but with a reduced sensitivity. Matrix and interference effects were systematically investigated. The method’s accuracy was evaluated using CRMs, and the reliability of the method was verified by comparative analysis using the YS/T 539.4-2009 standard. The method gave a LOD and LOQ of 0.05 and 0.10 μg mL−1, respectively, with an RSD of <2.5%. The method broadened the detection range for the determination of Cr content in nickel-based alloys by ICP-OES and lowered the detection limit from 2.0% to 0.1%.
This final section covers a range of alloy type and instrumental techniques. High-entropy alloys (HEAs) are those that consist of five or more principal elements in near-equiatomic proportions that tend to form simple solid solutions during solidification owing to high mixing entropy. Laurent et al.75 have used TOF-SIMS to analyse HEAs, offering detailed chemical information about the top layers of a sample. Samples of Ru–Pt–Pd–Ir–Rh-based porous HEAs with different mixing characteristics were analysed alongside seven simpler alloys. An automatic quantification process was developed to effectively deal with the large data sets obtained from the ToF-SIMS spectra of HEAs, allowing for accurate and reliable data interpretation. The complex chemical fingerprint obtained from the alloy surface was translated into a matrix showing the individual isotopic mass distributions of each polyatomic chemical species. From this, the authors introduced two key metrics: the cluster ratio and the oxide ratio, to quantify respectively the degree of atomic-level mixing and the surface oxidation. The results revealed that increasing elemental complexity (number of different elements in the alloy) enhanced atomic mixing. Moreover, HEAs with enhanced atomic mixing were shown to be more resistant to surface oxidation. Femtosecond LA spark-induced breakdown spectroscopy (fs-LA-SIBS) has been employed for the quantitative analysis of magnesium alloy samples.76 As with other metals discussed above, machine learning models: Random Forest, SVM, PLS, and k-Nearest Neighbours were used to evaluate their classification performance when identifying the Mg alloys. Performance evaluation was based on sensitivity, specificity, and accuracy. The results indicated that the Random Forest Regression model performed optimally for regression tasks, while the Random Forest Classification model outperformed other models in classification tasks. Several data processing strategies for the improved analysis of solder by ETV-ICP-OES were reported by Moghadam et al.77 Techniques for data processing have traditionally applied internal standardisation and some form of correction (e.g., blank subtraction) to compensate for positive bias from the background. However, a blank may not always be easily obtained in applications of solid-sampling research, and in some cases, degrade detection limits and signal integrations. Firstly, point-by-point internal standardisation with an Ar emission line was used to determine the sensitivity and accuracy of the solder analysis by ETV-ICP-OES. Then, in Method A, peak area with average blank subtraction from empty graphite boats was used. In Method B, peak area with integrated background correction was used, and finally, in Method C, peak height with averaged background correction was employed. Despite being the simplest to implement, subtracting the average background signal from the height of the peak produced during the vaporisation step, i.e. Method C, systematically yielded the lowest LODs without compromise in accuracy or precision.
4. Organic materials
4.1. Organic chemicals, paints and explosives
Several papers discussed the analysis of paint or have described methodology for the analysis of paint layers or for the monitoring of paint layer removal. In one example, Seney et al. used a portable XRF instrument to determine Pb in fresh paint samples.78 Traditional methods using ICP-OES would normally require an acid dissolution of the paint prior to analysis. This leads to protracted sample preparation protocols and the use of instrumentation that is out of budget for some laboratories. The potential to use a relatively cheap instrument that may analyse the sample directly, i.e. that does not require sample preparation, is therefore attractive. Results obtained using the portable XRF instrument were compared with those obtained using dissolution followed by ICP-OES analysis. A very good correlation between the data sets was obtained, but the XRF data were routinely 27% lower than those for the standard method. The authors therefore used the correlation equations to “recalibrate” the XRF instrument leading to an improvement in accuracy to give a 4.4% error when two CRMs were analysed. The approach was then applied to the analysis of 11 paint samples. Again, the XRF data were compared with those from ICP-OES and this time, agreement was within 10%. Another example was presented by Beck et al. who used PIXE and elastic backscattering spectrometry (EBS) simultaneously as well as computer simulation to demonstrate the influence of pigment grain size during the analysis of paint layers.79 Simple samples, in which well-characterised pigments in linseed oil were prepared and the elemental composition determined using PIXE and the particle size distribution by SEM. The data from these were input to the simulation software which provided elastic back-scattering spectrometry (EBS) spectra that were, in general, in reasonable agreement with experimental ones. Any discrepancies were attributed to the simplified model used that assumed the particles were spherical, whereas the SEM data showed that, in reality, they were not. Despite this obvious flaw, the authors concluded that the combination of the two ion beam techniques along with SEM data, enabled complete information on both the pigment and the binder to be obtained. It was also concluded that the simulation and quantitative understanding of EBS spectra from paintings requires the microstructure of the paint layer to be taken into account. The monitoring of paint layer removal is important if the underlying materials are not to be damaged. This is especially true when those materials, such as aircraft parts, are easily damaged. One example was presented by Liu et al. who described the laser cleaning of white automotive paint layers from their substrates using on-line monitoring methods.80 A white composite paint comprising a clear base, intermediate and epoxy primer coatings was used as an example. The power, spot overlap rate and number of laser cleanings of a 1064 nm pulsed fibre laser with 100 kHz repetition rate and 150 ns pulse width were optimised. For the outer coating combination comprising the clear and white colour coating, these values were 14 W, 50% and five times, whereas the inner coating combination comprising the mid and epoxy primer coatings required values of 20 W, 50% and seven times. Monitoring was achieved using LIBS which measured the changes in elemental peak intensity with the number of laser cleanings. Under these optimal conditions, a cleaning efficiency of 98.9% was achieved leading the authors to conclude that the methodology was rapid, efficient, safe and caused no underlying substrate damage. A similar topic was investigated by Yang et al. who reported the on-line monitoring of paint removal from aircraft skin aluminium alloy using LIBS followed by PCA – support vector regression (SVR) of the spectra and with the standard curve method.81 For a high-frequency nanosecond infrared pulsed laser and the standard curve method, five Ti lines were monitored and correlated with the paint layer thickness. The average coefficient of the segmented curve fitting of the TiII 589.088 nm spectral line was 0.89 and was the highest of the spectral lines monitored, leading to a rmse of 12.28 μm being obtained. The PCA-SVR methodology provided superior results, with a fitting coefficient of 0.97 being obtained yielding a rmse of 2.92 μm.The techniques of LIF, LIBS and ATR-FTIR were used by Wahl et al. to monitor laser cleaning to obtain enhanced adhesion of silicone adhesive in the automotive industry.82 The aluminium samples were contaminated with varying quantities of cooling lubricant and the LIBS and LIF were used as on-line tools for monitoring the laser cleaning process. Results were compared with those obtained using the off-line ATR-FTIR method and were in good agreement. The methodology enabled contamination levels of between 0.15 and 2 g m−2 to be determined.
The analysis of wood has been reported in two studies. In one by Betlej et al., LIBS was used to determine the extent of graphene oxide incorporation into wood samples.83 Wood was impregnated using two methods: vacuum and pressureless with ultrasound. The LIBS data (mainly C at 247 nm and O at 777 nm) were then used to characterise differences in the elemental composition between the surface layers of wood impregnated with graphene oxide and the untreated wood. Other microscopic techniques (stereomicroscope, confocal laser scanning microscope and SEM) were also used to characterise the surfaces of the wood. The LIBS data indicated an increase in C and O signal in the near-surface layers in the early wood cell zone although the overall signal was significantly lower for particles impregnated using the pressureless ultrasound method. This was attributed to disintegration processes. Once obtained, the LIBS spectra were subjected to a factorial analysis which was partially successful in separating differences between impregnated and un-impregnated samples and the impregnation process. The partial success was attributed to non-optimised laser parameters which, the authors stated, would be addressed in future work. The other paper to analyse wood was presented by Capela et al. who used LIBS to assess the quality of recycled wood.84 Numerous potential contaminants such as As, Ba, Cd, Cr, Cu, Hg, Pb, Sb and Ti were determined with three different analytical lines being used for each analyte. Samples were placed on translation stages where the LIBS laser (Q-switched, Nd-YAG operating at 1064 nm and using an energy of 51 mJ) produced spot sizes of 300 μm. The effects of a double plasma or air plasma during the measurements was greatly minimised in the system used by a simple extraction device. A total of 2500 shots were fired at each sample over an area of 9.6 × 9.6 cm using a step size of 2 mm, meaning that the entire surface was mapped in approximately an hour. The spectra obtained underwent a pre-processing to minimise the effects of the continuous background and Bremsstrahlung followed by asymmetric least squares smoothing. The use of three lines per analyte also helped minimise spectral interference since it is unlikely that each of the lines would be affected by any interference to the same degree. An algorithm developed in Python enabled the spectra to be interrogated so that the presence of any of the contaminants could be confirmed. Both handheld and micro-XRF were also used for comparison of data. After a preliminary sample set of 10 samples was analysed successfully in which the LIBS system and processing algorithm was optimised, a larger sample set of 100 was analysed with the LIBS data being compared with those obtained using only the handheld XRF system. Agreement was in excess of 88%. The methodology was said to contribute to more sustainable waste management practices and to facilitate the quick identification and remediation of contaminated materials.
A paper by Song et al. described a method by which the quality of spectra can be improved enabling a rapid characterisation and identification of fire-retardant coatings.85 Raman, LIBS and smartphone spectra were obtained from known samples (up to a maximum of 300) and then 85 “unknown” samples were identified using the database established. High quality Raman spectra were first obtained, and these were used to train two neural networks working in conjunction to generate useful spectral fingerprints. This combined neural network system was called Twin Spectral Reconstruction Network (TSR-Net). Seven brands of fire retardants were obtained and painted onto aluminium sheets. Some of these were artificially aged using the Chinese National standard GB 14907-2018 so that the test would be equally applicable to aged as well as new samples. The accuracy of the analysis of the test samples depended on the number of samples used to train the model, with 150 being sufficient in many instances. The accuracy of the identification was compared with those obtained using benchmark techniques such as SVM and partial least squares – discriminant analysis. The TSR-Net obtained a maximum accuracy of 0.977, which out-performed the benchmark techniques by between 0.024–0.047. The smartphone methodology developed used short videos of samples being illuminated by a colour-changing screen and converts them into spectral data. Although nowhere near as accurate as the LIBS method, it did provide an accuracy of approximately 87% and therefore was a useful pre-screening tool.
An extraction method for the evaluation of metal organic frameworks (MOFs) in impregnated samples was developed by Mukai et al.86 Traditionally, the concentration of the MOF was determined using an acid digestion followed by an atomic spectrometric method of detection. Unfortunately, this suffers from the drawback of also measuring metal that is un-associated with the MOF, hence leading to possible inaccuracy. The method developed in this paper used a two-step extraction method for copper(II) benzene-1,3,5-tricarboxylate (Cu-BTC MOF) impregnated materials to determine the MOF formed, free metals and ligands. Several extraction solvents were tested, with ethanol being the most efficient for the extraction of the free Cu2+ and the benzene-1,2,3-tricarboxylate ligand and 0.3 mol per L nitric acid was then used to extract the Cu-BTC MOF. Both sets of extractants were then analysed using FAAS for Cu determination (plus the Cu-BTC MOF) and HPLC for the unreacted ligand. Accuracy was assessed using spike/recovery experiments, with both the MOF-formed Cu and the BTC ligand yielding full recovery.
The determination of F in materials using atomic spectrometry is not simple because it has a very high ionization potential and therefore exhibits low sensitivity for most techniques. A paper by Simon et al. described the determination of PFAS using HR-CS-GF-molecular-AS.87 In general, the methodology suffers the drawback of potential losses of analyte through volatilisation prior to the molecular formation and hence, inaccuracy. In this manuscript, the authors optimised the methodology by determining the F response factor for numerous PFAS having different physical and chemical properties. Use of a magnesium modifier during the drying stage enabled the precision of the different F response factors to be improved from 55% to 27%. Since the response factors had become more similar, the accuracy could be improved. The instrumental LOD and LOQ were 1.71 and 5.13 μg L−1, respectively. Matrix effects from different water matrixes were evaluated using perfluorooctanoic acid as calibrant. A significant interference was observed from Cl at a concentration above 50 mg L−1 and therefore, this had to be separated from the analyte prior to determination.
4.2. Pharmaceuticals
The analysis of pharmaceutical materials continues to be a popular area of research. Although many methods of analysis continue to use “traditional” techniques, the use of atomic spectrometry is increasing. The analysis of illicit drugs is covered in the Forensics section of this review.A review (77 references) that discussed the emergence of ICP-MS as an alternative method of analysis to radiopharmaceutical approaches was presented by Klika et al.88 For some experiments, the use of radiochemicals is extremely useful. However, they are expensive, require specialist facilities and are prone to restrictions in use. The use of ICP-MS offers a rapid and reliable alternative. The practical applications of ICP-MS in radiopharmaceutical cancer research have only emerged in recent years. This paper focused on the development and implementation of nonradioactive ICP-MS-based assays in radiopharmaceutical research and aimed to inspire future research efforts in this area. A review (137 references) of the applications of total reflection X-ray fluorescence (TXRF) spectrometry in food, cosmetics and pharmaceutical research was presented by Marguí et al.89 The review provided sections discussing elemental analysis for quality and safety of consumer products covering the three topic areas and then went on to give an experimental introduction to transmission TXRF. This had sub-sections including elemental quantification and figures of merit as well as sample preparation and handling procedures. The bulk of the paper provided examples of numerous applications; many of which were tabulated for ease of use. The concluding remarks stated that TXRF has been used successfully for the determination of mid- and high atomic number analytes, but that further research was required to improve the detectability of the low atomic number ones.
The use of TOF-SIMS for the analysis of pharmaceuticals has been reported on this year. Two papers by Finsgar discussed the relative merits of using TOF-SIMS for the analysis of pharmaceutical products.90,91 In the first of the papers, signals associated with the spatial distribution of indapamide, amlodipine and perindopril in solid dosage form were obtained using TOF-SIMS combined with multivariate curve resolution. The spatial distribution was assessed in both 2 and 3D. The main analysis chamber was flooded with argon to a pressure of 5 × 10−7 torr and a flood gun was used to compensate for any charging effects during the measurement. Depth profiling was performed using a gas cluster ion beam using 10 keV Ar2000+ at a target current of 14 nA. This provided a crater area of approximately 500 by 500 μm, and the analysis was performed on an area of approximately 200 by 200 μm in the middle of the sputter crater. Numerous different TOF-SIMS modes were utilised in an attempt to obtain optimal data acquisition. Peaks unique to each of the analytes were identified and then used to determine the distribution of the analytes in individual pills. Distribution of the analytes was not homogeneous. Instead, it was noted that they were present as small agglomerates distributed throughout the tablet. Analysis of the excipients silica and magnesium stearate found that the silica was homogeneous, but that the magnesium stearate was also present as agglomerates. The distribution of active pharmaceutical ingredients in tablets is not often determined and has not been for these analytes and so this is an interesting application. In the other paper,90 paracetamol (also known as acetaminophen) tablets containing 500 mg of paracetamol along with methyl p-hydroxybenzoate, propyl p-hydroxybenzoate, gelatine, silica, talc and magnesium stearate were purchased from a local drugstore. They were then analysed using TOF-SIMS using Ar gas applied at 5 × 10−7 torr and a primary ion beam of 30 keV Bi3+ and XPS employing an Al Kα primary X-ray beam. Again, 3D profiling was undertaken, with the depths of the crater being measured using a stylus profilometer. The paracetamol’s characteristic signals were identified through high-mass resolution MS analysis, tandem (MS/MS) TOF-SIMS analysis and multivariate statistical analysis, utilising the multivariate curve resolution technique. Multivariate curve resolution analysis of the positive TOF-SIMS data identified the peak at 152.07 to be unique to paracetamol, whereas the negative mode identified 149.05 and 150.06 to be unique. The 2 and 3D imaging were undertaken using the positive mode because it offered greater sensitivity. The paracetamol was found not to be homogeneously distributed at the micron level. A paper presented by Lagator et al. also investigated the use of TOF-SIMS for pharmaceutical analysis, concentrating on the use of different gas cluster ion beams (GCIBs).92 Altering the carrier gas composition led to the formation of (Ar/CO2)(n) or (H2O)(n) GCIBs. The addition of a reactive species (CO2) to water GCIBs enhanced the secondary ion yield of small pharmaceutical compounds in the positive ion mode significantly. In negative ion mode, the results were less impressive, with either marginal enhancements or no change in sensitivity. It was noted that an excess of CO2 led to the formation of CO2 clusters, resulting in reduced yields compared with the water ones. As well as the amount of CO2, the cluster size was also identified as being critical for maximal sensitivity. An added advantage of using CO2 in the argon carrier gas was that the spot size decreased, i.e. better resolution was obtained. In a two-drug system (acetaminophen and diclofenac) the CO2-doped water clusters proved to be effective in overcoming matrix effects in positive ion mode compared with pure water clusters. However, this effect was not observed to a great extent in negative ion mode.
An interesting paper in which analytes that are not detectable directly using ICP-MS were first derivatised and then separated and quantified by HPLC-ICP-MS was presented by Liu et al.93 Acetaminophen tablets were first extracted and o-acetaminophenol, p-nitrophenol and p-aminophenol then derivatised. Unfortunately, the abstract of the paper did not specify what the species were derivatised with to make them detectable using ICP-MS. The results, however, were impressive, with LOD for the three species being 0.089, 0.097 and 0.161 μmol, respectively. Accuracy was described as being between 96 and 105% for all analytes and precision was better than 4.9% RSD. Results were compared with those obtained using HPLC with UV detection and these agreed within 20%.
Cauduro et al. described methodology whereby Os and other platinum group elements (PGEs) were determined in APIs.94 The samples were first acid digested with MAE ensuring that the conditions (sample mass (500 mg) and ratio of nitric and hydrochloric acids (1:1, 6 mL)) were optimal. The samples were then stabilised in 5% hydrochloric acid prior to ICP-MS analysis. This stabilising solution was suitable for all analytes except Os which required a mixture of 85 mmol per L acetic acid, 10 mmol per L thiourea and 0.6 mmol per L ascorbic acid. An evaluation of C-based interferences was then made using concentrations between 50 and 2000 mg L−1 and citric acid as the C source. The presence of C at concentrations above 800 mg L−1 increased the sensitivity of all PGEs except Os, which required a slightly higher concentration (>1500 mg L−1) for the same effect. This enhanced sensitivity could potentially lead to inaccuracies if standards are not matrix matched. Method validation through spiking experiments yielded recoveries in the range 97–105%. The LOQ values were 0.003, 0.013, 0.007, 0.001, 0.001 and 0.001 μg g−1 for Ir, Os, Pd, Pt, Rh and Ru, respectively. The advantages this method provided over existing ones were an improved LOQ (because of the high sample mass) and being able to determine Os reliably; something many other studies have struggled with.
Spray dried poly(lactic-co-glycolic acid) (PLGA)-based controlled-release injectables is a burst release material, where a significant amount of the drug is released prematurely within a short period of time following administration. This can lead to a deterioration in the performance and quality of the end product. A paper by Michaelides et al. discussed the use of the model API bovine serum albumin (BSA) added to the PLGA in different ratios to identify the sources of burst release.95 The materials produced were characterised in terms of their morphology, particle size, surface area, thermal properties, moisture content and chemical composition. Techniques such as XPS, hard XPS and argon cluster, sputtering-assisted TOF-SIMS analysis were used for the characterisations and revealed an enrichment of PLGA on particle surfaces with the BSA protein buried beneath. However, the PLGA surface was found to be prone to degradation and pore formation, and this was the cause of the initial burst release which comprised 85% of total release. Although the reason for the early release was identified, the authors did not prevent it, with increasing the polymer concentration failing to prevent the burst release.
4.3. Cosmetics
There have been several applications of the analysis of cosmetic samples published during this review period. An example was presented by Asadieraghi et al. who described a novel method for the determination of Au in cosmetics and natural water samples.96 The method was based on ion pair (IP) DLLME with detection using a quartz atom concentrator tube and FAAS. The method for cosmetic samples (skin and eye serums) involved a MAE Aqua Regia dissolution followed by dilution to 25 mL. An aliquot then underwent a syringe-to-syringe IP-DLLME protocol. Briefly, this involved sample being adjusted to pH 2.5 using hydrochloric acid, placing this in a syringe and then 213 μL of 1-octanol (extraction solvent) which contained 1550 μg Aliquat336 injected into it. A connection to a second syringe enabled the resulting solution to be passed from one syringe to another. After six transfers back and forth, the resulting solution became cloudy when the aqueous and organic phases became homogeneous. Centrifugation allowed the organic phase to be separated and then removed. A final dilution to 300 μL using MeOH followed before the Au was determined using the quartz tube FAAS method. The extraction protocol was optimised using Plackett–Burman Design followed by Box–Behnken Design and response surface methodology. Under optimal conditions, the calibration range was linear between 0.2 and 300 ng mL−1, the enrichment factor was 280, the LOD was 0.1 ng mL−1 and precision at 80 ng mL−1 was 1.89%. These figures of merit were favourable when compared with other preconcentration methods but were still insufficient to determine Au in natural water samples. Spike/recovery experiments of these samples indicated quantitative extraction. A paper by Ali and Abojassim described the FAAS analysis of 36 care product materials (baby powder, shampoo with soap and cream) and then used the data obtained for Cd, Cr and Pb to calculate chronic daily intake, hazard quotient and index, as well as carcinogen parameters such as daily intake, cancer risk and total cancer risk.97 All samples (both solid and liquid) were acid digested using a mixture of nitric and perchloric acids (10:1), diluted and then aspirated into the FAAS instrument. The equations required for the calculations were presented and results tabulated. All samples were within the safe limits specified by organisations such as Association of South-East Asian Nations, Health Canada, and the USA Environmental Protection Agency. A chemometric method entitled “Minimum Spanning Tree-Based Clustering” was used by Castello et al. to chemically evaluate commercial nail polish samples.98 A total of 45 samples were analysed for Cd, Co, Cr, Cu, Fe, Mn, Ni, Pb and Zn using LIBS and EDXRF and data inserted to the software. For the LIBS analysis, the nail polish was coated (three layers) onto glass substrate; whereas the EDXRF analysis was accomplished by placing the polish inside a polypropylene cuvette, covered with Mylar film and then dried to constant weight. Four distinct groupings were obtained from the data. Cluster one (four samples) had high levels of Fe (up to 45000 mg kg−1). Cluster two (two samples) had high levels of Co, Cr, Cu and Pb (1500–5000 mg kg−1). The other two clusters had much lower concentrations of analytes, with cluster three having concentrations of between 31 and 500 mg kg−1 and cluster four between 11 and 30 mg kg−1. The polish samples were also analysed using ICP-OES following an acid digestion procedure and these data were used to corroborate the LIBS and EDXRF data. The analytes Cd, Mn and Zn were not detected in any of the samples. The methodology allowed the rapid analysis and characterisation of the samples and was therefore regarded as being ideal for monitoring purposes. The concentrations of As, Cd, Cr, Cu and Pb in 12 skin whitening creams available in Pakistan were the subjects of a study undertaken by Ehsan et al. who determined them using FAAS following a nitric:perchloric (4:1) acid digestion.99 Analytical data were input to ANOVA in an attempt to differentiate between different samples. Arsenic had, by far, the highest concentrations (23–263 mg kg−1), whereas the Cd was not detected in any of the samples. The rudimentary statistical analysis indicated that it would be possible to differentiate between different samples. The authors hoped that their study would help consumers choose the samples with the lowest concentrations of toxic metals and hence, minimise the adverse health effects.
4.4. Fuels and lubricants
Papers for this subject were thin on the ground this year, this may be the product of petroleum fuel and coal being seen as past technology with the transfer to electric vehicles and global environmental concerns. However, petroleum fuel, and jet fuel in particular, are set to stay with us for some time due to the energy density and a range of issues with current battery and hydrogen technology. The near absence of papers on sustainable fuels, particularly sustainable jet fuels and biofuels was particularly perplexing. These subjects are a hot topics at the moment and make up a significant portion of the analytical work in commercial laboratories. There are problems around the analysis of these fuels, and they don’t just slot into the standard routine methods. It seems however, that funding for academic institutions is not available to tackle these new research possibilities, who are preferring to stick to re-inventing the wheel for standard routine methods that work well and have served the industry for many years. This seems a waste of resources and money which could be focussed better elsewhere to help make a more meaningful contribution to the industry and global emissions.A review which covers characterisation techniques employed in solid-state hydrogen storage research, giving their principles, advantages, limitations, and synergistic applications, has been published.106 In the paper conventional methods such as the Sieverts technique, gravimetric analysis, and SIMS, alongside composite and structure approaches including Raman spectroscopy, XRD, XPS, SEM, TEM and AFM were critically reviewed. The review highlighted the role of in situ and operando characterisation in unravelling the complex mechanisms of hydrogen sorption and desorption. Challenges associated with characterising metal-based solid-state hydrogen storage materials were also addressed and innovative strategies to overcome these obstacles discussed. The integration of advanced computational modelling and data-driven approaches with experimental techniques to enhance our understanding of hydrogen-material interactions at the atomic and molecular levels were also covered. This paper also provided a critical assessment of the practical considerations in characterisation, including equipment accessibility, sample preparation protocols, and cost-effectiveness. By synthesising recent advancements and identifying key research directions, this review aimed to guide future efforts in the development and optimisation of high-performance solid-state H storage materials, ultimately contributing to the broader goal of sustainable energy systems.
A paper by Endriss et al.107 described the evaluation and optimisation of an XRF analyser for the rapid analysis of chemical elements in solid biofuels. To optimise wood-fired heat and power plants, it is essential to rapidly determine the chemical composition of the solid biofuels on-site shortly before combustion. The standard procedures for chemical analysis, ICP-OES and ICP-MS, are time-consuming and expensive and XRF analysis is one of the most promising methods that may fulfil this need and has been evaluated in various fields like geology, coal, ash analysis, as well as biomass in general. The XRF instrument was calibrated using several wood chip samples. Measurements before and after calibration were compared with the reference ICP-OES method. Results show that XRF can be recommended for determining Al, Ca, Cr, Fe, K, Mg, Mn, P, Pb, Si and Zn whilst Cl, Ni, S and Ti can potentially be determined with element-specific calibrations. Three elements, As, Cd and Cu could not be measured satisfactorily as many measurements were below the limit of detection.
4.5. Polymers
Polymer analysis has been extensively documented during this review period. However, one topic stands out significantly, this being the analysis of micro and nano polymer particles (so-called micro- and nanoplastics). This aspect inevitably shifts the focus of this section towards the role that atomic spectrometry can play in addressing the challenges posed by the presence of these micro or nanoparticles in the environment and their potential toxicity to humans and wildlife – hence the increased attention devoted to this topic here. In addition, other key topics remain consistent with previous years, particularly the classification of polymers for waste management and the leaching of polymer materials from food packaging or healthcare products. These topics, considered relevant from the perspective of atomic spectrometry and polymer analysis, are discussed in detail below.Consistent with the key topic identified during this review period, two reviews – focusing on (1) the spectroscopic techniques available for the characterisation of micro-polymers and the assessment of metal adsorption, and (2) the analytical processes for nano-polymer analysis and remediation strategies – deserve attention in this ASU. Vasudeva et al. summarised recent advances in analytical methods using spectroscopic techniques for the characterisation of micro-polymers, along with the determination of adsorbed metals.108 The review included 265 references and focused on the types of micro-polymers, methods for extraction, identification (visual and microscopic examination), and characterisation techniques, which were classified into conventional chemical analysis methods, thermochemical, and spectroscopic. Among the spectroscopic techniques, the review provides information on the operating principles that enable the characterisation of micro-polymers using FTIR spectroscopy, Raman spectroscopy, fluorescence spectroscopy, and LIBS. A table summarising studies by several researchers using IR and Raman spectroscopy to detect micro-polymers from different sources, along with the techniques used and identified polymer classes, deserves special attention. The review also outlines the advantages and limitations of the different identification techniques utilised for micro-polymers characterisation and highlights the need for developing multi-modal analytical strategies to address the challenges posed by micro polymer particle analysis. A concluding section offers further insights into future prospects and pressing needs, in alignment with the Sustainable Development Goals set by the United Nations for 2030. Overall, the review concluded that there is a need to develop hyphenated techniques such as LIBS and Raman, which share similar optical configurations and instrumentation requirements. The other review of relevance to this section, this time by Moteallemi et al., focused on the challenges and perspectives of analytical methods and removal strategies for nano-polymers from aquatic environments.109 Therefore, this second review also addresses the issue of small polymer fragments – a topic of clear concern today – which explains its relevance in this ASU section. With the aid of 233 references, the review introduces the nano-polymer issue, from their origin to their impact on aquatic systems. A substantial section was dedicated to sample collection and isolation of nano-polymers from the sample matrix, including methods such as digestion, preconcentration, membrane filtration, ultrafiltration, cloud point extraction, and solvent evaporation. Separation techniques, including FFF, magnetic FFF, HPLC, SEC, hydrodynamic chromatography and electrophoresis are described in a subsequent section. In addition to microscopic and light-scattering techniques, the review provides information on spectroscopic techniques, namely FTIR, Raman microspectroscopy, surface-enhanced Raman spectroscopy (SERS), GC-MS, pyr-GC-MS, TD-GC-MS, and ICP-MS. Removal strategies for nano-polymers from water and wastewater are also discussed, with a relevant table highlighting the various approaches. In the conclusion section, the authors emphasise the need to combine different techniques – as already noted in the previously discussed review on micro-polymers. It is also clearly indicated that the field is still in its infancy, with a lack of standardisation, concerns about efficiency, and, therefore, implementation remains a work in progress.
A critical aspect of ICP-MS operated in single-particle (SP) mode for characterising micro-polymers is the ability of the sample introduction system to efficiently transport micrometer-sized particles into the ICP, as the transport of relatively large entities is limited when using traditional sample introduction systems. In this context, Sandro et al. carried out a fundamental study of four different sample introduction systems to evaluate the upper size limit of micro-polymer analysis via ICP-TOF-MS.110 Three different micro-polymer standards – polystyrene (PS), PMMA and PVC – within a size range of 3 to 20 μm were measured. Two different ICP-MS spray chamber and torch configurations were used: a downwards-pointing and a horizontal ICP-TOF-MS. The sample introduction systems consisted of: (1) a microdroplet generator followed by a falling-tube device with two He gas inlets and a gas exchange device using Ar as sweep gas – this setup was used with the downwards-pointing ICP-TOF-MS; (2) pneumatic nebulisation for particle introduction using online microdroplet calibration for quantification; (3) the same setup as (2), but with particles introduced through the microdroplet generator and the calibration solution via pneumatic nebulisation; and (4) an identical setup as (2) but with the cyclonic spray chamber replaced by a high-efficiency single cell introduction system. Setups 2, 3 and 4 were used with the horizontal ICP-TOF-MS configuration. After a systematic evaluation, the authors reported that pneumatic nebulisation with either a cyclonic or high-efficiency spray chamber in a conventional horizontal ICP configuration allowed for the successful introduction of polymer particle suspensions with sizes up to 8 μm, while larger particles were not, or were only poorly, transported (transport efficiencies ≤10%). The use of the downward-pointing ICP configuration substantially improved the transport efficiency for larger particles, with sizes up to 20 μm being successfully introduced. Similarly, using the falling tube setup with the horizontal ICP also allowed for the introduction of micro polymer particles up to 20 μm in size, which was attributed to improved desolvation of the generated droplets, enabling more efficient transport of particles by the carrier gas. However, the authors noted that determining transport efficiencies for the analysed standards remained challenging, as particle number concentrations in the suspensions appeared to be affected by sample preparation. It was also observed that the estimated transport efficiencies decreased with increasing particle sizes for the spray-chamber-based arrangements, a trend not seen with vertical particle introduction. Moreover, the vertical trajectories of droplets and micro-polymers enabled a greater number of particle events to be detected per measurement sequence for the largest particle size studied, which is considered advantageous for routine application. Continuing with the need to develop new strategies to facilitate the transport of micrometer-sized polymer particles to the ICP ionisation source, laser ablation has already demonstrated its potential in previous works, as described in this same section of the 2024 ASU review. This approach has continued to evolve, and in this case, Brunnbauer et al. focused on the development of novel calibration approaches for determining the size of micro-polymers via laser ablation SP-ICP-MS.111 In this work, an in-house created polystyrene (PS) thin film was used for size calibration of micro-polymers. Laser ablation was used as a method for sampling and introducing micro-polymers of different types and sizes into the ICP-MS. For calibration, defined amounts of carbon were introduced into the ICP-MS by quantitative ablation of a polymer thin film using different laser spot sizes, along with accurate measurements of the depths and areas of the resulting laser craters. To prepare the PS thin film used as a standard, PS flakes were dissolved in toluene to create a 5% (w/w) polymer solution. A volume of 70 μL of the PS solution was pipetted onto a silicon wafer and rotated for 90 seconds until the solvent had completely evaporated, resulting in a uniform PS film with a thickness of 150 nm, as determined via profilometry; the thickness variations were below 5 nm. The authors observed that using silicon wafers as a substrate proved more effective for preparing uniform films compared to glass microscope slides. A calibration curve was created by correlating the measured carbon signal intensities with the known introduced carbon content, yielding an LOD of 4.85 pg carbon, which is equivalent to a size of 2.12 μm for spheric PS particles. This calibration approach was successfully applied to determine the size of 2, 3 and 4.5 μm PS particles, with deviations of ≤6.3% from the certified diameters. When applying the PS calibration to determine the sizes of PVC and PMMA particles, good estimations were achieved despite differences in chemical composition. Therefore, the authors highlighted the universal applicability of the presented calibration strategy. Furthermore, the authors also investigated the transport efficiencies for differently sized particles and different polymer types, in line with the previously discussed study. In the case of laser ablation as a sampling method for micro-polymers, it was found that transport efficiency was significantly influenced by particle size, with larger particles showing lower efficiencies than smaller ones. It was also observed that gas flow and the properties of the transfer line significantly affected the transport efficiencies of micro-polymers, but that up to 95% of the sampled particles could be detected under optimum conditions.
While the two previously described studies focused on determining the size of micro-polymers by monitoring C as the main chemical component of polymers (a direct approach), other studies have addressed the detection of specific micro-polymers through one of their other main constituent elements, such as F in PTFE. It is worth noting that the degradation of fluoropolymers poses an additional concern compared to other polymer types, due to their potential to break down into PFAS under certain conditions, particularly during disposal or at extreme temperatures. For this reason, developing methodologies for the detection of fluoropolymers – with PTFE being the most common due to its unique properties is of utmost importance. Although, as indicated above, F is usually an inaccessible element for ICP-MS, Gonzalez de Vega et al. developed a method for detecting F-containing particles via SP-ICP-MS.112 This work built on previous studies that focused on F detection using tandem ICP-MS/MS with a Ba-based plasma modifier. Briefly, Ba2+ was added to induce in-plasma formation of the BaF+ molecular ion at a m/z of 157. MS/MS configuration was necessary to overcome the highly abundant Ba-based spectral interferences (e.g., 138Ba18O1H+). This study employed a design-of-experiments approach for the systematic optimisation of (1) plasma parameters, (2) ion optics, (3) mass filtering, and (4) collision/reaction cell conditions. Each group was optimised individually to achieve the highest signal-to-noise ratio and the torch z-position and nebuliser gas flow were found to be the most relevant plasma related parameters. For ion extraction and focusing, only hard extraction conditions enabled the detection of F-based particles. The collision/reaction cell gases (O2, H2, and He) and their settings were optimised. The role of the first quadrupole mass filter was also evaluated, with a bandpass of 4 amu providing the highest signal-to-noise ratio for BaF+. The optimised methodology could detect micron sized PTFE particles for determining both number concentrations and size distributions. Validation of the F-selective method was performed by comparing the results with those obtained via microscopy, Raman spectroscopy, and C-selective SP-ICP-MS. As a proof-of-concept, bulk PTFE material was stirred in simulated seawater under UV-light illumination for six days. After this incubation period, a micro-polymer number concentration of 2.35 × 105 F-based particles per gram of immersed bulk PTFE was determined. These PTFE particles had a mean mass and size of 28 pg and 2.7 μm, respectively. The authors highlighted that the method advances the analysis of fluoropolymer microparticles by targeting both C and F, allowing for the determination of number concentrations and size distributions even in complex matrices.
Although the ICP-MS technique has seen a remarkable increase in research activity for sizing micro-polymers by targeting constituent elements (e.g., C and F), other techniques also deserve attention in this regard. For example, LIBS and Raman spectroscopy remain important techniques to consider when discussing micro-polymer analysis. Furthermore, other studies have reported the use of alternative types of mass spectrometry techniques. Continuing with this latter example, Li et al., reported on the use of ambient microwave plasma torch desorption/ionisation mass spectrometry as a rapid and concise analytical approach for micro-polymer detection.113 Ambient mass spectrometry is rapidly evolving due to its ability to perform real-time, fast, and in situ analysis of diverse samples with minimal or no sample pretreatment. In this study, a system featuring a microwave plasma torch was coupled with an LTQ Orbitrap MS and was specifically constructed and refined for the rapid analysis of micro-polymers. Using the microwave excitation source, the micro-polymers did not undergo complete fragmentation, giving rise to characteristic peaks of monomer units without generating complex fingerprints or small-molecule products. The testing process for a single sample takes about 30 seconds, enabling faster analysis. The method was not limited by sample size, allowing for macroscale polymer analysis. Alumina ceramic was selected as the sample platform material, considering its high temperature stability, insulation properties, and low background interference. The distance between the ionisation zone and the MS inlet was also optimised. It was observed that at very short distances, the signal decreased and became unstable due to gas flow disturbances. However, at greater distances, few ions entered the MS, also reducing signal intensity. An optimal distance of 9 mm was identified. The power of the microwave plasma torch was also optimised, with 100 W determined to be sufficient to desorb and ionise micro-polymers without causing excessive fragmentation or signal loss. The Ar gas flow rate influenced the shape of the plasma torch, with an optimal flow of 1.4 L min−1 selected. Under optimised conditions, the method proved efficient for the desorption and ionisation of a wide range of micro-polymers (PA, PET, PMMA, PLA, PHB, PP, and PE). Linear relationships were established between sample mass and the intensity of characteristic ions, with R2 values exceeding 0.98. A simplified pretreatment process was also developed to rapidly extract micro-polymers from soil, enabling performance assessment in environmental samples. Detection was found to be nearly free from matrix interferences. The authors also demonstrated the compatibility of the developed method with SEM for multimodal characterisation, thereby complementing traditional MS analysis with additional insights into the size and morphology of micro-polymers.
As previously mentioned, LIBS and Raman spectroscopy are techniques that consistently appear in each edition of this annual review due to its growing applicability in polymer analysis, including the analysis of micro-polymers. In this context, Vasudeva et al., discussed the applicability of a custom-made hyphenated LIBS-Raman spectroscopic system for characterising micro-polymers.114 This system is a bimodal LIBS-Raman technique, combining elemental and molecular identification. The approach offers several advantages, as it is cost-effective, requires minimal sample preparation, and allows for rapid analysis. The study validated the system’s ability to identify both the polymer class and metal contaminants adsorbed onto micro-polymers in the size range of 1–5 mm. For this purpose, six micro-polymer samples collected from an estuary were characterised using the LIBS-Raman spectroscopy system. The analytical performance of this laser-based technique was compared with conventional methods such as FTIR spectroscopy, confocal Raman spectroscopy, and SEM-EDS. Raman analysis identified PE, PP and PET micro-polymers, with results confirmed by both confocal Raman and FTIR. However, the study did not focus on micro-polymer detection only, but as mentioned above, it also aimed to provide information on adsorbed metals. Analysis by LIBS enabled the detection of Al, Ni, Co, and Zn, along with trace elements such as Ca and Mg. Cross-examination with EDS validated the presence of these trace elements on the micro-polymer surfaces. Based on the observed results, the authors demonstrated the potential of the multimodal approach for identifying micro-polymers and surface-adsorbed heavy metals, thereby supporting the advancement of micro-polymer research through rapid and comprehensive characterisation.
The following two works focus on the on the potential toxicity of microplastics resulting from the presence and release of other elements, particularly under extreme degradation conditions, such as those caused by disinfection treatments with chlorine or by photo-aging. Ho et al. quantified the effects of chlorine disinfection on micro-polymers using SP-ICP-MS.115 A significant number of micro-polymers are released into aquatic environments and also appear in potable water following water treatment. However, the effects of disinfection processes on these microparticles remain largely unknown. In this study, changes in the Cl content of micro-polymers exposed to various water treatments were quantified via SP-ICP-MS. To overcome spectral overlap, H2 was introduced into the collision/reaction cell, enabling the monitoring of 35Cl1H2+ and 12C1H+ product ions for Cl and C quantification, respectively. It was found that PS microparticles exhibited increased reactivity to Cl disinfectant after pre-disinfection with UV light and in mildly acidic to neutral pH environments. The authors expressed concern that half of the PS particles exposed to 10 mg Cl2 per L, a typical Cl dose in water treatment, were chlorinated, with Cl content comparable to that of particles subjected to extreme conditions (50 and 100 mg Cl2 per L). Even more concerning, cell viability studies revealed that chlorinated micro-polymers induced significantly higher rates of cell death in both human A549 and Caco-2 cells, with effects found to be dependent on both Cl dose and polymer type. This study is the first to explore the potential of ICP-MS for quantifying the effects of Cl disinfection on small micro-polymers. The authors highlight the capability of the method to measure micro-polymer particles in suspension within a short acquisition time (300–900 particles per minute), in contrast to most surface-sensitive techniques that require analysis of individual particles. They also report limitations, such as the inability to distinguish between different chemical forms of Cl on the disinfected micro-polymers and the need to collect a sufficient number of signals for accurate data analysis. Based on the findings, it is likely that water treatment processes alter the surface properties of micro-polymers, thereby affecting their adsorption capacities for organic pollutants and heavy metals. In line with this work, Zjacić et al., evaluated the release of organic compounds and metals from micro-polymers resulting from the fragmentation of pristine and photo-aged PP films.116 In this study, the structural and morphological properties of pristine and photo-aged (14, 28, 42, and 56 days) PP films were characterised via FTIR, enabling the chemical transformation occurring after photodegradation to be evaluated. The structural changes in the polymer molecules were suggested to result in significant variations in polymer properties, such as the reduction in strength, which, in turn, allows easier and faster fragmentation. Analysis by SEM indicated the formation of surface cracks and scars on the photo-aged PP films. After photo-aging, the PP films were ground in a cryomill to assess their fragmentation into smaller particles. In the case of pristine PP, all particles were larger than 100 μm, while aged PP yielded a significant mass fraction of micro-polymers smaller than 100 μm. Leaching experiments were conducted in high purity water (HPW), adjusted with 0.1 mol per L H2SO4 to pH 6, with the polymer particles at a concentration of 100 g L−1 and incubation with shaking at 40 °C for 10 days. Prior to leaching experiments, the metal content present in the starting materials was quantified via ICP-MS, and the results were compared with those obtained from the leachates. The results for pristine samples showed low concentrations of monitored elements in the leachate, with up to 5% (Ni) of the total amount present in the solid sample. Several metals detected in the solid were not found in the leachate (Al, Ca, Cr, Fe, Mn, Pb, and Sr), indicating their stability in the polymer under slightly acidic conditions. However, in the leachate from the photo-aged PP irradiated for 56 days, all monitored elements were detected at significantly higher concentrations than in the pristine samples. In some cases, the concentrations in the leachate approached 100% of the initial content (Cu, Ni, Pb, and Zn), indicating that nearly all of the metals present in the PP material had leached out. These findings suggest that polymers remaining in the environment for extended periods undergo aging and may consequently release metals. However, the concentration of all detected metals in leachates were below the limits set by current legislation, except for Al. The leachates were also assessed for toxicity using representative aquatic species (Vibrio fischeri, Daphnia magna, and Pseudokirchneriella subcapitata). The most sensitive organism to the leached metals was the freshwater crustacean Daphnia magna. However, maximal inhibition did not exceed 90%. It was also found that the PP micro-polymer leachates strongly inhibited glucose biodegradability due to the presence of metals such as Al, Ba, Cu, Mn, Pb, Sr and Zn, and even Cr and Pb, although the latter were present in relatively low concentrations.
Determining the exact concentrations of regulated metals within polymeric materials is important for risk assessment. However, this knowledge has not yet been incorporated into standards concerning recycled materials. Kligenberg et al. quantified regulated metals in recycled post-consumer PP via ICP-MS, AAS and LIBS analyses.118 In this study, a strategic hierarchy for testing the heavy metal load in recycled plastics was established. A LIBS method, which enables rapid in-line analysis with minimal sample preparation, was used for preliminary screening of unusually high elemental loads and various metals including Cu, Fe, Ti, and Zn, could be qualitatively identified. Apart from Cu, none of the detected elements fall under current regulation. In addition, several alkali and alkaline earth metals, such as Ba, Ca, K, Mg, and Na, were also detected. Metal quantification was by ICP-MS and ETAAS. The contents of ten regulated elements (As, Cd, Co, Cr, Cu, Hg, Ni, Pb, Se, and Sn) were measured in different batches of post-consumer, post-industrial, and virgin PP. Concentrations in post-consumer PP were found to be 2–4 orders of magnitude below their thresholds for non-food applications. However, the possible accumulation of trace metals should be considered over multiple recycling cycles due to repeated cross-contamination of the material. The variability in metal concentrations within the same processing batch showed that contamination was relatively homogeneous, with a relative variability not exceeding 30%. Comparison of batches collected at different times confirmed batch-to-batch variations, especially for elements occurring at higher concentration levels, such as Cr, Cu, or Pb. In summary, the results demonstrated that the mechanically recycled polyolefins evaluated comply with regulatory standards and represent valuable raw materials suitable for a circular economy and the sustainable development of a wide range of polymer products. The final application of interest in this section focused on the characterisation of recyclable (bio)polymeric materials of various types and origins.119 The study aimed to develop and validate broadly applicable methods for the comprehensive assessment of As, Bi, Hg, and Sb by CV-HR-CS-AAS, and of total Cr, Cd, Co, Cu, Mn, Ni, Pb, Sr, and Zn using HR-CS-FAAS, in (bio)plastic recyclable materials. The methods were applied to a wide range of recyclable polymer materials based on PE, PP, PC, PS, PET, PVC, ABS, and ER, and were validated through the analysis of CRMs. For CV-HR-CS-AAS analyses, a CV generator was coupled to a heated quartz tube atomiser placed in an electric oven at a controlled temperature. The quartz tube atomiser was sealed at both ends with quartz windows. The spectrometer was equipped with a high-intensity Xe short-arc lamp (190–900 nm), a high-resolution double monochromator, and a CCD detector. For HR-CS-FAAS analyses, a Ti burner acetylene–air flame was mounted in place of the quartz tube. The (bio)polymeric recyclable material was mineralised via high-pressure MAE using a mixture of HNO3, H2SO4, and H2O2. The methods developed were more sensitive and had better LOD values than direct techniques commonly used to analyse polymeric materials, such as EDXRF, and LA-ICP-OES or LA-ICP-MS, FAAS and ICP-OES. The LOD values for Cu, Mn, Ni, and Zn were better in HR-CS FAAS than in ICP-MS, and poorer for Cd, Cr, and Pb than in GFAAS. Chemical vapor generation after digestion proved crucial for achieving the necessary analytical performance for the determination of As, Bi, Hg, and Sb in various polymeric materials at concentrations lower than the maximum regulated values and migration limits defined by European guidelines. According to the authors, attention must be paid to the selective collection and recycling of (bio)polymeric materials regardless of origin, due to the high variability in elemental content, even though the concentrations of most elements were low, often below the LOD. This survey is particularly relevant from the perspective of open science and citizen science, given the growing need to recycle increasing quantities of (bio)plastic materials.
5. Inorganic materials
5.1. Ceramics and refractories
As usual, this section will be relatively brief because of the paucity of relevant papers published. However, two papers were worthy of mention. An interesting paper presented by Emrick et al. was a good example of how portable instrumentation may be used to monitor the quality/contamination of a sample as it progresses through the manufacturing stages.123 In this paper, hand held versions of both XRF and of LIBS were employed to detect elemental impurities in additively manufactured ultra-high-temperature ceramics such as 3D-printed alumina. Analytical measurements were taken during several stages of the manufacturing process including high-temperature de-binding and sintering that were used to bake out organic impurities and improve grain cohesion of the ceramic. Impurities such as C, Ca, Fe, H, Na and Si were detected, but their concentrations decreased as the sample preparation process progressed.A multimodal Laser Opto-Ultrasonic Dual Detection system incorporating both LIBS and laser ultrasonics was described by Sattar et al. who used it to characterise BaTiO3 ceramics.124 The ceramics were prepared using a solid-state processed reaction method and were doped with Sn to concentrations between 0 and 20%. This preparation methodology was discussed in the paper, as was the methodology used for calculating Ba, Sn and Ti concentrations using calibration free LIBS. The laser ultrasonic component enabled the grain size to be determined when in acoustic attenuation mode and porosity when in velocity mode. This one instrument platform therefore enabled the compositional, structural and mechanical properties of the materials to be evaluated. Increasing Sn concentration led to a decrease in grain size from 14.39 μm in pure BaTiO3 to 4.08 μm when it was doped with 20% Sn. This also led to a corresponding increase in porosity from 9.25 to 13.7%. Also used in this study was XRD and XPS to investigate the electronic state of the analytes. It was concluded that one instrument being able to provide so much information was going to be of significant use throughout the ceramics industry.
5.2. Glass
This has been a surprisingly busy area of research during this review period. Significant effort is going into finding suitable materials to act as repositories for nuclear waste or to immobilise the waste itself and this is reflected in several publications. Included in this number is a paper presented by Tang et al. who reported the ICP-MS determination of Ru in borosilicate glasses used for nuclear waste immobilisation.125 Since ruthenium oxide is relatively volatile, the final amount present in the glass may not be that expected since it may volatilise away during the vitrification process. Treating the sample with sodium hydroxide at 350 °C for 8 h enabled a recovery of 95% to be obtained. An LOD of 0.4 mg kg−1 was not the lowest achievable but was adequate for the task required. Another application was described by Guo et al. who used spatially confined fibre-optic-LIBS to determine U in glass materials.126 Many methods are available to determine U, however, most cannot be employed in a radioactive environment. This is where the standoff ability of LIBS, especially when used in conjunction with fibre optics, comes into play. Unfortunately, there is an inherent decrease in sensitivity observed when using standoff LIBS because there is loss of light when being transmitted through the fibre optics and collection of light emitted tends to have a lower efficiency. Spatially confining the plasma goes a significant way in ameliorating these problems. The paper explained how this was achieved and stated that having plates that had a gap smaller than the plasma produced, led to an increase in U signal intensity by a factor of 3–4. The signal to noise ratio was enhanced by a factor of only 2–4. The enhancements were postulated to arise through the high density of the core plasma or a synergistic effect of plasma expansion and shockwave confinement. The ionic U lines at 409.01 nm and 367.01 nm as well as the atomic line at 358.48 nm were used for quantification. The lowest LOD achieved was 95 mg kg−1. In common with the previous paper discussed, this LOD is not very low, but is fit for purpose.One highly reactive fission product is Cs which is volatile and may be lost at high temperatures and methods for immobilising it are, therefore, an area of study. An example was presented by Shi et al. who prepared a series of lead, boron and zinc oxide (known as PBZ) glass mixtures as a sealant or immobilising agent for the Cs.127 The series contained five glasses with varying amounts of Pb, each prepared using a melt-quenching process. Once prepared, the materials were characterised using XRD, Raman and differential thermal analysis. As the Pb content increased, the melting temperature and viscosity decreases and the structural polymerisation increases. A glass of composition 35PbO–60B2O3–5ZnO that had been doped with Cs up to a concentration of 13.6% was then tested to evaluate its stability and ability to immobilise the Cs. This was assessed using static leaching tests in which the glass was crushed to give a particle size between 100 and 200 mesh, soaked in water and EtOH to clean it and then 3 g placed in a PTFE container with 30 mL of water and kept at 90 °C for 7 days. A longer-term experiment was also conducted in which a 10 × 10 × 10 cm cube was placed in a vessel and leached using water for 42 days. In all cases the leachates were analysed using ICP-MS which indicated that the Cs leaching rate was 1.6 × 10−1 g per m2 per day. This was considered to be sufficiently low for the material to be considered as having good stability. A paper presented by Kumar et al. assessed Cs volatility loss during glass melting and compared microwave and conventional heating.128 Glasses of different composition (borosilicate, borosilicate with high Cs and boron-free) were prepared using both techniques and the products were characterised using ICP-MS, XRD, SEM-EDX, Raman spectrometry, differential scanning calorimetry and Knudsen cell mass spectrometry. A chemical analysis in which the glass powder was mixed with K4BiI7, forming Cs–Bi–I, crystals, was used to estimate the Cs content. Several of these tests indicated that the microwave heating process led to significantly less volatilisation of the Cs. It was also noted that borosilicate glasses exhibited greater Cs evaporation than boron-free glasses, which was attributed to boron-assisted evaporation. The maximum amount of Cs that could be incorporated into the glasses was 35%, since a concentration of 40% inhibited the glass formation.
Microbially induced corrosion of glasses used for radioactive waste immobilisation is obviously a concern if radioactive materials are not to be released into the environment. This was the subject of a paper presented by Parker et al. who described a step-by-step protocol for preparing bacterial biofilms for TOF-SIMS analysis.129 Three biofilm preparation techniques: no desalination, centrifugal spinning and water submersion were compared, with the centrifugal spinning providing the best media removal efficiency (99% compared with 55% for the water submersion). The increased removal of the media had the added advantage of increasing the biological signals by over four times for fatty acids when compared with the water submersion method. The TOF-SIMS analysis of glass exposed to a Paenibacillus polymyxa SCE2 biofilm for a period of three months demonstrated significant compositional changes, with assorted phosphate-based molecules (especially iron phosphate) being detected that were not present in control samples. It was noted that, because of the complexity of the spectra produced, not all relevant signals could be identified. Despite this, the authors concluded that there was irrefutable evidence that glass corrosion was occurring.
A paper that investigated leaching rates from two different borosilicate glasses (Na-borosilicate and Zr–Na borosilicate) was presented by Sun et al.130 The effects of α particle and recoil nuclear irradiation on the leach rates were simulated by irradiating the materials using 2 MeV He ions or 5 MeV Xe ions. Following the irradiation, static leaching tests following the same standard protocol described previously (water at 90 °C for 42 days) were employed to assess what damage had been inflicted. The leached samples and leachate solutions were then analysed using a series of analytical techniques including Raman, FTIR, ICP-OES and TOF-SIMS. The He irradiation enhanced the interaction between water and borosilicate glass significantly, increasing the initial leaching rate by factors of between 1.1 and 1.7 compared with un-irradiated glass. The TOF-SIMS data confirmed that the alteration layer thickness formation was accelerated by a factor of 1.4 to 1.8 compared with the un-irradiated glass. The effects exerted by the Xe were even more pronounced than they were for the He irradiation. Overall, the radiation hitting the surface of the glasses was concluded to accelerate leaching from it and hence, when in the environment, would increase the risk of contamination of ground waters. However, the study helped enhance understanding of the leaching behaviour. A paper by Cöpoglu et al. also described the leaching properties of glasses, in this case sodium borosilicate ones used for the coating of food contact surfaces.131 The glasses were first sintered at 830 °C for 4.5 minutes to transform them into a crystalline phase. Once the glass constituents had been determined using XRF, the samples were immersed in 3% acetic acid for time periods between 15 and 960 minutes. The leachates were then analysed using ICP-MS and the surface of the glasses were examined using SEM and profilometry for any roughness. In all cases and for all immersion periods, the glasses remained compliant with ISO 4531:2022 food-contact standard. Surface roughness was observed to increase initially but then stabilised. Given the absence of constant leaching, the conclusion was that the materials are safe for the use as food packaging materials.
Window soda lime glass was analysed using LIBS employing a Transversely Excited Atmospheric pressure CO2 laser in a paper by Sajic et al.132 The glass components were spiked with the compounds Cr2O3, CoO, CdCO3 and CuO which colours the glass green, blue, yellow and red, respectively and then heated to 1100 °C for three hours. After gentle cooling, the metallic colourants were determined using a LIBS system made in house. The internal standards Si (for Cd and Cu) and Fe (for Co and Cr) were used to normalise the signals and to improve precision. Calibration was achieved with standards made in house and yielded R2 values ranging between 0.969 and 0.983. Limits of detection were 19, 15, 13 and 24 mg kg−1 for Cd, Co, Cr and Cu, respectively. Accuracy of the analysis was assessed by comparing the LIBS data with those obtained using XRF. In general, good agreement was obtained with R2 values of between 0.975 and 0.986 being obtained. The authors concluded that their LIBS system could be used to analyse glasses successfully, causing minimal damage to the sample in terms of the crater size and not causing the glass to crack or shatter.
A relatively straightforward application in which glass beads used for pavement marking were analysed for As content using HG-CS-AAS was reported by Wei et al.133 The digestion method, including the digestion mixture and the MAE conditions, was optimised using orthogonal design. The optimised digestion mixture was HNO3, HF and H2O2 in the ratio 5:6:2 and the temperature program was 10 minutes of heating time and then held at 210 °C for 40 min. The program finished by condensing the sample volume to 1 mL using a temperature of 160 °C. The hydride formation conditions were optimised using two factor ANOVA, with a concentration of NABH4 of 20 g L−1 and HCL concentration at 7% being optimal. The two factor ANOVA was also used to optimise gas transfer flow rate (6 L h−1) and atomisation temperature (950 °C). Under all of these optimised conditions, an excellent linear calibration (R2 = 0.9993) and low LOD (0.185 μg L−1) was achieved. Spiking experiments at the 6–10 μg L−1 level yielded recoveries of 97–103%. The conclusion was that the method developed was reliable and suitable for the required analysis.
The manufacture of fused silica optical surfaces has numerous steps including final polishing. The methods used differ between manufacturers and this can potentially lead to differences in the quality of the finish even for nominally the same products. Seeling et al. conducted a multi-technique (XPS, LIBS, IR, ellipsometry and atomic force microscopy) study of these materials obtained from different manufacturers.134 The study proved that the surfaces had different levels of contamination originating from the polishing materials, the water and techniques used, refractive index and surface roughness. A sample showed traces of Ce and Fe and another had some La. All three of these analytes were thought to originate from the polishing agents used. All four samples contained C, but at very different concentrations. It was concluded that workers should be wary of buying replacement parts from different manufacturers because these differences may result in a slightly different performance.
5.3. Electronic materials
This area is one of the most active and productive areas for atomic spectrometry applications, driven by the relentless demand for safer, longer-lasting batteries, higher-efficiency photovoltaic devices, more stable memory architectures, and environmentally sustainable manufacturing. The careful characterisation of these materials, especially via atomic spectrometry, remains essential for meeting performance targets and maintaining quality control in industrial settings. As in previous years, much of the work continues to rely on well-established methods, using atomic spectrometry in largely formulaic ways to verify elemental composition, map dopant levels, assess layer thicknesses, and determine uniformity. X-ray-based techniques, such as XPS, GIXRF, and GEXRF, as well as TOF-SIMS remain dominant for surface and depth profiling, thanks to their high sensitivity and spatial resolution. Other frequently used techniques include LIBS, ICP-MS, and XRF for bulk composition and, increasingly, for depth-resolved or interface-specific studies. However, despite the prevalence of these routine applications, the review period also featured a number of genuinely interesting analytical innovations. A prominent trend is the development of advanced sample preparation and microextraction approaches that expand the scope of laser-based spectroscopy. Collectively, these papers illustrate that although the majority of work applies atomic spectrometry in well-worn ways to characterise prepared samples, there remains room for meaningful innovation in sample preparation, microextraction, and speciation analysis. These advances not only improve analytical performance but also expand the range of sample types and matrices that can be addressed, whether in advanced semiconductor fabrication, environmental monitoring, or cross-disciplinary industrial settings.Li and Xia135 provided a timely review (21 references) of TOF-SIMS applications in Li-ion battery research, highlighting its roles in elucidating solid electrolyte interphase (SEI) composition, tracking electrolyte ageing, and mapping interfacial reactions. They emphasised depth-resolved and spectrally rich SIMS studies as essential for developing safer, longer-lasting batteries. Song et al.136 offered a tutorial review (164 references) on advanced detection techniques, including optical spectroscopy, TOF-SIMS, synchrotron X-ray methods, and neutron imaging, used to study electrocatalytic reactions in Li–S batteries. The review underscored the value of both ex and in situ analyses for clarifying Li2S nucleation and decomposition mechanisms and informing better cathode designs.
Finally in this opening section, Su et al.137 tackled the problem of trace metal speciation in strong acids and bases used in semiconductor manufacturing. By combining breakthrough curve theory with ICP-MS, they differentiated and quantified ng mL−1 level Cu, K, Mg and Na. Exploiting changes in breakthrough time with increasing acidity, they achieved precise speciation even at extremely low concentrations, a critical need for yield control in advanced semiconductor processes.
The improvement of depth resolution and controlling measurement artefacts remain central goals. Tröger et al.138 optimised TOF-SIMS sputter parameters to minimise measurement induced broadening when profiling silicon multilayers grown by molecular beam epitaxy with 2 nm bilayer periodicity. They demonstrated accurate atomic concentration-depth profiles in Si–SiGe heterostructures containing quantum wells, underscoring the method’s relevance for advanced device fabrication. Zhou et al.139 explored OrbiSIMS, a hybrid instrument combining Orbitrap and TOF mass analysers, for semiconductor depth profiling using Sb-implanted silicon. While offering superior mass resolution, OrbiSIMS showed slightly inferior depth resolution compared with magnetic sector SIMS, highlighting the trade-offs between mass accuracy and depth profiling fidelity. Zhang et al.140 applied LA-ICP-MS to depth profile multilayer thin films in PbS quantum dot photovoltaic devices. By optimising laser energy density and spot size, and employing a rapid washout ablation cell, they reduced interlayer mixing, achieving sub-6% deviation in thickness compared to SEM measurements, demonstrating practical relevance for device production quality control.
Several studies have advanced LIBS-based analysis for thin films and complex surfaces. Matsumoto et al.141 developed a surface-enhanced LIBS method for Nb detection in highly corrosive electropolishing solutions used in superconducting cavity production. By trapping μL scale samples on porous silicon supports, they achieved sensitive detection with a mean absolute error of 0.4 g L−1, offering an on-site monitoring strategy in challenging chemical matrices. Elsewhere, picosecond LIBS in a vacuum was applied to study B layers on W substrates,142 achieving high-resolution depth profiles that revealed subtle morphology changes tied to deposition conditions. Femtosecond LIBS was employed for real-time interface detection during laser processing of thin-film stacks,143 enabling on-the-fly identification of layer boundaries during ablation, offering a promising route to in situ process control. In situ LIBS was also used to monitor imidisation reactions of polyimide films during curing, tracking characteristic spectral lines in real-time to provide quality control during thin-film production.144 Kaplan et al.145 extended microextraction strategies to novel sampling surfaces. In this study, the analytical performance (LOD, accuracy, and precision) were determined and compared with unstructured surfaces. The dried droplet residue of aqueous heavy metal solutions of Pb, Cr, and Cu were analysed at their respective emission wavelengths under optimised experimental conditions. The results obtained from CYL-20 structured surfaces, in comparison to 300 nm thin-film coated surfaces, gave up to 17, 11, and 7-fold increases in LIBS signal strength for Cu, Cr, and Pb, respectively. The experiments were performed using single and multi-element standards, and water CRMs. Absolute detection limits, 0.8 pg Pb, 0.5 pg Cu, and 0.45 pg Cr, were obtained from the analysis of 500 nL standard solutions. Results with 70–75% accuracy and 95% precision were obtained from the repeated measurements. The authors state that although the results are promising, more extensive fabrication and application studies are required to find optimum microstructures of different sizes and shapes to improve the method figures of merit.
The tracing of enriched isotopes is valuable for studying surface reactions and diffusion. One notable application introduced labelled O (18O) and H (2H) into metal oxides and tracked their diffusion and exchange processes using SIMS.146 The paper considered two approaches: (i) ion beam analysis, which utilises specific nuclear reactions performed at tandem or van de Graaff accelerators, and (ii) SIMS. Despite the well-established use of isotopic tracing in silicon technology since the early 1970s, its application in applied electrochemistry or surface science to explore phenomena at the nanoscale remains relatively understated and unexplored by non-physicists. Examples of quantitative and qualitative analyses of an 18O tracer in thin Al2O3 and Ti films were presented through a comparison of nuclear reaction analysis, narrow resonance depth profiling, and TOF-SIMS and elastic recoil detection analysis applied to 2H tracing for optimising atomic layer deposition formation of titanium oxides. Such studies provide critical insights into oxidation resistance and catalytic properties relevant for electronic coatings and devices. Collectively, these studies highlight a balance between routine characterisation, confirming composition, thickness, and uniformity, and sustained innovation in depth profiling, calibration, in situ process monitoring, and sample preparation. Such developments are essential for supporting the increasingly complex requirements of next-generation semiconductor, display, photovoltaic, and advanced coating technologies.
Improving solid-state electrolytes and electrodes remains a central research challenge, given their promise for safer, high-energy-density batteries. Cressa et al.147 developed an operando neutron imaging approach, including a novel sample holder, to observe lithium dendrite formation in Li7La3Zr2O12 (LLZO) solid-state electrolyte half-cells during cycling. Post-mortem SEM and SIMS analyses revealed intergranular lithium accumulations and whisker-like dendrites in pores and cracks, offering a multi-scale picture of failure modes crucial for solid-state battery design. Cui et al.148 addressed air sensitivity in high-Ni layered oxide cathodes, quantifying residual alkaline compounds formed during storage. Using TOF-SIMS to map lithium species distribution, they demonstrated that a controlled recalcination process could restore electrochemical performance fully, providing an industry-relevant approach to stabilising these high-energy materials. Delfino et al.149 developed a correlative approach combining focused ion beam lift-out with SIMS to characterise Li-ion battery electrodes across their entire thickness down to sub-particle domains. By minimising air exposure, they preserved nanometric features of solid-electrolyte interphases and residual lithiation patterns.
The analysis of materials arising from electronic waste recycling remains a critical topic. D’Agostino et al.150 developed and validated SI-traceable ICP-MS and INAA methods for determining 20 technology-critical elements in printed circuit board materials from WEEE, producing well-characterised reference materials for industry. Lancaster et al.151 assessed XRF as a rapid screening tool for critical elements in printed circuit boards, LEDs, and Li-ion battery waste. Their inter-laboratory comparison highlighted challenges with sample preparation, spectral interference, and particle size effects, while confirming XRF’s potential for sorting and valuation in industrial recycling. Dommaschk et al.152 developed a direct solid sampling graphite furnace AAS method to analyse battery black-mass recyclates. Their method achieved better than 12% precision with results consistent with ICP-OES digestion, offering a rapid, robust alternative suited for industrial recycling workflows. Schnickmann et al.153 used LA-ICP-ToF-MS and EPMA to study lithium distribution in synthetic slags from pyrometallurgical battery recycling.21 They identified Li-bearing phases and fine structures that could retain lithium otherwise lost in slags, supporting improved resource recovery strategies.
The study of corrosion and electrochemical stability is also a recurring theme. One study combined hard XPS and ICP-MS to investigate Cr–Fe interfacial electrochemical processes,154 providing element-specific dissolution and redox data relevant to corrosion-resistant coatings. Reliability in memory devices also received attention. One investigation examined the role of hydrogen in resistive switching of ReRAM structures,155 correlating H incorporation with failure modes and underscoring the need for strict fabrication control. Ni–Cr alloy nanofilms, important for electronic contacts and heating elements, were studied using multi-dimensional LIBS156 to map spatial composition at high resolution, revealing subtle variations that could impact conductivity and reliability.
5.4. Nanomaterials
Atomic spectrometry, through techniques such as XRD, XPS, XRF, SP-ICP-MS and ICP-OES has a key role in the characterisation and detection of nanoparticles (NPs) with over 200 papers published in the period covered by this ASU. As this section focuses on the analysis of the NPs themselves papers covering NP detection in a wide variety of sample matrices are to be found in the other ASUs in this series as cited earlier in this review. In addition, many published articles only mention the technique(s) used without any further analytical detail and as such are not reported on here.Two papers report on the use of IDA for the determination of metrics associated with NPs. The first of these covered here quantified the nanoscale Si mass fraction in coffee creamer by AF4/ICP-MS using 29Si enriched Si NPs for the IDA.163 The Si content of the enriched Si NPs suspension was determined by reverse IDA with further characterisation by TEM, particle tracking analysis and AF4-multiangle light scattering. The coffee creamer/spike blend was prepared in HPW followed by vortex mixing and incubation for 2 hours at room temperature on a roller system to ensure spike/sample equilibration. The AF4 membrane was of regenerated cellulose and samples injection of 21.8 μL were made into the carrier flow of HPW. The outlet of the AF4 system was connected to an ICP-MS/MS instrument via a PEEK transfer line. The isotopes 28Si and 29Si were measured on mass in MS/MS mode with H2 used as the cell gas to minimise any polyatomic interferences. Mass bias correction was by the bracketing method, using IRMM-018a, which was introduced post-AF4 after each sample injection. The Si NP content of the coffee creamer was found to be 107 ± 8.6 mg kg−1. The main uncertainty contributions arose from the measurement of the isotope ratio in the sample blend (14.3%), blend-to-blend variability (70.4%), and the Si mass fraction in the spike (12.9%). Method performance was assessed by spiking food grade SiO2 into sample/spike blends giving a recovery of 95.6 ± 4.1%. The paper also covers the use of double particle specific IDA for the measurement of a separate, well characterised nanosilica sample. In this case, Si associated with the nanoscale fraction was 1094 ± 9.2 mg kg−1 which was in statistical agreement with the total Si mass fraction, 1108 ± 14.4 mg kg−1, found after the material was subjected to an MAE digestion procedure. The paper contains considerable detail on the work undertaken and is recommended reading by workers in this field. The second paper reported on here describes a method to quantify the particle size and number concentration of Pt NPs by suIDA-SP-ICP-MS.164 In this work an ionic 194Pt-enriched standard solution was used as a spike and the paper gives a good theoretical overview of this IDA approach, which can suffer from missing particle events due to the need to monitor two isotopes, and a good discussion on the optimisation of Pt spike concentration required for accurate results to be obtained. The particle transport efficiency to the plasma was measured using the DMF method. Values in good agreement with those previously found by AF4-ICP-MS were obtained by spiking the sample to 500 ng per L 194Pt for 30 nm Pt NPs and 1000 ng per L 194Pt for 50 nm giving recoveries of 100 ± 1% and 101 ± 2%. Using these spike concentrations the size determination was also found to be accurate, with mean values close to those of found by TEM and also previously reported. The developed method was then used to quantify the PNC and NPsize of both 30 and 50 nm spiked into a variety of complex matrices, urine, urine + TMAH, Dulbecco’s modified Eagle’s medium, synthetic seawater and seawater, and the results obtained compared with those obtained by conventional SP-ICP-MS. In all cases the PNC was underestimated by SP-ICP-MS, recoveries of approximately 40 to 80%, whilst the NPsize measurements obtained by both methods were in agreement. This reader could not find details of how the conventional SP-ICP-MS was undertaken, but if the calibrants were not matrix matched to the samples then this may be a reason the poor recoveries obtained for the PNC by this method.
Whilst SP-ICP-MS has now been in regular use for over 10 years there is still a need for the methodology to be improved and robustly validated. A particular area requiring that this is undertaken is the estimation of the transport efficiency (TE) of the sample solution to the plasma and two papers of note focus on this topic. Bolea and Laborda have published a paper which covers the transport efficiency (TE) methods used in SP-ICP-MS.165 The work gives a useful brief history of the topic and covers the theory of the various TE methods currently in use, particle size, particle frequency and dynamic mass flow, which are also tabulated along with potential sources of bias for ease of reference, along with the equations needed for the first two methods mentioned above. A separate section also discusses the limitation of each approach. The authors call for the terminology used for TE to be harmonised and note that there is a general consensus about the robustness of the particle size method, mainly because it is not affected by particle losses. There is a recommendation to use the particle size method when the measurement objective is the determination of the NPsize or the elemental mass per particle, whereas the particle frequency method should be used when the objective is for the determination of PNCs. With respect to measurement uncertainty, they also conclude that all three methods show similar values under optimal conditions, with the uncertainty associated to the use of a reference material the main contribution in both frequency and size methods. Bazo et al. have conducted a detailed study of intensity- and time-based approaches for sizing micro and nano particles (MNPs) using Au and SiO2 model particles and SP-ICP-MS.166 Two sizing methods using TE, particle size and particle frequency (both intensity-based) along with two TE independent methods, one based on direct external calibrations with MNPs and a relative approach which uses both MNP and ionic standards. All of these methods are described and discussed in considerable detail in the paper, along with the equations used and the uncertainty estimation approach. For the TE-dependent approaches, the particle frequency method was characterised by larger uncertainties than the particle size method. The results of the latter method were found to depend on the selection of appropriately sized reference MNPs, with the use of multiple sizes recommended, and gave the most accurate results. The TE-independent methods exhibited the lowest uncertainties of all the strategies evaluated. External calibrations benefited from simpler calculations, but their application could be hindered by a lack of reference MNPs within the desired size range or by the need for interpolations outside the calibration range. Finally, transit time signals are directly proportional to the MNP size rather than its mass. The time-based method demonstrated adequate performance for sizing AuNPs but failed when sizing the largest SiO2 MNPs (1000 nm). The conclusions and the results obtained in the work are usefully summarised in tabular form as well as the text. The authors also emphasise that the work focussed on the sizing of NPs only and that other considerations are needed for PNC measurements, for which a paper is cited. A second paper also compared the performance of different methods, particle size and gravimetric measurements, for determining the TE.167 The study used Au and Pt NPs (both 30 nm ∅) and no significant differences were found for each method. The work also investigated four different approaches for differentiating between the background signal and that arising from NPs with a background signal of up to 100 or 750 ng L−1 of added ionic Pt, 30 and 50 nm Pt NPs, respectively. The best results were obtained by applying a deconvolution to the data. The authors also call for the production of more reference materials, certified for size and PNC, to allow full method validation and state that further research and inter-laboratory comparisons would be helpful in the advancement of SP-ICP-MS. All three papers are recommended reading for workers in this field.
Most work using SP-ICP-MS use the standard instrumental set-up which means that the TE is usually between 2 and 5% which can be greatly increased by the use of microdroplet generators, total consumption nebuliser/spray chamber configurations and USNs. Dong et al. have published a paper which described the design and evaluation of a micro USN for use with NPs.168 A sheet of commercially available micro USN, which operates at 108 Hz, was bonded to a no waste spray chamber with samples delivered via a syringe pump. In theory the system should have given a TE of 100% but only 30% was obtained for Ag NPs of 40 nm ∅ whilst it was 60% for those of 60, 80 and 100 nm. The addition of 0.26 mmol per L EDTA to the samples improved these values to 80% with the increase being attributed to reduced adsorption of the NPs to the sample introduction system. Zhou et al. have taken a different approach to improving TE, reporting in a paper full of detail on the evaluation of eight modified cyclonic spray chambers, with a volume ranging from 25 to 125 mL.169 Each spray chamber was fitted with an IR emitter, inserted in a modified baffle, and the gap between the top of the baffle and the top of the spray chamber was varied. After optimisation of the sample flow rate and spray chamber temperature all of the modified spray chambers improved the TE and sensitivity, with a 50 mL volume, 2 mm baffle gap spray chamber IR-heated to 110 °C giving the best results (TE of 100%) and allowed the accurate characterisation of Au and Pt NPs without any measurement of the TE. Wang et al. have described the 3D printed total consumption spray chamber with the aim of improving TE with the device.170 The transport trajectories of individual particles passing through the spray chamber were simulated, using computational fluid dynamics, to providing theoretical guidance for the design and optimisation process. Statistical analysis of particle trajectories revealed that under the absorption boundary condition, particles between 20 and 100 nm achieved TEs exceeding 19%. A TE of 0% particles for NPS larger than 100 nm was thought to be due to increased deposition within the spray chamber. After optimisation of various operating parameters, including sample flow rate, carrier gas flow, rate, addition of a makeup gas and working temperature a TE of 61% was achieved. The set up was then used to determine the size of various sizes of Au NPs (15, 20, 30 and 40 nm ∅) and the results obtained were in good agreement with those from analysis by TEM.
Usually, ICP-MS or ICP-TOF-MS is used for NP characterisation, but Xing et al. have used MC-ICP-MS for this purpose.171 To achieve this the signal output from the instrument’s electron multipliers was directly routed to a high-speed oscilloscope, using 4 signal lines and recording at a rate of 0.8 ns per data point. The oscilloscope processed a single particle data burst according to two criteria: 3 ions arriving within a ∼75 ns interval, which indicated the beginning of a SP ion cloud, plus a count rate of more than 50 recorded within the span of the time window. This was termed as event-triggered signal capture and delivered the complete ion burst structure of a single particle event. This set-up was developed to allow high precision isotope ratios to be calculated for a single particle, which would not be possible using the electron multipliers alone due to the minimum dwell time available of 100 ms. The minimum detectable particle size was calculated to be 8 nm for AgNPs. The measured 109Ag:107Ag isotope ratios, which is 0.931 and was obtained from the data for 2000 particles, for NPs of 20, 40, 60, 80 and 100 nm ∅ were 0.921 ± 0.086, 0.937 ± 0.063, 0.926 ± 0.051, 0.936 ± 0.040, and 0.944 ± 0.029. Precision was limited by counting statistics of the isotopic signals. The precision of δ109Ag could be as low as 0.7‰, even for the measurement of 20 nm AgNPs (N = 17000). No mention of mass bias correction could be found in the paper so the lack of this correction could be the cause of the differences observed between the measured a true ratio of 109Ag:107Ag.
The SP-ICP-MS technique is usually used for suspensions of spherical NPs, however two papers have deviated from this year. Morales et al.47 used the technique, along with HR-TEM and SEM, to look at anisotropic structures, including solid Pt-nanorods and hollowed Fe2O3-nanotubes. Solid Pt-nanorods (191 ± 18 nm in diameter) showed important heterogeneity in their length, ranging from 42 to 72 nm, due to sample preparation difficulties. The analysis by SP-ICP-MS confirmed the presence of two different populations of Pt nanorods at 19 ± 4 fg and 41 ± 5 fg, respectively, yielding a mean value of 23 ± 12 fg Pt per rod and a length range from 38 to 67 nm, in agreement with TEM measurements. In the case of the two different sized double-walled Fe2O3-nanotubes of 900 nm and 1800 nm in length, the SP-ICP-MS measurements provided results of 16 ± 10 and 25 ± 4 fg Fe per nanotube, respectively. From this data, the layer thickness of the Fe2O3 nanotube wall was calculated and gave values ranging between 20 ± 6 and 17 ± 4 nm, respectively. This was in good agreement with the TEM estimations (18 ± 4 nm). The second paper used a commercially available microextraction (ME) probe in conjunction with SP-ICP-MS for the direct analysis of NPs on surfaces.172 The extractions were performed by depositing a 200 nL aliquot of an NP suspension (typically containing ∼500 NPs) was deposited on a Teflon surface. The probe head was lowered onto the sample surface, forming a seal in a 2 × 4 mm area and then a carrier solution was pumped onto the sealed area to mobilise the NPs and transport them to an ICP-MS instrument. Two probe heads were evaluated in this work, the standard fitment and a custom designed one machined from PEEK. After optimisation of various parameters, e.g. flow rate, carrier solution composition the system was used to compare the extraction efficiency (EE, defined as the ratio of particles measured to particles deposited on the surface) of the commercial and custom probe heads. The PEEK probe head gave an increased EE compared to the commercial probe head, 8.5 ± 3% and 3.9 ± 3%, respectively. The EEs for ME-SP-ICP-MS were typically 4–10%, which is similar to transport efficiencies (1–10%) for conventional SP-ICP-MS. The system was then employed for the analysis of nano- and micro-particles with the sizes of Au NPs, 30 ± 3 and 51 ± 1.9 nm (certified sizes), and Fe-based microparticles, 1000 ± 50 nm (certified size), were accurately determined to be 32.2 ± 2.5, 50.8 ± 3.4, and 1030 ± 57 nm, respectively. The authors see the device having applications with surfaces such as silicon wafers, vitreous carbon planchettes, and GSR tabs which are commonly used to collect NPs.
Two ubiquitous laboratory contaminants are Fe and Zn which presents a serious challenge when low LOD values are required. Two papers have described the development of methods to determine NPs of these elements. Boutry et al. focussed on Fe2O3 NPs, using ICP-MS/MS as the detector, optimising various parameters, collision/reaction gases, forward power, torch injector diameter and data treatment to determine the threshold between background and NP signals.173 After optimisation a LODsize of 24.1 ± 5.0 nm and 11.5 ± 0.4 nm with a dwell times of 3 and 0.1 ms, respectively and 30 nm NPs. The methodology was also applied to other Fe based NPs suspensions, commercial polydisperse (50 to 100 nm) and monodisperse (20 nm) and in-house synthesised (10 nm). In all cases the found sizes by SP-ICP-MS agreed with those determined by TEM. Although the dissolved Fe signal may arise from dissolved Fe in the reagents no mention is made of any cleaning protocols used for the labware involved, which is well known as a source of Fe, or any clean handling techniques. Applying both of these may have reduced the dissolved Fe signal and should be routinely applied when attempting to achieve low Fe LOD values. The final paper covered here concerns the measurement of ZnO NPs and dissolved Zn by SP-ICP-MS.174 Unusually, the waste collection method was used to determine sample flow rate, giving an RSD of 5%, when the gravimetric method by difference of the amount aspirated from a sample tube gives an RSD in the region of 1%. After method optimisation the LOD values for NPsize, dissolved Zn concentration, and particle concentration were 67 nm, 0.4 μg L−1, and 1.08 × 105 particles per mL, respectively. The method uncertainty was estimated by the bottom-up approach and for the PNC measurements the major contributions arose from sample preparation (58%) and the TE (25%).
:147Sm being 0.912, which is within 1% of the expected ratio. In the analysis of lead isotopes from galena particles the RSD for 208Pb:206Pb ratio ranged from 32 to 2% for particles with 1.4 to 80 fg of Pb. The developed method was then used for radiometric dating of monazite particles by measuring the 208Pb:232Th and 206Pb:238U ratios. These analyses, of individual particles that contained only 0.02–80 fg of Th and 0.03–30 fg of U, showed radiogenic Pb-isotope signatures and a median age of 550 Ma. Two approaches, SP-MC-ICP-MS and SP-ICP-TOF-MS have been compared for the measurement of Nd isotope ratios in NdVO4 particles of approximately 120 nm ∅.176 The particles were first characterised in terms of mass and PNC by SP-ICP-TOF-MS and then by SP-MC-ICP-MS, using a 50 ms dwell time, for isotopic precision. For the isotopic ratio measurements, the MC-ICP-MS performance was compared to the ICP-TOF-MS, and isotope ratio precision was found to be poorer (R−2 between 0.98 and 0.99) compared to ICP-TOF-MS (R−2 between 0.88 and 0.97). The accuracy attained on a single particle level, was compared to bulk digestion followed by MC-ICP-MS analysis, and the SP-MC-ICP-MS technique was able to determine the particle population average to a <4% relative differences for all the Nd ratios measured. The LOD for the SP-MC-ICP-MS approach was also assessed. With an all Faraday-cup based detection scheme the determined LOD for the measurements was 0.2 fg for Nd, per particle.In spICP-TOF-MS, element signals can only be recorded as “particles” if they are above the critical value, which is the threshold used to distinguish between particle-derived and background signals. To gain a better understanding of how biases and noise can alter the interpretation of data, Monte Carlo simulations have been used to model SP-ICP-TOF-MS signals as a function of measurement parameters, such as particle size distribution (PSD), multi-element composition, absolute sensitivities and measurement noise from ion-counting (Poisson) statistics.177 The simulations allowed for the systematic comparison of known (simulated) element mass distributions to experimental (measured) data. To demonstrate the accuracy of the model in predicting SP-ICP-TOF-MS signal structure, data from in-lab measurements and simulations for the detection of CeO2, ferrocerium mischmetal, and bastnaesite particles were compared. In all cases agreement between the measured and simulated data was obtained except when particle sizes were close to or below the NPsize LOD. The paper gives a detailed and readable account of the theory and discussion of the results and is well worth a read.
Solution based methods are more commonly used than LA in SP analyses, possibly due to difficulties in gaining accurate and reliable calibration. Kronlachner et al. have addressed this by embedding NPs in polymer thin films.178 For creating a calibration for mass and size investigations, defined amounts of the element of interest were introduced into the ICP-MS by quantitatively ablating polymer thin film spiked with a defined amount of liquid element standard with different laser spot sizes. The methodology was developed and optimised using Au NPs and SP-ICP-Q-MS. Using the proposed calibration approach deviations of ≤ 2.5% from the certified NPsize were achieved with an LOD for Au of 3 × 10−7 ng, which translates to a particle size of approximately 15 nm, which the authors state is comparable to values in the literature for suspension-based approaches. Multi-element NPs, Gd doped Ce2O3, were analysed using SP-ICP-TOF-MS and the thin-film-based calibration approach. Comparative measurements of the material confirmed the investigated sizes and composition of the particles. This developed alternative approach circumvents the need for certified particulate standard materials by using in-house-produced spiked polymer thin films as storage-stable calibration standards. Full details of the development and optimisation of the method are given in a comprehensive and readable paper.
The SPCal software, which is an open-source SP data processing platform, was developed for use with SP-ICP-MS data. This software has now been updated so that it can also be used with SP-ICP-TOF-MS data sets.179 Various tools have been incorporated to facilitate the handling, manipulation and calibration of large data sets and provide the required statistical fundament and models to promote accurate thresholding. Non-target screening tools are also included to pinpoint particulate elements in unknown samples without the requirement for a priori investigations or modelling. Along with functions for calculating NPsize, PMC and PNC, methods to carry out cluster analysis (PCA, HAC) are also now included to allow the interrogation of particle populations by selecting specific features.
6. Forensics
A forensic analysis of seized methamphetamine samples from Ankara, Türkiye using both inorganic (ICP-MS) and organic (GC-MS) methods of analysis was presented by Aksoy et al.180 The determination of Al, As, Au, Ba, Cr, Cu, Fe, Mn, Mo, Pb, Sb, Sn, V and Zn can potentially provide information on the manufacturing processes and precursor materials used. A MAE brought the analytes into solution for the ICP-MS analysis. Once the analytical data had been obtained, they were input to several chemometric tools including PCA and hierarchical cluster analysis (HCA). Five distinct sources of the drugs were identified. These were: Iran and Afghanistan (with ephedrine-based synthesis methods) and non-ephedrine-based synthesis methods based in Southeast Asia and Europe. The combination of both organic and inorganic analysis therefore provided the optimal methodology for obtaining maximum information about synthesis sites and methods.The analysis of counterfeit bank notes has been the subject of a paper by Bejjani et al. who used TOF-SIMS to obtain surface chemical imaging and depth profiling.181 Although steps have been made in many countries to prevent counterfeiting (watermarks, holograms, plasticised notes, metal strips, etc.), the counterfeiters have continued to adapt and hence there is a necessity to provide ever increasingly sophisticated methods of analysis. Five different counterfeit notes were examined, three 100000, one 10000 and one 5000 Lebanese Lira. The TOF-SIMS was used to examine different security features in each with the hologram being investigated in the 100000 Lebanese Lira notes, and the green and red stripes in the 10000 and 5000 Lira ones, respectively. Analysis of the coloured stripes would indicate if the correct pigment had been used whereas the holograms required depth profiling analysis. Using the dynamic mode of TOF-SIMS and with an Ar primary beam, it was demonstrated that the fake security threads consisted of an ink layer deposited onto an aluminium base, which was then cut into rectangular shapes and fixed onto the paper. This created the illusion of a colour changing woven thread. By determining the Na signal from different layers of the notes, (as well as two other methods) it was possible to determine that the hologram in the 100000 Lira notes was placed on the paper before the printed layer in the counterfeit notes. The presence of Na2SO4 in the pigment of the counterfeit money was also noted. One of the other main advantages of the technique is that TOF-SIMS is effectively non-sample destructive, with an area as small as 500 × 500 μm being studied.
Attacks on bank automated teller machines (ATMs) are relatively frequent and Zoon and Janssen have developed a database of glass from ATMs that can be used as a reference to compare against fragments found on suspects.182 Glass was collected from the scene of an attack and transported to the laboratory where the thickness and colour of each sample was noted. Chemical analysis using LA-ICPMS was also undertaken where a total of 18 analytes were determined with 29Si+ being used as an internal standard. Each sample was analysed six times, to ensure reliability of the data. The database now contains data from over 650 attacks in the Netherlands and over 3000 from Germany. Any samples suspected of being in contact with one of these glasses may then be analysed. First, the sample is beaten with a plastic batten to ensure the release of particulates. Those thought to be glass particulates were then analysed a single time. Their elemental compositions were then compared with the reference glass database using an overlap criterion and a calculated likelihood ratio. The methodology has been very successful in that over 50% of cases investigated since 2019 have found a match.
An interesting application was reported by Senra et al. who described the portable XRF analysis of cigarette ash followed by a chemometric analysis of the analytical data.183 A pack of cigarettes was bought from each of 10 different brands with one brand having a second pack bought so that an estimate of within-brand precision could be made. Each cigarette was then “smoked” under the identical ISO conditions and the ash from each collected with 0.4000–0.8000 g placed in sample holders. Data from 14 analytes were then analysed using one-way ANOVA to see if there were any significant differences between any of the brands. Once it had been established that there were significant differences, the data were input to HCA. This chemometric analysis was at least partially successful, with two main groups being identified; one containing four brands and the other six. However, these groupings could be separated further with most brands being clearly distinguishable from the others. Only in a couple of cases was potential confusion observed. It was concluded that the methodology had the advantage of being able to be performed at the crime scene, minimising the chances of sample loss, contamination, mis-labelling, etc. It also had the advantage of being non-destructive, enabling further analysis if required.
Kotrly et al. have presented a paper entitled “New robotic tools for multimodal non-destructive analysis and characterisation of 2D and 3D objects”.184 The paper reported the development of prototype devices that are capable of imaging and mapping both 2D and 3D artworks, including those that have complex curves, in a non-destructive manner. The devices were described within the paper and were stated to provide high quality pictures with a spatial resolution at the μm level. The basic version of the device enabled transmission X-ray imaging and mapping of the individual photons with high-sensitivity and high-resolution detectors. In addition to XRD, a new device also has XRF ability capable of point analysis and mapping.
The analysis of Pb-free solder alloys has been reported. A study by Moghadam et al.185 has validated the use of LIBS/LA-ICP-MS for the forensic evaluation of Pb-free solder alloys, which can form valuable evidence from post-blast crime scenes involving homemade and improvised explosive devices. The use of LIBS/LA-ICP-MS was seen as competitive with other spectroscopic-based forensic techniques as it analyses samples directly and requires minimal destruction of the exhibit. Following a one-standard calibration technique, nine major (alloying metals) and trace elements (impurities or additives) were quantified in lead-free solders. Optimising laser parameters and using Pb as a naturally occurring internal standard were shown to compensate for mass-dependent drift and matrix effects. The quantitative results from Pb-free CRMs aligned with the certified values and with results from two techniques in a cross-validation comparison, including ETV-ICP-OES and NAA. The utilisation of peak ratios in a model of PCA was presented to identify key compositional differences among solders and provide a visual model for solder discrimination. Outcomes of this approach demonstrated the potential for associating or discriminating lead-free solders, including different solders from the same manufacturer. Together, this technique can establish chemical concordance among known and questioned materials and offers a utilitarian approach for the forensic assessment of trace evidence.
7. Cultural heritage
This section produced a number of interesting papers this year, however the predominant subject appears to be the composition and provenance analysis of Chinese porcelain, glazes and pigments. There is also much work using portable XRF analysis instrumentation to understand the chemical characterisation of artifacts.7.1. Metallic cultural heritage samples
A paper de Palaminy et al.186 investigated the potential of femtosecond laser ablation coupled with multicollector ICP-MS for Cu isotopic analysis in Au matrices. Elemental analysis of Au artifacts, a commonly used method to pinpoint the Au source has failed in some cases. Isotopic analysis can offer a more accurate means of identifying the source of the metal, however, the application of this analysis to Au matrices remains a challenge. In this paper the authors successfully determined Cu isotope ratios in Au matrices and achieved a repeatability of 0.12 to 0.26‰ δ65Cu analyses carried out over 8 days. This work was conducted using an isotopically characterised in-house matrix-matched Au standard with copper concentrations varying from 4.5 to 9.6 wt%. This work opens new avenues of research for provenance studies of precious museum artefacts and archaeological finds with potential applications in authentication analyses on similar gold materials.The second paper in this section, by Duan et al.,187 measured the Pb isotopic composition of objects from the Western Han Dynasty These constitute a significant category of metal funeral artifacts, however, there has been an absence of scientific analyses on this subject. This paper describes in situ Pb isotope analysis of ten Pb objects sourced from the Dabaiyang sites in Xi’an using SEM-EDS, and LA-MC-ICP-MS methods. Those objects composed entirely of Pb, were found to be cast without annealing or forging, exhibited minor soil impurities and sporadic metallic elements introduced during incomplete separation from accompanying copper. The Pb isotopic characteristics of artifacts from the same tomb were very similar, indicating they may have been batch-produced. The Pb isotope ratios were divided into three groups, two likely from the Xiaoqinling area near Xi’an, and the other from Hunan matching the pattern between the Pb isotopes and different ages previously discovered. Based on archaeological findings and Pb isotope data, it is suggested that Pb minerals from Hunan were locally collected and smelted during the early Western Han Dynasty and then transported to Xi’an through Nanyang.
7.2. Organic cultural heritage samples
Kuehn et al.188 investigated the use of LIBS for the reassociation of co-mingled skeletal remains. When multiple sets of skeletonised human remains exist in the same context, they can become commingled due to multifactorial circumstances that affect the postmortem environment. This paper evaluates the potential of portable LIBS as a useful tool for reassociating commingled human skeletal remains in forensic contexts from their elemental signatures. In this study six skeletons drawn from the donated skeletal collection at the Florida Institute for Forensic Anthropology and Applied Science at the University of South Florida, were used to assess whether LIBS data could be used to reassociate multiple skeletal elements from the same individual. The LIBS data were collected at 206 anatomical locations from 28 individual bones across each skeleton in the sample. Data were reassigned to the individual with an accuracy of 91% using quadratic discriminant analysis of dimensionally reduced data (via PCA). The study demonstrated that portable LIBS has potential for reassociating co-mingled human skeletal remains from forensic contexts.The second paper by Anduze et al.189 used PIXE and EPR for the identification, quantification and sourcing of fossil hydrocarbons in Egyptian mummies. Egyptian mummies are often covered with black embalming matter which is made of complex mixtures of natural organic substances such as vegetable resins, beeswax, animal fats, gums and vegetable oils, as well as bitumen. The paper investigated the potential of certain transition metals, in particular V and Ni, for detecting the presence of bitumen and tracing its origin. The PIXE analysis showed that all the mummies studied in this work (bird, ram, crocodile, human) which span a period of about 1000 years and come from different sites in Egypt have a nearly constant Ni:V ratio close to that of bitumen from the Dead Sea suggesting a well-defined source of bitumen supply.
7.3. Ceramic cultural heritage samples
A review paper (82 references) by Lozada-Mendieta et al.190 reviewed ceramic archaeometric studies in the Amazon and Caribbean regions. This paper gives a brief historical overview of the topic, covers the technologies used e.g. thermoluminesence, NAA and XRF, article provenance, and suggests future developments for archaeometry in Latin America. The authors conclude that there is a diversity of methods that have been applied and continue to expand, used in complementary ways to confirm or contest more complex questions on the production and circulation of ancient ceramic materials and the social groups behind them. That the current existence of established local laboratories and an increasing number of trained archaeologists in archaeometric techniques is the product of a slow but steady rise of the discipline in the region and its recent consolidation. However, there is still a dependency that some countries have on foreign collaborations, due to the costs and limited access of certain type of instrumental analysis and that the research questions have been rather limited and there is a need to go beyond provenance studies and composition. A paper by Charlton et al.191 investigated the application of LIBS and LA-ICP-MS analysis to Hellenistic tableware. Tableware usually consists of the bulk material which is covered by a thin film, or slip. The slips are commonly analysed by X-ray microanalysis, offering point-by-point analysis of mostly major elements while different tools such as XRF, ICP-MS, and INAA, have been used for bulk body analysis. This can cause difficulties in comparing the bulk composition and thus in determining the similarities and differences in the preparation process of the clay paste for the slips and bodies of an object. Given the artistic value of these objects, museum curators tend to be reluctant to provide samples for invasive characterisation. Micro-destructive laser ablation methods offer a robust solution to addressing both the relationships between ceramic body and surface treatment chemistry. This paper provides a proof-of concept for LIBS and LA-ICP-MS analysis of the slips and bodies of Hellenistic fine wares from the Greek colony of Issa in modern Croatia. The results show remarkable diversity in the use of clay types and processing techniques.7.4. Glass cultural heritage samples
A paper by Delbey et al.192 describes a self-seal LA open cell for trace element analysis of full-size archaeological artefacts. LA-ICP-MS is increasingly becoming the standard analytical approach in archaeological science, especially for the analysis of glass, glaze, and metal. The low detection limits, high precision and wide element range twinned with a very small sample size makes it ideal for culturally valuable objects. However, in the past, the small sample chamber size in commercial laser systems has meant that the size of object that could be analysed is very limited. Ongoing work by various groups has demonstrated the potential of open architecture and portable systems to overcome this issue. This paper reports an example of this and the results of a validation on a self-seal open ablation cell coupled with a very large sample chamber developed by Cranfield University and Elemental Scientific Lasers. Comparison of the analysis of standard reference materials between the self-seal open cell and a conventional closed cell in a two-volume chamber show that the count rates for most elements drop by 40–70% in the open cell, however precision, accuracy, fractionation, and LOD values are barely affected. This means that the resulting outputs of both chambers are very similar and shows that the open cell is a very real solution to the problem of the small sample chambers in conventional LA-ICP-MS instruments. The functionality of the open cell is demonstrated using a case study of two Chinese polychrome enamelled copperwares dating from the late 18th or early 19th century Qianlong or Jiaqing periods. Due to conservation work on the pieces, five small enamel samples were available which meant that the results derived from these samples in the conventional laser chamber could be compared to results from the objects themselves in the large sample chamber. Time resolved analysis was carried out giving more information on the thickness of enamel layers and variability between colours and across the object demonstrating the usefulness of having large complete objects in the chamber.Work reported by Giachet et al.193 described the systematic chemical analysis of glass beads from post 15th century West African sites. The trading of glass beads towards and within sub-Saharan West Africa grew exponentially over time to culminate with the establishment of the Atlantic Trade. Although these artefacts are very commonly found in archaeological contexts dating after the 15th century CE, the assemblages are generally poorly studied from a chemical point of view. This study consisted of 916 glass beads found in five archaeological sites in Ghana, Mali, and Senegal dated to between the 15th and the mid-20th century CE. Besides the techno-stylistic classification of the whole assemblage, the compositional study of a sub-group of 578 monochrome and polychrome glass beads was performed. The use of LA-ICP-MS and the statistical analysis of the results by PCA led to the identification of the probable origin of the glass. Different suppliers were distinguished for the Ghanaian earlier beads and the Senegalese and Malian later ones in relation to the different European trade partners at different times.
Finally, a paper by Rasmussen et al.194 while not ground-breaking in technical innovation is however, interesting. It describes the chemical analysis of fragments of glass and ceramic ware from Tycho Brahe’s laboratory at Uraniborg. In addition to his astronomical observations the Renaissance astronomer was known also for his interest in alchemy, equipping castle Uraniborg with a state-of-the-art alchemical laboratory when it was erected around 1580. After Brahe’s death Uraniborg was demolished. In this study four glass shards and one ceramic shard most likely from the alchemical laboratory, and retrieved during an archaeological excavation, were analysed. Cross sections of the shards were analysed for 31 trace elements by LA-ICP-MS with the aim of detecting any traces of the chemical substances on the inside or outside of the shards. Four of the elements were found in excess on the exterior surfaces of the shards, Au, Cu, Hg and Sb and are in accordance with the reconstructed recipes of the three Paracelsian medicines for which Brahe was famous. This is the first experimental data casting light on the alchemical experiments that took place at Uraniborg in the 16th century.
8. Abbreviations
| 2D | two dimensional |
| 3D | three dimensional |
| AAS | atomic absorption spectrometry |
| ABS | acrylonitrile butadiene styrene |
| AF4 | asymmetric flow-field flow fractionation |
| AFM | atomic force microscopy |
| AI | artificial intelligence |
| ANN | artificial neural networks |
| ANOVA | analysis of variance |
| API | active pharmaceutical ingredients |
| AS | absorption spectrometry |
| ASU | Atomic spectrometry update |
| ATM | automated teller machine |
| ATR | attenuated total reflectance |
| BAM | Bundesamt für Materialforscung und Prüfung (Germany) |
| BDSAC | benzyldimethylstearylammonium chloride |
| BPNN | back-propagation neural network |
| BSA | bovine serum albumin |
| CCD | charge coupled device |
| CE | common era |
| CNN | convolutional neural network |
| CRM | certified reference material |
| CS | continuum source |
| CU-BTC | copper(II) benzene-1,3,5-tricarboxylate |
| CV | coefficient of variation |
| DFT | density functional theory |
| DLLME | dispersive liquid–liquid microextraction |
| DMF | dynamic mass flow |
| DRBM | discriminative restricted Boltzmann machine |
| EBS | elastic backscattering spectrometry |
| ED | energy dispersive |
| EDS | energy dispersive spectrometry |
| EDTA | ethylenediaminetetraacetic acid |
| EDX | energy-dispersive X-ray |
| EE | extraction efficiency |
| EE | epoxy resin |
| ENM | engineered nanomaterial |
| EPMA | electron probe microanalysis |
| ERM | European reference material |
| ETAAS | electrothermal atomic absorption spectrometry |
| EtOH | ethanol |
| ETV | electrothermal vaporisation |
| FAAS | flame atomic absorption spectrometry |
| FFF | field flow fractionation |
| FTIR | Fourier transform infrared |
| GA | genetic algorithm |
| GC | gas chromatography |
| GCIB | gas cluster ion beam |
| GE-XRF | grazing exit X-ray fluorescence |
| GF | graphite furnace |
| GI-XRF | grazing incidence X-ray fluorescence |
| GSR | gunshot residue |
| HCA | hierarchical cluster analysis |
| HDPE | high density polyethylene |
| HEA | high-entropy alloy |
| HPLC | high performance liquid chromatography |
| HPW | high purity water |
| HR | high resolution |
| HR-CS-AAS | high resolution continuum source atomic absorption spectrometry |
| ICP | inductively coupled plasma |
| IDA | isotope dilution analysis |
| INAA | instrumental neutron activation analysis |
| IP | ion pair |
| IRMM | Institute for Reference Materials and Measurements |
| ISO | International Organization for Standardisation |
| KELM | kernel extreme learning machine |
| LA | laser ablation |
| LASSO | least absolute shrinkage and selection operator |
| LDA | linear discriminant analysis |
| LDPE | low density polyethylene |
| LIBS | laser-induced breakdown spectroscopy |
| LIF | laser-induced fluorescence |
| LOD | limit of detection |
| LOQ | limit of quantification |
| MAE | microwave-assisted extraction |
| MALDI | matrix-assisted laser desorption ionisation |
| MC | multicollector |
| ME | microextraction |
| MeOH | methanol |
| MH | metal hydride |
| ML | machine learning |
| MNP | micro and nano particle |
| MOF | metal organic framework |
| MS | mass spectrometry |
| NIRS | near infra-red spectroscopy |
| NIST | National Institute of Standards and Technology |
| NP | nanoparticle |
| OES | optical emission spectrometry |
| PA | polyamide |
| PAES | plasma acoustic emission signal |
| PC | polycarbonate |
| PCA | principal component analysis |
| PE | polyethylene |
| PEEK | polyetheretherketone |
| PET | polyethyleneterephthalate |
| PFAS | perfluoroalkyl substance |
| PGE | platinum group element |
| PHB | poly(3-hydroxybutyrate) |
| PIGE | particle induced gamma ray emission |
| PIXE | particle-induced X-ray emission |
| PLA | polylactic acid |
| PLGA | poly(lactic-co-glycolic acid) |
| PLS | partial least squares |
| PMC | particle mass concentration |
| PMMA | polymethylmethacrylate |
| PNC | particle number concentration |
| PP | polypropylene |
| PS | polystyrene |
| PSO | particle swarm optimisation |
| PTFE | poly(tetrafluoroethylene) |
| PVC | poly(vinyl chloride) |
| PVP | polyvinylpyrrolidone |
| Pyr | pyrolysis |
| RBS | Rutherford backscattering spectrometry |
| REE | rare earth element |
| RM | reference material |
| Rmse | root mean square error |
| RSD | relative standard deviation |
| SAXS | small angle X-ray scattering |
| SDVS | spectral distance variable selection |
| SEC | size exclusion chromatography |
| SEM | scanning electron microscopy |
| SERS | surface-enhanced Raman spectroscopy |
| SIMS | secondary ion mass spectrometry |
| SP | single particle |
| SRM | standard reference material |
| SS | stainless steel |
| SSA | sparrow search algorithm |
| ST | S-transform |
| suIDA | species unspecific isotope dilution analysis |
| SVM | support vector machine |
| SVR | support vector regression |
| TD | thermal desorption |
| TDS | thermal desorption spectrometry |
| TE | transport efficiency |
| TEM | transmission electron microscopy |
| TMAH | tetramethylammonium hydroxide |
| TOF | time-of-flight |
| TSR-NET | twin spectral reconstruction network |
| TXRF | total reflection X-ray fluorescence |
| UHV | ultra-high vacuum |
| UV | ultraviolet |
| VHN | Vicker’s microhardness number |
| VIM | variable importance measurement |
| XPS | X-ray photoelectron spectroscopy |
| XRD | X-ray diffraction |
| XRF | X-ray fluorescence |
| XRR | X-ray reflectometry |
| XRS | X-ray spectrometry |
Conflicts of interest
There are no conflicts to declare.Data availability
There is no additional data associated with this article.References
- E. Bolea-Fernandez, R. Clough, A. Fisher, B. Gibson and B. Russell, J. Anal. At. Spectrom., 2024, 39(11), 2617–2693 RSC.
- J. R. Bacon, O. T. Butler, W. R. L. Cairns, O. Cavoura, J. M. Cook, C. M. Davidson and R. Mertz-Kraus, J. Anal. At. Spectrom., 2024, 39(1), 11–65 RSC.
- M. Patriarca, N. Barlow, A. Cross, S. Hill, A. Robson and J. Tyson, J. Anal. At. Spectrom., 2024, 39(3), 624–698 RSC.
- E. H. Evans, J. Pisonero, C. M. M. Smith and R. N. Taylor, J. Anal. At. Spectrom., 2024, 39(5), 1188–1211 RSC.
- R. Clough, C. F. Harrington, S. J. Hill, Y. Madrid and J. F. Tyson, J. Anal. At. Spectrom., 2024, 39(7), 1629–1664 RSC.
- C. Vanhoof, J. R. Bacon, U. E. A. Fittschen and L. Vincze, J. Anal. At. Spectrom., 2024, 39(9), 2152–2164 RSC.
- S. Hill, H. Goenaga-Infante and R. Garton, J. Anal. At. Spectrom., 2024, 39(9), 2150–2151 RSC.
- T. F. da Silva, J. Vac. Sci. Technol., A, 2025, 43, 020803 CrossRef.
- K. J. Robinson and H. Thissen, J. Vac. Sci. Technol., A, 2024, 42, 040801 CrossRef CAS.
- A. M. Spool and L. Finney, J. Vac. Sci. Technol., A, 2024, 42, 043204 CrossRef CAS.
- S. Y. Zhao, Y. C. Zhao, Y. J. Dai, Z. Y. Liu, Z. Y. Hou, X. Gao and Z. Wang, J. Anal. At. Spectrom., 2025, 40(3), 15 RSC.
- M. F. Churbanov, I. V. Skripachev, A. Y. Sozin and I. I. Evdokimov, J. Anal. Chem., 2024, 79(8), 1011–1019 CrossRef CAS.
- F. H. Alkallas, A. M. Mostafa, A. B. Trabelsi, M. A. Khater, M. M. Motawea, A. N. Al-Ahmadi, W. B. Elsharkawy and R. A. Rezk, Mater. Chem. Phys., 2024, 320, 129443 CrossRef CAS.
- Z. Z. Wang, S. J. Shen, Y. Arima, C. Li, W. Z. Zhou, S. J. Li, J. J. Yan and Y. Deguchi, Spectrochim. Acta, Part B, 2024, 220, 107016 CrossRef CAS.
- S. J. Li, Z. X. Qin, Y. Lu, R. D. Jia, Z. Z. Wang, Y. Deguchi and R. E. Zheng, Spectrochim. Acta, Part B, 2024, 219, 107002 CrossRef CAS.
- D. M. Qu, G. Yang, W. W. Zhou, H. H. Sun, J. X. Fang, D. Tian, C. S. Li and Q. K. Li, Rev. Sci. Instrum., 2024, 95(8), 083001 CrossRef CAS PubMed.
- S. Ahlawat, A. Singh, S. Sahu, P. K. Mukhopadhyay, R. Arya and S. K. Dixit, J. Anal. At. Spectrom., 2024, 39(7), 1927–1937 RSC.
- J. W. Zhang, D. S. Wu, Y. Zhou, Y. Li and L. X. Sun, Spectrosc. Spectral Anal., 2024, 44(10), 2834–2839 CAS.
- T. Chen, L. X. Sun, H. B. Yu, L. F. Qi, P. Zhang and H. Y. Dong, Analyst, 2024, 149(17), 4407–4417 RSC.
- X. H. Huang, X. C. Liu, Y. L. Liu, C. Song, Y. C. Sun and Q. J. Zhang, Spectrosc. Spectral Anal., 2024, 44(9), 2412–2419 CAS.
- K. Li, X. Wang, J. Wang, P. Yang, G. Tian and X. Li, J. Appl. Spectrosc., 2024, 91(5), 1149–1155 CrossRef CAS.
- Y. Jin, S. Liu, H. Min, C. L. Yan, P. Su, Z. M. Huang, Y. R. An and C. Li, Appl. Spectrosc., 2025, 79(3), 426–437 CrossRef CAS PubMed.
- X. M. Lin, X. Y. Duan, J. J. Lin, Y. T. Huang, J. F. Yang, Z. J. Zhang and Y. J. Dong, J. Appl. Spectrosc., 2024, 91(5), 1156–1166 CrossRef CAS.
- Q. D. Zeng, G. H. Chen, W. X. Li, J. L. Meng, G. Li, J. H. Tong, Z. H. Tian, X. L. Zhang, G. H. Li, L. B. Guo and Y. J. Xiao, Spectrosc. Spectral Anal., 2024, 44(6), 1559–1565 CAS.
- Z. Wu, X. H. Su, B. W. Fan, H. H. Zhu, Y. B. Zhang, B. Fang, Y. F. Wang and T. Lü, Spectrosc. Spectral Anal., 2024, 44(12), 3339–3346 CAS.
- J. B. Cai, M. R. Dong, F. Q. Tang, K. Q. Chen, Z. C. Li, W. J. Li, S. C. Yao and J. D. Lu, J. Anal. At. Spectrom., 2025, 40(3), 808–816 RSC.
- L. Retterath, P. Kohns and G. Ankerhold, Spectrochim. Acta, Part B, 2024, 219, 107003 CrossRef CAS.
- P. A. Sdvizhenskii, D. Y. Karpenkov, A. V. Rogachevskaya, S. V. Volokhov, K. E. Galeru, D. V. Kuznetsov and V. N. Lednev, Opt. Laser Technol., 2025, 183, 112378 CrossRef.
- S. L. Xiong, M. C. Cui, N. Yang, G. Y. Shi, Y. X. Pi, Y. Y. Mu, Y. T. Zhang and Y. Zhao, J. Anal. At. Spectrom., 2024, 39(11), 2737–2747 RSC.
- S. L. Xiong, N. Yang, H. Y. Guan, G. Y. Shi, M. Luo, Y. Deguchi and M. C. Cui, Anal. Chim. Acta, 2025, 1336, 343496 CrossRef CAS PubMed.
- F. Q. Tang, M. R. Dong, J. B. Cai, Z. C. Li, K. Q. Chen, W. J. Li, S. C. Yao and J. D. Lu, J. Anal. At. Spectrom., 2024, 39(12), 3025–3034 RSC.
- A. Riedo, P. K. Schmidt, N. J. Boeren, S. Gruchola, L. N. Knecht, M. Tulej and P. Wurz, J. Anal. At. Spectrom., 2025, 40(6), 1469–1474 RSC.
- X. M. Lin, J. F. Yang, Y. T. Huang, J. J. Lin and C. J. Che, J. Anal. At. Spectrom., 2024, 39(11), 2917–2928 RSC.
- C. Soto-Gancedo, A. Méndez-Vicente, I. García-Blanco, J. García-Cabezas, J. Orejas, N. Bordel and J. Pisonero, Appl. Surf. Sci., 2024, 665, 160306 CrossRef CAS.
- K. Asghar, S. Kim, M. Shabpiray, M. F. Ngulimi, B. K. Seo and C. Roh, Mater. Lett., 2024, 370, 136842 CrossRef CAS.
- S. J. Pan, S. D. Deng, Y. J. Qiang and X. H. Li, Mater. Today Commun., 2025, 44, 112103 CrossRef CAS.
- G. J. Liu, J. L. Li and Y. S. Zhang, Constr. Build. Mater., 2024, 440, 137445 CrossRef CAS.
- M. Alinezhadfar, P. Schmutz, F. E. Furcas, J. Reuteler, R. Crockett and U. Angst, Corros. Sci., 2025, 248, 112796 CrossRef CAS.
- A. V. Abramov, R. R. Alimgulov, A. I. Trubcheninova, V. A. Volkovich and I. B. Polovov, Metals, 2024, 14(11), 1215 CrossRef CAS.
- A. Kras, I. Milosev, A. Seyeux and P. Marcus, npj Mater. Degrad., 2024, 8(1), 65 CrossRef CAS.
- A. E. Hughes, C. D. Easton, T. R. Gengenbach, M. C. Biesinger and M. Laleh, J. Vac. Sci. Technol., A, 2024, 42(5), 053206 CrossRef CAS.
- S. Mändl and D. Manova, Metals, 2024, 14(6), 659 CrossRef.
- V. Ortolland, F. Martin, Q. Auzoux and K. Wolski, Int. J. Hydrogen Energy, 2024, 67, 577–591 CrossRef CAS.
- Q. Yang, J. N. Mei, W. Z. Ma, S. J. Feng, Y. H. Hang, Z. Y. Yu, F. J. Shi, H. Q. Qin, C. Ye, Z. M. Lu and S. C. Yao, At. Spectrosc., 2024, 45(6), 462–472 CrossRef CAS.
- D. Wang, H. Li, L. L. Guo, H. J. Guo, X. X. Li, S. S. Li and Z. Y. Zhang, At. Spectrosc., 2024, 45(6), 473–480 CrossRef CAS.
- L. Kalchgruber, L. L. E. Mears, H. Handerkas, M. Hahn, M. Rester, C. Weissensteiner and M. Valtiner, J. Electrochem. Soc., 2025, 172(2), 021502 CrossRef CAS.
- S. G. Morales, A. S. G. Garcia, V. V. Martínez, M. C. Rodriguez, V. M. D. Pidal and M. Montes-Bayón, Talanta, 2025, 287, 127600 CrossRef CAS PubMed.
- C. Walkner, G. Mukhametzianova, S. Wagner, J. C. Korp, A. Graf, J. Irrgeher, T. C. Meisel and T. Prohaska, Metals, 2025, 15(1), 67 CrossRef CAS.
- S. W. Raja, R. Acharya, A. D. Gandhi and J. B. Singh, J. Anal. At. Spectrom., 2024, 39(7), 1919–1926 RSC.
- K. O. Cooke, A. Mirza, J. L. Chen and A. Al Hausone, Crystals, 2024, 14(12), 1048 CrossRef CAS.
- M. Han, Y. H. Jia, L. S. Dai, J. Y. Hu, L. Zhao, X. Zhang, C. Wei and H. Z. Wang, Spectrosc. Spectral Anal., 2024, 44(12), 3406–3413 CAS.
- P. C. Singh and R. K. Singh, Opt. Mater., 2025, 162, 116869 CrossRef CAS.
- D. M. Qu, Z. Y. Zhang, J. X. Liang, H. W. Liao and G. Yang, Spectrosc. Spectral Anal., 2024, 44(11), 3222–3227 CAS.
- X. H. Wang, Y. Gong, Z. J. Zhu, S. Z. He, Y. L. Tian, G. Q. Guo, Y. T. Zhou, X. C. Sun, T. Gong, X. B. Qiu, X. H. He, H. Q. Chen, L. Zhang and C. L. Li, Opt. Laser Technol., 2025, 185, 112525 CrossRef.
- K. P. Lee, D. P. Tran, B. Y. Chen, Y. Q. Lin, C. N. Li, J. Y. Huang, H. C. Chen, R. Y. Chen and C. H. Chen, Electrochim. Acta, 2024, 500, 144769 CrossRef CAS.
- X. Y. He, J. C. Hu, X. Peng, J. Song, Y. F. Yuan and J. L. Qu, Anal. Chim. Acta, 2024, 1330, 343271 CrossRef CAS PubMed.
- M. J. F. Marques, J. Jaume, D. Mercier, A. Seyeux, S. Zanna, R. Basseguy and P. Marcus, Corros. Sci., 2024, 233, 112053 CrossRef CAS.
- R. H. Zhao, R. Hauert, R. Crockett, C. Cancellieri, S. Gaan, P. Schmutz, M. Heuberger and L. P. H. Jeurgens, Appl. Surf. Sci., 2024, 667, 160363 CrossRef CAS.
- O. Beyss, U. Breuer and D. Zander, Appl. Surf. Sci., 2025, 687, 162258 CrossRef CAS.
- A. C. Li, S. Chai, H. M. Peng, Z. Q. Zhao, J. Ren and W. D. Wu, Anal. Chem., 2025, 97(6), 3253–3262 CrossRef CAS PubMed.
- R. Hai, X. Y. Wang, X. Bai, Z. L. He, D. Wu, C. Li, H. B. Ding and X. F. Yao, Spectrochim. Acta, Part B, 2024, 217, 106976 CrossRef CAS.
- Y. J. Dai, Q. Ma, T. S. Zhang, S. Y. Zhao, L. Zhou, X. Gao and Z. Y. Liu, Chemometr. Intell. Lab. Syst., 2025, 258, 105342 CrossRef CAS.
- J. Ekar, S. Kos and J. Kovac, Surf. Interfaces, 2024, 49, 104408 CrossRef CAS.
- W. Q. Zhao, J. Y. Pei, J. S. Zhang, H. X. Xiao, S. P. Li, X. M. Peng and S. Y. Qiu, J. Alloys Compd., 2024, 997, 174925 CrossRef CAS.
- Y. P. Zhao, X. Y. Liu, H. Liu, K. Fang, J. Wang, X. F. Luo, N. Xu, X. L. Sun, Y. Liu, Y. H. Gao, Z. P. Wu, X. F. Li and X. Y. Zhang, Acta Phys. Sin., 2024, 73(21), 216103 CrossRef.
- G. Y. Chen, B. H. Su, D. M. Qu, X. Y. Jin, G. Yang, Q. K. Li and T. Wang, J. Anal. At. Spectrom., 2024, 39(9), 2290–2297 RSC.
- J. J. Zhou, S. F. Hu, X. D. Ren, M. G. Li, Y. Y. Xu, T. L. Zhang, H. S. Tang and H. Li, Spectrochim. Acta, Part B, 2024, 217, 106957 CrossRef CAS.
- J. B. Sirven, C. Quéré, J. Maisonneuve and B. Richebé, Spectrochim. Acta, Part B, 2025, 225, 107123 CrossRef CAS.
- H. G. Li, J. Y. Hu, R. Qi, X. F. Sun and H. Z. Wang, Microchem. J., 2024, 204, 110932 CrossRef CAS.
- K. Ronoh, J. Novotny, L. Mrña, A. Knápek and D. Sobola, Appl. Surf. Sci., 2025, 682, 161687 CrossRef CAS.
- P. R. Behera, R. Farzana and V. Sahajwalla, J. Environ. Manage., 2024, 369, 122371 CrossRef CAS PubMed.
- Q. X. Fan, J. Y. Li, L. W. Zheng, C. Y. Hao, Q. Q. Zhang and Y. Z. Wang, PLoS One, 2024, 19(9), e0309156 CrossRef CAS PubMed.
- M. Beyhaghi, Y. W. Zheng, J. Hobley, M. Rouhani and Y. R. Jeng, Appl. Surf. Sci., 2024, 665, 160276 CrossRef CAS.
- J. Liu, S. N. Tan, Z. N. Qi, L. L. Ren, J. J. Ge and Z. H. Sun, Spectrosc. Spectral Anal., 2024, 44(8), 2250–2255 CAS.
- O. F. Laurent, M. De Marco, M. Faustini, C. Boissière, C. Poleunis, A. Delcorte and D. P. Debecker, Anal. Chem., 2025, 97(6), 3328–3336 CrossRef CAS PubMed.
- J. Liu, J. Wang, X. P. Li, H. Lin, T. C. Liu, B. Y. Zhou and X. Y. He, RSC Adv., 2025, 15(3), 1549–1556 RSC.
- K. Moghadam and D. Beauchemin, At. Spectrosc., 2024, 45(2), 83–91 CrossRef CAS.
- C. S. Seney, A. M. Kiefer, S. T. Brown, E. R. Stair, D. G. Nelson, L. L. Bandoo, W. B. Stewart, N. Z. Lara and C. Donaldson, Anal. Methods, 2025, 17(4), 665–670 RSC.
- L. Beck, M. Mayer, T. F. Silva, C. Berthier and L. Pichon, Nucl. Instrum. Methods Phys. Res., Sect. B, 2024, 555, 165468 CrossRef CAS.
- X. C. Liu, Y. Bai, T. X. Bian, J. Y. Yang, Y. Li and W. R. Duan, J. Laser Appl., 2025, 37(2), 022001 CrossRef.
- W. F. Yang, G. Li, Z. R. Qian, Y. Cao, D. H. Lin, S. L. Li, X. Zheng, D. H. Zhu, M. Y. Xie and Y. K. Yang, Anal. Methods, 2024, 16(28), 4700–4709 RSC.
- M. Wahl, P. Doraciak, L. Römling, T. Scheffler, T. Lampke and F. Henning, Int. J. Adhes. Adhes., 2025, 138, 103938 CrossRef CAS.
- I. Betlej, W. Skrzeczanowski, B. Nasilowska, A. Bombalska, P. Borysiuk, M. Nowacka and P. Boruszewski, Coatings, 2025, 15(1), 34 CrossRef CAS.
- D. Capela, S. Pessanha, T. Lopes, R. Cavaco, J. Teixeira, M. F. S. Ferreira, P. Magalhaes, P. A. S. Jorge, N. A. Silva and D. Guimaraes, J. Hazard. Mater., 2025, 489, 137493 CrossRef CAS PubMed.
- W. R. Song, Z. Y. Song, X. Yue, Z. C. Zhu, J. Wang, H. Wang and Z. Wang, Measurement, 2025, 242, 116074 CrossRef.
- M. Mukai, R. Rani, N. Iwanaga, K. Saeki, K. Toda and S. I. Ohira, Anal. Sci., 2024, 40(9), 1793–1797 CrossRef CAS PubMed.
- F. Simon, M. von der Au, L. Gehrenkemper and B. Meermann, Talanta, 2025, 281, 126811 CrossRef CAS PubMed.
- K. D. Klika, J. L. Han, M. S. Busse, V. A. Soloshonok, R. Javahershenas, F. Vanhaecke and A. Makarem, J. Am. Chem. Soc., 2024, 146(45), 30717–30727 CrossRef CAS PubMed.
- E. Marguí, D. Eichert, J. Jablan, F. Bilo, L. E. Depero, A. Pejovic-Milic, A. Gross, H. Stosnach, A. Kubala-Kukus, D. Banas and L. Borgese, J. Anal. At. Spectrom., 2024, 39(7), 1700–1719 RSC.
- M. Finsgar, Microchem. J., 2024, 202, 112156 CrossRef.
- M. Finsgar, Microchem. J., 2024, 207, 110835 CrossRef.
- M. Lagator, B. Patel, S. Sheraz and N. Lockyer, Anal. Chem., 2024, 96(38), 15185–15193 CAS.
- Y. L. Liu, Y. R. Peng, Q. P. Wu, Z. E. Cheng, F. F. Xie and S. W. Li, Anal. Sci., 2025, 43(6), 813–824 CrossRef PubMed.
- V. H. Cauduro, A. S. Henn, C. A. Bizzi, M. F. Mesko and E. M. M. Flores, At. Spectrosc., 2024, 45(5), 373–380 CrossRef CAS.
- K. Michaelides, M. A. Al Tahan, Y. D. Zhou, G. F. Trindade, D. J. H. Cant, Y. W. Pei, P. Dulal and A. Al-Khattawi, Mol. Pharm., 2024, 21(12), 6245–6256 CrossRef CAS PubMed.
- M. Asadieraghi, T. M. Isfahani and M. A. Senejani, Microchem. J., 2024, 205, 111213 CrossRef CAS.
- B. H. Ali and A. A. Abojassim, Int. J. Environ. Anal. Chem., 2025 DOI:10.1080/03067319.2025.2463992.
- H. F. Castello, F. L. R. Silva, D. S. Ferreira, A. L. M. Levada, E. R. Pereira and F. M. V. Pereira, J. Chemom., 2024, 38, e3575 CrossRef CAS.
- A. Ehsan, N. Sultana, B. Munir, N. Assad, A. Ghaffar and M. Naeem-ul-Hassan, Int. J. Environ. Anal. Chem., 2024 DOI:10.1080/03067319.2024.2408636.
- D. Pruski and M. Sprynskyy, Energies, 2024, 17(17), 4267 CrossRef CAS.
- K. Madhavi, R. Shekhar, G. Venkateshwarlu and K. Dash, Talanta, 2025, 291, 127820 CrossRef CAS PubMed.
- J. B. Cai, M. R. Dong, H. J. Chen, Z. H. Shang, S. C. Yao and J. D. Lu, At. Spectrosc., 2024, 45(3), 200–208 CAS.
- H. K. Du, S. Y. Ke, W. Zhang, D. F. Qi and T. F. Sun, Anal. Sci., 2024, 40(9), 1709–1722 CrossRef CAS PubMed.
- R. Gao, J. X. Li, H. Z. Han, J. C. Song, J. Y. Huo, L. Dong, W. G. Ma, S. Q. Wang, Y. Zhang, L. Zhang, P. H. Zhang, Z. F. Ye, Z. J. Zhu, Y. Zhao, W. B. Yin and S. T. Jia, J. Anal. At. Spectrom., 2025, 17, 10.1039/d4ja00402g.
- M. Chbani, M. C. Boiron, C. Peiffert, J. Marin, C. Lorgeoux and R. Michels, Geostand. Geoanal. Res., 2024, 48(4), 823–835 CrossRef CAS.
- Y. H. Xu, Y. Zhou, Y. T. Li and Y. Zheng, Molecules, 2024, 29(21), 5014 CrossRef CAS PubMed.
- F. Endriss, D. Kuptz, D. Wissmann, H. Hartmann, E. Dietz, A. Kappler and H. Thorwarth, Energy Fuels, 2024, 38(17), 16426–16440 CrossRef CAS.
- M. Vasudeva, A. K. Warrier, V. B. Kartha and V. K. Unnikrishnan, TrAC, Trends Anal. Chem., 2025, 183, 118111 CrossRef CAS.
- A. Moteallemi, M. H. Dehghani, F. Momeniha and S. Azizi, Microchem. J., 2024, 207, 111884 CrossRef CAS.
- F. Sandro, H. Bodo and G. Detlef, J. Anal. At. Spectrom., 2025, 40(1), 276–285 RSC.
- L. Brunnbauer, L. Kronlachner, E. Foisner and A. Limbeck, J. Anal. At. Spectrom., 2025, 40(3), 753–761 RSC.
- R. Gonzalez de Vega, T. T. Moro, B. Grüner, T. D. Maranhao, M. J. Huber, N. P. Ivleva, E. Skrzypek, J. Feldmann and D. Clases, J. Anal. At. Spectrom., 2024, 39(8), 2030–2037 RSC.
- Q. Li, W. W. Chen, F. J. Chu, J. Luo, H. R. Feng and Y. J. Pan, Anal. Chem., 2025, 97(8), 4433–4439 CrossRef CAS PubMed.
- M. Vasudeva, U. K. Adarsh, A. K. Warrier, S. D. George and V. K. Unnikrishnan, Sci. Rep., 2024, 14(1), 19327 CrossRef CAS PubMed.
- W. K. Ho, J. T. S. Lum, T. K. Lam, T. N. Yip, C. H. H. Hor and K. S. Y. Leung, Sci. Total Environ., 2024, 954, 176132 CrossRef CAS PubMed.
- J. P. Zjacić, Z. Katančić, M. Kovacic, H. Kusic, Z. Hrnjak Murgić, D. D. Dionysiou, P. Karamanis and A. Loncaric Bozic, Sci. Total Environ., 2024, 935, 173344 CrossRef PubMed.
- U. K. Adarsh, A. Bankapur, A. K. Pai, V. B. Kartha and V. K. Unnikrishnan, Talanta, 2024, 277, 126393 CrossRef CAS PubMed.
- P. Klingenberg, C. G. Schirmeister, M. Kappeler, A. Calean, H. Biester, E. Licht and B. Rapp, Polym. Test., 2024, 136, 108480 CrossRef CAS.
- B. D. Szeredai, T. Frentiu, M. Ponta, N. Muntean and E. Covaci, Spectrochim. Acta, Part B: Atomic Spec., 2024, 218, 106995 CrossRef CAS.
- M. M. Ghuniem, Food Anal. Methods, 2024, 17(11), 1497–1510 CrossRef.
- A. R. López, G. Binda, G. Roncoroni, S. Recchia, D. Monticelli and D. Spanu, Molecules, 2024, 29(12), 2870 CrossRef PubMed.
- J. R. Ingham, G. L. Donati, L. Douvris, G. Bartzas, D. D. Bussan and C. Douvris, Sci. Total Environ., 2024, 930, 172790 CrossRef CAS PubMed.
- D. Emrick, A. Cochrane and A. P. Rao, Appl. Opt., 2024, 63(31), 8273–8279 CrossRef CAS.
- H. Sattar, S. H. Bakhtiar, J. Y. Zhou, S. Zafar, M. Imran, M. R. Abukhadra, B. Salah, L. B. Guo and W. Luo, Talanta, 2025, 292, 127914 CrossRef CAS PubMed.
- X. X. Tang, Y. L. Li, Y. Li, L. Q. Shi, Y. Qian and Z. D. Li, J. Radioanal. Nucl. Chem., 2024, 333(10), 5049–5056 CrossRef CAS.
- X. Y. Guo, J. Wu, J. H. Li, M. X. Shi, X. X. Zhu, Y. Zhou, D. Wu, Z. Y. Song, S. J. Huang and X. W. Li, J. Anal. At. Spectrom., 2024, 39(11), 13 RSC.
- J. P. Shi, L. L. Wang, H. Xie, S. D. Yao, X. Q. Pu, Z. Liu, X. J. Chen and Y. Ding, Ceram. Int., 2024, 50(12), 20851–20859 CrossRef CAS.
- B. Kumar, B. Das, P. Sinha, U. Jain, P. Sengupta and A. K. Mandal, J. Non-Cryst. Solids, 2025, 657, 123482 CrossRef CAS.
- G. D. Parker, A. Plymale, J. Hager, L. Hanley and X. Y. Yu, Biointerphases, 2024, 19(5), 051004 CrossRef CAS PubMed.
- Y. Q. Sun, X. F. Chen, N. Chen, B. Y. Zhang, Z. X. Jin, Y. Z. Liang, Y. H. Pan, K. Bai, T. S. Wang and H. B. Peng, J. Am. Ceram. Soc., 2025, 5, e20370 CrossRef.
- N. Cöpoglu, I. Ceylan, O. Karaahmet and B. Çiçek, Surf. Interfaces, 2024, 54, 105060 CrossRef.
- A. Sajic, D. Rankovic, M. Ristic, M. Markovic, J. Ciganovic and M. Kuzmanovic, Spectrochim. Acta, Part B, 2025, 224, 107113 CrossRef CAS.
- Z. H. Wei, W. Zhou, P. H. Xu, M. Lei and H. Yao, Constr. Build. Mater., 2025, 458, 139549 CrossRef CAS.
- S. Seeling, R. Köhler, D. Tasche and C. Gerhard, Surf. Interface Anal., 2025, 57(2), 111–120 CrossRef CAS.
- P. W. Li and X. N. Xia, Curr. Anal. Chem., 2025, 21(1), 1–4 CrossRef CAS.
- Y. Z. Song, X. J. Wei, L. X. Song, W. L. Cai, X. H. Liu and W. Zhang, Energy Storage Mater., 2024, 70, 103491 CrossRef.
- P. J. Su, W. Leung, C. J. Shen, M. H. Chen and Y. H. Lai, Talanta, 2025, 291, 127819 CrossRef CAS PubMed.
- J. Tröger, R. Kersting, B. Hagenhoff, D. Bougeard, N. V. Abrosimov, J. Klos, L. R. Schreiber and H. Bracht, J. Appl. Phys., 2025, 137(2), 025301 CrossRef.
- Y. D. Zhou, A. Franquet, V. Spampinato, A. Merkulov, M. R. Keenan, P. A. W. van der Heide, G. F. Trindade, W. Vandervorst and I. S. Gilmore, J. Vac. Sci. Technol., A, 2024, 42(5), 053208 CrossRef CAS.
- D. W. Zhang, H. L. Li, F. Li, W. Guo, L. L. Jin and S. H. Hu, Chin. J. Anal. Chem., 2024, 52(10), 1609–1618 CAS.
- A. Matsumoto, Y. Toyama, Y. Shimazu, K. Nii, Y. Ida and S. Yae, J. Anal. At. Spectrom., 2024, 39(10), 2532–2542 RSC.
- H. C. Wu, R. X. Yi, A. Houben, S. Brezinsek, M. Rasinski, C. Li, G. Sergienko, Y. F. Liang, T. Dittmar and H. B. Ding, Nucl. Mater. Energy, 2024, 41, 101812 CrossRef CAS.
- I. Jang, J. Lee and S. Jeong, Int. J. Precis. Eng. Manuf. Green Technol., 2025, 12(3), 785–798 CrossRef.
- B. Achleitner, T. Huber, S. Larisegger, M. Nelhiebel, P. Knaack and A. Limbeck, Polym. Test., 2024, 141, 108647 CrossRef CAS.
- D. Kaplan, I. Yaman and S. Yalçin, Spectrochim. Acta, Part B, 2024, 215, 106905 CrossRef CAS.
- A. Baron-Wiechec, G. C. Lin, B. B. Xia, A. Sotniczuk, J. J. Ganem and I. Vickridge, Appl. Surf. Sci., 2025, 693, 162645 CrossRef CAS.
- L. Cressa, P. Boillat, M. Gerard, Y. Y. Sun, S. Sharma, O. De Castro, M. Nojabaee, G. Schmitz, T. Wirtz and S. Eswara, Electrochim. Acta, 2024, 494, 144397 CrossRef CAS.
- Z. H. Cui, P. Zuo, Z. Z. Guo, C. M. Wang and A. Manthiram, Adv. Mater., 2024, 36, 2402420 CrossRef CAS PubMed.
- P. M. Delfino, M. Bofanova, E. De Vito, N. Dupré, G. Lamblin, W. Porcher, T. Wirtz and J. N. Audinot, ACS Appl. Mater. Interfaces, 2024, 16(42), 57141–57150 CrossRef CAS PubMed.
- G. D’Agostino, M. Oelze, J. Vogl, J. P. Ghestem, N. Lafaurie, O. Klein, D. Pröfrock, M. Di Luzio, L. Bergamaschi, R. Jacimovic, C. Oster, J. Irrgeher, S. T. Lancaster, A. Walch, A. Röthke, L. Michaliszyn, A. Pramann, O. Rienitz, T. Sara-Aho, O. Cankur, D. Kutan and J. Noireaux, J. Anal. At. Spectrom., 2024, 39(11), 2809–2823 RSC.
- S. T. Lancaster, E. Sahlin, M. Oelze, M. Ostermann, J. Vogl, V. Laperche, S. Touze, J. P. Ghestem, C. Dalencourt, R. Gendre, J. Stammeier, O. Klein, D. Pröfrock, G. Kosarac, A. Jotanovic, L. Bergamaschi, M. Di Luzio, G. D’Agostino, R. Jacimovic, M. Eberhard, L. Feiner, S. Trimmel, A. Rachetti, T. Sara-Aho, A. Roethke, L. Michaliszyn, A. Pramann, O. Rienitz and J. Irrgeher, Waste Manage., 2024, 190, 496–505 CrossRef CAS PubMed.
- M. Dommaschk, T. Sieber and J. Acker, J. Anal. At. Spectrom., 2024, 39(10), 2522–2531 RSC.
- A. Schnickmann, C. Stremtan, L. Schlatt, T. Schirmer and T. Ulrich, J. Anal. At. Spectrom., 2025, 40, 1049–1057 RSC.
- G. H. Li, Y. Y. Chang, V. Wieser and H. W. Cheng, Corros. Sci., 2024, 237, 112278 CrossRef CAS.
- H. R. J. Cox, M. K. Sharpe, C. McAleese, M. Laitinen, J. Dulai, R. Smith, J. England, W. H. Ng, M. Buckwell, L. F. Zhao, S. Fearn, A. Mehonic and A. J. Kenyon, Adv. Mater., 2024, 36, 2408437 CrossRef CAS PubMed.
- Y. Y. Liu, C. Q. Wu, L. L. Dong and J. S. Xiu, Spectrochim. Acta, Part B, 2024, 222, 107062 CrossRef CAS.
- A. J. Goodman, B. F. Benner and M. D. Montaño, Environ. Sci. :Nano, 2025, 12(3), 1789–1800 RSC.
- M. I. Chronakis, B. Meermann and M. von der Au, Anal. Bioanal. Chem., 2025, 417(1), 7–13 CrossRef CAS PubMed.
- L. Chayanun, S. Benter, A. Mikkelsen, R. Timm and J. Wallentin, Nanotechnology, 2025, 36(15), 152001 CrossRef PubMed.
- D. Bartczak, S. C. Nuñez, A. Kubicka, D. Ojeda, A. S. Cachero, S. Cowen, S. Ellison, G. Holcombe and H. Goenaga-Infante, Anal. Bioanal. Chem., 2025, 417, 2655–2667 CrossRef CAS PubMed.
- A. N. Geraldes, M. A. Hortellani, M. Rostelato and J. E. S. Sarkis, J. Braz. Chem. Soc., 2025, 36(5), e20250011 Search PubMed.
- J. Deumer, E. Andresen, C. Gollwitzer, R. Schürmann and U. Resch-Genger, Anal. Chem., 2024, 96(48), 19004–19011 CrossRef CAS PubMed.
- D. Bartczak, S. Cuello-Nuñez, M. Pálmai, S. Hill, P. Petrov, Z. Varga, R. Szalay and H. Goenaga-Infante, Anal. Chem., 2024, 96(32), 12966–12972 CrossRef CAS PubMed.
- M. Hernández-Postigo, A. Sánchez-Cachero, M. Jiménez-Moreno and R. C. R. Martín-Doimeadios, Microchim. Acta, 2025, 192(1), 28 CrossRef PubMed.
- E. Bolea and F. Laborda, Spectrochim. Acta, Part B, 2024, 216, 106941 CrossRef CAS.
- A. Bazo, E. Bolea-Fernandez, A. Rua-Ibarz, M. Aramendía and M. Resano, Anal. Chim. Acta, 2024, 1331, 343305 CrossRef CAS PubMed.
- A. Sánchez-Cachero, M. Jiménez-Moreno, N. R. Fariñas and R. C. R. Martín-Doimeadios, Microchim. Acta, 2023, 190(12), 476 CrossRef PubMed.
- J. H. Dong, Z. L. Zhu, L. J. Li, P. J. Xing, S. Y. Li, L. Ouyang, X. Liu, W. Guo, H. T. Zheng and R. Qian, J. Anal. At. Spectrom., 2024, 39(11), 2791–2798 RSC.
- Z. C. Zhou, M. J. Burgener, J. Burgener and D. Beauchemin, J. Anal. At. Spectrom., 2024, 39(8), 2078–2086 RSC.
- D. Y. Wang, J. H. Zhang, C. J. Fan, X. D. Li, L. H. Liu, X. T. Yan, Y. Y. Li, B. He, Y. G. Yin, L. G. Hu and G. B. Jiang, Anal. Chem., 2025, 97(3), 1792–1798 CrossRef CAS PubMed.
- P. J. Xing, N. S. Belshaw, J. H. Dong, L. J. Li, Y. H. Geng, H. T. Zheng, X. Liu and Z. L. Zhu, Talanta, 2024, 278, 126540 CrossRef CAS PubMed.
- J. S. Stanberry, H. B. Andrews, C. V. Thompson, B. W. Ticknor and B. T. Manard, Anal. Chem., 2025, 97(3), 1688–1694 CrossRef CAS PubMed.
- M. Boutry, J. Mistral, P. Oliveira, N. Baskali-Bouregaa, F. Bessueille-Barbier, N. Gilon, C. Ladaviere and L. Ayouni-Derouiche, J. Anal. At. Spectrom., 2024, 39(7), 1726–1735 RSC.
- L. M. G. dos Santos, C. Barata-Silva, S. A. V. Neto, F. S. Bazilio, A. L. O. da Silva, S. C. Jacob and J. C. Moreira, J. Nanopart. Res., 2024, 26(10), 233 CrossRef CAS.
- S. E. Szakas and A. Gundlach-Graham, J. Anal. At. Spectrom., 2024, 39(7), 1874–1884 RSC.
- B. T. Manard, V. C. Bradley, L. Hendriks, D. R. Dunlap, N. A. Zirakparvar, B. W. Ticknor, M. Toro-Gonzalez and H. B. Andrews, Talanta, 2025, 286, 127516 CrossRef CAS PubMed.
- R. L. Buckman and A. Gundlach-Graham, Anal. Methods, 2024, 16(34), 5802–5811 RSC.
- L. Kronlachner, Z. Gajarska, P. Becker, D. Günther and A. Limbeck, J. Anal. At. Spectrom., 2025, 40(2), 467–477 RSC.
- T. E. Lockwood, L. Schlatt and D. Clases, J. Anal. At. Spectrom., 2025, 40(1), 130–136 RSC.
- C. Aksoy, B. Yüksel, N. Sen, F. Ustaoglu, G. Söylemez and T. Öncü, Forensic Sci. Int., 2025, 368, 112407 CrossRef CAS PubMed.
- A. Bejjani, R. Tannous, G. Chalhoub, M. Hamdan and B. Nsouli, Forensic Chem., 2024, 41, 100617 CrossRef CAS.
- P. D. Zoon and M. E. Janssen, Forensic Sci. Int., 2024, 360, 112049 CrossRef CAS PubMed.
- B. P. Senra, H. Ribeiro and A. Guedes, Forensic Sci. Int., 2024, 361, 112083 CrossRef CAS PubMed.
- M. Kotrly, J. Uher, J. Jakubek, R. Skála, J. Bohácová, I. Turková and K. Starkbaumová, J. Appl. Crystallogr., 2025, 58, 168–179 CrossRef CAS.
- K. Moghadam, D. Beauchemin and C. Dalpé, J. Forensic Sci., 2025, 70, 980–994 CrossRef PubMed.
- L. de Palaminy, F. Poitrasson, C. Pecheyran, G. Barbotin, P. Louvat, S. Berail, A. L. Ronzani, L. Robbiola and S. Baron, J. Anal. At. Spectrom., 2025, 40(1), 226–237 RSC.
- C. G. Duan, T. Zuo, D. Chen and W. G. Luo, Eur. Phys. J. Plus, 2024, 139(8), 729 CrossRef.
- K. Kuehn, K. M. Livingston, J. D. Bethard and M. Baudelet, Forensic Sci. Int., 2025, 367, 112312 CrossRef CAS PubMed.
- O. Anduze, D. Gourier, L. Binet, A. Malergue, V. Grossi and A. Lattuati-Derieux, J. Anal. At. Spectrom., 2025, 40(2), 487–497 RSC.
- N. Lozada-Mendieta and X. S. Villagran, Archaeometry, 2024 DOI:10.1111/arcm.13003.
- M. Charlton and M. Mise, Eur. Phys. J. Plus, 2024, 139(10), 945 CrossRef CAS.
- T. Delbey, D. Norris, D. Douglas and A. Shortland, Microchem. J., 2024, 207, 112015 CrossRef CAS.
- M. T. Giachet, B. Gratuze, D. Genequand, Y. S. B. Loukou, É. Huysecom and A. Mayor, PLoS One, 2025, 20(2), e0318588 CrossRef PubMed.
- K. L. Rasmussen and P. Grinder-Hansen, Heritage Sci., 2024, 12(1), 228 CrossRef CAS.
| This journal is © The Royal Society of Chemistry 2025 |