No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Improving laser-induced breakdown spectroscopy regression models via transfer learning
Abstract
Laser-induced breakdown spectroscopy (LIBS) is a well-established analytical tool with relevance in extra-terrestrial exploration. Despite considerable efforts towards the development of calibration-free LIBS approaches, these are currently outperformed by calibration-based...
... Alternatively, the calibration model can be built in a low-dimensional embedding space, where data from both systems are aligned [19]. The CT was extensively studied in various spectroscopic branches (mostly in NIR and IR [18]) but emerged in the LIBS only relatively recently (see [19], [20]). While the data complexity varies between distinct spectroscopic techniques (e.g., the number of effective features in NIR spectra is considerably lower than in LIBS [21]), transfer approaches are analogical and extendable among the techniques. ...
... Due to the computational cost, only a randomly selected subset of 25,000 spectra (10%) was used for the optimization. K = 4 was selected as the optimal number from the considered options: 3,4,5,6,7,8,9,10,15,20. ...
The mutual incompatibility of distinct spectroscopic systems is among the most limiting factors in Laser-Induced Breakdown Spectroscopy (LIBS). The cost related to setting up a new LIBS system is increased,...
... TL [49] has been widely applied in computer vision for object detection [50], [51], image classification [52], [53], and face recognition [54], [55]. It has also been extended to LIBS applications in recent years, for the classification of rocks on Mars [56], prediction of major oxide content in soils [57], and determination of alkali in rocks [58]. TL has also been shown to be effective in correcting matrix effects in LIBS applications. ...
The application of machine learning (ML) has accelerated the development of laser-induced breakdown spectroscopy (LIBS) in soil analysis. However, analyzing remote LIBS data in real time using ML is challenging due to several factors. Firstly, building robust ML models requires extensive calibration datasets, which are not always possible with limited LIBS experimental data. Secondly, matrix effects can worsen LIBS performance, and changes in sample physical properties or the apparatus can impact the distribution and intensity of emission lines. These issues may lead to concept drift in real-time/online data streaming, causing the relationship between the input and the target spectra to change over time. Consequently, an ML model designed for one LIBS system may not apply to another. To conquer these challenges, we propose a framework based on transfer learning to use limited experimental data and adapt to the emission line variation in the LIBS streaming. A model is first pre-trained using a large labelled source dataset and then fine-tuned with new experimental measurements to classify soil samples. LIBS measurements are conducted with variations in sample properties and experimental parameters to simulate differences in remote LIBS sensors. The collected spectra are fed into the model by chunks, and data evolution is dynamically learned by self-balanced learning to self-adapt to the domain shift. The proposed framework is found effective in improving classification accuracy during data streaming by implementing transfer learning and supporting adaptation compared to the literature. The code of the proposed method is available in the GitHub at
https://github.com/kelci2017/LIBS_streaming
.
... Instead of considering the data, TL can also be performed using modelbased approaches. Namely, a trained model may be fine-tuned (trained further after the initial training) on the newly obtained data, as it has been done in several spectroscopic applications [26,27,28]. Alternatively, a common representation of the old and new datasets can be found by a transformation model. ...
The mutual incompatibility of distinct spectroscopic systems is among the most limiting factors in Laser-Induced Breakdown Spectroscopy (LIBS). The cost related to setting up a new LIBS system is increased, as its extensive calibration is required. Solving the problem would enable inter-laboratory reference measurements and shared spectral libraries, which are fundamental for other spectroscopic techniques. In this work, we study a simplified version of this challenge where LIBS systems differ only in used spectrometers and collection optics but share all other parts of the apparatus, and collect spectra simultaneously from the same plasma plume. Extensive datasets measured as hyperspectral images of heterogeneous specimens are used to train machine learning models that can transfer spectra between systems. The transfer is realized by a pipeline that consists of a variational autoencoder (VAE) and a fully-connected artificial neural network (ANN). In the first step, we obtain a latent representation of the spectra which were measured on the Primary system (by using the VAE). In the second step, we map spectra from the Secondary system to corresponding locations in the latent space (by the ANN). Finally, Secondary system spectra are reconstructed from the latent space to the space of the Primary system. The transfer is evaluated by several figures of merit (Euclidean and cosine distances, both spatially resolved; k-means clustering of transferred spectra). The methodology is compared to several baseline approaches.
With the increase in the number of types of spectrometers in use, calibration models cannot be shared among different instruments, however, this problem can be solved via calibration transfer (CT). In this study, a variety of modern process analysis technology (PAT) data are taken as the research object. After preprocessing the spectra data using principal component analysis (PCA) and cubic spline interpolation, the TrAdaBoost algorithm in transfer learning combined with extreme learning machine (ELM), i.e., TrAdaBoost-ELM, is used to transfer the master model to slave instruments and to make comparisons with the Transfer via an Extreme learning machine Auto-encoder Method (TEAM) and the semi-supervised parameter free framework for calibration enhancement (SS-PFCE) method. After the master model is transferred by the TrAdaBoost-ELM algorithm for the prediction dataset of slave instruments, the mean coefficient of determination of prediction (R
p
2
) increases from 0.7843 to 0.8707, and the mean root mean square error of prediction (RMSEP) decreases from 2.7508 to 2.3112. Furthermore, variable combination population analysis (VCPA) in combination with a genetic algorithm (VCPA-IGA) were used to select characteristic wavelengths in molecular and atomic spectra, respectively. For the same type of laser-induced breakdown spectroscopy (LIBS) instruments K1 and K2, after processing by the VCPA-IGA algorithm, the LIBS calibration model established on K1 was transferred successfully to K2, and for the major elements, the mean R
p
2
= 0.9563 and the mean RMSEP = 1.3796. After processing by the VCPA algorithm, the near-infrared (NIR) model for instrument L was transferred to a different instrument J, and the prediction results were R
p
2
= 0.9110 and RMSEP = 0.4044 °Brix. The results demonstrated that an appropriate variable selection method combined with the TrAdaBoost-ELM algorithm can be effectively used for CT for spectrometers of the same and different types, thus achieving model sharing between different spectrometers.
Laser-induced breakdown spectroscopy (LIBS) has been applied in coal analysis for advantages such as real-time online analysis. Fine-tuning is a transfer learning method that has been utilized in LIBS to improve accuracy in the target domain with a limited training set by introducing a model trained on a different but related source domain. This research proposed a hybrid transfer learning method (HTr-LIBS) to further enhance the performance of LIBS coal analysis by combining fine-tuning with sample reweighting. A neural network was pre-trained on the source domain and target domain training set. The sample weights of the source domain were iteratively adjusted according to the prediction errors. The pre-trained neural network with optimal sample weights was then fine-tuned using the target domain training set. The proposed method significantly improved the analytical accuracy compared to direct modeling using small training sets. When the training set size increased to 19, the R2P of direct modeling for ash content and volatile matter content were 0.8105 and 0.9440, respectively. HTr-LIBS increased the R2P for ash content and volatile matter content to 0.9029 and 0.9627, respectively. The improvements were more significant and stable than fine-tuning of the source domain model without sample reweighting. The introduction of target domain data during pre-training and the iterative adjustment of sample weights both contributed to the improvements.
Laser-induced breakdown spectroscopy (LIBS) instruments have gradually become an attractive technical tool in the field of rock chemical composition analysis due to their advantages of simplicity, rapid detection and simultaneous multi-element analysis. However, the analysis performance of LIBS instruments is prone to being affected by factors such as the changes in the measurement environment and differences between instruments. It is difficult to share the prediction models even among the same type of instruments. Thus, the generalization capability of LIBS analysis instruments is limited. Herein, we propose a transfer learning method based on the dynamic time warping (DTW) algorithm for constructing a transfer model that can correct the secondary instrumental spectrum. The method was tested by using 49 national standard samples and compared with the piecewise direct standardization (PDS) algorithm. The transfer learning method based on DTW was superior to the PDS algorithm in both the classification prediction of lithology and the quantitative prediction of elements. However, it was found that transfer learning could not fully correct the spectra in the quantitative predictions. Hence, the Pearson coefficient was also introduced in this work to select features. Eventually, the average prediction coefficient of determination of the model improved from 0.6844 to 0.9263, and the average root mean square error of prediction was reduced from 10.5621 to 2.3831. The result shows that the transfer learning method, combined with the feature selection and DTW, can not only enhance the potential of the model for a wide range of applications, but also simplify the model and improve the performance of the prediction.
The SuperCam instrument on the Perseverance Mars 2020 rover uses a pulsed 1064 nm laser to ablate targets at a distance and conduct laser induced breakdown spectroscopy (LIBS) by analyzing the light from the resulting plasma. SuperCam LIBS spectra are preprocessed to remove ambient light, noise, and the continuum signal present in LIBS observations. Prior to quantification, spectra are masked to remove noisier spectrometer regions and spectra are normalized to minimize signal fluctuations and effects of target distance. In some cases, the spectra are also standardized or binned prior to quantification. To determine quantitative elemental compositions of diverse geologic materials at Jezero crater, Mars, we use a suite of 1198 laboratory spectra of 334 well-characterized reference samples. The samples were selected to span a wide range of compositions and include typical silicate rocks, pure minerals (e.g., silicates, sulfates, carbonates, oxides), more unusual compositions (e.g., Mn ore and sodalite), and replicates of the sintered SuperCam calibration targets (SCCTs) onboard the rover. For each major element (SiO2, TiO2, Al2O3, FeOT, MgO, CaO, Na2O, K2O), the database was subdivided into five “folds” with similar distributions of the element of interest. One fold was held out as an independent test set, and the remaining four folds were used to optimize multivariate regression models relating the spectrum to the composition. We considered a variety of models, and selected several for further investigation for each element, based primarily on the root mean squared error of prediction (RMSEP) on the test set, when analyzed at 3 m. In cases with several models of comparable performance at 3 m, we incorporated the SCCT performance at different distances to choose the preferred model. Shortly after landing on Mars and collecting initial spectra of geologic targets, we selected one model per element. Subsequently, with additional data from geologic targets, some models were revised to ensure results that are more consistent with geochemical constraints. The calibration discussed here is a snapshot of an ongoing effort to deliver the most accurate chemical compositions with SuperCam LIBS.
With the ChemCam instrument, laser-induced breakdown spectroscopy (LIBS) has successively contributed to Mars exploration by determining the elemental compositions of soils, crusts, and rocks. The American Perseverance rover and the Chinese Zhurong rover respectively landed on Mars on February 18 and May 15, 2021, further increase the number of LIBS instruments on Mars. Such an unprecedented situation requires a reinforced research effort on the methods of LIBS spectral data analysis. Although the matrix effects correspond to a general issue in LIBS, they become accentuated in the case of rock analysis for Mars exploration, because of the large variation of rock compositions leading to the chemical matrix effect, and the difference in surface physical properties between laboratory standards (in pressed powder pellet, glass or ceramic) used to establish calibration models and natural rocks encountered on Mars, leading to the physical matrix effect. The chemical matrix effect has been tackled in the ChemCam project with large sets of laboratory standards offering a good representation of various compositions of Mars rocks. The present work more specifically deals with the physical matrix effect which is still lacking a satisfactory solution. The approach consists in introducing transfer learning in LIBS data treatment. For the specific application of total alkali-silica (TAS) classification of rocks (either with a polished surface or in the raw state), the results show a significant improvement in the ability to predict of pellet-based models when trained together with suitable information from rocks in a procedure of transfer learning. The correct TAS classification rate increases from 25% for polished rocks and 33.3% for raw rocks with a machine learning model, to 83.3% with a transfer learning model for both types of rock samples.
Mars Surface Composition Detector (MarSCoDe) is a scientific payload onboard the Mars rover of China’s Tianwen-1 mission. With the capability of 1.6∼7 m remote detection and analysis, MarSCoDe instrument suite consists of a two dimensional pointing mirror and an optical head outside the rover cabin, a Laser Induced Breakdown Spectroscopy (LIBS) spectrometer and a Short Wave Infrared (SWIR) spectrometer collectively covering 240∼2400 nm and a master controller unit inside the rover body, calibration target sets, optical fibers and power cables connecting the internal and external units. Combining the techniques of active LIBS, passive SWIR and micro-imaging, MarSCoDe provides functions including elemental composition discrimination and quantitative determination, classification of rock and soil characteristics, sample texture imaging and characterization of plasma-excited area. This paper introduces MarSCoDe mainly in terms of scientific objectives, design requirements, assembly and implementation, spectral and radiation calibration, and performance verification. The LIBS laser irradiance on the target can soundly exceed 10 MW/mm2, and the performance of the LIBS module operated at different temperatures has been tested. The field of view of the SWIR spectrometer is 36.5 mrad. The micro-imager can extract the central pixel area of 320 × 320 and 1024 × 1024, and the former is binned into 64 × 64. The 2D pointing mirror has a wide forward detection range and a narrow backward calibration range, with the pointing pitch accuracy better than 0.133°.
On the NASA 2020 rover mission to Jezero crater, the remote determination of the texture, mineralogy and chemistry of rocks is essential to quickly and thoroughly characterize an area and to optimize the selection of samples for return to Earth. As part of the Perseverance payload, SuperCam is a suite of five techniques that provide critical and complementary observations via Laser-Induced Breakdown Spectroscopy (LIBS), Time-Resolved Raman and Luminescence (TRR/L), visible and near-infrared spectroscopy (VISIR), high-resolution color imaging (RMI), and acoustic recording (MIC). SuperCam operates at remote distances, primarily 2–7 m, while providing data at sub-mm to mm scales. We report on SuperCam’s science objectives in the context of the Mars 2020 mission goals and ways the different techniques can address these questions. The instrument is made up of three separate subsystems: the Mast Unit is designed and built in France; the Body Unit is provided by the United States; the calibration target holder is contributed by Spain, and the targets themselves by the entire science team. This publication focuses on the design, development, and tests of the Mast Unit; companion papers describe the other units. The goal of this work is to provide an understanding of the technical choices made, the constraints that were imposed, and ultimately the validated performance of the flight model as it leaves Earth, and it will serve as the foundation for Mars operations and future processing of the data.
Regression models are constructed to predict glucose and lactate concentrations from near‐infrared spectra in culture media. The partial least‐squares (PLS) regression technique is employed, and we investigate the improvement in the predictive ability of PLS models that can be achieved using wavelength selection and transfer learning. We combine Boruta, a nonlinear variable selection method based on random forests, with variable importance in projection (VIP) in PLS to produce the proposed variable selection method, VIP‐Boruta. Furthermore, focusing on the situation where both culture medium samples and pseudo‐culture medium samples can be used, we transfer pseudo media to culture media. Data analysis with an actual dataset of culture media and pseudo media confirms that VIP‐Boruta can effectively select appropriate wavelengths and improves the prediction ability of PLS models, and that transfer learning with pseudo media enhances the predictive ability. The proposed method could reduce the prediction errors by about 61% for glucose and about 16% for lactate, compared to the traditional PLS model.
Accurately identifying pregnancy status is imperative for a profitable dairy enterprise. Mid-infrared (MIR) spectroscopy is routinely used to determine fat and protein concentrations in milk samples. Mid-infrared spectra have successfully been used to predict other economically important traits, including fatty acid content, mineral content, body energy status, lactoferrin, feed intake, and methane emissions. Machine learning has been used in a variety of fields to find patterns in vast quantities of data. This study aims to use deep learning, a sub-branch of machine learning, to establish pregnancy status from routinely collected milk MIR spectral data. Milk spectral data were obtained from National Milk Records (Chippenham, UK), who collect large volumes of data continuously on a monthly basis. Two approaches were followed: using genetic algorithms for feature selection and network design (model 1), and transfer learning with a pretrained DenseNet model (model 2). Feature selection in model 1 showed that the number of wave points in MIR data could be reduced from 1,060 to 196 wave points. The trained model converged after 162 epochs with validation accuracy and loss of 0.89 and 0.18, respectively. Although the accuracy was sufficiently high, the loss (in terms of predicting only 2 labels) was considered too high and suggested that the model would not be robust enough to apply to industry. Model 2 was trained in 2 stages of 100 epochs each with spectral data converted to gray-scale images and resulted in accuracy and loss of 0.97 and 0.08, respectively. Inspection on inference data showed prediction sensitivity of 0.89, specificity of 0.86, and prediction accuracy of 0.88. Results indicate that milk MIR data contains features relating to pregnancy status and the underlying metabolic changes in dairy cows, and such features can be identified by means of deep learning. Prediction equations from trained models can be used to alert farmers of nonviable pregnancies as well as to verify conception dates.
In the last few years, LIBS has become an established technique for the assessment of elemental concentrations in various sample types. However, for many applications knowledge about the overall elemental composition is not sufficient. In addition, detailed information about the elemental distribution within a heterogeneous sample is needed. LIBS has become of great interest in elemental imaging studies, since this technique allows to associate the obtained elemental composition information with the spatial coordinates of the investigated sample. The possibility of simultaneous multi-elemental analysis of major, minor, and trace constituents in almost all types of solid materials with no or negligible sample preparation combined with a high speed of analysis are benefits which make LIBS especially attractive when compared to other elemental imaging techniques. The first part of this review is aimed at providing information about the instrumental requirements necessary for successful LIBS imaging measurements and points out and discusses state-of-the-art LIBS instrumentation and upcoming developments. The second part is dedicated to data processing and evaluation of LIBS imaging data. This chapter is focused on different approaches of multivariate data evaluation and chemometrics which can be used e.g. for classification but also for the quantification of obtained LIBS imaging data. In the final part, current literature of different LIBS imaging applications ranging from bioimaging, geoscientific and cultural heritage studies to the field of materials science is summarized and reviewed.
The primary neural networks’ decision-making units are activation functions. Moreover, they evaluate the output of networks neural node; thus, they are essential for the performance of the whole network. Hence, it is critical to choose the most appropriate activation function in neural networks calculation. Acharya et al. (2018) suggest that numerous recipes have been formulated over the years, though some of them are considered deprecated these days since they are unable to operate properly under some conditions. These functions have a variety of characteristics, which are deemed essential to successfully learning. Their monotonicity, individual derivatives, and finite of their range are some of these characteristics. This research paper will evaluate the commonly used additive functions, such as swish, ReLU, Sigmoid, and so forth. This will be followed by their properties, own cons and pros, and particular formula application recommendations.
Near infrared (NIR) spectrum analysis technology has outstanding advantages such as rapid, nondestructive, pollution-free, and is widely used in food, pharmaceutical, petrochemical, agricultural products production and testing industries. Convolutional neural network (CNN) is one of the most successful methods in big data analysis because of its powerful feature extraction and abstraction ability, and it is especially suitable for solving multi-classification problems. CNN-based transfer learning is a machine learning technique, which migrates parameters of trained model to the new one to improve the performance. The transfer learning strategy can speed up the learning efficiency of the model instead of learning from scratch. In view of the difficulty in acquisition of drug NIR spectral data and high labeling cost, this paper proposes three simple but very effective transfer learning methods for multi-manufacturer identification of drugs based on one-dimensional CNN. Compared with the original CNN, the transfer learning method can achieve better classification performance with fewer NIR spectral data, which greatly reduces the dependence on labeled NIR spectral data. At the same time, this paper also compares and discusses three different transfer learning methods, and selects the most suitable transfer learning model for drug NIR spectral data analysis. Compared with the current popular methods, such as SVM, BP, AE and ELM, the proposed method achieves higher classification accuracy and scalability in multi-variety and multi-manufacturer NIR spectrum classification experiments.
In this review, we present a short, although comprehensive, review on the Industrial Applications of Laser-Induced Breakdown Spectroscopy (LIBS). Attention has been devoted to the most recent applications where LIBS can potentially make positive differences compared to other conventional analytical techniques. Of particular interest were energy, steel, and coal industries, besides new emerging applications, where the intrinsic features of LIBS are particularly interesting, such as sorting of waste for selective recycling, and food industries.
The rapid development in NIR and information technologies saw the development of various initiatives that have generated large scale databases of soil spectroscopy globally. Models generated within a specific spectral or geographical domain should be carefully used in other contexts since they may lose their validity. This includes the application of a global, continental or national spectral libraries to local areas or regions. Both, global and local models are valuable and, ideally, we would like to transfer some of the rules learnt by the more general global models to a local domain. In machine learning, the process of sharing intra-domain information is known as transfer learning. This paper aims to describe and evaluate the effectiveness of transfer learning to "localise" a general soil spectral model. The transfer process consists on, first, training a model with a big volume of data covering a diverse group of cases. Second, some layers of the trained neural network are used to build a local model, which is fine-tuned by using a smaller amount of local data. We demonstrated this method using the LUCAS database, an European dataset, comprising spectral data from 21 countries. For each country, we generated three models: a) Global, with data from all except the country of interest; b) Local, with data from the country; and c) Transfer, pre-trained as the Global model and fine-tuned with data from the country. The results showed that the Transfer model can lower the error (expressed as RMSE) 91% of the cases, with a mean reduction of RMSE: 10.5, 11.8, 12.0 and 11.5% for organic carbon, cation exchange capacity, clay content and pH, respectively. This paper demonstrates the usefulness of transfer learning for soil spectroscopy, which will enhance the use of global spectral libraries for local application.
Soil spectra are often measured in the laboratory, and there is an increasing number of large-scale soil spectral libraries establishing across the world. However, calibration models developed from soil libraries are difficult to apply to spectral data acquired from the field or space. Transfer learning has the potential to bridge the gap and make the calibration model transferrable from one sensor to another. The objective of this study is to explore the potential of transfer learning for soil spectroscopy and its performance on soil clay content estimation using hyperspectral data. First, a one-dimensional convolutional neural network (1D-CNN) is used on Land Use/Land Cover Area Frame Survey (LUCAS) mineral soils. To evaluate whether the pre-trained 1D-CNN model was transferrable, LUCAS organic soils were used to fine-tune and validate the model. The fine-tuned model achieved a good accuracy (coefficient of determination (R²) = 0.756, root-mean-square error (RMSE) = 7.07 and ratio of percent deviation (RPD) = 2.26) for the estimation of clay content. Spectral index, as suggested as a simple transferrable feature, was also explored on LUCAS data, but did not performed well on the estimation of clay content. Then, the pre-trained 1D-CNN model was further fine-tuned by field samples collect in the study area with spectra extracted from HyMap imagery, achieved an accuracy of R² = 0.601, RMSE = 8.62 and RPD = 1.54. Finally, the soil clay map was generated with the fine-tuned 1D-CNN model and hyperspectral data.
Collecting strong enough and repeatable signals from laser-induced plasmas is the primary goal of laser-induced breakdown spectroscopy optical detection systems. Typically, the light emitted from the plasma is refracted by the lens, collected by the fiber, and measured by the spectrometer. In the present work, we established a three-dimensional model to systematically evaluate the overall emission collected from different positions of the plasma for a typical optical collection system composed of a focus lens and a collection fiber, and sensitivity analyses were further performed. In addition, experiments were conducted and partially validated the model. Results showed that for the collection system with an optical fiber located on the focal point of the collection lens, the collection efficiency distribution is almost constant within a large cylindrical-shaped area, while for that located off the focal point, there is a rhombus-shaped area with higher collection efficiency than other areas. This much higher collection efficiency area is small in size but has a large impact on the detected spectral intensity. The spatially distributed collection efficiency on the lens parameters, such as size and position, was further discussed to clarify the impacts of the collection system. Furthermore, sensitivity analyses were performed to evaluate the impact of the collection system on the signal repeatability. Based on these calculations, recommendations for the design of the collection for optimized spectral intensity and stability were proposed.
T91 steel is a representative martensitic heat-resistant steel widely used in high temperature compression components of industrial equipment. During the service period, the operation safety and the service life of the equipment will be affected by the change of structure and mechanical properties of the steel components, which is called material aging. In order to develop a rapid in-situ aging estimation technology of high temperature compression components surface, laser-induced breakdown spectroscopy (LIBS) coupled with support vector machine (SVM) was employed in this paper. The spectral characteristics of 10 T91 steel specimens with different aging grades were analyzed. Line intensities and the line intensity ratios (ionic/atomic and alloying element/matrix element) that indicate the change of metallographic structure were used to establish SVM models, and the results using different variable sets were compared. The model was optimized by comparing different pulse number for practical effectiveness, and the robustness of the model was investigated in dealing with the inhomogeneity of steel composition. The study results show that the estimation model obtained the best performance using line intensities and line intensity ratios averaged from 31st–60th laser pulses as input variables. The estimation accuracy of validation set was greatly improved from 75.8% to 95.3%. In addition, the model showed the outstanding capacity for handling the fluctuations of spectral signals between measuring-points (spots), which indicated that the aging estimation based on a few measuring-points is feasible. The studies presented here demonstrate that the LIBS coupled with SVM is a new useful technique for the aging estimation of steel, and would be well-suited for fast safety assessment in industrial field.
This review is focused on a comparison of LIBS with the two most common plasma Optical Emission Spectroscopy (OES) techniques for analysis of metals; spark OES and glow discharge (GD) OES. It is shown that these two techniques have only minor differences in analytical performance. An important part of the paper reviews a direct comparison of the analytical figures of merit for bulk analysis of steels using spark and LIBS sources. The comparison was carried out using one instrument with interchangeable sources, eliminating differences related to the optical system and detectors. It was found that the spark provides slightly better analytical figures of merit. The spark analysis is considerably faster, the simple design of the spark stand has enabled complete automation, both properties of great importance in the metallurgical industry for routine analysis. The analysis of non-metallic inclusions (NMI) with spark and LIBS is presented, in the case of the spark this has become known as Pulse Distribution Analysis (PDA). A very significant difference between the techniques is that the electrical spark typically evaporates ~ 100 times more material than a single laser pulse, resulting in complete evaporation of an NMI present in the evaporated metal. The major advantage of LIBS is that it is localised with very good lateral resolution. The major advantages of spark is that it is much faster (can be done simultaneous with the bulk analysis) and easier to quantify. Compositional Depth Profiling (CDP) is compared for GD-OES and LIBS. It is shown that for applications where GD-OES is well suited, e.g. coated metallic sheet, GD-OES still performs slightly better than LIBS. Similar to the case of NMI analysis, the major advantage of LIBS is the great lateral resolution. This allows elemental surface mapping, as well as CDP of very small areas on μm scale. One further advantage of LIBS is that samples of almost any material, shape and size can be analysed, whereas GD-OES has only limited capabilities for non-flat and small samples. A general conclusion of this review is that LIBS is not likely to replace spark and GD-OES in the foreseeable future, for applications where these techniques are well suited. On the other hand several new applications, particularly in the field of on-line monitoring of industrial processes, are making great inroads for LIBS in the metallurgical and manufacturing industries.
As an important branch of laser-induced breakdown spectroscopy (LIBS), calibration-free LIBS (CF-LIBS) is a famous element quantitative analysis method without standards. This method has many advantages such as real-time, in-situ, on-site, single-point, and multi-elemental analysis, with excellent potential for geology, archaeology, industrial and environmental monitoring, and biomedicine. In this review, we summarized the development of CF-LIBS. It covered a brief description of the basic theory of CF-LIBS, several modified methods and variants, proposed to overcome the non-stoichiometric ablation, self-absorption effect, and high algorithmic complexity. Furthermore, the applications of CF-LIBS in a variety of fields were reviewed. Finally, the existing problems of CF-LIBS and its potential were discussed.
The accuracy of laser-induced breakdown spectroscopy (LIBS) methods for analyzing geological samples is improved when calibration standards and unknown targets are compositionally similar. A recent study suggests that customized submodels can be used to optimize calibration datasets to achieve more accurate predictions [1]. In practice, this is difficult to implement because the errors inherent in the methods used for sorting unknown targets by composition may affect how successfully this matching can occur. Moreover, creation of submodels intrinsically reduces the size of the dataset on which the model is trained, which has been shown to reduce prediction accuracy. This paper uses LIBS spectra of 2990 unique rock powder standards to compare the accuracy of 1) submodels generated for each element over its geochemical range, 2) submodels created using SiO2 content only, 3) submodels created using the ratio of Si(II)/Si(I) emission lines to group spectra by a proxy for approximate plasma temperature, and 4) models created using all data. Results indicate that prediction accuracies are not always improved by creating submodels because subdividing a dataset to optimize calibrations will always result in a smaller database available for each submodel, and the reduced training set size negatively affects accuracy. Customized LIBS standards for specific applications might overcome this problem in cases where the matrix is similar and the expected concentration range is known. But in a majority of geochemical applications, submodel approaches are only useful in improving prediction accuracies when the initial database is itself extensive enough to support large, robust submodel calibration suites.
In the ceramic production process, the content of Si, Al, Mg, Fe, Ti and other elements in the ceramic raw materials has an important impact on the quality of the ceramic products. Exploring a method that can quickly and accurately analyze the content of key elements in ceramic raw materials is of great significance to improve the quality of ceramic products. In this work, laser-induced breakdown spectroscopy (LIBS) is used for rapid analysis of ceramic raw materials. The chemical element composition and content of ceramic raw materials are quite different, which leads to serious matrix effects. Building an artificial neural network model is an effective way to solve the complex matrix effects, but model training can easily lead to overfitting due to the high number of spectral features and the limited number of samples. In order to solve this problem, we propose a feature extraction method that combines the linear regression (LR) and the sparse and under-complete autoencoder (SUAC) neural network. This LR + SUAC method performs nonlinear feature extraction and dimension reduction on high-dimensional spectral data. The spectral data dimension is reduced from 8188 to 100 through the LR layer, and further reduced to 32 through the SUAC encoding layer. Further, a quantitative analysis model for the elemental composition of ceramic raw materials is established by the combination of LR + SUAC and Back Propagation Neural Network (BPNN). Since the input data dimension and redundant information are greatly reduced by LR + SUAC, the overfitting problem of BPNN is greatly reduced. Experiment results showed that the LR + SUAC + BPNN method obtained the best quantitative analysis performance compared with several other methods in the cross-validation process.
Laser-induced breakdown spectroscopy (LIBS) is regarded as the future superstar for chemical analysis, but the relatively high measurement uncertainty and error remain the persistent challenges for its technological development as well as wide applications. In the present work, mechanisms of measurement uncertainty generation and basic principle of signal uncertainty and matrix effects impacting quantification performance were explained. Furthermore, methods for raw signal improvement including sample preparation, system optimization, and especially plasma modulation, which modulates the laser-induced plasma evolution process for higher signal repeatability and signal-to-noise ratio, were reviewed and discussed. Different LIBS mathematical quantification methods including calibration-free methods and calibration methods, which were classified into physical-principle based calibration model, data-driven based calibration model, and hybrid model, were discussed and compared. Overall, a framework of quantification improvement strategy including key steps and main way-out was summarized and recommended for LIBS future development.
Calibration regression models based on visible and near-infrared (Vis/NIR) spectroscopy are now widely used in the rapid non-destructive prediction of agricultural products' quality parameters. However, the distributions of products' quality parameters and spectral responses are different in various batches of products, so a calibration regression model built on one dataset might be ineffective when tested in another. How to improve the generalization ability of models is a crucial problem and is formulated as ‘Calibration Transfer’. Calibration transfer was firstly proposed to eliminate the difference of spectral responses between spectrometers, but now it has been applied in the transfer between different domains of spectral samples or different components. In this paper, we proposed two robust models with great generalization ability in the calibration transfer task, respectively using dimension reduction and transfer learning, namely SPRS (Standard normal variate, Partial least squares dimension reduction, Ridge regression, Slope/bias) and SNV-based Aug-TrAdaBoost.R2. We tested the two models in spectral datasets of tea leaves to predict the moisture content of samples, it was found that SPRS and SNV-based Aug-TrAdaBoost.R2 can reach great performance over both source and target domains across different batches, different varieties, and different classes of tea leaf samples. SPRS and SNV-based Aug-TrAdaBoost.R2 achieved R² values of 0.9314 and 0.9895 in cross-tea-class prediction whereas traditional calibration method PLSR + S/B only achieved 0.4874. SPRS had low computation complexity and was more robust while SNV-based Aug-TrAdaBoost.R2 had higher accuracy in target domain prediction but was computation-consuming. The two proposed models showed the potentials of online automatic quality parameters prediction and high-accuracy prediction across domains of various samples.
This work was designed to observe and further correct the physical matrix effect in analysis of solid materials with laser-induced breakdown spectroscopy (LIBS), effect arisen when a calibration model established...
Emission spectra yielded by laser-induced breakdown spectroscopy (LIBS) exhibit high dimensionality, redundancy, and sparsity. The high dimensionality is often addressed by principal component analysis (PCA) which creates a low dimensional embedding of the spectra by projecting them into the score space. However, PCA does not effectively deal with the sparsity of the analysed data, including LIBS spectra. Consequently, sparse PCA (SPCA) was proposed for the analysis of high-dimensional sparse data. Nevertheless, SPCA remains underutilized for LIBS applications. Thus, in this work, we show that SPCA combined with genetic algorithms offers marginal improvements in clustering and quantification using multivariate calibration. More importantly, we show that SPCA significantly improves the interpretability of loading spectra. Finally, by using the randomized SPCA (RSPCA) algorithm for carrying out SPCA, we indirectly demonstrate that the analysis of LIBS data can greatly benefit from the tools developed by randomized linear algebra: RSPCA offers a 20-fold increase in computation speed compared to PCA based on singular value decomposition.
During the near‐infrared (NIR) spectroscopy analysis process, most existing methods can carry out calibration transfer only between the same samples. In the machine learning area, transfer learning has the potential to achieve calibration transfer across different kinds of samples. This ability raises the following questions: Is this transfer process feasible in the field of NIR spectroscopy? How can this transfer process be realized? To solve these problems, on the basics of boosting extreme learning machine (ELM), the instance transfer learning method was applied. The TrAdaBoost for classification problems was improved to the TrAdaBoost for regression. Simulation verification of ten datasets (fuels and foods) from different instruments was performed. The results demonstrated that by applying this instance transfer model after principal component analysis (PCA) dimension reduction, the conditions of NIR spectroscopy analysis could be relaxed; in other words, the target attributes and sample types need not be the same.
In the past decades various categories of chemometrics for laser-induced breakdown spectroscopy (LIBS) analysis have been developed, among which an important category is that based on artificial neural network (ANN). The most common ANN scheme employed in LIBS researches so far is back-propagation neural network (BPNN), while there are also several other kinds of neural networks appreciated by the LIBS community, including radial basis function neural network (RBFNN), convolutional neural network (CNN), self-organizing map (SOM), etc. In this paper, we introduce the principles of some representative ANN methods, and offer criticism on their features along with comparison between them. Then we afford an overview of ANN-based chemometrics applied in LIBS analysis, involving material identification/classification, component concentration quantification, and some unconventional applications as well. Furthermore, a comprehensive discussion on ANN-LIBS methodologies is provided from four aspects. First, a few general progressing trends are displayed. Next we expound some specific implementation techniques, including variable selection, network construction, data set utilization, network training, model evaluation, and chemometrics selection. In addition, the limitations of ANN approaches are remarked, mainly concerning overfitting and interpretability. Finally a prospect of future development of ANN-LIBS chemometrics is presented. Throughout the discussion quite a few good practices have been highlighted. This review is expected to shed light on the further upgrade of ANN-based LIBS chemometrics in the future.
This work proposes a spectral convolutional neural network (CNN) operating on laser induced breakdown spectroscopy (LIBS) signals to learn to (1) disentangle spectral signals from the sources of sensor uncertainty (i.e., pre-process) and (2) get qualitative and quantitative measures of chemical content of a sample given a spectral signal (i.e., calibrate). Once the spectral CNN is trained, it can accomplish either task through a single feed-forward pass, with real-time benefits and without any additional side information requirements including dark current, system response, temperature and detector-to-target range. Our experiments demonstrate that the proposed method outperforms the existing approaches used by the Mars Science Lab for pre-processing and calibration for remote sensing observations from the Mars rover, ‘Curiosity’.
Laser-induced breakdown spectroscopy (LIBS) data acquired from 2959 geochemical standards allow the effects of training set size on LIBS accuracy in geochemical analyses to be evaluated. In addition, LIBS prediction accuracies are quantified for 65 elements based on a typical benchtop instrument. Analyses used two equivalent randomly selected subsets of the full data set to compare prediction accuracies of partial least squares models using 75, 50, 25, 10, 5, 2.5, 1, and 0.5% of the total data set for training and the remainder for testing. The number of components, a measure of complexity, in the PLS models was shown to increase with the size of the training set. Based on root mean square errors on unseen test data, our results show that the larger the training set, the better (lower) the prediction accuracy will be on unseen data. Calibration (training set) size was shown to have a first-order effect on prediction accuracy relative to spectral resolution and detector sensitivity. Different methods of assessing model accuracy using root mean square error (RMSE) are compared, including the error of the calibration (RMSE-C), the error of cross-validation (RMSE-CV), and the error of prediction (RMSE-P). Use of RMSE-C is inappropriate because the samples being predicted are those on which the model was trained. In data sets that are sufficiently large, use of test data (RMSE-P) provides the best measure of prediction accuracy, while RMSE-CV is useful only to provide an estimate of subsequent model performance. Increasing the number of cross-validation folds for our large dataset yields surprisingly comparable RMSE-CV values for models with five or more (up to 100) folds, but this result is likely not applicable to smaller data sets and needs further evaluation.
Ablation geometry significantly affects the plasma parameters and the consequent spectroscopic observations in laser-induced breakdown spectroscopy. Nevertheless, plasmas induced by laser ablation under inclined incidence angles are studied to a significantly lesser extent compared to plasmas induced by standard orthogonal ablation. However, inclined ablation is prominent in stand-off applications, such as the Curiosity Mars rover, where the orthogonality of the ablation laser pulse cannot be always secured. Thus, in this work, we characterize non-orthogonal ablation plasmas by applying plasma imaging, tomography, and spectral measurements. We confirm earlier observations according to which non-orthogonal ablation leads to a laser-induced plasma that consists of two distinct parts: one expanding primarily along the incident laser pulse and one expanding along the normal of the sample surface. Moreover, we confirm that the former emits mainly continuum radiation, while the latter emits mainly sample-specific characteristic radiation. We further investigate and compare the homogeneity of the plasmas and report that inclined ablation affects principally the ionic emissivity of laser-induced plasmas. Overall, our results imply that the decreased fluence resulting from inclined angle ablation and the resulting inhomogeneities of the plasmas must be considered for quantitative LIBS employing non-orthogonal ablation.
It is a challenge task to enhance the analysis accuracy of laser-induced breakdown spectroscopy (LIBS) in high temperature applications when certified standard samples used for building calibration curves at high temperature are limited or not available. A novel LIBS quantitative analysis method for alloy steel at high temperature via feature-based transfer learning (FTL) is proposed. The spectral data of calibration samples at room temperature and the spectral data of uncalibrated samples at high temperature are together transferred into a high-dimensional feature space using kernel function mapping where an LIBS regression model is trained and established. For testing samples, the measured spectra at high temperature are mapped into the high-dimensional feature space with the same kernel parameters used in the training process, and then the concentration results can be obtained by the regression model. Experiments on certified alloy steel standard samples were conducted, in which 12 samples with both the concentration information and the measured spectra at room temperature and 8 samples only with the spectra measured at high temperature were used to train the analysis model. The 8 samples at high temperature were used for testing. The experimental results of Cr concentration showed that with the feature-based transfer learning, the mean relative error decreased from 32.31% to 6.08%. The proposed method does not need the element concentration for samples at high temperature to build the regression model, which provides a feasible and effective way for LIBS analysis of samples at high temperature, such as fast industrial measurement in iron and steel smelting production processes.
Calibration transfer is a subtle issue in the practical application of near-infrared (NIR) spectroscopy technique. In this paper, a novel method to calibration transfer based on neighborhood preserving embedding (CTNPE) for correcting spectral differences was proposed. As a manifold learning method, neighborhood preserving embedding (NPE) can not only capture the nonlinear manifold structure, but also retain the linearity and show good generalization ability. Since this approach can reveal low dimensional manifold structure in high dimensional spectroscopic data, it is beneficial to construct the transform relationship between source and target spectra. The performance of CTNPE was assessed and compared to that of piecewise direct standardization (PDS) and other four dimensionality reduction-based methods, including transfer based on target factor analysis (TTFA), spectral space transformation (SST), calibration transfer based on canonical correlation analysis (CTCCA) and based on independent component analysis (CTICA), in two real cases. The results indicated that CTNPE was able to successfully transfer spectra between instruments and samples in different physical states. Furthermore, CTNPE provided lower prediction errors than PDS, TTFA, CTCCA, SST, CTICA and direct prediction without a transfer function. Therefore, the comprehensive investigation carried out in the presented work demonstrates that CTNPE is a promising calibration transfer method for NIR, especially for correcting the variations for samples in different physical states.
In recent years, important developments have been achieved in the application of laser-induced breakdown spectroscopy (LIBS) for elemental imaging. The aim of this chapter is to report recent instrumental configurations and applications related to LIBS-based imaging. In the first section, different instrumental alternatives for LIBS imaging measurements are presented. The second section reports the wide variety of laboratory applications in geological, industrial, and biomedical fields that have benefited from LIBS mapping techniques.
The review focuses on the most relevant advances and is reported in different sections relative to the analyzed objects (identification of rocks/minerals and sourcing; resources applications; slurry and drill cores; rare earth elements; light elements). Special sections report on the good practices for Laser Induced Breakdown Spectroscopy (LIBS) analysis and the most critical points that should be checked in order to validate any LIBS analysis on most common geological purposes. LIBS gives access to the most relevant elements in geosciences/geology and the typical detection limits fit usual requirements, with the advantage of permitting faster analyses than the other techniques classically used. Whether considering the case of metals of economic interest, that of critical elements, or that of light elements, LIBS has definitely been proved an adequate tool, and there is no need to do more on its evaluation in this field of application. Considering that LIBS measurements require limited sample pre-treatment, and considering also that LIBS is a fast all-optical multi-elemental technique, it is undoubtedly the optimal way to achieve a first quick screening and then provide valuable data prior to any further laboratory analyses. Therefore, the recent development of LIBS imaging should quickly lead to the implementation of LIBS imaging systems in the analytical laboratories worldwide in charge of analyzing geological samples.
Laser induced breakdown spectroscopy (LIBS) is a materials characterization technique that has been advanced and refined to provide in situ, real-time quality control of polymer matrix composite (PMC) surfaces. A LIBS system was designed and assembled at the NASA Langley Research Center in order to detect ultralow concentrations of silicone contamination on carbon fiber reinforced polymer (CFRP) materials. The LIBS instrument provides high sensitivity detection and nearly non-destructive inspection of PMC substrates prior to adhesive bonding. This review focuses on work conducted using LIBS as a characterization tool for surface contaminants for improved adhesive bonding and coating processes of CFRP materials. In addition, this work describes how the LIBS technique was advanced by analyzing the laser parameters, studying the laser-matter interactions, and comparing results to X-ray photoelectron spectroscopy. Discussion is also presented on LIBS instrumentation, recommendations for laser parameters and instrumentation components, LIBS technique maturity, applications, and limitations.
Laser‐induced breakdown spectroscopy (LIBS) has become a prominent analytical technique in recent years for real‐time characterization of soil properties. However, only a few studies of soil chemical and physical properties have been reported using LIBS until recently. The aims of this article are to: (a) provide the basic principles of LIBS for soil analysis and (b) present the use of LIBS for the analysis of soil pH, soil texture and the humification degree of soil organic matter (SOM). The second article will cover soil classification and soil elemental analysis, including plant nutrients, carbon (C) and toxic elements. LIBS is a multi‐element analytical technique based on atomic spectroscopy that employs a high‐energy laser pulse focused onto a sample surface to create a transient plasma. It is a spectroscopic analytical technique that requires very little or no sample preparation, examines each sample in seconds, and offers a flexible platform for the examination of a broad array of elements in the sample. LIBS also can be used to infer soil chemical and physical properties if a relationship exists between the chemical composition and the soil properties. With proper calibration, LIBS has a great potential for real‐time in‐field soil analysis and precision farming that could lead to improved soil management and agricultural production, and reduced agricultural environmental impacts.
Highlights
Laser‐induced breakdown spectroscopy (LIBS) is a fast, multi‐element analytical technique with great potential for soil characterization.
Basic principles of LIBS and general description of its use for soil analysis are provided.
Soil chemical and physical characterization by LIBS are reviewed and compared to other techniques.
LIBS advantages, limitations and challenges are discussed for soil chemical and physical characterization.
In‐field soil health assessments, including plant nutrients and toxic elements, are needed and could improve the sustainability of agriculture production. Among the available analytical techniques for these analyses, laser‐induced breakdown spectroscopy (LIBS) has become one of the most promising techniques for real‐time soil analysis at low cost and without the need for reagents. The first part of this two‐part review (Part I, Villas‐Boas, P.R., de Franco, M.A., Gollany, H.T., Martin‐Neto, L. & Milori, D.M.B.P. 2019. Applications of laser‐induced breakdown spectroscopy for soil characterization, Part I: Review of Fundamentals and Chemical and Physical Properties.) in this issue focused on the fundamentals of LIBS for soil analysis and its use for soil chemical and physical characterization. Our objectives in this review article (Part II) are to review (a) the main applications of LIBS in the determination of soil carbon (C), nutrients and toxic elements, spatial elemental mapping, and (b) its use in soil classification. Traditional and more recent techniques will be compared to LIBS, considering their advantages and disadvantages. Laser‐induced breakdown spectroscopy is a promising, versatile technique for detecting many elements in soil samples, requires little or no sample preparation, takes only a few seconds per sample, and has a low cost per sample compared to other techniques. However, overcoming matrix effects is a challenge for LIBS applications in soil analysis, because most studies are conducted with limited changes in the matrix. In spite of the limitation of matrix effects, a typical LIBS system has a limit of detection of 0.3, 0.6, 4, 7, 10, 18, 46 and 89 mg kg ⁻¹ for Mo, Cu, Mg, Mn, Fe, Zn, K and Ca, respectively. Laser‐induced breakdown spectroscopy holds potential for real‐time in‐field spatial elemental analysis of soils and practical applications in precision farming with proper calibration. This could lead to immediate diagnoses of contaminated soil and inefficient nutrient supplies and facilitate well‐informed soil management, increasing agricultural production while minimizing environmental impacts.
Highlights
Laser‐induced breakdown spectroscopy (LIBS) is a fast analytical technique with great potential for elemental mapping of soils.
Elemental analyses of soil and plants and soil classification by LIBS are reviewed.
Soil and rhizosphere spatial elemental analyses by LIBS are presented.
LIBS advantages, limitations and challenges are discussed for soil elemental analysis.
With the rapid development of NIR spectroscopy technology and chemometrics, many previous studies have focused on calibration transfer of quantitative analysis model and lots of effectively methods have been proposed, such as slope and bias correction (SBC), piecewise direct standardization (PDS) etc., by which we can implement calibration transfer between different spectrometers. Furthermore, whether it is possible to realize calibration transfer cross different components or not? To answer this question, this paper proposed a novel method which combines principal component analysis (PCA), weighted extreme learning machine (ELM) and TrAdaBoost algorithm. Two public NIR spectroscopy datasets (Corn and Gasoline) are applied to validate the possibility and effectiveness of proposed algorithm through four different experimental protocols. The experimental results show that while the objects of source and target domains are same, whatever calibration transfer between different instruments or components (experimental protocol #2, #3 and #4), the generalization performance of target domain model will improve a lot, especially while target domain contains fewer samples. Particularly, compared with experimental protocol #2 and #3 (only instruments or components between source and target domains are different), there is a significant improvement while the instruments and components are all different (experimental protocol #4). However, while the objects, components and instruments between source and target domains are all different, the generalization performance of quantitative analysis model can not be improved after calibration transfer. The experimental results indicate that in the area of NIR spectroscopy calibration transfer area, the assumption of original TrAdaBoost algorithm can be relaxed so that the labels between source and target domains can different (cross components), but the features must be same.
Near-infrared spectroscopy mainly reflects the frequency-doubled and total-frequency absorption information of hydrogen-containing groups (O‒H, C‒H, N‒H, S‒H) in organic molecules for near-infrared lights with different wavelengths, so it is applicable to testing of most raw materials and products in the field of petrochemicals. However, the modeling process needs to collect a large number of laboratory analysis data. There are many oil sources in China, and oil properties change frequently. Modeling of each raw material is not only unfeasible but also will affect its engineering application efficiency. In order to achieve rapid modeling of near-infrared spectroscopy and based on historical data of different crude oils under different detection conditions, this paper discusses about the feasibility of the application of transfer learning algorithm and makes it possible that transfer learning can assist in rapid modeling using certain historical data under similar distributions under a small quantity of new data. In consideration of the requirement of transfer learning for certain similarity of different datasets, a transfer learning method based on local similarity feature selection is proposed. The simulation verification of spectral data of 13 crude oils measured by three different probe detection methods is performed. The effectiveness and application scope of the transfer modeling method under different similarity conditions are analyzed.[Figure not available: see fulltext.] © 2019, Higher Education Press and Springer-Verlag GmbH Germany, part of Springer Nature.
Coal is one of the world’s most abundant primary energy sources. Real-time coal analysis technology is thirstily needed for coal blending, combustion optimization, pollution reduction, and pricing. Laser-induced breakdown spectroscopy (LIBS) has been a promising candidate for coal analysis since of its unique fast, in-situ, and online capabilities. Coal is a sedimentary rock with complex and heterogeneous composition, and therefore, laser-coal interaction owns multiplex phenomena. A systematic study of the experimental conditions required for stable coal-plasma formation and evolution is a headway for enhancing LIBS results. In coal power plants, LIBS offers three installation sets: inline, at-line, and offline, with minimal space requirements and ease of retrofit. Moreover, LIBS is a safer technique with lower installation and maintenance costs and fits the concern of coal-fired power plants for multi-elements detection in fast records. Coal analysis mainly includes: calorific heat value, proximate analysis, ultimate analysis, and other related analyses. LIBS data is handled with continuously-developing mathematical and statistical modeling to provide smart extraction of the required spectral information for coal analysis. In this tutorial review, we summarize the previous research contributions utilizing LIBS for coal analysis; including: fundamentals and key factors, operation modes, data processing, and finally analytical results. Over and above, LIBS contribution in the fly-ash analysis, likewise some literature about combustion diagnostics might be included to present a simple guideline for researchers interested in LIBS applications for coal-utilization.
Laser-induced breakdown spectroscopy (LIBS) in the last decades has become a promising analytical technique for a broad variety of archeological objects with great results obtained either alone or in combination with complementary techniques. It is mainly advantageous due to no sample preparation, minimally destructive, rapid analysis and depth profiling analysis spur LIBS technique to become a significant attractive technique for the characterization and conservation of archeological samples or artworks. The present paper describes in brief the basic principles and instrumentation of LIBS, and reviews several case studies on metallic alloys, ceramic, glass, painted artworks, historical buildings and biomaterials in the most recent 7 years (2011–2017) that demonstrate the applicability and prospects of LIBS in the field of archeological science.
In this paper, we present a critical review on the applications of the Laser-Induced Breakdown Spectroscopy (LIBS) technique to Cultural Heritage and Archaeology. The strategies used by the groups involved in this kind of research for the analysis of the typical materials of interest (metals, pigments, pottery, glasses, etc.) are discussed in detail, as well as the use of LIBS in combination with other techniques (LIBS and Raman, LIBS and XRF, LIBS and MS). Specific applications of LIBS as a support for Cultural Heritage restoration and the application of the technique for the analysis of underwater objects are treated in separated sessions. In conclusion, new trends of LIBS for Cultural Heritage and Archaeology (micro-LIBS analysis, 3D elemental imaging, Surface- and Nanoparticle-Enhanced LIBS) are introduced and discussed.
This paper demonstrates the new capabilities of laser-induced breakdown spectroscopy (LIBS) to perform fast high-resolution multi-elemental mappings of geological samples. A rock sample with a 5-mm surface roughness and one with a plane surface were laser scanned at a repetition rate of 1 kHz with a spatial resolution of 50 μm, using the Mission: Coriosity platform from ELEMISSION Inc. The elemental mappings of the few cm² surfaces for 6 elements were obtained, revealing high levels of detail and complexity in the element distribution. In addition, the distributions of 9 elements along linear segments on the rock surfaces reveal correlations between the elements which allow inferring the mineralogical content of the samples. The platform provides a novel approach for fast simultaneous mapping of several elements as it can acquire and process 1000 spectra per second, which was not technically possible with traditional LIBS instrumentation. In addition, the platform is coupled with advanced graphical algorithms allowing the direct visualization of the distribution of several elements on the same image.
The analysis accuracy of laser-induced breakdown spectroscopy (LIBS) in high temperature applications will decrease when certified standard samples used for building calibration curves are insufficient. A novel LIBS quantitative method based on transfer learning is proposed, in which information on the spectra at room temperature is transferred to the spectra at high temperature in order to assist in building a better regression model. An iterative weight adjusting scheme is used for different samples in model training and the concept of ensemble learning is involved when the results of testing samples are predicted. Experiments on certified alloy steel standard samples were conducted to analyze Cr concentrations. The calibration dataset consisted of 15 standard samples at room temperature and 4 standard samples at high temperature. Another 3 samples at high temperature were used for testing. The results showed that the average absolute and relative errors of 3 testing samples were reduced by 1.8% and 20.58%, respectively. The proposed method provides a feasible way for LIBS analysis of samples at high temperature with lower cost and enhances the potentiality of LIBS in online industrial measurement in high temperature production processes, such as iron and steel smelting.
Plastic recycling has been the key issue for reducing environmental problems and resolving resource depletion. To improve the recovery rate of plastics, the plastic wastes are correctly identified according to their resin type. However, the identification system, which is able to identify black plastics according to not only the type of black plastics but also the grade of resins correctly, has not been introduced. In this paper, laser-induced breakdown spectroscopy, intelligent algorithms and preprocessing algorithms are used to improve the identification of black plastics such as polypropylene, polystyrene (PS), and acrylonitrile butadiene styrene (ABS). The laser-induced breakdown spectroscopy is capable of obtaining the characteristic spectrum regardless of material’s physical state. To extract the new features which are very valuable to improving learning performance, increasing computational efficiency, and building better generalization models from the obtained spectra through laser-induced breakdown spectroscopy, the hybrid preprocessing algorithm, composed of principal component analysis and independent component analysis, is used. In addition, the intelligent algorithm named the extended radial basis function neural networks inheriting the advantages of fuzzy theory and neural networks is used to identify black plastic samples into several categories with respect to their resins. The proposed identification system, composed of three parts such as laser induced breakdown spectroscopy, hybrid preprocessing algorithms, and an efficient intelligent classification algorithm, is able to show the synergy effect on the black plastic identification problem. From several experimental results, it can be seen that the identification system based on laser-induced breakdown spectroscopy and the intelligent algorithm is used for identification of black plastics by resin type.
Food quality is related to geographic origins. The frequent occurrence of safety affairs in agricultural products makes it necessary to establish a rapid method for monitoring the quality and safety and classifying origins. In this work, 20 kinds of rice samples from different geographic origins were chosen as samples. Four different sample preparation methods, like rice powder pellet with boric acid (RPPBA), rice powder pellet (RPP), rice grain pellet (RGP) and rice grain (RG), were used to compare the results of rice origins classification using laser-induced breakdown spectroscopy (LIBS). Support vector machine (SVM) was applied to study if the information contained in LIBS spectra was able to classify different geographic origins. The results show that the classification accuracies of these four different sample preparation methods of RPPBA, RPP, RGP, and RG were 93.70%, 95.20%, 98.80%, and 99.20%, respectively; the 5-fold cross-validation classification accuracies were 94.50%, 97.35%, 99.25%, and 99.20%, respectively; and the sample preparation times were 15, 12, 10, and 1 min, respectively. It can be concluded that the RG method was found to be simpler and more efficient. The LIBS technique combined with chemometric method should be a promising tool to rapidly distinguish different rice geographic origins.
Data processing in the calibration-free laser-induced breakdown spectroscopy (LIBS) is usually based on the solution of the radiative transfer equation along a particular line of sight through a plasma plume. The LIBS data processing is generalized to the case when the spectral data are collected from large portions of the plume. It is shown that by adjusting the optical depth and width of the lines the spectra obtained by collecting light from an entire spherical homogeneous plasma plume can be least-square fitted to a spectrum obtained by collecting the radiation just along a plume diameter with a relative error of 10⁻¹¹ or smaller (for the optical depth not exceeding 0.3) so that a mismatch of geometries of data processing and data collection cannot be detected by fitting. Despite the existence of such a perfect least-square fit, the errors in the line optical depth and width found by a data processing with an inappropriate geometry can be large. It is shown with analytic and numerical examples that the corresponding relative errors in the found elemental number densities and concentrations may be as high as 50% and 20%, respectively. Safe for a few found exceptions, these errors are impossible to eliminate from LIBS data processing unless a proper solution of the radiative transfer equation corresponding to the ray tracing in the spectral data collection is used.
Laser-induced breakdown spectroscopy has become a popular tool for rapid elemental analysis of geological materials. However, quantitative applications of LIBS are plagued by variability in collected spectra that cannot be attributed to differences in geochemical composition. Even under ideal laboratory conditions, variability in LIBS spectra creates a host of difficulties for quantitative analysis. This is only exacerbated during field work, when both the laser-sample distance and the angle of ablation/collection are constantly changing. A primary goal of this study is to use empirical evidence to provide a more accurate assessment of uncertainty in LIBS-derived element predictions. We hope to provide practical guidance regarding the angles of ablation and collection that can be tolerated without substantially increasing prediction uncertainty beyond that which already exists under ideal laboratory conditions.
Spectra were collected from ten geochemically diverse samples at angles of ablation and collection ranging from 0° to ± 60°. Ablation and collection angles were changed independently and simultaneously in order to isolate spectral changes caused by differences in ablation angle from those due to differences in collection angle. Most of the variability in atomic and continuum spectra is attributed to changes in the ablation angle, rather than the collection angle. At higher angles, the irradiance of the laser beam is lower and produces smaller, possibly less dense plasmas. Simultaneous changes in the collection angle do not appear to affect the collected spectra, possibly because smaller plasmas are still within the viewing area of the collection optics, even though this area is reduced at higher collection angles.
A key observation is that changes in the magnitude of atomic and total emission are < 5% and 10%, respectively, in spectra collected with the configuration that most closely resembles field measurements (VV) at angles < 20°. In addition, variability in atomic and continuum emission is strongly dependent upon sample composition. Denser, more Fe/Mg-rich rocks exhibited much less variability with changes in ablation and collection angles than Si-rich felsic rocks.
Elemental compositions of our variable angle data that were predicted using a much larger but conventionally-collected calibration suite show that accuracy generally suffers when the incidence and collection angles are high. Prediction accuracy (for measurements acquired with varying collection and ablation angles) varies from ± 1.28–1.86 wt% for Al2O3, ± 1.25–1.66 wt% for CaO, ± 1.90–2.21 wt% for Fe2O3T, ± 0.76–0.94 wt% for K2O, ± 2.85–3.61 wt% MgO, ± 0.15–0.17 wt% for MnO, ± 0.68–0.78 wt% for Na2O, ± 0.33–0.42 wt% for TiO2, and ± 2.94–4.34 wt% SiO2. The ChemCam team is using lab data acquired under normal incidence and collection angles to predict the compositions of Mars targets at varying angles. Thus, the increased errors noted in this study for high incidence angle measurements are likely similar to additional, unacknowledged errors on ChemCam results for non-normal targets analyzed on Mars. Optimal quantitative analysis of LIBS spectra must include some knowledge of the angle of ablation and collection so the approximate increase in uncertainty introduced by a departure from normal angles can be accurately reported.
A review of the methods of signal enhancement in laser-induced breakdown spectroscopy (LIBS) is presented. Conventional LIBS suffers from disadvantages of low sensitivity and high limits of detection compared with other analytical techniques such as inductively coupled plasma mass spectrometry (ICP-MS). During the last two decades, various methods have been applied to LIBS in order to realize highly quantitative and qualitative analysis. Current approaches include double-pulse excitation, spatial or magnetic confinement, spark discharge, etc. Different configurations of experimental setups and conditions are suggested for the realization of these improved techniques, while various parameters influence significantly on the enhancement effect. With the aim to study the laser ablation process and characterize the effectiveness of each method, several parameters such as plasma volume and emission intensity are reviewed. Several suggestions are proposed to explain the mechanism of each enhancement method. These modified techniques have been applied on various materials and fields.
A detailed theoretical and experimental analysis is performed for a wide range of laser operating conditions, typical for laser induced breakdown spectroscopy (LIBS) and laser ablation (LA) experiments on copper metallic target. The plasma parameters were experimentally estimated from the line intensities ratio which reflects the relative population of neutral excited species in the plasma. In the case of LA experiments the highest temperature observed was 8210 ± 370 K. In case of LIBS measurements, a maximum temperature of 8123 K has been determined. The experimental results are in good agreement with a stationary, hydrodynamic model. We have theoretically investigated the plasma emission based on the generalized collisional-radiative model as implemented in the ADAS interconnected set of computer codes and data collections. The ionic population density distribution over the ground and excited states into the cooper plasma is graphically displayed as output from the code. The theoretical line intensity ratios are in good agreement with experimental values for the electron density and temperature range measured in our experiments.
The ChemCam Laser-Induced Breakdown Spectroscopy (LIBS) instrument onboard the Mars Science Laboratory (MSL) rover Curiosity has obtained > 300,000 spectra of rock and soil analysis targets since landing at Gale Crater in 2012, and the spectra represent perhaps the largest publicly-available LIBS datasets. The compositions of the major elements, reported as oxides (SiO2, TiO2, Al2O3, FeOT, MgO, CaO, Na2O, K2O), have been re-calibrated using a laboratory LIBS instrument, Mars-like atmospheric conditions, and a much larger set of standards (408) that span a wider compositional range than previously employed. The new calibration uses a combination of partial least squares (PLS1) and Independent Component Analysis (ICA) algorithms, together with a calibration transfer matrix to minimize differences between the conditions under which the standards were analyzed in the laboratory and the conditions on Mars. While the previous model provided good results in the compositional range near the average Mars surface composition, the new model fits the extreme compositions far better. Examples are given for plagioclase feldspars, where silicon was significantly over-estimated by the previous model, and for calcium-sulfate veins, where silicon compositions near zero were inaccurate. The uncertainties of major element abundances are described as a function of the abundances, and are overall significantly lower than the previous model, enabling important new geochemical interpretations of the data.