No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Prediction of Thermogravimetric Data in Bromine Captured from Brominated Flame Retardants (BFRs) in e-Waste Treatment Using Machine Learning Approaches
Abstract
The principal objective in the treatment of e-waste is to capture the bromine released from the brominated flame retardants (BFRs) added to the polymeric constituents of printed circuit boards (PCBs) and to produce pure bromine-free hydrocarbons. Metal oxides such as calcium hydroxide (Ca(OH) 2) have been shown to exhibit high debromination capacity when added to BFRs in e-waste and capturing the released HBr. Tetrabromobisphenol A (TBBA) is the most commonly utilized model compound as a representative for BFRs. Our coauthors had previously studied the pyrolytic and oxidative decomposition of the TBBA:Ca(OH) 2 mixture at four different heating rates, 5, 10, 15, and 20°C/min, using a thermogravimetric (TGA) analyzer and reported the mass loss data between room temperature and 800°C. However, in the current work, we applied different machine learning (ML) and chemometric techniques involving regression models to predict the TGA data at different heating rates. The motivation of this work was to reproduce the TGA data with high accuracy in order to eliminate the physical need of the instrument itself, so that this could save significant experimental time involving sample preparation and subsequently minimizing human errors. The novelty of our work lies in the application of ML techniques to predict the TGA data from e-waste pyrolysis since this has not been conducted previously. The significance of our work lies in the fact that e-waste is ever increasing, and predicting the mass loss curves faster will enable better compositional analysis of the e-waste samples in the industry. Three ML models were employed in our work, namely Linear, random forest (RF), and support vector regression (SVR), out of which the RF method exhibited the highest coefficient of determination (R 2) of 0.999 and least error of prediction as estimated by the root mean squared error (RMSEP) at all 4 heating rates for both pyrolysis and oxidation conditions. An 80:20 split was used for calibration and validation data sets. Furthermore, for showing versatility and robustness of the best-predicting RF model, it was also trained using all the data points in the lower heating rates of 5 and 10°C/min and predicted on all the data points for the higher heating rates of 15 and 20 continued...
... 20 A plethora of works in literature have focused on the copyrolysis of TBBA with various metal oxides, 21−25 but very few have focused on TBP. 1, 13 We would like to highlight a recent work by us, where thermogravimetric analysis (TGA) of TBBA combined with Ca(OH) 2 was investigated and various machine learning (ML) techniques were applied to reproduce the TGA data. 26 This work was an extension of a previous work also by Ali et al., 12 where TGA data were obtained for samples of both pure TBBA and TBBA combined with Ca(OH) 2 . Furthermore, these samples were also subjected to pyrolysis under both N 2 and O 2 environments, where the objective was to explore the debromination capability of Ca(OH) 2 from the pyrolytic products of TBBA. ...
... Predicting the TGA data for new input conditions such as HRs, time spent by the sample inside the chamber, heat supplied by the TGA instrument at different times, and temperatures of the sample and the chamber at different times is achieved by establishing linear and nonlinear relationships between these inputs and the output mass loss data. In our previous work, 26 we compared the prediction ability of random forest (RF) and support vector (SVR) regression techniques as with ordinary least-squares (OLS)-based multiple linear regression (MLR) for reproducing the TGA data of TBBA copyrolyzed with Ca(OH) 2 under both oxygen and nitrogen environments. These techniques are elaborated in the ML Methods Section for readers' reference. ...
... This was also the industry standard and was chosen due to the large size of our data set. 26 As can be seen in Figure 2, we applied an 80:20 split for calibrating and validating the models in scenario 1, where the ML models were trained by using the TGA data at each HR for both TBP and TBP + hematite samples. However, in scenario 2, two sets of investigations were conducted: (i) the entire data points in the lower HRs considering two sets at a time (5, 10°C/min) initially and then considering three sets at a time (5, 10, and 15°C/min) were used as calibration sets for training the ML models and the performance was tested on the data points from the remaining HRs two at a time (15, 20°C/min) and 20°C/min alone for the calibration set corresponding to three HRs at a time. ...
The release of bromine-free hydrocarbons and gases is a major challenge faced in the thermal recycling of e-waste due to the corrosive effects of produced HBr. Metal oxides such as Fe2O3 (hematite) are excellent debrominating agents, and they are copyrolyzed along with tetrabromophenol (TBP), a lesser used brominated flame retardant that is a constituent of printed circuit boards in electronic equipment. The pyrolytic (N2) and oxidative (O2) decomposition of TBP with Fe2O3 has been previously investigated with thermogravimetric analysis (TGA) at four different heating rates of 5, 10, 15, and 20 °C/min, and the mass loss data between room temperature and 800 °C were reported. The objective of our paper is to study the effectiveness of machine learning (ML) techniques to reproduce these TGA data so that the use of the instrument can be eliminated to enhance the potential of online monitoring of copyrolysis in e-waste treatment. This will reduce experimental and human errors as well as improve process time significantly. TGA data are both nonlinear and multidimensional, and hence, nonlinear regression techniques such as random forest (RF) and gradient boosting regression (GBR) showed the highest prediction accuracies of 0.999 and lowest prediction errors among all the ML models employed in this work. The large data sets allowed us to explore three different scenarios of model training and validation, where the number of training samples were varied from 10,000 to 40,000 for both TBP and TBP + hematite samples under N2 (pyrolysis) and O2 (combustion) environments. The novelty of our study is that ML techniques have not been employed for the copyrolysis of these compounds, while the significance is the excellent potential of enhanced online monitoring of e-waste treatment and extension to other characterization techniques such as spectroscopy and chromatography. Lastly, e-waste recycling could greatly benefit from ML applications since it has the potential to reduce total and operational costs and improve overall process time and efficiency, thereby encouraging more treatment plants to adopt these techniques, resulting in reducing the increasing environmental footprint of e-waste.
... The authors mentioned that managing e-waste through recycling, reusing, and reducing is acknowledged as the primary methods currently in use. The proposed method claims to achieve higher accuracy compared to existing approaches such as deep learning, TensorFlow deep learning, Cuckoo search-based neural network, and machine learning, with accuracy rates of 19.49 %, 18.05 %, 12.77 %, Ali et al. [5] discussed the treatment of e-waste with a focus on capturing bromine released from brominated flame retardants (BFRs) in printed circuit boards (PCBs) and producing bromine-free hydrocarbons. This study applied machine learning and chemometric techniques to predict the TGA data at various heating rates. ...
... The authors aimed to accurately reproduce the TGA data, eliminating the need for the physical instrument, saving experimental time, and minimizing errors. Ali et al. [5] reported a novel method of applying machine learning techniques to predict TGA data in e-waste pyrolysis, which has not been done previously [97]. The significance of this research is that as e-waste continues to increase, faster prediction of mass loss curves enables better compositional analysis in the industry. ...
In today's world, the proliferation of electronic devices has led to a significant increase in electronic waste (e-waste) generation, necessitating the development of innovative approaches for sustainable management. E-waste recycling, which involves the recovery of valuable materials from discarded electronic devices, has emerged as a promising solution to the growing e-waste problem. This article presents an analysis of the current state of research on e-waste management, encompassing various recycling approaches, including mechanical, chemical, and biological methods. The analysis revealed that most of the research on e-waste management has focused on the development of recycling technologies, with a significant emphasis on the use of chemical methods. However, there is a growing interest in the use of biological methods, such as bioreactors and microbial technologies, for e-waste management. Many challenges including lack of uniform regulations, inadequate infrastructure, and high cost of recycling technologies were initiated. The formation of product reuse through remanufacturing, and the deployment of effective recycling facilities are necessary for the management of e-waste. The challenge is to develop innovative and cost-effective solutions to e-waste management (plastic-based e-waste and metals-based e-waste). Several technologies are currently applied to plastic-based e-waste and metals-based e-waste management. primary, secondary, and tertiary recycling of plastic-based e-waste and metallurgical approaches for metals-based e-waste are ideal methods for e-waste management. Furthermore, the techno-economic feasibility of different e-waste recycling approaches was estimated. The analysis suggests that while some recycling approaches are economically viable, there is a need for more research to optimize the efficiency and cost-effectiveness of these methods.
... Recent co-pyrolytic studies utilizing metal oxides such as zincite [18], franklinite [18], La 2 O 3 [19], Sb 2 O 3 [20], CaO [19], CuO [19], Fe 2 O 3 [21] and lead oxide (PbO) have demonstrated varying debromination potential [22]. Likewise, our recent studies simulated the behavior of thermal degradation profiles of BFRs when mixed with Fe 2 O 3 [23] and Ca(OH) 2 [24] using machine learning approaches based on thermogravimetric analysis data. Lead is a semi-volatile heavy metal among the often-discharged compounds in the environment when e-waste is processed, stored, or disposed of using shoddy methods, including open burning causing high environmental toxicity [25]. ...
The Stockholm Convention treaty of the United Nations Environment Program has impeded the mass production of legacy brominated fire retardants (BFRs) paving the route to introduce novel BFRs (NBFRs) into the industrial market. Tetrabromobisphenol A diallyl ether (TBBPA-DAE), is a widely emerging NBFR with a high rate of production. The deleterious impacts, neurobehavioral consequences and toxic effects of NBFRs have been well-documented. Co-pyrolysis of BFRs with metal oxides has emerged as a potential de-bromination technique in e-waste recycling that curtails the bromine release into the environment. Herein, a multitude of characterization studies are done to probe into the debromination efficiency of lead oxide (PbO) during its co-pyrolysis with TBBPA-DAE via products (char, gas and condensates) analyses. The thermogravimetric analysis suggested a pyrolytic run up to 600 • C. The GCMS analysis showed that the release of brominated compounds was completely restricted in the condensate at 600 • C by increasing the phenol production. This was due to the capture of HBr by the PbO to form PbBr 2 , which was validated by the spectral, XRD and SEM-EDX analyses. The IC analysis also endorsed a better efficiency of PbO in HBr capture (80.04 %) in comparison to the HCl capture (45.57 %) proving that PbO is a good debromination agent envisaging further probes to other emerging NBFRs. The study also investigates the de-chlorination of PVC to establish the universal de-halogenation capacity of PbO toward mixed halogenated plastic wastes.
... In all cases, the first zone signifies the debromination zone releasing HBr gases (Ali et al., 2023e;Altarawneh et al., 2019;Kumagai et al., 2017). At this zone, the metallic cations start capturing the bromine free radicals to form respective metal bromides as validated in XRD and SEM-EDX analyses. ...
Brominated flame retardants (BFRs) are bromine-bearing additives added to the polymeric fraction in various
applications to impede fire ignition. The Stockholm Convention and various other legislations abolished legacy
BFRs usage and hence, the so-called novel BFRs (NBFRs) were introduced into the market. Recent studies
spotlighted their existence in household dust, aquifers and aquatic/aerial species. Co-pyrolysis of BFRs with
metal oxides has emerged as a potent chemical recycling approach that produces a bromine-free stream of hydrocarbon.
Herein, we investigate the debromination of two prominent two NBFRs; namely tetrabromobisphenol
A 2,3-dibromopropyl ether (TD) and tetrabromobisphenol A diallyl ether (TAE) through their co-pyrolysis with
zinc oxide (ZnO) and franklinite (ZnFe2O4). Most of the zinc content in electrical arc furnace dust (EAFD) exists
in the form of these two metal oxides. Conversion of these metal oxides into their respective bromides could also
assist in the selective extraction of the valuable zinc content in EAFD. The debromination potential of both oxides
was unveiled via a multitude of characterization studies to analyze products (char, gas and condensates). The
thermogravimetric analysis suggested a pyrolytic run up to 500 ◦C and the TAE treatment with ZnO produced
only a trivial amount of brominated compounds (relative area, 0.83%). Phenol was the sole common compound
in condensable products; potentially formed by the β-scission debromination reaction from the parental molecular
skeleton. Inorganic compounds and methane were the major constituents in the gaseous products. The
pyrochar analyses confirmed the presence of metal bromides retained in the residue, averting the bromine
release into the atmosphere. The ion chromatography analysis portrayed <8% of HBr gas release into the atmosphere
upon pyrolysis with ZnO. The ZnO dominance herein envisaged further probes into other spinel ferrites
in combating brominated polymers.
... On the other hand, SVM creates a decision boundary between two categories based on margin calculation principles, minimizing classi cation error. The classi cation function of SVM has been extended to cancer genomics, enabling the discovery of new biomarkers, drug targets, and deeper insights into cancer-inducing genes in cancer genome classi cation or typing [19][20][21]. In this study, we compare RF and SVM models to select the best model for predicting the probability of multiple indicators or disease onset or progression based on individual patient characteristics. ...
Osteoarthritis (OA) is a debilitating joint disorder characterized by the progressive degeneration of articular cartilage. Although the role of ion channels in OA pathogenesis is increasingly recognized, diagnostic markers and targeted therapies remain limited. In this study, we analyzed the GSE48556 dataset to identify differentially expressed ion channel-related genes (DEGs) in OA and normal controls. We identified a total of 47 DEGs, with the majority involved in transient receptor potential (TRP) pathways. To select potential diagnostic markers, we employed machine learning algorithms, LASSO and SVM-RFE, and identified seven genes (CHRNA4, GABRE, HTR3B, KCNG2, KCNJ2, LRRC8C, and TRPM5) as the best characteristic genes for distinguishing OA from healthy samples. The differential expression of these seven marker genes was validated, and gene set enrichment analysis (GSEA) and gene set variation analysis (GSVA) were performed to explore their involvement in biological pathways. We performed clustering analysis and identified two distinct subtypes of OA, C1 and C2, with differential gene expression and immune cell infiltration profiles. Using weighted gene co-expression network analysis (WGCNA), we identified three key genes (PPP1R3D, ZNF101, and LOC651309) associated with OA. We constructed a prediction model using these genes and validated it using the GSE46750 dataset, demonstrating reasonable accuracy and specificity. Our findings provide novel insights into the role of ion channel-related genes in OA pathogenesis and offer potential diagnostic markers and therapeutic targets for the treatment of OA.
This chapter explores the integration of AI and ML techniques in the design and manufacturing of green and flexible electronics. It emphasizes the significance of these technologies for sustainable development and discusses the challenges and opportunities they present. The chapter provides an overview of AI and ML algorithms applicable to the field, including materials selection, intelligent manufacturing, energy efficiency prediction, and design optimization. Case studies highlight successful applications of AI and ML in green and flexible electronics. Overall, this chapter demonstrates how AI and ML contribute to the sustainability, efficiency, and performance of electronic devices, including AI-assisted materials selection, intelligent manufacturing processes, predictive modeling for energy efficiency, and optimization of flexible electronic designs using machine learning. It demonstrates the effectiveness of AI and ML in achieving sustainable development in the field of green and flexible electronics.
This chapter provides a comprehensive overview of electronic waste (e-waste) generation, classification, and recycling strategies. The chapter outlines the various stages of e-waste recycling, including the collection, sorting, dismantling, and advanced recycling techniques. The use of alternate materials and solutions such as the shared economy model, products-as-a-service model, and product ownership model are also discussed. The chapter delves into organic electronics, including the organic field effect transistors, photovoltaics, memory devices, and LEDs. The role of IT-enabled electronics and global initiatives and policies on e-waste management are highlighted. The chapter concludes with case studies that provide insights into various business models and considerations. Overall, this chapter emphasizes the importance of e-waste management and sustainable solutions to mitigate the adverse effects of e-waste on the environment and public health.
Electronic waste (e-waste) is considered a major issue that our world is tackling nowadays. This electronic waste causes various health issues to animals as well as human beings which further results in environmental pollution in developing countries like India. To overcome these issues, proper e-waste collection is proposed by using the dynamic sine cosine-based neural network optimization (DSCNN) approach. The major objective of this approach involves collecting waste from the individual, hence handling the widespread adoption and use of smartphones. To enhance waste planning collection, residents upload a photograph of their waste to the waste collection company’s server, which mechanically recognizes and categorizes the image. A new classification and detection scheme using the DSCNN approach is proposed for efficient e-waste collection planning and correctly detects the type and quantity of waste components in images. The identification and classification accuracy of the uploaded images is very accurate; this method describes the e-waste collection process in various streets and buildings in Maharashtra, India. Experimental results describe that the proposed approach readily achieves the proper allocation of vehicle collection, vehicle routing plan, and household e-waste collection, resulting in reduced collection costs. Moreover, the proposed DSCNN method is compared to various other methods like random forest algorithm (RFA), fractional henry gas optimization (FHGO), behavior-based swarm model by the fuzzy controller (BSFC), and deep learning convolutional neural network (DL-CNN). The DSCNN approach yielded an e-waste collection detection accuracy of 97%. The accuracy rates of 94%, 95%, 93%, and 92.15% are obtained from the DL-CNN, FHGO, BSFC, and RFA.
Thermal treatment of bromine-contaminated polymers ( i.e. , as in e-waste) with metal oxides is currently deployed as a mainstream strategy in recycling and resources recovery from these objects.
Thermogravimetric analysis (TGA) was utilised to compare the thermal stability of pure phase change material (D-mannitol) to that of nano-enhanced PCM (NEPCM) (i.e., PCM containing 0.5% and 1% multiwalled carbon nanotubes (MWCNT)). Using model-free kinetics techniques, the kinetics of pure PCM and NEPCM degradation were analysed. Three different kinetic models such as Kissinger-Akahira-Sunose (KAS), the Flynn-Wall-Ozawa (FWO), and the Starink were applied to assess the activation energies of the pure and nano-enhanced PCM samples. Activation energies for pure PCM using the Ozawa, KAS, and Starink methods ranged from 71.10–77.77, 79.36–66.87, and 66.53–72.52 kJ/mol, respectively. NEPCM’s (1% MWCNT) activation energies ranged from 76.59–59.11, 71.52–52.28, and 72.15–53.07 kJ/mol. Models of machine learning were utilised to predict the degradation of NEPCM samples; these included linear regression, support vector regression, random forests, gaussian process regression, and artificial neural network models. The mass loss of the sample functioned as the output parameter, while the addition of nanoparticles weight fraction, the heating rate, and the temperature functioned as the input parameters. Experiment-based TGA data can be accurately predicted using the created machine learning models.
Complete and deep removal of bromine from polymeric constituents in waste printed circuit boards (WPCBs) signifies the prime aim of the commonly deployed catalytic/pyrolytic approach in the treatment of e-waste. To attain this aim, a wide array of debromination agents were utilized with a varying degree of effectiveness, scalability, and economic viability. Through the use of a multitude of techniques (flow reactor experiments, chemical analysis, surface characterization, and molecular simulation), this work presents conclusive evidence that co-pyrolysis of tetrabromobisphenol A (TBBA; as a model compound for the bromine content in WPCBs) with calcium hydroxide (Ca(OH)2) results in a complete fixation of the bromine content and produces bromine-free pyrolysis oil. For instance, measurements by IC and FTIR (gas sampling cell) confirms the disappearance of HBr from the gaseous stream during co-pyrolysis of a TBBA (1):Ca(OH)2 (2) mixture between 200 – 500 oC. Likewise, no brominated compounds could be detected from the analysis of the condensate and gas fractions. As the collected oil is dominated by n-alkanes, it is concluded that Ca(OH)2 displays both cracking and debromination capacity. DFT calculations maps out reaction pathways for the transformation of Ca(OH)2 into CaBr2 via dissociative addition of HBr molecules. Formed CaBr2 could open a venue for a complete recovery of bromine in thermal treatment of e-waste. Outcomes reported herein shall be instrumental in the design and operation of an effective chemical recycling facility for the ever-increasing loads of e-waste.
The geometrical and topological features of metal-organic frameworks (MOFs) play an important role in determining their ability to capture and store methane (CH4). Methane is a greenhouse gas that has been shown to be more dangerous in terms of contributing to global warming than carbon dioxide (CO2), especially in the first 20 years of its release into the atmosphere. Its accelerated emission increases the rate of global temperature increase and needs to be addressed immediately. Adsorption processes have been shown to be effective and efficient in mitigating methane emissions from the atmosphere by providing an enormous surface area for methane storage. Among all the adsorbents, MOFs were shown to be the best adsorbents for methane adsorption due to their higher favorable steric interactions, the presence of binding sites such as open metal sites, and hydrophobic pockets. These features may not necessarily be present in carbonaceous materials and zeolites. Although many studies have suggested that the main reason for the increased storage efficiencies in terms of methane in the MOFs is the high surface area, there was some evidence in certain research works that methane storage performance, as measured by uptakes and deliveries in gravimetric and volumetric units, was higher for certain MOFs with a lower surface area. This prompted us to find out the most significant property of the MOF, whether it be material-based or pore-based, that has the maximum influence on methane uptake and delivery, using a comprehensive statistical approach that has not previously been employed in the methane storage literature. The approach in our study employed various chemometric techniques, including simple and multiple linear regression (SLR and MLR), combined with different types of multicollinearity diagnostics, partial correlations, standardized coefficients, and changes in regression coefficient estimates and their standard errors, applied to both the SLR and MLR models. The main advantages of this statistical approach are that it is quicker, provides a deeper insight into experimental data, and highlights a single, most important, parameter for MOF design and tuning that can predict and maximize the output storage and capture performance. The significance of our approach is that it was modeled purely based on experimental data, which will capture the real system, as opposed to the molecular simulations employed previously in the literature. Our model included data from ~80 MOFs and eight properties related to the material, pore, and thermodynamics (isosteric adsorption energy). Successful attempts to model the methane sorption process have previously been conducted using thermodynamic approaches and by developing adsorption performance indicators, but these are either too complex or time-consuming and their data covers fewer than 10 MOFs and a maximum of three MOF properties. By comparing the statistical metrics between the models, the most important and statistically significant property of the MOF was determined, which will be crucial when designing MOFs for use in storing and delivering methane.
2,4,6-tribromophenol (TBP) signifies the most widely produced bromophenols with a direct application as a brominated flame retardant (BFRs). Bromophenols also arise as a major product from thermal degradation of various BFRs. Co-pyrolysis of BFRs with metal oxides represent a viable approach in the chemical recycling of polymers laden with bromine, in a process that aims to reduce or eliminate of brominated species. However, most literature studies have focused on the degradation of tetrabromobisphenol A (TBBA) in particular. This work investigates – via a wide array of techniques- co-pyrolysis of TBP with a notable metal oxide; namely hematite (Fe2O3). Hematite exhibits a good debromination capacity via reducing emission of HBr from degradation of neat TBP by 45% when a 1:1 mixing ratio is considered. Deploying various mixing ratios does not alter the apparent degradation steps. Between 150 to 500 °C, Fe2O3-assited conversion of TBP reaches ∼ 33%. Analysis of evolved products in the oil fraction reveals the formation of several non-brominated species, nonetheless, unreacted TBP dominate the profile of products. However, no brominated compounds are observed in the gaseous fraction where most products feature alkylated benzenes. XRD and SEM-EDX analysis of the solid residue confirms bromine's fixation by hematite. Conversion of hematite into FeBr3, and consequently FeBr2, ensues via dissociative uptake of HBr followed by a water elimination. Addition of hematite reduces the activation energy for the degradation process in reference to neat TBP. Outcomes reported herein shall be useful in further exploring the effectiveness of iron oxides species as debromination agents of BFRs.
Melamine (MEL) and its derivatives are increasingly applied as nitrogenous flame retardants in consumer products. Nevertheless, limited information is available on their environmental occurrence and subsequent human exposure via multiple exposure pathways. In this study, we analysed MEL and its derivatives in dust (indication of the dust ingestion route) and hand wipe samples (indication of the hand-to-mouth route) collected in various microenvironments. The levels of ∑MELs in both dust (median: 24,100 ng/g) and participant hand samples (803 ng/m²) collected in e-waste dismantling workshops were significantly higher than those in samples collected in homes (15,600 ng/g and 196 ng/m², respectively), dormitories (13,100 ng/g and 227 ng/m², respectively) and hotel rooms (11,800 ng/g and 154 ng/m², respectively). Generally, MEL dominated in dust samples collected in e-waste dismantling workshops, whereas cyanuric acid dominated in hand wipe samples. This may occur partly because the latter is an ingredient in disinfection products, which are more frequently employed in daily lives during the COVID-19 pandemic. Exposure assessment suggests that dust ingestion is an important exposure pathway among dismantling workers and the general population, whereas hand-to-mouth contact could not be overlooked in certain populations, such as children and dismantling workers not wear gloves at work.
It is common to split a dataset into training and testing sets before fitting a statistical or machine learning model. However, there is no clear guidance on how much data should be used for training and testing. In this article, we show that the optimal training/testing splitting ratio is p:1$$ \sqrt{p}:1 $$, where p$$ p $$ is the number of parameters in a linear regression model that explains the data well.
This paper uses techniques from Random Matrix Theory to find the ideal training-testing data split for a simple linear regression with m data points, each an independent n-dimensional multivariate Gaussian. It defines "ideal" as satisfying the integrity metric, i.e. the empirical model error is the actual measurement noise, and thus fairly reflects the value or lack of same of the model. This paper is the first to solve for the training and test size for any model in a way that is truly optimal. The number of data points in the training set is the root of a quartic polynomial Theorem 1 derives which depends only on m and n; the covariance matrix of the multivariate Gaussian, the true model parameters, and the true measurement noise drop out of the calculations. The critical mathematical difficulties were realizing that the problems herein were discussed in the context of the Jacobi Ensemble, a probability distribution describing the eigenvalues of a known random matrix model, and evaluating a new integral in the style of Selberg and Aomoto. Mathematical results are supported with thorough computational evidence. This paper is a step towards automatic choices of training/test set sizes in machine learning.
In this work, the application of Short-Wave Infrared (SWIR: 1000–2500 nm) spectroscopy was evaluated to identify plastic waste containing brominated flame retardants (BFRs) using two different technologies: a portable spectroradiometer, providing spectra of single spots, and a hyperspectral imaging (HSI) platform, acquiring spectral images. X-ray Fluorescence (XRF) analysis was preliminarily performed on plastic scraps to analyze their bromine content. Chemometric methods were then applied to identify brominated plastics and polymer types. Principal Component Analysis (PCA) was carried out to explore collected data and define the best preprocessing strategies, followed by Partial Least Squares—Discriminant Analysis (PLS-DA), used as a classification method. Plastic fragments were classified into “High Br content” (Br > 2000 mg/kg) and “Low Br content” (Br < 2000 mg/kg). The identified polymers were acrylonitrile butadiene styrene (ABS) and polystyrene (PS). Correct recognition of 89–90%, independently from the applied technique, was achieved for brominated plastics, whereas a correct recognition ranging from 81 to 89% for polymer type was reached. The study demonstrated as a systematic utilization of both the approaches at the industrial level and/or at laboratory scale for quality control can be envisaged especially considering their ease of use and the short detection response.
Support vector regression (SVR) models have been designed to predict the concentration of chemical oxygen demand in sequential batch reactors under high temperatures. The complex internal interaction between the sludge characteristics and their influent were used to develop the models. The prediction becomes harder when dealing with a limited dataset due to the limitation of the experimental works. A radial basis function algorithm with selected kernel parameters of cost and gamma was used to developed SVR models. The kernel parameters were selected by using a grid search method and were further optimized by using particle swarm optimization and genetic algorithm. The SVR models were then compared with an artificial neural network. The prediction results R2 were within >90% for all predicted concentration of COD. The results showed the potential of SVR for simulating the complex aerobic granulation process and providing an excellent tool to help predict the behaviour in aerobic granular reactors of wastewater treatment.
More than 200 kg real waste electrical and electronic equipment (WEEE) shredder residues from a German dismantling plant were treated at 650 °C in a demonstration scale thermochemical conversion plant. The focus within this work was the generation, purification, and analysis of pyrolysis oil. Subsequent filtration and fractional distillation were combined to yield basic chemicals in high purity. By means of fractional distillation, pure monocyclic aromatic fractions containing benzene, toluene, ethylbenzene, and xylene (BTEX aromatics) as well as styrene and α-methyl sty-rene were isolated for chemical recycling. Mass balances were determined, and gas chromatography mass spectrometry (GC-MS) as well as energy dispersive X-ray fluorescence (EDXRF) measurements provided data on the purity and halogen content of each fraction. This work shows that thermochemical conversion and the subsequent refining by fractional distillation is capable of recycling WEEE shredder residues, producing pure BTEX and other monocyclic aromatic fractions. A significant decrease of halogen content (up to 99%) was achieved with the applied methods.
The main objective of this study is to evaluate and compare the performance of different machine learning (ML) algorithms, namely, Artificial Neural Network (ANN), Extreme Learning Machine (ELM), and Boosting Trees (Boosted) algorithms, considering the influence of various training to testing ratios in predicting the soil shear strength, one of the most critical geotechnical engineering properties in civil engineering design and construction. For this aim, a database of 538 soil samples collected from the Long Phu 1 power plant project, Vietnam, was utilized to generate the datasets for the modeling process. Different ratios (i.e., 10/90, 20/80, 30/70, 40/60, 50/50, 60/40, 70/30, 80/20, and 90/10) were used to divide the datasets into the training and testing datasets for the performance assessment of models. Popular statistical indicators, such as Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), and Correlation Coefficient (R), were employed to evaluate the predictive capability of the models under different training and testing ratios. Besides, Monte Carlo simulation was simultaneously carried out to evaluate the performance of the proposed models, taking into account the random sampling effect. The results showed that although all three ML models performed well, the ANN was the most accurate and statistically stable model after 1000 Monte Carlo simulations (Mean R = 0.9348) compared with other models such as Boosted (Mean R = 0.9192) and ELM (Mean R = 0.8703). Investigation on the performance of the models showed that the predictive capability of the ML models was greatly affected by the training/testing ratios, where the 70/30 one presented the best performance of the models. Concisely, the results presented herein showed an effective manner in selecting the appropriate ratios of datasets and the best ML model to predict the soil shear strength accurately, which would be helpful in the design and engineering phases of construction projects.
We developed Europe-wide models of long-term exposure to eight elements (copper, iron, potassium, nickel, sulfur, silicon, vanadium, and zinc) in particulate matter with diameter <2.5 μm (PM 2.5) using standardized measurements for one-year periods between October 2008 and April 2011 in 19 study areas across Europe, with supervised linear regression (SLR) and random forest (RF) algorithms. Potential predictor variables were obtained from satellites, chemical transport models, land-use, traffic, and industrial point source databases to represent different sources. Overall model performance across Europe was moderate to good for all elements with hold-out-validation R-squared ranging from 0.41 to 0.90. RF consistently outperformed SLR. Models explained within-area variation much less than the overall variation, with similar performance for RF and SLR. Maps proved a useful additional model evaluation tool. Models differed substantially between elements regarding major predictor variables, broadly reflecting known sources. Agreement between the two algorithm predictions was generally high at the overall European level and varied substantially at the national level. Applying the two models in epidemiological studies could lead to different associations with health. If both between-and within-area exposure variability are exploited, RF may be preferred. If only within-area variability is used, both methods should be interpreted equally.
Atmospheric chemistry models are a central tool to study the impact of chemical constituents on the environment, vegetation and human health. These models are numerically intense, and previous attempts to reduce the numerical cost of chemistry solvers have not delivered transformative change.
We show here the potential of a machine learning (in this case random forest regression) replacement for the gas-phase chemistry in atmospheric chemistry transport models. Our training data consist of 1 month (July 2013) of output of chemical conditions together with the model physical state, produced from the GEOS-Chem chemistry model v10. From this data set we train random forest regression models to predict the concentration of each transported species after the integrator, based on the physical and chemical conditions before the integrator. The choice of prediction type has a strong impact on the skill of the regression model. We find best results from predicting the change in concentration for long-lived species and the absolute concentration for short-lived species. We also find improvements from a simple implementation of chemical families (NOx = NO + NO2).
We then implement the trained random forest predictors back into GEOS-Chem to replace the numerical integrator. The machine-learning-driven GEOS-Chem model compares well to the standard simulation. For ozone (O3), errors from using the random forests (compared to the reference simulation) grow slowly and after 5 days the normalized mean bias (NMB), root mean square error (RMSE) and R² are 4.2 %, 35 % and 0.9, respectively; after 30 days the errors increase to 13 %, 67 % and 0.75, respectively. The biases become largest in remote areas such as the tropical Pacific where errors in the chemistry can accumulate with little balancing influence from emissions or deposition. Over polluted regions the model error is less than 10 % and has significant fidelity in following the time series of the full model. Modelled NOx shows similar features, with the most significant errors occurring in remote locations far from recent emissions. For other species such as inorganic bromine species and short-lived nitrogen species, errors become large, with NMB, RMSE and R² reaching >2100 % >400 % and <0.1, respectively.
This proof-of-concept implementation takes 1.8 times more time than the direct integration of the differential equations, but optimization and software engineering should allow substantial increases in speed. We discuss potential improvements in the implementation, some of its advantages from both a software and hardware perspective, its limitations, and its applicability to operational air quality activities.
In this study, the spatial prediction of rainfall-induced landslides at the Pauri Gahwal area, Uttarakhand, India has been done using Aggregating One-Dependence Estimators (AODE) classifier which has not been applied earlier for landslide problems. Historical landslide locations have been collated with a set of influencing factors for landslide spatial analysis. The performance of the AODE model has been assessed using statistical analyzing methods and receiver operating characteristic curve technique. The predictive capability of the AODE model has also been compared with other popular landslide models namely Support Vector Machines (SVM), Radial Basis Function Neural Network (ANN-RBF), Logistic Regression (LR), and Naïve Bayes (NB). The result of analysis illustrates that the AODE model has highest predictability, followed by the SVM model, the ANN-RBF model, the LR model, and the NB model, respectively. Thus AODE is a promising method for the development of better landslide susceptibility map for proper landslide hazard management.
This paper investigates the use of least squares support vector machines and Gaussian process regression for multivariate spectroscopic calibration. The performances of these two non-linear regression models are assessed and compared to the traditional linear regression model, partial least squares regression on an agricultural example. The non linear models, least squares support vector machines, and Gaussian process regression, showed enhanced generalization ability, especially in maintaining homogeneous prediction accuracy over the range. The two non-linear models generally have similar prediction performance, but showed different features in some situations, especially when the size of the training set varies. This is due to fundamental differences in fitting criteria between these models.
The objective of this study was to investigated the use of chemometric modeling of thermogravimetric (TG) data as an alternative approach to estimate the chemical and proximate (i.e. volatile matter, fixed carbon and ash contents) composition of lignocellulosic biomass. Since these properties affect the conversion pathway, processing costs, yield and / or quality of products, a capability to rapidly determine these for biomass feedstock entering the process stream will be useful in the success and efficiency of bioconversion technologies. The 38-minute long methodology developed in this study enabled the simultaneous prediction of both the chemical and proximate properties of forest-derived biomass from the same TG data. Conventionally, two separate experiments had to be conducted to obtain such information. In addition, the chemometric models constructed with normalized TG data outperformed models developed via the traditional deconvolution of TG data. PLS and PCR models were especially robust in predicting the volatile matter (R²–0.92; RPD– 3.58) and lignin (R²–0.82; RPD– 2.40) contents of the biomass. The application of chemometrics to TG data also made it possible to predict some monomeric sugars in this study. Elucidation of PC loadings obtained from chemometric models also provided some insights into the thermal decomposition behavior of the chemical constituents of lignocellulosic biomass. For instance, similar loadings were noted for volatile matter and cellulose, and for fixed carbon and lignin. The findings indicate that common latent variables are shared between these chemical and thermal reactivity properties. Results from this study buttresses literature that have reported that the less thermally stable polysaccharides are responsible for the yield of volatiles whereas the more recalcitrant lignin with its higher percentage of elementary carbon contributes to the yield of fixed carbon.
An important question in constructing Cross Validation (CV) estimators of the generalization error is whether rules can be established that allow " optimal " selection of the size of the training set, for fixed sample size n. We define the resampling effectiveness of random CV estimators of the generalization error as the ratio of the limiting value of the variance of the CV estimator over the estimated from the data variance. The variance and the covariance of different average test set errors are independent of their indices, thus, the resampling effectiveness depends on the correlation and the number of repetitions used in the random CV estimator. We discuss statistical rules to define optimality and obtain the " optimal " training sample size as the solution of an appropriately formulated optimization problem. We show that in a broad class of smooth loss functions, and in particular for the q-class of loss functions, when the decision rule is the sample mean the problem of obtaining the optimal training sample size has a general solution independent of the data distribution. The analysis offered when the decision rule is regression illustrates the complexity of the problem.
Calcium hydroxide [Ca(OH)2] is one of the major constituents of hydrated portland cement paste. Its content can be used to trace the progress of cement hydration or serve as an indicator of the extent of pozzolanic reaction. The thermogravimetric analysis (TGA) method is often used to determine the Ca(OH)2 content because it is a relatively easy and fast procedure. However, no universally accepted method exists for the preparation of TGA specimens and for the interpretation of the resulting TGA curves. This paper presents an investigation on the contents of Ca(OH)2 in samples subjected to different preparation techniques. The results showed that a certain amount of calcium carbonate (CaCO3) was produced as a result of carbonation during the sample preparation process. The degree of carbonation was dependent on the sample preparation, and carbonated Ca(OH)2 was considered to determine the accurate total Ca(OH)2 content. In addition, a modified interpretation of the TGA curve for Ca(OH)2 was suggested. In this interpretation, the mass losses caused by the other hydration products, except for the Ca(OH)2 and the carbonated Ca(OH)2, were considered so that the accurate content of Ca(OH)2 could be determined. The interpretation technique was verified by comparing the results with those obtained by differential scanning calorimetry. Ultimately, the actual contents of Ca(OH)2 in pastes undergoing different sample preparation techniques were determined by using the modified interpretation of the TGA curve for the Ca(OH)2. The results showed that this interpretation yielded comparable contents of Ca(OH)2 in most of the sample preparation techniques used in this study.
Brominated flame retardants (BFRs) constitute a major load in the polymeric fraction of e-waste. Degradation of BFRs-laden plastics over transition metal oxides is currently deployed as a mainstream strategy in the disposal and treatment of the non-metallic segment of e-waste. However, interaction of pyrolysis’s products of BFRs with transition metal oxides is well-known to facilitate the formation of notorious pollutants. Despite recent progress to comprehend the germane chemistry of this interaction, several important pertinent aspects remain to be addressed. To fill in this gap, an integrated experimental and simulation account of the pyrolytic and oxidative decomposition of a gaseous stream of 2,4,6-tribromophenol (TBP) over hematite (Fe2O3) has been reported herein. TBP is utilized as a model compounds of BFRs as their most common formulations include brominated phenolic rings. Overall, hematite entails a rather low cracking capacity under pyrolytic conditions. Analysis of condensate products indicates that oxidative degradation of a gaseous stream of TBP results mainly in the formation of brominated alkanes such as bromoethane and bromo-pentane. Likewise, Ion chromatography (IC) measurements disclosed a noticeable reduction in the concentrations of escaped HBr. Transformation of iron oxides into iron bromides (possibly in the form of FeBr2) during pyrolysis and combustion operations is evident through XRD measurements. Density functional theory (DFT) calculations map out important reactions pathways that operate in the initial degradation of the TBP molecule. From a broader perspective, outlined results shall be instrumental to precisely assess the effectiveness of using iron oxides in thermal catalytic recycling of e-waste and the likely emission of brominated toxicants.
The surface tension at the critical micelle concentration (STCMC) is an important descriptor for surfactants in applications including cosmetics, pharmaceuticals and food. A predictive STCMC Random Forest model was trained with 691 conventional surfactants and 9 amino acids. The model was evaluated by fivefold cross-validation, and the prediction power was tested for peptides by direct comparison with the experimental STCMC values found in the literature or measured in this work. Predictions were also conducted for short peptide permutations. The estimated STCMC approached the experimental values as the peptide length decreased, suggesting a strong influence of secondary structures for longer sequences where the developed algorithm fails to make robust predictions. Concerning the short peptide permutations, the model estimated lower STCMC values as the carbon number increased within the hydrophobic portions. This highlights the importance of the hydrophobic amino acids leucine, isoleucine, and phenylalanine in peptides with attractive surfactant properties.
This study presents a novel recycling scheme for spent Li-ion batteries that involves the leaching of lithium in hot water followed by the dissolution of all transition metals in HCl solution and their separation using the ionic liquid Cyphos IL104. The parametric studies revealed that >84 % Li was dissolved while the cathode material was leached at 90 °C for 2 h. Approximately 98 % Li from the non-acidic solution was directly precipitated as Li2CO3 at a Li+:CO32- ratio of 1:1.5. The transition metals from the Li-depleted cathode mass were efficiently (>98 %) dissolved in 3.0 mol·L-1 HCl at 90 °C for a 3 h leaching process. Manganese from the chloride leach liquor was selectively precipitated by adding KMnO4 at a 1.25-fold higher quantity than the stoichiometric ratio, pH value 2.0, and temperature 80 °C. The remaining co-existing metals (Ni and Co) were separated from the chloride solution by contacting it with a phosphonium-based ionic liquid at an equilibrium pH value of 5.4 and an organic-to-aqueous phase ratio of 2/3. The loaded ionic liquid was quantitatively stripped in 2.0 mol·L-1 H2SO4 solution, which yielded high-purity CoSO4·xH2O crystals after evaporation of the stripped liquor. Subsequently, ∼99 % nickel was recovered as nickel carbonate [NiCO3·2Ni(OH)2] from the Co-depleted raffinate by the precipitation performed at Ni2+:CO32- ratio of 1:2.5, pH value of 10.8, and temperature of 50 °C. Finally, a process flow with mass and energy balances yielding a high recovery rate of all metals in the exhausted cathode powder of spent LiBs was proposed.
With light emitting diode (LED) devices popularizing, waste LEDs are expected to increase rapidly. Nevertheless, recycling policies and chains lag, largely due to scattered knowledge of waste generation and recycling benefits. Herein, Holt's method is used to estimate the sales of LED devices in China from 2012-2025. Then, we employ the market supply A model with Weibull lifetime distribution embedded to quantify the waste generation and in-use stock of LEDs. On this basis, the potential economic and environmental benefits of recycling are explored. Results show that: (1) From 2012 to 2025, the in-use stock of LEDs will increase from 0.13 to 66.17 billion pieces, while waste generation will increase from 1.57 million to 4.79 billion pieces; (2) In 2025, the potential economic benefits of recycling will reach $11.99 billion, mainly from precious metals; (3) In 2025, the carbon emission reductions of recycling will reach 7.54 billion tons (CO2-eq.), derived from the substitution of recycling for other waste disposal options. To keep technologies, capacities, and collection systems relatively in sync with the growing waste LEDs, priority should be given to lifetime extension and then sustainable designs. Nowadays, incentives for lifetime extension and the utilization of recycling systems for other e-waste are beneficial. In the medium-term, designs of modularity, standardization, and generalization, as well as the integration of disassembly, pre-treatment, and resource extraction, would be the focus. In the long run, the authority and stakeholders should spend efforts on clarifying recycling responsibilities, developing tailed business models, and promoting cost-effective recycling technologies.
The increasing amount of e-waste plastics needs to be disposed of properly, and removing the brominated flame retardants contained in them can effectively reduce their negative impact on the environment. In the present work, TBBPA-bis-(2,3-dibromopropyl ether) (TBBPA-DBP), a novel brominated flame retardant, was extracted by ultrasonic-assisted solvothermal extraction process. Response Surface Methodology (RSM) achieved by machine learning (support vector regression, SVR) was employed to estimate the optimum extraction conditions (extraction time, extraction temperature, liquid to solid ratio) in methanol or ethanol solvent. The predicted optimum conditions of TBBPA-DBP were 96 min, 131 mL g⁻¹, 65 °C, in MeOH, and 120 min, 152 mL g⁻¹, 67 °C in EtOH. And the validity of predicted conditions was verified.
Literature provides detailed mechanisms underpinning the formation of a wide array of bromine (Br)-containing molecules with a prime focus on dioxin-like compounds. However, from a more applied point of view, the practical deployment of attained thermo-kinetic parameters remains inadequate in the absence of a robust kinetic model that connects bromine transformation at the molecular level with pertinent experimental observations. Herein and to fill in this gap, this study constructs a chemical kinetic model to account for the “homogenous gas phase” emission of Br-aromatic pollutants from the oxidative thermal decomposition of a monobromobenzene molecule (MBZ). The latter serves as a model compound for brominated flame retardants (BFRs) present in e-waste. The model consists of sub-mechanisms (that include reaction rate constants and thermochemical T-dependent functions) for HBr oxidation, combustion mechanism of C1-C6 species, bromine transformation, and synthesis of Br dioxin-like compounds. Reaction rate parameters were obtained for a large array of reactions that constitute the core of the model. For instance, the obtained activation energies for the initial pathways in the formation of brominated biphenyls reside in the range of ~ 15–45 kJ/mol. Considering oxidation of 5000 ppm MBZ in a plug flow reactor, the model reasonably predicts the temperature-dependent profiles (between 500–1200 oC at atmospheric pressure) of a few PBDD/Fs (i.e., polybrominated dibenzo-p-dioxins)) isomers in reference to limited corresponding experimental measurements. Most Br dioxin-like compounds appear in the narrow temperature window of 600–1000 oC and achieve their highest abundance at molar yields in the range of 1.0–15 mmol/mol MBZ. A high load (100–120 mmol/mol MBZ) of brominated environmentally persistent free radicals (Br-EPFR) emerges and shifts from bromophenoxy radicals to bromocyclopentadienyl radicals around 700 oC. Oxidation of a 2-bromophenol molecule results in the formation of higher yields of Br-toxicants when compared with that of MBZ. The assembled model provides an informed hazards assessment into the potential emission inventories of bromine toxicants in the gas phase at conditions encountered in real scenarios, such as open burning and primitive treatment of e-waste. Via an atomic-base understanding of the complex bromine chemistry and speciation, the model allows the underlying operational conditions that reduce the emission of Br-notorious pollutants to be surveyed and fine-tuned.
This work presents the development of molecular-based mathematical model for the prediction of CO2 solubility in deep eutectic solvents (DESs). First, a comprehensive database containing 1011 CO2 solubility data in various DESs at different temperatures and pressures is established, and the COSMO-RS-derived descriptors of involved hydrogen bond acceptors and hydrogen bond donors of DESs are calculated. Afterwards, the efficiency of the input variables, i.e., temperature, pressure, COSMO-RS-derived descriptors of HBA and HBD as well as their molar ratio, is explored by a qualitative analysis of CO2 solubility in DESs using a simple multiple linear regression model. A machine learning method namely random forest is then employed to develop more accurate nonlinear quantitative structure-property relationship (QSPR) model. Combining the QSPR validation and comparisons with literature-reported models (i.e., COSMO-RS model, traditional thermodynamic models and equations of state methods), the developed QSPR model with COSMO-RS-derived parameters as molecular descriptors is suggested to be able to give reliable predictions of CO2 solubility in DESs and could be used as a useful tool in selecting DESs for CO2 capture processes.
Electric arc furnace dust (EAFD) signifies a major source of recyclable zinc. Most of the zinc load in EAFD exists as zincite (ZnO) and franklinite (ZnFe2O4). The heterogenous mixture of EAFD renders it technologically challenging to extract the valuable zinc content in EAFD via commonly utilized hydrometallurgical and pyrometallurgical operations. Co-pyrolysis of EAFD with halogen-containing polymers (most notably polyvinyl chloride, PVC, and brominated flame retardants, BFRs) is currently deployed as a potent approach in the selective extraction of zinc from EAFD. A robust optimization of this process necessitates acquiring accurate and representative kinetic parameters of involved chemical reactions. Herein, we construct a kinetic model that accounts the surface halogenation of zinc ions in franklinite into zinc halides (ZnCl2/ZnBr2). Governing reaction and activation energies for the dissociative adsorption of hydrogen halides and alkyl halides with a franklinite surface were computed with the DFT + U formalism. Products profiles from the constructed kinetic model are discussed in the context of literature available experimental measurements pertinent to transformation of zinc into zinc halides. The predicted temperature window for the synthesis of surface ZnCl2/ZnBr2 moieties coincides with analogous results inferred from pyrolysis experiments. Uptake of HCl and HBr by franklinite commences at 600 K and 500 K, respectively. The model satisfactorily illustrates chemical phenomena that dictate the mass loss curves in EAFD-PVC/BFRs formulations, most notably dehalogenation of halogenated alkanes, evaporation of zinc halides evaporation, and water evolution.
Electronic waste (e-waste) generation has been growing in volume worldwide, and the diversity of its material composition is increasing. Sustainable management of this material is critical to achieving a circular-economy and minimizing environmental and public health risks. This study's objective was to investigate the use of pyrolysis as a possible technique to recover valuable materials and energy from different components of e-waste as an alternative approach for limiting their disposal to landfills. The study includes investigating the potential environmental impact of thermal processing of e-waste. The mass loss and change in e-waste chemicals during pyrolysis were also considered.
The energy recovery from pyrolysis was made in a horizontal tube furnace under anoxic and isothermal conditions of selected temperatures of 300 oC, 400 oC, and 500 oC. Critical metals that include the rare earth elements and other metals such as In, Co, Li) and valuable metals (Au, Ag, Pt group were recovered from electronic components. Pyrolysis produced liquid and gas mixtures of organic compounds that can be used as fuels. Still, the process also emitted particulate matter and semi-volatile organic products, and the remaining ash contained leachable pollutants.
Furthermore, toxicity characteristics leaching procedure (TCLP) of e-waste and partly oxidized products were conducted to measure the levels of pollutants leached before and after pyrolysis at selected temperatures. TCLP result revealed the presence of heavy metals like As, Cr, Cd, and Pd. Lead was found at 160 mg/L in PCBs leachate, which exceeded the toxicity characteristics (TC) limit of 5 mg/L. Liquid sample analysis from TCLP also showed the presence of C10–C19 components, including benzene. This study's results contribute to the development of practical recycling alternative approaches that could help reduce health risks and environmental problems and recover materials from e-waste. These results will also help assess the hazard risks that workers are exposed to semi-formal recycling centers.
In the present study, adsorption of methylene blue dye in residual agricultural biomass (orange bagasse) was modelled using o machine learning algorithm Random Forest (RF) and compared with the traditional Artificial Neural Networks (ANN) approach. The Machine Learning was performed using Python, a free and open source programming language. The models were built and validated with a combination of 202 independent experiments aimed at separately predicting the final concentration of methylene blue (Cf), adsorption capacity (Q) and adsorbate percentage removal (R%), having as input variables: Temperature, pH, adsorbent dosage, contact time, salinity, initial methylene blue concentration and rotation. The validation process of the models was carried out using the coefficient of determination (R²) and the Mean Squared Error (MSE). According to the obtained results, both RF and ANN models exhibited similar performances, as shown by their respective R² values 0.9739 and 0.9734 for Cf; 0.9932 and 0.9919, for Q; 0.9318 and 0.9257 for R%, as well as their respective MSE values 0.0012 and 0.0016 for Cf; 0.0005 and 0.0007 for Q; 0.0015 and 0.0019 for R%. However, RF stood out due to its capacity to better capture data variation. Finally, it was possible to point out that both methods resulted in models able to satisfactorily predict all three response variables, thereby allowing less experimental effort.
Completely and deeply removed bromide from waste printed circuit boards (WPCBs) is necessary due to their toxicity and carcinogenicity. To achieve this purpose, calcium hydroxide (Ca(OH)2) as a debromination agent was added during pyrolysis process of WPCBs. The results showed that hydrogen bromide (HBr), 4-bromophenol, 2-bromophenol and 2,4-dibromophenol were the main bromide species in pyrolysis products. The Ca(OH)2 plays a significant role for removing HBr and organic bromide, but not affects products yield. Optimal removal efficiency for 4-bromophenol, 2-bromophenol and 2,4-dibromophenol reached 87.5%, 74.6% and 54.5%, respectively. And debromination efficiency was related to the steric hindrance caused by bromide atoms. The Ca(OH)2 can be activated by captured HBr and its thermal decomposition. And the newly-generated calcium bromide and calcium oxide significantly facilitate debromination due to their high surface energy and reactivity. The debromination mechanism was clarified by experiments coupled with computational chemistry: the coordination of bromide and calcium to form [Ph-Br···Ca²⁺] or [Ph-Br···Caatom]. Then, electrons were delivered form bromide atom to Ca²⁺ or Caatom, which resulted in the stretch and weaken the C-Br bond. Hence, the C-Br bond was more easily to break. This work can provide support for designing novel and efficient debromination agents applied for high-temperature system.
This paper reviews the latest research findings on the combined treatment of both electric arc furnace dust (EAFD) and halogenated plastic wastes, mainly polyvinyl chloride (PVC) and brominated flame-retardants (BFRs). EAFD contains heavy metals (Zn, Pb, Fe, Cd, etc.); its disposal using the traditional landfilling method threatens the environment. On the other hand, halogenated plastic wastes accumulate annually at an alarming rate due to their excessive production, consumption, and disposal. PVC, for example, does not decompose naturally; it remains one of the most dangerous plastics, as it contains high proportions of chlorine that is responsible for hazardous emissions of chlorinated organic compounds (dioxins) and hydrochloric acid vapour. Recent research have focused on the combined treatment of PVC/BFRs and EAFD. HCl/HBr acids produced from the pyrolysis of PVC/BFRs can react with the metal oxides in the EAFD to convert them into readily separable metal halides. Alternatively, several researches illustrated the advantages of using additives such as metal oxides during the incineration treatment of waste PVC/BFRs to fix gaseous HCl/HBr, and consequently, EAFD would be considered an excellent and cheap candidate for PVC dechlorination, as well as dehalogenation of other halogenated plastics during thermal recycling processes. In this review we critically discuss literature findings on thermal treatment of PVC/BFR materials under oxidative and pyrolytic environments, typically at temperatures of 200 –900 °C in presence of metal oxides or EAFD. We also discuss the treatment/disposal routes for both waste materials (EAFD and halogenated plastic wastes) and the environmental impact of these disposal options. The review, finally, proposes the research necessary to minimize the hazards of these waste materials; Several future research areas were identified including the need to study the behaviour of real EAFD-plastic waste mixtures under oxidative thermal conditions with focus on both the selective recovery of metals and identification, quantification, and minimization of halogenated organic compounds released during the combined thermal treatment.
Brominated flame retardants (BFRs) are bromine-bearing hydrocarbons added or applied to materials to increase their fire resistance. As thermal treatment and recycling are common disposal methods for BFR-laden objects, it is essential to precisely describe their decomposition chemistry at elevated temperatures pertinent to their thermal recycling. Laboratory-level and pilot-scale investigations have addressed the thermal decomposition of pure BFRs and/or BFR-laden polymers under oxidative and pyrolytic environments, typically at temperatures of 280–900 °C. These studies shed light on the effects of various factors influencing the decomposition behaviour of BFRs such as chemical character, polymer matrix, residence time, bromine input, oxygen concentration, and temperature. Although BFRs decomposition mainly occurs in a condensed phase, gas phase reactions also contribute significantly to the overall decomposition of BFRs. Exposing BFRs to temperatures higher than their melting points results in evaporation. Quantum chemical calculations have served to provide mechanistic and kinetic insights into the chemical phenomena operating in decomposition of BFRs and subsequent emissions of polybrominated dibenzo-p-dioxins and dibenzofurans (PBDD/Fs). Under thermal conditions such as smouldering, municipal waste incineration, pyrolysis, thermal recycling, uncontrolled burning and fires, BFRs degrade and form brominated products of incomplete combustion (BPICs). Thermal degradation of BFRs often proceeds in the presence of bromine atoms which inhibit complete combustion. Major BPICs comprise brominated benzenes and phenols in addition to a wide range of brominated aromatics. Pyrolytic versus oxidative conditions seems to have very little influence on the thermal stability and decomposition behaviour of commonly-deployed BFRs. Thermal degradation of BFRs produces potent precursors to PBDD/Fs. Experimental studies have established inventories of PBDD/F emissions with alarming high yields for many BFRs. Co-combustion of BFRs-containing objects with a chlorine source (e.g. polyvinyl chlorides) results in the emission of significant concentrations of mixed halogenated dibenzo-p-dioxins and dibenzofurans (i.e. PXDD/Fs). Formation of PBDD/Fs from incomplete BFRs decomposition occurs primarily due to the condensations of gas phase precursors, including unaltered structural entities of some BFRs in their own right. Complete destruction of BFRs promotes PBDD/Fs formation via de novo synthesis. Bromination of PBDD/Fs in gas phase reactions is more prevalent if compared with chlorination mechanisms of PCDD/Fs, which is largely dominated by heterogeneous pathways. In uncontrolled burning and in simulated fly ash experiments, a strong correlation between congeners patterns of polybrominated diphenyl ethers (PBDEs) and PBDD/Fs indicate that PBDEs function as direct precursors for PBDD/Fs, even in the de novo synthesis route. In this review, we critically discuss current literature on BFRs thermal decomposition mechanisms; gather information regarding the contribution of homogenous and heterogeneous routes to overall BFRs decomposition; survey all studies pertinent to the emission of PBDD/Fs and their analogous mixed halogenated counterparts from open burning of e-waste, and finally, highlight knowledge gaps and potential directions that warrant further investigations.
Oxygen availability was identified to play a key role in determining product selectivity of tetralin oxidation conducted at constant temperature and pressure in a microfluidic reactor. The current study concerns...
This work is concerned with the development of multivariate calibration models to establish spectrum-composition relationships for the hydrocarbon products in the H-ZSM-5 catalyzed oligomerization of propylene. Regression models based on two multivariate methods were investigated in this work: least squares-support vector machines (LS-SVM) and partial least squares (PLS) regression. The performance of two nonlinear kernels, radial basis function (RBF) and polynomial, is compared with PLS as well as its variant, interval-PLS regression (i-PLSR). For comparing with i-PLSR, the Fourier Transform Infrared (FTIR) spectra of the products served as inputs and the respective C1-C10 concentrations, obtained from Gas Chromatography (GC) were the outputs. The sensitivity of the product distribution to inlet operating conditions was also evaluated through the calibration methods. Spectral clusters having distinct chemical character were identified using principal component analysis (PCA) and hierarchical clustering analysis (HCA) and also used as inputs to the different regression techniques to compare with the full spectrum models. It was found that the best performing spectral regions from i-PLSR had chemical relevance and agreed with findings from HCA, which improved the predictive capabilities significantly. The decreasing order of performance of the chemometric methods evaluated was: LS-SVM-RBF > LS-SVM-Polynomial > i-PLS > PLS. The prediction accuracy of RBF kernel-based LS-SVM regression technique was the highest, indicating its suitability for effective online monitoring of moderately complex processes like acid catalyzed propylene oligomerization.
This work reports on the bromine fixing ability of a typical electric arc furnace dust (EAFD) upon its co-pyrolysis with tetrabormobisphenol (TBBA), the most widely used brominated flame retardant) both experimentally and theoretically following thermodynamic calculations. Experimentally, the following variables were considered in this investigation: EAFD: TBBA mass ratio (1:1, 1:2 and 1:3), pyrolysis heating rate and its final temperature and the effect of the NaCl and KCl presence in the dust. In the thermodynamic analysis the same parameters were studied excluding the heating rate. According to thermodynamic calculations, it was found that almost 100% of bromine, released as HBr during the thermal decomposition of TBBA, can be fixed by EAFD as metal bromides when 1:1 and 1:2 ratios where used. These metal bromides remain mainly in the solid form below 400 °C; above this temperature they commence evaporation leaving the reaction system. At 1:3 ratio almost 10% of the initial bromine contents is released in HBr gaseous form. Experimentally, it was found that about 70% of HBr is captured by EAFD when 1:1 and 1:2 ratios were used at temperatures below 350 °C, however, only 53% were captured when ratio 1:3 was used. At all conditions, the escaped gaseous HBr was as low as 6%. It was also found that high heating rates negatively affected the metal oxides ‘capacity to capture emitted HBr.
The catalytic upgrading of the vapor intermediates from the pyrolysis of brominated acrylonitrile-butadiene-styrene (Br-ABS) was investigated over the Fe/ZSM-5 catalysts in a two-stage fixed bed reactor. Results showed that HZSM-5 and Fe/ZSM-5 catalysts exhibited high catalytic cracking activities, resulting in the increased yield of oil from 62.8 wt% to 64.3 wt% and 66.7 wt%, respectively. When the higher amounts of Fe/ZSM-5 catalysts were applied, the oil yield decreased to 59–61.6 wt%, whereas high amounts of coke were deposited on the catalysts. On the other hand, the higher percentages of the single ring and 2 ring aromatic compounds in the oils was obtained by the Fe/ZSM-5 catalysts, compared to the thermal pyrolysis and the catalytic upgrading by the parent HZSM-5 catalyst. The Fe/ZSM-5 catalysts significantly promoted the formation of styrene monomer and dimer derivatives. It could be proposed that the Fe based materials was in favor of the depolymerization of the polymer matrix, providing the styrene sources for the secondary oligomerization over the parent HZSM-5 catalyst. In addition, the Fe/ZSM-5 catalyst exhibited effective debromination performance, by means of cracking the organobromine compounds and capturing the inorganic bromine in the catalyst. The possible catalytic cracking mechanism over the Fe/ZSM-5 catalyst was discussed.
Hydrogen halides (HCl/HBr) represent the major halogen fragments from thermal decomposition of halogen laden materials; most notably PVC and brominated flame retardants (BFRs). Co-pyrolysis of halogen-containing solid wastes with metal oxides is currently deployed as a main stream strategy to treat the halogen content as well as to recycle the valuable metallic fraction embedded in electric arc furnace dust (EAFD) and e-waste. However, designing an industrial-scale recycling facility necessitates accurate knowledge on mechanistic and thermo-kinetic parameters dictating the interaction between metal oxides and hydrogen halides. In this contribution, we investigate chemical interplay between HCl/HBr and zincite surfaces, as a representative model for structures of zinc oxides in EAFD by using different sets of functionals, unit cell size and energy cut-off. In the first elementary step, dissociative adsorption of the HCl/HBr molecules affords oxyhalides structures (Cl/Br−Zn, H−O) via modest activation barriers. Conversion of oxyhalides structure into zinc halides occurs through two subsequent steps, further dissociative adsorption of HCl/Br over the same surface Zn atom as well as the release of H2O molecule. Evaporation (or desorption of zinc halides molecules) signifies a bottleneck for the overall halogenation of ZnO. Our Simplified kinetic model on the HCl + ZnO system concurs very well with experimentally reported TGA weigh loss profiles on two grounds; accumulation of oxyhalides till ~ 700 K and desorption of ZnCl2 at higher temperatures. Thermo-kinetic and mechanistic aspects reported herein could be useful in the pursuit to design of a large-scale catalytic upgrading unit that operates to extract the valuable zinc loads from EAFD.
Bioreactors and associated bioprocesses are quite complex and non-linear in nature. A small change in initial condition can greatly alter the output product quality. It is pretty difficult at times to model the system mathematically. In this work, the fault detection problem is studied in the context of bioreactors, mainly, a reactor from penicillin production process. It is very important to identify the faults in a live process to avoid the product quality deterioration. We have focused on the process history based methods to identify the faults in a bioreactor. We want to introduce random forest, a powerful machine learning algorithm, to identify several types of faults in a bioreactor. The algorithm is simple, easy to use, shows very good generalization ability without compromising much on the classification accuracies and also an ability to give variable importance as a part of algorithm output. We compared its performance with two popular methods, namely support vector machines and artificial neural networks and found that the overall performance is superior in terms of classification accuracies and generalization ability.
Hydrates always are considered as a threat to petroleum industry due to the operational problems
it can cause. These problems could result in reducing production performance or even production
stoppage for a long time. In this paper, we were intended to develop a LSSVM algorithm for
prognosticating hydrate formation temperature (HFT) in a wide range of natural gas mixtures. A
total number of 279 experimental data points were extracted from open literature to develop the
LSSVM. The input parameters were chosen based on the hydrate structure that each gas species
form. The modeling resulted in a robust algorithm with the squared correlation coefficients (R2)
of 0.9918. Aside from the excellent statistical parameters of the model, comparing proposed
LSSVM with some of conventional correlations showed its supremacy. Particularly in the case
of sour gases with high H2S concentrations, where the model surpasses all correlations and
existing thermodynamic models. For detection of the probable doubtful experimental data, and
applicability of the model, the Leverage statistical approach performed on the data sets. This
algorithm showed that the proposed LSSVM model is statistically valid for HFT prediction and
the almost all the data points are in the applicability domain of the model.
Recycling of printed circuit boards (PCBs) is complicated by the presence of flame retardants containing halogen and phosphorus, as the degradation products of these retardants reduce the quality of the produced gases and liquids. Moreover, during thermal treatment, corrosive and toxic compounds are released and the volatilization of undesirable metals incorporated in the PCB matrix is enhanced. To combat this problem, we investigated the effects of calcium hydroxide (Ca(OH)2) on the thermal decomposition of both phenol and epoxy resin paper-laminated PCBs containing tetrabromobisphenol-A. Pyrolysis experiments revealed a maximum removal of 94 % HBr, 98 % brominated phenols, and 98 % phosphorus from the gaseous and liquid pyrolysis products. In addition, Br-induced metal volatilization was inhibited, improving the recovery amount in the solid fraction. Thermogravimetry–mass spectrometry revealed that Ca(OH)2 enhanced the evolution of phenolic compounds produced from the PCB matrix, mainly below 300 °C, while the fixation of brominated compounds took place primarily above 300 °C.
DSC, TG, and TG-MS techniques were used to investigate the reactivity of PbO and Fe2O3 with HBr from thermal degradation of tetrabromobisphenol A under inert and oxidizing atmospheres. The HBr acted as an excellent brominating agent for PbO and separated Pb as a volatile bromide (79 and 90% in He and He + 5 vol% 02, respectively) from the solid up to 580 degrees C. For Fe2O3, the amount of vaporized bromide was only 20 and 13% under inert and oxidizing atmospheres, respectively. In inert atmosphere the formed char acted as a reducing agent for converting the remaining oxides into metallic forms. For TBBPA + PbO, about 3% of metallic Pb remained in the residue as most of the oxide vaporized below 970 degrees C. The unreacted Fe2O3 underwent progressive reduction into metallic Fe (75%), which remained in the residue. In oxidizing atmosphere, the unreacted PbO vaporized completely, while the Fe2O3 remained unchanged in the residue. The organic char decomposed and vaporized as carbon mono- and di-oxides. Simultaneous TG-MS measurements indicated that the presence of PbO and Fe2O3 strongly accelerated TBBPA degradation and enhanced char formation.
Pyrolysis study of a printed circuit board (FR-4 type, epoxy resin reinforced by glass fibers) under the presence of metal oxide, such as ZnO, Fe2O3, La2O3, CaO and CuO, has been carried out in the view of the emission control of waste electrical and electronic equipment (WEEE) containing brominated flame retardants by metallurgical wastes like slags and dusts. The oxide content is of 2–10 mass% to the printed circuit board on the basis of the stoichiometric reaction of oxide and bromine to the corresponding metal bromide or oxybromide. It has been revealed that the formation of hydrogen bromide and brominated organic compounds is significantly suppressed by the addition of ZnO and La2O3. The bromine fixation ability of various oxides is compared and discussed.
This contribution is focused on the on-site determination of the bromine content in waste electrical and electronic equipment (WEEE), in particular waste plastics from television sets (TV) and personal computer monitors (PC) using a handheld X-ray fluorescence (XRF) device. The described approach allows the examination of samples in regards to the compliance with legal specifications for polybrominated biphenyls (PBBs) and polybrominated diphenyl ethers (PBDEs) directly after disassembling and facilitates the sorting out of plastics with high contents of brominated flame retardants (BFRs). In all, over 3000 pieces of black (TV) and 1600 pieces of grey (PC) plastic waste were analysed with handheld XRF technique for this study. Especially noticeable was the high percentage of pieces with a bromine content of over 50,000 ppm for TV (7%) and PC (39%) waste plastics. The applied method was validated by comparing the data of handheld XRF with results obtained by GC-MS. The results showed the expected and sufficiently accurate correlation between these two methods. It is shown that handheld XRF technique is an effective tool for fast monitoring of large volumes of WEEE plastics in regards to BFRs for on-site measurements.
Due to the health impacts caused by exposures to air pollutants in urban areas, monitoring and forecasting of air quality parameters have become popular as an important topic in atmospheric and environmental research today. The knowledge on the dynamics and complexity of air pollutants behavior has made artificial intelligence models as a useful tool for a more accurate pollutant concentration prediction. This paper focuses on an innovative method of daily air pollution prediction using combination of Support Vector Machine (SVM) as predictor and Partial Least Square (PLS) as a data selection tool based on the measured values of CO concentrations.The CO concentrations of Rey monitoring station in the south of Tehran, from Jan. 2007 to Feb. 2011, have been used to test the effectiveness of this method. The hourly CO concentrations have been predicted using the SVM and the hybrid PLS–SVM models. Similarly, daily CO concentrations have been predicted based on the aforementioned four years measured data. Results demonstrated that both models have good prediction ability; however the hybrid PLS–SVM has better accuracy. In the analysis presented in this paper, statistic estimators including relative mean errors, root mean squared errors and the mean absolute relative error have been employed to compare performances of the models. It has been concluded that the errors decrease after size reduction and coefficients of determination increase from 56 to 81% for SVM model to 65–85% for hybrid PLS–SVM model respectively. Also it was found that the hybrid PLS–SVM model required lower computational time than SVM model as expected, hence supporting the more accurate and faster prediction ability of hybrid PLS–SVM model.
A predictive method, based on quantitative structure-activity relationship (QSAR) techniques, has been developed for liquid viscosities of organic compounds. On the basis of the set of data including viscosity and other 18 physicochemical properties of 116 organic compounds of diverse structure over a viscosity range of 0.197–19.9 mPa·s, two basic models using both multiple linear regression and partial least squares regression have been developed and their prediction abilities compared. The results recommend the traditional multiple linear regression model. The basic model has been developed further to cover about 230 compounds. Required parameters for the predictive model can be readily available or calculated purely from structural information. The prediction result is critically compared with four existing approaches in versatility and reliability. This approach can be used for the reasonably accurate prediction of liquid viscosities for a wide variety of organic compounds based on chemical structure.
Data splitting is the act of partitioning available data into two portions, usually for cross-validatory purposes. One portion of the data is used to develop a predictive model and the other to evaluate the model's performance. This article reviews data splitting in the context of regression. Guidelines for splitting are described, and the merits of predictive assessments derived from data splitting relative to those derived from alternative approaches are discussed.
Random forest (RF) has been proposed on the basis of classification and regression trees (CART) with "ensemble learning" strategy by Breiman in 2001. In this paper, RF is introduced and investigated for electronic tongue (E-tongue) data processing. The experiments were designed for type and brand recognition of orange beverage and Chinese vinegar by an E-tongue with seven potentiometric sensors and an Ag/AgCl reference electrode. Principal component analysis (PCA) was used to visualize the distribution of total samples of each data set. Back propagation neural network (BPNN) and support vector machine (SVM), as comparative methods, were also employed to deal with four data sets. Five-fold cross-validation (CV) with twenty replications was applied during modeling and an external testing set was employed to validate the prediction performance of models. The average correct rates (CR) on CV sets of the four data sets performed by BPNN, SVM and RF were 86.68%, 66.45% and 99.07%, respectively. RF has been proved to outperform BPNN and SVM, and has some advantages in such cases, because it can deal with classification problems of unbalanced, multiclass and small sample data without data preprocessing procedures. These results suggest that RF may be a promising pattern recognition method for E-tongues. (c) 2012 Elsevier B.V. All rights reserved.
Traditionally, active compounds were discovered from natural product extracts by bioassay-guided fractionation, which was with high cost and low efficiency. A well-trained Support Vector Regression (SVR) model based on Mean Impact Value (MIV) was used to identify lead active compounds on inhibiting the proliferation of the HeLa cells in curcuminoids from Curcuma longa L.. Eight constituents possessing the high absolute MIV were identified to have significant cytotoxicity, and the cytotoxic effect of these constituents was partly confirmed by subsequent MTT (3-(4, 5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assays and previous reports. In the dosage range of 0.2-211.2, 0.1-140.2, 0.2-149.9 μM, 50% inhibiting concentrations (IC(50) ) of curcumin, demethoxycurcumin and bisdemethoxycurcumin were 26.99±1.11, 19.90±1.22, 35.51±7.29 μM, respectively. It was demonstrated that our method could successfully identify lead active compounds in curcuminoids from Curcuma longa L. prior to bioassay-guided separation. The use of a SVR model combined with MIV analysis could provide an efficient and economical approach for drug discovery from natural products. © 2013 John Wiley & Sons A/S.
A quantitative structure−property relationship (QSPR) approach was used to develop a predictive model for viscosities of pure organic liquids using a set of 403 compounds that belong to diverse classes of organic chemicals. A pool of 116 descriptors that encode topostructural, topochemical, electrotopological, geometrical, and quantum chemical properties of the organic compounds was used to develop QSPR models, based on the robust Random Forest (RF) regression algorithm. The performance of the algorithm, in terms of correlation coefficients and mean square errors, was determined to be good. The capability of the algorithm to build models and select the most-informative features simultaneously is very useful for several quantitative structure−activity/property relationship tasks. The eight most-dominant features selected by the RF regression algorithm primarily contained predictors that encode characteristics of atoms and groups that form hydrogen bonds, as well as factors involving molecular shape and size.
This study concerns the development of a new system to detect meat and bone meal (MBM) in compound feeds, which will be used to enforce legislation concerning feedstuffs enacted after the European mad cow crisis. Focal plane array near-infrared (NIR) imaging spectroscopy, which collects thousands of spatially resolved spectra in a massively parallel fashion, has been suggested as a more efficient alternative to the current methods, which are tedious and require significant expert human analysis. Chemometric classification strategies have been applied to automate the method and reduce the need for constant expert analysis of the data. In this work the performance of a new method for multivariate classification, support vector machines (SVM), was compared with that of two classical chemometric methods, partial least squares (PLS) and artificial neural networks (ANN), in classifying feed particles as either MBM or vegetal using the spectra from NIR images. While all three methods were able to effectively model the data, SVM was found to perform substantially better than PLS and ANN, exhibiting a much lower rate of false positive detection. Copyright © 2004 John Wiley & Sons, Ltd.