Analysis of the polymeric fractions of scrap from mobile phones using laser-induced breakdown spectroscopy: Chemometric applications for better data interpretation
Graphical abstract
Introduction
Currently, the management of obsolete electronic equipment or the scrap originating from these devices, generally named as WEEE (waste electrical and electronic equipment) or e-waste, is of great concern in many countries [1], [2]. Despite the advances arising from the implementation of norms, such as Directives 2002/95/EC [3] and 2002/96/EC [4] (created by the European Union), the different aspects related to the correct destination of e-waste are far from being fully developed, established and implemented in all countries [1], [2], [5], [6].
In Latin American cities, this problem has not reached the same magnitude as observed in other cities such as Lagos (Nigeria), New Delhi (India), Guiyu and Taizhou (China) [1], [2], [7]; nevertheless, it is obvious that the general development of policies on e-waste management in this region is not ideal [1]. In this context, the Brazilian situation illustrates this problem. Brazil is the second largest producer of e-waste among emerging countries and similarly to others, the complete recycling process of the e-waste does not occur properly. The exportation of printed circuit boards from Brazil to countries such as Canada, Belgium and Singapore, where they are recycled for precious metal refining, is an example of this issue [1], [8].
On the other hand, the mechanical recycling of WEEE plastics, which represents around 20–25% by weight of WEEE (around 50% by volume) [9], is the main polymer recycling process performed in Brazil. In this technique the plastics are essentially selected by their type (not solely in Brazil), and aspects like their origin, previous use and possible contaminants are frequently unconsidered for economic or analytical reasons. This practice is undesirable since hazardous substances such as the toxic elements and the brominated flame retardants can remain in the recycled plastics engendering danger to its users or contaminating the environment [10].
Therefore, independent of the development level of the country, actions aimed at dealing with e-waste problems must be conducted during all stages of the electronic device cycle: (1) from its assembly using components either with the absence or low levels of hazardous substances, (2) to its disposal, and (3) finally to its recycling and the prevention of illegal commerce of its generated scraps. Therefore, it is clear that chemical analysis is a critical step in ensuring the success of each of these actions. In this regard, the development of green, fast, simple and inexpensive analytical methods is desirable [11].
The combination of all these attributes into a unique analytical methodology is not an easy task; however, the development of laser-induced breakdown spectroscopy (LIBS) during the past few years has led to major steps in this direction [12], [13], [14]. Some advantages of LIBS include the elimination of expensive gases for plasma formation and the possibility to achieve analysis without sample preparation protocols with minimal amount of material. However, the proposition of calibration procedures for quantitative analysis using this method is not trivial. Despite this challenge, LIBS has been successfully applied in a wide range of analyses (e.g., biomedical, industrial, forensics, environmental, geological and archeological) [12], [13].
In plastics analysis, LIBS associated with chemometric tools has been used to identify the different polymer types found in diverse wastes (including the scraps of electronic devices) as well as to determine the brominated flame retardants (BFRs) and metals in these materials [15], [16], [17], [18], [19]. However, the use of LIBS to build chemometric models based on the spectral data of the polymers employed in electronic device manufacturing, which may assist in the characterization or traceability of this equipment and its scraps, is an application that remains to be further exploited.
Despite their relatively small size, mobile phones are significant contributors to the growth of electronic waste. This occurs in function of their short life span, which is reduced by the high number of models launched with new technologies and by the increasing number of consumers that quickly acquire new devices. Moreover, due to the relatively small size, mobile phones (or their broken components) can easily be added to the domestic waste or improperly exposed to the environment. This kind of disposal poses a serious problem once the hazardous elements or materials such as Ba, Ni, and carbon black [20], [21], [22], [23] used in the assembly of these devices can pollute the air when burnt in incinerators or contaminate the soil and drinking water sources through leaching when buried in landfills [6], [24].
From the analysis of a set of 50 mobile phones, this study aims to demonstrate that the combination of LIBS with chemometric strategies can provide a simple and useful tool (proposition of classification models) for the classification of the origin and manufacturer of polymeric scraps of these devices.
These classification models can be used in circumstances where determination of the traceability of these products is necessary. Examples of these situations are the investigations of illegal WEEE commerce, piracy and smuggling. In addition, this methodology can also be applied in cases where the chemical profile (fingerprint) of these scraps is an important factor for the correct choice of their management.
Section snippets
LIBS set-up and samples
The experiments were performed using an Applied Spectra (Fremont, California – USA) J200 commercial LIBS managed by the Axiom 2.5 software. This instrument is equipped with an ablation chamber with an HEPA air cleaner to purge ablated particles, an automated XYZ stage and a 1280×1024 CMOS color camera imaging system. The Nd:YAG laser was employed at 1064 nm, delivering a maximum of 100 mJ in a single laser pulse with 8 ns duration at a frequency of 10 Hz. The plasma light emission was guided into
Assessment of repeatability from the LIBS measurements
Prior to the sample analysis, the repeatability of the LIBS system was evaluated. For this purpose, spectra of the same polymer piece (randomly chosen from the sample set) were obtained over 10 subsequent days. On each day, 60 spectra were obtained from 20 laser shots at 3 distinct points. From these spectra, 14 emission lines were selected in an effort to cover the entire spectral range (186–1042 nm). The selected lines were (I and II mean atomic and ionic emission lines, respectively): Line
Conclusions
In this study, a procedure combining LIBS and chemometric tools for the analysis of e-waste samples obtained from obsolete or broken mobile phones was described. A strategy for the recording and treatment of data was also presented in which the elements related to the different types of samples were highlighted. Additionally, the proposed classification models demonstrated good predictive ability (excepted for PLS-DA). The analytical method proposed herein extends the applicability of the LIBS
Acknowledgments
The authors are grateful for the financial support of Conselho Nacional de Desenvolvimento Científico e Tecnológico, CNPq (Grants 304772/2012-7 and 474357/2012-0), to Grants 2012/01769-3, 2012/50827-6 and 2013/04688-7 São Paulo Research Foundation (FAPESP) and Thermo Scientific – Analítica for the Thermo iCE 3000 Atomic Absorption Spectrometer.
References (43)
- et al.
Waste Manag.
(2012) - et al.
Sci. Total Environ.
(2013) - et al.
Environ. Impact Assess. Rev.
(2010) - et al.
J. Hazard. Mater.
(2014) - et al.
Talanta
(2014) - et al.
Talanta
(2013) - et al.
Spectrochim. Acta Part B At. Spectrosc.
(2013) - et al.
J. Hazard. Mater.
(2010) - et al.
Environ. Res.
(2013) - et al.
Chemom. Intell. Lab. Syst.
(1987)
Spectrochim. Acta Part B At. Spectrosc.
Anal. Chim. Acta
Waste Manag.
Waste Manag.
Environ. Sci. Technol.
Quim. Nova
Circuit World
Environ. Sci. Technol.
Energy Environ. Sci.
Cited by (49)
Combining spectroscopy and machine learning for rapid identification of plastic waste: Recent developments and future prospects
2023, Journal of Cleaner ProductionRapid origin identification of chrysanthemum morifolium using laser-induced breakdown spectroscopy and chemometrics
2023, Postharvest Biology and TechnologyLaser-induced breakdown spectroscopy as enabling key methodology for inverse production of end-of-life electronics
2021, Spectrochimica Acta - Part B Atomic SpectroscopyCitation Excerpt :Spectra of PCBs (the board itself, not the electronic components mounted on the board) showed signals attributed to Cu, Ba, Si, and Mg. In [4] the same group studied polymeric fractions of scrap from mobile phones using LIBS and chemometric methods with the goal to classify the origin and manufacturer of polymeric scraps of these devices. Classification models based on a K-nearest neighbor approach (KNN) showed the highest degrees of correct classifications of ≥92% for black and white polymers.
Laser induced breakdown spectroscopy for plastic analysis
2021, TrAC - Trends in Analytical ChemistryIsotopic analysis and plasma diagnostics for lithium detection using combined laser ablation–tuneable diode laser absorption spectroscopy and laser-induced breakdown spectroscopy
2021, Spectrochimica Acta - Part B Atomic SpectroscopyCitation Excerpt :These drawbacks can be managed with control of the time delay and width of the acquisition window, but ultimately limit the potential for quantitative analytical data and isotopic information to be obtained from a measurement. Numerous attempts have been made to increase the spectral resolution and improve the detection limits which can be achieved by LIBS: these include the use of multiple laser pulses for the ablation process [9–12]; hardware additions such as spectrometer enhancements [13,14], magnetic fields [15], and electric discharges [16]; chemometric methods [17–22]. Many of chemometric techniques have become commonplace, particularly multivariate data analyses such as Partial Least Squares (PLS) and Principal Component Analysis (PCA).