Reference database design for the automated analysis of microplastic samples based on Fourier transform infrared (FTIR) spectroscopy.
Primpke S, Wirth M, Lorenz C, Gerdts G. · 2018
Read Reference database design for the automated analysis of microplastic samples based on Fourier transform infrared (FTIR) spectroscopy. by Primpke S, Wirth M, Lorenz C, Gerdts G. (2018) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/europmc-pmc6113679.txt.
Introduction
Primpke and co-authors present a methods article in Analytical and Bioanalytical Chemistry on a reference database for spotting microplastics. Microplastic particles are hard to identify as they get smaller and samples get more mixed. Infrared spectroscopy, called FTIR, is a flexible tool for the task. Automated analysis needs a well-built reference library.
The team designed a database from reference spectra grouped by cluster analysis. It covers the spectral range used to tell polymers apart. The database adapts to single-particle checks and to chemical imaging with FTIR microscopes. Those imaging methods produce large datasets that software must sort. The design aims to make automated identification more reliable across labs.
Catalog record
| CREATOR | Primpke S, Wirth M, Lorenz C, Gerdts G. |
|---|---|
| YEAR | 2018 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 41 KB |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Reference database design for the automated analysis of microplastic samples based on Fourier transform infrared (FTIR) spectroscopy By Sebastian Primpke, Marisa Wirth, Claudia Lorenz, Gunnar Gerdts Published in: Analytical and Bioanalytical Chemistry DOI: 10.1007/s00216-018-1156-x ABSTRACT Abstract The identification of microplastics becomes increasingly challenging with decreasing particle size and increasing sample heterogeneity. The analysis of microplastic samples by Fourier transform infrared (FTIR) spectroscopy is a versatile, bias-free tool to succeed at this task. In this study, we provide an adaptable reference database, which can be applied to single-particle identification as well as methods like chemical imaging based on FTIR microscopy. The large datasets generated by chemical imaging can be further investigated by automated analysis, which does, however, require a carefully designed database. The novel database design is based on the hierarchical cluster analysis of reference spectra in the spectral range from 3600 to 1250 cm−1. The hereby generated database entries were optimized for the automated analysis software with defined reference datasets. The design was further tested for its customizability with additional entries. The final reference database was extensively tested on reference datasets and environmental samples. Data quality by means of correct particle identification and depiction significantly increased compared to that of previous databases, proving the applicability of the concept and highlighting the importance of this work. Our novel database provides a reference point for data comparison with future and previous microplastic studies that are based on different databases. [Figure: ᅟ] Electronic supplementary material The online version of this article (10.1007/s00216-018-1156-x) contains supplementary material, which is available to authorized users. Keywords: Microplastics, Infrared, FTIR, Imaging, Database, Spectroscopy INTRODUCTION The pollution of aquatic systems with small plastic particles called microplastics (MP) [1] is an emerging topic in environmental and analytical science [2, 3]. These particles are defined as < 5 mm in size and often further divided into subcategories, e.g., large MP (5 mm–500 μm) and small MP (500–1 μm) as described by Hidalgo-Ruz et al. [4]. Two introduction pathways for MP into the environment are possible. The first is primary MP, to which the use and disposal of microbeads in cosmetic and cleaning products largely contribute [5]. The second is secondary MP formed by fragmentation of litter by mechanical or UV light-induced degradation. MP are ubiquitous in the environment [6] and their reliable monitoring is demanded within the European Marine Strategy Framework Directive (MSFD) by descriptor 10 [7]. To investigate and monitor MP pollution, it is necessary to identify the particles [2, 4]. One method is the visual identification without further chemical identification, which has a high potential of false counts. If MP are further investigated by chemical identification, up to 70% falsely assigned particles can be found [4]. Therefore, chemical identification is necessary for monitoring and different analytical methods are already in use for MP analysis. To determine the mass of plastic within the sample, mass spectrometry is combined with pyrolysis gas chromatography (Py-GC) [8] or thermal extraction desorption gas chromatography (TED-GC) [9]. Both allow the chemical identification of the polymer types as well as the determination of mass of MP in a sample. Nonetheless, through these processes, the sample is destroyed and particle sizes and numbers cannot be calculated, which is a major drawback, for example, for ecotoxicology studies. In contrast, spectroscopic methods like Fourier transform infrared (FTIR) and Raman spectroscopy enable the measurement of particle numbers and sizes as well as polymer identification. Both methods identify the MP polymers through their molecular vibrations in a complementary manner [10] and can be introduced into microscopic setups, which allows chemical imaging [11–13]. Single-element detectors, which are already frequently applied in MP analysis [14], were used for first setups for chemical imaging [15]. Their major drawback is the high measurement time necessary for large field sizes. The application of focal plane array (FPA) detectors enables fast measurements of large field sizes with high resolution [11], called FTIR imaging. Typically, the (in)organic matrix of environmental samples is reduced prior to measurement by chemical or enzymatic treatment and the residue concentrated onto filters [2, 16]. In earlier studies [13, 17, 18], the complete filter areas were measured followed by an analysis through integration of plastic polymer-specific band regions for the generation of false color images. The hereby pre-selected particles had to be compared via manual comparison to reference spectra, which is a time-consuming task [10, 11, 19] and prone to human bias. Through the development of an automated analysis pipeline [20], it was shown that the expenditure of time and human bias was reduced to a minimum, while large field sizes could be measured. Further, it was found that small microplastic particles, which were previously missed in manual analysis, could now be successfully identified. The automated analysis pipeline uses spectral correlation of the raw and first derivative of vector-normalized spectra against a reference library for chemical identification. Afterwards, both results are compared and if an identical result is found, the pixel is counted as identified. By image analysis, the size and number of particles for each polymer are determined. For all described FTIR-based analyses, the underlying database is crucial for the quality of the results. While different methods are available for data handling [20–22], commercial databases are unsuitable for these methods. Application of the automated analysis on different samples made it clear that for standardized analyses, a specialized database design is necessary to distinguish between different materials in specific spectral ranges [23–25]. In the case of the automated analysis, the spectra have to be sorted into clusters, which are necessary, as no categorization or too fine categorization would lead to errors in the assignment process. Large gaps or several small particles instead of one large particle would be assigned, falsifying the determined particle abundance derived by image analysis. Especially in a large database with several entries of similar materials, this is likely to happen, as the same material can be assigned to different database hits. For standardization of MP analysis, the database should be designed with adaptability for future research questions, while the original design can serve as a reference point for future versions and different spectral ranges and derivatives. In this study, we present a detailed novel approach for an adaptable database design (ADD) for the automated analysis based on statistical methods followed by validation. Therefore, we investigated the typical spectral range (3600–1250 cm−1) for Anodisc [11] filter material regarding differences within the reference spectra by cluster analysis. By manual evaluation of the generated clusters and further validation, an initial reference library in the spectral range 3600–1250 cm−1 was determined for ADD and evaluated, which can serve as a basis for future database adaptations. MATERIALS AND METHODS FTIR measurements To set up a general spectral database, polymer samples from different suppliers were measured via attenuated total reflection (ATR)-FTIR spectroscopy on a Bruker Tensor 27 System (Bruker Optics GmbH) with a diamond platinum ATR-unit (Bruker Optics GmbH). The spectra were recorded in absorbance mode within the range from 4000 to 400 cm−1 with a resolution of 4 cm−1 and 32 scans were co-added. Each measurement was performed in triplicate. Selected materials were additionally measured in transmission mode via a μFTIR microscope (see below) at a resolution of 8 cm−1 with six co-added scans. The