Automated Machine-Learning-Driven Analysis of Microplastics by TGA-FTIR for Enhanced Identification and Quantification
Bibliographic record
Abstract
Microplastics persist as ubiquitous environmental contaminants, and efficient methods to quantify and identify their presence are essential for assessing their environmental and health impacts. Common identification approaches typically fall under either vibrational spectroscopy or thermoanalytical techniques with thermogravimetric analysis (TGA) coupled with Fourier transform infrared (FTIR) spectroscopy bridging the intersection. Despite its potential, TGA-FTIR remains relatively underutilized for microplastic analysis, even though each thermogram is associated with approximately 200 FTIR spectra that can be rapidly assessed with targeted automated data analysis. This work explores the development of data analysis routines specialized in identifying plastic components from TGA-FTIR. A dedicated spectral library and a matching algorithm were created to identify polymers from their gas-phase FTIR spectra. The approach was further enhanced by utilizing machine learning (ML) classification techniques, including k-nearest neighbor, random forest, support vector classifier, and multilayer perceptron. The performance of these classifiers for complex data sets was evaluated using synthetic data sets generated from the spectral library. ML techniques offered precise and unambiguous identification compared with a custom spectral matching algorithm. By correlating polymer identities with mass loss in the thermogram, this approach combines qualitative insights with semiquantitative analysis, enabling a streamlined assessment of plastic content in samples.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".