EventsThe 1st International Online Conference on Recycling
Published
This submission belongs to the session S1. Advances in Recycling Technologies of the event The 1st International Online Conference on Recycling
Published date
02 Sep, 2026
Academic Editor
author-avatarHuijuan Dong
Citation
Miriam D. Guimarães, Samuel F. M. Costa, Dinis O. Abranches, Ana M. Ferreira, João A. P. Coutinho, Simão P. Pinho, Olga Ferreira, Mónia A. R. Martins, Dissolution-based recycling of 3D-printed PLA waste using green solvents: a combined computational and experimental approach, in Proceedings of The 1st International Online Conference on Recycling, 7 September–8 September 2026, MDPI: Basel, Switzerland
Share
Email
Facebook
Twitter
LinkedIn

Dissolution-based recycling of 3D-printed PLA waste using green solvents: a combined computational and experimental approach

Miriam D. Guimarães 1,2
Samuel F. M. Costa 1
image
image
1. Department Chemical and Biological Technology, School of Technology and Management, Campus Santa Apolónia, CIMO, LA SusTEC, Polytechnic Institute of Bragança, Bragança 5300-253, Portugal
2. Department of Chemistry, Campus Universitário de Santiago, University of Aveiro, CICECO − Aveiro Institute of Materials, Aveiro 3810-193, Portugal
Abstract

Polylactic acid (PLA) is often held up as a model for sustainable materials design because it is bio-based, biodegradable and increasingly used in additive manufacturing. However, as 3D printing expands, so does the generation of PLA waste, which comes from failed prints, support structures and products that have reached the end of their useful life. Global PLA production is expected to exceed 2.4 million tonnes per year by 2029, and the limitations of existing recycling routes are becoming increasingly apparent. Cycles of melt processing thermally degrade the polymer, compromising mechanical performance.

Given these limitations, this study proposes a dissolution-based recycling route for PLA waste, based on lower-hazard and bio-based solvent systems that enable the selective recovery of polymers and effective separation of additives and copolymers. The original contribution of this work is the integration of thermodynamic modeling with experimental validation and machine learning into a computational-experimental framework that selects solvents while reducing the time and cost associated with conventional trial-and-error approaches. Hansen solubility parameters and the predictive model COSMO-RS were applied to select and classify candidates, covering both conventional solvents and bio-derived alternatives. The most promising candidates underwent experimental validation through PLA dissolution tests conducted under moderate thermal conditions, at temperatures above and below the polymer glass transition temperature.

The experimental results were compiled into a dataset and used to train and evaluate machine learning models. The models classified solvents in agreement with the experimental results and identified new promising candidates beyond those initially selected. The convergence of thermodynamic modelling, experimental evidence, and data-driven prediction establishes a coherent methodology with the potential for wider application in the sustainable recycling of PLA and other polymers, which minimizes the use of hazardous chemicals and aligns with the broader objectives of green chemistry and the circular economy.

Keywords
additive manufacturing
alternative solvents
plastic recycling
machine learning
Extraction and Recovery of High-Value Chemical Compounds from Aqueous Phase Pyrolysis Bio-Oil
Intelligent Prediction of Compressive Strength for Cemented Tailings Backfill Using Bayesian Optimized Machine Learning