Breaking the "Greenwashing" Cloud: "AI + Analytical Chemistry" Identifies the Authenticity of Recycled Plastics

When the "dual carbon" strategy has become a global consensus, recycled plastics have fallen into the quagmire of "bad money driving out good": counterfeit virgin materials, certificate fraud, trust collapse... These issues have severely disrupted industry order and damaged the credibility of the green industry. How to scientifically and efficiently identify recycled plastics has become a pressing technical challenge for the industry.

Recently, Fang Yimin, the application manager for recycled plastics at Zhili Technology (Guangdong) Co., Ltd., delivered a speech on the topic of "Application of Recycled Plastic Identification Technology." He systematically introduced a set of identification systems that integrate analytical chemistry and artificial intelligence, providing a reliable technical pathway to address the chaos in the industry.

Fang YiminRecycled Plastic Application Manager at Zhili Technology (Guangdong) Co., Ltd.

— 1 —

Technological Breakthrough: From "Hard to Distinguish" to "Accurate Identification"

Traditional methods find it difficult to distinguish between native and recycled plastics that have a highly similar appearance. However, Zhili Technology has developed an efficient and accurate identification system through an interdisciplinary integration of "analytical chemistry + machine learning."

Interdisciplinary Integration: Analytical Chemistry × Machine Learning

Import data from more than 20 high-end instruments such as DSC, Raman, UV-vis, HS-GC-MS, HPLC, etc., into AI models like random forest, support vector machine, and logistic regression to establish the industry's largest "recycled plastic fingerprint database."

2. Three major optimizations to leave no place for false information to hide.

Solvent optimization: comparative analysis of several extraction reagents, UV spectral characteristic differences magnified 8 times.

Time Optimization: Dynamic Extraction Curve, 32% Improvement in Recycled Material Signal, Zero Interference in Virgin Material.

Algorithm Optimization: Second-order Derivative + Smoothing Denoising, 0.3 seconds to complete the determination of one spectrum, accuracy ≥98%.

3. Authoritative Recognition

In the 2023 blind sample assessment, 21 fiber samples were tested with an accuracy of 100%.

In 2025, it was continuously cited by the University of Manchester in the UK and the University of Murcia in Spain, becoming the international forefront "Chinese solution."

— 2 —

Five major scenarios implemented to promote the rebuilding of trust in the industry.

1. Standard Construction: Establishing a Technical Benchmark for the Industry

Zhili Technology, entrusted by the National Standardization Technical Committees and industry associations, has taken the lead in formulating the group standards "Identification of Recycled Plastics by Headspace Gas Chromatography-Mass Spectrometry" and "Identification Methods for Virgin and Recycled PET/PP".National Recommended Standard "Plastics - Identification of Recycled Plastic Components Part 1: Polyethylene Terephthalate (PET) Materials", "Plastics - Identification of Recycled Plastic Components Part 2: Polypropylene (PP) Materials"Block the institutional loopholes of "fake regeneration" from the source.

2. Enterprise Empowerment: Building Autonomous Identification Capability

Zhili Technology provides "Identification Training + Cloud Database" services for the demand of an international brand to authenticate each batch of recycled PET, enabling it to have independent identification capabilities and significantly improving supply chain management efficiency.

3. Fiber Anti-Counterfeiting: Defending the Reputation of "China Recycled"

Facing the challenge of "pseudo-recycling" impacting domestic recycled polyester fibers, Zhili Technology utilized machine learning models to complete multiple batches of fiber and fabric blind sample verifications, achieving accuracy rates that meet customer requirements and helping Chinese companies gain international trust.

4. Food Contact Materials: "Verification" Ensures Safety

Provide systematic sampling inspection services for a leading food brand using PET, PE, and PP materials to ensure that no recycled plastics are mixed into food contact materials, safeguarding the baseline of food safety.

5. Identification of Daily Chemical Products: Supporting the Credibility of Green Claims

Conduct identification tests on personal care products claiming "100% recycled" to provide technical endorsement for brands and enhance market integrity.

— 3 —

Chili Technology, next steps plan

1. Optimize the "semi-quantitative identification method" and accept blind sample verification from enterprises.Cross-validate using multiple methods to ensure quantitative/semi-quantitative accuracy, and plan to apply for the establishment of a group standard in the second half of 2025.）

2. Improve the identification database in the textile field to fully prepare for the revision of the national standard for the identification of recycled polyester (PET) fibers to be launched.Add more representative samples of fibers, textiles, and fabrics.）

3. Complete the optimization of decolorization pretreatment methods in identification experiments to expand the sample applicability range of identification services.Solve the problem of identifying recycled plastics in colored products, with plans to complete by the second half of 2025.）