273: Rapid Detection of Antioxidant Capacity and Total Phenolic Content in Maple Syrup by Raman Spectroscopy and Deep Learning

273: Rapid Detection of Antioxidant Capacity and Total Phenolic Content in Maple Syrup by Raman Spectroscopy and Deep Learning

Monday, July 14, 2025 10:00 AM to Wednesday, July 16, 2025 3:00 PM · 2 days 5 hr. (America/Chicago)
Exhibit Hall A - Posters
Expo OnlyTotal Access Registration

Information

Introduction

The growing consumer interest in food antioxidants and their health benefits has prompted extensive analysis of total phenolic content (TPC) and antioxidant capacity of maple syrup. Traditional chemical methods, including Folin-Ciocalteu (FC), 2,2-diphenyl-1-picrylhydrazyl (DPPH), oxygen radical absorbance capacity (ORAC), and ferric reducing antioxidant power (FRAP) assays, provide reliable results but have significant drawbacks. These methods are less eco-friendly, labor-intensive, and time-consuming, creating a pressing need for simple, rapid, high-throughput techniques for analyzing antioxidant profiles of maple syrup. Raman spectroscopy emerges as a promising alternative, offering rapid, non-destructive analysis by characterizing the vibrational modes of functional groups to generate unique fingerprints of chemical components.

Methods

This study combined Raman spectroscopy with deep learning to determine TPC and antioxidant capacity of maple syrup. Reference measurements were conducted using the FC assay for TPC quantification, while antioxidant capacity was determined through DPPH, ORAC, and FRAP assays. Raman spectra were collected from 36 maple syrup samples using a benchtop Raman spectrometer, and these spectral data were paired with their corresponding reference values to develop and cross-validate four deep learning models. To facilitate on-site analysis, four additional deep-learning models were established and cross-validated using spectra acquired by a portable Raman spectrometer.

Results

The deep learning models developed using Raman spectra acquired by benchtop Raman spectrometer demonstrated good prediction performance (R2 = 0.93, 0.93, 0.94, 0.97 for TPC, FRAP, DPPH, ORAC assays, respectively). In comparison, the models developed using portable Raman spectrometer exhibited comparable predictive capability (R2 = 0.88, 0.88, 0.92, 0.94 for TPC, FRAP, DPPH, ORAC assays, respectively).

Significance

Raman spectroscopy can be used as an effective tool for rapid, high-throughput, on-site analysis of antioxidant profiles of maple syrup.

Authors: Li Xiao, Jinxin Liu, Marti Z. Hua, Xiaonan Lu*

Short Description
This study combined Raman spectroscopy with deep learning to determine total phenolic content and antioxidant capacity of maple syrup. The developed deep learning models showed good prediction performance (R2 >0.88), indicating the potential of using Raman spectroscopy for rapid, high-throughput, on-site analysis of antioxidant profiles of maple syrup.
Track
Nutraceutical & Functional Foods

Log in

See all the content and easy-to-use features by logging in or registering!