Photonic Sensor Detects Soil Nutrients and Microplastics

Photonic sensor data, combined with explainable deep learning, classified soil nutrients and polyethylene microplastics with high accuracy, supporting portable environmental monitoring and precision agriculture following future field-testing validation.

Study: Integrating nano crystal sensor with explainable deep learning for nutrients and microplastic-toxicity detection. Image Credit: narong sutinkham/Shutterstock

A recent study published in Scientific Reports introduced a simulation-based artificial intelligence (AI) sensing framework for detecting soil macronutrients and low-density polyethylene microplastics. The system combined a two-dimensional (2D) dual-ring-cavity photonic crystal sensor with a customized Deep & Cross Network version 2 (DCN v2) model for environmental monitoring and precision agriculture applications. The goal was to address limitations of conventional optical classification and feature engineering methods.

The novel framework showed excellent classification accuracy while maintaining physics-informed analysis. According to researchers, it could support portable monitoring devices for tracking soil quality and microplastic contamination, facilitating rapid on-site assessments.

Limitations of Traditional Analysis Techniques

Traditional soil analysis has heavily relied on wet chemical extraction and laboratory spectroscopy to determine soil composition. Although these methods are accurate, they require extensive sample preparation, generate chemical waste, and are not suitable for rapid field-based measurements. In addition, existing diagnostic techniques often struggle to distinguish weak signals from low-atomic-number materials within complex soil mixtures.

To address these limitations, researchers have explored optical sensing techniques, mainly planar photonic crystal platforms. These structures use periodic refractive-index variations to trap and manipulate light within a photonic bandgap. When environmental conditions or material compositions change, the sensor produces a shift in resonance wavelength, enabling rapid and non-invasive detection of soil characteristics and contaminants.

Functionality of the Photonic Sensor System

Researchers developed a multi-stage framework that combines optical sensor physics with deep learning analysis. They designed a 2D photonic crystal sensor with a compact size of 11.0 μm x 8.85 μm. The structure consists of a square lattice containing 21 x 17 silicon rods embedded in air, with a lattice constant of 540 nm and a rod radius of 108 nm.

The sensing region uses a dual ring defect cavity composed of an outer silicon ring with a radius of 600 nm and an inner germanium core with a radius of 300 nm. The high refractive index of germanium strengthens light-matter interactions inside the cavity. When soil enters the sensor, they modify dielectric properties and shift the optical response. The resulting spectra were analyzed using the finite element method under transverse conditions.

For classification, the DCN v2 model using a parallel dual-tower architecture was implemented. Before training, a mutual information-based feature selection process removed redundant parameters and retained sensitivity, quality factor, figure of merit, power efficiency, sampling wavelength, peak resonance wavelength, and normalized power. The cross network modeled polynomial feature interactions using low-rank matrix factorization, while the deep neural network learned complex nonlinear relationships within the training dataset.

Validation and Interpretability of the AI Model

The sensor demonstrated strong optical and analytical performance during validation. The dual-ring-cavity design achieved a maximum quality factor of 18244 and an optical transmission efficiency of 99.39%. The system also showed sensitivities of 631 nm/RIU for nutrient detection and 432 nm/RIU for low-density polyethylene microplastic sensing. The sensor also reached a figure of merit of 3440 RIU-1 and a detection limit of 0.00003 RIU.

Stability testing under structural deviations of ±10 to 20 nm and temperature variations of ±5°C confirmed reliable operation. On the AI side, the DCN v2 model achieved a classification accuracy of 99.87%, along with near-perfect precision, recall, and F1 score.

To improve interpretability, researchers applied SHAP (SHapley Additive exPlanations) and LIME (Local Interpretable Model-agnostic Explanations) methods. These analyses showed that the model’s decisions were mainly influenced by physically meaningful parameters such as quality factor and power efficiency rather than irrelevant statistical patterns.

Applications of the Nano-Photonic Sensing Framework

The integrated AI framework has important implications in precision agriculture and environmental monitoring. Because the system detects materials through optical resonance shifts instead of large laboratory equipment, it can be integrated with compact spectrometers, miniature laser diodes, and wavelength-selective photodetectors. This allows the development of portable handheld devices for rapid on-site soil nutrient analysis.

Download the PDF of this page here

The platform also supports monitoring of microplastic contamination in agricultural soils and beyond. By detecting low-density polyethylene particles directly within soil samples, the system can help track pollution originating from degraded plastic mulch films and wastewater sources. According to the study, this combined capability could support real-time environmental risk assessment in contaminated agricultural regions.

Future Directions in Intelligent Soil Diagnostics

In summary, the combination of an optimized nano-crystal cavity with an explainable Deep & Cross Network algorithm represents an important advancement in intelligent environmental sensing. By avoiding direct dependence on raw refractive-index inputs, the framework can effectively differentiate highly similar target concentrations with high accuracy, thereby supporting automated multi-species detection and edge-based analysis.

Although this current study is limited to simulation, future work will focus on the physical fabrication of the sensor. Researchers plan to manufacture the silicon-germanium cavity structures using electron-beam and deep-ultraviolet lithography. They will also evaluate the system under realistic environmental conditions, including variations in moisture, ambient noise, and organic contamination, to improve long-term operational reliability.

Journal Reference

Magdy, A., Abd-Elsamee, S. & et al. (2026). Integrating nano crystal sensor with explainable deep learning for nutrients and microplastic-toxicity detection. Sci Rep 16, 15179. DOI: 10.1038/s41598-026-51368-3, https://www.nature.com/articles/s41598-026-51368-3

Disclaimer: The views expressed here are those of the author expressed in their private capacity and do not necessarily represent the views of AZoM.com Limited T/A AZoNetwork the owner and operator of this website. This disclaimer forms part of the Terms and conditions of use of this website.

Muhammad Osama

Written by

Muhammad Osama

Muhammad Osama is a full-time data analytics consultant and freelance technical writer based in Delhi, India. He specializes in transforming complex technical concepts into accessible content. He has a Bachelor of Technology in Mechanical Engineering with specialization in AI & Robotics from Galgotias University, India, and he has extensive experience in technical content writing, data science and analytics, and artificial intelligence.

Citations

Please use one of the following formats to cite this article in your essay, paper or report:

  • APA

    Osama, Muhammad. (2026, May 25). Photonic Sensor Detects Soil Nutrients and Microplastics. AZoSensors. Retrieved on May 26, 2026 from https://www.azosensors.com/news.aspx?newsID=16872.

  • MLA

    Osama, Muhammad. "Photonic Sensor Detects Soil Nutrients and Microplastics". AZoSensors. 26 May 2026. <https://www.azosensors.com/news.aspx?newsID=16872>.

  • Chicago

    Osama, Muhammad. "Photonic Sensor Detects Soil Nutrients and Microplastics". AZoSensors. https://www.azosensors.com/news.aspx?newsID=16872. (accessed May 26, 2026).

  • Harvard

    Osama, Muhammad. 2026. Photonic Sensor Detects Soil Nutrients and Microplastics. AZoSensors, viewed 26 May 2026, https://www.azosensors.com/news.aspx?newsID=16872.

Tell Us What You Think

Do you have a review, update or anything you would like to add to this news story?

Leave your feedback
Your comment type
Submit

Sign in to keep reading

We're committed to providing free access to quality science. By registering and providing insight into your preferences you're joining a community of over 1m science interested individuals and help us to provide you with insightful content whilst keeping our service free.

or

While we only use edited and approved content for Azthena answers, it may on occasions provide incorrect responses. Please confirm any data provided with the related suppliers or authors. We do not provide medical advice, if you search for medical information you must always consult a medical professional before acting on any information provided.

Your questions, but not your email details will be shared with OpenAI and retained for 30 days in accordance with their privacy principles.

Please do not ask questions that use sensitive or confidential information.

Read the full Terms & Conditions.