Titre de la dissertation: Research on Ternary Semiconductor Metal Oxide Sensor for Cooked Rice Quality Detection at Room Temperature

Promoteur: Madame Carla Bittencourt

Résumé de la dissertation: Rice, as a major staple grain in China, garners significant attention for its flavor and quality. The identified volatile organic compounds (VOCs) in cooked rice include aldehydes, ketones, acids, esters, alcohols, hydrocarbons, and heterocyclic compounds, with aldehydes and alcohols comprising the majority of the flavor profile. Various factors, including rice variety, storage time and temperature, degree of milling, amylose content, soaking time, and cooking method, influence the VOC components in cooked rice. Clarifying the correlation between the VOCs released and the quality of cooked rice can have profound implications for ensuring food security and optimizing strategic stockpiles. Traditional methods for testing rice quality, such as sensory analysis and physical or chemical evaluations, have notable limitations. These include their subjective nature, complex procedures, time-consuming processes, reliance on large equipment, and the need for skilled professionals. Consequently, current technologies fall short in delivering objective, real-time, and rapid testing solutions.With the growth of the agricultural economy and advancements in science and technology, intelligent detection is emerging as a key trend in the monitoring of the edible agricultural products industry. This highlights the urgent need to develop rapid and effective alternative technologies for evaluating the freshness of cooked rice under varying conditions. In this framework, gas sensor technology is gaining prominence in the qualitative and quantitative assessment of agricultural products. Among these, ternary semiconductor metal oxide resistive (TSMOX) sensors are particularly promising due to their low cost, compact size, ease of integration, and high sensitivity. However, the application of TSMOX sensors for detecting cooked rice quality still faces challenges, including high operating temperatures, insufficient detection limits, performance drift, and low accuracy. To address these issues, this thesis investigates VOCs emitted during cooked rice storage and identifies three characteristic VOCs as biomarkers of cooked rice quality. It further designs and develops three innovative room-temperature TSMOX-based gas sensors, explores their working mechanisms, and validates their feasibility.

By linking the odor characteristics of cooked rice (i.e., specific VOCs) to changes in sensor resistance signals, the study seeks to develop a reliable method for evaluating cooked rice quality through effective data analysis. Furthermore, it demonstrates the effectiveness of the room-temperature semiconductor gas sensor array. The ultimate objective is to establish a scientific foundation for the rapid, non-destructive testing of cooked rice quality. The specific research focuses are:

• Flavor generation mechanism and characteristic odor information extraction of cooked rice during storage:

The study explored the mechanisms of flavor generation and extracted characteristic odor information of cooked rice at different stages of freshness during storage. It briefly describes the main causes of cooked rice spoilage, considering factors such as rice variety, environmental conditions, and other contributing elements. The types of VOCs detectable in cooked rice, ranging from freshness indicators to spoilage markers, were summarized along with their biochemical production mechanisms. Specifically, 1-octen-3-ol, benzaldehyde, and nonanal were identified as the key indicators of cooked rice quality. This provides a foundational basis for designing room-temperature semiconductor gas sensors and sensor arrays for quality detection.

• Design and fabrication of a new room-temperature semiconductor nonanal sensor.

La-decorated Bi₂O₂CO₃ (BCO-La) microspheres were synthesized via a facile wet-chemical strategy for low-concentration nonanal sensing at room temperature. The BCO-6La sensor demonstrated significantly higher sensitivity compared to the pure BCO sensor, achieving a response of 174.6 to 30 ppm nonanal. It also exhibited faster response time (36 s) when exposed to 18 ppm of nonanal. Additionally, the sensor showed superior selectivity for nonanal gas detection (approximately 4-24 times higher) compared to interfering gases, including 1-octanol, geranyl acetone, linalool, hexanal, 2-pentyfuran, and 1-octen-3-ol, during cooked rice quality detection. The gas-sensing mechanism and the factors contributing to the enhanced sensing performance of the BCO-La microspheres were demonstrated in a realistic detection scenario.

• Design and fabrication of a new room-temperature semiconductor benzaldehyde sensor.

A highly efficient fabrication technique was introduced to produce carbon-functionalized cladding bismuth tungstate-based benzaldehyde sensors. A 3D twisted micro-flower structure was obtained, and a uniform carbon cladding was applied using a one-pot wet-chemistry strategy. The C_0.75/WO₃/Bi₂WO₆ sensor exhibited excellent benzaldehyde sensing performance across a wide concentration range, with remarkable sensitivity (33.7 @ 50 ppm), stability for over 15 days, and robust selectivity, all at room temperature. Even under high humidity conditions, the response value showed minimal degradation, decreasing by only 8.96% compared to performance in dry conditions. Density functional theory (DFT) calculations revealed that the Cx/WO₃/Bi₂WO₆ sensor exhibited higher benzaldehyde adsorption energy and stronger anti-humidity performance than WO₃/Bi₂WO₆, thereby enhancing its gas-sensing efficacy.

• Defect engineering strategy for 1-octen-3-ol detection using NiWO₄ microstructures:

A defect-engineering strategy was employed to modify the electronic and surface properties of NiWO₄ microstructures through controlled hydrogen reduction for 1-octen-3-ol detection. Systematic annealing protocols induced a high density of oxygen vacancy (O_V) via selective oxygen removal, as confirmed by structural and spectroscopic characterization. The engineered O_V enhanced bulk electron transport by improving conductivity and boosted surface reactivity by optimizing adsorption sites. The optimized NiWO₄ sensor demonstrated exceptional performance for 1-octen-3-ol detection at 40 ppm, with a high response value (Ra/Rg = 20.6), a fast response time (22 s), and strong resilience to relative humidity (19.7% response drift across 20-80% RH). The enhanced sensing performance was explained using an oxygen adsorption model and molecular dynamics (MD) simulations.

• Development of a gas sensor array system for cooked rice quality evaluation:

A high-sensitivity, highly selective gas sensor array was developed to evaluate cooked rice quality accurately. CuO/Bi2O2CO3 (Cu-BC) p-n heterostructure micro-flowers were used as substrates to create a four-channel gas sensor array that converts voltage signals into resistance. Individual Cu_x-BC sensors (x = 10, 20, 30, and 40) exhibited strong responses to nonanal, benzaldehyde, and 1-octen-3-ol, with detection specificity confirmed through principal component analysis (PCA) and linear discriminant analysis (LDA). This integrated gas sensor array effectively identified and distinguished the quality of cooked rice, ranging from freshly prepared rice to rice stored for 1 to 6 weeks. The sensing mechanism relies on increased O_V and improved electron mobility. Additionally, MD simulations were used to explore the adsorption and diffusion mechanisms of target gas molecules and oxygen on the materials’ surfaces.

The comprehensive research of this thesis lays the groundwork for developing advanced gas-sensing technologies that enable rapid, non-destructive evaluation of cooked rice quality.