Advances in gas sensors and electronic nose technologies for agricultural cycle applications

Highlights

• Gas sensors for the cultivation preparation applications are reviewed.

• Gas sensors for the crop production applications are reviewed.

• Gas sensors for the harvesting and storage applications are reviewed.

Abstract

Agricultural cycle is the annual cycle of activities for planting and harvesting. The global warming and climate change currently cause serious impacts on the agricultural cycle leading to reduced crop quantity and quality. In the past decade, the gas sensors and electronic nose (E-nose) technologies have shown great promise and utility in monitoring and prediction of important parameters related to the growth and harvest of a crop. This review summarizes overall the evolution of utilizing gas sensors and E-nose technologies for guideline of agricultural best management practices including applications on cultivation preparation, crop production, harvesting and storage of crops. Advantages and limitations of these technologies in agricultural applications are reviewed to highlight potential research directions.

Introduction

Gas sensor technology has made great advances in the past few decades and is expected to achieve even more. Conceptual approach of gas sensor development from the past to the present focuses on creating and inventing new gas sensing instruments to overcome some of the most common life problems/pain points using odor detection such as food and beverage contamination (Bonah et al., 2020, Wongchoosuk et al., 2010, Timsorn and Wongchoosuk, 2019, Viejo et al., 2020, Traiwatcharanon et al., 2020), air pollution (Ahmad et al., 2020, Wongchoosuk et al., 2014, Chaloeipote et al., 2021a, Arunragsa et al., 2020), human health problems (Binson et al., 2021, Xu et al., 2018a) as well as safety and security (Bakar et al., 2018, Jha et al., 2019, Adib et al., 2018). These research output will help to increase convenience, safety and improve human life quality. Although, several spectrometry-based methods such as mass and ion mobility spectrometry (Zheng et al., 2021), gas chromatography (Zhenhe et al., 2020), and infrared spectrometry (Junaedi et al., 2021) have been successfully to identify and analyze odors/ volatile organic compounds (VOCs) molecules, these instruments are very expensive, time-consuming processes, complex, and bulky. An alternative method in arrays or as individual sensors based on metal oxide semiconductor (MOS), conductive polymer, electrochemical, quartz crystal microbalances (QCMs), or micro/nanocomposites gas sensors that offer easy operation, rapid, real-time detection, low operation cost, reliable, non-destructive measurement has become of great interest in recent years.

The evolution of gas sensor technology in terms of design and construction started with a single-gas detector and continued to develop until it was a gas-sensor array system known as electronic nose (E-nose) or also called as artificial olfactory system (AOS) or machine olfaction. In 1982, Krishna Persaud and George Dodd from University of Warwick published pioneering research on “Analysis of discrimination mechanisms in the mammalian olfactory system using a model nose” that provide researchers with valuable knowledge on array of heterogeneous chemical gas sensors for E-nose system (Persaud et al., 1982). Approximately a decade later, the first commercial E-nose has emerged (Gardner et al., 1994). The basic working principle of E-nose is based on the human olfactory system that consists of three main parts; (I) sample delivery part, (II) detection system and (III) computing system that can be learned, recognized and distinguished odors (Shurmer et al., 1990, Shurmer and Gardner, 1992, Wongchoosuk et al., 2009). For the sample delivery part, the sample is injected into a clean air carrier to the instrument. The direct diffusion of sample into sensor chamber is also usually found. After sensor signals reach to equilibrium stage or a fixed time, the clean air/inert gases are served as gas flushing to clean the sensor response into the baseline values. In the detection part, there are multiple gas sensors embedded in the sensor chamber to interact with various types of aroma molecules at room/high operating temperature. The computing system includes the pattern recognition algorithms to eliminate the background interferences, analyze the electrical signal data and identify the desired aroma (Zhang et al., 2013, Mirshahi et al., 2017). Various statistical techniques such as principal component analysis, (PCA) (Bedoui et al., 2018, Dittrich et al., 2021), artificial neural network (ANN) (Fu et al., 2007, Zou and Lv, 2020, Timsorn et al., 2016, Sun et al., 2021), support vector machine (SVM) (Timsorn et al., 2017, Kok et al., 2021) and cluster analysis (CA) (Falasconi et al., 2012, Xu et al., 2020) have been usually employed. Based on these three main parts of E-nose system, the gas sensors in detection system may be considered as the crucial component that contributes to the response of target odors in various E-nose applications (Zohora et al., 2013, Nazemi et al., 2019). For this reason, researchers worldwide have been trying to investigate a new gas sensor array leading to the enhanced performances and high accuracy of E-nose system that can be applied to several different sectors (Karakaya et al., 2020). The typical resistive-type gas sensor structure consists of substrate, electrodes, and sensing films. The heater may be integrated into sensor structure in case of MOS sensing materials due to their working principle on high operating temperature. In case of other gas sensors such as electrochemical sensors, they include membrane to specify gas/VOCs diffusion and electrolyte for the electrochemical reaction between the sensitive elements and the detected gas/VOCs molecules to generate electrical signals. The type of sensing material plays an important role for design circuits and data analysis for E-nose system. For practical application in agriculture, the gas sensors in an E-nose should be high sensitivity to various functional groups, high stability, less effect on temperature and humidity, and wide range of detection.

The discovery of graphene in 2004 and its exceptional ability to detect even a single gas molecules attracted worldwide efforts to explore sensing nanomaterials for gas sensing applications (Schedin et al., 2007). The bar diagram in Fig. 1, depicts the exponentially increasing number of publications over the years. In 2004, only 9 articles were published on nanomaterial-based gas sensors and the number of publications rises to 186 articles in 2020. To normalize the number of nanomaterial-based gas sensors publications against number of papers based on agriculture sensors (searching by using keyword “agriculture sensor”) in the same year, the ratio of papers contribution on nanomaterial-based gas sensors is currently in a range of 10–12%. It should be noted that the MOS gas sensors own micro/nano structured gas sensors. The E-nose based on MOS gas sensors could potentially be used agriculture robotics, pattern recognition, quality control, gas chromatography and air quality applications as depicted by the network in Fig. 2. Moreover, there have been several researches about the integration of gas sensors with other technologies, including wearable e-textiles technology (Lee et al., 2021, Li et al., 2019, Seesaard et al., 2015, Yun et al., 2015) and wireless communication technology (Park et al., 2016, Deng et al., 2016, Chiou and Wu, 2017). However, until now, a comprehensive review of the E-nose technologies based on gas sensors for applications of agricultural cycle is still limited.

This paper provides a review of the recent advances in the development of gas sensors including MOS, conducting polymer, nanomaterials in the E-nose technologies related to the agricultural cycle. The topics can be classified according to three main sections behind smart farm management solutions include: (I) cultivation preparation section, (II) crop production section and (III) harvesting and storage section. Firstly, gas sensors for the cultivation preparation section have been presented with soil quality assessment, adding organic fertilizers and water management. Secondary, gas sensors for the crop production section were used to anticipate and prevent damage from diseases and pests. Additionally, gas sensors were employed to elevate the quality of products to be safe from chemical contamination in the crop production process. Finally, gas sensors for the harvesting and storage section have been revealed in context of the development of real-time odor detection storage containers to reduce post-harvest losses and improved quality of products for maturation process monitoring such as fruit and vegetables stored in different conditions and also extended shelf life and transportability of food supplies. Most of this content will be placing more emphasis on depth of using gas sensing technologies to reduce production costs and increase crop yield and preventive maintenance and handling. In addition, a combination of gas sensors device and internet of things (IoT) can be improved farm management using digital technology. The schematic diagram as shown in Fig. 3 provides an overview of the entire literature review of this article.