Detecting Quality of Green Coffee Using an Electronic Nose System

Abstract

Listen

Translate article

Introduction Electronic noses are widely used in food science such as food freshness detection [1] and fruit ripeness classification [2]. Odor detection is one of the potential methods to identify the quality of coffee. Some researchers have used gas sensors to identify different stages of coffee roasting [3]. Another researcher has employed an electronic nose as a “cup test” tool [4]. In the aforementioned applications, the electronic nose serves to provide an objective solution. The quality identification of green coffee is very important during storage. The green coffee quality is easily affected by water vapor, which causes the green coffee getting mold. The coffee made from mold coffee bean induces bad taste and affects people health. Since the storage environment for green coffee is strict, there is a requirement to find a method to check the green coffee quality during the long storage. This paper presents an electronic nose system that can detect the quality of green coffee. In this study, the Arabica green coffee from different places of origin are used. These different kinds of green coffee are stored in a moist environment to observe the change of the coffee beans. The odor of the green coffee is detected by the electronic nose, and the result shows that the electronic nose has the potential of being a useful tool to discriminate whether the green coffee is suitable for drinking. Method The electronic nose system consists of a flow meter, a three-way valve, a pump and the gas sensor array, which contains 14 metal oxide sensors. The three-way valve controls the constant flow by a flow meter between baseline gas and the target sample. The sensor signals are acquired by a data acquisition card (USB-6343) and recorded with a laptop with National Instrument®LabVIEW Signal Express program. Figure.1(a) depicts the experiment setup.For the experiment sample, the origin Arabica green coffee beans from different places were prepared. The coffee beans were stored in an open and damp environment where the beans would mold over time. To sample the coffee odor, 30g green coffee beans were placed in a 60ml vial, and the vial was sealed for 30 minutes before the experiment. Dry air was used as the baseline for the system. Moreover, an empty vial was used to avoid the influence with and without vial. After the sensor array reached a steady-state, the coffee odor was pumped into the sensor chamber for 1.5 minutes that was long enough for the gas sensor array to react with the sample. To see the coffee beans’ change, the experiments were performed every several days, and the total experiment lasted for two or three weeks depending on the status of the beans. Results In this study, the quality of coffee was divided into fresh, sub-fresh and mold stages. These three stages were evaluated by the professional coffee maker based on the color and the shape of green coffee. Fig. 1(b, c, d) shows the different response curves of these three stages. In the sensor array, the resistance variation rate (ΔRs)/Rs0 is considered as features, where Rs0 represents the resistance of the sensor array in the baseline environment, and ΔRs represents the resistance difference of the sensor array in the baseline environment and the target odor. The odor patterns obtained from sensor responses were analyzed with principal component analysis (PCA), by which a distinctive profile between initial and the following days of storage was obtained. Fig.1(e) shows the PCA result. The PCA result shows a good correlation between the changing of gas composition and the storage time. Conclusions In this paper, the response curves of the green coffee samples with different storage times were obtained using a self-made electronic nose. The PCA results show that the electronic nose could distinguish the fresh green coffee, and the green coffee’s quality decay when stored in an inappropriate environment. The result also shows the electronic nose technology has the potential to meet the demands for rapid, low-cost, and nondestructive detection for the storage of green coffee.