Prediction Of Some Quality Properties Of Rice And Its Flour By Near-Infrared Spectroscopy (NIRS) Analysis

Introduction

Rice's market value is greatly affected by its post-milling quality traits. Essential quality metrics, including amylose content (AC), gelatinization temperature, gel consistency, and pasting properties, are pivotal for determining the ultimate quality of rice. Near-Infrared Spectroscopy (NIRS) has emerged as a swift and prognostic technique for evaluating these quality metrics, providing a more efficient and economical solution than conventional measurements. NIRS is noted for its straightforward operation and rapid sample processing, unlike older analytical methods. NIRS-based calibration models have shown proficiency in predicting rice's protein content (PC). The accuracy of NIRS in forecasting rice's AC has been validated. Research indicates that NIRS's assessment of rice quality traits correlate strongly with moisture content and PC in brown rice, as well as with breakdown and peak viscosity in milled rice. NIRS has also been utilized in determining rice starch quality attributes, demonstrating its capability as a quick method for estimating PC combinations. Additionally, NIRS has been applied in measuring various rice quality parameters, such as AC, pasting measures like breakdown viscosity (BDV), setback viscosity (SBV), and pasting temperature, with notable predictive precision. White rice quality indices are especially important for consumer sales. Conventional quality testing in the food sector is time-consuming and expensive, necessitating specific equipment. In contrast, NIR spectroscopy provides a fast and non-invasive method to assess rice quality, concentrating on its chemical and physicochemical characteristics.

Method

The analysis of rice flour was carried out using a Near-Infrared (NIR) scanning device set to reflectance mode. The rice flour sample was placed into a designated container for scanning, which gathered spectral data indicative of the sample's characteristics. This data was then processed to represent the inverse of light reflectance, yielding a spectrum that detailed the sample's unique properties. To enhance precision, the spectrum was averaged out over multiple scans. This non-destructive testing method was performed at standard room temperature, ensuring consistent and accurate assessment of the rice flour's quality attributes. The reflectance readings were taken at intervals of 6.5 nm across a NIR wavelength range from 870 to 2450 nm, and the data was logged as Log(1/R).

Figure 1: The absorbance spectra of rice and its flour specimens.

Figure 1 illustrates that the primary absorption bands were detected at the following wavelengths: 997, 1199, 1457, 1566, 1752, 1938, 2100, 2180 and 2295 nm. In a similar pattern, grain samples showed absorption bands at 998, 1199, 1456, 1573, 1750, 1940, 2094, and 2200 nm. The Near-Infrared Spectroscopy (NIRS) range of 870–2450 nm revealed bands at 997 and 998 nm, which are associated with the second overtone of O-H or NH2 groups. Additionally, bands at 1199 nm correspond to the second overtone of NH2 or O-H groups, while 1457, 1566 and 1573 nm are linked to the first overtone of O-H and N-H groups. The first overtone of C-H groups was noted at 1752 nm. The combinations of O-H and N-H groups were seen at 1938, 2094, 2100 and 2180 nm, and the C-H group combinations were identified at 2295 nm.

The study identified that the absorption peaks observed at the wavelengths of 1566 and 2100 nm are indicative of the Setback Viscosity (SBV). Additionally, the absorption peak at the wavelength of 997 nm is associated with the Breakdown Viscosity (BDV). Furthermore, the primary absorption bands located at wavelengths of 1,199 nm, 1750 nm, 2094 nm, and 2295 nm are linked to the Protein Content (PC) of the rice.

Measurement system

Figure 2: NIRQuest spectrometer.

The spectral data for rice and its flour samples were gathered using the NIRQuest 256 spectrometer. This spectrometer is specifically designed for Near-Infrared (NIR) analysis, delivering high precision and dependability. Developed by Ocean Optics, this device is fitted with a sensitive InGaAs detector, which is essential for acquiring premium spectral data for a broad spectrum of uses. Its small form factor and adaptability make it exceptionally suitable for incorporation into a variety of measurement configurations, notably for the quality evaluation in the agricultural industry, with a focus on rice analysis.

The NIRQuest 256 spectrometer is exceptional in providing swift, non-invasive assessments, which permits the immediate appraisal of rice quality indicators, such as amylose content and protein levels, with extraordinary precision. This feature is vital for guaranteeing the uniformity and superiority of rice products in the contemporary marketplace.

Illumination during the analysis was provided by standard incandescent bulbs. The samples were positioned to ensure that the inner section of a cylindrical probe was in direct alignment with the light source. This arrangement allowed for the efficient reflection of light through the probe into an optical fiber, enabling the collection of reflected light data from the sample surfaces under varying conditions, which was then thoroughly analyzed to assess the quality characteristics of the rice.

Result

Figure 3: The scatter plots of predicted and measured parameters of rice and its flour.

The scatter plots from Figures 3a to 3h offer a visual comparison between the predicted and actual values of the chemical and physicochemical properties of rice flour and grain samples. The dots' closeness to the target line indicates that Near-Infrared Spectroscopy (NIRS) can accurately predict these properties, showing a strong correlation with the actual measurements. The results underscore the method's robustness in correlating predicted values with actual ones for key quality parameters of rice.

The performance of the NIR spectroscopy technique is highlighted by its precise predictions for critical quality indicators such as amylose and protein content, as well as viscosity measures. For instance, the amylose content predictions in flour were highly accurate, and the protein content predictions were even more precise. Similarly, the viscosity readings in rice were impressively close to the actual values. These high correlation coefficients, all surpassing 0.8, affirm the reliability of NIR spectroscopy in assessing rice quality.

Furthermore, the Standard Error of Prediction (SEP) values fell within acceptable ranges, validating the accuracy of the NIR spectroscopy method utilized in this study. The data strongly supports the efficacy of NIR spectroscopy in evaluating the quality attributes of rice and its flour, showcasing the strength of the predictive models used. This evidence points to NIR spectroscopy as a suitable and reliable method for rice quality analysis.

Conclusion

Traditional rice quality evaluation methods typically employed in food industry labs are not only costly but also time-intensive, needing specific testing equipment. However, Near-Infrared Spectroscopy (NIRS) offers a nondestructive approach to accurately measure the chemical and physicochemical characteristics of various rice flour and grain types. This study's findings indicate that NIRS is effective in forecasting Amylose Content (AC), Protein Content (PC), Brown Rice Density Value (BDV), and Starch Branching Value (SBV). Furthermore, the calibration models' reliability was confirmed through statistical analysis. Current research underscores the suitability of NIRS for assessing the quality parameters of both rice and its flour.