Rapid moisture determination for cup lump natural rubber by near infrared spectroscopy

• Main Authors: Sunisa Suchat, Parichat Theanjumpol, Seppo Karrila
• Format: Journal
• Published: 2018
• Subjects: Agricultural and Biological Sciences
• Online Access: https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84939639654&origin=inward; http://cmuir.cmu.ac.th/jspui/handle/6653943832/43989
• Institution: Chiang Mai University

© 2015. Natural rubber (Hevea brasiliensis) is an important economic crop in South-East Asia. The price paid for cup lump raw rubber depends principally on its moisture content (MC), which an examiner subjectively estimates visually and by touch. These inaccurate and subjective estimates are disputable and cause contentions. This study aimed to develop a near infrared spectroscopy (NIRS) method calibrated for rapid, accurate, inexpensive, and nondestructive moisture content determination of cup lump. We used NIRSystem 6500 to measure the reflectance spectra, and developed calibration equations for the moisture content by partial least squares (PLS) regression, restricted to that part of the spectrum known to depend on water. The calibration equation for moisture content had the coefficient of determination equal to 0.98, the standard error of calibration 1.68%, the standard error of prediction 1.48%, and an average difference between the actual and predicted value of 0.02%. The actual and the predicted values did not significantly differ at a 98% confidence level. The fitted model was then validated with a new batch of samples, by a comparison of NIR-based moisture content estimates to standard laboratory determinations. NIRS is a well-established technique for determining the moisture content in cup lump, but this study demonstrated its suitability for rapid determinations without preprocessing the samples, provided a fixed procedure is used in coagulation to form cup lump. However, the transfer of these techniques to field use will require technology development to ensure robust device performance.