Cotton micronaire is an essential fiber quality attribute that characterizes both fiber maturity and fineness components. Micronaire and other attributes are measured on fiber lint routinely in laboratories under controlled environmental conditions following a well-established high-volume instrument protocol. In this study, the attenuated total reflection Fourier transform infrared spectroscopy, characterizing fundamental group vibrations in fiber cellulose from 4000 to 400 cm−1, and using an attenuated total reflection device, was explored for fiber micronaire assessment, especially for seed cotton locule fibers that were mingled with nonlint materials, and varied in fiber maturity within a naturally variable sample. Partial least squares multivariate regression models and the algorithmic infrared maturity approach were developed and then applied to predict micronaire values of validation samples and independent seed cotton samples for comparison. Unlike partial least squares models that showed worse in the coefficient of determination, bias, and percentage of samples within the 95% agreement range for independent samples than for validation samples, the algorithmic infrared maturity approach indicated a similarity in the coefficient of determination, bias, and percentage of samples within the 95% agreement range between the validation samples and independent samples. In particular, the algorithmic infrared maturity approach avoided the need to re-calibrate the model with new samples. Therefore, the development of a robust and effective Fourier transform infrared technique combined with the infrared maturity approach for rapid laboratory micronaire assessment and distribution demonstrated a great potential for its extension to the early micronaire testing in remote/breeding locations, and also to regular cotton fibers, processed cotton yarns and fabrics.
Read full abstract