Model for predicting the moisture content of kiwifruit slices during hot air drying based on low-field nuclear magnetic resonance

An experiment has been designed to investigate the drying features of kiwi slices with the thickness of 3, 4, 5 mm, during hot-air drying at different drying temperatures (70, 80 and 90 ℃), in order to understand the internal migration of moisture in the drying kiwi slices for the food production. A low-field nuclear magnetic resonance (LF-NMR) method was used to characterize the internal distribution of moisture and its change rules in the drying kiwi slice during hot-air drying. The moisture in kiwi slices was also determined according to the national standard GB5009.3-2016 “Determination of moisture in foods”. A dynamic model was then established based on the obtained moisture-time data that verified and predicted in this experiment. The results indicated that the hot-air drying process of kiwi slices started with external control step, and then changed to internal diffusion control. The effective diffusion coefficient of moisture ranged from 1.58×10-7 to 4.18×10-7 m2/s, and the diffusion efficiency increased with the increase of temperature. The rise of temperature can significantly increase the drying rate of kiwi slices, and thereby accelerate the migration of combined water, immobilized water, and free water. Specifically, the free water and combined water changed before the immobilized water, and the content of free water gradually decreased in the early stage of drying. In this process, the contents of immobilized water and combined water both presented a trend of first increase, and then decrease. Upon the removal of the free water, the immobilized water and combined water successively reached the maximum. Thereafter, as the drying continued, the immobilized water was gradually removed, while the content of combined water started to decline until the end. The part of free water in kiwi slices was first converted into the immobilized and the combined water, which could be converted into each other, thereby to form a cycle reciprocates with the entire drying process. A multiple linear regression (MLR) model was established to quantitatively detect the moisture content in kiwi slices during the drying process. In this model, the correlation coefficient of prediction (Rp) and root mean square error of prediction (RMSEP) reached 0.981 and 0.51% respectively. In the NMR data, taking the slice thickness and drying temperature as the independent variables, and the moisture content of the kiwi slice as the dependent variable, a multiple linear regression analysis was carried out to establish a dynamic model of moisture content, where, the goodness of fit of the model was 0.982. The results demonstrated that the low-field NMR combined with mathematical model can be used to clarify the hot-air drying process of kiwi slices, in order to achieve rapid and non-destructive detection of moisture content during the drying process. The finding can provide a sound theoretical basis for the hot-air drying process to effectively improve the production design of kiwi slices.