Native starch granule has a complex crystalline structure. Controlled decrystallization cassava starch (CDCS) was prepared by the alcohol–alkali method. Polarized microscopy, scanning electron microscopy (SEM), X‐ray diffraction (XRD), and Fourier transform infrared spectroscopy (FTIR) analysis of the CDCS are carried out to investigate its structural information. The results show that the relative crystallinity (RC) of the processed cassava starches can be reduced from 25.56% to 17.90% by adjusting the concentration of sodium hydroxide while keeping the integrity of the starch granule. During the decrystallization process, as the sodium hydroxide concentration is increased from 0.25 to 1.25 mol L−1, the polarized light of the starch granules disappear from the inside to the outside, and the X‐ray diffraction peaks of native cassava starch (NCS) disappear sequentially from 17.16°, 23.04°, 15.08°, to 17.83°. On the other hand, the FTIR peaks of the NCS centered at 1645 cm−1 is shifted to about 1600 cm−1, and the intensity of the peaks at 3424, 2929, 1645, and 1017 cm−1 become weaker with decrystallization process. This study provides valuable data and a theoretical basis to understand the mechanism of the controlled decrystallization of cassava starch.

The rosin acid starch (RAS) is prepared under the catalysis of lipase, by esterification on the starch particles' surface in a two-phase reaction system, using rosin acid and native cassava starch as starting materials. The effects of reaction time, solvent types, substrate ratio, reaction temperature, and lipase dosage on the degree of substitution (DS) of RAS are investigated. Based on the single factor experiments, the reaction conditions are optimized according to the principle of the combination Box-behnken centered design and response surface methodology analysis. The mathematical model of the influence of reaction temperature, material ratio, and enzyme addition amount on DS is established. The optimum synthesis conditions are determined as follows: solvent, trichloromethane; reaction time, 10 h; reaction temperature, 59 °C; substrate ratio (m Rosin acid :m starch) , 4.1; and lipase dosage, 264 µL. Under these conditions, it is found that the DS is 0.0608. The analysis of variance and verification experiments show that the mathematical model has good fit degree and high credibility. Compared with the native cassava starch, RAS has improved contact angle, emulsion capacity, and emulsion stability.