An improved approach for evaluating the semicrystalline lamellae of starch granules by synchrotron SAXS

A fitting method combined with a linear correlation function was developed as an improved approach for the SAXS analysis of the semicrystalline lamellae of starch granules. Using a power-law function with two Gaussian plus Lorentz functions, the SAXS pattern was resolved into sub-patterns of the net lamellar peak and the power-law scattering plus scattering background (PL+B). The ratio of the net lamellar peak area (A_peak) to the total scattering area (A_total) was proposed equal to the proportion of the lamellae within the starch granule (P_SL). Along with this fitting method, we obtained a better profile of linear correlation function, with the elimination of the interference of non-lamellar amorphous starch (i.e., amorphous growth rings). Then, we could accurately calculate the lamellar parameters, e.g., P_SL, the thicknesses of semicrystalline (d), crystalline (d_c) and amorphous (d_a) lamellae, and the volume fraction (φ_c) of crystalline lamellae within semicrystalline lamellae. Quantitative analysis revealed that P_SL was positively correlated with the crystallinity (X_c) of starch. It was confirmed that the distribution of lamellar thickness was more important than the starch botanical origin in affecting the validity of the developed fitting method. We also proposed a criterion to test the validity of the proposed method. Specifically, the total SAXS pattern should be mostly tangent to the profile of PL+B at a high q tail (close to 0.2Å^-1).