The configuration and packing of the chain molecules of native starch as derived from X-ray diffraction of part of a single starch grain

Abstract

To elucidate the crystal structure of native starch, an X-ray fibre diagram of the substance is indispensable. On account of the smallness of the starch grains and the radial orientation of their crystallites, the usual techniques for X-ray diffraction are unsuitable for obtaining such a diagram from the grains. For that purpose a micro-technique enabling part of a grain to be irradiated by a narrow X-ray pencil is required. By means of a new micro-method for X-ray diffraction 3, 6—not described in this communication —a fibre diagram of a native starch grain could be obtained for the first time from part of a starch grain of the large type occurring in the orchid Phajus grandifolius. This result and the conclusions drawn from the diagram were published briefly in earlier papers 1, 2. The present paper is a more detailed account, and at the end an answer is given to the remarks made by Rundle 5 with regard to the above-mentioned papers. The fibre diagram obtained yields conclusive evidence that in starch grains, according to the assumptions made on the basis of their optical behaviour, the chain molecules are oriented radially. The powder diagram of Phajus starch is almost identical with that of potato starch. Its crystalline modification is therefore the B-modification according to Katz and Van Itallie 7. The fibre diagram obtained may be indexed provisionally on the basis of an orthorhombic unit cell with axes:: a = 9.0 A ̊ , b ( fibre period) = 10.6 A ̊ , c = 15.6 A ̊ . The dimensions of a and c axes are slightly different from those calculated by Rundle et al. 4 from a fibre diagram obtained from artificial threads of the amylose fraction of native starch yielding the same—but possibly on account of this difference not quite the same—X-ray spacings as native starch in the B-modification. The difficulty of understanding the fibre period from a chain of α-glucose residues, already mentioned by Rundle et al. 4, has been discussed. It appears that the period can be understood without any difficulty from a spiral chain with three α-glucose residues per turn, whereas important objections can be raised against the configuration with two glucose residues per period proposed before by Rundle et al. 4, and independently by the present author in 1946 1. Also spiral configurations for starch with six glucose residues per turn, as proposed by Hanes 15, Freudenberg et al. 14 and Caesar and Cushing 13 cannot be present in native starch in the B-modification. It is shown that four different spiral configurations of chains with three α- d-glucose residues per turn can be built up. A spiral configuration with three glucose residues per turn not only explains the fibre period, but also the hexagonal ratio of the axes in the basal plane of the provisional unit cell and the intrinsic birefringence of starch grains mentioned by Frey-Wyssling 17. The spiral chains cannot be packed in the provisional unit cell in such a manner that the X-ray density is consistent with observed experimental densities of native starch. Therefore, and on account of the hexagonal ratio of the axes in the basal plane, a hexagonal unit cell is suggested with a = 18 A ̊ and b = 10.6 A ̊ , through which 18 chains are running; i.e. the cell contains 54 glucose residues and water molecules. The X-ray density of such a packing is consistent with observed densities of starch grains.