## Abstract

The crystal structure of polyethylene terephthalate (Terylene) has been determined by the interpretation of the X-ray diffraction patterns given by drawn fibres. The unit cell, which contains one chemical unit, $\chembond{1,0} $CO.C$_{6}$H$_{4}$.CO.O.(CH$_{2}$)$_{2}$O$\chembond{1,0} $, is triclinic, with a = 4$\cdot $56 angstrom, b = 5$\cdot $94 angstrom, c = 10$\cdot $75 angstrom, $\alpha $ = 98$\frac{1}{2}$ degrees, $\beta $ = 118 degrees, $\gamma $ = 112 degrees. This gives the density of the crystals as 1$\cdot $455 g/cm$^{3}$. If it is assumed that the density of amorphous regions in partly crystalline material is the same as that of the entirely amorphous material (1$\cdot $335), the proportion by weight of crystals in ordinary drawn yarn (density 1$\cdot $39) is 48%, and in specimens of the highest recorded density (1$\cdot $41) it is 64$\cdot $5%. These figures are, however, lower limits; if, as is likely, the density of amorphous regions in partly crystalline specimens is lower than 1$\cdot $335, the proportions of crystalline material are higher than the figures given. The positions of atoms in the crystals have been deduced from the relative intensities of X-ray diffractions; approximate positions were found by trial-and-error methods, and refinement was then effected by calculations of electron densities, using the three-dimensional Fourier synthesis method. The molecules are nearly planar in configuration. Distances between atoms in neighbouring molecules are all normal Van der Waals contact distances; there is therefore no structural evidence for any unusually strong forces between the molecules. This conclusion is in agreement with that based on an estimate of the cohesion energy, which is very nearly the same as that of a comparable aliphatic polyester, polyethylene adipate. The high melting-point of polyethylene terephthalate in comparison with aliphatic polyesters is therefore not due to strong forces between the molecules; it is attributed to the rigidity of the aromatic ring with its attached $ \matrix \quad \quad \quad \ \quad \text{O}- \\ \quad \ \ \diagup \\ -\text{C} \\ \quad \quad \chembond{2,315} \\ \quad \quad \quad \ \ \ \text{O} \endmatrix $. In most drawn fibres of this polymer, the crystals do not have their c-axes (those parallel to the chain molecules) exactly parallel to the fibre axes; they are tilted in a precisely defined crystallographic direction, such that the $\overline{2}$30 plane remains vertical and the inclination of the 001 plane to the fibre axis increases. This is especially well defined in fibres heated to 210 degrees C and allowed to relax, the tilt angle being then about 5 degrees. In principle it is expected that polymer crystals of triclinic symmetry would be tilted, but it is not clear what determines the crystallographic direction of tilt. Tilted crystal X-ray diffraction photographs played an important part in the structure determination.