The diffractive scattering by crystals of thermal neutrons, which have the de Broglie wave-length of about 1.6 A, had been investigated by several authors. It gives many intersting facts, which may offer some means of supplementing the X-ray study of the solid state. In the present experiments the total cross section of thermal neutrons have been measured for Fe, Cu and Al as a function of the degree of cold-working. The thermal neutrons from a paraffin “howitzer” with 200_??_300 mC Rn-Be source are detected by a BCl3 ionization chamber and a linear amplifier as shown in Fig. 1. About 2000 counts per minute are obtained without a scatterer, and the residual counts due to higher energy neutrons above Cd cut-off are 5% or less. Completely annealed plates of electrolytic iron, Armco iron, pure Cu and pure Al are subjected to various degrees of cold-working (rolling). The final thickness of the scatterers is so regulated as to give about 50% transmission in each case. The experimental results are illustrated in Fig. 5, which shows the relative variation in the total cross. section as a function of the degree of working. It can be seen that the annealed samples of Fe and Cu show small cross sections which quickly increase to the saturation values by, a cold-working of an order of only a few per cent reduction of thickness. On the other hand, Al shows no marked change beyond the limit of experimental errors. After analysing the experimental results of Beyer and Whitaker on Fe, Halpern, Hamermesh and Johnson suggested, the existence of secondary extinction in the scattering of thermal neutrons by a single crystal. They also reaches the conclusions that the primary extinction must be negligible on account of the small interaction of neutron and nucleus, provided that the size of the mosaic block of Fe be of the order of 10-5cm. It would, however, be premature to conclude that Fe had such a small mosaic block, though-Miller and DuMond had determined the size of a block of Al to be of that order of magnitude from X-ray investigations; because, in general, the lower limit of the size of a mosaic block in metals-is considered to be in the order of 10-4cm after Bragg. The present results seem to point out the existence of the effect of primary extinction in annealed polycrystalline Fe and Cu. If we neglect the incoherent part of elastic scattering, the size of the mosaic block can be calculated by the same method as described in a previous paper by one of us (M. K.). According to Darwin, a measure of the degree of extinction is given by tanh mq/mq, where q is the, amplitude of the wave reflected by a certain net plane, and m is the number of net planes forming a mosaic block. By assuming that the primary extinction is completely removed by sufficient working, the size of a mosaic block is determined to be in the order of 5×10-4cm for annealed Armco iron, 3.6×10-4cm for annealed electrolytic iron and 5.8×10-4cm for annealed Cu. The reason why Al shows no marked decrease of cross section by annealing can be accounted for by the smallness of q and of the size of a mosaic block. Rasetti has found that in the case of Pb or Bi the cross sections are almost the same for the single and polycrystalline states and tried to attribute this fact to the isotope or spin disorder in crystal lattices. It is difficult, however, to explain our results on Al along these lines, because Al has only one isotope.
Read full abstract