COMPUTER-SIMULATION STUDY OF THE EFFECT OF GRAIN SIZE ON THE EFFICIENCY OF LATENT-IMAGE FORMATION

Latent-image formation in AgBr cubes from 0.1 to 0.95-mu m edge length has been studied by a computer simulation technique based on the nucl eation-and-growth model. Both the depth and density of electron traps were found to affect quantum sensitivity. As the trap depth increases, the trap density must decrease so as to maintain reasonably eficient latent-image formation. Because of the high irradiance conditions used in the simulations, a four-atom minimum developable size leads to an efficiency loss due to high irradiance reciprocity failure. In this ca se, to maintain reasonable efficiencies, the trap density had to be de creased relative to that for a three-atom minimum developable size. As the grain size increases, the trap density, at a fixed trap depth, mu st be decreased to maintain high efficiency. This reduction in trap de nsity is due to the effect of grain size on the partitioning between f ree and trapped states, and the resultant effect on the predominant pa thway for recombination. For a constant trap depth of 0.2 eV that may be typical for sulfur-plus-gold sensitization, it is not possible to f ind a fixed trap density that leads to a size-independent inefficiency . However, if different trap densities are used for different grain si zes, it is possible to find such a condition. This result indicates th at electron range limitations due to size alone are not important in e xplaining the experimentally observed sublinear dependence of speed on grain size for large grains, at least within the size range studied. When a lower hole mobility is used the dependence of quantum sensitivi ty on trap density is significantly reduced because the contribution o f free-hole/trapped-electron recombination is reduced.