Iterative discrete on-axis encoding of diffractive optical elements

Abstract

The recently developed iterative discrete on-axis (IDO) encoding method (based on iterative optimization routines such as simulated annealing) has previously been used to generate arrays of signal spots with extremely high diffraction.1 This encoding method has been extended to encode diffractive optical elements for high-resolution imaging. The encoding method is used to determine the location of the fringes in the computer-generated hologram (CGH) in order to minimize a particular cost function. Depending on the cost function, the CGH can be designed to minimize aberrations, maximize diffraction efficiency, and/or minimize noise due to undesired diffraction orders. These design parameters can be optimized for several points in the output plane and for a range of wavelengths. Standard diffractive element design procedure consists of two steps: (1) use of simulated annealing to calculate the desired continuous-phase function of the CGH, and (2) use of a deterministic encoding method to convert the phase function to a set of discrete, quantized fringes of constant phase. The modified IDO encoding method differs from the standard procedure in that the set of fringes are calculated directly by simulated annealing. The computation time can be decreased by restricting the hologram transmittance to contain particular symmetries (e.g., radial symmetry). Computer simulation results indicate that our method can produce diffractive optical elements with higher diffraction efficiency and larger minimum feature sizes than can conventional encoding methods. Experimental results will be presented.