Metallo-dielectric photonic crystal is a special class of photonic crystals that are known for the extreme robustness of the photonic gap. Simulation based on multiple scattering theories predicts the survival of photonic band gaps in the presence of disorder. For the particular case of fcc metallo-dielectric crystal, it was found theoretically that both the lower frequency directional gap and higher frequency absolute gap persist when randomness is introduced in the microwave regime....[ Read more ]

Metallo-dielectric photonic crystal is a special class of photonic crystals that are known for the extreme robustness of the photonic gap. Simulation based on multiple scattering theories predicts the survival of photonic band gaps in the presence of disorder. For the particular case of fcc metallo-dielectric crystal, it was found theoretically that both the lower frequency directional gap and higher frequency absolute gap persist when randomness is introduced in the microwave regime.

The goal of this research is to see experimentally whether the directional gap and the absolute gap survive positional disorder in a randomly-assembled sample. The fundamental building blocks are iron balls coated with a layer of insulating silicon rubber, and the frequency range of interest is from 1 to 18 GHz. Experiments for both ordered and randomly-assembled configurations are prepared separately. Many system configuration problems are encountered and solved during the course of the experiment. A best possible experiment configuration is identified.

For the ordered fcc sample in a particular horizontal configuration, the measured results are in quantitative agreement with the theory prediction. We can only find qualitative agreement between theory and experiment in other configurations, and our experimental results differ quantitatively in some aspects from the numerical simulation in the disordered sample. Both a directional gap and an absolute gap are indeed found in our results, but the position of the gaps deviates from the simulation. However, the qualitative feature of our experiments can be explained by the theory. We find that structural change in the sample due to its own weight is a factor that affects the result. Some suggestions for further improvements of the experiments are introduced.