Professor, Department of Electronic Science and Engineering, Graduate School of Engineering, Kyoto University
Manipulating light with photonic crystals
—Nanostructures lead to photonic bandgap formation—
Optical devices can be miniaturized to as small as one hundred thousandths of their current size using photonic crystals. Photonic crystals are a kind of photonic insulators, through which specific wavelengths of light cannot propagate due to a photonic bandgap. When two or more substances that have a large difference in their refractive indices are arranged alternately with a period of a half wavelength, a band gap structure, through which the light of the certain wavelength cannot propagate, is formed.
Prof. Noda fabricated the world’s first photonic crystals that have a lattice-like structure with stacked stripes of GaAs. Free-spaces in the stacked stripes structure form a diamond lattice structure, and light is reflected due to the difference in refractive index between air that fills the free-spaces and GaAs. To block light with wavelengths of 1.4 to 1.6 microns, which are used in optical communication devices, 200 nm wide stripes have to be stacked with a period of 700 nm, and the degree of accuracy must be within 30 nm. He developed alignment equipment that uses optical laser diffraction patterns when stacking stripes of GaAs. He, then, fabricated photonic crystals with sharp stripe edges by maintaining the temperature of wafer fusion which binds the surfaces of the stripes together at about 500°C. Light within a photonic bandgap cannot propagate through a photonic crystal, but when defects are introduced into the crystal to disturb the periodic structure, light can propagate through the photonic crystal via the defects. When a pair of crossed stripes is removed, a sharp bending waveguide is made, and therefore, light can be bent at a right angle. Point defect cavities act as resonators and confine light within the cavities. Light can be amplified, or oscillated, using point defect cavities, and the wavelength of confined light depends on the size of the point defect cavities. Prof. Noda showed how to obtain the desired wavelength of light by changing the sizes of the defects and free-spaces within the stacked stripe structures.
Making two-dimensional photonic crystals is easier than three-dimensional photonic crystals. When air holes forming triangular lattice patterns are made with a period of a half wavelength on a thin silicon plate with a thickness of 250 nm, a photonic bandgap is formed due to the periodic structure in the in-plane direction, and light within the bandgap cannot propagate through the crystals. In the vertical direction with a non-periodic structure, light is confined because of total internal reflection, which occurs at the interface between silicon and air. Introducing a line defect by removing a single row of air holes forms a linear waveguide. When point defect holes are introduced by increasing the hole radius, a specific wavelength of light, corresponding to the radius of the defect hole, is confined within the defects, and then light is emitted perpendicular to the surface of the two-dimensional photonic crystal. Thus, optical branching filters, which separate light with different wavelengths, which propagate through the waveguides, can be fabricated.
High-performance resonators should confine light over a long period of time. To increase the Q factor, which is an indication of the performance level of the resonators, the difference in refractive indices in the vertical direction must be large if point defect holes in two-dimensional photonic crystals are used as resonators. Prof. Noda fabricated point defects using the same method as line defects, i.e. he removed holes, and increased the Q factor from 450 to 3800. To obtain higher Q factors, he had to prevent light from leaking out of the crystal. “When I thought about where the leaks were, I realized that the edges of the defect holes, which reflect light, were the places that light leaked. Light leaks up and down due to the strong reflection as ocean waves hit a quay and the splash shoots up high. So, buffer materials must be placed on the edges of the defect holes to prevent light from leaking,” says Prof. Noda. When the positions of the holes adjacent to a defect formed by removing a hole are shifted to the outside, the phase of the propagating light is shifted by the periodic disturbance, and the first reflection becomes weak. A 60 nm shift caused the Q factor to increase to 45,000, which is larger than that of the current ultra-small optical resonators by two orders of magnitude. Now, the value has reached almost 1 million, and theoretically, the value has no upper limit. When he started studying photonic crystals, others said that he would not achieve anything from his research. “However, if you think too much about what will happen next, you will never do anything new. So, you should take the first step toward something new. It may be harsh at first, but if you make adjustments as needed, things will get better. If you don’t try out new things, you will never gain anything. All you need to do is to take the first step,” says Prof. Noda.
