Fundamental Study on Silicon Optical Switch for Total-Silicon Opto-Electric Integrated Circuits

本研究的目的是利用硅中的自由载流子实现红外光吸收光开关。这是从来没有提出过的，也是一个非常具有挑战性的目标。由于实现这一点的关键问题之一是如何产生大量的自由载流子，因此制作了梳状MOS晶体管。由于下面所述的原因，第一次试验失败了，但实现了一种新的梳状MOS晶体管，并根据对光波导结构的分析确定了未来的目标。第一次试验失败的原因1.入射的平行红外光直径为10μm，不能停留在1μm高和1 mm长的光束内，而是分散到硅衬底上。1.实现了有效沟道长度为2μm、高1μm、束宽50 nm的梳状晶体管。有31束光束的梳形晶体管实现了比平面晶体管高5倍的驱动能力。进一步的研究1.入射红外光应该利用SOI衬底锁定在具有较高折射率的硅光束内。应该利用伊藤门。但是，由于ITO的导电性较小，吸收较小，因此估计栅导电性应该得到优化。由于分析了铝等高反射率金属薄膜的适用范围，因此将采用铝栅MOS晶体管和ITO晶体管。在上述实验和分析的基础上，仍计划实现一种光硅开关MOS晶体管。

The purpose of this study is an implementation of optical switch based on infra-red light absorption with free carriers in silicon. This is not ever proposed and a very challenging target. Since one of key issues to realize that is how large amount of free carriers is generated, a comb-shaped MOS transistor is fabricated. Due to reasons described below, the first trial failed however, a new comb-shaped MOS transistor is realized and future targets can be fixed based on analyses on optical wave guide structure.<Reasons of failure of first trial>1. Incident parallel infra-red light of 10 μm in diameter does not stay within a beam of 1 μm in height and 1 mm in length but disperse into silicon substrate.2. Polysilicon with 10^<21>/cm^3 impurity, filled into comb gaps, strongly absorbs incident infrared light.<Realization of comb-shaped transistor>1. A comb-shaped transistor of 2 μm in effective channel length with 1-μm high and 50-nm wide beams is successfully obtained.2. A comb-shaped transistor having 31 beams realizes 5-times greater drivability compared to that of planar transistor.<Further investigations>1. Incident infra-red light should be locked up inside silicon beam having higher refractive index using SOI substrate.2. ITO gate should be utilized. However, since the less conductivity of the ITO gives less absorption, it is estimated that the gate conductivity should be optimized.3. Since it is analyzed that high reflective metal films like aluminum can be suitable, aluminum-gate MOS transistors will be implemented as well as the ITO transistors.4. Based on the experimental and the analyses described above, it is still planned to realize an optical silicon switch MOS transistor.