1.High speed silicon-based integrated optoelectronic circuit for short-wave infrared detection (矽基短波紅外光高速偵測器積體電路)：

近年來，由於網路的盛行，人類對於資料的交換、環境的識別與隱私權的需求爆炸性增加，使得紅外光感測元件的需求大幅增加。其中，光通訊元件、用於自動車輛三維環境建構的光達系統與用於隱私權人臉辨識，在效能與安全的考量下，將轉向1550 nm。但1550 nm 的光偵測器根基於Ge與InGaAs，基板與製程的成本相當高，以至於大量應用受到限制。本研究面對此一問題，提出以矽為基板，搭配奈米結構與新穎高折射率差二氧化矽-鍺錫週期性結構矽光子紅外偵測器技術。

起始研究工作將著重在元件製程的開發，與探討週期孔洞對多晶鍺錫吸光特性的關聯性。光波範圍將以1000-1600 nm分光光源為主。以光吸光能力來做為評估元件好壞的機制，同時進行實驗室等級光偵測器元件的製作，並與國際結果進行比較，達成在1550 nm外部量子效率大於 20 %，光響應大於 0.30 A/W。基於多晶鍺錫於紅外光吸收最佳化的研究成果，提升光偵測器元件的效能，並將著重於將多晶鍺錫光偵測器技術導入矽積體電路中，以積體電路公司的製程製作電晶體、反相器、電阻、電感、電容等元件，並以BiCMOS 技術製作接觸區，同時保留紅外光光偵測器區域。研究後製程對於轉阻放大器元件特性的影響。本年度將同時研究鍺錫紅外光偵測器的的動態行為，了解電荷產生、電荷轉移，與電荷消失的機制，尋找光電反應時間的主要因素，達成在1550 nm外部量子效率大於 40 %，光響應大於 0.50 A/W，元件頻寬大於25 GHz。最終研究將著重在達成高速矽積體電路化鍺錫高效能電紅外光偵測器電路，目標是10Gb/s 的光電積體電路。第二年研究將建立鍺錫紅外光偵測器製程對積體電路元件特醒的影響參數，將持續進行後製程溫度修正，以確保積體電路元件正常。最後，將利用所得到的參數，以現有的轉阻放大器電路與高折射率差週期性鍺錫結構紅外光偵測器整合成為短波紅外光高速光偵測器電路。藉由此一研究的完成，將加速矽光子紅外光偵測器的廣泛應用。

In recent years, due to the fast growing of of the world-wide web, the demanding on data exchange, 3D environment re-construction and smart phone security, which has greatly increased the demand for infrared photodetectors. Among these applications, the optical wavelength for component/subsystems such as fiber transceiver module, lidar for the automatic vehicle and face recognition for security, will migrate to 1550 nm under the consideration of performance and safety. However, the 1550 nm photodetector is based on Ge and InGaAs, in which the cost of the substrate and device process is too high to be ubiquitous. In this proposal, we propose a 1550 nm infrared photodetector comprise Si as the substrate and nanostructures for light-trapping and a novel, high refractive index contrast SiO2-GeSn materials for light absorption.

We will focus on the development of the device process for the infrared photodiode and explore the correlation between the periodic microholes and the light absorption characteristics of silicon dioxide/germanium-tin high refractive contrast periodic structure. The light source will be based on the 1000-1600 nm monochromatized light source. The light absorption capacity was used as the key parameter to evaluate the quality of the microhole array. At the same time, the photodetector was made and bench marked with international results. The external quantum efficiency was aim to larger than 20% at 1550 nm, and the responsivity higher than 0.30 A/W was expected. Based on the research results of the optimization of infrared absorption of Silicon dioxide/germanium-tin periodic structure in the first year, the performance of photodetector will be improved. The external quantum efficiency was aim to larger than 40% at 1550 nm, and the responsivity higher than 0.50 A/W, bandwidth larger than 25 GHz was expected. In addition, we will integrate Silicon dioxide/germanium-tin periodic photodetector into 0.18 um BiCMOS technology. MOSFETs, inverters, resistors, inductors, capacitors and other components that used in the transimpedance amplifier were designed and evaluated in the BiCMOS technology. Meanwhile, leave an area for infrared light detector in this design. The final goal of this research will focus on the realization of high-speed silicon integrated circuit-based on germanium-tin high-efficiency electrical infrared photodetector circuits with a target of 10 Gb/s.With this technology, we believe the integrated photonic circuit for 1550 nm will be realize for variety of application, especially the LiDAR in self-driving cars.

利用0.18 um BiCMOS製程驗證紅外光偵測器的積體電路化可行性元件截面圖。

Schematic structure for the proposed Si-based GeSn OEIC.

元件製作流程如圖。(a)週期孔洞製作，(b)矽材料選區再氧化，(c)透明氧化導電層製作，(d)金屬接觸點製作與(e)元件橫截面示意圖。
Process flow and schematic diagram of the GeSn photodetector.

2.Multiple Photon-generation from Uniformly Dispersed Silicon Quantum Dots in Phosphorescent OLED

Quantum dots is a special form of semiconductor material that exhibit three-dimensional confinement in structure. If the size of this confinement less than twice of the Bohr exciton radius, the quantum effect will cause the bandgap of the material to change, called quantum confinement. Therefore, this cause the optical emission properties correlate to the size of the QDs. The QDs has been made having multiple color emission by tuning it size. The QDs-based light-emitting diodes (LED) was then developed accordingly. Those QDs are compound based, CdSe in particular, is the most efficient luminescent material. Cadmium (Cd) is toxic that may cause environmental problem limiting the applications. Silicon is an alternative material. It is abandum in the earth, non-toxic, and wavelength tunable in optical emission. Silicon and SiGe based QDs-LED were published to emit either visible or infrared light. However, the quantum efficiency is too small to be practical. To enhance the luminescent efficiency, the Si QDs were then inserted into organic materials to have band alignment for carrier injection. The Si QDs were produced by the solid-state synthesize procedure. The HSiCl₃ was used as the starting material to form Si/SiO₂ glass layer. This layer was then reduced to form Si QDs with allylbenzene capping layer. The valence band and conduction band edge for the 1.8 nm large Si QDs are 5.7 eV and 3.9 eV, respectively. The Si QDs layer was inserted between poly(N,N′-bis(4-butylphenyl)-N,N’-bis(phenyl)benzidine (poly-TPD) and 1,3,5-tris(N-phenylbenzimidazol2,yl)benzene (TPBI), served as a hole transport layer and hole blocking layer, respectively. The highest external quantum efficiency (EQE)increases to 1.1 % with emission at around 700 nm. Further reduced the sized of the Si QDs to around 0.8 nm, the emission shifted from 700 nm to around 400 nm, obtaining blue-white spectrum in conjunction to the emission poly-TPD [16]. Replacing the poly-TPD and TPBI by 1,1-Bis[(di-4-tolylamino) phenyl] cyclohexane (TAPC) and ZnO nanoparticles in inverted structure with appropriated band alignment, the EQE could be enhanced to 2.7%. In addition to be a luminescent materials, the QDs could also be used as the band alignment layer in an organic light-emitting diode (OLED) due to the very efficient carrier transport through QDs. To enhance luminescent efficiency, combine organic and QDs is a suitable approach. The small molecular phosphorescent metal complex has very efficient luminescent ability due to the metal-ligand charge transfer that can convert all the electrons to photons by utilizing both singlet and triplet excitons. Among the phosphorescent materials, the Ir-based metal-complex molecule, 4,4’-Bis(N-carbazolyl)-1,1’-biphenyl (CBP) and Bis[2-(4,6-difluorophenyl) pyridinato-C2, N] (picolinato) iridium (III)(FIrpic) with sky-blue emission was used to explore how Si QDs help OLED. Sandwich FIrpic with hole and electron injection layers, a strong sky-blue emission was observed with quantum efficiency around 5 %. Further fine tuning the band alignment by the complicated injection and transport layer, the quantum efficiency can be increased to 22%.

3.Short-wavelength infrared LiDAR for Autonomous Driving

未來世界距離不是限制人類的理由，工作的地點不受限制，無人運輸工具將會成熟，解決人們行的安全問題。不論是送貨或是載人; 距離不論是數公尺到數百公尺的人工視野。達成這些需求的方法統稱為『自駕車』(Autonomous Car)或是『先進駕駛輔助系統』(Advanced Driver Assistance Systems, ADAS)。許多的公司或是研究機構正著手進行自駕車的研究與應用，著手於未來將商品做自動投遞業務。

自動駕駛汽車的原型看起來與一般汽車相差無幾。遍佈車身的內建感測器會發出訊號，遇到物體便會反彈，運作原理與轎車的停車感測器相同，讓自駕車知道自身與路緣、行人和其他車輛相距多遠。中央處理電腦與GPS系統都藏在車內，所以自動駕駛汽車的外觀與一般車輛主要不同之處，在於偵測系統。在討論偵測系統前，我們知道全無人駕駛非一蹴可及，需要相當時間的科技進展，系統驗證，道路實測等等。

在自駕車上，距離與環境的感測器系統主要為雷達(Radio-frequency Detection and Ranging, RADAR)，影像系統(Camera)以及光達系統(Light Detection and Ranging, LiDAR)，等三類。雷達系統所使用的微波波長較長，不受天候影響，具有短中長程優異的偵測能力，對於單一物體的速度與距離能有效偵測。但是雷達無法判斷物體之種類及大小。光學影像具有物體辨識能力，能分辨顏色，標線與交通號誌，但無法得知物體的距離與速度資訊。同時亦受大雨、濃霧與強光影響，夜晚辨識力有限，無法與雷達系統完全互補。光達系統係利用雷射光照射在物體上，利用反射光進行物體種類與距離，速度判斷。同時由於波長短，空間解析度較雷達佳，能高精度測得多個物體距離與速度，辨識物體外型並建立精細3D環境資訊。對於自動駕駛來說，可同時補足雷達與光學影像的不足。主要應用在數公尺至200公尺之內物體的偵測，得到速度資訊。並與影像系統融合，得到物體種類，將成為未來無人駕駛車的核心技術。

整個LiDAR系統可簡單分為光源、掃描系統、檢測訊號系統以及資料處理系統。光源系統，通常為雷射。雷射發出後經過透鏡組或光學模組，進行光束的校正與掃瞄後，於透鏡組發射至環境中，再將環境待側物所反射光導入系統中，藉由光路導引至光偵測器，進而進行訊號解析。因此，一個光達收發模組中，需要發光系統、光學掃瞄系統、光偵測系統、以及光電訊號解調系統來達成。

其中最基礎的是雷射光波長的選擇。雷射光必須不受日光干擾，對環境不影響，且不傷害人體，尤其是眼睛。以偵測100公尺遠的物體為例，雷射光發出後，經物體反射回到偵測器端，光強度衰減為發射光強度的在10^-7-10^-9倍[1]。因此，在光達的裝置上，為達成遠距離的偵測，需要應用一個分常高光功率的脈衝雷射當作發射源，在雷射光抵達待測物前被眼睛接收到，將會造成無可挽回的災難。目前車用光達大部分使用波長905-nm的半導體雷射，因為905-nm的半導雷射的製作技術較成熟、成本低，可以產生較高功率的短脈衝雷射，量測距離較遠。但是905-nm波段的人眼安全允許暴露值相當低，在波長超過高於1400-nm的光，例如1550-nm是相對較為安全的，這部分可見於IEC 60825 的Maximun Permeable Exposure (MEP)規範中。因為，在可見光與小於1400-nm的紅外光會聚焦在視網膜，再加上眼球的聚焦作用，容易對視網膜造成永久傷害，因此以一個ns的脈衝光雷射光而言，1550 nm的MEP 是905 nm 的10的5次方倍以上，相當驚人。目前車用光達模組都需要符合人眼安全規範，但未來無人車充滿整個街上，905奈米波段的雷射滿天飛的時候，人眼安全雷射的使用，將會是一個無可迴避的議題，光達裝置的雷射波長移至1550nm將是未來的主流趨勢[2]。因此本研究嘗試製作1550 nm光達系統，並討論其中可能的困難，與設計的重點。

所設計的光達(LiDAR)系統如下圖一所示，包含三個主要部件: 1,550 nm 紅外光收發與飛行時間法距離運算模組、1550 nm紅外光掃瞄系統、及三維點雲圖產生與顯示模組。透由脈衝紅外光發射至光掃描模組，投射到待測物上，反射光由光接收器接收。兩者訊號透由同步激發系統，算出時間差。以飛行時間法計算出待測物距離。光掃描模組的掃描訊號及距離同時送入微處理器進行三維點雲圖產生，此為光達工作原理。

1550 nm紅外光光達運作原理圖。

[1]. G. M. Williams, “Optimization of eyesafe avalanche photodiode lidar for automobile safety and autonomous navigation systems.” Optical Engineering 56(3), 031224, March, 2017

[2]. J.Y. Jhong and Zingway Pei, “Analysis on the 1550 nm Beam Steering Behavior in the Linear Optical-Phased Array Emitters for LiDAR Application,” SPIE Optics & Photonics, San Diego, Aug., 2019.