銻化合物半導體量子結構物理及元件應用

Abstract

含銻的化合物半導體是所有三五族半導體中之最奇特也最少被人所探討的半導體材料，它擁有所有半導體材料中最高的電子遷移率(InSb)，GaSb則具有與Ge類似的能隙與同樣高的電洞遷移率。由於銻化合物的electronic affinity較低，它與其它三五族半導体所形成的異質接面結構常會形成type II或type III的界面，這種奇特的界面不僅有許多有趣的物理現象也提供了許多特別元件應用的機會。 在過去銻化合物之所以鮮少為人所探討一個主要的原因是它所對應的發光波長落在中紅外線，這無論對光纖通訊的應用或display的應用而言都顯不出它的重要性，但近年來環保議題高漲，而空氣汙染的偵測需要中紅外線的光源及偵測，因此銻化物半導體的光電元件又重新顯得重要。不僅如此，矽元件在經過多年的發展之後，它的performance似乎已經走到了極致，要再進一步提升元件的速度則必需利用其它的材料。近年來Intel開始InSb量子井FET的研發。目的是將高速的InSbFET與Si元件integrate在同一矽基板上。初步的實驗結果顯示，在相同的channel length之下InSb FET的speed-power product將比Si MOSFET高出一到兩個order of magnitude。這是極為令人振奮的結果，這比傳統GaAsFET或HEMT又更向前邁進了一步。 上述的這些元件雖然有如此的potential，但要真正實現商業上的量產，還有許多技術上的挑戰需要克服也需要許多人力和物力的投入和努力。由於銻原子較大，它們的晶格常數比其它主要半導體的晶格常數要來的大。要成長銻化物磊晶在一些熟知的基板如Si、GaAs或InP上必需要克服晶格不批配的問題，以及發展將dislocation及impurity阻絕在外的長晶的技巧。 除了磊晶成長的問題外在元件的設計上也有很大的挑戰。在發光元件上，如要將所發的波長拉長，我們必須利用typeII的heterojunctuion。但是typeII的heterojunctuion的發光效率低，要得到好的及在室溫下可操作的雷射就必須要有很好的元件設計。至於在高速元件上，挑戰則是更大，要獲得高電子遷移率，必須要用narrow bandgap材料如InSb及InAs，但因為bandgap小極易產生impact ionization，元件很容易breakdown，而使得操作電壓變得很小。要克服這兩個彼此矛盾的問題，不是易事。它需要有不同於傳統的新concept和新idea。 我們在此proposed的計劃中將針對上述的問題加以探討，利用band engineering提出我們對元件的創新設計，我們希望在計劃結束時能有破性的成果，將無論是發光元件或高速元件推升到實用的地步。 在元件的應用外，銻化物與其他半導體間的heterostructure及奈米結構將是我們另一個研究主題，GaSb量子點在GaAs matrix中因為type II接面而形成電子的環狀結構，這將使我們有機會看到Aharonov–Bohm effect。我們已經建立一套在低溫高磁下的magneto-optical及micro photoluminenscence量測系統，這將使我們能單一的量子點或量子環的發光特性，不僅如此了我們所設計研發在低溫高磁系統下的橢圓儀將讓我們能量出任何奈米結構在低溫高磁下的EM response。這將使我們對銻化物及其量子結構有更深入的瞭解。

Antimonite compound semiconductors are probably the least well known semiconductors while possessing some of the most intriguing properties. InSb, the III-V material with the largest lattice constant, has the highest electron mobility of all semiconductors. GaSb, which has similar bandgap as Ge, has one of the highest hole mobility of all semiconductors. The combinations of III-Sb materials offer the possibilities of making light emitters and detectors operating in the wavelengths of mid-infrared, the wavelength range that is important for air pollution monitoring. With the ever growing concern on the carbon emission and environmental protection, the monitoring and the control of the air quality have become one of the highest priorities of all countries. So the development of Sb based devices for mid-infrared applications has generated a lot of interests recently. However to extend the wavelength beyond 2.5um using Sb compounds is still a challenge. One often has to use the type-II heterostructures for this purpose. But the reduced recombination rate due to the spatially separated electrons and holes causes the light emission less efficient. In this program, we plan to tackle this problem by using innovation design of the heterojunctions and the use of nano-cavities based on photonic crystals. Another interesting device that has drawn a lot of attentions recently is the Sb based FETs for high speed applications. Intel has recently put in a strong effort in the development InSb quantum well FETs for the purpose of making devices that can be integrated with Si technology but with performance beyond the capability of Si devices. Experimentally, they have demonstrated some very exciting result, which shows that the speed-power product of an InSb FET can be one or two orders of magnitude lower than the best Si devices. However, to realize such devices in a large scale, one still need to overcome many technical obstacles. One of the biggest challenges is to increase the breakdown voltage of such devices. Because of the narrow bandgap, the impact ionization happens at a relatively small voltage, and that causes the breakdown voltage to drop. The requirements for high mobility and high operation voltage are conflictive to each other. We plan, in this program, to use an innovative approach to overcome this problem. By proper band engineering, we should be able to obtain both high electron mobility and high breakdown voltage at same time. Besides device studies, we will also investigate the interesting physics behind the heterostructrures and nanostructures of Sb compounds. GaSb quantum dots in GaAs matrix offer the possibility to generate an ideal electron ring structure that we can observe the Aharonov–Bohm effect. We have set up a micro-PL system and an ellipsometer system that are able for us measure all the interesting magneto-optical and EM responses of the Sb compounds and the nanostructures that were not well known in the past.