氮化銦超快速載子釋能機制研究

Ultrafast Carrier Relaxation Study in Indium Nitride

蘇倚恩(Yi-En Su)

指導教授 ： 孫啟光

國立臺灣大學/電機資訊學院/光電工程學研究所/碩士(2010年)

https://doi.org/10.6342/NTU.2010.02785

摘要

氮化銦在近期因為其極小的直接能帶寬而廣泛受到重視(小於零點七個電子伏特)，相對而言，氮化鎵則有著大約三點四個電子伏特的直接能帶寬，氮化銦鎵因此被預期根據不同比例的銦和鎵可使其放射光譜波長範圍從三百七十奈米一直到一千七百奈米，涵蓋了可見光波長的區域(大約四百奈米到七百奈米之間)、，因此氮化銦被認為有極高的潛力應用在發光二極體(LED)燈以及太陽能電池材料的上面。對於氮化銦的各樣載子(帶電粒子)加熱後的動力與釋能的機制了解顯得格外重要。
近代有關氮化銦載子動力學的研究當中，發現氮化銦還擁有許多其他的特性，很小的電子有效質量(零點零四二個電子質量)、相對小的電洞有效質量(零點四二個電子質量)，非拋物線型的傳導帶，電子大量在表面累積的特性，很高的電子遷移率和很高的電子飽和漂移速度。其中有些特性使得氮化銦不只在光電領域，就算在各種需要快速反應的電子設備當中，也成為一種很有潛力的材料。而氮化銦極高的縱向光學聲子能量(73個毫電子伏特)，也使其被預期會有個極快的，低於一百個飛秒雷射的電子釋能時間。然而過去實驗的結果卻不符合理論上的預期，截至2010之前的期刊中所發表的氮化銦的電子釋能時間都高於四百個飛秒，甚至在早期一點的文章有量測到接近於十個皮秒的釋能時間。
對於電子釋能時間的延長有兩種不同的機制解釋，一為熱聲子效應，一為其他電子對於個別電子與縱向光學聲子作用所產生的屏蔽效應。對於氮化銦電子釋能時間遠高於理論預期的原因，早期的文章期刊將其歸因於熱聲子效應，一直到2006年溫氏等人刊登的一篇文章根據一些實驗結果反對了這種說法，並且將釋能時間延長的原因歸於屏蔽效應，至此關於氮化銦釋能時間延長原因的爭論沒有中斷過。此篇論文中將証實氮化銦過長的電子釋能時間是來自於屏蔽效應。並且將發表在藉由降低電子濃度而去除屏蔽效應的影響後卻可觀察到的，符合理論預期的低於一百個飛秒雷射的電子釋能時間。
本論文也將發表關於氮化銦電洞在雷射光激發後在價帶帶底的各種動力學行為的研究，包含其電洞吸能與釋能時間的理論預測和實驗比較。這些研究結果對於氮化銦載子動力學的全盤了解將會有很大的幫助，希望未來氮化銦在各個前述領域的實際應用上將因為這篇論文所做的研究，而能更快預測並且解決因氮化銦載子特性所產生的問題，並且進一步有效的應用其特殊的載子特性，在光電與固態電子領域產生有效的突破與發展。

關鍵字

氮化銦 ； 載子動力 ； 電子釋能 ； 電洞吸能 ； 参鎂氮化銦

並列摘要

Indium nitride (InN) has been considered important recently because of its small direct band gap (~0.6 electron volt (eV)), comparing to Gallium Nitride (GaN) with a direct band gap of ~3.4eV. With different fraction of indium and gallium, GaxIn1-xN was expected to have emitting spectral wavelength ranged from 370nm to 1700nm, covering the region of visible light (about 400nm ~ 700nm) and telecommunication wavelength. InN was regarded as a great potential material applied in light-emitting diode (LED) and solar cells. Therefore, the carrier thermalization dynamics and energy relaxation mechanism in InN is highly important.
Recent investigations of InN carrier dynamics found that InN had many other special properties, very light electron effective mass (0.042m0), relatively light hole effective mass (0.42m0), nonparabolic conduction band, high electron accumulation on surface, high electron mobility, and high electron saturation drift velocity. These special properties make InN not only prospective in the optoelectrical field, but also a potential material for high reactive rate electrical devices. The ultrahigh longitudinal optical phonon (LO-phonon) energy of InN (~73meV) makes it expected to have an ultrafast electron energy relaxation time below 100 femtoseconds (fs). However, experimental results were not consistent with theoretical expectation. All the paper published before 2010 reported an electron energy relaxation time longer than 400fs. Papers published earlier even reported an energy relaxation time near 10 picoseconds (ps).
There are two explanations for the postponed electron energy relaxation time, one is the hot phonon effect, and one is the screening effect between each electron and LO-phonons induced by other electrons. Earlier papers attributed the reason to the hot phonon effect. Until 2006, Wen et al reported some experimental results against the story, and attributed the postponed relaxation time to the screening effect. Since then, the argument between the reasons of relatively slow energy relaxation time has never ended. This thesis will prove that the main reason of the postponed electron energy relaxation time shown in the published papers is the screening effect. This thesis also show that as the screening effect and the hot phonon effect being removed by lowering the electron density, a sub-100 fs electron energy relaxation time can be directly observed, consistent with the theoretical expectation.
This thesis also report an investigation on hole thermalization dynamics at the edge of the valence band, including holes heating time with both theoretical expectation and experimental result. I hope that this thesis would make significant help in understanding InN carrier dynamics. While InN is applied in the electrical or optoelectrical field in the future, the investigation and understanding of InN carrier mechanism in this thesis could help to predict or to solve some carrier property problems, and further more to make breakthrough in the fields of optoelectronics and solid state electronics.

並列關鍵字

InN ； carrier dynamics ； electron cooling ； hole heating ； Mg doped InN

參考文獻

[1.1] S. Das Sarma, “Quantum Many-Body Aspects of Hot-Carrier Relaxation In Semiconductor Microstructures” in “Hot Carriers In Semiconductor Nanostructures” edited by Jagdeep Shah. New Jersey (1992).

[1.2] S. Z. Sun, Y. C. Wen, S. H. Guol, H. M. Lee, S. Gwo, and C.-K. Sun, “Observation of femtosecond carrier thermalization time in indium nitride”, Appl. Phys. Lett. 103, 123513 (2008)

[1.3] Kasic A, Schubert M, Saito Y, Nanishi Y and Wagner G, “Effective electron mass and phonon modes in n-type hexagonal InN”, Phys. Rev. B 65, 115206 (2002).

[1.4] Inushima T, Higashiwaki M and Matsui T, “Optical properties of Si-doped InN grown on sapphire (0001)”, Phys. Rev. B68 235204 (2003)

[1.5] C. L. Tang and D. J. Erskine, Phys. Rev. Lett. “Femtosecond Relaxation of Photoexcited Nonequilibrium Carriers in AlxGa1-xAs”, 51, 840 (1983)