氮化銦鎵有機金屬化學氣相沉積法熱力學磊晶模型分析

Abstract

本文主要探討氮化銦鎵有機金屬化學氣相沉積法的熱力學磊晶模型。在傳統模型中，磊晶溫度並不影響氮化銦鎵的薄膜組成。換句話說，高銦組成之氮化合物可在任何成長溫度加以合成。但事實上，研究學者發現利用有機金屬化學氣相沉積法在高溫製備薄膜時，銦原子極難溶入薄膜，這與傳統熱力學之磊晶模型完全不符。Koukitu教授為此提出NH3寄生反應機制，試圖解釋上述熱力學模型的缺失。但實驗數據顯示，NH3寄生反應機制並不明顯，顯然Koukitu教授修正後之氮化銦鎵熱力學模型仍未足以圓滿解釋氮化合物在有機金屬化學氣相沉積法的磊晶現象。 在我們氮化銦鎵之磊晶實驗中，我們發現除了NH3的解釋，另一最有可能的解釋應是銦原子的高溫脫逸或副反應現象。依據我們氮化銦鎵熱力學之高溫修正模型計算，理論值與實驗結果相當吻合。除此之外，我們亦能預測出銦滴點產生之磊晶條件，並確認在無氫的環境中，能完全避免銦滴點產生，因而可以有效控制氮化銦鎵之晶體品質。相信我們在氮化銦鎵的熱力學磊晶模型之研究，將有助於氮化合物在光電元件之製備與研究。

We have carried out a thermodynamic model for metalorganic vapor phase epitaxial (MOVPE) growth of InGaN tenary Using the conventional model without any modification, the InGaN solid composition is found to be independent of growth temperature, contrary to the results observed from InGaN MOVPE growth, such as difficulty of In incorporation at high growth temperatures. Because of this, Koukitu et al. introduced an ammonia parasitiic reaction to describe the growth of InGaN.However, no evidence of this reaction has been discovered for InGaN, which indicative of imperfection of Koukitu model. In our InGaN thermodynamic growth model, a high-temperature parasistic rerasitio reaction of In, such as high temperature effect and In droplet formation, instead of parasitionreaction of NH3 was used to express the low In incorporation efficiency and formation of In droplets for InGaN. It is found that our model can give a better description of InGaN MOVPE growth, and more importantly it takes into the account of In droplet formation so that a high-quality of InGaN without any metal droplets on the surface can be easily obtained in our InGaN growth.