Study on the polarity control of hexagonal-structure widegap compound semiconductors and its effect on the material control

宽禁带化合物半导体如GaN具有六方纤锌矿结构，沿c轴缺乏对称性。与这一特性相对应的是，生长过程中的极性/方向对外延过程本身和材料的许多性能都有很大的影响。本文首先研究了氮化镓MOVPE和分子束外延生长过程中的极性控制机制，提出了一种成功的从N-极性到Ga-极性转换的方法，其中Al双层起着重要的作用。结果表明，在生长过程中，覆盖层表面化学/热稳定，Al双层膜可以冻结在晶体中。其次，研究了氧化锌的极性控制。与氮化物相比，氧化物的极性控制比较复杂。这部分归因于氧化物倾向于无定形，特别是在低温下，低温沉积的氧化锌缓冲液有时不能保持衬底和外延层之间的外延关系。我们提出用氮化物作为缓冲层，成功地实现了极性控制。最后，我们首次将我们的工作扩展到InN外延，并阐明了N-极性生长区域在InN本身的外延及其材料控制中是可取的。这是因为InN在低于500℃的低温下很容易分解，但N-极性生长的外延温度可以高达600℃，比单极性生长高出约100℃。在此基础上，我们首次成功地生长了器件级的InN，InGaN和InAlN等InN基三元合金，以及由InN/InGaN和InN/InAlN异质结组成的非常精细的结构SQW/MQW。

Widegap compound semiconductors such as GaN have the hexagonal wurtzite structure and are lack in symmetry along the c-axis. Reflecting this property, the epitaxy process itself and many properties of these materials were greatly affected by the polarity/direction during growth. However, the mechanism for the polarity selection and its control during growth of these materials were not clear.In this research, first we investigated the polarity control mechanism in both MOVPE and MBE of GaN, and we proposed a successful method for the polarity inversion from N-polarity to Ga-polarity, where Al-bilayer played an important role. It was found that the All-covered surface was chemically/thermally stable and Al-bilayer could be frozen into the crystal during growth.Next, the polarity control of ZnO was investigated. The polarity control of oxides was complicated compared to that of nitrides. This was partly attributed to the fact that oxides tended to be amorphous, in particular at low temperatures, the ZnO buffer deposited at low-temperatures sometimes could not keep the epitaxy relationship between the substrate and epilayer. We proposed to use nitrides as buffer layers resulting in successful polarity control.Finally we extended our work to the InN epitaxy for the first time and we clarified that the N-polarity growth regime was preferable in the epitaxy of InN itself and its material control as well. This was because the InN easily decomposes at such low temperatures below 500 C but the epitaxy temperature for the N-polarity growth could be as high as 600 C, which was about 100 deg higher than that of In-polarity growth. On the basis of these results, we succeeded in growth of device-quality InN, InN-based ternary alloys, such as InGaN and InAlN, and very fine structure SQW/MQW consisted of InN/InGaN and InN/InAlN heterostructures for the first time.