Defect reduction in GaN using the in-situ growth of transition metal nitride layers

自20世纪60年代以来，基于砷化镓及相关材料的红色、黄色和绿色LED已经问世。然而，适合于蓝光LED的宽禁带半导体很难开发出来，直到1993年才制造出第一个高效、长寿命的GaN基蓝光LED。这使固态照明领域发生了革命性的变化，并启动了一项专注于GaN和相关材料的相当大的研究工作。然而，人们很快发现，GaN薄膜中高密度的线位错严重限制了在紫外区和紫外区发射的GaN基器件的性能。因此，许多研究工作都集中在降低位错密度上，以努力开发高效的紫外光和紫外光发光二极管和激光器。这些发射器本身对水消毒、数据存储和通信非常有用，但对于白色LED也是必要的，白色LED通常基于紫色或UV LED和荧光粉的组合，这些荧光粉将短波长光转换为适当的红、绿和蓝波长的混合物。如果可以通过减少缺陷来提高白光LED的性能，那么将产生比传统技术更好的发光效率和寿命的固态照明。鉴于欧盟最近宣布禁止销售传统灯泡，这一倡议尤其重要。我们认为，明亮的白色LED为英国照明提供了最佳解决方案；荧光节能灯泡含有汞等有害金属(造成废物处理问题)，尺寸较大，显色性往往较差。因此，我们建议的简短项目的目的是减少或消除(A)六方GaN(一种特别适合于LED的GaN形式)中的线位错和(B)立方GaN(一种特别适合于激光的GaN形式)中的堆积缺陷。这一雄心勃勃的目标是由我们的初步研究推动的，该研究表明，新型过渡金属氮化物中间层具有阻止这种位错传播到器件的有源发光区的能力。我们打算通过使用与生长GaN相同的技术，即六方GaN的MOCVD或立方体GaN的MBE，在单一的生长过程中原位生长这样的层来发展这一概念。预计这将提供实际的好处(不中断生长过程，如以前所要求的)，并通过改善层的粗糙度等因素在减少位错密度方面提供显著改善。然而，我们也打算这项工作为过渡金属氮化物与GaN器件的集成铺平道路，这将改进LED接触技术，改进缓冲层技术，并能够制造出高结构质量的超导Tm-氮化物/绝缘GaN异质结(约瑟夫森结)，从而导致在医疗保健和地球物理中使用的超灵敏磁强计等应用。

Red, yellow and green LEDs based on GaAs and related materials have been available since the 1960s. However, wide band gap semiconductors suitable for blue LEDs were difficult to develop and the first efficient, long-lifetime blue GaN-based LED was not fabricated until 1993. This revolutionised the field of solid-state lighting and kick-started a considerable research effort focusing on GaN and related materials. However, it soon became clear that high densities of threading dislocations in the GaN films were severely limiting the performance of GaN-based devices that emit in the violet and UV region. Much research effort has therefore been focused on dislocation density reduction, in an effort to develop efficient violet and UV-emitting LEDs and lasers. These emitters are useful in their own right for water sterilisation, data storage and communications, but are also needed for white LEDs, which are commonly based on the combination of a violet or UV LED and phosphors which convert the short-wavelength light into an appropriate mixture of red, green and blue wavelengths. If the performance of white LEDs can be improved via defect reduction, then solid-state lighting with a luminous efficacy and lifetime considerably better than conventional technology will result. Such an initiative is particularly important in light of a recent EU announcement regarding the banning of sale of conventional light bulbs. We believe that bright, white LEDs offer the best solution for lighting the UK; fluorescent energy-saving bulbs contain harmful metals such as mercury (creating a waste disposal problem), are large in size and often provide poor colour rendering. The aim of our short proposed project is therefore to reduce or eliminate (a) threading dislocations in hexagonal GaN (a form of GaN particularly suitable for LEDs) and (b) stacking faults in cubic GaN (a form of GaN particularly suitable for lasers). This ambitious aim is prompted by our initial research, which indicated that novel transition metal nitride interlayers have the ability to block such dislocations from propagating into the active light-emitting region of a device. We intend to develop this concept by growing such layers in-situ in a single growth process using the same technique that is used to grow the GaN, namely MOCVD for hexagonal GaN or MBE for cubic GaN. This is expected to provide both practical benefits (no interruption of the growth process, as was previously required) and to provide significant improvements in dislocation density reduction via improvement in the roughness of the layers, among other factors. However, we also intend this work to pave the way for integration of transition metal nitrides with GaN devices, which could improve LED contact technology, improve buffer layer technology and allow superconducting TM-nitride/insulating GaN heterostructures (Josephson junctions) of high structural quality to be fabricated, leading to application such as ultra-sensitive magnetometers for use in healthcare and geophysics.