Feasibility study of plasma-assisted electroepitaxy for the growth of GaN layers and bulk crystals

人们对氮化物半导体有很高的商业和科学兴趣。第三族氮化物(AlN、GaN和InN及其固溶体)被用于琥珀色、绿色、蓝色和白色发光二极管、蓝色/紫外线激光二极管以及大功率、高频和高温电子器件。用金属-有机气相外延(MOVPE)、氢化物气相外延(HVPE)和分子束外延(MBE)生长了用于器件制造的III族氮化物层。阻碍氮化物技术进步的最严重的问题之一是很少有合适的晶格匹配的衬底可以生长III族氮化物薄膜。第三类氮化物层通常生长在非晶格匹配的蓝宝石、砷化镓或碳化硅衬底上。然而，为了制造高质量的GaN基器件，需要在晶格常数和热膨胀特性上与外延氮化物层相匹配的块状GaN衬底。在低位错密度的标准III-V晶体的商业化生产中，通常采用溶液生长方法。众所周知，通过使用溶液生长技术，结晶过程非常接近平衡。目前，正在积极探索溶液生长方法来生长块状GaN晶体。为了实现高效的GaN溶液生长技术，我们需要满足3个主要要求：1)N在GaN基溶液中的高溶解度，2)N和Ga向生长界面的有效传输，3)防止溶液中的局部过饱和，否则将导致自发结晶。从液态Ga溶液中可以生长出高质量、低位错密度的纤锌矿GaN衬底。然而，氮气在液态镓中的溶解度很低，很难获得合理的生长速度，因此大面积块体GaN晶体仍然没有商业化。利用微波等离子体源产生的氮气与金属镓反应合成纤锌矿GaN晶体，避免了氮气所需的高平衡压力。射频等离子体源中的氮原子可以用来在镓熔体中产生高浓度的氮。不幸的是，这种高浓度的氮只存在于金属镓表面附近，通常会导致多晶GaN在金属镓表面自发结晶。为了实现高效的外延生长过程，需要开发一种技术来将N物种通过镓熔体输送到生长表面，同时最小化自发形核。液相电外延(LPEE)是一种晶体生长方法，在保持整个系统温度不变的情况下，通过通过溶液-衬底界面的直流电来启动和维持层的生长。电流通过LPEE生长池产生四个主要效应：1)液态金属溶液中溶剂的电迁移；2)异质界面上的Peltier效应：3)生长池的焦耳加热；4)溶液中对流的增加。电迁移和生长界面的Peltier冷却共同产生了生长界面所需的浓度梯度。在目前的应用中，我们建议开发一种全新的、廉价的技术来生长具有低位错密度的高质量块体GaN晶体--等离子体辅助电外延(PAEE)。我们将结合等离子体工艺在Ga熔体中产生高浓度N物种的优势和电外延将这些物种从Ga表面转移到生长界面的优势，而不会在表面或溶液中自发结晶。

There is a high level of commercial and scientific interest in nitride semiconductors. The group III-nitrides (AlN, GaN and InN and their solid solutions) are being used for amber, green, blue and white light emitting diodes, for blue/UV laser diodes and for high-power, high-frequency and high temperature electronic devices. Group III-nitride layers for device fabrication are grown by metal-organic vapour phase epitaxy (MOVPE), hydride vapour phase epitaxy (HVPE) and molecular beam epitaxy (MBE). One of the most severe problems hindering progress in the field of nitride technology is the rarity of suitable lattice-matched substrates onto which group III-nitride films can be grown. Group III-nitride layers are commonly grown on non-lattice matched sapphire, GaAs or SiC substrates. However, bulk GaN substrates, which are matched in lattice constant and thermal expansion properties to epitaxial nitride layers are needed for fabrication of the highest-quality GaN-based devices. Solution growth methods are normally used in the commercial production of standard III-V crystals with low dislocation density. It is well established that by using solution growth techniques, the crystallisation takes place very close to equilibrium. At present, solution growth methods are being actively explored to grow bulk GaN crystals. In order to achieve an efficient solution growth technique for GaN we need to fulfil 3 main requirements: 1) a high solubility of N in the Ga-based solution, 2) efficient transport of N and Ga to the growth interface and 3) prevent local supersaturation in the solution, which will otherwise result in spontaneous crystallisation. High quality low dislocation density bulk wurtzite GaN substrates can be grown from liquid Ga solutions. However, the solubility of N2 in liquid Ga is very low and it is difficult to obtain reasonable growth rates and therefore large area bulk GaN crystals are still not commercially available. Wurtzite GaN crystals have been synthesized by reacting gallium metal with atomic nitrogen produced by a microwave plasma source, which avoids the high equilibrium pressure needed for N2. Atomic nitrogen from an RF plasma source can be used to produce high concentrations of N in a Ga-based melt. Unfortunately, this high concentration of N only exists close to the surface of the metallic Ga and normally results in spontaneous crystallization of polycrystalline GaN on the surface of metallic gallium. In order to achieve an efficient epitaxial growth process one needs to develop a technique to transport the N species through the gallium melt to the growth surface and at the same time to minimize spontaneous nucleation. Liquid phase electroepitaxy (LPEE) is a crystal growth method, in which the layer growth is initiated and sustained by passing a direct electric current through the solution-substrate interface while the temperature of the overall system is maintained constant. An electric current passing through the LPEE growth cell causes four main effects: 1) electromigration of the solvents in the liquid metal solutions; 2) Peltier effect at hetero-interfaces: 3) Joule heating of the growth cell and 4) increased convection in the solution. Electromigration and Peltier cooling of the growth interface together produce the required concentration gradient to the growth interface. In this current application we are proposing to develop an entirely novel, inexpensive technique for the growth of high quality bulk GaN crystals with a low dislocation density - plasma assisted electroepitaxy (PAEE). We will combine advantages of the plasma process for producing high concentrations of N species in the Ga melt with the advantages of electroepitaxy in transferring these species from the Ga surface to the growth interface without spontaneous crystallisation on the surface or within the solution.