The design and manufacture of advanced, epitaxial semiconductor devices is critically dependent on the availability of single crystal substrates with negligibly low levels of extended defects.  The substrates available commercially which posses sufficiently low defect densities and sufficiently large wafer areas only include 3 lattice constants: GaP/Si (5.45 Å/5.43 Å), GaAs/Ge (5.65 Å/5.66 Å), and InP (5.87 Å).  Design of epitaxial devices is thus limited by this constraint, since the growth of a thin film with a lattice constant even a fraction of a percent different from that of the substrate soon accumulates significant numbers of defects once the ‘critical’ layer thickness is reached.  The ability to develop devices without regard for lattice constant would fundamentally change the manner in which they are designed, and expand the palate of materials and alloy systems available to crystal growers.  One of the most promising routes to new lattice constants is the metamorphic buffer layer (MBL), in which a ternary (or higher order) alloy is grown on a commercial substrate, starting at the lattice matching composition and grading until a composition corresponding to the desired lattice constant is reached. Instead of forcing a suboptimal device onto one of the commercially available substrates, the device itself can be optimized and the substrate tailored to that device.
Since dislocations are required to accommodate the misfit between the substrate and epilayer, the goal of MBLs is to relax the misfit strain by introducing dislocations in a controlled manner which confines them to the graded region.  An optimal MBL has zero residual strain and a threading dislocation (TD) density equal to, or even lower than, that of the original substrate.  Limitation of processes which impede dislocation glide is crucial as blocked dislocations provide less strain relief compared to dislocations which can lengthen without limitation. Then fewer dislocation half loops are required to achieve the same strain relaxation and the total number of accompanying threading segments, which intersect the surface, is minimized.  A common strategy in the literature involves the use of slow grading rates which space out dislocations, decreasing the likelihood that dislocations will interact and impede each other.  Phase separation also creates glide impediments which lead to increased dislocation densities, and the choice of growth conditions that avoid phase separation is necessary as well.
Hydride vapor phase epitaxy (HVPE) is an attractive technique for the growth of MBLs.  Its characteristically high growth rates makes the growth of thick grading layers at slow grading rates feasible, allowing ample spacing between dislocations.  Growth is common at temperatures higher than those employed by molecular beam epitaxy (MBE) and metal-organic vapor phase epitaxy (MOVPE), which promotes rapid dislocation glide.  Since HVPE growth temperatures are well above the spinodal temperatures of common ternary systems like GaAs_xP_1-x, In_xGa_1-xAs, and In_xGa_1-xP and growth proceeds much closer to equilibrium than in MBE and MOVPE, phase separation is potentially mitigated in HVPE layers.  Lastly, the high growth rates make the growth of a sacrificial polishing layer possible.  The use of chemical mechanical planarization techniques reduces the surface roughness inherent to mismatched growth by all types of vapor deposition allows the creation of an ‘epi-ready’ surface with a new lattice constant.
This work will demonstrate the growth and subsequent device integration of HVPE-grown In_xGa_1-xAs and GaAs_xP_1-x MBLs.  The relationships between grading style, grading rate, final x (in In_xGa_1-xAs or GaAs_xP_1-x), and deposition temperature on buffer layer properties were explored.  The suitability of the MBLs as a platform for device growth was characterized through the measurement of surface residual strain, threading dislocation density, and root mean square (RMS) surface roughness. Defect structure was analyzed by transmission electron microscopy and related to measured macroscopic layer parameters.  Guidelines for buffer layer design and optimization will be offered. The potential for device incorporation through the growth and polishing back of a sacrificial layer was demonstrated through the growth of quantum cascade laser type superlattices on planarized HVPE surfaces.

1. PRESENTATION: Thick metamorphic buffer layers grown by hydride vapor phase epitaxy as platforms for novel semiconductor devices, TIFF image data, little-endian, 0.5MB

• Legal notice:
  

  Copyright (c) Pielaszek Research, all rights reserved.
  The above materials, including auxiliary resources, are subject to Publisher's copyright and the Author(s) intellectual rights. Without limiting Author(s) rights under respective Copyright Transfer Agreement, no part of the above documents may be reproduced without the express written permission of Pielaszek Research, the Publisher. Express permission from the Author(s) is required to use the above materials for academic purposes, such as lectures or scientific presentations.
  In every case, proper references including Author(s) name(s) and URL of this webpage: https://science24.com/paper/28785 must be provided.

1. Characterization of thick InxGa1-xAs metamorphic buffer layers grown by HVPE
2. Metastable Epitaxial Alloys