Single-crystal – oriented-crystal – and epitaxy growth processes; – Forming from vapor or gaseous state – With decomposition of a precursor
Reexamination Certificate
2000-06-28
2002-09-10
Kunemund, Robert (Department: 1765)
Single-crystal, oriented-crystal, and epitaxy growth processes;
Forming from vapor or gaseous state
With decomposition of a precursor
C117S093000, C117S094000, C117S095000, C117S952000
Reexamination Certificate
active
06447604
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates to a method of achieving improved epitaxy quality of (Al,In,Ga)N films on corresponding free-standing substrates for fabrication of opto-electronic and electronic devices and device precursor structures.
2. Description of the Related Art
(Al,In,Ga)N (which term as used herein refers inclusively and alternatively to each of individual nitrides containing one or more of Al, In and Ga, thereby alternatively encompassing each of AlN, Al
x
In
1−x
N (or AlInN), Al
x
Ga
1−x
N (or AlGaN), Al
x
In
y
Ga
1−x−y
N (or AlInGaN), InN, In
y
Ga
1−y
N (or InGaN) and GaN where 0≦x≦1 and 0≦y≦1, as well as mixtures thereof and doped layers (n-type or p-type) or remaining undoped) has been extensively studied with respect to its epitaxial layer growth on heavily lattice-mismatched substrates such as sapphire and SiC.
A primary reason for the pervasive character of such research is that free-standing (FS), coefficient of thermal expansion (CTE)—matched, and lattice-matched GaN substrates of suitable quality and size are unavailable.
Without homoepitaxial or native substrates, misfit dislocations will form due to lattice mismatch at the epitaxy-substrate interface, and cracking and bowing will occur due to the CTE mismatch, thereby limiting the quality of the epi and device layers. The epitaxial layer quality on these non-optimal substrates (e.g., sapphire or SiC) is of reasonable quality for simple electronic devices if complicated interlayer techniques are used.
Typically, to make higher quality devices, very difficult and complicated overgrowth techniques such as ELOG (epitaxial lateral overgrowth) or LEO (lateral epitaxial overgrowth) or Pendeo-epitaxy are employed, but the resulting material is non-uniform in morphology and crystalline quality. Further, the resulting material typically has a high carrier concentration due to impurity incorporation from the masking material. Such overgrowth techniques employ the use of a masking material such as SiO
2
to inhibit growth in certain areas on the substrate material. The epitaxial material then grows between the masked region and then laterally over the masking material, thereby reducing dislocation propagation in the laterally grown area.
The lack of a suitable quality lattice-matched (Al,In,Ga)N substrate has impeded (Al,In,Ga)N device developers from realizing the full potential of the (Al,In,Ga)N device capabilities and slowed the development of this material system. The complexity and difficulties attendant the lateral overgrowth techniques have prevented such approach from being satisfactorily commercially used.
A small amount of work has been done to produce nitride substrates and an even smaller amount of epitaxial layer growth has been done on the limited amount of GaN material produced.
As a background to discussion of the problems with GaN epilayer growth on FS GaN, techniques for producing free-standing GaN are described below. The ensuing discussion also highlights how the properties of some of the substrates have inhibited the development of a suitable epitaxial process.
Substrate Production
Potential methods for producing lattice matched or nearly lattice matched substrates superior to sapphire and SiC that have been developed to date include high pressure GaN crystal growth, AlN bulk growth, lithium aluminate (LAO), lithium gallate (LGO), thick (>100 micron) HVPE GaN and lift-off, and HVPE GaN boule growth, as discussed more fully below.
High Pressure Crystal Growth
High pressure crystal growth has been successful in producing small platelets (<20 mm diameter and <1-2 mm thick) of less than 300 square millimeters area of single crystal GaN but the GaN crystals have several problems. This technique produces small platelets and the scalability is difficult and the cost of the process is quite large compared to other alternatives. Further, dopant and conductivity control of the crystal is very difficult due to the technique. Another disadvantage is that high unintentional impurity levels are present in the crystal including oxygen, which make the substrates conductive. These high levels of impurities limit the frequency range of devices produced on the substrates due to parasitic capacitances between device layers and charge in the substrate and may inhibit epitaxy nucleation on the substrate at sufficiently high impurity concentrations.
AlN (or GaN) Substrate Formation via Sublimation and Re-condensation
The production of bulk AlN by sublimation and re-condensation technique is being performed to produce suitable, high quality, nearly lattice-matched (2.5% difference from GaN) substrates for GaN epitaxial growth. Currently, the boule diameter is limited to 13 millimeters, severely limiting the production of lost cost, high volume devices.
Another issue with these substrates is the extremely high oxygen level, on the order of parts per million (ppm), which will likely reduce the thermal conductivity of the substrates, making them less advantageous for high frequency, high power devices.
In addition to affecting the thermal conductivity, the high impurity incorporation in these substrates inhibits the production of controlled electrical conductivity type substrates, namely p-type substrates. These substrates are difficult to dope heavily by conventional techniques, making them less advantageous for vertical opto-electronic device structures. In the case of AlN substrates, the substrate and associated devices are disadvantaged by high ionization or activation energy of acceptors and donors in the crystal, as compared to GaN substrates.
Lithium Aluminate (LAO) and Lithium Gallate (LGO)
LAO and LGO are closely lattice-matched substrates (compared to SiC and sapphire) and are available in reasonable quality and size, however, several issues exist that prevent their applicability to the GaN material system. Most importantly, LAO and LGO materials suffer from low decomposition temperatures preventing them from being easily used for GaN growth at typical growth temperatures. Li and Ga desorption and diffusion from the substrate into the epitaxial film and growth environment make nucleation and high quality, impurity-free growth very difficult, thus limiting the applicability of this substrate. Limited process conditions are employed to grow on these substrates due to their high susceptibility to decomposition under H
2
. Non-uniform polarity of the substrate surface is also an issue, typically causing mixed polarity domains in the GaN epitaxial film. The fabrication of vertical devices structures on such substrates also involves issues of doping and suppression of decomposition.
HVPE (Halide Vapor Phase Epitaxy) GaN Substrates via LILO (Laser Induced Lift-off) and HVPE GaN Based FS GaN Substrates via Boule Growth
The HVPE GaN method is the most preferred method to date to produce FS GaN substrates. It enables large-area freestanding GaN wafers to be produced of high quality and low dislocation density, on which high quality, smooth epitaxial films and high quality devices can be fabricated. The process has the ability to be easily scaled to the desired size of the wafer, and substrate conductivity type can be readily controlled. Precursor and growth process set-up is relatively inexpensive compared to other techniques (e.g. high pressure crystal growth) and can be easily controlled with conventional process controls. Impurity incorporation is minimal and can be controlled through precursor purity and gas-phase ambient purity as well as reactor leak integrity and construction.
Homo-epitaxial Growth on High Quality FS HVPE GaN Substrates
Because there have been no large area, freestanding GaN wafers commercially and readily available, there has been limited opportunity to develop the conditions to produce high quality epitaxial layer growth on FS GaN.
As discussed hereinafter in greater detail, the present invention enables growth of epitaxial films of crystalline quality at least as good as that of the substrate, resolv
Brandes George R.
Flynn Jeffrey S.
Keogh David M.
Landini Barbara E.
Vaudo Robert P.
Advanced Technology & Materials Inc.
Chappuis Margaret
Hultquist Steven J.
Zitzmann Oliver A.
LandOfFree
METHOD FOR ACHIEVING IMPROVED EPITAXY QUALITY (SURFACE... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with METHOD FOR ACHIEVING IMPROVED EPITAXY QUALITY (SURFACE..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and METHOD FOR ACHIEVING IMPROVED EPITAXY QUALITY (SURFACE... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2877451