1. Field of Invention
This invention relates generally to the growth of III-nitride films on mismatched substrates.
2. Description of Related Art
Semiconductor light-emitting devices (LEDs) are among the most efficient light sources currently available. Materials systems currently of interest in the manufacture of high-brightness LEDs capable of operation across the visible spectrum include Group III-V semiconductors, particularly binary, ternary, and quaternary alloys of gallium, aluminum, indium, and nitrogen, also referred to as III-nitride materials. Typically, III-nitride light emitting devices are fabricated by epitaxially growing a stack of semiconductor layers of different compositions and dopant concentrations on a sapphire, silicon carbide, or III-nitride substrate by metal-organic chemical vapor deposition (MOCVD), molecular beam epitaxy (MBE), or other epitaxial techniques. The stack often includes one or more n-type layers doped with, for example, Si, formed over the substrate, a light emitting or active region formed over the n-type layer or layers, and one or more p-type layers doped with, for example, Mg, formed over the active region.
Since III-nitride substrates are generally not commercially available, III-nitride devices are often fabricated on substrates such as sapphire and SiC, which are referred to as “mismatched” because of differences in lattice constant and thermal expansion coefficients between these substrates and III-nitride layers. The differences in lattice constant and thermal expansion coefficient cause strain which makes the fabrication of thick, flat, defect-free III-nitride layers difficult.
In the case of sapphire substrates in particular, one method to alleviate the problems caused by the lattice constant and thermal expansion coefficient mismatch is to grow the III-nitride device layers on a nucleation layer grown over the substrate. The nucleation layer is usually a thin amorphous or polycrystalline AIN or GaN layer formed by low temperature growth. In theory, since the amorphous or polycrystalline nucleation layer alleviates some strain caused by the lattice constant and thermal expansion coefficient mismatch, the quality of a crystalline III-nitride layer grown over the nucleation layer should be improved over a III-nitride layer grown without a nucleation layer.
The use of low temperature nucleation layers is not ideal for several reasons. Before growth of the semiconductor layers, the substrate is heated to a high temperature for cleaning, then the temperature in the reactor is reduced to grow the low temperature nucleation layer. The temperature must then be raised again to grow the device layers. Such temperature cycling is time consuming and therefore increases the likelihood of the accumulation of contaminants on the wafer surface that can affect yield or device performance. In addition, the heating that takes place between growth of the low temperature nucleation layer and growth of the next layer causes recrystallization in the low temperature nucleation layer that is necessary for growth of high quality layers over the nucleation layer. Accordingly, the heating must be carefully controlled. Deviations from the desired heating pattern can cause poor device performance or poor yields. Further, the materials quality of layers grown over the low temperature nucleation layer and the performance of the resulting device is sensitive to the thickness of the nucleation layer and the growth temperature of the nucleation layer. The thickness and temperature regimes that yield acceptable device performance can be narrow, making it difficult to maintain the required thickness and temperature regimes across the wafer and resulting in poor yield. Finally, employing a low temperature nucleation layer results in a high density of threading dislocations in the device layers, on the order of 1010 cm−2. Additional processing is required to reduce the density of the threading dislocations. Such additional processing increases the time and cost of fabrication and the risk of poor yield or poor device performance.
Another method to alleviate the problems caused by the lattice constant and thermal expansion coefficient mismatch is described by U.S. Pat. No. 5,990,495 to Ohba. Ohba teaches “growing a buffer layer directly on a monocrystal substrate at a first temperature . . . and growing an element-forming layer on said buffer layer at a second temperature which is lower than said first temperature . . . . The temperature for growing the buffer layer should preferably be higher than 1300° C.” See column 3, lines 54-64. After growth, the nucleation layer is annealed. “The temperature for annealing the buffer layer is preferably in the range of 1350° C. to 1500° C., and the time period for annealing the buffer layer should preferably be in the range of 10 minutes to 60 minutes. If the annealing temperature is less than 1350° C., it would be difficult to alleviate strain completely.” See column 6, line 64 to column 7, line 2. Ohba's method of fabricating a high temperature nucleation layer is undesirable because it requires high temperature growth and a growth pause in order to perform the anneal.
In accordance with embodiments of the present invention, a method of forming a light emitting device includes providing a sapphire substrate, growing an Al1−xGaxN first layer by vapor deposition on the substrate at a temperature between about 1000° C. and about 1180° C., and growing a III-nitride second layer overlying the first layer. The first layer may have a thickness between about 500 angstroms and about 5000 angstroms. In some embodiments, reaction between the group V precursor and the substrate is reduced by starting with a low molar ratio of group V precursor to group III precursor, then increasing the ratio during growth of the first layer, or by using nitrogen as an ambient gas.
The use of a high temperature nucleation layer according to embodiments of the present invention is preferable to a low temperature nucleation layer because it may eliminate the temperature cycling, growth pause, and additional processing required for low temperature nucleation layers, and because it eliminates problems with uniformity due to the narrow thickness and temperature regimes required by low temperature nucleation layers.
Nucleation layer 12 is Al1−xGaxN with a thickness that can range between about 500 angstroms to 5000 angstroms, usually ranges between about 1000 angstroms and about 3000 angstroms, and most preferably ranges between about 2000 angstroms and about 2500 angstroms. The composition of aluminum nitride may be between about 50% and 100% (i.e. x ranges from 0 to about 0.5). Nucleation layer 12 may be fabricated at a temperature ranging from about 1000° C. to about 1180° C., is usually fabricated at a temperature ranging from about 1060° C. to about 1150° C., and is most preferably fabricated at a temperature ranging from about 1080° C. to 1100° C. In some embodiments, coalescence layers in devices with nucleation layers grown at less than 1060° C. exhibited increased surface roughness. The suitable temperature, composition, and thickness of nucleation layer 12 may depend on the layers grown over nucleation layer 12. Typically, nucleation layer 12 is undoped, but in some embodiments it may be lightly doped.
After growth of nucleation layer 12, a coalescence layer 14 is formed to provide a smooth surface for later growth of device layers 15, 16, and 17. Coalescence layer 14 is often GaN, but may be GaN, InGaN, AlGaN, AIN, or AlInGaN. The composition of coalescence layer 14 is selected such that coalescence layer 14 has a desired strain state and is transparent to light emitted by the active region. In some embodiments, coalescence layer 14 is an additional layer separate from the device layers that provides a smooth uniform surface on which to grow the device layers. In other embodiments, coalescence layer is the first device layer and serves a purpose in the device in addition to providing a smooth surface for the layers overlying coalescence layer 14. Coalescence layer 14 may be undoped or doped n-type or p-type. If the dopant is Si, coalescence layer may be doped to a maximum dopant concentration of about 5×1018 cm−3. Coalescence layer 14 may be more heavily doped with n-type dopants other than Si. If the dopant is a p-type dopant such as Mg, coalescence layer 14 may be doped to a maximum dopant concentration of about 5×1019 cm−3. The thickness of coalescence layer 14 may range from about 500 angstroms to about 2 microns, and usually ranges from about 0.5 microns to about 1.5 microns. Coalescence layer 14 is usually grown at the same temperature as nucleation layer 12, or at a slightly lower temperature. In one example, nucleation layer 12 is grown at a temperature of 1090° C., and coalescence layer 14 is grown at a temperature ranging from about 1060° C. to about 1090° C.
The semiconductor layers in the device illustrated in
In some embodiments, the ratio of group V precursors to group III precursors is adjusted during fabrication of nucleation layer 12 in order to reduce reaction, for example nitridation, between the group V precursor and the substrate.
After the portion of the nucleation layer immediately adjacent to the substrate has finished growing, for example, after growth of the first 100 to 500 angstroms, the flow rate of the group V precursor is increased in stage 22 to a group V/group III molar ratio of 168. Typically the flow rate of the group III precursor remains constant during the entire growth of the nucleation layer and the group V precursor flow rate is adjusted to achieve the desired molar ratio, though the group V precursor flow rate can be held constant as the group III precursor flow rate is adjusted. The molar ratios given in
In stage 24, the group III precursor is changed or an additional group III precursor is added and growth of coalescence layer 14 begins. For example, if coalescence layer 14 is GaN, the flow of trimethyl aluminum is stopped and a gallium precursor such as trimethyl gallium is introduced into the reactor. If coalescence layer 14 is AlGaN, the flow rate of the aluminum precursor may be adjusted and a gallium precursor is introduced to the reactor. Coalescence layer 14 and the layers in n-type region 15 are usually grown at a molar ratio of about 1000, though the molar ratio can be as high as 2000.
In general, III-nitride layers are grown in a hydrogen ambient, because layers grown in a hydrogen ambient are observed to be smoother than layers grown in other ambients. In some embodiments of the present invention, nitrogen rather than hydrogen is used as ambient during growth of nucleation layer 12.
The light output, surface smoothness, dislocation density, and strain in devices fabricated with the high temperature nucleation layer of
Embodiments of the present invention also offer several advantages over the method described in U.S. Pat. No. 5,990,495 (hereinafter “Ohba”). First, the methods described herein do not require an anneal to create devices with sufficiently low strain to exhibit acceptable device performance. Second, nucleation layer 12 and coalescence layer 14 are grown at significantly lower temperature than Ohba's nucleation layer, reducing the time required to heat the substrate, the time required to cool the device to temperatures appropriate for forming the device layers, and the possibility of damage to the reactor.
Having described the invention in detail, those skilled in the art will appreciate that, given the present disclosure, modifications may be made to the invention without departing from the spirit of the inventive concept described herein. Therefore, it is not intended that the scope of the invention be limited to the specific embodiments illustrated and described.
Number | Name | Date | Kind |
---|---|---|---|
5393993 | Edmond et al. | Feb 1995 | A |
5585648 | Tischler | Dec 1996 | A |
5661074 | Tischler | Aug 1997 | A |
5888886 | Sverdlov et al. | Mar 1999 | A |
5903017 | Itaya et al. | May 1999 | A |
5990495 | Ohba | Nov 1999 | A |
5990496 | Kunisato et al. | Nov 1999 | A |
6106898 | Takamatsu et al. | Aug 2000 | A |
6229151 | Takeuchi et al. | May 2001 | B1 |
6303473 | Heffernan et al. | Oct 2001 | B1 |
6426512 | Ito et al. | Jul 2002 | B1 |
6440214 | Hooper et al. | Aug 2002 | B1 |
6440823 | Vaudo et al. | Aug 2002 | B1 |
6593016 | Chiyo et al. | Jul 2003 | B1 |
6610551 | Doverspike et al. | Aug 2003 | B1 |
6683327 | Krames et al. | Jan 2004 | B2 |
20030102482 | Saxler | Jun 2003 | A1 |
20030198837 | Craven et al. | Oct 2003 | A1 |
20030207518 | Kong et al. | Nov 2003 | A1 |
20040113163 | Steigerwald et al. | Jun 2004 | A1 |
Number | Date | Country |
---|---|---|
10-27947 | Jan 1989 | JP |
11-40893 | Jun 1989 | JP |
11-97803 | Aug 1989 | JP |
10-270368 | Oct 1998 | JP |
11-112029 | Jan 1999 | JP |
P2000-40858 | Feb 2000 | JP |
P2000-68498 | Mar 2000 | JP |
P2000-196143 | Jul 2000 | JP |
WO 9910936 | Mar 1999 | WO |
Number | Date | Country | |
---|---|---|---|
20040115853 A1 | Jun 2004 | US |