Coating apparatus – Gas or vapor deposition – With treating means
Reexamination Certificate
1997-03-24
2001-04-17
Bueker, Richard (Department: 1763)
Coating apparatus
Gas or vapor deposition
With treating means
C118S7230IR
Reexamination Certificate
active
06217662
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to semiconductor manufacturing processes, and in particular relates to an improved susceptor design for epitaxial growth on silicon carbide substrates.
BACKGROUND OF THE INVENTION
The present invention relates to the production of epitaxial layers of semiconductor materials on silicon carbide substrates. Silicon carbide offers a number of advantageous physical and electronic characteristics for semiconductor performance and devices. These include a wide bandgap, high thermal conductivity, high saturated electron drift velocity, high electron mobility, superior mechanical strength, and radiation hardness.
As is the case with other semiconductor materials such as silicon, one of the basic steps in the manufacture of a number of silicon-carbide based devices includes the growth of thin single crystal layers of semiconductor material on silicon carbide substrates. The technique is referred to as “epitaxy,” a term that describes crystal growth by chemical reaction used to form, on the surface of another crystal, thin layers of semiconductor materials with defined lattice structures. In many cases, the lattice structure of the epitaxial layers (or “epilayers”) are either identical, similar, or otherwise related to the lattice structure of the substrate. Thus, epitaxial growth of either silicon carbide epitaxial layers on silicon carbide substrates or of other semiconductor materials on silicon carbide substrates, is a fundamental technique for manufacturing devices based on silicon carbide.
Silicon carbide is, however, a difficult material to work with because it can crystallize in over 150 polytypes, some of which are separated from one another by very small thermodynamic differences. Furthermore, because of silicon carbide's high melting point (over 2700° C.), many processes for working silicon carbide, including epitaxial film deposition, often need to be carried out at much higher temperature than analogous reactions in other semiconductor materials.
Some basic reviews of semiconductor manufacturing technology can be found for example in Sze,
Physics of Semiconductor Devices,
2d Ed. (1981), Section 2.2, pages 64-73; or in Dorf,
The Electrical Engineering Handbook,
CRC Press, (1993) at Chapter 21 “Semiconductor Manufacturing,” pages 475-489; and particularly in Sherman,
Chemical Vapor Deposition for Microelectronics: Principles, Technologies and Applications,
(1987), ISBN O-8155-1136-1. The techniques and apparatus discussed herein can be categorized as chemical vapor deposition (CVD) or vapor phase epitaxy (VPE) in which reactant gases are exposed to an energy source (e.g. heat, plasma, light) to stimulate a chemical reaction, the product of which grows on the substrate.
There are several basic techniques for CVD epitaxial growth, the two most common of which are the hot (heated) wall reactor and cold wall reactor processes. A hot wall system is somewhat analogous to a conventional oven in that the substrate, the epitaxial growth precursor materials, and the surrounding container are all raised to the reaction temperature. The technique offers certain advantages and disadvantages.
The second common conventional technique is the use of a “cold wall” reactor. In such systems, the substrate to be used for epitaxial growth is placed on a platform within a container (typically formed of quartz or stainless steel). In many systems, the substrate is disk-shaped and referred to as a “wafer.” The substrate platform is made of a material that will absorb, and thermally respond to, electromagnetic radiations.
As is known to those familiar with such devices and techniques, the susceptor's response to electromagnetic radiation is an inductive process in which alternating frequency electromagnetic radiation applied to the susceptor generates an induced (inductive) current in the susceptor. The susceptor converts some of the energy from this inductive current into heat. In many systems, the electromagnetic radiation is selected in the radio frequency (RF) range because materials such as glass and quartz are transparent to such frequencies and are unaffected by them. Thus, the electromagnetic radiation passes through the container and is absorbed by the susceptor which responds by becoming heated, along with the wafer, to the temperatures required to carry out the epitaxial growth. Because the container walls are unaffected by the electromagnetic energy, they remain “cold” (at least in comparison to the susceptor and the substrate), thus encouraging the chemical reaction to take place on the substrate.
A thorough discussion of the growth of silicon carbide epitaxial layers on silicon carbide substrates is set forth for example in U.S. Pat. No. 4,912,063 to Davis et al. and U.S. Pat. No. 4,912,064 to Kong et al., the contents of both of which are incorporated entirely herein by reference.
The use of a cold wall reactor to carry out epitaxial growth, although satisfactory in many respects, raises other problems. In particular, because a semiconductor wafer rests on a susceptor, the wafer side in contact with the susceptor will become warmer than the remainder of the substrate. This causes a thermal gradient in the axial direction through the wafer. In turn, the difference in thermal expansion within the wafer caused by the axial gradient tends to cause the peripheral edges (typically the circumference because most wafers are disc-shaped) to curl away from, and lose contact with, the susceptor. As the edges lose contact with the susceptor, their temperature becomes lower than the more central portions of the wafer, thus producing a radial temperature gradient in the substrate wafer in addition to the axial one.
These temperature gradients, and the resulting physical effects, have corresponding negative affects on the characteristics of the substrate and the epitaxial layers upon it. For example, if the edges are placed in extreme tension, they have been observed to crack and fail catastrophically. Even if catastrophic failure is avoided, the epitaxial layers tend to contain defects. At silicon carbide CVD growth temperatures (e.g. 1300°-1800° C.), and using larger wafers (i.e. two inches or larger), wafer bending becomes a significant problem. For example,
FIG. 3
herein plots the values of wafer deflection (H) at various axial temperature gradients as a function of the wafer diameters.
Furthermore, because wafers have a finite thickness, the heat applied by the susceptor tends to generate another temperature gradient along the central axis of the wafer. Such axial gradients can both create and exacerbate the problems listed above.
Yet another temperature gradient typically exists between the rear surface of the substrate wafer and the front surface of the susceptor; i.e. a surface-to-surface gradient. It will thus be understood that both radiant and conductive heat transfer typically take place between susceptors and substrate wafers. Because many susceptors are formed of graphite coated with silicon carbide, the thermodynamic driving force created by the large temperature gradients between the susceptor and the silicon carbide wafers also causes the silicon carbide coating to undesirably sublime from the susceptor to the wafer.
Additionally, because such sublimation tends to promote pin hole formation in the susceptor coating, it can permit contaminants from the graphite to escape and unintentionally dope the substrates or the epilayers. This in turn ultimately leads to non-uniform doping levels in the semiconductor material, and reduces the lifetime of the susceptor. The problems created by susceptors which undesirably emit dopants is set forth for example in the background portion of U.S. Pat. No. 5,119,540 to Kong et al.
Nevertheless, a need still exists for susceptors that can operate at the high temperatures required for silicon carbide processing while minimizing or eliminating these radial, axial and surface to surface temperature gradients, and the associated physical changes and problems.
OBJECT AND SUMMARY OF THE IN
Carter, Jr. Calvin
Kong Hua-Shuang
Sumakeris Joseph
Bueker Richard
Cree Inc.
Fieler Erin
Summa, P.A. Philip
LandOfFree
Susceptor designs for silicon carbide thin films does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Susceptor designs for silicon carbide thin films, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Susceptor designs for silicon carbide thin films will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2503851