Chemistry: electrical and wave energy – Apparatus – Coating – forming or etching by sputtering
Reexamination Certificate
1998-11-23
2002-02-05
Nguyen, Nam (Department: 1753)
Chemistry: electrical and wave energy
Apparatus
Coating, forming or etching by sputtering
C204S298040, C204S298060, C204S298080, C250S492100, C250S492200
Reexamination Certificate
active
06344116
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates generally to integrated circuits (ICs), and more particularly to three-dimensional ICs, and still more particularly to three-dimensional (3-D), monocrystalline ICs, and to their associated technology.
2. Description of the Prior Art
The dominant current effort to achieve 3-D organization with integrated circuits is the “multichip module” approach, which places one conventional two-dimensional (2-D) integrated circuit on top of another, and another on top of that, and so forth. But the resulting volumetric packing density is low. Also, the extra procedures required to combine finished ICs in such an assembly are causes of yield loss.
A more ambitious but still brute-force effort to realize a three-dimensional integrated-circuit (3-D IC) employed a technique that became known as “stacking,” usually applied to CMOS (complementary metal-oxide-silicon) circuitry. Two or more layers of circuitry were sought in this general approach by cyclic repetition of essentially conventional fabrication-technology steps. But in spite of lavish investments in this concept for a period of about a decade, the technology by now has been largely abandoned. Among the numerous shortcomings of stacking are reliability penalties, yield problems, proliferation of process steps, and a thermal-conductivity penalty because of the multiple and extensive layers of insulating material of poor thermal-conductivity properties. “Thermal budget” problems in fabrication are severe, and inadequate crystalline quality in all but the substrate layer is endemic, and inadequate planarity in the advancing free surface causes pattern-definition problems.
SUMMARY OF THE INVENTION
A totally new approach to realizing 3-D organization is described in the present patent application and in the companion three issued patents and the two pending applications listed above in the opening paragraph. The three patents cited there deal with the structures of devices and circuits for 3-D implementation, and the two pending applications deal with methods for realizing such structures. The invention of the present application involves the assembly of sets of method and apparatus options for use at various stages and in various aspects of monocrystalline 3-D IC fabrication. Overall, the methods and apparatuses are amenable to start-to-finish automation.
Our monocrystalline 3-D approach brings together three technologies that are individually of intrinsic interest and are also applicable to mainstream IC fabrication. These are (1) sputter epitaxy, (2) real-time pattern generation, and (3) flash diffusion. The principles employed in implementing the three-way combination are described in Cases 4 and 5 cited above and will not be repeated here. Let us instead describe the three technologies one at a time and examine various options within each. After this we shall look in a similar way at other feature packets that are a part of our new technology.
Sputter Epitaxy. G. K. Wehner proposed this silicon-growth method in 1959 and demonstrated its feasibility in the late 1980s. The procedure he described is this: Form a mercury plasma in a Pyrex chamber in the vicinity of a doped-silicon “target” that is biased several hundred volts negative with respect to the plasma potential. The resulting mercury-ion bombardment of the target removes atoms from it via the sputtering mechanism. These ejected atoms display roughly a cosine distribution, and travel in straight lines in the reduced-pressure environment, held at about one millitorr. After an initial brief transient period, the atoms removed accurately replicate the target composition, because of a mass-conservation principle demonstrated by Wehner. The escaping atoms, silicon and dopant alike, are intercepted by a substrate that faces the target and that becomes the growth sample after deposition commences.
A crucial insight was delineated in 1959 by Wehner. By merely applying a small negative bias to the sample, less than 30 V in the silicon case, one can keep the sample under mild ion bombardment. This bias value is below the silicon sputtering threshold, and hence does not remove atoms from the sample, but does impart energy to atoms on the surface, permitting them to move readily across the surface in search of favorable crystallographic sites at which to lodge, thus contributing to the growth process. In conventional epitaxial procedures such as growth from the vapor, surface-atom energy is supplied by raising growth-sample temperature, typically well above 1000 C Wehner reasoned that sputter epitaxy would permit him to drop sample temperature significantly, and in his subsequent experimental verification he grew good-quality monocrystalline silicon at a sample temperature of 350 C . . . ! Boron contamination from the walls of his Pyrex system prevented him, however, from achieving the kind of growth-layer purity and the associated high values of carrier mobility he sought.
In the 1990s W. T. Peria made major improvements in the technology of sputter epitaxy, using a system he had devised and described publicly in previous years in connection with sputter epitaxy and also unrelated work. With these significant changes he has been able to achieve much higher sample purity and carrier mobility than before. First, he employed a stainless-steel chamber for reasons of strength and contamination avoidance. He chose a nonferromagnetic form of stainless steel so that a magnetic field can be created inside the chamber using external permanent magnets, for reasons described below. These magnets, the chamber, and several internal features can be seen in the schematic crosssectional diagram of FIG.
1
.
An important feature is plasma confinement, achieved here in part by the use of two anode-cathode pairs symmetrically disposed at the ends of what becomes a roughly rectangular volume of plasma. Its longest dimension of about 10 cm is fixed by the separation of the two anodes, and the next longest, by the cathode-and-shield length of about 8 cm. The two shields that can be seen facing each other in
FIG. 1
each contain a thoriated-tungsten welding-rod cathode, and each cathode has an accompanying anode taking the form of a rectangular loop of refractory-metal wire, with its long sides parallel to the cathode and its plane about 2 cm from the cathode. Also, the thickness of the plasma is about 2 cm. In the present system, the support members for the target, not shown, are protected by a shield, also not shown, that protects them from ion bombardment and that is held at a potential approximating the nearly common potential of a nodes and plasma, a very convenient reference potential in a system such as this.
Peria chose xenon as the bombarding species because of its conveniently low ionization potential of about 12 V. He uses a xenon pressure of about one millitorr, though it can be varied about an order of magnitude in either direction if desired. The cathodes are heated to about 1600 C by the passage of current through them, which requires a current of about 16 A. The cathodes are biased about 24 V negative with respect to the reference potential, causing electrons emitted by the in candescent thoriated tungsten to be accelerated by the associated anode. Most of the emitted electrons pass through the anode loop, which can of course have shapes other than rectangular; with its energy of 24 eV, each electron is able to ionize a xenon atom when such a collision occurs.
The permanent magnets shown serve to improve plasma density and confinement by creating ma genetic lines of force approximately parallel to the plane defined by the cathodes, and generally normal to the cathode rods. This causes a given electron to describe a helical path of small radius about a line of force, in the process appreciably lengthening the electron's path as compared to a straight path. This in turn increases the probability that a given electron will encounter a neutral xenon atom and ionize it, and thus results in increased plasma dens
MacCrisken John E.
Warner, Jr. Raymond M.
Jaeger Hugh D.
Nguyen Nam
VerSteeg Steven H.
LandOfFree
Monocrystalline three-dimensional integrated-circuit technology does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Monocrystalline three-dimensional integrated-circuit technology, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Monocrystalline three-dimensional integrated-circuit technology will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2968231