Forming of quantum dots

Details Forming of quantum dots Forming of quantum dots

2001-08-03

2003-07-22

Mulpuri, Savitri (Department: 2812)

Semiconductor device manufacturing: process

Making device or circuit emissive of nonelectrical signal

C438S478000, C438S962000

Reexamination Certificate

active

06596555

ABSTRACT:

TECHNICAL FIELD
The present invention relates to the manufacturing of “quantum dots” of a first material in a second semiconductor material. More specifically, the present invention relates to the forming of quantum dots in a silicon substrate.
BACKGROUND OF THE INVENTION
Such quantum dots are, for example, described in document “Ge/Si self-assembled quantum dots grown on Si (100) in an industrial high-pressure chemical vapor deposition reactor” by C. Hernandez, Y. Campidelli, D. Simon, D. Bensahel, I. Sagnes, G. Patriarche, P. Boucaus and S. Sauvage, published in J. Appl. Phys., 86/2, 1999, 1145-1148. They are used in a great number of fields of application. The telecommunication field is considered hereafter as a non-limiting example.
In the telecommunication field, systems based on optical fibers are increasingly used. For this purpose, transmitters adapted to converting a potential difference into a light signal and receivers adapted to converting a light signal transmitted by an optical fiber into an electric signal are used. The transmitted and/or received light signal is generally located in a wavelength range between 1.4 and 1.5 &mgr;m.
Up to a recent period, combinations of materials of columns III and V of the periodic table of elements have been used to form such receivers. For example, gallium arsenide (AsGa) or phosphorus indide (InP).
It would be however preferred to use silicon-based materials for optical transmitters and receivers, but the use of silicon poses different problems. First, the energy gap of silicon between its valence and conduction bands is relatively small and the transitions are of “indirect” type. “Indirect” means that the passing of an electron from the valence band to the conduction band occurs in several jumps and not in a single jump as in the case of III-V compounds. Silicon is then almost impossible to use as an emitter, that is, a converter of electric power into light power. Indeed, due to the indirect character of the electronic transitions, said transitions are highly dissipative and only slightly emissive. Further, the relatively small energy band gap, on the order of 1.1 eV, corresponds to an emission of photons having a wavelength smaller than 1 &mgr;m, seldom used in the field of telecommunications.
It has thus been provided to improve the optical properties, that is, the emission and reception properties, of silicon by forming quantum dots therein.
FIG. 1
illustrates in a partial simplified cross-section view a quantum dot such as described, for example, in the above-mentioned article of the Journal of Applied Physics. Dot B is formed of a germanium island formed on a single-crystal silicon substrate S. Although it is theoretically desired to form a perfect cube having an edge smaller than 50 nm, a dome- or drop-shaped nanostructure B having a base L and a height h is formed in practice. In practice, such dots with a base L ranging between 20 and 50 nm and a height ranging between 6 and 30 nm can be achieved. Dot B is encapsulated in a single-crystal silicon layer.
In many applications, to obtain acceptable performances, for example a transceiver with a satisfactory emissivity/receptivity, it is desirable to be able to form several superposed planes, each containing several dots similar to that in FIG.
1
.
As discussed in the previously-mentioned article, the forming of domes or drops results from a mechanism of stress between crystal meshes having different, but relatively close dimensions, of two semiconductors. This so-called Stranski-Krastanow growth process has been shown to cause, for example, the forming of germanium drops on silicon based on various deposition processes including molecular epitaxies, low-pressure vapor phase chemical depositions, or vapor phase chemical depositions under strong vacuum.
More specifically, to form germanium quantum dots in silicon, an epitaxy by continuous injection of germane (GeH
4
) on a single-crystal silicon substrate is for example performed. Then, the few first deposited atomic layers form a layer having a regular but non planar surface. Due to the stress associated with the crystal lattice differences, the surface has a “rippling” shape of sinusoidal type. In other words, the upper surface of a germanium layer of a few atomic layers, formed on silicon, has regularly distributed bumps and holes. As the injection of germane continues, the crystallographic stress—deformations of the natural germanium lattice—cause the growth of drops of dots similar to those in FIG.
1
. This injection must be interrupted when the drops or dots have reached a desired dimension, before occurrence of a coalescence of the drops and the forming of a continuous layer containing dislocations.
An epitaxial growth of a silicon layer that encapsulates the germanium dots is then performed.
FIG. 2
illustrates, in a partial simplified cross-section view, the result of the repeated implementation of such a method. For example, three substantially horizontal planes of germanium drops (quantum dots)
21
encapsulated in silicon
22
have been formed on a silicon substrate
20
.
FIG. 3
illustrates in a partial simplified top view one of the planes of drops
21
.
It is desired to obtain dot densities as high as possible (≧10
10
cm
−2
) and a size distribution as fixed as possible. However, as very schematically illustrated in
FIGS. 2 and 3
, the current implementation of the Stranski-Krastanow stress growth method results in heterogeneous drop structures as concerns the drop distribution as well as their size. To underline all the difficulties of implementation of this method, the main steps of the vapor phase epitaxy of germanium drops on a silicon substrate will be reminded.
First, the surface state of the silicon substrate has a significant role. It is currently believed that an optimal compromise between a perfect homogeneity and a certain distribution of defects has to be found.
The choice of the epitaxy conditions also appears to have to satisfy a compromise. Indeed, these conditions must be chosen so that the epitaxy is not too slow since, in this case, risks associated with the presence of inevitable impurities (brought in by the gas precursors and/or associated with the reactor leak rate) increase. However, if the germanium drop growth rate is too high, the process becomes difficult to control. Indeed, this growth must be precisely interrupted, as previously indicated.
Thus, an “optimal” temperature, which corresponds to a maximum “controllable” growth rate, that is, a rate as fast as possible to avoid the above-mentioned defects and sufficiently slow to enable interruption of the epitaxy in a precise manner at a given stage (for example, a few tens of monoatomic layers) is defined for given pressure, flow, and gas dilution conditions, in an epitaxy reactor of given type.
As an example, a typical sequence of implementation of a Stranski-Krastanow method at a temperature on the order of 650° C. (from 630 to 670° C.) and at a pressure ranging between 0.02 and 0.04.10
5
Pa (from 20 to 30 Torr) includes, starting from a single-crystal silicon substrate, the steps of:
injecting silane with a 10% dilution in hydrogen for substantially 5 s, which results in the growth of substantially from two to three single-crystal silicon monolayers (optional step);
injecting germane with a 10% dilution in hydrogen for substantially 30 s, which results in the growth of germanium “drops” of a height corresponding to a pile of some thirty atoms; and
reinjecting silane to form an encapsulation layer.
According to this conventional sequence, the silane and germane injections are successively performed at close flow rates. Hydrogen (the carrier gas) is injected at a rate of approximately 10 l/mn. The flow rate of the injected gases is generally maintained between 15 and 25 cm
3
/mn, typically on the order of 20 cm
3
/mn. This flow rate is chosen as previously explained to obtain a maximum controllable deposition rate.
SUMMARY OF THE INVENTION
An embodiment of the present invention provides a method of quantum dot grow

Affiliated with

Also associated with

Rating

Forming of quantum dots does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Forming of quantum dots, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Forming of quantum dots will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3020874