JP2001308465A - Semiconductor element - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve the characteristics of a semiconductor device by controlling and suppressing the delay in light emitting life of a semiconductor element using a quantum dot and, in addition, by realizing such a configuration that can eliminate phonon bottle neck phenomena. SOLUTION: A self-formed quantum dot 2 containing no impurity is formed in contact with a p-type wetting layer.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a semiconductor device using quantum dots, which enables control and improvement of carrier recombination efficiency and recombination life and high-speed carrier injection.

[0002]

2. Description of the Related Art With the progress of semiconductor processes, nano-scale growth technology or fine processing technology has been used for the fabrication of semiconductor devices. A semiconductor device utilizing a quantum mechanical effect, such as an HBT (heterojunction b), as well as improving the degree of integration of an integrated circuit.
An ipolar transistor and a quantum well laser have been put to practical use.

There is great expectation for quantum dots in this field as the ultimate material utilizing the quantum mechanical effect. For example, in the field of semiconductor lasers, improvement of the characteristics of quantum well lasers has been pointed out. It has been proposed to use quantum dots as one of the means for solving the problem.

In addition, researches for realizing next-generation semiconductor devices using quantum dots, which is a new material, have been actively conducted, such as a proposal of a quantum dot memory utilizing the hole burning effect. .

[0005] The quantum dot structure is a box of potential that is extremely fine enough to give the carrier three-dimensional quantum confinement. Since the density of states becomes a delta function, the conduction band basis of one quantum dot is Only two electrons can enter the position.

The use of such quantum dots in the active region of a laser has the advantage that the interaction between electrons and holes and light can be made as efficient as possible, and is expected as a next-generation semiconductor device technology. But big.

As a technique for forming a quantum dot structure, there can be mentioned a high-level artificial processing technique which is an extension of the development of a fine processing technique. For example, a method using lithography using an electron beam, a mask / A method of forming a quantum dot structure at the top of the pyramid-shaped crystal stacked in the opening of the pattern, a method of forming a quantum dot structure at the bottom of a tetrahedral hole formed by etching in the opening of the mask pattern,
A method using a lateral growth at an initial stage of growth on a vicinal substrate, a scanning tunneling (STM) method.
Atomic manipulation methods and the like that apply a microscopic technique, and these methods have an advantage that the formation position of the quantum dot structure can be arbitrarily controlled due to the common feature of processing artificially.

As a new technique realized recently, a method of self-forming quantum dots (self-organization phenomenon) is known. Specifically, a lattice-mismatched system (strain In this case, a semiconductor of (system) is grown by vapor phase epitaxial growth. In this case, not a two-dimensional layer but a three-dimensional fine structure, that is, a quantum dot structure is self-formed.

The method based on the self-assembly phenomenon is easy to carry out, and it is possible to obtain quantum dots of extremely high uniformity, high number density and high quality as compared with the case of processing artificially. it can.

In recent years, there have been reports that semiconductor lasers and the like using the self-formed quantum dots have been realized, and the possibility of quantum dot devices is becoming a reality.

Among the self-organization phenomena of quantum dots, the most frequently used means is SK (Transki-Kr).
This is a means called an “astanov” mode. When a high-strain semiconductor is grown by vapor phase epitaxial growth, a two-dimensional film (wetting layer) grows in the initial stage of the growth, and when the supply of the raw material is further continued, the three-dimensional microstructure is eventually formed. It is a self-forming structure, ie, a quantum dot structure.

Various methods have been developed for controlling the self-formed quantum dots. A typical method is to cover a quantum dot grown in SK mode with a quantum well,
It is a technique of embedding, and it is considered that this technique is effective in controlling the wavelength of a quantum well, improving uniformity, and the like.

As described above, progress in the technology of forming quantum dots is truly remarkable. On the other hand, problems in practical use as a device are becoming clearer. The other problem is the phonon bottleneck problem.

FIG. 2 is a diagram for explaining the problem of carrier injection into quantum dots. At the time of carrier injection into a semiconductor quantum dot device, the charge neutral condition is satisfied as a whole, but individual quantum dots are viewed. If so, this is not always the case.

That is, since electrons and holes are separately and randomly injected, quantum dots which do not immediately contribute to light emission exist with a certain probability ([B] and [C] in FIG. 2). ]).

In these cases, in order to emit light,
One has to wait for the oppositely charged carriers to diffuse from the other quantum dots, which means that the emission lifetime of the entire system apparently becomes longer.

The apparent light emission lifetime depends on the probability distribution of the events shown in FIG. 2. As a result, it is difficult to predict the characteristics of the device, and the carrier waiting for light emission has a non-intermediate time. If trapped in the emission level, the luminous efficiency is reduced, and furthermore, in fact, it is considered that the possibility that holes having a large effective mass are re-emitted from the quantum dots is not high, as shown in FIG. [B] is considered to hardly contribute to light emission.

Now, regarding the problem of the phonon bottleneck phenomenon, the phonon bottleneck phenomenon indicates that carrier relaxation to discrete levels is suppressed by the law of conservation of energy.

A physical feature of the quantum dot is that the density of states function becomes a delta function, which is the physical property on which the quantum dot device is expected to surpass the conventional semiconductor device. However, on the contrary, it is thought that the emission of optical phonons necessary for relaxation of interlevel carriers is suppressed due to these physical properties, making it difficult for carriers to be injected into the ground level. This phenomenon is called a phonon bottleneck phenomenon. It is.

Although the degree to which the actual phonon bottleneck phenomenon impairs the characteristics of the quantum box device is being discussed at the academic level, the carrier relaxation between levels is slower than in a quantum well. Has been obtained (if necessary, “Physical Review B5
4, R5243, 1996 "), it is clear that it is necessary to suppress the phonon bottleneck phenomenon as much as possible in order to improve the characteristics of the quantum box device.

[0021]

SUMMARY OF THE INVENTION In the present invention, it is possible to control the delay of light emission lifetime in a semiconductor device using quantum dots.
It is intended to realize a configuration capable of suppressing the phonon bottleneck phenomenon and improving characteristics of the semiconductor device.

[0022]

According to the present invention, by doping a semiconductor layer in contact with a quantum dot in advance, it is possible to eliminate a problem caused by local breaking of a neutral charge condition at the time of carrier injection. It is fundamental to improve the characteristics of a semiconductor device using quantum dots by avoiding the phonon bottleneck phenomenon.

FIG. 1 is an explanatory view of a semiconductor device for clarifying the principle of the present invention. FIGS. 1A and 1B show a semiconductor device having a preferred structure for the present invention, and FIGS. Represents a semiconductor element for comparison with (A) and (B).

In (A) and (B), 1 is a p-type wet layer of self-formed quantum dots, 2 is a self-formed quantum dot, 3
Is an n-type quantum well layer in which self-formed quantum dots are embedded,
Reference numeral 4 denotes a self-formed quantum dot. The combination of the conductivity type and the semiconductor layer to be doped is arbitrary.

In the semiconductor device having the structure shown in the figure, since abundant carriers due to doping fill the level of the quantum dot, when a carrier having the opposite polarity enters the quantum dot, it can be seen in FIG. 2 [A]. And the problem caused by carrier injection is eliminated.

At the same time, the state where the carrier fills the level of the quantum dot means that no further relaxation between levels is required, and the problem of the phonon bottleneck is eliminated.

When the quantum dots are also doped,
It is conceivable that the doping may cause a new problem that the crystallinity of the quantum dots themselves deteriorates, but this is done by doping only the semiconductor layer in contact with the quantum dots and not doping the quantum dots themselves. You can do it.

The quantum dots in the semiconductor device according to the present invention can be formed by applying a semiconductor fine processing technique, or can be formed by, for example, SK-mode, that is, a self-organization phenomenon.

When the self-assembly phenomenon is used, if the wetting layer is doped, the wetting layer and the self-formed quantum dots are simultaneously grown during crystal growth and have substantially the same composition.
Since the wetting layer is in contact with all the quantum dots, carrier injection into the quantum dots can be performed with high efficiency, and the effects expected of the present invention can be sufficiently obtained.

In the case where a structure in which quantum dots are embedded in a quantum well layer is adopted, carrier injection from a potential close to the quantum dots is realized by doping the quantum well layers, whereby the quantum dots are injected into the quantum dots. Since the carrier injection efficiency increases, the effect expected by the present invention can be sufficiently obtained in this case as well.

Further, in the present invention, it is only necessary that one of pn carriers is sufficiently injected into the quantum dot, so that it is not necessary to dope all the semiconductor layers from the electrode side to the quantum dot. It suffices if the semiconductor layer in contact with is doped.

Therefore, there may be another non-doped semiconductor layer between the doped semiconductor layer and the semiconductor layer on the electrode side having the same polarity, whereby the semiconductor crystal is over-doped. It is possible to prevent the property from lowering.

Regarding the phonon bottleneck problem, the present invention is based on the fact that an electron whose effective mass is light and whose level spacing is wide is selected as a doping species, and therefore, an n-type dopant is selected as a doping species. Effect can be enhanced.

FIG. 1C shows the case of non-doping.
The problem already described with reference to FIG. 2 occurs, and FIG.
Represents the case where the entirety from the electrode 7 to the inside of the semiconductor layer 5 and the quantum dot 6 is doped. In this case, the crystallinity is reduced due to excessive doping, and the device characteristics are deteriorated.

By employing the means of the present invention, it is possible to control and improve the light emission lifetime and efficiency due to the local violation of the charge neutrality condition, and to avoid the phonon bottleneck phenomenon. A high performance semiconductor device can be realized. In practicing the present invention, any conductive material that can generate quantum dots without depending on the type of semiconductor material may be used.

[0036]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS To illustrate the structure of a semiconductor device according to the present invention, a GaAs buffer layer having a thickness of 450 nm and an Al _0.3 layer having a thickness of 100 nm are formed on a GaAs substrate.
A Ga _0.7 As layer, a GaAs layer having a thickness of 100 nm,
Quantum dot layer by SK mode, thickness 100 [nm]
GaAs layer, 100 nm thick Al _0.3 Ga
A stacked body of a _0.7 As layer and finally a 50 [nm] GaAs layer is formed.

To form the laminate, MBE (mo
This method can be realized by applying a linear beam epitaxy method and setting the film formation temperature to 520 ° C.

FIG. 3 is a diagram showing the dependence of the photoluminescence emission intensity on the excitation light intensity, and FIG. 4 is a diagram showing the dependence of the emission lifetime on the excitation light intensity. Also, the results of theoretical calculations regarding the presence or absence of doping are shown.

The vertical axis in FIG. 3 shows the photoluminescence emission intensity, the horizontal axis shows the excitation light intensity (W / cm ² ), and the vertical axis in FIG. , And the horizontal axis indicates the intensity of the excitation light (relative intensity).

From the experimental results shown, it can be seen that both the photoluminescence emission intensity and the emission lifetime depend on the excitation light intensity, because the number of carriers to be excited depends on the number of carriers. This is due to the fact that the event generation probabilities seen in FIG. 2 are different. For example, when the number of excited carriers increases, the event generation probability of [A] in FIG. 2 increases.

The above experimental results are in good agreement with the theory, and the luminous efficiency is significantly improved by doping.
It is clear that the light emission lifetime does not fluctuate depending on the excitation light intensity, and the present invention is effective for improving the light emission intensity and controlling the light emission lifetime.

FIG. 5 is a diagram showing the result of calculating the rise of time-resolved photoluminescence with respect to the ground level of the quantum dot. Time is plotted on the horizontal axis, and light output is plotted on the vertical axis. The figure also shows the case where the present invention is implemented, that is, the case of n-type impurity doping, and the case where the present invention is not implemented, that is, the case of non-doping.

When the present invention is not used, the rise time is 2
In contrast, it takes about [nsec], whereas when the present invention is used, the rise delay is negligible, and the rise delay is caused by the electron phonon bottleneck phenomenon. It is clear that the present invention is effective in solving the phonon bottleneck problem.

FIG. 6 is a perspective view of an essential part of a stripe laser obtained by applying the present invention. In FIG.
-GaAs substrate, 12 is a 1.0 [μm] thick n-Al
The GaAs buffer / cladding layer 13 is an n-side SCH (s
eparte definition heater
14, an active layer, 15 a p-side SCH layer, and 16 a p-AlGaAs layer having a thickness of 1.0 μm.
The s cladding layer 17 is a 0.5 μm thick p-GaAs layer.
An s cap layer, 18 indicates a p-side electrode, and 19 indicates an n-side electrode.

FIG. 7 is a cutaway side view of a main part of a surface laser obtained by applying the present invention. In FIG.
s substrate, 22 is a mirror layer composed of InAs / GaAs,
23 is an SCH layer, 24 is an active layer, 25 is an SCH layer, 26
Are mirror layers made of InAs / GaAs, 27 and 28
Indicates an electrode.

In each of the lasers shown in FIGS. 6 and 7, the active layer 14 or the active layer 24 is constituted by a layer containing quantum dots formed by applying the present invention.

The present invention can be embodied in many forms, including the above-described embodiment, and will be exemplified below as additional notes.

(Supplementary Note 1) Semiconductor quantum dots that do not contain impurities (for example, self-formed quantum dots 2 or 4: see FIG. 1) are semiconductor layers that contain impurities (for example, p-type wet layer 1 of self-formed quantum dots or self-formed). A semiconductor element formed in contact with an n-type quantum well layer 3 in which quantum dots are embedded (see FIG. 1). (1)

(Supplementary Note 2) A semiconductor device characterized in that a semiconductor quantum dot containing no impurity is formed in contact with a quantum well layer containing an impurity. (2)

(Supplementary Note 3) A semiconductor layer containing an impurity in contact with a semiconductor quantum dot containing no impurity, and a semiconductor layer containing no impurity is interposed between an electrode having the same polarity as the semiconductor layer. Semiconductor device. (3)

(Supplementary Note 4) A semiconductor element characterized in that a semiconductor layer containing an impurity in contact with a semiconductor quantum dot containing no impurity is an n-type.

(Supplementary Note 5) A semiconductor characterized in that island-shaped semiconductor quantum dots which are generated by self-organization and do not contain impurities are formed in contact with a semiconductor layer which is a wetting layer containing impurities. element.

[0053]

In the semiconductor device according to the present invention, the semiconductor quantum dots containing no impurities are formed in contact with the semiconductor layer containing the impurities.

By adopting the above configuration, it is possible to control and improve the light emission lifetime and efficiency due to the local violation of the charge neutrality condition, to avoid the phonon bottleneck phenomenon, and to obtain a high performance. A semiconductor element can be realized.

[Brief description of the drawings]

FIG. 1 is an explanatory diagram of a semiconductor device for clarifying the principle of the present invention.

FIG. 2 is a diagram illustrating a problem of carrier injection into a quantum dot.

FIG. 3 is a diagram relating to excitation light intensity dependence of photoluminescence emission intensity.

FIG. 4 is a diagram relating to excitation light intensity dependence of emission lifetime.

FIG. 5 is a diagram showing the result of calculating the rise of time-resolved photoluminescence with respect to the ground level of quantum dots.

FIG. 6 is a perspective view of a principal part of a stripe laser obtained by applying the present invention.

FIG. 7 is a cutaway side view of a main part of a surface laser obtained by applying the present invention.

[Explanation of symbols]

 REFERENCE SIGNS LIST 1 p-type wet layer of self-assembled quantum dots 2 self-assembled quantum dots 3 n-type quantum well layer embedded with self-assembled quantum dots 4 self-assembled quantum dots

Claims (3)

[Claims] 2. A semiconductor device comprising: a semiconductor quantum dot containing no impurity formed in contact with a semiconductor layer containing an impurity. 2. A semiconductor device, wherein a semiconductor quantum dot containing no impurity is formed in contact with a quantum well layer containing an impurity. 3. A semiconductor comprising: a semiconductor layer containing an impurity in contact with a semiconductor quantum dot containing no impurity; and a semiconductor layer containing no impurity interposed between electrodes of the same polarity as the semiconductor layer. element.