JP3453265B2 - Doping method of nitride-based compound semiconductor and manufacturing apparatus therefor - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for doping a nitride-based compound semiconductor used in a blue to green light emitting device and the like and an apparatus for manufacturing the same.

[0002]

2. Description of the Related Art In recent years, gallium nitride, indium gallium nitride, aluminum gallium nitride, and the like have attracted attention as nitride-based compound semiconductors that are promising for realizing compound semiconductor light-emitting devices such as blue to green laser devices. Mg is used as a p-type dopant and Si is generally used as an n-type dopant for a compound semiconductor. For example, in JP-A-4-321280, Si is used as an n-type dopant and M is used as a p-type dopant.
g is used. Further, as another p-type dopant, JP-A-6-209121 describes that an organometallic compound containing Cd, Zn, Be, Ca, Sr, Ba and the like as a p-type dopant is used in the MOCVD method. There is.

These dopant sources are Mg, Si, when a nitride compound semiconductor is grown by the MBE method.
It is introduced in the form of a solid raw material such as Zn and Cd, and is supplied in the form of heated molecules.

When the MOCVD method is used,
It is supplied in the form of an organic metal such as (Cp ₂ ) Mg (cyclopentadienyl magnesium), DEZn (diethyl zinc), DECd (diethyl cadmium), SiH ₄ (silane) or a gas.

FIG. 10 shows a conventional MBE device.
Reference numeral 101 is a Knudsen cell (hereinafter referred to as K cell), and 102 is a nitrogen radical cell. By enclosing a dopant source such as Mg in the K cell 1 and heating the dopant source appropriately, a molecular beam is generated and a dopant component is supplied. Also, the nitrogen radical cell 1
A nitrogen gas is radicalized from 02 to supply a nitrogen component. Further, as group III raw materials, simple metals such as Ga, Al and In, TMG (tetramethylgallium), TM
An organometallic compound such as A (tetramethylaluminum) is used to fill the K cell 103 and supply the group III element component. A thin film is deposited on the substrate 104 by the supplied dopant component, nitrogen component, and group III component.

[0006]

However, when a nitride compound semiconductor is crystal-grown, even if an undoped film is formed without supplying a dopant, the conductivity type is n-type and its carrier concentration is 10 ¹⁷ to 10 ²⁰ cm
It became ^-3 . This is due to nitrogen atom vacancies formed during crystal growth.

Therefore, when a p-type conductive film is obtained, excessive Mg doping is required to compensate for n-type carriers due to nitrogen atom vacancies and generate more effective p-type carriers. The doping is efficient. Therefore, the conventional dopant has a problem that the carrier concentration cannot be increased.

Also, when obtaining an n-type conductive film,
In the mechanism of n-type conduction by nitrogen atom vacancies, crystallographically,
Alternatively, there is a problem that it is electrically unstable.
Therefore, it is necessary to effectively replace Si, which is an n-type dopant, with Ga sites while reducing the number of nitrogen atom vacancies. The same problem occurs when Zn or Cd is used as the p-type dopant.

In the case of using both n-type and p-type dopants, nitrogen atom vacancies are reduced and the dopant is changed to G.
It was necessary to substitute the a site.

The object of the present invention is to solve the above mentioned problems,
It is an object of the present invention to provide a method for doping a nitride-based compound semiconductor that can be more efficiently doped, and an apparatus for manufacturing the same.

[0011]

A method for doping a nitride compound semiconductor according to the present invention uses a nitride compound composed of a dopant atom and a nitrogen atom as a p-type dopant source or an n-type dopant source . Semiconductor
It is characterized in that it is grown by a sagittal epitaxy apparatus.

Further, Be _{3 is} used as the p-type dopant source.
At least 1 in N ₂ , Mg ₃ N ₂ , Zn ₃ N ₂ , and Cd ₃ N ₂ .
It is characterized by using one.

Si _{3 is} used as the n-type dopant source.
At least one of N ₄ and Ge ₃ N ₄ is used.

The apparatus for producing a nitride compound semiconductor of the present invention is a molecular beam epitaxy apparatus having at least two nitrogen radical cells and at least one.
It is characterized in that a dopant is enclosed in two nitrogen radical cells.

The dopant to be enclosed in the radical cell is at least one of Be, Mg, Zn, Cd, Si and Ge.

[0016]

DETAILED DESCRIPTION OF THE INVENTION

(Embodiment 1) A K-cell of an MBE device was filled with Mg ₃ N ₂ as a dopant source, and a p-type GaN epitaxial growth film with a film thickness of 1 μm was grown on a SiC substrate at a substrate temperature of 700 ° C. The growth conditions are as follows: Ga beam intensity 3 × 10 ⁻⁷ Torr, nitrogen radical cell input power 5
00W, nitrogen pressure in chamber of growth chamber 7 × 10 ^-5
Torr, growth rate 0.5 μm / h. As a comparative example, Mg was enclosed as a dopant source, and the film was formed under the same conditions under other conditions.

Photoluminescence of these two types of p-type GaN epitaxial growth films was measured, and the spectrum of the results is shown in FIG. The photoluminescence spectra (PL spectra) of both are almost the same.
It was confirmed that ₃ N ₂ functions as a p-type dopant.

FIG. 2 shows the relationship between the Mg concentration and the carrier concentration in the film evaluated by SIMS. Mg concentration is M
It is changed by raising and lowering the temperature of the K cell of g and Mg ₃ N ₂ . As seen in FIG. 2, when compared in the same carrier concentration, who who encapsulating Mg was shown enclosed Mg as dopant, Mg concentration in about one order of magnitude film has high, Mg ₃ N ₂ had higher doping efficiency. This is because when Mg is used as the dopant source, the number of nitrogen atom vacancies in the film is the same as that of the undoped film, or more than that of the undoped film due to the self-compensation effect. Therefore, the doping efficiency is lower when Mg is used as the dopant source. On the other hand, when Mg ₃ N ₂ is used as a dopant source, when Mg is replaced with a Ga site, nitrogen atoms bonded to Mg fill the nitrogen atom vacancies to improve crystallinity and increase doping efficiency. .

(Embodiment 2) Cd ₃ N ₂ is enclosed as a dopant source in a K cell of an MBE apparatus, and a substrate temperature of 600 is used.
A p-type InGaN epitaxial growth film having a film thickness of 1 μm was grown on a SiC substrate under the condition of ° C. The growth conditions are: Ga beam intensity of 3 × 10 ⁻⁷ Torr, In beam intensity of 6 × 10 ⁻⁸ Torr, input power of nitrogen radical cell of 500 W, and nitrogen pressure of 7 × 10 in the chamber of the growth chamber.
^-8 Torr, growth rate 0.55 μm / h. As a comparative example, Cd was encapsulated as a dopant, and the film was formed under the same conditions under other conditions.

Photoluminescence of these two types of p-type InGaN epitaxially grown films was measured, and the spectrum of the results is shown in FIG. The photoluminescence spectra (PL spectra) of both are almost the same although the emission spectra of the impurity levels are slightly different, and C
It was confirmed that d ₃ N ₂ functions as a p-type dopant.

(Embodiment 3) Si ₃ as a dopant for an electron beam gun (hereinafter referred to as an E gun) of an MBE apparatus.
Encapsulating N ₄ and film thickness 1 μm under the condition of substrate temperature 600 ° C
Was grown on a SiC substrate. The growth conditions are: Ga beam intensity 3 × 10
^-7 Torr, In beam intensity 8 × 10 ^-8 Torr, input power of nitrogen radical cell 500 W, nitrogen pressure in chamber in growth chamber 7 × 10 ^-5 Torr, growth rate 0.5
7 μm / h. As a comparative example, Si was sealed in a K cell as a dopant source, and the film was formed under the same conditions under other conditions.

The PL spectra of photoluminescence of these two types of n-type GaN epitaxial growth films are shown in FIG.
Shown in. The PL spectra obtained by measuring the photoluminescence of these two types of n-type GaN epitaxial growth films are almost the same, and the carrier concentration of both is 3 ×.
It was almost the same as 10 ¹⁷ cm ^−3, and it was confirmed that Si ₃ N ₄ functions as an n-type dopant.

FIG. 5 shows the relationship between the Si concentration and the carrier concentration in the film evaluated by SIMS. Si concentration is S
It is changed by changing the temperature of the K cell for i and the current value of the E gun for Si ₃ N ₄ . As seen in Figure 5,
When comparing with the same carrier concentration, the Si concentration in the film was about twice higher when Si ₃ N ₄ was used as the dopant. When Si is used as a dopant, this is obtained as the sum of the portion due to Si and the nitrogen atom vacancies in the film as the carrier concentration, but in the case of Si ₃ N ₄ , the portion due to the nitrogen atom vacancies is This is because it decreases.

(Embodiment 4) Next, a higher purity II
A method of producing a nitride of a group element in-situ and using it as a dopant will be described. The MBE apparatus used in this embodiment is shown in FIG. This MBE device has a plurality of nitrogen radical cells, and at least one of the nitrogen radical cells is a cell 61 used as a nitrogen source of a nitride compound semiconductor (hereinafter referred to as a film forming nitrogen radical cell). And a nitrogen radical cell 62 (hereinafter referred to as a dopant nitrogen radical cell) in which a dopant is enclosed. The film forming nitrogen radical cell 61 requires a large amount of nitrogen radicals, and therefore requires a large input power of about 500 W. On the other hand, in the nitrogen radical cell for dopant, p-BN
For example, Mg is enclosed as a p-type dopant inside the manufactured aperture, and since it is a nitrogen radical cell for impurity doping, it operates with an input power of about 200 W, which is smaller than that of the film forming nitrogen radical cell. In the present embodiment, the p-type dopant Mg is enclosed as the nitrogen radical cell for the dopant, but the same effect can be obtained by enclosing the n-type dopant.

Further, in the apparatus shown in FIG. 6, one nitrogen radical cell 61 for film formation, one nitrogen radical cell 62 for p-type dopant, and n-type dopant (for example, Si or Si ₃ N ₄ etc.) The K cell 63 for use is used. It has a K cell 64 capable of encapsulating a dopant such as In for supplying a Group III element, and a K cell 65 for supplying tetramethylgallium (TMG) to generate a molecular beam.

Further, as shown in FIG. 7, a nitrogen radical cell 71 for n-type dopant is provided instead of the K cell for n-type dopant, and Si and Ge are used as the n-type dopant source in the nitrogen radical cell 71 for n-type dopant. A device in which various dopants are enclosed may be used.

By encapsulating the dopant raw material in the dopant nitrogen radical cell 62 and generating nitrogen plasma,
Group II element nitrides are prepared and doped in-situ. When using this method, the purity of Group II elements is
Almost equal to the purity of Group II element, the doping efficiency was improved.

Using this MBE apparatus, a 1 μm-thick p-type GaN epitaxial growth film using Mg as a p-type dopant source is grown on a SiC substrate. At this time, the Ga beam intensity was 3 × 10 ⁻⁷ Torr, the input power of the nitrogen radical cell was 500 W, the nitrogen pressure in the chamber in the growth chamber was 7 × 10 ⁻⁵ Torr, and the growth rate was 0.
5 μm / h. For comparison, the case of using Mg ₃ N ₂ shown in Embodiment 1 and the in-situ of this embodiment are compared.
FIG. 8 shows the relationship between the concentration of Mg and the carrier concentration evaluated by SIMS when using Mg ₃ N ₂ produced by u. The Mg concentration in the GaN film is controlled by changing the input power and the nitrogen flow rate of the K cell for Mg ₃ N ₂ and the nitrogen radical cell for dopant. When Mg ₃ N ₂ produced in-situ by the nitrogen radical cell for dopant was used, the purity was higher, so that the doping efficiency was further improved.

(Fifth Embodiment) M shown in the fourth embodiment
Using a BE device, Si was enclosed in a nitrogen radical cell for dopant and an n-type GaN epitaxial growth film was grown. The conditions at this time are: Ga beam intensity 3 × 10
^-7 Torr, input power of nitrogen radical cell is 500W,
Nitrogen pressure in chamber of growth chamber 7 × 10 ^-5 Tor
r, the growth rate is 0.5 μm / h. As a comparison, the case where Si ₃ N ₄ is used as the dopant source shown in the first embodiment and the S produced in-situ of the present embodiment.
FIG. 8 shows the relationship between the Si concentration and the carrier concentration as a result of evaluation by SIMS in the case of using i ₃ N ₄ . G
The concentration of Si in the aN film is controlled by changing the input power and the nitrogen flow rate of the K cell for Si ₃ N ₄ and the nitrogen radical cell for dopant. When Mg ₃ N ₂ produced in-situ is used, the number of nitrogen atom vacancies can be further reduced as in the fourth embodiment.

[0030]

According to the present invention, since the nitrogen atom vacancies can be reduced by nitrogen combined with the doping material, the doping efficiency of the III-V group nitride compound semiconductor can be increased. Therefore, it is possible to improve the characteristics of the semiconductor light emitting device formed of the nitride compound semiconductor.

[Brief description of drawings]

FIG. 1 is a diagram showing a PL spectrum of a GaN epitaxial growth film doped with Mg.

FIG. 2 is a diagram showing a correlation between Mg concentration and carrier concentration.

FIG. 3 is a diagram showing a PL spectrum of a GaN epitaxial growth film doped with Cd.

FIG. 4 is a diagram showing a PL spectrum of a Si-doped GaN epitaxial growth film.

FIG. 5 is a diagram showing a correlation between Si concentration and carrier concentration.

FIG. 6 is a schematic diagram of an MBE device according to the present invention.

FIG. 7 is a schematic diagram of an MBE apparatus according to the present invention.

FIG. 8 is a diagram showing a correlation between Mg concentration and carrier concentration.

FIG. 9 is a diagram showing a correlation between Si concentration and carrier concentration.

FIG. 10 is a schematic diagram of a conventional MBE device.

[Explanation of symbols]

61 Nitrogen radical cell for film formation 62 Nitrogen radical cell for p-type dopant 63 K cell for n-type dopant 64, 65 K cells 71 Nitrogen radical cell for n-type dopant

Claims (5)

(57) [Claims] 1. Molecular beam epitaxy of a nitride-based compound semiconductor
A method for doping a nitride-based compound semiconductor, characterized in that a nitrogen compound composed of a dopant atom and a nitrogen atom is used as a p-type dopant source or an n-type dopant source in the method of growing with a See apparatus . 2. Be ₃ N ₂ as the p-type dopant source,
The method for doping a nitride compound semiconductor according to claim 1, wherein at least one of Mg ₃ N ₂ , Zn ₃ N ₂ and Cd ₃ N ₂ is used. 3. Si ₃ N ₄ as the n-type dopant source,
The method for doping a nitride-based compound semiconductor according to claim 1, wherein at least one of Ge ₃ N ₄ is used. 4. A molecule for growing a nitride compound semiconductor.
An apparatus for producing a nitride-based compound semiconductor, comprising at least two nitrogen radical cells in a line epitaxy apparatus , and encapsulating a dopant in at least one nitrogen radical cell. 5. The dopant is Be, Mg, Zn,
The device for manufacturing a nitride-based compound semiconductor according to claim 4, which is at least one of Cd, Si, and Ge.