JPH1022353A - Method for inspecting epitaxial semiconductor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a method for easily inspecting and judging crystallization of an epitaxial semiconductor expressed by an N type Alx Ga1-x As (0<=x<=0.35) in which Si is doped at a high concentration of 2.5×10<18> cm<-3> or more. SOLUTION: In the method, inspection is carried out over crystallization of an epitaxial semiconductor expressed by an N type Alx Ga1-x As (0<=x<=0.35) in which Si is doped at a high concentration of 2.5×10<18> cm<-3> or more. In this case, a ratio (I'/I0 ) of a peak intensity I' indicative of maximum of intensities present in a range of above 0.3eV and below 1.0eV on a lower energy side of the spectrum from the band end peak position to a band end peak intensity I0 in a photoluminescence spectrum of the eptitaxial semiconductor measured at a temperature of liquid nurogen (77K) is used as an index.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an n-type Al _x doped with Si at a high concentration used for manufacturing various electronic devices such as high electron mobility transistors and optical devices.
The present invention relates to a method for inspecting a Ga _1-x As (0 ≦ x ≦ 0.35) epitaxial semiconductor.

[0002]

2. Description of the Related Art In an epitaxial semiconductor used for a high-mobility heterojunction transistor, the conductance generally increases as the carrier concentration increases, and the performance increases as the crystallinity increases. Therefore, a semiconductor having a carrier concentration as high as possible and high crystallinity has been required. Here, high crystallinity of a certain semiconductor layer means that there are few intrinsic defects in the semiconductor layer, and when heat is applied to the semiconductor layer, the influence of heat on the semiconductor layer close to the semiconductor layer is small. That means. Further, when heat is applied to the semiconductor layer, a small influence of the heat on the semiconductor layer adjacent to the semiconductor layer is referred to as a high thermal stability of the semiconductor layer.

The concentration of intrinsic defects in a semiconductor layer corresponds to the concentration obtained by subtracting the carrier concentration from the impurity doping concentration in the semiconductor layer. However, n-type Al _x Ga _{1 -x} As (0 ≦
x ≦ 0.35) In the epitaxial semiconductor, Si
When doping is performed, in the region of 2.5 × 10 ¹⁸ cm ⁻³ or more, the proportional relationship is broken depending on the growth conditions, and when the amount of impurity added is increased, the carrier concentration is saturated without increasing from a certain point. Tend. It is known that, when x is 0.2 or more, the donor generally forms two kinds, a normal shallow donor level and a deep level called a DX center. The sum refers to the carrier concentration.

As described above, when Si is doped at a high concentration, it is difficult to accurately measure the carrier concentration of the semiconductor, so that the concentration of intrinsic defects cannot be obtained accurately. Therefore, after assembling a semiconductor manufactured under certain manufacturing conditions into a device, the device is evaluated,
The reality was to find appropriate semiconductor manufacturing conditions by trial and error.

Further, Al doped with Si at a high concentration
_{In x} Ga _1-x As (0 ≦ x ≦ 0.35) semiconductors, a large number of intrinsic vacancies related to Ga vacancies are generated at the same time, and the interdiffusion coefficient of the group III element increases abnormally through the defects. (Nao Nakajima, Applied Physics, Vol. 54, No. 11, 1167 (1985)). For example, in an AlAs / GaAs superlattice doped with Si at a high concentration, if the Si concentration exceeds 1 × 10 ¹⁹ cm ⁻³ , interdiffusion of Al and Ga occurs due to thermal annealing, and the superlattice is easily broken. (L. Pave
si et al. J. Appl. Phys. 71,
2225 (1992)). Conventionally, the thermal stability of such a semiconductor layer has been evaluated by secondary ion mass spectrometry (SIMS) or tunnel effect electron microscope (TEM). It could take more than a few days.

[0006]

The object of the present invention is to provide:
The crystallinity of an epitaxial semiconductor represented by n-type Al _x Ga _1-x As (0 ≦ x ≦ 0.35) obtained by doping Si at a high concentration of 5 × 10 ¹⁸ cm ⁻³ or more can be easily determined. It is an object of the present invention to provide an inspection method which can perform the inspection.

[0007]

In view of these problems, the present inventors considered that Al _x Ga doped with Si at a high concentration.
As a result of intensive studies on _1-x As (0 ≦ x ≦ 0.35) epitaxial semiconductors, n-type Al _x Ga _1-x As (0 ≦ x ≦ ³⁾ having a carrier concentration of 2.5 × 10 ¹⁸ cm ⁻³ or more 0.
35) It has been found that the ratio of I ′ to I ₀ in the PL spectrum at 77 K of the semiconductor increases with an increase in the amount of impurity added, and that the ratio sharply increases in the carrier concentration saturation region, leading to the present invention.

That is, the present invention relates to [1] 2.5 × 10
N-type Al _x G doped with ¹⁸ cm ^-3 or more of Si
a _1-x As (where 0 ≦ x ≦ 0.35) where a band in a photoluminescence spectrum of an epitaxial semiconductor measured at a liquid nitrogen temperature (77 K) is determined by a method of testing the crystallinity of the epitaxial semiconductor. 0.3 eV or more and 1.0 to the lower energy side from the band edge peak position with respect to the edge peak intensity (hereinafter referred to as I ₀ ).
The ratio (I '/ I) of the peak intensity (hereinafter, referred to as I') indicating the largest peak among the peaks existing in the range of eV or less.
_The present invention relates to a method for inspecting the crystallinity of an epitaxial semiconductor using ₀ ) as an index.

Further, according to the present invention, when the value of [2] I ′ / I ₀ is 5 or less, the crystallinity of the semiconductor is judged to be good,
If the ratio exceeds 5, the crystallinity of the semiconductor is determined to be defective. Further, the present invention relates to [3] n-type A
Inspection of the crystallinity of the epitaxial semiconductor according to [1] or [2], wherein the epitaxial semiconductor represented by l _x Ga _1-x As (0 ≦ x ≦ 0.35) is manufactured on a GaAs single crystal substrate. Pertains to the method. Here, for measurement of a photoluminescence spectrum (hereinafter, sometimes referred to as a PL spectrum), excitation light having energy equal to or higher than the band gap energy of a target semiconductor layer must be used.

[0010]

Next, the present invention will be described in detail.
The method for inspecting an epitaxial semiconductor according to the present invention is 2.5 ×
N-type Al doped with Si to 10 ¹⁸ cm ^-3 or more
_{In a} method for testing the crystallinity of an epitaxial semiconductor represented by _x Ga _1-x As (where 0 ≦ x ≦ 0.35), a photoluminescence spectrum of the epitaxial semiconductor is measured at a liquid nitrogen temperature (77 K), A peak showing the largest peak among peaks present in a range of 0.3 eV to 1.0 eV on the lower energy side from the band edge peak position with respect to the band edge peak intensity (hereinafter sometimes referred to as I ₀ ). It is characterized in that a ratio (I ′ / I ₀ ) of intensity (hereinafter, sometimes referred to as I ′) is obtained, and this value is used as an index. Here, the band edge peak intensity refers to the intensity of the photoluminescence peak corresponding to the energy substantially corresponding to the band gap energy of the semiconductor layer in the photoluminescence spectrum emitted from the semiconductor layer. The band edge peak position refers to a position of a light emission wavelength corresponding to the band gap energy of the semiconductor layer.

In order to obtain an epitaxial semiconductor having a high carrier concentration, a low defect density and excellent thermal stability, which can be used for manufacturing a high-speed electronic device and an optical device, the value of the ratio I '/ I ₀ is required. Is preferably 5 or less,
Since the crystal is exposed to a more severe environment depending on the operating conditions of the device, the ratio I ′ / I ₀ is more preferably 3 or less.

In addition, since an epitaxial semiconductor is formed on a lattice-matched substrate, it is possible to reduce the density of defects such as dislocations and to suppress lattice relaxation.
The l _x Ga _1-x As (0 ≦ x ≦ 0.35) epitaxial semiconductor is preferably formed on a GaAs single crystal substrate.

When the method for testing an epitaxial semiconductor of the present invention is applied to a semiconductor having a quantum well structure, the thermal stability of the quantum well structure can be easily studied. For example, as shown in Example 1, details are about 3 × 10
N-type Al _x Ga _{1 -x} A having a carrier concentration of ¹⁸ cm ^-3
s (0 ≦ x ≦ 0.35) and In _y Ga _1-y As (0.1
In ≦ y ≦ 0. 35) and the strained quantum well structure obtained by stacking, the ratio of I 'for I ₀ in PL spectra at 77K in said n-type _{Al x Ga 1- x As (0} ≦ x ≦ 0. 35) It can be seen that, in a crystal having a large value, the quantum well structure is easily deteriorated by thermal annealing. In other words, it can be seen that even if the carrier concentration is the same, in a semiconductor crystal having a large ratio of I ′ to I ₀ due to a difference in crystal growth conditions and an impurity addition amount, the quantum well structure easily undergoes deterioration.

[0014] These are because defects generated in n-type Al _x Ga _{1 -x} As (0 ≦ x ≦ 0.35) due to high concentration Si doping are the causes of saturation of the carrier concentration and at the same time, in the vicinity of the strained quantum well. It means that the diffusion of group 3 atoms is accelerated. As described above, by using the inspection method of the present invention, the quantum well stability can be easily studied.

[0015]

EXAMPLES The present invention will be described in more detail with reference to the following Examples, but it should not be construed that the invention is limited thereto. Example 1 An n-type Al _x Ga _{1 -x} As (0 ≦ x ≦ 0.
FIG. 1 shows a schematic cross-sectional view of an example of a heterojunction epitaxial semiconductor including 35). Fig. 1
5,000-nm thick n-type Al _0.2 Ga _0.8 As layer 1 (lower layer) on semi-insulating GaAs substrate 3 at 110 °
The thickness of the In _0.2 Ga _0.8 As layer 2,1000Å the thickness of the n-type Al _0.2 Ga _0.8 As layer 1 (upper layer) was grown in this order. The n-type Al _0.2 Ga _0.8 As layer 1 (lower layer) and the n-type Al _0.2 Ga _0.8 As layer 1 (upper layer) have 3 × 1
A Si doping of 0 ¹⁸ cm ^-3 was performed.

The carrier concentration of the n-type Al _0.2 Ga _0.8 As layer 1 (lower layer) of the heterojunction epitaxial semiconductor was determined by photo-Hall effect measurement. The sample for optical Hall effect measurement was obtained by cleaving the obtained epitaxial substrate into a square of about 8 mm, forming an In _0.2 Ga _0.8 As layer 2 and an n-type Al _0.2 G
a After removing _0.8 As layer 1 (upper layer),
A der Pauw method (clover leaf type) pattern was formed and fabricated. After irradiating the sample with white light at 77K, the light Hall effect was measured. As the irradiation time of the white light is increased, the electron concentration stops increasing, and a point appears to be saturated. This point corresponds to the case where electrons are emitted from all DX centers in the semiconductor by irradiation of white light,
The electron concentration at this time is the carrier concentration. That is,
The carrier concentration in the semiconductor corresponds to a saturation value of the electron concentration. As a result, n of the heterojunction epitaxial substrate
The carrier concentration of the type Al _0.2 Ga _0.8 As layer 1 is 3 × 1
0 ¹⁸ cm ^-3 , which was consistent with the set value due to doping.

In addition, In of the obtained epitaxial substrate
_0.2 Ga _0.8 As layer 2 and n-type Al _0.2 Ga _0.8 As
The layer 1 (upper layer) was removed, and the PL spectrum of the n-type Al _0.2 Ga _0.8 As layer 1 (lower layer) was measured. The spectrum is shown in FIG. The PL spectrum was measured at 77 K under the excitation of an argon ion laser (wavelength: 514.5 nm) in the wavelength range of 650 nm to 1100 nm. FIG.
Band edge peak (peak near 700 nm)
The ratio of the intensity (I ′) of the peak near 940 nm to the intensity (I ₀ ) was (I ′ / I ₀ ) = 1.05.

The obtained heterojunction epitaxial semiconductor containing n-type Al _0.2 Ga _0.8 As of In _0.2 Ga _0.8 A
FIG. 3 shows the PL spectrum of the s-strained quantum well. The PL spectrum was measured at 77 K in the wavelength range of 800 nm to 1100 nm under argon ion laser excitation. From the PL spectrum of the In _0.2 Ga _0.8 As strained quantum well shown in FIG. 3, it can be seen that the peak intensity is strong, the half width is small, and a good strained quantum well is formed.

On the other hand, the n-type Al _0.2 Ga _0.8
The heterojunction epitaxial semiconductor containing As is thermally annealed at 600 ° C. for 100 minutes in a hydrogen atmosphere to obtain In.
The PL spectrum of the _0.2 Ga _0.8 As strained quantum well was similarly measured. FIG. 4 shows the results. No change is observed between the obtained PL spectrum and the PL spectrum before thermal annealing shown in FIG. Thus before and after thermal annealing, the heterojunction epitaxial semiconductor an In _{0. 2} G
Since a _0.8 change in PL spectra of As strained quantum well is not observed, an In _0.2 Ga _0.8 As strain effect of n-type Al _0.2 Ga _{0. 8} As having a high carrier concentration in the quantum well it can be seen that no.

As described above, the properties of n-type Al _0.2 Ga _0.8 As doped with Si at a high concentration and its In
The effect on the _0.2 Ga _0.8 As strained quantum well is (I ′ / I
Similar results were obtained when ( ₀ ) was 5 or less. on the other hand,
When (I ′ / I ₀ ) exceeds 5, remarkable deterioration in crystallinity as shown in Comparative Example 1 below was observed.

COMPARATIVE EXAMPLE 1 A heterojunction epitaxial semiconductor having the same structure as the heterojunction epitaxial semiconductor in Example 1 and having a doping amount of Si of n-type Al _0.2 Ga _0.8 As doubled (6 × 10 ¹⁸ cm ⁻³ ). The same evaluation as in Example 1 was performed on the semiconductor.

The carrier concentration of the n-type Al _0.2 Ga _0.8 As layer 1 (lower layer) of the heterojunction epitaxial substrate was determined by measuring the optical Hall effect. The sample for measuring the optical Hall effect was processed in the same procedure as in Example 1, and the optical Hall effect of the sample was measured in the same manner. The carrier concentration of the n-type Al _0.2 Ga _0.8 As heterojunction epitaxial substrate are the same 3 × 10 ¹⁸ cm ^-3 and the carrier concentration of the n-type Al _0.2 Ga _0.8 As layer 1 of Example 1 (lower layer), doped Was about half of the set value.

Obtained n-type Al _0.2 Ga _0.8 As layer 1
FIG. 5 shows the PL spectrum of (lower layer). The PL spectrum was measured under the same conditions as in Example 1, except that the In _0.2 Ga _0.8 As layer 2 was removed by etching. The ratio between the intensity of the band edge peak (peak near 700 nm) and the intensity of the peak near 940 nm shown in FIG. 5 was 20.

The obtained heterojunction epitaxial semiconductor containing n-type Al _0.2 Ga _0.8 As of In _0.2 Ga _0.8 A
FIG. 6 shows the PL spectrum of the s-strained quantum well. The PL spectrum was measured under the same conditions as in Example 1. P in FIG.
As is clear from the comparison with the L spectrum,
The spectrum shows n-type Al with increased Si doping
It appears the effect of _{0.2 Ga 0.8 As, In 0. 2} Ga 0.8 As
The peak intensity of the strained quantum well was about one-eighth of the peak intensity in FIG. 3, and a short wavelength shift and an increase in the half width at the peak position were observed.

Further, the obtained n-type Al _0.2 Ga _0.8 A
thermal annealing at 600 ° C. for 100 minutes in a hydrogen atmosphere to the heterojunction epitaxial semiconductor containing
The PL spectrum of the _0.2 Ga _0.8 As strained quantum well was measured in the same manner as in Example 1. FIG. 7 shows the result. The peak intensity of the In _0.2 Ga _0.8 As strained quantum well after annealing was about 約 of the intensity of the peak before annealing. Note that the PL spectrum of FIG.
It is normalized by the peak intensity of the _0.2 Ga _0.8 As strained quantum well. Figure 6 and in longitudinal thermal annealing As is apparent from a comparison of FIG. 7, In _0.2 Ga 0 heterojunction epitaxial semiconductor including n-type Al _0.2 Ga _0.8 As in Comparative Example _1.
A remarkable change was observed in the PL spectrum of the 8As strained quantum well. The photoluminescence peak of the In _0.2 Ga _0.8 As strained quantum well is further shifted to the shorter wavelength side than before the thermal annealing, an increase in the half width and a decrease in the photoluminescence intensity are observed.
In, shows that lattice relaxation occurs in the interdiffusion, and In _0.2 Ga _0.8 As to Al.

As shown in Comparative Example 1, n-type Al _0.2 Ga _0.8 As with an increased amount of impurity added by using the method for inspecting an epitaxial semiconductor of the present invention was used in Example 1.
Has the same carrier concentration as n-type Al _0.2 Ga _0.8 As, but has an heterojunction with In _0.2 Ga _0.8 As.
It can be seen that the effect on the environment is large. It is predicted that it is difficult to maintain the shape of the quantum well and the steepness of the interface by using n-type Al _0.2 Ga _0.8 As in Comparative Example 1.
High-speed electronic devices, semiconductor lasers, light-emitting diodes,
It can be seen that it is difficult to give sufficient characteristics to various optical devices such as photodiodes.

On the other hand, the epitaxial half of the present invention
By using the conductor inspection method, it becomes clear from the first embodiment.
As expected, n-type Al doped with high concentration of Si_x
Ga _1-xAs (0 ≦ x ≦ 0.35) epitaxial substrate
Low defect density and little effect on other layers
N-type Al_xGa_1-xJudge As (0 ≦ x ≦ 0.35)
It turns out that it is possible to do. Therefore,
By using Ming's epitaxial semiconductor inspection method
With high carrier concentration, excellent in performance and reliability
Type Al_xGa_1-xHigh speed including As (0 ≦ x ≦ 0.35)
Electronic devices and semiconductor lasers and light emitting diodes,
Epitaxial for various optical devices such as photodiodes
It makes it easy to manufacture substrates industrially.
is there.

[0028]

The inspection method of the epitaxial semiconductor of the present invention
2.5 × 10 ¹⁸cm^-3More than high
N-type Al doped with Si_xGa_1-xA
epitaxial semiconductor represented by s (0 ≦ x ≦ 0.35)
The crystallinity of the body can be easily determined. The inspection method
Has high carrier concentration, low defect density and high speed performance
At the same time, high electron mobility with excellent thermal stability and reliability
Various electronic devices such as heterojunction transistors and light
Useful for device manufacturing.

[Brief description of the drawings]

FIG. 1 n-type Al _x Ga _{1 -x} As (0 ≦ x ≦ 0.35)
FIG. 2 is a schematic cross-sectional view of a heterojunction epitaxial substrate including:

FIG. 2 shows n-type Al _0.2 G at I ′ / I ₀ = 1.05.
a PL spectrum of _0.8 As layer.

FIG. 3 shows n-type Al _0.2 G at I ′ / I ₀ = 1.05.
4 is a PL spectrum before annealing of an In _0.2 Ga _0.8 As strained quantum well heterojunction with a _0.8 As.

FIG. 4 shows n-type Al _0.2 G at I ′ / I ₀ = 1.05.
10 is a PL spectrum after annealing of an In _0.2 Ga _0.8 As strained quantum well heterojunction with a _0.8 As.

FIG. 5 shows n-type Al _0.2 Ga at I ′ / I ₀ = 20.
PL spectrum of _0.8 As layer.

FIG. 6 shows n-type Al _0.2 Ga at I ′ / I ₀ = 20.
_0.8 As heterojunction to an In _0.2 Ga _0.8 As strain PL spectrum before annealing the quantum well.

FIG. 7 shows n-type Al _0.2 Ga at I ′ / I ₀ = 20.
_0.8 As and PL spectra after annealing of an In _0.2 Ga _0.8 As strained quantum well to the heterojunction.

[Explanation of symbols]

1 ... n-type Al _x Ga _1-x As (0 ≦ x ≦ 0.35) 2 ... In _y Ga _1-y As (0.1 ≦ y ≦ 0.35) 3 ... semi-insulating GaAs substrate

Claims (4)

[Claims] 1. A 2.5 × 10 ¹⁸ cm in ^-3 is formed by doping Si into n-type Al _x Ga _1-x As (wherein, 0 ≦ x ≦
In the method for testing crystallinity of an epitaxial semiconductor represented by 0.35), a band edge peak intensity (hereinafter, referred to as I ₀ ) in a photoluminescence spectrum of the epitaxial semiconductor measured at a liquid nitrogen temperature (77 K). A peak intensity (hereinafter, referred to as a peak intensity) indicating a maximum peak among peaks present in a range of 0.3 eV to 1.0 eV on the low energy side from the band edge peak position.
Notated as I '. A method for inspecting the crystallinity of an epitaxial semiconductor, wherein the ratio (I ′ / I ₀ ) is used as an index. 2. When the value of I ′ / I ₀ is 5 or less, the crystallinity of the semiconductor is determined to be good, and when the ratio exceeds 5, the crystallinity of the semiconductor is determined to be poor. 2. The method for testing the crystallinity of an epitaxial semiconductor according to claim 1, wherein: 3. When the value of I ′ / I ₀ is 3 or less, the crystallinity of the semiconductor is determined to be good, and when the ratio exceeds 3, the crystallinity of the semiconductor is determined to be poor. 2. The method for testing the crystallinity of an epitaxial semiconductor according to claim 1, wherein: 4. An n-type Al _x Ga _{1 -x} As (0 ≦ x ≦ 0.3
2. The method according to claim 1, wherein the epitaxial semiconductor represented by 5) is formed on a GaAs single crystal substrate.
4. The method for testing crystallinity of an epitaxial semiconductor according to 2 or 3.