JPS63278282A - Photoconductive photodetector - Google Patents

Abstract

PURPOSE:To reduce a dark current at a photoconductive device composed of a mutually laminated body of AlInAs and GaInAs layers by a method wherein a density of free electrons in the GaInAs layers is reduced while the AlInAs layers are doped with a p-type impurity. CONSTITUTION:AlInAs layers 12 doped with a p-type impurity and impurity-undoped GaInAs layers 13 are laminated alternately on an InP substrate 11 by an organometallic thermal decomposition method, a liquid-phase epitaxial growth method, a molecular beam epitaxial growth method or the like. As the p-type impurity, Zn, Be, Mg, Mn or the like is to be used. During this process, the concentration of the p-type impurity to be doped inside the AlInAs layers 12 is set at more than 10<16> cm<-3>; a layer thickness of each AlInAs layer 12 is set at more than 50 Angstrom ; the layer thickness of each GaInAs layer is set at less than 0.5 mum. Under these conditions, the GaInAs layers are depleted; a carrier contributing to the conduction is nearly zero when the layers are not irradiated with the light. The number (n) of layers to be laminated is decided in such a way that a product nt of the thickness (t) of the GaInAs layers 13 multiplied by the number (n) of layers to be laminated is more than 0.5 mum. The light which is incident under this condition is absorbed efficiently into the GaInAs layers. Electrodes 14, 14 are formed on this mutually laminated body by, e.g., an evaporation method.

Description

DETAILED DESCRIPTION OF THE INVENTION (1) Technical Field The present invention relates to improving the sensitivity of a photoconductive type light receiving element.

A photoconductive type light-receiving element is a light-receiving element that can detect the intensity of irradiated light by utilizing the fact that the resistance of a material changes when irradiated with light.

The materials used differ depending on the wavelength of the target light.

CdS cells have been widely used as photoconductive cells for a long time. This element is easy to use because it has high sensitivity to visible light. For visible light, in addition to this, CdS
e, ZnO, etc.

The most frequently used element for photodetection is S.
i is a photodiode. In the case of a photodiode, photocurrent is generated by irradiating it with light.

For this reason, a pn junction is essential.

However, the photoconductive effect is a change in the resistance of a material upon irradiation with light. This is a bulk property of the material. Therefore, a pn junction is not essential for a photoconductive effect element.

Semiconductors such as Si, Ge, and GaAs are P
type and n-type semiconductors can be easily produced. PN junctions are also easy to make.

However, many I-VI group semiconductors are difficult to form pn junctions with. In this way, it is not possible to make a photodiode for something that cannot make a pn junction. Therefore, we decided to create an element that takes advantage of the photoconductive effect.

The wavelength of light that can be detected differs depending on the band gap.

In general, ■-■ group compound semiconductors have a wide range of possible band gaps, so there are materials that are sensitive to light of any wavelength.

For example, for infrared light, PbS, InSb
.

Photoconductive cells such as CdHgTe and Ge are used.

The CdS cells mentioned at the beginning have high sensitivity, but their operating speed is extremely slow. However, photoconductive cells generally have slow operating speeds.

In addition to this, photoconductive type light receiving elements have a large dark current.
There is a drawback.

The present invention aims to reduce dark current in an AlInAs/GaInAs stacked photoconductive type device.

(a) Prior Art Conventionally, photoconductive light receiving elements have been manufactured using various semiconductor materials. For example, S, M.

“Physics of Sem1condu” by Sze
ctor Devices.

2nd edition PP 744-748 describes various photoconductive type light receiving elements.

The present invention relates to an improvement in a photoconductive light receiving element having a structure in which AJInAs layers and GaInAs layers are repeatedly laminated.

AI on the InP substrate! InAs layer and GaInA
This is made by growing multiple layers alternately with s-layers. A,eIn
A multilayer structure of As/GaInAs can be fabricated by molecular beam epitaxy (MBE), liquid phase epitaxy (LPE), or the like.

AlInAs mixed crystal and Ga InAs mixed crystal are InP
The mixed crystal ratio is determined by the condition of lattice matching with the substrate.

However, for simplicity, the mixed crystal ratio is omitted.

FIG. 2 is a schematic cross-sectional view of such a photoconductive type light receiving element.

On the InP substrate 11, an AlInAs layer 12, a GaI
It has an alternating multilayer structure including an nAs layer 13, an AlInAs layer 12, a GaInAs layer 13, . . . .

The bandgap of the AlInAs layer is higher than that of GaInAs.
wider than the bandgap of the layer.

The GaInAs layer is non-doped, and the AlInAs layer is n-doped.
Dope type impurities. Electrons donated from the n-type impurity move toward the GaInAs layer due to the difference in band gap.

The GaInAs layer is undoped. In other words, it is not intentionally doped with impurities. Since there are very few impurities, there is little scattering of electrons due to impurities. Since there is less electron scattering, electron mobility increases. This makes it possible to create a light-receiving element that responds quickly.

Devices in which the space in which impurities that emit electrons exist and the space in which electrons travel are separated in this way are used in high-speed MODFETs and the like that use two-dimensional electron gas.

(c) Problems to be Solved by the Invention However, photoconductive elements made of alternating AlInAs/GaInAs multilayers have the drawback of large dark current and, as a result, low sensitivity to light.

Dark current is the current that flows when the light receiving element is not exposed to light.

The reason for this is that even when no light is irradiated, there are
This is due to the presence of carriers that contribute to electrical conduction.

This point will be explained with reference to the band structure diagram in FIG. In this figure, the horizontal axis is the spatial coordinate in the vertical direction in the element of FIG. 2. The vertical axis is the energy of electrons and holes.

The upper curve ABCD is the lower end of the conduction band 1. The curve JK below this is the upper end of the valence band 2. Between conduction band 1 and valence band 2 is a band gap where no free carriers can exist.

Since this is a device in which the AlInAs layer 12 and the GaInAs layer 13 are alternately laminated, the band structures corresponding to the two alternate.

This structure is obtained because the AlInAs layer has a wider band gap than the GaInAs layer.

The straight line indicated by the dashed line is Fermi level 4.

In the figure, the "+" mark indicates an n-type impurity (donor) that has lost electrons and has become positively charged.

The n-type impurity is also applied to the AlInAs layer and the undoped GaI layer.
It also exists in the nAs layer.

AI! The n-type impurity 5 in the InAs layer forms a level just below the conduction band 1. It is a shallow impurity level.

The n-type impurity in the non-doped GaInAs layer also forms an impurity level just below the conduction band 1.

However, the bandgap is different and conduction band 1 is GaIn.
There is a pulse-like change in which the value is low for As and high for AlInAs. For this reason, GaInAs and AlI
The levels of n-type impurities 5 and 6 in nAs are different.

The level of n-type impurity 6 in the GaInAs layer is lower than the level of n-type impurity 5 in the AlInAs layer.

Although the GaInAs layer is non-doped, it is an n-type semiconductor due to impurities contained in the material. Since it is an n-type impurity that creates a shallow level, Fermi level 4 is located below the conduction band 1(7) t .

The A,/InAs layer is doped with n-type impurities and becomes an n-type semiconductor. Fermi level 4 is slightly above the center of the band gap.

In the vicinity of the center A and G of the GaInAs layer, the lower end of the conduction band is higher than the Fermi level 4. Since it is a semiconductor, this is a big deal.

However, both ends B and E of the band of the GaInAs layer.

In cases such as (2), the conduction band 1 is pushed downward.

Conduction band 1 is below Fermi level 4. Therefore, when the electrons leave the n-type impurity and enter the concave portions of conduction bands such as B, E, and I, the energy is lowered.

The band structure of the GaInAs layer is upwardly convex, and free electrons always exist at both ends B, E, and I.

This is surprising since it is below Fermi level 4, and exhibits conductivity similar to metal.

Even when no light is irradiated, B and E.

Free electrons 3 exist in the concave part of the conduction band, such as (2).

Therefore, if a voltage is applied, current will flow even if no light is irradiated.

In this way, the presence of free electrons at the edge portions at both ends of the GaInAs layer increases the dark current.

The lower the dark current, the better. This is because if the dark current is small, the ratio of the resistance value when no light is irradiated and when weak light is irradiated can be increased.

The presence of free electrons 3 causes dark current.

Then, why does the band of the Ga I nAs layer become convex upward and why are there free electrons 3 near the boundary?
That becomes a problem.

The AlInAs layer has a wider band gap. here n
Doped with type impurities. If it is only AlInAs, some of the electrons will go to the conduction band, and others will go to the n
It will be constrained by type impurities.

However, AI! There is not only an InAs layer, but an adjacent GaInAs layer. The lower end of the conduction band of the GaInAs layer is lower than the n-type impurity level of the AlInAs layer. Therefore, most of the electrons possessed by the n-type impurity in the AlInAs layer fall into the conduction band of GaInAs.

These electrons are bound by the n-type impurity of GaInAs and do not decrease. This is because the n-type impurity has the same number of electrons, and if the n-type impurity binds these electrons, there is no room to bind the electrons newly transferred from AlInAs.

Since the AlInAs layer has n-type impurities 5 that have lost electrons, it is positively charged.

This positive charge attracts electrons in the conduction band in the GaInAs layer, so the boundary points B and E are drawn to the electrons.

■ forms a potential that becomes lower in energy. In other words, the positive charges in the AlInAs layer cause the band of the GaInAs layer to curve downward at both ends due to the Coulomb force.

The band of the GaInAs layer therefore becomes upwardly convex.

The actual shape of this curve depends on the distribution of impurities in Ga I nAs. If the impurities are uniformly distributed, the Gaussian equation can be simply solved, and the band shape will be a quadratic curve. In other words, it becomes part of a parabola.

A depression in the band structure at both ends of the GaInAs layer is thus created. The electrons leave the impurity 5 of AlInAs and enter the conduction band of GaInAs, which has a narrow band gap, but a Coulomb force acts between them and the impurity 5 of AlInAs, so they are drawn toward the AlInAs and accumulate at the boundary. .

Next, the curvature of the band of the A,/InAs layer will be described.

The number of free electrons at both ends of the band of the GaInAs layer is greater than the number of n-type impurities 6 that have lost electrons in the GaInAs layer. Therefore, the GaInAs layer becomes negatively charged. A. Since the eInAs layer is positively charged, it is natural that GaInAs is negatively charged.

Since the GaInAs layer is negatively charged, AlInAs
When viewed from the layer, it exerts a repulsive force on electrons.

Therefore, a downwardly convex potential is formed for electrons in the AlInAs layer.

This is how the AlInAs band distortion occurs.

If there is an AlInAs layer, the Gaussian equation can be easily integrated, and the AlInAs band becomes a parabola.

2) Configuration What increases the dark current as a photodetector is Ga.
These are free electrons accumulated in the concave portions of the band that occur at both ends of the InAs layer.

In order to reduce dark current, it is sufficient to reduce the amount of free electrons accumulated.

In the present invention, the AlInAs layer is doped with n-type impurities. This reduces the free electron density of the GaInAs layer. If the free electron density decreases, the dark current decreases. This improves sensitivity.

In some cases, the free electron density can be set to O,
Free holes may also be introduced.

FIG. 1 is a band structure diagram of a photoconductive type light receiving element of the present invention in which an AlInAs layer is doped with an n-type impurity.

The upper curve LMNOPQ is the conduction band 1.

The lower curve R5TUV... is the valence band 2.

The region sandwiched between conduction band 1 and valence band 2 is a band gap (forbidden band).

I will explain the differences from the band structure in Figure 3.

(+) Fermi level 4 falls.

(il) The band structure of the Ga I nAs layer becomes convex downward.

(lli) AI! The band structure of the InAs layer becomes convex upward.

Since the band structure is different in this way, GaInAs
Free electrons do not accumulate in the layer.

First, the Fermi level 4 will decrease.

The Fermi level can also be said to be the minimum energy that allows all electrons to enter the electron level below that level. Introducing n-type impurities means increasing the level at which electrons can be captured, so naturally the Fermi level decreases.

Since the AlInAs layer is doped with n-type impurities, a shallow acceptor level is created just above the valence band 2.

This is the n-type impurity level 7 indicated by "-". Since there are more n-type impurities 7 than n-type impurities 5, AlI
Holes are predominantly supplied into the nAs layer.

If there is only the crystal structure of AlInAs, the hole is AlI
It either exists between the valence band ST of nAs or is captured by n-type impurities.

However, GaInAs with a narrow band gap exists next to AlInAs with a wide band gap.

The bandgap difference is large, and the valence band UV of the GaInAs layer is higher than the level "-" trapped in the n-type impurity 7 of the AlInAs layer.

Here, "high" means higher than electrons. Since holes have opposite charges, they will see a potential opposite to that of electrons.

Therefore, it is "low" with respect to holes.

Therefore, almost all of the holes in the A, gInAs layer enter the valence band of the adjacent GaInAs layer.

Although n-type impurities also exist in the AlInAs layer, there are more n-type impurities. The electrons donated by the n-type impurity are recombined with the holes donated by the n-type impurity and are substantially annihilated.

Excess holes enter the valence band of the GaInAs layer.

Then, the AlInAs layer will be negatively charged because n-type impurities that have lost holes remain. Since it is negatively charged, it attracts holes from the adjacent Ga InAs layer.

This is attracted by Coulomb force.

In this case, the potential for holes decreases in a portion close to the GaInAs layer. Since the band structure is written for electrons, it is the opposite for holes. For electrons, the potential increases. This is U
This is an upward curve at point V.

Holes may accumulate near point U and point V in some cases. U,
At point V, the potential is particularly low for holes. This is the singularity of the hole potential.

This hole is stable. This is because there is no n-type impurity in the GaInAs layer.

Overall, there are no excess electrons and no accumulation of electrons. In other words, there are no free electrons.

Depending on the amount of n-type impurity doped, in the present invention, free electrons can be excluded, but free holes are generated. Free holes are generated at the edges R, U, and V of the valence band, and dark current due to this may become a problem.

However, holes are not really a problem.

One reason is that the mobility of holes in GaInAs is an order of magnitude smaller than the mobility of electrons. Since the current is proportional to the mobility, the hole mobility is low, so the dark current associated with it is also small.

Another is that the amount of holes is a variable that can be controlled.
That's what it means. Therefore, the amount of remaining holes can be arbitrarily adjusted by adjusting the amount of n-type impurity doped into the AlInAs layer.

For these reasons, it can be seen that the problem of residual holes caused by the present invention can be easily overcome.

) Function As shown in Figure 2, AlInAs/
The electrode 14 is grown by epitaxially growing a GaInAs multilayer film.
The element with the numeral 14 was used as a photoconductive type light-receiving element.

Electrodes 14.14 make ohmic contact.

Since it is desirable to make ohmic contact with all of the multilayer film, a portion may be buried in the multilayer film.

A voltage is applied to electrode 14.14. Irradiates light from above. Light is efficiently absorbed by the Cya InAs layer.

The number of electrons and holes increases due to absorption of light.

This occurs in the GaInAs layer. Since the mobility of holes is slow, their contribution is small. Since the number of electrons with high mobility increases in proportion to the amount of light irradiation, the current also increases in proportion to the amount of light irradiation.

Since there are few impurities in the GaInAs layer, impurity scattering of electrons is small and electron mobility is high. Therefore, high-speed response can be obtained.

Although the Ga I nAs layer is described as an n-type semiconductor, it is not an intentional n-type impurity, but is due to residual impurities in the material. It is a non-doped GaInAs layer.

Since there are no free electrons near the boundary with AlInAs, there is no dark current caused by this.

)) Example 1: On the InP substrate 11, organic metal pyrolysis (OMVPE) was applied.
), liquid phase epitaxial growth (L'PE), molecular beam epitaxial growth (MBE), etc., the AlInAs layer 12 doped with an n-type impurity and the impurity-free G
Thirteen AlInAs layers were alternately laminated.

As n-type impurities, Zn, Be, Mg, M
n, etc. were used.

Apart from this, the n-type impurity concentration in the AlInAs layer 12 is 10
16z-3 or higher. The layer thickness of the AlInAs layer 12 is 5
Must be 0A or more.

Further, the thickness of the Ga InAs layer is 0.5 μm or less.

Under this condition, the Ga InAs layer is depleted. The carriers contributing to conduction become almost zero when no light is irradiated.

Regarding the number of stacked layers n, GaInAs layer l 3 (7)
The product n of the thickness t and the number of laminated layers n is set to be 0.5 Pm or more.

Under this condition, the incident light is efficiently absorbed into the GaInAs layer.

Electrodes 14.14 made of, for example, AuGe are formed on the alternating multilayer body by, for example, a vapor deposition method.

At this time, in order to reduce the contact resistance of the electrodes,
Alloying may be performed by heating to a certain degree.

When measuring the dark current of this element, under the same conditions,
It was found that the reduction was about 1/10 to 1/100 compared to the conventional one.

(e) Effects According to the present invention, a high-speed photoconductive light receiving element with little dark current can be produced.

It can be effectively used in the fields of measurement and optical communication.

[Brief explanation of drawings]

FIG. 1 is a diagram of the band structure of the photoconductive element of the present invention. FIG. 2 is a schematic front view of the photoconductive element of the present invention. FIG. 3 is a diagram of the band structure of a conventional photoconductive element. 1...Conduction band 2...Valence band 3...Free electrons 4...Fermi level 5...N-type impurity of AlInAs 6・・・・・・
・N-type impurity of GaInAs 7...AlInA
s P-type impurity 8...Hole 11...InP substrate 12--AlInAs layer 13...-GaInAs layer 14...Electrode inventor Sasaki Goro

Claims (1)

[Claims] An InP substrate 11, an AlInAs layer 12 epitaxially grown alternately on the InP substrate 11, and a GaInA layer.
The AlInAs layer 1 is composed of a laminated structure of the S layer 13 and a plurality of electrodes 14 provided on the laminated structure.
2 is a photoconductive type light-receiving element characterized in that a P-type impurity is added.