JPS6158279A - Electrostatic induction type semiconductor photodetector - Google Patents

Abstract

PURPOSE:To perform a detection of light by a method wherein the high-resistance regions between the gates and the drain are made narrower their forbidden band widths than that of the drain region, the sensitivity of the avalanche photo diode to the light of a long wavelength is increased, the optical amplification degree at the SIT part is increased, and furthermore, a function equivalent to the operation of the SCR, which is generated due to the forbidden band gap difference between the high-resistance regions and the drain, is let the photodetector have. CONSTITUTION:An n<+> type layer 21 and an n<-> type layer 24 are both an InGaAs layer, but the forbidden band widths thereof are wider than that of the InP layer of an n<-> the layer 22. p<+> type InP layers 25 are used as the gates of the photo SIT and an n<+> type InP layer 28 is used as the source of the SIT. The whole surface is covered with an insulative optical reflection preventing film 26, and Au-Zn-Ni electrodes 27 and Au-Ge-Ni electrodes 29 and 20 are attached. Each dimension and concentration of the layers 24 and 22 are selected and the layers 24 and 22 are made to deplete with voltage between the electrodes 27 and 29 and voltage between the electrodes 27 and 29 and the electrode 20 and the respective height of the potential barriers, which are generated in the layers 24 and 22 held between the gates 25, is made to change by the effect of the SIT. Moreover, when the dimension, voltage and impurity concentration are selected in such a way that an avalanche electric field is applied to the layer 22, the light of a long wavelength is received by the avalanche photo diode at a high speed and a high sensitivity, and furthermore, signal is amplified at the SIT part.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a high-speed and highly sensitive photodetector based on an electrostatic induction transistor, and is particularly suitable for long-wavelength optical communication because the long-wavelength sensitivity is improved. Used as a High-Speed and High-Sensitivity Photodetector (Prior Art) As a conventional photodetector that is particularly close to the present invention, there is an electrostatic induction phototransistor.

For example, an electrostatic induction phototransistor (Japanese Patent Application No. 56-192417 ``Semiconductor Photoelectric Conversion Device'') uses p -1-n diodes distributed around the gate as a photodetector. n
Taking a channel as an example, holes among the electron-hole pairs generated by light are accumulated in the gate region during p-mode, causing a change in the gate potential, which modulates the potential barrier height in the channel. Therefore, the amplified current flowing between the source and drain is modulated by this gate potential change. The optical amplification factor is extremely high, and the weaker the incident light intensity, the higher the optical amplification, which is the complete opposite of that of conventional bipolar phototransistors. The maximum optical amplification degree G mg,t is approximately expressed as. where nj i,in'' is the impurity density in the source region, p, cT is p'' is the impurity density in the gate region, D6
is the electron diffusion constant, DF7 is the hole diffusion constant, W9 is the effective thickness of the potential barrier, L7y is the hole diffusion distance, q is the unit charge layer, K is Boltzmann's constant, T is the absolute temperature, ■ .

1crff is the diffusion potential of the p1n diode between the gate and source, and Vbic4''J is the peak height of the potential barrier as seen from the source.

An example of its general structure is shown in FIG. n + wild root board 6
The p medium region 62 diffused into the n- or p- high resistance 11161 formed on the 0 is between the gates 62, and the n?
Region 67 represents the source or drain region. The n-type substrate 60 is a drain or source region. 64 and 65 represent a gate electrode, source or drain electrode, respectively. 63 is an insulator. The light penetrating by the incident light hν mainly generates electrons in the n-high resistance layer 61.
Generates hole pairs. Since these holes are accumulated in the p+ gate region 62, an amplified signal is obtained between the main electrodes 65 and 66.

In order to form a potential barrier in the n-high resistance layer, D”
Dimensions between gates 62, impurity density and n-111161
dimensions and impurity density are selected. In the conventional structure, the forbidden band width of the semiconductor in the high resistance 111[i region n-61 and the forbidden band width of the semiconductor forming the p+ gate region 11162 and the n++ plate 60 are all the same. That is, almost all regions are often formed of the same semiconductor. There are some examples of static induction transistors in which a Peter junction is provided in a gate region or a source region. In this case, gate region 6
If the semiconductor No. 2 is formed of a material having a wider forbidden band width than the semiconductor in the channel region, the gate 62 and the channel n-
A heterojunction occurs at the junction interface 61. If only the gate is formed of a semiconductor with a wide forbidden band width, and the other regions are made of the same semiconductor, the current amplification factor increases by integrating the potential difference at the heterojunction interface. That is,
The potential difference at the hetero interface is included in the exponential term in equation (1). Furthermore, even if a semiconductor with a wide forbidden band width is used in the source region, the current amplification factor increases. In FIG. 2, if the n'' substrate 60 is formed in the p medium region, it becomes an electrostatic induction type photothyristor.

[Problems that the invention attempts to solve]

However, the present inventors have discovered a static N-induced photodetector that is even faster and more sensitive.

[Means to be solved by the invention]

By using a semiconductor whose forbidden band width is narrower than that of the gate and drain regions in the high resistance m region between the gate and drain, the long wavelength sensitivity is increased, the optical amplification of the static induction transistor portion is increased, and the high resistance layer This results in a photodetector with a mounting surface equivalent to the Sirisk operation due to the bandgap difference between the region and the drain. Furthermore, compared to the conventional case where the semiconductor part with the narrow forbidden band width is formed with the same forbidden band width,
A stronger electric field can be generated with the same thickness and voltage, so if the dimensions, voltage, impurity density, etc. are selected so that an avalanche field is applied, long wavelength light can be received by the APD at high speed and with high sensitivity. Furthermore, signal amplification is performed in the electrostatic four-layer transistor section. Further, even if the electric field strength is not made so high as to cause an avalanche electric field, the 84-component device of the present invention has sufficiently high sensitivity and 1:) speed.

(Function) As mentioned above, the gate of the conventional static induction phototransistor
If the high resistance layer portion between the trains is formed of a semiconductor with a narrow forbidden band width, a strong electric field is likely to be generated between the gate and drain. In the case of an avalanche electric field, carriers multiplied by collision ionization of the Q17 rear travel in a strong electric field. Taking the case of an n-channel static fixes transistor as an example, the multiplied holes are accumulated in the p+ gate region, and some of the multiplied electrons are deposited at the heterojunction interface between the high resistance layer and the anode region. Accumulated. The holes accumulated in the p+ gate region positively charge the p+ gate region, thus forward biasing the gate of the static W1 induction transistor and lowering the potential P3 wall in the channel. Along with this, electrons are injected from the n+ source region to the high-resistance channel region, and the holes, which serve as optical signals multiplied by the avalanche electric field, amplify even more electrons and form a gap between the source and drain. Therefore, an optical signal output can be obtained.

(Example) FIG. 1 shows an embodiment of the invention.

Region 21 represents the drain or anode region, and is n+
The main electrode region is formed as a region or an n-medium region, and if it is an n-medium region, it operates as a static induction phototransistor, and if it is a pf region, it operates as a static induction photothyristor. For example, the material is n+InP or p
"InP is used. The ¥4 area 20 is the drain or anode electrode, and the electrode material is Au-Qe-Ni or pya for n+1nP.
An alloy such as LI-7n-Ni is used. Region 22 is a high resistance layer formed of a semiconductor with a small forbidden band width, which is a feature of the present invention.

For example, 1nQaAs is used. The region 24 is also a high resistance layer, but is a semiconductor layer different from the region L1i22 and having a wider forbidden band width than the region 22. For example, n = [n p
Use. The n-middle region 25 acts as a gate of an electrostatic induction phototransistor or an electrostatic four-layer photothyristor,
For example, it is formed of p+1nP. The n middle region 28 functions as a source region of an electrostatic induction phototransistor or a cathode mbX of an electrostatic induction photothyristor, and is formed of, for example, n''''InP. Region 26 is an insulating optical anti-reflection coating. The region 27 is a gate electrode, and is formed of Au-Zn-Ni etc. for p"tnp, for example. The region 29 is an electrode of the n-middle region 28, and is formed of ALI-Qe for, for example, n-middle InP. -Ni, etc. The light 11ν enters from the surface of the p+ gate region.The high resistance layer regions 24 and 22 are composed of the p+ gate region 25 and the n'' region 28 or the p+ gate region 25 and the high impurity density region 21.
This corresponds to the i-layer portion of the p-1-n diode formed between the two. High resistance layer regions 24 and 22 are gate electrodes 2
The dimensions and impurity density are selected so that depletion occurs due to the voltage applied between the main electrode 7, one main electrode 29, and the other main electrode 20. Furthermore, a potential barrier is formed in a portion of the high resistance layer portions 24 and 22 sandwiched between the p′+ gate 25 when viewed from the n+ region 28, and its height is equal to the voltage between the gate ff electrode 27 and the n electrode 29 and the electrode 20. Dimensions between p÷ gate WA area 25 and impurities as changed by electrostatic induction effect due to voltage of ! F/J! , the thickness of high resistance WI22.24 is chosen. High resistance W! Other area 2 as J area 22
By using a semiconductor whose forbidden band width is narrower than that of 4.25.21.28, the sensitivity in the long wavelength region of the wavelengths of light incident from the surface increases. Furthermore, since a strong electric field is likely to occur, the dimensions, impurity density, etc. should be adjusted so that an avalanche electric field is generated in the region 22 between p'' (25) n- (24) n- (22) n+ or p10 (21). If selected, an electrostatic dielectric 3# type photodetector in which avalanche photodiodes are distributed around the gate will be formed.Avalanche diodes are formed between the p+ gate 25 and the highly impurity-dense U region 11!21. A signal obtained by further amplifying the optical signal received at high speed and high sensitivity is extracted from between the main electrodes 29 and 20.Electron-hole pairs are mostly depleted in the high resistance layer ffa regions 22 and 24. The holes are generated, and among them, the holes are transferred to the p+ gate region 25, and the potential barrier height generated in the vicinity of the gate region [28] changes due to the static 1! induction effect of the gate 25. A p-type semiconductor may be used instead of the high resistance region 24. If the p-type region is completely depleted, the operation will be equivalent to that of a static induction phototransistor, but
If some neutral region of the bipolar base layer remains, it becomes a bipolar phototransistor or a photothyristor. Or high resistance temperature range 1m! All of 24 or part of area 22 is prohibited as in high resistance temperature area*22.
Semiconductors with narrow bandwidth (e.g. n-In Qa As
) may be formed. High impurity density region 21 is a region of the same conductivity type as gate region 25 (p+ in the example of FIG. 1)
In this case, the embodiment of FIG. 1 operates as a static induction photothyristor. In this case, the hetero interface between the region 21 and the narrow forbidden band region 22 becomes a region where electrons behind electrons and holes generated by light are accumulated. This accumulation of electrons causes more holes to be amplified and injected from the p-type region 21 to the region 22. Although InP-InGaAs is shown as an example of a semiconductor forming each part, the semiconductor is not limited to this, and GaAs-GaAt
Combinations of other binary, ternary, and quaternary compound semiconductors such as As, GaAs-InGaAsP, etc. may also be used. In the embodiments of the present invention, a planar gate structure is shown, but it goes without saying that the shape of the gate may be a buried gate, a cut gate, a MOS (Mis) gate, a hetero gate, or a Schottky gate.

〔Effect of the invention〕

The electrostatic induction photodetector according to the present invention is characterized in that a semiconductor with a narrow forbidden band width is mainly used between the gate and drain or between the gate and anode. increases long wavelength sensitivity. Furthermore, since it is easy to increase the electric field strength in this region, for example, if it becomes an avalanche electric field strength, it becomes possible to receive light at high speed and with high sensitivity using the avalanche diode between the gate and drain or between the gate and anode. ? l! Faster and more sensitive than inductive photodetectors.

The present invention is different from conventional! J manufacturing method such as MOCVD method, optical C
It is easy to manufacture using the VD method, photoepitaxial method, or molecular layer epitaxial method, and is of high industrial value because it can be used as a photodetector for high-speed, high-sensitivity, and long-wavelength optical communications.

[Brief explanation of drawings]

1 shows an embodiment of the present invention, and FIG. 2 shows a cross section 41 of a conventional electrostatic induction photodetector. 20... Main electrode serving as a drain or anode, 2
1...n+ or p10 high impurity density region, 22.
...High resistance layer with narrower forbidden band width than other regions, 2
4... High resistance layer, 25... Gate diffusion region, 28...
. . . Source or cathode region, 29 . . . Source or cathode electrode, 26 . . . Insulating optical antireflection film, 27 . . . Gate electrode.

Claims (1)

[Claims] (1) A semiconductor substrate is composed of first and second heterojunctions each consisting of a first and a second high resistance layer, and a first conductivity type heterojunction formed in the first high resistance layer. an impurity density gate diffusion region; and a high impurity density region of a second conductivity type opposite to that of the gate diffusion region, which is sandwiched between the gate diffusion region and formed in the first high resistance layer. A first or second conductivity type that forms a bonding surface with one main electrode region and a second high-resistance layer opposite to the first main surface on which the gate and the first main electrode region are formed. a second main electrode region consisting of a high impurity density region, first and second electrode regions formed on the first and second main electrode regions, and the first height region including the gate diffusion region. It has an insulating/optical antireflection film formed on the resistance layer and a third gate electrode formed on the gate diffusion region, and only the semiconductor forming the second high resistance layer is separated from the others. formed of a semiconductor having a forbidden band width smaller than the forbidden band width of the semiconductor in the region;
The gate diffusion region is configured such that an avalanche electric field can be easily generated by a voltage applied to the second main electrode region and the gate region, and the gate diffusion region is arranged in the first high resistance layer or in the first and second high resistance layers. A potential barrier is formed by the voltage between the gate diffusion region and the first main electrode region and the voltage between the gate diffusion region and the second main electrode region, and the first and second high resistance layer regions are almost depleted. The dimensions of each part are selected such that the height of the potential barrier changes depending on the potential of the gate diffusion region and the second main electrode region. An electrostatic induction photodetector characterized by controlling majority carriers flowing between electrodes.