LT6044B - High speed photodiode - Google Patents

Abstract

Uni-trevelling carrier type photodiode is made of semiconductor materiais, whereas the heterostructure comprises an absorber made of a materiai containing bismuth atoms.Such semiconductors having bismuth can be for example GaAsBi, InPBi or GalnAsBi.Introduction of Bi into such compounds of the A3B5group, raises the energy levels of the valence band, without making noticeable influence on the energy levels of the conduction band. Therefore, the conduction band offsets in heterojunctions, like GaAsBi/GaAs,InPBi/lnP and GalnAsBi/GalnAs, are rather small. In another embodiment, the concentration of bismuth atoms decreases uniformly with increasing distance from the junction between the absorber and the collector layer.

Description

This invention relates to the field of semiconductor photodetectors and can be used to detect, measure and translate high frequency repetition rates, including terahertz frequencies, into optical signals.

TECHNICAL LEVEL

The high-speed photodetector assembly and production technology described in the proposal can be used in the manufacture of broadband optoelectronic components and systems, including receivers of optical communication systems and terahertz emitters, illuminated by two laser wavelengths.

High-speed photodiodes have a particularly wide application in telecommunications. There they are used to detect signals transmitted by fiber. The faster the photodiode is used, the more information can be transmitted through the fiber-optic line. Fast photodiodes are also used in temporal resolution spectroscopy, pulsed laser characterization (pulse duration and temporal envelope detection) and other applications where ultra-short infrared signals are required.

High-speed photodiodes, which record infrared signals with a wavelength between 1 pm and 1.5 pm, usually have an interlayer made of a semiconductor with an appropriate energy gap ε ₉ = 1.24 / where λ ₉ is the longest wavelength the photodiode is still sensitive. A known example of such a device is a so-called PIN diode, in which an absorber layer (i-layer) made of a semiconductor with a smaller gap is inserted between two layers of a wider gap of the semiconductor doped with acceptor and donor impurities, respectively p-type and n-type. . The i-layer in the middle of the Pn junction generates an internal electric field that forces light-generated electrons and holes to move, thereby inducing photocurrent. The infrared PIN diode absorber layer is usually made from InGaAs or InGaAsP compounds. In these materials, as in other A ₃ B ₅ semiconductors, the motility of the holes is significantly less than that of the electrons, so the movement of the holes created by the light limits the speed of the PIN photodiodes and their limiting frequency, which in these devices is only tens of GHz. In order to improve these parameters, it is necessary to reduce the thickness of the i-layer while decreasing the sensitivity of the photodiode itself, but in this case only the threshold frequencies of about 100 GHz are reached. Another disadvantage of PIN diodes is that due to the slow removal of holes from the i-layer, it generates a spatial charge that causes saturation of the photocurrent and a decrease in photosensitivity at already relatively low levels of optical excitation.

Several solutions have been proposed to overcome the disadvantages of PIN diodes.

U.S. Pat. US5818096 describes a photodiode with a single type of carrier (UniTraveling Carrier - UTC - photodiode). The derivative of the UTC photodiode layers is shown in Figs. 1, its bar graph is shown in Figs. 2.

Basically, a UTC photodiode consists of:

I P-type electrical conductivity of the absorber layer from the energy gap of the narrow reserves e _g i oily substances (siauratarpio semiconductor) (1);

II. N-type conductive layer of a material having a wider resistor energy gap e _g2 (Wide-band semiconductor) (2);

III. A non-alloy or weakly doped collector layer of a wider _resistor energy range ε _ρ2 ( _Wide- band semiconductor material) (3);

IV Barrier layer of p-type conductivity wide-band semiconductor (4). The function of this layer is to stop the diffusion of electrons into the anode contact (5);

V Metal anode contact (5);

VI Metal cathode contact (6);

VII Crystal Tray (7).

The p-type layer (1) of the photodiode derivative is made of a semiconductor with a narrow reserve energy gap ε ₉ ι and the n-type layer (2) is made of a semiconductor with a wider reserve energy gap e _g2 . Both of these layers are separated by an unalloyed or weakly doped semiconductor layer of a wider _resistor energy gap ε _ρ2 (3). The narrow gap p-type semiconductor layer (1) is bordered by the same conductivity type layer from the wide-gap semiconductor layer (4). The photodiode is sensitive to radiation with a quantum energy ranging from ε _9ΐ to e _g2 .

The optical radiation is absorbed by the absorber (1) which satisfies the condition of electrical neutrality. The photosensitized electrons in the absorber (1) are diffused into a weakly doped wide-band semiconductor layer (3) (collector layer), which has a strong electric field transferring electrons to a donor-doped n-type layer (2) made of a wide-band semiconductor. In the absorber (1), the non-equilibrium holes created by the light signal mix with the equilibrium holes in this layer: additional holes over time equal to dielectric relaxation time T _d = ε / σ (where ε is the dielectric permeability of semiconductor and σ is its electrical conductivity). cathode contact. In a typical A ₃ B ₅ semiconductor with a hole concentration equal to 10 ¹⁸ cm ³ , xd has a order of several femtoseconds, so unlike the PIN photodiode, the dynamics of the unbalanced holes have no greater effect on the UTC darkness recovery rate. This parameter, and thus the cut-off frequency of the UTC photodiode, is determined by the duration of the non-equilibrium electron beam through the collector layer (3), which is equal to τβ = Lc / vs. where Lc is the length of the collector layer (3) and the saturation velocity of the electron drift in the collector layer (3) is predominant. With a collector layer length of 200 nm and an electron drift velocity of 2 10 ⁷ cm / s, we obtain τθ = 1 ps, so that the UTC photodiode's dark state recovery time could provide a very wide frequency range of 1 THz. Another factor limiting the speed of the device prevents this from the time it takes for the electrons to pass from the absorber (1) to the collector (3). When there is no internal electric field in the absorber layer (1), electrons from this layer to the collector (3) fall only due to diffusion; the characteristic duration of this process can be estimated as t _d «l ² s ·, / 4D _e (where Ls is the width of the absorber layer (1) and D _e is the electron diffusion coefficient of that layer material (by constructing in this expression a typical absorber layer, values from InGaAs: L _s = 200 nm and D _e = 100 cm ² / s will give t _d «10 ps, which is ten times the duration of the electron beam through the collector. This parameter can be reduced by narrowing the absorber region (1), however, this results in a decrease in the concentration of the light-excited charge carriers and, consequently, in the sensitivity of the photodiode.Another way of accelerating the transfer of electrons from the absorbent layer (1) to the collector layer (3) involves the formation of a variable composition absorbent layer.

The changing composition of the layer results in a coordinate dependent energy gap and the resulting quasi-electric field that forces electrons to move toward the collector layer (3). The electron energy bar diagram of such a photodiode is shown in Fig.3.

The sensitivity of the UTC photodiode is strongly influenced by the difference in the _conductivity difference ε _ρ2 - £ gi = Δε ₀ + Δε _ν in the heterosaturation between the narrow-band absorber layer (1) and the wide-band semiconductor layer (3). of the trūkε ₀ and Δε _ν fractures of the bands. In InGaAs / lnP heterosandel, the conductivity band between the absorber and collector layers in conventional UTC photodiodes accounts for ~ 80 percent of the total difference in the energy gap between the materials forming the heterosand. As a result, there is a high potential barrier for electrons moving from the absorber layer to the collector, interfering with the movement of the photosensitized electrons and impairing most of the device's performance.

FIG. 4 shows the spectral dependencies of the sensitivity of UTC photodiodes in structures with absorber and collector layers of materials with the same reserve energies (ε ₉ ι = 0.8 eV and ε & = 1.42 eV), but with different ratios of conduction and valence band breaks Δε ₀ and Δε _ν , .

FIG. 5 shows the spectral characteristics of UTC photodiodes with absorbent layers of different bonded energy gap materials. It can be seen that the sensitivity can be increased by using a material with a lower ε _9ΐ . On the other hand, the use of a lower resin energy gap material allows the thickness of the absorber to be reduced while improving the dynamic characteristics of the photodiode.

THE SUBSTANCE OF THE INVENTION

The present invention seeks to solve the above problems and to provide a photodiode with one type of moving charge carriers having better characteristics than conventional photodiode of this type with InGaAs absorber layers and InP collector layers.

A semiconductor photodiode consisting of a p-type conductivity absorber layer, an unalloyed collector layer, and two n-type and p-type conductivity contact layers is proposed, wherein the absorber layer is made of a semiconductor containing bismuth atoms. This reduces cracks in the conduction band levels.

Therefore, the use of a semiconductor material containing an absorber layer Bi can improve photodiode performance parameters such as its sensitivity and speed.

BRIEF DESCRIPTION OF THE DRAWING FIGURES

In order to better understand the invention and evaluate its practical applications, the following illustrative drawings are provided. The drawings are given by way of example only and in no way limit the scope of the invention.

1. Structure of a photodiode with one type of moving charge;

FIG. Electron energy bar diagram of a photodiode with one type of moving charge;

3. Diagram of electron energy bands of a photodiode with one type of moving charge and a variable absorber energy gap;

FIG. Sensitivity spectral dependence of photodiodes with moving one type of charge in structures with absorber and collector layers of materials with the same reserve energies (e _g i = 0.8 eV and = 1.42 eV) but with different conductivity and valence band fractures Δε ₀ and Δε _ν , ;

FIG. Spectral Characteristics of Photodiodes with Moving Single-Chargers and Absorption Layers of Different Reserve Energy Gap Materials. The breakage of the conductivity band was in all cases considered to be £ _c = 40%;

FIG. Comparison of the variation of the protected energy gap in GaAsBi and InGaAs semiconductors with increasing Bi and In concentrations, respectively;

FIG. The deviation of the lattice constants of the ternary InGaAs and GaAsBi compounds from the dependence of the GaAs lattice constant on the change in the bonding energy gap of the compound with respect to GaAs.

PREFERRED EMBODIMENTS

In a preferred embodiment of the invention, a single-mode, moving-charge photodiode is made of semiconductors having a bismuth-containing material in its heterostructure layer. Such bismuth-containing semiconductors may, for example, be GaAsBi, InPBi or GalnAsBi. In these A ₃ B ₅ semiconductor compounds, introduction of Bi atoms raises valence band energy levels with little effect on the position of the conduction band energy levels. Therefore, the band gap breaks in the respective heterosandes: GaAsBi / GaAs, InPBil InP and GalnAsBi / GalnAs, are small, for example, in the first of the listed heterosandes, only 20% to 30% of the GaAs and GaAsBi duct energy gap.

It should be emphasized that the use of a GaAsBi absorber layer provides a unique opportunity to produce infrared-sensitive photodiodes on GaAs substrates. Experimental studies of GaAsBi have shown that the energy gap between this compound decreases very rapidly during introduction of Bi - 90 meV for each percentage of Bi introduced. Therefore, ε ₉ ι = 0.8 eV can be achieved when only 10% Bi is introduced into GaAs, and GaAsBi with 15% Bi has a reduced energy gap even ε ₉ ι = 0.4 eV. For comparison, the energy gap of the ternary compound InGaAs decreases much slower as the proportion of In grows - with the introduction of GaAs by as much as 30% In, the energy gap decreases to only 1 eV. The concentration of Bi or its variation can be selected according to the desired characteristics of the photodiode. It should be apparent to one skilled in the art how varying the Bi impurity concentration can provide the desired photodiode speed and sensitivity.

Fig. The characteristics of the six ternary compounds, InGaAs and GaAsBi, are compared with each other, showing how the deviation of their respective lattice constants from the GaAs lattice constant depends on the change in the bonding energies of the compound with respect to GaAs. Five times smaller than in InGaAs, the deviation of the GaAsBi lattice constant from the GaAs substrate lattice allows for better crystalline quality of epitaxial layers, higher electron mobility and longer life in these layers. These parameters directly determine the sensitivity and speed of the photodiode.

In another embodiment of the present invention there is provided a semiconductor photodiode consisting of a p-type conductivity absorber layer, an unalloyed collector layer and two n-type and p-type conductivity contact layers, wherein the absorber layer is made of a semiconductor containing bismuth atoms in the compound decreases evenly along the boundary of the absorber layer with the collector layer.

In such a photodiode p-type doped absorber (1), the position of the valence band edge will be independent of the coordinate, while the conductor band edge energy will increase steadily from the absorber boundary with the collector (3) to the boundary layer (4). This growth will create a force acting only on the electrons and forcing them to move toward the collector. This increases the sensitivity of the photodiode (more light-excited electrons participate in the photocurrent) and increases the speed of this device (reducing the time it takes for electrons to reach the collector). Due to the fact that Bi Input alters the bar graph much more strongly than other atoms, variable-energy gap (varison) layers of semiconductors containing bismuth atoms are best suited for photodiode absorbers, since a high conductivity band edge energy gradient can be obtained with relatively little change in Bi concentration. .

Using the examples described above, one of ordinary skill in the art may propose more heterogeneous compositions for detecting light in the infrared band, but the protection of the present invention should encompass all variations in the structure of moving single-charge photodiodes employing bismuth atoms that reduce conductivity bands. level breaks, which improves the sensitivity and speed of such a photodiode.

Claims (5)

A moving single-charge photoelectric detector having a heterostructure comprising at least a barrier, a p-type absorber, a collector and a conductivity layer, characterized in that the absorber layer is composed of a semiconductor material containing bismuth atoms. 2. A photodetector according to claim 1, characterized in that the proportion of the bismuth atoms in the absorber compound decreases uniformly with respect to the boundary of the absorber layer with the collector layer. 3. A photodetector according to one of claims 1 to 2, characterized in that the heterostructure is one of GaAsBi / GaAs, InPBi / lnP, GalnAsBi / GalnAs, InGaAsBi / AllnAs. 4. Photo detector according to one of claims 1 to 3, characterized in that the percentage of bismuth atoms in the GaAsBi absorber material is between 5 and 15 percent or varies between any values in the range of 5 to 15 percent. 5. A photodetector as claimed in any one of claims 1 to 4, characterized in that it is adapted to operate in one of the applications of IR spectroscopy, time-resolution spectroscopy, data transmission, radiation pulse characterization, terahertz wave generation.