JPH0353789B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an optoelectronic switch that switches electrical signals using light as a medium.

As a conventional switch of this type, an optoelectronic switch has been proposed which combines an electric-to-optical conversion element 1 and an optical-to-electrical conversion element 2 and uses light 3 as a medium, as shown in FIG.
(Electronics Letters Vol.14, PP.502-503,
1978, “Optoelectronic Broadband Switching
That is, the light 3 corresponding to the input electric signal A is generated using an electro-optical conversion element 1 such as a semiconductor laser or a light emitting diode.
is incident on a photo-electric conversion element 2 such as a photodiode (hereinafter abbreviated as PD) or an avalanche photodiode (hereinafter abbreviated as APD). At this time, the polarity of the bias electrode applied to the PD or APD is changed to a bias power source 4a with a different polarity,
4b and the electrical switch 5, the light-to-electrical conversion effect is small under forward bias, achieving an OFF state, and the light-to-electrical conversion action is large under reverse bias, achieving an ON state. By switching the polarity of the bias voltage, a signal B is obtained at the terminal of the load resistor 6 in the on state, and no signal is output in the off state. Therefore,
This allows the switch to be performed.

However, this type of switch using conventional PD or APD has the following drawbacks.

That is, first, during forward bias, a current of several mA to several tens of mA flows, resulting in large power consumption. Second, the circuit for polarity switching becomes complicated and requires power supplies with different polarities. Third, impedance matching is difficult because the impedance of the PD and APD changes significantly due to forward/reverse polarity switching. Fourth, when switching the bias voltage from the forward direction to the reverse direction, there is a charge accumulation effect, which limits the switching speed.

These drawbacks have limited the range of applications of conventional optoelectronic switches.

In view of the above points, the present invention was made to eliminate these drawbacks, and its purpose is to provide an optoelectronic switch that has low power consumption, a simple circuit configuration, an ultra-high speed switch, and is capable of integration. Our goal is to provide the following.

In order to achieve such an object, the optoelectronic switch according to the present invention provides a pn junction in a first semiconductor of - group or Si as a photo-electrical conversion element, and furthermore, the absorption of light energy is carried out in this first semiconductor. It has a lattice constant that almost matches that of the semiconductor, and the energy on the minority carrier side is discontinuously smaller than that of the first semiconductor.
using PDs or APDs configured to operate in second semiconductors of the Group or Ge group, with a reverse bias voltage low enough that the depletion layer in these PDs or APDs does not reach the heterojunction consisting of the first semiconductor and the second semiconductor; PD or APD It is designed to switch.

Hereinafter, embodiments of the present invention will be described in detail based on the drawings.

FIG. 2 is a structural diagram of a PD used in the present invention, and is a sectional view showing the basic configuration of a photodiode constituting the optoelectronic switch of the present invention. Here, to simplify the explanation, we will use indium phosphorus InP as the first semiconductor and In _x Ga _1-x As _1-y P _y (indium gallium arsenic phosphorus) as the second semiconductor. In the - group compound semiconductor, a PD composed of x=0.53 and y=0, that is, indium, gallium, and arsenide In0.53Ga0.47As (hereinafter abbreviated as InGaAs) will be described.

In the figure, first, 7, 8, and 10 are semiconductors, each using a so-called p ⁺ type InP containing a large number of p-type impurities, for example, 10 ¹⁸ cm ^-3 or more (hereinafter referred to as p-type InP 7). ), a semiconductor 8 using n-type InP containing n-type impurities (hereinafter referred to as n-type InP 8) is provided to form a pn junction 9. Next, a semiconductor 10 (hereinafter referred to as
(referred to as InGaAs10) and heterojunction 1
form 1. Furthermore, electrodes 12 and 13 are formed on the surfaces of InGaAs 10 and p-type InP 7, respectively, to enable electrical connection.

In the photodiode configured in this manner, incident light is incident from either the InGaAs 10 side or the p-type InP 7 side, as shown by 3a or 3b in the figure. Almost the same effect can be obtained regardless of the direction of the incident light 3a, 3b.

In order to operate this as an optoelectronic switch, the voltage must be changed from V ₁ to V ₂ with respect to the input terminal 14 to which the power supply 12 is connected.
Apply a pulse voltage V _p that varies between In this case, voltage V ₁ is 0V or photodiode 16
The reverse voltage _V2 is lower than a certain constant voltage with respect to the photodiode 16, that is, the electrode 12 side is at a positive potential and the electrode 13 side is at a negative potential. voltage. Details regarding these voltages will be described later.

On the other hand, the wavelength of the incident light 3a or the incident light 3b is
The light energy is not absorbed in indium phosphorus InP7, and indium gallium arsenide
The wavelength is selected so that the light energy is absorbed in the InGaAs layer 10.

Next, the operation of the photodiode 16 in the embodiment shown in FIG. 2 will be described in detail with reference to FIG. Figure 3 shows the energy band structure of the photodiode 16, where a shows the structure at a bias voltage of 0V, and b shows the structure when a bias voltage corresponding to the above voltage _V2 is applied. be. In FIG. 3, the same symbols as in FIG. 2 indicate corresponding parts, 17 is the conduction band edge, 18 is the valence band edge, and 19 is the Fermi level. A pn junction 9 is formed between p ⁺ -InP 7 and n-InP 8, and a heterojunction 11 is formed between n-InP 8 and n-InGaAs 10.

Now, the photodiode 16 with such a configuration
On the other hand, when light of the above-mentioned wavelength is incident, the light energy is absorbed only by the InGaAs 10, and a hole 20 is generated. The hole 20 generated in this way moves in the direction of the arrow in FIG. 3a, but as shown in FIG. 3a, at a bias voltage of 0V,
The energy barrier 21 formed in the heterojunction 11 cannot be overcome, and no current according to the incident light flows. That is, even if light is incident, no output signal appears at the output terminal 15 in FIG. 2, and the switch operation is in an off state.

Next, when the applied bias voltage is increased so that the depletion layer spreading from the pn junction 9 reaches the heterojunction 11, the energy band structure becomes as shown in FIG. 3b, and the holes 20 generated by light overcome the energy barrier 21. You will get it. Therefore,
A current flows in accordance with the incident light, and an output signal B appears at the output terminal 15 in FIG. 2, realizing an on state as a switch operation.

As is clear from the foregoing, the off-state and on-state of the switch operation are achieved by either letting the depletion layer not reach the heterojunction 11 or by letting it reach the heterojunction 11. This is done by preventing or allowing the generated holes 20 to overcome the energy barrier 21. That is, the applied bias voltage V ₁ (see FIG. 2) for realizing the OFF state may include 0 V and is a reverse voltage as low as to prevent the depletion layer from reaching the heterojunction 11. On the other hand, the applied bias voltage to achieve the on-state
V ₂ (see FIG. 2) may be a reverse voltage sufficient to allow the depletion layer to reach the heterojunction 11 or a voltage higher than that.

In other words, the switching operation can be performed with reverse voltage magnitudes including 0V.

Figure 4 is an explanatory diagram showing the relationship between the impurity concentration of indium phosphide InP and the depletion layer width when biased near the avalanche breakdown voltage. There is. Then, from this figure, the thickness of n-InP8 is determined,
Furthermore, the minimum value of the voltage V ₂ that should be applied to achieve the on state can be read from the vertical axis on the right side of the figure. For example, when the impurity concentration of n-InP8 shown on the horizontal axis is 1×10 ¹⁶ cm ⁻³ , the depletion layer width shown on the vertical axis on the left side of the figure can be read as 3.3 μm when the electric field E is 0.

By extending the depletion layer beyond the heterojunction, an on-state with desirable characteristics can be obtained. Therefore, it is desirable that the electric field E has a value of approximately 1×10 ⁵ V/cm.

That is, from FIG. 4, since a depletion layer width of 2.5 μm is obtained at this time, this value is determined as the thickness of the n-InP layer. The voltage for switching between the OFF state and the ON state can be read as 47V from the vertical axis on the right, and the bias voltages V ₁ and V ₂ should be determined as 0≦V ₁ <47 [V], 47 [V] < V ₂ Accordingly, it is possible to realize an off state and an on state, respectively.

Next, the ratio of signal transmission in the OFF state and the ON state (hereinafter referred to as switching ratio) will be explained. First, the switching ratio is determined by the energy barrier ΔE (21 in FIG. 3) formed at the heterojunction 11, and its value is approximately expressed as exp(-ΔE/ _KT ). Here, K is Boltzmann's constant, T
is the absolute temperature. And the energy barrier ΔE
As a first approximation, is the difference in the energy forbidden band width of two semiconductors forming a heterojunction. InP and
In the case of In _0.53 Ga _0.47 As, ΔE=0.53eV. As is well known, In _x Ga _1-x
In As _1-y P _y , by changing the compositions x and y, the energy gap width changes, and therefore the energy barrier ΔE can be changed.

FIG. 5 is an explanatory diagram showing the relationship between the energy barrier ΔE and the switching ratio, and the switching ratio can be determined for any energy barrier ΔE. Using this figure, the composition of InGaAsP can be determined for the required switching ratio value.

An embodiment of the optoelectronic switch according to the present invention using the photodiode 16 described above is shown in FIG. Furthermore, the signal A to be switched is added, and the optical signal 3 modulated thereby is emitted. This optical signal 3 is then received by the photodiode (PD) 16 described above. This PD16
The voltage _V2 or 0V is applied to the switch 5 while being switched arbitrarily. Note that it is clear that this bias voltage circuit can be replaced with a pulse voltage circuit. 25 is a detector.

As mentioned above, in the case of a bias voltage of 0V, no signal appears at the output terminal 15, and the voltage V ₂
The signal appears only when .

7 to 9 are explanatory diagrams showing data of experimental results, and FIG. 7 shows the switching ratio measured by the embodiment shown in FIG. 6 with respect to the frequency of signal A. Here, a spectrum analyzer was used as the detector 25 shown in FIG.

The photodiodes used were indium phosphide InP and indium gallium arsenide In _0.53
It is composed of Ga _0.47 As. switching ratio
It can be seen that a value of 85 dB or more is achieved at a frequency of 500 MHz or less.

Figure 8 shows the voltage applied to the photodiode at 0V.
This is the result of observing the change in switch operation over time as a pulse voltage varying from V2 to _V2 . In addition,
In this data, although there is a transient response due to a defect in the circuit, the switching speed is 100 nsec or less. By improving the bias circuit, the optoelectronic switch according to the present invention can switch a high frequency signal of 500MHz at a high speed of 10nsec, and the switching ratio can be increased to 80dB.
This shows that it is possible to do the above.

FIG. 9 is an explanatory diagram showing the result of switching a high-speed pulse signal as the electrical signal A. It can be seen that the high-speed pulse is switched at a high-speed switching speed of 10 nsec. Note that FIG. 8 and FIG. 9 use an oscilloscope as the detector 25 shown in FIG. 6.

Next, a method for manufacturing a PD or APD used in the present invention will be explained with reference to FIG. As an example, the case of In _0.53 Ga _0.47 As will be described. First, FIG. 10a shows a cross section of a semiconductor in which n-InP 8 and n-InGaAs 10 are sequentially grown on p ⁺ -InP 7 by the well-known liquid phase epitaxial method. The n-type impurity concentration was 1×10 ¹⁶ cm ⁻³ , and the thicknesses of n-InP and n-InGaAs were 2.5 μm and 2.0 μm, respectively. Here, the relationship between thickness and impurity concentration can be obtained from FIG. 4, as already stated. Next, a photosensitive resist film is applied to the surface of n-InGaAs 10 by a commonly known method, and after exposing this to light using an appropriate mask, the resist is developed, as shown in Figure 10b. As shown, the resist 26 remains partially. Next, for n-InGaAs10, H ₂ O, H ₂ SO ₄ ,
Using an etching solution containing H ₂ O ₂ in a well-known ratio, and for n-InP8, using an etching solution containing bromine (Br) and methyl alcohol in a well-known ratio. After etching and further removing the resist film on the n-InGaAs 10, a so-called mesa shape with a cross section as shown in FIG. 10c is formed. Here, the diameter of the mesa portion was 140 μmφ.

Finally, on the p ⁺ -InP7 side, an Au-Zn alloy, n
- After depositing an Au-Ge-Ni alloy on the InGaAs 10 side in a vacuum, electrodes 12 and 13 are formed by alloying treatment at 400° C. for 5 minutes.
FIG. 10d is a cross-sectional view of the completed photodiode.

Although a mesa structure has been described here, a planar structure is also easily possible. It may also be a structure in which a semiconductor layer with a large energy bandgap, called a so-called cap layer, is provided on the light incident surface. Furthermore, the crystal growth method is not limited to the liquid phase epitaxial growth method, and vapor phase epitaxial growth methods and molecular beam epitaxial growth methods are also considered, and the point is that the first semiconductor and the second semiconductor are It is obvious that a photodiode can be constructed as long as it is formed so as to satisfy the size relationship of the respective energy band gaps and has the layer structure described here. Also, for simplicity, InP and
Although we have described in detail an optoelectronic switch using a photodiode composed of In _0.53 Ga _0.47 As, combinations of semiconductors other than the above,
For example, if the first semiconductor with a pn junction is InP or
In _x Ga _1-x As _1-y P _y (0≦x≦1, 0≦y≦1), and the second semiconductor is In _x ′Ga ₁ with a smaller energy bandgap than the first semiconductor. _-x ′As _1-y ′P _y ′(0
A high-performance optoelectronic switch can also be realized with a photodiode in which ≦x'≦1, 0≦y'≦1).

If the photodiode has an avalanche image multiplication effect, the switching ratio and signal-to-noise ratio will be improved, and more desirable switching characteristics will be obtained.

In addition, in the explanation so far, it is also possible to completely reverse the p-type and n-type, and in this case,
You can read holes as electrons.

As explained above, in the optoelectronic switch according to the present invention, by changing the magnitude of the reverse bias voltage applied to the photodiode, it is possible to perform a high-speed switch with a large switching ratio relative to the electrical signal to be switched. It becomes possible to do this. As mentioned above, preliminary experimental results show that a 500MHz high-frequency signal can be switched at a switching speed of 10nsec, and the signal transmission ratio (switching ratio) between the off state and on state is over 80dB. Ta. In principle, it is possible to switch from a DC signal to a high-frequency signal up to several GHz, considering the modulation characteristics of the light source and the high-frequency characteristics of the photodiode, and the switching speed is determined by the junction capacitance of the photodiode. , it is also easily possible to set it to 10 nsec or less. Also, high-speed digital signal switching is possible as well.

As is clear from the above description, according to the present invention, (1) low power consumption, (2) simple circuit configuration, (3) high-speed switching is possible, and (4) the photodiode always has high impedance. Therefore, it is possible to construct an optoelectronic switch with easy impedance matching (5) and a sufficiently large signal switching ratio as in the conventional case.

Therefore, various applications that could not be realized with conventional optoelectronic switches are facilitated. That is, high-density integration is easily possible due to low power consumption, and even in a complicated switch structure, the device can be made smaller and more economical, and reliability can be further improved.

Another important effect is that, as mentioned above, high-speed switching operations on the order of nanoseconds are possible, allowing high-speed switching of microwave lines, millimeter-wave intermediate frequency switching, digital line switching, telephone line exchanges, and satellite communication lines. It can be applied to high-speed phase switching used in time-division multiplexed high-speed signal switching switch lasers, etc. In addition, in the explanation so far, we have only talked about the switch operation, but if the reverse bias voltage is changed continuously, the magnitude of the signal output also changes continuously, so it can be used as a variable attenuator with signal directionality. There are also applications.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing an example of a conventional optoelectronic switch, FIG. 2 is a sectional view showing the basic structure of a photodiode constituting the optoelectronic switch according to the present invention, and FIG. A diagram showing the energy band structure of a photodiode. Figure 4 is an explanatory diagram showing the relationship between n-InP impurity concentration and depletion layer width in the photodiode of Figure 2. Figure 5 is an explanatory diagram showing the relationship between the energy barrier ΔE and the switching ratio. FIG. 6 is a block diagram showing the basic configuration of an embodiment of the optoelectronic switch according to the present invention, and FIGS. 7 to 9 show experimental results of the optoelectronic switch according to the present invention. FIG. 7 is an explanatory diagram showing the data. FIG. 7 is the frequency characteristic of the switching ratio, FIG. 8 is the high frequency signal switch operation characteristic, FIG. 9 is the digital signal switch operation characteristic, and FIG. 10 is the photo of FIG. FIG. 2 is an explanatory diagram showing an example of a method for manufacturing a diode. DESCRIPTION OF SYMBOLS 1... Electrical-optical conversion element, 2... Optical-electrical conversion element, 3... Light, 4a, 4b, 4... Bias power supply, 5... Switch circuit, 7, 8, 10... Semiconductor, 9... ... pn junction, 11 ... heterojunction, 16
...Photodiode.

Claims (1)

[Claims] 1. A light source that is modulated by an electrical signal, an avalanche photodiode that receives an optical signal from the light source and converts it into an electrical signal, and a bias power supply that applies to the avalanche photodiode. In the optoelectronic switch configured by, the avalanche photodiode is
a - group or Si first semiconductor having a p-n junction; and a lattice constant that approximately matches that of the first semiconductor, and the energy on the minority carrier side is more discontinuous than that of the first semiconductor. The structure has a heterojunction formed with a second semiconductor of the - group or Ge, which is small in size, and the applied voltage of the bias power supply is sufficiently low so that the depletion layer of the avalanche photodiode does not reach the heterojunction. When the reverse bias voltage (including 0V) is applied, the depletion layer reaches the heterojunction or the second
An optoelectronic switch characterized in that the avalanche photodiode is turned off and on by switching to a sufficiently high reverse bias voltage to extend into the semiconductor. 2 An avalanche photodiode is formed in which a pn junction is formed in a first semiconductor of InP (indium phosphide), and has a lattice constant that approximately matches that of the first semiconductor and an energy band gap that falls within the range. In _x Ga _1-x As _1-y P _y (indium, gallium, arsenic, phosphorus; 0≦x≦1, smaller than that of the first semiconductor)
2. The optoelectronic switch according to claim 1, wherein the optoelectronic switch is constituted by an avalanche photodiode formed with a second semiconductor in which 0≦y≦1. 3 Place the avalanche photodiode and the pn junction in the first semiconductor of In _1-x Ga _x As _1-y P _y (Indium-Gallium-Arsenic-Phosphorus; 0≦x≦1, 0≦y≦1) formed, has a lattice constant that substantially matches that of the first semiconductor, and has an energy forbidden band width smaller than that of the first semiconductor.
Avalanche formed with a second semiconductor of In _1-x ′Ga _x ′As _1-y ′P _y ′ (indium, gallium, arsenic, phosphorus; 0≦x′≦1, 0≦y′≦1) 2. The optoelectronic switch according to claim 1, characterized in that it is constituted by a photodiode.