JPH0990440A - Optical switch driving method and optical switch - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce the size of a device by dispensing with the making of the optical path lengths of two optical paths for control light being in a Mach- Zehnder type light control optical switch large. SOLUTION: A control light is inputted to nonlinear waveguides 6, 7 via an input port for control light 9, a 3db coupler 3 and couplers for control light 4, 5 to change refractive indexes of the waveguides. A signal light is inputted from an input port 8 and is bisected to be guided into arms A, B. After the signal light receives phase changes by being transmitted through nonlinear waveguides 6, 7, signals interfer and are multiplexed with each other in a 3db coupler 2 to be outputted from either of output ports 10, 11. The nonlinear waveguides 6, 7 are constituted of materials whose band edge energies are respectively E1 , E2 and are made so that relations of hν-E1 <3kBT/2 and hν-E2 >3kBT/2 are established among the band edge energies E1 , E2 , the temp. T of nonlinear waveguides and the frequency ν of the control light.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an optical switch driving method and an optical switch for controlling light with light, and more particularly to a Mach-Zehnder type optical switch used as an optical control element in the fields of optical fiber communication and optical information processing. Drive method and optical switch.

[0002]

2. Description of the Related Art In order to increase the speed of optical fiber communications and optical information processing systems, it is essential to increase the speed of elements that perform optical control. As a light control element that can be used for such purposes,
An all-optical switch that controls light by light is drawing attention.
In particular, the Mach-Zehnder type waveguide element that can control the route of the signal light by the control light and can be cascaded makes it possible to encode / decode ultrahigh-speed signals by combining (multiplex) / dividing (demultiplex). It is regarded as important as a device.

In order to realize such an all-optical device, the optical nonlinear characteristics of the medium are used. On this occasion,
Either the non-resonant optical nonlinearity in which the control light is in the non-absorption wavelength region of the optical nonlinear material or the resonant optical non-linearity in which the control light is in the absorption wavelength region of the optical nonlinear material is used.

Non-resonant optical non-linearity is switched on /
Although the switch-off time is short, a very high power is required for the control light, and there is a difficulty in downsizing the optical switch module. On the other hand, the resonance-type optical nonlinearity enables low-power operation, but the switch-off time is limited to the longitudinal relaxation time of the excited carriers (carrier life; usually about several nanoseconds), and there is a difficulty in increasing the speed. Therefore, an all-optical type optical switch that realizes a switch-off time that is not limited by the longitudinal relaxation time of excited carriers while making use of the characteristic that switching is possible with low control light power is disclosed in JP-A-7-20510.
Has been proposed.

This conventional example is based on the Mach-Zehnder interference system 2
At least a part of one optical path is composed of a semiconductor material exhibiting optical non-linearity, and a single control light is divided into two parts and introduced with a certain time difference. Shown in 1. As shown in FIG. 1, the signal light input from the input port 8 is divided into two by the 3 dB coupler 1, propagates in the arm A and the arm B, and is divided into the first and second arms provided in the respective arms. The nonlinear waveguides 6 and 7 undergo a phase change according to their respective refractive indexes.

On the other hand, the control light input from the control light input port 9 is divided into two by the 3 dB coupler 3, and then the first,
It is introduced into the arms A and B through the second control light input couplers 4 and 5. Here, the distance from the 3 dB coupler 3 to the first and second control light input couplers 4 and 5 is one so that the control light is input to the arms A and B with a certain time difference (for example, 1 ps). Has been made long. Arm A,
The control light introduced into B is then input to the first and second nonlinear waveguides 6 and 7.

The non-linear waveguides 6 and 7 are made of GaAs (wavelength corresponding to band edge energy: 870 nm). On the other hand, the signal light has a wavelength of about 900 nm and the control light has a wavelength of about 900 nm. Light having a wavelength of about 865 nm is used. Therefore, the signal light passes through the nonlinear waveguides 6 and 7 without being absorbed, and the control light is absorbed here and changes its refractive index. The signal lights propagating through the arms A and B interfere with each other in the 3 dB coupler 2 and are combined.

The destination of the signal light which is interfered and multiplexed in the 3 dB coupler 2 is determined by the relative phase of the light propagating through both arms. That is, when the optical lengths of both arms (propagation distance of light in consideration of the refractive index) are exactly the same, the signal light after the combination has the first and second output ports 10 and 11.
Is divided into two again. However, when the optical length of the arm A is ¼ wavelength shorter than that of the arm B, all the multiplexed signal lights are guided to the first output port 10. On the contrary, when the optical length of the arm A is ¼ wavelength longer than that of the arm B, all the signal light is guided to the second output port 11. Since the optical length is given by the actual waveguide length × refractive index, by changing the refractive index of the nonlinear waveguide with control light, 3d
The B coupler 2 can be caused to switch the signal light.

FIG. 4 schematically shows changes in control light intensity and refractive index in two nonlinear waveguides of the conventional Mach-Zehnder interference system, and changes in signal light intensity output from the first and second output ports with time. FIG. In the illustrated example, the input signal light is a direct current having a constant voltage for simplicity. Further, all the signal light is guided to the second output port 11 when the control light is not incident. When the control light is input, the control light is first input to the side of the first control light input coupler 4 having a short waveguide after the 3 dB coupler 3, and this light is absorbed by the first nonlinear waveguide 6 and refracted. Change the rate. As a result, the relative phase of the signal light from both arms input to the 3 dB coupler 2 changes, and all the signal light is switched so as to be output from the first output port 10.

The control light is input to the second control light input coupler 5 with a delay of a fixed time (for example, 1 ps), whereby the second control light is input.
Carriers are generated in the nonlinear waveguide 7, and the refractive index thereof changes. Thus, at this point, the carriers generated in the first nonlinear waveguide 6 remain without being almost extinguished, so that the refractive index also hardly changes from the value changed by the incident control light. Therefore, when the control light is incident on the second nonlinear waveguide, there is no difference in the refractive index between the arms A and B, and the phase difference between the two signal lights input to the 3 dB coupler 2 is the same as before the control light is incident. Will be the same. Therefore, the signal light is again output only from the second output port 11.

Since the refractive indices of the first and second nonlinear waveguides 6 and 7 return to their original values at the same rate of change, the signal light continues to be output from the second output port in the process. After the elapse of the longitudinal relaxation time (on the order of nanoseconds), the carriers generated by the incident control light disappear, and the refractive indices of the nonlinear waveguides 6 and 7 also return to their original values. As described above, in the above-described conventional example, by adjusting the timing of the control light introduced into each arm, a sufficiently fast switching operation can be performed regardless of the vertical relaxation time.

[0012]

However, in the above-mentioned conventional optical switch, the temperature state of the element and the temperature of the generated carrier are only changed by injecting the control light having the wavelength of the band edge energy into the nonlinear waveguide. Had no consideration, it had the following problems to be solved.

The first problem is that the optical non-linearity that generally occurs in an optical non-linear material, as assumed in the above publication, changes monotonically with time so that it can be described by a single time constant. Rather, it comes from having components with small transient time constants. It has been theoretically clarified that the component having the small time constant is mainly caused by the temperature change of carriers due to photoexcitation.

Due to this transient optical non-linear component, the output waveform of the optical switch from the optical switch does not have a smooth waveform as shown in the above publication, but has a distorted waveform. The partially enlarged view of FIG. 4 schematically shows this. The time domain in which this transient optical nonlinear response occurs is usually the time for temperature relaxation of the carrier after the control light is incident, and a transient response lasting for about the energy relaxation time (on the order of 1 ps) based on carrier-phonon scattering occurs. . This transient optical non-linearity causes distortion in the optical switch waveform of the above method.

A second problem to be solved is that in the above optical switch, it is necessary to divide a single control light into two arms and introduce them into two arms with a finite time difference. This finite time difference In order to provide the optical path difference, it is necessary to give an optical path difference to the two control light optical paths, but the physical size of the device becomes large because of the optical path difference. For example, it is difficult to reduce the physical size of the device to less than 100 microns because it is necessary to provide an actual optical path difference of about 100 microns to realize switching of 1 picosecond.

These problems can be solved by considering the temperature T of the optical nonlinear semiconductor material provided in the two optical paths of the Mach-Zehnder interference system, its band edge energy E, and the frequency ν of the control light to be input. That is, an optical switch driving method that sets these to appropriate conditions can prevent the switch output waveform from being distorted, or can perform switching without requiring a physical optical path difference. Make the switch feasible.

[0017]

In the optical switch driving method according to the present invention for solving the above-mentioned problems, at least a part of each of the two arms constituting the Mach-Zehnder system is made of an optical nonlinear material. A method of driving a Mach-Zehnder type optical switch for controlling switching of signal light by control light, which has a means for introducing one control light into the optical nonlinear material portions of both arms of a Mach-Zehnder system in approximately two halves. The relationship between the band edge energy E of the optical nonlinear material and the frequency ν of the control light is determined according to the temperature T of the optical nonlinear material.

More specifically, the optical non-linearity materials of the two arms are made of the same material having the band edge energy E ₀ , and the control light divided into two parts has the optical non-linearity of the two arms with a certain time difference. Between the band-edge energy E ₀ , the temperature T of the optically nonlinear material, and the frequency ν of the control light, which is introduced into the material, hν−E ₀ = 3k _B T / 2 (where h is Planck's constant , K _B are established such that the Boltzmann constant) is established.

Alternatively, the optical non-linear materials of the two arms are composed of different materials having band edge energies E ₁ and E ₂ , respectively, and band edge energies E ₁ and E 2.
Between the temperature T ₂ and the optical nonlinear material and the frequency ν of the control _{light, hν-E 1 <3k B} T / 2 and,, hν-E _2>
Relationship 3k _B T / 2 is made to stand by.

[0020]

BEST MODE FOR CARRYING OUT THE INVENTION Next, embodiments of the present invention will be described. FIG. 1 is a plan view showing a schematic configuration of a Mach-Zehnder type optical switch according to the present invention. In the first embodiment, the configuration of the optical switch itself is the same as in the case of the conventional example described above, so redundant description will be omitted. In the first embodiment, GaAs, for example, is used as the material of the nonlinear waveguides 6 and 7 that exhibit optical nonlinearity.

As described above, the transient optical nonlinear component is caused by the carrier temperature change caused by the optical excitation. Therefore, in order to overcome this, it is only necessary to prevent the temperature change of the carrier. For this purpose, the energy hν of the incident control light (h: Planck's constant, ν: optical frequency) is the average energy E ₀ + 3k of all the carriers.
_B T / 2 [k _B : Boltzmann's constant, E ₀ : bandgap energy of the nonlinear waveguide material, T: temperature of the nonlinear waveguide (or the optical switch)].

Therefore, the frequency ν of the control light, the temperature T of the nonlinear waveguide, and the band edge energy E ₀ of the semiconductor material of the nonlinear waveguide are controlled so that hν = E ₀ + 3k _B T / 2 (1) is satisfied. do it. Thermal energy 3
k _B T / 2 is about 40 meV at room temperature. The method according to the first embodiment is disclosed in
By adding the condition as shown in the equation (1) to the optical switch proposed in Japanese Patent No. 20510, it is possible to prevent the optical switch waveform distortion due to the transient optical nonlinearity based on the carrier temperature change.

Next, a second embodiment will be described. The optical switch used in the second embodiment is also the one shown in FIG. 1, but it is different from the case of the first embodiment in the following points. The optical path lengths from the 3 dB coupler 3 to the arms A and B are substantially equal (however, there may be an appropriate optical path length difference). As semiconductor materials forming the first and second nonlinear waveguides 6 and 7, materials having different bandgap energies are used.

The operating principle of the optical switch proposed in Japanese Unexamined Patent Publication No. 7-20510 is that two optical paths (arm A,
B) is to provide a short time difference for the phase change of the light due to the change in the nonlinear refractive index, and use the time difference of the phase change for the light switching. Therefore, even if the control light is incident on the arms A and B at the same time, it is sufficient to have a mechanism that automatically causes a transient difference in the nonlinear refractive index change occurring in the arms A and B.

In the second embodiment, as this mechanism, the above-mentioned optical non-linear characteristic due to the carrier temperature change due to light absorption is positively utilized. The band edge energies of the semiconductor materials in the arms A and B that exhibit optical nonlinearity are E ₁ and E ₂ , respectively. Now, a so that _{hν-E 1 <3k B T} / 2 (2) hν-E 2> 3k B T / 2 (3), the temperature T of the frequency [nu, the semiconductor material of the control light
When the band edge energies E ₁ and E ₂ are controlled, carrier cooling occurs in the semiconductor medium in the arm A and carrier heating occurs in the semiconductor medium in the arm B. The sign of the transient optical nonlinearity due to this, that is, the sign of the refractive index change is opposite.

After the energy relaxation, optical nonlinearity due to ordinary band-filling occurs in the semiconductor material in each optical path, which has the same sign and magnitude.
Therefore, the optical nonlinearity in both optical paths, that is, the difference component of the nonlinear refractive index change, lasts for the time period during which the transient optical nonlinearity associated with the carrier temperature change lasts.

As described above, the energy relaxation time is approximately 1 picosecond, so high-speed optical switching of approximately 1 picosecond is possible (see FIG. 3). The magnitude of the optical nonlinearity generated in this way is by no means large, but the optical phase change in both optical paths occurs in proportion to the optical path length in the nonlinear medium, so if the length of the nonlinear medium is selected appropriately. It is possible to obtain sufficient switching characteristics. In order to utilize the above principle, the band edge energy of the semiconductor medium in the optical path A and the band edge energy of the semiconductor medium in the optical path B need to be different from each other. It can also be controlled by the confinement effect.

[0028]

Next, embodiments of the present invention will be described with reference to the drawings. [First Embodiment] FIG. 1 is a schematic configuration diagram of a Mach-Zehnder type optical switch to which the method of the first embodiment of the present invention is applied. The configuration and the basic operation are the same as in the case of the conventional example, and the duplicated description will be omitted. Input / output ports (8 to 11), 3 dB couplers 1 to 3, control light input couplers 4 and 5 and waveguides connecting them are composed of optical fibers, and nonlinear waveguides 6 and 7 are GaAs core / AlGaAs cladding. It is composed of a semiconductor waveguide. Of course, all components can be made of semiconductors.

Here, the signal light has an absorption edge of 870.
Light having a wavelength of about 900 nm that does not cause optical absorption in an optical nonlinear material (GaAs) having a wavelength of about nm is used. On the other hand, as the control light, light having energy higher than the absorption edge energy at which light absorption occurs in the nonlinear waveguide material (GaAs) is used. In the method of the first embodiment, the light wavelength of the control light is selected not only to be shorter than the absorption edge wavelength, but also to satisfy the condition of the above formula (1). As mentioned earlier, since the thermal energy 3k _B T / 2 in the room is the size of about 40 meV, when operating the present device at room temperature, the optical wavelength of about 846nm as a light wavelength of the control light The light that the user has may be used. Actually, according to the result of the experiment, when the control light having the wavelength of 820 nm is used, the distortion of the switching waveform as shown in FIG. 4 is observed, whereas the control light having the wavelength of 846 nm is used. If so, no such distortion was observed. In this case, the wavelength of the control light is selected so as to satisfy the expression (1), but of course the element temperature and the band edge energy may be controlled so as to satisfy the expression (1).

[Second Embodiment] Next, a second embodiment of the present invention will be described. In the present embodiment, the mixed crystal ratios of the semiconductor materials forming the first and second nonlinear waveguides 6 and 7 are changed to make the band edge energies of both materials different (E ₁ and E ₂ respectively). ), And the light wavelength of the control light is selected so as to satisfy the above expressions (2) and (3). FIG. 2 shows the mutual relationship between the light energy of the control light and the band edge energy of both materials. Now, it is assumed that the first nonlinear waveguide 6 is made of GaAs. The band edge energy E ₁ of GaAs is about 1429 meV (wavelength: 870).
nm). Considering room temperature operation, the average energy of carriers in GaAs is E ₁ + (3k _B T / 2) = 14.
The size is about 69 meV (wavelength: 846 nm).
The other nonlinear waveguide 7 is made of an AlGaAs mixed crystal, and its band edge energy is larger than E ₁ by 20 meV, E _{2 to} 1449 meV (wavelength: 858 nm).
So that To this end, Al _x Ga _1-x As
The mixed crystal ratio x defined by is set to x = 0.016.

At this time, the average energy of the carriers in the AlGaAs mixed crystal is E ₂ + (3 k _B T / 2) = 1489 me.
The size is about V (wavelength: 835 nm). In this situation, when light having an optical energy of 1459 meV (wavelength: 840 nm) is used as the control light, the energy of the control light is larger than the average energy of carriers in one of the nonlinear waveguides (GaAs), which causes carrier heating, In the other nonlinear waveguide (AlGaAs), carrier cooling is caused because it is smaller than the average energy of carriers.

As described above, the sign of the transient nonlinear refractive index change accompanying this is opposite. In FIG. 3, the first nonlinear waveguide 6 is GaAs and the second nonlinear waveguide 7 is AlG.
An example of the time waveform of the refractive index change in the case of aAs is shown. In fact, with such a configuration, the nonlinear waveguides 6 and 7
It was confirmed that optical switching of about 1 picosecond was possible when the control light was simultaneously irradiated on.

Although the above embodiments have been described by taking a specific semiconductor material as an example, the method of the present invention is not limited to the materials of these embodiments. For example, in the second embodiment, the nonlinear refractive index change accompanying the actual excitation of GaAs or AlGaAs optical carriers has been described as an example.
It can also be applied to all-optical switches using other semiconductor materials such as P, InGaAs, and InGaAsP. In the second embodiment, the lengths of the optical paths from the 3 dB coupler 3 to the couplers 4 and 5 do not necessarily have to be equal. If there is a difference in the optical path lengths, the lengths of the first and second nonlinear waveguides 6 and 7 may be adjusted accordingly.

[0034]

As described above, the method for driving an optical switch according to the present invention, the band edge energy of the nonlinear material of the two optical paths of the Mach-Zehnder interference system, its temperature, and the wavelength of the control light incident on the material. And are controlled so as to satisfy a predetermined relationship, the method of the present invention can realize an optical switch in which the switch output waveform is not distorted and an optical switch in which the device size is not limited by the optical path difference.

[Brief description of drawings]

FIG. 1 is a schematic configuration diagram of a Mach-Zehnder type optical switch for explaining an example of the present invention and a conventional example.

FIG. 2 is a diagram showing a mutual relationship between optical energy of control light and band edge energy of an optical nonlinear material.

FIG. 3 is a diagram showing changes over time of respective amounts in two optical nonlinear material portions of a Mach-Zehnder type optical switch for explaining an example of the present invention.

FIG. 4 is a diagram showing an output waveform of each part of a conventional Mach-Zehnder type optical switch.

[Explanation of symbols]

 1, 2, 3 3 dB coupler 4 first control light input coupler 5 second control light input coupler 6 first nonlinear waveguide 7 second nonlinear waveguide 8 input port 9 control light input port 10 first output port 11th 2 output ports

Claims (5)

[Claims] 1. At least a part of each of the two arms constituting the Mach-Zehnder system is made of an optical non-linear material, and one control light is divided into approximately two minutes in the optical non-linear material portions of both arms of the Mach-Zehnder system. A method of driving a Mach-Zehnder type optical switch for controlling switching of signal light by control light, which comprises means for introducing the optical signal, and a band edge energy E of the optical nonlinear material according to a temperature T of the optical nonlinear material. And a frequency ν of the control light are defined. 2. The band edge energy of the optical nonlinear material of at least one of the two arms is adjusted by introducing strain into the optical nonlinear material or by exerting a quantum confinement effect. The method of driving an optical switch according to claim 1. 3. The two arms of the optically non-linear material are made of the same material with band edge energy E _0.
The divided control light is introduced into the optical nonlinear material of the two arms with a certain time difference, and the band edge energy E ₀ , the temperature T of the optical nonlinear material and the frequency ν of the control light are
Between, _{schematic, hν-E 0 = 3k B} T / 2 ( where, h is Planck's constant, k _B is Boltzmann's constant) optical switch according to claim 1, wherein a relationship is established Driving method. 4. The optical non-linearity material of the two arms is composed of materials of different configurations having band edge energies E ₁ and E ₂ , respectively, and band edge energies E ₁ and E ₂
Between the temperature T and the frequency of the control light ν optical nonlinearity material _{and, hν-E 1 <3k B} T / 2 and,, hν-E _2> 3
The method of driving an optical switch according to claim 1, wherein the relationship of k _B T / 2 is established. 5. At least a part of each of the two arms constituting the Mach-Zehnder system is made of different optical nonlinear materials having band edge energies E ₁ and E ₂ , respectively, and the optical components of both arms of the Mach-Zehnder system are provided. A Mach-Zehnder type optical switch for controlling switching of signal light by control light, which has a means for introducing one control light into the non-linear material part by dividing it into two parts. The length is selected so that the phase of the light propagating through one arm leads when the control light is not input, and the phase of the light propagating transiently through the other arm leads when the control light is input. Optical switch characterized by.