WO2001033287A1

WO2001033287A1 - Electroabsorption optical modulator and method for fabricating the same

Info

Publication number: WO2001033287A1
Application number: PCT/JP2000/003069
Authority: WO
Inventors: Masaki Kato; Yoshiaki Nakano
Original assignee: Center For Advanced Science And Technology Incubation, Ltd.
Priority date: 1999-10-28
Filing date: 2000-05-12
Publication date: 2001-05-10
Also published as: AU4432800A

Abstract

An electroabsorption optical modulator capable of achieving both negative chirping and polarization independence and having a simple structure. Barrier layers (21, 22) are provided in a position within the width of rectangular quantum wells (11, 12) and near one ends of the rectangular quantum wells (11, 12). As a result, the rectangular quantum wells (11, 12)are equivalently pre-biased. With this structure, negative chirping of an optical signal and polarization independence of the extinction characteristic are both achieved. Since the structure of the modulator is simple, the fabrication is relatively easy.

Description

Description Electro-absorption optical modulator and method for manufacturing the same

The present invention relates to a waveguide type electro-absorption optical modulator (hereinafter simply referred to as “electro-absorption optical modulator”) provided with a multiple quantum well, and in particular, it is possible to make a negative trap. The present invention relates to a simple electroabsorption type optical modulator. Background art

It has been proposed to use an electric field absorption type optical modulator as an optical modulator for modulating light transmitted through an optical fiber. Electroabsorption type optical modulator has the advantage of low-voltage driving, high extinction ratio, the modulation bandwidth of more than 5 0 GH _Z possible. Therefore, one of the applications of this optical modulator is to apply this modulator to a system that modulates the light transmitted through an optical fiber with a millimeter-wave band radio signal and sends it out through an optical fiber. Has been proposed. As another application example, it has been proposed to use it as a repeater for shaping and retiming of a pulse transmitted in an optical fiber.

Characteristics desired for an optical modulator for such an application include:

1. Independent of polarization in extinction characteristics (Sufficiently dependent on polarization direction)

2. Negative chirp can be applied to light passing through the optical modulator

There are two things. The reason that the former property is preferable is that light transmitted through ordinary optical fiber cannot maintain polarization, so if the extinction characteristic fluctuates excessively due to polarization, it is used as a modulator for the transmitted light. It is difficult. The latter characteristic is desired because the negative chirp in the optical modulator reduces the transmission distance of light as information (pulse) in a normal optical fiber. It can be extended.

Examples of the former technology that is oriented toward polarization independence include, for example, Japanese Patent Application Laid-Open No. 11-144,799 (Reference 1), “LE〇S'9710th annual meeting, Proceedings, P146--14. 7 (Literature 2), "ECOC '99 Proceedings II-I 72-73" (Literature 3), IEICE Technical Report PS 98-36 P49-54 (Literature 4), PS 9 9-11 P63-68 (Reference 5) and PS99-1 P1-6 (Reference 6). Examples of technologies aimed at low chirp include, for example, Japanese Patent Application Laid-Open No. H10-272732 (Reference 7), Proceedings of the 4th Federated Lecture Meeting of the Japan Society of Applied Physics P1 13 7 (Reference 8) , `` 1992 OSA Technical Digest Series Volume 4 320 / Itu 3 '' (Reference 9), Proceedings of the 59th Annual Meeting of the Japan Society of Applied Physics P1 13 7 (Reference 10). Something was done.

However, in the former technology, which aims at polarization independence, low or negative capping is not intended. Also, the latter technology, which aims at low chirp, has a problem that the structure is complicated because two thin barrier layers are provided inside each of the multiple rectangular quantum wells. In addition, the latter technique has a problem that it is difficult to obtain polarization-independent characteristics. Disclosure of the invention

The electro-absorption optical modulator capable of being negatively trapped according to claim 1, wherein the rectangular quantum well is located inside the width of the rectangular quantum well and near one end of the rectangular quantum well. A barrier layer for applying a pre-bias to the well is provided. As a result, it is possible to achieve both negative capping and polarization independence, and furthermore, there is an effect that the configuration is simple and the manufacturing is easy.

An electroabsorption optical modulator according to claim 2 is the electroabsorption optical modulator according to claim 1, wherein a portion corresponding to the rectangular quantum well in the first semiconductor is subjected to an extensional strain. Thereby, the polarization-independent characteristics can be further improved. The method for manufacturing an electro-absorption optical modulator according to claim 3 is to manufacture the electro-absorption optical modulator by changing a position of a barrier layer to select a chirp parameter from a range over a negative value. Configuration. This makes it possible to obtain an electro-absorption optical modulator that matches the system design conditions.

An electro-absorption optical modulator according to claim 4, wherein in the electro-absorption optical modulator according to claim 1 or 2, the amount of compressive strain applied to the semiconductor constituting the barrier layer or the second semiconductor is selected. Thus, the chirp parameter is selected. Thereby, the same effect as in claim 3 can be obtained.

A method of selecting a chirp parameter according to claim 5, comprising a rectangular quantum well formed by stacking a first semiconductor and a second semiconductor having a higher energy order. A barrier layer for applying a pre-bias to the rectangular quantum well is provided at a position inside the width of the rectangular quantum well and near one end of the rectangular quantum well; A chirp parameter is selected by applying a compressive strain to the semiconductor or the second semiconductor. Thereby, the same effect as in claim 3 can be obtained.The electroabsorption optical modulator according to claim 6 is the electroabsorption optical modulator according to claim 1 or 2, wherein By selecting the energy level of the second semiconductor, the cap parameter is selected. Thereby, the same effect as in claim 3 can be obtained.

A method of selecting a chirp parameter according to claim 7, comprising: a rectangular quantum well formed by stacking a first semiconductor and a second semiconductor having a higher energy order. A barrier layer for applying a pre-bias to the rectangular quantum well at a position inside the width of the rectangular quantum well and near one end of the rectangular quantum well; The configuration is such that the cap parameter is selected by selecting the level. This allows The same effect as in claim 3 can be obtained.

9. The method for manufacturing an electro-absorption optical modulator according to claim 8, wherein the manufacturing method according to claim 1 or 2, wherein the energy level of the barrier layer or the barrier layer is formed. By selecting the amount of compressive strain of the semiconductor to be manufactured, a configuration is adopted in which the chaper parameter is selected from a range over a negative value to manufacture the electroabsorption optical modulator. Thereby, the same effect as in claim 3 can be obtained. BRIEF DESCRIPTION OF THE FIGURES

FIG. 1 (a) schematically shows the energy band structure of the optical modulator according to the first embodiment of the present invention, and FIGS. (B) to (d) show the equivalents. This is the energy band diagram. FIG. 2 is an explanatory view corresponding to the enlarged view of FIG. (A) and illustrating the wave function of the carrier. FIG. 3 to FIG. 5 are explanatory diagrams for explaining the advantages of the optical modulator according to the present embodiment. FIG. 6 is a diagram schematically illustrating an energy band structure of an optical modulator according to a second embodiment of the present invention. FIG. 6 (a) is a comparative example, and FIG. 6 (b) is that of the second embodiment. Is shown. FIG. 7 is an explanatory diagram for explaining a carrier wave function when a voltage is applied. FIG. 8 is an explanatory diagram showing an absorption spectrum of the optical modulator. FIG. 9 is an explanatory diagram showing the modulation characteristics of the optical modulator. FIG. 10 is an explanatory diagram showing a result of calculation of a trap parameter using the amount of compressive strain applied to the semiconductor constituting the barrier layer as a parameter. FIG. 11 is an explanatory diagram showing a relationship between an insertion loss in an optical modulator and a capture parameter. FIG. 12 shows the modulation characteristics of the optical modulator. BEST MODE FOR CARRYING OUT THE INVENTION

An electro-absorption optical modulator according to a first embodiment of the present invention will be described below with reference to the accompanying drawings.

This optical modulator is composed of a first semiconductor (a so-called well layer) composed of InGaAs. A second semiconductor having a higher energy order and a composition of InAlAs (a semiconductor that constitutes a layer serving as a barrier on both sides of the well, and is referred to as a second semiconductor in this specification). It mainly has a laminated structure of. The composition of the first and second semiconductors may be, but not limited to, InGaAs / InP, InGaAlAs / InAlAs, and InGaAsP / InGaAsP. The first and second semiconductors form rectangular quantum wells 11 and 12 (existing for electrons and holes, respectively) in the energy band structure shown in FIGS. 1 (a) and 2. ing. Inside the widths of the rectangular quantum wells 1 1 and 1 2 (that is, the portion corresponding to the well layer) and near one end (the right end in FIG. 1) of each of the rectangular quantum wells 1 1 and 1 2, Thin energy barrier layers 21 and 22 are provided. The barrier layers 21 and 22 have a shape when grasped as an energy level or a potential. As a specific material configuration, a semiconductor serving as a barrier (for example, having the same composition as the second semiconductor) may be used. Semiconductor) is formed by laminating them sufficiently thinly. Pre-bias is applied to the rectangular quantum wells 11 and 12 by these barrier layers 21 and 22. Here, “pre-biased” means, as shown in Figs. 1 (b) to (d), that when the electric field F = 0, equivalently, it is simply a “rectangular (no barrier) rectangle. This means that the state is equivalent to the presence of an electric field in the quantum well. In this specification, “near one end of the rectangular quantum wells 11 and 12” means a position where such pre-bias can be applied. In this embodiment, as shown in FIG. 1 (a), the width (the thickness of each layer) of the barrier layers 21 and 22 is 0.9 nm, and the width of the rectangular quantum wells 11 and 12 is The width from the inner end to the barrier layers 21 and 22 is 9.4 nm for the wide one (ie the width of the wide well layer in Figure 1a) and 1.5 nm for the narrow one. And

The first semiconductor (well layer) constituting the rectangular quantum well 11 is composed of a semiconductor substrate used. A 0.36% elongation strain is applied to the plate (this configuration is the same as the conventional technology and is not shown. In this embodiment, the substrate is an InP substrate). Further, in the present embodiment, a portion corresponding to the rectangular quantum well 11 in the second semiconductor is subjected to 0.45% compressive strain with respect to the semiconductor substrate.

Other specific configurations in the present embodiment are basically the same as those described in References 1 to 10 described as the prior art, and the fabrication thereof is performed by the molecular beam epitaxy method (MBE). The detailed description is omitted because it is possible based on a known technique using the above method.

Next, the operation of the optical modulator according to the present embodiment will be described. In this optical modulator, since an equivalent pre-bias is applied, as shown in Literatures 2 to 6, the shift of the absorption edge wavelength with respect to the applied electric field regardless of the polarization direction is TE or TM. Can be almost equal. This reduces the dependence of the extinction characteristic on the polarization direction. That is, polarization-independent light modulation becomes possible. Further, in the optical modulator of the present embodiment, a portion corresponding to the rectangular quantum well 11 in the first semiconductor is given 0.36% of extensional strain. In addition, whether the polarization direction is TE or TM, the absorption edge wavelength can be made uniform regardless of the magnitude of the electric field. This makes it possible to achieve polarization independence over a wide wavelength range.

Further, in the present embodiment, 0.45% compressive strain is applied to the portion corresponding to the rectangular quantum well 11 in the second semiconductor, so that the average strain of the first and second semiconductors is reduced. This has the advantage that it can be made smaller and many quantum wells can be stacked. Further, in the optical modulator according to the present embodiment, it is possible to select a chirp parameter (sometimes referred to as a chirp amount) and to provide a negative chirp based on the following principle. First, as is known, the chirp parameter in this type of optical modulator can be expressed by the following equation (1).

β = (4π / λ) (Δη / Δα) (1) Here, λ: operating wavelength of the modulator, Δ η: change in refractive index, Δ Δ: change in loss.

Η and αα are linked by Kramers-Kronig and cannot be designed independently. As can be seen from equation (1), in order to obtain a negative chirp, at the operating wavelength, (1) Δα is positive (Δα> 0) and (2) the refractive index changes negatively (Δη <0). What is necessary is just to design a quantum well structure that satisfies the two conditions simultaneously. Here, the condition (1) is used because if the loss is not positive, the extinction function, which is the premise of the optical modulator, cannot be obtained. Also, in order to realize condition (2), it is necessary that the absorption coefficient in the region slightly shorter than the operating wavelength be significantly smaller than the increase in the operating wavelength. Although this fact is already known, using FIG. 3 (a) as an example, it will be further described that the operating wavelength is set to 1550 nm and slightly shorter than that (for example, 15 (0 to 154 nm), the decrease in the absorption coefficient when the applied voltage is changed from 0 V to 2 V is larger than the increase in the operating wavelength. You.

The present inventor has studied the wave function of the quantum well structure according to the present embodiment, and has obtained the following knowledge. This will be described with reference to FIG. First, when the electric field F = 0, the wave function is localized on the side of the wide well layer (the well layer on the left side of the barrier layers 21 and 22 in the figure). In Fig. 2, the wave function of electrons is denoted by E, and the wave function of heavy holes is denoted by HH. LH in Fig. 1 indicates the light function of the light hole. When a small electric field is applied to the quantum well with this configuration, the electron wave function shifts to the left in the figure. On the other hand, the wave function of the heavy hole seeps out of the well layer on the opposite side (ie, the narrower side) of the barrier layer 22 (see the wavy line in Fig. 2). As a result, the value of the overlap integral of the electron wave function and the hole wave function becomes smaller, and the exciton absorption peak sharply decreases. Furthermore, the relationship between the absorption coefficient and the wavelength was determined in the TE and TM modes by calculation. The results are shown in Fig. 3 (a) (b ). From this result, as described above, it is understood that the condition of the above-mentioned ② “the refractive index changes negatively” is satisfied. Since the absorption coefficient is positive, the condition (Δhi> 0) is satisfied. In addition, the change in the absorption coefficient in the TE and TM modes is almost the same, indicating that polarization independence has been achieved. Furthermore, the relationship between the applied voltage and the insertion loss and the chirp parameters was determined as shown in Fig. 3 (c) and (d). From this, it can be seen that if the applied voltage is 0.3 V or more, the chirp parameter can be negatively encoded. The insertion loss at an applied voltage of 0.3 V is about 7 dB. Note that the calculation conditions here are based on the assumption that the quantum well of the present embodiment is used to form an optical modulator having 10 periods of wells and a total well thickness of 0.16 / xm. is there. For Figures 3 (c) and (d), a high-mesa waveguide with a width of 2.5 μπι and an element length of 75 μm (confinement coefficient of 0.278 for TE, 0.234 for TM) is assumed. ing.

Negative trapping is possible even when a voltage is applied to a normal rectangular quantum well, but the insertion loss in that case depends on conditions, but is generally around 10 to 12 dB. There is a problem that it gets bigger. On the other hand, according to the optical modulator of the present embodiment, a negative trap can be realized while keeping the insertion loss relatively small at 7 dB.

Examples of the advantages of the negative cap will be further described with reference to FIGS. 4 and 5, which are known characteristics. According to Fig. 4, it can be seen that the signal transmission distance by the optical fiber can be extended by the negative cap. FIG. 4 shows the relationship between the chirp parameters and the 3 dB bandwidth (“Mitigation of Chromatic Dispersion Effects Employing Electroabsorption Modulator-Based TransnuttersJ IEEE Photoron. Technol. Lett., Vol. 11 p. (See 883-885, 1999.) Also, according to Fig. 5, it is possible to shift the position (dip) where the loss increases due to the increase in frequency to the higher frequency side, facilitating application to high-frequency signals. ("Simple Measurement of Fiber Dispersion and Chirp Parameter or Intensity Modulated Li ght Emitter J J. Li ghtwave Technol., Vol. 11, p. 1937-1940, 1999). Note that, in FIG. 5, α represents a cap parameter.

Further, according to the optical modulator of the present embodiment, since the basic property of being polarization independent is maintained, there is an advantage that the negative chirp can be achieved together with the polarization independent.

Further, in the optical modulator of the present embodiment, by adjusting the positions of the barrier layers 2 1 and 2 2 at the time of manufacturing, it is possible to adjust the degree of the wave function of the heavy hole extruding into a narrow well, The cap parameter can be controlled from a positive value to a negative value. In other words, there is an advantage that the cap parameter can be selected according to the system design conditions.

Furthermore, the conventional quantum well structure directed to a low-cap or negative-cap was provided with two barrier layers and had a complicated configuration. However, the optical modulator of the present embodiment has a more complicated configuration than the conventional one. In addition, since it is possible to achieve negative cap and polarization independence with a simple configuration, there is an advantage that it is easy to manufacture and can exhibit excellent characteristics.

Next, a second embodiment of the present invention will be described with reference to FIGS. Since the basic configuration of the optical modulator according to the second embodiment is the same as that of the first embodiment, the same reference numerals are used for the same elements, and the description will not be repeated. In the second embodiment, InGaAs with 0.4% tensile strain is used as the first semiconductor, and InAlAs with 0.7% compressive strain is used as the second semiconductor. Have been. The composition of the semiconductor for forming the barrier layer 21 is the same as that of the second semiconductor and has the same compressive strain. The energy level diagram for this configuration is shown in Fig. 6 (b). The actual widths (thicknesses) of the rectangular quantum wells 1 1 and 1 2 and the barrier layers 2 1 and 2 2 are as shown in FIG. For comparison, FIG. 6 (a) shows the energy level in a state where the semiconductor constituting the barrier layer and the second semiconductor are both lattice-matched (that is, the compressive strain is eliminated). Comparing the two, Fig. 6 (b) shows the energy level for heavy holes, in particular. It can be seen that the bell is low (that is, the well is shallow). Table 1 below shows an example of the relationship between the amount of compressive strain and the well depth for electrons, heavy holes, and light holes. It can be seen that the well becomes shallower (ie, the barrier becomes lower) as the compressive strain increases.

Table “! 0.4% 3 Long strain InGaAs / compressive strain In A 1 As (0%, 0.35

0.7%) quantum well electrons (E), heavy holes (HH),

Tohole (LH) well depth. The unit is eV.

The effect of the shallow well (lower barrier) will be explained with reference to FIG. FIG. 7 (a) corresponds to the configuration of FIG. 6 (a), and FIG. 7 (b) corresponds to the configuration of FIG. 6 (b). Figures 7 (a) and 7 (b) show the wave functions (before application: solid line, after application: broken line) of carriers (electrons E, heavy holes HH, light holes LH) before and after voltage application. As can be seen from the above, in the second embodiment, the wave function of the hole (especially the snake hole) greatly extrudes beyond the barrier layer 21 in the direction of the narrow well (to the right in the figure). Then, the overlap of electron-hole wave functions becomes smaller. This is because the depth of the well with respect to the hole was reduced. FIG. 8 (b) shows the absorption spectrum of the optical modulator of the present embodiment. Figure 8 (a) is shown for comparison, and shows the absorption spectrum of the optical modulator with the configuration of Figure 6 (a). It can be seen that, compared to the case of lattice matching, in the present embodiment, for both TE and TM polarized light, the absorption increases near 1500 nm compared to the increase in absorption near the operating wavelength of 1550 nm. A large decrease in absorption is seen. Therefore, it is considered that negative capping is possible. FIG. 9 shows the modulation characteristics of the optical modulator according to the present embodiment. put it here And _d show the amount of wavelength detuning from the exciton absorption peak. The solid line is the characteristic for TE polarized light, and the broken line is the characteristic for TM polarized light. Although the extinction characteristic is slightly lower than that in the case of lattice matching (not shown), it is in a range where there is no practical problem. In addition, the polarization dependency does not deteriorate. Therefore, even in the present embodiment, the modulation characteristics themselves are hardly affected.

FIG. 10 shows the results of calculation of the chip parameter using the amount of compressive strain applied to the semiconductor constituting the barrier layer as a parameter. The numbers shown with% in the figure are the amounts of compressive distortion. FIG. 10 (a) shows an example of wavelength detuning _d = 50 nm, FIG. 10 (b) shows an example of fly _d = 60 nm, and FIG. 10 (c) shows an example of fly _d = 70 nm. It can be seen that the curve of the chirp parameter shifts to the left for both TE polarized light and TM polarized light as the amount of compressive strain increases. Therefore, it can be seen that a small (some negative) trap parameter can be realized with a small applied voltage (ie, a small insertion loss) by applying compressive strain. As described above, by selecting the amount of compression distortion, the cap parameter can be selected from a negative range. Also, based on this selection, an optical modulator can be designed and manufactured.

Figure 11 shows the relationship between the insertion loss and the cap parameter in the optical modulator. Fig. 11 (a) shows an example of wavelength detuning _d = 60 nm, and Fig. 11 (b) shows an example of wavelength _d = 70 nm. The numbers shown in% in the figure are the amounts of compressive strain applied to the semiconductor constituting the barrier layer. The characteristics indicated by RQW in the figure indicate that in a rectangular quantum well without a barrier layer, 0.4% tensile strain was applied to the first semiconductor and 0.7% compressive strain was applied to the second semiconductor. This is for reference only. It can be seen that as the amount of compression is increased, the insertion loss required to realize a negative chirp decreases. For example, as indicated by the arrow in FIG. 11 (b), the insertion loss for zero-chip is improved by 9 dB from 12 dB to 3 dB. FIG. 11 shows the characteristics of the pre-biased quantum well and the RQW of the present embodiment. Comparing the characteristics of rectangular quantum wells, at zero-cap, RQW still requires a large insertion loss of 12 dB for zero-cap. The reason is that in the RQW, the phenomenon of localization of the hole wave function as shown in Fig. 7 is unlikely to occur, and the overlap between the electron wave function and the hole wave function is large.

Fig. 12 shows the modulation characteristics. FIG. 7A shows the case of the present embodiment, and FIG. 7B shows the case of the above-described RQW. Obviously, this embodiment has a good polarization independence. Further, in the present embodiment, a large extinction ratio of 15 dB is obtained with an applied voltage of 2 V over a wide wavelength band of 20 nm.

In this embodiment, the chirp parameter can be selected by selecting the amount of compressive strain to be applied to the semiconductor constituting the barrier layer, whereby the desired chirp parameter (a negative value is also possible. It is possible to obtain an optical modulator having:

In the above-described second embodiment, compressive strain is applied to the semiconductor constituting the barrier layer. However, the present invention is not limited to this. By lowering the energy level of the semiconductor constituting the barrier layer, Negative chirp associated with the localization of heavy holes due to the application becomes possible. As an example of a means for lowering the energy level of a semiconductor, use of InAlGaAs as the semiconductor constituting the barrier layer can be considered. That is, by selecting the energy level of the semiconductor constituting the barrier layer, the selection of the chirp parameter becomes possible.

Furthermore, in the second embodiment described above, attention has been paid to the semiconductor constituting the barrier layer. However, the present invention is not limited to this. For example, compressive strain may be applied to the second semiconductor (parts other than the barrier layer) or the energy level may be reduced. As a result, the localization of the heavy holes accompanying the voltage application is promoted, and a negative chirp can be achieved. However, the operation of only the portion other than the barrier layer has relatively little effect on the chip parameters. It is thought that it is good to carry out the operation with the layer.

Each of the embodiments described above is merely an example, and the scope of the invention is not limited thereto. Industrial applicability

According to the present invention, it is possible to provide an electro-absorption optical modulator capable of achieving both a negative gap and polarization independence.

Claims

The scope of the claims

1. An electro-absorption optical modulator having a rectangular quantum well formed by laminating a first semiconductor and a second semiconductor having a higher energy order, wherein the inside of the width of the rectangular quantum well is And a barrier layer for applying a pre-bias to the rectangular quantum well is provided at a position near one end of the rectangular quantum well. Modulator.

2. The electroabsorption optical modulator according to claim 1, wherein a tensile strain is applied to a portion of the first semiconductor corresponding to the rectangular quantum well.

3. The method for manufacturing an electro-absorption optical modulator according to claim 1 or 2, wherein a position of the barrier layer is selected to select a chirp parameter from a range over a negative value. A method of manufacturing an electro-absorption optical modulator, comprising manufacturing the electro-absorption optical modulator.

4. The cap parameter is selected by selecting the amount of compressive strain applied to the semiconductor constituting the barrier layer or the second semiconductor, and the cap parameter is selected. Electro-absorption optical modulator.

5. An electro-absorption optical modulator having a rectangular quantum well formed by laminating a first semiconductor and a second semiconductor having a higher energy order, wherein an inner side of the width of the rectangular quantum well is provided. And a barrier layer for applying a pre-bias to the rectangular quantum well is provided at a position near one end of the rectangular quantum well, and compressive strain is applied to a semiconductor or the second semiconductor constituting the barrier layer. Selecting a cap parameter by selecting the

6. The cap parameter is selected by selecting an energy level of the barrier layer or the second semiconductor. Item 3. The electro-absorption optical modulator according to item 2 or 2.

7. An electro-absorption optical modulator having a rectangular quantum well formed by stacking a first semiconductor and a second semiconductor having a higher energy order, wherein the width of the rectangular quantum well is within the width of the rectangular quantum well. And providing a barrier layer for applying a pre-bias to the rectangular quantum well at a position near one end of the rectangular quantum well, and selecting an energy level of the barrier layer to select a chirp parameter. The selection method of the cap parameter which features.

8. The method for manufacturing an electro-absorption optical modulator according to claim 1 or 2, wherein the energy level of the barrier layer or the amount of compressive strain of a semiconductor forming the barrier layer is selected. A method for manufacturing an electro-absorption optical modulator, wherein the electro-absorption optical modulator is manufactured by selecting a cap parameter from a range over a negative value.