JP2023126083A - Semiconductor optical element and method for manufacturing the same - Google Patents

Abstract

To provide a semiconductor optical element that is easy to manufacture and can increase coupling efficiency, and a method for manufacturing the same.SOLUTION: A semiconductor optical element comprises a substrate having a first optical waveguide formed of silicon, and a semiconductor element having a second optical waveguide formed of a III-V compound semiconductor, which is bonded to the top surface of the substrate, and the first optical waveguide and the second optical waveguide form a directional coupler.SELECTED DRAWING: Figure 1B

Description

The present disclosure relates to a semiconductor optical device and a method for manufacturing the same.

There is a known technique for bonding a semiconductor element made of a III-V compound semiconductor to a substrate such as an SOI (Silicon On Insulator) substrate (so-called silicon photonics) on which an optical waveguide is formed (for example, as described in Non-Patent Document 1). ).

R. Kou et al. “Inter-layer light transition in hybrid III-V/Si wavguides integrated by μ-transfer printing” Optic Express 28 (13), 19772-19 782, June 2020

In order to increase the coupling efficiency between the optical waveguide provided on the substrate and the optical waveguide of the III-V group semiconductor element, the tip of the optical waveguide of the III-V group semiconductor may be tapered. However, it has been difficult to reduce the width of the tapered tip to 400 nm or less by dry etching, for example. Therefore, it is an object of the present invention to provide a semiconductor optical device that is easy to manufacture and can improve coupling efficiency, and a method for manufacturing the same.

A semiconductor optical device according to the present disclosure includes a substrate having a first optical waveguide made of silicon, and a semiconductor device having a second optical waveguide bonded to the upper surface of the substrate and made of a III-V compound semiconductor. , wherein the first optical waveguide and the second optical waveguide form a directional coupler.

A method for manufacturing a semiconductor optical device according to the present disclosure includes the steps of: bonding a semiconductor element made of a III-V compound semiconductor to an upper surface of a substrate having a first optical waveguide made of silicon; forming a second optical waveguide, the first optical waveguide and the second optical waveguide forming a directional coupler.

According to the present disclosure, it is possible to provide a semiconductor optical device that is easy to manufacture and can increase coupling efficiency, and a method for manufacturing the same.

FIG. 1A is a plan view illustrating the semiconductor optical device according to the first embodiment. FIG. 1B is a partially enlarged plan view of the semiconductor optical device. FIG. 2A is a cross-sectional view taken along line AA in FIG. 1B. FIG. 2B is a cross-sectional view taken along line BB in FIG. 1B. FIG. 3A is a diagram illustrating the effective refractive index. FIG. 3B is a diagram illustrating the effective refractive index. FIG. 4 is a diagram illustrating transmittance. FIG. 5A is a diagram illustrating the dependence of transmittance on the amount of overlap. FIG. 5B is a diagram illustrating the wavelength dependence of transmittance. FIG. 6 is a cross-sectional view illustrating a semiconductor optical device according to a modification. FIG. 7A is a plan view illustrating a semiconductor optical device according to the second embodiment. FIG. 7B is a cross-sectional view taken along line CC in FIG. 7A. FIG. 8A is a plan view illustrating a semiconductor optical device according to the third embodiment. FIG. 8B is a partially enlarged plan view of the semiconductor optical device. FIG. 9A is a cross-sectional view taken along line DD in FIG. 8A. FIG. 9B is a cross-sectional view taken along line EE in FIG. 8A. FIG. 10A is a plan view illustrating a method for manufacturing a semiconductor optical device. FIG. 10B is a cross-sectional view taken along line DD in FIG. 10A. FIG. 10C is a cross-sectional view taken along line EE in FIG. 10A. FIG. 11A is a plan view illustrating a method for manufacturing a semiconductor optical device. FIG. 11B is a cross-sectional view taken along line DD in FIG. 11A. FIG. 11C is a cross-sectional view taken along line EE in FIG. 11A. FIG. 12A is a plan view illustrating a method for manufacturing a semiconductor optical device. FIG. 12B is a cross-sectional view taken along line DD in FIG. 12A. FIG. 13A is a plan view illustrating a semiconductor optical device according to a fourth embodiment. FIG. 13B is a cross-sectional view taken along line EE in FIG. 13A. FIG. 14 is a plan view illustrating a semiconductor optical device according to the fifth embodiment. FIG. 15A is a diagram illustrating a calculation result of coupling efficiency. FIG. 15B is a diagram illustrating a calculation result of coupling efficiency.

[Description of embodiments of the present disclosure]
First, the contents of the embodiments of the present disclosure will be listed and explained.

One form of the present disclosure provides (1) a semiconductor element having a substrate having a first optical waveguide made of silicon, and a second optical waveguide bonded to the upper surface of the substrate and made of a III-V compound semiconductor; The first optical waveguide and the second optical waveguide are semiconductor optical devices forming a directional coupler. Since a directional coupler is formed by bringing the first optical waveguide and the second optical waveguide close to each other, manufacturing is easy. By forming a directional coupler, coupling efficiency can be increased.
(2) In (1) above, the first optical waveguide may have a bent shape so as to approach the second optical waveguide. A directional coupler is formed by the first optical waveguide approaching the second optical waveguide. By forming a directional coupler, coupling efficiency can be increased.
(3) In (1) or (2) above, the first optical waveguide has a first part and a second part, and the distance between the second part and the second optical waveguide is The distance between the first portion and the second optical waveguide may be smaller than the distance between the first portion and the second optical waveguide, and the second portion and the second optical waveguide may form the directional coupler. By forming a directional coupler, coupling efficiency can be increased.
(4) In any one of (1) to (3) above, the second optical waveguide is located above one end of the first optical waveguide in the width direction, and the second optical waveguide is located above one end of the first optical waveguide in the width direction. It is not necessary to stretch to one end. When the first optical waveguide and the second optical waveguide form a directional coupler, coupling efficiency can be increased.
(5) In any one of (1) to (4) above, in the width direction of the first optical waveguide and the second optical waveguide, the center of the second portion of the first optical waveguide is located in the second optical waveguide. It may be spaced apart from the center of the wave path. When the first optical waveguide and the second optical waveguide form a directional coupler, coupling efficiency can be increased.
(6) In (5) above, at least a portion of the second portion of the first optical waveguide may not overlap the second optical waveguide in the direction in which the substrate and the semiconductor element are bonded. good. When the first optical waveguide and the second optical waveguide form a directional coupler, coupling efficiency can be increased.
(7) In any one of (3) to (6) above, the optical waveguide may include a phase adjustment section provided in the first portion of the first optical waveguide. The phase of light can be adjusted.
(8) In any one of (1) to (7) above, the first optical waveguide and the second optical waveguide form a plurality of the directional couplers, and the plurality of directional couplers include: The first optical waveguide and the second optical waveguide may be lined up along the extending direction. Multiple directional couplers can increase coupling efficiency.
(9) In any one of (1) to (8) above, the semiconductor element has a first semiconductor layer and a mesa, the first semiconductor layer is bonded to the upper surface of the substrate, and the mesa is The second optical waveguide may protrude from the first semiconductor layer to a side opposite to the substrate. When the first optical waveguide and the second optical waveguide form a directional coupler, coupling efficiency can be increased.
(10) In (9) above, the mesa includes a second semiconductor layer, a third semiconductor layer, and a fourth semiconductor layer, and the second semiconductor layer, the third semiconductor layer, and the fourth semiconductor layer are The third semiconductor layer may be stacked in this order on the first semiconductor layer and have a multiple quantum well structure. The third semiconductor layer becomes the core of the second optical waveguide, and light can be confined in the third semiconductor layer.
(11) In any one of (1) to (10) above, the substrate has a first layer, a second layer, and a third layer stacked in order, and the first layer and the third layer may be formed of silicon, the second layer may be formed of silicon oxide, and the semiconductor element may be bonded to the third layer. The first optical waveguide provided in the third layer and the second optical waveguide provided in the semiconductor element form a directional coupler, thereby increasing coupling efficiency.
(12) In any one of (1) to (11) above, the substrate has two first optical waveguides, and the two first optical waveguides and the second optical waveguide are connected to the directional coupler. may be formed. The coupling length between the first optical waveguide and the second optical waveguide can be shortened.
(13) In any one of (1) to (12) above, the semiconductor element may have an optical gain, and the semiconductor element may function as a laser element. Light generated by the semiconductor element can propagate through the second optical waveguide and be transitioned between the second optical waveguide and the first optical waveguide in the directional coupler.
(14) In any one of (1) to (13) above, the first optical waveguide has a tapered portion, and the tapered portion becomes thinner as it approaches the tip of the first optical waveguide. The tapered portion of the waveguide and the second optical waveguide may form the directional coupler. The coupling efficiency between the first optical waveguide and the second optical waveguide is increased. Improved tolerance to dimensional errors.
(15) In the above (14), the tapered portion of the first optical waveguide may have an asymmetrical shape with respect to the direction in which the first optical waveguide extends. The coupling efficiency between the first optical waveguide and the second optical waveguide is increased. Improved tolerance to dimensional errors.
(16) In (15) above, the first end of the first optical waveguide is parallel to the extending direction of the first optical waveguide, and the second end of the first optical waveguide is parallel to the second optical waveguide. The tapered portion may form the asymmetrical shape. The coupling efficiency between the first optical waveguide and the second optical waveguide is increased. Improved tolerance to dimensional errors.
(17) In the above (14), the tapered portion of the first optical waveguide may have a shape that is symmetrical with respect to the direction in which the first optical waveguide extends. The coupling efficiency between the first optical waveguide and the second optical waveguide is increased. Improved tolerance to dimensional errors.
(18) A step of bonding a semiconductor element formed of a III-V group compound semiconductor to the upper surface of a substrate having a first optical waveguide formed of silicon, and a step of forming a second optical waveguide on the semiconductor element. , wherein the first optical waveguide and the second optical waveguide form a directional coupler. Since a directional coupler is formed by bringing the first optical waveguide and the second optical waveguide close to each other, manufacturing is easy. By forming a directional coupler, coupling efficiency can be increased.

[Details of embodiments of the present disclosure]
Specific examples of semiconductor optical devices and methods of manufacturing the same according to embodiments of the present disclosure will be described below with reference to the drawings. Note that the present disclosure is not limited to these examples, but is indicated by the scope of the claims, and is intended to include all changes within the meaning and range equivalent to the scope of the claims.

<First embodiment>
FIG. 1A is a plan view illustrating a semiconductor optical device 100 according to the first embodiment. FIG. 1B is a partially enlarged plan view of the semiconductor optical device 100. FIG. 2A is a cross-sectional view taken along line AA in FIG. 1B. FIG. 2B is a cross-sectional view taken along line BB in FIG. 1B.

As shown in FIG. 1A, two sides of the semiconductor optical device 100 extend parallel to the X-axis direction. The other two sides extend parallel to the Y-axis direction. The Z-axis direction is the normal direction of the XY plane and the layer stacking direction. The X-axis direction, Y-axis direction, and Z-axis direction are orthogonal to each other. The semiconductor optical device 100 may be a single rectangular chip, or may be a rectangular region of a part of a large chip (such as a wavelength tunable laser) in which a plurality of devices are integrated.

The semiconductor optical device 100 is a hybrid optical device having a substrate 10 and a semiconductor element 40. The substrate 10 has an upper surface parallel to the XY plane. The semiconductor element 40 is bonded to the upper surface of the substrate 10. In the plan view, the semiconductor element 40 is seen through and the upper surface of the substrate 10 is illustrated.

The substrate 10 has an optical waveguide 20 (first optical waveguide). The semiconductor element 40 has an optical waveguide 41 (second optical waveguide). The optical waveguide 20 and the optical waveguide 41 extend from one end of the semiconductor optical device 100 in the X-axis direction to the other end.

As shown in FIG. 1A, the optical waveguide 41 extends, for example, in a straight line and is parallel to the X-axis direction. The optical waveguide 20 has a wavy shape in the XY plane, and approaches the optical waveguide 41 at the antinode of the wave. The straight portion of the optical waveguide 20 is parallel to the X-axis direction. The optical waveguide 20 bends and approaches the optical waveguide 41. The optical waveguide 20 has a portion 30 (first portion) far from the optical waveguide 41 and a portion 32 (second portion, a portion corresponding to the antinode of the wave) close to the optical waveguide 41.

As shown in FIGS. 1B and 2A, a portion of the portion 32 of the optical waveguide 20 extends below and overlaps the optical waveguide 41. The distance between the optical waveguide 20 and the optical waveguide 41 becomes closer in the portion 32, and the optical waveguide 20 and the optical waveguide 41 form a directional coupler (DC) 21. The number of directional couplers 21 may be plural or one. In the first embodiment, the number of directional couplers 21 is plural, for example, three or more, five or more. The plurality of directional couplers 21 are arranged in the X-axis direction.

A phase adjustment section 23 is provided in a portion of the optical waveguide 20 that extends in the X-axis direction. The number of phase adjustment sections 23 may be one or more. In the phase adjustment section 23, a heater of a predetermined length is provided along the optical waveguide 20. The heater changes the temperature of the phase adjustment section 23. The refractive index of the phase adjustment section 23 changes due to temperature change, and the phase of light passing through the phase adjustment section 23 changes. The heater is made of metal such as tantalum (Ta), for example. The length of the phase adjustment section 23 in the X-axis direction is, for example, 100 μm.

As shown in FIGS. 1B and 2A, the substrate 10 has an optical waveguide 20, a groove 22, and a terrace 24. The grooves 22 are located on both sides of the optical waveguide 20 and extend along the optical waveguide 20 . As shown in FIG. 1B, the optical waveguide 20 is linear and has a wave-like planar shape. The support body 26 is linear and parallel to the X-axis direction. The support 26 is formed from the Si layer 16 and is spaced apart from the optical waveguide 20 . The terrace 24 is a planarly expanding portion of the Si layer 16.

As shown in FIGS. 2A and 2B, the substrate 10 is an SOI substrate having a substrate 12 (first layer), a box layer 14 (second layer), and a silicon (Si) layer 16 (third layer). The substrate 12 is made of Si and has a thickness of 500 μm, for example. The box layer 14 is formed of silicon oxide (SiO ₂ ) with a thickness of 3 μm, for example, and is laminated on the upper surface of the substrate 12 . The Si layer 16 is made of Si and has a thickness of 220 nm, for example, and is laminated on the upper surface of the box layer 14.

The groove 22 is a portion of the Si layer 16 that is depressed from the top surface in the Z-axis direction, and is formed, for example, by etching the Si layer 16. In the Z-axis direction, the groove 22 may extend halfway through the Si layer 16, or may extend through the Si layer 16 to the box layer 14. That is, the depth of the groove 22 is, for example, 220 nm or less (thickness of the Si layer 16). The refractive index of the substrate 12 and the Si layer 16 is, for example, 3.48 at a wavelength of 1.55 μm. The refractive index of the box layer 14 is lower than the refractive index of the substrate 12 and the Si layer 16, and is, for example, 1.44 at a wavelength of 1.55 μmm.

The optical waveguide 20, the terrace 24, and the support body 26 are formed in the Si layer 16, and are portions that protrude from the bottom surface of the groove 22 in the Z-axis direction. In the etching process for forming the groove 22, the portions that are not etched become the optical waveguide 20, the terrace 24, and the support 26. The top surfaces of the optical waveguide 20, the terrace 24, and the support 26 are parallel to the XY plane and form one plane, that is, the top surface of the substrate 10. The thickness T1 of the optical waveguide 20 shown in FIG. 2A is equal to the depth of the groove 22, and is, for example, 220 nm. The width W1 of the optical waveguide 20 is, for example, 880 nm.

The semiconductor element 40 has a bonding layer 42 (first semiconductor layer), cladding layers 44 and 48, optical confinement layers 45 and 47, an active layer 46 (third semiconductor layer), and a contact layer 49. The bonding layer 42 covers the upper surface of the Si layer 16 of the substrate 10 and is bonded to the upper surface. The bonding layer 42 may be in contact with the upper surface of the Si layer 16, or another layer may be provided between the bonding layer 42 and the Si layer 16.

The semiconductor element 40 has a mesa 43. As described in the third embodiment, the mesa 43 is formed on the semiconductor element 40 by bonding the semiconductor element 40 to the substrate 10 and etching it. Mesa 43 protrudes from the upper surface of Si layer 16 in the Z-axis direction. Mesa 43 includes cladding layers 44 and 48, optical confinement layers 45 and 47, active layer 46, and contact layer 49. A cladding layer 44 (second semiconductor layer), an optical confinement layer 45, an active layer 46, an optical confinement layer 47, a cladding layer 48 (fourth semiconductor layer), and a contact layer 49 are arranged on the bonding layer 42 in the Z-axis direction. They are stacked in this order. The mesa 43 extends in parallel to the X-axis direction from one end of the substrate 10 to the other end in the X-axis direction, and functions as the optical waveguide 41 . The active layer 46 becomes the core layer of the optical waveguide 41.

As shown in FIGS. 2A and 2B, the bonding layer 42 of the semiconductor element 40 is bonded to and supported by the terraces 24 of the substrate 10. In the region shown in FIG. 1B, the central part of the mesa 43 of the semiconductor element 40 in the X-axis direction (the part close to the part 32 of the optical waveguide 20) is located so as to overlap the optical waveguide 20, and The parts on both sides in the direction are located so as to overlap on the support body 26. Mesa 43 is supported by support 26 .

The insulating film 25 covers the side surfaces of the mesa 43 and the top surface of the Si layer 16. The insulating film 25 is made of an insulator such as silicon nitride (SiN), silicon oxide (SiO ₂ ), silicon oxynitride (SiON), or the like. The thickness of the insulating film 25 is, for example, 100 nm or more and 600 nm or less. The insulating film 25 has an opening on the upper surface of the mesa 43. An electrode 27 is provided on the upper surface of the contact layer 49. The electrode 27 is made of metal such as gold (Au), for example.

The bonding layer 42 is made of indium phosphide (InP) with a thickness of 182 nm, for example. The cladding layer 44 is made of n-type indium phosphide (n-InP) with a thickness of 180 nm, for example. For example, Si is used as the n-type dopant. The cladding layer 48 is made of p-type indium phosphide (p-InP) with a thickness of 1700 nm, for example. The refractive index of the bonding layer 42, cladding layers 44, and 48 is lower than the refractive index of the active layer 46, for example, 3.17 at a wavelength of 1.55 μm. The contact layer 49 is formed of, for example, p+ type gallium indium arsenide ((p+)-GaInAs). For example, zinc (Zn) is used as the p-type dopant.

The active layer 46 has a multi-quantum well structure (MQW) and includes a plurality of well layers and a plurality of barrier layers. The plurality of well layers and the plurality of barrier layers are alternately stacked. One well layer is formed of, for example, GaInAsP with a thickness of 6 nm. One barrier layer is made of GaInAsP with a thickness of 10 nm, for example. The thickness of the active layer 46 is, for example, 90 nm. The refractive index of the active layer 46 is, for example, 3.44 at a wavelength of 1.55 μm.

The optical confinement layers 45 and 47 are made of, for example, undoped gallium indium arsenide phosphide (i-GaInAsP). The thickness of the optical confinement layer 45 is, for example, 80 nm. The thickness of the optical confinement layer 47 is, for example, 100 nm. The bandgap wavelength of the optical confinement layers 45 and 47 is, for example, 1.2 μm, which is shorter than the wavelength of light input and output by the semiconductor optical device 100. The refractive index of the optical confinement layers 45 and 47 is, for example, 3.34 at a wavelength of 1.55 μm. Each of the semiconductor layers of the semiconductor element 40 is formed of a III-V group compound semiconductor, and may be formed of a semiconductor other than the above.

The width W2 of the mesa 43 is, for example, 500 nm, 550 nm, 600 nm, or several hundred nm. The distance between mesa 43 and portion 30 of optical waveguide 20 is greater than the distance between mesa 43 and portion 32 of optical waveguide 20 . The distance D1 in the Y-axis direction between the phase adjustment section 23 and the portion 32 of the optical waveguide 20 shown in FIG. 1B is, for example, 0.9 μm. Portion 32 of optical waveguide 20 extends below mesa 43 . The width direction of the optical waveguide is perpendicular to the extending direction of the optical waveguide. In FIG. 1B, the width direction is parallel to the Y-axis direction. In the portion 32 of the optical waveguide 20, the mesa 43 is located above one end in the width direction of the optical waveguide 20 and does not extend to the other end. That is, one end of the optical waveguide 20 is located below the mesa 43 and the other end is located outside the optical waveguide 20. Mesa 43 and part of portion 32 of optical waveguide 20 overlap when viewed from the Z-axis direction. The width W3 of the overlapping portion (overlap amount, see FIG. 2A) is, for example, 300 nm or more and 400 nm or less, and is several hundred nm.

A line C1 in FIG. 2A represents the center of the optical waveguide 20 in the Y-axis direction. Line C2 represents the center of the optical waveguide 41 in the Y-axis direction. The center of the optical waveguide 20 and the center of the optical waveguide 41 do not overlap and are spaced apart from each other.

The semiconductor element 40 and the substrate 10 are evanescently optically coupled. Portion 32 of optical waveguide 20 and optical waveguide 41 of semiconductor element 40 form directional coupler 21 and are optically coupled. The coupling length L in one directional coupler 21 shown in FIG. 1B, that is, the length of the portion where the optical waveguide 20 and the optical waveguide 41 overlap and are parallel, is, for example, 50 μm. The distance D2 from the coupling portion to the phase adjustment section 23 is, for example, 20 μm.

For example, one end of the optical waveguide 20 is assumed to be an input port IN, and one end of the optical waveguide 41 is assumed to be an output port OUT. Light is input to the input port IN of the optical waveguide 20. The light propagates through the optical waveguide 20 and transitions from the optical waveguide 20 to the optical waveguide 41 at the directional coupler 21 . The light that has transitioned to the optical waveguide 41 is emitted from the end of the optical waveguide 41. The heater of the phase adjustment section 23 provided in the optical waveguide 20 changes the refractive index of the optical waveguide 20. By changing the refractive index, it is possible to adjust the phase of light and achieve phase matching. The semiconductor optical device 100 functions as a Mach-Zehnder interferometer. By applying a voltage to the electrode 27 provided on the optical waveguide 41, the optical waveguide 41 comes to have an optical gain.

3A and 3B are diagrams illustrating the effective refractive index. The horizontal axis represents the width of the optical waveguide. The vertical axis represents the effective refractive index of the optical waveguide at a wavelength of 1.55 μm.

FIG. 3A shows the effective refractive index of the optical waveguide 20. FIG. 3B shows the effective refractive index of the optical waveguide 41. As shown in FIGS. 3A and 3B, as the width of the optical waveguide increases, the effective refractive index increases. It is preferable to adjust the widths of the optical waveguides 20 and 41 so that the effective refractive index of the optical waveguide 20 and the refractive index of the optical waveguide 41 are approximately the same. By making the refractive indexes of the two substantially equal, the coupling efficiency between the optical waveguide 20 and the optical waveguide 41 in the directional coupler 21 can be increased.

FIG. 4 is a diagram illustrating the transmittance, and represents the calculation result of the transmittance of light from the optical waveguide 20 to the optical waveguide 41. The horizontal axis represents the coupling efficiency X in one directional coupler 21. The vertical axis represents the transmittance T (maximum transmittance) of light from the optical waveguide 20 to the optical waveguide 41 when light with a wavelength of 1.55 μm is guided. The number of directional couplers 21 is assumed to be three. It is assumed that the coupling efficiencies X of the three directional couplers 21 are equal to each other.

The coupling efficiency X is designed by adjusting the effective refractive index, coupling length L, and overlap amount W3 of the optical waveguide 20 and the optical waveguide 41. Transmittance T is expressed as a function of coupling efficiency X by the following equation.
T=16X ³ -24X ² +9X
The range of coupling efficiency X that can achieve the maximum transmittance T=100% is expressed by the following equation. However, in the following formula, X is treated as 100%.
If n is an odd number, 1≧X≧sin ² (π/(2n))
If n is an even number, 1-sin ² (π/(2n))≧X≧sin ² (π/(2n))
n is the number of directional couplers 21, and in the example of FIG. 4, n=3. When specific numerical values are input into the above two formulas, X (%) becomes as follows.
When n=1, X=100%.
When n=2, X=50%.
When n=3, 100%≧X≧25%
When n=4, 85.35%≧X≧14.64%
When n=5, 100%≧X≧9.54%
When n=6, 93.30%≧X≧6.69%
That is, when n is set to 3 or more, the range of the coupling rate X in which the transmittance T can be 100% is wider when n is an odd number than when n is an even number. As shown in FIG. 4, when n=3, when the coupling efficiency X becomes 25% or more, the transmittance T becomes 100%. That is, light can be input into the optical waveguide 20 and all of the light can be output from the optical waveguide 41.

FIG. 5A is a diagram illustrating the dependence of transmittance on the amount of overlap. The horizontal axis represents the overlap amount W3. The vertical axis represents the maximum transmittance from the optical waveguide 20 to the optical waveguide 41. The transmittance was calculated using the wavelength of light as 1550 nm and the overlap amount W3 as a variable.

The broken line in FIG. 5A is an example in which the width W2 of the optical waveguide 41 is 550 nm. The solid line is an example in which the width W2 is 600 nm. The dotted line is an example in which the width W2 is 650 nm. In all three examples, the optical loss can be suppressed to less than 1 dB when the overlap amount W3 is in the range of 290 nm to 360 nm.

FIG. 5B is a diagram illustrating the wavelength dependence of transmittance. The horizontal axis represents the wavelength of light. The vertical axis represents transmittance. In the example of the solid line, the width W2 of the optical waveguide 41 is 600 nm, and the amount of overlap W3 is 330 nm. In the example of the broken line, the width W2 is 650 nm and the overlap amount W3 is 330 nm. In the example of the dotted line, the width W2 is 550 nm and the overlap amount W3 is 330 nm. In the example of the dashed-dotted line, the width W2 is 600 nm and the overlap amount W3 is 300 nm. In the example of the two-dot chain line, the width W2 is 600 nm and the overlap amount W3 is 360 nm. Optical loss is suppressed to less than 1.8 dB over the entire C band (1530 nm to 1565 nm).

Due to misalignment of the resist pattern during the manufacturing process, variations occur in the width W1 of the optical waveguide 20, the width W2 of the optical waveguide 41, and the amount of overlap W3. As shown in FIG. 5A, for example, if the width W2 of the mesa 43 is within the range of 600±50 nm and the overlap amount W3 is within the range of 290 nm to 360 nm, the optical loss is suppressed to less than 1 dB. As shown in FIG. 5B, when the width W2 of the optical waveguide 41 is within the range of 600±50 nm and the overlap amount W3 is between 300 nm and 330 nm, the optical loss is suppressed to less than 1.8 dB in the entire C band. be done.

According to the first embodiment, the substrate 10 has an optical waveguide 20 made of Si. The semiconductor element 40 is bonded to the upper surface of the substrate 10 and has an optical waveguide 41 made of a III-V compound semiconductor. When the optical waveguide 20 and the optical waveguide 41 form the directional coupler 21, high coupling efficiency can be obtained. Light transfers between the optical waveguide 20 and the optical waveguide 41 in the directional coupler 21 . As shown in FIGS. 5A and 5B, it is possible to suppress light loss and propagate light to two optical waveguides.

The coupling efficiency can be increased by providing the optical waveguide 41 with a thin taper having a width of, for example, several hundred nm or less. However, it is difficult to form a taper by etching with a high aspect ratio (high thickness to width ratio). According to the first embodiment, it is not necessary to form a taper in the optical waveguide 41. By bringing the optical waveguide 20 and the optical waveguide 41 close to each other, a directional coupler 21 is formed, and the coupling efficiency can be increased.

As shown in FIGS. 1A and 1B, the optical waveguide 20 has a wavy planar shape and is bent to approach the optical waveguide 41. When the optical waveguide 20 approaches the optical waveguide 41, a directional coupler 21 is formed.

Specifically, the optical waveguide 20 has a portion 30 and a portion 32. The distance between portion 32 and optical waveguide 41 is smaller than the distance between portion 30 and optical waveguide 41. Since the portion 32 is close to the optical waveguide 41, a directional coupler 21 is formed. For example, a wavy optical waveguide 20 is formed in the Si layer 16 by etching. A mesa 43 is formed in the semiconductor element 40 at a position close to the optical waveguide 20 by etching. The semiconductor optical device 100 having the directional coupler 21 can be manufactured through a simple process.

As shown in FIG. 2A, the center of the portion 32 of the optical waveguide 20 (line C1) is spaced apart from the center of the optical waveguide 41 (line C2). That is, the optical waveguide 20 is spaced apart from the optical waveguide 41 in the width direction (Y-axis direction) and also spaced apart from the optical waveguide 41 in the thickness direction (Z-axis direction). In other words, the optical waveguide 20 and the optical waveguide 41 are arranged obliquely to each other in a plane parallel to the YZ plane. The diagonally arranged optical waveguide 20 and optical waveguide 41 form a directional coupler 21 . It is possible to increase the coupling efficiency and transition light between two optical waveguides.

For example, as shown in FIG. 2A, in the width direction (Y-axis direction), the optical waveguide 41 is located above one end of the optical waveguide 20 and does not extend to the other end. That is, in the Z-axis direction, a part of the optical waveguide 20 overlaps with the optical waveguide 41, and another part does not overlap with the optical waveguide 41. With respect to the width W1 of the optical waveguide 20 (eg, 880 nm), the amount of overlap W3 between the optical waveguide 20 and the optical waveguide 41 is, for example, 300 nm±30 nm. The amount of overlap W3 may be less than or equal to half of the width W1, or may be more than half of the width W1.

The bonding layer 42 of the semiconductor element 40 is bonded to the upper surface of the substrate 10. The mesa 43 protrudes from the bonding layer 42 in the Z-axis direction and has an optical waveguide 41 . The optical waveguide 41 is located above the substrate 10. The diagonally arranged optical waveguide 20 and optical waveguide 41 form a directional coupler 21 . It is possible to increase the coupling efficiency and transition light between two optical waveguides.

In the mesa 43, a cladding layer 44, an optical confinement layer 45, an active layer 46, an optical confinement layer 47, a cladding layer 48, and a contact layer 49 are laminated in this order. The active layer 46 has a multiple quantum well structure and functions as a core layer of the optical waveguide 41. Active layer 46 is sandwiched between cladding layers 44 and 48. Light can be confined in the active layer 46 and loss can be suppressed.

Substrate 10 is an SOI substrate and includes a substrate 12, a box layer 14, and a Si layer 16. An optical waveguide 20 is provided in the Si layer 16. By forming the directional coupler 21 by the Si optical waveguide 20 and the III-V group compound semiconductor optical waveguide 41, coupling efficiency can be increased.

As shown in FIGS. 1B and 2B, the substrate 10 has a support 26. As shown in FIGS. The support body 26 is spaced apart from the optical waveguide 20, is located below the mesa 43, extends in the same direction as the mesa 43 (X-axis direction), and supports the mesa 43. The mechanical strength of the semiconductor optical device 100 is improved. The width of the support body 26 is preferably smaller than the width W1 of the optical waveguide 20, for example, 100 nm or less. Since the support body 26 is thin, the effective refractive index of the support body 26 is lower than the effective refractive index of the optical waveguide 20 and the optical waveguide 41. Light becomes difficult to spread to the support body 26. The support body 26 may not be provided on the substrate 10, and the groove 22 may be provided in the portion of the substrate 10 below the mesa 43. Light loss can be suppressed.

For example, by providing the terrace 24 under the mesa 43, mechanical strength can be increased. However, there is a possibility that the discontinuity in the refractive index between the directional coupler 21 and the terrace 24 will increase, leading to an increase in light loss. By arranging the mesa 43 on the support body 26, mechanical strength can be increased and discontinuity in the refractive index can be reduced.

As shown in FIG. 1A, the semiconductor optical device 100 has a plurality of directional couplers 21. Since the plurality of directional couplers 21 are arranged along the X-axis direction, the coupling efficiency is increased. The number of directional couplers 21 is, for example, three or more, five or more, and preferably an odd number.

A phase adjustment section 23 is provided in a portion 30 of the optical waveguide 20. When the heater provided in the phase adjustment section 23 generates heat, the temperature of the phase adjustment section 23 changes, and the refractive index changes. The phase of light propagating through the optical waveguide 20 can be adjusted. Light can be generated in the active layer 46 by passing a current through the mesa 43 using an electrode provided on the mesa 43. The generated light is optically coupled from the optical waveguide 41 to the optical waveguide 20 in the directional coupler 21 . The phase of the light transferred to the optical waveguide 20 is adjusted by the phase adjustment unit 23, so that the light can be optically coupled to the optical waveguide 41 with high coupling efficiency in the next directional coupler 21. Furthermore, the refractive index of the optical waveguide 41 can also be changed using the electrode 27 provided on the mesa 43. By setting the effective refractive index of the optical waveguide 20 and the effective refractive index of the optical waveguide 41 to be approximately the same, the coupling efficiency can be increased.

(Modified example)
FIG. 6 is a cross-sectional view illustrating a semiconductor optical device 100A according to a modification, and shows a cross-section corresponding to FIG. 2A. In the example of FIG. 6, the entire optical waveguide 41 does not overlap the optical waveguide 20. In other words, the amount of overlap is zero. The distance between the optical waveguide 20 and the optical waveguide 41 in the Y-axis direction is, for example, about 500 nm. A directional coupler 21 is formed by bringing the optical waveguide 20 and the optical waveguide 41 closer together.

As shown in FIGS. 2A and 6, at least a portion of the optical waveguide 20 does not overlap the optical waveguide 41. The optical waveguide 20 and the optical waveguide 41 are brought close to each other, and a directional coupler 21 is formed.

<Second embodiment>
FIG. 7A is a plan view illustrating a semiconductor optical device 200 according to the second embodiment. FIG. 7B is a cross-sectional view taken along line CC in FIG. 7A. Description of the same configuration as in the first embodiment will be omitted.

As shown in FIGS. 7A and 7B, the substrate 10 has two optical waveguides 20. The two optical waveguides 20 are symmetrical with respect to an axis extending along the widthwise center of the optical waveguide 41 (an axis parallel to the X-axis). Each of the two optical waveguides 20 and the optical waveguide 41 form a directional coupler 21 .

According to the second embodiment, since the two optical waveguides 20 and the optical waveguide 41 form the directional coupler 21, the coupling efficiency can be increased. By using two optical waveguides 20, the coupling length between the optical waveguides (length L in FIG. 1B) can be increased by 1/(√2) compared to when there is only one optical waveguide 20. can.

In the example of FIG. 7B, the two optical waveguides 20 do not overlap the optical waveguide 41. Parts of the two optical waveguides 20 may overlap the optical waveguide 41.

<Third embodiment>
FIG. 8A is a plan view illustrating a semiconductor optical device 300 according to the third embodiment. FIG. 8B is a partially enlarged plan view of the semiconductor optical device 300. Description of the same configuration as in the first embodiment will be omitted.

As shown in FIG. 8A, the substrate 10 has an optical waveguide 20, two ring resonators 50, and two loop mirrors 52. The optical waveguide 20 extends from one end of the substrate 10 to the other end in the X-axis direction. Two ring resonators 50 and two loop mirrors 52 are provided in the middle of the optical waveguide 20. Like the optical waveguide 20, the ring resonator 50 and the loop mirror 52 are provided on the Si layer 16 of the substrate 10 and have a Si core.

A semiconductor element 40 is bonded to the center of the upper surface of the substrate 10 . The optical waveguide 41 of the semiconductor element 40 and the optical waveguide 20 of the substrate 10 form a plurality of directional couplers 21. The optical waveguide 20 is separated from the optical waveguide 41 in a portion other than the directional coupler 21 . The end of the semiconductor element 40 is spaced apart from the end of the substrate 10.

As shown in FIG. 8B, the semiconductor element 40 has tapered portions 54 at both ends in the X-axis direction. The tapered portion 54 is spaced apart from the optical waveguide 41 and located above the optical waveguide 20. The tapered portion 54 has a tapered shape that protrudes outward from the semiconductor element 40 and becomes thinner as the distance from the semiconductor element 40 increases.

As shown in FIG. 8A, the ring resonator 50 and the loop mirror 52 are spaced apart from the semiconductor element 40 in the X-axis direction. One ring resonator 50 and one loop mirror 52 are arranged in this order from one end of the semiconductor element 40 to one end of the substrate 10. One ring resonator 50 and one loop mirror 52 are arranged in this order from the other end of the semiconductor element 40 to the other end of the substrate 10. Ring resonator 50 and loop mirror 52 are optically coupled to optical waveguide 20.

FIG. 9A is a cross-sectional view taken along the line DD in FIG. 8A, and illustrates a cross-section including the directional coupler 21. FIG. As shown in FIGS. 8A and 9A, the optical waveguide 20 and the optical waveguide 41 overlap. Semiconductor element 40 has electrodes 60 and 62. Electrode 60 and electrode 62 are spaced apart from each other.

As shown in FIG. 9A, the insulating film 25 has an opening above the bonding layer 42 and an opening above the mesa 43. The electrode 60 is an n-type electrode, is provided in the opening of the insulating film 25 , contacts the bonding layer 42 , and is electrically connected to the bonding layer 42 . The electrode 62 is a p-type electrode and has a pad 62a and a connection portion 62b. The pad 62a and the connecting portion 62b are formed of one metal layer and are electrically connected. The pad 62a is spaced apart from the optical waveguide 41 and provided on the insulating film 25. The connecting portion 62b is provided on the mesa 43, contacts the contact layer 49 through the opening of the insulating film 25, and is electrically connected to the contact layer 49.

The electrode 60 is formed of, for example, an alloy of gold, germanium, and Ni (AuGeNi). The electrode 62 is formed of, for example, a laminate of titanium, platinum, and gold (Ti/Pt/Au). The electrodes 60 and 62 further include a gold (Au) wiring layer.

FIG. 9B is a cross-sectional view taken along the line EE in FIG. 8A, and shows a cross-section including the tip of the mesa 43 and not including the directional coupler 21. FIG. As shown in FIG. 9B, the optical waveguide 20 and the optical waveguide 41 are separated from each other. The tip portion of the mesa 43 is not provided with an electrode 62 and is covered with an insulating film 25 . Tapered portion 54 is formed from bonding layer 42 . The tapered portion 54 does not include the optical confinement layer 45, the active layer 46, the optical confinement layer 47, the cladding layer 48, and the contact layer 49. Therefore, the etching aspect ratio when forming the tapered portion 54 can be kept low.

The semiconductor element 40 has an optical gain and is evanescently optically coupled to the substrate 10. By applying a voltage to the semiconductor element 40 using the electrode 60 and the electrode 62, a current flows through the mesa 43. By injecting carriers into the active layer 46, the active layer 46 generates light. In the directional coupler 21 , light transits from the optical waveguide 41 of the semiconductor element 40 to the optical waveguide 20 of the substrate 10 .

The light propagating through the optical waveguide 20 is reflected by the loop mirror 52. The light is repeatedly reflected by the two loop mirrors 52 and oscillates as a laser. The laser light is emitted to the outside of the semiconductor optical device 300.

(Production method)
10A, FIG. 11A, and FIG. 12A are plan views illustrating a method for manufacturing the semiconductor optical device 300. 10B, 11B and 12B are cross-sectional views along line DD of the corresponding plan view. FIGS. 10C and 11C are cross-sectional views along line EE of the corresponding plan view.

Before the steps shown in FIGS. 10A to 10C, dry etching is performed on the Si layer 16 of the substrate 10 to form grooves 22. An optical waveguide 20, a terrace 24, a ring resonator 50, and a loop mirror 52 are formed in the unetched portion. For example, a bonding layer 42, a cladding layer 44, an optical confinement layer 45, an active layer 46, an optical confinement layer 47, A cladding layer 48 and a contact layer 49 are epitaxially grown. Semiconductor elements 40 are formed by dicing the wafer. No mesa 43 or electrodes are formed on the semiconductor element 40.

As shown in FIGS. 10A to 10C, the semiconductor element 40 is bonded to the substrate 10. Specifically, the upper surface of the Si layer 16 of the substrate 10 and the surface of the bonding layer 42 of the semiconductor element 40 are activated by performing nitrogen (N ₂ ) plasma treatment on the upper surface of the substrate 10 . The activated surface is ultrasonically cleaned in water. The semiconductor element 40 is temporarily bonded to the upper surface of the substrate 10 by bringing the activated surfaces into contact with each other. After temporary bonding, annealing is performed at, for example, 300° C. for 2 hours to remove moisture and strengthen the bonding strength (O ₂ bonding). Two ring resonators 50 and two loop mirrors 52 are located outside the semiconductor element 40. A wavy portion of the optical waveguide 20 is covered with a semiconductor element 40 .

An insulating film serving as an etching mask is formed on the substrate 10 and the semiconductor element 40. A resist pattern is formed on the insulating film by photolithography or the like, and the pattern is transferred to the insulating film by etching (the etching mask and resist pattern are not shown). Etching is performed using an etching mask. For example, RIE using a mixed gas of methane and hydrogen (CH ₄ /H ₂ ) or a chlorine-based gas and wet etching are performed to form the mesa 43 in the semiconductor element 40 . The bonding layer 42 is exposed in a portion other than the mesa 43. The insulating film used as a mask is removed by wet etching using buffered hydrogen fluoride (BHF). Further, the bonding layer 42 is etched to form a tapered portion 54. Since the tapered portion 54 does not include the optical confinement layer 45, the active layer 46, the optical confinement layer 47, the cladding layer 48, and the contact layer 49, the etching aspect ratio when the tapered portion 54 is formed is low. Therefore, the shape of the narrow tip of the tapered portion 54 can be formed with high precision.

As shown in FIGS. 11A to 11C, the insulating film 25 is formed by, for example, chemical vapor deposition (CVD). Insulating film 25 covers substrate 10 and semiconductor element 40 . For example, an opening is formed in the insulating film 25 by wet etching. As shown in FIG. 11B, one opening 25a is provided at a position spaced apart from mesa 43, and one opening 25b is formed above mesa 43.

As shown in FIGS. 12A and 12B, electrodes 60 and 62 are formed, for example, by vacuum deposition and lift-off. For example, a wiring layer of Au may be formed by plating. A semiconductor optical device 300 is formed by dicing the substrate 10 in a wafer state.

According to the third embodiment, the semiconductor optical device 300 functions as a laser device. A semiconductor element 40 with optical gain generates light. Since the optical waveguide 41 of the semiconductor element 40 and the optical waveguide 20 of the substrate 10 form the directional coupler 21, light transitions between the two optical waveguides. The light propagates through the optical waveguide 20, is reflected by the two loop mirrors 52, and oscillates as a laser. The semiconductor optical device 300 can emit laser light outward from the edge of the substrate 10.

In the semiconductor optical device 300, the optical waveguide 20 and the optical waveguide 41 may be separated from each other as shown in FIG. Two optical waveguides 20 may be provided. Optical elements other than the ring resonator 50 and the loop mirror 52 may be provided on the substrate 10.

<Fourth embodiment>
FIG. 13A is a plan view illustrating a semiconductor optical device 400 according to the fourth embodiment. FIG. 13B is a cross-sectional view taken along line EE in FIG. 13A. Descriptions of configurations that are the same as any of the first to third embodiments will be omitted. As shown in FIG. 13A, the semiconductor optical device 400 has an optical waveguide 20 and an optical waveguide 41. The optical waveguide 20 and the optical waveguide 41 form one directional coupler 21.

Optical waveguide 20 and optical waveguide 41 extend in the X-axis direction. The optical waveguide 20 extends from one end of the substrate 10 in the X-axis direction to the middle of the substrate 10. One end of the optical waveguide 20 is located at the end of the substrate 10 and functions as an input port IN. The other end of the optical waveguide 20 has a tapered portion 70 . The portion of the optical waveguide 20 other than the tapered portion 70 is linear. A portion of the optical waveguide 20 near the input port IN is exposed from the bonding layer 42 of the semiconductor element 40. The tapered portion 70 and a portion of the optical waveguide 20 near the tapered portion 70 are covered with the bonding layer 42 .

The optical waveguide 41 extends from the middle of the substrate 10 to the end of the substrate 10 opposite to the input port IN. The end of the optical waveguide 41 is the output port OUT.

The optical waveguide 20 has a tapered portion 70 at the tip opposite to the input port IN. The tapered portion 70 of the optical waveguide 20 and the optical waveguide 41 form a directional coupler 21. The tapered portion 70 has a shape that is symmetrical with respect to the X-axis. One end 20a (first end) and another end 20b (second end) of the optical waveguide 20 in the Y-axis direction are inclined from the X-axis and approach the optical waveguide 41. End portion 20a and end portion 20b form a tapered portion 70. The tapered portion 70 becomes thicker as it gets farther from the tip of the optical waveguide 20, and thinner as it gets closer to the tip.

The coupling length L1 between the optical waveguide 20 and the optical waveguide 41 in the directional coupler 21 shown in FIG. 13A is, for example, 200 μm. The distance L2 from the end of the bonding layer 42 to the tapered portion 70 of the optical waveguide 20 is, for example, 15 μm. Among the two ends of the tapered portion 70, the width W4 at the end closer to the input port IN is, for example, 1200 nm. The width W5 of the optical waveguide 20 at the other end (tip) of the tapered portion 70 is smaller than the width W4, and is, for example, 400 nm. The thickness T1 of the optical waveguide 20 shown in FIG. 13B is, for example, 220 nm. The distance g between the center of the optical waveguide 20 (line C1) and the center of the optical waveguide 41 (line C2) is 0 nm or more, and may be several hundred nm. When the distance g is 0 nm, the center of the optical waveguide 20 and the center of the optical waveguide 41 overlap.

Light enters the optical waveguide 20 from the input port IN. The light propagates through the optical waveguide 20 and transitions from the optical waveguide 20 to the optical waveguide 41 at the directional coupler 21 . The light that has transitioned to the optical waveguide 41 is emitted from the output port OUT of the optical waveguide 41.

According to the fourth embodiment, the tapered portion 70 of the optical waveguide 20 and the optical waveguide 41 form the directional coupler 21. Therefore, high coupling efficiency can be obtained. The light transfers from the optical waveguide 20 to the optical waveguide 41 in the directional coupler 21 . It is possible to suppress light loss and emit light from the output port OUT.

The Si layer 16 may be etched to provide the optical waveguide 20 with a tapered portion 70 . It is not necessary to form a multi-stage taper in the semiconductor element 40. The process is simplified. Discrepancies may occur in dimensions such as the widths W4 and W5 of the optical waveguide 20, the width W2 of the optical waveguide 41, and the distance g between the optical waveguides. According to the fourth embodiment, since the optical waveguide 20 has the tapered portion 70, tolerance to dimensional deviations is improved. Even if an error occurs in dimensions, high coupling efficiency is maintained and deterioration of characteristics is suppressed.

<Fifth embodiment>
FIG. 14 is a plan view illustrating a semiconductor optical device 500 according to the fifth embodiment. Descriptions of configurations that are the same as any of the first to fourth embodiments will be omitted. The tapered portion 70 of the optical waveguide 20 and the optical waveguide 41 form a directional coupler 21. The tapered portion 70 has an asymmetric shape with respect to the X-axis. One end 20a (first end) of the optical waveguide 20 in the Y-axis direction is parallel to the X-axis and extends linearly. Another end 20b (second end) of the optical waveguide 20 is inclined from the X-axis and approaches the optical waveguide 41. End portion 20a and end portion 20b form a tapered portion 70.

According to the fifth embodiment, the tapered portion 70 of the optical waveguide 20 and the optical waveguide 41 form the directional coupler 21. Therefore, high coupling efficiency can be obtained. The light transfers from the optical waveguide 20 to the optical waveguide 41 in the directional coupler 21 . It is possible to suppress light loss and emit light from the output port OUT. According to the fifth embodiment, since the optical waveguide 20 has the tapered portion 70, tolerance to dimensional deviations is improved.

FIGS. 15A and 15B are diagrams illustrating calculation results of coupling efficiency. It is assumed that the TE0 mode enters from the optical waveguide 20 and is excited to the optical waveguide 41. The worst value of the coupling efficiency at three wavelengths (1530 nm, 1547.5 nm, and 1565 nm) is calculated by the full vector beam propagation method. The coupling efficiency in FIGS. 15A and 15B is normalized by the maximum value of the coupling efficiency in the fifth embodiment (an example of one-sided taper). The numbers (0.3, 0.8, 0.9, etc.) in FIGS. 15A and 15B represent normalized coupling efficiencies. The maximum value of the normalized coupling efficiency is 1. As the coupling efficiency decreases from 1, the characteristics deteriorate.

The horizontal axis in FIGS. 15A and 15B represents the distance g between the centers of the optical waveguides. The vertical axis represents the width W2 of the optical waveguide 41. The coupling efficiency with respect to changes in distance g and width W2 is evaluated.

FIG. 15A represents the coupling efficiency in the fourth embodiment. As shown in FIG. 15A, the distance g is approximately varied in the range of −800 nm to 800 nm. When the distance g is 0, the center of the optical waveguide 20 overlaps the center of the optical waveguide 41. When the distance g is a positive value, the center of the optical waveguide 20 is located to the left of the center of the optical waveguide 41 in FIG. 13B. When the distance g is a negative value, the center of the optical waveguide 20 is located to the right of the center of the optical waveguide 41 in FIG. 13B. The width W2 is varied within a range of 450 nm to 700 nm.

In the example of FIG. 13A, the tapered portion 70 has a symmetrical shape, so the coupling efficiency in FIG. 15A is symmetrical with respect to the distance g=0. When the distance g is around 0 and the width W2 is from 450 nm to 600 nm, the normalized coupling efficiency is 0.8 or more. When the distance g is 500 nm to 600 nm and the width W2 is around 550 nm to 600 nm, the coupling efficiency is 0.8 or more. When the distance g is about 100 nm to 400 nm and the width W2 is about 450 nm to 550 nm, the coupling efficiency is 0.5 or less. It is considered that unnecessary mode conversion occurs.

FIG. 15B represents the coupling efficiency in the fifth embodiment. As shown in FIG. 15B, the distance g is varied in the range of −600 nm to 600 nm. The width W2 is varied within a range of 450 nm to 700 nm.

When the distance g is from around 0 to 200 nm and the width W2 is from 450 nm to 550 nm, the coupling efficiency is 0.5 or less. When the distance g is 200 nm or more, high coupling efficiency can be obtained over a wide range. When the distance g is around 400 nm and the width W2 is from 450 nm to 650 nm, the coupling efficiency is 0.9 or more.

As shown in FIGS. 15A and 15B, the coupling efficiency can be improved by setting the distance g and the width W2 within appropriate ranges. In the example of FIG. 15A, even if the distance g has an error of about 100 nm around 0, the coupling efficiency can be increased to 0.8 or more by setting the width W2 within the range of 150 nm from 450 nm to 600 nm. can. In the example of FIG. 15B, even if the distance g has an error of about ±150 nm around 400 nm, and the width W2 has an error of about ±100 nm around 550 nm, the coupling efficiency can be 0.8 or more. High coupling efficiency can be obtained throughout the C band.

Although the embodiments of the present disclosure have been described in detail above, the present disclosure is not limited to such specific embodiments, and various modifications and variations may be made within the scope of the gist of the present disclosure as described in the claims. Changes are possible.

10, 12 Substrate 14 Box layer 16 Si layer 20, 41 Optical waveguide 21 Directional coupler 22 Groove 23 Phase adjustment section 24 Terrace 25 Insulating film 25a, 25b Opening 26 Support 27, 60, 62 Electrode 30, 32 Portion 40 Semiconductor element 42 Bonding layer 43 Mesa 44, 48 Cladding layer 45, 47 Optical confinement layer 46 Active layer 49 Contact layer 50 Ring resonator 52 Loop mirror 54 Tapered part 62a Pad 62b Connection part 70 Tapered part 20a, 20b End part 100, 100A , 200, 300, 400, 500 Semiconductor optical device IN Input port OUT Output port

The insulating film 25 covers the side surfaces of the mesa 43 and the top surface of the bonding layer 42 . The insulating film 25 is made of an insulator such as silicon nitride (SiN), silicon oxide (SiO ₂ ), silicon oxynitride (SiON), or the like. The thickness of the insulating film 25 is, for example, 100 nm or more and 600 nm or less. The insulating film 25 has an opening on the upper surface of the mesa 43. An electrode 27 is provided on the upper surface of the contact layer 49. The electrode 27 is made of metal such as gold (Au), for example.

As shown in FIGS. 10A to 10C, the semiconductor element 40 is bonded to the substrate 10. Specifically, the upper surface of the Si layer 16 of the substrate 10 and the surface of the bonding layer 42 of the semiconductor element 40 are activated by nitrogen (N ₂ ) plasma treatment. The activated surface is ultrasonically cleaned in water. The semiconductor element 40 is temporarily bonded to the upper surface of the substrate 10 by bringing the activated surfaces into contact with each other. After temporary bonding, annealing is performed at, for example, 300° C. for 2 hours to remove moisture and strengthen the bonding strength (O ₂ bonding). Two ring resonators 50 and two loop mirrors 52 are located outside the semiconductor element 40. A wavy portion of the optical waveguide 20 is covered with a semiconductor element 40 .

Claims (18)

a substrate having a first optical waveguide made of silicon;
a semiconductor element bonded to the upper surface of the substrate and having a second optical waveguide formed of a III-V compound semiconductor;
The first optical waveguide and the second optical waveguide form a directional coupler in a semiconductor optical device. 2. The semiconductor optical device according to claim 1, wherein the first optical waveguide has a bent shape so as to approach the second optical waveguide. The first optical waveguide has a first part and a second part,
The distance between the second portion and the second optical waveguide is smaller than the distance between the first portion and the second optical waveguide,
3. The semiconductor optical device according to claim 1, wherein the second portion and the second optical waveguide form the directional coupler. The second optical waveguide is located above one end of the first optical waveguide in the width direction, and does not extend to the other end of the first optical waveguide. The semiconductor optical device described above. According to claim 1 or 2, the center of the second portion of the first optical waveguide is spaced apart from the center of the second optical waveguide in the width direction of the first optical waveguide and the second optical waveguide. The semiconductor optical device described above. 6. The semiconductor optical device according to claim 5, wherein at least a portion of the second portion of the first optical waveguide does not overlap the second optical waveguide in the direction in which the substrate and the semiconductor device are bonded. The semiconductor optical device according to claim 3, further comprising a phase adjustment section provided in the first portion of the first optical waveguide. The first optical waveguide and the second optical waveguide form a plurality of the directional couplers,
3. The semiconductor optical device according to claim 1, wherein the plurality of directional couplers are arranged along the extending direction of the first optical waveguide and the second optical waveguide. The semiconductor element has a first semiconductor layer and a mesa,
the first semiconductor layer is bonded to the top surface of the substrate;
3. The semiconductor optical device according to claim 1, wherein the mesa protrudes from the first semiconductor layer to a side opposite to the substrate and includes the second optical waveguide. The mesa has a second semiconductor layer, a third semiconductor layer, and a fourth semiconductor layer,
A second semiconductor layer, a third semiconductor layer, and a fourth semiconductor layer are stacked in this order on the first semiconductor layer,
10. The semiconductor optical device according to claim 9, wherein the third semiconductor layer has a multiple quantum well structure. The substrate has a first layer, a second layer, and a third layer stacked in order,
The first layer and the third layer are made of silicon,
The second layer is formed of silicon oxide,
3. The semiconductor optical device according to claim 1, wherein the semiconductor element is bonded to the third layer. The substrate has two first optical waveguides,
3. The semiconductor optical device according to claim 1, wherein the two first optical waveguides and the second optical waveguide form the directional coupler. the semiconductor element has an optical gain;
3. The semiconductor optical device according to claim 1, wherein the semiconductor device functions as a laser device. The first optical waveguide has a tapered portion,
The tapered portion becomes thinner as it approaches the tip of the first optical waveguide,
3. The semiconductor optical device according to claim 1, wherein the tapered portion of the first optical waveguide and the second optical waveguide form the directional coupler. 15. The semiconductor optical device according to claim 14, wherein the tapered portion of the first optical waveguide has an asymmetric shape with respect to a direction in which the first optical waveguide extends. A first end of the first optical waveguide is parallel to the extending direction of the first optical waveguide, a second end of the first optical waveguide is close to the second optical waveguide, and the tapered portion is parallel to the asymmetrical direction. 16. The semiconductor optical device according to claim 15, which forms a shape. 15. The semiconductor optical device according to claim 14, wherein the tapered portion of the first optical waveguide has a shape symmetrical with respect to a direction in which the first optical waveguide extends. a step of bonding a semiconductor element formed of a III-V group compound semiconductor to an upper surface of a substrate having a first optical waveguide formed of silicon;
forming a second optical waveguide in the semiconductor element,
A method of manufacturing a semiconductor optical device in which the first optical waveguide and the second optical waveguide form a directional coupler.

Images