JP2008103600A - Semiconductor optical device - Google Patents

Abstract

A structure of a surface output type semiconductor optical device is simplified to facilitate manufacture. Further, in addition to the light emitting device, it can be used as an optical amplification device.
An active layer region formed on a semiconductor substrate in which an optical transition occurs, a waveguide formed in the active layer region, and an electrode for supplying a current to the active layer region, are provided along the surface of the semiconductor substrate. A semiconductor optical device having a waveguide and outputting light in the vertical direction, wherein the electrodes are made of conductive materials electrically connected to each other, have a two-dimensional periodic shape, and the semiconductor Formed along the waveguide in the vicinity of the waveguide of the optical element, electrically connected to the active layer region, and acting as a photonic crystal on the light wave propagating through the waveguide, and perpendicular to the waveguide The target light is guided in the direction, and its period is almost the wavelength length of light propagating in the waveguide, and a current is passed between the electrode and the semiconductor substrate to feed the active layer region. To do.
[Selection] Figure 1

Description

  The present invention relates to a semiconductor optical device that can be used as a light emitting element or an optical amplifying element that outputs light from the surface of a semiconductor substrate.

  In general, it is known that a semiconductor optical device is small, has high conversion efficiency, and has a long lifetime. For example, semiconductor light-emitting diodes (LEDs) are low-cost, long-life solid-state light sources that can be widely used for lighting. Also, surface emitting LEDs and surface emitting laser diodes are attracting attention because of the great advantages that can be obtained by integrating them with a single chip together with other electronic circuits.

  However, usually, for example, most of the light inside the LED is confined in the semiconductor because the refractive index between the semiconductor and air is greatly different, which reduces the light extraction efficiency and becomes an obstacle to improving the output. ing.

  Several methods have been reported to overcome this problem of reducing light extraction efficiency. One of them is to use Bragg diffraction in a periodic structure such as a photonic crystal (PhC). Further, two-dimensional (2D) PhC with second-order Bragg diffraction is used, and in this method, a surface emitting laser (SEL) can be realized (Patent Document 1).

  Therefore, it can be said that 2D-PhCs is a promising candidate for the next generation of LEDs and SELs. However, the disadvantage is that the manufacturing process for 2D-PhC devices is complex. This is because a light-emitting element using 2D-PhC needs to be operated by current injection. That is, as is well known, dry etching is generally used in the manufacturing process. However, process damage occurs when producing PhCs, and wasteful current increases, resulting in a decrease in luminous efficiency.

  As a method for avoiding this problem, a method of forming 2D-PhC completely separated from the active layer and then fusing the wafer with the active layer has been reported. However, since the present invention does not use wafer fusion and the structure is different, it will not be described in detail here.

[Conventional example 1]
The semiconductor light emitting device disclosed in Patent Document 1 includes a first conductive type semiconductor layer, a second conductive type semiconductor layer, an active layer, a photonic band layer, and a light emission surface. It has a photonic band structure. The photonic band layer is optically coupled to the active layer and is provided along the direction in which the main surface of the substrate extends.

[Conventional example 2]
Patent Document 2 includes a two-dimensional photonic crystal surface emitting laser, particularly a photonic crystal periodic structure in which a refractive index period is two-dimensionally arranged in or near an active layer that emits light by carrier injection. A two-dimensional photonic crystal surface emitting laser that emits surface light by resonating with a nick crystal is disclosed.

  However, the semiconductor light emitting device disclosed in the above-mentioned Patent Document 1 is such that an active layer is further epitaxially grown on the photonic band layer, and a complicated process is used.

Further, in the two-dimensional photonic crystal surface emitting laser disclosed in Patent Document 2, an electrode is provided in a region surrounding the photonic crystal, and current cannot be supplied uniformly to the active layer.
JP 2000-332351 A JP 2003-273456 A

  The structure of a conventional surface output type semiconductor optical device using a photonic crystal is simplified and easily manufactured, and a uniform current is supplied to the active layer region. Furthermore, it is used not only as a light emitting device but also as an optical amplification device.

  According to the present invention, the manufacturing process of the photonic crystal portion can be simplified as compared with the conventional manufacturing process. Further, since the photonic crystal portion is used as an electrode, the current supply to the active layer region can be made uniform. In addition, because of its structure, it is possible to easily input light from the outside, so that it can be used as an optical amplification device.

  First, the present invention relates to a semiconductor substrate, an active layer region formed on the semiconductor substrate and causing an optical transition, a waveguide formed in the active layer region, and an electrode for supplying current to the active layer region. And a semiconductor optical device having a waveguide along the surface of the semiconductor substrate and outputting light including a component in a direction perpendicular to the waveguide.

  In the above-mentioned semiconductor optical device, the above-mentioned electrode also constitutes a photonic crystal. This is composed of conductive substances electrically connected to each other, and has a two-dimensional periodic shape, and is formed along the waveguide in the vicinity of the waveguide of the semiconductor optical device. . The electrode is electrically connected to the active layer region. This electrode acts as a photonic crystal to guide light in the direction perpendicular to the waveguide to the light wave propagating through the waveguide, and the period in the two-dimensional periodic shape is The wavelength of the secondary gamma point of the photonic band realized by the photonic crystal is determined to be the same as the emission wavelength. Further, a current is supplied between the electrode and the semiconductor substrate to supply power to the active layer region.

The design guideline for the period Δ in the shape having the above two-dimensional periodicity may be approximately the same as the used wavelength (λ _o ). However, it must be calculated in consideration of the equivalent refractive index ( _ne ) of the waveguide at the wavelength used. It is therefore Δ ≒ λ _o / n _e. When this period Δ is satisfied, the light guided through the waveguide forms a standing wave having the period Δ. However, this guideline is the same as the design of the distributed feedback type resonator using the second-order diffraction in one dimension, and does not strictly handle all the waveguiding and diffraction phenomena that occur in two dimensions. Therefore, when calculating strictly, it is performed by a method that can describe the interference of all the light waves developed on the two-dimensional plane. For example, it can be calculated with high accuracy by photonic band structure calculation in wave number space by plane wave expansion method, electromagnetic field analysis by FDTD method (Finite Difference Time Domain Method) and the like.

  In addition, it is desirable to form the electrode with a material having a high conductivity in order to supply a current. However, since the conductivity of a material with a relatively low dielectric loss in the frequency region of light is generally low, A multi-layered electrode using a material having a relatively low conductivity at a portion relatively close to the waveguide and a material having a relatively high conductivity at a relatively far portion. Consists of.

  In addition, a metal is used as the material of the electrode.

  In addition, the active layer region is a region mainly including electron transition by stimulated emission in order to perform laser oscillation and optical amplification.

  Alternatively, it is assumed that the active layer region is a region mainly composed of electron transition due to spontaneous emission in order to form a light emitting diode.

  In addition, the present invention further includes an incident waveguide for inputting light from the outside to the above-described waveguide to constitute an optical amplifier, and the input light is amplified and output by the above-mentioned active layer region. To do.

  Moreover, this invention is equipped with the structure which inputs the light from said incident waveguide into said waveguide through said electrode.

  Embodiments of the present invention will be described below in detail with reference to the drawings. In the following description, devices having the same function or similar functions are denoted by the same reference numerals unless there is a special reason.

FIG. 1 shows a schematic diagram of a semiconductor optical device of the present invention. This is intended for LEDs (light emitting diodes) and semiconductor lasers. This semiconductor optical device includes an active layer region 2 and an electrode 4 for supplying current to the active layer region 2 on an N-type semiconductor substrate 1. The active layer region 2 includes a P-type region 2a, an N-type region 2b, which are cladding layers, and a waveguide region 3. N-type region 2 b is electrically connected to the N-type substrate, and P-type region 2 a is electrically connected to electrode 4. A wiring pad 5 is connected to the electrode 4, and a forward voltage is applied between the wiring pad 5 and the N-type substrate 1. Here, the electrode 4 constitutes a two-dimensional photonic crystal spreading on the surface, and the period may be set to be approximately the same as the wavelength of light to be output as described above. However, it must be calculated in consideration of the equivalent refractive index ( _ne ) of the waveguide at the wavelength used. It is therefore Δ ≒ λ _o / n _e. When this period Δ is satisfied, the light guided through the waveguide forms a standing wave having the period Δ.

  It is obvious that the same function can be provided by switching the PN polarity of the semiconductor and the voltage application direction of the power source 6. If the function of the N-type region 2b can be performed by the N-type semiconductor substrate 1, this need not be provided.

  The waveguide region 3 may be a PN junction as shown in FIG. 2 (a), or may be configured using a quantum well as shown in FIG. 2 (b). Further, a configuration using quantum dots as shown in FIG. 2C or a configuration using multiple quantum wells as shown in FIG. The point is to have a structure having a light emitting action or a light amplifying action.

  The end surface of the active layer region 2 is in contact with air and has a function of reflecting light due to the difference in dielectric constant. However, when more complete reflection is required, a means for reflecting light may be provided on the end surface. Good. It is known that metal films and multilayer dielectric films are effective for reflecting light.

The electrode 4 can be formed by, for example, a lift-off process as shown in FIGS. The feature of the lift-off process is that the process damage is small.
(A) First, a resist pattern 10 is formed in the P-type region 2a.
(B) The Ti (titanium) layer 11 and the gold layer 12 are formed to a thickness capable of a lift-off process.
(C) The resist layer is removed to leave an electrode pattern.
(D) A protective film 13 is formed as necessary.
Here, the Ti layer 11 functions as an adhesion layer. The gold layer 12 serves as a current supply layer. If necessary, a silver layer or an aluminum layer may be formed thereon or in place of the gold layer.

  The electrode formed as described above slightly changes the effective refractive index of the waveguide. The important point of the present invention is that the effect as a photonic crystal reaches the waveguide by changing the effective refractive index. According to the simulation using the plane wave expansion method, a significant band gap can be confirmed in the photonic band structure if the calculated effective refractive index difference of the waveguide is 0.1% or more with and without metal. For this reason, it can be seen that it is sufficient to design at least 0.1% or more of the effective refractive index difference when calculating the waveguide.

  The semiconductor optical device of FIG. 3 is directed to the optical amplifying element, and includes a matching region 8 for introducing light from the optical fiber 9 that is an incident waveguide into the active layer region 2 with low loss. The matching region 8 shown in FIG. 3 is a grating formed at the same time as the electrode 4 and includes a wiring 7 for supplying a current to the matching region. In order to introduce light from the optical fiber 9 into the active layer region 2 with low loss, the light is incident obliquely on the surface so as to be captured by the waveguide region 3. Here, the matching region 8 is for improving the incident efficiency of light, and it is not necessary to use this when the loss at the time of incidence can be ignored. Even in this case, since it is sufficient that the incidence is performed within the critical angle, the incidence can be performed with a gentler limit as compared with the surface emitting laser of the conventional vertical resonator. Further, it is clear that optical amplification can be performed even if light is incident from the end face of the active layer region 2.

  The present invention can be operated by light excitation in addition to the current excitation shown in FIG. The case of a waveguide region in which a semiconductor heterojunction having a slab type (one-dimensional type) waveguide mode is produced by crystal growth will be described below.

A multi-quantum well (MQWs) of 6 cycles of In _0.2 Ga _0.8 Sb / Al _0.3 Ga _0.7 Sb (thickness = 4.5 nm / 20 nm) was used as the material for the active layer. MQWs were grown on the thick AlSb lower cladding layer, and then a 100 nm Al _0.3 Ga _0.7 Sb upper cladding layer was grown on the MQW layer by molecular beam epitaxy (MBE). At some MQW locations, it has a slab waveguide structure and the optical mode is slightly oozing out to the surface. After MBE growth, a 100 μm × 100 μm 2D metal PhCs pattern was fabricated by electron beam lithography (EBL). Ti / Au (thickness = 50 mm / 500 mm) was deposited on the EBL pattern, and 2D metal PhCs was fabricated by a lift-off process. FIG. 5A shows a scanning electron microscopic photograph of the produced 2D metal PhCs.

Table 1 shows the average lattice constant (Lc) of the 2D metal PhCs thus produced.

  The size of the 2D metal PhCs was determined by matching the frequency of the secondary gamma edge of the 2D metal PhCs with the frequency of the photoluminescence (PL) spectrum from the MQWs. The optical properties of 2D metal PhCs were investigated with a 532 nm continuous wave laser focused by an objective mirror with a numerical aperture of 0.5 PL. PL was measured at room temperature.

  The PhCs consists of a square lattice having a size 0.3 times the lattice constant. Here, the TE-mode photonic band structure was calculated based on the effective refractive index n of the slab waveguide depending on whether Ti / Au was present or not. That is, when there is a metal on the surface, n is slightly smaller than when there is no metal. In realizing PhCs, the two large and small refractive indexes become the refractive index contrast. Further, a photonic band diagram of the TE mode is shown in FIG. This is based on calculation, and this calculation was performed using the plane wave expansion method.

  The measurement result of PL is shown in FIG. FIG. 6 shows the PL results for 2D metal PhCs with 10 mW excitation according to the presence / absence of the pattern (indicated as W / O PhC in FIG. 2). When the period of 2D metal PhCs is changed by about 0.28-0.282 (insertion in FIG. 6), the PL peak position from 2D metal PhCs is the reference frequency (Lc / La) (where Lc is 2D metal PhCs). The average lattice constant of La changed with respect to the wavelength of light in free space. These values are slightly lower than the calculated second order gamma edge of the photonic band structure of 0.287 (Lc / La). However, the discrepancy is within a few percent. This means that the above calculation method is appropriate, and it shows that a photonic crystal using a metal periodic structure is effective. In addition, 2D metal PhCs reduces the actual excitation power by approximately half. This is because 2D metal PhCs occupies 51% per unit pattern. However, the PL peak intensity is stronger when 2D metal PhCs is present than when it is absent. These phenomena indicate that Bragg diffraction has increased external quantum efficiency.

  The increase rate is approximately 4 for sample e. (PL intensity ratio × actual excitation power ratio). In addition, optical amplification from the edge of the secondary gamma band was observed. This is because light has a slow group velocity at the band edge, and laser operation is possible with optical amplification that can be expected from now on. Therefore, for sample e, PL that depends on the excitation intensity was measured in order to clarify the optical amplification. The results are shown in FIG. When the input power is 16 to 17 mW, there is a jump in the slope of the output characteristic in FIG. 7, and a threshold value at which an amplification action appears can be obtained. From this, it was possible to observe the PL intensity amplification according to the slow group velocity at the second-order gamma band edge with the 2D metal PhC. On the other hand, broad PL spectra of all samples were obtained. These broad PL spectra are probably related to variations in the structure of 2D metal PhC, as shown in FIG. 5 (a). The fluctuation of the period of 2D metal PhC is about 8%. This period variation affected the photonic band structure that approximately matched the full width at half maximum (FWHM) of PL. Another possibility is the metal effect. Although 2D metal PhC treated as metal operated as an optical cavity, the metal caused loss of light. In the optical cavity, the FWHM reflects the quality (Q) factor. The Q factor is defined by the proportion of field energy stored in the cavity of power dissipated in the cavity. Thus, 2D metal PhC exhibits a lower Q factor than conventional ones that do not include materials that are optically lossy.

  Although the optical mode (TE) overlap with the metal region is only 0.07%, the loss of light by the metal is low. In practice, the loss of light due to the metal can be ignored. Therefore, if an accurate 2D metal PhC is made, the FWHM can be narrowed, thereby enabling laser operation.

  Also, if the MQWs position can be close to the metal, the optical mode overlap increases and the effective refractive index n of the metal waveguide decreases. This effect (overlap) increases the loss of light at the metal, but the later effect (decrease in the effective refractive index n) results in a greater n contrast between the slab waveguides with or without the metal, and more Create good optical cavities and increase photonic band gap. Therefore, it is desirable to optimize the balance for the laser operation.

It is a schematic diagram which shows the semiconductor optical device of this invention. It is a schematic diagram which shows the structure of an active layer area | region. It is a schematic diagram which shows the formation process of the electrode which is a photonic crystal. It is a schematic diagram which shows the semiconductor optical device which performs optical amplification. (A) is a scanning electron micrograph of the produced 2D metal PhCs, (b) is an index, and (c) is a band diagram. It is a figure which shows the measurement result of photoluminescence (PL). It is a figure which shows the amplification effect | action in photoluminescence.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 N type semiconductor substrate 2 Active layer area | region 2a P type area | region 2b N type area | region 3 Waveguide area | region 4 Electrode 5 Wiring pad 6 Power supply 7 Wiring 8 Matching area | region 9 Optical fiber 10 Resist pattern 11 Ti (titanium) layer 12 Gold layer

Claims (7)

A semiconductor substrate;
An active layer region formed on the semiconductor substrate and causing optical transition;
A waveguide formed in the active layer region,
An electrode for supplying current to the active layer region, a semiconductor optical device having a waveguide along the surface of the semiconductor substrate and outputting light including a component in a direction perpendicular to the waveguide,
The electrode is composed of conductive substances electrically connected to each other, has a two-dimensional periodic shape, and is formed along the waveguide in the vicinity of the waveguide of the semiconductor optical element. It is electrically connected to the active layer region, and acts as a photonic crystal on the light wave propagating through the waveguide to guide the target light in a direction perpendicular to the waveguide.
A semiconductor optical device, wherein a current is passed between the electrode and the semiconductor substrate to supply power to the active layer region.   The electrode has a structure including two conductive layers, and a material having a relatively low conductivity is used in a portion relatively close to the waveguide, and a material having a relatively high conductivity is used in a relatively far portion. 2. The semiconductor optical device according to claim 1, wherein the semiconductor optical device is composed of an electrode having a multilayer structure.   3. The semiconductor optical device according to claim 1, wherein the electrode material is a metal.   4. The semiconductor optical device according to claim 1, wherein the active layer region is a region mainly composed of electron transition by stimulated emission.   4. The semiconductor optical device according to claim 1, wherein the active layer region is a region mainly including electron transition by spontaneous emission.   5. The semiconductor optical device according to claim 4, further comprising an incident waveguide for inputting light from the outside to the waveguide, and amplifying the input light in the active layer region and outputting the amplified light.   The semiconductor optical device according to claim 6, further comprising a configuration for inputting light from the incident waveguide to the waveguide through the electrode.