JP2006323210A - Spot size converter - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a spot size converter in which device miniaturization is achieved and transmission characteristic is improved. <P>SOLUTION: On an underclad 86, a first core 81 whose refractive index is higher than that of the underclad 86 is formed. The first core 81 has a projecting portion 87 and an opposite waveguide. A second core 83 having a straight portion 84 and a tapering portion 85 are formed on the first core 81. In the projecting portion 87, the first core 81 and the second core 83 are almost equal in width. The tapering portion 85 is formed on the opposite waveguide. In addition, the straight portion 84 is formed at least on a slab 82. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a spot size converter, and more particularly to a spot size converter that enables low-loss connection of optical components having different mode fields.

  Spot size converters (SSC) have been developed for a long time from the idea of introducing light oscillated from a semiconductor laser (LD) into a fiber with high efficiency, and their application to devices has become active. In particular, an LD with SSC, in which a semiconductor laser is provided with SSC, is well known. The factor that determines the coupling efficiency between the fiber and the LD is the size of the spot. The spot size of single mode fiber is about 5 um, whereas the spot size of LD is as small as about 1 um. If these are simply combined, a large coupling loss occurs. Therefore, by providing a spot size converter in the LD, the spot size of light emitted from the LD is increased. In this way, the coupling efficiency between the LD and the fiber is improved. However, when the LD has an SSC structure, the characteristics of the LD are often adversely affected, and the yield is also reduced at the manufacturing stage, which increases the cost.

  On the other hand, the quartz-based planar optical waveguide circuit is made of the same quartz-based material as that of the fiber, and is designed to be almost the same as that of the fiber even in the spot size, and originally has good coupling with the fiber. However, these optical waveguides have been downsized in recent years. In order to reduce the size, the relative refractive index difference of the waveguide (also simply referred to as “Δ” in the present specification) is reduced, and the waveguide bending radius is reduced, thereby realizing a reduction in circuit size. At that time, not only the circuit becomes small, but also the spot size becomes small, and the deterioration of the coupling with the fiber becomes a problem.

  Therefore, various SSCs have been proposed in the past for quartz-based waveguides. Basically, the configuration is the same as that shown in the semiconductor field. FIG. 1 is a diagram showing a configuration of a conventional spot size converter. FIG. 1A shows an SSC 2 having a tip portion with a tapered structure 1 as shown in Non-Patent Document 1, and FIG. 1B actually shows an SSC 2 as shown in Non-Patent Document 2. FIG. 1C shows an SSC 5 having a structure in which the waveguide core end face is enlarged. As shown in Non-Patent Documents 3 and 4, the SSC 5 has two waveguides with different spot sizes stacked or included.

  In the structure shown in FIG. 1A, in the tapered structure 1, when the mode field of the propagating light is narrowed, if the width becomes narrower than a certain width, the light cannot be confined in the core, and the field starts to expand and the spot size Is an SSC utilizing the fact that the In FIG. 1B, the end size of the core 3 is actually enlarged to widen the field and convert the spot size. Further, in FIG. 1C, the spot size is obtained by eliminating the waveguide core 7 having a different Δ from the waveguide core 6 and transferring the light to the waveguide core 7 which is the other core. Convert.

  Among the above, in the structure of FIG. 1A, the spot size increases as the distance from the tip of the SSC 2 increases. Low-loss coupling is realized by positioning a core such as a fiber at a position where a desired spot size can be obtained. For this reason, another waveguide circuit cannot be manufactured after the tip disappears in this SSC. Also in FIG. 1B, there is no problem when the core 3 is enlarged to increase the spot size and couple with a fiber or the like, but it is difficult to construct another optical circuit using the expanded core. This is because the core is large and cannot achieve single mode.

  On the other hand, the case of FIG. 1C is different from the SSC of FIG. 1A and FIG. That is, it is composed of two kinds of waveguide cores 6 and 7 having Δ different from each other, and each waveguide has no problem even if it constitutes each circuit, and the SSC portion has a structure as shown in FIG. Just connect. In FIGS. 1A and 1B, the position is limited to the periphery of the end face of the circuit board, but the structure shown in FIG. 1C can be distinguished from others in that it can be formed at an arbitrary location.

  In the following description, what is simply described as SSC is composed of at least two types of Δ waveguides, and a part of the high Δ waveguide is formed as a taper structure on or inside the low Δ waveguide. The SSC as shown in FIG.

  In Non-Patent Document 3, when coupling an LD to a waveguide end face, coupling efficiency is improved by coupling to a waveguide with a small spot size of high Δ, and spot size conversion is performed by performing spot size conversion via SSC. The spot size is enlarged by shifting to a waveguide. A low loss connection between the low Δ waveguide and the fiber is realized.

  This can also be applied to a planar optical waveguide type optical transceiver using a thin film filter. FIG. 2A shows a configuration of an optical transceiver 20 using a conventional planar optical waveguide circuit and a thin film filter. FIG. 2B is a diagram illustrating an enlarged portion Q of a connection portion between the SSC and the LD included in the waveguide in FIG.

  As shown in FIG. 2 (a), a groove is formed at the intersection 22 of the intersecting waveguide 21 embedded in the clad composed of the over clad and the under clad, and light is demultiplexed in the groove. A thin film filter 23 is inserted, and SSCs are respectively provided at two ends of the waveguide 21. A semiconductor light receiving element and an LD are mounted on each of the SSCs. FIG. 2B shows an enlarged view of a connection portion between the SSC 24 and the LD 25. In FIG. 2B, the SSC 24 includes a high Δ waveguide 26 and a low Δ waveguide 27, and the low Δ waveguide 27 has a structure that is widened by a taper in a direction opposite to the connection portion with the LD 25. The high Δ waveguide 26 having a structure in which the waveguide is narrowed by the taper in the above direction is provided. In such a configuration, the light that leaks due to the decrease in the core size at the tapered portion shifts to the low Δ waveguide 27 that is the lower waveguide.

  The signal light incident from the input / output fiber passes through the thin film filter 23 and enters the light receiving element. Further, the light output from the LD 25 is coupled to the high Δ waveguide 26 with high efficiency and guided to the waveguide 21 through the structure shown in Non-Patent Document 3 described above. When the LD 25 is directly incident on the low Δ waveguide 27 (Δ = 0.3%), the coupling loss is about 11 dB, but according to Non-Patent Document 3, the improvement is as much as 8 dB up to about 3 dB through the SSC 24. ing.

  On the other hand, the coupling portion between the waveguide and the fiber can realize the coupling with the fiber with low loss because of the connection with the low Δ waveguide. Since the filter portion is formed with a groove, there is no waveguide, and the propagation light is diffracted and spread, and a loss occurs when the light transmitted or reflected by the thin film filter 23 is coupled to the waveguide again. In order to reduce this loss, it is preferable to use a waveguide having a large spot size, that is, a low Δ so as to suppress diffraction, in the filter insertion portion.

  Thus, by using the structure as shown in FIG. 1C, it is possible to improve both the coupling efficiency with the LD, the insertion loss of the filter, and the coupling with the fiber in the optical transceiver. Become.

“Compact and low-loss arrayed waveguide grating module with tolerance-relaxed spot-size converter” PJOTO. TECH. LETT., 5, (2), p239, 2003 “High density and low-loss arrayed waveguide gratings composed of high index difference PLCs”, IPR’02 (Vancouver, CANADA), IThG1-1-3, Jul, 2002 “Efficient coupling of a semiconductor laser to an optical fiber by means of a tapered waveguide on silicon”, Appl. Phys. Lett. 55 (23), 1989 "Prototype of laminated optical branching device for optical gate array" The 51st Applied Physics Related Conference Lecture Proceedings 28p-N-1 Masao Kawachi “Planar Lightwave Circuit” Denki Theory C, Vol.113, No.6, 440-445, 1993

  FIG. 3 shows the transmission characteristics of an optical waveguide circuit including an SSC having a structure in which conventional heterogeneous Δ waveguides are stacked. This is a transmission characteristic when light is incident from the high Δ waveguide side using a fiber having a very high Δ and a small spot size, and a normal single mode fiber is connected to the low Δ waveguide. As described above, it is very useful to couple optical components having different spot sizes with a low loss by using SSC for different types of Δ waveguides. However, as a result of actually producing and evaluating the SSC as shown in FIG. 1 (c), it is clear by the inventors that a problem such as a ripple in the wavelength characteristic as shown in FIG. 3 occurs. It became. It has also been found that the amplitude of this ripple depends on the taper length of the SSC transition portion, and the amplitude increases as the taper becomes shorter.

  Such a ripple in the transmission characteristics is a problem that is not regarded as important in SSCs such as semiconductor lasers that have been developed in advance, but is not grasped. However, the planar optical waveguide circuit mainly has a passive function, requires a stable transmittance in a wide wavelength range, and strongly requires that the transmission characteristics have no wavelength dependency.

  FIG. 4 shows the frequency of LD coupling efficiency when the optical transceiver shown in FIG. 2 is manufactured using the above-described planar optical waveguide on which the conventional SSC is mounted. In this case, the average coupling efficiency was 4.8 dB, and the minimum and maximum values were 3.7 dB and 5.8 dB, respectively.

  Despite the devices fabricated in the same way, the device characteristics had large variations. The cause is ripples appearing in the transmission characteristics shown in FIG. Even if the SSC of the same design is measured, the wavelength at which the ripple takes the maximum and the minimum is not the same wavelength due to a manufacturing error. The wavelength varied from element to element. In some elements, the transmissivity becomes maximum at the oscillation wavelength of the LD, and in some elements, the transmissivity becomes minimum. Therefore, the variation from element to element increases.

  As described above, the conventional SSC has a problem that ripples that cannot be controlled due to transmission characteristics are generated. Further, if the taper length of the SSC transition portion is increased in order to suppress the ripple, there is a problem that the length of the SSC on the circuit is several millimeters and the circuit cannot be reduced in size.

  FIG. 5 (a) shows a conventional SSC-equipped array diffraction grating (AWG) 51 fabricated with a high Δ waveguide and reduced in size. Such an AWG has a problem that when it is manufactured using a high Δ waveguide, the coupling with the fiber is deteriorated. Therefore, the coupling efficiency with the fiber is improved by mounting the SSC at the input / output. In FIG. 5A, the AWG 51 includes an arrayed waveguide 52 and a waveguide 53 as an input / output port including a plurality of waveguides. These waveguides are embedded in a clad composed of an over clad and an under clad. Each waveguide of the waveguide 53 is provided with an SSC 54 as shown in FIG. The SSC 54 includes a high Δ waveguide 55 and a low Δ waveguide 56, and the high Δ waveguide 55 is provided on the low Δ waveguide 56. The low Δ waveguide 56 and the optical fiber 57 are optically connected.

  Even in such an AWG circuit, the SSC section causes ripples due to the transmission characteristics, so that the transmittance of each channel of the AWG transmission spectrum is affected by the ripples as shown in FIG.

  FIG. 7 is a diagram showing a comparison of the chip sizes of the AWG 71 without the SSC and the AWG 72 with the SSC 73 mounted. As described above, if the length of the SSC transition portion is increased in order to prevent the influence of the ripple of the SSC portion, as shown in FIG. 7, the AWG can be realized with a small size using a high Δ waveguide. However, the SSC 73 becomes large, and the entire circuit including the SSC 73 cannot be reduced in size. There is a problem that if the Δ is further increased in the future, the SSC unit becomes larger than the AWG circuit, which may lead to an overturned result.

  The present invention has been made in view of such problems, and an object of the present invention is to provide a spot size converter capable of realizing downsizing of the apparatus and improved transmission characteristics.

  In order to achieve such an object, the present invention provides a first waveguide formed on an under clad and having a higher refractive index than that of the under clad. A waveguide, a slab region provided at one end of the second waveguide, and a third provided on the side of the slab region opposite to the side on which the second waveguide is provided. A first waveguide having a first waveguide, and a fourth waveguide and a tapered waveguide formed on the first waveguide, and being refracted more than the first waveguide. A fifth waveguide having a high rate, and either the fourth waveguide or the tapered waveguide is formed on the second waveguide, and the third waveguide is formed. On the waveguide, the other of the fourth waveguide and the tapered waveguide is formed, and at least On the slab region, and wherein the fourth waveguide are formed.

  The invention according to claim 2 is a first waveguide formed on the undercladding and having a refractive index higher than that of the undercladding, the second waveguide and one end of the second waveguide A first waveguide having a slab region provided on the second slab region and a third waveguide provided on a side of the slab region opposite to the side on which the second waveguide is provided; A fourth waveguide formed in a first waveguide and having a tapered waveguide, and a second waveguide having a higher refractive index than that of the first waveguide; In the second waveguide, either the fourth waveguide or the tapered waveguide is embedded and formed in the third waveguide. And the other of the tapered waveguides is embedded and formed at least In the Love region, wherein said quaternary waveguide being embedded in form.

  The invention according to claim 3 is the invention according to claim 1 or 2, wherein the second waveguide and the third waveguide are a longitudinal direction of the second waveguide and the third waveguide. It is provided in the slab region so as to substantially coincide with the longitudinal direction.

  According to a fourth aspect of the present invention, in the first or second aspect of the invention, the fifth waveguide is bent at a bending radius such that a part of the fourth waveguide does not cause a bending loss. It is a bending waveguide, and the third waveguide is provided in a position different from a position facing the second waveguide in the slab region.

  According to a fifth aspect of the present invention, in the invention according to any one of the first to fourth aspects, when the fourth waveguide is formed on the second waveguide, the second waveguide The width and the width of the fourth waveguide are substantially equal.

  The invention according to claim 6 is the invention according to any one of claims 1 to 5, wherein the angle at which the light incident from the second waveguide spreads in the slab region is φ, and the length of the slab region Is L and the width of the slab region is W, the width W of the slab region is set so as to satisfy W> Ltanφ.

  The invention according to claim 7 is the invention according to any one of claims 1 to 5, wherein the slab region is a multimode interference waveguide, and the second waveguide and the third waveguide are included. The optical waveguide in which the tapered waveguide is formed has a light intensity higher than that of the optical branching region of light incident on the slab region from the waveguide in which the tapered waveguide is not formed. It is provided in a weak area.

  According to an eighth aspect of the present invention, in the seventh aspect of the present invention, the region where the light intensity is weak is a region between adjacent optical branching regions of the optical branching region branched into N (N is an integer of 2 or more). It is characterized by being.

  The invention according to claim 9 is the invention according to claim 7 or 8, characterized in that the region where the light intensity is weak is the region where the light intensity is the weakest compared to the light intensity of the light branching region. .

The invention of claim 10, wherein, in the invention of claim 8 or 9, wherein the refractive index of the under-cladding and n _L, the refractive index of the slab region and n _c, the length of the slab region is L, The wavelength of incident light to the slab region is λ _o , which corresponds to the polarization of the incident light, is 0 when the polarization is TE, and σ is 1 when the polarization is TM. When the width of the slab region with respect to the fundamental mode in the slab region is W _e , W _e = (λ _o · N · L / n _L ) (1), and the tapered waveguide is formed. The width W _{s of the} slab region when the distance from the boundary between the unguided waveguide and the slab region to the N-branched optical branch region is the shortest is W _s = W _e − (λ _{o / π) (n L / n} c) 2σ (n L 2 -n c 2) - Is expressed as ^{0.5) (2),} the width of the slab region, depending on the length and the desired N of the slab region, the equation (1) and (2) that is set so as to satisfy the It is characterized by.

  As described above, according to the present invention, since the slab region is provided in the first waveguide having a refractive index lower than that of the fifth waveguide, the generation of ripple can be suppressed, and the transmission characteristics can be reduced. Improvement is possible. In addition, since it is not necessary to lengthen the tapered waveguide of the fifth waveguide in order to suppress the generation of ripples, it is possible to reduce the size of the device.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In the drawings described below, components having the same function are denoted by the same reference numerals, and repeated description thereof is omitted.
In this specification, a waveguide with a high Δ such as “high Δ waveguide” or a waveguide with a low Δ such as “low Δ waveguide” means relative to the two target waveguides. In terms of two waveguides having different Δs, a waveguide having a higher Δ becomes a “high Δ waveguide”, and a waveguide having a lower Δ becomes a “low Δ waveguide”.

Hereinafter, the cause of the ripple in the SSC in which the high Δ waveguide is formed on the low Δ waveguide will be described.
In order to obtain a desired coupling efficiency between the fiber and the LD, each waveguide is desirably single mode. However, when a waveguide with a high Δ is fabricated on or inside a waveguide with a low Δ, when viewed as a single waveguide, each waveguide is stacked and included even if it has a shape that achieves a single mode. Depending on the size of the core and Δ, the single mode is not achieved.

  For example, considering the case where the spot size is converted from the second waveguide core having a high Δ to the first waveguide core having a low Δ, it is only necessary that the propagation light ideally enters only the second waveguide core. In practice, however, even when most of the light is coupled into the second waveguide core, it is also coupled to higher-order modes in the first waveguide core and propagates into the first waveguide core. There is light. In this state, when light reaches the spot size transition portion formed by the taper of the SSC, the modes are coupled. The coupling efficiency gives a wavelength dependence, resulting in ripples on the transmission spectrum. That is, if the light propagating in the first waveguide core is removed by some method and then guided to the transition portion composed of the second waveguide core, the ripple can be suppressed.

(First embodiment)
In the present embodiment, a quartz-based planar optical waveguide circuit (PLC) manufactured by depositing quartz-based glass on a silicon substrate is used.
Quartz-based PLC is fabricated on a silicon substrate by flame deposition and reactive ion etching, and an optical waveguide circuit with low loss and good compatibility with a single mode optical fiber is realized (see Non-Patent Document 5).

  In the present embodiment, for the purpose of improving the coupling efficiency between the LD and the PLC, an SSC is provided at the PLC input / output section (the end of the optical waveguide of the PLC). FIG. 8A is a bird's-eye view of the SSC configuration of the optical circuit according to this embodiment, and FIG. 8B is a top view of the SSC configuration of FIG.

  8A and 8B, the refractive index is lower than that of the second core 83, which is a core that will later be an upper layer, on the under clad 86 formed on the silicon substrate, and the refractive index is lower than that of the under clad 86. A high first core 81 is provided. The first core 81 has a slab (slab waveguide) 82 having a length (slab region length) L parallel to the longitudinal direction of the first core 81. The slab 82 is provided away from the end face of the first core 81 by a predetermined distance.

  On the first core 81, a second core 83 having a refractive index higher than that of the first core 81, which includes a linear portion 84 and a tapered portion 85, is formed. The tapered portion 85 becomes a spot size transition portion. Further, as described above, in the present embodiment, due to the refractive index relationship between the first core 81 and the second core 83, the first core 81 is a low Δ waveguide and the second core 83 is a high Δ waveguide. Become.

  In the present embodiment, as described above, the first core 81 has the slab 82 having an infinite width in a region where at least a part of the linear portion 84 of the second core 83 located in the upper layer is formed, In the portion where the LD is coupled (near the end face of each core), a protruding portion 87 is formed in which the widths of the first core 81 and the second core 83 are substantially equal. That is, the second core 83 is formed so that the tapered portion 85 is formed after the slab region is finished at least from the end surface, and the linear portion 84 of the second core 83 is always formed on at least the slab 82. ing. The position where the slab 82 is provided is set so that at least the protruding portion 87 can be formed. These cores are covered with an overclad (not shown).

  As described above, the first core 81 is formed with the slab 82 so as to have the projecting waveguide (corresponding to the projecting portion 87), and the slab is opposed to the projecting waveguide. In this configuration, the waveguides are connected. In this specification, the “opposing waveguide” is a waveguide included in the first core, and is provided on the surface facing the surface of the slab on which the protruding waveguide is provided. It refers to the waveguide. In the present embodiment, as shown in FIGS. 8A and 8B, the protruding waveguide and the opposing waveguide are arranged on a substantially straight line with respect to the length direction of the slab 82. The slab 82 is provided.

  First, the effect that the widths of the first core 81 and the second core 83 are almost equal in the input portion of the LD, that is, in the protruding portion 87 including the end face will be described.

  When the LD is incident from the end face, a clad having a refractive index as low as possible is arranged around the second core 83 that is a high Δ waveguide so that the effective refractive index difference of the second core 83 is increased. For this reason, in the LD coupling portion, the clad is arranged around the second core 83 by making the widths of the first core 81 and the second core 83 the same, that is, by providing the protruding portion 87. . As a result, since the effective refractive index difference of the second core 83 can be increased, the spot size approaches that of the LD, and there is an effect of increasing the coupling efficiency.

Next, the effect that the second core 83 provides the region where the first core 81 below the straight portion 84 is the slab 82 will be described.
If the LDs are all coupled to the second core 83 which is a high Δ waveguide, there is no problem, but in fact, coupling also occurs to the first core 81 which is a low Δ waveguide located in the lower layer. Then, for the reason described above, a ripple is generated on the wavelength characteristics. Therefore, in order to prevent this, the light that has been coupled into the first core 81 is emitted in the slab region (slab 82). After radiating, the light in the second core 83 is shifted to the first core 81 by the SSC transition portion (tapered portion 85) to convert the spot size. By doing so, it is possible to suppress the occurrence of ripples on transmission characteristics due to light interference in the upper and lower waveguides. In this manner, since the generation of ripples can be suppressed without increasing the spot size transition portion (taper portion 85), the apparatus can be downsized.

  Here, the size of the slab will be described. It is preferable to make the length L of the slab region as large as allowed in the circuit. In coupling of waveguides facing each other across the slab, the light radiated into the slab from one waveguide (in the configuration of FIG. 8, the protruding portion 87 of the first core 81) is the same waveguide (opposing) In the case of coupling to a waveguide), the coupling efficiency deteriorates in proportion to the inverse of the square of the distance L in calculation. The purpose of the slab is to radiate the light that has been coupled to the first core 81 having the slab as described above. Therefore, it is effective for suppressing the ripple if the length L of the slab region is as long as possible. is there.

The width of the slab is defined as infinite length, but it is finite in actual production. Here, in the SSC according to the present embodiment shown in FIG. 8, a top view of a configuration in which the slab 82 having an infinite length is replaced with a slab 88 having a finite length W ₁ is shown in FIG. As can be seen from FIG. 9, the light that has entered the slab 88 incident from the protruding portion 87 of the first core 81 spreads at an angle φ. Therefore, emitted light 90 emitted from the projecting portion 87 enters the slab of the first core 81, if the slab width W ₁ is narrow, it becomes reflected light 91 is reflected by the slab sides, opposing guide of the first core 81 It may be incident on the waveguide 89. Then because weakens the effect of providing the slab 88, it is desirable in this case that the width W _1. That is, it is possible to suppress the ripple by providing the slab 88, by increasing the width W _1, it can be realized inhibition of further ripples.

It is possible to easily calculate how much the slab width W is necessary. For simplicity, it is assumed that there is no second core 83, and only the first core 81 is considered. When the relative refractive index difference delta ₁ with respect to the cladding of the first core, the light enters the slab from one waveguide extends at an angle phi. This angle φ is obtained using Δ ₁

It becomes. Since the length of the slab area (slab 88) allowed in the circuit is L, at least the width W of the slab is approximately
W> Ltanφ
If present, the light reflected from the side surface of the slab will not be coupled to the other waveguide.

  In the present embodiment, the case where the second core is located on the first core has been described. However, the waveguide structure is not limited to this. For example, the shape which the 2nd core encloses in the 1st core can be considered. FIGS. 10A and 10B are bird's-eye views of an SSC according to the present embodiment in which a high Δ waveguide is embedded in a low Δ waveguide.

  In FIG. 10A, the refractive index is lower than that of the second core 103, which is the core provided inside, on the underclad (not shown) formed on the silicon substrate, and higher than that of the underclad. A first core 101 is provided. The first core 101 has a slab 102 having a length L parallel to the longitudinal direction of the first core 101 (the length of the slab region) L and a width that can be regarded as an infinite length. The slab 102 is provided away from the end face of the first core 101 by a predetermined distance. That is, a convex portion 106 is formed at a predetermined distance from the end surface of the first core 101 along the longitudinal direction, and the slab 102 is connected to the convex portion 106. Of course, the convex portion 106 is a waveguide included in the first core 101.

  In the first core 101, a second core 103 having a refractive index higher than that of the first core 101, which includes a straight portion 104 and a tapered portion 105, is formed. The tapered portion 105 becomes a spot size transition portion. Further, as described above, in the present embodiment, due to the refractive index relationship between the first core 101 and the second core 103, the first core 101 is a low Δ waveguide and the second core 103 is a high Δ waveguide. Become.

In the configuration of FIG. 10A, as described above, the first core 101 has the slab 102 having an infinite width in the region where at least a part of the linear portion 104 of the second core 103 located inside exists. Even in the portion where the LD is coupled (near the end face of each core), the first core 101 includes the second core 103. That is, the second core 103 is formed so that the tapered portion 105 is formed after at least the slab region ends from the end surface, and the straight portion 104 of the second core 103 is always formed in the slab 102. Yes. The position where the slab 102 is provided is set so that at least the convex portion 106 can be formed. These cores are covered with an overclad (not shown).
In the configuration of FIG. 10B, a slab 107 having a finite length is provided in place of the slab 102 having an infinite length in FIG.

  Even in this case, the effect of the slab described above can be obtained. Ideally, when the second core 103 is positioned at the center of the first core 101, the waveguide can achieve complete single mode propagation, and therefore, it is considered that the occurrence of ripples that have been a problem does not occur. However, in reality, it is difficult to manufacture completely in the center due to manufacturing errors and the like, and the problem of generation of ripple that is a problem in this specification remains unchanged. Therefore, even in the case where the second core 103 is included in the first core 101, there is an effect of the structure in which the slab region of the present invention is provided. Further, when the second core 103 is provided in the first core 101, the length L of the slab can be shortened, and the circuit can be miniaturized.

  In addition, when using several waveguides provided with SSC based on this embodiment, you may make it provide a slab area | region separately in SSC with which each waveguide is provided. Moreover, you may make it share a slab between SSC with which each waveguide is provided. That is, a form in which the adjacent low Δ waveguide is formed on the end face of the slab region of a certain low Δ waveguide may be employed.

(Second Embodiment)
In this embodiment, an optical transceiver including the SSC described in the first embodiment will be described.
FIG. 11 is a diagram illustrating a configuration of the optical transceiver according to the present embodiment. FIG. 11B is a diagram illustrating an enlarged portion S of a connection portion between the SSC and the LD included in the waveguide in FIG.

  In FIG. 11A, the optical transceiver 110 includes a crossed waveguide 111 embedded in a clad composed of an over clad and an under clad. A groove is formed at the intersection 112 of the intersecting waveguide 111, and a thin film filter 113 for demultiplexing light is inserted into the groove. Each of the two ends of the waveguide 111 is described with reference to FIG. SSCs (indicated by reference numeral 114 in the present embodiment) are provided, and each of the SSCs 114 is provided with a semiconductor light receiving element and an LD. FIG. 11B shows how the SSC 114 and the LD 115 are coupled.

  The signal light incident from the input / output fiber passes through the thin film filter 113 and enters the light receiving element. Further, the light output from the LD 115 is coupled with the second core 83 which is a high Δ waveguide of the SSC 114 with high efficiency and is guided to the waveguide 111.

  FIG. 12 is a diagram illustrating transmission characteristics in the optical transceiver illustrated in FIG. As shown in FIG. 12, the ripple on the transmission characteristics is greatly reduced. Therefore, the transmission characteristics can be improved by using the SSC 114 according to the present embodiment. Further, since it is not necessary to lengthen the spot size transition portion (tapered portion 85) in order to suppress ripples, it is possible to reduce the size of the apparatus.

  FIG. 13 is a diagram showing the frequency of LD coupling efficiency in the optical transceiver shown in FIG. According to the present embodiment, as can be seen from FIG. 12, the occurrence of ripple can be suppressed. As a result, as can be seen from FIG. 13, the variation in coupling efficiency for each element can be reduced.

  In addition, although this embodiment demonstrated SSC which provided the 2nd core on the 1st core, SSC which provided the 2nd core in the 1st core may be sufficient.

(Third embodiment)
In the present embodiment, a configuration in which a multi-mode interference waveguide (MMI) is applied to the slab in the SSC described in the first embodiment will be described. The optimization of the finite slab width to shorten the slab length will also be described.

  As shown in the first and second embodiments, it is useful and highly effective to provide a slab region in the SSC for the purpose of LD coupling. However, in order to radiate the light remaining in the first core to such an extent that a sufficient effect can be obtained in the slab region, a sufficient length is required. With regard to the slab width, when a quartz optical waveguide is assumed, even in the case of a narrow-pitch optical fiber, there is a margin because the gap is as wide as 127 μm. When the slab width W = 127 um and Δ = 0.3%, if the slab length L is shorter than 1.6 mm, the light radiated in the slab region and reflected on the side surface of the slab is coupled to the opposing waveguide. Absent. However, the slab length is one factor that increases the circuit area. Therefore, the length of the slab region should be as short as possible.

  In the configuration of FIG. 8 of the first embodiment, the slab width W is long enough to be considered as infinite, and a sufficient width is provided so that the light reflected from the slab side face does not enter the opposing waveguide. However, if the reflection is actively used, the slab length can be shortened by making the width of a certain slab. Or if the length is the same, the ripple suppression effect can be further enhanced.

  In this embodiment, MMI is used as means for removing light remaining in the first core (low Δ waveguide) using a slab having a finite width. The MMI is a light multiplexing / demultiplexing circuit that uses a multimode waveguide capable of propagating higher-order modes and performs light multiplexing / demultiplexing using interference between modes in the waveguide. FIG. 14 shows the result of calculating the propagation state in the MMI by the beam propagation method (BPM) when Δ = 0.3%, the input waveguide width is 8.0 μm, and the slab width is 60 μm. FIG. From FIG. 14, it can be seen that the propagating light is branched into four, three and two. Usually, the MMI is used to produce a branch device by providing an output waveguide at a place where the intensity of the branched light is strong.

  It is confirmed that there is a region 141 in which the light intensity is weak between the two branches in the region 140 near the center of FIG. In the first embodiment, the light remaining in the first core (low Δ waveguide) is radiated and attenuated, but in this embodiment, the width of the slab region is finite, and the slab region of the first core is By operating as an MMI, interference is utilized and light in the first core is attenuated.

  Also, assuming the coupling with the LD, if the LD is coupled directly to the second core (high Δ waveguide) located on the end surface of the slab, the operation as an MMI becomes difficult. This is because even if the LDs are manufactured from the same substrate, there is a difference between solids, and it can be said that the control of the relative position between the LD and the second core is not uniquely determined due to the problem of mounting accuracy. Therefore, it is difficult to operate the MMI as designed. Therefore, it is necessary to insert a waveguide composed of the first and second cores at the entrance end even if it is short. That is, as shown in FIGS. 8A and 8B, the first core is provided with a protrusion (protrusion portion 87).

It is desirable that the length of the slab region is long as long as it is allowed by the circuit to be constructed. This is because the longer the interval between the two branches in the MMI, the larger the distance, and the lower the recombination rate in the output waveguide (opposing waveguide). Let L ₁ be the length allowed in the circuit configuration. When the length of the slab region allowed in the circuit is L ₁ = 500 μm, the refractive index n _L = 1.4491 of the first core constituting waveguide and the refractive index n of the underclad n = 1.444, and the wavelength of the incident light Assuming a case where λ _p = 1.31 μm is considered, it is confirmed whether there is an attenuation effect of light in the first core using BPM as described above.

For simplicity, the second core was ignored for the sake of simplicity, and the calculation was performed for the case where the fundamental mode input was given in the first core. For example, if the actual circuit has a refractive index n _L = 1.449 and a core shape of 8 × 8 um ² with a first core, and a waveguide mode analysis is performed by a computer, a second-order mode as shown in FIG. 25 is observed. The At this time, the intensity distribution in the first core is very similar to that of the fundamental mode when there is no second core above. For this reason, it can be said that it is sufficient to discuss the propagation of the light remaining in the first core in consideration of only the fundamental mode of the first core without considering the second core.

FIG. 16A is a diagram showing a result of evaluating recombination of residual light in the first core with respect to the width of the slab when the length of the slab is 500 μm. In FIG. 16A, the light intensity in the first core coupled to the output waveguide (opposing waveguide) is shown as 1 with the slab width as a function. According to this, it can be seen that there is a minimum value in a specific slab width W _s , in this case, around 30 μm. In the minimum state, the coupling efficiency of the residual light in the first core at the exit of the slab is 7.8 dB.

  FIG. 16B shows a calculation result when the slab width is 200 μm and the slab length L is a function. In this case, the slab width 200 um is a width that can be regarded as an infinite length. In FIG. 16B, it can be seen that in order to obtain the extinction of 7.8 dB corresponding to the minimum in FIG. 16A, the length of the slab is about 1.2 mm. In other words, rather than simply radiating to the slab and attenuating the light remaining in the first core, attenuating with the above method has the effect of shortening the slab length and reducing the size of the circuit. Is possible. Further, if a slab length of 1.2 mm is allowed in the circuit, extinction can be further obtained by using the same method. If the same calculation is performed at 1.2 mm, extinction of 18 dB or more is possible.

The slab width that takes such a minimum value can be obtained by the following calculation. If determining the length of the length L ₁ of the slab acceptable on the circuit configuration described above, then the points of N branches propagating light in FIG. 14, the slab region given by the shortest distance from the light incident side The width Ws is given by the following equation.
_{W s = W e - (λ} o / π) (n L / n c) 2σ (n L 2 -n c 2) - (0.5) ( Equation 1)
Here, W _e corresponds to the width of the fundamental mode in the slab, λ _o corresponds to the wavelength of light incident on the slab, and n _L and n _c are the refractions of the waveguides constituting the slab, respectively. Is the refractive index of the underclad of the slab. σ corresponds to the polarization of incident light, σ = 0 at TE, and σ = 1 at TM.

In addition, _{W e} are,
W _e = (λ _o · N · L / n _L ) (Formula 2)
Given in.

  As can be seen from FIG. 14, when N = 2, the region where the light is weakened between the light branch regions is the widest.

  As can be seen from the above, in this embodiment, in order to suppress the residual light propagating through the slab composed of the MMI from being incident on the output waveguide (opposite waveguide) of the first core, The counter waveguide is connected to a region where the light intensity is lower than that of the light branching region, preferably a region where the light intensity is weakest. That is, in FIG. 15, a protrusion 153 is formed by a waveguide at one end of the slab 152, and an opposing waveguide (output waveguide) 154 is connected to the other end of the slab 152. In this way, the first core 151 is formed. At this time, by using (Equation 1), it becomes possible to set the slab width according to the slab length L appropriate for circuit design. Note that an MMI waveguide may be used for the slab 152.

FIG. 17 is a bird's-eye view showing an example of the configuration of the SSC using the first core 151 described in FIG.
In FIG. 17, a second core 155 having a refractive index higher than that of the first core 151, which includes a linear portion 156 and a tapered portion 157, is formed on the first core 151 formed on the underclad (not shown). Yes. The tapered portion 157 serves as a spot size transition portion. As described above, in the present embodiment, the first core 151 is a low Δ waveguide and the second core 155 is a high Δ waveguide because of the refractive index relationship between the first core 151 and the second core 155. Become.

  In the present embodiment, as described above, the first core 151 includes the slab 152 configured by the MMI waveguide in a region where at least a part of the linear portion 156 of the second core 155 located in the upper layer is formed. In the portion where the LD is coupled (near the end face of each core), a straight portion 158 in which the widths of the first core 151 and the second core 155 are substantially equal is formed. That is, the second core 155 is formed so that the tapered portion 157 is formed on the opposing waveguide 154, and the linear portion 156 of the second core 155 is always formed on the slab 152. Further, the position where the slab 152 is provided is set so that at least the straight portion 158 can be formed. That is, the slab 152 is provided so that the protrusion 153 is formed. These cores are covered with an overclad (not shown).

  In the present embodiment, as in the first embodiment, the residual light in the slab 152 can be reduced from entering the counter waveguide 154, so that the generation of ripple can be suppressed, and the slab When the length is determined, the optimum width can be uniquely determined, and downsizing of the apparatus can be realized.

  In addition, although this embodiment demonstrated SSC which provided the 2nd core on the 1st core, SSC which provided the 2nd core in the 1st core may be sufficient.

(Fourth embodiment)
In the present embodiment, a method of removing residual light in the first core by using a bending waveguide and a slab waveguide together will be described.

18A and 18B are top views illustrating the configuration of an SSC having a bent waveguide according to this embodiment.
Since the first core (low Δ waveguide) has a lower refractive index than the second core (high Δ waveguide), it exhibits radiation against a sharp bend. As shown in FIG. 18A, the second core 182 has a bending radius that does not emit light and can be bent. Then, the light in the first core 181 is radiated into the clad without being bent, and as a result, the residual light in the first core 181 is attenuated. As shown in FIG. 18 (a), it is effective to produce only a bend in the first core and the second core, but the light in the first core 181 does not bend at all with respect to the bend, and the loss is large. However, some of the residual light propagates to the tapered portion 183, which is a transition portion.

  Accordingly, as shown in FIG. 18B, a slab 185 is provided on the first core 184 formed in the underclad (not shown), and the bending radius is such that bending loss does not occur in the second core 182 in the slab 185. Give the curve part at. These cores are covered with an overclad (not shown). At this time, the light remaining in the lower part, that is, the first core 184 radiates through the slab 185 at an angle φ. Therefore, bending is performed so that the tapered portion 183 that is the transition portion of the second core 182 that is the output portion can be disposed outside the radiation angle. At this time, the tapered portion 183 is not formed on the slab 185. By doing so, the residual light in the first core 184 can be efficiently removed, and the occurrence of ripples can be suppressed as a result.

  For example, even if there is a coupling portion between the slab 185 and the opposing waveguide 186 in the first core 184 within the radiation range of the light emitted at the radiation angle φ, the first core 184 propagates in the first core by adding an offset. The degree to which the light that has been recombined can be surely reduced as compared to the case of facing each other.

  18A and 18B show a form in which the second core is formed on the first core, but a form in which the second core is provided in the first core may be used.

  EXAMPLES Hereinafter, although this invention is demonstrated in detail based on an Example, this invention is not limited to the following Example, It cannot be overemphasized that it can change variously in the range which does not deviate from the summary.

(Example)
In this example, an arrayed waveguide grating having a structure as shown in FIG. 19A was manufactured. The first core 194, Δ = 0.3%, the core shape is 8 × 8um ^2, the second core 195 is delta = 1.5%, core shape is a 4 × 4 um ^2, those 100G, 16 channels Produced. The length of the tapered portion 196 of the second core 195 was 1 mm.

  In FIG. 19A, an AWG 191 includes an arrayed waveguide 192 and a waveguide 193 as an input / output port made up of a plurality of waveguides. These waveguides are embedded in a clad made of an over clad and an under clad. ing. Each waveguide of the waveguide 193 is provided with an SSC as shown in FIG. The SSC includes a second core 195 as a high Δ waveguide and a first core 194 having a slab region 197 as a low Δ waveguide, and the second core 195 is provided on the first core 194. Yes. The first core 194 and the optical fiber 198 are optically connected.

  This AWG 191 has a shape in which the second core 195 formed on the slab 197 of the first core 194 is bent as described with reference to FIG. The light incident from the single mode fiber is first coupled to the first core 194, which is a low Δ waveguide, and is inserted into the PLC. As a result of separately examining the connection loss at this time, the connection loss between the optical fiber and the low Δ waveguide formed of the first core 194 was 0.1 dB / location.

  The light propagated to the rectangular core of the first core 194 reaches the transition portion of the SSC composed of the tapered portion 196 of the second core 195, and gradually shifts to the second core 195 having a high refractive index, thereby converting the spot size. . Thereafter, the second core 195 reaches the curved portion constituting the AWG 191. At this time, the light in the first core 194 that could not be transferred into the second core 195 is radiated and attenuated to the slab region. As a result of separately examining in detail the conversion loss caused by spot conversion in the SSC part of AWG191, it was very low loss of 0.1 dB / location. Thereafter, the light that has been propagated through the AWG and dispersed is reached to the output waveguide, is spot-converted again, and is guided to the output fiber.

  At this time, the light emitted into the slab by the slab 197 included in the first core 194 on the input side is less than that because the spot size conversion loss is estimated to be 0.1 dB. When the light is radiated in the slab 197 and reaches the output side, it is considerably attenuated. As a result, in FIG. 6, the transmissivity between channels fluctuated due to the ripple caused by the SSC unit. However, in the AWG of this embodiment, the transmissivity between channels is changed as shown in FIG. The difference was 0.2 dB or less. This is a range of variation between channels of a normal AWG, and is not affected by the fact that the SSC is installed.

  In the present embodiment, the second core 195 is provided on the first core 194. However, the present invention is not limited to this, and the second core 195 is provided in the first core 194 as shown in FIGS. May be provided.

(Fifth embodiment)
Conventionally, as a technique for enhancing the performance of AWG (polarization independence, athermalization), a groove is formed in the array waveguide of the AWG, and a wave plate is inserted into the groove or filled with athermal resin as desired. Etc. are done. That is, the arrayed waveguide is cut at the groove, and the wave plate and the athermal groove are inserted into the cut region. Therefore, at this time, it is necessary to couple the cut waveguide through a region other than the arrayed waveguide, such as a space, a wave plate, or an athermal resin.

  However, if the arrayed waveguide is constituted by a high Δ waveguide, that is, a waveguide having a small mode field, the diffraction loss at the groove portion increases. This is because the diffraction loss increases as the waveguide mode field decreases, that is, as the waveguide Δ increases.

  For example, if Δ of a high Δ waveguide is 1.5%, a groove is formed in the arrayed waveguide composed of the high Δ waveguide, the waveguide is divided, and a wave plate is inserted into the groove, the loss is It will be about 1 dB. In the future, when the Δ of the waveguide is further increased, this excess loss will be further increased.

  Therefore, in the present embodiment, the SSC according to an embodiment of the present invention is used in a place where a space is skipped, that is, a coupling portion between a waveguide and a waveguide through a certain space. In this way, where the space is skipped, the diffraction loss can be greatly reduced by increasing the mode field using the low Δ waveguide.

  22 is a diagram showing an AWG according to this embodiment, FIG. 23 is a diagram showing an enlarged portion V of one waveguide of the array waveguide in FIG. 22, and FIG. 24 is an array waveguide in FIG. FIG.

  In FIG. 22, the AWG 200 includes an arrayed waveguide 201 embedded in a clad composed of an over clad and an under clad. As shown in FIGS. 23 and 24, the arrayed waveguide 201 is formed with a wave plate insertion groove 206, and the wave plate 207 is inserted into the wave plate insertion groove 206. At this time, the arrayed waveguide 201 is divided by the wave plate insertion groove 206. In the present embodiment, the SSC of the present invention is provided to face each of the end portions of the divided waveguide. That is, the SSC is provided so as to sandwich the wave plate 207.

  Specifically, as can be seen from FIGS. 23 and 24, the first core (low Δ waveguide) having the protruding portion 202 and the slab 203 at the end of the divided array waveguide on the wave plate insertion groove 206 side. ), A second core (high Δ waveguide) having a straight line portion 204 and a tapered portion 205 is formed. At this time, the tapered portion 205 is included in the protrusion 202, and the tapered portion 205 and the straight portion 204 are disposed in the first core so that the straight portion 204 is always included in the slab 203. Since the first core and the second core can have different circuit configurations, the second core constitutes an actual array waveguide.

  By adopting such a configuration, the light passing through the arrayed waveguide (second core) is converted from a small mode field to a large mode field by moving from the tapered portion 205 to the protruding portion 202, and the wave plate 207. And enters the opposite projecting portion 202. The light incident on the protrusion 202 is incident on the second core via the tapered portion 205.

  Therefore, the light incident on the AWG passes through the high Δ waveguide in the arrayed waveguide, and when passing through the space, is performed by a waveguide having a large mode field (low Δ waveguide). It becomes possible to reduce the diffraction loss in the insertion groove 205.

  22 to 24, if Δ of the first core is 0.3%, the diffraction loss when the wave plate 206 is inserted into the arrayed waveguide can be reduced to 0.1 dB or less.

  In addition, although the form which embedded the 2nd core in the 1st core was demonstrated in this embodiment, the form which provides a 2nd core on a 1st core may be sufficient.

(A)-(c) is the figure which showed the structure of the conventional spot size converter. (A) is a figure which shows the structure of the conventional optical transceiver using a planar optical waveguide circuit and a thin film filter, (b) is an expansion of the connection part of SSC and LD which a waveguide has in (a). It is a figure which shows the part Q. It is a figure which shows the transmission characteristic in the optical waveguide circuit containing SSC of the structure where the different type | formula (DELTA) waveguide was laminated | stacked conventionally. It is a figure which shows the frequency of coupling efficiency of LD at the time of producing using the planar optical waveguide carrying the conventional SSC. (A) is a figure which shows the structure of the conventional array diffraction grating (AWG) mounted with SSC, (b) is the enlarged part R of the connection part of SSC and optical fiber which a waveguide has in (a). FIG. It is a figure which shows the transmission characteristic of each channel of the conventional AWG. It is a figure which shows the structure of the conventional AWG without SSC and AWG mounted with SSC. (A) is a bird's-eye view of the SSC configuration of the optical circuit according to one embodiment of the present invention, and (b) is a top view of the SSC configuration of (a). 1 is a top view of an SSC configuration, according to one embodiment of the present invention. FIG. (A) And (b) is a bird's-eye view of SSC which provided the high (DELTA) waveguide inside the low (DELTA) waveguide based on one Embodiment of this invention. (A) is a figure which shows the structure of the optical transceiver using the planar optical waveguide circuit and thin film filter based on one Embodiment of this invention, (b) is SSC which the waveguide in (a) has, It is a figure which shows the enlarged part S of the connection part with LD. It is a figure which shows the permeation | transmission characteristic in the optical waveguide circuit containing SSC of the structure which laminated | stacked the different type | formula (DELTA) waveguide based on one Embodiment of this invention. It is a figure which shows the frequency of coupling efficiency of LD at the time of producing using the planar optical waveguide which mounts SSC based on one Embodiment of this invention. It is a figure which shows the result of having calculated the mode of propagation in MMI by the beam propagation method based on one Embodiment of this invention. It is a figure which shows the mode of propagation at the time of implementing optimization of slab width based on one Embodiment of this invention. (A) And (b) is a figure which shows the result of having evaluated the recombination of the residual light in a 1st core with respect to the width | variety and length of a slab based on one Embodiment of this invention. 1 is a bird's-eye view of an SSC configuration according to an embodiment of the present invention. (A) And (b) is a top view explaining the structure of SSC which has a bending waveguide based on one Embodiment of this invention. (A) is a figure which shows the structure of the array diffraction grating (AWG) mounted in SSC based on one Embodiment of this invention, (b) is SSC and optical fiber with which the waveguide in (a) has It is a figure which shows the enlarged part T of a connection part. It is a figure which shows the transmission characteristic of each channel of AWG based on one Embodiment of this invention. (A) is a figure which shows the structure of the array diffraction grating (AWG) mounted in SSC based on one Embodiment of this invention, (b) is SSC and optical fiber with which the waveguide in (a) has It is a figure which shows the enlarged part U of a connection part. It is a figure which shows AWG which concerns on one Embodiment of this invention. It is a figure which shows the enlarged part V of one waveguide of the arrayed waveguide in FIG. FIG. 24 is a top view of the arrayed waveguide in FIG. 23. It is a figure which shows the result of the mode analysis of the waveguide based on one Embodiment of this invention.

Explanation of symbols

81, 101 First core 82, 88, 102, 107 Slab 83, 103 Second core 84, 104 Linear portion 85, 105 Tapered portion 86 Under cladding 87 Protruding portion 89 Opposite waveguide 106 Convex portion

Claims (10)

A first waveguide formed on the underclad and having a higher refractive index than that of the underclad, the second waveguide, and a slab region provided at one end of the second waveguide; A first waveguide having a third waveguide provided on the side of the slab region facing the side on which the second waveguide is provided;
A fourth waveguide formed on the first waveguide and a tapered waveguide, and a fifth waveguide having a higher refractive index than the first waveguide;
Either the fourth waveguide or the tapered waveguide is formed on the second waveguide, and the fourth waveguide or the tapered waveguide is formed on the third waveguide. An optical spot size converter, wherein the other one of the tapered waveguides is formed, and the fourth waveguide is formed at least on the slab region. A first waveguide formed on the underclad and having a higher refractive index than that of the underclad, the second waveguide, and a slab region provided at one end of the second waveguide; A first waveguide having a third waveguide provided on the side of the slab region facing the side on which the second waveguide is provided;
A fourth waveguide and a tapered waveguide embedded in the first waveguide, the second waveguide having a higher refractive index than the first waveguide; Prepared,
Either the fourth waveguide or the tapered waveguide is embedded in the second waveguide, and the fourth waveguide has the fourth waveguide. A spot size characterized in that the other of the waveguide or the tapered waveguide is embedded and formed, and at least in the slab region, the fourth waveguide is embedded. converter.   The second waveguide and the third waveguide are provided in the slab region so that the longitudinal direction of the second waveguide substantially coincides with the longitudinal direction of the third waveguide. The spot size converter according to claim 1 or 2, wherein the spot size converter is provided. The fifth waveguide is a bending waveguide in which a part of the fourth waveguide is bent at a bending radius such that bending loss does not occur.
3. The spot size converter according to claim 1, wherein the third waveguide is provided at a position different from the position facing the second waveguide in the slab region. 4.   When the fourth waveguide is formed on the second waveguide, the width of the second waveguide is substantially equal to the width of the fourth waveguide. The spot size converter according to any one of claims 1 to 4.   The angle at which the light incident from the second waveguide spreads in the slab region is φ, the length of the slab region is L, and the width of the slab region is W, so that W> Ltanφ is satisfied. 6. The spot size converter according to claim 1, wherein a width W of the slab area is set. The slab region is a multimode interference waveguide;
Of the second waveguide and the third waveguide, the waveguide in which the tapered waveguide is formed is the waveguide from which the tapered waveguide is not formed. 6. The spot size converter according to claim 1, wherein the spot size converter is provided in a region where light intensity is weaker than a light branch region of light incident on the slab region.   8. The spot size converter according to claim 7, wherein the light intensity region is a region between adjacent light branch regions of the light branch region branched into N (N is an integer of 2 or more). .   9. The spot size converter according to claim 7, wherein the region having the low light intensity is a region having the weakest light intensity as compared with the light intensity of the light branching region. The refractive index of the under-cladding and n _L, and the refractive index of the slab region and n _c, the length of the slab region is L, the wavelength of the incident light to the slab region and lambda _o, of the incident light It corresponds to the polarization, and when the polarization is TE, it is 0 when the polarization is TM, and when the polarization is TM, σ is defined as σ, and the width of the slab region with respect to the fundamental mode in the slab region is W _e. Then, W _e = (λ _o · N · L / n _L ) (1), and from the boundary between the waveguide in which the tapered waveguide is not formed and the slab region, The width W _{s of the} slab region when the distance to the branched optical branch region is the shortest is W _s = W _e − (λ _o / π) (n _L / n _c ) ^2σ (n _L ² −n _c ^{2) - (0.5)} (2) and is expressed, the width of the slab region, the slab territory Depending on the length and the desired N of the above formula (1) and (2) the spot size converter of claim 8 or 9, wherein it is set so as to satisfy.