JP7008316B2 - Optical connection structure - Google Patents

Description

The present invention relates to an optical connection structure, and more specifically, to an optical connection structure capable of realizing low-loss crossing between a plurality of optical waveguides in an optical circuit such as an optical switch or an optical integrated circuit.

Conventionally, in optical circuits such as optical switches and optical integrated circuits (hereinafter referred to as "optical circuits"), in-plane intersections in which a plurality of optical waveguides on the same plane intersect have been used. The larger the optical circuit, the more the number of in-plane intersections tends to increase. As the number of in-plane intersections increases, the influence of insertion loss and crosstalk generated in in-plane intersections on the optical propagation performance in optical circuits and the like increases. Therefore, there is a demand for reducing insertion loss and crosstalk at in-plane intersections. As a method of responding to this demand, there is a method of forming an optical circuit or the like into two layers and using a grade separation of an optical waveguide.

Patent Document 1 couples an optical signal from a first waveguide to bridge waveguides in different planes through directional coupling in order to reduce optical loss and crosstalk in the cross waveguide, from the bridge waveguide. Disclosed is a waveguide crossover that directionally couples to a second waveguide. Further, FIG. 6 of Patent Document 2 has Si cores 3 and 30 having separated tapered regions 30A above the grade separation core 300 (for example, Si core), and tapered regions 5A and 5C at both ends of the upper layer thereof. A three-dimensional optical waveguide in which a SiN core 5 is provided and light is sterically propagated between the Si core 3 and the SiN core 5 and the Si core 30 is disclosed. Further, Non-Patent Document 1 reduces optical loss and crosstalk at a waveguide intersection by optical coupling between a Si waveguide having a taper and a SiN waveguide having a taper on the upper layer at the waveguide intersection on the SOI substrate. Disclose what you want to do.

Special Table 2006-525556 Japanese Unexamined Patent Publication No. 2014-157210

WD Sacher, Y. Huang, GQ Lo, and JKS Poon, “Multilayer silicon nitride-on-silicon integrated photonic platforms and devices,” J. Light. Technol., Vol. 33, no. 4, pp. 901-910, 2015

The waveguide crossover of Patent Document 1 does not disclose an optical waveguide structure using a Si semiconductor process (for example, a CMOS process). Further, Patent Document 2 does not disclose the design method (specification) of each core constituting the three-dimensional optical waveguide. Further, in Non-Patent Document 1, crosstalk cannot be sufficiently suppressed because the distance between the Si waveguide and the SiN waveguide is smaller than the mode field diameter (<100 nm).

INDUSTRIAL APPLICABILITY The present invention can greatly reduce the insertion loss at the three-dimensional intersection of the optical waveguide in a two-layered optical circuit or the like, and can be manufactured by using a Si semiconductor process (for example, a CMOS process), and the manufacturing variation thereof. It is an object of the present invention to provide a two-layer optical connection structure capable of improving tolerance to (size fluctuation of an optical waveguide).

The optical connection structure in the region where the optical waveguides of one aspect of the present invention intersect includes a first optical waveguide and a second optical waveguide that intersects the first optical waveguide. The second optical waveguide is between the first and second regions that are in the same plane as the first optical waveguide and separated in the horizontal direction, and the first region and the second region that are in a plane different from the first optical waveguide and are separated in the horizontal direction. Includes a third region in which it is located. The end of the first region and one end of the third region, and the end of the second region and the other end of the third region have overlapping regions, respectively, and the first region, the second region, and the first region are used. The three regions have a tapered region leading to the overlap region. Each of the overlapping regions has a length approximately equal to the perfect bond length in the optical transition from the first region to the third region or from the third region to the second region.

The optical connection structure in the region where the optical waveguides of another aspect of the present invention intersect includes a first optical waveguide and a second optical waveguide intersecting the first optical waveguide. The second optical waveguide is a third region located between the first region and the second region on the same plane as the first optical waveguide and separated from each other, and the first region and the second region on a plane different from the first optical waveguide. Includes areas. The ends of the first region and the second region have a first tapered region and a constant width region, and both ends of the third region have a second tapered region and a third tapered region, and a constant width region. And the third tapered region form an overlapping region and have substantially the same length.

The optical connection structure in the region where the optical waveguides of another aspect of the present invention intersect includes a first optical waveguide and a second optical waveguide that intersects the first optical waveguide. The second optical waveguide is a third region located between the first region and the second region on the same plane as the first optical waveguide and separated from each other, and the first region and the second region on a plane different from the first optical waveguide. Includes areas. The end of the first region and one end of the third region, and the end of the second region and the other end of the third region each have an overlap region. In the overlap region, the first region and the second region of the second optical waveguide have a first tapered region, a constant width region, and a second tapered region at the tip, and both ends of the third region are second. It has a 3 taper region, a 4th taper region, and a constant width region or a 5th taper region at the tip.

According to the present invention, in a region where two optical waveguides intersect, one optical waveguide is formed into a bridge structure (overpass) while utilizing a tapered structure, and light is spatially bypassed and propagated to perform optical connection. It is possible to reduce the optical loss and crosstalk at the optical waveguide intersection while shortening the length required for the optical waveguide.

It is (a) top view and (b) side view which show the whole structure of the optical connection structure of one Embodiment of this invention. It is (a) top view and (b) side view which show the structure of the optical connection structure of one Embodiment of this invention. It is a figure which shows the relationship of each parameter in the optical connection structure of one Embodiment of this invention. It is a figure which shows the tolerance to the variation of the width of the SiN optical waveguide in the optical connection structure of one Embodiment of this invention. It is (a) top view and (b) side view which show the structure of the optical connection structure of another embodiment of this invention. It is (a) top view and (b) side view which show the structure of the optical connection structure of another embodiment of this invention. It is a figure which shows the relationship of each parameter in the optical connection structure of another embodiment of this invention. It is a figure which shows the tolerance to the variation of the width of the SiN optical waveguide in the optical connection structure of one Embodiment of this invention. It is (a) top view and (b) side view which show the whole structure of the optical connection structure of another embodiment of this invention.

An embodiment of the present invention will be described with reference to the drawings. The optical connection structure described below is described by taking the case where one set (two) of optical waveguides intersects as an example, but the number of intersecting optical waveguides is not limited to one set (two), and a plurality of optical waveguides intersect. The optical connection structure of the present invention can be applied to the optical waveguide intersection of the present invention. For example, as will be described later with reference to FIG. 9, one optical waveguide may intersect a plurality of optical waveguides in different layers, more broadly, an optical circuit or the like. Further, in the following description, only the intersecting optical waveguides, that is, mainly the core silicon (Si) and silicon nitride (SiN) are described, but in reality, the optical waveguide (core) is surrounded by the core. There is also a clad layer with a low refractive index (n) (eg silicon oxide), which is configured to propagate light primarily within the core. Further, the optical connection structure of the present invention is used as a part of an optical circuit such as an optical switch and an optical integrated circuit, but the description thereof is omitted because other circuit configurations such as the optical switch are the same as the conventional configurations. is doing.

FIG. 1 is a top view (a) and a side view (b) showing the overall configuration of an optical connection structure according to an embodiment of the present invention. In the top view of FIG. 1A, the first optical waveguide 1 running in the horizontal direction and the second optical waveguide 2 running in the vertical direction intersect at the intersection 3. The first optical waveguide 1 is made of, for example, silicon (Si). The second optical waveguide 2 is composed of three regions, that is, a separated first region 2A and a second region 2B, and a third region 2C between them. The first region 2A and the second region 2B of the second optical waveguide 2 are on the same plane as the plane in which the first optical waveguide is located. As shown in the side view of FIG. 1 (b), the third region 2C of the second optical waveguide 2 is a plane having the first region 2A and the second region 2B of the first optical waveguide 1 and the second optical waveguide 2. It is on a plane (layer) above the distance D from the layer). That is, the third region 2C forms a bridge structure and forms a grade separation with the first optical waveguide 1 at the intersection 3. The first region 2A and the second region 2B of the second optical waveguide 2 are made of, for example, silicon (Si), and the third region 2C is made of silicon nitride (SiN).

The end of the first region 2A and one end of the third region 2C of the second optical waveguide 2 have a region 4A that overlaps in the vertical direction, and similarly, the end of the second region 2B and the third region 2C. The other end of the waveguide has a vertically overlapping region 4B. In the overlap region 4A, one end of the first region 2A and one end of the third region 2C of the second optical waveguide 2 has a tapered tapered shape (region). Similarly, in the overlap region 4B, the end of the second region 2B and the other end of the third region 2C have a tapered tapered shape (region). In the two regions 4A and 4B, the first region 2A and the third region 2C, the third region 2C and the second region 2B function as so-called directional couplers, as shown by the arrow 6 in FIG. 1 (b). , Light transitions from the first region 2A to the upper third region 2C, and further transitions from the third region 2C to the lower second region 2B. As a result, light propagates on the first optical waveguide 1 through the grade separation of the second optical waveguide, so that reflection and scattering of light in the intersection region of the first and second optical waveguides are suppressed, and insertion loss and insertion loss occur. It is possible to significantly suppress the occurrence of crosstalk.

Here, the distance D, which is one of the parameters of the optical connection structure of the embodiment of the present invention of FIG. 1, that is, the plane (layer) having the first region 2A and the second region 2B of the second optical waveguide 2 and the first A method of determining the distance (hereinafter, referred to as “interlayer distance”) D from a plane (layer) having the three regions 2C will be described. In the following description, the case where the first region 2A and the second region 2B of the second optical waveguide 2 are made of Si and the third region 2C is made of SiN, that is, the case of a Si / SiN waveguide will be described as an example. Generally, the photoelectric field propagating through the optical waveguide has a spatial expanse. Interactions between Si / SiN waveguides occur where the hem of the spatially spread photoelectric field overlaps the other waveguide. As a result, scattering loss occurs at grade separation where each waveguide is orthogonal, and light energy transfer occurs at a coupling region where each waveguide is parallel (such as a directional coupler).

The optical connection structure of the present invention makes good use of the tapered structure, reduces the interaction at the overpass, and increases the interaction at the coupling region, so that the loss and crosstalk at the intersection are reduced to a negligible level. It is possible to make a short optical connection while suppressing it. To achieve this goal, it is important to set an appropriate Si / SiN interlayer distance D, which can vary depending on the width or thickness of the Si / SiN waveguide. Specifically, considering the photoelectric field distribution of the SiN waveguide in the basic width and thickness, when the electric field strength at the center of the SiN waveguide is 1, the location is separated by a distance where the electric field strength is about 1% or less. The center of the Si waveguide should be present (although the electric field should be the polarization directional electric field of light). At this time, the loss at the grade separation can be suppressed to a negligible extent. As an example, when the width and thickness of the SiN waveguide are 1 μm and 400 nm, respectively, and the width and thickness of the Si waveguide are 430 nm and 220 nm, respectively, the electric field strength of the Si waveguide portion in TM (Transverse Magnetic) -like mode is The interlayer distance D, which is about 1%, is 1.5 μm. At this time, it has been obtained by the FDTD calculation by the present inventors that a very small loss of 0.0005 dB can be realized at the intersection. This loss of 0.0005 dB means that, for example, in a large-scale and small silicon photonics optical switch having a 32 × 32 input / output port, the number of optical waveguide intersections is close to 1000 in one form of a practical switch circuit. Therefore, it is a loss level required because it is necessary to reduce the loss between optical waveguides to a negligible level (0.0005 dB).

The minimum value (1.5 μm) of the interlayer distance D is determined by the above criteria, and the maximum value is specified by the manufacturing process in the CMOS manufacturing plant. As an example, the film thickness of the clad layer (for example, SiO ₂ ) constituting the interlayer distance D which can be produced by the CVD process and the subsequent process can be carried out is about 2 μm, and the upper limit is defined by this value. Further, when the interlayer distance D becomes larger than that, it is necessary to further reduce the SiN waveguide width in order to sufficiently shorten the length of the Si / SiN waveguide (for example, <300 μm). However, since the minimum dimension of the SiN waveguide width in general KrF lithography is about 200 nm, even if the above-mentioned film thickness problem can be solved, the problem (restriction) of this minimum dimension arises as the next problem. It ends up. From the above studies, it is desirable that the interlayer distance D in the optical connection structure of one embodiment of the present invention is in the range of about 1.5 to 2.0 μm.

The minimum value (1.5 μm) of the interlayer distance D is obtained when the width and thickness of the SiN waveguide are standard sizes (1 μm, 400 nm), respectively, and when this standard size changes (increases or decreases). In the same FDTD calculation, the minimum value of the interlayer distance D at which the loss at the intersection is 0.0005 dB can be obtained. Further, the upper limit value (2.0 μm) of the interlayer distance D is obtained from the fact that the minimum dimension of the SiN waveguide width in the current general KrF lithography is about 200 nm, and it is further improved by the progress of lithography technology. If a short SiN waveguide width can be obtained stably and easily, the interlayer distance D can be set to a value larger than 2.0 μm.

When the interlayer distance D becomes as large as about 1.5 to 2.0 μm, the interaction between the layers becomes smaller. Therefore, in the configuration including the taper length similar to the conventional technique, the light that can reach the SiN waveguide or the Si waveguide is very small. It will be slight and the efficiency will be very poor. In the tapered structure, efficiency can be improved by increasing the taper length, but if the same structure as before is used, a taper length of 2.7 cm is required, and a large-scale and small optical switch, etc. It is not realistic as an optical connection structure used for. In order to realize a highly efficient optical connection while suppressing the length of the optical connection structure to about 200 to 300 μm, it is necessary to devise a method for strengthening the interaction even if the interlayer distance D is large. Therefore, in the present invention, the design of the optical connection structure as described below and each parameter constituting the optical connection structure is adopted.

A method (design method) for determining other parameters of the optical connection structure according to the embodiment of the present invention in FIG. 1 will be described. In the following description, the case where the first region 2A and the second region 2B of the second optical waveguide 2 are made of Si and the third region 2C is made of SiN, that is, the case of a Si / SiN waveguide will be described as an example. In the overlap regions 4A and 4B of FIG. 1 (b), that is, in the optical transition between the optical waveguides in the optical coupler, when the effective refractive indexes n _Si and n _Si N of the two waveguides are equal, the light is transmitted from the Si waveguide. It is possible to completely move to the SiN waveguide and vice versa from the SiN waveguide to the Si waveguide. The length of the optical waveguide required for the optical transition at this time is called the perfect bond length. That is, when the length of the bond portion is shorter than the perfect bond length, or when the effective refractive index of the Si / SiN waveguide is different, the optical transition efficiency is lowered. Therefore, in order to realize an interlayer optical connection using a directional coupler, it is necessary to match the effective refractive indexes of the Si waveguide and the SiN waveguide in the optical coupling region.

FIG. 2 shows a configuration example (direction) of one end of the first region 2A (Si) and one end of the third region 2C (SiN) of the second optical waveguide 2 of the optical connection structure according to the embodiment of the present invention of FIG. It is (a) top view and (b) side view which shows the example using a coupler). The same applies to the configuration of the end of the second region 2B (Si) and the other end of the third region 2C (SiN) of the second optical waveguide 2, which is the other end of the bridge. In FIG. 2, the whole is divided into three regions A, C, and B, and in the tapered regions A and B, the effective refractive indexes of the Si / SiN waveguides are matched in the overlap region C, so that each waveguide is matched. The width is gradually narrowed. The length of the overlap region _C is equal to the perfect bond length LC between Si / SiN waveguides with the narrowed minimum width. If the lengths LA and LB of the tapered regions _A and _B are short, the refractive index changes sharply for light and reflection and scattering occur. Therefore, the lengths _LA and _LB need to be, for example, about 100 μm. The width W ₁ of the first region 2A (Si) is, for example, 430 nm, and the thickness d1 thereof is, for example, 220 nm. The width W ₂ of the third region 2C (SiN) is, for example, 1 μm, and the thickness d2 thereof is, for example, 400 nm. The values of these parameters (length _LA , _LB , width W ₁ , W ₂ , thickness d1, d2) are just examples, and other values should be appropriately selected and adopted for each parameter. Can be done.

The width W Si of the first region 2A (Si), the width W _Si N of the third region 2C ( _SiN ), and the perfect bond length L _C in the overlap region C of FIG. 2 can be designed as follows, for example. can. The C-band wavelength band (1530 nm-1565 nm) is one of the most commonly used wavelength bands in optical information communication because it is characterized by a small propagation loss in an optical fiber. Here, the parameter design method in the C band wavelength band will be described. FIG. 3 shows the results of calculating the relationship between each parameter when the center wavelength is 1.55 μm by FEM (Finite Element Method). FIG. 3A shows the relationship between the width W _SiN of the third region 2C (SiN) and the width W _Si of the first region 2A (Si), and FIG. 3B shows the width W _SiN of the third region 2C (SiN). The relationship between and the perfect bond length L _C is shown.

The dotted line in FIG. 3 is the result of linear fitting of each point, and it can be seen that linear approximation can be performed in a narrow range. From this figure, if the width W _SiN of the SiN waveguide is determined, the Si waveguide width W _Si and the perfect bond length _LC can be uniquely determined. More specifically, the Si waveguide width W _Si and the perfect bond length _LC can be obtained from the following equations (1) and (2) representing the fitting curve. The dotted line indicated by the lower limit in FIG. 3 exemplifies the case where the lower limit of the SiN waveguide width is, for example, 350 nm due to the limitation of the manufacturing process, and the lower limit of the perfect bond length L _C in that case is 58 μm. It turns out that
W _Si = 0.307 x W _SiN + 50.3 nm (1)
LC = 118 _x 0.001 x W _SiN + 15.5 μm (2)

Next, with respect to manufacturing variations (size fluctuation of the optical waveguide) when the optical connection structure of the embodiment of the present invention described with reference to FIGS. 1 to 3 is manufactured by using a Si semiconductor process (for example, a CMOS process). Consider tolerance. In the Si / SiN waveguide, if the waveguide width fluctuates due to a manufacturing error, the effective refractive indexes _nSi and _nSiN also change, and the optical transition efficiency decreases. Therefore, it is necessary to consider increasing this tolerance also when designing / using the optical connection structure of one embodiment of the present invention. Now, when the TE (Transverse Electric) -like mode and the TM _- like mode are designed so that the perfect bond length LC is about the same (for example, 70 μm), FIGS. 3, equations (1) and (2) are used. ), Each parameter is as shown in the table below.

FIG. 4 shows the results of calculating how the insertion loss when using each design parameter in the above table changes with respect to the width of the SiN waveguide. In FIG. 4, in the TE-like mode, the insertion loss increases by 1 dB when the width of the SiN waveguide changes by ± 15 nm, whereas in the TM-like mode, this width is ± 26 nm. Therefore, it is expected that the tolerance to the width change of the SiN waveguide will increase by 79% by using the TM-like mode. Although it is assumed here that the width of the SiN waveguide changes, the same tendency can be obtained even if the width of the Si waveguide changes. Therefore, it is desirable to use the TM-like mode in order to increase the tolerance to the fluctuation of the waveguide width. In that case, for example, the above-mentioned design values (complete bond length LC is 70 μm, taper regions A and _B length L) are used. It can be seen that a compact optical connection structure with a total length of about 270 μm can be realized by adopting 100 μm each for _A and _LB ).

Next, the optical connection structure of another embodiment of the present invention will be described with reference to FIGS. 5 to 9. FIG. 5 shows a configuration example of the end of the first region 2A (Si) and one end of the third region 2C (SiN) of the other optical waveguide 2 of the optical connection structure of the embodiment of the present invention of FIG. It is (a) top view and (b) side view which shows the example) which used the heat insulation taper. The same applies to the configuration of the end of the second region 2B (Si) of the second optical waveguide 2 and the other end of the third region 2C (SiN). In FIG. 5, the whole is divided into three regions (i), (ii), and (iii). The region (i) is a tapered portion for narrowing the width of the Si waveguide (first region 2A) having a normal width W ₁ and thickness d 1 to W _Si and expanding the mode size. In region (ii), Si / SiN waveguides are arranged one above the other, the width of the Si waveguide is constant within the region, and the width of the SiN waveguide (third region 2C) is from W _SiN1 to W _SiN2. By adopting a tapered structure that gently spreads up to, light can be transferred between the Si waveguide and the SiN waveguide. Region (iii) is a tapered portion that extends the width of the SiN waveguide from W _SiN 2 to the normal width W ₂ . The reason why the length portion _LB of the region (iii) is shorter than the length _LA of the region (i) is that the SiN waveguide has a smaller effective refractive index, so even if the width is changed more rapidly, the light is emitted. This is because the change in the refractive index felt by is suppressed to a small extent.

FIG. 6 shows a configuration example of the end of the first region 2A (Si) and one end of the third region 2C (SiN) of the other optical waveguide 2 of the optical connection structure of the embodiment of the present invention of FIG. It is (a) top view and (b) side view which shows the example) which used the heat insulation taper. The same applies to the configuration of the end of the second region 2B (Si) of the second optical waveguide 2 and the other end of the third region 2C (SiN). In FIG. 6, as in the configuration example of FIG. 5, the whole is divided into three regions (i), (ii), and (iii). Compared with the configuration example of FIG. 5, the Si waveguide and the SiN waveguide overlap each other in the regions (i) and (iii). That is, the SiN waveguide has a tip taper structure that extends to region (i), the tip of which has a width W _SiN0 . Similarly, the Si waveguide has a tip taper structure that extends to region (iii), the tip of which has a width W _{Si 0} .

In the configuration example of FIG. 6, the width of the Si waveguide is narrowed to W _Si in the region (i) so that the light distribution spreads toward the clad side. As a result, the interaction between Si / SiN waveguides becomes stronger, and higher efficiency can be obtained even with a shorter taper. Next, in the region (ii), the width W _Si of the Si waveguide is fixed, and only the width of the SiN waveguide is gradually changed from W _SiN1 to W _SiN2 , so that the light is transferred to the SiN waveguide. Let me. Finally, in region (iii), the width of the SiN waveguide is widened from W _SiN2 to the basic width W2. The reason why the length portion _LB of the region (iii) is shorter than the length _LA of the region (i) is that the SiN waveguide has a smaller effective refractive index, as in the configuration of FIG. This is because even if the width is changed, the change in the refractive index felt by light can be suppressed to a small value.

Here, in order to explain the merit that the tip taper structure exists in the regions (i) and (iii), for example, consider the case where light enters from the left side of FIG. At this time, in the configuration of FIG. 5, since the tip of the SiN waveguide appears at the place where the Si waveguide becomes thin and the mode is widened, light scattering loss due to the tip occurs. On the other hand, in the configuration of FIG. 6, since the tip of the SiN waveguide appears in a place where the Si waveguide is thick (light is localized in Si), light scattering by the tip is suppressed. The tip taper structure of the Si waveguide has the same merit. Based on the calculation by the FDTD method, if the width W _SiN0 of the tip taper of the SiN waveguide is 350 nm and the width W _Si0 of the tip taper of the Si waveguide is 100 nm in the configuration of FIG. 6, the Si / SiN interlayer distance D is It has been confirmed that the scattering loss at 1.5 μm is 0.009 dB and 0.006 dB, respectively, which are negligibly small.

In the configurations of FIGS. 5 and 6, the length of the region (i) is based on the design guideline that the spatial change rate of the effective refractive index is kept constant in order to shorten the length of the structure while suppressing the light scattering loss as much as possible. _{LA can be, for example, 100 μm, and the length L B of} region (iii) can be, for example, 20 μm. It is known that the tapered portion of these regions has a role of expanding the mode of light toward the clad side as a pre-stage of optical connection, but if the change in the refractive index of this portion is too rapid, scattering loss occurs. By adopting these values for the lengths _LA and _LB , the spatial change of the effective refractive index in these regions can be set to about 0.01 (μm) ^-1 . The length _LB of the region (iii) is as short as 20 μm because the change in the effective refractive index with respect to the change in the waveguide width is smaller in the SiN waveguide.

The width W ₁ of the Si waveguide is, for example, 430 nm, and the thickness d 1 thereof is, for example, 220 nm. The width W ₂ of the SiN waveguide is, for example, 1 μm, and its thickness d 2 is, for example, 400 nm. The values of these parameters (W ₁ , W ₂ , d1, d2) are merely examples, and other values can be appropriately selected and adopted for each parameter. As already described above, the interlayer distance D in FIG. 5B is preferably in the range of about 1.5 to 2.0 μm due to the reduction of the light loss level and the limitation of the manufacturing process.

The width W _Si of the Si waveguide in FIGS. 5 and 6, the width W _SiN1 , W _SiN2 of the SiN waveguide, and the length L _taper of the tapered portion of the region (ii) can be designed, for example, as follows. .. Here, the parameter design method in the C band wavelength band will be described. Figure 7 shows the results of FEM calculation of the relationship between each parameter when the center wavelength is 1.55 μm. FIG. 7A shows the relationship between the width W _Si of the Si waveguide and the widths W _SiN1 and W _SiN2 of the SiN waveguide, and FIG. 7B shows the relationship between the width W _Si of the Si waveguide and the length L _taper of the tapered portion. Shows the relationship. The dotted line in FIG. 7 is a curve obtained by fitting each point. From FIG. 7, when the width W _Si of the Si waveguide is determined, the SiN waveguide widths W _SiN1 , W _SiN2 and the length L _taper of the tapered portion can be uniquely determined.

More specifically, the SiN waveguide widths W _SiN1 , W _SiN2 and the length L _taper of the tapered portion can be obtained from the following equations (3) to (5) representing the fitting curve. The horizontal dotted line in FIG. 7A exemplifies a case where the lower limit of the SiN waveguide width is, for example, 350 nm due to restrictions on the manufacturing process, and the Si waveguide width W _Si is 210 nm and the SiN waveguide width is 210 nm. It can be seen that the minimum device length is when the width W _SiN1 is 350 nm (the same applies to W _SiN0 in FIG. 6), the W _SiN2 is 510 nm, and the taper length L _taper is 80 μm.
W _SiN1 = 0.0429 x W _Si ² -12.5 x W _Si + 1090 nm (3)
W _SiN2 = 0.0429 × W _Si ² -12.5 × W _Si +1250 nm (4)
L _taper = 1560 x 0.001 x W _Si -248 μm (5)

Here, in the configuration of FIG. 6, the tolerance for manufacturing variation (size variation of the optical waveguide) will be examined. The result of calculating the loss when the width of the SiN waveguide changes using the mode coupling theory is shown as the dotted line of the plurality of tapers in FIG. However, in the calculation, the upper and lower limits of the SiN waveguide width are set to 1 μm and 350 nm, and the taper length L _taper is set to 280 μm. Further, as a reference, the tolerance when a directional coupler having the same configuration as that of the embodiment of FIG. 2 is designed using the TE-like mode is also shown as a dotted line in FIG. 8 as a conventional example. Comparing these results, when the multi-step taper structure of the present invention is used, a tolerance improvement of about 4.7 times is obtained as compared with the conventional directional coupler, and the width in which the efficiency is reduced by 1 dB is ± 68 nm. You can see that.

FIG. 9 shows a configuration example in which the above-mentioned optical connection structure of the present invention is applied to a plurality of optical waveguide intersections as the optical connection structure of another embodiment of the present invention. As shown in the top view of FIG. 9A, there may be a plurality of first optical waveguides (1A, 1B, 1C) running in the lateral direction, and these exist on the same plane, and the number thereof is arbitrary. be. Further, the second waveguide running in the vertical direction is divided into three regions (2A, 2B, 2C) as in FIG. 1, and in the bridge structure 2C, all the first optical waveguides (1A, 1B, 1C) and a plurality of them. Overpass at the intersection (3A, 3B, 3C). In the present embodiment, while crossing over with a plurality of first optical waveguides (1A, 1B, 1C), as shown by arrow 6 in FIG. 9B, between the first region 2A and the third region 2C. Since the optical transition and the optical transition between the third region 2C and the second region 2B need to be performed once each, it can be used as a short and low-loss optical connection structure. The number of second optical waveguides running in the vertical direction may be plurality, and in a broader sense, the optical waveguides in the two layers may each form a complicated optical circuit or the like.

An embodiment of the present invention has been described with reference to the drawings. However, the present invention is not limited to these embodiments. For example, in each of the above-described embodiments, as the tapered structure, a tapered structure having a curve defined by a parabola or any other smooth function can be adopted instead of the linear taper. Further, the width of the tapered structure is not limited to being changed in two or three steps, and may be changed in four or more steps. Further, the present invention can be carried out in a mode in which various improvements, modifications and modifications are added based on the knowledge of those skilled in the art without departing from the spirit of the present invention.

The optical connection structure of the present invention can be widely used industrially as an optical connection structure that realizes low-loss intersection of optical waveguides in optical circuits such as optical switches and optical integrated circuits.

1, 1A, 1B, 1C: Optical Waveguide (Si Optical Waveguide)
2: Optical Waveguide 2A, 2C: Si Optical Waveguide 2C: SiN Optical Waveguide 3, 3A, 3B, 3C: Intersection 4A, 4B: Overlap Region 6: Light Propagation (Transition)

Claims (4)

It is an optical connection structure in the area where optical waveguides intersect,
With the first optical waveguide,
Includes a second optical waveguide that intersects the first optical waveguide
The second optical waveguide includes a first region and a second region which are in the same plane as the first optical waveguide and are separated from each other, and a plane different from the first optical waveguide and which are the first region and the second region. Including the third area in between
The ends of the first region and the second region have a first tapered region and a first constant width region arranged at the tip thereof.
Both ends of the third region have a second tapered region and a third tapered region arranged at the tip thereof.
The width of the third taper region changes more slowly than the second taper region.
An optical connection structure in which the first constant width region and the third tapered region form an overlapping region. It is an optical connection structure in the area where optical waveguides intersect,
With the first optical waveguide,
Including a second optical waveguide that intersects the first optical waveguide.
The second optical waveguide includes a first region and a second region which are in the same plane as the first optical waveguide and are separated from each other, and a plane different from the first optical waveguide and which are the first region and the second region. Including the third area in between
The end of the first region and one end of the third region, and the end of the second region and the other end of the third region each have an overlap region.
In the overlap region, the first region and the second region of the second optical waveguide have a first tapered region, a first constant width region, and a fifth tapered region at the tip, and the first Both ends of the three regions have a second tapered region, a third tapered region, and a second constant width region or a fourth tapered region at the tip.
The first tapered region and the second constant width region or the fourth tapered region of the tip overlap , the first constant width region and the third tapered region overlap, and the first of the tips. An optical connection structure in which the 5 taper region and the second taper region overlap . In the second optical waveguide, the first region and the second region are made of a silicon optical waveguide, and the third region is made of a silicon nitride optical waveguide.
The light according to claim 1 or 2 , wherein the distance between the first region and the second region and the third region of the second optical waveguide in the thickness direction is in the range of 1.5 to 2.0 μm. Connection structure. In the overlap region, the width W _si (nm) of the first constant width region of the silicon optical waveguide and the tip width W _siN1 (nm) and the rear end of the third tapered region of the silicon nitride optical waveguide. The width W _siN2 (nm) and the length L _taper (μm) of the first constant width region of the silicon optical waveguide and the third taper region of the silicon nitride optical waveguide are
W _siN1 = 0.043W _si ^2-12.5 × W _si +1090
W _siN2 = 0.043W _si ^2-12.5 × W _si +1250
L _taper = 1560 x 0.001 x W _si -250
The optical connection structure according to claim 3 , which has the above-mentioned relationship.

Images