Optical multiplexer and method for manufacturing the same

Abstract

This suppresses the deterioration of multiplexing characteristics caused by variations in the gap width between a pair of optical waveguides due to the miniaturization of the multiplexer. [Solution] An optical multiplexer having a first optical waveguide and a second optical waveguide, the first and second optical waveguides each having a first mode coupling region and a second mode coupling region of a predetermined length adjacent to a predetermined gap width, the optical waveguides of the first and second mode coupling regions constitute a multiplexing section, in the multiplexing section light is transferred between the optical waveguides of the first and second mode coupling regions, the first light in the optical waveguide of the first mode coupling region and the second light in the optical waveguide of the first or second mode coupling region having a different wavelength from the first light are combined in the multiplexing section into one of the optical waveguides of the first and second mode coupling regions, the core cross-sections of at least the optical waveguides of the first and second mode coupling regions in the first and second optical waveguides are asymmetric in the height direction, and the optical multiplexer has at least one multiplexing section.

Description

Technical Field

[0001] The present invention relates to an optical waveguide type optical combiner used in an optical beam scanning type video projection device or the like, and a method for manufacturing the same.

Background Art

[0002] In an optical beam scanning type video projection device that scans an optical beam and projects a video onto a screen or the like, in order to project a color video, three primary colors of red, blue, and green light are combined into one beam, and an optical combiner is used. In particular, in a glasses-type display that projects a video onto the retina, miniaturization of this optical combiner is required. For example, Patent Documents 1 to 3 and Non-Patent Document 1 describe miniaturized optical waveguide type combiners. These optical combiners have red, blue, and green light incident from the incident ends of different optical waveguides respectively, and after combining these lights, the combined light is emitted from the output end of one optical waveguide.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Patent Document 2

Non-Patent Documents

[0004]

Non-Patent Document 1

Summary of the Invention

[0005] Figure 1 is a plan view and a side view showing the configuration of an optical waveguide type multiplexer of Patent Document 1. The side view in Figure 1 is, for example, a side view of the leftmost part of the plan view. The optical waveguide type multiplexer 100 shown in Figure 1 has first to third optical waveguides 101 to 103 and first to third multiplexing sections 110 to 130. The first to third optical waveguides are usually embedded in a cladding layer 112 formed on a substrate 111. The first multiplexing section 110 and the third multiplexing section 130 are optical directional couplers composed of optical waveguides of mode coupling regions 103c that are close together with predetermined gap widths gw1 and gw3 over predetermined lengths L1 and L3 within the second optical waveguide 102 and the third optical waveguide 103. Similarly, the second multiplexing section 120 is composed of optical waveguides of mode coupling regions 103c that are close together with predetermined gap widths gw2 over a predetermined length L2 within the first optical waveguide 101 and the second optical waveguide 102. This is an optical directional coupler composed of an optical waveguide with a coupled region 101c.

[0006] Taking the second multiplexer 120 as an example, this optical directional coupler is the S-shaped guide of the first optical waveguide 101. The optical waveguide 101 has a wave path 101s, a linear mode-coupled region 101c of a predetermined length L2, and another S-shaped waveguide 101s. Furthermore, it has an optical waveguide of a linear mode-coupled region of a predetermined length L2 of a second optical waveguide 102. A region of length L2 adjacent to the waveguide 101 of the mode-coupled region 101c of the first optical waveguide 101 via a gap width gw2 is the mode of the second optical waveguide 102. This corresponds to the coupling region. The same applies to the first multiplexing section 110 and the third multiplexing section 130.

[0007] Hereafter, optical waveguides will sometimes be referred to simply as waveguides. Unless otherwise specified, optical waveguides and waveguides will be considered the same thing. Also, optical directional couplers will sometimes be referred to simply as directional couplers.

[0008] The optical multiplexing method for the optical multiplexer in Figure 1 is as follows: First, the first waveguide 101 is incident. The first visible light (usually red light R) incident on end 101a is transferred to the second waveguide 102 by mode coupling in the second multiplexer 120, then transferred to the third waveguide 103 by mode coupling in the third multiplexer 130 and propagated thereafter before returning to the second waveguide 102 and being emitted from the exit end 102b of the second waveguide 102.

[0009] Next, the second visible light (usually green light G) incident on the incident end 102a of the second waveguide 102 is After propagating to the third waveguide 103 via mode coupling at the first multiplexer 110, the waves propagate to the second waveguide 102 via mode coupling at the third multiplexer 130, and are emitted from the exit end 102b of the second waveguide 102.

[0010] Furthermore, the third visible light (usually blue light B) incident on the incident end 103a of the third waveguide 103 is At the first multiplexer 110, a portion of the propagating light is transferred to the second waveguide 102 by mode coupling and propagates there, while the remainder propagates through the third waveguide 103 as is. At the third multiplexer 130, it is transferred to the second waveguide 102 and combined with the third visible light B that propagated through the second waveguide 102, and then emitted from the exit end 102b of the second waveguide 102. In this way, all the light travels through the same second waveguide 102 It is emitted from the exit end 102b.

[0011] Figure 2 is a plan view and a cross-sectional view showing the configuration of the optical waveguide type multiplexer of Patent Document 2. The side view of Figure 2 is, for example, a side view of the left end of the plan view. The optical waveguide type multiplexer 200 of Figure 2 has first to third waveguides 101 to 103 and first and second multiplexing sections 110 and 120. The first to third waveguides are embedded in a cladding layer 112 formed on a substrate 111. The first multiplexing section 110 and the second multiplexing section 120 are also optical directional couplers having a pair of optical waveguides with predetermined length mode coupling regions that are close together with gap widths gw1 and gw2, respectively.

[0012] The optical multiplexing method in the case of the optical multiplexer shown in Figure 2 is as follows. First, the first visible light (usually red light R) incident on the incident end 102a of the second waveguide 102 is modally coupled in the first multiplexing unit 110. It then moves onto the third waveguide 103, and within the first multiplexer 110, it returns to the second waveguide and propagates. Next, the flux transfers to the first waveguide 101 in the second flux combiner 120, returns to the second waveguide again within the second flux combiner 120, and exits from the exit end 102b of the second waveguide 102.

[0013] Next, the second visible light (usually green light G) incident on the incident end 103a of the third waveguide 103 is Then, at the first multiplexing section 110, the waves are transferred to the second waveguide 102 by mode coupling and propagate, and the second multiplexing section After transferring to the first waveguide 101 in section 120, the device returns to the second waveguide again within the second combined section 120 and exits from the exit end 102b of the second waveguide 102.

[0014] Then, the third visible light (usually blue light B) is incident on the incident end 101a of the first waveguide 101. The light then transfers to the second waveguide 102 via mode coupling in the second multiplexer 120 and exits from the exit end 102b of the second waveguide 102. In this way, all the light travels through the same second waveguide 102 It is emitted from the exit end 102b.

[0015] Here, the multiplexing section of these optical waveguide type multiplexers is mainly composed of an optical directional coupler, as shown in Non-Patent Document 1. An optical directional coupler brings two optical waveguides, each carrying different types of light, close together and has a gap width between the two optical waveguides and a mode coupling region length that allows light propagating in one optical waveguide to transfer to the other optical waveguide. It utilizes the phenomenon of propagating light transferring between optical waveguides (mode coupling) to change the path of the light.

[0016] This optical directional coupler has the following two problems. First, in the optical directional coupler, Figure 1 As shown in the cross-sectional view in Figure 2, the optical waveguides 101-103 that constitute the optical directional coupler are embedded in the cladding layer 112 formed on the substrate 111. In this case, the width of the gap between the two waveguides is on the order of microns in order to facilitate the transfer of light. It needs to be made smaller. Therefore, when manufacturing an optical directional coupler, the variation in the gap width due to the miniaturization of the gap becomes large. As a result, there is a problem that the magnitude of mode coupling changes and the multiplexing characteristics of the optical waveguide type multiplexer vary. As a result, there is a major industrial problem of reduced productivity during manufacturing due to the deterioration of the yield during the production of the optical waveguide type multiplexer.

[0017] Second, there is a problem that the mode coupling efficiency of the optical directional coupler itself as the multiplexing section is poor. That is, even between red light and blue light, which have the largest wavelength difference among the three primary colors of red, green, and blue, the difference in mode coupling of light in the optical directional coupler is small, meaning that the multiplexing performance where one propagating light transfers between waveguides and the other propagating light does not transfer is low. In other words, there is a problem that the difference in mode coupling between red light and blue light is small, the difference in the transfer characteristics between waveguides in the multiplexing section is small, and the multiplexing performance of the optical directional coupler is not high.

[0018] Therefore, an object of the first aspect of the present embodiment is to provide an optical multiplexer and a method for manufacturing the optical multiplexer that solve any of these problems of the optical multiplexer. Solving means for the problem

Means for solving the problem

[0019] The first aspect of this embodiment is an optical multiplexer having a first optical waveguide and a second optical waveguide, the first and second optical waveguides each having a first mode coupling region and a second mode coupling region of a predetermined length adjacent to a predetermined gap width, the optical waveguides of the first and second mode coupling regions constitute a multiplexing section, in the multiplexing section light is transferred between the optical waveguides of the first and second mode coupling regions, the first light in the optical waveguide of the first mode coupling region and the second light in the optical waveguide of the first or second mode coupling region having a different wavelength from the first light are combined in the multiplexing section into one of the optical waveguides of the first and second mode coupling regions, the core cross-sections of at least the optical waveguides of the first and second mode coupling regions in the first and second optical waveguides are asymmetric in the height direction, and the optical multiplexer has at least one multiplexing section. [Effects of the Invention]

[0020] According to the first aspect, variations in the multiplexing characteristics of waveguide-type multiplexers caused by variations in the gap width between waveguides of optical multiplexers due to variations in the manufacturing process can be reduced. In addition, the efficiency of mode coupling of the directional coupler itself (the multiplexing characteristics of the optical multiplexer) can be improved. [Brief explanation of the drawing]

[0021] [Figure 1] These are a plan view and a side view showing the configuration of an optical waveguide type multiplexer as specified in Patent Document 1. [Figure 2] These are a plan view and a cross-sectional view showing the configuration of an optical waveguide type multiplexer as stipulated in Patent Document 2. [Figure 3] This figure shows the main operation examples of the optical directional coupler (combining section) that forms the optical waveguide type multiplexer in this embodiment. [Figure 4] This figure shows the cross-sectional shapes of a pair of waveguides in a conventional optical multiplexer (directional coupler) and a pair of waveguides in this embodiment. [Figure 5] This figure shows the simulation results of the relationship between the ratio of the bond lengths of the blue and red sections (B / R ratio) and the ratio of the lengths of the top and bottom sides of the trapezoidal cross-section (top side / bottom side ratio). [Figure 6A]This diagram shows the relationship between seven types of waveguide cross-sectional shapes and their top / bottom ratios. [Figure 6B] This diagram shows the relationship between seven types of waveguide cross-sectional shapes and their top / bottom ratios. [Figure 7] This figure uses the same top / bottom ratio as in Figure 5 for the horizontal axis, and the vertical axis represents the amount of change in the B / R ratio associated with changes in the gap width gw, ranging from 3.1 to 0.3 μm: (1) the difference between the maximum and minimum values, and (2) the standard deviation. [Figure 8] This figure corresponds to Figure 7, showing optical directional couplers with different refractive index differences between the core and cladding, waveguide spacing, and gap width ranges. [Figure 9] This figure corresponds to Figure 7, showing optical directional couplers with different waveguide heights and gap width ranges. [Figure 10] This figure corresponds to Figure 7 of an optical directional coupler, in which the core cross-section has a shape in which two rectangular sections of different widths are stacked in layers. [Figure 11] This diagram corresponds to Figure 7 of an optical directional coupler, in which the core cross-section is a trapezoid with one side perpendicular and the other oblique. [Figure 12A] This figure shows the change in the B / R ratio with respect to the ratio of gap embedding height to waveguide height for an optical directional coupler in which a portion of the gap between waveguides is embedded in the core cross-section. [Figure 12B] This figure shows the change in the B / R ratio with respect to the ratio of gap embedding height to waveguide height for an optical directional coupler in which a portion of the gap between waveguides is embedded in the core cross-section. [Figure 13] Figure 5 shows the relationship between the B / R ratio and the gap width (gw) for a typical directional coupler, with trapezoidal and rectangular cross-sections. [Figure 14] Figure 5 shows the characteristics of a typical directional coupler, plotted with the horizontal axis as (difference between the length of the base and the length of the top of the trapezoid) / (length of the base of the trapezoid) and the vertical axis as the difference between the maximum and minimum values ​​of the B / R ratio. [Figure 15]Figure 5 shows the relationship between the coupling length and the top / bottom ratio for a typical directional coupler (with a fixed refractive index difference of 0.8% between the core and cladding, a waveguide height (h) of 1.6 μm, and a waveguide spacing (d) of 4 μm), represented by the red and blue lines. [Figure 16] Figure 5 shows the characteristics of a typical directional coupler, plotted with the waveguide spacing (d) on the horizontal axis, the maximum B / R ratio on the vertical axis, and the gap width (g) as a parameter. [Figure 17] This figure shows an example of a pseudo-trapezoidal shape in the core cross-section of a waveguide in an optical directional coupler. [Figure 18] This is a plan view showing an example of a three-primary-color optical waveguide type multiplexer in Example 6. [Figure 19] This is a plan view of the multiplexer in Example 7. [Figure 20] This figure shows the configuration of the first optical waveguide type multiplexer in the third embodiment. [Figure 21] This figure shows the simulation results of how the three colors of light propagate in the first multiplexer of the third embodiment. [Figure 22] This figure shows the configuration of the second optical waveguide type multiplexer in the third embodiment. [Figure 23] This figure shows the simulation results of the propagation of three colors of light in the second multiplexer of the third embodiment. [Figure 24] This figure shows the third and fourth optical waveguide type multiplexers of the third embodiment. [Figure 25] This figure shows the simulation results of the propagation of three colors of light in the fourth multiplexer 800 of the third embodiment. [Modes for carrying out the invention]

[0022] This embodiment will be described below with reference to the drawings.

[0023] [Main operation examples of optical directional couplers (combining section) forming optical waveguide type multiplexers] FIG. 3 is a diagram showing main operation examples of an optical directional coupler (combining section) that forms an optical waveguide type combiner in the present embodiment. The operation examples of the optical directional coupler are summarized for red and blue of the three primary colors, and there are three types: operation examples (1), (2), and (3) in FIG. 3.

[0024] The optical directional coupler constituting the combining section has two linear optical waveguides WG1 and WG2 of a predetermined length L arranged adjacent to each other with a short gap width gw. When the gap width between the two optical waveguides is short, the field of light propagating through the optical waveguide can penetrate beyond the interface between the core and the cladding of the optical waveguide and reach the adjacent optical waveguide. This phenomenon is a characteristic due to the wave nature of light and is called an evanescent wave. That is, while light is propagating in one optical waveguide, it gradually transfers to the adjacent optical waveguide, and finally, all of it transfers. The length of the optical waveguide where all this light transfers is the coupling length.

[0025] If the optical waveguide length L of the optical directional coupler is set to half of the coupling length, due to the mode coupling of light, only 50% of the propagating light transfers between the optical waveguides, and the light is distributed 50:50 between the two optical waveguides. If the gap width between the two optical waveguides outside the optical directional coupler is made sufficiently long, the light distributed to both optical waveguides propagates through each optical waveguide while remaining 50:50. If the optical waveguide length of the optical directional coupler is set to the coupling length, the light propagating in one optical waveguide all transfers to the other optical waveguide. Furthermore, if it is set to twice the coupling length, after the light propagating in one optical waveguide transfers to the other optical waveguide, it returns to the original optical waveguide. Also, if it is set to three times the coupling length, it transfers three times, and the light propagating in one optical waveguide repeats the transfer three times and finally transfers to the other optical waveguide. Moreover, the longer the wavelength of light, the wider the field of light and the longer the length that penetrates beyond the boundary of the optical waveguide, the stronger the optical mode coupling, and the shorter the coupling length. Therefore, since the wavelengths λ of the three lights, red R, green G, and blue B, are λR > λG > λB, the coupling lengths L have the relationship LR < LG < LB.

[0026]

[0027] ​As described above, mode coupling is related to the spreading (leakage) of light propagating from an optical waveguide. The smaller the gap between adjacent optical waveguides, the easier it is for light to reach the adjacent waveguide, and as a result, light is more likely to transfer to the adjacent waveguide. As the wavelength of light increases, this leakage amount increases, so red light exhibits greater mode coupling than blue light. Therefore, light with a longer wavelength can transfer to an adjacent optical waveguide even with a shorter mode coupling region.

[0028] In operation example (1), red R and blue B light are incident on separate optical waveguides WG1 and WG2. Red R travels from optical waveguide WG1, where it was incident, to the adjacent optical waveguide WG2, while blue B travels straight through optical waveguide WG2, where it was incident. Both red R and blue B are then combined and emitted from optical waveguide WG2, where the blue light was incident. This operation example (1) corresponds to the operation of the second wave combiner 120 in Figure 1.

[0029] In operation example (2), red R and blue B light are incident on separate optical waveguides WG1 and WG2. Red R temporarily transfers to the other optical waveguide WG2, and in some cases, it may transfer between the two optical waveguides several times before finally returning to the original optical waveguide WG1 where it was incident. Blue B transfers from the optical waveguide WG2 where it was incident to the other optical waveguide WG1. Then, both red R and blue B are combined and emitted from the optical waveguide WG1 where the red light was incident. This operation example (2) corresponds to the operation of the second wave combiner 120 in Figure 2.

[0030] In operation example (3), red R and blue B light are incident on the same optical waveguide WG1. Red R temporarily moves to the other optical waveguide WG2, and in some cases, it may move between the two optical waveguides several times before finally moving to the other optical waveguide WG2. Blue B also moves from the optical waveguide WG1 in which it was incident to the other optical waveguide WG2. Then, both red R and blue B are combined and emitted from an optical waveguide WG2 that is different from the optical waveguide WG1 in which both red and blue were incident. This operation example (3) is not used in the multiplexing section shown in Figures 1 and 2, but it is one of the main operations used in the multiplexing section of the embodiment described later. By combining these operation examples (1), (2), and (3), an optical waveguide type multiplexer can be formed.

[0031] According to the main operation examples of the multiplexer described above, in this specification, the multiplexer is defined as a system in which at least two light beams of different wavelengths, one or both of which are incident (propagate) into a pair of optical waveguides (a pair of optical waveguides in a mode-coupled region) of a directional coupler, are combined into one of the optical waveguides of the pair of optical waveguides in the first and second mode-coupled regions by transferring from the optical waveguide of the first mode-coupled region to the optical waveguide of the second mode-coupled region, and then the combined light is emitted (propagated) from that one optical waveguide. The two light beams incident and propagating into the directional coupler may be incident separately into the pair of optical waveguides (operation examples (1) and (2) in Figure 3), or they may be incident together into one of the pair of optical waveguides (operation example (3) in Figure 3).

[0032] Note that while the operation examples (1), (2), and (3) in Figure 3 were explained using a directional coupler consisting of two straight waveguides, in reality, as shown in Figures 1 and 2, a pair of S-shaped optical waveguides 101s, each consisting of a curve, are connected to the left and right ends of the pair of optical waveguides (optical waveguides 101 in the mode-coupled region 101c), respectively. Therefore, the transfer of light between the two optical waveguides does not occur only in the directional coupler consisting of two straight waveguides, but also partially in these S-shaped optical waveguides in actual multiplexers. In other words, the directional coupler consisting of two straight waveguides shown here is for the purpose of explaining the operating principle, and in actual multiplexer manufacturing, a directional coupler that effectively incorporates an S-shaped optical waveguide is used. The structure needs to be determined.

[0033] Furthermore, while the mode coupling region has been described here as a straight waveguide, it can certainly be a curve that deviates slightly from a straight line; the point is that any waveguide in which mode coupling occurs between two adjacent optical waveguides is acceptable. The mode coupling regions described below are all assumed to be straight lines, but they are not limited to straight lines.

[0034] In this embodiment, the directional coupler is composed of an optical waveguide that propagates fundamental mode light. Alternatively, the directional coupler in this embodiment is composed of an optical waveguide for a pseudo-fundamental mode in which the magnitude of the fundamental mode light is larger than that of other higher-order mode light.

[0035] In the optical waveguide type multiplexers described above, as shown in the side view of the multiplexer viewed from the left in Figures 1 and 2, optical waveguides with a rectangular core cross-section are mainly used (Patent Document 1, Patent Document 2). However, firstly, in optical directional couplers of optical waveguides with these rectangular cross-sections, if the gap width variation increases due to the miniaturization of the gap between optical waveguides, the magnitude of mode coupling in the multiplexing section changes significantly, resulting in a problem of variation in the multiplexing characteristics of the waveguide type optical multiplexer.

[0036] Secondly, there is the problem that the mode coupling efficiency of the directional coupler itself as a multiplexing section is poor, resulting in low multiplexing performance for the directional coupler.

[0037] Therefore, the inventors investigated optical waveguides with a core cross-section that is asymmetrical in the height direction, instead of optical waveguides with a rectangular core cross-section, and found that the two problems mentioned above could be resolved. Thus, in the optical multiplexer of this embodiment, the S-shaped waveguide of the optical directional coupler and the mode coupling region The core cross-sectional shape of the optical waveguide in the region is made asymmetrical in the height direction.

[0038] First, as an example of an optical waveguide with a core cross-section asymmetric in the height direction, we will describe the case of an optical waveguide with a trapezoidal cross-section. Figure 4 shows the cross-sectional shapes of a pair of waveguides in a conventional optical multiplexer (directional coupler) and a pair of waveguides in this embodiment. These structures can be defined by the waveguide height (h), waveguide width (w), waveguide spacing (d), and gap width (gw). In the case of a trapezoid, the gap width (gw) is defined as the gap width at the base of the trapezoid. The interval (d) is defined as the distance between the centers of the optical waveguides.

[0039] The mode coupling efficiency of a directional coupler as a multiplexer is usually expressed by the coupling length. As mentioned above, the coupling length represents the distance in the waveguide direction over which light propagating in one optical waveguide transfers to the other optical waveguide in a directional coupler. Therefore, a longer coupling length results in smaller mode coupling, and a shorter coupling length results in larger mode coupling. This coupling length varies depending on the wavelength of light, and generally, shorter wavelengths such as blue light have a larger value than longer wavelengths such as red light. This is because the length over which longer wavelength red light seeps out from the core of the field is longer than that of shorter wavelength blue light. The coupling length also depends on the gap width of the optical waveguides in the pair of mode coupling regions of the directional coupler. A narrower gap results in a shorter coupling length, and a wider gap results in a longer coupling length. Therefore, as shown in Figures 3(1), (2), and (3)... The operation is determined by the ratio of the blue coupling length (LB) to the red coupling length (LR) (LB divided by LR, denoted as the B / R ratio (=LB / LR)). Furthermore, the ratio of the coupling lengths depends on the gap width of the pair of optical waveguides, and further, on the wavelengths of the two light sources.

[0040] As a result of their investigation, the inventors of this invention found that the coupling length, particularly the ratio (B / R ratio) of red and blue light, which have the largest wavelength difference among the three primary colors of light, depends greatly on the waveguide cross-sectional shape.

[0041] Figure 5 shows the simulation results of the relationship between the ratio of the coupling lengths of the blue and red cores (B / R ratio) and the ratio of the lengths of the top and bottom sides of the trapezoidal cross-section (top / bottom ratio). In this case, the refractive index difference between the core and cladding is 0.8%, the waveguide height (h) is 1.6 μm, and the waveguide spacing (d) is 4 μm. The value of is a typical value for the directional coupler, which is the multiplexing section of the multiplexer shown in Non-Patent Document 1. Note that the refractive index difference between the core and the cladding is calculated by setting the refractive index of the core to n. core The refractive index of the cladding is n clad When this is the case, (n core -n clad ) / n core It is defined as ×100 (%). In addition, the gap width (gw) is varied from 3.1 to 0.3 μm as a parameter. However, since the waveguide spacing (d) is fixed at 4 μm, the waveguide width (w) varies from 0.9 to 3.7 μm.

[0042] Figures 6A and 6B show the relationship between seven types of waveguide cross-sections and their top / base ratios. In Figure 5, a top / base ratio of 1 on the horizontal axis corresponds to a rectangular cross-section (cross-section in Figure 6A(1)), a top / base ratio less than 1 but greater than 0 corresponds to a trapezoidal cross-section (Figure 6A(2)), a top / base ratio of 0 or less corresponds to a triangular cross-section (Figure 6A(3)), and a top / base ratio greater than 1 corresponds to an inverted trapezoidal cross-section where the length of the top side is greater than the length of the base side (Figure 6A(4)).

[0043] Furthermore, Figure 6B(5) shows a two-stage core cross-section where multiple rectangles are stacked, Figure 6B(6) shows a trapezoidal cross-section with one side perpendicular, and Figure 6B(7) shows a cross-section where the gap between waveguides is filled with core material. In the cross-section of Figure 6B(7), the gap embedded height (geh) is added as a parameter.

[0044] Furthermore, in Figure 5, the wavelength of red light is 0.633 μm and the wavelength of blue light is 0.448 μm. However, the embodiments of this application are not limited to these wavelengths for both red and blue light; any wavelength range that is actually perceived as blue or red is acceptable. Specifically, the respective wavelength ranges are approximately 0.6 to 0.8 μm for red and approximately 0.4 to 0.5 μm for blue.

[0045] Note that Figure 5 shows the simulation results for a directional coupler consisting of two straight waveguides, but if an S-shaped optical waveguide consisting of curves is connected to these directional couplers... However, the coupling lengths of the red and blue light themselves change. If it is connected to the waveguide of a directional coupler, the bending radius of the S-shaped optical waveguide is this curve. If a practical S-shaped optical waveguide does not have radiation loss to the outside due to turbulence, then the blue and red connections... The ratio of combined lengths (B / R ratio) has been confirmed to be approximately the same as that of a directional coupler consisting of two straight waveguides. This result applies not only to Figure 5 but to all cases described later.

[0046] [Optical waveguide of the first embodiment that reduces the amount of variation (variation) in the B / R ratio, which is the first challenge] type multiplexer] Figure 5 shows the B / R ratio on the vertical axis corresponding to the top / bottom ratio = (w-2x) / w (see Figure 6A(3)) on the horizontal axis, for each of the eight gap widths. According to this, when the top / bottom ratio is changed by changing the x value, the B / R ratio peaks when the top / bottom ratio is between 0 and 1, and the position of each peak shifts according to the gap width. In other words, Figure 5 shows that the B / R ratio is highly dependent on the top / bottom ratio and also highly dependent on the gap width.

[0047] Therefore, to quantify the change in the B / R ratio associated with the change in gap width, the horizontal axis is set to the same top / bottom ratio as in Figure 5, and the vertical axis shows the difference between the maximum and minimum values ​​of the B / R ratio in the gap width range of 3.1 to 0.3 μm.

[0048] Figure 7 shows the difference between the maximum and minimum values, with the horizontal axis having the same top / bottom ratio as in Figure 5, and the vertical axis representing the change in the B / R ratio associated with the change in gap width gw in the range of 3.1 to 0.3 μm (1) (2) The figure shows the standard deviation. Figure 7 is an example of a waveguide spacing w = 4 μm. The range of gap width gw from 3.1 to 0.3 μm covers commonly used gap widths, as shown in Figure 7. The result in (1) can be considered as the difference between the maximum and minimum values ​​that generally hold true for directional couplers that are typically fabricated. As an example, Figure 5 shows the magnitude of the difference between the maximum and minimum values ​​D_max-min when the top / bottom ratio is 1. The following can be seen from Figure 7.

[0049] (1) Where the cross section changes from a rectangular cross section with a top-to-base ratio of 1 to a trapezoidal cross section, the difference between the maximum and minimum values ​​of the B / R ratio changes significantly, resulting in a trapezoid with a top-to-base ratio less than 1 and greater than 0. The difference between the maximum and minimum values ​​decreases rapidly. The straight line in the diagram is a guide line to illustrate this effect.

[0050] (2) The difference between the maximum and minimum B / R ratio in a trapezoid is the difference between the maximum and minimum values ​​of the trapezoid (where the top side / bottom side ratio is less than 1). When the value is full and greater than 0, it is significantly smaller compared to the case of a rectangle (top side / base side ratio is 1). This means that the range of variation in the B / R ratio when the gap width changes due to process variations is small for trapezoids. Note that when the top side / base side ratio is 0 or less, the difference between the maximum and minimum values ​​of the B / R ratio is Similarly small, this region has a triangular cross-section.

[0051] (3) The same effect as the difference between the maximum and minimum values ​​of these B / R ratios is obtained in the case of the standard deviation of the B / R ratio obtained by changing the gap width in Figure 7(2).

[0052] From the above results, it can be seen that the magnitude of the change in the B / R ratio due to the change in gap width is smaller compared to a rectangular cross-section, with the top side / bottom side ratio being less than 1.

[0053] As shown above, even if variations (fluctuations) in gap width occur due to the miniaturization of the gap between waveguides, in the case of a trapezoidal cross section with a top-to-bottom ratio of less than 1, the magnitude of the variation in the B / R ratio with respect to the gap width variation is smaller compared to a rectangular cross section. As a result, making the core cross section of the waveguide of an optical directional coupler trapezoidal reduces the variation in the multiplexing characteristics of industrially important optical waveguide type multiplexers. Since the light transfer operation shown in Figures 3(1), (2), and (3) is determined by the B / R ratio, even if the gap width between waveguides in the multiplexing section varies in the manufacturing process, a small variation in the B / R ratio means that the operating characteristics of the multiplexing section will also be small. In other words, in the optical multiplexing section, the ratio of the gap width to the optical path length which depends on the wavelength of light, and the B / R ratio, will be at the desired optimal value. In this case, the desired combined wave characteristics are obtained. Therefore, if the B / R ratio fluctuates little with respect to gap width variations due to manufacturing variations, the desired combined wave characteristics are more easily obtained, and the yield improves.

[0054] Therefore, it is desirable that the top / base ratio of one or both of the waveguide cross-sections in the combined section (optical directional coupler) of the first embodiment be less than 1, for example, 0.95 or less. If the top / base ratio is 0.95 or less, the waveguide core cross-section will have a significant trapezoidal shape, which is substantially different from a rectangular shape (top / base ratio of 1 ± 0.05) which is subject to manufacturing variations. Furthermore, it is desirable that the top / base ratio be less than 1 (0.95 or less) and greater than or equal to 0, which is the range of a shape (in this case, the trapezoidal range) in which the light intensity of the light emitted from the waveguide does not deviate much from a symmetrical (circular) distribution. Most preferably, for example, the top / bottom ratio is less than 1 (0.95 or less) and preferably 0.5 or more. Furthermore, it is desirable that the top / bottom ratio be less than 1 (0.85 or less) and 0.5 or greater. If the value is less than 1 and greater than or equal to 0.5, the light intensity distribution of the emitted light will remain close to a circle. Even if the top / bottom ratio is 0 or less, the light intensity of the emitted light has a symmetrical (circular) distribution. If not required, the top / bottom ratio may be 0 or less, as shown in Figure 6A(3). A top-to-base ratio of 0 or less means a triangle in which the height of the core cross-section is lower than the height h of the waveguide. ru.

[0055] The top / bottom ratio characteristics of the waveguide cross-section described above have been confirmed not only for the structural parameters of the typical directional couplers shown in Figures 5 and 7, but also for the structural parameters of other commonly used directional couplers, as will be discussed later. This indicates that it is a universal effect that holds true for commonly used multiplexers. For example, in addition to the structure of a core layer surrounded by a cladding layer in an optical directional coupler, a ridge-type waveguide with a ridge-shaped core layer also exhibits similar B / R ratio characteristics with respect to the top / bottom ratio.

[0056] [Regarding the asymmetrical shape of the core's cross-section in the height direction] In the first embodiment, the asymmetrical shape of the core cross-section of a pair of waveguides in the optical directional coupler in the height direction includes trapezoidal, triangular, inverted trapezoidal, two-tiered, trapezoidal with one side perpendicular, and gap-embedded shapes in which part of the gap between rectangular waveguides is filled with core material, as shown in Figures 6A(2)-(4) and 6B(5)-(7).

[0057] [The core cross-section is trapezoidal or pseudo-trapezoidal] As already explained in Figures 5 and 7, a trapezoidal core cross-section results in less variation in the B / R ratio due to manufacturing process variations. A trapezoidal core cross-section includes a pseudo-trapezoidal shape.

[0058] For example, even if the side walls of the core cross-section are not straight but consist of curves that deviate from a straight line, if the shape can be approximately considered a pseudo-trapezoid and the characteristics of the trapezoid are almost the same as those of the pseudo-trapezoid, then the pseudo-trapezoid can also be considered a trapezoid.

[0059] Furthermore, even if the top and bottom edges of the core cross-section are not straight lines but gentle curves, the shape can be considered a pseudo-trapezoid, and if the characteristics of the trapezoid are almost the same as those of the pseudo-trapezoid, then the pseudo-trapezoid can also be considered a trapezoid. In addition, shapes in which the corners between the top edge and the side are rounded, and shapes in which the outer corners between the side and the bottom are rounded, are also pseudo-trapezoids.

[0060] Here, we have described the case where the core cross-section is trapezoidal, but the same applies to waveguide cross-sectional shapes other than trapezoidal, as described below ((5) two-stage core shape in Figure 6B, (6) trapezoidal shape with one side vertical and the other oblique, and (7) gap-embedded shape where the gap between waveguides is embedded). In other words, even in the case of these shapes, if the top, bottom, and side walls are curves that deviate from straight lines, they can be approximately considered as the corresponding cross-sectional shape where the top, bottom, and side walls are straight lines, and if the characteristics of the corresponding cross-sectional shape consisting of straight lines are approximately the same as those of the case where the top, bottom, and side walls are curves that deviate from straight lines, then the case where the top, bottom, and side walls are curves that deviate from straight lines can also be considered as the corresponding cross-sectional shape consisting of straight lines.

[0061] Figure 17 shows examples of pseudo-trapezoidal shapes in the core cross-section of a waveguide of an optical directional coupler. Figure 17 shows three types of pseudo-trapezoidal examples: (1) rounded corners, (2) fluctuations in the side walls, and (3) fluctuations in the upper base.

[0062] In Figure 17(1), an example of corner blunting, the corner between the top side Lt and the side Ls of the trapezoid is rounded off, and the distance between the ideal trapezoidal corner and the actual shape is shown as the amount of blunting D. Also, the exterior angle between the base side Lb and the side Ls of the trapezoid is filled in, and the distance between the ideal trapezoidal exterior angle and the actual shape is shown as the amount of blunting D. The range of blunting amounts D that are considered to be pseudo-trapezoids is, for example, as follows: D <MAX(Lt,Ls,Lb)*0.2 In other words, if the degree of corner blunting D is less than 20% of the longest of the top side Lt, bottom side Lb, and side Ls adjacent to that corner, the trapezoid is considered a pseudo-trapezoid. In this embodiment, even if corner blunting of this degree is formed due to the manufacturing process, etc., it is considered a pseudo-trapezoid.

[0063] In the example of sidewall fluctuations in Figure 17(2), the trapezoidal sidewall is curved, and the distance between it and an ideal trapezoidal sidewall is shown as the sidewall fluctuation amount D. The range of fluctuation amount D that is considered to be pseudo-trapezoidal is, for example, as follows: MAX(D) <Ls*0.2 In other words, if the maximum value of the sidewall fluctuation D is less than 20% of the sidewall length Ls, the trapezoid is considered a pseudo-trapezoid. In this embodiment, even if this level of sidewall fluctuation is formed due to the manufacturing process, it is considered a pseudo-trapezoid.

[0064] In the example of upper base fluctuation in Figure 17(3), the upper and lower sides of the trapezoid are curved, and the distance between the upper and lower sides of an ideal trapezoid is shown as the upper base fluctuation amount D. The range of upper base fluctuation amount D that is considered a pseudo-trapezoid is, for example, as follows: MAX(D)<(Lt or Lb)*0.2 In other words, if the maximum value of the fluctuation D of the top or bottom edge is less than 20% of the length Lt or Lb of that top or bottom edge, the trapezoid is considered a pseudo-trapezoid. In this embodiment, even if fluctuations of the top and bottom edges due to the manufacturing process, etc., are formed to this extent, it is considered a pseudo-trapezoid.

[0065] [Other examples of trapezoids (1)] Figure 8 corresponds to Figure 7 for optical directional couplers with different refractive index differences between the core and cladding, waveguide spacing, and gap width ranges. In Figure 8, the waveguide height (h) is 1.6 μm, the same as in Figures 5 and 7, but the refractive index difference between the core and cladding is 1.1%, and the waveguide spacing (d) is 3.4 μm. The gap width is varied from 1.2 μm to 2 μm. Figure 8(1) shows the relationship between the difference between the maximum and minimum values ​​of the B / R ratio and the upper / lower ratio, and Figure 8(2) shows the relationship between the standard deviation of the B / R ratio and the upper / lower ratio. In this case as well, the three results (1), (2), and (3) explained in Figure 7 are obtained.

[0066] [Other examples of trapezoids (2)] Figure 9 is a diagram corresponding to Figure 7 for optical directional couplers with different waveguide height and gap width ranges. In Figure 9, the refractive index difference between the core and cladding is 0.8%, and the waveguide spacing (d) is 4 μm and Figure 5. The same results as in Figure 7 are obtained, but with a waveguide height (h) of 1.2 μm. The gap width is varied from 1.5 μm to 2.4 μm. Figure 9(1) shows the relationship between the difference between the maximum and minimum B / R ratios and the upper / lower ratio, and Figure 9(2) shows the relationship between the standard deviation of the B / R ratio and the upper / lower ratio. In this case as well, the three results (1), (2), and (3) explained in Figure 7 are obtained.

[0067] The three examples of directional couplers shown in Figures 7, 8, and 9 above, namely (1) Refractive index difference between core and cladding: 0.8%, waveguide height (h): 1.6 μm, waveguide spacing (d): Example of 4μm (2) Difference in refractive index between core and cladding: 1.1%, waveguide height (h): 1.6 μm, waveguide spacing (d): Example of 3.4 μm and (3) Refractive index difference between core and cladding: 0.8%, waveguide height (h): 1.2 μm, waveguide spacing (d): As shown in the 4μm example, the three results (1), (2), and (3) explained in Figure 7 are It is shown that these three examples of directional couplers hold true. Since these three examples are representative of directional couplers commonly used in trichromators, the three results (1), (2), and (3) explained in Figure 7 can be said to generally hold true for commonly used directional couplers. Here, commonly used directional couplers are defined as those with a refractive index difference of 5% or less between the core and cladding, and a waveguide spacing of 5% or less. (d) is 15 μm or less, the waveguide height (h) is 10 μm or less, and the main propagating through the waveguide is This is the case when the essential mode becomes the basic mode.

[0068] [Example of a two-stage core cross-section] Similar results were obtained in a directional coupler using a waveguide whose core cross-section, as shown in Figure 6B(5), consists of two layers of rectangular sections of different widths stacked together (a two-layer core cross-section).

[0069] Figure 10 is a diagram corresponding to Figure 7 of an optical directional coupler with a core cross-section consisting of two stacked rectangular sections of different widths. In this case, the waveguide has an upper core with a smaller width than the lower core, and the upper and lower cores are the same height. The refractive index difference between the core and cladding is 0.8%, the waveguide height (h) is 1.6 μm (sum of the heights of the upper and lower cores), the waveguide spacing (d) is 4 μm, and the gap width (gap width between the lower cores) is 1.5 μm. The range is varied within 2.7 μm. Figure 10(1) shows the relationship between the difference between the maximum and minimum values ​​of the B / R ratio and the upper / lower ratio, and Figure 10(2) shows the relationship between the standard deviation of the B / R ratio and the upper / lower ratio. Here, the upper / lower ratio is the ratio of the width of the upper core to the width of the lower core. The following can be seen from Figure 10.

[0070] (1a) When the top / bottom ratio becomes 1, that is, when the cross section changes from a rectangular cross section to a two-stage core cross section, the difference between the maximum and minimum values ​​of the B / R ratio changes significantly, and as a result of the two-stage core cross section, the difference between the maximum and minimum values ​​decreases rapidly. The straight lines in the figure are guide lines to illustrate this effect.

[0071] (2a) The difference between the maximum and minimum B / R ratio in the two-stage core section is significantly smaller across the entire range of the two-stage core section (top / bottom ratio less than 1 and greater than or equal to 0) compared to the case of a rectangle (top / bottom ratio of 1). It's blooming.

[0072] (3a) The same effect as the difference between the maximum and minimum values ​​of these B / R ratios is obtained in the case of the standard deviation of the B / R ratio obtained by changing the gap width, as shown in Figure 10(2).

[0073] Modified examples of the two-stage core cross-section are as follows. Above, we illustrated a waveguide in which the core cross-section of the waveguide has a shape in which two rectangles of different widths are stacked in stages. However, even when the waveguide core cross-section has a shape in which two or more rectangles of different widths are stacked in stages, and the width decreases as you go to the upper stages, the results of (1a), (2a), and (3a) above are obtained. Note that here, "the width decreases as you go to the upper stages" means that the core width of most stages decreases as you go to the upper stages among a large number of stages, and does not mean that the width of the multi-stage core cross-section decreases monotonically from the lower stages to the upper stages. In such a multi-stage core cross-section, the limit when the number of stages is large corresponds to a trapezoid.

[0074] Furthermore, in the case of a trapezoidal cross-section where the side walls are not straight but consist of curves or other shapes that deviate from straight lines, this can be considered the limit when the number of stages in a multi-stage core cross-section becomes large, and it is natural that a trapezoid consisting of curves or other shapes that deviate from straight lines will also exhibit the results of (1a), (2a), and (3a) above. Note that the heights of each stage of the multi-stage core do not necessarily have to be the same.

[0075] [Example of a trapezoid with a cross-section where one side is vertical and the other is oblique] Similar results were obtained in a directional coupler using a waveguide with a trapezoidal cross-section where one side is vertical and the other is oblique, as shown in Figure 6B(6).

[0076] Figure 11 is a diagram corresponding to Figure 7 of an optical directional coupler in which the core cross-section is a trapezoid with one side perpendicular and the other oblique. In this case, the refractive index difference between the core and cladding is 0.8%, and the waveguide height (h) The gap width is 1.6 μm, the waveguide spacing (d) is 4 μm, and the gap width is varied in the range of 1.6 μm to 2.8 μm. Figure 11(1) shows the relationship between the difference between the maximum and minimum B / R ratio and the upper / lower ratio, and Figure 11(2) shows the relationship between the standard deviation of the B / R ratio and the upper / lower ratio. The following can be seen from Figure 11.

[0077] (1b) Where the top / base ratio is 1, that is, where the cross-section changes from a rectangular to a trapezoidal shape, the difference between the maximum and minimum values ​​of the B / R ratio changes significantly, and as it becomes a trapezoid, the difference between the maximum and minimum values ​​decreases rapidly. The straight lines in the figure are guide lines to illustrate this effect.

[0078] (2b) The difference between the maximum and minimum B / R ratio in a trapezoid is the difference between the maximum and minimum values ​​of the trapezoid (where the top side / bottom side ratio is 1 (less than 0 and greater than or equal to 0), the values ​​are significantly smaller compared to the case of a rectangle.

[0079] (2c) The same effect as the difference between the maximum and minimum values ​​of these B / R ratios is obtained in the case of the standard deviation of the B / R ratio obtained by changing the gap width, as shown in Figure 11(2). Note that these results are obtained in the same way whether the slanted side of the trapezoid is on the adjacent waveguide core side or the opposite side.

[0080] [Example where the gap between waveguides is filled] Similar results were obtained in the directional coupler shown in Figure 6B(7), where a portion of the gap between waveguides is embedded.

[0081] Figures 12A and 12B show the change in the B / R ratio with respect to the ratio of gap embedding height to waveguide height for an optical directional coupler in which the core cross-section has a shape in which part of the gap between waveguides is embedded. In this case, the refractive index difference between the core and cladding is 0.8%, and the waveguide height (h) The gap width is 1.6 μm, the waveguide spacing (d) is 4 μm, and the gap width is varied in the range of 1.6 μm to 2.8 μm. According to the study, the B / R ratio is the gap filling relative to the waveguide height h. It was shown that this strongly depends on the ratio of the height geh, and also strongly depends on the gap width gw. Therefore, Figure 12A(1) quantifies the change in the B / R ratio associated with the change in gap width, with the horizontal axis representing the ratio of the gap embedding height to the waveguide height, and the vertical axis representing the gap width when it is 1.6 The difference between the maximum and minimum B / R ratios in the 2.8 μm range is plotted. The following can be seen from Figure 12A(1).

[0082] (4) The difference between the maximum and minimum B / R ratio is smaller when the gap is filled than when it is not. Furthermore, the difference between the maximum and minimum B / R ratio decreases monotonically from 0 to 0.3 as the ratio of gap filling height to waveguide height increases, and above 0.3, the difference between the maximum and minimum B / R ratio becomes small and almost constant.

[0083] (5) The same effect as the difference between the maximum and minimum values ​​of these B / R ratios is also obtained in the case of the standard deviation of the B / R ratio obtained by changing the gap width shown in Figure 12A(2).

[0084] From the above results, it can be seen that in an optical directional coupler with a core cross-section in which a portion of the gap between waveguides is embedded, the magnitude of the variation in the B / R ratio due to the variation in gap width is such that the difference between the maximum and minimum values ​​of the B / R ratio decreases monotonically, and the ratio of the gap embedding height to the waveguide height is small and almost constant at 0.3 or higher.

[0085] According to Figure 12A and Figure 12B (described later), the ratio of the gap embedding height geh to the waveguide height h in the core cross-section, geh / h, is preferably in the range of 0.02 to 0.5, which is significantly different from the case where the gap embedding height is 0. If the ratio geh / h is 0.02 or higher, it is significantly different from the case where the gap embedding height is 0, and the range of variation in the B / R ratio becomes significantly narrower. 0.02 is the value of the ratio geh / h that corresponds to 3.8, which is about 90% of the difference between the maximum and minimum values ​​of the B / R ratio, 4.2. Also, if the ratio geh / h is 0.5 or lower, In an optical directional coupler, light propagating in one waveguide can be transferred to the other waveguide with almost 100% efficiency. Beyond this upper limit, the rate of light transfer deteriorates.

[0086] Alternatively, the ratio geh / h is preferably in the range of 0.02 to 0.9, which is significantly different from the case where the gap filling height is 0. The above applies to ratios geh / h of 0.02 or higher. The above applies to ratios geh / h of 0.9 or lower. If sufficient gap exists, light propagating through one waveguide can transfer to the other waveguide. If this upper limit is exceeded, the gap is filled too much, effectively merging the two waveguides, and the transfer of light becomes impossible.

[0087] The results shown in (4) and (5) above have also been obtained for directional couplers of representative structures other than those shown here. Figure 12B(3) shows the relationship between the standard deviation of the B / R ratio and the ratio of the gap filling height to the waveguide height. In the figure, G1 represents the structure described above (refractive index difference between core and cladding: 0.8%, waveguide height (h) :1.6μm, waveguide spacing (d):4μm), G2 is defined as having a refractive index difference between the core and cladding of 0.8%, a waveguide height (h) of 1.2 μm, and a waveguide spacing (d) of 4 μm. G3 has a refractive index difference between the core and cladding of 1.1%, waveguide height (h) of 1.6 μm, and between waveguides. This is the case where the gap (d) is 3.4 μm.

[0088] In all cases, as shown by the guide lines, the trend of the standard deviation of the B / R ratio changes when the ratio of gap embedding height to waveguide height is 0.3. From 0 to 0.3, the difference in the standard deviation of the B / R ratio decreases monotonically, and above 0.3, the difference in the standard deviation of the B / R ratio becomes small and almost constant. The same trend is obtained for the difference between the maximum and minimum values ​​of the standard deviation, indicating that the results shown in (4) and (5) above are not only valid for the special structure of directional couplers, but are universal properties that hold true for general directional tangential couplers.

[0089] In this example, where the gap between waveguides is filled with core material, the structure is the same as the two-stage core cross-section described above. That is, it corresponds to the case where the lower core of the two-stage core is connected between the two waveguides.

[0090] The three examples shown above (the example of a two-stage core cross-section, the example of a trapezoidal cross-section with one side vertical and the other oblique, and the example where the gap between waveguides is filled) are representative examples of optical waveguides with asymmetric core cross-sections in the height direction, including the usual trapezoidal shape. It can be seen that these optical waveguides have the advantage of smaller B / R ratio fluctuations compared to rectangular cross-sections, even if there is variation (fluctuation) in the gap width due to gap miniaturization. It should also be noted that, importantly, even if two or more of the above structures that are superior to rectangular cross-sections are combined, the superior characteristics compared to rectangular cross-sections are naturally maintained. For example, even when the gap between waveguides in a directional coupler consisting of trapezoidal cross-section waveguides is filled with core material, the superior characteristics compared to rectangular cross-sections are maintained. Thus, using optical waveguides with asymmetric core cross-sections in the height direction inherently reduces the magnitude of B / R ratio fluctuations compared to rectangular cross-sections.

[0091] [Relationship between gap width variation and assimilar characteristics in the assimilar manufacturing process] Figure 13 shows the relationship between the B / R ratio and the gap width (gw) for a typical directional coupler whose characteristics were shown in Figure 5, in both trapezoidal and rectangular cross-sections. The B / R ratio at this point is the maximum B / R ratio, and represents the maximum B / R ratio for each gap width (gw). This refers to the B / R ratio value when this occurs. As shown in Figure 5, the maximum value of this B / R ratio is the maximum B / R ratio obtained when the top / base ratio falls within the range of a trapezoid (the range where the top / base ratio is 0.95 or less and greater than 0). The maximum B / R ratio within the range where the top / base ratio of the trapezoidal cross-section is 0.95 or less and greater than 0 is selected as a representative value, and the variation (variation) of the gap width is considered. This shows the variation in the B / R ratio relative to ). On the other hand, the B / R ratio in a rectangular cross-section is the B / R ratio when the top / bottom ratio in Figure 5 is 1.

[0092] From Figure 13, the trapezoidal cross section has a B / R ratio gap width (gw) compared to the rectangular cross section. The dependence is clearly small, and even if there is manufacturing variation (fluctuation) in the gap width due to gap miniaturization, the magnitude of the B / R ratio fluctuation is smaller in the case of a trapezoidal cross section compared to a rectangular cross section. As a result, it can be inferred that the manufacturing variation in the multiplexing characteristics of industrially important waveguide-type multiplexers is reduced.

[0093] [Regarding the manufacturing process of multiplexers and variations (fluctuations) in gap width] Here, we will discuss the relationship between the variation (fluctuation) in gap width associated with gap miniaturization and the manufacturing process of the multiplexer. Multiplexers are typically manufactured by the following process. (1) A first SiO2 glass film is deposited on a Si substrate or SiO2 substrate, followed by a second SiO2 glass film containing a dopant to increase the refractive index. The second SiO2 glass film containing the dopant is etched using a mask pattern formed by exposure and development of a resist applied on the second SiO2 glass film, except for the portion that will become the waveguide core. (2) Subsequently, a third SiO2 glass film is deposited as an overcladding layer, forming a multiplexer consisting of a waveguide core made of second SiO2 glass containing a dopant to increase the refractive index, surrounded by the first and third SiO2 glass cladding layers.

[0094] In this case, minute variations in the etching conditions of the second SiO2 glass containing the dopant cause variations, or inconsistencies, in the gap width between waveguides in the directional coupler of the multiplexer.

[0095] However, although the gap width between waveguides varies, the distance between the centers of the two waveguides forming the directional coupler (waveguide spacing (d)) is determined solely by the mask pattern and does not depend on minute variations in etching conditions. Therefore, the gap width (gw) dependence of the characteristics shown in Figures 5, 7, 8, 9, 10, 11, 13, and Figure 16 (described later) can be determined by fixing the waveguide spacing (d). They are looking at it.

[0096] However, the characteristics shown in Figures 5, 7, 8, 9, 10, 11, 13, and 16 are based on this waveguide spacing. (d) shows the same trend even if there are some fluctuations. Therefore, the results obtained in these figures are naturally valid even for multiplexers manufactured using manufacturing processes other than those described in (1) and (2) above. For example, this applies to all cases where a structure is formed in close proximity to the other waveguides, such as when the core is formed by compression molding of a second SiO2 glass film using a nanoimprint mold.

[0097] In compression molding using molds such as nanoimprint, a core shaped like a mountain peak is created by pressing a mold with the core shape of the waveguide onto an SiO2 glass core layer. Variations in the waveguide spacing d and gap width gw during mold manufacturing and when pressing the mold onto the glass core layer result in variations in the gap width between waveguides.

[0098] Furthermore, here we have shown an example where a second SiO2 glass core containing a dopant to increase the refractive index is completely surrounded by the first and third SiO2 glass cladding materials (channel-type optical waveguide). However, naturally, even if the second SiO2 glass core is not completely surrounded by the SiO2 glass cladding material, if the light is confined within the core, it will exhibit the same waveguide cross-sectional shape dependence. One example of this is the ridge-type waveguide. In a ridge-type waveguide, a first SiO2 glass film and a second SiO2 glass film containing a dopant to increase the refractive index are formed, and both glass films are etched into a mountain-back shape using a mask pattern to form a mountain-back shaped core. No cladding layer is formed on the sides and top of the core, and it is in contact with the air. Therefore, Even in the case of ridge-type waveguides, slight variations in manufacturing conditions can cause fluctuations in the gap width between waveguides in the directional coupler of the multiplexer.

[0099] Furthermore, while this example shows a waveguide multiplexer with a core made of SiO2 glass containing a dopant to increase the refractive index, surrounded by an SiO2 glass cladding, the core and cladding are not limited to these materials. For example, the refractive index of the core only needs to be greater than that of the cladding, and the composition of the glass is not limited to (1) SiO2 glass containing a dopant to increase the refractive index or (2) SiO2 glass for the cladding. Also, while this example shows a material mainly composed of SiO2 glass, other materials such as organic compounds like plastics may also be used.

[0100] Here, we discuss the advantages of a cross-sectional shape that is asymmetrical in the height direction, including trapezoids, in relation to the relationship between red and blue light, among the three primary colors of light. However, similar advantages regarding cross-sectional shapes exist for red and green, and green and blue light. However, as explained later in the section on improving the efficiency of mode coupling in the directional coupler itself, the wavelength difference between the two colors is smaller in the case of red and green, and green and blue light compared to the case of red and blue light. Therefore, the improvement effect on the overall characteristics of the multiplexer is relatively smaller compared to the case of red and blue light. For this reason, the improvement in the characteristics of the directional coupler in the case of red and blue light will mainly contribute to the improvement of the overall characteristics of the multiplexer.

[0101] [Relationship between B / R ratio and the difference between the length of the base and the length of the top of a trapezoid] Up to this point, we have evaluated the B / R ratio of a multiplexer using the top / base ratio as the cross-sectional shape of the waveguide forming the multiplexer. Next, we will evaluate the B / R ratio of the multiplexer from a different perspective, using the difference between the length of the base and the length of the top of a trapezoid as an indicator of the cross-sectional shape of the waveguide.

[0102] Figure 14 shows the difference between the maximum and minimum B / R ratios of the directional coupler, and (the length of the base and top of the trapezoid). This shows the relationship between (difference in length) / (length of trapezoid base). According to Figure 6A(3), the top side is (w-2x) and the base side is w, so the horizontal axis is (difference between the length of the trapezoid base and the length of the top side) / (length of the trapezoid base) = 2x / w. The directional coupler used here is the same as the typical directional coupler whose characteristics are shown in Figure 5. Similarly, the gap width (gw) as a parameter. The fluctuation range and the fixed waveguide spacing (d) were assumed to be the same.

[0103] On the horizontal axis of Figure 14, the position where 2x / w=0 corresponds to a rectangular cross-section, the region where 2x / w is greater than 0 corresponds to a trapezoidal cross-section, and the region where 2x / w is less than 0 corresponds to an inverted trapezoidal cross-section where the top side is greater than the base side. From Figure 14, it can be seen that the difference between the maximum and minimum values ​​of the B / R ratio becomes significantly smaller as the cross-section changes from rectangular to trapezoidal. This is shown by guide line 20. This result also leads to the conclusion that, similar to the case examined using the top / base ratio in Figures 7-12, directional couplers formed using a trapezoidal cross-section waveguide exhibit a smaller variation in the B / R ratio with respect to gap width variations than rectangular cross-section waveguides, thus reducing the variation in the multiplexing characteristics of industrially important waveguide-type multiplexers.

[0104] [The second challenge is the efficiency of mode coupling of the directional coupler itself (multiplexing performance of the optical directional coupler)] [A second embodiment of an optical waveguide multiplexer that improves upon the above.] The first embodiment describes a configuration that improves the gap width variation caused by variations in the manufacturing process of an optical waveguide multiplexer. In the second embodiment, the efficiency of mode coupling of the directional coupler itself (the multiplexing performance of the optical directional coupler) can be improved by making the core shapes of the two optical waveguides in the multiplexing section asymmetric in the height direction, such as by making them trapezoidal. Below, we describe the structure of a directional coupler as the multiplexing section in an optical waveguide multiplexer that can improve the characteristics of the optical waveguide multiplexer itself.

[0105] Figure 15 shows a typical directional coupler whose characteristics were shown in Figure 5 (refractive index of core and cladding). The difference is 0.8%, with a fixed waveguide height (h) of 1.6 μm and a fixed waveguide spacing (d) of 4 μm. This diagram shows the relationship between the coupling length and the top / bottom ratio for the red and blue sections, respectively. In this case, the gap width (gw) is 1.9 μm and the waveguide width (w) is 2.1 μm.

[0106] As can be seen from Figure 15, in the range of 0 to 1 for the top / bottom ratio (the trapezoidal range), the coupling length of blue B increases significantly, while the coupling length of red R does not increase as much and remains low. This means that the length of the waveguide over which blue light propagating in one waveguide can transfer to the adjacent waveguide (the blue coupling length) increases. The reason for this is thought to be that the field of red light propagating in a waveguide is wide, while the field of green light is narrow. Therefore, if the core cross-section of the waveguide is trapezoidal, the field of blue light with a shorter wavelength is confined near the bottom, making the field spread narrower, but the field spread of red light with a longer wavelength is only limited by the trapezoidal shape.

[0107] In the case of the transfer of red and blue light shown in Figure 3(1), the combined wave characteristics of the optical directional coupler can be improved such that the red light R is completely transferred from waveguide WG1 to WG2, while the blue light B is hardly transferred from waveguide WG2 to WG1.

[0108] Figure 16 plots a typical directional coupler, whose characteristics were shown in Figure 5, with the waveguide spacing (d) on the horizontal axis, the maximum B / R ratio on the vertical axis, and the gap width (g) as a parameter. Similar to Figure 13, the maximum B / R ratio is the value of the B / R ratio that is maximized for each gap width (gw), and this maximum value is obtained when the top / bottom ratio falls within the range of a trapezoid.

[0109] As can be seen from Figure 16, the maximum B / R ratio depends on the waveguide spacing (d) and spans a wide range from around 3 to about 9. On the other hand, the dependence on the gap width (gw) as a parameter is small. From this, it can be seen that the way red and blue light transfer as shown in Figures 3(2) and (3) is achievable when the waveguide cross-section is trapezoidal, and has the advantage of being less dependent on the gap width.

[0110] To explain in more detail, in Figure 3 (2), blue light B transfers once while red light R transfers four times. Therefore, if the B / R ratio is around 4, the ratio of the coupling lengths LB and LR between blue light B and red light R becomes LB:LR = 4:1, making this type of transfer possible.

[0111] Furthermore, in Figure 3 (3), since blue light transfers once for every three transfers of red light, if the B / R ratio is near 3, the ratio of the coupling lengths LB and LR of blue light B and red light R becomes LB:LR = 3:1, making this type of transfer possible. In summary, it can be seen that both B / R ratios near 4 and near 3 fall within the range shown in Figure 16.

[0112] Furthermore, as mentioned above, in Figure 3(1), a larger B / R ratio is better, and in Figure 16, it is possible to make the B / R ratio sufficiently large, up to about 9. Therefore, all of the basic configurations of the optical waveguide type multiplexer, shown in Figures 3(1), (2), and (3), can be realized with a trapezoidal waveguide.

[0113] While the advantages of an optical waveguide with a core cross-section that is asymmetrical in the height direction (trapezoidal) have been described for the relationship between red and blue light, as previously stated, similar advantages exist for red and green light, and for green and blue light.

[0114] [Propagation modes of waveguides forming directional couplers] Up to this point, we have mainly shown examples of optical directional couplers using waveguides that propagate single-mode light. However, the waveguides used here are not strictly limited to those that propagate only single-mode light. Even in waveguides capable of propagating lower-order modes of light other than the fundamental mode, if the dominant mode is the fundamental mode (the mode with the highest light intensity among the propagating modes is the fundamental mode), the multiplexing characteristics will be slightly degraded, but all the multiplexing characteristics and effects described so far will still apply, just as in the example using a waveguide that propagates single-mode light. Furthermore, even if one of the waveguides in an optical directional coupler is a waveguide that propagates multi-mode light, the mechanism of light transfer is basically the same as in the case of a single-mode waveguide, and similarly, all the results described so far will still apply.

[0115] Next, examples of the first and second embodiments will be described.

[0116] [Example 1: An example of a multiplexer in Figure 1 with a trapezoidal core cross-section] Example 1 is a three-primary-color optical waveguide type multiplexer shown in Figure 1. The transfer operation of the three light waves in the multiplexing section is as described in Figure 1, but further details are as follows. In the first multiplexing section 110, the blue light B that has propagated through the third optical waveguide 103 is separated into the third and second optical waveguides. That is, because the length L1 of the mode coupling region of the first multiplexing section 110 is half the distance over which the blue light completely transfers to the second waveguide 102, the blue light is separated into the third and second optical waveguides in 50% in equal proportions. The separated blue light then begins to transfer again to the second waveguide 102 in the third multiplexing section 130, and eventually all of the blue light propagates into the second waveguide. In the case of blue light, the optical directional coupler is interrupted midway due to the presence of the second multiplexer 120, but the first multiplexer 110 and the third multiplexer 130 can be considered together as a single directional coupler. On the other hand, since the magnitude of mode coupling for green light G is twice that of blue light, it can completely transfer to the adjacent third waveguide 103 with only the length L1 of the mode coupling region of the first multiplexer 110.

[0117] In the third multiplexer 130, the red light R moves from the second waveguide 102 to the third waveguide 103, and then back to the second waveguide 102. That is, the length L of the mode coupling region of the third multiplexer 130. 3 is twice the coupling length required for red light to transfer to the adjacent optical waveguide. Therefore, the red light R transfers from the second waveguide 102 to the third waveguide 103, then returns from the third waveguide to the second waveguide, and finally propagates back through the original second waveguide 102. This transfer of light is based on the fact that the strength of the mode coupling is approximately red:green:blue = 4:2:1. The strength of this mode coupling increases as the wavelength of light increases. This is because the spreading (leakage) of light propagating from the optical waveguide is wavelength-dependent.

[0118] The substrate 111 is made of Si, the cladding layer 112 is made of SiO2 glass, and the cores constituting the optical waveguides 101, 102, and 103 are made of SiO2 glass with a higher refractive index than the cladding layer. The first to third waveguides 101 to 103 are waveguides for the fundamental mode or the pseudo-fundamental mode, in which the propagation of the fundamental mode is the largest component.

[0119] The first multiplexer 110, the second multiplexer 120, and the third multiplexer 130 are directional couplers. The wavelength of red light is 0.633 μm, the wavelength of green light is 0.521 μm, and the wavelength of blue light is 0.448 μm. Waveguide 101, Waveguide 202, and Waveguide 302 The waveguide width (w) of 03 is 2.1 μm, and the waveguide height (h) is 1.6 μm. The waveguide spacing (d) is The thickness is 3.2 μm in the first and third multiplexing sections 110 and 130, and 4 μm in the second multiplexing section 120. Furthermore, the core cross-section of at least the mode-coupled waveguide among the first to third waveguides is trapezoidal (top / base ratio of 0.75). The core cross-sections of the first to third waveguides may also be trapezoidal (top / base ratio of 0.75). The refractive index difference between the core and the cladding is 0.8%.

[0120] This multiplexer, with its trapezoidal core cross-section, reduces the overall length of the multiplexer by approximately 10% compared to multiplexers using optical waveguides with a rectangular core cross-section, resulting in improved mode coupling efficiency. Furthermore, it improved the yield in multiplexer manufacturing by approximately 20%.

[0121] [Example 2: An example of a multiplexer in Figure 1 with a core cross-section consisting of two stacked rectangular sections of different widths.] Example 2 is an example of a three-primary-color optical waveguide type multiplexer shown in Figure 1. The structure of the multiplexer in Example 2 is the same as in Example 1, except for the cross-sectional structure of the waveguides. The first to third waveguides 101 to 103 are also the same as in Example 1, being waveguides for the fundamental mode or the pseudo-fundamental mode, where the propagation of the fundamental mode is the maximum component. The core cross-section of the waveguide in Example 2 is a shape of two stacked rectangles of different widths, as shown in Figure 6(5), where the width of the upper core is half the width of the lower core, and the heights of the upper and lower cores are the same. The waveguide height (h) is 1.6 μm (sum of the heights of the upper and lower cores), and the waveguide spacing (d) is 3.2 μm between the first multiplexer 110 and the third multiplexer 130. The gap width is μm, and 4 μm in the second multiplexing section 120. Each multiplexing section consists of a directional coupler. The gap width (gap width between the upper cores) is 0.8 μm between the first multiplexing section 110 and the third multiplexing section 130. The second multiplexing section 120 is 1.9 μm. The top / bottom ratio is 0.85. The core cross-section of at least the mode-coupled region waveguide among the first to third waveguides may be in the shape of two stacked rectangles of different widths.

[0122] This multiplexer reduces the overall length of the multiplexer by approximately 8% compared to multiplexers using optical waveguides with a rectangular core cross-section, resulting in improved mode coupling efficiency. Furthermore, it improved the yield in multiplexer manufacturing by approximately 30%.

[0123] [Example 3: An example of a trapezoidal cross-section with one side angled, using the multiplexer shown in Figure 1] Example 3 is an example of a three-primary-color optical waveguide type multiplexer shown in Figure 1. The structure of the multiplexer in Example 3 is the same as in Example 1, except for the cross-sectional structure of the waveguides. The core cross-section of the waveguide in Example 3 is a trapezoidal trapezoid with one side vertical and the other oblique, as shown in Figure 6(6), and the oblique sides face each other between the waveguides. First waveguide 101, second waveguide 102, and third waveguide 1 The waveguide width (w) of 03 is 2.1 μm, and the waveguide height (h) is 1.6 μm. The waveguide spacing (d) is The thickness is 3.2 μm in the first and third multiplexing sections 110 and 130, and 4 μm in the second multiplexing section 120. Furthermore, the top / bottom ratio of the waveguide cross-section is 0.75. The core cross-section of at least the waveguide in the mode coupling region of the first to third waveguides may be a trapezoidal shape with one side sloped.

[0124] This multiplexer reduces the overall length of the multiplexer by approximately 5% compared to multiplexers using optical waveguides with a rectangular cross-section, resulting in improved mode coupling efficiency. Furthermore, it improved the yield in multiplexer manufacturing by approximately 15%.

[0125] [Example 4: An example of a directional coupler in which the gap between waveguides is filled in the multiplexer shown in Figure 1] Example 4 is an example of a three-primary-color optical waveguide type multiplexer shown in Figure 1. The structure of the multiplexer in Example 4 is the same as in Example 1, except for the cross-sectional structure of the waveguide. The core cross-section of the waveguide in Example 4 has the shape in which the gap between waveguides shown in Figure 6(7) is filled. First waveguide 10 1. The waveguide width (w) of the second waveguide 102 and the third waveguide 103 is 2.1 μm, and the waveguide height (h) is 1.6 μm. The waveguide spacing (d) is 3.2 μm between the first multiplexer 110 and the third multiplexer 130, and 4 μm in the second multiplexer 120. The ratio of the gap filling height to the waveguide height is 0.2. The core cross-section of at least the waveguides in the mode coupling region of the first to third waveguides may have a shape in which the inter-waveguide gap is filled.

[0126] This multiplexer reduces the overall length of the multiplexer by approximately 15% compared to multiplexers using optical waveguides with a rectangular cross-section, resulting in improved mode coupling efficiency. Furthermore, it improved the yield in multiplexer manufacturing by approximately 30%.

[0127] [Example 5: An example of a multiplexer in Figure 2 with a trapezoidal core cross-section] Example 5 is an example of a three-primary-color optical waveguide type multiplexer shown in Figure 2. In this case, the multiplexer is formed from the first to third waveguides 101 to 103 and the first and second multiplexing sections 110 and 120. The first to third waveguides 101 to 103 are for the fundamental mode or when the propagation of fundamental mode light is at its maximum. These are waveguides for the pseudo-fundamental modes, which are the components of the optical spectrum. These optical waveguides are embedded within a cladding layer 112 formed on a substrate 111. The substrate 111 is made of Si, and the cladding layer 112 is made of SiO2 glass. The transfer operation at the intersection of the three optical waves is as explained in Figure 2.

[0128] Here, the first multiplexer 110 and the second multiplexer 120 are directional couplers. The wavelength of red light is 0.633 μm, the wavelength of green light is 0.521 μm, and the wavelength of blue light is 0.448 μm. The waveguides of the first waveguide 101, the second waveguide 102, and the third waveguide 103 The path width (w) is 2.1 μm, and the waveguide height (h) is 1.6 μm. The waveguide spacing (d) is 3.2 μm between the first multiplexing section 110 and the second multiplexing section 120. The waveguide cross-section is trapezoidal (top / base ratio is 0.75). At least the core cross-section of the waveguide in the mode coupling region of the first to third waveguides must be trapezoidal. The refractive index difference between the core and the cladding is 0.8%.

[0129] This multiplexer improved the yield rate in multiplexer manufacturing by approximately 20% compared to multiplexers using optical waveguides with a rectangular cross-section.

[0130] [Example 6: An example of a multiplexer in Figure 18 with a trapezoidal core cross-section] Figure 18 is a plan view showing an example of a three-primary-color optical waveguide type multiplexer of Example 6. In this multiplexer 300, the planar structure of the first to third waveguides 101 to 103 and the first and second multiplexing sections 110 and 120 is the same as in Figure 2. Also the same as in Figure 2 is that the first to third waveguides 101 to 103 are waveguides for the fundamental mode or the pseudo-fundamental mode, where the propagation of the fundamental mode is the maximum component. However, unlike in Figure 2, in this example, green G, blue B, and red R are incident on the first to third waveguides 101 to 103, respectively. Also, the side view of the left end of the substrate 111 is the same as in Figure 2, but the core cross-section of each waveguide is trapezoidal. These optical waveguides are embedded in a cladding layer 112 formed on the substrate 111. The substrate 111 is made of Si, and the cladding layer 112 is made of SiO2 glass.

[0131] Furthermore, the blue light B incident on the incident end 102a of the second waveguide 102 travels straight through the first combiner 110, then moves to the first waveguide 101 in the second combiner 120, and exits from the exit end 101b of the first waveguide 101. The red light R incident on the incident end 103a of the third waveguide 103 travels straight through the first combiner In the wave section 110, the light propagates to the second waveguide 102 via mode coupling, repeatedly moves between the second waveguide 101 and the first waveguide 101 in the second multiplexing section 120, and finally moves back to the first waveguide 101 before emitting from the exit end 101b of the first waveguide 101. Green light G incident on the incident end 101a of the first waveguide 101 moves to the second waveguide 102 via mode coupling in the second multiplexing section 120, returns to the first waveguide 101 again in the second multiplexing section 120, and then emits from the exit end 101b of the first waveguide 101. Thus, all light is emitted from the same exit end 101b of the first waveguide 101.

[0132] The B / R ratio of the second wave combiner 120 is 3. The number of times red light R transfers in the second wave combiner 120 is 3, the number of times green light G transfers is 2, and the number of times blue light B transfers is 1. The light transfer in the first wave combiner 110 is shown in Figure 3(1), and the light transfer in the second wave combiner 120 is equivalent to that shown in Figure 3(3).

[0133] Here, the first multiplexer 110 and the second multiplexer 120 are directional couplers. The wavelength of red light is 0.633 μm, the wavelength of green light is 0.521 μm, and the wavelength of blue light is 0.448 μm. The waveguides of the first waveguide 101, the second waveguide 102, and the third waveguide 103 The path width (w) is 2.1 μm, and the waveguide height (h) is 1.6 μm. The waveguide spacing (d) is 4 μm at the first multiplexing section 110 and 3.2 μm at the second multiplexing section 120. The waveguide cross-section is trapezoidal (top / base ratio is 0.75). At least the core cross-section of the waveguide in the mode coupling region of the first to third waveguides must be trapezoidal. The refractive index difference between the core and cladding is 0.8%.

[0134] This multiplexer reduces the overall length of the multiplexer by approximately 10% compared to multiplexers using optical waveguides with a rectangular cross-section, resulting in improved mode coupling efficiency. Furthermore, it improved the yield in multiplexer manufacturing by approximately 20%.

[0135] [Example 7: An example of a multiplexer in Figure 19 with a trapezoidal core cross-section] Figure 19 is a plan view of the multiplexer of Embodiment 7. This multiplexer 400 is formed from first to third waveguides 101 to 103, a first multiplexing section 110, and a second multiplexing section 120. The first to third waveguides 101 to 103 are waveguides for the fundamental mode or the pseudo-fundamental mode, where the propagation of the fundamental mode is the largest component. Here, the first multiplexing section 110 and the second multiplexing section 120 are not simply two optical waveguides placed adjacent to each other, but rather multimode optical waveguides 105 and 106 with a large waveguide width are placed between the two optical waveguides. These multimode optical waveguides have the effect of improving the efficiency of mode coupling between the first multiplexing section 110 and the second multiplexing section 120.

[0136] An example of the wave multiplexing method in Example 7 is as follows: The incident end 103a of the third waveguide 103 The second visible light (usually green light G) incident on the first multiplexer 110 is transferred to the second waveguide 102 via the multimode optical waveguide 105 through mode coupling, and then continues straight through the second multiplexer 120, emitting from the exit end 102b of the second waveguide 102.

[0137] The first visible light (usually red light R) incident on the input end 101a of the first waveguide 101 is transferred to the second waveguide 102 via the multimode optical waveguide 106 by mode coupling in the second multiplexer 120 and propagates there, and is emitted from the output end 102b of the second waveguide 102. The third visible light (usually blue light B) incident on the ray end 102a is directed to the first multiplexer 110 and the second Mode coupling occurs in the multiplexer 120, but most of the light continues straight ahead and exits from the exit end 102b of the second waveguide 102. In this way, all the light is in the same second waveguide 102. It is fired from the exit end 102b.

[0138] In this case, the red light is transferred and the blue light travels in a straight line, as shown in Figure 3(1), and this operation occurs in the multiplexing section 120. However, since the red light is transferred via the multimode optical waveguide 106, the transfer is somewhat more complex, but in principle, it can be explained by the operation shown in Figure 3(1). From this, the first to third waveguides 101 to 103 and the multimode optical waveguides 105 and 106 By using an optical waveguide with a core cross-section that is asymmetric in the height direction, improvements in multiplexing characteristics can be obtained in the same way as the multiplexers shown in Figures 1, 2, and 18.

[0139] [Optical waveguide type multiplexer of the third embodiment] In the multiplexers shown in Figures 1, 2, 18, and 19, multiple multiplexers are arranged at different positions in the optical propagation direction (horizontal direction from left to right in the figures) from the incident side to the exit side. This results in a long optical propagation length of the multiplexer, making it difficult to miniaturize the multiplexer. In contrast, the optical waveguide type multiplexer of the third embodiment concentrates multiple multiplexers at one point in the optical propagation direction, thereby shortening the optical propagation length of the multiplexer.

[0140] [First multiplexer of the third embodiment] Figure 20 shows the configuration of the first optical waveguide multiplexer in the third embodiment. Figure 20 shows a plan view and a cross-sectional view of the first multiplexer 500 at a cross section CS1. This first optical waveguide multiplexer 500 has first to fourth waveguides 101 to 104 and first to third multiplexing sections 110 to 130. The first to fourth waveguides 101 to 104 are waveguides for the fundamental mode or the pseudo-fundamental mode, in which the propagation of the fundamental mode is the largest component. These multiplexing sections are usually formed by optical directional couplers. That is, the first multiplexing section 110 is a pair of waveguides of a predetermined length that are adjacent to each other via a gap width gw1 in the third waveguide 103 and the second waveguide 102. This is an optical directional coupler composed of waveguides with mode coupling regions 103c and 102c. The second multiplexer 120 consists of a pair of waveguides with mode coupling regions 102c and 101c of predetermined length that are adjacent to each other via a gap width gw2, in the second waveguide 102 and the first waveguide 101. This is a directional optical coupler. The third coupling section 130 consists of a pair of predetermined length molars adjacent to each other via a gap width gw3 in the waveguides of the first waveguide 101 and the fourth waveguide 104. This is an optical directional coupler composed of waveguides with coupling regions 101c and 104c.

[0141] Furthermore, the optical directional coupler in the multiplexing section has a waveguide with a pair of linear mode coupling regions of a predetermined length. These waveguides are arranged in close proximity with a predetermined gap width between them, and the predetermined length and gap width are selected so that the propagating light can transfer from one waveguide to the other. As mentioned above, the gap width between waveguides that light can transfer to and the predetermined length of the straight waveguide differ depending on the wavelength of the propagating light.

[0142] This first multiplexer 500 has a fourth waveguide 104 on the opposite side (lower side in the figure) of the first waveguide 101 from the second waveguide 102, in addition to the first to third waveguides 101 to 103, and the ranges of the optical propagation directions of the first to third multiplexing sections 110 to 130 overlap each other in a direction perpendicular to the optical propagation direction. As shown in the plan view of Figure 20, these optical waveguides are usually embedded in a cladding layer 112 formed on the substrate 111, similar to the multiplexers described above.

[0143] The multiplexing of the multiplexer 500 can be done as follows by adjusting, for example, the lengths of each multiplexing section 110-130. That is, in one example of a multiplexing method, the first waveguide 101 The first visible light (usually red light R) incident on the incident end 101a travels back and forth several times through the fourth waveguide 104, first waveguide 101, second waveguide 102, and third waveguide 103 through mode coupling of the three multiplexers: the third multiplexer 130, the second multiplexer 120, and the first multiplexer 110, and ultimately... The emission is emitted from the exit end 102b of the second waveguide 102.

[0144] The second visible light (usually green light G) incident on the incident end 102a of the second waveguide 102 is the first After propagating in the third waveguide 103 via mode coupling at the multiplexing section 110, the signal returns to the second waveguide 102 and propagates, departing from the exit end 102b of the second waveguide 102.

[0145] The third visible light (usually blue light B) incident on the incident end 103a of the third waveguide 103 is the first At the multiplexing section 110, mode coupling occurs, and the signal moves to the second waveguide 102, and is emitted from the exit end 102b of the second waveguide 102.

[0146] Furthermore, some of the second visible light (usually green light G) and third visible light (usually blue light B) are transferred to the first waveguide 101 and the fourth waveguide 104 via the second and third multiplexers 120 and 130, but the amount of transfer can be reduced, for example, by adjusting the length of each multiplexer. In this way, all colors of light are emitted from the same exit end 102b of the second waveguide 102.

[0147] In the optical multiplexer structure shown in the plan view of Figure 20(1), the ranges of the first to third multiplexing sections 110 to 130 in each optical propagation direction (horizontal direction in the figure) overlap in the direction perpendicular to the optical propagation direction (vertical direction in the figure). When this overlapping portion is cut by the cross-section plane CS1, simulations show that the multiplexing efficiency improves when the four optical waveguides 103, 102, 101, and 104 constituting the multiplexing section are made symmetrical in both the structure of the optical waveguides (height and width of the core cross-section) and the arrangement (waveguide spacing and gap width) in the cross-sectional view, as shown in the cross-sectional view of Figure 20(2). Therefore, in this multiplexer, it is desirable that the straight portion (mode coupling region) of the optical waveguides constituting the multiplexing section be symmetrical in the cross-sectional view, as shown in the cross-sectional view of Figure 20(2). However, even if the structure and arrangement of the optical waveguides are not perfectly symmetrical, for example, 95% symmetry is sufficient. Also, if the characteristics are not a concern, this symmetry is acceptable. It is not limited to structure.

[0148] Furthermore, among the four optical waveguides 103, 102, 101, and 104 that constitute the multiplexing section where the ranges of the optical propagation directions of the first to third multiplexing sections 110 to 130 overlap, if the gap width gw2 between the central optical waveguide 102 and 101 is made larger than the other gap widths gw1 and gw3, Simulations show that wave efficiency can be further improved. Therefore, it is desirable to make the gap width gw2 between optical waveguides 102 and 101 larger than the other gap widths gw1 and gw3. It's nice.

[0149] The technical significance of the first multiplexer 500 of the third embodiment having a fourth waveguide 104 is as follows, for example. When the multiplexer shown in Figures 1, 2, and 18 is composed of three waveguides corresponding to red, blue, and green, and multiple multiplexing sections (directional couplers) are overlapped at least partially in the longitudinal direction (optical propagation direction) of the multiplexer, it was not possible to emit multiplexed light from the output end of one waveguide. However, when the fourth waveguide was placed outside the three waveguides, it became possible to emit multiplexed light from the output end of one waveguide in a configuration where multiple multiplexing sections were overlapped in the optical propagation direction.

[0150] Specifically, in a multiplexer composed of three waveguides, even if the multiplexing sections 110 and 120 partially overlap as shown in Figure 20, the blue and green waves can be combined, and from the output end 102b of the second waveguide... The combined light is emitted. However, red cannot be combined into blue and green combined light, and the light other than at the emission end 102b is emitted. The light would otherwise be emitted from the exit end. Therefore, by adding a fourth waveguide 104 outside the first waveguide into which the red light is incident, and forming a multiplexer 130, the red light propagating in the first waveguide 101 splits into the second waveguide 102 and the fourth waveguide 104, and then returns to the first waveguide, repeating this process in the multiplexers 120 and 130. Ultimately, most of the red light will be emitted from the exit end 102b. .

[0151] Figure 21 shows the simulation results of the propagation of three colors of light in the first multiplexer 500 of the third embodiment. The blue light B incident on the third waveguide 103 is transferred to the second waveguide 102 in the first multiplexer 110, propagates there, and is emitted from the exit end 102b. This is the same as the transfer of blue B in 3(2). The green light G incident on the second waveguide 102 temporarily transfers to the third waveguide 103 at the first multiplexer 110, then transfers to the second waveguide 102, propagates there, and is emitted from the exit end 102b. Coupling length based on the wavelengths of blue and green light This phenomenon is due to the difference between them.

[0152] On the other hand, the red light R incident on the first waveguide 101 repeatedly transfers to the fourth waveguide 104 in the third multiplexer 130, and then gradually transfers to the second and third waveguides 102 and 103, which are located a little further away in the second multiplexer 120, and finally transfers to the second waveguide 102 (at this position, the fourth waveguide 104 has disappeared or is far away, and the third multiplexer 130 does not exist). This transfer of red light R is due to the fact that, as mentioned above, the linear portions (mode coupling regions) of the optical waveguides constituting the three multiplexers within the multiplexer have a symmetrical structure and arrangement in the cross-sectional view, so that the propagation mode of the four waveguides as a whole can be represented by a first-order combination (addition) of the independent and unique propagation modes (fundamental modes) of each optical waveguide, and ultimately becomes the unique propagation mode of a single waveguide 102.

[0153] Furthermore, by making the gap width gw2 between the first waveguide 101 and the second waveguide 102 larger than the other gap widths gw1 and gw3, it is possible to prevent blue and green light, which have shorter wavelengths than red light, from transferring to the first waveguide 101 and the fourth waveguide 104 via the second combiner 120 as much as possible. Blue and green light, which have shorter wavelengths than red light, have less light leaking laterally from the waveguide (evanescent light), making it less likely for them to transfer through the enlarged gap width gw2. In this way, the gap width gw2 between the first waveguide 101 and the second waveguide 102 is set to the other gear It is desirable to make the top width larger than gw1 and gw3 due to the characteristics of optical combined waves.

[0154] Regarding the left-right symmetry of the waveguides in the overlapping multiplexing section mentioned above, the multiplexer is composed of three waveguides, and the waveguide distance between the first waveguide 101 and the second waveguide 102 is set to the second waveguide If the distance between waveguide 102 and the third waveguide 103 is made larger, the symmetry is broken. It can be said that the fourth waveguide 104 was added to resolve this symmetry breakdown.

[0155] The technical significance of adjusting the length of each multiplexer in the third embodiment is as follows: As shown in the propagation of red light in Figure 21, the length of the third multiplexer 130, where the red light incident on the third waveguide transfers to the second waveguide 102 for emission, is determined, and at the same time, the lengths of the first and second multiplexers 110 and 120, where the blue and green light also transfer to the second waveguide 102 for emission, are determined. The length of each multiplexer can be determined by simulation. It can also be determined by numerical calculation using mode coupling theory, but the calculation becomes extremely complex.

[0156] In Figure 20, only the straight portion of the fourth waveguide 104 exists within the multiplexer, but of course, on the inlet or outlet side of the fourth waveguide 104, there is an S-shaped waveguide consisting of a curve and an inlet / outlet Waveguides may be connected to form input and output ends, similar to the other optical waveguides 101, 102, and 103. The input and output sides are shown by dashed lines in Figure 20. By providing an S-shaped waveguide and an output waveguide on the output side of the fourth waveguide 104, the small amount of red light propagating through the fourth waveguide 104 can be reduced. This suppresses the propagation of colored light as stray light within the cladding layer. Furthermore, it is possible to introduce another light source from the incident end of the waveguide on the dashed side of the fourth waveguide 104.

[0157] Furthermore, the core cross-sectional structure of optical waveguides 103, 102, 101, and 104 may be symmetrical in the height direction, such as having a rectangular core cross-section, or the core cross-section may be asymmetrical in the height direction to improve the characteristics. In this case, the operation in Figure 3(2) where the red light returns to its original state and the blue light takes over occurs in the first multiplexer 110. However, since the red light takes over to the second optical waveguide 102 via the third and second multiplexers 130 and 120, the transfer is somewhat more complex, but in principle, it can be explained by the operation shown in Figure 3(2).

[0158] In the first multiplexer 500 of the third embodiment shown in Figure 20, the length of the entire multiplexer can be shortened compared to the multiplexers shown in Figures 1 and 2, because the longitudinal ranges of all the multiplexing sections overlap.

[0159] [Second multiplexer of the third embodiment] Figure 22 shows the configuration of the second optical waveguide multiplexer in the third embodiment. Figure 22 shows a plan view and a cross-sectional view of the second multiplexer 600 at the cross-section CS2. This second optical waveguide multiplexer 600 has first to fourth waveguides 101 to 104A and first to third multiplexing sections 110 to 130, similar to the structure shown in Figure 20. The first to fourth waveguides 101 to 104A are waveguides for the fundamental mode or the pseudo-fundamental mode, where the propagation of the fundamental mode is the maximum component. These multiplexing sections are usually formed by directional couplers. This second multiplexer 600 has a fourth waveguide 104A in addition to the first to third waveguides 101 to 103, with red light incident on the fourth waveguide 104A, and there is no waveguide on the incident side of the first waveguide 101, and the ranges of the optical propagation directions of the first to third multiplexing sections 110 to 130 overlap. In other words, the second multiplexer 600 in Figure 22 differs from the configuration of the first multiplexer 500 in Figure 20 in that red light is not incident on the first waveguide 101, and red light R is incident on the fourth waveguide 104A. The pair of waveguides that make up the first to third multiplexing sections 110 to 130 are the same as those of the first multiplexer 500 in Figure 20. As shown in the cross-sectional view in Figure 22(2), these optical waveguides are usually embedded in a cladding layer 112 formed on the substrate 111, similar to the multiplexers described above.

[0160] The multiplexing of the second multiplexer 600 is done, for example, by adjusting the length of each multiplexing section 110-130. This can be done as follows. That is, in one example of a wave multiplexing method, the first visible light (usually red light R) incident on the incident end 104a of the fourth waveguide 104A travels back and forth several times through the fourth waveguide 104A, the first waveguide 101, the second waveguide 102, and the third waveguide 103 by mode coupling of the three wave multiplexers: the third wave multiplexer 130, the second wave multiplexer 120, and the first wave multiplexer 110. Ultimately, the signal is emitted from the exit end 102b of the second waveguide 102.

[0161] The second visible light (usually green light G) incident on the incident end 102a of the second waveguide 102 is the first After propagating in the third waveguide 103 via mode coupling at the multiplexing section 110, the signal returns to the second waveguide 102 and propagates, emitting from the exit end 102b of the second waveguide 102. Third waveguide 1 The third visible light (usually blue light B) incident on the incident end 103a of 03 is then processed in the first multiplexer 110. Mode coupling transitions to the second waveguide 102, and the signal is emitted from the exit end 102b of the second waveguide 102. In this way, some of the light of the second visible light (usually green light G) and the third visible light (usually blue light B) is transferred to the first waveguide 101 and the fourth waveguide 104 via the second and third multiplexers 120 and 130, but the amount of transfer can be reduced by adjusting the length of each multiplexer. In this way, light of all colors is emitted from the same exit end 102b of the second waveguide 102.

[0162] In the structure of the second multiplexer 600 shown in Figure 22, the ranges of the optical propagation directions of the first to third multiplexing sections 110 to 130 overlap. Furthermore, when the overlapping portion is cut along the cross-sectional plane CS2, simulations show that the multiplexing efficiency improves if the four optical waveguides 103, 102, 101, and 104A constituting the multiplexing section are symmetrical in both the structure of the optical waveguides (height and width of the core cross-section) and the arrangement (waveguide spacing and gap width), as shown in the cross-sectional view of Figure 22(2). For this reason, in this multiplexer 600 as well, it is desirable that the linear portions (mode coupling regions) of the four optical waveguides 101 to 104A constituting the first to third multiplexing sections are symmetrical, as shown in the cross-sectional view of Figure 22(2). However, even if the structure and arrangement of the optical waveguides are not perfectly symmetrical, for example, 95% symmetry is sufficient. Also, if the characteristics of multiplexing efficiency are not a concern, the structure is not limited to this symmetrical structure.

[0163] Furthermore, among the four optical waveguides 103, 102, 101, and 104A that constitute the overlapping section where the longitudinal ranges of the first to third overlapping sections 110 to 130 overlap, if the gap width gw2 between the central optical waveguide 102 and 101 is made larger than the other gap widths gw1 and gw3, Simulations show that wave efficiency can be further improved.

[0164] Figure 23 shows the simulation results of the propagation of three colors of light in the second multiplexer 600 of the third embodiment. The blue light B incident on the third waveguide 103 and the green light G incident on the second waveguide 102 are the same as the transfer of light in the first multiplexing section 110 and the second multiplexing section 120 of the first multiplexer 500 shown in Figure 21. On the other hand, the red light R incident on the fourth waveguide 104A repeatedly transfers light to the first waveguide 101 in the third multiplexing section 130, and then gradually transfers light to the second and third waveguides 102 and 103, which are located at a slightly distant position in the second multiplexing section 120 and the first multiplexing section 110, and finally transfers light to the second waveguide 102 (at this position, the fourth multiplexer 104 has disappeared or is far away, and the third multiplexing section 130 does not exist).

[0165] In Figure 22, the fourth optical waveguide 104A is connected to the incident waveguide and the S-shaped waveguide. The first optical waveguide 101 has an S-shaped waveguide and an output waveguide connected, but of course, the fourth The S-shaped waveguide and the output waveguide may also be connected to the output side of the optical waveguide 104A (dashed line in Figure 22). (See reference), in that case, it is possible to prevent the small amount of red propagating light propagating through the fourth optical waveguide 104A from becoming stray light within the cladding layer. Also, on the incident side of optical waveguide 101, the incident waveguide and the S-shaped waveguide are provided. They may be connected, or, like the other optical waveguides 102 and 103, they may form input and output ends.

[0166] Similarly, the second multiplexer 600 also combines the optical waveguides 103, 102, 101, and 104A. The cross-sectional structure may have a rectangular core cross-section that is symmetrical in the height direction, or, to improve the characteristics, the optical waveguide may have a core cross-section that is asymmetrical in the height direction. In this case, the operation in which the red light returns to its original state and the blue light is transferred, as shown in Figure 3(2), occurs in the first multiplexer 110. However, the red light incident on the fourth waveguide 104A is transferred to the second waveguide 102 via the third and second multiplexers 130 and 120, resulting in a somewhat more complex transfer process, but in principle, it can be explained by the operation shown in Figure 3(2).

[0167] [Third and fourth multiplexers of the third embodiment] Figure 24 shows the third and fourth optical waveguide multiplexers of the third embodiment. Figure 24(1) shows a plan view of the third optical waveguide multiplexer 700. The basic structure of this third optical waveguide multiplexer 700 is the same as that of the first multiplexer 500 shown in Figure 20, so no explanation is given. However, the third multiplexer 700 differs from the first multiplexer 500 in Figure 20 in that blue light B is incident on the second waveguide 102 and green light G is incident on the third waveguide 103, and the combined light of the three lights is emitted from the exit end 103b of the third waveguide 103.

[0168] The multiplexing of this third multiplexer 700 can be performed as follows by adjusting the length of each multiplexing section. That is, in one example of a multiplexing method, the first visible light (usually red light R) incident on the incident end 101a of the first waveguide 101 is combined with the third multiplexing section 130, the second multiplexing section 120 and The mode coupling of the three multiplexers of the first multiplexer 110 results in the fourth waveguide 104 and the first waveguide 1 The device travels back and forth through waveguide 01, the second waveguide 102, and the third waveguide 103 several times, and finally exits from the exit end 103b of the third waveguide 103.

[0169] The third visible light (usually blue light B) incident on the incident end 102a of the second waveguide 102 is the first At the multiplexer 110, the light is transferred to the third waveguide 103 by mode coupling and exits from the exit end 103b of the third waveguide 103. The second visible light (usually green light G) incident on the incident end 103a of the third waveguide 103 is transferred to the second waveguide 102 by mode coupling at the first multiplexer 110 and propagates there, then returns to the third waveguide 103 and propagates there, exiting from the exit end 103b of the third waveguide 103. ru.

[0170] Furthermore, some of the light of the second visible light (usually green light G) and the third visible light (usually blue light B) is transferred to the first waveguide 101 and the fourth waveguide 104 via the second and third multiplexers 120 and 130, but the amount of transfer can be reduced by adjusting the length of each multiplexer. In this way, light of all colors is transferred to the same third waveguide 103b or It will be fired.

[0171] Furthermore, in the case of the structure of the third multiplexer 700 shown in Figure 24(1), the overlapping range of the optical propagation direction of each multiplexer, the relationship between the structure and arrangement of each optical waveguide, the installation of input and output waveguides in the fourth waveguide, the cross-sectional structure of each waveguide, and the characteristics of the multiplexer are the same as in the case of the structure shown in Figure 20.

[0172] Figure 24(2) shows a plan view of the fourth optical waveguide type multiplexer 800. The basic structure of this fourth optical waveguide type multiplexer 800 is the same as that of the second multiplexer 600 shown in Figure 22, so no explanation is given. However, the fourth multiplexer 800 differs from the second multiplexer 600 in Figure 22 in that blue light B is incident on the second waveguide 102, green light G is incident on the third waveguide 103, and the combined light of the three lights is emitted from the exit end 103b of the third waveguide 103.

[0173] The multiplexing of this fourth multiplexer 800 can be performed as follows by adjusting the length of each multiplexing section. That is, in one example of the multiplexing method, the first visible light (usually red light R) incident on the incident end 104a of the fourth waveguide 104 is combined with the third multiplexing section 130, the second multiplexing section 120, and the first multiplexing section 110 through mode coupling, resulting in the fourth waveguide 104A and the first waveguide. The device travels back and forth through waveguide 101, the second waveguide 102, and the third waveguide 103 several times, and finally exits from the exit end 103b of the third waveguide 103.

[0174] The third visible light (usually blue light B) incident on the incident end 102a of the second waveguide 102 is the first At the multiplexer 110, the light is transferred to the third waveguide 103 by mode coupling and exits from the exit end 103b of the third waveguide 103. The second visible light (usually green light G) incident on the incident end 103a of the third waveguide 103 is transferred to the second waveguide 102 by mode coupling at the first multiplexer 110 and propagates there, then returns to the third waveguide 103 and propagates there, exiting from the exit end 103b of the third waveguide 103. ru.

[0175] Figure 25 shows the simulation results of the propagation of three colors of light in the fourth multiplexer 800 of the third embodiment. Green light G incident on the third waveguide 103 is transferred to the second waveguide 102 in the first multiplexer 110, then transferred to the third waveguide 103, and propagates and is emitted from there. Blue light B incident on the second waveguide 102 is transferred to the third waveguide 103 in the first multiplexer 110, and propagates and is emitted from there. Then, the red light R incident on the fourth waveguide 104A repeatedly transfers between the first waveguide 101 and the third waveguide 130, gradually transferring to the second and third waveguides 102 and 103, which are located at a slightly distant position in the second waveguide 120 and the first waveguide 110, and finally transferring to the third waveguide 103 (at this position, the fourth waveguide 104A has disappeared or is separated, and the third waveguide 130 does not exist).

[0176] Furthermore, some of the light of the second visible light (usually green light G) and the third visible light (usually blue light B) is transferred to the first waveguide 101 and the fourth waveguide 104A via the second and third multiplexers 120 and 130, but the amount of transfer can be reduced by adjusting the length of each multiplexer. In this way, light of all colors is transferred to the same third waveguide 103b exit end 103b It is emitted from there.

[0177] Furthermore, in the case of the structure of the fourth multiplexer 800 shown in Figure 24(2), the overlapping range of the optical propagation direction of each multiplexer, the relationship between the structure and arrangement of each optical waveguide, the installation of input and output waveguides in the first and fourth waveguides, the cross-sectional structure of each waveguide, and the characteristics of the multiplexer are the same as in the case of the structure shown in Figure 22.

[0178] Next, we will describe an example of the third embodiment.

[0179] This embodiment has the configurations of four types of multiplexers 500 to 800 described in Figures 20, 22, and 24, respectively, and has first to fourth waveguides 101 to 104, 104A, and first to third multiplexing sections 110 to 130. These optical waveguides are embedded in a cladding layer 112 formed on a substrate 111. The substrate 111 is made of silicon Si, and the cladding layer 112 is made of SiO2 glass. The first multiplexing section 110, the second multiplexing section 120, and the third multiplexing section 130 are optical This is a directional coupler. The wavelength of red light is 0.633 μm, the wavelength of green light is 0.521 μm, and the wavelength of blue light is 0.448 μm. Waveguide 101, Waveguide 202, The waveguide width (w) of the third waveguide 103 and the fourth waveguides 104 and 104A is 2.1 μm, and the waveguide height (h) is 1.6 μm. The waveguide spacing (d) is 3.2 μm between the first multiplexer 110 and the third multiplexer 130, and 4 μm in the second multiplexer 120. The refractive index difference between the core and the cladding is 0.8%.

[0180] The characteristics of each multiplexer fabricated using the above structure are as follows: (1) In the case of the first multiplexer 500 in Figure 20: When the waveguide cross-section is rectangular, a multiplexer length of 1.6 mm yields a 3-color average multiplexing efficiency of 70%. Furthermore, by using a trapezoidal waveguide cross-section, the yield in multiplexer manufacturing can be reduced by approximately 20%. We were able to improve it by %. (2) In the case of the second multiplexer 600 with the structure shown in Figure 22: When the waveguide cross-section is rectangular, a multiplexing efficiency of 87% (average of three colors) is obtained with a multiplexer length of 2 mm. Furthermore, by changing the waveguide cross-section to trapezoidal, the yield in multiplexer manufacturing was improved by approximately 20%. (3) In the case of the third multiplexer 700 with the structure of Figure 24(1): When the waveguide cross-section is rectangular, a multiplexer length of 1.8 mm yields an average multiplexing efficiency of 92% for three colors. Furthermore, by using a trapezoidal waveguide cross-section, the yield in multiplexer manufacturing was improved by approximately 20%. (4) In the case of the fourth multiplexer 800 with the structure of Figure 24(2): When the waveguide cross-section is rectangular, a multiplexer length of 1.5 mm yields an average multiplexing efficiency of 85% for three colors. Furthermore, by using a trapezoidal waveguide cross-section, the yield in multiplexer manufacturing was improved by approximately 20%. [Explanation of symbols]

[0181] 100~800: Optical waveguide type multiplexer 101~104, 104A: 1st to 4th optical waveguide 101a~104a: Incidence end 101b~103b: Output end 101c~104c: Waveguides in the mode-coupled region 110~140: 1st to 4th multiplexing section 111: Circuit board 112: Cladding layer gw: Gap width

Claims

1. The first optical waveguide and It has a second optical waveguide, The first and second optical waveguides each have a first mode-coupled region and a second mode-coupled region of a predetermined length adjacent to a predetermined gap width, The optical waveguides in the first and second mode coupling regions constitute a multiplexing section. In the multiplexing section, light is transferred between the optical waveguides of the first and second mode coupling regions. The first light in the optical waveguide of the first mode-coupled region and the second light in the optical waveguide of the first or second mode-coupled region, which has a different wavelength from the first light, are combined in the multiplexing section into one of the optical waveguides of the first and second mode-coupled regions. The core cross-sections of the optical waveguides in at least the first and second mode coupling regions within the first and second optical waveguides are asymmetric in the height direction. An optical multiplexer having at least one of the multiplexing sections.

2. The optical multiplexer according to claim 1, wherein the core cross-section of the optical waveguide in the first and second mode coupling regions has an upper side and a lower side, the lengths of the upper side and the lower side are different, and the ratio of the length of the upper side to the length of the lower side is 0.95 or less.

3. The optical multiplexer according to claim 2, wherein the core cross-section of the optical waveguide in the first and second mode coupling regions is one of the following: a trapezoid, a pseudo-trapezoid, a shape formed by stacking a plurality of rectangles of different widths, or a shape in which a portion of the gap between adjacent cores is filled with core material.

4. The optical multiplexer according to claim 3, wherein the pseudo-trapezoid is either a trapezoid in which one of the sides, top, or bottom is curved, or a trapezoid in which the interior angle between the top and side or the exterior angle between the side and bottom is rounded.

5. The first and second optical waveguides are optical waveguides in which the propagation of the fundamental mode of light is the maximum component. The optical multiplexer according to claim 1, wherein a multimode optical waveguide is arranged between the optical waveguide of the first mode coupling region and the optical waveguide of the second mode coupling region.

6. The second wavelength of the second light is shorter than the first wavelength of the first light. The optical multiplexer according to claim 1, wherein the first light moves from the optical waveguide of the first mode-coupled region to the optical waveguide of the second mode-coupled region, the second light travels straight through the optical waveguide of the second mode-coupled region, and the first and second light are combined into the optical waveguide of the second mode-coupled region.

7. The second wavelength of the second light is shorter than the first wavelength of the first light. The optical multiplexer according to claim 1, wherein the first light moves from the optical waveguide of the first mode-coupled region to the optical waveguide of the second mode-coupled region and then finally returns to the optical waveguide of the first mode-coupled region, the second light moves from the optical waveguide of the second mode-coupled region to the optical waveguide of the first mode-coupled region, and the first and second light are combined into the optical waveguide of the first mode-coupled region.

8. The second wavelength of the second light is shorter than the first wavelength of the first light. The optical multiplexer according to claim 1, wherein the first light moves from the optical waveguide of the first mode-coupled region to the optical waveguide of the second mode-coupled region, then returns to the optical waveguide of the first mode-coupled region, and finally moves to the optical waveguide of the second mode-coupled region, and the second light moves from the optical waveguide of the first mode-coupled region to the optical waveguide of the second mode-coupled region, and the first and second light are combined into the optical waveguide of the second mode-coupled region.

9. The optical multiplexer according to any one of claims 1 to 8, wherein the first light is red and the second light is blue.

10. A method for manufacturing an optical multiplexer according to claim 1, A method for manufacturing an optical multiplexer, comprising the step of patterning a core layer containing the core material into a first optical waveguide and a second optical waveguide.

11. A first optical waveguide having a first mode coupling region, A second optical waveguide adjacent to the first optical waveguide and having a second mode coupling region, A third optical waveguide is located on the opposite side of the first optical waveguide and adjacent to the second optical waveguide, and has a third mode coupling region, On the opposite side of the second optical waveguide and adjacent to the first optical waveguide, a fourth mode coupling region is provided. It has a fourth optical waveguide, The third mode coupling region and the second mode coupling region are arranged in close proximity over a first length with a first gap width to form a first multiplexing section. The second mode coupling region and the first mode coupling region are brought close together over a second length with a second gap width to form a second multiplexing section. The first mode coupling region and the fourth mode coupling region are connected by a third gear over a third length. A third wave-combining section is formed by arranging them in close proximity with a width of tip, A first light (R) having a first wavelength is incident on the first or fourth optical waveguide. A second light (G,B) having a second wavelength shorter than the first wavelength is incident on the second optical waveguide. A third light having a third wavelength shorter than the first wavelength and different from the second wavelength is incident on the third optical waveguide. The first to third light is combined into the second mode-coupled region of the second optical waveguide or the third mode-coupled region of the third optical waveguide, and emitted from the optical waveguide connected to the second or third mode-coupled region. An optical multiplexer in which the first to third multiplexing sections are arranged to overlap each other in the optical propagation direction from the incident side to the exit side of the first to third light.

12. The optical multiplexer according to claim 11, wherein in a cross-sectional view taken in a direction perpendicular to the optical propagation direction at the position where the first to third multiplexing sections overlap, the shape and arrangement of the first to fourth optical waveguides are symmetrical.

13. The gap widths of the first and third gaps are approximately equal. The optical multiplexer according to claim 11 or 12, wherein the second gap width is greater than the first and third gap widths.

Images