JP2014132314A - Spot size converter - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a spot size converter capable of suppressing reduction in optical coupling efficiency.SOLUTION: A spot size converter 1B includes a semiconductor substrate 3b, and a spot size conversion part 6 formed on the semiconductor substrate 3b. The spot size conversion part 6 includes: a silicon oxide layer S1 formed on the semiconductor substrate 3b; a silicon waveguide W1 formed on the silicon oxide layer S1; a silicon oxide layer S2 having a portion formed on the silicon waveguide W1 and a portion formed on the silicon oxide layer S1; a silicon waveguide W2 formed on the silicon oxide layer S1; a silicon oxide layer S3 having a portion formed on the silicon waveguide W2 and a portion formed on the silicon oxide layer S2. An enlarged region R10 including the silicon oxide layers S1-S3 is formed on a fourth region. The enlarged region R10 is optically coupled to the silicon waveguide W2.

Description

  The present invention relates to a spot size converter.

  Non-Patent Document 1 describes a mode size converter. This mode size converter is formed on a silicon substrate, a silicon oxide layer formed on the silicon substrate, a silicon wire core formed on the silicon oxide layer, and a silicon oxide layer so as to cover the tip of the silicon wire core. And a polymer core. The silicon wire core has a dimension of about 0.3 μm. The polymer core has a dimension of about 3 μm.

T. Shoji, T. Tsuchizawa, T. Watanabe, K. Yamada and H. Morita, "Low loss mode size converter from 0.3 mmsquare Si wire waveguides to single mode fibers", Electronics Letters, 24no. 25. vol. 38, pp 1669-1670, 2002.

  In the mode size converter of Non-Patent Document 1, the mode diameter (spot size) of the light from the silicon wire core is expanded by the polymer core to achieve optical coupling to a single mode fiber having a mode diameter of 4.3 μm. Yes.

  By the way, for example, when optical coupling to an optical fiber having a mode diameter of about 10 μm is required, it is conceivable to increase the size of the polymer core of the mode size converter described above. However, in the mode size converter of Non-Patent Document 1, the silicon wire core and the polymer core are formed on the same silicon oxide layer. For this reason, when the dimension of the polymer core is increased as described above, the deviation between the optical axis of the silicon wire core and the optical axis of the polymer core increases. In such a situation, when light travels from the silicon wire core to the polymer core, light fluctuation may occur in the direction intersecting the light traveling direction. When light fluctuation occurs, the optical coupling efficiency between the optical fiber and the mode size converter may be reduced.

  This invention is made | formed in view of the said situation, and it aims at providing the spot size converter which can suppress the fall of optical coupling efficiency.

  A spot size converter according to the present invention is formed on a semiconductor substrate including a first region, a second region, a third region, and a fourth region arranged in order in a predetermined direction, and the semiconductor substrate. A spot size conversion unit, and the spot size conversion unit includes a first region, a second region, a third region, a fourth region, and a first silicon oxide layer formed on the semiconductor substrate; A first silicon waveguide formed on the first region, the second region, and the first silicon oxide layer; and formed on the first region, the second region, and the first silicon waveguide. And a second silicon oxide layer having a third region, a fourth region, and a portion formed on the first silicon oxide layer, a second region, a third region, and a second region A second silicon waveguide formed on the second silicon oxide layer, a second region, And a portion formed on the second silicon waveguide, and a third silicon oxide layer having a first region, a fourth region, and a portion formed on the second silicon oxide layer And a silicon oxide waveguide including a first silicon oxide layer, a second silicon oxide layer, and a third silicon layer is formed on the fourth region. Optically coupled to two silicon waveguides.

  In this spot size converter, light traveling in a predetermined direction through the first silicon waveguide transitions to the second silicon waveguide via the second silicon oxide layer. The light that has transitioned to the second silicon waveguide travels through the second silicon waveguide and is propagated to the silicon oxide waveguide that is optically coupled to the second silicon waveguide. The light propagated to the silicon oxide waveguide proceeds in a predetermined direction while the spot size is enlarged. As described above, in this spot size converter, the spot size is converted while moving the optical axis by shifting light from the first silicon waveguide to the second silicon waveguide. Therefore, according to this spot size converter, the spot size can be converted while suppressing the fluctuation of the light caused by the deviation of the optical axis.

  In the spot size converter according to the present invention, the first silicon waveguide has a first taper portion whose width decreases in a predetermined direction on the second region, and the second silicon waveguide. May have a second tapered portion whose width increases in a predetermined direction on the second region. According to this configuration, light can be efficiently transitioned from the first silicon waveguide to the second silicon waveguide.

  In the spot size converter according to the present invention, the length of the overlapping portion of the first taper portion and the second taper portion in the direction intersecting the predetermined direction in the predetermined direction is the first taper portion and the second taper portion. The length of the taper portion in a predetermined direction may be equal to or shorter than the length. According to such a configuration, light can be more efficiently transitioned from the first silicon waveguide to the second silicon waveguide, so that optical loss can be suppressed.

  In the spot size converter according to the present invention, the optical axis of the second silicon waveguide may be positioned substantially at the center of the silicon oxide waveguide in a direction intersecting the predetermined direction. According to such a configuration, it is possible to further suppress fluctuation of light traveling through the silicon oxide waveguide.

  The spot size converter which concerns on this invention WHEREIN: The space | gap may be formed in the outer periphery of a spot size conversion part. According to such a structure, since the space | gap part acts as a clad, light can be confined suitably.

  ADVANTAGE OF THE INVENTION According to this invention, the spot size converter which can suppress the fall of optical coupling efficiency can be provided.

FIG. 1 is a diagram illustrating an optical device according to the present embodiment. FIG. 2 is a view showing a cross section taken along line II-II in FIG. FIG. 3 is a diagram showing a configuration of the spot size converter shown in FIG. FIG. 4 is a view showing a cross section of the spot size converter shown in FIG. FIG. 5 is a view showing a cross section of the spot size converter shown in FIG. FIG. 6 is a diagram illustrating a configuration of the spot size conversion unit illustrated in FIG. 1. FIG. 7 is a diagram showing main manufacturing steps of the spot size converter. FIG. 8 is a diagram showing main manufacturing steps of the spot size converter. FIG. 9 is a diagram showing main manufacturing steps of the spot size converter. FIG. 10 is a diagram showing a cross section in the main manufacturing process of the spot size converter 1B. FIG. 11 is a diagram for explaining the calculation result. FIG. 12 is a diagram for explaining the calculation result. FIG. 13 is a diagram for explaining a spot size converter according to a comparative example.

  DESCRIPTION OF EMBODIMENTS Hereinafter, embodiments for carrying out the present invention will be described in detail with reference to the accompanying drawings. In the description of the drawings, the same elements are denoted by the same reference numerals, and redundant description is omitted.

  FIG. 1 is a diagram illustrating an optical device according to the present embodiment. FIG. 2 is a view showing a cross section taken along line II-II in FIG. In the following drawings, an orthogonal coordinate system M is shown. The optical device 80 shown in FIGS. 1 and 2 includes a spot size converter 1 </ b> A, an optical device arranged in order in a predetermined direction (direction D in the orthogonal coordinate system M: a direction from an optical fiber 81 to an optical fiber 82 described later) A functional unit 2 and a spot size converter 1B are provided. The spot size converters 1A and 1B and the optical function unit 2 are formed integrally with each other.

  The optical device 80 is disposed between the optical fiber 81 and the optical fiber 82. The optical device 80 has an end face 80 a optically coupled to the optical fiber 81 and an end face 80 b optically coupled to the optical fiber 82. The optical device 80, for example, guides light input from the optical fiber 81 through the end face 80a and outputs the light to the optical fiber 82 through the end face 80b. At that time, the optical device 80 performs predetermined processing such as light modulation by the optical function unit 2.

  The optical fibers 81 and 82 are quartz single mode optical fibers (SMF) as an example, and propagate light having a wavelength of 1.55 μm as an example. The optical fibers 81 and 82 have a mode diameter (spot size) of 10 μm, for example.

  Spot size converter 1A includes an end face 80a. For example, the spot size converter 1 </ b> A guides the light input from the optical fiber 81 while converting the spot size and outputs the light to the optical function unit 2. Spot size converter 1B includes an end face 80b. The spot size converter 1B, for example, guides the light input from the optical function unit 2 while converting the spot size and outputs the light to the optical fiber 82.

  The optical function unit 2 includes an optical circuit that performs predetermined processing on the light input to the optical device 80. The optical function unit 2 is formed between the spot size converter 1A and the spot size converter 1B. The optical function unit 2 includes a semiconductor substrate 3a and a main body 2a having an optical circuit. The main body 2a is formed on the semiconductor substrate 3a and joined to the semiconductor substrate 3a. The main body 2a has a waveguide 2b. The waveguide 2b optically couples the spot size converter 1A and the spot size converter 1B to each other.

  According to such an optical device 80, for example, light having a wavelength of 1.55 μm emitted from the optical fiber 81 is incident on the end surface 80 a of the optical device 80. The light incident on the end face 80a is propagated to the optical function unit 2 by being guided through the spot size converter 1A. The light propagated to the optical function unit 2 is propagated to the spot size converter 1B via the waveguide 2b. The light propagated to the spot size converter 1B is converted so as to match the optical axis position and spot size of the optical fiber 82 while being guided through the spot size converter 1B, and is emitted from the end face 80b. The light emitted from the end face 80 b enters the optical fiber 82.

  Subsequently, the spot size converters 1A and 1B will be described in detail. Since spot size converter 1A and spot size converter 1B have the same configuration, spot size converter 1B will be mainly described in the following description.

  FIG. 3 is a diagram showing a configuration of the spot size converter shown in FIG. 3A is a plan view of a spot size converter, and FIG. 3B is a diagram showing a cross section taken along line III-III in FIG. 3A. As shown in FIG. 3, the spot size converter 1B includes a semiconductor substrate 3b and a main body 4 formed on the semiconductor substrate 3b.

  The semiconductor substrate 3b is a semiconductor substrate made of silicon, for example. Here, the semiconductor substrate 3b included in the spot size converter 1B (and the spot size converter 1A) and the semiconductor substrate 3a included in the optical function unit 2 are integrally formed. The semiconductor substrate 3b includes regions R1 to R8 arranged in order in the direction D. The region R1 is continuous with the semiconductor substrate 3a of the optical function unit 2. On the region R1 of the semiconductor substrate 3b, the semiconductor substrate 3b and the main body 4 are joined to each other. On the other hand, on the regions R2 to R8 of the semiconductor substrate 3b, the semiconductor substrate 3b and the main body 4 are separated from each other by a predetermined distance L1. That is, a gap G1 having a distance L1 is formed between the semiconductor substrate 3b and the main body portion 4.

  The main body 4 includes a spot size conversion unit 6 and terraces 7a and 7b. The spot size conversion unit 6 is formed between the terrace 7a and the terrace 7b. A gap G2a is formed between the spot size conversion unit 6 and the terrace 7a. A gap G2b is formed between the spot size conversion unit 6 and the terrace 7b. The width L2 of the gaps G2a and G2b is 2 μm as an example.

The spot size conversion unit 6 will be described. The spot size conversion unit 6 is formed on the semiconductor substrate 3b. The spot size conversion unit 6 includes silicon oxide layers S1 to S4 and silicon waveguides W1 to W3. Silicon oxide layer S1~S4 is made of, for example, of SiO _2. The silicon waveguides W1 to W3 are made of Si, for example. The silicon oxide layers S1 to S4 are stacked in the direction Z orthogonal to the direction D in the order of the silicon oxide layer S1, the silicon oxide layer S2, the silicon oxide layer S3, and the silicon oxide layer S4. A silicon waveguide W1 is sandwiched between the silicon oxide layer S1 and the silicon oxide layer S2. A silicon waveguide W2 is sandwiched between the silicon oxide layer S2 and the silicon oxide layer S3. A silicon waveguide W3 is sandwiched between the silicon oxide layer S3 and the silicon oxide layer S4.

  The silicon waveguide W1 has an optical axis A1. The silicon waveguide W2 has an optical axis A2. The silicon waveguide W3 has an optical axis A3. The optical axes A1 to A3 are separated from each other in the direction in which the silicon waveguides W1 to W3 are stacked (Z direction). On the other hand, the optical axes A1 to A3 overlap on the same line when viewed from the direction Z. As an example, the silicon waveguides W2 to W3 are waveguides made of single crystal silicon or polycrystalline silicon.

  The silicon oxide layer S1 is formed on the regions R1 to R8 and the semiconductor substrate 3b. The silicon oxide layer S1 is bonded to the semiconductor substrate 3b on the region R1 of the semiconductor substrate 3b. The silicon oxide layer S1 is separated from the semiconductor substrate 3b in the direction Z in the regions R2 to R8 of the semiconductor substrate 3b. As an example, the silicon oxide layer S1 has a thickness of 2 μm. Note that the silicon oxide layer S1 corresponds to a first silicon oxide layer.

  As shown in FIGS. 3 and 4, the silicon waveguide W1 is formed on the regions R1 to R4 and on the silicon oxide layer S1. The silicon waveguide W1 has a waveguide portion W1a and a waveguide portion W1b that are integrally formed. The regions R1 to R3 correspond to the first region. The silicon waveguide W1 corresponds to a first silicon waveguide. FIG. 4 is a view showing a cross section of the spot size converter shown in FIG. 4A is a sectional view taken along line IVa-IVa in FIG. 1, and FIG. 4B is a sectional view taken along line IVb-IVb in FIG. FIG. 4C is a cross-sectional view taken along line IVc-IVc in FIG. 1, and FIG. 4D is a cross-sectional view taken along line IVd-IVd in FIG.

  The waveguide portion W1a is formed on the regions R1 and R2 and the silicon oxide layer S1. As an example, the waveguide portion W1a has a thickness of 0.22 μm and a width of 0.5 μm. One end side of the waveguide portion W1a is continuous with the waveguide 2b of the optical function portion 2. A tapered portion T1 is formed on the other end side of the waveguide portion W1a. The width of the tapered portion T1 decreases in the direction D. The length L3 along the direction D of the tapered portion T1 is 150 μm as an example.

  The waveguide portion W1b is formed on the regions R3 and R4 and on the silicon oxide layer S1. As an example, the waveguide portion W1b has a thickness of 0.11 μm and a width of 0.5 μm. One end side of the waveguide portion W1b is continuous with the waveguide portion W1a. A tapered portion T2 is formed at the other end of the waveguide portion W1b. The tapered portion T2 has the same configuration as the tapered portion T1 of the waveguide portion W1a. That is, the width of the tapered portion T2 decreases in the direction D. As an example, the length along the direction D of the tapered portion T2 is 150 μm. Note that the region R4 corresponds to a second region.

  The silicon oxide layer S2 includes portions formed on the regions R1 and R2 and the waveguide portion W1a of the silicon waveguide W1, and portions formed on the regions R3 and R4 and the waveguide portion W1b of the silicon waveguide W1. And portions formed on the regions R5 to R8 and the silicon oxide layer S1. As an example, the silicon oxide layer S2 has a thickness of 1.28 μm on the regions R1 and R2. As an example, the silicon oxide layer S2 has a thickness of 1.39 μm on the regions R3 and R4. The silicon oxide layer S2 has a thickness of 1.5 μm as an example on the regions R5 to R8. The silicon oxide layer S1 is also formed on the side surface of the silicon waveguide W1. That is, the silicon oxide layer S1 embeds the silicon waveguide W1. Accordingly, the silicon oxide layers S1 and S2 function as a cladding of the silicon waveguide W1. Note that the silicon oxide layer S2 corresponds to a second silicon oxide layer.

  As shown in FIGS. 3 to 5, the silicon waveguide W2 is formed on the regions R4 to R6 and the silicon oxide layer S2. As an example, the silicon waveguide W2 has a thickness of 0.11 μm and a width of 0.5 μm. The silicon waveguide W2 has a tapered portion T3 formed at one end on the region R4 and a tapered portion T4 formed at the other end on the region R6. A waveguide portion W2a is formed between the tapered portion T3 and the tapered portion T4 along the direction D. The silicon waveguide W2 corresponds to a second silicon waveguide. FIG. 5 is a view showing a cross section of the spot size converter shown in FIG. 5A is a cross-sectional view taken along the line Va-Va in FIG. 1, and FIG. 5B is a cross-sectional view taken along the line Vb-Vb in FIG. FIG. 5C is a cross-sectional view taken along the line Vc-Vc in FIG. 1, and FIG. 5D is a cross-sectional view taken along the line Vd-Vd in FIG.

  The taper portion T3 on the region R4 increases in width from the sharpened one end toward the direction D, and the other end has the same width as the waveguide portion W2a. As an example, the length in the direction D of the tapered portion T3 is 150 μm. The taper portion T4 on the region R6 decreases in width from the one end portion having the same width as the waveguide portion W2a in the direction D, and is sharpened at the tip of the other end. As an example, the length in the direction D of the tapered portion T4 is 150 μm. For example, the width of the waveguide portion W2a is 0.5 μm, and the length along the direction D is 50 μm. Therefore, the length in the direction D of the silicon waveguide W2 is 350 μm as an example. Note that the region R5 and the region R6 correspond to a third region.

  The silicon oxide layer S3 includes a portion formed on the regions R1 to R3 and the silicon oxide layer S2, a portion formed on the regions R4 to R6 and the silicon waveguide W2, a region R7 to R8, and the silicon oxide. And a portion formed on the layer S2. The silicon oxide layer S2 has a thickness of 1.5 μm as an example on the regions R1 to R3. As an example, the silicon oxide layer S2 has a thickness of 1.5 μm on the regions R7 to R8. As an example, the silicon oxide layer S2 has a thickness of 1.39 μm on the regions R4 to R6. The silicon oxide layer S3 is also formed on the side surface of the silicon waveguide W2. That is, the silicon oxide layer S3 embeds the silicon waveguide W2. Accordingly, the silicon oxide layers S2 and S3 function as a cladding of the silicon waveguide W2. Note that the silicon oxide layer S3 corresponds to a third silicon oxide layer. The region R8 corresponds to a fourth region.

  The silicon waveguide W3 is formed on the regions R6 and R7 and the silicon oxide layer S3. For example, the silicon waveguide W3 has a thickness of 0.11 μm and a width of 0.5 μm. The silicon waveguide W3 has a tapered portion T6 formed at one end on the region R6 and a tapered portion T7 formed at the other end on the region S7. A waveguide portion W3a is formed between the tapered portion T6 and the tapered portion T7. The length in the direction D of the silicon waveguide W3 is 350 μm as an example. The tapered portion T6 has the same configuration as the tapered portion T3 of the silicon waveguide W2. Further, the tapered portion T7 has the same configuration as the tapered portion T4 of the silicon waveguide W2.

  The silicon oxide layer S4 includes a portion formed on the regions R1 to R5 and the silicon oxide layer S3, a portion formed on the regions R6 and R7 and the silicon waveguide W3, a region R8, and the silicon oxide layer S3. And a portion formed thereon. The silicon oxide layer S3 has a thickness of 5 μm as an example on the regions R1 to R5. As an example, the silicon oxide layer S3 has a thickness of 5 μm on the region R8. For example, the silicon oxide layer S3 has a thickness of 4.89 μm on the regions R6 and R7. The silicon oxide layer S4 is also formed on the side surface of the silicon waveguide W3. That is, the silicon oxide layer S4 embeds the silicon waveguide W3. Accordingly, the silicon oxide layers S3 and S4 function as a cladding of the silicon waveguide W3.

  In the spot size conversion unit 6 configured as described above, a transition region R9 for transitioning light from the silicon waveguide W1 to the silicon waveguide W3 is formed on the regions R3 to R7. The transition region R9 includes silicon oxide layers S1 to S4 and silicon waveguides W1 to W3. The transition region R9 has a substantially rectangular parallelepiped shape extending in the direction D. The cladding portion of the transition region R9 constituted by the silicon oxide layers S1 to S4 has a rectangular cross section with the direction Z as the longitudinal direction (see (c) in FIG. 4). The cross section of the cladding part has a width of 3.5 μm and a height in the direction Z of 10 μm. Thus, according to the clad portion having the cross-sectional shape with the direction Z as the longitudinal direction, the mode shape spreading around the silicon waveguides W1 to W3 can be made vertically long, and light can be efficiently emitted in the direction Z. Transition can be made.

  Gaps G2a and G2b are formed between the side surface R9a of the transition region R9 included on the outer periphery of the spot size conversion unit 6 and the terraces 7a and 7b. In addition, a gap G1 is formed between the lower surface R9b of the transition region R9 included in the outer periphery of the spot size conversion unit 6 and the semiconductor substrate 3b. Therefore, the outer periphery of the transition region R9 is in contact with air. Accordingly, in the transition region R9, the core is the silicon waveguides W1 to W3, the first cladding is the silicon oxide layers S1 to S4, and the second cladding is air. The gaps G1, G2a, and G2b may be filled with a material that can form a predetermined refractive index difference with silicon oxide.

  In such a transition region R9, the transition portion C1 is formed by the taper portion T2 of the silicon waveguide W1, the taper portion T3 of the silicon waveguide W2, and the silicon oxide layer S2 sandwiched between the taper portion T2 and the taper portion T3. Is configured. On the region R4, the silicon waveguide W1 and the silicon waveguide W2 overlap in the direction Z with the silicon oxide layer S2 interposed therebetween. By overlapping the silicon waveguide W1 and the silicon waveguide W2, it is possible to suppress light from being emitted from one end of the silicon waveguide W1 at a wide angle. Accordingly, it is possible to suppress a decrease in light transition efficiency.

  FIG. 6 is a diagram illustrating a configuration of the spot size conversion unit illustrated in FIG. 1. 6A is a plan view of the spot size conversion unit, and FIG. 6B is a diagram showing a cross section taken along line VI-VI in FIG. 6A. As shown in FIG. 6, the length L4 in the direction D of the overlapping portion of the taper portion T2 and the taper portion T3 in the direction Z is not more than the length L5 in the direction D of the taper portions T2 and T3. That is, the ratio of the length L4 of the overlapping portion of the taper portions T2 and T3 to the length L5 of the taper portions T2 and T3 is 5% or more. Further, the ratio of the length L4 of the overlapping portion of the taper portions T2 and T3 to the length L5 of the taper portions T2 and T3 is 100% or less.

  The length L5 in the direction D of the tapered portions T2 and T3 is 50 μm to 500 μm, and is 50 μm as an example. Further, the length L4 of the overlapping portion of the taper portions T2 and T3 is 50 μm or less. As an example, when the length L4 of the overlapping portion between the tapered portions T2 and T3 is set to 50% of the length L5 of the tapered portions T2 and T3, the length L4 of the overlapping portion between the tapered portions T2 and T3 is 25 μm. is there. Further, the interval between the tapered portions T2 and T3 in the direction Z is defined by the thickness of the silicon oxide layer S2 sandwiched between the tapered portions T2 and T3. In the present embodiment, the interval between the tapered portions T2 and T3 is 1.39 μm as an example.

  Further, in the transition region R9, the transition portion C2 is configured by the taper portion T4 of the silicon waveguide W2 and the taper portion T6 of the silicon waveguide W3. In the region R6, the silicon waveguide W2 and the silicon waveguide W3 overlap in the direction Z with the silicon oxide layer S3 interposed therebetween. Since the transition part C2 has the same configuration as the transition part C1, a detailed description thereof will be omitted.

  On the other hand, as shown in FIG. 3, on the region R8, an enlarged region R10 for expanding the light emitted from the tapered portion T7 of the silicon waveguide W3 is formed. The enlarged region R10 is optically coupled to the silicon waveguide W2 via the silicon waveguide W3. One end of the enlarged region R10 is continuous with the transition region R9. The other end of the enlarged region R10 is included in the end face 80a. The enlarged region R10 is a silicon oxide waveguide constituted by silicon oxide layers S1 to S4. The enlarged region R10 has a portion where the cross-sectional area is enlarged along the direction D.

  One end of the enlarged region R10 that is continuous with the transition region R9 has the same cross-sectional shape as the transition region R9 (see FIG. 5C). That is, the width of one end of the enlarged region R10 is, for example, 3.5 μm, and the height is, for example, 10 μm. On the other hand, the width of the other end including the end face 80a in the enlarged region R10 is larger than that of the one end. That is, the width of the other end of the enlarged region R10 is, for example, 10 μm, and the height is, for example, 10 μm, and has a square cross-sectional shape (see FIG. 5D). In the enlarged region R9, the cross-sectional area of the enlarged region R10 sandwiched between the gaps G2a and G2b is enlarged by increasing the distance between the gap G2a and the gap G2b. Therefore, since the mode diameter of the light propagating through the enlarged region R10 and emitted from the end face 80a is close to the mode diameter of the optical fiber 81, the optical coupling efficiency between the spot size converter 1B and the optical fiber 81 can be increased. .

  As described above, the gap G2a is formed between the side surface R10a of the enlarged region R10 and the terrace 7a, and the gap G2b is formed between the side surface R10a of the enlarged region R10 and the terrace 7b. Further, a gap G1 is formed between the bottom surface R10b (see FIG. 4C, etc.) of the enlarged region R10 and the semiconductor substrate 3b. Therefore, a waveguide is formed with the enlarged region R10 as the core and the air in the air gap as the cladding.

  Here, the optical axis A3 of the silicon waveguide W3 is located substantially at the center of the enlarged region R10 in the direction Z. Specifically, the height of the enlarged region R10 is, for example, 10 μm. The silicon waveguide W3 is sandwiched between a clad having a thickness of, for example, 5 μm made of the silicon oxide layers S1 to S3 and a clad having a thickness of, for example, 4.89 μm, made of the silicon oxide layer S4. Furthermore, the optical axis A3 of the silicon waveguide W3 is located at the approximate center of the enlarged region R10 in the direction H perpendicular to the direction D and the direction Z (see FIG. 3A). Accordingly, the optical axis A3 of the silicon waveguide W3 is arranged at the approximate center of each of the direction Z and the direction H of the enlarged region R10 in the cross section intersecting with the direction D.

  Next, a manufacturing method of the spot size converter 1B according to the present embodiment will be described. FIG. 7 is a diagram showing main manufacturing steps of the spot size converter. FIG. 7A shows a cross section of the SOI substrate. FIG. 7B is a plan view showing the SOI substrate. FIG. 7C is a view showing a cross section in the process of forming the silicon waveguide. FIG. 7D is a plan view in the process of forming the silicon waveguide. FIG. 7E is a diagram showing a cross section in the step of forming the silicon oxide layer. FIG. 7F is a plan view in the step of forming the silicon oxide layer. In addition, the material and dimension of each part in the following manufacturing methods are examples.

As shown in FIGS. 7A and 7B, an SOI substrate 10 is prepared. The optical device 80 including the spot size converter 1 </ b> B is manufactured using an SOI (Silicon on Insulator) substrate 10. The SOI substrate 10 is generally used in a CMOS process. Further, a CMOS process is applied to the manufacturing method of the spot size converter 1B. The SOI substrate 10 has a structure in which a silicon substrate 11, a BOX (Buried Oxide) layer 12, and a device layer 13 are stacked. The silicon substrate 11 is made of single crystal silicon and has a thickness of 2 μm. The BOX layer 12 is made of silicon oxide (SiO ₂ ) and has a thickness of 0.22 μm. The device layer 13 is made of single crystal silicon and has a thickness of 2 μm. Here, the silicon substrate 11 is used as the semiconductor substrate 3b of the spot size converter 1B. The BOX layer 12 is used as the silicon oxide layer S1 of the spot size converter 1B. The device layer 13 is used as the silicon waveguide W1 of the spot size converter 1B.

  As shown in FIGS. 7C and 7D, the device layer 13 is etched to form the silicon waveguide W1. More specifically, first, a mask for the waveguide portion W1a is formed on the device layer 13. Then, the device layer 13 is etched using the mask. The etching depth at this time is 0.11 μm. The etching depth is controlled based on the etching time. By this step, a waveguide portion W1a having a width of 0.5 μm and a tapered portion T1 having a tip width of 0.1 μm is formed. Subsequently, a mask for the waveguide portion W1b is formed on the device layer 13. Then, the device layer 13 is further etched using the mask. The etching depth at this time is 0.11 μm, and etching is performed until the BOX layer 12 is exposed. By this step, the waveguide portion W1b having a width of 0.5 μm and a tapered portion T2 having a tip width of 0.1 μm is formed.

  As shown in FIGS. 7E and 7F, a silicon oxide layer S2 is formed on the silicon oxide layer S1 and the silicon waveguide W1. More specifically, a silicon oxide layer is grown on the silicon oxide layer S1 and the silicon waveguide W1 by using a PECVD (plasma-enhanced chemical vapor deposition) method. At this time, the silicon oxide layer is grown so as to have a thickness of 2 μm. After the silicon oxide layer is grown, the silicon oxide layer is polished using a CMP (Chemical Mechanical Polishing) method. By this polishing, the silicon oxide layer is removed from the surface by 0.5 μm. Through the above steps, the silicon oxide layer S2 having a thickness of 1.5 μm is formed. Further, since the surface of the silicon oxide layer S2 is formed by the CMP method, the surface roughness (surface roughness) is reduced. By reducing the roughness of the surface of the silicon oxide layer S2, the propagation loss of the waveguide can be reduced.

  FIG. 8 is a diagram showing main manufacturing steps of the spot size converter. FIGS. 8A and 8C are views showing a cross section in the process of forming the silicon waveguide. FIGS. 8B and 8D are plan views in the process of forming the silicon waveguide. FIG. 8E is a diagram showing a cross section in the step of forming the silicon oxide layer. FIG. 8F is a plan view in the step of forming the silicon oxide layer.

  As shown in FIGS. 8A and 8B, an amorphous silicon layer AS is grown on the silicon oxide layer S2. The amorphous silicon layer AS is grown by the LPCVD method and has a thickness of 0.2 μm. Next, heat treatment is performed on the amorphous silicon layer AS by an RTA (Rapid Thermal Annealing) method. By this heat treatment, the amorphous silicon layer AS is converted into the polysilicon layer PS, so that the absorption loss in the silicon waveguide W2 can be reduced. Then, the polysilicon layer PS is polished using a CMP method. By this polishing, the polysilicon layer PS is removed by 0.09 μm. Through the above steps, a polysilicon layer PS having a thickness of 0.11 μm is formed.

  As shown in FIGS. 8C and 8D, the polysilicon layer PS is etched to form the silicon waveguide W2. More specifically, a mask for the silicon waveguide W2 is formed on the polysilicon layer PS. Then, the polysilicon layer PS is etched using the mask. The etching depth at this time is 0.11 μm. The etching depth is controlled based on the etching time. By this etching, the silicon waveguide W2 is formed. In the silicon waveguide W2, the width of the waveguide portion W2a is 0.5 μm, and the tip widths of the tapered portions T3 and T4 are 0.1 μm. Further, the silicon waveguide W2 has a portion overlapping with a part of the previously formed silicon waveguide W1 when viewed from the direction Z. Moreover, since the surface of the silicon waveguide W2 is formed by the CMP method, the roughness of the surface is reduced. By reducing the roughness of the surface of the silicon waveguide W2, the propagation loss of the waveguide can be reduced.

  As shown in FIGS. 8E and 8F, a silicon oxide layer S3 is formed on the silicon oxide layer S2 and the silicon waveguide W2. The silicon oxide layer S3 is formed by the same process as the silicon oxide layer S2. That is, a silicon oxide layer having a thickness of 2 μm is grown on the silicon oxide layer S1 and the silicon waveguide W1 by PECVD. After the silicon oxide layer is grown, the silicon oxide layer is polished using a CMP method. By this polishing, a silicon oxide layer S3 having a thickness of 1.5 μm is formed. Further, since the surface of the silicon oxide layer S3 is formed by the CMP method, the roughness of the surface is reduced. By reducing the roughness of the surface of the silicon oxide layer S3, the propagation loss of the waveguide can be reduced.

  FIG. 9 is a diagram showing main manufacturing steps of the spot size converter. FIGS. 9A and 9C are cross-sectional views in the process of forming the silicon waveguide. FIGS. 9B and 9D are plan views in the process of forming the silicon waveguide. FIG. 9E is a diagram showing a cross section in the step of forming the silicon oxide layer. FIG. 9F is a plan view in the step of forming the silicon oxide layer.

  Next, a step of forming the silicon waveguide W3 is performed. The step of forming the silicon waveguide W3 is performed in the same manner as the method of forming the silicon waveguide W2, except that the position of forming the silicon waveguide W3 is different from that of the silicon waveguide W2.

  As shown in FIGS. 9A and 9B, an amorphous silicon layer AS having a thickness of 0.2 μm grown by the LPCVD method is grown on the silicon oxide layer S3. Next, the amorphous silicon layer AS is subjected to a heat treatment to be converted into a polysilicon layer PS. Then, the polysilicon layer PS is polished by CMP to form a polysilicon layer PS having a thickness of 0.11 μm.

  As shown in FIGS. 9C and 9D, the polysilicon layer PS is etched to form a silicon waveguide W3. More specifically, after forming a mask for the silicon waveguide W3, the polysilicon layer PS is etched using the mask. In the silicon waveguide W3 formed by these steps, the width of the waveguide portion W3a is 0.5 μm, the tip widths of the tapered portions T6 and T7 are 0.1 μm, and the silicon waveguide W2 formed earlier is formed. It has a part that overlaps with part when viewed from the direction Z.

  As shown in FIGS. 9E and 9F, a silicon oxide layer S4 is grown on the silicon oxide layer S4 and the silicon waveguide W3. More specifically, after a silicon oxide layer having a thickness of 5.5 μm is grown on the silicon oxide layer S3 by PECVD, the silicon oxide layer is polished using CMP. By this polishing, a silicon oxide layer S4 having a thickness of 5 μm is formed. Moreover, since the surface of the silicon oxide layer S4 is formed by the CMP method, the roughness of the surface is reduced. By reducing the roughness of the surface of the silicon oxide layer S4, it is possible to reduce the propagation loss of the waveguide.

  FIG. 10 is a diagram showing a cross section in the main manufacturing process of the spot size converter 1B. (A) of FIG. 10 is a figure which shows the end surface 80a in the process of forming a space | gap. FIG. 10B is a view showing a cross section taken along the line Vc-Vc in FIG. 1 in the step of forming the air gap. FIG. 10C is a view showing a cross section taken along the line Vd-Vd in FIG. 1 in the step of forming the air gap.

  As shown in FIGS. 10A to 10C, first, the silicon oxide layers S1 to S4 are etched to form gaps G2a and G2b having a width of 2 μm. Thereby, the terraces 7a and 7b and the silicon oxide layers S1 to S4 are formed, and a confinement structure in which the air in the gaps G2a and G2b is clad in the direction H is formed. Next, the silicon substrate 11 is etched to form a recess 11a. By this step, the gap G1 is formed. Thereby, the semiconductor substrate 3b and the spot size conversion part 6 are formed. At this time, the spot size conversion unit 6 has a cantilever structure. According to such a cantilever structure, a confinement structure in which air is clad in the direction Z is formed.

  As described above, in the spot size converter 1B according to the present embodiment, the light traveling in the direction D through the silicon waveguide W1 is transitioned to the silicon waveguide W2 via the silicon oxide layer S2. The light that has transitioned to the silicon waveguide W2 is guided through the silicon waveguide W2 and propagates to the enlarged region R10 that is optically coupled to the silicon waveguide W2 via the silicon waveguide W3. The light propagated to the enlarged region R10 proceeds in the direction D while the spot size is enlarged. Thus, in the spot size converter 1B, the spot size is converted while moving the optical axes A1 to A3 by shifting light from the silicon waveguide W1 to the silicon waveguide W2 and the silicon waveguide W3. Therefore, according to the spot size converter 1B, the spot size can be converted while suppressing the fluctuation of the light caused by the deviation of the optical axis.

  In the spot size converter 1B according to the present embodiment, the silicon waveguide W1 has a tapered portion T2 whose width decreases in the direction D in the region R4, and the silicon waveguide W2 faces in the direction D in the region R4. And has a tapered portion T3 whose width increases. According to this configuration, the light traveling through the silicon waveguide W1 has its light mode expanded at the tapered portion T2, and is efficiently optically coupled to the tapered portion T3 of the silicon waveguide W2 via the silicon oxide layer S2. The Therefore, light can be efficiently transitioned from the silicon waveguide W1 to the silicon waveguide W2.

  In the spot size converter 1B according to the present embodiment, the length L4 in the direction D of the overlapping portion of the tapered portion T2 and the tapered portion T3 in the direction Z orthogonal to the direction D is the tapered portion T2 and the tapered portion T3. The length L in the direction D is equal to or less than L5. According to such a configuration, light can be more efficiently transitioned from the silicon waveguide W1 to the silicon waveguide W2, and therefore, light loss can be suppressed and light can be further transmitted from the silicon waveguide W1 to the silicon waveguide W2. It is possible to make transition efficiently.

  In the spot size converter 1B according to the present embodiment, the optical axis A3 of the silicon waveguide W3 is located at the center of the enlarged region R10 in the direction Z orthogonal to the direction D. Furthermore, the optical axis A3 of the silicon waveguide W3 is located at the center of the enlarged region R10 with respect to the direction H perpendicular to the direction D and the direction Z. According to such a configuration, when the light emitted from the silicon waveguide W3 travels in the direction D in the expansion region R10 and expands in the direction including the direction Z and the direction H, from the position where the light is emitted. Since the distances to the side surfaces, the upper surface, and the lower surface of the enlarged region R10 are substantially the same, fluctuations in light intensity are suppressed.

  In the spot size converter 1B according to the present embodiment, since the gaps G1, G2a, and G2b are formed on the outer periphery of the spot size conversion unit 6, a good light confinement structure can be formed.

  In the spot size converter 1B according to the present embodiment, the waveguide is made of single crystal or polycrystalline silicon. As described above, according to the optical transition structure in which the waveguides made of silicon are optically coupled to each other, the waveguide made of silicon has a function of strongly confining light, so that the transition efficiency between the waveguides is increased. be able to.

  In the spot size converter 1B according to the present embodiment, the mode diameter of the light emitted from the end face 80a of the enlarged region R10 can be made close to the mode diameter of the optical fiber, for example, about 10 μm. According to such a configuration, it is possible to suppress a decrease in coupling efficiency between the spot size converter 1B and the optical fiber.

  The spot size converter 1B according to the present invention is not limited to the specific configuration disclosed in the present embodiment. In the above embodiment, the spot size converter 1B includes the three silicon waveguides W1 to W3, and the silicon waveguide W2 is optically coupled to the enlarged region R10 via the silicon waveguide W3. However, it is not limited to this form. The number of silicon waveguides provided in the spot size converter 1B may be two. That is, the silicon waveguide W2 may be directly optically coupled to the enlarged region R10. Further, the number of silicon waveguides provided in the spot size converter 1B may be three or more. As described above, by appropriately selecting the number of silicon waveguides and the interval between the silicon waveguides, the silicon waveguides optically coupled to the enlarged region R10 can be disposed at a desired position.

  In the spot size converter 1B according to the present embodiment, the state of light transition between the silicon waveguides W1 to W3 was confirmed by calculation. For this calculation, a three-dimensional beam propagation method (3D-BPM method) was used. 11 and 12 are diagrams for explaining the calculation results. In particular, FIG. 11A is a diagram showing a cross section of the spot size converter unit 6, and FIG. 11B is a diagram showing a calculation result in the DZ plane. 12A is a diagram showing a calculation result in the DZ plane, and FIG. 12B is a diagram showing a calculation result in a cross section along the line XIIa-XIIa in FIG. (C) of FIG. 12 is a diagram showing a calculation result in the cross section along the XIIb-XIIb line of FIG. 12, and (d) of FIG. 12 is a diagram showing a calculation result in the cross section along the XIIc-XIIc line of FIG. FIG. 12E is a diagram showing the calculation result in the cross section along the XIId-XIId line in FIG. 12, and FIG. 12F is the calculation result in the cross section along the XIIe-XIIe line in FIG. FIG.

The following numerical values were used for this calculation. In this calculation, light was incident on the spot size converter 1B from the end face 80a, and it was confirmed that the light transits from the silicon waveguide W3 to the silicon waveguide W1 via the silicon waveguide W2.
The wavelength of incident light is 1.55 μm.
The width of the silicon waveguides W1 to W3 is 0.5 μm.
The length of the silicon waveguides W1 to W3 is 350 μm.
The thickness of the silicon waveguides W1 to W3 is 0.11 μm.
The thickness of the silicon oxide layer S1 is 2 μm.
The thickness of the silicon oxide layer S2 is 1.5 μm.
The thickness of the silicon oxide layer S3 is 1.5 μm.
The thickness of the silicon oxide layer S4 is 5 μm.
Length of taper portions T1 to T6 is 150 μm.
Distance between taper portions T2 and T3 1.39 μm.
Distance between taper portions T4 and T6 1.39 μm.
The height of the enlarged region R10 is 10 μm.
The width of the enlarged region R10 is 10 μm.

  As shown in FIG. 11B, the light incident from the end face 80a is first coupled to the silicon waveguide W3. Subsequently, the light guided through the silicon waveguide W3 transitions to the silicon waveguide W2. And it turns out that the light which guided the silicon waveguide W2 has changed to the silicon waveguide W1.

  As shown in FIGS. 12B to 12F, the light incident on the end face 80a is confined in a waveguide having the silicon oxide layers S1 to S4 as the core and air as the cladding in the enlarged region R10. (See (b) of FIG. 12). Then, when the width of the enlarged region R10 becomes narrow and the cross-sectional shape becomes vertically long, the mode of guided light changes to vertically long (see FIG. 12C). Subsequently, during the transition from the silicon waveguide W3 to the silicon waveguide W2 and the transition from the silicon waveguide W2 to the silicon waveguide W1, light is confined in each silicon waveguide, and the position of the optical axis is changed. .

[Example 1]
A spot size converter corresponding to the spot size converter 1B according to the present embodiment was produced as Example 1. The spot size converter according to the first embodiment has the numerical values as described above. Then, the coupling loss between the spot size converter and the single mode optical fiber having a flat end face was evaluated. In this evaluation, the coupling loss in the TE mode was evaluated. As a result, it was found that the coupling loss between the spot size converter and the single mode optical fiber was 0.8 dB. Therefore, according to this spot size converter, it was found that the coupling loss can be suppressed as compared with a spot size converter (coupling loss: 2.3 dB) of a comparative example described later.

  The relationship between the amount of displacement between the spot size converter according to Example 1 and the optical fiber and the coupling efficiency was confirmed. First, the position of the optical fiber with respect to the spot size converter according to Example 1 was adjusted, and the spot size converter and the optical fiber were arranged at a position where the coupling efficiency was maximized. Then, the position of the optical fiber was shifted in the direction Z or the direction H, and the shift position of the optical fiber where the coupling efficiency was -1 dB from the maximum value was confirmed. As a result, in the direction Z and the direction H, it was found that when the optical fiber was moved by 2.4 μm, the coupling efficiency became −1 dB from the maximum value.

[Example 2]
A spot size converter corresponding to the spot size converter 1B according to the present embodiment was produced as Example 2. The spot size converter according to the second embodiment has the numerical values as described above. However, the spot size converter according to the second embodiment has a silicon waveguide made of amorphous silicon instead of the silicon waveguides W2 and W3 made of polysilicon. The coupling loss was evaluated by the same method as described above. As a result, it was found that the coupling loss was 1.4 dB in the spot size converter including the silicon waveguide made of amorphous silicon. It has been found that the absorption loss in the silicon waveguide can be reduced by performing heat treatment by the RTA method in the process of forming the silicon waveguide to convert amorphous silicon into polysilicon. Therefore, it was found that the spot size converter provided with the silicon waveguide made of polysilicon can suppress the coupling loss more than the spot size converter provided with the silicon waveguide made of amorphous silicon.

[Comparative Example 1]
A spot size converter 100 according to a comparative example was produced. FIG. 13 is a diagram for explaining a spot size converter according to a comparative example. In particular, FIG. 13A is a diagram illustrating a cross section of a spot size converter according to a comparative example. The spot size converter 100 includes only a silicon waveguide W1 as a waveguide. That is, the spot size converter 100 does not include the silicon waveguides W2 and W3. And the coupling loss between optical fibers was confirmed using the method similar to Example 2. FIG. As a result, the coupling loss between the spot size converter 100 and the optical fiber was 2.3 dB.

  Furthermore, in the spot size converter 100 according to Comparative Example 1, the light distribution in the cross section orthogonal to the direction D was confirmed by calculation. Each numerical value mentioned above was used for calculation. FIG. 13B is a diagram showing the light intensity distribution in the cross section along the line XIII-XIII. The thickness of the silicon oxide layer S1 located below the silicon waveguide W1 is 2 μm, and the thickness of the silicon oxide layers S2 to S4 located above the silicon waveguide W1 is 7 μm. For this reason, the light mode is not a circle and has a flat shape. It has been found that such a mode is greatly different from the mode shape of the optical fiber, so that the coupling loss increases.

  DESCRIPTION OF SYMBOLS 1A, 1B ... Spot size converter, 2 ... Optical function part, 3 ... Base part, 4 ... Main part, 6 ... Spot size conversion part, 7a, 7b ... Terrace, 10 ... Semiconductor substrate, 80 ... Optical apparatus, 81, 82 ... Optical fiber, A1 to A3 ... Optical axis, AS ... Amorphous silicon layer, C1, C2 ... Transition part, D ... Predetermined direction, G1, G1, G2a ... Gap, M ... Cartesian coordinate system, R1 to R8 ... Region, R9 ... transition region, R10 ... enlarged region (silicon oxide waveguide), S1 to S4 ... silicon oxide layer, T1 to T6 ... taper portion, W1 to W3 ... silicon waveguide.

Claims (5)

A semiconductor substrate including a first region, a second region, a third region, and a fourth region arranged in order in a predetermined direction;
A spot size converter formed on the semiconductor substrate;
With
The spot size converter
A first silicon oxide layer formed on the first region, the second region, the third region, the fourth region, and the semiconductor substrate;
A first silicon waveguide formed on the first region, the second region, and the first silicon oxide layer;
On the first region, the second region, and the portion formed on the first silicon waveguide, and on the third region, the fourth region, and the first silicon oxide layer A second silicon oxide layer having a portion formed on;
A second silicon waveguide formed on the second region, the third region, and the second silicon oxide layer;
On the second region, the third region, and the portion formed on the second silicon waveguide, and on the first region, the fourth region, and the second silicon oxide layer A third silicon oxide layer having a portion formed on;
Have
A silicon oxide waveguide including the first silicon oxide layer, the second silicon oxide layer, and the third silicon layer is formed on the fourth region,
The silicon oxide waveguide is optically coupled to the second silicon waveguide;
A spot size converter characterized by that. The first silicon waveguide has a first taper portion whose width decreases toward the predetermined direction on the second region,
The second silicon waveguide has a second taper portion whose width increases toward the predetermined direction on the second region.
The spot size converter according to claim 1. The length of the overlapping portion of the first taper portion and the second taper portion in the direction crossing the predetermined direction in the predetermined direction is the first taper portion and the second taper portion. Or less than the length in the predetermined direction of
The spot size converter according to claim 2. The optical axis of the second silicon waveguide is located substantially at the center of the silicon oxide waveguide in a direction intersecting the predetermined direction.
The spot size converter as described in any one of Claims 1-3 characterized by the above-mentioned. A void is formed on the outer periphery of the spot size conversion unit,
The spot size converter as described in any one of Claims 1-4 characterized by the above-mentioned.