JP5135003B2 - Optical element, wavelength dispersion correction element, and phase modulation element - Google Patents

Description

  The present invention relates to an optical element capable of performing chromatic dispersion correction and phase modulation by applying a voltage or current applied to an optical waveguide section, and a chromatic dispersion correcting element and a phase modulation element using such an optical element.

  Patent Documents 1 to 3 below describe structures for performing chromatic dispersion correction, phase modulation, and the like in an optical waveguide including a photonic crystal.

Patent Document 1 (particularly FIG. 1) discloses a voltage or current applied to an optical waveguide using chromatic dispersion including a photonic crystal layer in which media having different refractive indexes are periodically arranged on a nanometer scale. A small chromatic dispersion correction element that controls chromatic dispersion by changing the above is described. In this wavelength dispersion correction element, a core layer is provided on a photonic crystal layer in which holes are periodically arranged, and a voltage is applied to an internal electrode on the top of the core layer, or a photonic crystal layer is applied to the photonic crystal layer. By applying a current or a voltage between the contacted terminal electrodes, the chromatic dispersion can be electrically controlled. For example, the material of the photonic crystal layer is silicon (Si), the hole is silicon dioxide (SiO ₂ ), the core layer is silicon nitride (Si ₃ N ₄ ), and the upper cladding and the lower cladding are SiO ₂ .

Patent Document 2 (particularly FIG. 1) discloses a photonic crystal part that applies predetermined phase modulation to an optical transmission line, and an energy application part that applies energy (voltage, current, etc.) to the photonic crystal part. The photonic crystal part is described as an optical element formed above or below the transmission line. In addition, a configuration example in which a photonic crystal part is arranged under the core of each branched waveguide of the Mach-Zehnder type waveguide as a small-sized and low-voltage optical modulator using this optical element is shown. Has been. The material constituting the Mach-Zehnder type waveguide is, for example, Si ₃ N ₄ , and the photonic crystal part is formed by arranging SiO ₂ disks in a triangular lattice pattern in the Si film. Since the photonic crystal portion has conductivity by adding impurities to the Si base material, when the applied voltage is changed, the refractive index changes due to the change in the concentration of electrons or holes in the base material. Thereby, the dispersion curve governing the characteristics of the photonic crystal part can be changed, and as a result, a phase shift can be generated.

  Patent Document 3 (particularly FIG. 3) describes a wavelength dispersion correction element that uses a spectral band in a band gap region of a photonic crystal and uses the photonic crystal as a cladding. When a linear defect is provided in a photonic crystal under a band gap condition, a gap mode is generated in a spectral region within the photonic band. That is, light is confined in the linear defect and guided along the linear defect. By connecting a plurality of regions having different periods of the photonic crystal in the waveguide direction, the sign of chromatic dispersion and the lowest-order and higher-order chromatic dispersion can be made variable by the action of electricity or heat.

  Patent Documents 4 to 6 below describe structures for generating phase modulation by changing the refractive index in a silicon-based optical waveguide that does not include a photonic crystal.

  Patent Document 4 (particularly FIG. 5) describes a structure in which a PN junction is formed in the plane of an SOI layer (Silicon On Insulator: a silicon layer formed on an insulating film) constituting a part of an optical waveguide. Has been. The SOI layer is located on the insulating layer, and an oxide layer and an SiN rib are formed thereon. At both ends of the PN junction, an electrical contact portion to the N-type Si region and an electrical contact portion to the P-type Si region are respectively provided.

  In Patent Document 5 (particularly FIG. 1), an SOI layer and a silicon layer are formed on the SOI layer, and each silicon layer has conductivity (P type or N type) having different polarities. A structure in which a PN junction is formed by sandwiching a dielectric film is described. By providing the dielectric layer between the P-type conductive layer and the N-type conductive layer, the current between the PN junction is cut off, and the power consumption is reduced. At the same time, the diffusion of holes and electrons is blocked, and the response speed of the electro-optic effect based on the change in density of holes and electrons is not limited by the diffusion speed of holes and electrons.

Patent Document 6 (particularly FIG. 1) has a structure in which a PN junction is formed of a semiconductor film (for example, the upper side is N-type and the lower side is P-type), and an insulating layer is interposed between them. Have been described. An example of the material of the lower semiconductor layer is an SOI layer, and an example of the insulating layer is SiON or SiO ₂ . By providing an insulating layer between the P-type semiconductor layer and the N-type semiconductor layer, the current between the PN junctions is cut off, and the power consumption is reduced. At the same time, the diffusion of holes and electrons is blocked, and the response speed of the electro-optic effect based on the change in density of holes and electrons is not limited by the diffusion speed of holes and electrons.

In Non-Patent Document 1, the wavelength dependence of the refractive index due to free carrier absorption is derived based on absorption spectrum data in silicon having P-polarity and N-polar conductivity. Further, an empirical formula is given as to how much the refractive index changes at wavelengths of 1550 nm and 1300 nm when the carrier concentration changes.
Non-Patent Document 2 describes a light intensity modulator using a photonic gap waveguide as a phase modulation unit and incorporating it into both branch portions of a Mach-Zehnder type waveguide. Here, the photonic gap waveguide uses a spectral band in the band gap region of the photonic crystal, and has a linear defect provided in the photonic crystal as a core and the photonic crystal as a cladding. A P-I-P junction is formed across the waveguide, and phase modulation is performed electrically. Since the slope of the dispersion curve of the gap mode is nearly horizontal, the phase change becomes large, and light intensity modulation with a short waveguide length becomes possible.
US Pat. No. 6,065,280 JP 2007-171422 A International Publication No. 2004/063797 US Patent Application Publication No. 2006/0133754 US Pat. No. 6,063,301 US Pat. No. 7,034,487 R. A. Soref and B.M. A. Bannett, “Electronic effects in silicon”, IEEE Journal of Quantum Electronics, 1987, No. 1, p. 123-129 Y. Jiang et al., “80-micron interaction length silicon photonic modulator,” Applied Physics Letters, 2005, Vol. 87, No. 22, p. 221105-1 to 221105-3

In the technique described in Patent Document 1, the internal electrode for applying a voltage is above the core layer. Further, the base medium on which the photonic crystal is formed has conductivity of only one of P-type and N-type. Accordingly, a voltage is applied to the internal electrode on the upper part of the core layer or the terminal electrodes on both ends of the photonic crystal layer, and the refraction of the base material accompanying the change in the density of holes (in the case of P type) or electrons (in the case of N type). When trying to use the rate change for the electro-optic element, there are the following problems.
-Since the upper electrode and the photonic crystal layer are separated from each other, the field effect is weakened and the voltage cannot be lowered.
-Since the base medium has conductivity of only one of P-type and N-type, current always flows when voltage is applied, and power consumption cannot be reduced. Even after the voltage is reduced to zero, the speed of density change is slowed by the diffusion of holes or electrons. Therefore, the rate of change in the refractive index accompanying the change in the density of holes or electrons becomes slow.

In the technique described in Patent Document 2, the base medium on which the photonic crystal is formed has conductivity of only one of P-type and N-type. Therefore, when a voltage is applied to the electrodes at both ends of the photonic crystal and the refractive index change of the base material accompanying the change in hole or electron density is used for the electro-optic element, there are the following problems.
-Since the base medium has conductivity of only one of P-type and N-type, current always flows when voltage is applied, and power consumption cannot be reduced. Even after the voltage is reduced to zero, the speed of density change is slowed by the diffusion of holes or electrons. Therefore, the rate of change in the refractive index accompanying the change in the density of holes or electrons becomes slow.

In the technique described in Patent Document 3, since the gap mode generated in the spectral region of the photonic gap is used, there are the following problems.
The waveguide mode exists only on the low frequency side of the light line in the dispersion characteristic, and the bandwidth is expected to be about 0.3 THz. On the other hand, when the C band is taken as an example of the frequency band used in the optical communication, the bandwidth is 4 THz. Therefore, it can be used only in a very limited spectral region.
Moreover, the junction structure regarding electrical control is not specified. Therefore, there are the following problems.
・ Current cannot be reduced and power consumption cannot be reduced.
・ Density of density change is slowed by diffusion of holes, electrons or both. Therefore, it is not possible to speed up the electro-optic response using the refractive index change accompanying the density change of holes, electrons, or both.

In the technique described in Patent Document 4, no photonic crystal is formed in the SOI layer. Further, in the PN junction formed in the plane of the SOI layer, the boundary between the P region and the N region where the junction bias is zero becomes a transition region in which the density of holes and electrons is gradually switched. For this reason, there are the following problems.
Since the photonic crystal is not formed, the efficiency of the electro-optic effect cannot be increased, and the size cannot be reduced (for example, the phase modulation unit is 1 mm or less).
-Holes and electrons are distributed because the boundary between the P region and the N region is a transition region. Therefore, when the refractive index change due to the density change of holes and electrons is generated, the junction current cannot be reduced and the power consumption cannot be reduced. Even if the voltage applied to the PN junction is changed at high speed from the forward bias condition to the reverse bias condition, the speed of density change is slowed by the diffusion of holes and electrons. Therefore, it is not possible to speed up the electro-optic response using the refractive index change accompanying the hole and electron density changes.

In the techniques described in Patent Documents 5 and 6, no photonic crystal is formed in regions (P region and N region) where the electro-optic effect is generated. The PN junction is formed in the vertical direction (direction perpendicular to the substrate surface). In the case of forming a junction perpendicular to the substrate surface, assuming that the element is made of a silicon-based material, the upper conductive layer is formed by laminating polysilicon and performing heat treatment. As described above, there are the following problems.
Since the photonic crystal is not formed, the efficiency of the electro-optic effect cannot be increased, and the size cannot be reduced (for example, the phase modulation unit is 1 mm or less).
-Light scattering by the grain boundary in the upper conductive layer is large, and the optical loss of the waveguide cannot be reduced. In the chromatic dispersion correction element, it is necessary to increase the waveguide length in order to increase the correction fee. However, since the loss is large, it cannot be applied to the chromatic dispersion correction element.

  In the technique described in Non-Patent Document 1, in silicon, only a refractive index change is given when the density of electrons or holes changes, and what function is realized using any configuration, It is not described at all.

The technique described in Non-Patent Document 2 has the following problems because it uses a gap mode generated in the spectral region of the photonic gap.
The waveguide mode exists only on the low frequency side of the light line in the dispersion characteristic, and the bandwidth is expected to be about 0.3 THz. On the other hand, when the C band is taken as an example of the frequency band used in the optical communication, the bandwidth is 4 THz. Therefore, it can be used only in a very limited spectral region.
The slope of the dispersion curve of the waveguide mode varies greatly even within the above 0.3 THz. For this reason, the chromatic dispersion is large, and the waveform of the transmission signal is distorted to deteriorate the transmission quality.
In addition, since the P-I-P junction or the P-I-N junction is used, there are the following problems.
-Since the boundary between the P region and the P region or the boundary between the P region and the N region is a transition region, holes, electrons, or both are distributed. Therefore, even if the bias is changed at a high speed, the speed of density change becomes slow due to diffusion of holes, electrons, or both. Therefore, it is not possible to speed up the electro-optic response using the refractive index change accompanying the density change of holes, electrons, or both.

  The present invention has been made in view of the above circumstances, and in an optical element utilizing the electro-optic effect, the power consumption is reduced, the electro-optic response is accelerated, the applied voltage is lowered, and the wavelength used is widened. It is an object of the present invention to provide an optical element that can be achieved, and a chromatic dispersion correction element and a phase modulation element using such an optical element.

In order to solve the above-described problems, the present invention provides an optical element in which an optical waveguide core includes a flat photonic crystal layer and a rectangular waveguide provided in a rib shape on one side of the optical waveguide core, and the photonic crystal layer Has a P-type region having P-polar conductivity and an N-type region having N-polar conductivity, and the P-type region and the N-type region are arranged in the plane of the photonic crystal layer. A gap region made of an insulating medium is provided along a line that overlaps the center line of the rectangular waveguide, the rectangular waveguide is made of an insulating medium, and the gap region overlaps the center in the width direction of the rectangular waveguide. The rectangular waveguide is directly laminated on the P-type region and the N-type region on both sides in the width direction, respectively .

  The photonic crystal layer is formed by periodically arranging cylindrical portions having a refractive index different from that of the matrix medium in the matrix medium, and the gap region is the same as the cylinder portion in the photonic crystal layer. The cylindrical part in the said gap area | region shall be distribute | arranged so that the center axis may cross | intersect the centerline of a gap area | region including the cylindrical part arrange | positioned with a period.

The cylindrical part in the photonic crystal layer may be made of the same material as the insulating medium constituting the gap region.
The photonic crystal layer may have a photonic gap on a higher frequency side than a frequency band of light guided to the optical waveguide core.
The photonic crystal layer may have an in-plane PN junction through the gap region.

The optical waveguide core has an oxide film provided as an insulating layer on a silicon substrate as a lower clad, the photonic crystal layer and the gap region are provided on the lower clad, and the rectangular waveguide Provided on the nick crystal layer and the gap region, an upper clad is provided on the optical waveguide core, a base medium of the photonic crystal layer is made of silicon, and a cylindrical portion in the photonic crystal layer and the insulating medium constituting the gap region is composed of SiO _2, the rectangular waveguide is made of _Si 3 _{N 4,} the upper cladding can be comprised of SiO _2.

The present invention also provides a chromatic dispersion correction element having the above-described optical element.
The present invention also provides a phase modulation element including the above-described optical element.

The present invention is also an optical element in which the optical waveguide core is a flat slab part and a rectangular waveguide provided in a rib shape on one side of the slab part, and the slab part has P-polar conductivity. A gap region made of an insulating medium separating the P-type region and the N-type region in the plane of the slab portion; and a rectangular region having a P-type region and an N-type region having N-polar conductivity. Provided along a line that overlaps the center line of the waveguide, the rectangular waveguide is made of an insulating medium, and the gap region is arranged to overlap the center in the width direction of the rectangular waveguide, An optical element is provided in which the rectangular waveguide is directly laminated on the P-type region and the N-type region .

  According to the present invention, in an optical element that changes a refractive index of a medium by applying a voltage or current applied to an optical waveguide portion, power consumption is reduced, electro-optical response is increased, applied voltage is decreased, and operating wavelength is used. It is possible to increase the bandwidth. The optical element of the present invention can be applied to a wavelength dispersion correction element, a phase modulation element, a light intensity modulation element, and the like.

The present invention will be described below with reference to the drawings based on the best mode.
<First embodiment>
FIG. 1 shows a first embodiment of the optical element of the present invention. In this figure, only a portion near the center in the longitudinal direction of the optical waveguide is shown, and there are voltage application electrodes (see FIG. 5) at both ends in the width direction of the optical waveguide core and at both ends in the longitudinal direction. The incident part and the emission part are not shown. In addition, in FIG. 1A, the rectangular waveguide only shows the outline, and the illustration of the cladding is omitted.

  The optical element 10 according to this embodiment includes a photonic crystal layer 11 having a flat optical waveguide core and a rectangular waveguide 14 provided in a rib shape on one side of the optical waveguide core. The photonic crystal layer 11 has P polarity. A P-type region 11A having N conductivity and an N-type region 11B having N-polar conductivity, and further, in the plane of the photonic crystal layer 11, a P-type region 11A and an N-type region 11B are provided. A gap region 13 made of an insulating medium is provided along a line overlapping the center line of the rectangular waveguide 14.

  In the example shown in FIG. 1B, the optical waveguide structure is formed on the substrate 17. Specifically, the lower clad 15 is provided on the substrate 17, the photonic crystal layer 11 and the gap region 13 are provided on the lower clad 15, and the rectangular waveguide 14 is further provided on the photonic crystal layer 11. An upper clad 16 is provided on the crystal layer 11 and the rectangular waveguide 14. The rectangular waveguide 14 may be disposed below the photonic crystal layer 11, but if the rectangular waveguide 14 is disposed above the photonic crystal layer 11, the flatness of the photonic crystal layer 11 is ensured. Is easy.

  The photonic crystal has a structure in which two substances having different refractive indexes are periodically arranged. The P-type region 11A and the N-type region 11B constituting the photonic crystal layer 11 are each provided with P-type or N-type conductivity by adding appropriate impurities to the photonic crystal medium. .

The base medium of the photonic crystal layer 11 has a higher refractive index than the lower cladding 15 and the upper cladding 16 in the wavelength band of the guided light, and is given conductivity of P polarity or N polarity by addition of impurities. Materials are used.
The P-type region 11A and the N-type region 11B may have opposite polarities regardless of the propagation direction of light guided by the optical waveguide core. Therefore, in FIG. 1, the left region is P-type and the right region is N-type. However, the polarities are reversed so that the left region is N-type and the right region is P-type. Does not affect.

  The impurity imparting conductivity (dopant) can be appropriately selected and used depending on the base medium. For example, when the base medium is a group IV semiconductor such as silicon, a group III element such as boron (B) is used as an additive that imparts P-type polarity, and phosphorus (P) or the like is used as an additive that imparts N-type polarity. Group V elements such as arsenic (As) can be used.

The portions 12 periodically arranged in the base medium of the photonic crystal layer 11 may be made of a material having a refractive index different from that of the base medium, and may be a conductive material or an insulating medium. The shape of the periodic array portion is not particularly limited, but is preferably cylindrical (cylindrical portion 12). Hereinafter, this portion will be referred to as a cylindrical portion 12 and will be described.
The arrangement of the columnar portions 12 may be any as long as it constitutes an appropriate two-dimensional lattice. For example, as shown in FIG. 1A and FIG. The two-dimensional lattice is continuously arranged across the P-type region 11A and the N-type region 11B.
The cylindrical portion 12 can be made of a material different from the insulating medium that forms the gap region 13, but is preferably made of the same material as the insulating medium that forms the gap region 13.

The gap region 13 is provided along a line that overlaps the center line of the rectangular waveguide 14. Here, the center line of the rectangular waveguide 14 is a line formed by connecting the center points of the cross section of the rectangular waveguide 14 (see FIG. 1B) along the longitudinal direction of the rectangular waveguide 14. Similarly, the center line of the gap region 13 is defined as a line formed by connecting the center points of the cross section of the gap region 13 (see FIG. 1B) along the longitudinal direction of the gap region 13.
At this time, the center line of the rectangular waveguide 14 and the center line of the gap region 13 are arranged parallel to each other along the longitudinal direction, and are perpendicular to the surface of the photonic crystal layer 11 (up and down in FIG. 1B). Direction). Therefore, in FIG. 1A, the gap region 13 is disposed so as to overlap the center of the rectangular waveguide 14 in the width direction.

  The gap region 13 has a lower refractive index than that of the photonic crystal layer 11 and the rectangular waveguide 14 in the wavelength band of the guided light, and separates the P-type region 11A and the N-type region 11B. ). The photonic crystal layer 11 has an in-plane PN junction via the gap region 13. The gap region 13 is disposed in the plane of the photonic crystal layer 11 and has the same thickness as the photonic crystal layer 11.

  In the present embodiment, the gap region 13 includes a cylindrical portion 18 arranged at the same period as the cylindrical portion 12 in the photonic crystal layer 11. As shown in FIG. 1B, the cylindrical portion 18 in the gap region 13 is arranged so that the central axis thereof intersects the center line of the gap region 13. Preferably, the central axis of the cylindrical portion 18 is arranged perpendicular to the surface of the photonic crystal layer 11.

The rectangular waveguide 14 has a rectangular cross section and is disposed at a position overlapping the gap region 13 in the vertical direction. As the constituent material of the rectangular waveguide 14, a material having a higher refractive index than the lower cladding 15 and the upper cladding 16 in the wavelength band of the guided light is used. The upper clad 16 is provided on the rectangular waveguide 14 in the region where the rectangular waveguide 14 exists, and on the P-type region 11A or the N-type region 11B in the region where the rectangular waveguide 14 does not exist.
The rectangular waveguide 14, the lower cladding 15 and the upper cladding 16 are made of an insulating medium.

  In the optical element 10 of the present embodiment, a composite core that confines and propagates light includes a flat plate (slab-like) photonic crystal layer 11 and a rectangular waveguide 14 having a rectangular cross section. That is, in the present invention, since the photonic crystal layer 11 is used as a part of the core, the photonic gap of the photonic crystal layer 11 is lower than the frequency band of light guided to the optical waveguide core. Moreover, it is necessary to be on the high frequency side. However, as will be described later, the frequency region of the dispersion curve of light guided to the photonic crystal layer has a tendency that the higher the mode, the higher the frequency. Therefore, in order to perform single mode propagation in the fundamental mode, the frequency band of light guided to the optical waveguide core needs to be on the lower frequency side than the photonic gap of the photonic crystal layer. In other words, the photonic crystal layer is designed so that the photonic gap is on the higher frequency side than the frequency band of light guided to the optical waveguide core. Thereby, as will be described later, the light guided to the optical waveguide core can be more effectively modulated by the change in the refractive index due to the application of voltage or the like to the photonic crystal layer 11.

However, the confinement efficiency of the photonic crystal layer 11 is low and the slab mode is set. Then, since light is mainly guided to the rectangular waveguide 14, the rectangular waveguide 14 is formed on the photonic crystal layer 11 in order to give a larger modulation based on the refractive index change of the photonic crystal layer 11. It is desirable to laminate directly on top.
The width of the photonic crystal layer 11 may be such that several or more cylindrical portions 12 are arranged on both sides of the gap region 13. In FIG. 1A, the cylindrical portion 12 is shown for 3.5 periods on both sides of the gap region 13, but is not particularly limited to this number.

An example based on silicon will be given for the material constituting each part and the dimensions of each part. In this embodiment, the base material constituting the photonic crystal layer 11 is silicon (Si), and the design value of the refractive index is 3.45. The material of the cylindrical portion 12 and the gap region 13 (including the cylindrical portion 18) was silicon oxide (SiO ₂ ), and the design value of the refractive index was 1.45. The length of one side of the regular triangle that is a unit cell (shown in the upper right of FIG. 1), that is, the pitch a was 403 nm, and the diameter 2r of the cylindrical portion 12 was 250 nm (radius r was 125 nm).

In order to generate P-type polarity conductivity in the P-type region 11A, boron (B) was ion-implanted into the silicon base material to generate holes. In order to produce N-type conductivity in the N-type region 11B, phosphorus (P) was ion-implanted into the silicon base material to generate electrons. The density of holes and electrons was 10 ¹⁹ / cm ³ at room temperature. The thicknesses of the P-type region 11A, the gap region 13 and the N-type region 11B were 100 nm.
The material of the rectangular waveguide 14 was silicon nitride (Si ₃ N ₄ ), the width was 1000 nm, and the thickness was 400 nm. The thickness of the lower clad 15 was 2000 nm, and the thickness of the upper clad 16 was 400 nm immediately above the rectangular waveguide 14.

  The factor that determines the gap width Wgap of the gap region 13 is that Wgap is narrowed as long as the current between the PN junctions is cut off by providing an insulating gap between the P-type conductive layer and the N-type conductive layer. Is preferred. As Wgap increases, the field effect weakens and the operating voltage increases, so it is preferable to reduce Wgap to the limit of the processing dimension. In the silicon device manufacturing process, the limit value of the processing dimension is generally defined by the resolution of optical drawing and is about 180 nm. Therefore, in this example, Wgap was set to 180 nm.

  As far as the present inventors know, it is unknown how the chromatic dispersion of the optical waveguide is affected by the presence of the insulating gap region 13. Therefore, in the above-described embodiment, the optical pulse propagation analysis of the waveguide structure including the gap region 13 was performed based on the time-domain difference method (finite-difference time-domain method) to elucidate the influence on the chromatic dispersion. As a result, it was confirmed that the presence of the gap region 13 does not change the sign of chromatic dispersion compared to the case where the gap region 13 does not exist (when the P-type region and the N-type region are directly joined). In addition, the absolute value of chromatic dispersion at that time decreases, but at Wgap = 180 nm, the reduction rate of chromatic dispersion is within 30%. Then, by designing the waveguide length to be extended by about 30% at the maximum compared to the case where the gap region 13 does not exist, chromatic dispersion equivalent to the case where the gap region 13 does not exist is generated. Can be used as

In this embodiment, the longitudinal direction of the rectangular waveguide 14 is parallel to the unit vector 1 in FIG. Here, the unit vector 1 is directed from the vertex 3 of any one of the equilateral triangles constituting the unit cell to the midpoint 5 of the base opposite to the vertex 3. Note that the characteristics of the optical element 10 are not affected even if the direction of the unit vector 1 is reversed and defined.
Note that there is another unit vector 2 in a two-dimensional lattice having equilateral triangles as unit cells. Unit vector 2 goes from vertex 3 to another vertex 4. In the present embodiment, the longitudinal direction of the rectangular waveguide 14 may be parallel to the unit vector 2.

The graph of FIG. 3 shows the result of calculating the properties of the waveguide mode obtained for the pitch a and the diameter 2r of the cylinder. This dispersion curve is obtained by the plane wave expansion method with the optical waveguide core regarded as a slab (flat plate). The vertical axis of the graph in FIG. 3 is the frequency of light, and the horizontal axis is the wave number. Both ends of the horizontal axis correspond to points (k _Γ ) where the wave number is zero, k _M indicates a point on the wave number space corresponding to the end point of the unit vector 2, and k _K indicates the wave number corresponding to the end point of the unit vector 1. A point in space. The light propagation direction is parallel to the unit vector 2 in the region (I) of FIG. 3 and parallel to the unit vector 1 in the region (III).

Therefore, in this embodiment in which the optical waveguide core is parallel to the unit vector 1, the region (III) is the corresponding wavenumber region, and the photonic gap is on the boundary line k _K between the region (II) and the region (III). Thus, it corresponds to the frequency region of the gap between the dispersion curve of the fundamental mode and the dispersion curve of the first-order mode. On the lower frequency side than the photonic gap, the total reflection condition is not satisfied for the primary mode, and the optical waveguide of the optical element 10 is a single mode waveguide. In the case of light propagating in the fundamental mode, the closer the frequency is to the photonic gap, the greater the chromatic dispersion and phase change, so the waveguide length required for desired chromatic dispersion correction and phase modulation is shortened or the applied voltage is reduced. can do.

  Therefore, in order for the entire target frequency band (for example, C band) to exist on the lower wavelength side than the photonic gap, the photonic gap of the photonic crystal layer is the frequency band of light guided to the optical waveguide core. The pitch a or the diameter 2r is determined so that it exists on the higher frequency side. For example, if a = 403 nm, 2r ≧ 240 nm. Further, when r / a = 0.31, a ≧ 390 nm.

  The frequency of the photonic gap may be set as appropriate from the viewpoint of shortening the waveguide length required for realizing optical functions such as chromatic dispersion correction and phase modulation. For example, when the wavelength band of the light to be used is the C band (1530 to 1560 nm), it is preferably about 192 to 196 THz when converted to the frequency band, and therefore the frequency of the photonic gap is preferably in the vicinity of, for example, about 200 THz. As an indication that the waveguide length can be reduced to about half that of the conventional C-band device, the frequency of the photonic gap is close to about 208 THz (about 1440 nm in wavelength). The difference between the band of light to be used and the photonic gap is approximately within 12 THz in frequency (within approximately 90 nm in wavelength).

When the optical waveguide core is parallel to the unit vector 2, the region (I) is a corresponding wavenumber region, and the photonic gap is on the boundary line k _M between the region (I) and the region (II). This corresponds to the frequency range of the gap between the dispersion curve of the fundamental mode and the dispersion curve of the first-order mode. In this case, according to the results of FIG. 3, the photonic gap on k _M is positioned on the lower frequency side than on the k _K. Thus, the entire frequency band of interest (e.g., C-band) in order to present more photonic gap in the low wavelength side, as compared with the case of the k _K, may be reduced pitch a or diameter 2r. However, as the dimensions become smaller, the effects of dimensional errors and defects become stronger due to processing technology limitations. The ratio (r / a) between the pitch a and the diameter 2r is selected from a range in which any cylindrical portion 12 does not overlap the adjacent cylindrical portion 12 or the gap region 13.

FIG. 4 is an explanatory diagram of a method for applying a voltage to the photonic crystal layer 11. In FIG. 4, the cylindrical portions 12 and 18, the lower clad 15, the upper clad 16, and the substrate 17 are not shown.
In the present embodiment, in order to apply a voltage to the photonic crystal layer 11, as shown in FIG. 4, electrodes 19A and 19B are provided at both ends of the P-type region 11A and the N-type region 11B, respectively. It is only necessary to connect 19B to the power supply 20A, 20B for applying voltage to generate a potential difference between the electrodes 19A, 19B. Examples of the material constituting the electrodes 19A and 19B include good conductors such as aluminum (Al). In order to reduce contact resistance, a material subjected to alloying by heat treatment is preferable. Power 20A, 20B, the electrode 19A, respectively to apply a voltage _{V 1} and _{V 2} to 19B.

Voltage _{V 1} of the P-type region 11A side is positive, if the voltage _{V 2} of the N-type region 11B side is negative, become forward biased condition, holes h are towards the N-type region 11B side from the P-type region 11A side The electrons e tend to flow from the N-type region 11B side toward the P-type region 11A side. However, since it is blocked by the gap region 13, it cannot flow to the opposite region. As a result, holes and electrons are accumulated in the region directly below the rectangular waveguide 14 with the gap region 13 therebetween, and the density of each increases. As the density of holes, electrons, or both increases, the refractive index decreases according to empirical rules, as described in Non-Patent Document 1. Since the region directly under the rectangular waveguide 14 is a region where the guided light exists, when a voltage is applied under a forward bias condition, the refractive index of the region where the guided light exists can be reduced.

On the other hand, in the voltage V ₁ of the P-type region 11A side is negative and when the voltage V ₂ of the N-type region 11B side is positive, becomes reverse biased condition, holes h and electrons e tended away from the gap region 13, In the region directly below the rectangular waveguide 14, the density of holes, electrons or both decreases and the refractive index increases. That is, when a voltage is applied under a reverse bias condition, the refractive index of the region where the guided light exists can be increased.

As described above, by changing the voltage value applied to the electrodes 19A and 19B, the refractive index of the base medium of the photonic crystal layer 11 existing in the region where the light is guided can be increased or decreased. As a result, the slope and curvature of the dispersion curve of the fundamental mode change, and the phase and chromatic dispersion can be made variable. In addition, even if either _{one of} the voltages V ₁ and V ₂ is set to zero, a potential difference can be generated between the P-type region 11A and the N-type region 11B, so that any one of the power supplies 20A and 20B may be omitted. . For example in the case of omitting the supply 20B of N-type region 11B side, a case of applying by increasing the absolute value of the voltage V ₁ of the P-type region 11A side as needed, both electrodes 19A, the voltage at 19B An equivalent increase / decrease in refractive index can be obtained. If neither the power 20A of the P-type region 11A side conversely increases as necessary absolute value of the voltage V ₂ of the N-type region 11B side. When one of the power sources is omitted, the electrode on the side where the power source is omitted may be grounded, or the electrode on the side where the power source is omitted may be omitted.

  As described above, when the photonic crystal layer 11 is composed of two regions having different polarities such as P-polarity and N-polarity, there are two types of carriers, that is, holes and electrons, compared to the case of unipolarity. Therefore, since it becomes easy to change the voltage of each region to attract or separate carriers from the waveguide region, the voltage that causes the same refractive index change as in the case of unipolarity can be lowered, and the voltage can be reduced. It becomes easy.

  When the optical element 10 is manufactured using a silicon-based material, an oxide film provided as an insulating layer on the substrate 17 is used as the lower clad 15, and a silicon layer formed as an SOI layer on the lower clad 15 is used. Thus, the photonic crystal layer 11 is preferably formed. In this case, since it is not necessary to laminate a polysilicon film as a base medium of the photonic crystal layer 11, there is no light scattering at the grain boundary in the polysilicon film, and there is no possibility of increasing the light loss.

Further, if the insulating gap region 13 is made of the same material as the cylindrical portion 12 in the photonic crystal layer 11, these can be formed simultaneously. For example, optical drawing and dry etching are performed on the SOI layer to remove the silicon in a groove shape and a cylindrical shape, and an insulating medium such as SiO ₂ is embedded in the groove and the cylindrical shape by vapor phase chemical growth. The portion where the insulating medium is laminated thicker than the silicon layer is flattened to the same height.

For other production methods, reference can be made to Patent Document 1 or Japanese Patent No. 3917170 which is a family patent thereof. For example, the substrate 17 may be formed of, for example, silicon (Si) and doped with impurities to provide conductivity.
The rectangular waveguide 14 is formed by laminating a layer of silicon nitride (Si ₃ N ₄ ) on the photonic crystal layer 11, and then patterning with a predetermined width and height, and Si ₃ N _{4 at a} position overlapping the gap region 13. It can be formed by leaving a layer. If the upper clad 16 has a shape raised on the rectangular waveguide 14 after the silicon oxide film (SiO ₂ ) is laminated on the photonic crystal layer 11 and the rectangular waveguide 14, unnecessary portions are subjected to CMP. It can be formed by removing it by polishing or the like and planarizing the SiO ₂ film.

In the present embodiment, the photonic crystal layer 11 having a “flat plate shape” includes a case where the photonic crystal layer 11 has a curved surface shape or a shape in which a plane and a curved surface are combined in a manufacturing process. The lattice includes an arrangement along the shape of the photonic crystal layer 11. However, considering the linearity of propagating light, the photonic crystal layer 11 is desirably formed in a planar shape as an example of a flat plate shape.
Further, the electrodes 19A and 19B are made of titanium / nickel alloy (TiNi) which is a relatively hard alloy at the contact surface with the photonic crystal layer 11, and the contact surface with a lead wire (not shown) is made of relatively soft aluminum / It is also preferable to form with a copper alloy (AlCu). In this case, the reliability of adhesion between the photonic crystal layer 11 and the electrodes 19A and 19B and adhesion between the electrodes 19A and 19B and the lead wires can be improved.

  In FIG. 5, the typical perspective view of the optical element 10 of a present Example is shown. In FIG. 5, illustration of the lower clad 15, the upper clad 16, and the substrate 17 is omitted. The electrodes 19A and 19B are assumed to exist over the entire length of the optical waveguide.

In the present embodiment, the change in chromatic dispersion has been described above. However, the optical element 10 of the present embodiment can cause a phase change by a change in applied voltage as described below, so that it can also be applied to a phase modulation element. Is possible.
The voltage _{V 1} applied to the P-type region 11A side + and _{V A,} when the voltage _{V 2} applied to the N-type region 11B side -V _A (i.e. _V 1 = _-V 2), the light propagated through the waveguide FIG. 6 shows the result of plotting the change in the phase with respect to the voltage _VA . FIG. 6A shows the voltage dependence of the phase at wavelengths of 1530 nm (□), 1545 nm (◯), and 1560 nm (Δ). The phase decreases as V _A increases. This is because the refractive index of the medium decreases as V _A increases, as described above with reference to FIG. When _VA is changed in the range from -5V to + 5V for each wavelength, the phase change becomes 3.14 rad or more in absolute value, and the phase change more than half wavelength required for light intensity modulation. was gotten. Since the phase change at 1560 nm is the smallest, operation as a light intensity modulator is possible in the entire C band of 1530 to 1560 nm. By using the photonic crystal layer 11 as a part of the waveguide core, the phase change is increased by about one digit compared to the waveguide without the photonic crystal.

The voltage dependence of the phase shown in FIG. 6 (a) is obtained from the wavelength dependence of the phase shown in FIG. 6 (b). The wavelength dependence of the phase is calculated by calculating the carrier density distribution based on the finite element method (FEM), and further based on the finite difference time-domain method (fine-difference time-domain method, FDTD). It was obtained by calculating the phase change. In FIG. 6B, _VA was changed at 2.5V intervals in the range from -5V to + 5V, and the values of _VA were plotted. In this case, the wavelength dependence of the phase is a parabolic shape having a positive curvature. This is a result reflecting the dispersion characteristics of the photonic crystal. Changing the voltage _VA changes the parabolic curvature. Since the refractive index of the medium decreases as V _A increases, the parabolic curvature decreases. Since the amount of chromatic dispersion is proportional to the curvature of the parabola representing the wavelength dependence of the phase, the chromatic dispersion can be changed by changing the applied voltage, and control by the chromatic dispersion voltage becomes possible.
6A and 6B, the phase origin is not an absolute phase origin, but is arbitrary.

As described above, in this example, it was shown that the phase and chromatic dispersion can be changed by changing the voltage applied to each of the electrodes 19A and 19B. Next, the fact that the current flowing at the time of voltage application is low and the low voltage / low current operation, that is, the low power operation is possible will be described based on the experimental result of the direct current-voltage characteristic shown in FIG. The horizontal axis in FIG. 7 is the potential difference (V ₁ −V ₂ ) between the electrodes 19A and 19B, and is based on the voltage V ₂ applied to the electrode 19B. Since V ₁ = V _A and V ₂ = −V _A as described above, the horizontal axis is 2V _A. Even when the potential difference changed from −30V to + 30V, the flowing current was 0.2 nA at the maximum in absolute value. In the voltage range (−5 to +5 V) shown on the horizontal axis of FIG. 6A, the maximum potential difference is 10 V, and the current is 0.03 nA or less in absolute value. Therefore, the power consumption in this voltage range is 0.3 nW or less, and it can be seen that the operation with very low power is possible.

<Second Embodiment>
FIG. 8 shows a second embodiment of the optical element of the present invention. The optical element 20 of the second embodiment has the same configuration as the optical element 10 of the first embodiment except that the cylindrical portion is omitted from the gap region 28. FIG. 8 is similar to FIG. It is represented. For example, when the gap width Wgap is 180 nm and the diameter 2r of the cylindrical part 18 is 260 nm in the first embodiment, about 80% of the cross-sectional area of the cylindrical part 18 is included in the gap width Wgap of the gap region 13. Therefore, depending on the exposure resolution and processing accuracy of the optical drawing apparatus in the manufacturing process, the shape of the cylindrical portion 18 at the boundary between the gap region 13 and the P-type or N-type regions 11A and 11B is not clear, and the cylindrical portion 18 The presence or absence of has no significant effect on chromatic dispersion. In that case, as shown in FIG. 8, the cylindrical portion may be removed from the gap region 23.

<Third and Fourth Embodiment>
Although the phase change increases according to the photonic crystal, the wavelength dependence of the phase also increases. In application to an optical modulator, it may be necessary to reduce the wavelength dependency of the phase in order to avoid fluctuations in the applied voltage due to wavelength differences in order to simplify the configuration of the power supply. In this case, there is a case where it is desired to reduce the influence of the photonic crystal even if the amount of change in phase due to fluctuations in the applied voltage is reduced. For this purpose, as shown in FIG. 3, one method is to reduce the number of cylinders constituting the photonic crystal.

  The optical element 30 of the third embodiment shown in FIG. 9 has the same configuration as the optical element 20 of the second embodiment except that the cylindrical portion 32 adjacent to the gap region 38 is omitted. In addition, FIG. 9 is represented similarly to FIG. Even in this case, the photonic crystal exists up to a region directly below the rectangular waveguide 14. If it is desired to further reduce the wavelength dependency, the outer cylindrical portion 32 may be removed in another row.

  The optical element 40 of the fourth embodiment shown in FIG. 10 has the same configuration as that of the optical element 30 of the third embodiment except that all the cylindrical portions are omitted. In this optical element 40, a homogeneous layer having P-polarity or N-type conductivity is provided instead of the photonic crystal. That is, the optical waveguide core of the optical element 40 is composed of a flat slab portion 41 and a rectangular waveguide 44 provided in a rib shape on one side of the slab portion 41. The slab portion 41 is a P-type region having P-polar conductivity. 41A and an N-type region 41B having N-polar conductivity, and a gap region 43 made of an insulating medium separating the P-type region 41A and the N-type region 41B is rectangular in the plane of the slab portion 41. It is provided along a line that overlaps the center line of the waveguide 44. In this case, the wavelength dependency can be minimized.

<Configuration example of wavelength dispersion correction element>
FIG. 11 shows a configuration example of the wavelength dispersion correcting element of the present invention using the optical element of the present invention. The chromatic dispersion correcting element 50 is provided in a form in which the above-described optical element 51 is enclosed in a package. An input optical fiber 52 is connected to the incident side of the optical element 51, and an output optical fiber is connected to the output side. 53 is connected. While propagating through a long-distance optical fiber transmission line, an optical pulse 54 whose time width is widened by the chromatic dispersion is incident on the optical element 51 through the input optical fiber 52. The optical element 51 is connected to power sources 56A and 56B for controlling the wavelength dispersion. These power, which corresponds to the power supply 20A, 20B in FIG. 4, one power 56A is a voltages _{V 1} is applied to the electrode of the P-type region side of the optical element 51, the other power supply 56B, an optical A voltage V ₂ is applied to the electrode on the N-type region side of the element 51. If the voltages V ₁ and V ₂ have different signs, the potential difference (V ₁ −V ₂ ) can be increased without increasing the absolute values of the voltages V ₁ and V ₂ . The two power supplies 56A and 56B may be incorporated into a single power supply device. One of the two power sources 56A and 56B may be omitted and grounded. When the optical element 10 of the first embodiment is used as the optical element 51, when the waveguide length of the optical element 51 is 100 mm, the time required for the chromatic dispersion to change due to a voltage change is 0.1 ms or less. Can do.

<Configuration example of phase modulation element and light intensity modulation element>
FIG. 12 shows a configuration example of the phase modulation element and the light intensity modulation element of the present invention using the optical element of the present invention. The optical intensity modulation element 60 shown in FIG. 12 is provided with the optical elements of the first to third embodiments in one branching waveguide 61A of the Mach-Zehnder type interference waveguide, and the first phase modulation section (phase modulation element) 62 and To do. The other branching waveguide 61 </ b> B is provided with the optical elements of Examples 1 to 3 to form a second phase modulation unit (phase modulation element) 63. The Mach-Zehnder type interference waveguide and both phase modulation sections are configured on the same substrate. As the optical waveguide in the Mach-Zehnder type interference waveguide portion, as shown in the fourth embodiment, one having no cylindrical portion constituting the photonic crystal is used.

An electrical signal is applied to the first phase modulation unit 62 from the first electrical signal source 64, and an electrical signal is applied to the second phase modulation unit 63 from the second electrical signal source 65. The electrical signal may be applied to both electrodes of the optical element constituting the phase modulation unit, or may be applied only to one electrode.
When continuous light is incident on the input waveguide 60A of the Mach-Zehnder type interference waveguide, the light is branched into the branch waveguides 61A and 61B, and the intensity of each is modulated by an electric signal. When the sign of the electrical signal is inverted between the first electrical signal source 64 and the second electrical signal source 65, the emitted light emitted from the output waveguide 60B does not generate chirp and has no distortion. Modulated light is generated. If the waveguide lengths of both phase modulation sections 62 and 63 are 1 mm, the voltage applied as an electric signal can be 4 V or less at maximum, and intensity modulation can be performed at a transmission rate of 40 Gbps (gigabit per second).

  On the other hand, in the document related to the above-mentioned Patent Document 6 (L. Liao et al., “High speed silicon Mach-Zehnder modulator”, Optics Express, 2005, Vol. 13, p. 3129-3135) Is 3.45 mm or more. Compared with this, according to the present embodiment in which the waveguide length is 1 mm, the waveguide length can be reduced to one third or less.

  The optical element of the present invention can be applied to a wavelength dispersion correction element, a phase modulation element, a light intensity modulation element, and the like.

A diagram showing a first embodiment of an optical element of the present invention, (a) is a top view, (b) is a sectional view taken along the P ₁ -P ₁ line in (a). It is explanatory drawing of the unit vector which prescribes | regulates the unit cell of a photonic crystal. It is a graph which shows an example of the dispersion characteristic of the waveguide mode in 1st Example of this invention. It is explanatory drawing of the density change of the carrier by the voltage application to a photonic crystal. It is a perspective view which shows the example of arrangement | positioning of the electrode for applying a voltage to a photonic crystal. It is a graph which shows an example of the voltage dependency of (a) phase in the example of 1st Embodiment of this invention, and the wavelength dependency of (b) phase. It is a graph which shows an example of the direct current-voltage characteristic in 1st Example of this invention. A diagram showing a second embodiment of an optical element of the present invention, (a) is a top view, (b) is a sectional view taken along the P ₂ -P _2-wire (a). A diagram showing a third embodiment of an optical element of the present invention, (a) is a top view, (b) is a sectional view taken along the P ₃ -P _3-wire (a). It is sectional drawing which shows 4th Example of the optical element of this invention. It is a schematic block diagram which shows an example of the wavelength dispersion correction element of this invention. It is a schematic block diagram which shows an example of the phase modulation element of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10, 20, 30, 40 ... Optical element 11, 21, 31 ... Photonic crystal layer, 41 ... Slab part, 11A, 21A, 31A, 41A ... P-type area | region, 11B, 21B, 31B, 41B ... N-type area | region , 12, 22, 32 ... cylindrical portion, 13, 23, 33, 43 ... gap region, 14, 24, 34, 44 ... rectangular waveguide, 15, 25, 35, 45 ... lower cladding, 16, 26, 36, 46 ... upper clad, 17, 27, 37, 47 ... substrate, 18 ... cylindrical portion, 19A, 19B ... electrode, 20A, 20B ... power source, 50 ... wavelength dispersion correction element, 60 ... light intensity modulation element.

Claims (8)

The optical waveguide core is an optical element composed of a planar photonic crystal layer and a rectangular waveguide provided in a rib shape on one side of the surface,
The photonic crystal layer has a P-type region having P-polar conductivity and an N-type region having N-polar conductivity, and the P-type region and the P-type region are disposed in the plane of the photonic crystal layer. A gap region made of an insulating medium separating the N-type region is provided along a line overlapping the center line of the rectangular waveguide, the rectangular waveguide is made of an insulating medium, and the gap region is formed of the rectangular waveguide. An optical element , wherein the rectangular waveguide is directly laminated on the P-type region and the N-type region on both sides in the width direction, respectively, overlapping with the center in the width direction .   The photonic crystal layer is formed by periodically arranging cylindrical portions having a refractive index different from that of the matrix medium in the matrix medium, and the gap region is the same as the cylinder portion in the photonic crystal layer. 2. The optical element according to claim 1, wherein the optical element includes a cylindrical portion arranged in a cycle, and the cylindrical portion in the gap region is arranged so that a central axis thereof intersects a center line of the gap region. .   3. The optical element according to claim 2, wherein the cylindrical portion in the photonic crystal layer is made of the same material as the insulating medium constituting the gap region.   The optical according to any one of claims 1 to 3, wherein the photonic crystal layer has a photonic gap on a higher frequency side than a frequency band of light guided to the optical waveguide core. element. The optical waveguide core has an oxide film provided as an insulating layer on a silicon substrate as a lower clad, the photonic crystal layer and the gap region are provided on the lower clad, and the rectangular waveguide Provided on the nick crystal layer and the gap region, an upper cladding is provided on the optical waveguide core,
The base medium of the photonic crystal layer is made of silicon, the cylindrical portion in the photonic crystal layer and the insulating medium constituting the gap region are made of SiO ₂ , the rectangular waveguide is made of Si ₃ N ₄ , The optical element according to claim 3, wherein the upper clad is made of SiO ₂ . Wavelength dispersion compensation element comprising an optical element according to any one of claims 1 to 5. Phase modulating element comprising an optical element according to any one of claims 1 to 5. The optical waveguide core is a flat slab portion and an optical element comprising a rectangular waveguide provided in a rib shape on one side of the surface,
The slab portion includes a P-type region having P-polar conductivity and an N-type region having N-polar conductivity, and the P-type region and the N-type region are disposed in a plane of the slab portion. A gap region made of an insulating medium is provided along a line overlapping a center line of the rectangular waveguide, the rectangular waveguide is made of an insulating medium, and the gap region is at the center in the width direction of the rectangular waveguide. An optical element characterized in that the rectangular waveguide is directly laminated on the P-type region and the N-type region on both sides in the width direction, respectively .

Images