Optical Modulator

Abstract

The prevent invention provides an optical modulator which enables adjustment of an optimum bias condition even in a dual electrode-type structure. Provided is an optical modulator which includes a semiconductor layer that has a p-n junction in an optical waveguide core and a high-frequency (RF) electrode and a ground electrode that are for applying an RF signal to the semiconductor layer and which modulates an optical signal by applying the RF signal and a bias voltage to the semiconductor layer, said optical modulator comprising a bias electrode that is disposed so as to face the ground electrode in a substrate, wherein the RF electrode and the bias electrode are connected to the semiconductor layer.

Description

[Technical Field]

[0001] The present invention relates to an optical modulator, and more particularly to an optical modulator that is used in optical communication systems, optical information processing systems, etc., performs high-speed optical modulation, has excellent frequency characteristics and waveform quality, and is capable of long-distance optical communication. [Background technology]

[0002] The spread of high-definition video distribution services and mobile communications has led to a massive increase in the amount of traffic flowing through networks, and this volume continues to grow every year. To build high-speed, high-capacity optical networks that can meet this traffic demand, vigorous efforts are being made to develop basic devices capable of high-speed operation to be used at each node. Optical modulators, which directly modulate optical signals with wideband baseband signals, are one of the key devices.

[0003] A Mach-Zehnder (MZ) optical modulator has a structure in which light incident on an optical waveguide is split into two waveguides at a 1:1 intensity ratio, and the split light is propagated a certain distance before being recombined. In an MZ optical modulator, the phase of the two split lights is changed by a phase modulation section provided in each of the two split optical waveguides. The intensity and phase of the light can be modulated by changing the optical interference conditions when the two phase-shifted lights are combined.

[0004] The materials that make up the optical waveguide of an MZ-type optical modulator are dielectrics such as LiNbO3, and semiconductors such as InP, GaAs, and Si. By inputting a modulating electrical signal to an electrode placed near an optical waveguide made of these materials and applying a modulating voltage to the optical waveguide, the phase of the light propagating through the optical waveguide is changed.

[0005] The mechanism for changing the phase of light in MZ-type optical modulators is the Pockels effect when the material is LiNbO3, the Pockels effect and the quantum confined Stark effect (QCSE) when the material is InP or GaAs, and the carrier plasma effect when the material is Si.

[0006] High-speed, low-power optical communications require optical modulators with high modulation speeds and low drive voltages. Specifically, optical modulation at speeds of 10 Gbps or higher and with amplitude voltages of several volts is required. To achieve this, traveling-wave electrodes are required that match the speed of high-speed electrical signals with the speed of light propagating through the phase modulator and allow the light and electrical signals to interact with each other while propagating. Optical modulators using traveling-wave electrodes, such as those disclosed in Non-Patent Document 1, have been put into practical use, with electrode lengths ranging from several millimeters to several tens of millimeters.

[0007] Optical modulators using traveling-wave electrodes require low-loss and low-reflection electrode and optical waveguide structures so that the amplitude of the electrical signal and the intensity of the light propagating through the waveguide can be propagated without any loss. That is, for electrical signals, an electrode structure with low reflection and propagation loss over a wide frequency range is required, and for light, a waveguide structure with low reflection that can efficiently confine light and propagate it without loss is required.

[0008] Among MZ-type optical modulators, the Si optical modulator, which has an optical waveguide made of Si, is a promising option from the perspective of substrate material and manufacturing process. Si optical modulators are manufactured from an SOI (Silicon on Insulator) substrate, which has a thin Si film attached to an oxide film (BOX: Buried Oxide) layer formed by thermally oxidizing the surface of a Si substrate. The optical waveguide is manufactured by processing the Si thin film into fine lines so that light can be guided through the SOI layer, and then injecting impurities to form p-type and n-type semiconductors. Finally, SiO2, which will become the optical cladding layer, is deposited, and electrodes are formed, etc., to complete the manufacturing process.

[0009] In this case, the optical waveguide must be designed and processed to minimize optical loss. Specifically, the p-type and n-type impurity doping and electrode fabrication must be designed and processed to minimize optical loss as well as reflection loss and propagation loss of high-speed electrical signals.

[0010] Figure 1 shows the cross-sectional structure of an optical waveguide, which is the basis of a conventional Si optical modulator. Figure 1 shows a cross section (xz plane) of an optical waveguide 200 constructed on an SOI substrate, cut perpendicular to the light propagation direction (y-axis). Light propagates in the direction perpendicular to the plane of the paper (y-axis direction). The optical waveguide 200 of the Si optical modulator consists of a Si layer 2 sandwiched between upper and lower SiO2 cladding layers 1 and 3. The Si nanowire formed in the center of Figure 1 for confining light has a structure called a rib waveguide, which has different thicknesses. That is, as shown in Figure 1, a rib waveguide consists of a thick central Si layer 201 and thin slab regions 202a and 202b on either side of it. The thick central Si layer 201 of the Si layer 2 serves as the core, and the optical waveguide 200 confines light propagating in the direction perpendicular to the plane of the paper by utilizing the refractive index difference between the central thick Si layer 201 and the surrounding SiO2 cladding layers 1 and 3.

[0011] Thin slab regions 202a and 202b on both sides of the optical waveguide core of the thick Si layer 201 are provided with a heavily doped p-type semiconductor layer 211 and a heavily doped n-type semiconductor layer 214, respectively. Furthermore, a pn junction structure consisting of a moderately doped p-type semiconductor layer 212 and a moderately doped n-type semiconductor layer 213 is formed in and near the optical waveguide core of the Si layer 201. As will be described later, a modulated electrical signal and a bias voltage are applied from both the left and right ends of the Si layer 2 in FIG. 1 via electrodes (not shown). Instead of the pn junction in the center of the core, a pin structure may be used in which an undoped i-type (intrinsic) semiconductor is sandwiched between the pn junction structure consisting of the moderately doped p-type semiconductor layer 212 and the moderately doped n-type semiconductor layer 213.

[0012] The phase modulation operation in the optical waveguide 200 of the Si optical modulator can be explained as follows. Although not shown in FIG. 1 , two metal electrodes are provided in contact with the high-concentration p-type semiconductor layer 211 and the high-concentration n-type semiconductor layer 214 at both ends of the Si layer 2, respectively. A reverse bias voltage is applied to the pn junction at the center of the core via the two metal electrodes, along with a radio frequency (RF) modulation electrical signal. That is, a voltage with a positive potential on the high-concentration n-type semiconductor layer 214 side and a negative potential on the high-concentration p-type semiconductor layer 211 side is applied from the right end to the left end (x-axis direction) of the optical waveguide 200. The reverse bias voltage and the modulation electrical signal change the carrier density inside the core of the thick Si layer 201. The change in carrier density changes the refractive index of the core 201 of the optical waveguide due to the carrier plasma effect, thereby modulating the phase of light propagating through the core of the optical waveguide.

[0013] The dimensions of the optical waveguide in a Si optical modulator depend on the refractive index of each material that makes up the core and cladding. An example is shown below for a rib-type silicon waveguide structure, as shown in Figure 1, which has a core portion made of a thick Si layer 201 and slab regions 202a and 202b on both sides of it. The width (x-axis direction) of the optical waveguide core 201 is 400 to 600 nm, the height (z-axis direction) of the core portion is 150 to 300 nm, the thickness of the slab region is 50 to 200 nm, and the length (y-axis direction) of the optical waveguide is several mm.

[0014] One of the outstanding features of Si optical modulators is that they can be made compact due to the large difference in refractive index between the Si core through which light propagates and the SiO2 cladding layer. This large difference in refractive index makes it possible to confine light tightly, allowing the bending radius of the optical waveguide to be made extremely small, at around 10 μm. This makes it possible to make the optical multiplexing / demultiplexing circuit section of the Si optical modulator, which will be explained next, compact.

[0015] (Conventional single-electrode Mach-Zehnder optical modulator) 2 and 3 show a Si optical modulator constituting a conventional single-electrode Mach-Zehnder optical modulator (see, for example, Non-Patent Document 2). FIG. 2 shows a planar structure of a Si (SOI) substrate surface (xy plane) seen from above, and FIG. 3 shows a cross-sectional structure taken along line III-III' in FIG. 2. In the plan view of FIG. 2, optical input from the left end of the optical modulator is branched into two optical waveguides 7a and 7b, and is phase-modulated by a modulating electric signal (RF signal) applied between the upper and lower RF electrodes 5a and 5b and the central DC electrode 6. After being modulated, the optical signals are recombined and output as modulated light from the right end of the optical modulator.

[0016] In the cross-sectional view of Fig. 3, the Si optical modulator has a basic structure in which two optical waveguides, each having a cross-sectional structure similar to that of the optical waveguide 200 shown in Fig. 1, are arranged symmetrically. Two high-frequency lines (RF electrodes 5a, 5b) for inputting a pair of differential modulation electrical signals (RF signals) are provided on both sides, and a DC electrode 6 for applying a common bias voltage between them is provided. Two optical waveguides 7a, 7b are provided between the two RF electrodes 5a, 5b, sandwiching the DC electrode 6, and pn junction structures are formed symmetrically within the optical waveguides 7a, 7b. The RF electrodes 5a, 5b are in contact with high-concentration p-type semiconductor layers 211a, 211b, respectively.

[0017] The DC electrode 6 is in contact with the central high-concentration n-type semiconductor layer 214, and a reverse bias can be applied to the two pn junctions on the left and right by applying a positive voltage to the DC electrode 6 relative to the RF electrodes 5a and 5b. As will be the case hereinafter, these electrodes and the semiconductor layers are each electrically connected by one or more vias 4 (through electrodes).

[0018] In such a single-electrode Si optical modulator, the RF electrode and DC electrode are electrically independent, and there is no need to actively apply a bias voltage to the RF electrode to apply a reverse bias to the p-n junction. Here, we have explained an example in which the RF electrode is in contact with the p-type semiconductor layer and the DC electrode is in contact with the n-type semiconductor layer, but it is also possible for the RF electrode to be in contact with the n-type semiconductor layer and the DC electrode to be in contact with the p-type semiconductor layer. In this case, the bias voltage applied to the DC electrode can be a negative voltage relative to the RF electrode, thereby applying a reverse bias to the p-n junction.

[0019] (Conventional dual-electrode Mach-Zehnder optical modulator) Figure 4 shows a Si optical modulator constituting a conventional dual-electrode Mach-Zehnder optical modulator. It shows a planar structure seen through the top surface (xy plane) of a Si (SOI) substrate. Optical input from the left end of the optical modulator is branched into two optical waveguides 7a and 7b, modulated, and then recombined to be output as modulated light from the right end of the optical modulator. While propagating in the y-axis direction through the two branched optical waveguides 7a and 7b, the input light is phase-modulated by modulation electrical signals (RF signals) applied to RF electrodes 15a and 15b, respectively. The optical modulator has a coplanar waveguide (CPW) for optical waveguide 7a, consisting of two ground electrodes 16a and 17 sandwiching RF electrode 15a. Similarly, a CPW for optical waveguide 7b, consisting of two ground electrodes 16b and 17 sandwiching RF electrode 15b, is also provided.

[0020] This is called a dual electrode structure because one Mach-Zehnder (MZ) optical modulator has two RF signal input sections. The MZ optical modulator shown in Figure 4 has a symmetrical structure with respect to the center line that passes through the center of the ground electrode 17 and is parallel to the y-axis.

[0021] FIG. 5 shows the cross-sectional structure along V-V′ in FIG. 4 , illustrating only one optical waveguide 7a to be modulated and a phase modulation section including a corresponding CPW. One phase modulation section is an optical waveguide having a cross-sectional structure similar to that of the optical waveguide 200 shown in FIG. 1 . It includes an RF electrode 15a, which is a high-frequency line for inputting one of a pair of differential modulated electrical signals (RF signals), and two ground electrodes 16a and 17 disposed on either side of the RF electrode 15a. One optical waveguide core 7a is disposed between the RF electrode 15a and the ground electrode 16a, and a pn junction structure consisting of a moderately doped p-type semiconductor layer 212 and a moderately doped n-type semiconductor layer 213 is formed within the optical waveguide 7a. The RF electrode 15a is in contact with a heavily doped n-type semiconductor layer 214 via a via 19b. The ground electrode 16a is in contact with a heavily doped p-type semiconductor layer 211 via a via 19a.

[0022] The ground electrode 17 is not in contact with any of the semiconductor layers, but together with the ground electrode 16a, it forms a high-frequency transmission line (CPW) with a GSG (Ground Signal Ground) structure for the RF electrode 15a. This transmission line structure adjusts the characteristic impedance of the RF electrode as a transmission line, improving transmission characteristics. Furthermore, because the signal line of the RF electrode 15a is surrounded by the two ground electrodes 16a and 17, it is possible to form an optical modulator with little signal leakage, crosstalk, or propagation loss.

[0023] Note that Figure 5 shows a phase modulation unit including an RF electrode 15a, which is a high-frequency line that inputs one of the modulated electrical signals (RF signals) of a differential configuration, but the phase modulation unit including the other RF electrode 15b also has a configuration similar to that of Figure 5, except that the arrangement order of multiple semiconductor regions in the x-axis direction is reversed with the z-axis as the axis of symmetry.

[0024] The characteristic impedance of the RF electrodes 15a and 15b of the Si optical modulator as a high-frequency transmission line is greatly affected by the capacitance of the pn junction of the optical waveguide cores 7a and 7b in the Si layer. However, since the capacitance between the RF electrode and the ground electrode also has an effect, in a Si modulator with a dual electrode structure, it is relatively easy to adjust the characteristic impedance by adjusting the capacitance between the RF electrode 15a and the ground electrode 17. It is possible to set the characteristic impedance to about 50 Ω in a single-ended drive configuration and to about 100 Ω in a differential drive configuration.

[0025] Here, an example configuration has been described in which the RF electrode 15a is in contact with the high-concentration n-type semiconductor layer 214, and the ground electrode 16a is in contact with the high-concentration p-type semiconductor layer 211. However, the orientation of the pn junction may be reversed, with the RF electrode 15a in contact with the high-concentration p-type semiconductor layer, and the ground electrode 16a in contact with the high-concentration n-type semiconductor layer. In this case, the pn junction can be reverse-biased by applying a negative voltage to the ground electrode 16a as the bias voltage superimposed on the RF signal and applied to the RF electrode 15a.

[0026] In such dual-electrode Si optical modulators, DC bias voltages are applied to the RF electrodes in duplicate, which requires ingenuity compared to single-electrode Si modulators, such as the need for a bias T when connecting to the driver IC. However, this has the advantage that it is relatively easy to adjust the characteristic impedance by controlling the capacitance between the RF electrode and the ground electrode. In addition, because it is surrounded by ground electrodes, there is little signal leakage, making it possible to create an optical modulator with little crosstalk and propagation loss.

[0027] (High frequency signal propagation) To achieve high-capacity optical communications, optical modulators capable of modulating light at high speeds are required. To achieve high-speed optical modulation, optical modulators must have frequency characteristics that enable them to operate over a wide frequency band, from several hundred kHz to several tens of GHz.

[0028] Referring to Figure 6, we will explain the ideal propagation state of a modulation signal in a dual-electrode Si optical modulator with a CPW structure. As described above, the RF electrode 15a is configured as a CPW sandwiched between two ground electrodes 16a and 17. In the CPW, a single RF electrode 15a is sandwiched between the ground electrodes 16a and 17 on both sides, and a high-frequency signal propagates over the RF electrode 15a. Electromagnetically, the propagation of a high-frequency signal can be explained as a model in which dense and sparsely charged areas move like waves on a high-frequency transmission line, such as the RF electrode 15a. In this model, when the densely charged area propagates, Coulomb interaction induces dense areas of opposite polarity on the ground electrode or the RF electrode that forms the differential line, and these dense areas move in the same way as the high-frequency signal.

[0029] A CPW is an unbalanced transmission line in which the charge on the RF line balances with charges of opposite polarity induced in two ground electrodes. Figure 6 shows a schematic representation of the propagation of a modulated signal using the charge transfer model described above, using positive and negative signs. When an RF signal 21 is input, positive and negative charges 22 and 23 appear alternately on the RF electrode 15a at a given time along the y-axis, which is the propagation direction of the electrical signal. Corresponding to the dense areas of positive charges 22 on the RF electrode 15a, dense areas of negative charges of opposite polarity 24a and 24b appear on the two ground electrodes 16a and 17, respectively.

[0030] In an ideal CPW, as shown in Figure 6, the charges on the two ground electrodes 16a and 17 propagate in the y-axis direction, aligned with the charges on the RF electrode 15a. For example, the charge distribution in the x-axis direction at the output terminal of the optical modulator at a given time is maximum positive at the center of the RF electrode 15a and negative on the ground electrodes 16a and 17 near the RF electrode 15a. At the output terminal, the polarity of the charges on each electrode changes over time. The RF signal 21 propagates through the optical waveguide 7a while modulating it and is terminated by a termination resistor (not shown) connected to the right end of the RF electrode 15a. However, in an actual CPW in a Si optical modulator, the asymmetric nature of the transmission line causes asymmetric charge movement, which makes it difficult to efficiently apply the modulation signal to the RF electrode 15a.

[0031] 7 and 8 show a dual-electrode Si optical modulator with a conventional CPW structure (see, for example, Patent Document 1). Fig. 7 shows a planar structure of a Si (SOI) substrate surface (xy plane) seen from above, Fig. 8(a) shows a cross-sectional structure taken along line VIIIa-VIIIa' in Fig. 7, and Fig. 8(b) shows a cross-sectional structure taken along line VIIIb-VIIIb' in Fig. 7. In order to eliminate the phase difference between the high-frequency signals of the two ground electrodes 16a and 17, a wiring electrode 41 is provided in the cladding layer 3 to connect the two electrodes.

[0032] Fig. 8(a) shows a structure similar to the conventional dual-electrode Si optical modulator shown in Fig. 5. That is, one optical waveguide core 7a is provided between an RF electrode 15a and a ground electrode 16a, and a pn junction structure made of a moderately doped p-type semiconductor layer 212 and a moderately doped n-type semiconductor layer 213 is formed within the optical waveguide core 7a. The RF electrode 15a is in contact with the heavily doped n-type semiconductor layer 214 through vias 42b and 44b. The ground electrode 16a is in contact with the heavily doped p-type semiconductor layer 211 through vias 42a and 44a.

[0033] FIG. 8(b) shows a cross section including the wiring electrode 41. The linear wiring electrode 41 connects the two ground electrodes 16a and 17 without forming a via connecting the RF electrode 15a to the Si semiconductor layer 214. Therefore, the wiring electrode 41 does not contact the RF electrode 15a or the via 42b. The wiring electrode 41 equalizes the potential between the two ground electrodes 16a and 17. Therefore, the phase difference between the return currents induced on the ground electrodes 16a and 17 by the high-frequency electrical signal on the RF electrode 15a and propagating through the ground electrodes 16a and 17, i.e., the phase difference between the charges propagating through the ground electrodes, is eliminated. This makes it possible to fabricate a wideband Si optical modulator with reduced degradation in the transmission and reflection characteristics of the RF signal 21 input to the RF electrode 15a.

[0034] In semiconductor optical modulators made of InP, Si, or the like, a reverse bias voltage is applied to the pn junction or pin junction at the center of the core along with a radio frequency (RF) modulating electrical signal. That is, a voltage with a positive potential on the high-concentration n-type semiconductor layer 214 side and a negative potential on the high-concentration p-type semiconductor layer 211 is applied from the right end to the left end (x-axis direction) of the optical waveguide 200. The reverse bias voltage and modulating electrical signal change the carrier density inside the core of the thick Si layer 201. The change in carrier density changes the refractive index of the optical waveguide core 201 through the carrier plasma effect, thereby modulating the phase of the light propagating through the core of the optical waveguide.

[0035] Fig. 9 shows a drive circuit for a conventional single-electrode Mach-Zehnder optical modulator, and Fig. 10 shows a drive circuit for a conventional dual-electrode Mach-Zehnder optical modulator. A driver IC 300 having an open-collector output stage is connected to one end of the RF electrode of the optical modulator, and a termination resistor R L is connected. The open collector driver does not have a sending end resistor and can be connected to an optical modulator with any impedance. Therefore, by using an open collector driver, the power consumed by the sending end resistor can be reduced, making it possible to provide an optical modulator with high power efficiency and low power consumption.

[0036] Also, since there is no need to place a DC blocking capacitor in the RF wiring between the driver IC 300 and the optical modulator 100, the mounting configuration can be made compact. Furthermore, power is supplied to the driver IC 300 via the RF electrodes 15a and 15b of the optical modulator, and the power supply voltage (VCC) to the driver IC 300 can be used as the bias voltage for the optical modulator. Therefore, a single power supply can supply the bias voltage and power to the driver IC.

[0037] However, because the optimal drive conditions for the driver IC are determined by adjusting the gain amount, the frequency band peaking amount, power consumption, etc., the value of the power supply voltage (VCC) is also determined by the optimal conditions for the driver IC. For this reason, if the power supply voltage (VCC) of the driver IC and the optimal bias voltage of the Si optical modulator are not the same, a voltage adjustment function is required for each. When advanced performance is required for the optical modulator, such as when higher speeds and long-distance transmission are required, it is necessary to adjust the bias voltage to the pn junction.

[0038] This problem can be solved with a single-electrode optical modulator, in which the RF electrode and DC bias electrode are provided separately.However, with a dual-electrode optical modulator, in which a DC bias voltage is applied to the RF electrode in an overlapping manner, the inability to adjust the optical modulator to its optimal bias conditions has been an issue. [Prior art documents] [Patent documents]

[0039] [Patent Document 1] Patent No. 6499804 [Non-patent literature]

[0040] [Non-Patent Document 1] Kazuhiro Goi, Kenji Oda, Hiroyuki Kusaka, Kensuke Ogawa, Tsung-Yang Liow, Xiaoguang Tu, Guo-Qiang Lo, Dim-Lee Kwong, "20Gbps Binary Phase Modulation Characteristics of Si Mach-Zehnder Push-Pull Modulator," 2012 IEICE Society Conference, C-3-50, 2012 [Non-patent document 2] Po Dong, Long Chen, and Young-kai Chen, "High-speed low-voltage single-drive push-pull silicon Mach-Zehnder modulators" Opt. Express vol.20, no.6, pp.6163-6169, 2012. Summary of the Invention

[0041] An object of the present invention is to provide an optical modulator that can adjust to an optimum bias condition even in a dual electrode structure.

[0042] In order to achieve the above object, one embodiment of the present invention provides an optical modulator including a semiconductor layer having a pn junction in an optical waveguide core, and an RF electrode and a ground electrode for applying a radio frequency (RF) signal to the semiconductor layer, and modulating an optical signal by applying a bias voltage to the semiconductor layer together with the RF signal, characterized in that the optical modulator further includes a bias electrode disposed opposite the ground electrode inside a substrate, and the RF electrode and the bias electrode are connected to the semiconductor layer. [Brief explanation of the drawings]

[0043] [Figure 1] FIG. 1 is a diagram showing the cross-sectional structure of an optical waveguide that is the basis of a conventional Si optical modulator. [Figure 2] FIG. 2 is a plan view showing a Si optical modulator constituting a conventional single-electrode Mach-Zehnder optical modulator. [Figure 3] FIG. 3 is a cross-sectional view showing a Si optical modulator constituting a conventional single-electrode Mach-Zehnder optical modulator. [Figure 4] FIG. 4 is a plan view showing a Si optical modulator constituting a conventional dual-electrode Mach-Zehnder optical modulator; [Figure 5] FIG. 5 is a cross-sectional view showing a Si optical modulator constituting a conventional dual-electrode Mach-Zehnder optical modulator. [Figure 6] FIG. 6 is a diagram for explaining the propagation state in a dual-electrode Si optical modulator with a CPW structure. [Figure 7] Fig. 7 is a plan view showing a dual-electrode Si optical modulator with a conventional CPW structure. [Figure 8] Figure 8 is a cross-sectional view of a dual-electrode Si optical modulator with a conventional CPW structure. [Figure 9] FIG. 9 is a diagram showing a driving circuit of a conventional single-electrode Mach-Zehnder optical modulator. [Figure 10] FIG. 10 is a diagram showing a driving circuit of a conventional dual-electrode Mach-Zehnder optical modulator. [Figure 11] FIG. 11 is a plan view showing a dual-electrode Si optical modulator according to a first embodiment of the present invention; [Figure 12] FIG. 12 is a cross-sectional view showing a dual-electrode Si optical modulator according to a first embodiment of the present invention; [Figure 13] FIG. 13 is a diagram showing the distribution of electric charges when a modulating electric signal is applied to the Si optical modulator of Example 1; [Figure 14] FIG. 14 is a cross-sectional view showing a dual-electrode Si optical modulator according to the first embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0044] Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. The optical modulator of this embodiment is a dual-electrode Si optical modulator with a CPW structure, and is provided with an electrode for applying a bias voltage to the RF electrode, separate from the bias voltage applied to the optical modulator together with an RF signal. Even with a dual-electrode structure, it is possible to adjust the bias conditions to the optimum, enabling higher-speed operation and providing a high-quality, high-performance optical modulator applicable to long-distance optical communications. Furthermore, the optical modulator of this embodiment can be fabricated using a CMOS-compatible process suitable for mass production, making it possible to provide a low-cost optical modulator. [Example]

[0045] 11 and 12 show a dual-electrode Si optical modulator according to a first embodiment of the present invention. Fig. 11(a) shows a planar structure of the Si (SOI) substrate surface (xy plane) seen from above, and Fig. 11(b) shows a planar structure with only the bias electrode extracted. Fig. 11(c) shows a planar structure excluding the bias electrode, which is the same as the conventional dual-electrode Si optical modulator shown in Fig. 4. Optical input from the left end of the optical modulator is branched into two optical waveguides 7a and 7b, modulated, and then recombined to be output as modulated light from the right end of the optical modulator. While propagating in the y-axis direction through the two branched optical waveguides 7a and 7b, the input light is phase-modulated by radio frequency (RF) modulation electrical signals applied to RF electrodes 15a and 15b, respectively. The optical modulator has a CPW consisting of two ground electrodes 16a and 17 sandwiching the RF electrode 15a for the optical waveguide 7a. Similarly, the optical waveguide 7b has a CPW consisting of two ground electrodes 16b and 17 sandwiching an RF electrode 15b. The optical modulator of Example 1 differs from conventional optical modulators in the structure of the ground electrodes, and will be described in detail with reference to a cross-sectional view.

[0046] Fig. 12(a) shows the cross-sectional structure taken along line XIIa-XIIa' in Fig. 11(a), and Fig. 12(b) shows the cross-sectional structure taken along line XIIb-XIIb' in Fig. 11(a). One optical waveguide core 7a is provided between the RF electrode 15a and the ground electrode 16a, and a pn junction structure made of a moderately doped p-type semiconductor layer 212 and a moderately doped n-type semiconductor layer 213 is formed within the optical waveguide core 7a. The RF electrode 15a is in contact with the heavily doped n-type semiconductor layer 214 through vias 42b and 44b.

[0047] Like the ground electrode 17, the ground electrode 16a is not in contact with any semiconductor layer and forms a CPW with the RF electrode 15a. In addition, a bias electrode 18a, which is close to the ground electrode 16a and disposed opposite the ground electrode 16a inside the substrate, is in contact with the high-concentration p-type semiconductor layer 211 through vias 42a and 44a. Furthermore, a bias electrode 19, which is also close to the ground electrode 17 and disposed opposite the ground electrode 16a inside the substrate, is provided. Thus, this optical modulator differs from conventional optical modulators in that the ground electrodes sandwiching the RF electrode 15a have a structure in which they are divided into a ground electrode and a bias electrode.

[0048] Furthermore, in the optical modulator of Example 1, a via is not formed to connect the RF electrode 15a to the Si semiconductor layer 214, and a linear wiring electrode 41 connects the two bias electrodes 18a and 19. Therefore, the wiring electrode 41 is not in contact with the RF electrode 15a or the via 42b. The two ground electrodes 16a and 17 are not connected by a wiring electrode, which also differs from conventional optical modulators.

[0049] The ground electrodes 16a and 17 are connected to a ground pad (not shown) of the driver IC or a ground pad of the package that houses the optical modulator. The bias electrodes 18a and 19 are connected to a pad (not shown) for supplying a bias voltage, and can apply a reverse bias voltage to each of the optical modulators individually.

[0050] The optical modulator of Example 1 is connected to a driver IC having an open-collector output stage. At this time, the reverse bias voltage (Vpn) applied to the pn junction of the optical modulator is the difference between the power supply voltage (VCC) applied to the RF electrodes 15a and 15b and the voltage (Vbias) applied to the bias electrodes 18a and 18b. Vpn=VCC-Vbias First, the power supply voltage (VCC) to the driver IC is determined so that the driving conditions of the driver IC are optimized. Next, the reverse bias voltage (Vpn) is set by adjusting the voltage (Vbias) applied to the bias electrodes 18a and 18b so that the optimal bias voltage is applied to the pn junction of the Si optical modulator.

[0051] This makes it possible to set optimal driving conditions for both the driver IC and the optical modulator, enabling faster operation and providing a high-quality, high-performance optical modulator that can be applied to long-distance optical communications.

[0052] FIG. 13 shows the charge distribution when a modulated electrical signal is applied to the Si optical modulator of Example 1. FIG. 13(a) shows the charge distribution in the cross section XIIa-XIIa' in FIG. 11(a), and FIG. 13(b) shows the charge distribution in the cross section XIIb-XIIb' in FIG. 11(a). The CPW is an unbalanced transmission line in which the charge on the RF transmission line balances with charges of opposite polarity induced in two ground electrodes. The charge is concentrated near the pn junction in the Si optical waveguide, where the distance between the electrodes is narrowest, but is also distributed in the electrodes themselves. FIG. 13 shows a state in which a positive charge exists on the RF electrode 15a at a certain point in time due to the input of an RF signal.

[0053] At this time, the charge induced in the bias electrode 18a via the pn junction is negative. Because the bias electrode 18a and the ground electrode 16a are adjacent and face each other across the thin SiO2 cladding layer 3, they function as a capacitor, inducing a positive charge in the ground electrode 16a. When an RF signal propagates through the electrode 15a, the induced charge also propagates to the bias electrode 18a and the adjacent ground electrode 16a. Therefore, the bias electrode 18a and the ground electrode 16a are separate in terms of direct current (DC) but can be considered as a single electrode in terms of alternating current (AC). In other words, the core of the optical modulator of Example 1 is AC-connected to the ground electrode for the radio frequency (RF) modulating electrical signal and DC-connected to the bias electrode for the bias voltage. Therefore, modulation by an RF signal is possible, and voltages based on separate potentials, the bias potential and the ground potential, can be applied.

[0054] If the bias electrode 18a and the ground electrode 16a are not close enough to each other or if the overlapping area is small, they do not function well as a capacitor and no charge is induced. In this case, the bias electrode 18a and the ground electrode 16a cannot be viewed as a single electrode in terms of RF voltage. The bias electrode 18a resonates with the high-frequency signal induced by the RF electrode 15a, resulting in poor RF signal propagation at certain frequencies. This can lead to performance degradation in optical modulators that require light modulation over a wide frequency range. Therefore, the more closely the bias electrode 18a and the ground electrode 16a are positioned on a Si (SOI) substrate, the more effective the structure becomes, not only in the Z-axis direction but also in the XY plane, the greater the overlapping area. Resonance due to RF signals is more likely to occur when the distance between the ground electrodes 16a and 17 and the ground pad, and the distance between the bias electrodes 18a and 19 and the pad for supplying bias voltage are 1 / 2 or 1 / 4 of the wavelength of the propagating RF signal. [Example]

[0055] FIG. 14 shows a dual-electrode Si optical modulator according to a second embodiment of the present invention. The optical modulator of the second embodiment differs from the conventional optical modulator and the optical modulator of the first embodiment in the structure of the ground electrode. Similar to FIG. 13(a), FIG. 14(a) shows a cross-sectional structure without the wiring electrode 41 connecting the two bias electrodes 18a and 19. Similarly to FIG. 13(b), FIG. 14(b) shows a cross-sectional structure with the wiring electrode 41. The bias electrode 18a and the ground electrode 16a are provided on the upper surface of the Si (SOI) substrate and are connected to the bias electrode 18c and the ground electrode 16c formed inside the SiO2 cladding layer 3 by vias, respectively. Although FIG. 14 shows one pair of electrodes formed in the z-axis direction, multiple pairs of electrodes may be formed. In this way, the area of ​​the capacitor formed by the bias electrode 18c and the ground electrode 16c can be increased.

[0056] This widens the frequency range over which the bias electrode 18a and the ground electrode 16a can be viewed as a single electrode in terms of RF voltage. That is, the larger the capacitor area and capacitance, the lower the frequency range that can follow the movement of charge. Also, the larger the capacitance of the capacitor, the greater the amount of charge stored between the electrodes, and the stronger the Coulomb force between the positive and negative charges, allowing the bias electrode 18a and the ground electrode 16a to be viewed as a single electrode in terms of RF voltage.

[0057] Furthermore, in the optical modulator of Example 2, the ground electrode 17 is omitted, and only the bias electrode 19 that was adjacent to the ground electrode 17 is provided. In this way, the bias electrode 19 and the ground electrode 17 can be regarded as one in terms of RF voltage, so it is possible to provide one of them and omit the other. [Industrial Applicability]

[0058] The present invention can be generally applied to optical communication systems, and in particular to optical modulators in optical transmitters of optical communication systems.

Claims

1. An optical modulator including a semiconductor layer having a pn junction in an optical waveguide core, an RF electrode and a ground electrode for applying a radio frequency (RF) signal to the semiconductor layer, and modulating an optical signal by applying a bias voltage to the semiconductor layer together with the RF signal, the ground electrodes include a first ground electrode and a second ground electrode that are electrically isolated from each other and disposed on either side of the RF electrode; The optical modulator comprises: a first bias electrode disposed inside the substrate and facing the first ground electrode; a second bias electrode disposed inside the substrate and facing the second ground electrode; the RF electrode and the first bias electrode are connected to the semiconductor layer; and an optical modulator, wherein the first bias electrode and the second bias electrode are electrically connected to each other by a wiring electrode;

2. a pair of the first ground electrode and the first bias electrode, and a pair of the second ground electrode and the second bias electrode each form a capacitor; 2. The optical modulator according to claim 1, wherein the semiconductor layer is connected to the first ground electrode and the second ground electrode in an alternating current manner for the RF signal, and is connected to the first bias electrode and the second bias electrode in a direct current manner for the bias voltage.

3. 3. The optical modulator according to claim 1, wherein the first ground electrode and the second ground electrode, together with the RF electrode, form a coplanar waveguide (CPW).

4. The optical modulator according to claim 1 or 2, characterized in that the set of the first ground electrode and the first bias electrode, and the set of the second ground electrode and the second bias electrode are each arranged opposite to each other in multiple layers stacked in a direction perpendicular to the surface (x-y plane) of the substrate.

Images