Low-driving-voltage silicon electro-optical modulator

Abstract

The invention discloses a silicon electro-optical modulator with low driving voltage. The modulator comprises an input end directional coupler, upper and lower arm modulation waveguides, an output end directional coupler, a micro-ring resonant cavity and an electrode. The upper and lower arm modulation regions adopt PN doped structures, and the change of the effective refractive index of an optical waveguide mode relative to the modulation voltage under reverse bias of PN junctions is improved through cross doping, so that the driving voltage is reduced. When voltage is applied, the refractive index difference is generated between the upper arm and the lower arm, energy entering the micro-ring through coupling of the directional coupler is different, the coupling state of the micro-ring resonant cavity is changed, and then the intensity of output light is changed. The change amplitude of the refractive index along with the modulation voltage is improved through cross doping, the high-speed electro-optical modulator with the low driving voltage is achieved, and the high-speed electro-optical modulator has huge application prospects in the field of high-speed optical communication.

Description

Technical Field

[0001] The present invention relates to the field of integrated optoelectronic devices, and in particular to a low-driving-voltage silicon electro-optical modulator. Background Art

[0002] Silicon-based integrated platforms, with their CMOS (complementary metal oxide semiconductor) compatibility, wide transparency window, and high refractive index contrast, are becoming one of the most promising platforms for next-generation on-chip optical communications and optical interconnects. They have already enabled the development of a variety of high-performance active and passive devices, including modulators, detectors, filters, and beam splitters. Modulators are one of the most core components in silicon photonics. Traditional optical communications use lithium niobate (LNbO) modulators, which offer advantages such as wide bandwidth and low loss. However, due to LNbO's relatively weak electro-optical coefficient, LNbO modulators exhibit a large voltage-length product, requiring high drive voltages, and are incompatible with CMOS processes, limiting their application in large-scale integrated optical circuits.

[0003] Silicon modulators based on the carrier dispersion effect have advantages such as CMOS process compatibility and high modulation efficiency. Generally, to ensure that the modulator has a large electrical bandwidth, silicon modulators usually adopt a carrier depletion type. That is, the modulator PN junction operates in a reverse bias state. The change in carrier concentration caused by an external voltage can modulate the real and imaginary parts of the refractive index of the silicon material. The change in the refractive index of the material causes a change in the effective refractive index of the light field mode, which in turn causes a change in phase. Finally, the phase information is converted into intensity information through the interference structure. The structures currently used by mainstream silicon modulators include Mach-Zehnder interferometer structure and microring resonator structure. Both of these mainstream solutions currently have certain advantages and disadvantages. For the Mach-Zehnder interferometer structure, because the refractive index of the carrier depletion type changes little with voltage, its voltage-length product is large and the power consumption is high, which limits its application in large-scale integrated chips. The microring resonant cavity structure has a compact size, large bandwidth, and high modulation efficiency, but it has an inherent Q factor that limits its bandwidth. The larger the Q factor, the smaller the driving voltage applied during operation. However, due to the long photon lifetime, the photon oscillation time in the cavity is prolonged, and its electrical bandwidth will be limited. Summary of the Invention

[0004] This invention addresses the shortcomings of existing technologies and proposes a low-drive-voltage silicon electro-optic modulator. By utilizing cross-doped PN junctions and high-Q microrings to reduce the drive voltage, and by utilizing coupling-coefficient modulation, the resonant wavelength of the microring resonant cavity remains essentially unchanged, thus reducing the degree to which its Q factor limits the electrical bandwidth.

[0005] The technical solution adopted in the present invention is: A low-drive-voltage silicon electro-optic modulator comprises an input-end directional coupler, upper and lower-arm modulation waveguides, an output-end directional coupler, a microring resonant cavity, and electrodes. The upper and lower-arm modulation waveguides are ridge waveguides, and their modulation regions employ a PN doping structure. A PN junction is formed by periodic cross-doping, thereby increasing the amplitude of the change in the effective refractive index of the optical waveguide mode with the modulation voltage under reverse bias. The non-modulation region of the microring resonant cavity employs a wide waveguide to reduce losses, and a curved waveguide consisting of Euler bends and Bessel bends to reduce bending losses. The electrodes are arranged in a single-ended push-pull configuration, with a DC reverse bias voltage applied to the middle electrode. The upper and lower electrodes are a pair of high-speed GS electrodes. The input-end directional coupler splits incident light into the upper and lower arms. The refractive index difference between the upper and lower arms changes the coupling coefficient of the microring resonant cavity. This change in the coupling coefficient causes a change in the output light intensity of the microring resonant cavity. The output light is output through the output-end directional coupler, thereby modulating the intensity of the incident light.

[0006] The low-driving-voltage silicon electro-optic modulator, the input-end directional coupler, and the output-end directional coupler are used to realize beam splitting and beam combining, and the beam splitting ratio is 50 / 50.

[0007] In the low-driving-voltage silicon electro-optic modulator, the lower modulation arm of the modulator is part of a micro-ring resonant cavity, and the micro-ring resonant cavity has a high Q value, thereby achieving low driving voltage by reducing the energy required for critical coupling.

[0008] In the low-drive-voltage silicon electro-optic modulator, a wide waveguide is used in the non-modulation region of the micro-ring resonator of the modulator to reduce loss and achieve a high Q value; and a curved waveguide composed of Euler bending or Bessel bending is used in the 180° curved portion to reduce bending loss.

[0009] The electrodes of the low-drive-voltage silicon electro-optic modulator are arranged in a single-ended push-pull configuration. The middle electrode is applied with a DC reverse bias voltage of a depletion-mode modulator, and the upper and lower electrodes are a pair of high-speed GS electrodes.

[0010] The low-drive-voltage silicon electro-optic modulator has three different concentrations of P-type and N-type doping. The light doping concentration is used to form a PN junction; the heavy doping concentration has higher loss and is far away from the waveguide ridge, used to form an ohmic contact; and the medium doping concentration is used to connect the lightly doped part and the heavily doped part.

[0011] In the low-driving-voltage silicon electro-optic modulator, the lightly doped waveguide portion is cross-doped to form a periodic PN junction, thereby increasing the change of the effective refractive index of the light field with respect to the applied voltage and reducing the driving voltage; the medium-doping and heavy-doping portions are not periodically doped.

[0012] The beneficial effects of the present invention are: 1. The present invention is compatible with existing CMOS (complementary metal oxide semiconductor) processes and can be mass-produced at low cost compared to lithium niobate modulators.

[0013] 2. The present invention adopts a PN cross-doping scheme, which helps to improve modulation efficiency and reduce driving voltage.

[0014] 3. The present invention uses coupling coefficient modulation instead of resonant wavelength modulation. The resonant wavelength basically does not change, so the restriction of the resonant cavity Q factor on the electrical bandwidth will be weakened. 4. The present invention adopts a wide waveguide to reduce waveguide loss and realize a high-Q micro-ring cavity, which helps to reduce modulator loss and driving voltage. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 A structural diagram of a low-driving-voltage silicon electro-optic modulator.

[0016] Figure 2 It is a PN cross-doping structure in the modulation area.

[0017] In the figure: 1 is the input 3 dB directional coupler, 2 is the upper modulation arm, 3 is the lower modulation arm, 4 is the output 3 dB directional coupler, 5 is the microring resonator, 6 is the first high-speed GS electrode, 7 is the electrode, 8 is the second high-speed GS electrode, and 9 is a PN junction doping period. DETAILED DESCRIPTION

[0018] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0019] like Figure 1 As shown, a low driving voltage silicon electro-optic modulator includes an input directional coupler (input 3 The invention relates to a 3dB directional coupler 1, an upper and lower arm modulation waveguide (upper modulation arm 2, lower modulation arm 3), an output end directional coupler (output end 3dB directional coupler 4), a microring resonator 5 and an electrode 7 for providing a reverse bias voltage. The upper and lower arm modulation waveguides are ridge waveguides, and their modulation regions adopt a PN doping structure. PN junctions are formed by periodic cross-doping, thereby increasing the amplitude of the change of the effective refractive index of the optical waveguide mode with the modulation voltage under reverse bias. The non-modulation region of the microring resonator adopts a wide waveguide to reduce loss, and adopts a curved waveguide composed of Euler bending and Bessel bending to reduce bending loss. The electrodes adopt a single-ended push-pull arrangement, with a DC reverse bias voltage applied to the middle electrode. The upper and lower electrodes are a pair of high-speed GS electrodes (a first high-speed GS electrode 6 and a second high-speed GS electrode 8). The input end directional coupler splits the incident light into the upper and lower arms. The refractive index difference between the upper and lower arms changes the coupling coefficient of the microring resonator. The change in the coupling coefficient causes the output light intensity of the microring resonator to change. The output light is output through the output end directional coupler, thereby modulating the intensity of the incident light.

[0020] The present invention is based on a 220 nm thick silicon-on-insulator wafer. The silicon substrate is 525 μm thick, the buried oxide layer is 2 μm thick, and the ridge waveguide is etched to a medium thickness of 150 nm, leaving a 70 nm thick slab layer. The refractive index of silicon is approximately 3.47, and the refractive index of silicon dioxide is approximately 1.45. P doping is done with boron, and N doping is done with phosphorus. Light doping in the waveguide region forms a PN junction structure, such as Figure 2 As shown, this includes a complete doping period 9. The period size depends on the foundry's minimum doping feature size. Here, a doping period of 0.6 μm is selected, corresponding to a minimum doping feature size of 0.3 μm. At the same doping concentration, a smaller period results in more PN junctions and a stronger modulation capability of the effective refractive index of the optical waveguide mode. In this example, the operating band is the O band, the waveguide width is 0.38 μm, and the input optical field mode is the TE0 mode. The input optical field is split into two beams by a 3 dB directional coupler 1 at the input, entering the upper arm 2 and lower arm 3 of the modulator, respectively. Electrode 7 applies a voltage to reverse-bias the PN junction. A pair of high-speed GS electrodes (first high-speed GS electrode 6 and second high-speed GS electrode 8) form a single-ended push-pull structure. The electrodes are connected to heavily doped silicon via a via. The signal amplitude of the S electrodes changes the phase difference between the upper arm 2 and lower arm 3. This phase difference causes the intensity of the optical field entering the microring resonator 5 to change, thereby altering the coupling coefficient of the entire microring resonator. However, the resonant wavelength remains unchanged throughout this process. Finally, the output signal of the microring resonator is output via a 3 dB directional coupler 4 at the output port. The waveguide width in the non-modulation region of the microring resonator 5 is set to 1.5 μm. The multimode wide waveguide reduces the overlap between the transmission mode and the waveguide sidewalls, thereby reducing scattering losses caused by rough waveguide sidewalls. This is extremely important for achieving a high-Q microring cavity. The 180° bend uses an Euler bend with a minimum radius of 10 μm. The curvature of the Euler bend varies linearly with length, reaching zero curvature at the connection with the straight waveguide, where the curvature radius is infinite, minimizing losses and inter-mode crosstalk caused by mode mismatch.

[0021] Due to the optimized cross-doping structure and high-Q microring, the present invention can significantly reduce the driving voltage, achieving low insertion loss and low driving voltage for the modulator. At the same time, due to the inherent advantages of the coupled modulator, the resonant wavelength remains essentially unchanged, and the degree to which the Q factor of the resonant cavity limits the electrical bandwidth is weakened. The reason why the Q factor limits the bandwidth is that the larger the Q factor, the longer the photon lifetime at the resonant wavelength, resulting in a longer oscillation time of the light field in the cavity, thereby limiting the response speed of the modulator.

[0022] The embodiments described above may be further combined or replaced, and the embodiments are merely descriptions of preferred embodiments of the present invention and do not limit the concept and scope of the present invention. Various changes and improvements made to the technical solutions of the present invention by those skilled in the art without departing from the design concept of the present invention are within the scope of protection of the present invention. The scope of protection of the present invention is given by the appended claims and any equivalents thereof.

Claims

1. A low-drive-voltage silicon electro-optic modulator, characterized in that: The modulator includes an input-end directional coupler, upper and lower-arm modulation waveguides, an output-end directional coupler, a microring resonator, and electrodes. The upper and lower-arm modulation waveguides are ridge waveguides, and their modulation regions use a PN doping structure. A PN junction is formed by periodic cross-doping, which increases the amplitude of the change in the effective refractive index of the optical waveguide mode with the modulation voltage under reverse bias. The non-modulation region of the microring resonator uses a wide waveguide to reduce losses, and a curved waveguide consisting of Euler bending and Bessel bending to reduce bending losses. The electrodes are arranged in a single-ended push-pull configuration, with a DC reverse bias voltage applied to the middle electrode. The upper and lower electrodes are a pair of high-speed GS electrodes. The directional coupler at the input end splits the incident light into the upper and lower arms. The refractive index difference between the upper and lower arms changes the coupling coefficient of the microring resonator. The change in the coupling coefficient will cause the output light intensity of the microring resonator to change. The output light is output through the directional coupler at the output end, realizing the modulation of the incident light intensity.

2. The low driving voltage silicon electro-optic modulator according to claim 1, wherein: The input-end directional coupler and the output-end directional coupler are used to realize beam splitting and beam combining, and the beam splitting ratio is 50 / 50.

3. The low driving voltage silicon electro-optic modulator according to claim 1, wherein: The lower modulation arm of the modulator is part of a microring resonator with a high Q value, which reduces the energy required for critical coupling to achieve low driving voltage.

4. The low driving voltage silicon electro-optic modulator according to claim 1, wherein: The non-modulation area of ​​the modulator's micro-ring resonator uses a wide waveguide to reduce loss and achieve a high Q value; the 180° curved part uses a curved waveguide composed of Euler bending or Bessel bending to reduce bending loss.

5. The low driving voltage silicon electro-optic modulator according to claim 1, wherein: The modulator's electrode arrangement is single-ended push-pull; the middle electrode applies the DC reverse bias voltage of the depletion-mode modulator, and the upper and lower electrodes are a pair of high-speed GS electrodes.

6. The low driving voltage silicon electro-optic modulator according to claim 1, wherein: There are three different concentrations of P-type and N-type doping. The light doping concentration is used to form a PN junction; the heavy doping concentration has higher loss and is far away from the waveguide ridge, used to form an ohmic contact; the medium doping is used to connect the lightly doped part and the heavily doped part.

7. The low driving voltage silicon electro-optic modulator according to claim 1, wherein: The lightly doped waveguide part is cross-doped to form a periodic PN junction, which increases the change of the effective refractive index of the light field with the applied voltage and reduces the driving voltage; the medium-doped and heavily doped waveguides are not periodically doped.

Images