A modulator based on thin film lithium niobate-silicon hetero integration and a control method thereof

Abstract

This invention discloses a modulator and its control method based on thin-film lithium niobate-silicon heterointegration. The modulator includes a silicon substrate, a buried oxide layer, a silicon waveguide layer, and an upper cladding layer. The silicon waveguide layer contains two interferometer arms constituting a Mach-Zehnder interferometer. The phase modulation region of each interferometer arm is divided into at least three segments, including silicon-based phase modulation segments at both ends and a thin-film lithium niobate-silicon heterointegrated phase modulation segment in the middle. The silicon-based modulation segments utilize the plasmon dispersion effect for phase modulation, while the lithium niobate modulation segment utilizes the linear electro-optic effect for phase modulation. An inverse driving signal is applied to the silicon-based modulation segments of both arms, and an in-phase driving signal is applied to the lithium niobate modulation segment. This invention significantly improves modulation linearity and flexibly controls chirp parameters by synergistically utilizing the modulation characteristics of the two materials and independently controlling the amplitude and phase of the driving signal, while achieving low driving voltage and high modulation efficiency, and is compatible with CMOS processes.

Description

Technical Field

[0001] This invention belongs to the field of optoelectronic device technology, specifically relating to a modulator and its control method based on thin-film lithium niobate-silicon heterostructure integration. Background Technology

[0002] With the rapid development of technologies such as data center interconnection, 5G, and microwave photonics, the demand for high-speed, high-linearity, and low-power optical modulators is becoming increasingly urgent. Mach-Zehnder modulators have become core optical modulation devices due to their simple structure and ease of integration. Traditional silicon-based Mach-Zehnder modulators mainly utilize the plasmonic dispersion effect for phase modulation. Although compatible with CMOS processes, they have the following inherent drawbacks:

[0003] 1) High driving voltage and low efficiency: Due to the weak plasma dispersion effect of silicon, its V L (the product of half-wave voltage and length) is typically around 2V. A length of about cm often requires an arm length of millimeters or a driving voltage of tens of volts, resulting in high power consumption, which contradicts the low-power requirements of modern communication.

[0004] 2) Inherent nonlinear distortion: The relationship between carrier concentration and refractive index is nonlinear, which leads to severe signal distortion, especially for high-order modulation formats (such as 64-QAM) and analog optical links, severely limiting the spurious-free dynamic range (SFDR) of the system.

[0005] 3) Chirp: Under ideal and symmetrical operating conditions, the Mach-Zehnder modulator can achieve zero chirp or controllable chirp. However, in practical applications, due to device manufacturing deviations and operating point drift, it is not absolutely zero chirp. In high-speed, long-distance optical fiber transmission, this chirp will cause dispersion and degrade signal quality.

[0006] 4) High transmission loss: The high concentration of doping required to achieve electro-optic modulation introduces significant optical loss, reducing the optical output power of the modulator.

[0007] To overcome the weaknesses of silicon, existing technologies have employed heterogeneous integration of silicon and thin-film lithium niobate. Lithium niobate possesses strong Pockels effect, high linearity, and low optical loss; however, these solutions typically simply replace the silicon phase modulation arm entirely with lithium niobate, failing to fully leverage the synergistic optimization of the two materials' properties. For example, lithium niobate modulators produce zero chirp under ideal push-pull operation, but in some applications (such as dispersion management links), a certain degree of negative chirp may be beneficial, and lithium niobate modulators are difficult to adjust flexibly. Furthermore, existing solutions lack sophisticated structural design in terms of how to systematically compensate for nonlinear distortion and how to achieve independent optimization of drive voltage, linearity, and chirp within a single device. Therefore, a novel modulator structure capable of comprehensively addressing these issues is urgently needed. Summary of the Invention

[0008] The purpose of this invention is to overcome the shortcomings of the prior art and to simultaneously achieve extremely low driving voltage, high linearity (low signal distortion), near-zero or controllable chirp, and maintain a compact structure compatible with CMOS processes, thereby providing a modulator and its control method based on thin-film lithium niobate-silicon heterointegration.

[0009] To achieve the above objectives, the technical solution provided by this invention is as follows:

[0010] A modulator based on thin-film lithium niobate-silicon heterostructure integration and its control method are disclosed. The modulator includes an input waveguide, a beam splitter, a first interferometer arm, a second interferometer arm, a beam combiner, and an output waveguide. The modulator comprises two phase modulation segments made of different electro-optic materials, wherein at least one phase modulation segment is a silicon-based modulation segment based on carrier dispersion effect, and at least the other phase modulation segment is a lithium niobate-based modulation segment based on linear electro-optic effect. The silicon-based modulation segment and the lithium niobate-based modulation segment are configured with independent electrodes for applying independent electrical drive signals.

[0011] Multiple phase modulation segments composed of different electro-optic materials are alternately integrated on the two interferometer arms of an MZM. By applying asymmetric and independently controllable driving signals to these segments, the modulation characteristics of different materials produce a synergistic and compensating effect as a whole.

[0012] The basic structure includes an input waveguide, a 1x2 optical power splitter (such as a Y-branch or multimode interference coupler), two parallel interferometer arms, a 2x1 optical power combiner, and an output waveguide; these passive components are implemented using standard silicon waveguides to take advantage of their low transmission loss and mature fabrication technology.

[0013] Heterogeneous integrated phase modulation region: This is the core of this invention application. Each interferometer arm is divided into at least three phase modulation regions, preferably three (but can be more odd-numbered, such as five), forming a symmetrical "sandwich" structure: The first and third segments (end segments) are silicon-based phase modulation segments. In these segments, electrical modulation structures, such as PN junctions, PIN junctions, or MOS capacitor structures, are formed on the silicon waveguide. Phase modulation is achieved by injecting or depleting charge carriers using the plasma dispersion effect. The second segment (middle segment) is a thin-film lithium niobate-silicon heterogeneous integrated phase modulation segment. In this segment, a thin-film lithium niobate layer is integrated on the silicon waveguide using technologies such as wafer bonding. This segment utilizes the strong linear electro-optic effect (Pockels effect) of lithium niobate for phase modulation.

[0014] Electrode arrangement: Independent traveling wave electrodes are provided for the silicon modulation section and the lithium niobate modulation section respectively. The RF drive signals applied to the silicon-based modulation section of the two interferometer arms are out of phase (i.e., differential or push-pull drive, +V_RF and -V_RF), and the RF drive signals applied to the lithium niobate modulation section of the two interferometer arms are in phase (i.e., common-mode drive, +V_RF or +V_RF, with the same phase).

[0015] Adjustable electrical network: connected between the RF signal source and the electrodes, used to independently control the amplitude ratio of the RF signal applied to the silicon segment and the lithium niobate segment. ) and relative phase ( The network can be composed of adjustable attenuators and adjustable phase shifters.

[0016] The unique structure and driving scheme described above result in the following synergies and advantages:

[0017] Synergistic compensation for linearity: Phase variation introduced by silicon modulation segment _Si is a nonlinear function of voltage V; the phase change introduced by the lithium niobate modulation segment. _LN is highly linear. _LN V. Under the driving scheme of this invention, the total phase change of one interferometer arm is: _arm1 = _LN(V) + _Si(-V), the other arm is _arm2 = _LN(V) + _Si(+V) (assuming the lithium niobate segment is driven in phase and the silicon segment is driven in reverse phase). This is achieved by optimizing the amplitude ratio. This can cause a change in the total phase ( _arm1 The second derivative of _arm2) (d²(ΔΦ_arm1 - ΔΦ_arm2) / dV²) approaches zero near the operating point, thereby systematically compensating for nonlinear distortion, achieving extremely high linearity, and significantly improving SFDR.

[0018] Flexible control of chirp: In an ideal push-pull MZM, the phase changes of the two arms are equal in magnitude and opposite in direction, resulting in intensity modulation (zero chirp). This invention can precisely introduce a small phase imbalance by independently controlling the DC bias of the lithium niobate segment or fine-tuning the amplitude of its drive signal; for example, by applying an additional DC bias to the lithium niobate segment of one arm. The DC modulation of the lithium niobate in both arms results in a phase change that is no longer perfectly symmetrical. This controllable asymmetry allows for precise tuning of the chirp parameters of the output optical pulse, switching between positive chirp, zero chirp, and negative chirp to accommodate transmission fibers with different dispersion characteristics and optimize transmission performance.

[0019] High efficiency and low power consumption: The lithium niobate intermediate segment provides highly efficient linear modulation, which reduces overall power consumption. The silicon end segment only undertakes part of the modulation task, reducing the requirements for its driving voltage and power consumption; at the same time, the compactness of silicon devices allows the overall device size to remain small.

[0020] Process compatibility and robustness: This structure makes full use of the mature processing technology of silicon and the excellent electro-optic properties of lithium niobate, avoiding the high cost and integration difficulties of the all-lithium niobate solution, and combining high performance and manufacturability.

[0021] To further overcome the inherent nonlinearity of the cosine transfer function of the Mach-Zehnder modulator, this invention also provides an adaptive optical domain linearization scheme: the modulator system further includes: an auxiliary Mach-Zehnder modulator, arranged in parallel with the main segmented Mach-Zehnder modulator; an optical beam combiner for combining the output optical signals of the main modulator and the auxiliary modulator; a real-time monitoring feedback unit for detecting the degree of nonlinear distortion of the output optical signal; and an adaptive control unit for dynamically adjusting the bias point, drive signal amplitude, and relative phase of the auxiliary modulator according to the monitoring results; wherein, the auxiliary modulator is configured to generate an anti-nonlinear optical signal complementary to the cosine transfer function of the main modulator, and the nonlinear distortion is directly canceled in the optical domain through optical synthesis.

[0022] Compared with the prior art, the present invention has the following beneficial technical effects:

[0023] 1. The modulator of the present invention can simultaneously achieve extremely low drive voltage, high linearity (low signal distortion), near-zero or controllable chirp, and maintain a compact structure compatible with CMOS technology.

[0024] 2. The modulator proposed in this invention alternately integrates multiple phase modulation segments made of different electro-optic materials on the two interferometer arms of an MZM, and by applying asymmetric and independently controllable driving signals to these segments, the modulation characteristics of different materials produce a synergistic and compensating effect as a whole. Attached Figure Description

[0025] Figure 1 This is a three-dimensional structural diagram of a modulator based on thin-film lithium niobate-silicon heterostructure integration and its control method, according to an embodiment of the present invention.

[0026] Figure 2 This is a top view along the light propagation direction of an embodiment of the present invention.

[0027] Figure 3 The figures show cross-sections of the modulation section in an embodiment of the present invention, wherein (a) is a cross-sectional view of the silicon modulation section, and (b) is a cross-sectional view of the lithium niobate-silicon heterojunction modulation section.

[0028] Figure 4 This is a schematic diagram illustrating the linearity compensation principle of an embodiment of the present invention.

[0029] Figure 5 This is a comparison of the linear responses of embodiments of the present invention.

[0030] Figure 6 This is a comparison of the Q values ​​of the transmitted signals in an embodiment of the present invention.

[0031] The annotations in the attached figures are explained as follows:

[0032] 1-Silicon substrate, 2-BOX, 3-Clad layer, 4-First silicon modulation section, 5-Lithium niobate modulation section, 6-Second silicon modulation section. Detailed Implementation

[0033] To make the objectives, advantages, and features of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. Those skilled in the art should understand that these embodiments are merely used to explain the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0034] like Figure 1 , Figure 2As shown, this embodiment provides a modulator and its control method based on thin-film lithium niobate-silicon heterointegration. The modulator is fabricated on an SOI platform, consisting of a silicon substrate 1, a BOX 2, and a cladding 3 from bottom to top. Input light is introduced into the input waveguide via an edge coupler or grating coupler, and then split into two paths by a 1x2 multimode interferometer (MMI), entering the first and second interferometer arms. The phase modulation region of each interferometer arm is divided into three segments:

[0035] First silicon modulation section 4: A ridge-type silicon waveguide is used, with P-type and N-type doping on both sides to form a PN junction.

[0036] Lithium niobate modulation section 5: Thin-film lithium niobate is bonded to the silicon waveguide through a silicon dioxide dielectric layer to form a hybrid waveguide mode; electrodes are deposited on the lithium niobate layer.

[0037] The second silicon modulation segment 6 has the same structure as the first silicon modulation segment 4.

[0038] like Figure 3 As shown in (a), the cross-section of the silicon modulation section shows that the silicon waveguide is located on the BOX 2 layer, with P and N doped regions on both sides, and electrodes connected to it.

[0039] like Figure 3 As shown in (b), the cross-section of the lithium niobate modulation section 5 shows that the silicon waveguide is located on the BOX 2 layer, and a thin film of lithium niobate is integrated on top through a silicon dioxide bonding layer. The top electrode is used to apply an electric field.

[0040] One signal passes through an adjustable attenuator and an adjustable phase shifter and is input to the first set of electrodes, which drive the silicon modulation segments of the two arms in an inverted manner; the other signal passes through another adjustable attenuator and is input to the second set of electrodes, which drive the lithium niobate modulation segments 5 of the two arms in an in-phase manner; a DC bias source can be applied to the lithium niobate segment electrodes for chirp fine-tuning.

[0041] Finally, the light from both arms is interfered with by another MMI and output through the output waveguide.

[0042] The manufacturing process mainly includes: SOI wafer preparation, silicon waveguide etching, silicon modulation region doping, silicon dioxide cladding deposition and planarization, thin-film lithium niobate wafer bonding and thinning, lithium niobate waveguide patterning, electrode region opening, and metal electrode deposition.

[0043] like Figures 4 to 6 As shown, linearity optimization includes the following steps:

[0044] Step 1: Input a two-tone radio frequency signal (e.g., two single-tone signals with frequencies around 20 GHz and an interval of 100 MHz) as the modulation signal V_RF;

[0045] Step 2: Send the output optical signal of the modulator into a photodetector to convert it into an electrical signal, and observe its spectrum using a spectrum analyzer;

[0046] Step 3: Ensure that the adjustable phase shifter is set correctly so that the radio frequency signal applied to the lithium niobate modulation section 5 and the radio frequency signal applied to the silicon modulation section maintain the required in-phase / out-of-phase relationship;

[0047] Step 4: Adjust the adjustable attenuator that controls the RF drive power of the silicon-based modulation segment (i.e., adjust the scaling factor α), and observe the power change of third-order intermodulation distortion (IMD3) in the spectrum.

[0048] Step 5: Fine-tune the attenuator until the power of IMD3 is suppressed to a minimum; record the value of the adjustable attenuator at this time. This state is the optimal operating point of the modulator for linearity.

[0049] Adaptive optical domain linearization example:

[0050] 1) Optical path configuration:

[0051] The input light is split into two paths by a 1×2 optical beam splitter. One path enters the main segmented MZM, and the other enters the auxiliary MZM. The auxiliary MZM can adopt a simplified structure, and a thin-film lithium niobate phase modulator is preferred to obtain low latency. High linearity; the main and auxiliary optical signals are combined and output through a beam combiner.

[0052] 2) Electrical control system:

[0053] The main RF signal is split into two paths by a power divider. The main path signal drives the segmented electrodes of the main MZM through an adjustable attenuator and a phase shifter, while the auxiliary path signal drives the auxiliary MZM through an adjustable attenuator and a nonlinear transformation circuit. The adaptive control processor receives the monitoring feedback signal in real time and outputs control signals to each adjustable element.

[0054] 3) Adaptive optimization process:

[0055] Initialization phase: Set the auxiliary MZM bias to the minimum transmission point and the main MZM bias to the orthogonal operating point;

[0056] Distortion monitoring: Inject a two-tone test signal into the system and monitor the IMD3 power of the output signal;

[0057] Parameter scan: fine-tune the bias, amplitude, and phase parameters of the auxiliary MZM sequentially;

[0058] Gradient calculation: Calculate the gradient direction of each parameter based on the distortion change;

[0059] Iterative update: Update parameters along the negative gradient direction until IMD3 is below the target threshold;

[0060] Locked state: Enters tracking mode, periodically fine-tuning to compensate for environmental drift.

[0061] Chirp regulation:

[0062] 1) After completing linearity optimization (i.e., determining the scaling factor) After that, the drive signal is switched to a 40 GBaud PAM4 pseudo-random sequence.

[0063] 2) Test the modulator output optical signal in the following two scenarios:

[0064] a) Back-to-back connection;

[0065] b) Through a standard single-mode fiber of a specific length (e.g., 1 kilometer).

[0066] 3) Use a high-speed oscilloscope to observe and record the eye diagram after transmission through the optical fiber.

[0067] 4) Fine-tune the DC bias voltage of the lithium niobate modulation segment 5 on the first and second interferometer arms to introduce a controllable phase imbalance.

[0068] 5) For each bias voltage setting, record the vertical opening and bit error rate (BER) of the eye diagram after transmission.

[0069] 6) Select the DC bias point that maximizes the eye diagram opening and minimizes the bit error rate. This point represents the optimal chirp operating state for this specific transmission link. This method enables continuous and precise control of the output optical chirp (from negative chirp, zero chirp to positive chirp) to compensate for fiber dispersion and optimize long-distance transmission performance.

[0070] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the present invention.

Claims

1. A modulator based on thin-film lithium niobate-silicon heterogeneous integration, comprising an input waveguide, a beam splitter, a phase modulation region, a beam combiner, and an output waveguide, characterized in that: The phase modulation region includes a first interferometer arm, a second interferometer arm, and at least two phase modulation segments made of different electro-optic materials, wherein at least one phase modulation segment is a silicon-based modulation segment based on carrier dispersion effect, and at least another phase modulation segment is a lithium niobate-based modulation segment based on linear electro-optic effect; and the silicon-based modulation segment and the lithium niobate-based modulation segment are configured with independent electrodes for applying independent electrical drive signals.

2. The modulator based on thin-film lithium niobate-silicon heterogeneous integration according to claim 1, characterized in that: The first and second interferometer arms each contain three phase modulation segments, which are, in sequence, a first silicon-based modulation segment, a lithium niobate modulation segment, and a second silicon-based modulation segment along the light propagation direction.

3. A modulator based on thin-film lithium niobate-silicon heterostructure integration according to claim 2, characterized in that, The electrode is configured to: apply radio frequency signals with opposite phases to the silicon-based modulation segments on the first and second interferometer arms; apply radio frequency signals with the same phases to the lithium niobate modulation segments on the first and second interferometer arms; and the electrode is further configured to apply independent DC bias voltages to the lithium niobate modulation segments on the first and second interferometer arms to control the output optical chirp characteristics.

4. A modulator based on thin-film lithium niobate-silicon heterostructure integration according to claim 1, 2, or 3, characterized in that: It also includes an adjustable electrical network connected between the radio frequency signal source and the electrode. The adjustable electrical network is used to dynamically adjust the amplitude ratio and relative phase of the radio frequency signals applied to the silicon-based modulation segment and the lithium niobate modulation segment. The adjustable electrical network includes at least one adjustable attenuator and at least one adjustable phase shifter.

5. A modulator based on thin-film lithium niobate-silicon heterogeneous integration according to claim 1, characterized in that: The lithium niobate modulation section is a heterogeneous integrated waveguide structure of thin-film lithium niobate and silicon; the silicon-based modulation section is a silicon waveguide containing a PN junction, PIN junction or MOS capacitor structure.

6. A modulator based on thin-film lithium niobate-silicon heterogeneous integration according to any one of claims 4, characterized in that: Also includes: An auxiliary Mach-Zehnder modulator is arranged in parallel with the segmented Mach-Zehnder modulator; an optical beam combiner combines the output optical paths of the main modulator and the auxiliary modulator. The monitoring feedback unit is used to monitor the nonlinear distortion of the output optical signal in real time. The adaptive control unit dynamically adjusts the operating parameters of the auxiliary modulator based on the monitoring results.

7. The thin film lithium niobate-silicon hybrid integrated modulator of claim 6, wherein: The adaptive control unit is configured to optimize the bias voltage, RF drive amplitude, and relative phase of the auxiliary modulator using a gradient descent algorithm or an iterative learning control algorithm to minimize third-order intermodulation distortion of the output signal.

8. A modulator based on thin-film lithium niobate-silicon heterogeneous integration according to claim 6, characterized in that: The auxiliary modulator is a phase modulator based on thin-film lithium niobate, and the half-wave voltage of the auxiliary modulator is lower than the equivalent half-wave voltage of the main modulator.

9. The thin film lithium niobate-silicon hybrid integrated modulator of claim 6, wherein: The monitoring feedback unit includes a photodetector, a spectrum analysis module, and a harmonic detection algorithm, which are used to quantify the level of nonlinear distortion in real time.

10. A modulator control method based on thin-film lithium niobate-silicon heterostructure integration as described in any one of claims 1 to 9, characterized in that, Includes the following steps: Electrical control system: The main radio frequency signal is divided into two paths by a power divider. The main path signal drives the segmented electrodes of the main MZM through an adjustable attenuator and a phase shifter. The auxiliary path signal drives the auxiliary MZM through an adjustable attenuator and a nonlinear conversion circuit. The adaptive control processor receives monitoring feedback signals in real time and outputs control signals to each adjustable component. Adaptive optimization process: Initialization phase: Set the auxiliary MZM bias to the minimum transmission point and the main MZM bias to the orthogonal operating point; Distortion monitoring: Inject a two-tone test signal into the system and monitor the IMD3 power of the output signal; Parameter scan: fine-tune the bias, amplitude, and phase parameters of the auxiliary MZM sequentially; Gradient calculation: Calculate the gradient direction of each parameter based on the distortion change; Iterative update: Update parameters along the negative gradient direction until IMD3 is below the target threshold; Locked state: Enters tracking mode, periodically fine-tuning to compensate for environmental drift; Chirp regulation: 1) After completing the linearity optimization, i.e., determining the scaling factor Then, the drive signal is switched to a 40 GBaud PAM4 pseudo-random sequence; 2) Test the modulator output optical signal in the following two scenarios: a) Back-to-back connection; b) Through a 1-kilometer-long standard single-mode optical fiber; 3) Use a high-speed oscilloscope to observe and record the eye diagram after transmission through the optical fiber; 4) Fine-tune the DC bias voltage of the lithium niobate modulation section on the first and second interferometer arms to introduce a controllable phase imbalance; 5) For each bias voltage setting, record the vertical opening and bit error rate of the eye diagram after transmission; 6) Select the DC bias point that maximizes the eye diagram opening and minimizes the bit error rate. This point represents the optimal chirp operating state for this specific transmission link. This method enables continuous and precise control of the output optical chirp to compensate for fiber dispersion and optimize long-distance transmission performance.

Images