这是用户在 2025-3-7 9:59 为 https://arxiv.org/html/2502.03001v1 保存的双语快照页面,由 沉浸式翻译 提供双语支持。了解如何保存?
thanks: These authors contributed equally to this work.
感谢:这些作者对这项工作做出了同等贡献。
thanks: These authors contributed equally to this work.
感谢:这些作者对这项工作做出了同等贡献。
thanks: These authors contributed equally to this work.
感谢:这些作者对这项工作做出了同等贡献。

Universal Kerr-thermal dynamics of self-injection-locked microresonator dark pulses
自注入锁定微谐振器暗脉冲的普适克尔热动力学

Shichang Li International Quantum Academy, Shenzhen 518048, China Shenzhen Institute for Quantum Science and Engineering, Southern University of Science and Technology, Shenzhen 518055, China    Kunpeng Yu International Quantum Academy, Shenzhen 518048, China Hefei National Laboratory, University of Science and Technology of China, Hefei 230088, China    Dmitry A. Chermoshentsev Russian Quantum Center, Moscow 143026, Russia    Wei Sun sunwei@iqasz.cn International Quantum Academy, Shenzhen 518048, China    Jinbao Long International Quantum Academy, Shenzhen 518048, China    Xiaoying Yan International Quantum Academy, Shenzhen 518048, China    Chen Shen International Quantum Academy, Shenzhen 518048, China Qaleido Photonics, Shenzhen 518048, China    Artem E. Shitikov Russian Quantum Center, Moscow 143026, Russia    Nikita Yu. Dmitriev Russian Quantum Center, Moscow 143026, Russia    Igor A. Bilenko Russian Quantum Center, Moscow 143026, Russia Faculty of Physics, Lomonosov Moscow State University, Moscow 119991, Russia    Junqiu Liu liujq@iqasz.cn International Quantum Academy, Shenzhen 518048, China Hefei National Laboratory, University of Science and Technology of China, Hefei 230088, China

Microcombs, formed in optical microresonators driven by continuous-wave lasers, are miniaturized optical frequency combs with small size, weight and power consumption. Leveraging integrated photonics and laser self-injection locking (SIL), compact and robust microcombs can be constructed via hybrid integration of a semiconductor laser with a chip-based microresonator. While the current linear SIL theory has successfully addressed the linear coupling between the laser cavity and the external microresonator, it fails to manage the complicated nonlinear processes, especially regarding to dark-pulse microcomb formation. Here, we investigate theoretically, numerically and experimentally the Kerr-thermal dynamics of a semiconductor laser self-injection-locked to an integrated silicon nitride microresonator. We unveil intriguing yet universal dark-pulse formation and switching behaviour with discrete steps, and establish a theoretical model scrutinizing the synergy of laser-microresonator mutual coupling, Kerr nonlinearity, photo-thermal effect. Numerical simulation confirms the experimental result and identifies the origins. Exploiting this unique phenomenon, we showcase an application on low-noise photonic microwave generation with phase noise purified by 23.5 dB. Our study not only add critical insight of pulse formation in laser-microresonator hybrid systems, but also enables all-passive, photonic-chip-based microwave oscillators with high spectral purity.
微梳是在连续波激光器驱动的光学微谐振器中形成的,是一种小型化的光学频率梳,体积小、重量轻、功耗低。利用集成光子学和激光自注入锁定 (SIL) 技术,可以通过半导体激光器与芯片式微谐振器的混合集成来构建紧凑而坚固的微梳。虽然当前的线性 SIL 理论已经成功解决了激光腔和外部微谐振器之间的线性耦合问题,但它无法管理复杂的非线性过程,尤其是暗脉冲微梳的形成。在这里,我们从理论、数值和实验的角度研究了半导体激光器自注入锁定到集成氮化硅微谐振器的克尔热动力学。我们揭示了有趣但普遍的暗脉冲形成和离散步骤切换行为,并建立了一个理论模型来研究激光微谐振器互耦、克尔非线性和光热效应的协同作用。数值模拟证实了实验结果并确定了起源。利用这一独特现象,我们展示了一种低噪声光子微波生成应用,相位噪声净化了 23.5 dB。我们的研究不仅增加了激光微谐振器混合系统中脉冲形成的关键见解,而且还实现了具有高光谱纯度的全无源、基于光子芯片的微波振荡器。

Optical frequency combs (OFCs) Udem et al. (2002); Cundiff and Ye (2003); Fortier and Baumann (2019); Diddams et al. (2020) constitute broadband frequency rulers with equidistant lines of continuous-wave (CW) components, and have revolutionized timing, spectroscopy, precision measurement, and testing fundamental physics. While conventional OFCs are constructed with solid-state or fiber mode-locked lasers, microcombs harnessing enhanced Kerr nonlinearity in ultrahigh-QQitalic_Q optical microresonators driven by CW pumps have allowed miniaturized OFCs with small size, weight and power consumption Del’Haye et al. (2007); Kippenberg et al. (2018). With the emergence and quick maturing of ultralow-loss photonic integrated circuits Moss et al. (2013); Gaeta et al. (2019); Zhang et al. (2017); Liu et al. (2021), as well as heterogeneous and hybrid integration with semiconductor lasers Stern et al. (2018); Xiang et al. (2021), today microcombs can be built entirely on-chip and manufactured in large volume with low cost, catalyzing wide deployment outside laboratories and in space.
光频率梳 (OFC) Udem 等人 ( 2002);Cundiff 和 Ye ( 2003);Fortier 和 Baumann ( 2019);Diddams 等人 ( 2020) 构成具有等距连续波 (CW) 分量线的宽带频率标尺,并彻底改变了定时光谱学、精密测量和测试基础物理学传统 OFC 是用固态或光纤锁模激光器构造的,而微梳 - 利用由 CW 泵驱动的超高 QQitalic_Q 光学微谐振器中增强的克尔非线性 - 允许小型化 OFC,具有小尺寸、重量和功耗 Del'Haye 等人 ( 2007);Kippenberg 等人 ( 2018)随着超低损耗光子集成电路的出现和快速成熟 Moss 等人 ( 2013);Gaeta 等人 ( 2019);Zhang 等人(2017 年);刘等人(2021 年)以及与半导体激光器的异质和混合集成 Stern 等人(2018 年);Xiang 等人(2021 年)如今,微梳可以完全在芯片上构建,并以低成本进行大批量生产,从而在实验室外和太空中广泛部署

Depending on the microresonator’s group velocity dispersion (GVD), there are two types of coherent microcombs. The bright dissipative soliton microcombs Herr et al. (2013); Yi et al. (2015); Brasch et al. (2016a); Liang et al. (2015); Joshi et al. (2016); He et al. (2019) require anomalous GVD, while the dark-pulse (also termed “platicon”) microcombs require normal GVD Xue et al. (2015); Lobanov et al. (2015); Huang et al. (2015); Parra-Rivas et al. (2016); Nazemosadat et al. (2021); Wang et al. (2022). Compared to solitons, platicons exhibit remarkably higher CW-to-pulse power conversion efficiency Xue et al. (2017); Jang et al. (2021), thus are advantageous for coherent optical communication Fülöp et al. (2018) and photonic microwave generation Sun et al. (2024a). To initiate platicons, laser self-injection locking (SIL) Kondratiev et al. (2017, 2023) offers the most robust and effective form, and permits seamless integration of the pump laser and the microresonator Voloshin et al. (2021); Jin et al. (2021); Lihachev et al. (2022). However, while the current linear SIL theory and model Kondratiev et al. (2017) have addressed the linear coupling between the laser cavity and the external microresonator, it fails to manage the complicated nonlinear processes in the laser gain media and in platicon formation.
根据微谐振器的群速度色散 (GVD),相干微梳有两种类型。亮耗散孤子微梳Herr等人( 2013 ); Yi等人( 2015 ); Brasch等人( 2016a ); Liang等人( 2015 ); Joshi等人( 2016 ); He等人( 2019 )需要异常 GVD,而暗脉冲 (也称为“platicon”) 微梳需要正常 GVD Xue等人( 2015 ); Lobanov等人( 2015 ); Huang等人( 2015 ); Parra-Rivas等人( 2016 ); Nazemosadat等人( 2021 ); Wang等人( 2022 ) 。与孤子相比,扁子表现出明显更高的连续波到脉冲功率转换效率Xue2017 );Jang2021 ,因此有利于相干光通信Fülöp2018和光子微波产生Sun2024a 。 为了启动扁晶格,激光自注入锁定 (SIL) Kondratiev等人2017 年2023 年提供了最稳健、最有效的形式,并允许泵浦激光器和微谐振器Voloshin等人2021 年);Jin等人2021 年);Lihachev等人2022 年的无缝集成。然而,虽然当前的线性SIL 理论和模型Kondratiev等人2017 年已经解决了激光腔和外部微谐振器之间的线性耦合,但它无法管理激光增益介质和扁晶格形成中的复杂非线性过程。

Refer to caption
Figure 1: Principle and schematic of self-injection-locking, Kerr nonlinearity and photo-thermal effect in the laser-microresonator coupled system. a. Frequency-domain picture illustrating the SIL process with the Kerr-thermal effect. b. Frequency- and time-domain pictures describing the synergy of SIL, photo-thermal effect and Kerr nonlinearity, which yields narrowing of the pump laser’s linewidth, red-shift of the microresonator’s resonance grid, and formation of a platicon microcomb. c. Experimental setup showing the layout of a DFB laser chip edge-coupled to a Si3N4 microresonator chip. A home-made printed circuit board (PCB) is used to provide stable current to the laser and to stabilize the laser temperature. Output light from the Si3N4 microresonator is collected by a lensed fiber. d. Schematic of SIL. ELE_{\text{L}}italic_E start_POSTSUBSCRIPT L end_POSTSUBSCRIPT: light amplitude in the laser cavity. EE^{-}italic_E start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT/E+E^{+}italic_E start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT: light amplitude in the clockwise/counter-clockwise direction in the microresonator. EoutE_{\text{out}}italic_E start_POSTSUBSCRIPT out end_POSTSUBSCRIPT: light amplitude at the microresonator output. φ\varphiitalic_φ: feedback phase of EE^{-}italic_E start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT.
图1:激光微谐振器耦合系统中自注入锁定、克尔非线性和光热效应的原理及示意图。a . 频域图片展示了具有克尔热效应的 SIL 过程。 b . 频域和时域图片描述了 SIL、光热效应和克尔非线性的协同作用,这导致泵浦激光器的线宽变窄、微谐振器谐振网格的红移以及扁平微梳的形成。 c . 实验装置显示了 DFB 激光器芯片边缘耦合到 Si 3 N 4微谐振器芯片的布局。自制的印刷电路板 (PCB) 用于为激光器提供稳定的电流并稳定激光器温度。来自 Si 3 N 4微谐振器的输出光由透镜光纤收集。 d . SIL 示意图。 ELsubscriptE_{\text{L}}italic_E start_POSTSUBSCRIPT L end_POSTSUBSCRIPT :激光腔内的光振幅。 EsuperscriptE^{-}italic_E start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT / E+superscriptE^{+}italic_E start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT :微谐振器中顺时针/逆时针方向的光振幅。 EoutsubscriptE_{\text{out}}italic_E start_POSTSUBSCRIPT out end_POSTSUBSCRIPT :微谐振器输出处的光振幅。 φ\varphiitalic_φEsuperscriptE^{-}italic_E start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT 的反馈阶段。

Here, we investigate – theoretically, numerically and experimentally – the Kerr-thermal dynamics of platicon formation using a semiconductor laser self-injection-locked to an integrated silicon nitride (Si3N4) microresonator. We unveil an intriguing platicon switching behaviour with discrete steps, allowing operation of platicons in noise-quenched states and immunity to laser noise. Figure 1a illustrates the principle of our study. When a laser-cavity mode is tuned into a resonance mode of the Si3N4 microresonator and light is coupled into the microresonator, SIL occurs that locks the laser frequency to the resonance Kondratiev et al. (2017, 2023) and suppresses the laser linewidth. The linewidth suppression ratio is proportional to Qr/QdQ_{\text{r}}/Q_{\text{d}}italic_Q start_POSTSUBSCRIPT r end_POSTSUBSCRIPT / italic_Q start_POSTSUBSCRIPT d end_POSTSUBSCRIPT, where Qr/dQ_{\text{r/d}}italic_Q start_POSTSUBSCRIPT r/d end_POSTSUBSCRIPT is the quality factor of the microresonator/laser cavity. Meanwhile, photo-thermal effect Carmon et al. (2004); Brasch et al. (2016b); Yi et al. (2016); Gao et al. (2022) induces a global frequency shift of the microresonator’s resonance grid. With sufficient intracavity power, Kerr nonlinearity induces four-wave mixing (FWM) Kippenberg et al. (2004) that translates photons to other resonances. In microresonators of normal GVD, the synergy of SIL, photo-thermal and Kerr effects ultimately yield platicon formation, as shown in Fig. 1b.
在这里,我们从理论、数值和实验上研究了使用自注入锁定到集成氮化硅( Si3N4 ) 微谐振器的半导体激光器形成平晶格的克尔热动力学。我们揭示了一种有趣的平晶格开关行为,它具有离散步骤,允许平晶格在噪声猝灭状态下工作,并且不受激光噪声的影响。图1a说明了我们研究的原理。当激光腔模式调谐到Si3N4微谐振的谐振模式并将光耦合到微谐振器中时,会发生 SIL,将激光频率锁定到谐振Kondratiev等人2017 年2023 年并抑制激光线宽。线宽抑制比与 Qr/QdsubscriptsubscriptQ_{\text{r}}/Q_{\text{d}}italic_Q start_POSTSUBSCRIPT r end_POSTSUBSCRIPT / italic_Q start_POSTSUBSCRIPT d end_POSTSUBSCRIPT 成正比,其中 Qr/dsubscriptQ_{\text{r/d}}italic_Q start_POSTSUBSCRIPT r/d end_POSTSUBSCRIPT 是微谐振器/激光腔的品质因数。同时,光热效应Carmon等人2004 年);Brasch等人。2016b );Yi等人2016 );Gao等人2022在微谐振器的谐振网格中引起全局频率偏移。在腔内功率足够的情况下,克尔非线性会引起四波混频 (FWM) Kippenberg等人2004 ),从而将光子转换为其他谐振。在正常 GVD 的微谐振器中,SIL、光热和克尔效应的协同作用最终导致形成扁晶格,如图1b所示。

Experimental result. We use a commercial DFB laser chip edge-coupled to a Si3N4 microresonator chip Ye et al. (2023); Sun et al. (2024a), as shown in Fig. 1c. The DFB laser operates at 1549 nm in the telecommunication C band. It has 1 nm wavelength tuning range and 159 mW output CW power with driving current up to 500 mA. The Si3N4 microresonator has 10.7 GHz free spectral range (FSR) and normal GVD. (D2<0D_{2}<0italic_D start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT < 0). The microresonator is critically coupled Pfeiffer et al. (2017), with intrinsic quality factor Q0=23×106Q_{0}=23\times 10^{6}italic_Q start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT = 23 × 10 start_POSTSUPERSCRIPT 6 end_POSTSUPERSCRIPT. Actual photographs and characterization data of the DFB laser chip and the Si3N4 microresonator chip are found in Supplementary Information Note 1.
实验结果。我们使用边缘耦合到Si3N4谐振器芯片Ye2023 );Sun2024a )的商用 DFB 激光器芯片,如图1c所示。DFB 激光器在电信 C 波段的 1549 nm 下工作。它具有 1 nm 的波长调谐范围和 159 mW 输出 CW 功率,驱动电流高达 500 mA。Si3N4谐振器具有 10.7 GHz 的自由光谱范围 (FSR) 和正常 GVD。( D2<0subscript20D_{2}<0italic_D start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT < 0 )。微谐振器是临界耦合的Pfeiffer2017 ,具有固有品质因数 Q0=23×106subscript023superscript106Q_{0}=23\times 10^{6}italic_Q start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT = 23 × 10 start_POSTSUPERSCRIPT 6 end_POSTSUPERSCRIPT 。DFB 激光器芯片和Si3N4谐振器芯片的实际照片和特性数据可参见补充信息注释 1。

Figure 1d illustrates the schematic of SIL. Light ELE_{\text{L}}italic_E start_POSTSUBSCRIPT L end_POSTSUBSCRIPT in the laser cavity is emitted and coupled into the microresonator’s counter-clockwise direction E+E^{+}italic_E start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT. Rayleigh scattering in the microresonator reflects a portion of light to the clockwise direction EE^{-}italic_E start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT and to the laser cavity. To facilitate SIL, we optimize the edge coupling and the gap distance (thus the feedback phase ϕ\phiitalic_ϕ) between the two chips. In the SIL regime, we continuously tune the laser current, and monitor the output optical power and frequency from the Si3N4 chip. The output light is beaten against a reference laser, and the beat frequency is recorded. The forward (backward) tuning corresponds to increasing (decreasing) laser current, which decreases (increases) laser frequency.
1d显示了 SIL 的示意图。激光腔中的光 ELsubscriptE_{\text{L}}italic_E start_POSTSUBSCRIPT L end_POSTSUBSCRIPT 被发射并耦合到微谐振器的逆时针方向 E+superscriptE^{+}italic_E start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT 。微谐振器中的瑞利散射将一部分光反射到顺时针方向 EsuperscriptE^{-}italic_E start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT 并反射到激光腔。为了促进 SIL,我们优化了两个芯片之间的边缘耦合和间隙距离(因此优化了反馈相位 ϕ\phiitalic_ϕ )。在 SIL 状态下,我们不断调整激光电流,并监测 Si 3 N 4芯片的输出光功率和频率。输出光与参考激光器进行拍频,并记录拍频。前向(后向)调整对应于增加(减少)激光电流,从而降低(增加)激光频率。

Figure 2(a, b) shows the experimental data. When SIL occurs, the output power experiences a sudden drop, and the output frequency is shifted. In the SIL regime, we observe decreasing (increasing) output power and frequency with backward (forward) tuning, due to the photo-thermal effect Carmon et al. (2004); Brasch et al. (2016b); Yi et al. (2016); Gao et al. (2022). For example, the decreasing output power corresponds to decreasing transmitted power and increasing intra-cavity power. The latter enhances the photo-thermal effect and causes increasing resonance red-shift (towards longer wavelength) Carmon et al. (2004). As the laser frequency is locked to the resonance, the red-shifted resonance drags the laser frequency, leading to decreasing output optical frequency.
2 (a,b)显示了实验数据。当发生SIL时,输出功率会突然下降,并且输出频率会发生偏移。在SIL范围内,我们观察到由于光热效应导致输出功率和频率随着后向(前向)调谐而降低(增加) Carmon等人2004 );Brasch等人2016b );Yi等人2016 );Gao等人2022 。例如,输出功率的降低对应于透射功率的降低和腔内功率的增加。后者增强了光热效应并导致共振红移增加(朝向更长的波长) Carmon等人2004 。由于激光频率锁定在共振频率上,红移的共振会拖动激光频率,导致输出光频率降低。

Besides, it is apparent that the output optical power and frequency exhibit discrete step features. This is contrary to the conventional, linear SIL model Kondratiev et al. (2017) showing continuous, nearly anchored tuning curves of power and frequency. Here, the appearance of these steps are attributed to the Kerr nonlinearity in the microresonator with sufficient intra-cavity power. Note that similar discontinuous curves have been observed and characterized in self-injection-locked solitons Voloshin et al. (2021), while here we observe and characterize these curves for the first time for self-injection-locked platicons. Figure 2c shows the zoom-in profile of the gray-shaded zoom in Fig. 2b. Steps in backward tuning with 320.4 to 321.3 mA laser current are marked with 1 to 4. Typical optical spectra within each step are shown in Fig. 3a, evidencing formation and switching of different platicon states. Distinct fringes are marked with arrows on the spectral envelopes. Numerical simulation of time-domain pulse shapes in Fig. 3c (discussed later) confirms that all these platicon states are “single platicon” comprising only one dark pulse in the microresonator.
此外,很明显输出光功率和频率呈现离散阶跃特征。这与传统的线性 SIL 模型Kondratiev等人2017 年相反,该模型显示了连续的、几乎锚定的功率和频率调谐曲线。在这里,这些步骤的出现归因于具有足够腔内功率的微谐振器中的克尔非线性。请注意, Voloshin等人2021 年在自注入锁定孤子中已经观察并表征了类似的不连续曲线,而在这里我们第一次观察和表征了自注入锁定扁晶格的这些曲线。图2c显示了图2b中灰色阴影缩放的放大轮廓。使用 320.4 至 321.3 mA 激光电流进行后向调谐的步骤用 1 到 4 标记。图3a显示了每个步骤内的典型光谱,证明不同扁晶格状态的形成和切换。光谱包络上用箭头标记了不同的条纹。图3c中时域脉冲形状的数值模拟(稍后讨论)证实,所有这些平面状态都是“单个平面”,微谐振器中仅包含一个暗脉冲。

Moreover, identical Kerr-thermal platicon dynamics has also been observed in other two independent SIL setups, where different DFB lasers and 21.3-GHz-FSR Si3N4 microresonators are used. Details are found in Supplementary Information Note 1 and 2. These parallel experiments suggest that, our observation of the Kerr-thermal dynamics and discrete steps is universal and independent of the particular lasers or microresonators.
此外,在其他两个独立的 SIL 装置中也观察到了相同的克尔热平台动力学,其中使用了不同的 DFB 激光器和 21.3-GHz-FSR Si 3 N 4微谐振器。详细信息请参阅补充信息注释 1 和 2。这些平行实验表明,我们对克尔热动力学和离散步骤的观察是通用的,并且与特定的激光器或微谐振器无关。

Refer to caption
Figure 2: Experimental result in comparison with numerical simulation. a. Experimentally measured output laser power with forward (green) or backward (yellow) tuning. b. Experimentally measured beat frequency between the output laser and a frequency-fixed reference laser, with forward (blue) or backward (red) tuning. c. Zoom-in profiles of the gray-shaded region in b with forward (blue) or backward (red) tuning. Discrete steps with backward tuning are numbered with 1 to 4. d, e. Numerical simulation results corresponding to experimental data in a, b. Horizontal axes are the frequency detuning ξ\xiitalic_ξ (in the unit of κ/2\kappa/2italic_κ / 2) between the free-running laser frequency and the cold resonance frequency. Effective (Eff.) detuning is the frequency detuning (in the unit of κ/2\kappa/2italic_κ / 2) between the locked laser frequency and the cold resonance frequency. f. Zoom-in profile of the gray-shaded zoom in d, e. Gray curves outlines the full detuning range of platicon steps.
图2:实验结果与数值模拟的比较一个。 通过前向(绿色)或后向(黄色)调整实验测量的输出激光功率。 b . 通过实验测量输出​​激光器和频率固定参考激光器之间的拍频,进行前向(蓝色)或后向(红色)调整。 c . 放大b中灰色阴影区域的剖面图,并进行前向(蓝色)或后向(红色)调整。 具有向后调整的离散步骤以 1 至 4 进行编号。 d , e . 与a , b中实验数据对应的数值模拟结果。横轴为自由运行激光频率与冷谐振频率之间的频率失谐 ξ\xiitalic_ξ (单位为 κ/22\kappa/2italic_κ / 2 )。有效(Eff.)失谐为锁定激光频率与冷谐振频率之间的频率失谐(单位为 κ/22\kappa/2italic_κ / 2 )。 f . 灰色阴影放大的放大轮廓de 。 灰色曲线勾勒出了 platicon 台阶的全部失谐范围。

Theory and numerical simulation. To better understand our experimental result, we establish a comprehensive theoretical model harmonizing SIL, Kerr nonlinearity, and photo-thermal effect. The equations describing our laser-microresonator nonlinear system are written as
理论与数值模拟。为了更好地理解我们的实验结果,我们建立了一个综合的理论模型,协调了 SIL、克尔非线性和光热效应。描述我们的激光微谐振器非线性系统的方程式如下

τgN\displaystyle\partial_{\tau}g_{N}∂ start_POSTSUBSCRIPT italic_τ end_POSTSUBSCRIPT italic_g start_POSTSUBSCRIPT italic_N end_POSTSUBSCRIPT =JNκ~N(gN+g0)|aL|2\displaystyle=J_{N}-\tilde{\kappa}_{N}(g_{N}+g_{0})|a_{\text{L}}|^{2}= italic_J start_POSTSUBSCRIPT italic_N end_POSTSUBSCRIPT - over~ start_ARG italic_κ end_ARG start_POSTSUBSCRIPT italic_N end_POSTSUBSCRIPT ( italic_g start_POSTSUBSCRIPT italic_N end_POSTSUBSCRIPT + italic_g start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT ) | italic_a start_POSTSUBSCRIPT L end_POSTSUBSCRIPT | start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT (1)
τaL\displaystyle\partial_{\tau}a_{\text{L}}∂ start_POSTSUBSCRIPT italic_τ end_POSTSUBSCRIPT italic_a start_POSTSUBSCRIPT L end_POSTSUBSCRIPT =12(1+iαg)gNaL+iei(αLτ+φ)κ~Lb0\displaystyle=\frac{1}{2}\left(1+i\alpha_{g}\right)g_{N}a_{\text{L}}+i\mathrm{% e}^{i(\alpha_{\text{L}}\tau+\varphi)}\tilde{\kappa}_{\text{L}}b_{0}= divide start_ARG 1 end_ARG start_ARG 2 end_ARG ( 1 + italic_i italic_α start_POSTSUBSCRIPT italic_g end_POSTSUBSCRIPT ) italic_g start_POSTSUBSCRIPT italic_N end_POSTSUBSCRIPT italic_a start_POSTSUBSCRIPT L end_POSTSUBSCRIPT + italic_i roman_e start_POSTSUPERSCRIPT italic_i ( italic_α start_POSTSUBSCRIPT L end_POSTSUBSCRIPT italic_τ + italic_φ ) end_POSTSUPERSCRIPT over~ start_ARG italic_κ end_ARG start_POSTSUBSCRIPT L end_POSTSUBSCRIPT italic_b start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT (2)
τaμ\displaystyle\partial_{\tau}a_{\mu}∂ start_POSTSUBSCRIPT italic_τ end_POSTSUBSCRIPT italic_a start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT =(1+in=2dnn!μniT)aμ+i([|a|2a]μ+2aμμ|bμ|2)+iβ~bμ+iei(αLτ+φ)κ¯RaLδ0μ\displaystyle=-\left(1+i\sum_{n=2}^{\infty}\frac{d_{n}}{n!}\mu^{n}-iT\right)a_% {\mu}+i\left(\mathcal{F}[|a|^{2}a]_{\mu}+2a_{\mu}\sum_{\mu^{\prime}}\left|b_{% \mu^{\prime}}\right|^{2}\right)+i\tilde{\beta}b_{\mu}+i\mathrm{e}^{i(-\alpha_{% \text{L}}\tau+\varphi)}\bar{\kappa}_{\text{R}}a_{\text{L}}\delta_{0\mu}= - ( 1 + italic_i ∑ start_POSTSUBSCRIPT italic_n = 2 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT ∞ end_POSTSUPERSCRIPT divide start_ARG italic_d start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT end_ARG start_ARG italic_n ! end_ARG italic_μ start_POSTSUPERSCRIPT italic_n end_POSTSUPERSCRIPT - italic_i italic_T ) italic_a start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT + italic_i ( caligraphic_F [ | italic_a | start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT italic_a ] start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT + 2 italic_a start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT ∑ start_POSTSUBSCRIPT italic_μ start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT end_POSTSUBSCRIPT | italic_b start_POSTSUBSCRIPT italic_μ start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT end_POSTSUBSCRIPT | start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ) + italic_i over~ start_ARG italic_β end_ARG italic_b start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT + italic_i roman_e start_POSTSUPERSCRIPT italic_i ( - italic_α start_POSTSUBSCRIPT L end_POSTSUBSCRIPT italic_τ + italic_φ ) end_POSTSUPERSCRIPT over¯ start_ARG italic_κ end_ARG start_POSTSUBSCRIPT R end_POSTSUBSCRIPT italic_a start_POSTSUBSCRIPT L end_POSTSUBSCRIPT italic_δ start_POSTSUBSCRIPT 0 italic_μ end_POSTSUBSCRIPT (3)
τbμ\displaystyle\partial_{\tau}b_{\mu}∂ start_POSTSUBSCRIPT italic_τ end_POSTSUBSCRIPT italic_b start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT =(1+in=2dnn!μniT)bμ+i([|b|2b]μ+2bμμ|aμ|2)+iβ~aμ\displaystyle=-\left(1+i\sum_{n=2}^{\infty}\frac{d_{n}}{n!}\mu^{n}-iT\right)b_% {\mu}+i\left(\mathcal{F}[|b|^{2}b]_{\mu}+2b_{\mu}\sum_{\mu^{\prime}}\left|a_{% \mu^{\prime}}\right|^{2}\right)+i\tilde{\beta}a_{\mu}= - ( 1 + italic_i ∑ start_POSTSUBSCRIPT italic_n = 2 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT ∞ end_POSTSUPERSCRIPT divide start_ARG italic_d start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT end_ARG start_ARG italic_n ! end_ARG italic_μ start_POSTSUPERSCRIPT italic_n end_POSTSUPERSCRIPT - italic_i italic_T ) italic_b start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT + italic_i ( caligraphic_F [ | italic_b | start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT italic_b ] start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT + 2 italic_b start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT ∑ start_POSTSUBSCRIPT italic_μ start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT end_POSTSUBSCRIPT | italic_a start_POSTSUBSCRIPT italic_μ start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT end_POSTSUBSCRIPT | start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ) + italic_i over~ start_ARG italic_β end_ARG italic_a start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT (4)
τT\displaystyle\partial_{\tau}T∂ start_POSTSUBSCRIPT italic_τ end_POSTSUBSCRIPT italic_T =(kTμ(|aμ|2+|bμ|2)T)/th.\displaystyle=\left(k_{T}\sum_{\mu}\left(\left|a_{\mu}\right|^{2}+\left|b_{\mu% }\right|^{2}\right)-T\right)/t_{\text{h}}.= ( italic_k start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT ∑ start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT ( | italic_a start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT | start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT + | italic_b start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT | start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ) - italic_T ) / italic_t start_POSTSUBSCRIPT h end_POSTSUBSCRIPT . (5)

Derivation of Eq. (1–5) is found in Supplementary Information Note 3. In short, Eq. (1) describes the time evolution of the electron carrier inside the laser cavity, with τ=κt/2\tau=\kappa t/2italic_τ = italic_κ italic_t / 2, gN=2Gn(NN0)/κg_{N}=2G_{n}(N-N_{0})/\kappaitalic_g start_POSTSUBSCRIPT italic_N end_POSTSUBSCRIPT = 2 italic_G start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT ( italic_N - italic_N start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT ) / italic_κ, JN=2I/(κeVL)J_{N}=2I/(\kappa eV_{\text{L}})italic_J start_POSTSUBSCRIPT italic_N end_POSTSUBSCRIPT = 2 italic_I / ( italic_κ italic_e italic_V start_POSTSUBSCRIPT L end_POSTSUBSCRIPT ), κ~N=3Gn/(gVL)\tilde{\kappa}_{N}=3G_{n}/(gV_{\text{L}})over~ start_ARG italic_κ end_ARG start_POSTSUBSCRIPT italic_N end_POSTSUBSCRIPT = 3 italic_G start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT / ( italic_g italic_V start_POSTSUBSCRIPT L end_POSTSUBSCRIPT ), g0=2G~0/κg_{0}=2\tilde{G}_{0}/\kappaitalic_g start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT = 2 over~ start_ARG italic_G end_ARG start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT / italic_κ, and aL=gϵ0nL2VL/(ωLκ)ELa_{\text{L}}=\sqrt{g\epsilon_{0}n_{\text{L}}^{2}V_{\text{L}}/(\hbar\omega_{% \text{L}}\kappa)}E_{\text{L}}italic_a start_POSTSUBSCRIPT L end_POSTSUBSCRIPT = square-root start_ARG italic_g italic_ϵ start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT italic_n start_POSTSUBSCRIPT L end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT italic_V start_POSTSUBSCRIPT L end_POSTSUBSCRIPT / ( roman_ℏ italic_ω start_POSTSUBSCRIPT L end_POSTSUBSCRIPT italic_κ ) end_ARG italic_E start_POSTSUBSCRIPT L end_POSTSUBSCRIPT. Here, κ\kappaitalic_κ is the loaded linewidth of the Si3N4 microresonator, ttitalic_t is the real time, GnG_{n}italic_G start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT is the gain coefficient of the active area in the laser cavity, NNitalic_N is the electron carrier density, N0N_{0}italic_N start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT is the electron carrier density of the free-running laser with a fixed driving current, IIitalic_I is the driving current on the laser diode, eeitalic_e is the electron charge, VLV_{\text{L}}italic_V start_POSTSUBSCRIPT L end_POSTSUBSCRIPT is the effective volume of the laser diode’s active area, G~0\tilde{G}_{0}over~ start_ARG italic_G end_ARG start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT is the net gain at N0N_{0}italic_N start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT of the gain media in the laser, ϵ0\epsilon_{0}italic_ϵ start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT is the vacuum permittivity, nLn_{\text{L}}italic_n start_POSTSUBSCRIPT L end_POSTSUBSCRIPT is the refractive index of the active area, ωL\omega_{\text{L}}italic_ω start_POSTSUBSCRIPT L end_POSTSUBSCRIPT is the laser cavity resonance frequency, and ELE_{\text{L}}italic_E start_POSTSUBSCRIPT L end_POSTSUBSCRIPT is the laser amplitude in the laser cavity. The nonlinear coefficient of the Si3N4 microresonator is g=ω02cn2/(n02Veff)g=\hbar\omega_{0}^{2}cn_{2}/(n_{0}^{2}V_{\text{eff}})italic_g = roman_ℏ italic_ω start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT italic_c italic_n start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT / ( italic_n start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT italic_V start_POSTSUBSCRIPT eff end_POSTSUBSCRIPT ), where \hbarroman_ℏ is the reduced Plank constant, ω0\omega_{0}italic_ω start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT is the microresonator resonance frequency, ccitalic_c is the speed of light in vacuum, n2n_{2}italic_n start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT is the nonlinear refractive index of Si3N4, n0n_{0}italic_n start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT is the refractive index of Si3N4, and VeffV_{\text{eff}}italic_V start_POSTSUBSCRIPT eff end_POSTSUBSCRIPT is the effective volume of the Si3N4 microresonator.
方程(1–5)的推导见补充材料注释3简而言之,方程(1)描述了激光腔内电子载流子的时间演化,其中 τ=κt/22\tau=\kappa t/2italic_τ = italic_κ italic_t / 2gN=2Gn(NN0)/κsubscript2subscriptsubscript0g_{N}=2G_{n}(N-N_{0})/\kappaitalic_g start_POSTSUBSCRIPT italic_N end_POSTSUBSCRIPT = 2 italic_G start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT ( italic_N - italic_N start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT ) / italic_κJN=2I/(κeVL)subscript2subscriptJ_{N}=2I/(\kappa eV_{\text{L}})italic_J start_POSTSUBSCRIPT italic_N end_POSTSUBSCRIPT = 2 italic_I / ( italic_κ italic_e italic_V start_POSTSUBSCRIPT L end_POSTSUBSCRIPT )κ~N=3Gn/(gVL)subscript3subscriptsubscript\tilde{\kappa}_{N}=3G_{n}/(gV_{\text{L}})over~ start_ARG italic_κ end_ARG start_POSTSUBSCRIPT italic_N end_POSTSUBSCRIPT = 3 italic_G start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT / ( italic_g italic_V start_POSTSUBSCRIPT L end_POSTSUBSCRIPT )g0=2G~0/κsubscript02subscript0g_{0}=2\tilde{G}_{0}/\kappaitalic_g start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT = 2 over~ start_ARG italic_G end_ARG start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT / italic_κaL=gϵ0nL2VL/(ωLκ)ELsubscriptsubscript0superscriptsubscript2subscriptPlanck-constant-over-2-pisubscriptsubscripta_{\text{L}}=\sqrt{g\epsilon_{0}n_{\text{L}}^{2}V_{\text{L}}/(\hbar\omega_{% \text{L}}\kappa)}E_{\text{L}}italic_a start_POSTSUBSCRIPT L end_POSTSUBSCRIPT = square-root start_ARG italic_g italic_ϵ start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT italic_n start_POSTSUBSCRIPT L end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT italic_V start_POSTSUBSCRIPT L end_POSTSUBSCRIPT / ( roman_ℏ italic_ω start_POSTSUBSCRIPT L end_POSTSUBSCRIPT italic_κ ) end_ARG italic_E start_POSTSUBSCRIPT L end_POSTSUBSCRIPT 其中 κ\kappaitalic_κ 是Si3N4微谐振器的加载线宽 ttitalic_t 是实时 GnsubscriptG_{n}italic_G start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT 是激光腔内有源区的增益系数 NNitalic_N 是电子载流子密度 N0subscript0N_{0}italic_N start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT 是固定驱动电流下自由运行激光器的电子载流子密度 IIitalic_I 是激光二极管上的驱动电流 eeitalic_e 是电子电荷 VLsubscriptV_{\text{L}}italic_V start_POSTSUBSCRIPT L end_POSTSUBSCRIPT 是激光二极管有源区的有效体积 G~0subscript0\tilde{G}_{0}over~ start_ARG italic_G end_ARG start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT 是激光器中增益介质在 N0subscript0N_{0}italic_N start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT 处的净增益 ϵ0subscript0\epsilon_{0}italic_ϵ start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT 是真空介电常数 nLsubscriptn_{\text{L}}italic_n start_POSTSUBSCRIPT L end_POSTSUBSCRIPT 是有源区的折射率 ωLsubscript\omega_{\text{L}}italic_ω start_POSTSUBSCRIPT L end_POSTSUBSCRIPT 是激光腔谐振频率, ELsubscriptE_{\text{L}}italic_E start_POSTSUBSCRIPT L end_POSTSUBSCRIPT 是激光腔内的激光振幅Si3N4 微谐振器为 g=ω02cn2/(n02Veff)Planck-constant-over-2-pisuperscriptsubscript02subscript2superscriptsubscript02subscriptg=\hbar\omega_{0}^{2}cn_{2}/(n_{0}^{2}V_{\text{eff}})italic_g = roman_ℏ italic_ω start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT italic_c italic_n start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT / ( italic_n start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT italic_V start_POSTSUBSCRIPT eff end_POSTSUBSCRIPT ) ,其中 Planck-constant-over-2-pi\hbarroman_ℏ 是减小的普朗克常数, ω0subscript0\omega_{0}italic_ω start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT 是微谐振器谐振频率, ccitalic_c 是真空中的光速, n2subscript2n_{2}italic_n start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT 是 Si3N4 的非线性折射率, n0subscript0n_{0}italic_n start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT 是 Si3N4 的折射率, VeffsubscriptV_{\text{eff}}italic_V start_POSTSUBSCRIPT eff end_POSTSUBSCRIPT 是 Si3N4 微谐振器的有效体积

Equation (2) describes the laser dynamics in the laser cavity, with αL=2δωL/κ\alpha_{\text{L}}=2\delta\omega_{\text{L}}/\kappaitalic_α start_POSTSUBSCRIPT L end_POSTSUBSCRIPT = 2 italic_δ italic_ω start_POSTSUBSCRIPT L end_POSTSUBSCRIPT / italic_κ, φ=ω0ts\varphi=\omega_{0}t_{\text{s}}italic_φ = italic_ω start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT italic_t start_POSTSUBSCRIPT s end_POSTSUBSCRIPT, κ~L=2κexτRTLTcnLVL/(κτLn0Veff)\tilde{\kappa}_{\text{L}}=2\sqrt{\kappa_{\text{ex}}\tau_{\text{R}}T_{\text{L}}% T_{\text{c}}n_{\text{L}}V_{\text{L}}}/(\kappa\tau_{\text{L}}\sqrt{n_{0}V_{% \text{eff}}})over~ start_ARG italic_κ end_ARG start_POSTSUBSCRIPT L end_POSTSUBSCRIPT = 2 square-root start_ARG italic_κ start_POSTSUBSCRIPT ex end_POSTSUBSCRIPT italic_τ start_POSTSUBSCRIPT R end_POSTSUBSCRIPT italic_T start_POSTSUBSCRIPT L end_POSTSUBSCRIPT italic_T start_POSTSUBSCRIPT c end_POSTSUBSCRIPT italic_n start_POSTSUBSCRIPT L end_POSTSUBSCRIPT italic_V start_POSTSUBSCRIPT L end_POSTSUBSCRIPT end_ARG / ( italic_κ italic_τ start_POSTSUBSCRIPT L end_POSTSUBSCRIPT square-root start_ARG italic_n start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT italic_V start_POSTSUBSCRIPT eff end_POSTSUBSCRIPT end_ARG ). Here, αg\alpha_{g}italic_α start_POSTSUBSCRIPT italic_g end_POSTSUBSCRIPT is the linewidth enhancement factor, δωL\delta\omega_{\text{L}}italic_δ italic_ω start_POSTSUBSCRIPT L end_POSTSUBSCRIPT is the frequency difference between the cold-cavity resonance mode and the microresonator resonance mode, tst_{\text{s}}italic_t start_POSTSUBSCRIPT s end_POSTSUBSCRIPT is the time delay between the laser cavity and the microresonator, κex\kappa_{\text{ex}}italic_κ start_POSTSUBSCRIPT ex end_POSTSUBSCRIPT is the external coupling rate of the microresonator, τR\tau_{\text{R}}italic_τ start_POSTSUBSCRIPT R end_POSTSUBSCRIPT is the light round-trip time in the microresonator, TLT_{\text{L}}italic_T start_POSTSUBSCRIPT L end_POSTSUBSCRIPT is the light transmission rate at the laser cavity’s output surface, TcT_{\text{c}}italic_T start_POSTSUBSCRIPT c end_POSTSUBSCRIPT is the light coupling rate into the laser cavity from outside, τL\tau_{\text{L}}italic_τ start_POSTSUBSCRIPT L end_POSTSUBSCRIPT is the light round-trip time in the laser cavity, and b0b_{0}italic_b start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT is the back-scattered light in the microresonator at the 0th0^{\text{th}}0 start_POSTSUPERSCRIPT th end_POSTSUPERSCRIPT resonance mode.
方程(2)用 αL=2δωL/κsubscript2subscript\alpha_{\text{L}}=2\delta\omega_{\text{L}}/\kappaitalic_α start_POSTSUBSCRIPT L end_POSTSUBSCRIPT = 2 italic_δ italic_ω start_POSTSUBSCRIPT L end_POSTSUBSCRIPT / italic_κφ=ω0tssubscript0subscript\varphi=\omega_{0}t_{\text{s}}italic_φ = italic_ω start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT italic_t start_POSTSUBSCRIPT s end_POSTSUBSCRIPTκ~L=2κexτRTLTcnLVL/(κτLn0Veff)subscript2subscriptsubscriptsubscriptsubscriptsubscriptsubscriptsubscriptsubscript0subscript\tilde{\kappa}_{\text{L}}=2\sqrt{\kappa_{\text{ex}}\tau_{\text{R}}T_{\text{L}}% T_{\text{c}}n_{\text{L}}V_{\text{L}}}/(\kappa\tau_{\text{L}}\sqrt{n_{0}V_{% \text{eff}}})over~ start_ARG italic_κ end_ARG start_POSTSUBSCRIPT L end_POSTSUBSCRIPT = 2 square-root start_ARG italic_κ start_POSTSUBSCRIPT ex end_POSTSUBSCRIPT italic_τ start_POSTSUBSCRIPT R end_POSTSUBSCRIPT italic_T start_POSTSUBSCRIPT L end_POSTSUBSCRIPT italic_T start_POSTSUBSCRIPT c end_POSTSUBSCRIPT italic_n start_POSTSUBSCRIPT L end_POSTSUBSCRIPT italic_V start_POSTSUBSCRIPT L end_POSTSUBSCRIPT end_ARG / ( italic_κ italic_τ start_POSTSUBSCRIPT L end_POSTSUBSCRIPT square-root start_ARG italic_n start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT italic_V start_POSTSUBSCRIPT eff end_POSTSUBSCRIPT end_ARG ) 描述激光腔内的激光动力学,其中 αgsubscript\alpha_{g}italic_α start_POSTSUBSCRIPT italic_g end_POSTSUBSCRIPT 为线宽增强因子, δωLsubscript\delta\omega_{\text{L}}italic_δ italic_ω start_POSTSUBSCRIPT L end_POSTSUBSCRIPT 为冷腔谐振模式与微谐振器谐振模式之间的频率差, tssubscriptt_{\text{s}}italic_t start_POSTSUBSCRIPT s end_POSTSUBSCRIPT 为激光腔与微谐振器之间的时间延迟, κexsubscript\kappa_{\text{ex}}italic_κ start_POSTSUBSCRIPT ex end_POSTSUBSCRIPT 为微谐振器的外部耦合率, τRsubscript\tau_{\text{R}}italic_τ start_POSTSUBSCRIPT R end_POSTSUBSCRIPT 为光在微谐振器中的往返时间, TLsubscriptT_{\text{L}}italic_T start_POSTSUBSCRIPT L end_POSTSUBSCRIPT 为光在激光腔输出表面的传输速率, TcsubscriptT_{\text{c}}italic_T start_POSTSUBSCRIPT c end_POSTSUBSCRIPT 为从外部进入激光腔的光耦合率, τLsubscript\tau_{\text{L}}italic_τ start_POSTSUBSCRIPT L end_POSTSUBSCRIPT 为光在激光腔内的往返时间, b0subscript0b_{0}italic_b start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT0thsuperscript00^{\text{th}}0 start_POSTSUPERSCRIPT th end_POSTSUPERSCRIPT 谐振模式下微谐振器中的背向散射光

Equations (3, 4) describe the dynamics of the forward-propagating light aμa_{\mu}italic_a start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT and the back-scattered light bμb_{\mu}italic_b start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT at the μth\mu^{\text{th}}italic_μ start_POSTSUPERSCRIPT th end_POSTSUPERSCRIPT resonance mode, with aμ(bμ)=gϵ0n02Veff/(ω0κ)Eμ+(Eμ)a_{\mu}(b_{\mu})=\sqrt{g\epsilon_{0}n_{0}^{2}V_{\text{eff}}/(\hbar\omega_{0}% \kappa)}E_{\mu}^{+}(E_{\mu}^{-})italic_a start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT ( italic_b start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT ) = square-root start_ARG italic_g italic_ϵ start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT italic_n start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT italic_V start_POSTSUBSCRIPT eff end_POSTSUBSCRIPT / ( roman_ℏ italic_ω start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT italic_κ ) end_ARG italic_E start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT ( italic_E start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT ), dn=2Dn/κd_{n}=2D_{n}/\kappaitalic_d start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT = 2 italic_D start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT / italic_κ, β~=2β/κ\tilde{\beta}=2\beta/\kappaover~ start_ARG italic_β end_ARG = 2 italic_β / italic_κ, and κ~R=2κexTLTcn0Veff/(κτRnLVL)\tilde{\kappa}_{\text{R}}=2\sqrt{\kappa_{\text{ex}}T_{\text{L}}T_{\text{c}}n_{% 0}V_{\text{eff}}}/(\kappa\sqrt{\tau_{\text{R}}n_{\text{L}}V_{\text{L}}})over~ start_ARG italic_κ end_ARG start_POSTSUBSCRIPT R end_POSTSUBSCRIPT = 2 square-root start_ARG italic_κ start_POSTSUBSCRIPT ex end_POSTSUBSCRIPT italic_T start_POSTSUBSCRIPT L end_POSTSUBSCRIPT italic_T start_POSTSUBSCRIPT c end_POSTSUBSCRIPT italic_n start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT italic_V start_POSTSUBSCRIPT eff end_POSTSUBSCRIPT end_ARG / ( italic_κ square-root start_ARG italic_τ start_POSTSUBSCRIPT R end_POSTSUBSCRIPT italic_n start_POSTSUBSCRIPT L end_POSTSUBSCRIPT italic_V start_POSTSUBSCRIPT L end_POSTSUBSCRIPT end_ARG ). Here, Eμ+(Eμ)E_{\mu}^{+}(E_{\mu}^{-})italic_E start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT ( italic_E start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT ) is the laser amplitude of the forward-propagating (back-scattered) light at the μth\mu^{\text{th}}italic_μ start_POSTSUPERSCRIPT th end_POSTSUPERSCRIPT resonance mode, DnD_{n}italic_D start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT is the nthn^{\text{th}}italic_n start_POSTSUPERSCRIPT th end_POSTSUPERSCRIPT-order dispersion of the microresonator, β\betaitalic_β is the back-scattering ratio between the back-scattered light and the forward-propagating light, and [|a|2a]\mathcal{F}[|a|^{2}a]caligraphic_F [ | italic_a | start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT italic_a ] ([|b|2b])\mathcal{F}[|b|^{2}b])caligraphic_F [ | italic_b | start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT italic_b ] ) represents the Kerr interaction.
方程 (34) 描述了 μthsuperscript\mu^{\text{th}}italic_μ start_POSTSUPERSCRIPT th end_POSTSUPERSCRIPT 谐振模式下前向传播光 aμsubscripta_{\mu}italic_a start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT 和反向散射光 bμsubscriptb_{\mu}italic_b start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT 的动力学,其中 aμ(bμ)=gϵ0n02Veff/(ω0κ)Eμ+(Eμ)subscriptsubscriptsubscript0superscriptsubscript02subscriptPlanck-constant-over-2-pisubscript0superscriptsubscriptsuperscriptsubscripta_{\mu}(b_{\mu})=\sqrt{g\epsilon_{0}n_{0}^{2}V_{\text{eff}}/(\hbar\omega_{0}% \kappa)}E_{\mu}^{+}(E_{\mu}^{-})italic_a start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT ( italic_b start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT ) = square-root start_ARG italic_g italic_ϵ start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT italic_n start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT italic_V start_POSTSUBSCRIPT eff end_POSTSUBSCRIPT / ( roman_ℏ italic_ω start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT italic_κ ) end_ARG italic_E start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT ( italic_E start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT )dn=2Dn/κsubscript2subscriptd_{n}=2D_{n}/\kappaitalic_d start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT = 2 italic_D start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT / italic_κβ~=2β/κ2\tilde{\beta}=2\beta/\kappaover~ start_ARG italic_β end_ARG = 2 italic_β / italic_κκ~R=2κexTLTcn0Veff/(κτRnLVL)subscript2subscriptsubscriptsubscriptsubscript0subscriptsubscriptsubscriptsubscript\tilde{\kappa}_{\text{R}}=2\sqrt{\kappa_{\text{ex}}T_{\text{L}}T_{\text{c}}n_{% 0}V_{\text{eff}}}/(\kappa\sqrt{\tau_{\text{R}}n_{\text{L}}V_{\text{L}}})over~ start_ARG italic_κ end_ARG start_POSTSUBSCRIPT R end_POSTSUBSCRIPT = 2 square-root start_ARG italic_κ start_POSTSUBSCRIPT ex end_POSTSUBSCRIPT italic_T start_POSTSUBSCRIPT L end_POSTSUBSCRIPT italic_T start_POSTSUBSCRIPT c end_POSTSUBSCRIPT italic_n start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT italic_V start_POSTSUBSCRIPT eff end_POSTSUBSCRIPT end_ARG / ( italic_κ square-root start_ARG italic_τ start_POSTSUBSCRIPT R end_POSTSUBSCRIPT italic_n start_POSTSUBSCRIPT L end_POSTSUBSCRIPT italic_V start_POSTSUBSCRIPT L end_POSTSUBSCRIPT end_ARG ) 其中 Eμ+(Eμ)superscriptsubscriptsuperscriptsubscriptE_{\mu}^{+}(E_{\mu}^{-})italic_E start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT ( italic_E start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT )μthsuperscript\mu^{\text{th}}italic_μ start_POSTSUPERSCRIPT th end_POSTSUPERSCRIPT 谐振模式下前向传播(反向散射)光的激光振幅, DnsubscriptD_{n}italic_D start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT 是微谐振器的 nthsuperscriptn^{\text{th}}italic_n start_POSTSUPERSCRIPT th end_POSTSUPERSCRIPT 阶色散, β\betaitalic_β 是反向散射光与前向传播光之间的反向散射比,并且 [|a|2a]delimited-[]superscript2\mathcal{F}[|a|^{2}a]caligraphic_F [ | italic_a | start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT italic_a ][|b|2b])\mathcal{F}[|b|^{2}b])caligraphic_F [ | italic_b | start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT italic_b ] ) 表示克尔相互作用

Equation (5) describes the evolution of the normalized frequency shift TTitalic_T induced by the temperature in the microresonator, with kT=KT/gk_{T}=K_{T}/gitalic_k start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT = italic_K start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT / italic_g, and th=κτr/2t_{\text{h}}=\kappa\tau_{\text{r}}/2italic_t start_POSTSUBSCRIPT h end_POSTSUBSCRIPT = italic_κ italic_τ start_POSTSUBSCRIPT r end_POSTSUBSCRIPT / 2. Here, KTK_{T}italic_K start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT is the thermal-induced resonance shift coefficient, and τr\tau_{\text{r}}italic_τ start_POSTSUBSCRIPT r end_POSTSUBSCRIPT is the thermal relaxation time.
方程 (5) 描述了微谐振器中温度引起的归一化频率偏移 TTitalic_T 的演变,其中 kT=KT/gsubscriptsubscriptk_{T}=K_{T}/gitalic_k start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT = italic_K start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT / italic_gth=κτr/2subscriptsubscript2t_{\text{h}}=\kappa\tau_{\text{r}}/2italic_t start_POSTSUBSCRIPT h end_POSTSUBSCRIPT = italic_κ italic_τ start_POSTSUBSCRIPT r end_POSTSUBSCRIPT / 2 。其中, KTsubscriptK_{T}italic_K start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT 是热致谐振偏移系数, τrsubscript\tau_{\text{r}}italic_τ start_POSTSUBSCRIPT r end_POSTSUBSCRIPT 是热弛豫时间。

Refer to caption
Figure 3: Optical spectra of different platicon states and the resulted noise-quenching effect. a, b. Experimentally measured a and simulated b optical spectra of the platicon states labelled in Fig. 2c, f. They agree not only on the spectral envelopes but also on the number of fringes (marked with arrows). c. Simulated time-domain pulse shapes of the corresponding platicon states. The fringes, marked with arrows, are related to the oscillating tails at the bottom of the dark pulse. d. Measured platicon’s repetition rate frepf_{\text{rep}}italic_f start_POSTSUBSCRIPT rep end_POSTSUBSCRIPT (blue) and phase noise SϕS_{\phi}italic_S start_POSTSUBSCRIPT italic_ϕ end_POSTSUBSCRIPT at 10 kHz Fourier offset frequency (red) of the 10.7-GHz microwave, with backward tuning. f0=10.68545f_{0}=10.68545italic_f start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT = 10.68545 GHz. Vertical dashed lines highlights that the local minimum of SϕS_{\phi}italic_S start_POSTSUBSCRIPT italic_ϕ end_POSTSUBSCRIPT coincides with the local maximum of frepf_{\text{rep}}italic_f start_POSTSUBSCRIPT rep end_POSTSUBSCRIPT. Horizontal dashed lines highlights that this coincidence is associated with dfrep/dI=0\text{d}f_{\text{rep}}/\text{d}I=0d italic_f start_POSTSUBSCRIPT rep end_POSTSUBSCRIPT / d italic_I = 0. e. Phase noise SϕS_{\phi}italic_S start_POSTSUBSCRIPT italic_ϕ end_POSTSUBSCRIPT spectra of the local maximum and minimum points marked with green and black dots in d. Phase noise quenching up to 23.5 dB is observed. Inset shows the beatnote of the lowest SϕS_{\phi}italic_S start_POSTSUBSCRIPT italic_ϕ end_POSTSUBSCRIPT (black data) with 10 Hz RBW.
图3:不同铂离子状态的光谱和产生的噪声抑制效应ab 。图2c 、f 中标记的铂态的实验测量a和模拟b光谱。 他们不仅对光谱包络达成一致,而且对条纹数量(用箭头标记)也达成一致。 c . 模拟相应平面状态的时域脉冲形状。 用箭头标记的条纹与暗脉冲底部的振荡尾部有关。 d . 在 10.7 GHz 微波的 10 kHz 傅里叶偏移频率(红色)处测量平板天线的重复率 frepsubscriptf_{\text{rep}}italic_f start_POSTSUBSCRIPT rep end_POSTSUBSCRIPT (蓝色)和相位噪声 SϕsubscriptS_{\phi}italic_S start_POSTSUBSCRIPT italic_ϕ end_POSTSUBSCRIPT ,并进行后向调谐。 f0=10.68545subscript010.68545f_{0}=10.68545italic_f start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT = 10.68545 GHz。垂直虚线突出显示 SϕsubscriptS_{\phi}italic_S start_POSTSUBSCRIPT italic_ϕ end_POSTSUBSCRIPT 的局部最小值与 frepsubscriptf_{\text{rep}}italic_f start_POSTSUBSCRIPT rep end_POSTSUBSCRIPT 的局部最大值重合。水平虚线突出显示此重合与 dfrep/dI=0subscript0\text{d}f_{\text{rep}}/\text{d}I=0d italic_f start_POSTSUBSCRIPT rep end_POSTSUBSCRIPT / d italic_I = 0 相关。 e . d中用绿点和黑点标记的局部最大点和最小点的相位噪声 SϕsubscriptS_{\phi}italic_S start_POSTSUBSCRIPT italic_ϕ end_POSTSUBSCRIPT 频谱。观察到相位噪声抑制高达 23.5 dB。插图显示了最低 SϕsubscriptS_{\phi}italic_S start_POSTSUBSCRIPT italic_ϕ end_POSTSUBSCRIPT (黑色数据)的拍音,RBW 为 10 Hz。

We further perform numerical simulation of Eq. (1–5) using experimentally obtained or realistic values for each parameter. Details are found in Supplementary Information Note 3. Figure 2(d,e) presents the simulated output optical power and frequency, which agree with the experimental data on both the overall trend (due to the photo-thermal effect) and the platicon steps (due to the Kerr nonlinearity). Figure 2f shows the zoom-in profile of the gray-shaded zoom in Fig. 2(d,e), and highlights the full detuning range of platicon steps with gray curves. Steps corresponding to those in Fig. 2c are marked. We emphasize that, previous efforts Kondratiev et al. (2020); Lihachev et al. (2022); Wang et al. (2022) to model platicon formation in the SIL regime fail to investigate and analyze the transient behaviour, and thus fail to reveal the step features. Meanwhile, they have also omitted the photo-thermal effect.
我们进一步使用实验获得的每个参数的实际值对公式 (1-5) 进行数值模拟。详细信息见补充信息注释 3。图2 (d,e) 展示了模拟的输出光功率和频率,它们与实验数据在总体趋势(由于光热效应)和平板台阶(由于克尔非线性)上都一致。图2 f 显示了图2 (d,e) 中灰色阴影缩放的放大轮廓,并用灰色曲线突出显示了平板台阶的完整失谐范围。标记了与图2 c 中的台阶相对应的台阶。我们强调, Kondratiev等人2020 年);Lihachev等人2022 年);Wang等人2022 年在 SIL 范围内模拟平板形成的努力未能研究和分析瞬态行为,因此未能揭示台阶特征。同时,他们也忽略了光热效应。

Figure 3b presents the simulated platicon spectra for each labelled steps in Fig. 2f. They conform the experimental data in Fig. 3a, not only on the spectral envelopes but also on the number of fringes (marked with arrows). Figure 3c presents the simulated time-domain pulse shapes, revealing that the fringes are related to the oscillating tails at the bottom of the dark pulse (marked with arrows). Unlike the steps of bright dissipative solitons Guo et al. (2017); Voloshin et al. (2021), we demonstrate here that platicon steps can occur even in the single-platicon state, i.e. only one dark pulse in the microresonator.
3b展示了图2f中每个标记台阶的模拟平板光谱。它们不仅在光谱包络上,而且在条纹数量(用箭头标记)上都与图3a中的实验数据相符。图3c展示了模拟的时间域脉冲形状,表明条纹与暗脉冲底部的振荡尾部有关(用箭头标记)。与 Guo等人2017 年);Voloshin等人2021 年的亮耗散孤子台阶不同,我们在此证明,即使在单平板状态下,即微谐振器中只有一个暗脉冲,也可以发生平板台阶。

Low-noise photonic microwave generation. Finally, we showcase an immediate application of our findings, i.e. to generate low-noise photonic microwave which are ubiquitously deployed in modern information systems for wireless communication, timing and radar. Photonic microwave – microwave synthesized via photonics – allows unrivalled noise performance and bandwidth breaking the bottleneck of their electronic counterparts Fortier et al. (2011); Li et al. (2014); Xie et al. (2016). Among various photonics approaches, microcombs, which coherently channel microwave and terahertz frequency to optical frequency, offer an appealing solution for low-noise photonic microwave and millimeter-wave generation Liu et al. (2020); Zhang et al. (2019); Wang et al. (2021); Tetsumoto et al. (2021); Yao et al. (2022). Photodetection of the microcomb’s pulse stream creates a microwave whose carrier frequency corresponds to the microcomb’s repetition rate frepf_{\text{rep}}italic_f start_POSTSUBSCRIPT rep end_POSTSUBSCRIPT. Several methods have been implemented to further improve the microwave’s spectral purity, including the use of an external microwave Weng et al. (2019), an auxiliary laser Liu et al. (2024), a transfer comb Lucas et al. (2020), or two-point optical frequency division Kudelin et al. (2024); Sun et al. (2024b); Zhao et al. (2024).
低噪声光子微波产生。最后,我们展示了我们的研究成果的直接应用,即产生低噪声光子微波,这种微波广泛用于现代信息系统中的无线通信、授时和雷达。光子微波——通过光子学合成的微波——具有无与伦比的噪声性能和带宽,突破了电子微波的瓶颈Fortier等人2011 年);Li等人2014 年);Xie等人2016 年 。在各种光子学方法中,微梳将微波和太赫兹频率相干地引导到光频率,为低噪声光子微波和毫米波产生提供了一种有吸引力的解决方案Liu等人2020 年);Zhang等人2019 年);Wang等人2021 年);Tetsumoto等人2021 年);Yao等人2022 年 。对微梳脉冲流的光电检测产生了一个微波,其载波频率对应于微梳的重复率 frepsubscriptf_{\text{rep}}italic_f start_POSTSUBSCRIPT rep end_POSTSUBSCRIPT 。 为了进一步提高微波的光谱纯度,已经实施了多种方法,包括使用外部微波Weng2019 、辅助激光器Liu2024 、传输梳Lucas2020或两点光频分Kudelin2024 );Sun2024b );Zhao2024

Here we harness the platicon switching dynamics, and demonstrate a unique noise-quenching effect allowing suppression of microwave’s phase noise. In contrast to methods mentioned above, our system is free-running, all passive (i.e. without any active locking), and elegantly simple (i.e.without extra RF or laser sources). Experimentally, we collect the output light from the microresonator with a commercial photodetector (PD). The PD converts the dark pulse stream of frep=10.7f_{\text{rep}}=10.7italic_f start_POSTSUBSCRIPT rep end_POSTSUBSCRIPT = 10.7 GHz to a 10.7-GHz microwave in the critical X-band which is dedicated for radar, wireless networks, and satellite communication.
在这里,我们利用 platicon 开关动力学,并展示出独特的噪声抑制效果,从而抑制微波的相位噪声。 与上述方法相比,我们的系统是自由运行的、完全被动的(即没有任何主动锁定)并且非常简单(即没有额外的射频或激光源)。 实验中,我们利用商用光电探测器(PD)收集微谐振器的输出光。 PD 将 frep=10.7subscript10.7f_{\text{rep}}=10.7italic_f start_POSTSUBSCRIPT rep end_POSTSUBSCRIPT = 10.7 GHz 的暗脉冲流转换为关键 X 波段的 10.7 GHz 微波,该波段专用于雷达、无线网络和卫星通信。

By varying the laser current IIitalic_I within the SIL regime, we monitor the platicon’s frepf_{\text{rep}}italic_f start_POSTSUBSCRIPT rep end_POSTSUBSCRIPT and phase noise SϕS_{\phi}italic_S start_POSTSUBSCRIPT italic_ϕ end_POSTSUBSCRIPT at 10 kHz Fourier offset frequency of the 10.7-GHz microwave, as shown in Fig. 3d. For each step, a local minimum of SϕS_{\phi}italic_S start_POSTSUBSCRIPT italic_ϕ end_POSTSUBSCRIPT is observed. Meanwhile, it coincides with the local maximum of frepf_{\text{rep}}italic_f start_POSTSUBSCRIPT rep end_POSTSUBSCRIPT, as highlighted with vertical dashed lines in Fig. 3d. This coincidence is due to that, at the local maximum of frepf_{\text{rep}}italic_f start_POSTSUBSCRIPT rep end_POSTSUBSCRIPT, we have dfrep/dI=0\text{d}f_{\text{rep}}/\text{d}I=0d italic_f start_POSTSUBSCRIPT rep end_POSTSUBSCRIPT / d italic_I = 0, as highlighted with horizontal dashed lines in Fig. 3d. Therefore, frepf_{\text{rep}}italic_f start_POSTSUBSCRIPT rep end_POSTSUBSCRIPT becomes insensitive to current-noise-induced laser frequency jitter, resulting in the lowest SϕS_{\phi}italic_S start_POSTSUBSCRIPT italic_ϕ end_POSTSUBSCRIPT. More analysis of this coincidence is found in Supplementary Information Note 4. In fact, this effect is similar to the “quiet point” effect observed in bright dissipative solitons without SIL Yi et al. (2017); Triscari et al. (2023). The underlying mechanism is attributed to the multi-mode coupling between the laser cavity and the Si3N4 microresonator, which causes asymmetric comb-line enhancement or suppression, as shown in Fig. 3a. As a result, frepf_{\text{rep}}italic_f start_POSTSUBSCRIPT rep end_POSTSUBSCRIPT depends on the laser frequency (be more specifically, the detuning of the free-running laser frequency to the cold resonance frequency). In our theoretical model, only the pump resonance is considered coupled, i.e. b0b_{0}italic_b start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT in Eq. (2). A more precise simulation on the optical spectra should include multi-mode coupling by substituting b0b_{0}italic_b start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT with μbμeiμd1(ττs)\sum_{\mu}b_{\mu}e^{i\mu d_{1}(\tau-\tau_{\text{s}})}∑ start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT italic_b start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT italic_e start_POSTSUPERSCRIPT italic_i italic_μ italic_d start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_τ - italic_τ start_POSTSUBSCRIPT s end_POSTSUBSCRIPT ) end_POSTSUPERSCRIPT.
通过在 SIL 范围内改变激光电流 IIitalic_I ,我们监测了 10.7 GHz 微波 10 kHz 傅里叶偏移频率处的平面波 frepsubscriptf_{\text{rep}}italic_f start_POSTSUBSCRIPT rep end_POSTSUBSCRIPT 和相位噪声 SϕsubscriptS_{\phi}italic_S start_POSTSUBSCRIPT italic_ϕ end_POSTSUBSCRIPT ,如图3 d 所示。对于每一步,都可以观察到局部最小值 SϕsubscriptS_{\phi}italic_S start_POSTSUBSCRIPT italic_ϕ end_POSTSUBSCRIPT 。同时,它与局部最大值 frepsubscriptf_{\text{rep}}italic_f start_POSTSUBSCRIPT rep end_POSTSUBSCRIPT 重合,如图3 d 中垂直虚线突出显示的那样。这种巧合是由于在局部最大值 frepsubscriptf_{\text{rep}}italic_f start_POSTSUBSCRIPT rep end_POSTSUBSCRIPT 处,我们有 dfrep/dI=0subscript0\text{d}f_{\text{rep}}/\text{d}I=0d italic_f start_POSTSUBSCRIPT rep end_POSTSUBSCRIPT / d italic_I = 0 ,如图3 d 中水平虚线突出显示的那样。因此, frepsubscriptf_{\text{rep}}italic_f start_POSTSUBSCRIPT rep end_POSTSUBSCRIPT 对电流噪声引起的激光频率抖动不敏感,从而导致最低的 SϕsubscriptS_{\phi}italic_S start_POSTSUBSCRIPT italic_ϕ end_POSTSUBSCRIPT 。关于这种巧合的更多分析见补充信息注释 4。事实上,这种效应类似于在没有 SIL 的亮耗散孤子中观察到的“安静点”效应Yi等人2017 年);Triscari等人2023 年 。其根本机制归因于激光腔和 Si 3 N 4微谐振器之间的多模耦合,这会导致不对称梳状线增强或抑制,如图3a所示。因此, frepsubscriptf_{\text{rep}}italic_f start_POSTSUBSCRIPT rep end_POSTSUBSCRIPT 取决于激光频率(更具体地说,是自由运行激光频率与冷谐振频率的失谐)。在我们的理论模型中,仅考虑泵浦谐振耦合,即方程 (2) 中的 b0subscript0b_{0}italic_b start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT 。更精确的光谱模拟应包括多模耦合,方法是用 μbμeiμd1(ττs)subscriptsubscriptsuperscriptsubscript1subscript\sum_{\mu}b_{\mu}e^{i\mu d_{1}(\tau-\tau_{\text{s}})}∑ start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT italic_b start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT italic_e start_POSTSUPERSCRIPT italic_i italic_μ italic_d start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ( italic_τ - italic_τ start_POSTSUBSCRIPT s end_POSTSUBSCRIPT ) end_POSTSUPERSCRIPT 代替 b0subscript0b_{0}italic_b start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT

The SϕS_{\phi}italic_S start_POSTSUBSCRIPT italic_ϕ end_POSTSUBSCRIPT spectra of the local maximum and minimum points, marked with green and black dots in Fig. 3d, are measured and compared in Fig. 3e, showing 23.5 dB noise reduction. The lowest SϕS_{\phi}italic_S start_POSTSUBSCRIPT italic_ϕ end_POSTSUBSCRIPT reaches 45/75/105-45/-75/-105- 45 / - 75 / - 105 dBc/Hz at 0.1/1/10 kHz Fourier offset frequency, and is limited by the shot noise floor at higher frequency offset Savchenkov et al. (2008). Figure 3e inset shows the beatnote of the lowest SϕS_{\phi}italic_S start_POSTSUBSCRIPT italic_ϕ end_POSTSUBSCRIPT (black data) with 10 Hz resolution bandwidth (RBW). Supplementary Information Note 5 illustrates that, such a noise quenching phenomenon has been also observed in parallel experiments with different Si3N4 microresonators and DFB lasers, thus is universal.
在图3e中测量并比较了图3d中用绿点和黑点标记的局部最大值和最小值点的 SϕsubscriptS_{\phi}italic_S start_POSTSUBSCRIPT italic_ϕ end_POSTSUBSCRIPT 光谱,显示噪声降低了 23.5 dB。最低的 SϕsubscriptS_{\phi}italic_S start_POSTSUBSCRIPT italic_ϕ end_POSTSUBSCRIPT 在 0.1/1/10 kHz 傅里叶偏移频率时达到 45/75/105-45/-75/-105- 45 / - 75 / - 105 dBc/Hz,并在更高频率偏移处受到散粒噪声基底的限制Savchenkov等人2008 年 。图3e插图显示了最低 SϕsubscriptS_{\phi}italic_S start_POSTSUBSCRIPT italic_ϕ end_POSTSUBSCRIPT (黑色数据)的拍音,分辨率带宽(RBW)为 10 Hz。补充信息注释 5 表明,这种噪声猝灭现象也在使用不同的 Si 3 N 4微谐振器和 DFB 激光器的平行实验中观察到,因此具有普遍性。

In conclusion, we unveil an intriguing yet universal Kerr-thermal dynamics of a semiconductor laser self-injection-locked to a Si3N4 microresonator where platicon microcombs are formed. We experimentally observe and characterize the platicon switching dynamics with discrete steps. We further establish a comprehensive theoretical model cooperating the laser-cavity dynamics, the platicon formation dynamics, the photo-thermal dynamics, and the mutual coupling between them. Numerical simulation confirms the experimental result, and illuminates that the platicon switching phenomenon is originated from the synergy of SIL, Kerr nonlinearity and the photo-thermal effect. Exploiting this finding, we showcase low-noise microcomb-based microwave generation. Via operation of platicons with specific laser current, we achieve 23.5 dB phase noise quenching of the 10.7-GHz microwave carrier. Our study not only add critical insight of pulse formation in linear-and-nonlinear-coupled laser-microresonator systems, but also offer a neat solution for photonic-chip-based microwave oscillators with high spectral purity, ideal for microwave photonics, coherent optical communication, analog-to-digital conversion, wireless links, and radar.
总之,我们揭示了一种有趣而又具有普适性的克尔热动力学,该动力学自注入锁定在 Si 3 N 4微谐振器上,其中形成了扁梳微梳。我们通过实验观察并以离散步骤表征了扁梳开关动力学。我们进一步建立了一个综合的理论模型,该模型结合了激光腔动力学、扁梳形成动力学、光热动力学以及它们之间的相互耦合。数值模拟证实了实验结果,并阐明了扁梳开关现象源于 SIL、克尔非线性和光热效应的协同作用。利用这一发现,我们展示了基于低噪声微梳的微波产生。通过以特定激光电流操作扁梳,我们实现了 10.7 GHz 微波载波的 23.5 dB 相位噪声抑制。我们的研究不仅增加了线性和非线性耦合激光微谐振器系统中脉冲形成的关键见解,而且还为基于光子芯片的高光谱纯度微波振荡器提供了一个完美的解决方案,非常适合微波光子学、相干光通信、模数转换、无线链路和雷达。

Acknowledgments: We thank Baoqi Shi for characterizing the Si3N4 chips, and Zhiyang Chen for assistance in the experiment. We acknowledge support from the National Natural Science Foundation of China (Grant No.12261131503, 12404436), Innovation Program for Quantum Science and Technology (2023ZD0301500), Guangdong-Hong Kong Technology Cooperation Funding Scheme (Grant No. 2024A0505040008), Shenzhen-Hong Kong Cooperation Zone for Technology and Innovation (HZQB-KCZYB2020050), and Shenzhen Science and Technology Program (Grant No. RCJC20231211090042078). Russian Quantum Center was supported by RSF grant (23-42-00111). The DFB lasers were fabricated by Shenzhen PhotonX Technology Co. Ltd. and Henan Shijia Photons Technology Co. Ltd. . Silicon nitride chips were fabricated by Qaleido Photonics.
致谢:感谢石宝琪对 Si 3 N 4芯片进行表征,感谢陈志阳对实验的协助。我们感谢国家自然科学基金(批准号 12261131503、12404436)、量子科学技术创新计划(2023ZD0301500)、粤港科技合作资助计划(批准号 2024A0505040008)、深港科技创新合作区(HZQB-KCZYB2020050)和深圳市科技计划(批准号 RCJC20231211090042078)的支持。俄罗斯量子中心得到了 RSF 拨款(23-42-00111)的支持。DFB 激光器由深圳市光子科技有限公司和河南世嘉光子科技有限公司制造。氮化硅芯片由 Qaleido Photonics 制造。

Author contributions: W. S. and J. Liu conceived the experiment. S. L., W. S., J. Long and X. Y. built the experimental setup. S. L. and W. S. performed the experiments and analyzed the data, with the assistance from A. E. S. and N. Y. D.. K. Y., D. A. C., and W. S. developed the theory and performed the simulation, with the assistance from A. E. S., N. Y. D. and I. A. B.. C. S. fabricated the Si3N4 chips. W. S., S. L., K. Y. and J. Liu wrote the manuscript, with input from others. J. Liu supervised the project.
作者贡献:WS 和 J. Liu 构思了实验。SL、WS、J. Long 和 XY 建立了实验装置。SL 和 WS 在 AES 和 NYD 的帮助下进行了实验并分析了数据。KY、DAC 和 WS 在 AES、NYD 和 IAB 的帮助下开发了理论并进行了模拟。CS 制造了 Si 3 N 4芯片。WS、SL、KY 和 J. Liu 撰写了手稿,并得到了其他人的帮助。J. Liu 负责监督该项目。

Data Availability Statement: The code and data used to produce the plots within this work will be released on the repository Zenodo upon publication of this preprint.
数据可用性声明:本研究中使用的绘图代码和数据将在本预印本发布后Zenodo存储库中发布。

References  参考