传输极限皮秒脉冲光纤放大的理论和实验研究

Study on picosecond-pulse amplification is lagged behind that of femtosecond-/nanosecond-pulse amplification. Due to the narrow bandwidth of the picosecond pulses, there is no suitable method to amplify the pulse especially when the transform-limited pulse is required. However, high-energy picosecond pulses are needed for a variety of applications, such as micromachining. In the design of high peak power picosecond fiber amplifiers, it is crucial to avoid excessive nonlinear phase shift, which leads to spectral broadening and eventually to distortion of the temporal profile. The spectrum will broaden when the pulse propagates a distance longer than the nonlinear length. Ideally, the nonlinear length should exceed the length of the amplifier. To this end, one can decrease the peak power, and/or use gain fibers with large mode area. In this project, we study the dynamics of picosecond pulse amplification and propose a method to achieve transform-limited picosecond-pulse amplification. The key technique is the division of the seed pulse before the amplification and the recombination of the individual amplified pulses to obtain a pulse with high peak power. Apart from the using of large-mode-area gain fibers, circular polarization instead of linear polarization is considered in the amplification to reduce the accumulation of the nonlinear phase shift as the nonlinear refraction coefficient of circularly-polarized light is 2/3 of that of linearly-polarized light. Direct linearly-polarized output is intrinsically included in this method as the output comes from a polarizing beam splitter. This avoids the degradation of the polarization extinction ratio that is observed at high power in direct amplification with non-polarization-maintaining gain fibers. Furthermore, the use of the Faraday rotator mirror automatically cancels any birefringence perturbations to the gain fiber. At the end of the project, a practical method for transform-limited picosecond-pulse amplification in fiber amplifier should be matured. Theoretical study will be carried out during the project to predict the limitation and the scale up of the novel method on the achievable pulse peak power level of the amplified picosecond pulses. We believe the technique we develop in this project will become a classical scheme for the picosecond-pulse amplification, which is similar to the role of the well-known chirped-pulse amplification for femtosecond pulses.

本项目针对皮秒脉冲放大难以兼顾大功率放大和获得传输极限放大脉冲的问题，提出和研究从根本上克服非线性效应对放大脉冲的光谱展宽、大幅提高放大脉冲的脉冲能量和峰值功率并同时获得好的光束输出质量的新技术方案。其核心是在放大前将皮秒种子脉冲等分复制成小脉冲串并加以时延，经光纤放大后再补偿时延并叠加所有放大脉冲从而获得具有高峰值功率的传输极限皮秒脉冲，除使用大模场面积增益光纤降低非线性效应外，还通过使用圆偏振放大替代线偏振放大以减少非线性相位积累，并且使用法拉第旋转镜自动抵消增益光纤中的双折射扰动，采用直接线性偏振输出避免在非保偏增益光纤中放大输出的偏振弱化效应。这一方案根本上解决了传输极限皮秒脉冲放大的问题。通过本项目对传输极限皮秒脉冲放大的深入研究，以期对皮秒量级脉冲的放大有一突破性的认识从而使对皮秒脉冲的利用可以赶上纳秒和飞秒脉冲。

针对皮秒脉冲放大难以兼顾大功率放大和获得传输极限放大脉冲的问题，提出和研究克服非线性效应对放大脉冲的光谱展宽、大幅提高放大脉冲的脉冲能量和峰值功率并同时获得好的光束输出质量的技术方案。其核心是在放大前将皮秒种子脉冲等分复制成小脉冲串并加以时延，经光纤放大后再补偿时延并叠加所有放大脉冲从而获得具有高峰值功率的传输极限皮秒脉冲。在基金支持下，研究内容涉及种子源产生脉冲的性质研究、矢量孤子的偏振调控、耗散孤子共振型脉冲的特性以及皮秒脉冲放大系统的参数确定等方面。通过数值模拟揭示了耗散孤子陡峭光谱边缘的形成机理；实验报道并解释了全正色散光纤激光器中耗散孤子的光谱变化和光谱跳变特性；在L波段用非线性偏振旋转锁模技术实现飞秒脉冲锁模；理论分析了光纤激光器中初始条件对暗孤子脉冲形成的影响；证实了光纤激光器中由腔导致的峰值功率钳位效应的存在，激光器性能依赖于具体的腔结构；对“类噪声”脉冲的自相关特性进行研究；耗散孤子的泵浦磁滞回线和双稳态得到证实；数值模拟确认周期分叉在总色散为零的色散管理光纤激光器中的存在；首次在掺铥光纤激光器中实验观测并理论复现矢量孤子的产生，证明了矢量孤子的生成与波长无关；通过数值模拟研究了2μm波段正色散光纤激光器中的耗散矢量孤子的产生及其性质；观测到矢量耗散孤子的束缚态；通过偏振操控群速度锁定的矢量孤子实现伪高阶群速度锁定矢量孤子；首次确认耗散孤子共振型脉冲具有特殊的啁啾结构；基于时分复制皮秒脉冲光纤放大的理论模型，数值研究了五级双折射晶体级联时分复制光纤放大系统的各参数对放大脉冲特性的影响。确定了在光纤放大器非线性系数已知并且最大增益有限的条件下，为获得10 MW峰值功率且保持变换极限的皮秒放大脉冲，提出了种子脉冲峰值功率和增益光纤长度的选择标准。