光与物质的相互作用是凝聚态物理领域的一个研究热门，研究该方向一方面可以得到对材料微观性质的表征手段，另一方面也可以从中获得更强的激光光源。20世纪80年代开始，人们发现了在强光场作用下的高次谐波现象，即可以探测到一系列整数倍入射光频率的光谱峰。近二十年来，固体中的高次谐波已经成为了一个新兴的研究领域，这是因为，固体高次谐波被报导具有诸多应用前景，比如重构电子能带结构和作为一种先进光源等等。

近年来，高次谐波的机理研究正在飞速发展。迄今为止，主流的高次谐波理论为三步模型：1. 隧穿激发电子空穴对；2.在能带中受外电场作用加速；3. 电子空穴对复合。基于三步模型主要发展出了三种相对完善的模拟计算方法，可以对固体中的高次谐波进行模拟计算，并解释部分实验现象。然而，三步模型只能考虑导带和价带电子行为的影响，对于固体材料的本征固有属性的影响则无法考虑，比如，材料对称性或者缺陷态等等。这些属性的存在使得三步模型无法适用于许多材料中。为了填补这部分的空白，本文的主要研究内容为材料的物理特性对高次谐波的产生和调控的影响。本文主要的工作和创新点如下：

(1) 研究了量子限域效应对高次谐波行为的影响。这是一种常见于低维材料中的效应，然而三步模型无法准确描述其影响。本人首次发现了硒化镉/锌镉硫纳米板的静态高次谐波信号有着较为独特的现象：当壳层厚度处于一个中间值时，高次谐波的信号是最强的。通过势阱建模，用模拟计算结果复现出了该实验结果。分别从能量角度和空间角度探索量子限域效应对高次谐波信号的影响，发现能量角度和空间角度的影响是相反的。能量限制在壳层厚度越大时使得高次谐波信号越强，而空间限制则相反。这两个竞争机制的共同作用导致了这一现象。

(2) 研究了载流子动力学对高次谐波行为的影响。本人观测到高次谐波在泵浦光的作用下强度被抑制的现象，并且不同波长的泵浦光会引起不同的高次谐波动力学过程。对于各向同性氧化锌，400 nm泵浦光会引起皮秒量级的快过程，而800 nm泵浦光会引起百飞秒量级的快过程。通过一系列补充实验，将400 nm泵浦的快过程归因于导带底连续带的弛豫，800 nm泵浦的快过程归因于混频辐射过程，而慢过程则归因于导带电子的弛豫过程。载流子动力学行为影响高次谐波主要机理为泵浦光诱导的非相干电子的注入，本人的工作改进了这一机理。尤其是800 nm泵浦条件下，红外光与泵浦光共同作用注入了非相干电子，并消耗了产生高次谐波的红外光能量。此外，对于各向异性氧化锌，发现红外光的偏振主要影响载流子动力学过程，而泵浦光的偏振主要影响载流子激发数量。对于800 nm泵浦时反常的慢过程信号，这应为多光子激发与彩斑混频辐射竞争导致。

(3) 研究了缺陷态对高次谐波行为的影响。缺陷态一般不参与高次谐波过程，本人采用了三能级系统的受激辐射过程作为主要机理，通过氧化锌样品中天然的缺陷态形成了三能级系统，在实验上观测到特定阶次的高次谐波能被缺陷态增强的现象，其最大倍数为1.8。并且，氧化锌的两个缺陷态都能有效增强高次谐波的信号。同时本人也将高次谐波发展为一种缺陷探测的手段。实验中发现，不同缺陷态的寿命有较大的差别，且会随着泵浦通量的增大会逐渐增大，其中2.20 eV缺陷态载流子的时间常数为几十皮秒量级，而2.64 eV缺陷态载流子的时间常数为几皮秒量级。两者时间上的差别归因于缺陷态之间的载流子弛豫过程。

The interaction between light and matter is a hot research topic in the field of condensed matter physics, which can not only expand the means of characterizing the microscopic properties of materials, but also develop the laser sources with more intense pulses. Since 1980s, the phenomena of high harmonic generation (HHG) have been discovered with intense light field, i.e., integer multiple peaks of the incident light frequency can be detected in spectra. In the past twenty years, HHG in solid is becoming an emerging research area. The main reason is that HHG in solid has been reported to have many application prospects, e.g., reconstructing electronic band structure and developing advanced light source.

In recent years, research on the mechanism of HHG has been rapidly developing. Up to now, the mainstream mechanism of HHG can be described by the three-step model: (i) tunneling, (ii) acceleration and (iii) recombination. Based on this model, three relatively complete calculation methods have been put forward to calculate HHG in solids, and some experimental phenomena can be explained. However, the three-step model can only consider the effects of conduction and valence band electronic behavior, and cannot consider the effects of intrinsic properties of solid materials, e.g., material symmetry or defect states. The existence of these properties makes the three-step model unsuitable for many materials. In order to fill this gap, the main research content of this article is the impact of material physical properties on the generation and modulation of HHG. The main work and innovation points of this article are as follows:

(1) The influence of quantum confinement effect on HHG behavior has been studied. This is a common effect in low dimensional materials, while the three-step model cannot accurately describe its impact. The HHG in CdSe/ZnCdS core/shell nanoplatelet has been first discovered a unique phenomenon: HHG intensity would be strong when the shell thickness is specific. Via modeling the potential well, the calculation results can recurrent the experimental results. Analyzing the quantum-confinement-effect influenced HHG with both energy perspective and spatial perspective, it is found that the influence with energy perspective and spatial perspective are contrary. Energy constraints result in stronger HHG signals with greater thickness, while spatial constraints are the opposite. The coupling of these two competitive mechanisms should be the explanation of this phenomenon.

(2) The influence of carrier dynamics on HHG behavior has been studied. It has been observed that pump pulse would suppress the intensity of HHG. For isotropic zinc oxide, the pump pulse with different wavelengths would induce different dynamic processes, i.e., the pump with 400 nm would induce fast process of a few picoseconds while the pump with 800 nm would induce fast process of a few hundreds of femtoseconds. Through a series of supplementary experiments, the fast process of 400 nm pump is attributed to the decay process of the quasi-continuum broad defect state, while the fast process of 800 nm pump is attributed to mixing-frequency spots radiation. The slow process is attributed to the decay process of the conduction band. The main mechanism of the influence of carrier dynamics on higher harmonics is the pump-induced injection of incoherent electrons, which should be improved by this work. Especially under 800 nm pump conditions, the combination of infrared light and pump light would inject incoherent electrons and consume the energy of infrared light which can generate HHG. For anisotropic zinc oxide, it has been found that the polarization of infrared light mainly affects the carrier dynamics process, while the polarization of pump light mainly affects the carrier excitations. Meanwhile, abnormal slow process signals were also found when pumping at 800 nm, which should be caused by competition between multiphoton excitation and color spot mixing radiation.

(3) The influence of defect state on HHG behavior has been studied. The defect state generally does not participate in the HHG process. I have adopted the stimulated emission process of a three-level system as the main mechanism, and formed a three-level system through the inherent defect state in zinc oxide samples. In the experiments, it has been observed that the specific order of HHG can be enhanced by the defect state, with a maximum multiple of 1.8. Furthermore, both two native defects can be used to effectively enhance HHG. At the same time, I have also developed HHG as a means of defect detection. In the experiment, the lifetimes of the different defects are quite different and they would be larger when the pump fluence is higher. The time constant of the defect in 2.20 eV is tens of picoseconds while it is several picoseconds for that of 2.64 eV. The difference in time between the two is attributed to the carrier relaxation process between defect states.