德鲁德模型
激发态
硅
电介质
含时密度泛函理论
激发
热的
极化(电化学)
材料科学
原子物理学
物理
计算物理学
凝聚态物理
密度泛函理论
电子
碰撞频率
激光器
极化率
航程(航空)
频率响应
介电常数
介电响应
等离子体
分子物理学
有效质量(弹簧-质量系统)
热平衡
热导率
作者
Sato, S. A.,Yabana, K.,Shinohara Y,Otobe, T.,Bertsch, G. F.
出处
期刊:Cornell University - arXiv
日期:2013-03-13
标识
DOI:10.48550/arxiv.1303.3249
摘要
We calculate the dielectric response of crystalline silicon following irradiation by a high-intensity laser pulse, modeling the dynamics by time-dependent density functional theory (TDDFT). The pump-probe measurements are numerically simulated by solving the time-dependent Kohn-Sham equation with the pump and probe fields included as external fields. As expected, the excited silicon shows features of a particle-hole plasma in its response. We compare the calculated response with a thermal model and with a simple Drude model. The thermal model requires only a static DFT calculation to prepare electronically excited matter and agrees rather well with the TDDFT for the same particle-hole density. The Drude model with two fitted parameters (electron effective mass and collision time) also shows fair agreement at the lower excitation energies; the fitted effective masses are consistent with carrier-band dispersions. The extracted Drude lifetimes range from 6 fs at weak pumping fields to much lower values at high fields. However, we find that the Drude model does not give a good fit to the imaginary dielectric function at the highest fields. Comparing the thermal model with the Drude, we find that the extracted lifetimes are in the same range, 1-13 fs depending on the temperature. These short Drude lifetimes show that strong damping is possible in the TDDFT, despite the absence of electron scattering. One significant difference between the TDDFT response and the other models is that the response to the probe pulse depends on the polarization of the pump pulse. We also find that the imaginary part of the dielectric function can be negative, particularly for the parallel polarization of pump and probe fields.
科研通智能强力驱动
Strongly Powered by AbleSci AI