置换阻塞路径积分Monte Carlo方法研究有限温度下均匀电子气
Permutation blocking path integral Monte Carlo approach to the uniform electron gas at finite temperature
摘要:
有限温度下的均匀电子气(UEG)由于其与包括稠密等离子体和激光激发固体在内的许多应用的关键相关性而受到高度关注。特别是,密度泛函理论在很大程度上依赖于准确的热力学数据的UEG。直到最近,唯一现有的第一原理的结果已经获得了$N=33$电子与限制路径积分蒙特卡罗(RPIMC),为低到中等密度,$r_s = \overline{r}/a_B \gtrsim 1$。该数据已通过$r_s \leq 1$的配置路径积分蒙特卡罗(CPIMC)模拟得到补充,该模拟基本上偏离RPIMC,朝向更小的$r_s$和低温。在这项工作中,我们提出了一个独立的第三种方法--最近发展的排列阻塞路径积分蒙特卡罗方法(PB-PIMC)[T。Dornheim \textit{et al.},NJP \textbf{17},073017(2015)],我们将其扩展到UEG。有趣的是,PB-PIMC允许我们在整个密度范围内进行模拟,直到费米温度的一半($θ=k_BT/E_F=0.5$),因此,将我们的结果与上述两种方法进行比较。虽然我们发现与CPIMC的结果非常一致,但我们观察到RPIMC的偏差超出了统计误差,并随密度增加而增加。
英文摘要:
The uniform electron gas (UEG) at finite temperature is of high current interest due to its key relevance for many applications including dense plasmas and laser excited solids. In particular, density functional theory heavily relies on accurate thermodynamic data for the UEG. Until recently, the only existing first-principle results had been obtained for $N=33$ electrons with restricted path integral Monte Carlo (RPIMC), for low to moderate density, $r_s = \overline{r}/a_B \gtrsim 1$. This data has been complemented by Configuration path integral Monte Carlo (CPIMC) simulations for $r_s \leq 1$ that substantially deviate from RPIMC towards smaller $r_s$ and low temperature. In this work, we present results from an independent third method---the recently developed permutation blocking path integral Monte Carlo (PB-PIMC) approach [T. Dornheim \textit{et al.}, NJP \textbf{17}, 073017 (2015)] which we extend to the UEG. Interestingly, PB-PIMC allows us to perform simulations over the entire density range down to half the Fermi temperature ($θ=k_BT/E_F=0.5$) and, therefore, to compare our results to both aforementioned methods. While we find excellent agreement with CPIMC, where results are available, we observe deviations from RPIMC that are beyond the statistical errors and increase with density.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。