476 lines
12 KiB
TeX
476 lines
12 KiB
TeX
\documentclass[xcolor=x11names,compress]{beamer}
|
|
|
|
%% General document %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|
|
\usepackage{graphicx}
|
|
\usepackage{mathpazo}
|
|
\usepackage[english]{babel}
|
|
\usepackage[T1]{fontenc}
|
|
\usepackage[utf8]{inputenc}
|
|
\usepackage{xcolor}
|
|
\usepackage{siunitx}
|
|
\usepackage{graphicx}
|
|
\usepackage{physics}
|
|
\usepackage{multimedia}
|
|
\usepackage[absolute,overlay]{textpos}
|
|
\usepackage{ragged2e}
|
|
\usepackage{amssymb}
|
|
\usepackage[version=4]{mhchem}
|
|
\usepackage[style=verbose,backend=bibtex]{biblatex}
|
|
\bibliography{SlideToulouse}
|
|
|
|
|
|
|
|
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|
|
|
|
|
|
%% Beamer Layout %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|
|
\useoutertheme[subsection=false,shadow]{miniframes}
|
|
\useinnertheme{default}
|
|
|
|
|
|
\setbeamerfont{title like}{shape=\scshape}
|
|
\setbeamerfont{frametitle}{shape=\scshape}
|
|
\setbeamerfont{framesubtitle}{size=\normalsize}
|
|
\setbeamerfont{caption}{size=\scriptsize}
|
|
\setbeamercolor*{lower separation line head}{bg=DeepSkyBlue4}
|
|
\setbeamercolor*{normal text}{fg=black,bg=white}
|
|
\setbeamercolor*{alerted text}{fg=red}
|
|
\setbeamercolor*{example text}{fg=black}
|
|
\setbeamercolor*{structure}{fg=black}
|
|
|
|
|
|
\setbeamercolor*{palette tertiary}{fg=black,bg=black!10}
|
|
\setbeamercolor*{palette quaternary}{fg=black,bg=black!10}
|
|
\setbeamercolor{caption name}{fg=DeepSkyBlue4}
|
|
\setbeamercolor{title}{fg=DeepSkyBlue4}
|
|
\setbeamercolor{itemize item}{fg=DeepSkyBlue4}
|
|
\setbeamercolor{frametitle}{fg=DeepSkyBlue4}
|
|
\renewcommand{\(}{\begin{columns}}
|
|
\renewcommand{\)}{\end{columns}}
|
|
\newcommand{\<}[1]{\begin{column}{#1}}
|
|
\renewcommand{\>}{\end{column}}
|
|
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|
|
|
|
|
|
\setbeamertemplate{navigation symbols}{}
|
|
\setbeamertemplate{footline}[frame number]
|
|
\setbeamertemplate{caption}[numbered]
|
|
\setbeamertemplate{section in toc}[ball]
|
|
\setbeamertemplate{itemize items}[circle]
|
|
\beamerboxesdeclarecolorscheme{clair}{Coral4}{Ivory2}
|
|
\beamerboxesdeclarecolorscheme{foncé}{DarkSeaGreen4}{Ivory2}
|
|
|
|
\title[Title]{Perturbation theories in the complex plane}
|
|
\author[]{Antoine \textsc{Marie}}
|
|
\setbeamersize{text margin left=5mm}
|
|
\setbeamersize{text margin right=5mm}
|
|
\institute{Supervised by Pierre-François \textsc{LOOS}}
|
|
|
|
\begin{document}
|
|
|
|
|
|
|
|
|
|
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|
|
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|
|
|
|
\begin{frame}[plain]
|
|
|
|
\date{30th June 2020}
|
|
\titlepage
|
|
\end{frame}
|
|
|
|
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|
|
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|
|
|
|
\begin{frame}{Why do we use perturbation theories in computational chemistry?}
|
|
|
|
\pause[1]
|
|
|
|
The Hartree-Fock theory is \textcolor{Green4}{computationally cheap} and can be applied even to \textcolor{Green4}{large systems}.
|
|
|
|
But this method is missing the \textcolor{red}{correlation energy}...
|
|
|
|
\vspace{0.5cm}
|
|
|
|
\pause[2]
|
|
|
|
$\rightarrow$ We need methods to get this correlation energy!
|
|
|
|
\vspace{0.5cm}
|
|
|
|
\pause[3]
|
|
|
|
\begin{beamerboxesrounded}[scheme=foncé]{\centering A general method}
|
|
In physics perturbation theory is often a good way to improve the obtained results with an approximated Hamiltonian.
|
|
\end{beamerboxesrounded}
|
|
|
|
|
|
\end{frame}
|
|
|
|
\section{\textsc{Strange behaviors of the MP series}}
|
|
|
|
\begin{frame}{The Møller-Plesset perturbation theory}
|
|
|
|
\pause[1]
|
|
|
|
\begin{beamerboxesrounded}[scheme=foncé]{\centering Partitioning of the Hamiltonian}
|
|
|
|
\begin{equation}
|
|
H = H_0 + \lambda V
|
|
\end{equation}
|
|
|
|
\end{beamerboxesrounded}
|
|
|
|
\begin{itemize}
|
|
\centering
|
|
\item $H_0$: Unperturbed Hamiltonian
|
|
\item $V$: Perturbation operator
|
|
\end{itemize}
|
|
|
|
\pause[2]
|
|
|
|
\begin{beamerboxesrounded}[scheme=foncé]{\centering The Fock operator}
|
|
|
|
\begin{equation}
|
|
F = \sum\limits_{i=1}^{n} f(i) \hspace{0.3cm} ; \hspace{0.3cm} f(i) = h(i) + \sum\limits_{i=1}^{n/2} \left[2J_j(i) - K_j(i)\right]
|
|
\end{equation}
|
|
|
|
\end{beamerboxesrounded}
|
|
|
|
\begin{itemize}
|
|
\centering
|
|
\item $f(i)$: Fock operator
|
|
\item $h(i)$: One electron Hamiltonian
|
|
\item $J_j(i)$: Coulomb operator
|
|
\item $K_j(i)$: Exchange operator
|
|
\end{itemize}
|
|
|
|
\pause[3]
|
|
|
|
\begin{beamerboxesrounded}[scheme=foncé]{}
|
|
\centering
|
|
Full Configuration Interaction gives access to high-order terms of the perturbation series !
|
|
\end{beamerboxesrounded}
|
|
|
|
\end{frame}
|
|
|
|
\begin{frame}{Deceptive or slow convergences\footcite{gill_deceptive_1986}}
|
|
|
|
\begin{figure}
|
|
\centering
|
|
\includegraphics[width=0.45\textwidth]{gill1986.png}
|
|
\caption{\centering Barriers to homolytic fission of \ce{He2^2+} at MPn/STO-3G level ($n = 1$--$20$).}
|
|
\label{fig:my_label}
|
|
\end{figure}
|
|
|
|
|
|
\end{frame}
|
|
|
|
\begin{frame}{Multi-reference and spin contamination\footcite{gill_why_1988}}
|
|
\begin{table}
|
|
\centering
|
|
\begin{tabular}{c c c c c c c}
|
|
\hline
|
|
$r$ & UHF & UMP2 & UMP3 & UMP4 & $\expval{S^2}$ \\
|
|
\hline
|
|
0.75 & 0.0\% & 63.8\% & 87.4\% & 95.9\% & 0.00\\
|
|
1.35 & 0.0\% & 15.2\% & 26.1\% & 34.9\% & 0.49\\
|
|
2.00 & 0.0\% & 01.0\% & 01.8\% & 02.6\% & 0.95\\
|
|
2.50 & 0.0\% & 00.1\% & 00.3\% & 00.4\% & 0.99\\
|
|
\hline
|
|
\end{tabular}
|
|
\caption{\centering Percentage of electron correlation energy recovered and $\expval{S^2}$ for the \ce{H2} molecule as a function of bond length (r,\si{\angstrom}) in the STO-3G basis set.}
|
|
\label{tab:my_label}
|
|
\end{table}
|
|
|
|
\end{frame}
|
|
|
|
\begin{frame}{Divergent cases}
|
|
|
|
\begin{figure}
|
|
\centering
|
|
\includegraphics[width=0.6\textwidth]{The-energy-corrections-for-HF-at-stretched-geometry-in-the-cc-pVDZ-basis.png}
|
|
\caption{The energy corrections for HF at stretched geometry in the cc-pVDZ basis. \footcite{olsen_divergence_2000}}
|
|
\label{fig:my_label}
|
|
\end{figure}
|
|
|
|
\end{frame}
|
|
|
|
\section{The complex plane}
|
|
|
|
\begin{frame}{A simple example}
|
|
|
|
\begin{columns}
|
|
|
|
\column{0.48\textwidth}
|
|
|
|
\begin{beamerboxesrounded}[scheme=foncé]{An example function}
|
|
|
|
\begin{equation*}
|
|
\frac{1}{1 + x^4}
|
|
\end{equation*}
|
|
|
|
\end{beamerboxesrounded}
|
|
|
|
\vspace{1cm}
|
|
|
|
\begin{itemize}
|
|
\item Smooth for $x \in \mathbb{R}$
|
|
|
|
\item Infinitely differentiable in $\mathbb{R}$
|
|
\end{itemize}
|
|
|
|
\column{0.48\textwidth}
|
|
|
|
\begin{figure}
|
|
\centering
|
|
\includegraphics[width=0.6\textwidth]{exemplesingu.pdf}
|
|
\caption{Plot of $1/(1+x^4)$}
|
|
\label{fig:my_label}
|
|
\end{figure}
|
|
|
|
\end{columns}
|
|
|
|
But the Taylor expansion of this function does not converge for $x\geq1$...
|
|
\vspace{0.3cm}
|
|
\centering Why ?
|
|
|
|
\end{frame}
|
|
|
|
\begin{frame}{And if we look in the complex plane?}
|
|
|
|
\begin{columns}
|
|
|
|
\column{0.48\textwidth}
|
|
|
|
\centering The function has 4 singularities in the complex plane!
|
|
|
|
\vspace{1cm}
|
|
|
|
$x = e^{i\pi/4}, e^{-i\pi/4}, e^{i3\pi/4}, e^{-i3\pi/4}$
|
|
|
|
\column{0.48\textwidth}
|
|
|
|
\begin{figure}
|
|
\centering
|
|
\includegraphics[width=0.6\textwidth]{possingu.pdf}
|
|
\caption{\centering Singularities of the function $1/(1+x^4)$}
|
|
\label{fig:my_label}
|
|
\end{figure}
|
|
|
|
\end{columns}
|
|
|
|
The \textcolor{red}{radius of convergence} of the Taylor expansion of a function is equal to the distance of the \textcolor{red}{closest singularity} to the origin in the \textcolor{red}{complex plane}.
|
|
|
|
\end{frame}
|
|
|
|
\begin{frame}{Extending chemistry in the complex plane}
|
|
|
|
\begin{beamerboxesrounded}[scheme=foncé]{\centering $\lambda$ a complex variable}
|
|
|
|
\begin{equation*}
|
|
H(\lambda) = H_0 + \lambda V
|
|
\end{equation*}
|
|
|
|
\end{beamerboxesrounded}
|
|
|
|
\begin{columns}
|
|
|
|
\column{0.48\textwidth}
|
|
|
|
\begin{itemize}
|
|
\item $n$ Riemann sheets
|
|
\vspace{0.3cm}
|
|
\item Exceptional points interconnecting the sheets
|
|
\vspace{0.3cm}
|
|
\item No ordering property in the complex plane
|
|
\end{itemize}
|
|
|
|
\column{0.48\textwidth}
|
|
|
|
\begin{figure}
|
|
\centering
|
|
\includegraphics[width=0.7\textwidth]{riemannsheet.png}
|
|
\label{fig:my_label}
|
|
\end{figure}
|
|
|
|
\end{columns}
|
|
|
|
\end{frame}
|
|
|
|
\section{Classifying the singularity}
|
|
|
|
\begin{frame}{Which features of the system localize the singularities?}
|
|
|
|
\begin{itemize}
|
|
\item Partitioning of the Hamiltonian: Møller-Plesset, Epstein-Nesbet, \ldots
|
|
\item Zeroth-order reference: weak or strong correlation.
|
|
\item Finite or complete basis set.
|
|
\item Localized or delocalized basis functions.
|
|
\end{itemize}
|
|
|
|
\end{frame}
|
|
|
|
\begin{frame}{A two-state model\footcite{olsen_divergence_2000}}
|
|
|
|
\begin{columns}
|
|
|
|
\column{0.48\textwidth}
|
|
|
|
\begin{figure}
|
|
\centering
|
|
\includegraphics[width=0.8\textwidth]{avoidedcrossing.pdf}
|
|
\caption{Example of an avoided crossing.}
|
|
\label{fig:my_label}
|
|
\end{figure}
|
|
|
|
\column{0.48\textwidth}
|
|
|
|
\begin{beamerboxesrounded}[scheme=foncé]{A 2x2 matrix}
|
|
\centering \small{$\mqty(\alpha & \delta \\ \delta & \beta) =$}
|
|
|
|
\vspace{0.15cm}
|
|
|
|
\small{$\mqty(\alpha + \alpha_s & 0 \\ 0 & \beta + \beta_s ) + \mqty(- \alpha_s & \delta \\ \delta & - \beta_s)$}
|
|
|
|
\end{beamerboxesrounded}
|
|
\vspace{1cm}
|
|
\end{columns}
|
|
|
|
\end{frame}
|
|
|
|
\begin{frame}{Two-state model\footcite{olsen_divergence_2000}}
|
|
|
|
\begin{figure}
|
|
\centering
|
|
\includegraphics[width=0.6\textwidth]{figure-fig14.png}
|
|
\caption{\centering The energy corrections for HF at stretched geometry in the aug'-cc-pVDZ basis with the two-state model.}
|
|
\label{fig:my_label}
|
|
\end{figure}
|
|
|
|
\end{frame}
|
|
|
|
\begin{frame}{The Møller-Plesset Hamiltonian}
|
|
|
|
\begin{equation}
|
|
H(\lambda)=H_0 + \lambda (H_\text{phys} - H_0)
|
|
\end{equation}
|
|
|
|
\begin{equation}
|
|
H_\text{phys}=\sum\limits_{j=1}^{n}\left[ -\frac{1}{2}\grad_j^2 - \sum\limits_{k=1}^{N} \frac{Z_k}{|\vb{r}_j-\vb{R}_k|}+\sum\limits_{j<l}^{n}\frac{1}{|\vb{r}_j-\vb{r}_l|}\right]
|
|
\end{equation}
|
|
|
|
\begin{equation}
|
|
H_0=\sum\limits_{j=1}^{n}\left[ -\frac{1}{2}\grad_j^2 - \sum\limits_{k=1}^{N} \frac{Z_k}{|\vb{r}_j-\vb{R}_k|}+V_j^{(scf)}\right]
|
|
\end{equation}
|
|
|
|
\end{frame}
|
|
|
|
\begin{frame}{Existence of a critical point\footcite{stillinger_mollerplesset_2000}}
|
|
|
|
For $\lambda<0$:
|
|
|
|
\begin{equation*}
|
|
H(\lambda)=\sum\limits_{j=1}^{n}\left[ \underbrace{-\frac{1}{2}\grad_j^2 - \sum\limits_{k=1}^{N} \frac{Z_k}{|\vb{r}_j-\vb{R}_k|}}_{\text{Independant of }\lambda} + \overbrace{(1-\lambda)V_j^{(scf)}}^{\textcolor{red}{Repulsive}}+\underbrace{\lambda\sum\limits_{j<l}^{n}\frac{1}{|\vb{r}_j-\vb{r}_l|}}_{\textcolor{blue}{Attractive}} \right]
|
|
\end{equation*}
|
|
|
|
\end{frame}
|
|
|
|
\begin{frame}{Critical point in a finite basis set}
|
|
|
|
\pause[1]
|
|
|
|
\begin{beamerboxesrounded}[scheme=foncé]{\centering Exact energy $E(z)$}
|
|
$E(z)$ has a critical point on the negative real axis and $E(z)$ is continue for real values below $z_{crit}$.
|
|
\end{beamerboxesrounded}
|
|
|
|
\vspace{0.5cm}
|
|
|
|
\pause[2]
|
|
|
|
\begin{beamerboxesrounded}[scheme=foncé]{\centering In a finite basis set}
|
|
The singularities occur in complex conjugate pairs with non-zero imaginary parts and the energies are discrete.
|
|
\end{beamerboxesrounded}
|
|
|
|
\vspace{0.5cm}
|
|
|
|
\pause[3]
|
|
|
|
\centering \Large{How is this connected???}
|
|
|
|
\end{frame}
|
|
|
|
\begin{frame}{Singularities $\alpha$ and $\beta$ \footcite{sergeev_singularities_2006}}
|
|
|
|
\pause[1]
|
|
|
|
\begin{beamerboxesrounded}[scheme=foncé]{\centering Observation}
|
|
We can separate singularities in two parts.
|
|
\end{beamerboxesrounded}
|
|
|
|
\pause[2]
|
|
|
|
\begin{beamerboxesrounded}[scheme=foncé]{\centering Singularity $\alpha$}
|
|
\begin{itemize}
|
|
\item Large avoided crossing
|
|
\item Non-zero imaginary part
|
|
\item Interaction with a low-lying doubly excited states
|
|
\end{itemize}
|
|
\end{beamerboxesrounded}
|
|
|
|
\pause[3]
|
|
|
|
\begin{beamerboxesrounded}[scheme=foncé]{\centering Singularity $\beta$}
|
|
\begin{itemize}
|
|
\item Sharp avoided crossing
|
|
\item Very small imaginary part
|
|
\item Interaction with a diffuse function
|
|
\end{itemize}
|
|
\end{beamerboxesrounded}
|
|
|
|
\end{frame}
|
|
|
|
|
|
|
|
\begin{frame}{Modeling the critical point}
|
|
|
|
\pause[1]
|
|
|
|
\begin{beamerboxesrounded}[scheme=foncé]{\centering Stillinger}
|
|
\begin{quote}
|
|
\textit{"One might expect that $E_{FCI}(z) $ would try to model a continuum at $z_c$ with a grouping of discrete but closely spaced eigenstates that undergo sharp avoided crossing with the ground states."}
|
|
\end{quote}
|
|
\end{beamerboxesrounded}
|
|
|
|
\vspace{0.5cm}
|
|
|
|
\pause[2]
|
|
|
|
\begin{beamerboxesrounded}[scheme=foncé]{\centering Sergeev et al.}
|
|
Proof of the existence of this group of sharp avoided crossings for Ne, He and HF when the basis set contains diffuse functions.
|
|
\end{beamerboxesrounded}
|
|
|
|
\end{frame}
|
|
|
|
\section{Conclusion}
|
|
|
|
\begin{frame}{Conclusion}
|
|
|
|
\pause[1]
|
|
|
|
\begin{beamerboxesrounded}[scheme=foncé]{\centering Møller-Plesset perturbation theory}
|
|
By understanding how the singularities are localized in the complex plane we hope that it will gives us a deep understanding of the strengths and weaknesses of the Møller-Plesset method to get the correlation energy.
|
|
\end{beamerboxesrounded}
|
|
|
|
\vspace{0.5cm}
|
|
|
|
\pause[2]
|
|
|
|
\begin{beamerboxesrounded}[scheme=foncé]{\centering Spherium: a theoretical playground}
|
|
We will use the spherium model (two opposite-spin electrons restricted to remain on a surface of a sphere of radius $R$) to investigate the effects of symmetry breaking on singularities.
|
|
\end{beamerboxesrounded}
|
|
|
|
\end{frame}
|
|
|
|
\end{document} |