Mapping diverse hysteresis dynamics in scaled MoS$_2$ FETs using the universal method derived from TCAD modeling
Extracted source text · arXiv 2507.13002
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%% This is a (brief) model paper using the achemso class
%% The document class accepts keyval options, which should include
%% the target journal and optionally the manuscript type.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\documentclass[journal=jacsat,manuscript=article]{achemso}
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%% Place any additional packages needed here. Only include packages
%% which are essential, to avoid problems later. Do NOT use any
%% packages which require e-TeX (for example etoolbox): the e-TeX
%% extensions are not currently available on the ACS conversion
%% servers.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\usepackage[version=3]{mhchem} % Formula subscripts using \ce{}
\usepackage[]{mathrsfs}
\usepackage[]{listings}
\usepackage[latin1]{inputenc}
\usepackage{amssymb}
\usepackage{epsfig}
%\usepackage{wrapfig}
\usepackage{xspace}
\usepackage{pstricks}
\usepackage{psfrag}
\usepackage{textgreek}
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%% If issues arise when submitting your manuscript, you may want to
%% un-comment the next line. This provides information on the
%% version of every file you have used.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%\listfiles
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%% Place any additional macros here. Please use \newcommand* where
%% possible, and avoid layout-changing macros (which are not used
%% when typesetting).
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\input{./reliability/symsRel}
\newcommand*\mycommand[1]{\texttt{\emph{#1}}}
\newcommand{\figdir} {./figures/paper}
\newcommand{\VDirac} {V_\mathrm{Dirac}}
\newcommand{\IDirac} {I_\mathrm{Dirac}}
\newcommand{\dVDirac} {\Delta V_\mathrm{Dirac}}
\newcommand{\dIDirac} {\Delta I_\mathrm{Dirac}}
\newcommand*{\citen}[1]{%
\begingroup
\romannumeral-`\x % remove space at the beginning of \setcitestyle
\setcitestyle{numbers}%
\cite{#1}%
\endgroup
}
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%% Meta-data block
%% ---------------
%% Each author should be given as a separate \author command.
%%
%% Corresponding authors should have an e-mail given after the author
%% name as an \email command. Phone and fax numbers can be given
%% using \phone and \fax, respectively; this information is optional.
%%
%% The affiliation of authors is given after the authors; each
%% \affiliation command applies to all preceding authors not already
%% assigned an affiliation.
%%
%% The affiliation takes an option argument for the short name. This
%% will typically be something like "University of Somewhere".
%%
%% The \altaffiliation macro should be used for new address, etc.
%% On the other hand, \alsoaffiliation is used on a per author basis
%% when authors are associated with multiple institutions.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\author{\vspace*{-0.1cm}\it Y.Z. Lv}
\author{\vspace*{-0.1cm}\it H. Cai}
\author{\vspace*{-0.1cm}\it Y. Wu}
\author{\vspace*{-0.1cm}\it Yu.Yu. Illarionov}
\affiliation{Laboratory of 2D Optoelectronics and Nanoelectronics (L2DON), MSE Department, Southern University of Science and Technology, 1088 Xueyuan Blvd, 518055 Shenzhen, China}
\email{illarionov@sustech.edu.cn}
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%% The document title should be given as usual. Some journals require
%% a running title from the author: this should be supplied as an
%% optional argument to \title.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\title[An \textsf{achemso} demo]
{Mapping diverse hysteresis dynamics in scaled MoS$_2$ FETs using the universal method derived from TCAD modeling}
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%% Some journals require a list of abbreviations or keywords to be
%% supplied. These should be set up here, and will be printed after
%% the title and author information, if needed.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%\keywords{MoS$_2$, CaF$_2$, transistor, insulator scaling, MBE, CVD}
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%% The manuscript does not need to include \maketitle, which is
%% executed automatically.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\begin{document}
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%% The "tocentry" environment can be used to create an entry for the
%% graphical table of contents. It is given here as some journals
%% require that it is printed as part of the abstract page. It will
%% be automatically moved as appropriate.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%% The abstract environment will automatically gobble the contents
%% if an abstract is not used by the target journal.
%Field-effect transistors (FETs) based on 2D materials are now considered as a possible alternative or supplement to Si technologies. However, they may suffer from locally increased defects due to not properly adjusted processing. In some positions defects may cause a localized hysteresis of both signs. Here, we introduce the universal hysteresis mapping method to properly benchmark this behavior. Then we also consider the reliability of 2D FETs with fundamental defect bands which may add more understanding to the experimental performances and improve the 2D device application in the future.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\begin{abstract}
Field-effect transistors (FETs) based on 2D materials have already reached the stage of trial FAB integration. However, reliability limitations caused by various defects present a serious obstacle for their smooth way forward, especially when scaling the device geometries. Still the ongoing research is mostly focused on pure technology aspects, while reliability is often recalled only when showing a randomly measured gate transfer curve to manifest that the hysteresis is ``negligible''. In fact the hysteresis dynamics contain unique fingerprints of various mechanisms which may coexist or cancel each other, being more complex in scaled FETs, for instance because of simultaneous interaction of defects with the channel and top gate in thin insulators. To fill this gap, here by doing TCAD modeling for nanoscale MoS$_2$/HfO$_2$ FETs we introduce the universal hysteresis mapping method which can correctly capture commonly measured diverse hysteresis dynamics such as conventional clockwise (CW) and counterclockwise (CCW) hysteresis, as well as CW/CCW switching and time separation. Next we extend this method to bias-temperature instabilities (BTI) and show a clear correlation between complex hysteresis dynamics and abnormal BTI recovery. Finally, we validate our mapping method using available experimental data for MoS$_2$ FETs and demonstrate that it provides far more accurate results than a conventional constant current extraction of the hysteresis width, being also usable if a CCW hysteresis is caused by mobile ions.
\end{abstract}
%\textbf{Keywords:} Black phosphorus, MoS$_2$, graphene, transistor, 2D electronics, reliability improvement
%Fabrication of different electronic devices with atomically thin two-dimensional (2D) layers has achieved a considerable progress in the past few years. In particular, these emerging technologies have been
%suggested for applications in flexible electronics~\cite{LEE13,AKINWANDE14}, optoelectronic devices~\cite{AVOURIS14,BABLICH16} and sensors~\cite{SHAVANOVA16}. Also,
%Graphical abstract
%\begin{figure}[!h]
%\vspace{0mm}
%\begin{minipage} {\textwidth} %{\sminipagewidth}
%\hspace{2cm}
% \includegraphics[width=12cm]{\figdir/graa.eps} %???
%\end{minipage}
%\end{figure}
%\begin{figure}[!h]
%\vspace{0mm}
%\begin{minipage} {\textwidth} %{\sminipagewidth}
%\hspace{4.5cm}
% \includegraphics[width=9cm]{\figdir/graa.eps} %???
%\end{minipage}
%\end{figure}
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%% Start the main part of the manuscript here.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\section{Introduction}
Recent research advances in FETs with atomically thin 2D channels~\cite{WACHTER17,LEMME22,HUANG22,TAN23,PENDURTHI24} have made it possible to start with their trial FAB integration~\cite{ASSELBERGHS20,DOROW22,CHUNG24,MORTELMANS24}, while considering these new device technologies as a potential option to continue scaling of modern electronics following the Moore's law~\cite{SCHWIERZ15,DAS21,LEMME22}. However, the way of 2D FETs to the market is delayed by severe reliability limitations caused by preexisting and process-induced defects in gate insulators~\cite{ILLARIONOV20A} which still make it hard to compete with Si technologies in stable long-term operation. The material combination MoS$_2$/high-k oxide~\cite{JAYACHANDRAN24,CHUNG24,MORTELMANS24} which is now used as the core for 2D n-FETs is unfortunately not an exception since the upper defect bands of HfO$_2$ and Al$_2$O$_3$ are rather close to the conduction band edge of MoS$_2$~\cite{ILLARIONOV20A,ILLARIONOV24A}, not to mention possible process-induced defects~\cite{ILLARIONOV24A}. Furthermore, the reliability problem significantly complicates when the device geometries are scaled down~\cite{KACZER10A,GRASSER12A,MARKOV13}. This is because in nanoscale devices the electrostatic impact of every particular defect is larger, and thus even a relatively small number of defects or few localized defect clusters still present in mature 2D FETs could result in serious perturbations of stable device operation. As the real potential of 2D FETs could be fully exploited only for sub-10$\,$nm channel lengths~\cite{IRDS21}, further progress in 2D devices requires proper understanding of new physical mechanisms of reliability limitations which would accompany aggressive scaling. This can be achieved only if certain standard metrics for the reliability are used in the future analysis, while being easily benchmarkable using universal experimental techniques.
The commonly observed hysteresis of the gate transfer characteristics can be considered as an important metric for the reliability analysis as it is directly related to the gate bias stability. Still, the absolute majority of previous studies are restricted to fragmentary observation of hysteresis at a random sweep rate and often speculate that it is ``negligible'', ``near-zero'' or ``small''~\cite{ROH16,VU18,CHO21E,VENKA24,VENKA24,FAN25,LAN25} while substituting real physical analysis by generation of random ill-defined numbers for the hysteresis width. In reality, the hysteresis in 2D FETs can have multiple contributions which may cancel each other~\cite{PROVIAS23} and become more crucial in scaled devices, thus requiring a complex analysis. Though being disregarded in most studies, the information obtained from comprehensive hysteresis measurements could be of key importance for understanding the physical mechanisms such as simultaneous interaction of charge traps with the channel and top gate which perturb stable operation of 2D FETs. Furthermore, complementing correctly extracted hysteresis data by the results measured for bias temperature instabilities (BTI)~\cite{GRASSER11A} on the same device could significantly strengthen the reliability analysis.
Here we first perform a comprehensive TCAD modeling of hysteresis dynamics in scaled MoS$_2$/HfO$_2$ FETs and explain the changes observed with varied distance of a trap distribution from the MoS$_2$/HfO$_2$ interface. Based on these findings we introduce a universal hysteresis mapping method which can capture conventional clockwise (CW) and counterclockwise (CCW) hysteresis, as well as CW/CCW switching and time separation. Next we simulate positive BTI (PBTI) for the same trap distributions and demonstrate that complex hysteresis dynamics such as switching come together with abnormal PBTI recovery, while extending our mapping method towards the BTI analysis. Finally, we apply our universal mapping method to available experimental datasets including those for imec FAB MoS$_2$ FETs with scaled top gates~\cite{ILLARIONOV24A} and verify its functionality to capture experimentally observed hysteresis dynamics also for the case when the drift of mobile ions in oxide is involved in addition to charge trapping.
\section{TCAD setup and modeling approach}
We use a well-established TCAD simulator GTS Minimos-NT~\cite{MINIMOSNT22} which describes carrier transport in the channel with the drift-diffusion model and employs the four-state non-radiative multiphonon (NMP) model~\cite{GRASSER12A,ALKAUSKAS14} for charge trapping by oxide defects. This professional device modeling tool originally developed for Si technologies has been successfully adapted to FETs with 2D channels and used in many previous works~\cite{ILLARIONOV16B,KNOBLOCH18,KNOBLOCH23}.
\begin{figure}[!h]
\centering
\vspace{0mm}
\begin{minipage} {\textwidth} %{\sminipagewidth}
\hspace{0.2cm}
\centering
\includegraphics[width=9cm]{\figdir/Fig.1.eps} %???
\caption{\label{Fig.1} (a) Schematic layout of MoS$_2$ FETs used in our TCAD simulations with trap distributions D1, D2 and D3 centered at $y_{\mathrm{T}}\,=\,0.25\,d_{\mathrm{ox}}$, 0.5$\,d_{\mathrm{ox}}$ and 0.75$\,d_{\mathrm{ox}}$ from the MoS$_2$/HfO$_2$ interface, respectively. (b) The band diagram of the channel cross-section showing the key simulation parameters in a flat band condition. We consider three defect bands: acceptor-type at $E_{\mathrm{T}}\,=\,1.2\,\pm\,$0.2$\,$eV (upper fundamental), donor-type at $E_{\mathrm{T}}\,=\,-0.1\pm\,$0.2$\,$eV (lower fundamental) and acceptor-type at $E_{\mathrm{T}}\,=\,-0.3\pm\,$0.2$\,$eV (lower artificial). (c) Schematics of top gate voltage sweeps with logarithmically increased sweep times used in our transient NMP-based simulation setup. (d) Fragmentary schematics of subsequent PBTI stress/recovery rounds with logarithmically increased stress times implemented into our TCAD tool. In reality, $\sim$10 intermediate recovery sweeps and up to 8 rounds with different $t_{\mathrm{s}}$ are used. A stabilizing stage at $V_{\mathrm{TGR}}$ is always used before the first fresh sweep. }
\end{minipage}
\end{figure}
In line with modern scaling requirements~\cite{IRDS21}, we consider top-gated single-layer MoS$_2$ FETs ($L\,=\,10\,$nm, $W\,=1\,\mu$m) with HfO$_2$ insulators of $\,d_{\mathrm{ox}}$ =5.1 nm, i.e. the equivalent oxide thickness (EOT) of 1 nm. In order to reveal the impact of trap depth on the charge trapping in a scaled oxide, we set three oxide trap distributions D1, D2, D3 of $\sim$2000 traps centered at $y_{\mathrm{T}}\,=\,0.25\,d_{\mathrm{ox}}$, 0.5$\,d_{\mathrm{ox}}$ and 0.75$\,d_{\mathrm{ox}}$ from the MoS$_2$/HfO$_2$ interface, respectively. The schematic layout of the MoS$_2$ FETs with the positions of trap distributions in HfO$_2$ is shown in Fig.1a. The corresponding band diagram with barrier parameters taken from the literature~\cite{RASMUSSEN15} is shown in Fig.1b, with zero energy being in the middle of MoS$_2$ bandgap. It is well-known that HfO$_2$ has two fundamental defect bands~\cite{ILLARIONOV20A}, namely the upper acceptor-type (i.e. the charge -1 below $E_{\mathrm{F}}$ and 0 above $E_{\mathrm{F}}$) centered at $E_{\mathrm{T}}\,=\,1.2\,$eV and the lower donor-type (i.e. the charge +1 above $E_{\mathrm{F}}$ and 0 below $E_{\mathrm{F}}$) centered at $E_{\mathrm{T}}\,=\,-0.1\,$eV. However, since here we are targeting to get as much complex hysteresis features as possible, we also consider an artificial lower acceptor-type defect band at $E_{\mathrm{T}}\,=\,-0.3\,$eV which could express, for instance, additional process-induced defects in HfO$_2$. For all defect bands we consider the same Gaussian half-width of 0.2$\,$eV.
The most essential ingredient of our TCAD approach is the time-dependent analysis of charge trapping which is performed using the NMP model parameters~\cite{GRASSER12A} provided in Fig.S1 in the Supplementary Information (SI). While implementing real time as a variable~\cite{KNOBLOCH18}, we first obtain the hysteresis dynamics by simulating the forward/reverse $I_{\mathrm{D}}$-$V_{\mathrm{TG}}$ curves with logarithmically spaced sweep times $t_{\mathrm{sw}}$ from 10$^{-7}\,$s to 10$^{8}\,$s, thereby going far beyond the typical measurement ranges. The schematics of these $V_{\mathrm{TG}}$ sweeps versus time is shown in Fig.1c. Next we simulate the typically used BTI measure-stress-measure loops~\cite{PROVIAS23}, as schematically shown in Fig.1d. First the device is subjected to a relaxation voltage $V_{\mathrm{TGR}}$, followed by the fast sweep of the fresh $I_{\mathrm{D}}$-$V_{\mathrm{TG}}$ curve. Then a stress voltage $V_{\mathrm{TGS}}$ is applied for a certain stress time $t_{\mathrm{s}}$, and after this the recovery is monitored using intermediate $I_{\mathrm{D}}$-$V_{\mathrm{TG}}$ sweeps while placing the device back to $V_{\mathrm{TGR}}$ in between these sweeps. For the convenience, the fresh and control sweeps are performed between $V_{\mathrm{TGR}}$ and $V_{\mathrm{TGS}}$ which also gives better convergence. Just like in real BTI experiments, subsequent stress/recovery rounds with logarithmically increased $t_{\mathrm{s}}$ were implemented in our modeling setup.
\section{Universal mapping method: concept and verification}
In our previous studies the hysteresis width $\Delta V_{\mathrm{H}}$ was extracted using a constant current method near the threshold voltage $V_{\mathrm{th}}$ and plotted against the reciprocal sweep time 1/$t_{\mathrm{sw}}$~\cite{ILLARIONOV16A}. However, recent experimental results~\cite{PROVIAS23,KNOBLOCH23,ILLARIONOV24A} suggest that hysteresis dynamics may be more complex due to simultaneous contribution of multiple mechanisms. In line with these findings, we first consider the lower artificial defect band in HfO$_2$ and in Fig.2a-c show the $I_{\mathrm{D}}$-$V_{\mathrm{TG}}$ curves simulated for our nanoscale MoS$_2$ FETs with representative features such as pure CW hysteresis for the trap distribution D1 (Fig.2a), CW/CCW hysteresis switching for D2 (Fig.2b) and pure CCW hysteresis for D3 (Fig.2c). More $I_{\mathrm{D}}$-$V_{\mathrm{TG}}$ curves simulated for the slow and fast sweeps can be found in Fig.S2 in the SI. These results suggest that the hysteresis width $\Delta V_{\mathrm{H}}$ may depend not only on the sweep time $t_{\mathrm{sw}}$, but also on the constant current $I_{\mathrm{Dconst}}$ used for extraction. Therefore, the value of $\Delta V_{\mathrm{H}}$ obtained at a randomly selected $I_{\mathrm{Dconst}}$ may be ill-defined, for instance for the case of Fig.2b which has a switching point with zero hysteresis separating CW and CCW branches.
\begin{figure}[!h]
\centering
\vspace{0mm}
\begin{minipage} {\textwidth} %{\sminipagewidth}
\hspace{0.2cm}
\centering
\includegraphics[width=14cm]{\figdir/Fig.2.eps} %???
\caption{\label{Fig.2} The $I_{\mathrm{D}}$-$V_{\mathrm{TG}}$ curves ($V_{\mathrm{D}}\,=\,0.05\,$V, $V_{\mathrm{TG}}$ from -0.5$\,$V to 3$\,$V) simulated for MoS$_2$ FETs with lower artificial defect band in HfO$_2$ which show: (a) CW hysteresis at slow $t_{\mathrm{sw}}$ with the trap distribution D1; (b) switching between CW and CCW hysteresis at slow $t_{\mathrm{sw}}$ with the trap distribution D2; (c) CCW hysteresis at fast $t_{\mathrm{sw}}$ with the trap distribution D3. (d) Schematic illustration of our universal hysteresis mapping for the $I_{\mathrm{D}}$-$V_{\mathrm{TG}}$ curve with localized hysteresis branches of different signs.
(e) The $\Delta V_{\mathrm{H}}$(1/$t_{\mathrm{sw}}$) dependences concluded in between upper and lower UHFs extracted from simulated $I_{\mathrm{D}}$-$V_{\mathrm{TG}}$ curves. For D1 defects only CW hysteresis is present. Mid-HfO$_2$ defects from D2 may result in a switching between CW and CCW hysteresis. D3 defects cause pure CCW hysteresis. (f) Bias dependences of the capture and emission times of the traps which are active near the hysteresis maximum as extracted from TCAD. Proximity to the top gate makes $\tau_{\mathrm{c}}$ increasing vs. $V_{\mathrm{TG}}$. (g) Variation of the charge state for these active traps during forward and reverse sweeps with the mean active charge state shown with thick lines. As it is marked in (f), those of D1 traps which are neutral can capture an electron from MoS$_2$ during forward sweep but cannot emit it during reverse sweep, thereby causing CW hysteresis. For D2 traps the CW/CCW hysteresis switching appears due to complex interaction with both MoS$_2$ and top gate. Finally, charged D3 traps emit electrons to the top gate which causes pure CCW hysteresis.}
\end{minipage}
\end{figure}
To address these limitations, here we introduce a more systematic universal hysteresis mapping method which suggests scanning of $\Delta V_{\mathrm{H}}$ from the current value slightly above $I_{\mathrm{OFF}}$ to near-$I_{\mathrm{ON}}$ values as schematically shown in Fig.2d. The exact mapping range has to be adjusted depending on the presence of obvious artifacts in a particular set of the gate transfer curves simulated or measured for different $t_{\mathrm{sw}}$, for instance noise at low currents. Using a very fine step on $I_{\mathrm{Dconst}}$, we can obtain a serie with thousands of $\Delta V_{\mathrm{H}}$(1/$t_{\mathrm{sw}}$) rather than just one. Then next we can express the hysteresis dynamics as a distribution of these curves concluded in between the upper and lower universal hysteresis functions (UHFs) constructed as a piecewise maximum and minimum from the obtained set of $\Delta V_{\mathrm{H}}$(1/$t_{\mathrm{sw}}$) dependences, respectively. Below we will show that this is the only accurate way to capture complex hysteresis dynamics such as CW/CCW switching.
In Fig.2e we show the hysteresis mapping results for the trap distributions D1, D2 and D3 extracted from the simulated $I_{\mathrm{D}}$-$V_{\mathrm{TG}}$ curves ($t_{\mathrm{sw}}$ from 10$^{-7}$ to 10$^{8}$ s) by using 3000 $I_{\mathrm{Dconst}}$ steps. For D1 a pure clockwise hysteresis is observed for relatively slow sweeps (Fig.2e, top), which is consistent with the capture and emission time constants ($\tau_{\mathrm{c}}$ and $\tau_{\mathrm{e}}$, respectively) of active traps extracted from our TCAD simulations (Fig.2f, top). Namely, when the sweep time approaches $\tau_{\mathrm{c}}$, active traps capture an electron from the MoS$_2$ channel and change their charge state from neutral (0) to negative (-1) during the forward sweep (Fig.2e, top). Then since $\tau_{\mathrm{e}}$ is large and barely dependent on $V_{\mathrm{TG}}$, these traps remain charged during the reverse sweep. The resulting negative charge difference between forward and reverse sweeps explains the CW hysteresis which we observe for D1.
For mid-HfO$_2$ traps in the D2 distribution, the mapping results show a CW/CCW hysteresis switching at slow $t_{\mathrm{sw}}$ (Fig.2e, center), with CW and CCW contributions being expressed by the upper and lower UHFs, respectively. The capture time $\tau_{\mathrm{c}}$ linearly increases with $V_{\mathrm{TG}}$ between -0.5$\,$V and 2$\,$V due to the impact of top gate, and then tends to decrease between 2$\,$V and 3$\,$V (Fig.2f, center) which is similar to D1 and related to the impact of MoS$_2$ channel. As a result, we can see that $\tau_{\mathrm{c}}$ and $\tau_{\mathrm{e}}$ bias dependences have two obvious intersection points which results in a more complicated charge trapping dynamics as compared to D1. Namely, active are those traps which are initially in the negative charge state (i.e. below $E_{\mathrm{F}}$). During the forward sweep as soon as $t_{\mathrm{sw}}$ becomes close to $\tau_{\mathrm{e}}$, these traps start to emit an electron to the top gate, thereby approaching neutral state (Fig.2g, center). Then because of the decay of $\tau_{\mathrm{c}}$ in the end of the forward sweep they can again capture an electron from the channel, and in the beginning of reverse sweep emit it to the top gate while remaining neutral until the OFF state in which capture of electron from the top gate becomes possible, though already with no impact on the hysteresis. As a result, the charge difference between the forward and reverse sweeps changes from positive near $V_{\mathrm{th}}$ to negative in ON state, which explains the observed CW/CCW switching of hysteresis as the same traps can simultaneously exchange charges with the MoS$_2$ channel (CW mechanism) and the top gate (CCW mechanism). However, for a $t_{\mathrm{sw}}$ range which lies below the two intersections of $\tau_{\mathrm{c}}$ and $\tau_{\mathrm{e}}$ curves but above the minimum of $\tau_{\mathrm{c}}$ at $V_{\mathrm{TG}}\,=\,3\,$V pure CW hysteresis is present just like for D1 as no interaction with the top gate is possible.
Finally, for D3 we observe a moderately fast pure CCW hysteresis represented by lower UHF (Fig.2e, bottom). Since these traps are far from the MoS$_2$ channel, they can interact only with the top gate and thus $\tau_{\mathrm{c}}$ monotonically increases vs. $V_{\mathrm{TG}}$, with both time constants being smaller as compared to D1 and D2 cases. With the sweep times around the hysteresis maximum at $t_{\mathrm{sw}}$= 10$^{-2}\,$s which is close to $\tau_{\mathrm{e}}$, the charge state of active traps changes from negative to neutral during the forward sweep due to emission of electrons to the top gate, and then remains neutral during the reverse sweep as no capture is possible until the OFF state. As a result, positive charge difference between forward and reverse sweeps results in a CCW hysteresis. Obviously, these traps will not be active if $t_{\mathrm{sw}}$ is too large or to small, which explains why the hysteresis has a maximum which more or less matches the intersection point of time constant vs. $V_{\mathrm{TG}}$ curves. Furthermore, based on the analysis of the three cases we clearly see that intersection of rising $\tau_{\mathrm{c}}$($V_{\mathrm{TG}}$) with $\tau_{\mathrm{e}}$($V_{\mathrm{TG}}$) grounds the CCW hysteresis and intersection of decaying $\tau_{\mathrm{c}}$($V_{\mathrm{TG}}$) with $\tau_{\mathrm{e}}$($V_{\mathrm{TG}}$) is related to the CW hysteresis. We also note that when applying the suggested universal mapping method it is necessary to pay attention to possible presence of numerical artifacts when analyzing TCAD simulated data (e.g. Fig.2e, bottom), or to the noise features in the experimental data.
\begin{figure}[!h]
\centering
\vspace{0mm}
\begin{minipage} {\textwidth} %{\sminipagewidth}
\hspace{0.2cm}
\centering
\includegraphics[width=16cm]{\figdir/Fig.3.eps} %???
\caption{\label{Fig.3}The band diagrams of MoS$_2$ FETs extracted from our TCAD simulations for the forward sweep $V_{\mathrm{TG}}\,=\,2.5\,$V, $V_{\mathrm{TG}}\,=\,3\,$V and reverse sweep $V_{\mathrm{TG}}\,=\,2.5\,$V showing how the NMP trap charge state in the lower artificial defect band changes during the sweep. (a,b,c) For the distribution D1 which is closer to the MoS$_2$/HfO$_2$ interface, electron trapping from the channel results in a more negative charge state of defects during reverse sweep, which causes CW hysteresis. (d,e,f) For the mid-HfO$_2$ distribution D2, electron trapping from the channel can be followed by electron emission to the top gate, which explains the interplay between CW and CCW hysteresis. (g,h,i) The defects from distribution D3 are far from the channel, which makes electron trapping from MoS$_2$ unlikely. Thus, electron emission to the top gate makes the charge state during reverse sweep more positive and results in CCW hysteresis.}
\end{minipage}
\end{figure}
The band diagrams with all defects of the lower artificial defect band shown in Fig.3 further support our findings from Fig.2. Namely, the pure CW hysteresis is due to electron trapping from MoS$_2$ channel which saturates when all defects become negatively charged as the defect band is completely below $E_{\mathrm{F}}$ of the channel, thereby leading to a maximum at slow $t_{\mathrm{sw}}$. However, for D2 case defects can not only capture electrons from MoS$_2$ but also emit them to the top gate which has $E_{\mathrm{F}}$ crossing the lower part of the defect band which results in a CW/CCW hysteresis switching due to this complex charge trapping dynamics shown in Fig.2g, center. More band diagrams for this most complicated case can be found in Fig.S3 in the SI. For D3 case, as the traps are far from the channel, electron capture from MoS$_2$ becomes unlikely. However, since the defect band is mostly above the top gate $E_{\mathrm{F}}$ and the distance to the top gate is small, electron emission is possible which leads to the CCW hysteresis.
\begin{figure}[!h]
\centering
\vspace{0mm}
\begin{minipage} {\textwidth} %{\sminipagewidth}
\hspace{0.1cm}
\centering
\includegraphics[width=16.3cm]{\figdir/Fig.4.eps} %???
\caption{\label{Fig.4}$I_{\mathrm{D}}$-$V_{\mathrm{TG}}$ curves simulated for slow (a) and fast (b) sweeps assuming that both D1 and D3 are present in the oxide. (c) Mapping results show time-separated CW and CCW hysteresis on UHFs. The corresponding forward/reverse sweep band diagrams extracted from TCAD for slow (d,e) and fast (g,h) sweeps. (f) $\Delta V_{\mathrm{H}}$(1/$t_{\mathrm{sw}}$) curves calculated using the NMP charge values extracted from our TCAD simulations. D1+D3 curve is very smilar to the one obtained using a constant current method around $V_{\mathrm{TG}}\,=\,2\,$V (marked in Fig4(c)).}
\end{minipage}
\end{figure}
In Fig.4 we show that for $\sim$2000 traps equally split between D1 and D3 the CW hysteresis is present at slow sweeps (Fig.4a) and the CCW hysteresis is dominant at faster sweeps (Fig.4b). The mapping results (Fig.4c) clearly reveal time separation of these two phenomena, both having a distinct maximum in the corresponding UHF, i.e. upper for CW and lower for CCW. This case is indeed an excellent example which shows that near-zero hysteresis measured at a randomly selected single sweep time is meaningless for reliability analysis. Then we analyze the impact of NMP charges extracted from TCAD at $t_{\mathrm{sw}}$ from both peaks. Indeed, at slow $t_{\mathrm{sw}}$ the charge difference between reverse ($Q_{\mathrm{R}}$) and forward ($Q_{\mathrm{F}}$) sweeps is negative due to the dominant electron trapping from MoS$_2$ by D1 traps (Fig.4d,g). In contrast, at fast $t_{\mathrm{sw}}$ a positive $\Delta Q$ is created by dominant emission of electrons to the top gate by D3 traps (Fig.4e,h), even though their contributions to $Q_{\mathrm{R}}$ and $Q_{\mathrm{F}}$ are $\sim$3 times smaller due to (1 - $y_{\mathrm{T}}$/$d_{\mathrm{ox}}$) factor. Remarkably, the $\Delta V_{\mathrm{H}}$(1/$t_{\mathrm{sw}}$) trace calculated as -$\Delta Q$/$C_{\mathrm{ox}}$ is similar to the one extracted from the $I_{\mathrm{D}}$-$V_{\mathrm{TG}}$ curves around $V_{\mathrm{TG}}\,=\,2\,$V (Fig.4f), which justifies the validity of our approach. We note that a minor difference is because our TCAD setup considers charges in a self-consistent manner rather than a simple summing up which we perform when calculating $\Delta Q$. Finally, we have repeated these simulations assuming 450$\,$K temperature and demonstrated that both CW and CCW maxima shift towards faster sweep rates, which is consistent with the thermal activation of charge trapping~\cite{ILLARIONOV16A} (see Fig.S4 in the SI).
\begin{figure}[!h]
\centering
\vspace{0mm}
\begin{minipage} {\textwidth} %{\sminipagewidth}
\hspace{0.02cm}
\centering
\includegraphics[width=16.5cm]{\figdir/Fig.5.eps} %???
\caption{\label{Fig.5} (a) $I_{\mathrm{D}}$-$V_{\mathrm{TG}}$ characteristics simulated by TCAD with D1+D3 trap distributions assuming the initially stabilized condition (i.e. fresh), as well as the device after PBTI stress with $t_{\mathrm{s}}\,=\,100\,$ks, intermediate recovery points (thinner lines) and in the end of the recovery round ($t_{\mathrm{r}}\,=\,3\times 10^8\,$s). (b) The recovery traces extracted from the results simulated with six log-spaced $t_{\mathrm{s}}$ using the mapping method and constant current approach. Abnormal dynamics is due to contributions from fast charge exchange with the top gate (D3 traps) and slow charge exchange with the MoS$_2$ channel. (c) Stress time dependences of the recoverable and permanent components of PBTI as extracted from the first and last points of the traces from (b), the mapping method results are used.}
\end{minipage}
\end{figure}
In Fig.5 we show that complex hysteresis dynamics from Fig.4 is consistent with PBTI dynamics simulated using the same D1+D3 trap distributions while assuming the subsequent stress/recovery rounds (Fig.1d) with $V_{\mathrm{TGS}}\,=\,3\,$V and $V_{\mathrm{TGR}}\,=\,0\,$V. Already from the typical $I_{\mathrm{D}}$-$V_{\mathrm{TG}}$ curves (Fig.5a) of a representative stress/recovery round it is clear that PBTI drift is initially negative (i.e. NBTI-like) and becomes positive only after a long recovery. The corresponding recovery traces obtained with six log-spaced stress times $t_{\mathrm{s}}$ indeed show abnormal recovery dynamics which is not common for a usual PBTI degradation (Fig.5b). Namely, the fast component of PBTI is related to strongly recoverable negative $\Delta V_{\mathrm{th}}$. This is due to electron emission to the top gate by D3 defects and capture back during recovery, being consistent with fast CCW hysteresis in Fig.4. However, slow D1 defects simultaneously capture electrons from the MoS$_2$ channel with no emission back during recovery, and thus the permanent component is still PBTI-like which agrees with slow CW hysteresis in Fig.4. At the same time, in Fig.5c we show that both NBTI-like recoverable and PBTI-like permanent components of the degradation saturate vs. $t_{\mathrm{s}}$, which is also clear from Fig.5b. This is because for $t_{\mathrm{s}}$ above 1$\,$ks most active traps change their charge states, which is consistent with the hysteresis maxima in Fig.4c.
We also note that just like the hysteresis width $\Delta V_{\mathrm{H}}$, the BTI shift $\Delta V_{\mathrm{th}}$ may be sensitive to the constant current point at which it is extracted. Therefore, the mapping method can be introduced for BTI in a similar way to the hysteresis, see Fig.S5 and the related discussion in the SI. While here we simulate PBTI loops with $\sim$20$\mu$s measurement delay caused by control sweeps from 0 to 3$\,$V, the recoverable component is larger if extracted closer to the subthreshold region due to increased contribution of faster traps. However, in real measurements much narrower sweep ranges just around $V_{\mathrm{th}}$ are used if the degradation is expected to be small, to reduce the delay which is still as large as hundreds of milliseconds at best. Thus we suggest that in most cases the extracted $\Delta V_{\mathrm{th}}$ should not depend on $I_{\mathrm{Dconst}}$ that much. Still in this work we show PBTI traces extracted using both approaches. For instance, it may be important to understand how the trace was extracted when trying to fit the experimental results using TCAD tools as in some cases a good fit may be obtained by lucky combination of wrong parameters.
\section{Mapping the hysteresis and PBTI caused by fundamental HfO$_2$ defect bands}
After verifying the validity of our universal mapping method with the lower artificial defect band, we next use it for the analysis of hysteresis caused by the fundamental defect bands known for HfO$_2$~\cite{ILLARIONOV20A} while considering the same trap distributions D1, D2 and D3 (Fig.6a). The upper and lower UHFs shown in Fig.6b, top and the corresponding PBTI recovery traces for $t_{\mathrm{s}}\,=\,100\,$ks in Fig.6c, top suggest that the upper acceptor-type defect band can cause only CW hysteresis and conventional PBTI degradation. Since D1 traps are closer to the MoS$_2$/HfO$_2$ interface and thus faster, the corresponding hysteresis maximum is at faster sweep times and PBTI is more recoverable as compared to D2. We also note that the magnitude of $\Delta V_{\mathrm{th}}$ caused by PBTI is directly correlated with the fraction of hysteresis maximum which is concluded between 1/$t_{\mathrm{sw}}$ points corresponding to $t_{\mathrm{s}}\,=\,100\,$ks and delay time $t_{\mathrm{d}}\,=\,20\,\mu$s as marked in Fig.6b. This explains why D2 traps cause larger PBTI degradation as compared to D1 traps. Furthermore, the distances between PBTI traces will start saturating vs. $t_{\mathrm{s}}$ corresponding to the position of the hysteresis maximum (e.g. Fig.5b,c). At the same time, the traps from D3 cannot interact neither with the channel being far from the MoS$_2$/HfO$_2$ interface nor with the top gate being too much above the top gate Fermi level. As a result, there is no hysteresis and PBTI induced by D3 traps.
As for the lower donor-type defect band, we can see that being far from the conduction band of MoS$_2$, it has overall smaller contribution to both hysteresis (Fig.6b, bottom) and PBTI (Fig.6c, bottom). However, while D1 traps cause pure CW hysteresis at slow sweeps, the traps from D2 and D3 distributions can also interact with the top gate thus causing time separation of the CW/CCW hysteresis (D2) or pure CCW hysteresis (D3). Remarkably, in the latter case a purely NBTI-like PBTI degradation is observed. These findings suggest that the lower defect band may be still crucial in MoS$_2$/HfO$_2$ FETs if the defect densities are locally increased near the top gate~\cite{ILLARIONOV24A}, though being negligible as compared to the upper defect band for the channel-side defects. More detailed TCAD results for fundamental defect bands such as full hysteresis mapping with bias dependences of time constants and charged states (similar to Fig.2), band diagrams and PBTI $I_{\mathrm{D}}$-$V_{\mathrm{TG}}$ curves and traces for six $t_{\mathrm{s}}$ can be found in Fig.S6-8 and Fig.S9-11 for the upper and lower fundamental defect bands, respectively. Note that for the lower fundamental defect band interaction with the top gate is possible just like for the lower artificial band because they are both close to the top gate Fermi level, while the upper fundamental defect band is energetically far above and thus can interact only with the channel.
\begin{figure}[!h]
\centering
\vspace{0mm}
\begin{minipage} {\textwidth} %{\sminipagewidth}
\hspace{0.02cm}
\centering
\includegraphics[width=16.5cm]{\figdir/Fig.6.eps} %???
\caption{\label{Fig.6} (a) Schematic band diagram showing our three trap distribution within the upper and lower fundamental defect bands in HfO$_2$. The typical features of the hysteresis and BTI dynamics are noted for each case. (b) The upper and lower UHFs obtained from TCAD simulations with both defect bands. (c) The corresponding PBTI recovery traces extracted using mapping method for a representative stress time $t_{\mathrm{s}}\,=\,100\,$ks which are obviously correlated with hysteresis dynamics. The upper defect band has a considerably larger contribution because it is closer to the conduction band of MoS$_2$.}
\end{minipage}
\end{figure}
\section{Experimental validation}
We finally verify our universal hysteresis mapping method by revisiting previously published experimental results for different 2D devices. In Fig.7 we analyze the dataset obtained for the top-gated MoS$_2$/Al$_2$O$_3$ FETs in~\cite{IllARIONOV17A} which show a large CW hysteresis (Fig.7a). When applying the method to our TCAD results, we have noticed that for oxides with relatively broad defect bands the CW hysteresis width behaves as schematically shown in Fig.7b (see Fig.S12 in the SI for more details). Namely, even if a broad maximum is observed near $V_{\mathrm{th}}$, a much narrower though smaller maximum is present closer to the ON state which perfectly matches the corresponding mapping results shown in Fig.7c. There we can see that a clear maximum of $\Delta V_{\mathrm{H}}$ extracted at higher currents gets broader when moving down to $V_{\mathrm{th}}$ and finally turns to just a slight increase of the hysteresis for slow sweeps. This behaviour should be a typical feature of charge trapping in oxides with wide defect bands in which slower traps come into play during the reverse sweep, thus broadening the maximum and moving it outside the measurement range. However, previously it has been disregarded when using just a constant current extraction of $\Delta V_{\mathrm{H}}$. Thus it was typically assumed that experimental observation of the $\Delta V_{\mathrm{H}}$ maximum for 2D FETs with oxide insulators would require very long sweep times which cannot be reached in most experiments. In reality, the maximum is still there and could be used to get more information about charge trapping dynamics if the full mapping of hysteresis is performed.
\begin{figure}[!h]
\centering
\vspace{0mm}
\begin{minipage} {\textwidth} %{\sminipagewidth}
\hspace{0.7cm}
\centering
\includegraphics[width=13cm]{\figdir/Fig.7.eps} %???
\caption{\label{Fig.7} New understanding of the hysteresis results obtained for the top-gated MoS$_2$/Al$_2$O$_3$ FETs in~\cite{IllARIONOV17A} using our universal hysteresis mapping method. (a) The $I_{\mathrm{D}}$-$V_{\mathrm{TG}}$ characteristics showing large CW hysteresis which actually has a maximum near the ON state as schematically shown in (b). (c) The results obtained for this dataset using our universal hysteresis mapping method confirm this schematic interpretation, i.e. a clear maximum at higher currents which becomes broader closer to $V_{\mathrm{th}}$. }
\end{minipage}
\end{figure}
Next we apply our universal hysteresis mapping method to the experimental results for imec FAB MoS$_2$ FETs~\cite{ILLARIONOV24A}, in which clockwise and counterclockwise hysteresis are present simultaneously at slow $t_{\mathrm{sw}}$ (Fig.8a), with the latter being due to top gate edge defects. A single point extraction of $\Delta V_{\mathrm{H}}$ previously allowed us to reveal this interplay only at high temperatures when it becomes more pronounced~\cite{ILLARIONOV24A}. In contrast, here we can see a sizable counterclockwise contribution in the lower UHF already at 25$^{\mathrm{o}}$C (Fig.8b), which confirms the high sensitivity of our TCAD-inspired mapping method even to those hysteresis features which are barely visible.
Furthermore, in Fig.8c,d we examine the CW/CCW hysteresis switching which appears at high temperatures in back-gated MoS$_2$/SiO$_2$ FETs. As established in~\cite{KNOBLOCH23} by comprehensive TCAD modeling, this is due to the interplay between conventional charge trapping responsible for the CW hysteresis and drift of mobile ions in the oxide which introduce the CCW contribution. Our mapping results shown in Fig.8d clearly separate the CW contribution which appears as a standard increase of $\Delta V_{\mathrm{H}}$ for slower sweeps in the upper UHF, and the maximum of the CCW hysteresis in the lower UHF with the time position defined by the ion migration barrier~\cite{KNOBLOCH23}. The intermediate $\Delta V_{\mathrm{H}}$(1/$t_{\mathrm{sw}}$) curves such as the one extracted near the switching point are obviously affected by both hysteresis mechanisms. Therefore, the validity of our mapping method is not limited only to the hysteresis dynamics with charge trapping by insulator defects.
\begin{figure}[!h]
\centering
\vspace{0mm}
\begin{minipage} {\textwidth} %{\sminipagewidth}
\hspace{0.02cm}
\centering
\includegraphics[width=16.5cm]{\figdir/Fig.8.eps} %???
\caption{\label{Fig.8} (a) The $I_{\mathrm{D}}$-$V_{\mathrm{TG}}$ characteristics measured using different $t_{\mathrm{sw}}$ for FAB MoS$_2$ FETs~\cite{ILLARIONOV24A}. (b) The hysteresis curves obtained using our universal mapping method. A clear counterclockwise feature is present in the lower UHF, which reflects slow sweep hysteresis above $V_{\mathrm{th}}$. (c) The $I_{\mathrm{D}}$-$V_{\mathrm{BG}}$ characteristics measured using different $t_{\mathrm{sw}}$ for back-gated MoS$_2$/SiO$_2$ FETs~\cite{KNOBLOCH23} which show a clear CW/CCW hysteresis switching. (d) The corresponding mapping results obtained with 300 $I_{\mathrm{Dconst}}$ steps which clearly split the CW contribution coming from charge trapping by SiO$_2$ defects (upper UHF) and CCW hysteresis caused by mobile ions in SiO$_2$ (lower UHF).}
\end{minipage}
\end{figure}
\section{Conclusion}
In summary, by doing TCAD modeling for scaled MoS$_2$ FETs with different distributions of oxide traps we have derived and justified the universal method for mapping of hysteresis. The use of this method allowed us to get in depth understanding of the correlation between hysteresis and PBTI dynamics caused by HfO$_2$ defect bands while considering bias dependences of the time constants and the band diagrams. Experimental validation of our approach suggests that it can be used to explain complex hysteresis features which are observed in modern 2D FETs and often result from scaling, such as CW/CCW switching and time separation caused by an interplay between different physical mechanisms such as charge trapping from the channel and top gate, as well as drift of mobile ions in gate insulators. Our results considerably extend previous knowledge gained using the constant current extraction of hysteresis and could simplify the analysis of experimental data with TCAD, since they suggest the use of standardized hysteresis distributions rather than a single ill-defined curve. We expect that these findings will be useful to strengthen the maturity of reliability analysis in 2D FETs which are currently under trial FAB integration but cannot move forward in part due to the lack of reliability expertise in the community.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%% The "Acknowledgement" section can be given in all manuscript
%% classes. This should be given within the "acknowledgement"
%% environment, which will make the correct section or running title.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\section{Methods}
\textit{TCAD modeling approach}
Our TCAD setup is based on a well-established simulator GTS Minimos-NT~\cite{MINIMOSNT22} which describes carrier transport in the channel with the drift-diffusion model and employs the four-state non-radiative multiphonon (NMP) model~\cite{GRASSER12A,ALKAUSKAS14} for charge trapping by oxide defects. This professional device modeling tool originally developed for Si technologies has been successfully adapted to FETs with 2D channels and used in many previous works~\cite{ILLARIONOV16B,KNOBLOCH18,KNOBLOCH23}. For a comprehensive reliability analysis it is important that Minimos-NT allows to extract the time constants and charge states for the whole ensemble of defects, as well as the band diagrams.
We consider top-gated MoS$_2$ FETs with scaled (EOT$\,=\,$1$\,$nm) HfO$_2$ gate insulators and place the trap distributions in different sections of the oxide. Thus the most essential ingredient of our TCAD approach is the time-dependent analysis of charge trapping which is performed using the NMP model parameters~\cite{GRASSER12A} provided in Fig.S1 in the SI. While implementing real time as a variable~\cite{KNOBLOCH18}, we first obtain the hysteresis dynamics by simulating the forward/reverse $I_{\mathrm{D}}$-$V_{\mathrm{TG}}$ curves with logarithmically spaced sweep times $t_{\mathrm{sw}}$ from 10$^{-7}\,$s to 10$^{8}\,$s, thereby going far beyond the typical measurement ranges. The same approach is also employed to simulate the typically used BTI measure-stress-measure loops~\cite{PROVIAS23}, in which the device is first subjected to a relaxation voltage $V_{\mathrm{TGR}}$, followed by the fast sweep of the fresh $I_{\mathrm{D}}$-$V_{\mathrm{TG}}$ curve, and then a stress voltage $V_{\mathrm{TGS}}$ is applied for a certain stress time $t_{\mathrm{s}}$. Then the recovery is monitored using intermediate $I_{\mathrm{D}}$-$V_{\mathrm{TG}}$ sweeps while placing the device back to $V_{\mathrm{TGR}}$ in between these sweeps. Just like in real BTI experiments, subsequent stress/recovery rounds with logarithmically increased $t_{\mathrm{s}}$ were implemented in our TCAD modeling setup.
\textit{Universal hysteresis mapping method}
Our universal hysteresis mapping method suggests scanning of the hysteresis width $\Delta V_{\mathrm{H}}$ from the current value slightly above $I_{\mathrm{OFF}}$ to near-$I_{\mathrm{ON}}$ values. The exact mapping range has to be adjusted depending on the presence of obvious artifacts in a particular set of the gate transfer curves simulated or measured for different $t_{\mathrm{sw}}$, for instance noise at low currents. Using a very fine step on $I_{\mathrm{Dconst}}$, we can obtain a serie with thousands of $\Delta V_{\mathrm{H}}$(1/$t_{\mathrm{sw}}$) curves. Then next we can express the hysteresis dynamics as a distribution of these curves concluded in between the upper and lower universal hysteresis functions (UHFs) constructed as a piecewise maximum and minimum from the obtained set of $\Delta V_{\mathrm{H}}$(1/$t_{\mathrm{sw}}$) dependences, respectively.
\begin{acknowledgement}
The authors acknowledge the financial support from National Natural Science Foundation of China (NSFC) grant number W2432040, Guangdong Basic and Applied Basic Research Foundation 2024A1515010179 and Shenzhen Science and Technology Program 20231115150611001. We also thank Thomas Mueller (TU Wien), Eric Pop (Stanford), Ben Kaczer (imec) and Quentin Smets (imec) for previous collaboration which resulted in gaining the experimental datasets used for validation of the universal hysteresis mapping method.
\end{acknowledgement}
\section{Competing interests}
The authors declare no competing financial or non-financial interests.
\section{Data availability}
The data that support the findings of this work are available from the corresponding author upon reasonable request.
\section{Author contributions}
Y.Z.L. performed TCAD modeling, analyzed the results and wrote the manuscript together with Y.Y.I. who also supervised the research. H.C. and Y.W. contributed to data analysis and preparation of figures.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%% The appropriate \bibliography command should be placed here.
%% Notice that the class file automatically sets \bibliographystyle
%% and also names the section correctly.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%\bibliography{./bib/diss}
%\bibliography{./bib/diss1}
\providecommand{\latin}[1]{#1}
\makeatletter
\providecommand{\doi}
{\begingroup\let\do\@makeother\dospecials
\catcode`\{=1 \catcode`\}=2 \doi@aux}
\providecommand{\doi@aux}[1]{\endgroup\texttt{#1}}
\makeatother
\providecommand*\mcitethebibliography{\thebibliography}
\csname @ifundefined\endcsname{endmcitethebibliography}
{\let\endmcitethebibliography\endthebibliography}{}
\begin{mcitethebibliography}{35}
\providecommand*\natexlab[1]{#1}
\providecommand*\mciteSetBstSublistMode[1]{}
\providecommand*\mciteSetBstMaxWidthForm[2]{}
\providecommand*\mciteBstWouldAddEndPuncttrue
{\def\EndOfBibitem{\unskip.}}
\providecommand*\mciteBstWouldAddEndPunctfalse
{\let\EndOfBibitem\relax}
\providecommand*\mciteSetBstMidEndSepPunct[3]{}
\providecommand*\mciteSetBstSublistLabelBeginEnd[3]{}
\providecommand*\EndOfBibitem{}
\mciteSetBstSublistMode{f}
\mciteSetBstMaxWidthForm{subitem}{(\alph{mcitesubitemcount})}
\mciteSetBstSublistLabelBeginEnd
{\mcitemaxwidthsubitemform\space}
{\relax}
{\relax}
\bibitem[{Wachter, S. and Polyushkin, D.K. and Bethge, O. and Mueller,
T.}(2017)]{WACHTER17}
{Wachter, S. and Polyushkin, D.K. and Bethge, O. and Mueller, T.} {A
Microprocessor Based on a Two-Dimensional Semiconductor}. \emph{Nat.
Commun.} \textbf{2017}, \emph{8}, 14948\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[Lemme \latin{et~al.}(2022)Lemme, Akinwande, Huyghebaert, and
Stampfer]{LEMME22}
Lemme,~M.; Akinwande,~D.; Huyghebaert,~C.; Stampfer,~C. 2D Materials for Future
Heterogeneous Electronics. \emph{Nat. Commun.} \textbf{2022},
\emph{13}, 1392\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[Huang \latin{et~al.}(2022)Huang, Wan, Shi, Zhang, Wang, Wang, Yang,
Liu, Lin, Guan, \latin{et~al.} others]{HUANG22}
Huang,~J.-K. \latin{et~al.} {High-$\kappa$ Perovskite Membranes as
Insulators for Two-Dimensional Transistors}. \emph{Nature} \textbf{2022},
\emph{605}, 262--267\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[Tan \latin{et~al.}(2023)Tan, Yu, Tang, Gao, Yin, Zhang, Wang, Gao,
Zhou, Zheng, Liu, Jiang, Ding, and Peng]{TAN23}
Tan,~C. \latin{et~al.} {2D Fin Field-Effect Transistors Integrated with Epitaxial High-k
Gate Oxide}. \emph{Nature} \textbf{2023}, \emph{616}, 66--72\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[Pendurthi \latin{et~al.}(2024)Pendurthi, Sakib, Sadaf, Zhang, Sun,
Chen, Jayachandran, Oberoi, Ghosh, Kumari, \latin{et~al.}
others]{PENDURTHI24}
Pendurthi,~R. \latin{et~al.} Monolithic
Three-dimensional Integration of Complementary Two-Dimensional Field-Effect
Transistors. \emph{Nat. Nanotechnol.} \textbf{2024}, \emph{19},
970--977\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[Asselberghs \latin{et~al.}(2020)Asselberghs, Smets, Schram, Groven,
Verreck, Afzalian, Arutchelvan, Gaur, Cott, Maurice, Brems, Kennes,
Phommahaxay, Dupuy, Radisic, de~Marneffe, Thiam, Li, Devriendt, Huyghebaert,
lin, Caymax, Morin, and Radu]{ASSELBERGHS20}
Asselberghs,~I. \latin{et~al.} Wafer-Scale Integration of Double Gated WS$_2$
Transistors in 300mm Si CMOS Fab. IEEE International Electron Devices Meeting
(IEDM). 2020; pp 40--2\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[Dorow \latin{et~al.}(2022)Dorow, Penumatcha, Kitamura, Rogan,
OBrien, Lee, Ramamurthy, Cheng, Maxey, Zhong, \latin{et~al.}
others]{DOROW22}
Dorow,~C. \latin{et~al.} Gate Length
Scaling beyond Si: Mono-Layer 2D Channel FETs Robust to Short Channel
Effects. IEEE International Electron Devices Meeting (IEDM). 2022; pp 7--5\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[Chung \latin{et~al.}(2024)Chung, Chou, Yun, Hsu, Yu, Chang, Lee, Chou,
Ho, Wei, \latin{et~al.} others]{CHUNG24}
Chung,~Y.-Y. \latin{et~al.} Stacked Channel
Transistors with 2D Materials: An Integration Perspective. IEEE International
Electron Devices Meeting (IEDM). 2024; pp 1--4\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[Mortelmans \latin{et~al.}(2024)Mortelmans, Buragohain, Kitamura,
Dorow, Rogan, Siddiqui, Ramamurthy, Lux, Zhong, Harlson, \latin{et~al.}
others]{MORTELMANS24}
Mortelmans,~W. \latin{et~al.} Gate
Oxide Module Development for Scaled GAA 2D FETs Enabling SS$<$ 75mV/d and
Record Idmax$>$ 900$\mu$A/$\mu$m at Lg$<$ 50nm. IEEE International Electron
Devices Meeting (IEDM). 2024; pp 1--4\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[Schwierz \latin{et~al.}(2015)Schwierz, Pezoldt, and
Granzner]{SCHWIERZ15}
Schwierz,~F.; Pezoldt,~J.; Granzner,~R. {Two-Dimensional Materials and Their
Prospects in Transistor Electronics}. \emph{Nanoscale} \textbf{2015},
\emph{7}, 8261--8283\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[Das \latin{et~al.}(2021)Das, Sebastian, Pop, McClellan, Franklin,
Grasser, Knobloch, Illarionov, Penumatcha, Appenzeller, Cheng, Zhu,
Asselberghs, Li, Avci, Bhat, Anthopoulos, and Singh]{DAS21}
Das,~S. \latin{et~al.} {Transistors Based on Two-Dimensional Materials for
Future Integrated Circuits}. \emph{Nat. Electron.} \textbf{2021},
\emph{4}, 786--799\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[Illarionov \latin{et~al.}(2020)Illarionov, Knobloch, Jech, Lanza,
Akinwande, Vexler, Mueller, Lemme, Fiori, Schwierz, and
Grasser]{ILLARIONOV20A}
Illarionov,~Y. \latin{et~al.} {Insulators for
2D nanoelectronics: the gap to bridge}. \emph{Nat. Commun.}
\textbf{2020}, \emph{11}, 3385\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[Jayachandran \latin{et~al.}(2024)Jayachandran, Pendurthi, Sadaf,
Sakib, Pannone, Chen, Han, Trainor, Kumari, Mc~Knight, \latin{et~al.}
others]{JAYACHANDRAN24}
Jayachandran,~D. \latin{et~al.} Three-dimensional integration of two-dimensional field-effect transistors.
\emph{Nature} \textbf{2024}, \emph{625}, 276--281\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[Illarionov \latin{et~al.}(2024)Illarionov, Knobloch, Uzlu,
Banshchikov, Ivanov, Sverdlov, Otto, Stoll, Vexler, Waltl, Wang, Manna,
Neumaier, M.C., N.S., and Grasser]{ILLARIONOV24A}
Illarionov,~Y. \latin{et~al.} {Variability and high temperature reliability of
graphene field-effect transistors with thin epitaxial CaF$_2$ insulators}.
\emph{npj 2D Mater. and Appl.} \textbf{2024}, \emph{8}, 23\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[Kaczer \latin{et~al.}(2010)Kaczer, Roussel, Grasser, and
Groeseneken]{KACZER10A}
Kaczer,~B.; Roussel,~P.; Grasser,~T.; Groeseneken,~G. {Statistics of Multiple
Trapped Charges in the Gate Oxide of Deeply Scaled MOSFET Devices-Application
to NBTI}. \emph{IEEE Electron Dev. Lett.} \textbf{2010}, \emph{31},
411--413\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[Grasser(2012)]{GRASSER12A}
Grasser,~T. {Stochastic Charge Trapping in Oxides: From Random Telegraph Noise
to Bias Temperature Instabilities}. \emph{Microelectron. Reliab.}
\textbf{2012}, \emph{52}, 39--70\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[Markov \latin{et~al.}(2013)Markov, Amoroso, Gerrer, Adamu-Lema, and
Asenov]{MARKOV13}
Markov,~S.; Amoroso,~S.; Gerrer,~L.; Adamu-Lema,~F.; Asenov,~A. {Statistical
Interactions of Multiple Oxide Traps under BTI Stress of Nanoscale MOSFETs}.
\emph{IEEE Electron Dev. Lett.} \textbf{2013}, \emph{34}, 686--688\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[IEEE(2021)]{IRDS21}
IEEE IRDS More Moore. 2021\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[Roh \latin{et~al.}(2016)Roh, Lee, Jin, and Lee]{ROH16}
Roh,~J.; Lee,~J.-H.; Jin,~S.~H.; Lee,~C. Negligible Hysteresis of Molybdenum
Disulfide Field-Effect Transistors through Thermal Annealing. \emph{Journal
of Information Display} \textbf{2016}, \emph{17}, 103--108\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[Vu \latin{et~al.}(2018)Vu, Fan, Lee, Joo, Yu, and Lee]{VU18}
Vu,~Q.; Fan,~S.; Lee,~S.; Joo,~M.-K.; Yu,~W.; Lee,~Y. {Near-Zero Hysteresis and
Near-Ideal Subthreshold Swing in h-BN Encapsulated Single-Layer MoS$_2$
Field-Effect Transistors}. \emph{2D Mater.} \textbf{2018}, \emph{5},
031001\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[Cho \latin{et~al.}(2021)Cho, Pujar, Choi, Kang, Hong, Park, Baek, Kim,
Lee, and Kim]{CHO21E}
Cho,~H.~W.; Pujar,~P.; Choi,~M.; Kang,~S.; Hong,~S.; Park,~J.; Baek,~S.;
Kim,~Y.; Lee,~J.; Kim,~S. Direct Growth of Orthorhombic Hf$_{0.5}$Zr$_{0.5}$O$_2$ Thin
Films for Hysteresis-free MoS$_2$ Negative Capacitance Field-Effect Transistors.
\emph{npj 2D Mater. and Appl.} \textbf{2021}, \emph{5}, 46\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[Venkatakrishnarao \latin{et~al.}(2024)Venkatakrishnarao, Mishra, Tarn,
Bosman, Lee, Das, Mukherjee, Talha-Dean, Zhang, Teo, \latin{et~al.}
others]{VENKA24}
Venkatakrishnarao,~D. \latin{et~al.} Liquid Metal
Oxide-Assisted Integration of High-k Dielectrics and Metal Contacts for
Two-Dimensional Electronics. \emph{ACS Nano} \textbf{2024}, \emph{18},
26911--26919\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[Fan \latin{et~al.}(2025)Fan, Yi, Deng, Zhou, Zhang, Yu, Li, Li, Wu,
Zhou, \latin{et~al.} others]{FAN25}
Fan,~X. \latin{et~al.} 2D Edge-seeded Heteroepitaxy of Ultrathin High-$\kappa$
Dielectric CaNb$_2$O$_6$ for 2D Field-Effect Transistors. \emph{Nat.
Commun.} \textbf{2025}, \emph{16}, 2585\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[Lan \latin{et~al.}(2025)Lan, Yang, Kantre, Cott, Tripathi,
Appenzeller, and Chen]{LAN25}
Lan,~H.-Y.; Yang,~S.-H.; Kantre,~K.-A.; Cott,~D.; Tripathi,~R.;
Appenzeller,~J.; Chen,~Z. Reliability of high-performance monolayer MoS$_2$
transistors on scaled high-$\kappa$ HfO$_2$. \emph{npj 2D Mater. and
Appl.} \textbf{2025}, \emph{9}, 5\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[Provias \latin{et~al.}(2023)Provias, Knobloch, Kitamura, OBrien,
Dorow, Waldhoer, Stampfer, Penumatcha, Lee, Ramamurthy, \latin{et~al.}
others]{PROVIAS23}
Provias,~A. \latin{et~al.} Reliability Assessment of Double-Gated Wafer-Scale MoS$_2$ Field Effect
Transistors through Hysteresis and Bias Temperature Instability Analyses.
IEEE International Electron Devices Meeting (IEDM). 2023; pp 1--4\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[Grasser \latin{et~al.}(2011)Grasser, Kaczer, G{\"o}s, Reisinger,
Aichinger, Hehenberger, Wagner, Franco, Toledano-Luque, and
Nelhiebel]{GRASSER11A}
Grasser,~T.; Kaczer,~B.; G{\"o}s,~W.; Reisinger,~H.; Aichinger,~T.;
Hehenberger,~P.; Wagner,~P.-J.; Franco,~J.; Toledano-Luque,~M.; Nelhiebel,~M.
{The Paradigm Shift in Understanding the Bias Temperature Instability: From
Reaction-Diffusion to Switching Oxide Traps}. \emph{IEEE Trans. Electron
Devices} \textbf{2011}, \emph{58}, 3652--3666\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[al (2022)]{MINIMOSNT22}
{GTS Minimos-NT User Manual}. Global TCAD Solutions, Vienna, Austria,
2022\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[{Alkauskas, A. and Yan, Q. and Van de Walle, C.G.}(2014)]{ALKAUSKAS14}
{Alkauskas, A. and Yan, Q. and Van de Walle, C.G.} {First-Principles Theory of
Nonradiative Carrier Capture via Multiphonon Emission}. \emph{Phys. Rev. B}
\textbf{2014}, \emph{90}, 075202\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[Illarionov \latin{et~al.}(2016)Illarionov, Waltl, Rzepa, Kim, Kim,
Dodabalapur, Akinwande, and Grasser]{ILLARIONOV16B}
Illarionov,~Y.; Waltl,~M.; Rzepa,~G.; Kim,~J.-S.; Kim,~S.; Dodabalapur,~A.;
Akinwande,~D.; Grasser,~T. {Long-Term Stability and Reliability of Black
Phosphorus Field-Effect Transistors}. \emph{ACS Nano} \textbf{2016},
\emph{10}, 9543--9549\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[Knobloch \latin{et~al.}(2018)Knobloch, Rzepa, Illarionov, Waltl,
Schanovsky, Stampfer, Furchi, Mueller, and Grasser]{KNOBLOCH18}
Knobloch,~T.; Rzepa,~G.; Illarionov,~Y.; Waltl,~M.; Schanovsky,~F.;
Stampfer,~B.; Furchi,~M.; Mueller,~T.; Grasser,~T. {A Physical Model for the
Hysteresis in MoS$_2$ Transistors}. \emph{IEEE J. Electron Dev. Soc.}
\textbf{2018}, \emph{6}, 972--978\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[Knobloch \latin{et~al.}(2023)Knobloch, Waldhoer, Davoudi, Karl,
Khakbaz, Matzinger, Zhang, Smithe, Nazir, Liu, \latin{et~al.}
others]{KNOBLOCH23}
Knobloch,~T \latin{et~al.} Modeling the Performance
and Reliability of Two-Dimensional Semiconductor Transistors. IEEE
International Electron Devices Meeting (IEDM). 2023; pp 1--4\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[Rasmussen and Thygesen(2015)Rasmussen, and Thygesen]{RASMUSSEN15}
Rasmussen,~F.; Thygesen,~K. {Computational 2D Materials Database: Electronic
Structure of Transition-Metal Dichalcogenides and Oxides}. \emph{J.
Phys. Chem. C} \textbf{2015}, \emph{119}, 13169--13183\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[Illarionov \latin{et~al.}(2016)Illarionov, Rzepa, Waltl, Knobloch,
Grill, Furchi, Mueller, and Grasser]{ILLARIONOV16A}
Illarionov,~Y.; Rzepa,~G.; Waltl,~M.; Knobloch,~T.; Grill,~A.; Furchi,~M.;
Mueller,~T.; Grasser,~T. {The Role of Charge Trapping in MoS$_2$/SiO$_2$ and
MoS$_2$/hBN Field-Effect Transistors}. \emph{2D Mater.} \textbf{2016},
\emph{3}, 035004\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\bibitem[Illarionov \latin{et~al.}(2017)Illarionov, Knobloch, Waltl, Rzepa,
Pospischil, Polyushkin, Furchi, Mueller, and Grasser]{IllARIONOV17A}
Illarionov,~Y.; Knobloch,~T.; Waltl,~M.; Rzepa,~G.; Pospischil,~A.;
Polyushkin,~D.; Furchi,~M.; Mueller,~T.; Grasser,~T. {Energetic Mapping of
Oxide Traps in MoS$_2$ Field-Effect Transistors}. \emph{2D Mater.}
\textbf{2017}, \emph{4}, 025108\relax
\mciteBstWouldAddEndPuncttrue
\mciteSetBstMidEndSepPunct{\mcitedefaultmidpunct}
{\mcitedefaultendpunct}{\mcitedefaultseppunct}\relax
\EndOfBibitem
\end{mcitethebibliography}
\end{document}
%\input{/export/users/grasser/book/syms2}
\input{./reliability/syms2}
%\newcommand{\textmu} {\mu}
\newcommand{\HIP} {\ensuremath{\mathrm{HIP}}}
\newcommand{\HIPs} {\ensuremath{\mathrm{HIPs}}}
\newcommand{\Hyd} {\ensuremath{\mathrm{H}}}
\newcommand{\HydT} {\ensuremath{\mathrm{H_T}}}
\newcommand{\Hydi} {\ensuremath{\mathrm{H_i}}}
\newcommand{\Hydip} {\ensuremath{\mathrm{H_i^+}}}
\newcommand{\Hydp} {\ensuremath{\mathrm{H^+}}}
\newcommand{\Hydm} {\ensuremath{\mathrm{H^-}}}
\newcommand{\Hydz} {\ensuremath{\mathrm{H^0}}}
\newcommand{\Hydt} {\ensuremath{\mathrm{H_2}}}
\newcommand{\Deut} {\ensuremath{\mathrm{D}}}
\newcommand{\Ox} {\ensuremath{\mathrm{O}}}
\newcommand{\N} {\ensuremath{\mathrm{N}}}
\newcommand{\Si} {\ensuremath{\mathrm{Si}}}
\renewcommand{\P} {\ensuremath{\mathrm{P}}}
\renewcommand{\B} {\ensuremath{\mathrm{B}}}
\newcommand{\aSi} {\ensuremath{a\text{-}\mathrm{Si}}}
\newcommand{\cSi} {\ensuremath{c\text{-}\mathrm{Si}}}
\newcommand{\pSi} {\ensuremath{p\text{-}\mathrm{Si}}}
\newcommand{\Al} {\ensuremath{\mathrm{Al}}}
\newcommand{\AltOt} {\ensuremath{\mathrm{Al_2 O_3}}}
\newcommand{\SiN} {\ensuremath{\mathrm{Si_3 N_4}}}
\newcommand{\SiOt} {\ensuremath{\mathrm{SiO_2}}}
\newcommand{\SiON} {\ensuremath{\mathrm{SiON}}}
\newcommand{\HfOt} {\ensuremath{\mathrm{HfO_2}}}
\newcommand{\aSiOt} {\ensuremath{a\text{-}\mathrm{SiO_2}}}
\newcommand{\MoSt} {\ensuremath{\mathrm{MoS_2}}}
\newcommand{\X} {\mathrm{X}}
\newcommand{\K} {\mathrm{K}}
\newcommand{\Xit} {X_\mathrm{it}}
\newcommand{\Xbulk} {X_\mathrm{bulk}}
\newcommand{\Dim}[1] {[#1]}
\newcommand{\conc}[1] {[#1]}
\newcommand{\hp} {\mathrm{h}^+}
\newcommand{\eminus} {\mathrm{e}^-}
\newcommand{\dij} {d_{ij}}
\newcommand{\Dox} {D_\mathrm{ox}}
\newcommand{\dDox} {\Delta \Dox}
\newcommand{\Dit} {D_\mathrm{it}}
\newcommand{\dDit} {\Delta \Dit}
\newcommand{\Ditz} {D_\mathrm{it,0}}
\newcommand{\Qf} {Q_\mathrm{f}}
\newcommand{\Qit} {Q_\mathrm{it}}
\newcommand{\dQit} {\Delta \Qit}
\newcommand{\QitA} {Q_\mathrm{it,A}}
\newcommand{\QitD} {Q_\mathrm{it,D}}
\newcommand{\Qitsimple} {Q_\mathrm{it}^\mathrm{simple}}
\newcommand{\Qot} {Q_\mathrm{ot}}
\newcommand{\dQot} {\Delta \Qot}
\newcommand{\Qbt} {Q_\mathrm{bt}}
\newcommand{\Qtot} {Q_\mathrm{tot}}
\newcommand{\rQ} {r_Q}
\newcommand{\tz} {t_0}
\newcommand{\tox} {t_\mathrm{ox}}
\newcommand{\Cox} {C_\mathrm{ox}}
\newcommand{\Qox} {Q_\mathrm{ox}}
\newcommand{\dQox} {\Delta \Qox}
\newcommand{\Eox} {E_\mathrm{ox}}
\newcommand{\Ecrit} {E_\mathrm{crit}}
\newcommand{\EA} {E_\mathrm{A}}
\newcommand{\EAc} {E_\mathrm{Ac}}
\newcommand{\EAe} {E_\mathrm{Ae}}
\newcommand{\EAm} {E_\mathrm{Am}}
\newcommand{\EAz} {E_\mathrm{A0}}
\newcommand{\Ea} {E_\mathrm{a}}
\newcommand{\Eaf} {E_\mathrm{af}}
\newcommand{\Ear} {E_\mathrm{ar}}
\newcommand{\Ez} {E_0}
\newcommand{\ER} {\nrg_\mathrm{R}}
\newcommand{\ERs} {\nrg'_\mathrm{R}}
\newcommand{\ERAB} {\nrg_\mathrm{R12}}
\newcommand{\ERBA} {\nrg_\mathrm{R21}}
\newcommand{\Nint} {\mathrm{N}_\mathrm{int}} % for nitrogen at the interface
\newcommand{\Ct} {C_\mathrm{t}}
\newcommand{\tstress} {t_\mathrm{s}}
\newcommand{\trelax} {t_\mathrm{r}}
\newcommand{\tstressMax} {t_\mathrm{s}^\mathrm{max}}
\newcommand{\tstressMin} {t_\mathrm{s}^\mathrm{min}}
\newcommand{\trelaxMax} {t_\mathrm{r}^\mathrm{max}}
\newcommand{\trelaxMin} {t_\mathrm{r}^\mathrm{min}}
\newcommand{\nit} {n_\mathrm{it}}
\newcommand{\pit} {p_\mathrm{it}}
\newcommand{\Nit} {N_\mathrm{it}}
\newcommand{\Nitz} {N_\mathrm{it0}}
\newcommand{\Notz} {N_\mathrm{ot0}}
\newcommand{\dNit} {\Delta N_\mathrm{it}}
\newcommand{\dNitMax} {\Delta N_\mathrm{it,max}}
\newcommand{\Nitstress} {N_\mathrm{it,s}}
\newcommand{\Nith} {\tilde{N}_\mathrm{it}}
\newcommand{\Nbth} {\tilde{N}_\mathrm{bt}}
\newcommand{\Nitk} {N_\mathrm{it}^k}
\newcommand{\dNot} {\Delta N_\mathrm{ot}}
\newcommand{\Ntk} {N_\mathrm{t}^k}
\newcommand{\tNt} {\tilde{\Nt}}
\newcommand{\tg} {\tilde{g}}
\newcommand{\ck} {c^k}
\newcommand{\rk} {r^k}
\newcommand{\ci} {c_i}
\newcommand{\cj} {c_j}
\newcommand{\ri} {r_i}
\newcommand{\rj} {r_j}
\newcommand{\nnot} {n_\mathrm{ot}}
\newcommand{\Not} {N_\mathrm{ot}}
\newcommand{\Nf} {N_\mathrm{f}}
\newcommand{\Nits} {\widetilde{N_\mathrm{it}}}
\newcommand{\fs} {\tilde{f}}
\newcommand{\Gs} {\tilde{G}}
\newcommand{\Gsz} {\tilde{G}_\mathrm{it}}
\newcommand{\Gc} {G_\mathrm{c}}
\newcommand{\Gz} {G_\mathrm{0}}
\newcommand{\Git} {G_\mathrm{it}}
\newcommand{\Giti} {G_{\mathrm{it},i}}
\newcommand{\Nz} {N_0}
\newcommand{\PhiX} {\Phi_X}
\newcommand{\phiF} {\phi_\mathrm{F}}
\newcommand{\phiB} {\phi_\mathrm{B}}
\newcommand{\Ep} {\ensuremath{E'}}
\newcommand{\Epg} {\ensuremath{E'_\gamma}}
\newcommand{\Epd} {\ensuremath{E'_\delta}}
\newcommand{\Eps} {\ensuremath{E'_\mathrm{s}}}
\newcommand{\Epp} {\ensuremath{E'^+}}
\newcommand{\SSi} {\ensuremath{S_\Si}}
\newcommand{\SSip} {\ensuremath{S_\Si}^+}
\newcommand{\SOx} {\ensuremath{S_\Ox}}
\newcommand{\SOxp} {\ensuremath{S_\Ox}^+}
\newcommand{\Epbeta} {\ensuremath{E'_\beta}}
\newcommand{\KN} {\ensuremath{K_\mathrm{N}}}
\newcommand{\Pb} {\ensuremath{P_\mathrm{b}}}
\newcommand{\Pbp} {\ensuremath{P_\mathrm{b}^+}}
\newcommand{\Pbz} {\ensuremath{P_\mathrm{b0}}}
\newcommand{\Pbo} {\ensuremath{P_\mathrm{b1}}}
\newcommand{\sX} {s_h}
\newcommand{\Fh} {F_h}
\newcommand{\Fhi} {F_{h1}}
\newcommand{\Fhj} {F_{h2}}
\newcommand{\vFh} {\vect{F}_h}
\newcommand{\vEox} {\vect{E}_\mathrm{ox}}
\newcommand{\vFc} {\vect{F}_\mathrm{c}}
\newcommand{\vFcin} {\vect{F}_\mathrm{c}^\mathrm{B}}
\newcommand{\avFcin} {\vect{\overline{F}}_\mathrm{c}^\mathrm{B}}
\newcommand{\vFin} {\vect{F}^\mathrm{B}}
\newcommand{\vFz} {\vect{F}_\mathrm{0}}
\newcommand{\vFtot} {\vect{F}_\mathrm{tot}}
\newcommand{\vFhk} {\vect{F}^{(k)}_h}
\newcommand{\vFhij} {\vect{F}_{hij}}
\newcommand{\DHeff} {D_\mathrm{H,eff}}
\renewcommand{\DH} {D_\mathrm{H}}
\newcommand{\DX} {D_h}
\newcommand{\DXit} {D_{h,\mathrm{it}}}
\newcommand{\DXz} {D_{h,0}}
\newcommand{\NH} {N_\mathrm{H}}
%\newcommand{\NX} {N_X}
%\newcommand{\NXtot} {N_{X\mathrm{,tot}}}
\newcommand{\NX} {h}
\newcommand{\NXtot} {h_{\mathrm{tot}}}
%\newcommand{\NXi} {N_{Xi}}
%\newcommand{\NXj} {N_{Xj}}
\newcommand{\NXi} {h_{\mathrm{i}1}}
\newcommand{\NXj} {h_{\mathrm{i}2}}
\newcommand{\NXieq} {N_{Xi}^\mathrm{eq}}
\newcommand{\NXjeq} {N_{Xj}^\mathrm{eq}}
\newcommand{\NXz} {N_{X,0}}
\newcommand{\NXit} {h_{\mathrm{it}}}
\newcommand{\NXkit} {N^{(k)}_{X,\mathrm{it}}}
\newcommand{\NXk} {N^{(k)}_{X}}
\newcommand{\ZX} {Z_h}
\newcommand{\dH} {\Delta H}
\newcommand{\mobX} {\mu_h}
\newcommand{\mobc} {\mu_\mathrm{c}}
\newcommand{\deh} {\delta/2}
\newcommand{\Deh} {\frac{\delta}{2}}
\newcommand{\dEt} {\, \d \Et}
\newcommand{\dz} {\, \d z}
\newcommand{\dA} {\, \d A}
\newcommand{\dvA} {\, \d \vA}
\newcommand{\dV} {\, \d V}
\newcommand{\pa} {p_\mathrm{a}}
\newcommand{\pd} {p_\mathrm{d}}
\newcommand{\ka} {k_\mathrm{a}}
\newcommand{\kd} {k_\mathrm{d}}
\newcommand{\kaz} {k_\mathrm{a0}}
\newcommand{\kdz} {k_\mathrm{d0}}
\newcommand{\kf} {k_\mathrm{f}}
\newcommand{\kfi} {k_{\mathrm{f}1}}
\newcommand{\kfj} {k_{\mathrm{f}2}}
\newcommand{\kfe} {k_\mathrm{f,e}}
\newcommand{\kfz} {k_\mathrm{f0}}
\newcommand{\kr} {k_\mathrm{r}}
\newcommand{\krz} {k_\mathrm{r0}}
\newcommand{\kri} {k_{\mathrm{r}1}}
\newcommand{\krj} {k_{\mathrm{r}2}}
\newcommand{\kkr} {k^{(k)}_\mathrm{r}}
\newcommand{\fk} {f^{(k)}}
\newcommand{\gk} {g^{(k)}}
\newcommand{\gm} {g_\mathrm{m}}
\newcommand{\gmz} {g_\mathrm{m0}}
\newcommand{\texp} {n}
\newcommand{\zindeh} {-\deh \le z \le \deh}
\newcommand{\intdeh}[1] {\int_{-\deh}^{\deh} #1 \dz}
\newcommand{\ADiv}[1] {\mathrm{Div}\, #1}
\newcommand{\vez} {\vect{e}_z}
\newcommand{\Jn} {J_n}
\newcommand{\gn} {g_n}
\newcommand{\gp} {g_p}
\newcommand{\fn} {f_n}
\newcommand{\fp} {f_p}
\newcommand{\NEQ} {n_0}
\newcommand{\PEQ} {p_0}
\newcommand{\h} {h}
\newcommand{\hc} {h_\mathrm{c}}
\newcommand{\hz} {h_0}
\newcommand{\htrapped} {h_\mathrm{t}}
\newcommand{\Hz} {H_0}
\newcommand{\Hc} {H_\mathrm{c}}
\newcommand{\Ht} {H_\mathrm{t}}
\newcommand{\Htot} {H_\mathrm{tot}}
\newcommand{\Pc} {P_\mathrm{c}}
\newcommand{\Pck} {P_{\mathrm{c},k}}
\newcommand{\Pcn}[1] {P_{\mathrm{c},#1}}
\newcommand{\Pd} {P_\mathrm{d}}
\newcommand{\Pe} {P_\mathrm{e}}
\newcommand{\Ptr} {P_\mathrm{r}}
\newcommand{\pc} {p_\mathrm{c}}
\newcommand{\ps} {p_\mathrm{s}}
\newcommand{\pt} {p_\mathrm{t}}
\newcommand{\pz} {p_\mathrm{0}}
\newcommand{\ptk} {p_\mathrm{t}^k}
\newcommand{\pref} {p_\mathrm{ref}}
\newcommand{\Eref} {E_\mathrm{ref}}
\newcommand{\Eth} {E_\mathrm{th}}
\newcommand{\Xc} {X_\mathrm{c}}
\newcommand{\Ed} {E_\mathrm{d}}
\newcommand{\Eam} {E_\mathrm{am}}
\newcommand{\Edm} {E_\mathrm{dm}}
\newcommand{\Edi} {E_\mathrm{di}}
\newcommand{\Em} {E_\mathrm{m}}
\newcommand{\EmA} {E_\mathrm{A}}
\newcommand{\EmD} {E_\mathrm{D}}
\newcommand{\sigmaA} {\sigma_\mathrm{A}}
\newcommand{\sigmaEd} {\sigma_{E_\mathrm{d}}}
\newcommand{\sigmaD} {\sigma_\mathrm{D}}
\newcommand{\Emin} {E_\mathrm{min}}
%\newcommand{\Emid} {E_\mathrm{mid}}
\newcommand{\tmeas} {t_\mathrm{M}}
\newcommand{\Smeas} {S_\mathrm{M}}
\newcommand{\nmeas} {n_\mathrm{M}}
\newcommand{\ximeas} {\xi_\mathrm{M}}
\newcommand{\xz} {x_0}
\newcommand{\xm} {\overline{x}}
\newcommand{\tp} {t'}
\newcommand{\dtp} {\d \tp}
\newcommand{\intt}[1] {\int_0^t #1 \, \, \dtp}
\newcommand{\inttzt}[1] {\int_{\tz}^t #1 \, \, \dtp}
\newcommand{\intx}[1] {\int_0^\infty #1 \, \, \d x}
\newcommand{\Es} {E^*(t)}
\newcommand{\intzEs}[1] {\int_0^{\Es} #1 \, \, \dEt}
\newcommand{\intEsi}[1] {\int_{\Es}^\infty #1 \, \, \dEt}
\newcommand{\intEcEd}[1] {\int_{\Ec}^{\Ed(t)} #1 \, \, \dEt}
\newcommand{\intEdEc}[1] {\int_{\Ed(t)}^{\Ec} #1 \, \, \dEt}
\newcommand{\intEdi}[1] {\int_{\Ed(t)}^{-\infty} #1 \, \, \dEt}
\newcommand{\intiEd}[1] {\int_{-\infty}^{\Ed(t)} #1 \, \, \dEt}
\newcommand{\VVa} {\d v_\mathrm{a}}
\newcommand{\VVb} {\d v_\mathrm{b}}
\newcommand{\VVc} {\d v_\mathrm{c}}
\newcommand{\VVd} {\d v_\mathrm{d}}
\newcommand{\VVaz} {\d v_\mathrm{a,0}}
\newcommand{\VVbz} {\d v_\mathrm{b,0}}
\newcommand{\VVcz} {\d v_\mathrm{c,0}}
\newcommand{\VVdz} {\d v_\mathrm{d,0}}
\newcommand{\cns} {c_{n}}
\newcommand{\cps} {c_{p}}
\newcommand{\cn} {c_{n}}
\newcommand{\en} {e_{n}}
\newcommand{\cp} {c_{p}}
\newcommand{\ep} {e_{p}}
\newcommand{\Cn} {C_{n}}
\newcommand{\Cp} {C_{p}}
\newcommand{\Sn} {S_{n}}
\newcommand{\Sp} {S_{p}}
\newcommand{\Snk} {S_{n}^k}
\newcommand{\Spk} {S_{p}^k}
\newcommand{\ft} {f_\mathrm{t}}
\newcommand{\fit} {f_\mathrm{it}}
\newcommand{\fox} {f_\mathrm{ox}}
\newcommand{\fbt} {f_\mathrm{bt}}
\newcommand{\fith} {\tilde{f}_\mathrm{it}}
\newcommand{\fitheq} {\tilde{f}_\mathrm{it}^\mathrm{eq}}
\newcommand{\fbth} {\tilde{f}_\mathrm{bt}}
\newcommand{\fteq} {f_\mathrm{t}^\mathrm{eq}}
\newcommand{\ftk} {f_\mathrm{t}^k}
\newcommand{\ftkOld} {f_\mathrm{t}^{k,\mathrm{old}}}
\newcommand{\fpt} {f_{p,\mathrm{t}}}
\newcommand{\fptk} {f_{p,\mathrm{t}}^k}
\newcommand{\Dith} {\tilde{D}_\mathrm{it}}
\newcommand{\Dc} {D_\mathrm{c}}
\newcommand{\gDot} {D_\mathrm{ot}}
\newcommand{\Dbt} {D_\mathrm{bt}}
\newcommand{\Dbth} {\tilde{D}_\mathrm{bt}}
%\newcommand{\st} {s_\mathrm{t}}
\newcommand{\Eit} {E_\mathrm{it}}
\newcommand{\Eith} {\tilde{E}_\mathrm{it}}
\newcommand{\Ebt} {E_\mathrm{bt}}
\newcommand{\Etk} {E_\mathrm{t}^k}
\newcommand{\Emid} {E_\mathrm{mid}}
\newcommand{\Mt} {M_\mathrm{t}}
\newcommand{\nt} {n_\mathrm{t}}
\newcommand{\ns} {n_\mathrm{s}}
\newcommand{\tauz} {\tau_0}
\newcommand{\tauzz} {\tau_{00}}
\newcommand{\tauc} {\tau_\mathrm{c}}
\newcommand{\tauci} {\tau_{\mathrm{c},i}}
\newcommand{\tauej} {\tau_{\mathrm{e},j}}
\newcommand{\tauck} {\tau_{\mathrm{c},k}}
\newcommand{\tauek} {\tau_{\mathrm{e},k}}
\newcommand{\taucmin} {\tau_\mathrm{c,min}}
\newcommand{\taucmax} {\tau_\mathrm{c,max}}
\newcommand{\taucH} {\tau_\mathrm{c}^\mathrm{H}}
\newcommand{\taucL} {\tau_\mathrm{c}^\mathrm{L}}
\newcommand{\taueH} {\tau_\mathrm{e}^\mathrm{H}}
\newcommand{\taueL} {\tau_\mathrm{e}^\mathrm{L}}
\newcommand{\tauH} {\tau^\mathrm{H}}
\newcommand{\tauL} {\tau^\mathrm{L}}
\newcommand{\fH} {f^\mathrm{H}}
\newcommand{\fL} {f^\mathrm{L}}
\newcommand{\pH} {p^\mathrm{H}}
\newcommand{\pL} {p^\mathrm{L}}
\newcommand{\FH} {F^\mathrm{H}}
\newcommand{\FL} {F^\mathrm{L}}
\renewcommand{\taue} {\tau_\mathrm{e}}
\newcommand{\atauc} {\bar{\tau}_\mathrm{c}}
\newcommand{\ataue} {\bar{\tau}_\mathrm{e}}
\newcommand{\atau} {\bar{\tau}}
\renewcommand{\dtheta} {\Delta \theta}
\newcommand{\dtau} {\Delta \tau}
\newcommand{\thetaz} {\theta_0}
\newcommand{\thetac} {\theta_\mathrm{c}}
\newcommand{\thetacH} {\theta^\mathrm{H}_\mathrm{c}}
\newcommand{\thetaeL} {\theta^\mathrm{L}_\mathrm{e}}
\newcommand{\thetas} {\theta_\mathrm{s}}
\newcommand{\thetar} {\theta_\mathrm{r}}
\newcommand{\thetaci} {\theta_{\mathrm{c},i}}
\newcommand{\thetacipo} {\theta_{\mathrm{c},i+1}}
\newcommand{\thetack} {\theta_{\mathrm{c},k}}
\newcommand{\thetaej} {\theta_{\mathrm{e},j}}
\newcommand{\thetaejpo} {\theta_{\mathrm{e},j+1}}
\newcommand{\thetaek} {\theta_{\mathrm{e},k}}
\newcommand{\thetacmin} {\theta_\mathrm{c,min}}
\newcommand{\thetacmax} {\theta_\mathrm{c,max}}
\newcommand{\thetae} {\theta_\mathrm{e}}
\newcommand{\thetaemin} {\theta_\mathrm{e,min}}
\newcommand{\thetaemax} {\theta_\mathrm{e,max}}
\newcommand{\gtau} {g} %_\tau} %_\mathrm{c}}
\newcommand{\gtauij} {g_{i,j}} %_\mathrm{c}}
\newcommand{\gtauc} {g_\mathrm{c}}
\newcommand{\gtaue} {g_\mathrm{e}}
\newcommand{\Gtau} {G} %_\mathrm{c}}
\newcommand{\Gtauc} {G_\mathrm{c}}
\newcommand{\Gtaue} {G_\mathrm{e}}
\newcommand{\gtauP} {p} % ,\mathrm{P}}} %_\mathrm{c}}
\newcommand{\GtauP} {P} % ,\mathrm{P}}} %_\mathrm{c}}
\newcommand{\gtheta} {g_\theta} %_\mathrm{c}}
\newcommand{\gthetac} {g_\theta} %_\mathrm{c}}
\newcommand{\gthetae} {g_\theta} %_\mathrm{c}}
\newcommand{\Gthetac} {G_\theta} %_\mathrm{c}}
\newcommand{\Gthetae} {G_\theta} %_\mathrm{c}}
\newcommand{\nuz} {\nu_0}
\newcommand{\ftz} {f_\mathrm{t,0}}
%\newcommand{\RSRH} {R^\mathrm{SRH}}
\newcommand{\RSRH} {R}
\newcommand{\Rit} {R_\mathrm{it}}
\newcommand{\EFt} {E_\mathrm{Ft}}
\newcommand{\VFB} {V_\mathrm{FB}}
\newcommand{\rhos} {\rho_\mathrm{s}}
\newcommand{\rhosz} {\rho_\mathrm{s,0}}
\newcommand{\rhod} {\rho_\mathrm{d}}
\newcommand{\rhodz} {\rho_\mathrm{d,0}}
\newcommand{\ccite}[1] {[#1]}
\newcommand{\refcite}[2] {[#1] in [#2]}
\newcommand{\figcite}[2] {Fig.~#1 in [#2]}
\newcommand{\pcite}[2] {p.~#1 in [#2]}
\newcommand{\DIV}[1] {\ensuremath{\Nabla \inpr #1}}
\newcommand{\GRAD}[1] {\ensuremath{\Nabla \, #1}}
%\newcommand{\reactionS} {\Leftrightarrow}
\newcommand{\reaction} {\ \rightleftharpoons \ }
\newcommand{\reactionA}[2] {\ \overset{#1}{\underset{#2}{\rightleftharpoons}} \ }
\newcommand{\oreaction} {\ \rightarrow \ }
\newcommand{\oreactionA}[1] {\ \overset{#1}{\rightarrow} \ }
\newcommand{\bondtri} {\ensuremath{\!\!\equiv\!\!}}
\newcommand{\bondtwo} {\text{=}}
\newcommand{\bondone} {\text{--}}
\newcommand{\rplus} {\,+\,}
\newcommand{\trap} {^\bullet}
\newcommand{\atproc} {\,\text{at.}\,\%}
\newcommand{\Prob}[1] {P\{#1\}}
\newcommand{\CondProb}[2] {P\{#1\,|\,#2\}}
\newcommand{\Exp}[1] {\mathrm{E}\{#1\}}
\newcommand{\StdDev}[1] {\mathrm{StdDev}\{#1\}}
\newcommand{\Var}[1] {\mathrm{Var}\{#1\}}
%\newcommand{\degC} {\,^\mathrm{o}\text{C}}
%\newcommand{\A} {A\!\!\!^\mathrm{\raisebox{1mm}{o}}} %% use \AA instead
\newcommand{\ams} {austriamicrosystems}
\newcommand{\infineon} {infineon}
\newcommand{\amsi} {\ams{} and \infineon}
\newcommand{\kHdimF} {k_\mathrm{0f}} % dimerization of H to H_2, forward
\newcommand{\kHdimR} {k_\mathrm{0r}} % dimerization of H to H_2, reverse
\newcommand{\kHdimFz} {k_\mathrm{0f,0}} % dimerization of H to H_2, forward
\newcommand{\kHdimRz} {k_\mathrm{0r,0}} % dimerization of H to H_2, reverse
\newcommand{\EHdimF} {E_\mathrm{0f}} % dimerization of H to H_2, forward
\newcommand{\EHdimR} {E_\mathrm{0r}} % dimerization of H to H_2, reverse
\newcommand{\kHcrackEpbetaF} {k_\mathrm{1f}} % cracking of H_2 on Epbeta, forward
\newcommand{\kHcrackEpbetaR} {k_\mathrm{1r}} % cracking of H_2 on Epbeta, reverse
\newcommand{\kPassEpbetaF} {k_\mathrm{2f}} % passivation of Epbeta
\newcommand{\kPassEpbetaR} {k_\mathrm{2r}} % activation of Epbeta
\newcommand{\kActEpbeta} {k_x} % activation of Epbeta
\newcommand{\kActPbF} {k_\mathrm{3f}} % direct depassivation of Pb, forward
\newcommand{\kActPbR} {k_\mathrm{3r}} % direct depassivation of Pb, reverse
\newcommand{\kHActPbF} {k_\mathrm{4f}} % depassivation of Pb through H, forward
\newcommand{\kHActPbR} {k_\mathrm{4r}} % depassivation of Pb, reverse (passivation with H2)
\newcommand{\kHpActPbF} {k_\mathrm{5f}} % depassivation of Pb through H+, forward
\newcommand{\kHpActPbR} {k_\mathrm{5r}} % depassivation of Pb, reverse (passivation with H2)
\newcommand{\kpChargeSSiHF} {k_\mathrm{6f}} % charging of SSiH by a hole
\newcommand{\kpChargeSSiHR} {k_\mathrm{6r}} % donation of a hole by SSiHp
\newcommand{\knChargeSSiHF} {k_\mathrm{7f}} % donation of an electron by SSiH
\newcommand{\knChargeSSiHR} {k_\mathrm{7r}} % neutralization of SSiHp by an electron
\newcommand{\kSSiHpDonateHpF}{k_\mathrm{8f}} % donation of Hp from SSiHp
\newcommand{\kSSiHpDonateHpR}{k_\mathrm{8r}} % capture of Hp through SSi
\newcommand{\kpChargeSSiF} {k_\mathrm{9f}} % charging of SSi by a hole
\newcommand{\kpChargeSSiR} {k_\mathrm{9r}} % donation of a hole by SSip
\newcommand{\knChargeSSiF} {k_\mathrm{10f}} % donation of an electron by SSi
\newcommand{\knChargeSSiR} {k_\mathrm{10r}} % neutralization of SSip by an electron
\newcommand{\kHcrackSSiF} {k_\mathrm{11f}} % cracking of H_2 on SSi, forward
\newcommand{\kHcrackSSiR} {k_\mathrm{11r}} % cracking of H_2 on SSi, reverse
\newcommand{\tls} {|_\mathrm{1}}
\newcommand{\trs} {|_\mathrm{2}}
\newcommand{\tis} {|_i}
\newcommand{\tjs} {|_j}
\newcommand{\KF}[2] {k_{\mathrm{f}#1#2}}
\newcommand{\KR}[2] {k_{\mathrm{r}#1#2}}
%\newcommand{\R}[1] {r_{#1}}
\newcommand{\concl}[1] {[#1]_1}
\newcommand{\concr}[1] {[#1]_2}
\newcommand{\conci}[1] {[#1]_i}
\newcommand{\concj}[1] {[#1]_j}
\newcommand{\EI} {E_\mathrm{I}}
\newcommand{\HB} {H_\mathrm{B}}
\newcommand{\Hit} {H_\mathrm{it}}
%\newcommand{\Eit} {E_\mathrm{it}}
\newcommand{\Vstress} {V_\mathrm{stress}}
\newcommand{\Vrelax} {V_\mathrm{relax}}
\newcommand{\Vread} {V_\mathrm{r}}
\newcommand{\rect} {\mathrm{rect}}
\newcommand{\us} {H} %% unit step
\newcommand{\gs} {\tilde{g}}
\newcommand{\gse} {\tilde{g}_\mathrm{e}}
\newcommand{\gsc} {\tilde{g}_\mathrm{c}}
\newcommand{\Gsc} {\tilde{G}_\mathrm{c}}
\newcommand{\gammac} {\gamma} %_\mathrm{c}}
\newcommand{\nrgc} {\mathcal{E}_\mathrm{c}}
\newcommand{\nrge} {\mathcal{E}_\mathrm{e}}
\newcommand{\anrgc} {\bar{\mathcal{E}}_\mathrm{c}}
\newcommand{\snrgc} {\sigma_\mathrm{c}}
\newcommand{\fG} {f_\mathrm{g}}
\newcommand{\FG} {F_\mathrm{g}}
\newcommand{\fLog} {f_\mathrm{l}}
\newcommand{\FLog} {F_\mathrm{l}}
\newcommand{\pdf} {p.d.f.}
\newcommand{\cdf} {c.d.f.}
\newcommand{\tsrelax} {\tilde{t}_\mathrm{r}}
\newcommand{\tsz} {\tilde{t}_\mathrm{0}}
\newcommand{\UA} {V_1}
\newcommand{\UB} {V_2}
\newcommand{\UBV} {V_2^\mathrm{V}}
\newcommand{\UBC} {V_2^\mathrm{C}}
\renewcommand{\UC} {V_3}
\newcommand{\Ui} {V_i}
\newcommand{\Uj} {V_j}
\newcommand{\qi} {q_i}
\newcommand{\qj} {q_j}
\newcommand{\qA} {q_1}
\newcommand{\qAs}{q_{1'}}
\newcommand{\qB} {q_2}
\newcommand{\qBA}{q_{21}}
\newcommand{\qBs} {q_{2'}}
\newcommand{\qC} {q_{1'}}
\newcommand{\EAB} {E_{12}}
\newcommand{\EBA} {E_{21}}
\newcommand{\EABB} {E_{12}^\mathrm{B}}
\newcommand{\EBAB} {E_{21}^\mathrm{B}}
\newcommand{\lci} {\alpha_{i}}
\newcommand{\lcA} {\alpha_{1}}
\newcommand{\lcC} {\alpha_{3}}
\newcommand{\eBi} {\nrg_{\mathrm{B}i}}
\newcommand{\eji} {\nrg_{ji}}
\newcommand{\eBA} {\nrg_\mathrm{B1}}
\newcommand{\eBC} {\nrg_\mathrm{B3}}
\newcommand{\eA} {\nrg_{1}}
\newcommand{\eC} {\nrg_{3}}
\newcommand{\domegaji} {\Delta \omega_{ji}}
\newcommand{\domegaBA} {\Delta \omega_{21}}
\newcommand{\domegaBC} {\Delta \omega_{23}}
\newcommand{\Tstar} {T^*}
\newcommand{\Tc} {T_\mathrm{c}}
\newcommand{\Te} {T_\mathrm{e}}
\newcommand{\epsAB} {\nrg_{12}}
\newcommand{\epsABo}{\nrg_{12}^\circ}
\newcommand{\epsABs}{\nrg_{12'}}
\newcommand{\epsBsA}{\nrg_{2'1}}
\newcommand{\epsBsB}{\nrg_{2'2}}
\newcommand{\epsBBs}{\nrg_{22'}}
\newcommand{\epsBA} {\nrg_{21}}
\newcommand{\epsBAo}{\nrg_{21}^\circ}
\newcommand{\epsBC} {\nrg_{21'}}
\newcommand{\epsCB} {\nrg_{1'2}}
\newcommand{\epsAC} {\nrg_{11'}}
\newcommand{\epsCA} {\nrg_{1'1}}
\newcommand{\epsij} {\nrg_{ij}}
\newcommand{\epsTi} {\nrg_{\mathrm{T}i}}
\newcommand{\epsTj} {\nrg_{\mathrm{T}j}}
\newcommand{\epsTA} {\nrg_{\mathrm{T}1}}
\newcommand{\epsTBs}{\nrg_{\mathrm{T}2'}}
\newcommand{\epsTB} {\nrg_{\mathrm{T}2}}
\newcommand{\epsTC} {\nrg_{\mathrm{T}1'}}
\newcommand{\epsTAV} {\nrg_{\mathrm{T}1}^\mathrm{V}}
\newcommand{\epsTBV} {\nrg_{\mathrm{T}2}^\mathrm{V}}
\newcommand{\epsTCV} {\nrg_{\mathrm{T}1'}^\mathrm{V}}
\newcommand{\epsTAC} {\nrg_{\mathrm{T}1}^\mathrm{C}}
\newcommand{\epsTBC} {\nrg_{\mathrm{T}2}^\mathrm{C}}
\newcommand{\epsTCC} {\nrg_{\mathrm{T}1'}^\mathrm{C}}
\newcommand{\VGL} {V_\mathrm{G}^\mathrm{L}}
\newcommand{\VGH} {V_\mathrm{G}^\mathrm{H}}
\newcommand{\rL} {r^\mathrm{L}}
\newcommand{\rH} {r^\mathrm{H}}
\newcommand{\tmax} {t_\mathrm{max}}
\newcommand{\eps} {\epsilon}
\newcommand{\epsz} {\epsilon_0}
\newcommand{\epsr} {\epsilon_\mathrm{r}}
\newcommand{\muc}{\mu_\mathrm{c}}
\newcommand{\mue}{\mu_\mathrm{e}}
\newcommand{\mucz}{\mu_\mathrm{c0}}
\newcommand{\muez}{\mu_\mathrm{e0}}
%\newcommand{\fmuc}{f_\mu} % _\mathrm{c}}
\newcommand{\fmuc}{k} % _\mathrm{c}}
\newcommand{\fmue}{\fmuc} % \Delta \mu_\mathrm{e}}
\newcommand{\dmue}{\Delta \mu_\mathrm{e}}
\newcommand{\dmuez}{\Delta \mu_\mathrm{e0}}
\newcommand{\sigmac}{\sigma_\mathrm{c}}
\newcommand{\sigmae}{\sigma_\mathrm{e}}
\newcommand{\mucr}{\mu_\mathrm{cr}}
\newcommand{\muer}{\mu_\mathrm{er}}
\newcommand{\dmuer}{\Delta \mu_\mathrm{er}}
\newcommand{\sigmacr}{\sigma_\mathrm{cr}}
\newcommand{\sigmaer}{\sigma_\mathrm{er}}
\newcommand{\mucp}{\mu_\mathrm{cp}}
\newcommand{\muep}{\mu_\mathrm{ep}}
\newcommand{\dmuep}{\Delta \mu_\mathrm{ep}}
\newcommand{\sigmacp}{\sigma_\mathrm{cp}}
\newcommand{\sigmaep}{\sigma_\mathrm{ep}}
\newcommand{\Ar}{A_\mathrm{r}}
\newcommand{\tauzr}{\tau_\mathrm{0r}}
\newcommand{\Ap}{A_\mathrm{p}}
\newcommand{\tauzp}{\tau_\mathrm{0p}}
\newcommand{\Fr}{F_\mathrm{r}}
\newcommand{\Fp}{F_\mathrm{p}}
\newcommand{\frec} {f_\mathrm{r}}
\newcommand{\fperm}{f_\mathrm{p}}
\newcommand{\Fg}{F_\mathrm{\gamma}}
\newcommand{\Gg}{G_\mathrm{\gamma}}
\newcommand{\Fgr}{F_\mathrm{\gamma r}}
\newcommand{\Fgp}{F_\mathrm{\gamma p}}
\newcommand{\ag}{a_\gamma}
\newcommand{\vgs} {\vect{g}_\mathrm{s}}
\newcommand{\vgr} {\vect{g}_\mathrm{r}}
\newcommand{\vhs} {\vect{h}_\mathrm{s}}
\newcommand{\vhr} {\vect{h}_\mathrm{r}}
\newcommand{\tKs} {\tens{K}_\mathrm{s}}
\newcommand{\tKr} {\tens{K}_\mathrm{r}}
\newcommand{\tPs} {\tens{P}_\mathrm{s}}
\newcommand{\tPr} {\tens{P}_\mathrm{r}}
\newcommand{\tLambdas} {\tens{\Lambda}_\mathrm{s}}
\newcommand{\tLambdar} {\tens{\Lambda}_\mathrm{r}}
\newcommand{\tpulse} {t_\mathrm{p}}
\newcommand{\Vp} {V_\mathrm{p}}
\newcommand{\taup} {\tau_\mathrm{p}}
\newcommand{\unitmu}{\hbox{\textmu}}
\newcommand{\persname}[1]{#1}
\newcommand{\srh}{\persname{Shockley-Read-Hall}}
\newcommand{\CT}{\persname{Caughey} and \persname{Thomas}}
% References to Equations, Tables, Figures, Sections
\newcommand{\Eq}[1] {(\ref{#1})}
\newcommand{\eq}[1] {(\ref{#1})}
\newcommand{\longeq}[1] {equation (\ref{#1})}
\newcommand{\longeqs}[2] {equations (\ref{#1}) and (\ref{#2})}
\newcommand{\eqs}[2] {equations (\ref{#1}) and (\ref{#2})}
\newcommand{\Eqs}[2] {Equations (\ref{#1}) and (\ref{#2})}
\newcommand{\Fig}[1] {Fig.~\ref{#1}}
\newcommand{\Figt}[2] {Figs.~\ref{#1}--\ref{#2}}
\newcommand{\Figs}[2] {Figs.~\ref{#1} and \ref{#2}}
\newcommand{\Figab}[2] {Fig.\ \ref{#1}(#2)}
\newcommand{\Sec}[1] {Section~\ref{#1}}
\newcommand{\Chapter}[1] {Chapter~\ref{#1}}
\newcommand{\Appendix}[1]{Appendix~\ref{#1}}
\newcommand{\Table}[1] {Table~\ref{#1}}
\newcommand{\ea}[1] {#1 \textit{et al.}}
\newcommand{\eac}[2] {\ea{#1}~\cite{#2}}
\newcommand{\Lg}[1] {\ensuremath{L_{\mathrm{g}} =#1\,\munit{{\unitmu}m}}}
\newcommand{\Lgn}[1] {\ensuremath{L_{\mathrm{g}} =#1\,\munit{nm}}}
\newcommand{\LD} {L_{\mathrm{D}}}
\newcommand{\ds} {\displaystyle}
\newcommand{\ts} {\textstyle}
\newcommand{\sfrac}[2] {{\ts \frac{#1}{#2}}}
\renewcommand{\d}{\ensuremath{\mathrm{d}}}
% Nabla - do NOT use directly
\newcommand{\Nabla} {\ensuremath{\vect{\nabla}}}
\newcommand{\Varnabla}[1] {\ensuremath{\vect{\nabla}_{\negmedspace \vect{#1}}}}
\newcommand{\Nablak} {\begin{pmatrix} \frac{\partial}{\partial k_x} \\ \frac{\partial}{\partial k_y} \\ \frac{\partial}{\partial k_z} \end{pmatrix}}
\newcommand{\totder}[2] {\ensuremath{\frac{\d #2}{\d #1}}}
\newcommand{\fermii}{{\cal F}}
%% % Divergenz
%% \newcommand{\DIV}[1]{\ensuremath{\NABLA \inpr #1}}
%% \newcommand{\VARDIV}[2]{\ensuremath{\VARNABLA{#1} \inpr #2}}
%% % Rotation
%% \newcommand{\ROT}[1]{\ensuremath{\NABLA \expr #1}}
%% \newcommand{\VARROT}[2]{\ensuremath{\VARNABLA{#1} \expr #2}}
%% % Gradient of a scalar
%% \newcommand{\GRAD}[1]{\ensuremath{\NABLA \, #1}}
%% \newcommand{\VARGRAD}[2]{\ensuremath{\VARNABLA{#1} \, #2}}
%% % Gradient of a vector or tensor
%% \newcommand{\TENSGRAD}[1]{\ensuremath{\NABLA \otimes #1}}
%% \newcommand{\VARTENSGRAD}[2]{\ensuremath{\VARNABLA{#1} \otimes #2}}
%% % Laplace Operator
%% \newcommand{\LAPLACE}[1]{\ensuremath{\nabla^2} \, #1}
%% \newcommand{\VARLAPLACE}[2]{\ensuremath{\nabla^2_{\negmedspace #1} \, #2}}
\newcommand{\Div} {{\mathrm{div}}}
\newcommand{\Divd} {{\mathrm{div_d}}}
\newcommand{\Grad} {{\mathrm{grad}}}
\newcommand{\Curl} {{\mathrm{curl}}}
\newcommand{\Vargrad}[2]{\ensuremath{\Varnabla{#1} #2}}
\newcommand{\Vartensgrad}[2]{\ensuremath{\Varnabla{#1} \otimes #2}}
\newcommand{\Gradd} {{\mathrm{grad_d}}}
\newcommand{\dt} {\partial_t}
\newcommand{\D} {\Delta}
\newcommand{\erf} {\mathrm{erf}}
\newcommand{\erfc} {\mathrm{erfc}}
\newcommand{\inpr} {\cdot}
% vector product
\newcommand{\expr} {\ensuremath{\times}}
% tensor product
\newcommand{\tenspr} {\ensuremath{\otimes}}
%\newcommand{\T} {{\mathbf{T}}}
\newcommand{\T} {\mathrm{T}}
\newcommand{\FD}[2] {\frac{{\mathrm{d}} #1}{{\mathrm{d}} #2}}
\newcommand{\PD}[2] {\frac{\partial #1}{\partial #2}}
\newcommand{\PDS}[2] {\frac{\partial^2 #1}{\partial {#2}^2}}
\newcommand{\PDt}[1] {\partial_t #1}
\newcommand{\PDdot}[1] {\dot{#1}}
% Vector
%\newcommand{\vect}[1] {\ensuremath{{\boldsymbol{{\mathrm{#1}}}}}}
%\newcommand{\vect}[1] {\ensuremath{{\boldsymbol{{\mathrm{#1}}}}}}
\newcommand{\col}[1] {\mathrm{col}_{#1} \, }
\newcommand{\vect}[1] {{\mathbf{#1}}}
% some frequently used vectors
\newcommand{\vnrg} {\pmb{\nrg}}
\newcommand{\vpsi} {\pmb{\psi}}
\newcommand{\vLambda} {\vect{\Lambda}}
\newcommand{\vlambda} {\pmb{\lambda}}
\newcommand{\vomega} {\pmb{\omega}}
\newcommand{\lambdaZ} {\lambda^{(0)}}
\newcommand{\lambdaT} {\lambda^{(2)}}
\newcommand{\vPsi} {\vect{\Psi}}
\newcommand{\vA} {\vect{A}}
\newcommand{\vB} {\vect{B}}
\newcommand{\vBh} {\vect{B}_\mathrm{h}}
\newcommand{\vBc} {\vect{B}_\mathrm{c}}
\newcommand{\vD} {\vect{D}}
\newcommand{\vH} {\vect{H}}
\newcommand{\vc} {\vect{c}}
\newcommand{\vE} {\vect{E}}
\newcommand{\vF} {\vect{F}}
\newcommand{\vFnu} {\vect{F}_{\nu}}
\newcommand{\vFn} {\vect{F}_{n}}
\newcommand{\vFp} {\vect{F}_{p}}
\newcommand{\vGam} {\vect{\Gamma}}
\newcommand{\vg} {\vect{g}}
\newcommand{\vh} {\vect{h}}
\newcommand{\vk} {\vect{k}}
\newcommand{\vkz} {\vect{k}_0}
\newcommand{\vn} {\vect{n}}
\newcommand{\vp} {\vect{p}}
\newcommand{\vpd} {\vect{p}_\mathrm{d}}
\newcommand{\vpc} {\vect{p}_\mathrm{c}}
\newcommand{\vq} {\vect{q}}
\newcommand{\vKh} {\vect{K}_\mathrm{h}}
\newcommand{\vKc} {\vect{K}_\mathrm{c}}
%\newcommand{\vr} {\vect{r}}
\newcommand{\vu} {\vect{u}}
\newcommand{\vU} {\vect{U}}
\newcommand{\vuc} {\vect{u}_\mathrm{c}}
\newcommand{\vv} {\vect{v}}
\newcommand{\vvi} {\vect{v}\!}
\newcommand{\vvd} {\vect{v}_\mathrm{d}}
\newcommand{\vVd} {\vect{V}_\mathrm{d}}
\newcommand{\vvn} {\vect{v}_n}
\newcommand{\vvp} {\vect{v}_p}
\newcommand{\vVn} {\vect{V}_n}
\newcommand{\vVp} {\vect{V}_p}
\newcommand{\vV} {\vect{V}}
\newcommand{\vVi} {\vect{V}\!}
\newcommand{\Vi} {{V}\!}
\newcommand{\vViO}[1] {\vect{V}^{(#1)}_{\!i}}
\newcommand{\vP} {\vect{P}}
\newcommand{\vM} {\vect{M}}
\newcommand{\vPi} {\vect{P}\!}
\newcommand{\vvnu} {\vect{v}_{\nu}}
\newcommand{\vx} {\vect{x}}
\newcommand{\vxi} {\pmb{\xi}}
\newcommand{\vy} {\vect{y}}
\newcommand{\vX} {\vect{X}}
\newcommand{\vsat} {v_\mathrm{sat}}
\newcommand{\vth} {v_\mathrm{th}}
\newcommand{\pth} {p_\mathrm{th}}
\newcommand{\VT} {V_\mathrm{T}}
\newcommand{\Vth} {V_\mathrm{th}}
\newcommand{\IDth} {I_\mathrm{D,th}}
\newcommand{\ICP} {I_\mathrm{CP}}
\newcommand{\ICPz} {I_\mathrm{CP,0}}
\newcommand{\ICPMax} {I_\mathrm{CP}^\mathrm{max}}
\newcommand{\dICP} {\Delta \ICP}
\newcommand{\dVth} {\Delta \! \Vth}
\newcommand{\dVthOTF} {\Delta \Vth^\mathrm{OTF}}
\newcommand{\dVthMax} {\Delta \Vth^\mathrm{max}}
\newcommand{\dVthit} {\Delta \Vth^\mathrm{it}}
\newcommand{\dVthox} {\Delta \Vth^\mathrm{ox}}
\newcommand{\Vthz} {V_\mathrm{th0}}
\newcommand{\VD} {V_\mathrm{D}}
\newcommand{\VDD} {V_\mathrm{DD}}
\newcommand{\VS} {V_\mathrm{S}}
\newcommand{\VDS} {V_\mathrm{DS}}
\newcommand{\VF} {V_\mathrm{F}}
\newcommand{\VG} {V_\mathrm{G}}
\newcommand{\VB} {V_\mathrm{B}}
\newcommand{\VH} {V_\mathrm{H}}
\newcommand{\VL} {V_\mathrm{L}}
\newcommand{\Vbi} {\varphi_\mathrm{bi}}
\newcommand{\ID} {I_\mathrm{D}}
\newcommand{\IS} {I_\mathrm{S}}
\newcommand{\dID} {\Delta \ID}
\newcommand{\dIS} {\Delta \IS}
\newcommand{\IG} {I_\mathrm{G}}
\newcommand{\IDsat} {I_\mathrm{Dsat}}
\newcommand{\IDlin} {I_\mathrm{Dlin}}
\newcommand{\IDz} {I_\mathrm{D0}}
\newcommand{\dIDlin} {\Delta I_\mathrm{Dlin}}
\newcommand{\IB} {I_\mathrm{B}}
\newcommand{\IC} {I_\mathrm{C}}
\newcommand{\VC} {V_\mathrm{C}}
%\newcommand{\vect}[1]{\ensuremath{\vec{#1}}}
% Matrix
\newcommand{\matr}[1] {\ensuremath{{\boldsymbol{\mathrm{#1}}}}} % \underline oder \mathbf
% Tensor
%\newcommand{\tens}[1] {\ensuremath{{\mathbb{#1}}}} % \widetilde oder \mathcal
\newcommand{\tens}[1] {\ensuremath{{\mathrm{\hat{#1}}}}} % \widetilde oder \mathcal
%\newcommand{\tens}[1] {\ensuremath{{\mathbb{#1}}}} % \widetilde oder \mathcal
% statistical average
\newcommand{\statav} [1]{\ensuremath{\langle #1 \rangle}}
\newcommand{\statavO} [2]{\ensuremath{\langle #1 \rangle}^{(#2)}}
\newcommand{\statavOh} [2]{\ensuremath{\langle #1 \rangle}^{(#2)}_\mathrm{h}}
\newcommand{\sstatav} [1]{\ensuremath{\langle \! \langle #1 \rangle \! \rangle}}
\newcommand{\sstatavO} [2]{\ensuremath{\langle \! \langle #1 \rangle \! \rangle}^{(#2)}}
\newcommand{\sstatavOh}[2]{\ensuremath{\langle \! \langle #1 \rangle \! \rangle}^{(#2)}_\mathrm{h}}
\newcommand{\sstatavOc}[2]{\ensuremath{\langle \! \langle #1 \rangle \! \rangle}^{(#2)}_\mathrm{c}}
\newcommand{\stataveq} [1]{\ensuremath{\langle #1 \rangle_\mathrm{eq}}}
\newcommand{\Statav} [1]{\ensuremath{\Bigl\langle #1 \Bigr\rangle}}
\newcommand{\SStatav} [1]{\ensuremath{\Bigl\langle \! \! \Bigl\langle #1 \Bigr\rangle \! \! \Bigr\rangle}}
% intergration over k-space
\newcommand{\ddk} {\,\, \mathrm{d}^3 \vect{k}}
\newcommand{\ddp} {\,\, \mathrm{d}^3 \vect{p}}
%\newcommand{\ddk} {\,\, \mathrm{d} \vect{k}}
%\newcommand{\ddp} {\,\, \mathrm{d} \vect{p}}
\newcommand{\kint}[1]{\ensuremath{\int #1 \ddk}}
% Bereichsintegral "uber den k'-Raum
\newcommand{\ksint}[1]{\ensuremath{\int #1 \,\, \d^3 \vk'}}
% Bereichsintegral "uber den k''-Raum
\newcommand{\kssint}[1]{\ensuremath{\int #1 \,\, \d^3 \vk''}}
% intergration over E-space
\newcommand{\Eint}[1]{\ensuremath{\int #1 \,\, {\mathrm{d}} \nrg}}
\newcommand{\infint} {\mathop{\int}_{-\infty}^{\infty}}
\newcommand{\gint}[2] {\ds \mathop{\int}_{#1}^{#2}}
% bestimmtes 3fach Integral "uber dk_x dk_y dk_z von -8 bis 8
\newcommand{\bestkiiint}[1]{\ensuremath{\iiint\limits_{-\infty}^{\ \ \ \infty} #1 \,\, \d k_x \, \d k_y \, \d k_z}}
\newcommand{\Bestkiiint}[1]{\ensuremath{\iiint\limits_{-\infty}^{\ \ \ \infty} #1 \,\, \d^3 \vk}}
\newcommand{\sign} {\, \, \mathrm{sign}\,}
\newcommand{\FA} {F_\mathrm{\!A}}
\newcommand{\FAh} {F_\mathrm{\!A,h}}
\newcommand{\FAc} {F_\mathrm{\!A,c}}
\newcommand{\FAI} {F_\mathrm{\!AI}}
\newcommand{\FI} {F_\mathrm{I}}
\newcommand{\UTens} {\tens{I}}
\newcommand{\tA} {\tens{A}}
\newcommand{\tI} {\tens{I}}
\newcommand{\tJ} {\tens{J}}
\newcommand{\tL} {\tens{L}}
\newcommand{\tLambda} {\tens{\Lambda}}
\newcommand{\tF} {\tens{F}}
\newcommand{\tFh} {\tens{F}_\mathrm{h}}
\newcommand{\tFc} {\tens{F}_\mathrm{c}}
\newcommand{\tK} {\tens{K}}
\newcommand{\tG} {\tens{G}}
\newcommand{\tH} {\tens{H}}
\newcommand{\tB} {\tens{B}}
\newcommand{\tC} {\tens{C}}
\newcommand{\tP} {\tens{P}}
\newcommand{\tQ} {\tens{Q}}
\newcommand{\tS} {\tens{S}}
\newcommand{\tY} {\tens{Y}}
\newcommand{\tZ} {\tens{Z}}
\newcommand{\tEps} {\tens{\epsilon}}
%\newcommand{\Tr}[1] {{\mathrm{tr}\, (#1)}}
\newcommand{\Tr}[1] {{\mathrm{tr}\, #1}}
\newcommand{\F}[2] {F_{#1/2}(#2)} % Fermi Integral
\newcommand{\NS} {{N_{\mathrm{I}}}}
\newcommand{\NA} {{N_{\mathrm{A}}}}
\newcommand{\ND} {{N_{\mathrm{D}}}}
\newcommand{\NAm} {{N_{\mathrm{A}}^-}}
\newcommand{\NDp} {{N_{\mathrm{D}}^+}}
\newcommand{\Ni} {n_{\mathrm{i}}}
\newcommand{\NI} {N_{\mathrm{I}}}
\newcommand{\Nc} {N_{\mathrm{c}}}
\newcommand{\Nv} {N_{\mathrm{v}}}
\newcommand{\Nt} {N_{\mathrm{t}}}
\newcommand{\gc} {g_{\mathrm{c}}}
\newcommand{\dc} {d_{\mathrm{c}}}
\newcommand{\EF} {E_{\mathrm{F}}}
\newcommand{\Evac} {E_\mathrm{vac}}
\newcommand{\EFone} {E_{\mathrm{F1}}}
\newcommand{\EFtwo} {E_{\mathrm{F2}}}
\newcommand{\EFn} {E_{\mathrm{F}n}}
\newcommand{\EFp} {E_{\mathrm{F}p}}
\newcommand{\EFM} {E_{\mathrm{FM}}}
\newcommand{\EFS} {E_{\mathrm{FS}}}
\newcommand{\Ei} {E_{\mathrm{i}}}
\newcommand{\Eiz} {E_{\mathrm{i,0}}}
\newcommand{\Et} {E_{\mathrm{t}}}
\newcommand{\Eg} {E_{\mathrm{g}}}
\newcommand{\Ecz} {E_{\mathrm{c,0}}}
\newcommand{\Etot} {E_{\mathrm{tot}}}
\newcommand{\Ec} {E_{\mathrm{c}}}
\newcommand{\EcF} {E_{\mathrm{cF}}}
\newcommand{\Eci}[1] {E_{\mathrm{c}#1}}
\newcommand{\Ev} {E_{\mathrm{v}}}
\newcommand{\EvF} {E_{\mathrm{vF}}}
\newcommand{\Evz} {E_{\mathrm{v,0}}}
\newcommand{\gv} {g_{\mathrm{v}}}
\newcommand{\Evi}[1] {E_{\mathrm{v#1}}}
\newcommand{\Enu} {E_{\nu}}
\newcommand{\Enuz} {E_{\nu,0}}
\newcommand{\Emax} {E_{\mathrm{max}}}
\newcommand{\Is} {I_{\mathrm{s}}}
\newcommand{\Hnu} {H_{\nu}}
\newcommand{\bn} {\beta_n}
\newcommand{\bnu} {\beta_\nu}
% tensors
\newcommand{\tU} {\tens{U}}
\newcommand{\tUO}[1] {\tens{U}^{(#1)}}
\newcommand{\UO}[1] {U^{(#1)}}
\newcommand{\tUPO}[1] {\tens{U}^{\mathrm{P}(#1)}}
\newcommand{\tUs} {\tens{U}^*}
\newcommand{\tD} {\tens{D}}
\newcommand{\tFlux} {\tens{\Psi}}
\newcommand{\tR} {\tens{R}}
\newcommand{\tRm} {\tens{R}^\mathrm{m}}
\newcommand{\tT} {\tens{T}}
\newcommand{\tM} {\tens{M}}
\newcommand{\tV} {\tens{V}}
\newcommand{\tw} {\tens{w}}
\newcommand{\tW} {\tens{W}}
\newcommand{\tX} {\tens{X}}
\newcommand{\tTO}[1] {\tens{T}^{(#1)}}
\newcommand{\nO}[1] {n^{(#1)}}
\newcommand{\wO}[1] {w^{(#1)}}
% vectors
\newcommand{\vC} {\vect{C}}
\newcommand{\va} {\vect{a}}
\newcommand{\vb} {\vect{b}}
\newcommand{\ve} {\vect{e}}
\newcommand{\vex} {\vect{e}_x}
\newcommand{\vf} {\vect{f}}
\newcommand{\vfz} {\vect{f}_0}
\newcommand{\vz} {\vect{z}}
\newcommand{\vzero} {\vect{0}}
\newcommand{\vQ} {\vect{Q}}
\newcommand{\vQs} {\vQ^\mathrm{*}}
\newcommand{\vQdiff} {\vect{Q}_{\mathrm{diff}}}
\newcommand{\vQconv} {\vect{Q}_{\mathrm{conv}}}
\newcommand{\vS} {\vect{S}}
\newcommand{\vs} {\vect{s}}
\newcommand{\vr} {\vect{r}}
\newcommand{\vSn} {\vect{S}_{n}}
\newcommand{\vSnu} {\vect{S}_{\nu}}
\newcommand{\vJ} {\vect{J}}
\newcommand{\vJn} {\vect{J}_{n}}
\newcommand{\vJp} {\vect{J}_{p}}
\newcommand{\vJnDrift} {\vect{J}_{n}^\mathrm{Drift}}
\newcommand{\vJpDrift} {\vect{J}_{p}^\mathrm{Drift}}
\newcommand{\vJnDiff} {\vect{J}_{n}^\mathrm{Diff}}
\newcommand{\vJpDiff} {\vect{J}_{p}^\mathrm{Diff}}
\newcommand{\vJnu} {\vect{J}_{\nu}}
\newcommand{\vK} {\vect{K}}
\newcommand{\vKnu} {\vect{K}_{\nu}}
\newcommand{\vR} {\vect{R}}
\newcommand{\four}[1] {{\mathscr{F}\{#1\}}}
\newcommand{\fourI}[1] {{\mathscr{F}^{-1}\{#1\}}}
\newcommand{\Four}[1] {{\mathscr{F}\Bigl\{#1\Bigr\}}}
\newcommand{\FourI}[1] {{\mathscr{F}^{-1}\Bigl\{#1\Bigr\}}}
\newcommand{\Bern} {{\mathscr{B}}}
\newcommand{\Jnuk} {{J_{\nu}}_k}
\newcommand{\Snuk} {{S_{\nu}}_k}
\newcommand{\E} {\vect{E}}
\newcommand{\Tt} {T_{e}} %% thermodynamic temperature after tg0167
\newcommand{\Tn} {T_{n}}
\newcommand{\Tnu} {T_{\nu}}
\newcommand{\Tnui} {{\Tnu}_i}
\newcommand{\Tnuj} {{\Tnu}_j}
\newcommand{\nui} {\nu_i}
\newcommand{\nuj} {\nu_j}
\newcommand{\kL} {\kappa_{\mathrm{L}}}
\newcommand{\TL} {T_{\mathrm{L}}}
\newcommand{\TC} {T_{\mathrm{C}}}
\newcommand{\kB} {{\mathrm{k_B}}}
\newcommand{\kb}[1] {\mathrm{k}^{#1}_\mathrm{B}}
\newcommand{\Thnu} {\Theta_\nu}
\newcommand{\Thnui} {{\Thnu}_i}
\newcommand{\Thnuj} {{\Thnu}_j}
\newcommand{\mnu} {m_\nu}
\newcommand{\mn} {m_n^*}
\newcommand{\mz} {m_0}
\newcommand{\pot} {\psi}
\newcommand{\potL} {\psi_\mathrm{L}}
\newcommand{\potR} {\psi_\mathrm{R}}
\newcommand{\poti} {\psi_i}
\newcommand{\potj} {\psi_j}
\newcommand{\nrg} {{\mathcal{E}}}
\newcommand{\nrgref} {{\mathcal{E}}_\mathrm{ref}}
\newcommand{\nrgnu} {{\mathcal{E}_\nu}}
\newcommand{\taue} {\tau_\nrg}
\newcommand{\av}[1] {\ensuremath{\tilde #1}}
\newcommand{\atxi} {\Bigl|_\vect{x}_i \Bigr.}
\newcommand{\atyk} {\Bigl|_\vect{y}_k \Bigr.}
\newcommand{\mc} {m}
%\newcommand{\tm} {\tau_m}
%\newcommand{\tS} {\tau_S}
\newcommand{\snu} {s_\nu}
\newcommand{\munu} {\mu_\nu}
\newcommand{\mus} {\mu_S}
\newcommand{\q} {{\mathrm{q}}}
\newcommand{\funct}[1] {{\mathcal{#1}}}
\newcommand{\B} {\funct{B}} %% Bernoulli function
\newcommand{\HA} {\funct{H}} %% Harmonic average
\newcommand{\HTnu} {\overline{\Tnu}}
\newcommand{\HThnu} {\overline{\Thnu}}
\newcommand{\Y} {\mathrm{Y}} %% Y function
\newcommand{\Xf} {\mathrm{X}} %% X function
\newcommand{\Z} {\mathrm{Z}} %% Z function
\newcommand{\Zone} {\Z_1} %% Z1 function
\newcommand{\Xone} {\Xf_1} %% X1 function
\newcommand{\Xtwo} {\Xf_2} %% X2 function
\newcommand{\CY} {C_y} %% Y function
\newcommand{\Ord} {O}
\newcommand{\fdfull} {f_\nu(\vect{k}, \vect{r}, t)}
\newcommand{\fdfullp} {f_\nu(\vect{p}, \vect{r}, t)}
\newcommand{\fd} {{f_\nu}}
\newcommand{\fdz} {f_{\nu,0}}
\newcommand{\fdS} {f_{\nu,S}}
\newcommand{\fdA} {f_{\nu,A}}
\newcommand{\feq} {f_\mathrm{eq}}
\newcommand{\weq} {w_\mathrm{eq}}
\newcommand{\fcol} {\frac{\partial f}{\partial t}\Bigr|_\mathrm{coll}}
\newcommand{\wf} {\phi}
\newcommand{\Wf} {\Phi}
\newcommand{\PhiMS} {\Phi_\mathrm{MS}}
\newcommand{\wfi} {\wf_{i}}
\newcommand{\Wfu}[1] {\Wf_{#1}^\vu}
\newcommand{\Wfp}[1] {\Wf_{#1}^\vp}
\newcommand{\Wfui} {\Wfu{i}}
\newcommand{\Wfpi} {\Wfp{i}}
\newcommand{\fh} {f_\mathrm{h}}
\newcommand{\fc} {f_\mathrm{c}}
\newcommand{\Ah} {A_\mathrm{h}}
\newcommand{\Ac} {A_\mathrm{c}}
\newcommand{\fA} {f_\mathrm{A}}
\newcommand{\QA} {Q_\mathrm{A}}
\newcommand{\cA} {c_\mathrm{A}}
% \newcommand{\ac} {a_\mathrm{c}} clashes with some packages?
\newcommand{\fO}[1] {f^{(#1)}}
\newcommand{\fAO}[1] {f_\mathrm{A}^{(#1)}}
\newcommand{\fSO}[1] {f_\mathrm{S}^{(#1)}}
\newcommand{\fnrg} {f_\mathrm{\nrg}}
\newcommand{\fnrgarg} {f_\mathrm{\nrg}(\nrg)}
\newcommand{\fnrgform}{\exp \Bigl[ - \Bigl( \frac{\nrg}{a} \Bigr)^b \Bigr]}
\newcommand{\fM} {f_\mathrm{M}}
\newcommand{\fMform}[1]{\exp \Bigl[ - \frac{\nrg}{#1} \Bigr]}
\newcommand{\fS} {f_\mathrm{S}}
\newcommand{\fSarg} {f_\mathrm{S}(\nrg)}
\newcommand{\fSform} {AA \frgform}
\newcommand{\fI} {f_\mathrm{I}}
\newcommand{\fIarg} {f_\mathrm{I}(\nrg)}
\newcommand{\dnrg} {\,\, \mathrm{d} \nrg}
\newcommand{\dk} {\,\, \mathrm{d} k}
%\renewcommand{\dp} {\,\, \mathrm{d} p}
\newcommand{\dtheta} {\,\, \mathrm{d} \vartheta}
\newcommand{\dphi} {\,\, \mathrm{d} \phi}
\newcommand{\dvp} {\,\, \mathrm{d} \vp}
\newcommand{\dvk} {\,\, \mathrm{d} \vk}
\newcommand{\dvr} {\,\, \mathrm{d} \vr}
\newcommand{\QS} {Q_\mathrm{S}}
\newcommand{\DT} {D^\mathrm{T}}
\newcommand{\mass} {m^*}
\newcommand{\massl} {m^*_\mathrm{l}}
\newcommand{\masst} {m^*_\mathrm{t}}
\newcommand{\masssq} {m^{*2}}
\newcommand{\tmass} {\tens{m}^{-1}}
\newcommand{\anrg} {\alpha \nrg}
\newcommand{\Hu} {H_\mathrm{u}}
\newcommand{\Huarg} {H_\mathrm{u}(\nrg)}
\newcommand{\HU} {H_\mathrm{U}}
\newcommand{\HUarg} {H_\mathrm{U}(\nrg)}
\newcommand{\HUs} {H^*_\mathrm{U}}
\newcommand{\HUsarg} {H^*_\mathrm{U}(\nrg)}
\newcommand{\Hg} {H_\mathrm{g}}
\newcommand{\Hgarg} {H_\mathrm{g}(\nrg)}
\newcommand{\HS} {H_\mathrm{S}}
\newcommand{\HSarg} {H_\mathrm{S}(\nrg)}
\newcommand{\Hadp} {H_\mathrm{adp}}
\newcommand{\Hadparg} {H_\mathrm{adp}(\nrg)}
\newcommand{\Hivspm} {H_\mathrm{ivs}^{\pm}}
\newcommand{\Hivsp} {H_\mathrm{ivs}^+}
\newcommand{\Hivsm} {H_\mathrm{ivs}^-}
%\newcommand{\Hivsarg} {H_\mathrm{ivs}^{\pm}(\anrg)}
\newcommand{\Hnrg} {H_\nrg}
\newcommand{\Hnrgarg} {H_\nrg(\nrg)}
\newcommand{\Hf} {H\!_f}
\newcommand{\Hfarg} {H\!_f(\nrg)}
\newcommand{\HM} {H_\mathrm{M}}
\newcommand{\HMarg} {H_\mathrm{M}(\nrg)}
\newcommand{\Himp} {H_\mathrm{imp}}
\newcommand{\Himparg} {H_\mathrm{imp}(\nrg)}
\newcommand{\e} {\mathrm{e}}
\newcommand{\UC} {U_\mathrm{C}}
\newcommand{\VMS} {V_\mathrm{MS}}
\newcommand{\phiC} {\varphi_\mathrm{C}}
% Names to avoid spelling errors
\newcommand{\Boltzmann} {Boltzmann}
\newcommand{\Hamilton} {Hamilton}
\newcommand{\Hansch} {H{\"a}nsch}
\newcommand{\Cassi} {Cassi}
\newcommand{\Ricco} {Ricc{\`o}}
\newcommand{\Blo} {Bl{\o}tekj{\ae}r}
\newcommand{\Schro} {Schr{\"o}dinger}
\newcommand{\Goes} {Goes}
\newcommand{\rhs} {rhs}
\newcommand{\DD} {drift-diffusion}
\newcommand{\HD} {hydrodynamic}
\newcommand{\ET} {energy-transport}
\newcommand{\NP} {non-parabolic}
% for figures
\newcommand{\fsi}{\textstyle} %% very important text
\newcommand{\fsl}{\scriptstyle} %% less important text
\newcommand{\fsu}{\scriptscriptstyle} %% unimportant text
\newcommand{\unit}[1]{\ensuremath{{\mathrm{#1}}}}
\newcommand{\munit}[1]{\unit{#1}}
\newcommand{\unitS}[1]{\ensuremath{[#1]}}
% Program name
%\newcommand{\progname}[1]{\textsf{#1}}
\newcommand{\progname}[1]{\textsc{#1}}
% Some frequently used program names
\newcommand{\alib}{\progname{Algorithm Library}}
\newcommand{\cgg}{\progname{cgg}}
\newcommand{\cpp}{\progname{C{\tiny++}}}
\newcommand{\emacs}{\progname{Emacs}}
\newcommand{\geo}{\progname{geo2ps}}
\newcommand{\geotops}{\progname{geo2ps}}
\newcommand{\inp}{\progname{INP}}
\newcommand{\ipl}{\progname{IPL}}
\newcommand{\make}{\progname{make}}
\newcommand{\makedevice}{\progname{makedevice}}
\newcommand{\MDL}{\progname{Model Definition Language}}
\newcommand{\mdl}{\progname{MDL}}
\newcommand{\mkdev}{\progname{makedevice}}
\newcommand{\minimos}{\progname{MINIMOS}}
\newcommand{\minimossix}{\progname{MINIMOS~6}}
\newcommand{\mmnt}{\progname{Minimos-NT}}
\newcommand{\vsp}{\progname{VSP}}
\newcommand{\vasp}{\progname{VASP}}
\newcommand{\cptk}{\progname{CP2K}}
\newcommand{\mmtont}{\progname{mm62nt}}
\newcommand{\pbf}{\progname{PBF}}
\newcommand{\pbfm}{\progname{pbfm}}
\newcommand{\pif}{\progname{PIF}}
\newcommand{\pifcopy}{\progname{pifcopy}}
\newcommand{\pifrm}{\progname{pifrm}}
\newcommand{\pifmaid}{\progname{pifmaid}}
\newcommand{\prost}{\progname{prost2d}}
\newcommand{\punch}{\progname{punch}}
\newcommand{\rul}{\progname{RUL}}
\newcommand{\Siesta}{\progname{siesta}}
\newcommand{\sketch}{\progname{sketch}}
\newcommand{\spice}{\progname{Spice}}
\newcommand{\hspice}{\progname{HSpice}}
\newcommand{\spicethree}{\progname{SPICE3}}
\newcommand{\splitseg}{\progname{splitseg}}
\newcommand{\astap}{\progname{ASTAP}}
\newcommand{\svg}{\progname{SVG}}
\newcommand{\svgtops}{\progname{svg2ps}}
\newcommand{\tif}{\progname{TIF}}
\newcommand{\tifwrap}{\progname{tifwrap}}
\newcommand{\tiftopif}{\progname{tif2pbf.sh}}
\newcommand{\tsuprem}{\progname{Tsuprem4}}
\newcommand{\xpif}{\progname{xpif2d}}
\newcommand{\xsvg}{\progname{xsvg}}
\newcommand{\vmake}{\progname{VMAKE}}
\newcommand{\eas}{\progname{EAS}}
\newcommand{\Dessis}{\progname{Dessis}}
\newcommand{\Medici}{\progname{Medici}}
\newcommand{\Atlas}{\progname{Atlas}}
\newcommand{\VMC}{\progname{vmc}}
\newcommand{\VSP}{\progname{vsp}}
\newcommand{\FORTRAN} {\progname{FORTRAN}}
%\newcommand{\C} {\progname{C}}
%\newcommand{\CPP} {\progname{C++}}
\newcommand{\const}{\mathop{\rm const}}
\newcommand{\sio} {SiO$_2$}
\newcommand{\sige}{Si$_\mathrm{1-x}$Ge$_\mathrm{x}$}
\newcommand{\arsinh}{\mathop{\rm arsinh}}
\newcommand{\artan}{\mathop{\rm artan}}
% Name of a person (e.g. Maxwell)
%%\newcommand{\persname}[1]{\textsl{#1}}
\newcommand{\nrgF} {\nrg_\mathrm{F}}
\newcommand{\tauf} {\tau_f}
\newcommand{\taueff} {\tau_\mathrm{eff}}
\newcommand{\fFD} {f_\mathrm{FD}}
\newcommand{\statnrg}[1] {\statav{#1}_\nrg}
\newcommand{\nin} {$n^+$-$n$-$n^+$ test-structure}
\newcommand{\nm}[1] {#1\,\mathrm{nm}}
\newcommand{\dx} {\Delta x}
\newcommand{\dy} {\Delta y}
\newcommand{\mueff} {\mu_\mathrm{eff}}
\newcommand{\mueffz} {\mu_\mathrm{eff0}}
\newcommand{\dmueff} {\Delta \mu_\mathrm{eff}}
\newcommand{\Eeff} {E_\mathrm{eff}}
\newcommand{\Epeff} {E_\mathrm{p,eff}}
\newcommand{\Eneff} {E_\mathrm{n,eff}}
\renewcommand{\deg} {\ensuremath{^\circ}}
\newcommand{\degC} {\ensuremath{^\circ\text{C}}}
\newcommand{\bfdegC} {\ensuremath{^{\pmb\circ}\!\text{C}}}
\newcommand{\U}[2] {\ensuremath{#1\,\mathrm{#2}}}
\newcommand{\Vect}[1] {\{#1\}}
\newcounter{myitem}
\newcommand{\mycnt} {{\it (\roman{myitem})}\addtocounter{myitem}{1}}
\newcommand{\mybcnt} {\textbf{\emph{(\roman{myitem})}\addtocounter{myitem}{1}}}
\newcommand{\resetmycnt}{\setcounter{myitem}{1}}
\resetmycnt
\newcommand{\LFig} {{\bf Left}}
\newcommand{\MFig} {{\bf Middle}}
\newcommand{\RFig} {{\bf Right}}
\newcommand{\TFig} {{\bf Top}}
\newcommand{\BFig} {{\bf Bottom}}
\newcommand{\TLFig} {{\bf Top/Left}}
\newcommand{\BLFig} {{\bf Bottom/Left}}
\newcommand{\TRFig} {{\bf Top/Right}}
\newcommand{\BRFig} {{\bf Bottom/Right}}
\newcommand{\csch} {\text{csch}}
\newcommand{\psiswi} {\psi_\mathrm{s,wi}}
\newcommand{\psissi} {\psi_\mathrm{s,si}}
\newcommand{\psisa} {\psi_\mathrm{s,a}}
\newcommand{\psip} {\psi_\mathrm{p}}
\newcommand{\psistar} {\psi^*}
\newcommand{\VGs} {\VG^*}
\newcommand{\epsSi} {\varepsilon_\mathrm{si}}
\newcommand{\epsOx} {\varepsilon_\mathrm{ox}}
\newcommand{\EOx} {E_\mathrm{ox}}
\newcommand{\ESi} {E_\mathrm{si}}
\newcommand{\Ekin} {{\mathcal{E}}_\mathrm{kin}}
\newcommand{\Etrans} {{\mathcal{E}}_\mathrm{trans}}
\newcommand{\Epot} {\phi} %{\mathcal{E}}_\mathrm{pot}}
\newcommand{\Eelec} {{\mathcal{E}}_\mathrm{elec}}
\newcommand{\Evib} {{\mathcal{E}}_\mathrm{vib}}
\newcommand{\Erot} {{\mathcal{E}}_\mathrm{rot}}
\newcommand{\bra}[1] {\langle #1 |}
\newcommand{\ket}[1] {| #1 \rangle}
\newcommand{\op}[1] {\hat{#1}}
\newcommand{\opH} {\op{H}}
\newcommand{\opT} {\op{T}}
\newcommand{\opTe} {\op{T}_\mathrm{e}}
\newcommand{\opTN} {\op{T}_\mathrm{N}}
\newcommand{\opVee} {\op{V}_\mathrm{ee}}
\newcommand{\opVNN} {\op{V}_\mathrm{NN}}
\newcommand{\opVeN} {\op{V}_\mathrm{eN}}
\newcommand{\opW} {\op{W}}
\newcommand{\opx} {\op{x}}
\newcommand{\opa} {\op{a}}
\newcommand{\opap} {\op{a}^+}
\newcommand{\opam} {\op{a}^-}