\makeatletter
\@ifundefined{HCode}
{\documentclass[screen,CRCHIM,Unicode,biblatex,published]{cedram}
\addbibresource{crchim20250763.bib}
\newenvironment{noXML}{}{}
\def\tsup#1{$^{{#1}}$}
\def\tsub#1{$_{{#1}}$}
\def\thead{\noalign{\relax}\hline}
\def\tbody{\noalign{\relax}}
\def\endthead{\noalign{\relax}\hline}
\RequirePackage{etoolbox}
\usepackage{upgreek}
\def\hyphen{\text{-}}
\def\jobid{crchim20250763}
%\graphicspath{{/tmp/\jobid_figs/web/}}
\graphicspath{{./figures/}}
\newcounter{runlevel}
\let\MakeYrStrItalic\relax
\csdef{Seqnsplit}{\\}
\def\botline{\\\hline}
\def\refinput#1{}
\def\back#1{}
\def\ubreak{\unskip\break}
\def\0{\phantom{0}}
\def\bdot{\pmb{\cdot}}
\def\inlinefig#1{\includegraphics{#1}}
\def\tabnote#1{\vskip4pt\parbox{.83\linewidth}{#1}}
\def\mathbi#1{\text{\textit{\textbf{#1}}}}
\newenvironment{inftab}{}{}
\usepackage{multirow} 
\def\nrow#1{\@tempcnta #1\relax%
\advance\@tempcnta by 1\relax%
\xdef\lenrow{\the\@tempcnta}}
\def\morerows#1#2{\nrow{#1}\multirow{\lenrow}{*}{#2}}
\DOI{10.5802/crchim.429}
\datereceived{2025-09-22}
\daterevised{2025-10-17}
\dateaccepted{2025-10-20}
\ItHasTeXPublished
\makeatletter
\g@addto@macro{\UrlBreaks}{\UrlOrds}
\gappto{\UrlBreaks}{\UrlOrds}
\makeatother
}
{\documentclass[crchim]{article}
\def\CDRdoi{10.5802/crchim.429}
\let\refinput\input
\def\href#1#2{\url[#1]{#2}}
\let\ubreak\relax
\makeatletter
\def\CDRsupplementaryTwotypes#1#2{}
}
\makeatother

\usepackage{upgreek}

\dateposted{2026-01-22}
\begin{document}

\begin{noXML}

\CDRsetmeta{articletype}{research-article}

\title{From deprotometalation of ferrocenyl ketones to fused ferrocene
structures}

\alttitle{De la d\'{e}protom\'{e}tallation des c\'{e}tones
ferroc\'{e}niques aux structures ferroc\`{e}nes condens\'{e}es}

\author{\firstname{Madani} \lastname{Hedidi}}
\address{Univ Rennes, CNRS, ISCR (Institut des Sciences Chimiques de
Rennes) -- UMR 6226, F-35000 Rennes, France}
\address{Laboratoire de Chimie des Mat\'{e}riaux, Catalyse et
R\'{e}activit\'{e}, D\'{e}partement de Chimie, Facult\'{e} des Sciences
Exactes et d'Informatique, Universit\'{e} Hassiba Benbouali de Chlef,
Ouled Fares Chlef, BP 78C, 02180 Chlef, Alg\'{e}rie}
%\email{hedidi.m@gmail.com}

\author{\firstname{William} \lastname{Erb}\CDRorcid{0000-0002-2906-2091}\IsCorresp}
\addressSameAs{1}{Univ Rennes, CNRS, ISCR (Institut des Sciences
Chimiques de Rennes) -- UMR 6226, F-35000 Rennes, France}
\email[W. Erb]{william.erb@univ-rennes.fr}

\author{\firstname{Sirine} \lastname{Boussandel}}
\addressSameAs{1}{Univ Rennes, CNRS, ISCR (Institut des Sciences
Chimiques de Rennes) -- UMR 6226, F-35000 Rennes, France}
\address{University of Carthage, Faculty of Sciences of Bizerte,
Laboratory of Hetero-Organic Compounds and Nanostructured Materials
(LR18ES11), Zarzouna, 7021 Bizerte, Tunisia}
%\email{sirine.boussandel@univ-rennes.fr}

\author{\firstname{Yury Siarheevich} \lastname{Halauko}\IsCorresp}
\address{Department of Chemistry, Belarusian State University, 14
Leningradskaya St., 220030 Minsk, Belarus}
\email[Y. S. Halauko]{hys@tut.by}

\author{\firstname{Vadim Edvardovich} \lastname{Matulis}}
\addressSameAs{4}{Department of Chemistry, Belarusian State University,
14 Leningradskaya St., 220030 Minsk, Belarus}
%\email{matulisvad@gmail.com}

\author{\firstname{Jean-Pierre} \lastname{Hurvois}}
\addressSameAs{1}{Univ Rennes, CNRS, ISCR (Institut des Sciences
Chimiques de Rennes) -- UMR 6226, F-35000 Rennes, France}
%\email{jean-pierre.hurvois@univ-rennes.fr}

\author{\firstname{Marielle} \lastname{Blot}}
\addressSameAs{1}{Univ Rennes, CNRS, ISCR (Institut des Sciences
Chimiques de Rennes) -- UMR 6226, F-35000 Rennes, France}
%\email{marielle.blot@univ-rennes.fr}

\author{\firstname{Thierry} \lastname{Roisnel}\CDRorcid{0000-0002-6088-4472}}
\addressSameAs{1}{Univ Rennes, CNRS, ISCR (Institut des Sciences
Chimiques de Rennes) -- UMR 6226, F-35000 Rennes, France}
%\email{thierry.roisnel@univ-rennes.fr}

\author{\firstname{Ali} \lastname{Samarat}\CDRorcid{0000-0002-8334-7565}}
\addressSameAs{3}{University of Carthage, Faculty of Sciences of Bizerte,
Laboratory of Hetero-Organic Compounds and Nanostructured Materials
(LR18ES11), Zarzouna, 7021 Bizerte, Tunisia}
%\email{ali.samarat@gmail.com}

\author{\firstname{Florence} \lastname{Mongin}\CDRorcid{0000-0003-3693-8861}\IsCorresp}
\addressSameAs{1}{Univ Rennes, CNRS, ISCR (Institut des Sciences
Chimiques de Rennes) -- UMR 6226, F-35000 Rennes, France}
\email[F. Mongin]{florence.mongin@univ-rennes.fr}

\shortrunauthors

\keywords{\kwd{Ferrocene}\kwd{Ketone}\kwd{Deprotometalation}\kwd{Selectivity}\kwd{Polycyclic compounds}}

\altkeywords{\kwd{Ferroc\`ene}\kwd{C\'etone}\kwd{D\'eprotom\'etallation}\kwd{S\'electivit\'e}\kwd{Compos\'es polycycliques}}

\editornote{Article submitted by invitation}
\alteditornote{Article soumis sur invitation}

\thanks{Algerian Direction G\'{e}n\'{e}rale de la Recherche
Scientifique et du D\'{e}veloppement Technologique, Rennes
M\'{e}tropole, University of Carthage, Tunisian Ministry of Higher
Education and Scientific Research, Fonds Europ\'{e}en de
D\'{e}veloppement R\'{e}gional, Universit\'{e} de Rennes, Centre
National de la Recherche Scientifique, BASF, Thermofisher.}

\begin{abstract}
Although ferrocene ketones have been known since the early days of
ferrocene chemistry, their behavior in deprotometalation has never been
studied in detail. Here, we have optimized this reaction using lithium
2,2,6,6-tetramethylpiperidide in tetrahydrofuran containing
ZnCl\tralicstex{\textsubscript{2}\textperiodcentered\textit{N},\textit{N},\textit{N\textquotesingle},\textit{N\textquotesingle}}{$_{2}{\cdot}N$,$N$,$N'$,$N'$}-tetramethylethylenediamine as an
in-situ trap. Numerous 2-iodoferrocene derivatives were obtained, while
a change in regioselectivity was observed for certain aroyl- and
heteroaroylferrocenes, in good agreement with our DFT calculations. The
development of an enantioselective deprotometalation was also attempted
using lithium di[(\textit{S})-1-phenylethyl]amide, affording the
desired compounds in up to 60\% enantiomeric excess (\textit{ee}). A
selection of iodinated derivatives was finally subjected to
post-functionalizations, leading to original ferrocene-fused
heterocycles, including an enantiopure tetracyclic derivative. A
selection of compounds was studied in both electrochemical reduction
and oxidation, and weak interactions such as halogen--halogen,
halogen--oxygen and chalcogen--chalcogen bonds were identified in the
solid state for some new derivatives.
\end{abstract}

\begin{altabstract}
Bien que connues depuis le d\'{e}but de l'histoire de la chimie du
ferroc\`{e}ne, les c\'{e}tones ferroc\'{e}niques n'ont jamais fait
l'objet d'une \'{e}tude d\'{e}taill\'{e}e de
d\'{e}protom\'{e}tallation. Nous avons ici optimis\'{e} cette
r\'{e}action en pr\'{e}sence de
2,2,6,6-t\'{e}tram\'{e}thylpip\'{e}ridure de lithium et de
ZnCl\tralicstex{\textsubscript{2}\textperiodcentered\textit{N},\textit{N},\textit{N\textquotesingle},\textit{N\textquotesingle}}{$_{2}{\cdot}N$,$N$,$N'$,$N'$}-t\'{e}tram\'{e}thyl\'{e}thyl\`{e}nediamine
comme pi\`{e}ge \textit{in situ}. De nombreuses c\'{e}tones
ferroc\'{e}niques iod\'{e}es en position C-2 ont pu \^{e}tre obtenues
tandis qu'un changement de r\'{e}gios\'{e}lectivit\'{e} a \'{e}t\'{e}
observ\'{e} pour certains aroyl- et h\'{e}t\'{e}roaroylferroc\`{e}nes,
en accord avec nos calculs DFT. Une version \'{e}nantios\'{e}lective a
\'{e}galement \'{e}t\'{e} d\'{e}velopp\'{e}e en pr\'{e}sence de lithium
di[(\textit{S})-1-ph\'{e}nyl\'{e}thyl]amidure, avec un exc\`{e}s
\'{e}nantiom\'{e}rique (\textit{ee}) atteignant 60 \%. Plusieurs
compos\'{e}s ont \'{e}t\'{e} engag\'{e}s dans des
post-fonctionnalisations conduisant \`{a} de nouveaux
h\'{e}t\'{e}rocycles fusionn\'{e}s au ferroc\`{e}ne dont un
d\'{e}riv\'{e} t\'{e}tracyclique \'{e}nantiopur. Une \'{e}tude de
r\'{e}duction et d'oxydation \'{e}lectrochimiques a \'{e}t\'{e}
r\'{e}alis\'{e}e sur plusieurs d\'{e}riv\'{e}s tandis que des
interactions faibles telles que des liaisons halog\`{e}ne et
chalcog\`{e}ne ont pu \^{e}tre identifi\'{e}es \`{a} l'\'{e}tat solide.
\end{altabstract}

\maketitle
\vfill\pagebreak

\twocolumngrid

\end{noXML}

\section{Introduction}

Since the discovery and the structure determination of the parent
molecule~\cite{1,2,3,4}, ferrocenes have attracted considerable
interest from synthetic chemists due to their numerous
applications~\cite{5,6,7,8,9,10,11,12,13}, such as
catalysis~\cite{14,15,16,17}, materials science~\cite{18,19}, molecular
sensing~\cite{20,21}, and bioorganometallic
chemistry~\cite{22,23,24,25}. In addition to monosubstituted
ferrocenes, which can be synthesized from ferrocene via aromatic
electrophilic substitution~\cite{26} or deprotometalation/trapping
sequences~\cite{27,28}, polysubstituted structures may be required for
targeted applications~\cite{29,30,31,32,33}. Among the most commonly
used motifs, 1,1$'$-disubstituted derivatives are often obtained by
double deprotolithiation of ferrocene~\cite{34}, while
1,2-disubstituted derivatives are typically prepared by directed
functionalization of monosubstituted ferrocenes~\cite{35,36,37,38}.

As a bulky electron-rich aromatic, ferrocene tends to reduce the
electrophilicity of the attached functional group~\cite{39,40}.\ 
However, although ferrocene ketones are readily accessible by
Friedel--Crafts acylation of ferrocene~\cite{41}, exploiting this
change in reactivity to promote their functionalization by
deprotometalation/trapping sequences has rarely been explored.\ In
2000, Enders and coworkers converted ferrocene ketones to enantiopure
hydrazones{\break} (e.g., {\fontsize{9.9}{11.9}\selectfont using
(\textit{S})-1-amino-2-methoxymethylpyrrolidine},
SAMP) in order to achieve their diastereoselective functionalization by
deprotolithiation/trapping sequences~\cite{42}. A similar approach was
later developed by Top and coworkers using chiral imines derived from
ferrocene \mbox{ketones}~\cite{43}.

To our knowledge, the only study concerning the direct
functionalization of ferrocene ketones using lithium bases was reported
by Enders and coworkers, who subjected enantiopure diferrocenyl ketones
to \textit{s}-BuLi in the presence of
$N$,$N$,$N'$,$N'$-tetramethylethylenediamine (TMEDA) in toluene at
${-}$78~{\textdegree}C for 9~h, prior to interception with various
electrophiles~\cite{44}. However, the reactivity of the function toward
nucleophiles was probably affected by the presence of the two
organometallic cores, and the behavior of this particular substrate may
not reflect the behavior of aroylferrocenes in general. Thus, when
benzoylferrocene was treated with sodium zincate
(TMEDA)Na(TMP)Zn($t$-Bu)$_2$ [TMP ${=}$ 2,2,6,6-tetramethylpiperido] in
hydrocarbon solvents at room temperature, mixtures of products were
observed, some resulting from deprotonation (mediated by TMP) of the
ferrocene core at the site adjacent to the ketone, others from
competitive 1,2- and 1,6-addition reactions~\cite{45}.

In 2010, we reported a synthesis of 1-benzoyl-2-iodoferrocene in 36\%
yield using a base generated in tetrahydrofuran (THF) from LiTMP 
(1.5~equiv) and CdCl$_2{\cdot}$TMEDA (0.50~equiv)~\cite{46}.\ Here, our goal
is to evaluate an approach using the less toxic ZnCl$_2{\cdot}$TMEDA
for the functionalization of ferrocene ketones and to investigate the
development of an enantioselective version.

\section{Results and discussion} \label{sec2}
\subsection{Preliminary considerations} \label{sec2.1}

Our aim in this study was twofold: firstly, to identify suitable
conditions to functionalize ferrocene ketones by deprotometalation, and
secondly, to understand the regioselectivity observed during this
reaction. Indeed, most of the ferrocene ketones studied here have two
to four sites that can be functionalized in the presence of
organometallic bases.

Our approach involved the use of a hindered lithium amide, LiTMP, in
THF containing ZnCl$_2{\cdot}$TMEDA as an in-situ trap~\cite{47}.\ 
While many deprotometalations using lithium dialkylamides are under
thermodynamic control~\cite{48}, those using the strong base LiTMP can
also lead to derivatives functionalized next to a coordinating element,
probably via pre-metalation complexes (complex-induced proximity
effect)~\cite{49,50} or/and rate-limiting transition states
(kinetically enhanced metalation)~\cite{51,52}, or benefiting from
favorable hydrogen charges (overriding base
mechanism)~\cite{53,54,55,56}. This is particularly true when using an
in-situ trap, as we intended to do.

\begin{figure*}%%%
\includegraphics{fig01}
\caption{\label{fig1}Natural-population-analysis charges on hydrogen
atoms in the isolated molecule of benzoylferrocene (\textbf{1-Ph}) and
impact of coordination to lithium on these charges (complex of
\textbf{1-Ph} with LiNMe$_2$).}
{\vspace*{.1pc}}
\end{figure*}

Therefore, to get an idea of the kinetic acidity of the ferrocene
ketones selected for the present study, we calculated (see
computational details in Supplemantary material) the atomic charges of
a few representative examples by natural population analysis (NPA),
with and without oxygen coordination to lithium of LiNMe$_2$ chosen as
model base.\ For benzoylferrocene (\textbf{1-Ph}), the most acidic
hydrogens were found on the substituted cyclopentadienyl ring of
ferrocene, on either side of the benzoyl group
(\mbox{Figure}~\ref{fig1}, \textbf{1-Ph}). Further calculations of NPA
charges for all hydrogen atoms of the two substituted rings of the
complex \textbf{1-Ph}${\bdot}$\textbf{Li} showed comparable results
(Figure~\ref{fig1}, \textbf{1-Ph}${\bdot}$\textbf{Li}). A similar
tendency was observed for the other substrates (Figure~S1, in
Supplementary material).

However, formation of the thermodynamic product can also be observed in
reactions using strong lithium dialkylamides such as
LiTMP~\cite{57,58,59}.\ We therefore also calculated significant
p$K_{\mathrm{a}}$ values for several aroyl ketones in THF within the
DFT framework~\cite{60,61,62,63} (see Figure~S2 in Supplementary
material for a complete list). As shown on Figure~\ref{fig2}a, the
lowest p$K_{\mathrm{a}}$ value for \textbf{1-Ph} corresponds to the
ferrocene site next to the carbonyl group. Coordinating the ketone to
LiNMe$_2$ (\textbf{1-Ph}${\bdot}$\textbf{Li}, to mimic the stabilizing
effect that the function exerts on the lithiated derivative formed)
also seemed to show that the most stable lithium species was the one
deprotonated next to the carbonyl group on the ferrocene side
(\mbox{Figure}~\ref{fig2}a, \textbf{1-Ph}${\bdot}$\textbf{Li}).

\begin{figure*}
{\vspace*{.15pc}}
\includegraphics{fig02}
{\vspace*{.15pc}}
\caption{\label{fig2}Values of p$K_{\mathrm{a}}$ calculated for
ferrocene ketones in THF (an asterisk means that deprotonation in the
corresponding position is predicted to lead to interring rotation in
order to reduce electron repulsion), and impact of coordination to
lithium on these values (complexes with LiNMe$_2$).}
{\vspace*{.15pc}}
\end{figure*}

The introduction of electron-withdrawing substituents such as Br or
CF$_3$ on the phenyl group (with compounds \textbf{1-\textit{p}BrPh}
and \textbf{1-\textit{p}CF}$_{\mathbf{3}}$\textbf{Ph}) significantly
decreased the p$K_{\mathrm{a}}$ values at their adjacent positions
(Figure~\ref{fig2}b). However, coordination to LiNMe$_2$ restored the
desired reactivity, and the species deprotonated on the ferrocene side
appeared to be the most stable. A similar, although less pronounced,
effect can be noticed for \textbf{1-\textit{p}OMePh}
(Figure~\ref{fig2}c). Finally, in the case of (2-pyridoyl)- and
(2-benzothienoyl)ferrocenes (\textbf{1-2Py} and \textbf{1-2BTh}), the
calculations showed different behaviors, with lithium derivatives being
more stable when deprotonation occurs on the heterocycle side
(Figure~\ref{fig2}d). Furthermore, unlike the other benzoylferrocene
compounds considered, the conformation adopted by these heterocyclic
compounds has a clear impact on their computed p$K_{\mathrm{a}}$ values
(see Figures~S3 and~S4 in Supplementary material). 

All these data suggest that it should be possible to functionalize most
of these derivatives next to the ketone on the ferrocene core, while
(heteroaroyl)ferrocenes could afford derivatives functionalized on the
heterocycle side. With these predictions in mind, we then subjected our
selected compounds to the planned deprotometalation/trapping sequences.

\subsection{Functionalization of ferrocene ketones} \label{sec2.2}

Various ferrocene ketones have been prepared in the frame of this study
(for details, see Supplementary material). Most of the aroylferrocenes
were obtained by Friedel--Crafts acylation, by adapting reported
procedures~\cite{64,65,66}. A few other ferrocene ketones were
synthesized either by action of lithioferrocene onto the Weinreb
2,2,2-trifluoroacetamide (compound \textbf{1-CF}$_{\mathbf{3}}$; see
Scheme~\ref{sch7})~\cite{67} or by reacting heteroarylmetals with the
Weinreb amide of ferrocene~\cite{68} (compounds \textbf{1-2Py} and
\mbox{\textbf{1-2BTh}).} \looseness=1

Benzoylferrocene (\textbf{1-Ph}) was chosen to optimize the reaction
using LiTMP in the presence of an in-situ trap~\cite{47}. We first
tested the use of the putative Zn(TMP)$_2$ (generated from 1~equiv of
ZnCl$_2{\cdot}$TMEDA and 2~equiv of LiTMP) as an in-situ trap~\cite{69}
in the reaction of \textbf{1-Ph} with LiTMP (1.1~equiv) in THF at
${-}$20~{\textdegree}C. After subsequent trapping with iodine, the
expected product \textbf{2-Ph} was obtained in a moderate 55\% yield
although almost complete conversion was obtained (${<}$4\% \textbf{1-Ph}
recovered). Replacing Zn(TMP)$_2$ with ZnCl$_2{\cdot}$TMEDA (1.1~equiv)
favored the formation of \textbf{2-Ph}, isolated in 72\% yield (91\%
using 2.2~equiv of LiTMP) (Scheme~\ref{sch1}, Equation~(1)). The
intermediate ferrocenylzinc was also engaged in a Negishi
cross-coupling~\cite{70,71} with 2-chloropyridine. Thus, the use of
catalytic amounts of PdCl$_2$ and
1,1$'$-bis(diphenylphosphino)ferrocene (dppf)~\cite{72} led to the
expected heteroarylated product \textbf{3-Ph} (37\% yield)
(Scheme~\ref{sch1}, Equation~(2)).

\begin{scheme*}
{\vspace*{.15pc}}
\includegraphics{sc01}
{\vspace*{.75pc}}
\caption{\label{sch1}Functionalization of benzoylferrocene
(\textbf{1-Ph}) by deprotolithiation with in-situ trap, followed by
either iodination or Negishi cross-coupling.}
{\vspace*{-.6pc}}
\end{scheme*}

We next attempted to use chlorotrimethylsilane instead of
ZnCl$_2{\cdot}$TMEDA as the in-situ trap, but we failed to observe the
formation of the expected ferrocenylsilane under these conditions. We
also tried to replace benzoylferrocene (\textbf{1-Ph}) by
(phenylcarbonothioyl)ferrocene (\textbf{4-Ph}), prepared by reacting
the former with Lawesson's reagent~\cite{73}. However, as already
observed in the case of thionoesters~\cite{74}, the iodinated
derivative was not detected, while we recovered 20\% of \textbf{4-Ph}
and 29\% of \textbf{1-Ph}.

Functionalization at position C-2 of aroylferrocenes was then carried
out starting from three different (methoxybenzoyl)ferrocenes, e.g.,
\textbf{1-\textit{o}OMePh}, \textbf{1-\textit{m}OMePh}, and
\textbf{1-\textit{p}OMePh}. As shown in Scheme~\ref{sch2}, the amount
of base was advantageously increased from 1.1 to 1.5~equiv, enabling
the isolation of iodides \textbf{2-\textit{o}OMePh},
\textbf{2-\textit{m}OMePh}, and \textbf{2-\textit{p}OMePh} in yields
ranging from 82 to 94\%. Functionalization of
(methoxybenzoyl)ferrocenes at C-2 is therefore consistent with the
calculated p$K_{\mathrm{a}}$ values of \textbf{1-\textit{p}OMePh} after
coordination to lithium (Figure~\ref{fig2}c). In the case of
(4-\textit{tert}-butylbenzoyl)ferrocene (\textbf{1-\textit{pt}BuPh}),
derivative \textbf{2-\textit{pt}BuPh} was produced in a high 94\% yield
using 2~equiv of base. Even for the (chlorobenzoyl)ferrocenes
\textbf{1-\textit{o}ClPh}, \textbf{1-\textit{m}ClPh}, and
\textbf{1-\textit{p}ClPh}, the functionalization took place at the
ferrocene site next to the carbonyl group, producing iodides
\textbf{2-\textit{o}ClPh}, \textbf{2-\textit{m}ClPh}, and
\textbf{2-\textit{p}ClPh} in yields of 51--64\%.

\begin{scheme*}
\includegraphics{sc02}
{\vspace*{.9pc}}
\caption{\label{sch2}Regioselective functionalization of
aroylferrocenes at C-2 by deprotolithiation with in-situ trap, followed
by iodination.}
\end{scheme*}

The results were different for the other (halo)ben\-zoylferrocenes.\ In
the (bromoaroyl)ferrocene series, all the derivatives obtained resulted
from the functionalization next to the ketone function.\ However, the
iodoferrocenes were formed less efficiently
(\mbox{\textbf{2-\textit{o}BrPh}} ${>}$ \textbf{2-\textit{m}BrPh} ${>}$
\textbf{2-\textit{p}BrPh}), and iodination of the aryl ring was even
mainly observed in the case of \textbf{2-\textit{p}BrPh}
(\textbf{2}$'$\textbf{-\textit{p}BrPh}, ${>}$50\% yield;
Scheme~\ref{sch3}). This outcome could be rationalized by the combined
effects of the ketone function, which promotes deprotolithiation at its
neighboring sites, and bromine, which exerts a long-range acidifying
effect~\cite{75}. This is consistent with the p$K_{\mathrm{a}}$ values
calculated for \mbox{\textbf{1-\textit{p}BrPh}${\bdot}$\textbf{Li}},
which are similar for positions C-2 and C-2$'$ (Figure~\ref{fig2}b). A
similar trend was noticed in the (trifluoromethyl)ferrocene series,
which probably results from the acidifying properties of the
trifluoromethyl group in the \textit{ortho}, \textit{meta}, and
\textit{para} positions~\cite{56,76} (Scheme~\ref{sch3}). For
\textbf{1-\textit{p}CF}$_{\mathbf{3}}$\textbf{Ph}${\bdot}$\textbf{Li},
the p$K_{\mathrm{a}}$ difference between positions C-2 and C-2$'$ is
slightly greater than in the case of
\textbf{1-\textit{p}BrPh}${\bdot}$\textbf{Li} (Figure~\ref{fig2}b), which
could explain the predominance of
\textbf{2-\textit{p}CF}$_{\mathbf{3}}$\textbf{Ph}, functionalized on
the ferrocene side.

\begin{scheme*}
\includegraphics{sc03}
{\vspace*{.9pc}}
\caption{\label{sch3}Functionalization of (bromobenzoyl)- and
[(trifluoromethyl)benzoyl]ferrocenes by deprotolithiation with in-situ
trap, followed by iodination.}
\end{scheme*}

In contrast to bromine and trifluoromethyl, fluorine is known to be a
short-range acidifying group~\cite{76,77}. As a consequence, when
(2-fluoroben\-zoyl)ferrocene (\textbf{1-\textit{o}FPh}) was treated under
the same conditions, the expected product \textbf{2-\textit{o}FPh}
functionalized on the ferrocene ring (41\% yield) was accompanied by
(2-fluoro-3-iodobenzoyl)ferrocene
(\textbf{2}$'$\textbf{-\textit{o}FPh}, 6\% yield) and
1-(2-fluoro-3-iodobenzoyl)-2-iodoferrocene
(\textbf{2}$''$\textbf{-\textit{o}FPh}, 18\% yield), both due to
competitive reactions next to the halogen (Scheme~\ref{sch4}).\looseness=-1

\begin{scheme*}
\includegraphics{sc04}
{\vspace*{.9pc}}
\caption{\label{sch4}Functionalization of (2-fluorobenzoyl)ferrocene by
deprotolithiation with in-situ trap, followed by iodination.}
{\vspace*{-.2pc}}
\end{scheme*}

\looseness=-1
Coordination of (2-pyridoyl)- and (2-benzo\-thienoyl)ferrocenes
(\textbf{1-2Py} and \textbf{1-2BTh}) to lithium decreases their
p$K_{\mathrm{a}}$ values at position C-3$'$ by several units, making
these sites more prone to deprotometalation than those in position C-2
(Figure~\ref{fig2}d). For ferrocene \textbf{1-2Py}, the calculated
p$K_{\mathrm{a}}$ difference is about three units, while it reaches six
units for ferrocene \textbf{1-2BTh}. As a result, in these two cases,
the main products obtained are iodinated on the heterocycle with
\textbf{2}$'$\textbf{-2Py} (26\% yield) and above all
\textbf{2}$'$\textbf{-2BTh} (${>}$50\% yield). Products monoiodinated on
ferrocene were not observed, but the diiodides
\mbox{\textbf{2}$''$\textbf{-2Py}} and \textbf{2}$''$\textbf{-2BTh},
probably resulting from double deprotometalation/trapping, were
isolated in $\sim$10\% yield (Scheme~\ref{sch5}).

\begin{scheme*}
\includegraphics{sc05}
{\vspace*{.9pc}}
\caption{\label{sch5}Functionalization of (2-pyridoyl)- and
(2-benzothienoyl)ferrocenes by deprotolithiation with in-situ trap,
followed by iodination.}
{\vspace*{-.5pc}}
\end{scheme*}

Cinnamoyl- and (phenylpropioloyl)ferrocenes
(\mbox{\textbf{1-CH}${=}$\textbf{CHPh}} and
\textbf{1-C}${\equiv}$\textbf{CPh}) were also involved in~the reaction
for the purpose of comparison with~\mbox{\textbf{1-Ph}}.\ The expected
products were obtained in both cases, but with lower yields than those
observed~with \textbf{1-Ph}.\ Compound \textbf{2-C}${\equiv}$\textbf{CPh}
was produced in a higher yield (48\%) than
\textbf{2-CH}${=}$\textbf{CHPh} (only 25\%), probably due to higher
sensitivity of \textbf{1-CH}${=}$\textbf{CHPh} (only 17\% recovered) to
nucleophilic attacks, leading to unidentified decomposition products
(Scheme~\ref{sch6}).

\begin{scheme}
\includegraphics{sc06}
{\vspace*{.8pc}}
\caption{\label{sch6}Functionalization of cinnamoyl- and
(phenylpropioloyl)ferrocenes by deprotolithiation with in-situ trap,
followed by iodination.}
{\vspace*{-1.5pc}}
\end{scheme}

Finally, we compared the reactivities of
[(trifluoromethyl)carbonyl]ferrocene (\textbf{1-CF}$_{\mathbf{3}}$) and
(\textit{tert}-butylcarbonyl)ferrocene (\textbf{1-\textit{t}Bu}) under
these conditions. As the question of regioselectivity does not arise,
\textbf{1-\textit{t}Bu} was treated with 1.1 or 2.2~equiv of LiTMP to
provide, after subsequent trapping, the expected iodide
\textbf{2-\textit{t}Bu} in high yield (75--95\%). However, while the
use of 1.1~equiv of LiTMP on \textbf{1-CF}$_{\mathbf{3}}$ afforded
\textbf{2-CF}$_{\mathbf{3}}$ (as an inseparable mixture with
\textbf{1-CF}$_{\mathbf{3}}$) in an estimated 38\% yield, using an
excess of base was detrimental to the reaction outcome
(Scheme~\ref{sch7}).

\begin{scheme}
\includegraphics{sc07}
{\vspace*{.7pc}}
\caption{\label{sch7}Functionalization of [(trifluoromethyl){\ubreak}carbonyl]-
and (\textit{tert}-butylcarbonyl)ferrocenes by deprotolithiation with
in-situ trap, followed by iodination.}
{\vspace*{-1.5pc}}
\end{scheme}

Ferrocene ketones being prochiral substrates, their enantioselective
functionalization is expected to deliver enantio-enriched
1,2-disubstituted derivatives~\cite{78}. Enantioselective
deprotolithiation of ferrocenes substituted by aminomethyl~\cite{79,80}
and dimethylamino~\cite{81,82} groups, as well as hindered tertiary
carboxamide~\cite{83,84,85,86} or even triflone~\cite{87}, using
alkyllithium-chiral ligand chelates (e.g., \textit{n}-BuLi-sparteine),
represents an important achievement in this field. However, as such
chiral nucleophilic bases would be barely compatible with a sensitive
ketone, we evaluated another approach involving a chiral,
non-nucleophilic lithium dialkylamide \{lithium
di[(\textit{S})-1-phenylethyl]amide, (\textit{S})-PEALi\} in the
presence of an in-situ trap, as initially documented by
Simpkins~\cite{88} and later extended to sensitive ferrocene
carboxamides and esters~\cite{74,89,90}. Regarding the nature of the
in-situ trap~\cite{91}, we selected ZnCl$_2{\cdot}$TMEDA as well as the
putative zinc diamide \{(\textit{S})-PEA\}$_2$Zn, obtained in situ from
(\textit{S})-PEALi and ZnCl$_2{\cdot}$TMEDA in a 2:1 ratio, given their
superiority in studies already carried out in the group~\cite{87,92}.

\looseness=-1
To optimize the reaction, a solution of ketone \mbox{\textbf{1-Ph}} and
ZnCl$_2{\cdot}$TMEDA (1~equiv) in THF was treated with
(\textit{S})-PEALi (2.2~equiv) at various temperatures (0, ${-}$20,
${-}$50, and ${-}$80~{\textdegree}C) before iodolysis (Scheme~\ref{sch8},
Equation~(1)). Whereas lower yields of \textbf{2-Ph} were recorded at
${-}$50~{\textdegree}C and even lower at ${-}$80~{\textdegree}C, the best
enantioselectivities were observed at 0 or ${-}$20~{\textdegree}C but did
not exceed 38\% enantiomeric excess (\textit{ee}) in favor of the $R_{\mathrm{P}}$
enantiomer, as revealed by the growing of crystals suitable for X-ray
diffraction (XRD) analysis (Figure~\ref{fig3}). No change in
\textit{ee} was noticed by using either ($R$)-PEALi instead of
(\textit{S}) or a shorter contact time. We next attempted the reaction
using our alternative in-situ trap, \{(\textit{S})-PEA\}$_2$Zn, instead
of ZnCl$_2{\cdot}$TMEDA (Scheme~\ref{sch8}, Equation~(2)). The
reactions were thus repeated at ${-}$20 and ${-}$80~{\textdegree}C,
affording the major $R_{\mathrm{P}}$ enantiomer in a slightly improved 44\%
\textit{ee} and yields of 60 and 35\%, respectively, due to important
recovery of \textbf{1-Ph} at the lowest temperature. In a last attempt
to improve the enantioselectivity, THF was replaced by
2-methyltetrahydrofuran (2-MeTHF), which was found helpful in processes
involving sensitive species in asymmetric
\mbox{transformations}~\cite{93}. However, whether using
ZnCl$_2{\cdot}$TMEDA at ${-}$20~{\textdegree}C or
\{(\textit{S})-PEA\}$_2$Zn at ${-}$80~{\textdegree}C as an in-situ trap,
significantly lower yields (from 69 to 40\% and from 35 to 3\%,
respectively) and enantioselectivities (from 38 to 18\% \textit{ee} and
from 44 to 4\% \textit{ee}, respectively) were\break recorded.

\begin{scheme*}
{\vspace*{.1pc}}
\includegraphics{sc08}
{\vspace*{.9pc}}
\caption{\label{sch8}Enantioselective functionalization of
benzoylferrocene (\textbf{1-Ph}) by deprotolithiation using the lithium
di(1-phenylethyl)amide (\textit{S})-PEALi with zinc-based in-situ
traps. (a)~67\% yield and 38\% \textit{ee} ($S_{\mathrm{P}}$) using
(\textit{R})-PEALi. (b)~62\% yield and 30\% \textit{ee} ($R_{\mathrm{P}}$) after
5~min time. (c)~40\% yield and 18\% \textit{ee} ($R_{\mathrm{P}}$) using 2-MeTHF.
(d)~3\% yield and 4\% \textit{ee} ($R_{\mathrm{P}}$) using 2-MeTHF.}
{\vspace*{.25pc}}
\end{scheme*}

\begin{figure}
\includegraphics{fig03}
\caption{\label{fig3}Molecular structure of compound
$\mathbi{R}_{\mathbf{P}}$\textbf{-2Ph} in the solid
state. Thermal ellipsoids shown at the 30\% probability level. Selected
bond lengths ({\AA}) and angles ({\textdegree}): C10--C11 ${=}$ 1.475(3),
C6--I1 ${=}$ 2.082(2), C10--Cg2${\cdots}$Cg1--C3 ${=}$ 2.74 (Cg1 being the
centroid of the C1--C2--C3--C4--C5 ring and Cg2 the one of the
C6--C7--C8--C9--C10 ring), Cg2--C6--I1 ${=}$ 175.62, C6--C10--C11--O12
${=}$ 15.8(4), O12--C11--C13--C14 ${=}$ 25.9(3).}
{\vspace*{-.3pc}}
\end{figure}

Reactions using ZnCl$_2{\cdot}$TMEDA as an in-situ trap, being
considerably easier to implement, were performed at
${-}$20~{\textdegree}C to explore the behavior of a selection of
ferrocene ketones in asymmetric deprotolithiation (Scheme~\ref{sch9}).
The yields recorded for these reactions using (\textit{S})-PEALi were
generally lower than those obtained with the same amount of LiTMP,
which can be explained in most cases by the lower reactivity of
(\textit{S})-PEALi. Using these conditions, ketones
\mbox{\textbf{1-CH}}${=}$\textbf{CHPh} and
\textbf{1-C}${\equiv}$\textbf{CPh} only afforded mixtures of unidentified
products. The ferrocenyl versus aryl deprotonation selectivities were
the same, notably with the competitive formation of derivatives
iodinated on the aryl group in the case of \textbf{1-\textit{p}BrPh},
\mbox{\textbf{1-\textit{o}FPh}},
\mbox{\textbf{1-\textit{m}CF}$_{\mathbf{3}}$}\textbf{Ph}, and
\textbf{1-\textit{p}CF}$_{\mathbf{3}}$\textbf{Ph}. The
enantioselectivities remained modest at best, ranging from 0\% for the
sterically hindered ketone
\textbf{1-\textit{o}CF}$_{\mathbf{3}}$\textbf{Ph} to 60\% for the
trifluoromethylketone \textbf{1-CF}$_{\mathbf{3}}$. Pleasingly,
crystallization of \textbf{2-\textit{o}BrPh} afforded an enantiopure
product (see Supplementary material), and XRD analysis validated the
expected $R_{\mathrm{P}}$ configuration.

\begin{scheme*}
\includegraphics{sc09}
{\vspace*{.9pc}}
\caption{\label{sch9}Enantioselective functionalization of ferrocene
ketones by deprotolithiation using the lithium di(1-phenylethyl)amide
(\textit{S})-PEALi with ZnCl$_2{\cdot}$TMEDA as an in-situ trap.
(a)~\textbf{2}$'$\textbf{-\textit{p}BrPh} also obtained in 50\% yield.
(b)~\textbf{2}$'$\textbf{-\textit{m}CF}$_{\mathbf{3}}$\textbf{Ph} also
formed in 8\% yield.
(c)~\textbf{2}$'$\textbf{-\textit{p}CF}$_{\mathbf{3}}$\textbf{Ph} also
formed in 16\% yield. (d)~\textbf{2}$'$\textbf{-\textit{o}FPh} (23\%
yield) and \textbf{2}$''$\textbf{-\textit{o}FPh} (15\% yield, 28\%
\textit{ee}) also formed. (e)~Reaction carried out by using
(\textit{S})-PEALi (1.1~equiv) and I$_2$ (1.1~equiv).}
{\vspace*{-.7pc}}
\end{scheme*}

Slightly disappointed by these results, we next turned our attention to
a diastereoselective approach and prepared the enantiopure ferrocene
ketone $\mathbi{R}_{\mathbf{P}}$\textbf{-5} from
[(\textit{S})-1-(dimethylamino)ethyl]ferrocene~(Ugi's
amine)~\cite{94,95} (Scheme~\ref{sch10}).\ Further treatment of
$\mathbi{R}_{\mathbf{P}}$\textbf{-5} with LiTMP (2~equiv) in THF
containing ZnCl$_2{\cdot}$TMEDA (1~equiv) at ${-}$20~{\textdegree}C
before iodolysis afforded the expected iodoferrocene
$\mathbi{R}_{\mathbf{P}}$\textbf{-6} as a single diastereoisomer,
albeit in a low 20\% yield due to recovery of
$\mathbi{R}_{\mathbf{P}}$\textbf{-5} (about~20\%) and formation of
unidentified byproducts (Scheme~\ref{sch10}).

\begin{scheme*}
\includegraphics{sc10}
{\vspace*{.7pc}}
\caption{\label{sch10}Diastereoselective conversion of Ugi's amine to
the ketone $\mathbi{R}_{\mathbf{P}}$\textbf{-5} and further conversion
to the enantiopure iodoketone $\mathbi{R}_{\mathbf{P}}$\textbf{-6}.}
{\vspace*{-.6pc}}
\end{scheme*}

\begin{scheme*}
\includegraphics{sc11}
{\vspace*{.7pc}}
\caption{\label{sch11}Ring \textit{tert}-butylation of
$\mathbi{R}_{\mathbf{P}}$\textbf{-5} and \textbf{1-Ph}.}
{\vspace*{-.7pc}}
\end{scheme*}

In 1991, Olah and co-workers reported the ring \textit{tert}-butylation
of benzophenones by successive action of \textit{t}-BuLi in THF at very
low temperature and SOCl$_2$~\cite{96}. Inspired by these results, we
similarly treated $\mathbi{R}_{\mathbf{P}}$\textbf{-5} with
\textit{t}-BuLi at ${-}$80~{\textdegree}C for 15~min before addition of
I$_2$ to either intercept a deprotometalated species or oxidize a
1,6-adduct. Pleasingly, we were able to isolate the
1,6-addition/rearomatization product
$\mathbi{R}_{\mathbf{P}}$\textbf{-7} in 61\% yield (Scheme~\ref{sch11},
Equation~(1)). A similar result was recorded from benzoylferrocene
(\textbf{1-Ph}), with \textbf{1-\textit{pt}BuPh} obtained in 80\% yield
(Scheme~\ref{sch11}, Equation~(2)).

In the course of their studies on the reactivity of organolithium
compounds toward ketones, Yamataka and coworkers obtained both the
products coming from the 1,2-addition (65\%) and the
1,6-addition/rearomatization (28\%) by reacting benzophenone with
\textit{t}-BuLi in Et$_2$O at 0~{\textdegree}C~\cite{97}. The use of
various \textit{tert}-butylzinc species in order to achieve
\textit{para}-selective \textit{tert}-butylation of
benzophenone~\cite{98,99,100} and other (hetero)aromatic
ketones~\cite{101} has also been extensively studied. In our case, the
major 1,6-addition observed by simply using \textit{t}-BuLi could be
explained by the lower electrophilicity of the ketone and the higher
steric hindrance generated by the ferrocene core when compared with
benzophenone.

Whether using \textit{t}-BuLi~\cite{97} or \textit{tert}-butylzinc
species~\cite{102}, all the studies of this reaction have pointed out a
single electron transfer mechanism from the alkylmetal to the ketone.
Applied to \textbf{1-Ph}, the formation of \textbf{1-\textit{pt}BuPh}
could unfold as depicted in Scheme~\ref{sch12}.

\begin{scheme*}
\includegraphics{sc12}
{\vspace*{.65pc}}
\caption{\label{sch12}Proposed mechanism for the
\textit{tert}-butylation of \textbf{1-Ph}.}
{\vspace*{-.5pc}}
\end{scheme*}

\begin{scheme*}
\includegraphics{sc13}
{\vspace*{.65pc}}
\caption{\label{sch13}Suzuki--Miyaura and Sonogashira cross-coupling
reactions from \textbf{2-Ph}.}
{\vspace*{-.5pc}}
\end{scheme*}

\begin{scheme*}
\includegraphics{sc14}
{\vspace*{.65pc}}
\caption{\label{sch14}Conversion of \textbf{2-CH}${=}$\textbf{CHPh} to
the 2,3-dihydrothiopyrano[2,3]ferrocen-4-one \textbf{10}.}
{\vspace*{-.6pc}}
\end{scheme*}

\subsection{Post-functionalization toward polycyclic{\hfill\break}
compounds} \label{sec2.3}

\looseness=-1
Some of our iodinated ferrocene ketones were next engaged in
metal-promoted transformations, and post-functionalization by
Suzuki--Miyaura cross-coupling~\cite{103,104} was first considered from
\textbf{2-Ph}.\ The reactions were carried out with 4-methoxyphenyl- and
3-thienylboronic acid, under conditions previously
tested~\cite{105,106} (2~equiv CsF~\cite{107}, 5~mol\% Pd(dba)$_2$ (dba
${=}$ dibenzylidene\-acetone) and 20~mol\% SPhos
(2-(dicyclohexyl\-phosphino)-2$'$,6$'$-dimethoxybiphenyl)~\cite{108}~in
refluxing toluene), to afford the derivatives \textbf{8a} and
\textbf{8b} (Scheme~\ref{sch13}, Equation~(1) and~(2)).\ A Sonogashira
cross-coupling~\cite{109} was next attempted with
(trimethylsilyl)acetylene, under conditions previously
reported~\cite{40,110} [4~mol\% Pd(P\textit{t}Bu$_3$)$_2$ and 4~mol\%
CuI in THF-iPr$_2$NH at rt]. Although the \mbox{expected} product \textbf{9}
was obtained in a low 21\% yield due to 55\% recovery of \textbf{2-Ph}
(Scheme~\ref{sch13}, Equation~(3)), it could be a promising substrate
for accessing biologically active ferrocene derivatives such as
prostaglandin analogues~\cite{111}.

In the last decade, ferrocenes fused with heterocycles such as
pyridine~\cite{112}, 4-pyridone~\cite{112}, and
4-piperidinone~\cite{113} have appeared as privileged structures for
different applications~\cite{114}. With the aim of preparing a
sulfur-containing related compound and inspired by similar reaction in
the benzene series~\cite{115}, we reacted
\mbox{(\textit{E})-1-cinnamoyl-2-iodoferrocene}
(\textbf{2-CH}${=}$\textbf{CHPh}) with potassium ethyl thioxanthate in
the presence of Cu(OAc)$_2$ in dimethylsulfoxide (DMSO) at
70~{\textdegree}C. Pleasingly, we were able to isolate the expected
2,3-dihydrothiopyrano[2,3]ferrocen-4-one \textbf{10} in 47\% yield and
70\% \textit{de} (Scheme~\ref{sch14}).

Ferrocene-fused quinoline derivatives are also documented, and notably
ferroceno[\textit{c}]quinolines, which can be synthesized from
2-iodoferrocene carboxaldehyde~\cite{116}. As starting from a ketone
would provide a substituent in position C-6 of this building block, we
were keen to prepare such \mbox{derivatives} from the iodinated ketone
\textbf{2-\textit{p}CF}$_{\mathbf{3}}$\textbf{Ph}.\ Inspired by
previous results, it was treated with 4~equiv of 2-aminophenylboronic
acid, 5~mol\% Pd(dba)$_2$, 20~mol\% PPh$_3$, and 2~equiv of
CsF~\cite{107}, to which 2~equiv of Cs$_2$CO$_3$ were added for
subsequent \mbox{cyclization} in refluxing dioxane. This afforded the
original compound \textbf{11} in high yield (Scheme~\ref{sch15}). 

\begin{scheme*}
{\vspace*{.2pc}}
\includegraphics{sc15}
{\vspace*{.5pc}}
\caption{\label{sch15}Suzuki--Miyaura cross-coupling from
\textbf{2-\textit{p}CF}$_{\mathbf{3}}$\textbf{Ph} and subsequent
cyclization into ferroceno[\textit{c}]quinoline.}
{\vspace*{-.2pc}}
\end{scheme*}

\begin{scheme*}
\includegraphics{sc16}
{\vspace*{.6pc}}
\caption{\label{sch16}Conversion of \textbf{2}$'$\textbf{-2BTh} to the
polycyclic compound \textbf{12} by CH-functionalization.}
{\vspace*{-.7pc}}
\end{scheme*}

Condensed systems in which the cyclopentadienyl ring of the ferrocene
is annulated with a (hetero)aromatic moiety have seen renewed interest
with the advent of palladium-catalyzed enantioselective C--H bond
activation~\cite{114}. Inspired by the \mbox{synthesis} of ferrocene analogues
of fluorenone documented by You and coworkers~\cite{117}, we finally
attempted the synthesis of the original thiophene-containing derivative
\textbf{12} from the iodinated compound \textbf{2}$'$\textbf{-2BTh}. To
this purpose, our substrate was treated with 5~mol\% Pd(OAc)$_2$,
5~mol\% (\textit{S})-BINAP [BINAP ${=}$
2,2$'$-bis(diphenylphosphino)-1,1$'$-binaphthyl], 1.5~equiv of
Cs$_2$CO$_3$, and 0.3~equiv of pivalic acid in xylene at
60~{\textdegree}C. Pleasingly, the expected tetracyclic compound
\textbf{13} was isolated in 92\% yield and ${>}$99\% \textit{ee}, the
$R_{\mathrm{P}}$ absolute configuration being confirmed by XRD analysis
(Scheme~\ref{sch16}).

\subsection{Electrochemical characterization of selected compounds}
\label{sec2.4}

Although ferrocene ketones have been known since the early days of
ferrocene history~\cite{41}, and indeed helped Woodward and his team to
coined the name ferrocene~\cite{118,119}, they have been scarcely
studied from an electrochemical point of view. Kutal et~al.\ reported
an in-depth study of the link between the spectroscopic properties and
their electronic structure~\cite{120}, while Gao et~al.\ reported
third-order non-linear optical properties of some benzoylferrocenes,
supported by DFT calculations~\cite{121}. However, electrochemical
analyses were not included in these studies.\ The redox potential of a
few ferrocene \mbox{ketones} has been reported from time to
time~\cite{122,123,124}, most of the work on benzoylferrocenes coming
from the work of Kleinberg et~al.~\cite{125}. We were \mbox{therefore}
interested in investigating the electrochemical properties of some
ferrocene ketones prepared during this work, initially focusing our
attention on the ketone reduction. Cyclic voltammetry (CV) and
differential pulse voltammetry (DPV) were therefore realized in dry,
oxygen-free, \mbox{dimethylformamide}, \mbox{using}
\textit{n}-Bu$_4$NPF$_6$ (0.1~M) as the supporting electrolyte with a
glassy carbon disk as working electrode, an Ag/AgCl reference
electrode, and a glassy carbon rod as counter electrode
(Table~\ref{tab1}).

\begin{table*}%tab1
\tabcolsep21pt
\caption{\label{tab1}Electrochemical data (in~V) for the monoelectronic
reduction of selected ferrocene ketones{\vspace*{-.2pc}}}
\begin{tabular}{ccccc}
\thead
Compound & ${E_{\mathrm{pc}}}^{\mathrm{a}}$ &
${E_{\mathrm{pa}}}^{\mathrm{a}}$ &
${i_{\mathrm{pc}}/i_{\mathrm{pa}}}^{\mathrm{a}}$ & 
${E_{1/2}}^{\mathrm{b}}$\\
\endthead
\textbf{Benzophenone} & ${-}$1.74 & ${-}$1.64 & 1.00 & ${-}$1.68\\
\textbf{1-Ph} & ${-}$1.88 & ${-}$1.79 & 0.99 & ${-}$1.84\\
\textbf{1-\textit{p}OMePh} & ${-}$2.00 & ${-}$1.87 & 1.36 & ${-}$1.95\\
\textbf{1-\textit{p}CF}$_{\mathbf{3}}$\textbf{Ph} & ${-}$1.67 & ${-}$1.51 & nd$^{(\mathrm{c})}$ & ${-}$1.75\\
\textbf{1-2Py} & ${-}$1.68 & ${-}$1.60 & 1.43 & ${-}$1.64\\
\textbf{1-\textit{t}Bu} & ${-}$2.29 & ${-}$2.15 & 1.20 & ${-}$2.23\\
\textbf{2-Ph} & ${-}$1.70 & nd$^{(\mathrm{d})}$ & nd$^{(\mathrm{d})}$ & ${-}$1.68\\
\textbf{12} & ${-}$1.36 & ${-}$1.26 & 0.92 & ${-}$1.30{\vspace*{.2pc}}\\
\multicolumn{5}{c}{\protect\inlinefig{fx01}}
\botline
\end{tabular}
\tabnote{Potential values given relative to Ag/AgCl, scan rate ${=}$
100~mV$\cdot$s$^{-1}$. ${}^{\mathrm{a}}$From CV experiments.
${}^{\mathrm{b}}$From DPV experiments. ${}^{\mathrm{c}}$A complex
reduction process was observed. ${}^{\mathrm{d}}$An irreversible
dielectronic reduction process was observed.}
{\vspace*{-.2pc}}
\end{table*}

From benzophenone to \textbf{1-Ph}, \textbf{1-\textit{p}OMePh}, and
\textbf{1-\textit{t}Bu}, the monoelectronic reduction of the ketone,
forming the radical anion species, was found reversible and
increasingly difficult to achieve, in agreement with the increased
electron-donating properties of the group attached to the ketone.
Regarding the effect of electron-withdrawing groups, reduction is
easier from the 2-pyridyl derivative \mbox{\textbf{1-2Py}} than from the
compounds \textbf{1-\textit{p}CF}$_{\mathbf{3}}$\textbf{Ph} and
\textbf{1-Ph}, as one can have expected regarding the electronic
properties of these aromatics.\ We also attempted to push the reduction
of \textbf{1-\textit{p}CF}$_{\mathbf{3}}$\textbf{Ph} forward to form
the corresponding dianion species. However, although we did observe a
second reduction in DPV (see Supplementary material), it did not appear
to be the expected formation of the dianion species, but rather the
reduction of the \mbox{trifluoromethyl}~group,{\break}
as~\mbox{reported} by Perichon et~al.\ and Sav\'{e}ant et~al.
\cite{126,127}.\ Due to the inductive effect of iodine,
\mbox{\textbf{2-Ph}} was more easily reduced than benzoylferrocene
\mbox{(\textbf{1-Ph})} by 0.16~V. However, the first dielectronic
irreversible process was immediately followed by a second reduction at
${-}$1.84~V versus FcH/FcH$^+$, corresponding to that of \textbf{1-Ph}.
We suggest that the initial formation of the radical anion of
\textbf{2-Ph} was followed by the cleavage of the carbon--iodine bond
toward the neutral radical \textbf{1-Ph}$^{\bullet}$, which was more
easily reduced than the parent compound \textbf{2-Ph}
(Scheme~\ref{sch17}). The resulting anion \textbf{1-Ph}$^{-}$ would
then be protonated in situ to generate \textbf{1-Ph}, which exhibited
the expected monoelectronic reduction at ${-}$1.84~V. The fused
tetracycle \textbf{12} was found to undergo an easy and reversible
monoelectronic \mbox{reduction} first, followed by two irreversible
reduction peaks, which might be attributed to the reduction of the
radical anion to the corresponding dianion and to the reduction of the
benzothiophene core~\cite{128}.

\begin{scheme*}
{\vspace*{-.2pc}}
\includegraphics{sc17}
{\vspace*{.5pc}}
\caption{\label{sch17}(a)~Cyclic voltammetry of compounds \textbf{2-Ph}
(red) and \textbf{1-Ph} (blue). (b)~Putative mechanism for the
electrochemical conversion of \textbf{2-Ph} into \textbf{1-Ph}.}
{\vspace*{-.8pc}}
\end{scheme*}

We next investigated the oxidation of the organometallic core for the
same compounds but working in dichloromethane this time, in line with
our previous studies~\cite{92,129} (Table~\ref{tab2}). The redox
potential of \textbf{1-Ph} was measured first, and the 0.25~V value
versus FcH/FcH$^+$ was found in good agreement with the results of
Herberhold and Jahn, \mbox{although} their measurements were done in
acetonitrile instead of dichloromethane~\cite{123,124}. As expected,
the redox potential of \textbf{1-Ph} falls between the ones of
\textbf{1-\textit{p}OMePh} and
\textbf{1-\textit{p}CF}$_{\mathbf{3}}$\textbf{Ph} due to their
respective electron-donating and electron-withdrawing groups attached
to the phenyl ring.\ The thioketone \textbf{4-Ph} was more easily
oxidized than the parent \mbox{ketone} ($E_{1/2}$~0.22 versus 0.25~V),
probably due to the stronger electron-withdrawing properties of the
former~\cite{130}. Surprisingly, the redox potential of the pyridyl
derivative (\textbf{1-2Py}) was 50~mV lower than that of
benzoylferrocene (\textbf{1-Ph}). Indeed, due to the more pronounced
electron-withdrawing properties of the pyridine ring, the reverse order
was expected. However, electronic repulsion between the lone pairs of
the ketone and the pyridine's nitrogen might promote a conformation
similar to the one observed for \textbf{2-2Py} (see below), avoiding
the full transmission of the pyridine's electronic effect to the
ferrocene core. The ferroceno[\textit{c}]quinolines \textbf{11} was
found easier to oxidize than any of the ferrocene ketones studied,
probably due to the planar structure of the tricyclic core and to the
presence of an imine instead of a ketone. In the ferrocene series, it
is usually possible to link the $E_{1/2}$ value of monosubstituted
derivatives with the Hammett's parameter $\alpha_{\mathrm{p}}$~\cite{122,131,132}
and the $E_{1/2}$ of polysubstituted compounds with the sum of
$\alpha_{\mathrm{p}}$ or $\alpha_{\mathrm{p}} + \alpha_{\mathrm{m}}$, depending on the
substation pattern~\cite{133,134,135}.\ Unfortunately, only a
limited number of Hammett's parameters is known for ketones~\cite{130}.
However, it was still possible to find a correlation between the
recorded $E_{1/2}$ values and the sum of $\alpha_{\mathrm{p}}$ parameters for
compounds \textbf{1-\textit{t}Bu}, \textbf{1-Ph}, \textbf{2-Ph}, and
\textbf{1-CF}$_{\mathbf{3}}$ with the equation $E_{1/2} =
1.8466~\Sigma\alpha_{\mathrm{p}} - 0.026$ ($R^2 = 0.9857$) (see Supplemtary
material).\looseness=1

\begin{figure*}
\includegraphics{fig04}
\caption{\label{fig4}Molecular structure of compounds
\textbf{1-\textit{p}CF}$_{\mathbf{3}}$\textbf{Ph} (left),
\textbf{1-\textit{m}CF}$_{\mathbf{3}}$\textbf{Ph} (middle), and
\textbf{1-\textit{o}CF}$_{\mathbf{3}}$\textbf{Ph} (right) in the solid
state. Thermal ellipsoids shown at the 30\% probability level. Selected
bond lengths ({\AA}) and angles ({\textdegree}) for
\textbf{1-\textit{p}CF}$_{\mathbf{3}}$\textbf{Ph}: C10--C11 ${=}$
1.469(4), C10--Cg2${\cdots}$Cg1--C1 ${=}$ 4.71 (Cg1 being the centroid of
the C1--C2--C3--C4--C5 ring and Cg2 the one of the C6--C7--C8--C9--C10
ring), C6--C10--C11--O12 ${=}$ 12.9(4), O12--C11--C13--C14 ${=}$ 27.7(4);
for \textbf{1-\textit{m}CF}$_{\mathbf{3}}$\textbf{Ph}: C10--C11 ${=}$
1.469(3), C10--Cg2${\cdots}$Cg1--C1 ${=}$ 3.19, C9--C10--C11--O12 ${=}$
18.0(3), O12--C11--C13--C18 ${=}$ 20.2(3); for
\textbf{1-\textit{o}CF}$_{\mathbf{3}}$\textbf{Ph}: C10--C11 ${=}$
1.471(8), C10--Cg2${\cdots}$Cg1--C3 ${=}$ 7.58, C6--C10--C11--O12 ${=}$
2.5(8), O12--C11--C13--C18 ${=}$ 53.0(7).}
{\vspace*{.25pc}}
\end{figure*}

\begin{table*}%tab2
\tabcolsep23pt
\caption{\label{tab2}Electrochemical data (in~V) for the oxidation of
selected ferrocene ketones}
\begin{tabular}{ccccc}
\thead
Compound & ${E_{\mathrm{pa}}}^{\mathrm{a}}$ &
${E_{\mathrm{pc}}}^{\mathrm{a}}$ &
${i_{\mathrm{pa}}/i_{\mathrm{pc}}}^{\mathrm{a}}$ & 
${E_{1/2}}^{\mathrm{b}}$\\
\endthead
\textbf{1-Ph} & 0.30 & 0.20 & 0.88 & 0.25\\
\textbf{1-\textit{p}OMePh} & 0.27 & 0.17 & 0.84 & 0.21\\
\textbf{1-\textit{p}CF}$_{\mathbf{3}}$\textbf{Ph} & 0.34 & 0.24 & 0.75 & 0.29\\
\textbf{1-2Py} & 0.26 & 0.17 & 0.89 & 0.20\\
\textbf{1-\textit{t}Bu} & 0.25 & 0.16 & 0.90 & 0.22\\
\textbf{1-CF}$_{\mathbf{3}}$ & 0.50 & 0.40 & 0.81 & 0.44\\
\textbf{4-Ph} & 0.29 & nd$^{(\mathrm{c})}$ & nd$^{(\mathrm{c})}$ & 0.22\\
\textbf{2-Ph} & 0.38 & 0.29 & 0.84 & 0.33\\
\textbf{10} & 0.30 & 0.20 & 0.89 & 0.25\\
\textbf{11} & 0.24 & 0.14 & 0.97 & 0.17\\
\textbf{12} & 0.37 & 0.28 & 0.86 & 0.33{\vspace*{.2pc}}\\
\multicolumn{5}{c}{\protect\inlinefig{fx02}}
\botline
\end{tabular}
\tabnote{Potential values given relative to FcH/FcH$^+$, scan rate ${=}$
100~mV$\cdot$s$^{-1}$. ${}^{\mathrm{a}}$From CV experiments.
${}^{\mathrm{b}}$From DPV experiments. ${}^{\mathrm{c}}$Irreversible
oxidation was observed.}
\end{table*}

\subsection{Solid-state structures and weak interactions of some
ferrocenyl ketones} \label{sec2.5}

\begin{figure*}
\includegraphics{fig05}
\caption{\label{fig5}Halogen--oxygen bond observed in the solid state
for compound \textbf{2}$'$\textbf{-2Py}. Thermal ellipsoids shown at
the 30\% probability level. Selected bond lengths ({\AA}) and angles
({\textdegree}) O12${\cdots}$I2 ${=}$ 3.051, O32${\cdots}$I1 ${=}$ 3.061,
C11--O12${\cdots}$I2 147.99, C31--O32${\cdots}$I1 146.03,
O12${\cdots}$I2--C38 156.03, O32${\cdots}$I1--C18 160.76.}
\end{figure*}

In the frame of this work, many ferrocene ketone derivatives were found
to produce crystals suitable for XRD analysis, some of them deserving
\mbox{additional} comments.\ Regarding unsubstituted ferrocene ketones, an
eclipsed conformation of the ferrocene core was identified in most
cases and the \mbox{substitution} pattern of the phenyl ring was found
to have an impact on the solid-state structure.\ For \textit{para}-
and \mbox{\textit{meta}-substituted} aromatics, the C${=}$O bond was
found tilted above the substituted cyclopentadienyl (Cp) ring while the
phenyl ring was slightly inclined compared to the C${=}$O bond. However,
for \textit{ortho}-substituted aromatics, the carbonyl was found
aligned with the Cp ring while the aromatic was much more tilted,
probably to reduce the steric clash between the carbonyl group and the
\textit{ortho} substituent. The case of the three trifluoromethylated
derivatives is especially representative of this general trend
(Figure~\ref{fig4}).

The iodinated derivative \textbf{2}$'$\textbf{-2Py} further features an
intermolecular halogen--oxygen bond resulting from the interaction
between the region of positive electrostatic potential ($\sigma$-hole)
of iodine, acting as the donor, and a lone pair of the oxygen of the
ketone, acting as the acceptor~\cite{136}. While such interactions are
not usually observed with bare iodopyridines, they are more common from
iodopyridinium derivatives in which the electron-withdrawing pyridinium
ring increases the electrostatic potential of the iodine and thus the
strength of the interaction~\cite{137,138}. In
\textbf{2}$'$\textbf{-2Py}, the ketone adjacent to the iodine atom is
expected to have a similar effect, leading to a zigzag halogen-bond
network connecting all the molecules in the solid state
(Figure~\ref{fig5}).

As expected, the introduction of the iodine next to the ketone induced
some structural changes in the solid state, as observed between
\textbf{1-\textit{p}CF}$_{\mathbf{3}}$\textbf{Ph} and
\textbf{2-\textit{p}CF}$_{\mathbf{3}}$\textbf{Ph} (Figure~\ref{fig6}).
Indeed, to accommodate the iodine atom, which was inclined by
4.9{\textdegree} above the Cp ring, the C${=}$O bond was forced to move
from its tilted position to be aligned with the Cp ring. A large change
in the orientation of the phenyl ring was also observed between the two
structures.

\begin{figure}
{\vspace*{.3pc}}
\includegraphics{fig06}
{\vspace*{.3pc}}
\caption{\label{fig6}Molecular structure of compounds
\mbox{\textbf{1-\textit{p}CF}$_{\mathbf{3}}$\textbf{Ph}} (top) and
\textbf{2-\textit{p}CF}$_{\mathbf{3}}$\textbf{Ph} (bottom) in the solid
state. Thermal ellipsoids shown at the 30\% probability level. Selected
bond lengths ({\AA}) and angles ({\textdegree}) for
\textbf{2-\textit{p}CF}$_{\mathbf{3}}$\textbf{Ph}: C10--C11 ${=}$
1.49(1), C9--I1 ${=}$ 2.082(7), C10--Cg2${\cdots}$Cg1--C2 ${=}$ 3.35 (Cg1
being the centroid of the C1--C2--C3--C4--C5 ring and Cg2 the one of
the C6--C7--C8--C9--C10 ring), Cg2--C9--I1 ${=}$ 175.11,{\break}
C9--C10--C11--O12 ${=}$ ${-}$3(1), O12--C11--C13--C14 ${=}$ ${-}$138.0(8).}
{\vspace*{-.6pc}}
\end{figure}

The substitution pattern of the phenyl ring was also found to influence
the solid-state structures of the iodoferrocene ketones. Indeed, except
for \mbox{derivative} \textbf{2-\textit{p}CF}$_{\mathbf{3}}$\textbf{Ph}, the
C${=}$O bond was found tilted from the Cp ring but aligned with the
phenyl ring for all the \textit{para}-substituted derivatives, while
the contrary was identified for all the \textit{ortho}-substituted
derivatives, probably for steric encumbrance reasons. The
\textbf{2-\textit{p}OMePh} and \textbf{2-\textit{o}OMePh} structures
depicted in Figure~\ref{fig7} nicely illustrate this trend. The
opposite behavior of \textbf{2-\textit{p}CF}$_{\mathbf{3}}$\textbf{Ph}
might be rationalized in terms of substituent effects. Indeed, due to
its strong electron-withdrawing inductive effect, the trifluoromethyl
group might favor the resonance between the Cp ring of the
organometallic and the C${=}$O bond, which therefore need to be aligned.
For the other \textit{para}-substituted aryl derivatives studied, all
substituents have a positive mesomeric effect which might override the
donating effect of the \mbox{ferrocene} core, explaining why the C${=}$O
bond is therefore aligned with the phenyl ring.

\begin{figure}
\includegraphics{fig07}
\caption{\label{fig7}Molecular structure of compounds
\mbox{\textbf{2-\textit{p}OMePh}} (top) and \textbf{2-\textit{o}OMePh}
(bottom) in the solid state. Thermal ellipsoids shown at the 30\%
probability level. Selected bond lengths ({\AA}) and angles
({\textdegree}) for \textbf{2-\textit{p}OMePh}: C10--C11 ${=}$ 1.490(3),
C6--I1 ${=}$ 2.089(2), C10--Cg2${\cdots}$Cg1--C2 ${=}$ 3.12 (Cg1 being the
centroid of the C1--C2--C3--C4--C5 ring and Cg2 the one of the
C6--C7--C8--C9--C10 ring), Cg2--C6--I1 ${=}$ 178.23, C6--C10--C11--O12
${=}$ ${-}$33.5(3), O12--C11--C13--C18 ${=}$ ${-}$1.1(3); for
\textbf{2-\textit{p}OMePh}: C30--C31 ${=}$ 1.492(7), C26--I2 ${=}$
2.096(4), C30--Cg4${\cdots}$Cg3--C23 ${=}$ 22.47 (Cg3 being the centroid of
the C21--C22--C23--C24--C25 ring and Cg4 the one of the
C26--C27--C28--C29--C30 ring), Cg4--C26--I2 ${=}$ 176.92,
C26--C30--C31--O32 ${=}$ ${-}$10.1(8), O32--C31--C33--C38 ${=}$ ${-}$105.7(7).}
{\vspace*{-.7pc}}
\end{figure}

From racemic \textbf{2-\textit{o}ClPh}, preferential crystallization
of the enantiopure $S_{\mathrm{P}}$ enantiomer was pleasingly observed~\cite{139},
allowing us to identify the formation of a halogen--oxygen bond in the
solid state (Figure~\ref{fig8}).\ Due to the electron-richness of the
organometallic core, iodoferrocenes are usually not able to develop
such interactions. However, the presence of electron-withdrawing
substituents can \mbox{substantially} increase the positive electrostatic
potential of iodine's $\sigma$-hole.\ As a result, halogen bond
{networks} have been recently identified for various
iodoferrocenes substituted with sulfonamides~\cite{140},
sulfonates~\cite{92}, sulfoxides~\cite{141}, sulfonyl
fluoride~\cite{134}, and triflones~\cite{87}. Similar bonds were also
observed for ferrocene iodoalkyne derivatives~\cite{142}. In the case
of $\mathbi{S}_{\mathbf{p}}$\textbf{-2-\textit{o}ClPh},
having the iodine and the ketone groups in a same plane and pointing in
the same direction led to a zigzag chain of halogen--oxygen bonds with
bonds lengths and angles being in the range of classical
values~\cite{136}.

\begin{figure*}
{\vspace*{-.2pc}}
\includegraphics{fig08}
{\vspace*{-.2pc}}
\caption{\label{fig8}Halogen--oxygen bonds observed in the solid state
for compound
$\mathbi{S}_{\mathbf{P}}$\textbf{-2-\textit{o}ClPh}.
Thermal ellipsoids shown at the 30\% probability level. Selected bond
lengths ({\AA}) and angles ({\textdegree}) O12${\cdots}$I1 ${=}$ 2.999,
C11--O12${\cdots}$I1 139.63, O12${\cdots}$I1--C9 173.64.}
{\vspace*{-.05pc}}
\end{figure*}

\looseness=-1
We were pleased to identify the two types of halogen--halogen
interactions in three of our iodoferrocene ketones~\cite{143,144}. In
the solid state, both enantiomers of \textbf{2-\textit{p}ClPh} were
identified in the crystal structure, with the $R_{\mathrm{P}}$ enantiomer
interacting with the $S_{\mathrm{P}}$ enantiomer via a type~II iodine--iodine
interaction, characterized by two different C--I${\cdots}$I angle
values (Figure~\ref{fig9}a). However, while the two enantiomers of
\textbf{2}$''$\textbf{-2BTh} crystallized together in a similar manner,
a type~I iodine--iodine interaction with similar \mbox{C--I${\cdots}$I}
angles was observed between the iodine atoms \mbox{attached} to the
benzothiophene moiety and not between those linked to the ferrocene
core (\mbox{Figure}~\ref{fig9}b).

\begin{figure*}
\includegraphics{fig09}
{\vspace*{-.2pc}}
\caption{\label{fig9}Iodine--iodine interaction network observed in the
solid state for compounds \textbf{2-\textit{p}ClPh} (top) and
\textbf{2}$''$\textbf{-2BTh} (bottom). Thermal ellipsoids shown at the
30\% probability level. Selected bond lengths ({\AA}) and angles
({\textdegree}) for \textbf{2-\textit{p}ClPh}: I1${\cdots}$I2 ${=}$ 3.882,
I1${\cdots}$I2--C26 173.07, C6--I1${\cdots}$I2 73.43; for
\textbf{2-\textit{p}ClPh}: I2${\cdots}$I2 ${=}$ 3.772, I2${\cdots}$I2--C14
137.38, C14--I2${\cdots}$I2 137.38.}
{\vspace*{-.2pc}}
\end{figure*}

\begin{figure*}%%%
{\vspace*{.2pc}}
\includegraphics{fig10}
{\vspace*{.1pc}}
\caption{\label{fig10}Iodine--bromine bond network observed in the
solid state for compounds
$\mathbi{R}_{\mathbf{P}}$\textbf{-2-\textit{o}BrPh}
(a) and
$\mathbi{S}_{\mathbf{P}}$\textbf{-2-\textit{o}BrPh}
(b) and sense of the helix formed. Thermal ellipsoids shown at the 30\%
probability level; hydrogen atoms omitted for clarity. Selected bond
lengths ({\AA}) and angles ({\textdegree}) for
$\mathbi{R}_{\mathbf{P}}$\textbf{-2-\textit{o}BrPh}:
Br1${\cdots}$I1 ${=}$ 3.594, C18--Br1${\cdots}$I1 171.16, C9--I1${\cdots}$Br1
98.79; for
$\mathbi{S}_{\mathbf{P}}$\textbf{-2-\textit{o}BrPh}:
Br1${\cdots}$I1 ${=}$ 3.599, C17--Br1${\cdots}$I1 170.96, C9--I1${\cdots}$Br1
99.35.}
{\vspace*{.1pc}}
\end{figure*}

From racemic \textbf{2-\textit{o}BrPh}, spontaneous resolution was
observed and the two enantiomers were separately isolated as
conglomerates.\ As expected, single molecules of each enantiomer were
almost similar in the solid state, both crystallizing in the same
\mbox{tetragonal} system, although in different space groups. However, the
most interesting \mbox{characteristics} were observed having a closer
look at intermolecular halogen--halogen bonds for each enantiomer.
Indeed, the bromine of one molecule was found to develop a type~I
interaction with the \mbox{iodine} atom of \mbox{another} molecule,
leading to the helix arrangement of all molecules. While ferrocene
derivatives have previously been involved in the formation of such
structures~\cite{145,146,147,148}, the involvement of
\mbox{halogen--halogen} interactions to structure the arrangement is
pretty unusual. The planar chirality of the ferrocene core was further
found to influence the axial chirality of the helix, the $R_{\mathrm{P}}$
enantiomer giving rise to a \textit{M} helix (Figure~\ref{fig10}a),
while a \textit{P} helix was observed for the $S_{\mathrm{P}}$ enantiomer
(Figure~\ref{fig10}b).

It was finally possible to grow crystals of the enantiopure tetracycle
\textbf{12} suitable for XRD \mbox{analysis} (Figure~\ref{fig11}).\ Not
only the solid-state structure validated the expected $R_{\mathrm{P}}$
configuration of the compound but it also allowed the identification of
chalcogen--chalcogen interactions~\cite{149,150}. Indeed, a
sulfur${\cdots}$sulfur interaction between two molecules was likely to
happen due to the S1${\cdots}$S2 distance below the van der Waals radii
(3.54 versus 3.60~{\AA})~\cite{151}, although the various C--S${\cdots}$S
angles are shorter than expected, probably due to the rigid
fused-thiophene unit.\ The tetracyclic systems of the two molecules
composing the dimer were found perpendicular and, although this
arrangement is not usual, it was previously identified in other
benzothiophene \mbox{derivatives}~\cite{152}.

\begin{figure}
\includegraphics{fig11}
\vspace*{-2pt}
\caption{\label{fig11}Chalcogen--chalcogen bond observed in the solid
state for compound
$\mathbi{R}_{\mathbf{P}}$\textbf{-12}. Thermal
ellipsoids shown at the 30\% probability level.\ Selected bond lengths
({\AA}) and angles ({\textdegree}): S1${\cdots}$S2 ${=}$ 3.547,
C13--S1${\cdots}$S2 124.90, C20--S1${\cdots}$S2 144.94, C63--S2${\cdots}$S1
105.11, C70--S2${\cdots}$S1 76.54, angle plane
(C9--C8--C7--C6--C10--C11--C13--S1--C20--C19--C18--C17--C16--C15--C14)-(C56--C57--C58--C59--C60--C61--C63--S2--C70--C69--C68--C67--C66--C65--C64)
89.24.}
\vspace*{-2pt}
\end{figure}

\vspace*{-2pt}

\section{Conclusion} \label{sec3}

\vspace*{-2pt}

Here we have presented the first in-depth study of the
deprotolithiation of ferrocene ketones using a bulky, non-nucleophilic
lithium amide, as well as an in-situ trap to prevent nucleophilic
attack of the function by the ferrocenyllithium formed. The expected
iodoferrocene derivatives were obtained in most cases although the
presence of electron-acceptor substituents on the aryl moiety was found
to reroute the functionalization on this cycle, in agreement with our
DFT calculations (p$K_{\mathrm{a}}$ values after coordination to
lithium).

The development of an enantioselective version of the reaction was next
attempted using a chiral lithium amide. Although our best \textit{ee}
did not exceed 60\%, the feasibility of this approach was demonstrated.
Finally, we took advantage of the installation of the iodine on the
(hetero)aryl ring to reach\break {original} ferrocene-fused heterocycles,
including a tetracycle obtained by enantioselective C--H
functionalization.

Given the significant potential of both ferrocene ketones~\cite{65} and
ferrocene-fused heterocycles~\cite{153}, the current study is expected
to pave the way for future work toward such compounds to promote their
application in various fields.

\vspace*{-2pt}

\section*{Acknowledgments}

\vspace*{-2pt}

We acknowledge BASF (generous gift of
di[(\textit{S})-1-phenylethyl]amine and
di[(\textit{R})-1-phenylethyl]amine) and Thermofisher (generous gift of
2,2,6,6-tetramethylpiperidine).\ We acknowledge Olivier Perez, Carmelo
Prestipino and Jean-Fran\c{c}ois Lohier (CRISMAT, UMR CNRS 6508) and
well as Magali \mbox{Allain} (Moltech-Anjou, UMR CNRS 6200) for their
help in X-ray data collection.

\vspace*{-2pt}

\section*{Declaration of interests}

\vspace*{-2pt}

The authors do not work for, advise, own shares in, or receive funds
from any organization that could benefit from this article, and have
declared no affiliations other than their research organizations.

\vspace*{-2pt}

\section*{Funding} 

\vspace*{-2pt}

This work was supported by the Direction G\'{e}n\'{e}rale de la
Recherche Scientifique et du D\'{e}veloppement Technologique (MH),
Rennes M\'{e}tropole (WE), the University of Carthage and the Tunisian
Ministry of Higher Education and Scientific Research (SB), the Fonds
Europ\'{e}en de D\'{e}veloppement R\'{e}gional (FEDER; D8 VENTURE
Bruker AXS diffractometer), the Universit\'{e} de Rennes and the Centre
National de la Recherche Scientifique (WE, J-PH, MB, TR, FM). 

\vspace*{-2pt}

\section*{Supplementary materials}

\vspace*{-2pt}

Supporting information for this article is available on the journal's
website under \printDOI\ or from the author.

\CDRsupplementaryTwotypes{supplementary-material}{\cdrattach{crchim-429-suppl.pdf}}

The CCDC files 2490197 (\textbf{1-\textit{o}OMePh}), 2490198
(\textbf{1-\textit{p}ClPh}), 2490199 (\textbf{1-\textit{m}BrPh}),
2490200 (\textbf{1-\textit{o}CF}$_{\mathbf{3}}$\textbf{Ph}), 2490201
(\textbf{1-\textit{m}CF}$_{\mathbf{3}}$\textbf{Ph}), 2490202
(\textbf{1-\textit{p}CF}$_{\mathbf{3}}$\textbf{Ph}), 2490203
(\textbf{1-\textit{o}FPh}), 2490204 (\textbf{1-2BTh}), 2490205
(\textbf{1-C}${\equiv}$\textbf{CPh}), 2490206 [$
\mathbi{R}_{\mathbf{P}}$\textbf{-2-Ph}], 2490207
(\textbf{2-\textit{o}OMePh}), 2490208 (\textbf{2-\textit{p}OMePh}),
2490209 (\textbf{2-\textit{o}ClPh}), 2490210
(\textbf{2-\textit{p}ClPh}), 2490211
[$\mathbi{R}_{\mathbf{P}}$\textbf{-2-\textit{o}BrPh}],
2490212
[$\mathbi{S}_{\mathbf{P}}$\textbf{-2-\textit{o}BrPh}],
2490213 (\textbf{2-\textit{p}BrPh}), 2490214
(\textbf{2-\textit{p}CF}$_{\mathbf{3}}$\textbf{Ph}), 2490215
(\textbf{2}$'$\textbf{-2Py}), 2490216 (\textbf{2}$'$\textbf{-2BTh}),
2490217 (\textbf{2}$''$\textbf{-2BTh}) and 2490218 (\textbf{12})
contain the supplementary crystallographic data. These data can be
obtained free of charge from the Cambridge Crystallographic Data Center
via \href{https://www.ccdc.cam.ac.uk/structures}
{www.ccdc.cam.ac.uk/structures}.

\back{}

\printbibliography
\refinput{crchim20250763-reference.tex}

\end{document}