\makeatletter
\@ifundefined{HCode}
{\documentclass[screen,CRCHIM,Unicode,biblatex,published]{cedram}
\addbibresource{crchim20250567.bib}
\newenvironment{noXML}{}{}
\def\tsup#1{$^{{#1}}$}
\def\tsub#1{$_{{#1}}$}
\def\ndash{\text{--}}
\usepackage[T1]{fontenc}
\def\thead{\noalign{\relax}\hline}
\def\tbody{\noalign{\relax}\hline}
\def\endthead{\noalign{\relax}\hline}
\def\tabnote#1{\vskip4pt\parbox{.77\linewidth}{#1}}
\RequirePackage{etoolbox}
\usepackage{stackengine}
\def\jobid{crchim20250567}
%\graphicspath{{/tmp/\jobid_figs/web/}}
\graphicspath{{./figures/}}
\newcounter{runlevel}
\let\MakeYrStrItalic\relax
\csdef{Seqnsplit}{\\}
\def\refinput#1{}
\def\back#1{}
\def\botline{\\\hline}
\def\dollar{\$}
\def\og{\guillemotleft}
\def\fg{\guillemotright}
\def\mn{\phantom{$-$}}
\def\0{\phantom{0}}
\def\ubreak{\unskip\break}
\def\xmorerows#1#2{#2}
\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}}
\usepackage{hyperref}
\makeatletter
\g@addto@macro{\UrlBreaks}{\UrlOrds}
\gappto{\UrlBreaks}{\UrlOrds}
\DOI{10.5802/crchim.415}
\datereceived{2025-07-11}
\daterevised{2025-07-28}
\dateaccepted{2025-08-19}
\ItHasTeXPublished
}
{\documentclass[crchim]{article}
\usepackage[T1]{fontenc}
\usepackage{stackengine}
\def\CDRdoi{10.5802/crchim.415}
\def\DJ{\unient{00D0}}
\let\refinput\input
\let\ubreak\relax
\makeatletter
\def\CDRsupplementaryTwotypes#1#2{}
}
\makeatother

\usepackage{upgreek}

\dateposted{2025-11-04}
\begin{document}

\begin{noXML}

\CDRsetmeta{articletype}{research-article}


\title{In search of a viable route to high-melting olefinic block
copolymers}

\alttitle{\`{A} la recherche d'une voie viable vers des copolym\`{e}res
\`{a} blocs d'ol\'{e}fines \`{a} point de fusion \'{e}lev\'{e}}

\author{\firstname{Rafael} \lastname{Sablong}\CDRorcid{0000-0002-5718-124X}}
\address{PTG Eindhoven B.V., De Lismortel 31, 5612 AR Eindhoven, The Netherlands}

\author{\firstname{Wojciech} \lastname{Szot}\CDRorcid{0009-0000-1829-0999}}
\address{SABIC Technology \& Innovation, STC Geleen, Urmonderbaan 22,
6167 RD, Geleen, The Netherlands}
\address{Laboratory of Applied Macromolecular Materials, Chemical
Faculty, Gdansk University of Technology, G. Narutowicza Str. 11/12,
80-233 Gdansk, Poland}

\author{\firstname{Rob} \lastname{Duchateau}\CDRorcid{0000-0002-2641-4354}\IsCorresp}
\addressSameAs{2}{SABIC Technology \& Innovation, STC Geleen, Urmonderbaan 22,
6167 RD, Geleen, The Netherlands}
\address{Chemical Product Engineering, Department of Chemical
Engineering, University of Groningen, Nijenborgh 4, 9747 AG Groningen,
The Netherlands}
\email[R. Duchateau]{Rob.Duchateau@SABIC.com}

\shortrunauthors

\thanks{SABIC}

\keywords{\kwd{Olefinic block copolymer}
\kwd{Coordinative chain transfer polymerization}
\kwd{4-methyl-1-pentene (4M1P)}
\kwd{1-octene}}

\altkeywords{\kwd{Copolym\`{e}re \`{a} blocs d'ol\'{e}fines}
\kwd{Polym\'{e}risation coordinative par transfert de cha\^{i}ne}
\kwd{4-m\'{e}thyl-1-pent\`{e}ne (4M1P)}
\kwd{1-oct\`{e}ne}}

\begin{abstract}
This paper describes the coordinative chain transfer polymerization
(CCTP) of 1-octene and 4-methyl-1-pentene using three selected
catalysts. Special attention was devoted to the reversibility and the
rate of the polymeryl group transfer between the catalyst and the
zinc-based chain transfer agent (CTA) as this determines whether the
system could be used to produce corresponding olefinic block copolymers
(OBCs) based on higher \tralicstex{α}{${\upalpha}$}-olefins, 1-octene and
4-methyl-1-pentene. All the three tested catalysts undergo chain
transfer to zinc during the polymerization of 1-octene and
4-methyl-1-pentene. Depending on the catalyst and monomer used, chain
transfer to zinc appears to be either reversible or irreversible and,
in all cases, the chain transfer and chain growth rate constants are of
the same order of magnitude. All catalysts suffer from concurrent
\tralicstex{β}{${\upbeta}$}-H transfer processes, which can be attributed to the steric
hindrance of the applied \tralicstex{α}{${\upalpha}$}-olefins. For one of the
catalysts, \tralicstex{β}{${\upbeta}$}-H transfer after 2,1-misinsertion appeared to be
the prevalent competing chain transfer process. Adding a small amount
of ethylene clearly reactivated the dormant site formed after
2,1-misinsertion without terminating the growing polymer chain, which
opens the route to the formation of higher \tralicstex{α}{${\upalpha}$}-olefin-based
OBCs.
\end{abstract}

\begin{altabstract}
Cet article d\'{e}crit la polym\'{e}risation coordinative par transfert
de cha\^{i}ne (CCTP) du 1-oct\`{e}ne et du 4-m\'{e}thyl-1-pent\`{e}ne
en utilisant trois catalyseurs s\'{e}lectionn\'{e}s. Une attention
particuli\`{e}re a \'{e}t\'{e} accord\'{e}e \`{a} la
r\'{e}versibilit\'{e} et \`{a} la vitesse du transfert de la chaine
polym\`{e}re entre le catalyseur et l'agent de transfert de cha\^{i}ne
(CTA) \`{a} base de zinc, car cela d\'{e}termine si le syst\`{e}me peut
\^{e}tre utilis\'{e} pour produire des copolym\`{e}res \`{a} blocs
d'ol\'{e}fines (OBCs) correspondants bas\'{e}s sur des
\tralicstex{α}{${\upalpha}$}{-ol\'{e}fines sup\'{e}rieures, 1-oct\`{e}ne et
4-m\'{e}thyl-1-pent\`{e}ne. Les trois catalyseurs test\'{e}s subissent
un transfert de cha\^{i}ne vers le zinc lors de la polym\'{e}risation
du 1-oct\`{e}ne et du 4-m\'{e}thyl-1-pent\`{e}ne. Selon le catalyseur
et le monom\`{e}re utilis\'{e}s, le transfert de cha\^{i}ne vers le
zinc semble \^{e}tre soit r\'{e}versible soit irr\'{e}versible et, dans
tous les cas, les constantes de transfert de cha\^{i}ne et de
croissance de cha\^{i}ne sont du m\^{e}me ordre de grandeur. Tous les
catalyseurs souffrent de processus de transfert} \tralicstex{β}{${\upbeta}$}{-H
simultan\'{e}s, qui peuvent \^{e}tre attribu\'{e}s \`{a} l'encombrement
st\'{e}rique des} \tralicstex{α}{${\upalpha}$}{-ol\'{e}fines utilis\'{e}es. Pour l'un
des catalyseurs, le transfert} \tralicstex{β}{${\upbeta}$}{-H apr\`{e}s une mauvaise
insertion 2,1 semble \^{e}tre le processus de transfert de cha\^{i}ne
concurrent dominant. L'ajout d'une petite quantit\'{e}
d'\'{e}thyl\`{e}ne a clairement r\'{e}activ\'{e} le site dormant
form\'{e} apr\`{e}s la mauvaise insertion 2,1 sans mettre fin \`{a} la
cha\^{i}ne polym\`{e}re en croissance, ce qui ouvre la voie \`{a} la
formation d'OBCs bas\'{e}s sur des} \tralicstex{α}{${\upalpha}$}-ol\'{e}fines
sup\'{e}rieures.
\end{altabstract}

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

\maketitle

\twocolumngrid

\end{noXML}

\section{Introduction}\label{sec1}

Olefinic block copolymers (OBCs) is a general description for di-~or
multiblock copolymers composed of distinct polyolefin 
segments~\cite{1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16}. For example,
OBCs can be stereoblock copolymers~\cite{4}, or they can consist of
different semicrystalline blocks such as polyethylene (PE) and
isotactic \mbox{polypropylene} (iPP)~\cite{5,6,7,8,9,10,11}, but the OBCs
attracting the most interest are those composed of low glass transition
temperature ($T_{\mathrm{g}}$) soft blocks and high $T_{\mathrm{g}}$ or high melting
temperature ($T_{\mathrm{m}}$) hard blocks~\cite{12,13,14,15,16,17}. Depending on
the proportion of hard and soft segments, typically polyolefin
plastomers (POPs) or elastomers (POEs) are obtained. The major
difference between OBCs and corresponding random {copolymers} lies in the
fact that, for the same crystallinity, the $T_{\mathrm{m}}$ and hence the heat
deflection temperature of OBCs is \mbox{considerably} higher. As a result,
these materials have a wider service temperature window and
consequently a much better elastic recovery and creep resistance at
ambient temperature than random copolymers having the same
crystallinity~\cite{18,19,20,21}. There are various ways to produce
OBCs. Although not economically viable, the most straightforward
approach toward OBCs consists in using a living catalyst system under
changing feed  conditions~\cite{2,3,5,6,7,8,9,10,14}. With the
development of coordinative chain transfer polymerization
(CCTP)~\cite{22,23,24}, a pseudo-living polymerization process
accompanied by reversible chain transfer, the commercial synthesis of
OBCs has come within reach. Using CCTP, OBC's can be produced either
using two different catalysts (Dow Chemical's chain shuttling
polymerization)~\cite{3,4,11,12} or employing changing feed 
conditions~\cite{13,15,17}. As an example, ethylene/${\upalpha}$-olefin
(i.e., propylene, 1-octene)-based OBCs have been produced using CCTP on
commercial scale by Dow Chemical under the tradenames
Infuse{\texttrademark}~and Intune{\texttrademark}. \looseness=-1

As the upper service temperature of thermoplastic elastomers is
governed by the material's $T_{\mathrm{m}}$ or high $T_{\mathrm{g}}$, even OBCs containing
hard segments of polyethylene and polypropylene have their thermal
limitations. OBCs based on high-melting polyolefin blocks, such as
poly(4-methyl-1-pentene) (P4M1P, having a $T_{\mathrm{m}}$ above 
250~\textdegree C), would be
desirable as such thermoplastic elastomers could compete with
thermoplastic polyurethanes, silicones, and fluorocarbon elastomers in
various temperature-demanding applications. 

Whereas ethylene/${\upalpha}$-olefin OBCs, having ethylene-rich
crystallizable blocks, can be prepared using chain shuttling
polymerization, ethylene/${\upalpha}$-olefins, having crystallizable
poly(${\upalpha}$-olefin) sequences, such as P4M1P,
cannot be synthesized by this approach as the
ethylene/${\upalpha}$-olefin reactivity ratio is typically considerably
larger than one. To produce the desired block copolymers using chain
shuttling polymerization would require a catalyst that prefers
${\upalpha}$-olefins such as 4-methyl-1-pentene (4M1P) over ethylene,
which is very unlikely to find. Yet, the random copolymerization of
ethylene and 4M1P does not necessarily give statistical random
microstructures. Depending on the \mbox{catalyst} structure,
poly(ethylene-\textit{co}-4M1P)s with blocky microstructures have been
obtained, though the sequence length was typically too short to obtain
elastomeric properties~\cite{25,26,27}. Elastomeric multiblock
copolymers consisting of segmented high-melting crystallizable
P4M1P and amorphous or low-crystalline
poly(ethylene-\textit{co}-4M1P) blocks could potentially be produced
using sequential feed under CCTP conditions.

With the final objective in mind of finding a catalytic system that
could produce OBCs composed of high-$T_{\mathrm{m}}$ hard and low-$T_{\mathrm{g}}$ soft
blocks, the CCTP behavior of three selected catalysts regarding 4M1P
was studied.  To evaluate the effect of steric hindrance on the CCTP
behavior of the catalysts and the occurrence of competitive ${\upbeta}$-H
transfer side reaction, comparative experiments using non-branched
monomer 1-octene were used as benchmark. 

\section{Results and discussion}

To form multiblock copolymers using CCTP, conditions must be found
under which the polymerization proceeds in a highly controlled fashion.
${\upbeta}$-Hydrogen transfer or related chain termination processes
should be reduced to a minimum and the catalyst system should undergo
fast and reversible chain transfer to the chain transfer agent (CTA)
such that a pseudo-living polymerization process is obtained. Such
conditions not only depend on the catalyst applied, but also on the
type of monomer used~\cite{28,29}.


\begin{figure*}
\vspace*{-2pt}
\includegraphics{fig01}
\vspace*{-2pt}
\caption{\label{fig1}Structures of the catalyst precursors used
[Ph$_{2}$C(3-tBu-Cp)(2,7-tBu$_{2}$Flu)]ZrMe$_{2}$ (\textbf{1}),
[1-C$_{10}$H$_{6}$-5-(CMe$_2$-2,6-iPr$_{2}$-C$_{6}$H$_{3}
$)-C$_{6}$H$_{3}$N]HfMe$_{2}$ (\textbf{2}) and
\{CH$_{2}$[CH(Me)O-C$_{6}$H$_{4}$-2-(2-NC$_{12}$H$_{8}
$)-4-Me-C$_{6}$H$_{2}$O]$_{2}$\}HfCl$_{2}$ 
(\textbf{3}).}
\vspace*{-1pt}
\end{figure*}

Based on their potential as described in literature, the following
three catalysts have been selected for the polymerization of 4M1P and
1-octene under CCTP conditions: (\textbf{1}) a zirconocene:
[Ph$_{2}$C(3-tBu-Cp)(2,7-tBu$_{2}$-Flu)]ZrMe$_{2}$, (\textbf{2}) a
pyridylamido hafnium system:
[1-C$_{10}$H$_{6}$-5-(CMe$_2$-2,6-iPr$_{2}$-C$_{6}$H$_{3}$)-C$_{6}$H$_{3}$N]HfMe$_{2}$, 
and (\textbf{3}) a bis(phenolate)diether hafnium system:
\{CH$_{2}$[CH(Me)O-C$_{6}$H$_{4}$-2-(2-NC$_{12}$H$_{8}$)-4-Me-C$_{6}$H$_{2}$O]$_{2}$\}HfCl$_{2}$ 
(Figure~\ref{fig1}). Complex \textbf{1} was found to give
high-molecular-weight P4M1P~\cite{30}, and this type of catalyst is
known to undergo chain transfer with diethyl zinc (DEZ) as
CTA~\cite{31}. It was reported that \textbf{2} is able to produce PE/PP
multiblock copolymers in a living fashion under sequential feed
conditions and this type of catalyst---albeit with a slightly
different ligand structure---has been used commercially to
produce ethylene-based OBCs~\cite{8,12,13}. In addition, this
\mbox{catalyst} family is also known to 
\mbox{incorporate} \mbox{sterically} \mbox{hindered}
monomers like 4M1P, 3-methyl-1-butene, and vinyl
cyclohexane~\cite{32}. Finally, the third catalyst (\textbf{3}) has
shown excellent higher ${\upalpha}$-olefin \mbox{incorporation} abilities and
has also been applied in the production of ethylene-based
OBCs~\cite{33,34,35}.\looseness=-1

\subsection{Cyclopentadienyl-fluorenyl zirconocene (\textbf{1})}

The \textit{ansa}-zirconocene complex
[Ph$_{2}$C(3-tBu-Cp)(2,7-tBu$_{2}$Flu)]ZrMe$_{2}$ (\textbf{1};
Figure~\ref{fig1}), in combination with methylaluminoxane (MAO) as
activator, is known to polymerize higher ${\upalpha}$-olefins including
4M1P to high-molecular-weight products. This incited us to evaluate its
potential for the polymerization of 4M1P under CCTP conditions to
obtain a pseudo-living process in the presence of DEZ as CTA. The
influence of the steric hindrance during the polymerization was studied
by comparing the results of 4M1P and 1-octene polymerizations performed
under similar conditions.

The initial 4M1P polymerizations performed (Table~S1) demonstrated that
\textbf{1} activated with MAO indeed generates an active catalyst for
the polymerization of 4M1P and the activity increases with the presence
of increasing amounts of DEZ (Figure~S1A). The high-temperature
size-exclusion chromatography (HT-SEC) analyses of the polymer samples
show a dramatic effect of the increasing amounts of DEZ on the
polymer's molecular weight: it drops from 63~kDa in the absence of DEZ
to only 4~kDa for a Zn:Zr ratio of 200. Although the decrease in the
polydispersity ($\textit{\DJ}_{\mathrm{M}}$) from 2.4 to 1.8 was limited, 
a \textit{\DJ}$_{\mathrm{M}} < 2$
and a rough estimation of ${k}_{\mathrm{CT}}/{k}_{\mathrm{CG}}$ around 1.5
(Figure~S2)~\cite{36} indicate that the chain transfer
(${k}_{\mathrm{CT}}$) and chain growth (${k}_{\mathrm{CG}}$) rate constants are of the
same order of  magnitude~\cite{37}. As the molecular weight did not
increase with increasing polymerization time (Figure~S1B),
chain transfer is most likely quasi-irreversible, meaning that the
reversible exchange of ${\upbeta}$-branched polymeryls between catalyst
and CTA is considerably slower than the exchange of the initial ethyl
group of DEZ~\cite{28}.

\begin{figure*}
\vspace*{1pt}
\includegraphics{fig02}
\caption{\label{fig2}Molecular weight development versus conversion for
1-octene polymerization at 80~\textdegree C using \textbf{1}/MAO in the
(A) absence and (B) presence of DEZ (Table~\ref{tab1}).}
\end{figure*}

%tab1
\begin{table*}
\caption{\label{tab1}1-Octene polymerization with \textbf{1}/MAO(/DEZ)
at 80~\textdegree C}
\begin{tabular}{ccccccccc}
\thead
Exp. & \parbox[t]{1cm}{\centering DEZ  eq. Zr} & 
\parbox[t]{.9cm}{\centering \textit{t}  (min)} & 
\parbox[t]{.9cm}{\centering Yield  (g)} & 
\parbox[t]{1.7cm}{\centering Conversion$^{\mathrm{a}}$ (\%)}& 
\parbox[t]{1.2cm}{\centering ${M_{\mathrm{n}}}^{\mathrm{b}}$ (kDa)} & 
\parbox[t]{1.2cm}{\centering ${M_{\mathrm{w}}}^{\mathrm{b}}$ (kDa)} & 
{\textit{\DJ}$_{\mathrm{M}}$}$^{\mathrm{b}}$ &  
\parbox[t]{1.3cm}{\centering Chains per Zn}\vspace*{2pt}\\
\endthead
1 & - & 15 & 0.21 & \07 & 38.5 & 74.8 & 1.9 & - \\
2 & - & 30 & 0.54 & 19 & 38.4 & 73.0 & 1.9 & - \\
3 & - & 45 & 0.79 & 28 & 37.5 & 72.7 & 1.9 & - \\
4 & - & 60 & 1.05 & 37 & 35.6 & 70.9 & 2.0 & - \\
5 & 200 & 15 & 0.22 & \08 & \02.1 & \03.1 & 1.4 & 0.5 \\
6 & 200 & 30 & 0.32 & 11 & \02.3 & \03.4 & 1.5 & 0.7 \\
7 & 200 & 45 & 0.46 & 16 & \02.6 & \04.2 & 1.6 & 0.9 \\
8 & 200 & 60 & 0.63 & 22 & \03.0 & \05.2 & 1.8 & 1.1
\botline
\end{tabular}
\tabnote{Reaction conditions: 20~mL crimp-cap vial; 1-octene: 4~mL
(2.86 g); solvent: toluene (4~mL; ${V}_{\mathrm{reaction}}
\sim 8$~mL); \textbf{1}: 1.0~${\upmu}$mol; MAO: 500 equiv.;
stirring rate: 500 rpm; quenching: MeOH/HCl.}
\tabnote{$^{\mathrm{a}}$Determined by $^{1}$H NMR.
$^{\mathrm{b}}$Determined by size-exclusion chromatography.}
\end{table*}

Although no pseudo-living behavior was observed for 4M1P polymerization
using \textbf{1}/MAO/DEZ, the strong effect of DEZ on the molecular
weight encouraged us to perform polymerization experiments using the
less sterically hindered 1-octene at elevated temperatures to increase
the chain transfer rate. Table~\ref{tab2} shows the results for 1-octene
polymerization performed using \textbf{1}/MAO as catalyst system both
in the absence and presence of DEZ. In the presence of DEZ, the
catalytic activity was clearly lower, suggesting that heterobimetallic
catalyst--CTA adducts with a certain stability are 
formed~\cite{38,39,40}. This negative effect of DEZ on the activity is
opposite to what was observed during 4M1P polymerization. Most likely,
the larger steric hindrance of 4M1P reduces the stability of the
corresponding intermediate catalyst--CTA adduct. In the absence of DEZ,
the molecular weight slightly decreases over time, whereas in the
presence of DEZ, the molecular weight clearly increases with
progressing polymerization time (Figure~\ref{fig2}). The polydispersity
of the polymer obtained in the presence of DEZ is significantly lower
(\textit{\DJ}$_{\mathrm{M}}< 2$) than in the absence of DEZ, which suggests that reversible
chain transfer to zinc takes place for the less sterically hindered
1-octene. The fact that the molecular weight does not increase
proportionally with conversion agrees with a ${k}_{\mathrm{CT}}$ being of the
same order of magnitude as ${k}_{\mathrm{CG}}$~\cite{37}. Moreover, the
calculated number of chains per Zn for poly(1-octene) indicates that
only up to one ethyl group is involved in polymer chain exchange during
the polymerization at the given conditions (Table~\ref{tab1}). The limited
exchange of a second ethyl group is most likely caused by the enhanced
steric interactions between Zn(Et)-polymeryl species~\cite{28,29}.

%tab2
\begin{table*}
\caption{\label{tab2}Sequential and random 4M1P/ethylene
copolymerization with \textbf{2}/B(C$_{6}$F$_{5}$)$_{3}$/DiBAP\vspace*{2pt}}
\begin{tabular}{cccccccccc}
\thead
Exp. & \parbox[t]{1.1cm}{\centering 4M1P  (min)} & 
\parbox[t]{1.1cm}{\centering {$\mathrm{C}_{2}^{=}$}${}^{\mathrm{a}}$  (min)} & 
\parbox[t]{1.1cm}{\centering Yield  (g)} & \#$^{\mathrm{b}}$ & 
\parbox[t]{1.1cm}{\centering ${{M}_{\mathrm{n}}}^{\mathrm{c}}$  (kDa)} & 
\parbox[t]{1.1cm}{\centering ${{M}_{\mathrm{w}}}^{\mathrm{c}}$  (kDa)} & 
\textit{\DJ}$_{\mathrm{M}}^{\mathrm{c}}$ & 
\parbox[t]{1.1cm}{\centering ${T_{\mathrm{g}}}^{\mathrm{d}}$ (\textdegree C)} &
\parbox[t]{1.1cm}{\centering ${T_{\mathrm{m}}}^{\mathrm{d}}$ (\textdegree C)}\vspace*{2pt} \\
\endthead
\xmorerows{1}{1} & 1:10$^{\mathrm{e}}$ &  &  & 1 & 150.2 & 222.3 & 1.48 & - & - \\
 & 2:20$^{\mathrm{e}}$ &  & 2.11 & 2 & 318.0 & 533.8 & 1.68 & ${-}$37.2 & -\vspace*{3pt} \\
\xmorerows{1}{2} & 1:40\0 &  &  & 1 & 43.8 & 123.0 & 2.81 & \mn36.7 & 215.8 \\
 &  & 2:20 & 3.70 & 2 & 73.4 & 294.2 & 4.01 & \mn36.8 & 235.2\vspace*{3pt} \\
\xmorerows{2}{3} & 1:40\0 &  &  & 1 & 37.4 & 115.8 & 3.10 & \mn36.8 & 226.0 \\
 &  & 2:20 &  & 2 & 51.6 & 248.9 & 4.80 & \mn33.3 & 230.9 \\
 & 3:60\0 &  & 5.88 & 3 & 46.8 & 308.1 & 6.59 & \mn33.3 & 232.2 
\botline
\end{tabular}
\tabnote{Reaction conditions: 3-neck 200~mL round bottom flask; 2.5
${\upmu}$mol \textbf{2}; 1.0 equiv. of B(C$_{6}$F$_{5}$)$_{3}$; 8.0
equiv. of DiBAP (diisobutylaluminum
2,6-di-\textit{tert}-butyl-4-methylphenoxide); solvent: toluene (90
mL); 10~mL 4M1P; stirring rate: 200 rpm; $T = 25$~\textdegree C.}
\tabnote{$^{\mathrm{a}}${$\mathrm{C}_{2}^{=}$}\ ${=}$ ethylene,
pressure ${=}$ 1.3 bar.
$^{\mathrm{b}}$1, 2, 3: samples taken after feed \#1, 2 or 3.
$^{\mathrm{c}}$Determined by size-exclusion chromatography.
$^{\mathrm{d}}$Determined by differential scanning calorimetry.
$^{\mathrm{e}}$Ethylene was added together with 4M1P.}
\vspace*{2pt}
\end{table*}

Although CCTP conditions seem to have been achieved for the
polymerization of 1-octene, the existence of concurrent chain
terminations reactions was checked. The $^{1}$H NMR spectra of the
poly(1-octene)s produced clearly demonstrated the presence of
unsaturated-end-group resonances (Figure~S3). Besides the most
prominent vinylidene and vinylene end groups, trisubstituted olefin
resonances and small amounts of vinyl end-group resonances could be
observed. This indicates that ${\upbeta}$-hydrogen transfer forms the
prime concurrent chain termination process~\cite{41}. The same
unsaturated end-group resonances were found for P4M1Ps produced using
\textbf{1}/MAO/DEZ. 

In summary, although the catalyst clearly undergoes chain transfer to
zinc and it appears to be reversible in the case of 1-octene
polymerization, it is quasi-irreversible for 4M1P polymerization.
Furthermore, the competing chain termination by \mbox{${\upbeta}$-hydrogen}
transfer renders this catalyst unsuitable to produce OBCs from higher
${\upalpha}$-olefins like 1-octene and 4M1P.

\subsection{Pyridylamido hafnium system (\textbf{2})} 

The pyridylamido hafnium
catalyst system has been reported to be highly prone to undergoing fast
and reversible chain transfer to zinc and has been used in DOW
Chemicals' chain shuttling polymerization to produce 
OBCs~\cite{10,11},
Furthermore, Coates and coworkers reported the synthesis of
well-defined iPP/HDPE (isotactic polypropylene/high-density
polyethylene) di-, tri- and tetrablock copolymers using the
living-catalyst system \textbf{2}/B(C$_{6}$F$_{5}$)$_{3}$/DiBAP (DiBAP:
diisobutylaluminum 2,6-di-\textit{tert}-butyl-4-methylphenoxide) in
combination with a sequential feed of propylene and ethylene~\cite{8}.
Therefore, we hoped that this catalyst system would be a suitable
candidate for producing OBCs consisting of P4M1P hard segments and
poly(ethylene-\textit{co}-4M1P) soft segments under CCTP conditions.
Catalyst precursor \textbf{2} was selected for its high symmetry.
Keeping in mind that the first ${\upalpha}$-olefin insertion takes place
in the Hf--aryl bond, selecting \textbf{2} with a symmetrical CMe$_{2}$
bridge in the backbone of the ligand system most likely yields only one
major active species~\mbox{\cite{42,43,44,45}.}

Before studying the CCTP behavior of \textbf{2}, we were curious to
know if the catalyst, when activated with
B(C$_{6}$F$_{5}$)$_{3}$/DiBAP, would also be living for the sequential
feed copolymerization of ethylene and 4M1P. First, we reproduced
abovementioned iPP/HDPE di-, tri- and tetra-block copolymers reported
by Coates and coworkers as benchmarks (Table~S2). Both the gradual
increase in molecular weight and the molecular-weight-dependent average
branching density agree with a living catalyst system in combination
with a sequential feed of ethylene and propylene (Figure~S4).

Next, the copolymerization by sequential feeding of 4M1P and ethylene
was attempted using the same \textbf{2}/B(C$_{6}$F$_{5}$)$_{3}$/DiBAP
catalyst system as used for synthesizing the iPP/HDPE block copolymers
(Table~\ref{tab2}). As a benchmark, the random ethylene/4M1P
copolymerization was performed, which showed living-like behavior:
doubling the polymerization time resulted in a doubling of ${M}_{\mathrm{n}}$
and the polydispersity remained relatively low and below 2
(Figure~\ref{fig3}A).

\begin{figure*}
\vspace*{1pt}
\includegraphics{fig03}
\vspace*{2pt}
\caption{\label{fig3}Molecular weight distributions (MWD) and branching
densities per 1000 total nr.\ of carbons (SCB/1000TC) obtained by
high-temperature size exclusion chromatography of: (A)
poly(ethylene-\textit{co}-4M1P) random copolymer (Table~\ref{tab2},
entry 1), (B) poly(4M1P-\textit{block}-(ethylene-\textit{co}-4M1P))
diblock copolymer (Table~\ref{tab2}, entry 2) and (C)
poly(4M1P-\textit{block}-(ethylene-\textit{co}-4M1P)-\textit{block}-4M1P)
triblock copolymer (Table~\ref{tab2}, entry 3).}
\vspace*{1pt}
\end{figure*}

The situation is quite different when the polymerization is performed
using sequential feed polymerization of 4M1P and ethylene.
Figure~\ref{fig3}B shows the HT-SEC molecular weight distribution of the
first P4M1P block formed and the final product where ethylene was added
after 60~min \mbox{polymerization} time. The bimodal behavior of the final
product suggests that a significant part of the initially formed P4M1P
chains had terminated, and only a relatively small fraction of
poly(4M1P-\textit{block}-(ethylene-\textit{co}-4M1P)) diblock copolymer
had been produced. The formation of the diblock copolymer is supported
by the drop in the branching density of the diblock product as compared
to the branching density of the initial P4M1P precursor. For the 3-step
(4M1P ${-}$ (ethylene ${+}$ 4M1P) ${-}$ 4M1P) sequential feed experiment
(Figure~\ref{fig3}C), the same behavior was observed.

Comparing SEC data of samples taken after ({i})~40~min 4M1P
homopolymerization, ({ii}) after addition of ethylene and an additional
polymerization time of 20~min, and finally ({iii}) after 2~h total
polymerization time again shows that only a small fraction of the
initially formed P4M1P is transformed into the
poly(4M1P-\textit{block}-(ethylene-\textit{co}-4M1P)) diblock copolymer
after adding ethylene. Upon extending the polymerization for another 1
hour, the molecular weight of the high-molecular-weight fraction
further increased, and the branching density increased somewhat, which
suggests that after all the added ethylene has been consumed, 4M1P
continues to be incorporated in the polymer, \mbox{forming} the final
poly(4M1P-\textit{block}-(ethylene-\textit{co}-4M1P)-\textit{block}-4M1P) 
triblock copolymer. Although some triblock copolymer is clearly formed,
the bimodal molecular weight distribution obtained for the sequential
feed \mbox{polymerizations} and the fact that the molecular weight of
the low-molecular-weight fraction does not increase during the progress
of the polymerization reveals a non-living nature of the catalyst
system \textbf{2}/B(C$_{6}$F$_{5}$)$_{3}$/DiBAP during 4M1P
homopolymerization. Increasing the polymerization temperature from 25
to 60~\textdegree C led to deactivation of the catalyst system. These
results---in comparison with the living sequential ethylene/propylene
copolymerization---clearly show that the living nature of a catalyst is
strongly influenced by the nature of the monomers to be polymerized.
For the sterically hindered 4M1P, propagation is considerably slower
than for ethylene or propylene, which allows ${\upbeta}$-H transfer to
occur as a competing reaction, preventing the selective formation of
OBCs.

The low thermal stability of the
\textbf{2}/B(C$_{6}$F$_{5}$)$_{3}$/DiBAP catalyst system, its
non-living behavior for 4M1P polymerization, and its incompatibility
with DEZ as CTA~\cite{46} make B(C$_{6}$F$_{5}$)$_{3}$ an unsuitable
cocatalyst in combination with \textbf{2} for the polymerization of
4M1P under CCTP conditions. Therefore, B(C$_{6}$F$_{5}$)$_{3}$ was
replaced by borate cocatalysts. First, precatalyst \textbf{2} activated
with \textit{N},\textit{N}-dimethylanilinium
tetrakis(pentafluorophenyl)borate
([C$_{6}$H$_{5}$N(H)Me$_{2}$]$^{+}${\ubreak}[B(C$_{6}$F$_{5}$)$_{4}$]$^-$, 
AB) using DiBAP as scavenger was applied for the polymerization of 4M1P
(Table~\ref{tab3}, Table~S3, Figure~\ref{fig4}).

\begin{figure*}
\vspace*{-1pt}
\includegraphics{fig04}
\vspace*{-1pt}
\caption{\label{fig4}Molecular weight and polydispersity index (PDI) as
a function of conversion for 4M1P polymerization at 60~\textdegree C
using \textbf{2}/AB/DiBAP as catalyst system: (A) without DEZ and (B)
with DEZ (Table~\ref{tab3}).}
\end{figure*}

%tab3
\begin{table*}
\caption{\label{tab3}4M1P polymerization using \textbf{2}/AB/DiBAP with
different amounts of DEZ as CTA\vspace*{-1pt}}
\begin{tabular}{ccccccccc}
\thead
Exp. & \parbox[t]{1.1cm}{\centering Hf  (${\upmu}$mol)} &
\parbox[t]{1.1cm}{\centering DEZ  eq. Hf} & 
\parbox[t]{1cm}{\centering $t$  (min)} & 
\parbox[t]{1cm}{\centering Yield  (g)} &
\parbox[t]{1.6cm}{\centering Conversion$^{\mathrm{a}}$ (\%)} & 
\parbox[t]{1.2cm}{\centering ${{M}_{\mathrm{n}}}^{\mathrm{b}}$ (kDa)} & 
\parbox[t]{1.2cm}{\centering ${{M}_{\mathrm{w}}}^{\mathrm{b}}$ (kDa)} &
{\textit{\DJ}$_{\mathrm{M}}$}$^{\mathrm{b}}$\vspace*{2pt} \\
\endthead
1$^{\mathrm{c}}$ & 0.15 & - & 15 & 0.12 & \09 & 33.3 & 198.7 & \05.9 \\
2$^{\mathrm{c}}$ & 0.15 & - & 30 & 0.47 & 35 & 45.6 & 216.4 & \04.9 \\
3$^{\mathrm{c}}$ & 0.15 & - & 45 & 0.71 & 53 & 56.9 & 224.9 & \03.9 \\
4$^{\mathrm{c}}$ & 0.15 & - & 60 & 0.94 & 70 & -$^{\mathrm{f}}$ & -$^{\mathrm{f}}$ & -$^{\mathrm{f}}$ \\
5$^{\mathrm{d}}$ & 0.15 & 200 & 15 & 0.35 & 13 & 10.9 & \077.0 & \07.1 \\
6$^{\mathrm{d}}$ & 0.15 & 200 & 30 & 1.47 & 55 & 12.5 & 144.8 & 11.6 \\
7$^{\mathrm{d}}$ & 0.15 & 200 & 45 & 1.97 & 74 & 23.8 & 165.0 & \06.9 \\
8$^{\mathrm{d}}$ & 0.15 & 200 & 60 & 2.07 & 78 & \08.9 & 149.0 & 16.6 \\
9$^{\mathrm{d}}$ & 0.15 & 400 & 15 & 0.06 & \02 & \04.8 & \024.9 & \05.2 \\
10$^{\mathrm{d}}$\0 & 0.15 & 400 & 30 & 1.09 & 41 & \06.1 & 151.4 & 24.9 \\
11$^{\mathrm{d}}$\0 & 0.15 & 400 & 45 & 1.64 & 62 & 16.3 & 161.6 & \09.9 \\
12$^{\mathrm{d}}$\0 & 0.15 & 400 & 60 & 1.89 & 71 & 24.2 & 168.8 & \06.9 \\
13$^{\mathrm{d},\mathrm{e}}$ & \00.075 & 800 & 120\0 & 0.46 & 17 & \08.3 & \098.6 & 11.9 
\botline
\end{tabular}
\tabnote{Reaction conditions: 20~mL crimp-cap vial; 1.5 equiv. AB; 230
equiv. DiBAP; solvent: MCH/toluene ($3+1$~mL; ${V}_{\mathrm{reaction}}
\sim 8$~mL); stirring rate: 500 rpm; $T = 60$~\textdegree C;
quenching: MeOH/HCl.}
\tabnote{$^{\mathrm{a}}$Determined by $^{1}$H NMR.
$^{\mathrm{b}}$Determined by size-exclusion chromatography.
$^{\mathrm{c}}$2~mL of 4M1P.
$^{\mathrm{d}}$4~mL of 4M1P.
$^{\mathrm{e}}$460 equiv. DiBAP.
$^{\mathrm{f}}$Polymer not fully soluble in $o$-DCB.}
\vspace*{-1pt}
\end{table*}

While the activity was very low at 25~\textdegree C, the system is much
more active at 60~\textdegree C than \mbox{compared} to the borane-activated
system. However, raising the temperature further to 100~\textdegree C
resulted in a \mbox{significant} drop in the activity (Table~S3). The
molecular weight of the polymer also strongly depends on the
polymerization temperature and ${M}_{\mathrm{w}}$ values dropped by two orders
of magnitude when the temperature was raised from 25 to 100~\textdegree
C. Interestingly, the system shows an unusually long initiation period
of about 10~min (Figure~S5). Recently, a similar long initiation
time has been reported for the \mbox{1-octene} \mbox{polymerization} of the closely
related pyridylamido hafnium catalyst, $\mathbf{2}'$, containing a
CH(2-iPr-C$_{6}$H$_{4}$) bridge instead of a CMe$_{2}$ bridge between
the pyridine and the amido functionalities of the ancillary
ligand~\cite{3,47}. Already at low conversion, rather high molecular
weights are obtained (Figure~\ref{fig4}A), but the molecular weight
increases somewhat further with polymerization time. However, it does
not \mbox{increase} proportionally to the 4M1P conversion, indication that the
catalyst system is non-living. The~\mbox{polydispersity} is unusually broad
for a single-site catalyst system and is considerably broader than when
B(C$_{6}$F$_{5}$)$_{3}$ was used as cocatalyst (see above). 

The presence of DEZ has a slightly positive effect on the catalytic
activity with the highest activity for 200 equiv. of DEZ (Figure~S5).
Although the molecular weight values obtained are somewhat scattered
and there is little difference between adding 200 or 400 equiv. of DEZ
(Table~\ref{tab3}), the molecular weights do increase with conversion
(Figure~\ref{fig4}B). A rough estimation of the chain transfer rate
constant based on the Mayo equation demonstrates that chain transfer to
zinc is of the same \mbox{order} of magnitude as chain growth
(${k}_{\mathrm{CT}}/{k}_{\mathrm{CG}} = 6.3$, Figure~S6)~\cite{37}. 
However, the very
broad polydispersity of the products---even broader than observed in
the absence of DEZ---suggests that the chain transfer is
quasi-irreversible. Analyses of the products by $^{1}$H NMR indeed show
the existence of unsaturated vinylidene and vinylene end groups, which
indicates that the presence of DEZ cannot prevent concurrent chain
termination by ${\upbeta}$-H transfer. These observations agree with
results obtained by Landis and coworkers~\cite{28}, who found similar
fast and essentially irreversible chain transfer to zinc
$({k}_{\mathrm{CT}}/{k}_{\mathrm{CG}} = 7.27)$ 
accompanied by concurrent ${\upbeta}$-H
transfer for 1-octene polymerization using the closely related
pyridylamido hafnium catalyst $\mathbf{2}'$.

Next, 4M1P polymerization experiments with trityl borate
([C(C$_{6}$H$_{5}$)$_{3}$]$^{+}$[B(C$_{6}$F$_{5}$)$_{4}$]$^-$, TTB)
were carried out at different temperatures, with and without DEZ as CTA
(Table~S4). The \textbf{2}/TTB/DiBAP system resembles the
\textbf{2}/AB/DiBAP system: the activity increases with increasing
temperature from 25 to 60~\textdegree C and decreases once the
temperature is increased further to 100~\textdegree C. In the absence
of DEZ, there is no significant change in molecular weight and
polydispersity with increasing polymerization time as was observed in
the AB-activated system. Like for the AB-activated system, the
temperature has a dramatic effect on the molecular weight, which
clearly indicates that chain termination takes place (Figures~S7, S8).
The polydispersity of the polymer also drops with increasing
temperature, but it remains high for a single-site catalyst with a
value around 4.

Adding DEZ as CTA resulted in a clear reduction in molecular weight.
For polymerizations performed in the presence of DEZ at 25~\textdegree
C, a clear correlation between molecular weight and monomer conversion
could be observed (Figure~S9). Due to interference from the competitive
chain termination (see above), the molecular weight does not show a
clear trend as a function of polymerization time at 60~\textdegree C
(Figure~S10). The addition of DEZ resulted in a significant broadening
rather than a narrowing of the polydispersity, which hints at
irreversible chain transfer. 

\begin{figure*}
\vspace*{1pt}
\includegraphics{fig05}
\vspace*{1pt}
\caption{\label{fig5}(A) Conversion versus polymerization time and (B)
molecular weight as a function of conversion for 1-octene
polymerization at 100~\textdegree C, without and with 600 equiv. of DEZ
(Table~S7).}
\vspace*{1pt}
\end{figure*}


To investigate how steric hindrance at the active site affects the
chain transfer to zinc and to be able to directly compare the results
for \textbf{2} with the results of Landis et al. on $\mathbf{2}'$,
polymerization experiments using AB/DiBAP as activator system were
performed with the sterically less hindered 1-octene as the monomer
(Table~S5). The same unusual initiation period of around 10~min as
during the 4M1P polymerization was observed (Figure~S11A), in agreement
with Landis' results~\cite{3,47}. As expected, the final molecular
weight increases linearly with the initial 1-octene concentration both
with and without DEZ (Table~S6, Figure~S12). The difference in steric
hindrance between 1-octene and 4M1P is reflected in a higher molecular
weight of the poly(1-octene) compared to P4M1P obtained under the same
polymerization conditions (cf.\ Table~\ref{tab3}, Table~S3). Like for
4M1P polymerization, the presence of DEZ leads to an increase in
catalytic activity (Figure~S11A). Without DEZ, the molecular
weight remains constant during the whole polymerization period in
agreement with a non-living polymerization system. In the presence of
DEZ the molecular weight starts significantly lower and increases with
polymerization time, although it does not follow the 1-octene
conversion as would be expected for a pseudo-living system
(Figure~S11B). Chain transfer to zinc is of the same order of magnitude
as chain growth (${k}_{\mathrm{CT}}/{k}_{\mathrm{CG}} = 3.1$, 
Figure~S13). However, the fact that the polydispersity is three to four
times broader than without DEZ, in tandem with the number of chains per
zinc being greater than 2, suggest that it is quasi-irreversible, as
was observed for the 1-octene polymerization using
$\mathbf{2}'$~\cite{28}.\looseness=-1

Furthermore, $^{1}$H NMR clearly shows the presence of various
unsaturated end-group resonances indicating that competitive
${\upbeta}$-H and ${\upbeta}$-alkyl transfer takes place after both
1,2-insertion and 2,1-misinsertion (Figure~S14), which also contributes
to the broad polydispersity. These results are similar to those
reported by Landis and coworkers on 1-octene polymerization using the
closely related $\mathbf{2}'$, who also found fast but irreversible
chain transfer to zinc accompanied by ${\upbeta}$-H transfer~\cite{28}.

It can be concluded that the behavior of \textbf{2} activated with
borates is broadly comparable for \mbox{1-octene} and 4M1P
polymerization. The addition of DEZ clearly leads to a dramatic drop in
molecular weight and a broadening of the polydispersity. Whereas chain
transfer to zinc takes place, the high polydispersity suggests that it
is quasi-irreversible. Furthermore, ${\upbeta}$-H and ${\upbeta}$-alkyl
transfer is a seriously competitive reaction, which renders this
catalyst system unsuitable to produce multiblock copolymers consisting
of sterically hindered higher ${\upalpha}$-olefins such as 4M1P.

\subsection{Bis(phenolate)diether hafnium system (\textbf{3})}

Another contender in our quest for 4M1P-based OBCs is the hafnium
bis(phenolate)diether \mbox{complex} \textbf{3}, originally developed
by SYMYX and later \mbox{employed} by for example DOW Chemicals and
ExxonMobil for its excellent ability to incorporate higher
${\upalpha}$-olefins~\cite{33,34,35}. We decided to focus first on
1-octene polymerization and subsequently to apply the optimized
conditions to 4M1P polymerization. 

\begin{figure*}
\vspace*{-2pt}
\includegraphics{fig06}
\vspace*{-2pt}
\caption{\label{fig6}(A) 1-Octene conversion as a function of time
using \textbf{3}/1.5 TTB/100 TMA/600 DEZ at 100~\textdegree C. Addition
of ethylene after 120~min. (B) Molecular weight as a function of yield,
without and with addition of ethylene after 120~min (Table~S8).}
\vspace*{-2pt}
\end{figure*}

The 1-octene polymerizations were performed using TTB as cocatalyst and
trimethylaluminum (TMA) as scavenger~\cite{35}. It is worth mentioning
that the catalyst is thermally considerably more stable than 
\textbf{1}
and \textbf{2} as it shows a surprisingly long lifetime of several
hours at 100~\textdegree C. This is important in view of commercial
solution polymerization processes, which are run at elevated
temperatures. Under the polymerization conditions applied, the catalyst
system follows first order kinetics in monomer, gives reasonably high
1-octene conversions, and produces polymers with moderate molecular
weights and polydispersities around 2 as expected for a single-site
catalyst (Table~S7, Figure~\ref{fig5}A). In the absence of DEZ, the
molecular weight slightly drops with conversion, in line with the
decreasing monomer concentration (Figure~\ref{fig5}A). The addition of
DEZ leads to a slight increase in catalytic activity, clearly
indicating that there is no negative effect of catalyst--CTA adduct
formation (Figure~\ref{fig5}B)~\cite{38,39,40,41,42,43,44,45}.
Furthermore, the molecular weight is significantly lower and increases
linearly albeit not proportionally with 1-octene conversion. The
polydispersity around 2 is in line with moderately fast and reversible
chain transfer and agrees with the observed
${k}_{\mathrm{CT}}/{k}_{\mathrm{CG}} =13.5$ 
(Figure~S15). This is further supported by the number of polymer
chains per zinc, which gradually increases with increasing 1-octene
conversion, reaching a maximum value around 2 (\mbox{Table~S7}). This shows
that \textbf{3} is more tolerant to the steric hinderance of the
Zn(Et)-polymeryl species during chain transfer than \textbf{1} and
\textbf{2}.


To verify the existence of concurrent processes such as ${\upbeta}$-H
transfer, $^{1}$H NMR spectroscopy was used (Figure~S16). The $^{1}$H
NMR spectrum revealed the presence of several low-intensity signals in
the olefinic region (4.5--6.0 ppm) attributed to unsaturated end
groups. The major resonances were assigned to vinylene (${\sim}80\%$,
${\updelta}$ 5.37 ppm) and vinylidene (${\sim}15\%$, ${\updelta}$ 4.67 and
4.73~ppm) species. The other peaks of even lower intensity were
assigned to trisubstituted unsaturated and vinyl end groups. Thus,
contrarily to complexes \textbf{1} and \textbf{2}, complex \textbf{3}
hardly undergoes ${\upbeta}$-H elimination after 1,2-insertion, and the
main termination reaction proceeds via ${\upbeta}$-H transfer after
2,1-misinsertion~\cite{30}. Obviously, concurrent chain termination
processes are detrimental for the attempted formation of OBCs via CCTP.
However, as the prevalent concurrent chain transfer process \mbox{consists} in
chain termination after 2,1-misinsertion and its occurrence is low,
there might be a way to overcome this. 

A possible approach to reactivate the dormant site formed after
2,1-misinsertion without chain scission is the addition of sterically
unhindered ethylene that might be able to insert after a
2,1-misinsertion~\cite{51,52}. Figure~\ref{fig6} shows the result of
bubbling ethylene through the polymerization product mixture after the
catalytic activity had ceased (Experiment 8, Table~S8). A significant
increase in the 1-octene conversion, polymer yield, and in molecular
weight (${\Delta}{M}_{\mathrm{n}} = 30\%$, ${\Delta}M_{\mathrm{w}} = 66\%$) could
be observed, which clearly indicates that addition of \mbox{ethylene} 
\mbox{indeed}
reactivates the catalyst. Experiments where 1-octene was polymerized in
the presence of a small amount of ethylene added at the beginning of
the polymerization (Table~S9, Figure~S17) showed an initial
increase in activity, which subsequently leveled off to the same rate
as observed in the absence of ethylene. Most likely, the ethylene is
consumed very rapidly after which the 1-octene homopolymerization takes
place. This is supported by the nearly identical molecular weight build
up with 1-octene conversion for both systems, with and without
ethylene. Although the presence of ethylene truly seems to reactivate
the catalyst after 2,1-misinsertion, addition of small amounts of
ethylene only at the beginning of a batch polymerization does not have
the desired effect. A continuous flow of a minute amount of ethylene
will be needed to ({i}) continuously reactivate the catalyst and
({ii}) avoid incorporation of too much ethylene. Unfortunately,
this was not possible in the currently applied reactor setup.

Using the optimized conditions for 1-octene polymerization, 4M1P
polymerization experiments have been performed (Table~S10). The
conversion \mbox{versus} time plots for 4M1P and 1-octene polymerization
(Figure~\ref{fig7}A) reflect the difference in reactivity of the two
monomers: where the conversion of 1-octene in the presence of 600
equiv. of DEZ is 89\% after a polymerization time of 120~min, it is
only 39\% for 4M1P, and the catalyst lost most of its activity after
30~min at 100~\textdegree C. Possibly, for the more sterically
hindered 4M1P, the 2,1-misinsertion deactivates the catalyst to a large
extent leading to a low overall conversion. Additionally, it should be
noted that commercial-grade 4M1P, contrary to commercial-grade
1-octene, contains isomers with internal double bond, like
\mbox{4-methyl-2-pentene}, that could lead to catalyst poisoning and
low-molecular-weight products.

\begin{figure*}
\vspace*{1pt}
\includegraphics{fig07}
\vspace*{1pt}
\caption{\label{fig7}(A) Comparative conversion versus polymerization
time plot for the 1-octene (Table~S7) and 4M1P (Table~S10)
polymerizations and (B) molecular weight as a function of conversion
for 4M1P polymerization with 100 and 600 equiv. of DEZ versus catalyst
(Table~S10), using the \textbf{3}/TTB/TMA/DEZ system at 100~\textdegree
C.}
\end{figure*}


The molecular weight clearly drops with increasing DEZ concentration,
and the molecular weight increases linearly with increasing conversion
(Figure~\ref{fig7}B), clearly indicating chain transfer to zinc. The
most prominent concurrent chain transfer process is ${\upbeta}$-H
transfer after 2,1-misinsertion also for 4M1P polymerization as
revealed by $^{1}$H NMR. 

\begin{figure*}
\vspace*{1pt}
\includegraphics{fig08}
\vspace*{1pt}
\caption{\label{fig8}(A) 4M1P conversion versus polymerization time for
the 4M1P(/ethylene) (Table~S11) and (B) molecular weight as a function
of polymerization time for 4M1P(/ethylene), using the
\textbf{3}/MAO/DEZ system at 100~\textdegree C.}
\vspace*{1pt}
\end{figure*}

As experience has taught us that MAO effectively activates precatalyst
\textbf{3} providing a long-living active species during propylene and
1-octene polymerization, we performed 4M1P polymerizations in the
presence of DEZ and MAO as cocatalyst. The 4M1P conversion turned out
to be somewhat higher than with the TTB-activated system (52\% versus
39\%; Figure~\ref{fig8}), but a comparable deactivation due to the
{formation} of dormant sites after 2,1-misinsertion is expected.

Indeed, addition of ethylene after 1 hour reaction time clearly
demonstrated an increase in 4M1P conversion (${>}$20\%) and a
considerably higher (${>}$30\%) polymer molecular weight (Table~S11,
Figure~\ref{fig8},~\ref{fig9}). The size-exclusion-chromatography
traces (Figure~\ref{fig9}) show that after the addition of ethylene the
initial monomodal distribution shifts to higher molecular weight but
also transforms into a more bimodal distribution having a high
molecular weight tail. This suggests that to some extent ${\upbeta}$-H
transfer does occur, which is supported by the fact that the number of
polymer chains per zinc is higher than 2. Nevertheless, the high
branching density measured for the P4M1P homopolymer, followed by a
clear drop in branching density a few minutes after ethylene addition,
and finally an increase in the branching density 2~h after ethylene
addition (Figure~\ref{fig9}) supports the formation of a triblock
copolymer. The thus formed triblock copolymer, consisting of
({i}) a P4M1P first block, ({ii}) a random/gradient
copolymer poly(ethylene-\textit{co}-4M1P) block followed by
({iii}) a second P4M1P block formed after all ethylene has been
consumed, would be expected to have a melting temperature similar to
that of a P4M1P homopolymer but a lower crystallinity. This is indeed
observed (Table~S11). 

\begin{figure}
\includegraphics{fig09}
\caption{\label{fig9}Molecular weight distributions (MWD) and
short-chain branching (SCB) obtained by high-temperature size-exclusion
chromatography of (i) P4M1P after 60~min (Table~S11, entry 5),
(ii) P4M1P after 240~min (Table~S11, entry 7) and (iii)
poly(4M1P-\textit{block}-(ethylene-\textit{co}-4M1P)-\textit{block}-4M1P) 
triblock copolymer (Table~S11, entry 8) applying the \textbf{3}/MAO/DEZ
system.}
\end{figure}

Unfortunately, the molecular weight of the P4M1P homopolymers and the
corresponding putative block copolymer are clearly too low to afford
materials with good mechanical properties. Most likely,
2,1-misinsertion is a rather common process for this catalyst resulting
in short P4M1P blocks before the active site is deactivated. Thus, the
production of higher molecular weight material was attempted by
reducing [DEZ]:[complex \textbf{3}] molar ratio down to 50, reducing
the polymerization time before ethylene addition to only 10~min,
and limiting the total reaction time to 40~min. Nevertheless,
unfortunately only materials with a maximum ${M}_{\mathrm{n}}$ around 4.7~kDa
(\textit{\DJ}$_{\mathrm{M}} = 2.3$) were achieved.

\section{Conclusions}

The three catalysts that were tested in the polymerization of 1-octene
and 4M1P showed both similarities and clear differences. All three
catalysts clearly show chain transfer to zinc and the
${k}_{\mathrm{CT}}$ and ${k}_{\mathrm{CG}}$ rate constants are in the
same order of magnitude for catalyst \textbf{1} and \textbf{2}, whereas
for catalyst \textbf{3} the ${k}_{\mathrm{CT}}$ rate constant is around
one order of magnitude higher than the ${k}_{\mathrm{CG}}$ rate
constant. Furthermore, whereas catalyst \textbf{1} demonstrates
reversible chain transfer for the polymerization of 1-octene but
irreversible for the polymerization of 4M1P, catalyst \textbf{2} shows
irreversible chain transfer for both 1-octene and 4M1P polymerization.
Catalyst \textbf{3} undergoes reversible chain \mbox{transfer} to zinc for
both 1-octene and 4M1P polymerization but suffers severely from
2,1-misinsertion leading to dormant sites. All catalysts suffer from
concurrent ${\upbeta}$-H transfer processes, which can be 
\mbox{attributed} to the steric hindrance of the applied
\mbox{${\upalpha}$-olefins.} This steric hindrance lowers the chain growth
rate constant and hence increases the probability of chain scission.
Interestingly, for catalyst \textbf{3}, ${\upbeta}$-H transfer after
2,1-misinsertion appeared to be the prevalent competing chain transfer
process. Adding a small amount of ethylene clearly reactivated the
dormant site formed after 2,1-misinsertion of \textbf{3} without
terminating the growing polymer chain, which opens a route toward the
desired olefinic block copolymers. Nevertheless, under the batch
conditions used, concurrent ${\upbeta}$-H transfer could not be fully
prevented. 

\section*{Declaration of interests}

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.

\section*{Funding}
The work was financially supported by SABIC, which is gratefully
acknowledged. 

\section*{Underlying data}

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

Supporting information (experimental details, additional tables and
graphs and NMR spectra of selected samples) for this article is
available at \url{https://hal.science/hal-05264753}.

\CDRsupplementaryTwotypes{supplementary-material}{\cdrattach{crchim-415-suppl.pdf}}

\back{}

\printbibliography
\refinput{crchim20250567-reference.tex}

\end{document}