Kinetic Treatments for Catalyst Activation and Deactivation Processes based on Variable Time Normalization Analysis

Article abstract of DOI:10.1002/anie.201903878

Progress reaction profiles are affected by both catalyst activation and deactivation processes occurring alongside the main reaction. These processes complicate the kinetic analysis of reactions, often directing researchers toward incorrect conclusions. We report the application of two kinetic treatments, based on variable time normalization analysis, to reactions involving catalyst activation and deactivation processes. The first kinetic treatment allows the removal of induction periods or the effect of rate perturbations associated with catalyst deactivation from kinetic profiles when the quantity of active catalyst can be measured. The second treatment allows the estimation of the activation or deactivation profile of the catalyst when the order of the reactants for the main reaction is known. Both treatments facilitate kinetic analysis of reactions suffering catalyst activation or deactivation processes.

Full text of DOI:10.1002/anie.201903878

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Kinetic Treatments for Catalyst Activation and Deactivation
Processes based on Variable Time Normalization Analysis
Authors: Alicia Martínez-Carrión, Michael G. Howlett, Carla Alamillo-
Ferrer, Adam D. Clayton, Richard A. Bourne, Anna Codina,
Anton Vidal-Ferran, Ralph W. Adams, and Jordi Bures
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.201903878
Angew. Chem. 10.1002/ange.201903878
Link to VoR: http://dx.doi.org/10.1002/anie.201903878
http://dx.doi.org/10.1002/ange.201903878

10.1002/anie.201903878
Angewandte Chemie International Edition
COMMUNICATION
Kinetic Treatments for Catalyst Activation and Deactivation
Processes based on Variable Time Normalization Analysis
Alicia Martínez-Carrión,a,b Michael G. Howlett,a Carla Alamillo-Ferrer,a Adam D. Clayton,c Richard A.
Bourne,c Anna Codina,d Anton Vidal-Ferran,*b,e Ralph W. Adams,*a and Jordi Burés*a
Abstract: Progress reaction profiles are affected by both catalyst
activation and deactivation processes occurring alongside the main
reaction. These processes complicate the kinetic analysis of
reactions, often directing researchers towards incorrect conclusions.
We report the application of two kinetic treatments based on the
variable time normalization analysis (VTNA) to reactions involving
catalyst activation and deactivation processes. The first kinetic
treatment allows the removal of induction periods or the effect of rate
perturbations associated with catalyst deactivation from kinetic
profiles when the quantity of active catalyst can be measured. The
second treatment allows the estimation of the activation or deacti-
vation profile of the catalyst when the order of the reactants for the
main reaction is known. Both treatments facilitate kinetic analysis of
reactions suffering catalyst activation or deactivation processes.
Processes of catalyst activation and deactivation occur at
the same time as the main reaction.[1] As a consequence, the
concentration of active catalyst[2] varies throughout the course of
the reaction, affecting the reaction’s intrinsic kinetic profile. This
perturbation of the kinetic profile adds a layer of complexity to its
analysis. Often, this complication limits the quantitative kinetic
analysis to those sections of the reaction with no significant
variation of catalyst concentration. Herein, we describe two
treatments which facilitate a quantitative analysis of reactions
involving catalyst activation or deactivation processes. One or
other of these treatments can be used depending on the
available information. We show the feasibility of both methods in
the kinetic analysis of real reactions with severe changes of
active catalyst concentration during the course of the reaction: a
Figure 1. Variable time normalization analysis (VTNA) is a useful tool for the
kinetic analysis of reactions with catalyst activation or deactivation processes.
The two kinetic treatments herein presented are based on
the recently described variable time normalization analysis
(VTNA).[5] This analysis allows the removal of the kinetic effect of
any component of a reaction from the temporal concentration
profiles. Therefore, if both the concentration of active catalyst
and the progress of the reaction can be measured
simultaneously by any appropriate method, then the intrinsic
profile of the main reaction can be obtained (Figure 1a). The
resulting reaction profile of this first kinetic treatment is much
simpler to analyze than the original profile and facilitates the
extraction of mechanistic information, such as the orders of
reaction or intrinsic turnover frequency (TOF) of the catalyst.
The second kinetic treatment herein presented uses the reaction
progress profile and the orders of reaction to extract the catalyst
activation or deactivation profile (Figure 1b). This profile is in-
formative of the pathways of catalyst activation and deactivation,
and their kinetics. This knowledge can help to rationally modify
reaction conditions in order to maximize the turnover number
(TON) of the reaction.
hydroformylation reaction catalyzed by
a
supramolecular
rhodium complex[3] and an aminocatalytic Michael reaction.[4]
[a]
Dr. J. Burés, Dr. R. W. Adams, Dr. C. Alamillo-Ferrer, M. G. Howlett,
A. Martínez-Carrión
The University of Manchester, School of Chemistry
Oxford Road, Manchester, M13 9PL, UK
E-mail: jordi.bures@manchester.ac.uk,
ralph.adams@manchester.ac.uk
[b]
[c]
Prof. A. Vidal-Ferran, A. Martínez-Carrión
Institute of Chemical Research of Catalonia (ICIQ) &
The Barcelona Institute of Science and Technology (BIST)
Av. Països Catalans 16, 43007 Tarragona, Spain
E-mail: avidal@iciq.cat
Dr. R. A. Bourne, A. D. Clayton
University of Leeds
Institute of Process Research and Development
School of Chemistry & School of Chemical and Process Engineering
Leeds, LS2 9JT, UK
The first data treatment allows the intrinsic reaction profile of
a reaction altered by catalyst activation or deactivation process-
es to be uncovered. To demonstrate its potential, we chose two
reactions with large changes in catalyst concentration. The first
[d]
[e]
Dr. A. Codina
Bruker UK Ltd., Banner Lane, Coventry, CV4 9GH, UK
ICREA, Passeig Lluïs Companys, 23, 08010 Barcelona, Spain
is an asymmetric hydroformylation catalyzed by
a
su-
pramolecular rhodium complex, which requires three different
units to come together in order to form an active catalyst:
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/anie.201903878
Angewandte Chemie International Edition
COMMUNICATION
rhodium as active center, an enantiopure bisphosphite as ligand,
and a rubidium salt to regulate the geometry of the catalyst
(Figure 2).[3] As the catalyst formation process is not immediate,
the active catalyst concentration increases over the course of
the reaction, and the overall reaction profile shows a clear
induction period in the formation of the product (Figure 2a).
profile with no trace of any induction period, revealing the real
first-order profile of the original reaction. These observations
imply that, under the chosen reaction conditions, the olefin-
hydride insertion is the rate-determining step.[3], [9]
The second reaction we investigated is the enantioselective
aminocatalytic Michael addition of aldehyde to trans-β-
nitrostyrene (Figure 3). This reaction is generally performed with
10-20 mol% catalyst loading[4] due to the deactivation processes
that the catalyst can suffer. Indeed, when the reaction was run at
high substrate concentration and as little as 0.5 mol% catalyst,
most of the catalyst was deactivated before the reaction reached
completion (Figure 3a). As a consequence, the reaction profile is
curved with an apparent overall order close to one.[8] Due to the
overlap of signals in the NMR spectra of several deactivated
catalytic species, it was impossible to measure the amount of
active catalyst during the last part of the reaction (Figure 3b).
Nevertheless, when the measured amount of active catalyst was
used to normalize the time scale of the original reaction, the
kinetic profile became an almost perfect straight line (Figure 3c),
indicative of an overall zero order reaction. This overall reaction
order is in agreement with previous mechanistic studies
performed at higher catalyst loadings.[10] In addition, the slope of
the resulting straight line is approximately the TOF of the
reaction, in this case 1.86 min–1.
As the hydroformylation reaction is performed in
a
pressurized vessel with a constant supply of syngas, continuous
monitoring of the evolution of the reaction components was
challenging.[6] To overcome this challenge, we used a Bruker
InsightMR flow tube.[7] This device continuously recirculates a
small volume of the liquid reaction mixture through the reaction
vessel and
a modified NMR tube, enabling the on-line
monitoring of the reaction by NMR under these challenging
reaction conditions. In this case, we were able to simultaneously
monitor both the concentration of product and the amount of
rhodium hydride of the assembled supramolecular complex,
which is the resting state of the catalyst ([RhH] in Figure 2b).[8]
The catalyst profile was used to normalize the time scale of the
original progress reaction profile using VTNA.[8] The resulting
reaction profile (Figure 2c) is much simpler than the original
The second kinetic treatment for reactions with activation or
deactivation processes serves to estimate the concentration
profile of active catalyst when it cannot be measured during the
reaction. The amount of active catalyst is estimated by
Figure 3. The apparent first order of the original reaction profile of an
aminocatalyzed Michael addition (a) is corrected by the profile of active
catalyst (b) to reveal a zero-order reaction profile (c).
Figure 2. The original reaction profile of
catalysed by a supramolecular rhodium complex (a) is corrected by the profile
of active catalyst (b) to reveal a simple first-order reaction profile (c).
a hydroformylation reaction
This article is protected by copyright. All rights reserved.

10.1002/anie.201903878
Angewandte Chemie International Edition
COMMUNICATION
deconvolving its effect on the shape of the reaction profile using
VTNA. The VTNA subtracts the kinetic effect of a component
from the reaction profile when its concentration to the power of
the correct order of the reaction is used to normalize the time
scale.[5] When the time scale is normalized by all the kinetically
relevant components of the reaction, whose concentrations
change during the reaction, the reaction profile becomes a
straight line. Therefore, the amount of catalyst can be estimated
by maximizing the linearity of the resulting VTNA profile. This
process can be easily performed by many algorithms accessible
to all chemists. For generality, we have used the universally
available Microsoft Excel add-in Solver.[11]
To check the viability of our method, we chose the same re-
actions discussed above as it was possible to compare the
estimated profiles of active catalysts with the measured ones. In
the case of the hydroformylation reaction, there is a clear induc-
tion period in the original reaction profile (Figure 4a). This
phenomenon is indicative of the amount of active catalyst
building up, so the only constraint imposed on Solver was that
the amount of active catalyst could not decrease with time.[8] The
search started from values of 0% of active catalyst at all time
points to avoid any bias in the solution. The solution
automatically found by Solver provided a straight line with R2 =
0.99995 for the progress reaction profile when the time was
normalized against the concentration of starting material, 1, and
of variable active catalyst (Figure 4b). As the only evaluation
parameter is the straightness of the VTNA plot, the profile of
estimated active catalyst has the correct shape, but not
necessarily the correct magnitude. Therefore, the obtained
profile is better presented as percentage of active catalyst
(Figure 4c). The estimated profile for the activation of the
catalyst (red curve in Figure 4c) is in a reasonable agreement
with the measured profile (blue curve in Figure 4c). The small
discrepancy after the induction period could be due to the fact
that we only measured the hydride of the catalyst, but the total
amount of catalyst could be distributed among other species of
the catalytic cycle. Indeed, the estimated profile seems more
likely, as the very slow formation of the active catalyst in the last
section is not consistent with the profile for a standard kinetic
behavior.
The aminocatalytic Michael addition does not reach comple-
tion due to catalyst deactivation during the reaction (Figure 5a).
Both same-excess experiments[12] and a second addition of
fresh catalyst proved that the incomplete reaction was due to
catalyst deactivation.[8] The profile of deactivation was estimated
Figure 5. The original reaction progress profile (a) becomes an almost
perfect straight line (b) when applying the VTNA using the estimated catalyst
profile. The estimated profile of catalyst deactivation (red points in c) is in
agreement with the measured one (blue points in c).
Figure 4. The original reaction progress profile (a) becomes an almost
perfect straight line (b) when applying the VTNA using the estimated catalyst
profile. The estimated profile of catalyst activation (red points in c) overlays
perfectly with the measured one (blue points in c).
This article is protected by copyright. All rights reserved.

10.1002/anie.201903878
Angewandte Chemie International Edition
COMMUNICATION
using Solver starting from 100% of catalyst at all time points and
with the only restriction that the amount of catalyst could not
increase with time. The solution found by Solver converts the
original reaction profile (Figure 5a) into a straight line with an
R2=0.999995 (Figure 5b) when the time scale is normalized by
the amount of estimated active catalyst. The profile of estimated
active catalyst against time (red curve in Figure 5c) is in good
agreement with the measured profile (blue curve in Figure 5c),
showing that the approach also works well for estimation of
deactivation profiles. In addition, the estimated profile is able to
provide information for the last part of the reaction, when it was
impossible to confidently measure the amount of active catalyst
experimentally.[8]
In conclusion, VTNA is a powerful tool for the treatment of
reaction profiles altered by activation and deactivation processes.
It serves to simplify the profiles of reactions with variable catalyst
concentrations by normalization using the instantaneous catalyst
concentration measured throughout the reaction. Also, it
provides
a satisfactory estimation of the percentages of
activated or deactivated catalyst during the reaction by looking
for the best linearization of the variable time normalized reaction
profile. Both treatments render good results when applied to
experimental data, as proven by their application to different
reactions. These treatments are useful to facilitate the kinetic
analysis of catalytic reactions and obtain accurate and useful
mechanistic information.
There are some caveats to have in mind when using VTNA
to estimate the temporal profile of active catalyst. The first is that
the values of the resulting catalyst profile are relative because
the evaluation of the best profiles is based only on the R2 value
of the resulting VTNA plot. This means that there is an infinite
number of profiles with the same shape but different magnitudes
that would yield a straight line with an identical R2 value. There-
fore, the solution found by Solver should be treated as percent-
age of the maximum amount of active catalyst at each time and
if a concentration profile is desired, the concentration of active
catalyst at one time point must be known. Another caveat is that
if the order of the substances whose concentration is changing
during the reaction is not accurate, the VTNA plot and therefore
the estimated profile of active catalyst will be affected. To
minimize this effect, the order of kinetically relevant reactants
and catalyst should be known before applying the method.
Acknowledgements
The research leading to these results has received funding from
EPSRC (EP/S005315/1, EP/P007589/1, EP/K039547/1 and
EP/N509681/1), RSC (Mobility Grant for A. M.-C.), MINECO
(CTQ2014-60256-P, CTQ2017-89814-P), ICIQ Foundation
(predoctoral fellowship to A. M.-C.). The authors thank Bruker
(UK) for the generous loan of an InsightMR flow tube, and Dr. M.
Giménez (ICIQ) for advice in the experimental setup.
Keywords: kinetics • catalyst activation • catalyst deactivation •
reaction mechanisms • concentration reaction profiles
[1]
[2]
R. H. Crabtree, Chem. Rev. 2015, 115, 127.
Active catalyst refers only to on-cycle catalytic species. Therefore the
presented methods are applicable to catalytic systems regardless of the
reversible or irreversible nature of the off-cycle processes.
A. Vidal-Ferran, I. Mon, A. Bauzà, A. Frontera, L. Rovira, Chem.- Eur. J.
2015, 21, 11417.
[3]
[4]
[5]
Y. Hayashi, H. Gotoh, T. Hayashi, M. Shoji, Angew. Chem. Int. Ed.
2005, 44, 4212; Angew. Chem. 2005, 117, 4284.
a) J. Burés, Angew. Chem. Int. Ed. Engl. 2016, 55, 16084; Angew.
Chem. 2016, 128, 16318; b) J. Burés, Angew. Chem. Int. Ed. Engl.
2016, 55, 2028; Angew. Chem. 2016, 128, 2068; c) J. Burés, D.-T. C.
Nielsen, Chem. Sci. 2019, 10, 348.
[6]
[7]
a) A. C. Brezny, C. R. Landis, J. Am. Chem. Soc. 2017, 139, 2778; b) N.
J. Beacha, S. M. M. Knappa, C. R. Landis, Rev. Sci. Instrum. 2015, 86,
104101.
a) D. A. Foley, E. Bez, A. Codina, K. L. Colson, M. Fey, R. Krull, D.
Piroli, M. T. Zell, B. L. Marquez, Anal. Chem. 2014, 24, 12008; b)
www.bruker.com/InsightMR, accessed 27 Oct 2018.
[8]
[9]
See Supporting Information.
P. W. N. M. van Leeuwen, C. Claver, Eds. Rhodium Catalyzed
Hydroformylation, Kluwer Academic Publishers, Dordrecht, 2000, 70-72.
[10] a) J. Burés, A. Armstrong, A. D. G. Blackmond, J. Am. Chem. Soc.
2011, 133, 8822; b) K. Patora-Komisarska, M. Benohoud, H. Ishikawa,
D. Seebach, Y. Hayashi, Helv. Chim. Acta 2011, 94, 719; c) J. Burés, A.
Armstrong, A. D. G. Blackmond, J. Am. Chem. Soc. 2012, 134, 6741;
d) J. Burés, A. Armstrong, A. D. G. Blackmond, J. Am. Chem. Soc.
2012, 134, 14264; e) D. Seebach, X. Sun, M.-O. Ebert, W. B.
Schweizer, N. Purkayastha, A. K. Beck, J. Duschmalé, H. Wennemers,
T. Mukaiyama, M. Benohoud, Y. Hayashi, M. Reiher, Helv. Chim. Acta
2013, 96, 799; f) J. Burés, A. Armstrong, A. D. G. Blackmond, Acc.
Chem. Res. 2016, 49, 214.
Figure 6. The main pathways of catalyst deactivation are the reactions of the
transient zwitterionic iminium nitronate 7 with either propanal (3) or trans-β-
nitrostyrene (4) and the reaction of the catalyst 5 with the side product 10.
[11] S. Walsh, D. Diamond, Talanta 1995, 42, 561.
[12] D. G. Blackmond, Angew. Chem. Int. Ed. 2005, 44, 4302; Angew.
Chem. 2005, 117, 4374.
This article is protected by copyright. All rights reserved.