Organic &

Biomolecular Chemistry
View Article Online
PAPER View Journal | View Issue

Probing the mechanism of thermally driven thiol-

Published on 15 March 2018. Downloaded by University of New England on 20/04/2018 20:46:40.

Cite this: Org. Biomol. Chem., 2018,

Michael dynamic covalent chemistry†
16, 2725
Borui Zhang,‡ Progyateg Chakma,‡ Max P. Shulman, Jun Ke, Zachary A. Digby and
Dominik Konkolewicz *

Received 14th February 2018,
Accepted 15th March 2018
DOI: 10.1039/c8ob00397a
rsc.li/obc
The kinetics and mechanism of the thermally activated dynamic covalent exchange of thiol-Michael
adducts is investigated. A model system of thiol-Michael adducts between thiophenol and phenylvinyl-
ketone derivatives and adducts between 2-mercaptoethanol phenylvinylketone derivatives in N,N-di-
methylformamide (DMF) at elevated temperatures is used to probe the underlying exchange mechanism.
The kinetic data show negligible free Michael acceptor, which is consistent with the highly efficient thiol-
Michael reaction being a “click”-like reaction that significantly favors the adduct form. At elevated tem-
peratures of 90 °C in DMF the thiol-Michael adducts reach equilibrium after 24 h, although equilibration
did not occur within 24 h at 60 °C or 75 °C, and negligible exchange occurs under ambient conditions. A
kinetic model was developed to describe the dynamic covalent exchange and equilibration. The experi-
mental and simulation kinetic data of dynamic covalent exchange are consistent with the thiol-Michael
adducts undergoing a retro-Michael reaction, followed by subsequent addition of a free thiol to the liber-
ated Michael acceptor. Kinetic analysis is consistent with the fragmentation, or retro-Michael reaction,
being the rate-determining step in the dynamic covalent exchange. This suggests that the key step in
dynamic covalent exchange is not enhanced by addition of free thiol or free Michael acceptor, since the
addition reaction is much faster than the retro-Michael reaction. This fundamental study will guide the
design of organic compounds, materials, and bioconjugates that utilize the thermally activated dynamic
covalent thiol-Michael linkages.

Introduction moieties capable of dynamic covalent exchange can lead to

Dynamic covalent chemistry is an important field of organic
chemistry with significant applications from equilibrium
based synthetic reactions and complex molecular libraries to
materials science.1–9 Unlike kinetically controlled reactions,
where transformations with the lowest activation energy
barrier dominate the process, a dynamic covalent process is
governed by thermodynamics.1 A dynamic covalent reaction
typically involves two or more species that are in chemical
equilibrium with each other, and able to participate in forward
and reverse direction transformations to maintain this
dynamic equilibrium.1 There are many species that participate
in dynamic covalent processes, including Diels–Alder adducts,
boronic esters and boronates, imines, disulfides, olefins and
alkynes in the presence of a metathesis catalyst, and thiol-
Michael adducts, to name a few.1,10–19 Incorporation of these
Department of Chemistry and Biochemistry, Miami University, 651 E High St,
Oxford, OH, 45056, USA. E-mail: d.konkolewicz@miamiOH.edu
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
c8ob00397a
‡ Authors contributed equally.
well-defined organic molecules, self-directed assembly,
complex dynamically associated macrocycles, and polymers
with interesting properties such as self-healing or malleable
properties.4,5,20–22 However, to optimize the organic com-
pounds made by dynamic covalent approaches, or the per-
formance of organic materials containing dynamic covalent
groups, it is necessary to understand the mechanism of the
underlying dynamic covalent exchange.
The thiol-Michael reaction is a highly efficient reaction
between a thiol nucleophile and an electron poor alkene.23–26
Typically, the reaction is catalyzed with a small amount of base
or nucleophile, although the reaction can proceed in solvent
as well.27 The thiol-Michael reaction has found applications
from organic synthesis, materials chemistry and bioconjuga-
tion due to its facile implementation and high yield of the
desired organic products.7 A typical thiol-Michael reaction is
shown in Scheme 1A. The high yield, tolerance to functionality
and simple implementation have made the thiol-Michael reac-
tion a “click”-like reaction,23,28 and interestingly this reaction
also satisfies the dynamic covalent criteria under appropriate
stimulus.17,29–34 Typically, elevated pH or elevated tempera-
tures are used as stimuli for the activation of thiol-Michael

This journal is © The Royal Society of Chemistry 2018 Org. Biomol. Chem., 2018, 16, 2725–2734 | 2725

Paper
Scheme 1 (A) Thiol-Michael adduct formation reaction, (B) dynamic
Published on 15 March 2018. Downloaded by University of New England on 20/04/2018 20:46:40.
covalent exchange of thiol-Michael adducts.
adducts towards dynamic covalent exchange. This is an inter-
esting and important feature of the thiol-Michael adduct,
since it is essentially static under ambient conditions, and
typically only undergoes dynamic covalent exchange under
appropriate stimulus. This has important implications for
applications ranging from bioconjugation to material stabi-
lity.29,35,36 This dynamic covalent exchange of two thiol-
Michael adducts is given in Scheme 1B.
The dynamic nature of the thiol-Michael adduct has been
known for over 50 years,30,31 although it has only recently
gained significant attention.23,34 The pH or base driven thiol-
Michael adduct exchange has been studied as a dynamic
covalent process,17,37,38 in the context of bioconjugation pro-
cesses35,38,39 and polymer materials,29,32,40 or in a theoretical
context.34 In general the thermally driven dynamic thiol-
Michael exchange has received significantly less attention,
except for the early studies,30,31 and recent work from the past
two years.29,32,41,42 The thermally driven thiol-Michael
dynamic covalent chemistry has been used in polymer
materials;29,32,41,42 however, the details of this process remain
under-studied. This is a particularly interesting topic since the
reaction appears to be essentially static under ambient temp-
erature, although capable of significant exchange at elevated
temperatures. Studying both the kinetics of this reaction and
the process of this reaction has the potential to guide the
development of the next generation of bioconjugation reac-
tions and synthetic materials. Here, a model system of thio-
phenol (TPSH) and mercaptoethanol (MESH) adducts with
phenylvinylketone (PVK) based Michael acceptors in N,N-di-
methylformamide (DMF) is used to study the nature of the
Scheme 2 Dynamic covalent exchange of thiophenol (TPSH), thiophenol-phenylvinylketone adduct (TP-PVK), mercaptoethanol (MESH), mercap-
toethanol-phenylvinylketone adduct (ME-PVK), and phenylvinylketone (PVK). Rate coefficients of addition (ka) and fragmentation of the adducts (kf )
are used to describe the data.
2726 | Org. Biomol. Chem., 2018, 16, 2725–2734
View Article Online
Organic & Biomolecular Chemistry
thermally activated dynamic covalent exchange of thiol-
Michael adducts. This model system is shown in Scheme 2.
Experimental
Materials
All materials were purchased from commercial suppliers
unless otherwise specified. All materials were used as received
unless otherwise specified.
Nuclear magnetic resonance (NMR)
All nuclear magnetic resonance experiments were performed
on a Bruker 500 MHz NMR spectrometer.
High-resolution mass spectrometry (HRMS)
The HRMS data were recorded on a Thermo Scientific LTQ
Orbitrap XL mass spectrometer equipped with an Ion Max ESI
source. It was operated in the positive mode at 60 000 resolu-
tion, using direct infusion from a syringe at 5 μl min−1.
Synthesis of phenylvinylketone (PVK)
The synthesis of PVK was adapted from the literature.43,44
A solution of 3-chloropropiophenone (3.00 g, 17.79 mmol, 1.0
equiv.) in chloroform (50.00 g) was cooled to 0 °C. To this
stirred solution, triethyl amine (6.00 ml, 43.05 mmol, 2.4
equiv.) was added dropwise. The reaction mixture was stirred
for 18 h at room temperature, followed by washing with 0.1 M
HCl aq. (2 × 50 ml), distilled water (2 × 50 ml), saturated
NaHCO3 aq. (2 × 50 ml), and brine (1 × 50 ml). The organic
layer was dried over Na2SO4 and concentrated under reduced
pressure, giving the product PVK as a colorless oil in 99%
yield. The product was confirmed by 1H NMR spectroscopy,
agreeing with the literature.44,45 1H NMR (500 MHz, CDCl3)
δ 7.97–7.95 (m, 2H), 7.59–7.56 (t, 1H, J = 7.67 Hz), 7.50–7.46 (t,
2H, J = 7.84 Hz), 7.20–7.14 (m, 1H), 6.47–6.43 (dd, 1H, J = 1.61,
17.03 Hz), 5.95–5.92 (dd, 1H, J = 1.64, 10.59 Hz) ppm.
This journal is © The Royal Society of Chemistry 2018

Organic & Biomolecular Chemistry
Synthesis of para-chlorophenylvinylketone (PVK-Cl)
The synthesis of PVK-Cl followed the same process as PVK,
except that 3,4′-dichloropropiophenone was used instead of
3-chloropropiophenone. The product was confirmed by 1H NMR
spectroscopy, agreeing with the literature.45 1H NMR (500 MHz,
CDCl3) δ 7.95–7.84 (d, 2H, J = 8.57 Hz), 7.50–7.42 (d, 2H, J = 8.63
Hz), 7.18–7.06 (dd, 1H, J = 10.57, 17.52 Hz), 6.49–6.40 (dd, 1H, J =
1.65, 17.12 Hz), 5.99–5.92 (dd, 1H, J = 1.58, 10.58 Hz) ppm.
Published on 15 March 2018. Downloaded by University of New England on 20/04/2018 20:46:40.
Synthesis of para-methoxyphenylvinylketone (PVK-OMe)
3-Chloro-1-(4-methoxyphenyl)-1-propanone was synthesized
according to the literature.46 The product was used for the next
step without any purification. para-Methoxyphenylvinylketone
was synthesized following the same procedure as PVK, except
that 3-chloro-1-(4-methoxyphenyl)-1-propanone was used
instead of 3-chloropropiophenone. The product was confirmed
by 1H NMR spectroscopy, agreeing with the literature.45 1H
NMR (500 MHz, CDCl3) δ 7.99–7.95 (d, 2H), 7.21–7.14 (dd, 1H,
J = 10.52, 17.22 Hz), 6.98–6.94 (d, 2H), 6.45–6.39 (dd, 1H, J =
1.79, 17.18 Hz), 5.89–5.85 (dd, 1H, J = 1.85, 10.51 Hz),
3.98–3.82 (s, 3H) ppm.
Synthesis of 1-phenyl-3-( phenylthio)propan-1-one (TP-PVK)
A 25 ml glass vial equipped with a magnetic stir bar was
charged with PVK (0.2 g, 1.51 mmol, 1 equiv.), thiophenol
(TPSH) (0.1667 g, 1.51 mmol, 1 equiv.), 5% TEA (15 mg) and
dimethylformamide (DMF) (1.4260 g). The reaction was
allowed to continue at room temperature (25–30 °C) for 2 days
until all the alkene protons were consumed as monitored by
1
H NMR spectroscopy, giving the TP-PVK adduct in DMF as a
light yellow solution. The product was confirmed by 1H NMR
spectroscopy, agreeing with the literature.47 1H NMR
(500 MHz, CDCl3) δ 7.93–7.91 (dd, 2H, J = 1.45, 8.62 Hz),
7.60–7.55 (t, 1H, J = 7.45 Hz), 7.49–7.44 (t, 2H, J = 7.79 Hz),
7.40–7.36 (m, 2H), 7.33–7.28 (t, 2H, J = 7.78 Hz), 7.23–7.18 (t,
1H, J = 7.33 Hz), 3.33 (m, 4H) ppm.
Synthesis of 3-((2-hydroxyethyl)thio)-1-phenylpropan-1-one
(ME-PVK)
A 25 mL glass vial equipped with a magnetic stir bar was
charged with PVK (0.2 g, 1.51 mmol, 1 equiv.), 2-mercaptoetha-
nol (MESH) (0.1182 g, 1.51 mmol, 1 equiv.), 5% TEA (15 mg)
and dimethylformamide (DMF) (1.3775 g). The reaction was
This journal is © The Royal Society of Chemistry 2018
View Article Online
Paper
allowed to continue at room temperature (25–30 °C) for 2 days,
giving the ME-PVK adduct in DMF as a colorless solution. The
product was confirmed by 1H NMR spectroscopy, agreeing with
the literature.48 1H NMR (500 MHz, CDCl3) δ 7.99–7.97 (dd, 2H,
J = 1.48, 8.52 Hz), 7.63–7.58 (t, 1H, J = 7.44 Hz), 7.52–7.48 (t, 2H,
J = 7.70 Hz), 3.80–3.75 (t, 2H, J = 6.85 Hz), 3.37–3.31 (t, 2H, J =
7.18 Hz), 2.98–2.94 (t, 2H, J = 7.13 Hz), 2.78–2.73 (m, 2H) ppm.
Synthesis of 1-(4-methoxyphenyl)-3-( phenylthio)propan-1-one
(TP-PVK-OMe)
A 25 mL glass vial equipped with a magnetic stir bar was
charged with PVK-OMe (0.53 g, 0.00326 mol), thiophenol
(TPSH) (0.3780 g, 0.003423 mol), and a few drops of 5% TEA in
DMF solution. The reaction was allowed to continue at room
temperature (25–30 °C) for 1 day giving the TP-PVK-OMe
adduct. The product was purified by dissolving the reaction
mixture in dichloromethane (10 mL) and then washed with
water (3 × 50 mL), saturated sodium bicarbonate (2 × 50 mL),
and brine (50 mL). The organic solution was dried over mag-
nesium sulfate and the solvent was removed to give a solid
product (0.55049 g, 0.00202 mol, 62% yield). The product was
confirmed by 1H NMR, 13C NMR and HRMS. 1H NMR
(500 MHz, CDCl3) δ 7.91–7.88 (m, 2H), 7.38–7.36 (m, 2H),
7.31–7.28 (m, 2H), 7.21–7.18 (t, 1H, J = 7.35 Hz), 6.93–6.90 (m,
2H), 3.87 (s, 3H), 3.33–3.30 (m, 2H), 3.27–3.24 (m, 2H) ppm.
13
C NMR (126 MHz, CDCl3) δ 196.7, 163.7, 137.0, 136.0, 130.3,
129.7, 129.4, 129.1, 129.0, 127.5, 127.2, 126.2, 113.8, 55.5, 38.0,
28.1 ppm (Fig. S1†). HRMS (ESI, M + H+) calcd For C16H17O2S+:
273.0944, found: 273.0937.
Synthesis of 3-((2-hydroxyethyl)thio)-1-(4-methoxyphenyl)
propan-1-one (ME-PVK-OMe)
A 25 mL glass vial equipped with a magnetic stir bar was
charged with PVK-OMe (0.53 g, 0.00326 mol), 2-mercaptoetha-
nol (MESH) (0.26740 g, 0.003423 mol), and a few drops of 5%
TEA solution. The reaction was allowed to continue at room
temperature (25–30 °C) for 1 day giving the ME-PVK-OMe
adduct. The product was purified by dissolving the reaction
mixture in dichloromethane (10 mL) and then washed with
water (3 × 50 mL), saturated sodium bicarbonate (2 × 50 mL),
and brine (50 mL). The organic solution was dried over mag-
nesium sulfate and the solvent was removed to give a yellow oil
(0.336 g, 0.0014 mol, 43% yield). The product was confirmed by
1
H NMR, 13C NMR and high resolution mass spectrometry
(HRMS), agreeing with the literature.48 1H NMR (500 MHz,
CDCl3) δ 7.93–7.90 (m, 2H), 6.93–6.90 (m, 2H), 3.85 (s, 3H),
3.77–3.74 (t, 2H, J = 5.80 Hz), 3.23–3.20 (t, 2H, J = 7.20 Hz),
2.93–2.90 (t, 2H, J = 7.11 Hz), 2.77–2.74 (t, 2H, J = 5.94 Hz), 2.50
(s, 1H) ppm. 13C NMR (126 MHz, CDCl3) δ 196.7, 163.7, 130.4,
129.6, 113.8, 60.5, 55.5, 38.5, 35.8, 26.0 ppm (Fig. S2†). HRMS
(ESI, M + H+) calcd For C12H17O3S+: 241.0893, found: 241.0887.
Synthesis of 1-(4-chlorophenyl)-3-( phenylthio)propan-1-one
(TP-PVK-Cl)
The synthesis of TP-PVK-Cl followed the same process as for
TP-PVK-OMe. The isolated yield was 45% and the product was
Org. Biomol. Chem., 2018, 16, 2725–2734 | 2727

Paper
confirmed by 1H NMR, 13C NMR and HRMS. 1H NMR
(500 MHz, CDCl3) δ 7.86–7.83 (m, 2H), 7.43–7.41 (m, 2H),
7.38–7.36 (m, 2H), 7.32–7.29 (m, 2H), 7.23–7.19 (m, 1H),
3.33–3.25 (m, 4H) ppm. 13C NMR (126 MHz, CDCl3) δ 197.0,
139.8, 135.6, 134.8, 129.6, 129.5, 129.1, 129.0, 126.4, 38.4,
27.9 ppm (Fig. S3†). HRMS (ESI, M + H+) calcd for
C15H14ClOS+: 277.0448, found: 277.0438.
Synthesis of 1-(4-chlorophenyl)-3-((2-hydroxyethyl)thio)propan-
Published on 15 March 2018. Downloaded by University of New England on 20/04/2018 20:46:40.
1-one (ME-PVK-Cl)
The synthesis of TP-PVK-Cl followed the same process as for
ME-PVK-OMe. The isolated yield was 67% and the product was
confirmed by 1H NMR, 13C NMR and HRMS, agreeing with the
literature.48 1H NMR (500 MHz, CDCl3) δ 7.88–7.87 (m, 2H),
7.43–7.41 (m, 2H), 3.77–3.75 (t, 2H, J = 5.88 Hz), 3.26–3.23 (t,
2H, J = 7.14 Hz), 2.94–2.91 (t, 2H, J = 7.13 Hz), 2.77–2.74 (t, 2H,
J = 5.91), 2.47 (s, 1H) ppm. 13C NMR (126 MHz, CDCl3) δ 197.1,
139.9, 134.8, 129.5, 129.0, 60.6, 38.9, 35.7, 25.8 ppm (Fig. S4†).
HRMS (ESI, M + H+) calcd for C11H14ClO2S+: 245.0398, found:
245.0398.
Typical kinetic analysis
For 1 : 1 ratio, to the TP-PVK adduct solution was added 2-mer-
captoethanol (MESH) (0.1182 g, 1.51 mmol, 1 equiv.) and DMF
(0.1182 g). To the ME-PVK adduct solution was added thiol-
phenol (TPSH) (0.1667 g, 1.51 mmol, 1 equiv.) and DMF
(0.1667 g). 0.25 mL of each of the two mixtures was diluted
into a vial containing 1 mL DMF. A small volume of the
diluted reaction mixture was retained as a 0 h time point. The
rest of the diluted solutions were capped, protected by argon
during the reaction and heated to 60, 75 or 90 °C, or kept at a
room temperature of 20 °C for 24 h with 0.1 mL samples taken
at different lengths of time. The fraction of each adduct was esti-
mated using the ratio of integrals in the region 7.44–7.1 ppm
to the integral of peaks at 3.37–3.31 ppm. In the systems
initiated from the TP-PVK adduct and the ME-PVK adduct, the
integral in the region 7.44–7.1 ppm is 5 protons. However, the
TP-PVK adduct has an integral of 4 protons in the region
3.37–3.31 ppm, while the ME-PVK adduct has an integral of
2 protons in the region 3.37–3.31 ppm. Therefore, the mole
fraction of TP-PVK and ME-PVK can be estimated as follows:
I3:373:31 ppm  2
fTP‐PVK ¼
2 TP-PVK adduct or the ME-PVK adduct. However, since the reac-
f ME-PVK ¼ 1  f TP-PVK
A similar procedure was used for the dynamic exchange of
para-substituted phenylvinylketone derivatives. A similar pro-
cedure was performed where the ratio of MESH to TP-PVK
adduct was varied by changing the initial concentration of
2728 | Org. Biomol. Chem., 2018, 16, 2725–2734
View Article Online
Organic & Biomolecular Chemistry
MESH. Since the protons in the region 7.44–7.1 ppm all derive
from the TPSH fragment’s aromatic protons, not from the
PVK fragment’s aromatic protons, an integral in the region
7.44–7.1 ppm was assigned to 2.5 or 20 when the ratio of TPSH
to ME-PVK was 0.5 or 4. In these systems eqn (1) and (2) are
still used to calculate fTP-PVK and fME-PVK.
Results and discussion
The dynamic covalent system between thiophenol (TPSH),
thiophenol-phenylvinylketone adduct (TP-PVK), mercaptoetha-
nol (MESH), mercaptoethanol-phenylvinylketone adduct
(ME-PVK), and phenylvinylketone (PVK) given in Scheme 2 can
be represented through an overall equilibrium (KTP-ME-PVK).
This TP-ME-PVK system is chosen as a model system to deter-
mine the behavior of the thiol-Michael dynamic equilibrium
since it enables a range of analyses including Hammett ana-
lysis.49 In all cases, the NMR spectra show negligible signal in
the vinyl region (5–6.5 ppm),50 which indicates that the con-
centration of free PVK is very low, and it is likely a reactive
intermediate in this system with a low, steady state concen-
tration below the detection limit of NMR spectroscopy. Since
the free PVK cannot be quantified in these experiments, the
overall equilibrium is given below in terms of TPSH, MESH,
TP-PVK, and ME-PVK:
kf‐TP‐PVK ka‐ME‐PVK ½TPSHeq ½ME‐PVKeq
KTP‐ME‐PVK ¼ ¼ ;
ka‐TP‐PVK kf‐TP‐PVK ½TP‐PVKeq ½MESHeq
where the subscript eq refers to the equilibrium concentration.
Typical NMR exchange data given in Fig. S5† are given for
the TP-ME-PVK system starting from both the TP-PVK adduct
with 1 equivalent of MESH (Fig. S5A†) and starting from the
ME-PVK adduct with 1 equivalent of TPSH (Fig. S5B†). These
NMR data can be processed to extract the mole fraction of the
ME-PVK adduct and the mole fraction of the TP-PVK adduct.
These two fractions are plotted consistently in all figures in
this paper, and mole fraction specifically refers to the mole
fraction of TP-PVK and mole fraction of ME-PVK, with the sum
of these two mole fractions equaling 1. The TP-PVK adduct
with 1 equivalent of MESH in DMF, or the ME-PVK adduct
with 1 equivalent of TPSH in DMF, were heated at 90 °C for
24 h with kinetic samples taken at the allotted time point over
8 h, and the final point at 24 h. Fig. 1 gives the time evolution
of the TP-PVK adduct and the ME-PVK adduct. The first clear
conclusion is that the system under these conditions reached
equilibrium after 24 h, with [TP-PVK]eq/([TP-PVK]eq +
[ME-PVK]eq) ≈ 0.55 and [ME-PVK]eq/([TP-PVK]eq + [ME-PVK]eq)
≈ 0.45, regardless of whether the system was initiated from the
ð1Þ
tion is a dynamic covalent equilibrium process, involving
ð2Þ
forward and reverse reactions across 5 species, a simple first or
second order kinetic model cannot be used to describe the
system. Instead a kinetic model that explicitly accounts for
forward and reverse reactions was developed as outlined in the
ESI.†
This journal is © The Royal Society of Chemistry 2018
 View Article Online

Organic & Biomolecular Chemistry Paper

reaction initialized from ME-PVK). Further, knowledge of the

overall equilibrium constant KTP-ME-PVK obtained from the
experiment and the estimated values of kf-TP-PVK and kf-ME-PVK
gives the ratio ka-TP-PVK/ka-ME-PVK for systems that reached equi-
librium. These values are reported in Table 1. Fig. S6† shows
that the absolute values of the addition rate coefficients are
not important in the kinetic evolution, as long as the ratio of
the two addition rate coefficients is the same.
However, for systems that did not reach equilibrium,
Published on 15 March 2018. Downloaded by University of New England on 20/04/2018 20:46:40.

KTP-ME-PVK is not known. The data obtained for TP-PVK and

Fig. 1 Kinetics of thiol-Michael dynamic covalent exchange over 24 h,
and simulations of the data at 90 °C. System initialized from the thio-
phenol-phenylvinylketone adduct (TP-PVK-In) or the mercaptoethanol-
phenylvinylketone adduct (ME-PVK-In).
A sensitivity analysis of the kinetic model was performed to
evaluate the confidence of estimated parameters. As indicated
in Fig. S6–S13,† the kinetic model is very sensitive to the frag-
mentation rate coefficient (kf ) corresponding to the initial
adduct. Changes in the fragmentation rate coefficient (kf ) of
the adduct formed in exchange (not initialized from) lead to
minimal change in the fitted data, except at very long times.
Similarly, variations in absolute value and ratio of the addition
rate coefficients lead to minimal changes in the predicted
kinetic evolution.
The sensitivity analysis and the fact that the system is initia-
lized from both the TP-PVK and ME-PVK adducts indicates
that there is a high confidence over the estimated values
kf-TP-PVK (which can be estimated from the reaction initialized
from TP-PVK) and kf-ME-PVK (which can be estimated from the
ME-PVK indicate that the equilibrium constant is close to 1;
therefore for systems that did not reach equilibrium, but
rather only a small fraction of the initial adducts were
exchanged, a value of KTP-ME-PVK = 1 is used for these simu-
lations. The sensitivity analysis in Fig. S6–S13† suggests that
these systems are dominated by the fragmentation rate
coefficients (kf ) and small changes in addition rate coefficients
for either adduct (ka) will not lead to substantial changes in
the kinetic profiles.
Table 1 gives the kinetic parameters used to describe the
experimental dynamic covalent exchange data in Fig. 1. Fig. 1
also includes the theoretical (Th) data predicted from the
kinetic model with parameters given in Table 1, rows 1 and
2. In general the kinetic model provides a good description of
the experimental data with discrepancies between the theore-
tical and experimental data being less than the typical error of
NMR analysis of 1–5% and up to 10%.51–53 The good agree-
ment between experimental and kinetic modeling data
suggests that the transformation pathway in Scheme 2 is a
plausible pathway for dynamic covalent exchange of thiol-
Michael adducts. However, to rule out alternative pathways
such as direct attack by the free thiol, further kinetic studies
are needed.
Fig. 2 and 3 show the fitted and experimental data with
different ratios of the initial adduct (TP-PVK or ME-PVK) with
different ratios of the free thiol (MESH or TPSH). Note that the
same parameters are used to fit all experiments in this
TP-ME-PVK system as given in Table 1 rows 1 and 2. The
experimental kinetic data in Fig. 2 indicates that the initial
ratio of MESH : TP-PVK dictates the equilibrium position (after

Table 1 Summary of rate coefficients and obtained equilibrium constants used in all simulations

Initial adduct T (°C) Xa KTP-ME-PVK kf-TP-PVK (s−1) kf-ME-PVK (s−1) ka-TP-PVK/ka-ME-PVK (−)
−5 −5
TP-PVK 90 H 0.74 1.6 × 10 1.3 × 10 1.7
ME-PVK 90 H 0.58 1.6 × 10−5 1.3 × 10−5 2.1
TP-PVK 75 H 1b 4.8 × 10−6 4.0 × 10−6 1.2c
ME-PVK 75 H 1b 4.8 × 10−6 4.0 × 10−6 1.2c
TP-PVK 60 H 1b 1.7 × 10−6 2.0 × 10−6 0.9c
ME-PVK 60 H 1b 1.7 × 10−6 2.0 × 10−6 0.9c
TP-PVK 90 OMe 1b 4.8 × 10−6 4.3 × 10−6 1.1c
ME-PVK 90 OMe 1b 4.8 × 10−6 4.3 × 10−6 1.1c
TP-PVK 90 Cl 0.79 3.0 × 10−5 2.3 × 10−5 1.7
ME-PVK 90 Cl 0.83 3.0 × 10−5 2.3 × 10−5 1.6
a
X is the substituent para on the benzene ring of PVK. b Equilibrium not observed experimentally in 24 h, and value KTP-ME-PVK = 1 used in simu-
lation. c Estimated based on equilibrium constant of 1.

This journal is © The Royal Society of Chemistry 2018 Org. Biomol. Chem., 2018, 16, 2725–2734 | 2729

Paper
Published on 15 March 2018. Downloaded by University of New England on 20/04/2018 20:46:40.
Fig. 2 Kinetics and equilibrium position of dynamic covalent exchange
experiments (Ex) using three ratios of mercaptoethanol (MESH) to the
TP-PVK adduct at 90 °C. Kinetic simulations (Th) and predicted equili-
brium (Eq) presented. (A) Gives the mole fraction of TP-PVK, (B) gives
the mole fraction of ME-PVK. Insets of A and B give data for the first
2 hours of reaction.
24 h), with higher ratio of MESH : TP-PVK leading to less
TP-PVK at equilibrium. The dotted lines give the equilibrium
mole fraction of TP-PVK or ME-PVK based on the initial ratio
of MESH : TP-PVK and the mean equilibrium constant
KTP-ME-PVK = 0.66, taken as the average of Table 1 rows 1 and
2. The predicted equilibrium concentrations (Eq) of TP-PVK
and ME-PVK (dotted lines in Fig. 2) are in good agreement
with the experimentally observed equilibrium values.
In addition, the kinetic model (Th) provides an acceptable
agreement with the experimental values, especially since no
2730 | Org. Biomol. Chem., 2018, 16, 2725–2734
View Article Online
Organic & Biomolecular Chemistry
Fig. 3 Kinetics and equilibrium position of dynamic covalent exchange
experiments (Ex) using three ratios of thiophenolethanol (TPSH) to the
ME-PVK adduct at 90 °C. Kinetic simulations (Th) and predicted equili-
brium (Eq) presented. (A) Gives the mole fraction of TP-PVK, (B) gives
the mole fraction of mercaptoethanol-ME-PVK. Insets of A and B give
data for the first 2 hours of reaction.
additional fitting was performed. An important observation is
that in the early phase of the reaction (first two hours, inset
Fig. 2A and inset Fig. 2B), there is essentially no experi-
mentally observed difference in the kinetic evolution of
TP-PVK and ME-PVK adduct mole fractions, despite the ratio
of MESH : TP-PVK being changed by almost an order of magni-
tude. This provides evidence that the exchange of adducts
occurs through a fragmentation (or retro-Michael reaction)
pathway, followed by subsequent addition of the Michael
acceptor with a free thiol. If direct reaction of the free thiol, or
This journal is © The Royal Society of Chemistry 2018
 View Article Online

Organic & Biomolecular Chemistry Paper

its associated thiolate, on the adduct was significant, increases

in the free thiol concentration should increase the rate of
exchange, which is in contrast to the early phase kinetic data
obtained with different free thiol concentrations.
Fig. 3 provides similar data to Fig. 2, except that the ratio of
TPSH : ME-PVK is varied from 0.5 to 4 equivalents of thiol to
adduct. Again at 24 h, the concentration of TP-PVK and ME-PVK
agrees well with the values predicted from the mean equili-
brium constant KTP-ME-PVK = 0.66 for each system studied (dotted
Published on 15 March 2018. Downloaded by University of New England on 20/04/2018 20:46:40.

line for each system). Further, there is acceptable agreement

between the kinetic model predictions (Th, dashed lines), using
parameters from Table 1 rows 1 and 2, and the experimental
data (Ex) for each system. The agreement is notable, especially
considering that no further fitting was performed in describing
these data. The early phase data of each system (Fig. 3A inset
and Fig. 3B inset) indicate that there is minimal difference in
the rate of initial exchange with the different free thiol concen-
trations. This again provides evidence that the retro-Michael
fragmentation of adducts is involved and is rate determining in
the exchange of adducts. This similarity in the early phase of
the kinetics occurs over almost an order of magnitude in thiol-
concentration, which is inconsistent with a process where the
free thiol, or its thiolate, reacts with a thiol-Michael adduct.
Overall the data in Fig. 1–3 and the kinetic modeling corrobo-
rate that the thermally driven thiol-Michael reaction is consist-
ent with the mechanism proposed in Scheme 2.
To further understand the underlying thermally activated
thiol-Michael dynamic covalent process, the kinetics of
exchange were studied at 4 different temperatures 90, 75, 60 °C
and room temperature (363, 348, 333, and 293 K). Fig. 4 gives
the kinetic evolution of each elevated temperature system, as
well as experimentally fitted data. No measurable exchange of
thiol-Michael adducts was observed in the NMR spectra at
room temperature. Parameters used to fit the data are given in
Table 1. One key observation in Fig. 4A and B is that even after
24 h, the system has not approached equilibrium for the data
obtained at 75 and 60 °C, since the mole fraction of TP-PVK
and ME-PVK is not the same when initialized from TP-PVK or
ME-PVK. This is due to the clearly slower rate of exchange at
the lower temperatures of 75 and 60 °C, compared to the data
obtained at 90 °C. Due to the highly efficient thiol-Michael
adduct forming reaction, with rate coefficients of ca. 1000
M−1 s−1 for thiol-Maleimide coupling,54,55 it seems likely that
the fragmentation is the rate determining step in the dynamic
exchange. This is consistent with the kinetic modeling and the
experimental data in Fig. 1–3. Therefore, the lack of equili-
brium established even after 24 h at the lower temperatures is
likely due to a slow rate of fragmentation, or retro-Michael
reaction. Fitting these experimental data with the kinetic
model yields fragmentation rate coefficients for each adduct
(TP-PVK and ME-PVK) for each temperature studied. These
rate coefficients are given in Fig. 4C as an Arrhenius like plot
of log10 of the rate coefficient vs. 1/Temperature. The slope of
Fig. 4 Experimental and simulated (Th) kinetics of dynamic covalent
the Arrhenius plot can be used to estimate the activation exchange experiments at temperatures 363, 348 and 333 K. (A)
energy of the fragmentation, or retro-Michael, reaction. The Initialized from TP-PVK, (B) initialized from ME-PVK, and (C) Arrhenius
activation energy of TP-PVK is estimated as: Ea-TP-PVK = 75 plot of fitted adduct fragmentation rate coefficients.

This journal is © The Royal Society of Chemistry 2018 Org. Biomol. Chem., 2018, 16, 2725–2734 | 2731
 View Article Online

Paper Organic & Biomolecular Chemistry

kJ mol−1 while the activation energy of ME-PVK is estimated as:

Ea-ME-PVK = 63 kJ mol−1. These values of the activation energy
are similar to each other, within the uncertainty of the under-
lying experiments, and suggest a relatively slow process with a
high activation energy barrier, consistent with the observed
data. The pre-exponential factors (A) of the Arrhenius equation
are: ln(ATP-PVK/1 s−1) = 13.8 ln(AME-PVK/1 s−1) = 9.3.
Fig. 4C also extrapolates the Arrhenius plot to 1/T =
1/298 K, giving a predicted fragmentation rate coefficient of
Published on 15 March 2018. Downloaded by University of New England on 20/04/2018 20:46:40.

kf ≈ 10−7 s−1 for either adduct under ambient conditions. This

would correspond to a half-lifetime of exchange at this temp-
erature of t1/2 = ln(2)/kf ≈ 7 × 106 s ≈ 80 days. Therefore, under
ambient conditions of T = 25 °C there would be essentially no
exchange of thiol-Michael adducts observed even after ca.
2 months due to the very slow rate of retro-Michael reaction,
which is consistent with the experimental data. However, it is
important to note that elevated pH can induce retro-Michael
reactions and promote dynamic exchange even at ambient
temperature.33
In addition to temperature and concentration based kinetic
experiments, a Hammett analysis can be performed to provide
mechanistic insights. Fig. 5 compares the dynamic exchange
kinetics for a range of para substituted (X group) vinyl ketone
adducts. In particular, the methoxy group (OMe) and the
chloro group (Cl) are used to generate a system with an elec-
tron donating (OMe) group and an electron withdrawing group
(Cl) relative to the unsubstituted (H) system. The data in
Fig. 5A and B clearly indicate that the withdrawing Cl group
accelerates the rate of dynamic exchange, whereas the donat-
ing OMe group decelerates the dynamic exchange. These
kinetic data are described using the kinetic model using para-
meters outlined in Table 1. The sensitivity analysis in Fig. S6–
S13† indicates that the kinetic model is most sensitive to the
fragmentation rate coefficient (kf ) of the adduct used to initia-
lize the system. Therefore, the increasing trend kf-TP-PVK-X and
kf-ME-PVK-X with electron withdrawing character (increasing
Hammett σp parameter) can be reliably used to construct a
Hammett plot.49 Fig. 5C gives a Hammett plot of kf-TP-PVK-X
and kf-ME-PVK-X against the Hammett σp parameter,49 giving a
strong positive slope and good linear fit. The Hammett slopes
are 1.6 for kf-TP-PVK-X and 1.5 for kf-ME-PVK-X, which are indicative
of strong negative charge character in the retro-Michael reac-
tion’s rate determining step.
This negative charge character seems counterintuitive due
to the neutral nature of all compounds shown in Scheme 2.
However, close inspection of the proposed mechanism of the
thiol-Michael adduct formation reaction shows a chain reac-
tion in thiolate and an anionic thioether-enolate intermedi-
ate.23 This thiol-Michael adduct formation reaction is given in
Scheme 3. The anionic intermediate RS-PVK-X− is likely to
explain the strong negative charge character suggested by the
Hammett analysis in Fig. 5C. Of particular relevance to this
analysis is the principle of microscopic reversibility, which
Fig. 5 Experimental and simulated (Th) kinetics of dynamic covalent
states that if a reaction pathway is followed in the forward exchange experiments at 90 °C with different substituents. (A) Initialized
direction in a chemical process, the microscopic reverse of from TP-PVK, (B) initialized from ME-PVK, and (C) Hammett plot of
this pathway is also traversed in the reverse direction.56 fitted adduct fragmentation rate coefficients across the substituents.

2732 | Org. Biomol. Chem., 2018, 16, 2725–2734 This journal is © The Royal Society of Chemistry 2018

Organic & Biomolecular Chemistry
Scheme 3 Proposed thiol-Michael and retro thiol-Michael reaction
mechanism.
Published on 15 March 2018. Downloaded by University of New England on 20/04/2018 20:46:40.
Therefore, if the anionic intermediate RS-PVK-X− is an inter-
mediate in the adduct forming thiol-Michael reaction, it is
also an intermediate in the retro-Michael reaction. In the con-
jugation reaction thiolate attack is the rate determining step,
implying that loss of thiolate is also significant in the elimin-
ation process.57 The Hammett analysis in Fig. 5 suggests that
formation of the RS-PVK-X− intermediate, with its strong
anionic character, is a major contributor to the rate-determin-
ing step of the dynamic exchange of thiol-Michael adducts.
These results combined together suggest that dynamic
exchange of thiol-Michael adducts is primarily dictated by the
fragmentation, or retro-Michael reaction. Therefore, dynamic
properties in materials and bioconjugates based on the thiol-
Michael adduct will be primarily dictated by the rate of the
unimolecular retro-Michael reaction. This implies that the rate
of dynamic exchange is unlikely to be significantly accelerated
by the addition of large quantities of free thiol or free-Michael
acceptor, although this could lead to a change in material
composition, especially if a large excess of thiol or Michael
acceptor were used. Additionally, this implies that a material
containing a thermally labile thiol-Michael adduct can be acti-
vated to dynamic exchange, even in the absence of free thiol or
free Michael acceptor, since the intrinsic retro-Michael reac-
tion can drive the dynamic exchange.
Conclusions
A mechanistic study of the thermally activated dynamic
covalent chemistry of thiol-Michael adducts was performed
using a model system of thiophenol/mercaptoethanol dynamic
equilibrium with phenylvinylketone based Michael acceptors
in N,N-dimethylformamide. The dynamic equilibrium of thiol-
Michael adducts was consistent with a retro-Michael reaction
followed by subsequent addition of a free thiol to the generated
Michael adduct. The reaction rate was primarily dictated by the
rate of the retro-Michael reaction, or fragmentation of thiol-
Michael adducts. The early phase kinetics of dynamic thiol-
Michael exchange was essentially independent of the thiol con-
centration, suggesting that direct reaction of a free thiol with
the thiol-Michael adduct was negligible. Additionally, these
data are consistent with the retro-Michael step being rate deter-
mining in the dynamic exchange of thiol-Michael adducts.
Temperatures of at least 90 °C were needed to observe equili-
bration in 24 h, for thiol-phenylvinylketone adducts, and at
This journal is © The Royal Society of Chemistry 2018
View Article Online
Paper
least 60 °C were needed in DMF to observe greater than 10%
exchange of thiol-phenylvinylketone adducts in 24 h. Hammett
analysis indicated that a donating group on the Michael accep-
tor decreased the rate of dynamic exchange, while a withdraw-
ing group accelerated the reaction, suggesting that the retro-
Michael reaction goes through an intermediate with strongly
anionic character. This work provided insights into the
dynamic thermally driven thiol-Michael reaction, and can facili-
tate the optimization of the reaction in dynamic covalent
libraries and (bio)materials that utilize dynamic bonding.
Conflicts of interest
There are no conflicts to declare.
Acknowledgements
We thank Hannah Lacy for contribution to the modeling of
the reaction kinetics. We thank Theresa Ramelot for experi-
mental assistance with the HRMS data. Acknowledgment is
made to the Donors of the American Chemical Society
Petroleum Research Fund (57243-DNI7) for support of this
research.
References
1 S. J. Rowan, S. J. Cantrill, G. R. Cousins, J. K. Sanders and
J. F. Stoddart, Angew. Chem., Int. Ed., 2002, 41, 898.
2 G. Gasparini, M. Dal Molin, A. Lovato and L. J. Prins,
Supramolecular Chemistry: From Molecules to Nanomaterials,
2012.
3 Y. Jin, C. Yu, R. J. Denman and W. Zhang, Chem. Soc. Rev.,
2013, 42, 6634.
4 Y. Jin, Q. Wang, P. Taynton and W. Zhang, Acc. Chem. Res.,
2014, 47, 1575.
5 A. Granzhan, T. Riis-Johannessen, R. Scopelliti and
K. Severin, Angew. Chem., Int. Ed., 2010, 122, 5647.
6 J.-M. Lehn, Prog. Polym. Sci., 2005, 30, 814.
7 C. J. Kloxin, T. F. Scott, B. J. Adzima and C. N. Bowman,
Macromolecules, 2010, 43, 2643.
8 A. Wilson, G. Gasparini and S. Matile, Chem. Soc. Rev.,
2014, 43, 1948.
9 W. Zhang and Y. Jin, Dynamic Covalent Chemistry:
Principles, Reactions, and Applications, John Wiley & Sons,
2017.
10 G. Hizal, U. Tunca and A. Sanyal, J. Polym. Sci., Part A:
Polym. Chem., 2011, 49, 4103.
11 P. J. Boul, P. Reutenauer and J.-M. Lehn, Org. Lett., 2005, 7,
15.
12 R. Nishiyabu, Y. Kubo, T. D. James and J. S. Fossey, Chem.
Commun., 2011, 47, 1124.
13 M. E. Belowich and J. F. Stoddart, Chem. Soc. Rev., 2012, 41,
2003.
Org. Biomol. Chem., 2018, 16, 2725–2734 | 2733

Paper
14 M. Schuster and S. Blechert, Angew. Chem., Int. Ed. Engl.,
1997, 36, 2036.
15 H. Otsuka, S. Nagano, Y. Kobashi, T. Maeda and
A. Takahara, Chem. Commun., 2010, 46, 1150.
16 Y.-X. Lu and Z. Guan, J. Am. Chem. Soc., 2012, 134, 14226.
17 B. Shi and M. F. Greaney, Chem. Commun., 2005, 886, DOI:
10.1039/B414300K.
18 S. P. Black, J. K. Sanders and A. R. Stefankiewicz, Chem.
Soc. Rev., 2014, 43, 1861.
Published on 15 March 2018. Downloaded by University of New England on 20/04/2018 20:46:40.
19 M. Ortiz, C. Yu, Y. Jin and W. Zhang, Top. Curr. Chem.,
2017, 375, 69.
20 N. Roy, B. Bruchmann and J.-M. Lehn, Chem. Soc. Rev.,
2015, 44, 3786.
21 C. J. Kloxin and C. N. Bowman, Chem. Soc. Rev., 2013, 42,
7161.
22 T. Maeda, H. Otsuka and A. Takahara, Prog. Polym. Sci.,
2009, 34, 581.
23 D. P. Nair, M. Podgórski, S. Chatani, T. Gong, W. Xi,
C. R. Fenoli and C. N. Bowman, Chem. Mater., 2014, 26, 724.
24 J. W. Chan, C. E. Hoyle, A. B. Lowe and M. Bowman,
Macromolecules, 2010, 43, 6381.
25 S. Chatani, T. Gong, B. A. Earle, M. Podgórski and
C. N. Bowman, ACS Macro Lett., 2014, 3, 315.
26 C. E. Hoyle, A. B. Lowe and C. N. Bowman, Chem. Soc. Rev.,
2010, 39, 1355.
27 B. H. Northrop, S. H. Frayne and U. Choudhary, Polym.
Chem., 2015, 6, 3415.
28 H. C. Kolb, M. G. Finn and K. B. Sharpless, Angew. Chem.,
Int. Ed., 2001, 40, 2004.
29 P. Chakma, L. H. Rodrigues Possarle, Z. A. Digby, B. Zhang,
J. L. Sparks and D. Konkolewicz, Polym. Chem., 2017, 8, 6534.
30 C. F. H. Allen, J. O. Fournier and W. J. Humphlett,
Can. J. Chem., 1964, 42, 2616.
31 C. F. H. Allen and W. J. Humphlett, Can. J. Chem., 1966, 44,
2315.
32 B. Zhang, Z. A. Digby, J. A. Flum, P. Chakma, J. M. Saul,
J. L. Sparks and D. Konkolewicz, Macromolecules, 2016, 49,
6871.
33 Y. Zhong, Y. Xu and E. V. Anslyn, Eur. J. Org. Chem., 2013,
5017.
34 E. H. Krenske, R. C. Petter and K. N. Houk, J. Org. Chem.,
2016, 81, 11726.
35 S. D. Fontaine, R. Reid, L. Robinson, G. W. Ashley and
D. V. Santi, Bioconjugate Chem., 2015, 26, 145.
2734 | Org. Biomol. Chem., 2018, 16, 2725–2734
View Article Online
Organic & Biomolecular Chemistry
36 A. D. Baldwin and K. L. Kiick, Bioconjugate Chem., 2011, 22,
1946.
37 G. Joshi and E. V. Anslyn, Org. Lett., 2012, 14, 4714.
38 J. Chen, X. Jiang, S. L. Carroll, J. Huang and J. Wang, Org.
Lett., 2015, 17, 5978.
39 D. Kalia, S. P. Pawar and J. S. Thopate, Angew. Chem., Int.
Ed., 2017, 56, 1885.
40 P. M. Kharkar, K. L. Kiick and A. M. Kloxin, Polym. Chem.,
2015, 6, 5565.
41 N. Kuhl, R. Geitner, R. K. Bose, S. Bode, B. Dietzek,
M. Schmitt, J. Popp, S. J. Garcia, S. van der Zwaag,
U. S. Schubert and M. D. Hager, Macromol. Chem. Phys.,
2016, 217, 2541.
42 N. Kuhl, R. Geitner, J. Vitz, S. Bode, M. Schmitt, J. Popp,
U. S. Schubert and M. D. Hager, J. Appl. Polym. Sci., 2017,
134, 44805.
43 F. J. Golemba and J. E. Guillet, Macromolecules, 1972, 5,
212.
44 S. Chanthamath, S. Takaki, K. Shibatomi and S. Iwasa,
Angew. Chem., Int. Ed., 2013, 52, 5818.
45 S.-H. Guo, S.-Z. Xing, S. Mao, Y.-R. Gao, W.-L. Chen and
Y.-Q. Wang, Tetrahedron Lett., 2014, 55, 6718.
46 H. A. Khan, K. G. M. Kou and V. M. Dong, Chem. Sci., 2011,
2, 407.
47 D. Das, S. Pratihar and S. Roy, J. Org. Chem., 2013, 78, 2430.
48 E. Unluer, H. I. Gul, A. Demirtas, H. Sakagami,
N. Umemura, M. Tanc, C. Kazaz and C. T. Supuran,
J. Enzyme Inhib. Med. Chem., 2016, 31, 105.
49 C. Hansch, A. Leo and R. W. Taft, Chem. Rev., 1991, 91,
165.
50 M. Balci, Basic 1H-and 13C-NMR spectroscopy, Elsevier,
2005.
51 M. Bauer, A. Bertario, G. Boccardi, X. Fontaine, R. Rao and
D. Verrier, J. Pharm. Biomed. Anal., 1998, 17, 419.
52 K. McLeod and M. B. Comisarow, J. Magn. Reson., 1989, 84,
490.
53 S. E. Cremer and B. C. Trivedi, J. Am. Chem. Soc., 1969, 91,
7200.
54 S. Davey, Nat. Chem. Biol., 2015, 11, 241.
55 F. Saito, H. Noda and J. W. Bode, ACS Chem. Biol., 2015, 10,
1026.
56 S. Chandrasekhar, Res. Chem. Intermed., 1992, 17, 173.
57 E. H. Krenske, R. C. Petter, Z. Zhu and K. N. Houk, J. Org.
Chem., 2011, 76, 5074.
This journal is © The Royal Society of Chemistry 2018