Tuning the reactivity and efficiency of copper catalysts for atom transfer radical polymerization by synthetic modification of tris(2-methylpyridyl)amine

Article abstract of DOI:10.1016/j.polymer.2017.09.018

Copper complexes bearing ligands based on tris(2-methylpyridyl)amine (TPMA) are leading catalysts in atom transfer radical polymerization (ATRP). In recent years, the TPMA structure has been finely modified to enhance the catalytic activity of the corresp

Full text of DOI:10.1016/j.polymer.2017.09.018

Accepted Manuscript
Tuning the reactivity and efficiency of copper catalysts for atom transfer radical
polymerization by synthetic modification of tris(2-methylpyridyl)amine
Nadia Alessandra Carmo dos Santos, Francesca Lorandi, Elena Badetti, Klaus Wurst,
Abdirisak Ahmed Isse, Armando Gennaro, Giulia Licini, Cristiano Zonta
PII:
S0032-3861(17)30880-7
10.1016/j.polymer.2017.09.018
JPOL 19987
DOI:
Reference:
To appear in: Polymer
Received Date: 19 July 2017
Revised Date: 8 September 2017
Accepted Date: 9 September 2017
Please cite this article as: Carmo dos Santos NA, Lorandi F, Badetti E, Wurst K, Isse AA, Gennaro
A, Licini G, Zonta C, Tuning the reactivity and efficiency of copper catalysts for atom transfer radical
polymerization by synthetic modification of tris(2-methylpyridyl)amine, Polymer (2017), doi: 10.1016/
j.polymer.2017.09.018.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
Tuning the reactivity and efficiency of copper catalysts for
atom transfer radical polymerization by synthetic
modification of tris(2-methylpyridyl)amine
Nadia Alessandra Carmo dos Santos,a‡ Francesca Lorandi,a‡ Elena Badetti,a Klaus Wurst,b
Abdirisak Ahmed Isse,a* Armando Gennaro,a Giulia Licinia and Cristiano Zontaa*
aDepartment of Chemical Sciences, University of Padova, via Marzolo 1, 35131 Padova, Italy.
bInstitut für Allgemeine, Anorganische und Theoretische Chemie, Innrain 80/82, A-6020
Innsbruck, Austria
‡These authors contributed equally
Keywords: Tris(2-methylpyridyl)amine ligand, Atom Transfer Radical Polymerization, catalyst
activity tuning, electrochemically mediated polymerization, controlled aqueous polymerizations
Abstract
Copper complexes bearing ligands based on tris(2-methylpyridyl)amine (TPMA) are leading
catalysts in atom transfer radical polymerization (ATRP). In recent years, the TPMA structure
has been finely modified to enhance the catalytic activity of the corresponding Cu complex. In
this study, we report the synthesis of eight novel copper TMPA complexes, and the evaluation of
the catalytic activity for electrochemically mediated ATRP in both organic and aqueous media.
The TPMA skeleton was modified by inserting differently m-functionalised phenyl substituents.
Complete characterizations and electrochemical investigations were performed to define the
structural and redox properties of the new complexes and compare them to the archetype
1

ACCEPTED MANUSCRIPT
Cu/TPMA. Well-controlled polymerizations of some hydrophobic, hydrophilic and acidic
monomers are presented, with conversions > 80%, low dispersity and predetermined molecular
weights, thus confirming that these compounds are suitable ATRP catalysts for different systems.
1. Introduction
Over the past years, atom transfer radical polymerization (ATRP) strongly impacted the
world of polymer sciences thanks to its capability to generate macromolecules with
predictable architectures, functionalities and compositions [1-3]. This versatile process
was successfully applied to a large variety of monomers [4], showing great flexibility
towards solvents, functional groups and potential impurities. ATRP is a controlled radical
polymerization technique catalysed by transition metal complexes, mainly copper with
polyamine ligands. The process is initiated by the transfer of a halogen atom from an
alkyl halide initiator (R-X) to the active form of the catalyst, [CuIL]+. This generates a
free radical and the metal complex at a higher oxidation state, [XCuIIL]+ (Scheme 1)
[5,6]. The radical propagates for a short period, adding only to few monomer units, before
it is converted back to a dormant state (Pn-X) by a deactivation reaction with [XCuIIL]+.
ATRP equilibrium is strongly shifted towards the dormant species [7], so that the
concentration of propagating radicals is very low and radical-radical termination events
are predominantly suppressed. Therefore, the key-feature of the process is the continuous
activation-(propagation)/fast deactivation cycle, which enables concurrent growth of all
chains and hence very low molecular weight dispersity (Mw/Mn < 1.5).
2

ACCEPTED MANUSCRIPT
Scheme 1. General ATRP mechanism.
The catalyst plays a fundamental role, regulating the equilibrium between dormant and
active species along the whole process. A considerable disadvantage of ATRP, as it was
originally conceived, was the large amount of copper catalyst needed. Recently, several
methods using low-ppm levels of copper catalysts have been developed. These include
activators re-generated by electron transfer (ARGET) ATRP [8,9], initiators for
continuous activator regeneration (ICAR) ATRP [9,10], supplemental activator and
reducing agent (SARA) ATRP [11] also known as SET-LRP [12], photochemically
mediated ATRP [13,14] and electrochemically mediated ATRP (eATRP) [15-18].
Besides these advanced techniques, the quest to decrease the catalyst loading led to the
synthesis of novel ligands for the preparation of more active copper catalysts. Copper
complexes with ligands based on tris(2-methylpyridyl)amine (TPMA) are among the
most active catalysts in ATRP [19-21]. In particular, Cu/TPMA is the best performing
catalyst in aqueous media [17]; it is also a fundamental catalyst in dispersed media
combined with anionic surfactants [22]. The reactivity of ATRP catalysts can be directly
related to the standard reduction potential (Eo) of the Cu(II)/Cu(I) couple. Both the
activation rate constant and the equilibrium constant of ATRP increase by enhancing the
reducing power of the catalyst, i.e. decreasing Eo [21,23]. In this context, the ligand plays
a fundamental role in regulating the catalyst activity; it can modify the charge density on
the metallic centre, influencing the relative stability of Cu(II) and Cu(I). Therefore, the
ligand structure can be modified in order to tune the electronic properties of the metallic
3

ACCEPTED MANUSCRIPT
centre. Copper complexes with TPMA bearing methyl and methoxy groups on one or
more pyridine rings have been recently prepared to enhance catalyst activity in ATRP
[19,20]. However, catalysts with extremely high activity are not necessarily required and
can be even detrimental in particular environments, like in the presence of water, which
dramatically increases the activity of traditional Cu complexes [15,24]. In fact, one of the
major drawbacks to controlling aqueous ATRP has been identified as high KATRP
=
kact kdeact, resulting from the extremely fast activation, combined with low stability of
/
[XCuIIL]+ deactivators [16,17,25]. More important in aqueous media is the stability of the
catalyst over a broad range of pH, a requirement that is not fulfilled by many aliphatic
ligands, thus TPMA and its analogues are generally used [15,22,26,27].
In this study, we report eight novel TPMA copper complexes in which the skeleton has
been modified by adding functionalised phenyl substituents. The redox properties of these
complexes, in the absence and presence of halide ions, were analysed by cyclic
voltammetry and compared with the unmodified TPMA system, henceforth referred to as
TPMA, in order to predict their reactivity in
ATRP techniques that use low-ppm levels of catalyst,
interest, mainly because it allows to gain a deep understanding of the process besides
introducing additional tools for control [28]. Herein, we used ATRP on various
e
ATRP systems. Among the advanced
eATRP is receiving a growing
e
monomers with [CuIIL]2+ as catalyst, where L is TPMA or one of our modified ligands.
By comparing redox and catalytic features, it was possible to identify systems in which
phenyl-substitution can be successful and even beneficial.
2. Materials and Methods
4

ACCEPTED MANUSCRIPT
2.1. Chemicals
Acetonitrile (VWR) was distilled over CaH2 and stored under N2. Milli-Q ultrapure water
was used to prepare aqueous solutions. Methyl methacrylate (MMA, Sigma-Aldrich,
99%), oligo(ethylene glycol) methyl ether methacrylate (OEOMA, Sigma-Aldrich, Mn
=
500) were purified by passing through a column filled with active basic aluminium oxide
(Al2O3, VWR chemicals) to remove polymerization inhibitors. Methacrylic acid (MAA,
Sigma-Aldrich, 99%) was distilled through a 10 cm Vigreux column to remove
polymerization inhibitors. Tetraethylammonium tetrafluoroborate ((C2H5)4NBF4, Alfa
Aesar, 99%), used as supporting electrolyte, was recrystallized from ethanol and dried in
a vacuum oven at 70 °C for 48 h. Tetraethylammonium bromide ((C2H5)4NBr, Sigma-
Aldrich, 99%) and tetraethylammonium chloride ((C2H5)4NCl, Sigma-Aldrich, ≥ 98%),
used as source of halide ions, were recrystallized from ethanol and ethanol/acetone (2/1
v/v), respectively, and dried at 70 °C under vacuum for 48 h. All other commercial
products were used without further purification.
2.2. Experimental techniques
1H and 13C{1H} NMR spectra were recorded at 301 K on a Bruker AC-400, AC-300,
1
and AC-200 MHz instruments. The H NMR spectra were referenced to the solvent
residual peak of MeOD-d4 (3.31 ppm) or CD3CN-d3 (1.94 ppm); the 13C NMR spectra (50
MHz) were referenced to MeOD (49.00
118.26
±
0.02 ppm) or CD3CN peaks (1.32±0.02 and
1
±
0.02 ppm). Monomer conversion was determined by H NMR at room
temperature. Spectra of samples collected in organic solvents were recorded in CDCl3,
whereas D2O was used for water-soluble monomers. Cyclic Voltammetry (CV) and
polymerization experiments were carried out in a three-electrode cell by means of an
5

ACCEPTED MANUSCRIPT
Autolab PGSTAT 30 potentiostat/galvanostat (EcoChemie, The Netherlands) run by a PC
with Autolab GPES 4.9 software. In CV experiments, a Pt ring was used as counter
electrode, whereas the working electrode was a glassy carbon (GC) disk, fabricated from
a 3 mm diameter GC rod (Tokai GC-20). Before each experiment, the GC surface was
cleaned by polishing with a 0.25-
µm diamond paste, followed by ultrasonic rinsing in
ethanol for 5 min. In ATRP experiments, a Pt mesh and a Pt foil were used as working
e
and counter electrodes, respectively. The working electrode was electrochemically
activated by cycling the potential between -0.7 and 1 V vs. Hg/Hg2SO4 in 0.5 M H2SO4,
for 30 min at a scan rate of 0.05 V/s. The counter electrode was separated from the
cathodic compartment through a glass frit and a methylcellulose gel saturated with
(C2H5)4NBF4. In aqueous media, the reference electrode was a saturated calomel
electrode (SCE), whereas Ag|AgI|0.1 M
n-Bu4NI in DMF was used as a reference
electrode in organic solvents. In the latter case, the reference electrode was calibrated
versus the ferrocenium/ferrocene couple (Fc+/Fc) after each experiment and potentials
were converted to the SCE scale by using the known values of Eo(Fc+/Fc) vs. SCE: -0.390
in CH3CN and-0.476 V in DMF [29]. During electrolysis, the cathodic compartment was
maintained under vigorous magnetic stirring and Ar atmosphere. The cell had a double
wall jacket through which water was circulated from a thermostated bath (Thermo
Scientific, HAAKE SC100), to keep the temperature constant at 25 °C. Activation rate
constants were measured by means of a Rotating Disk Electrode, following a procedure
previously reported in full details [5,6]. During polymerizations, samples were withdrawn
periodically and analysed by GPC to determine number average molecular weight (Mn
)
and dispersity (Mw Mn). The GPC instrument (Agilent 1260 Infinity) was equipped with
/
6

ACCEPTED MANUSCRIPT
PLGel Columns and a refractive index detector. The mobile phase was either THF
(PMMA analysis) or DMF containing 1% v/v triethylamine and 1% v/v glacial acetic acid
for the analysis of POEOMA. The flow rate was 1 mL/min, and the columns and detector
were thermostated at 70 °C and 50 °C, respectively for POEOMA, and at 35 °C for
PMMA. The column system was calibrated with 12 linear PMMA (Mn = 634-1944000).
GPC analysis of polymethacrylic acid was carried out on PL aquagel-OH columns at 30
°C; the mobile phase was aqueous 0.1 M Na2HPO4 at a flow rate of 1 mL/min.
Ligands structures were characterized by ESI-MS experiments, carried out in positive
mode with an Agilent Technologies LC/MSD Trap SL AGILENT instrument. The mobile
phase was either methanol or acetonitrile. MS peaks are reported as monoisotopic mass.
Microanalyses were performed with a Flash 2000 Thermo Scientific Analyser.
X-ray analysis of each structure was performed, and supplementary crystallographic
data were deposited as CCDC-1542046-49 (2a 1542046, 2d 1542047, 3a 1542048 and 3c
1542049). Copies of the data can be obtained, free of charge, at the Cambridge
Crystallographic Data Centre.
2.3. General procedure of eATRP
All electrodes, electrolyte, solvent and reagents, except the initiator RX, were introduced
into the cell, and thermostated at 25 °C. After degassing with Ar for 20 min, cyclic
voltammetry of [XCuIIL]+ was recorded to measure the standard reduction potential of the
catalyst and to select the suitable Eapp value. The initiator was then added and the cell was
carefully degassed for 20 min. Last, the selected potential was applied, under continuous
vigorous magnetic stirring. Samples were withdrawn periodically to determine polymer
molecular weight and conversion.
7

ACCEPTED MANUSCRIPT
2.4. General synthetic procedures of copper complexes
0.25 mmol of ligand 1a-d were dissolved in anhydrous acetonitrile (15 mL) and 43 mg
(0.25 mmol) of the desired copper salt (Cu(ClO4)2•6H2O or CuCl2•2H2O) were added.
The solution was left at room temperature for 15 min, then diethyl ether was added to
give a precipitate that was filtered off. The residual solvent was then removed under
reduced pressure to give the desired copper complex. Caution! Perchlorate salts of metal
complexes with organic ligands are potentially explosive. They should be handled in
small quantity and with caution.
2a Dark blue crystal (85%). IR (KBr, cm-1): 3057, 2333, 1605, 1449, 1074, 769, 619.
ESI+ MS (m/z) Calc. C24H22N4CuClO4 528.1, Found 528.0 (M+). Elemental analysis
Calc. C24H22N4Cu. 2 ClO4. CH3CN: C= 46.61 %, H = 3.76 %, N = 10.45 %, Found: C
=46.25 %, H = 3.93 %, N = 10.45 %.
2b Dark green solid (97%). IR (KBr, cm-1): 3482, 2353, 1693, 1606, 1448, 1172, 776,
622. ESI+ MS (m/z) Calc. C25H22N4CuClO5 556.1, Found 556.1 (M+). Elemental analysis
Calc. C25H22N4Cu. 2 ClO4. 6 H2O: C= 45.7 %, H = 3.38 %, N = 8.53 %, Found: C =45.26
%, H = 3.66 %, N = 8.79 %.
2c Blue solid (95%). IR (KBr, cm-1): 3357, 2354, 1660, 1610, 1572, 1442, 1082, 764,
620. ESI+ MS (m/z) Calc. C25H23N5CuClO5 571.1, Found 571.0 (M+). Elemental analysis
Calc. C25H23N5OCu. 2 ClO4. 3H2O: C= 41.36 %, H = 4.03 %, N = 9.65 %, Found: C
=41.71 %, H = 3.60 %, N = 9.98 %.
2d Dark blue solid (80%). IR (KBr, cm-1): 3467, 3086, 1601, 1449, 1094, 786, 619. ESI+
MS (m/z) Calc. C25H24N4CuClO5 558.1, Found 558.0 (M+). Elemental analysis Calc.
8

ACCEPTED MANUSCRIPT
C25H24N4OCu. 2 ClO4. CH3CN: C= 46.33 %, H = 3.89 %, N = 10.01 %, Found: C =46.07
%, H = 4.07 %, N = 10.27 %.
3a Green crystals (87%). IR (KBr, cm-1): 1583, 1445, 765, 692. ESI+ MS (m/z) Calc.
C24H22N4CuCl 464.1, Found 464.1 (M+). Elemental analysis Calc. C24H22N4CuCl2: C=
57.55 %, H = 4.43 %, N = 11.19 %, Found: C = 57.82 %, H = 4.59 %, N = 11.42 %.
3b Green solid (89%). IR (KBr, cm-1): 3401, 3065, 1662, 1573, 1448, 801, 771. MS (m/z)
ESI+ MS (m/z) Calc. C25H22N4OCuCl 492.1, Found 492.1 (M+). Elemental analysis Calc.
C25H22N4OCuCl2. CH3CN: C= 56.77 %, H = 4.19 %, N = 10.59 %, Found: C =55.97 %,
H = 4.27 %, N = 10.57 %.
3c Light green needle shape crystal (93%). IR (KBr, cm-1): 3404, 3064, 2353, 1601, 1439,
1382, 763. ESI+ MS (m/z) Calc. C25H23N5OCuCl 507.1, Found 507.0 (M+). Elemental
analysis Calc. C25H23N5OCuCl2. CH3CN: C= 55.44 %, H = 4.48 %, N = 14.37 %, Found:
C =55.54 %, H = 4.52 %, N = 14.57 %.
3d Dark green crystal (85%). IR (KBr, cm-1): 3382, 3020, 2915, 1589, 1439, 1288, 1186,
1005, 799. ESI+ MS (m/z) Calc. C25H24N4OCuCl 494.1, Found 494.0(M+). Elemental
analysis Calc. C25H24N4OCuCl2: C= 56.56 %, H = 4.56 %, N = 10.55 %, Found: C =
56.33 %, H = 4.78 %, N = 10.72 %.
3. Results and discussion
Tris(2-methylpyridyl)amine complexes are widely used in catalysis and molecular
recognition. This is due to their polydentate nature, which allows the formation of stable
metal complexes also under turnover conditions. In the recent years, we have been
interested in the study of the catalytic and molecular recognition properties of this class of
9

ACCEPTED MANUSCRIPT
complexes [30-33]. In particular, Co(II) complexes in which the ligand bears different
functional groups have been synthesized and studied as hydrogen evolving catalysts, in
light-assisted conditions [34].
Scheme 2. General synthetic procedures for compounds 2a-d and 3a-d
.
The initial findings on the catalytic properties of these newly developed systems have
driven our interest to study ATRP activity of a series of Cu(II) complexes. Our
e
investigation was directed to evaluate if the phenyl substitution and/or the presence of
extra functionalities in the ligand backbone were affecting the catalytic properties of the
complexes. This goal has been achieved by
i) synthetizing and studying the Cu(II)
coordination chemistry of a series of ligands having different functionalities, ii) studying
their electrochemistry in solution and iii) performing eATRP in organic and aqueous
solvents.
3.1. Coordination Chemistry
Ligands 1a-d, as well as the corresponding copper complexes 2a-d and 3a-d (Scheme 2),
have been prepared according to a recently reported methodology [34]. The method
consists in the synthesis of a bromo derivative via reductive amination of commercially
available 6-bromo-2-pyridinecarboxaldehyde and di(2-picolyl)amine, followed by a
Suzuki coupling with appropriate boronic acid. The general procedure for the synthesis of
10

ACCEPTED MANUSCRIPT
Cu(II) complexes requires the mixing of an equimolar amount of copper salt with the
corresponding ligand 1a-d in acetonitrile. The solution is stirred at room temperature for
10 min and the desired compounds are obtained as crystalline powders or by
recrystallization with diethyl ether.
2a R = H; X = ClO4
= Cl
2d R = CH2OH; X = ClO4
3a R = H; X = Cl
3c R = CONH2 X
Fig. 1. X-ray crystal structures of complexes 2a,d and 3a,c
.
11

ACCEPTED MANUSCRIPT
Table 1
Thermodynamic data for Cu(II) and Cu(I) complexes in CH3CN + 0.1 M (C2H5)4NBF4 at 25 °C.
[CuIIL]2+ a
Eo (V vs. SCE) βII
1.056
[ClCuIIL]+ b
[BrCuIIL]+ c
Eo (V vs. KIIBr/KIBr
SCE)
e
e
Entry L
R
/
βI d
Eo (V vs. SCE) KIICl/KICl
f
1
2
3
4
5
6
-
-
-0.019
0.043
0.071
0.039
0.045
1.49
1.33
4.48
1.56
1.23
×
×
×
×
×
1018 -0.363
1017 -0.197
1016 -0.203
1017 -0.218
1017 -0.200
6.53
1.14
4.28
2.21
1.39
×
×
×
×
×
105
-0.282
-0.130
-0.157
-0.136
-0.131
2.79
8.40
7.15
9.08
9.44
×
×
×
×
×
104
TPMA -
104
104
104
104
102
103
102
102
H
1a
1b
1c
1d
CHO
CONH2
CH2OH
a Obtained by dissolving crystalline complexes in acetonitrile. b Obtained by dissolving crystalline
complexes 2a-d in acetonitrile, followed by the addition of 2 equivalents of (C2H5)4NCl. c Obtained by
dissolving crystalline complexes 2a-d in acetonitrile, followed by the addition of 2 equivalents of
(C2H5)4NBr. dβII and βI are the stability constants of [CuIIL]2+ and [CuIL]+, respectively. eKIIX and KIX are
the association constants of X- with [CuIIL]2+ and [CuIL]+, respectively. fNo added ligand, Eo of CuII(OTf)2
taken from ref. 36.
The obtained complexes have been characterized by ESI-MS, FT-IR spectroscopy and/or
elemental analysis, confirming the purity of the desired compounds (see supporting
information). Among the different structures, it was possible to obtain crystals suitable for
X-ray diffraction of complexes 2a,d and 3a,c. All the monocationic structures (Fig. 1)
show a distinctive Jahn-Teller effect with a strong distorted octahedral coordination for
the Cu(II) centre. The four short bonds of 2a,d and 3a,c, arranged in a square planar
manner, come from the amino group N(1) in the range between 203 and 206 pm, the two
unsubstituted pyridine moieties N(3), N(4) with bond lengths between 197 and 202 pm
and finally from acetonitrile N(5) at 2a,d with equal distances of 198 pm and Cl anions at
12

ACCEPTED MANUSCRIPT
3a,c around 225 pm, respectively. The elongated bonds at complexes 2a and 2d (N(2):
234, O(1): 281 and N(2): 249, O(2): 254 pm) are comparable to well described Cu(II)
complexes of pentadentate bispidine ligands with two amino groups and three pyridine
moieties, and a coordinated acetonitrile [35]. The molecules of 3a,c form a dimer,
whereas the chloride anions generate additional elongated bonds to the other Cu(II) centre
with lengths of 295 at 3c and 328 pm at 3a. The distances of the nitrogen atoms are
elongated to 245 and 250 pm (N(2) and N(7)) at 3c and 252 pm at N(2) of 3a. The
dimerization could be favoured by additive π-π interactions between the pyridine rings
with carbon-carbon distances of 342 (3c) and 354 pm at (3a).
It should be noted that while the reported complexes display octahedral geometries in the
solid state, and in two cases are dinuclear, previous solution studies of related complexes
suggest a monomeric character and C3 trigonal-bypiramidal symmetry [20].
3.2. Cyclic voltammetry of copper complexes
The redox properties of the overall complexes were investigated in acetonitrile by cyclic
voltammetry. Examples of typical voltammetric pattern for binary [CuIIL]2+ and ternary
[XCuIIL]+ complexes are reported in Fig. 2. A reversible peak couple, corresponding to
one electron transfer was observed for each complex.
[CuIIL]2+ + e-
[XCuIIL]+ + e-
[CuIL]+
(1)
(2)
[XCuIL]
The peak-to-peak separation,
Nernstian electron transfer (
∆
Ep, was greater than the typical value for a fast or
∆
Ep = 60 mV) and increased with the scan rate. These results
agree with previous reports on the electrochemical behaviour of copper-amine complexes
13

ACCEPTED MANUSCRIPT
employed as catalysts for ATRP, confirming that electron transfer is accompanied by
significant reorganizations [23]. The standard potential (Eo) of the Cu(II)/Cu(I) couple
was determined as Eo
≈ E1/2 = (Epc + Epa)/2, where Epc and Epa are the cathodic and anodic
peak potentials, respectively. The average of E1/2 values measured with scan rates ranging
from 0.02 V/s to 0.2 V/s is reported in Table 1 as an estimate of Eo for each complex.
As reported in Fig. 2a and Table 1, small differences on Eo values were observed
within the class of compounds formed by Cu(II) and ligands 1a-d.
Eo is weakly affected
by the substituent on the phenyl ring, probably because of the large distance separating it
from the metallic centre. Crystal structures (Fig. 1) support this statement, evidencing that
the biarylic system can rotate, increasing the distance between the substituent and the
metal.
Although the substituent effect is small, Eo of these compounds is 58-90 mV more
positive than the redox potential of [CuIITPMA]2+. It appears that the presence of the
phenyl ring significantly decreases the charge density on the metal centre, making it more
easily reducible. The reactivity of Cu(I) complexes with these ligands is also expected to
be smaller in comparison to [CuITPMA]+.
14

ACCEPTED MANUSCRIPT
Fig. 2. Cyclic voltammetry of 1 mM Cu(II) complexes in CH3CN + 0.1 M (C2H5)4NBF4, recorded on a GC
electrode at a scan rate of 100 mV s-1.
The relative stability of Cu(II) and Cu(I) can be calculated from the standard potentials
of [CuIIL]2+/[CuIL]+ and solvated Cu2+/Cu+ [36]:
II
ꢁ
ln ꢀꢀI
=
ꢂꢃ ꢄꢅCou2+
− ꢅo
Cu+
ꢆꢇ
[CuIIL]2+ [CuIL
/ ]
(3)
/
where βI and βII are the stability constants of [CuIL]+ and [CuIIL]2+. Calculated βII/βI
values are reported in Table 1. As previously found for similar copper-amine complexes
[36], [CuIIL]2+ is much more stable than [CuIL]+. It is however important to stress that
βII/βI decreased by about one order of magnitude when ligands 1a-d replaced TPMA. As
15

ACCEPTED MANUSCRIPT
a consequence of enhanced Cu(I) stabilization, a reduced reactivity of [CuIL]+ (L = 1a-d
towards RX activation is anticipated, compared to [CuITPMA]+.
)
When halide ions were added to a solution of [CuIIL]2+, ternary complexes [XCuIIL]+
were readily formed. Cyclic voltammetry of these compounds (Fig. 2b) showed a
negative shift of the Cu(II)/Cu(I) peak couple (Eq. 2). Eo values calculated from the peak
potentials as previously described are reported in Table 1. The trend of Eo values is quite
similar to that observed for the binary complexes: all [ClCuIIL]+ complexes have similar
Eo values that are 145-163 mV more positive than Eo of [ClCuIITPMA]+. Analogously, Eo
values of [BrCuIIL]+ are 125-152 mV more positive than Eo of [BrCuIITPMA]+.
Nonetheless, the standard potentials of [XCuIIL]+ are more negative (by 0.18-0.34 V) than
those of the corresponding binary complexes [CuIIL]2+. This negative shift of Eo can be
used to determine the relative affinity of [CuIIL]2+ and [CuIL]+ for halide ions:
II
ꢁ
ln ꢈꢈI
=
ꢄꢅ[oCuIIL]
ꢆ − ꢅ[oXCuIIL]+ [XCuIL]ꢇ
[CuIL
]
(4)
X
2+
/
/
ꢂꢃ
X
I
II
where KX and KX are the equilibrium constants of X- binding to [CuIL]+ and [CuIIL]2+,
respectively. The calculated values of KX
/
II KXI are in the range from 103 to 4 104 (Table
×
1). With all ligands 1a-d, Cu(II) shows much better affinity than Cu(I) for halide ions.
This is a fundamental prerequisite for a good ATRP catalyst: Cu(II) must be
predominantly present in the deactivator form, [XCuIIL]+, whereas Cu(I) is desired to give
a less stable [XCuIL] complex that readily dissociates to regenerate the activator [CuIL]+
[5, 37].
3.3. General considerations on copper-catalysed electrochemically mediated ATRP
16

ACCEPTED MANUSCRIPT
As shown in Table 1 the new complexes have similar Eo values that are all more positive
than Eo of the corresponding TPMA-based complexes. Of particular relevance to ATRP
are the redox properties of the ternary complexes [ClCuIIL]+ and [BrCuIIL]+, since usually
they are the predominant species present in solution under typical ATRP conditions [36].
Eo of these complexes is only slightly affected by the substituent on the phenyl ring with
overall variations of 21 mV for chlorides and 27 mV for bromides. On the other hand, it
has recently been shown that the catalyst activity increases with decreasing standard
reduction potential of [XCuIIL]+ [21, 37]. It is therefore reasonable to assume that the new
complexes have comparable reactivity and are less active than the copper complex with
unmodified TPMA. For this reason, all polymerizations were performed with L = 1a
chosen as a model catalyst for the new series of copper complexes. Electrochemically
mediated atom transfer radical polymerization ( ATRP) of different monomers was tested
,
e
in both organic solvents and water. The experimental conditions and the measured Eo
values of the catalyst in different solvent/monomer mixtures are reported in Table 2. All
e
ATRP experiments were performed under potentiostatic control, setting Eapp to tune the
reactivity of the catalyst according to the particular system.
17

ACCEPTED MANUSCRIPT
Table 2
Potentiostatic eATRP of different monomers catalysed by [XCuIIL]+ (L = 1a or TPMA).a
Entry Solvent Monomerb
L
X-
Eo
RX
Eapp
t
Conv.
Q
Mn
Mn,th Mw/Mn
ꢉꢊꢋꢊꢊ
(V)c
(V)c
(h) (%)
(C)
10-3
10-3
(h-1)d
1
2
3
4
5
6
7f
8
9i
CH3CNe MA
(50%)
MA
Br- -0.030 MBiB
Br- -0.091 MBiB
Br- -0.196 MBiB
Br- -0.134 EBPA
Br- -0.214 EBPA
-0.09
-0.15
-0.26
-0.19
-0.15
6
6
6
0
-
-
-
-
-
-
-
-
-
-
1a
DMFe
DMFe
DMFe
DMFe
H2O
0
1a
(50%)
MA
92
4.43 0.40 51.7 43.8 1.07
8.71 0.12 44.6 38.7 1.21
8.93 0.11 45.9 36.0 1.12
0.77 0.84 90.9 96.6 1.16
TPMA
1a
(50%)
MMA
(50%)
MMA
(50%)
OEOMA
(20%)
OEOMA
(10%)
MAA
14 83
12 77
TPMA
1a
Br- -0.133 HEBiB -0.07
Br- -0.260 HEBiB -0.20
3
4
5
4
88
95
81
96
H2O
3.10
63.0 51.0 1.17
TPMA
1a
H2O
Cl-
Cl-
-0.120h BiBA
-0.120h BiBA
-0.18
-0.18
3.56 0.35 21.4 13.9 1.49
3.20 0.84 18.0 16.8 1.42
(10%)g
MAA
H2O
TPMA
(10%)g
a
General conditions: ꢌꢍꢎꢏꢆ = 10-3 M,
T
= 25 °C, Vtot = 15 mL, CM
/
CRX
/
ꢌꢍꢎꢏꢆ
/
CL/
ꢌꢐ : 555/1/0.1/0.1/0.2
ꢑ
b
(entries 1-3); 467/1/0.1/0.1/0.2 (entries 4-5); 225/1/0.5/0.5/50 (entry 6); 200/1/0.1/0.4/50 (entry 8). Vol%.
d
cvs. SCE. Slope of the ln(ꢌꢒꢓ
/
CM
)
vs. time plot. e0.1 M (C2H5)4NBF4 as supporting electrolyte. fTaken from
g
h
ref. [17]. pH = 0.9 by addition of HCl. Cathodic peak potential, Ep,c (irreversible voltammetric response).
iTaken from ref. [40].
3.4. eATRP in organic solvents
Polymerization of methyl acrylate (MA) in CH3CN was unsuccessful: no polymer
formation was observed after 6 h, even by using Eapp
=
Eo - 60 mV to enhance the amount
of the activator, Cu(I). Indeed, this Eapp value should generate in solution a concentration
of Cu(I) that is 10-times higher than ꢌꢍꢎꢔꢔ, as can be derived from Equation 5, thus
maximizing the extent of the activation process.
18

ACCEPTED MANUSCRIPT
+
II
ꢅꢋꢊꢊ = ꢅꢕ/ꢖ
+
ꢂꢃ ln [XCu L]
(5)
ꢁ
[XCuIL]
A similar result was obtained also in DMF, which usually accelerates ATRP with respect
to CH3CN [38]. In contrast, when TPMA was used as ligand (Table 2, entry 3) very high
conversion was achieved in 6 h; the polymerization was well-controlled with linear first-
order kinetics and very low dispersity (Fig. S2-S3). A possible explanation for the failure
of polymerization with the catalytic system Cu/1a/Br- is the inefficient activation of the
secondary alkyl bromide dormant species (R′CH(Br)C(O)OCH3) by the activator complex
[CuIL]+ (L = 1a). This hypothesis is corroborated by activation rate constants (kact
measured for the initiator methyl 2-bromoisobutyrate (MBiB) by [CuIL]+ in CH3CN/MA
103 mol-1dm3s-1 for L = TPMA and
10-2 mol-1dm3s-1 for L = 1a (Fig. S1). It is noted that kact depends on the structure of
)
(50% v/v). The obtained values were kact= 1.1
×
1.3×
RX, increasing in the order primary < secondary < tertiary [21]. The dormant species
formed during ATRP of MA is a secondary alkyl halide of the same type of MBiB,
therefore its activation rate constant is expected to be at least one order of magnitude
lower than kact of MBiB [21].
When [CuII(1a)]2+ is used as catalyst, its reactivity toward the initiator is already low and
it becomes even lower when propagating chains are converted to dormant species, Pn-X.
The polymerization rate of ATRP, Rp, depends on several factors (Eq. 6) among which an
important role is played by the ratio between the activation and deactivation rate
constants, kact
/
kdeact, that is KATRP:
ꢔ
ꢆ
ꢘ
]
ꢝꢐ ꢘꢍꢎ ꢞ]
•
ꢘ ]ꢘ ]
ꢘ ]
ꢗꢊ = ꢉꢊ M P = ꢉꢊ M ꢙꢚꢛꢜꢝ
(6)
ꢔꢔ
ꢆ
ꢘꢐꢟꢍꢎ ꢞ]
19

ACCEPTED MANUSCRIPT
Both KATRP and kact are generally at least one order of magnitude higher for methacrylates
than for acrylates, mainly because of the higher stability of tertiary radicals arising from
methacrylates than secondary radicals from acrylates. Therefore, a fourth
eATRP
experiment was run in DMF using methyl methacrylate (MMA) in place of MA and ethyl
1-phenyl-1-bromoacetate (EBPA) as initiator. Note that homopolymerization of MMA by
e
ATRP catalysed by Cu complexes has never been reported. So far, only the Fe(II)/Fe(III)
redox couple was used as catalyst [39]. EBPA was selected as initiator because of its
extremely high reactivity and proved suitability for methacrylate polymerizations [8].
This reaction yielded good results (Fig. 3 and Table 2, entry 4): linear kinetic plot and
low dispersity indicated a well-controlled process, reaching 83% conversion in 14 h.
Cu/TPMA also performed quite well for eATRP of MMA but the huge activity of
[CuITPMA]+ required a careful tuning of the applied potential to achieve a controlled
polymerization under analogous conditions (supporting information). When Eo - 60 mV
was used, dispersity grew with time, reaching Mw
/Mn > 2 after only 40% conversion. Poor
control was still observed with Eapp
and linear first-order kinetics were obtained when the CuI to CuII ratio was reduced to ca.
Eo + 60 mV (Table 2, Entry 5). In fact, molecular weights
distribution depends on the amount of CuII deactivator, among several parameters,
is the conversion and DP = ꢌꢒꢓ ꢠꢡꢠꢢꢣꢤ/ꢌꢜꢓꢐ). Relatively
=
Eo, which implies ꢌꢍꢎ ≈ ꢌꢍꢎꢔꢔ, while Mw
/Mn = 1.12
ꢔ
0.1 by applying Eapp
=
according to Equation 7 (where
p
high ꢌꢍꢎ can help improve control over molecular weight distribution (i.e. lower
ꢔꢔ
Mw/Mn), particularly in the case of fast propagation (high kp) or slow deactivation (low
kdeact). kp of MMA at ambient temperature is quite low (ca. 300 M-1 s-1), however,
[CuITPMA]+ is much more active than [CuI1a]+, thus more radicals are generated and the
20

ACCEPTED MANUSCRIPT
instantaneous amount of dormant chains, ꢌꢘꢝ ꢐ], is lower. Therefore, when L = 1a was
ꢥ
used, it was not necessary to raise ꢌꢍꢎꢔꢔ, in contrast to the experiment with TPMA
,
whereby enhanced ꢌꢍꢎ was needed to deactivate radicals and control the process.
ꢔꢔ
ꢪ ꢬ
ꢦ
ꢕ
= 1 + ꢨꢝ + ꢩꢪ
ꢷ ꢄꢸꢖ − 1ꢇ
ꢆ
(7)
ꢫ
ꢘꢭ ꢮ]
ꢥ
ꢧ
ꢦ
ꢬ
ꢔꢔ
ꢘꢮꢴꢵ ꢶ]
ꢥ
ꢯꢰꢱꢲꢳ
When Cu/TPMA was used, low dispersity was achieved at the expense of the
polymerization rate, which decreased by increasing ꢌꢍꢎ (Eq. 6), and the process was
ꢔꢔ
slightly slower than with L = 1a. Moreover, measured MWs exceeded the theoretical
values at high conversions, probably because of termination via radical coupling, as
proved by a small shoulder at high MWs in the GPC traces (Fig. S6). A better accordance
between experimental and theoretical MWs was observed with L = 1a, despite the slightly
higher dispersity, Mw/Mn = 1.2 (Fig. 3).
21

ACCEPTED MANUSCRIPT
Fig. 3.
e
ATRP of MMA 50% v/v in DMF + 0.1 M (C2H5)4NBF4. [MMA]:[EBPA]:[CuII]:[Br-] =
467:1:0.1:0.2, L = 1a, Eapp = T= 25 °C, Vtot = 15 mL. a) First-order kinetic
E1/2 – 0.060 V, ꢹꢺꢻ = 10-3 M,
ꢼꢆ
plot and b) molecular weights and dispersity evolution as function of conversion.
3.5. eATRP in aqueous media
Challenging features of ATRP in aqueous media include: i) very high activation rate
leading to high concentration of radicals, ii) poor deactivation efficiency due to the facile
dissociation of [X-CuIIL]+, iii) high tendency of Cu(I) species towards disproportionation,
(iv) instability of the amine ligands in acidic media.
Recently, these problems have been overcome by using TPMA as ligand, in combination
with the advantages of eATRP [17].
e
ATRP of oligo(ethylene glycol) methyl ether methacrylate (OEOMA) was performed
with [CuII1a]2+ according to the guidelines defined for aqueous ATRP: large X- excess,
Eo [17]. The results are reported in Table 2, entry 6,
e
low monomer percentage, and Eapp
>
and Fig. 4. Polymerization of OEOMA was well-controlled showing linear first-order
kinetics and good agreement between experimental and theoretical molecular weights.
22

ACCEPTED MANUSCRIPT
The reaction reached 88% conversion in 3 h with very low dispersity (Mw/Mn < 1.2). The
result well matches with that obtained using [CuIITPMA]2+ as catalyst (Table 2, entry 7):
95% conversion in 4 h with Mw Mn = 1.17 [17]. Also notice that the amount of monomer
/
used herein is twice as that of the previously reported electrosynthesis with
[CuIITPMA]2+.
Very recently, aqueous eATRP was successfully applied to methacrylic acid [40]. The
harmful reaction preventing ATRP of this acidic monomer was identified as an
intramolecular cyclization process, leading to the formation of a lactone, with the loss of
chain-end functionalities. Low pH, excess of Cl− instead of Br−, and Eapp
<
Eo allowed to
control the polymerization of MAA using TPMA as ligand. 96% conversion in 4 h and
Eo - 120 mV (Table
Mw/Mn = 1.49 were reported at pH = 0.9, [NaCl] = 0.3 M and Eapp =
2
, entry 9). By replacing TPMA with 1a, a comparable result was obtained with 83%
conversion in 5 h and Mw Mn = 1.49 (Table 2, entry 8 and Fig. 4).
Overall, the copper complex with 1a is a suitable catalyst for aqueous
/
e
ATRP, even in
acidic conditions, providing a catalytic efficiency comparable to the one observed with
copper/TPMA in identical conditions.
23

ACCEPTED MANUSCRIPT
Fig. 4. eATRP of OEOMA 20% v/v in H2O + 0.1 M (C2H5)4NBr (circles) and MAA 10% v/v in H2O + 0.1 M NaCl
(+HCl, pH = 0.9) (squares). a) Conversion and first-order kinetic plots and b) molecular weights and dispersity
evolution with conversion. Polymerization conditions: Table 1, entries 6 and 8.
4. Conclusions
In summary, eight novel copper complexes with modified TPMA ligands were
synthesized and fully characterized. Electrochemical studies have shown a small
influence of the substituents in the phenyl ring on the redox properties of the catalyst and
hence on its reactivity, the dominant effect being the electron-withdrawing nature of the
phenyl substituent itself. The small influence of the substituents can be explained
considering that the possible rotation of the biarylic system moves the substituents away
from the metal centre.
Surprisingly, the presence of one phenyl ring on the TPMA backbone reduced
significantly the activity of [CuIL]+, while substituents on that ring had virtually no
effect on catalyst activity. Nevertheless, the novel copper complexes are robust and
suitable catalysts for
eATRP in organic solvents with reactive monomers such as methyl
24

ACCEPTED MANUSCRIPT
methacrylate or in aqueous media. Indeed, PMMA with Mn = 44600 and Mw/Mn = 1.21
was obtained in DMF, while low dispersity POEOMA and poly(methacrylic acid) were
produced in water reaching high conversions in less than 5 h. Although the new catalyts
are less reactive than the parent TPMA complex, their lower catalytic activity may be
exploited to gain easy control of extremely reactive systems, where the activation rate is
not an issue.
Acknowledgements
Financial support from the University of Padova (PRAT CPDA150001 and
CPDA153122, Progetto Attrezzature Scientifiche finalizzate alla Ricerca 2014) is
gratefully acknowledged.
References
[1]
and Future Perspectives, Macromolecules 45 (2012) 4015-4039.
[2] K. Matyjaszewski, N. V. Tsarevsky, Macromolecular Engineering by Atom Transfer
Radical Polymerization, J. Am. Chem. Soc. 136 (2014) 6513-6533.
[3] C. Boyer, N. A. Corrigan, K. Jung, D. Nguyen, T.-K. Nguyen, N. N. M. Adnan, S.
K. Matyjaszewski, Atom Transfer Radical Polymerization (ATRP): Current Status
Oliver, S. Shanmugam, J. Yeow, Copper-Mediated Living Radical Polymerization (Atom
Transfer Radical Polymerization and Copper(0) Mediated Polymerization): From
Fundamentals to Bioapplications, Chem. Rev. 116 (2016) 1803–1949.
[4]
K. Matyjaszewski, J. Xia, Atom Transfer Radical Polymerization, Chem. Rev. 101
(2001) 2921-2990.
25

ACCEPTED MANUSCRIPT
[5]
P. De Paoli, A. A. Isse, N. Bortolamei, A. Gennaro, New insights into the mechanism
of activation of atom transfer radical polymerization by Cu(I) complexes, Chem. Commun.
47 (2011) 3580-3582.
[6]
A. A. Isse, N. Bortolamei, P. De Paoli, A. Gennaro, On the mechanism of activation
of copper-catalyzed atom transfer radical polymerization, Electrochim. Acta 110 (2013)
655−662.
[7]
W. Tang, Y. Kwak, W. Braunecker, N. V. Tsarevsky, M. L. Coote, K.
Matyjaszewski, Understanding Atom Transfer Radical Polymerization: Effect of Ligand and
Initiator Structures on the Equilibrium Constants, J. Am. Chem. Soc. 130 (2008) 10702-
10713.
[8]
W. Jakubowski, K. Matyjaszewski, Activators Regenerated by Electron Transfer for
Atom-Transfer Radical Polymerization of (Meth)acrylates and Related Block Copolymers,
Angew. Chem. Int. Ed. 45 (2006) 4482−4486.
[9]
K. Matyjaszewski, W. Jakubowski, K. Min, W. Tang, J. Huang, W. A. Braunecker,
N. V. Tsarevsky, Diminishing catalyst concentration in atom transfer radical polymerization
with reducing agents, Proc. Natl. Acad. Sci. U. S. A. 103 (2006) 15309−15314.
[10] D. Konkolewicz, A. J. D. Magenau, S. E. Averick, A. Simakova, H. He, K.
Matyjaszewski, ICAR ATRP with ppm Cu Catalyst in Water, Macromolecules 45 (2012)
4461-4468.
[11] D. Konkolewicz, Y. Wang, P. Krys, M. Zhong, A. A. Isse, A. Gennaro, K.
Matyjaszewski, SARA ATRP or SET-LRP. End of controversy?, Polym. Chem. 5 (2014)
4396−4417.
26

ACCEPTED MANUSCRIPT
[12] B. M. Rosen, V. Percec, Single-Electron Transfer and Single-Electron Transfer
Degenerative Chain Transfer Living Radical Polymerization, Chem. Rev. 109 (2009)
5069−5119.
[13] A. Anastasaki, V. Nikolaou, Q. Zhang, J. Burns, S. R. Samanta, C. Waldron, A. J.
Haddleton, R. McHale, D. Fox, V. Percec, P. Wilson, D. M. Haddleton, Copper(II)/Tertiary
Amine Synergy in Photoinduced Living Radical Polymerization: Accelerated Synthesis of
ω-Functional and α,ω-Heterofunctional Poly(acrylates), J. Am. Chem. Soc. 136 (2014)
1141-1149.
[14] X. Pan, C. Fang, M. Fantin, N. Malhotra, W. Y. So, L. A. Peteanu, A. A. Isse, A.
Gennaro, P. Liu, K. Matyjaszewski, Mechanism of Photoinduced Metal-Free Atom Transfer
Radical Polymerization: Experimental and Computational Studies
(2016) 2411-2425.
, J. Am. Chem. Soc. 138
[15] A. J. D. Magenau, N. C. Strandwitz, A. Gennaro, K. Matyjaszewski,
Electrochemically mediated atom transfer radical polymerization, Science 332 (2011)
81−84.
[16] N. Bortolamei, A. A. Isse, A. J. D. Magenau, A. Gennaro, K. Matyjaszewski,
Controlled Aqueous Atom Transfer Radical Polymerization with Electrochemical
Generation of the Active Catalyst, Angew. Chem. Int. Ed. 50 (2011) 11391−11394.
[17] M. Fantin, A. A. Isse, A. Gennaro, K. Matyjaszewski, Understanding the
Fundamentals of Aqueous ATRP and Defining Conditions for Better Control,
Macromolecules 48 (2015) 6862-6875.
27

ACCEPTED MANUSCRIPT
[18] F. Lorandi, M. Fantin, A. A. Isse, A. Gennaro, Electrochemically mediated atom
transfer radical polymerization of n-butyl acrylate on non-platinum cathodes, Polym. Chem.
7 (2016) 5357-5365.
[19] K. Schröder, R. T. Mathers, J. Buback, D. Konkolewicz, A. J. D. Magenau, K.
Matyjaszewski, Substituted Tris(2-pyridylmethyl)amine Ligands for Highly Active ATRP
Catalysts, ACS Macro Lett. 1 (2012) 1037−1040.
[20] A. Kaur, T. G. Ribelli, K. Schröder, K. Matyjaszewski, T. Pintauer, Properties and
ATRP Activity of Copper Complexes with Substituted Tris(2-pyridylmethyl)amine-Based
Ligands, Inorg. Chem. 54 (2015) 1474-1486.
[21] M. Fantin, A. A. Isse, N. Bortolamei, K. Matyjaszewski, A. Gennaro,
Electrochemical approaches to the determination of rate constants for the activation step in
atom transfer radical polymerization, Electrochim. Acta 222 (2016) 393–401.
[22] M. Fantin, P. Chmielarz, Y. Wang, F. Lorandi, A. A. Isse, A. Gennaro, K.
Matyjaszewski, Harnessing the Interaction between Surfactant and Hydrophilic Catalyst To
Control eATRP in Miniemulsion, Macromolecules 50 (2017) 3726-3723.
[23] J. Qiu, K. Matyjaszewski, L. Thouin, C. Amatore, Cyclic voltammetric studies of
copper complexes catalyzing atom transfer radical polymerization, Macromol. Chem. Phys.
201 (2000) 1625-1631.
[24] J. P Mendes, P. V Mendonça, P. Maximiano, C. M. R. Abreu, T. Guliashvili, A. C.
Serra, J. F. J. Coelho, Getting faster: low temperature copper-mediated SARA ATRP of
methacrylates, acrylates, styrene and vinyl chloride in polar media using sulfolane/water
mixtures, RSC Advances 6 (2016) 9598-9603.
28