Synthesis of functionalized iron N-heterocyclic carbene complexes and their potential application as flame behavior modifier in cross linked epoxy resins

Article abstract of DOI:10.1016/j.ica.2021.120273

The design of new flame retardants (FR) that avoid the use of halogen and phosphorus additives is challenging and urgent. Herein we report on the synthesis of bis-amino functionalized N-heterocyclic carbene cyclopentadienone iron complexes aimed at promot

Full text of DOI:10.1016/j.ica.2021.120273

Inorganica Chimica Acta 519 (2021) 120273
Contents lists available at ScienceDirect
Inorganica Chimica Acta
journal homepage: www.elsevier.com/locate/ica
Research paper
Synthesis of functionalized iron N-heterocyclic carbene complexes and their
potential application as flame behavior modifier in cross linked
epoxy resins
,b,*
Andrea Cingolania, Valerio Zanottia, Cristiana Cesaria, Martina Ferria, Laura Mazzocchettia
,
,
,
b
,
,c,*
Tiziana Benellia b, Stefano Merighia, Loris Giorginia c, Rita Mazzonia
a Department of Industrial Chemistry “Toso Montanari”, University of Bologna, Viale Risorgimento 4, 40136 Bologna, Italy
b Interdepartmental Center for Industrial Research on Advanced Applications in Mechanical Engineering and Materials Technology, CIRI-MAM, University of Bologna,
Viale Risorgimento 2, 40136 Bologna, Italy
c Interdepartmental Centre for Industrial Research, Renewable Sources Environment, Sea, Energy (CIRI-FRAME), viale Risorgimento, 4 40136 Bologna, Italy
A R T I C L E I N F O
A B S T R A C T
The design of new flame retardants (FR) that avoid the use of halogen and phosphorus additives is challenging
and urgent. Herein we report on the synthesis of bis-amino functionalized N-heterocyclic carbene cyclo-
pentadienone iron complexes aimed at promoting the production of iron containing epoxy resins. Iron complexes
are successfully employed to obtain high Tg thermosets with as low as 5% hardener content. Moreover the
obtained resins display an impressive charring ability that paves the way to the application of such systems for
material with improved flame behavior.
This work is a contribution for the special issue
dedicated to Maurizio Peruzzini on the occa-
sion of his 65th birthday.
Keywords:
Iron complexes
N-heterocyclic carbene
NHC
Cyclopentadienone
Epoxy resin
Flame behaviour modification
Flame retardant
1. Introduction
have been withdrawn due to toxicity and bioaccumulation issues
respectively [7–10]. Thus the design of new flame retardants that avoid
A wide range of applications make use of epoxy resins (EP) due to
their outstanding properties such as high tensile strength and modulus,
excellent chemical and corrosion resistance, adhesion, and dielectric
properties. While their use is spreading to an increasing number of
fields, it should be also noted that epoxies suffer, as well renowned for
plastic materials, from intrinsic high flammability [1–2] that makes
them inappropriate for application in critical areas where strict rule
about flame behavior are issued. Since flammability is mainly a surface
phenomenon, insertion of an outer layer can delay the flame without
impacting on the substrate intrinsic properties [3–4], Nevertheless a
more common method involves the addition of some specific additives
known as flame retardants (FRs) to the bulk of the plastic [5–6]. This
latter approach is suffering however of all the drawbacks involved with
polymer mixing. In recent years, halogen and phosphorus containing FR
the use of halogen and phosphorus additives appears urgent. As a po-
tential alternative, organometallic complexes such as ferrocene-based
co-monomers have been applied. In this context our group reported on
the synthesis of functionalized ferrocenes as flame retardants [11].
Furthermore it has been demonstrated that ferrocene derivatives can
catalyze the formation of char acting as smoke suppressors [12–14]. This
behavior is likely to be further improved in the presence of aromatics.
Nevertheless, within the field of organometallic complexes only the
ferrocene based materials cited above have been developed for such
flame modification applications. Considering the previously described
advantages and drawbacks, the development of novel organometallic
iron based complexes able to act as resin hardener, thus entering in the
main polymer backbone, still represents a challenge toward the pro-
duction of new flame safer materials.
* Corresponding authors at: Department of Industrial Chemistry “Toso Montanari”, University of Bologna, Viale Risorgimento 4, 40136 Bologna, Italy.
E-mail addresses: laura.mazzocchetti@unibo.it (L. Mazzocchetti), rita.mazzoni@unibo.it (R. Mazzoni).
https://doi.org/10.1016/j.ica.2021.120273
Received 29 November 2020; Received in revised form 18 January 2021; Accepted 20 January 2021
Available online 4 February 2021
0020-1693/© 2021 Elsevier B.V. All rights reserved.

A. Cingolani et al.
Inorganica Chimica Acta 519 (2021) 120273
After getting involved for several years in functionalized ferrocene
chemistry [11,15–19], our group has recently extended the research
interests to cyclopentadienone iron complexes bearing N-heterocyclic
carbene ligands [20–21]. Since these complexes show similarity with
ferrocenes in term of stability and electrochemistry, it follows that, in
principle, they are good candidates to play a role as flame retardants.
With this objective in mind here we describe the design and the
synthesis of bis-amino functionalized N-heterocyclic carbene (NHC)
cyclopentadienone iron(0) complexes and their successful use as epoxy-
based network promoters. Charring ability is also discussed as a
parameter for the flame modification behavior of the hybrid epoxy
resins obtained.
383 m/z [M]+. Anal. Calcd (%) for C19H35BrN4O4: C, 49.24; H, 7.61; N,
12.09. Found: C, 49.09; H, 7.45; N, 11.86. Characterization is compa-
rable with the literature, where 1 was obtained under different condi-
tions [24].
2.3. Synthesis of di-carbonyl(1,3-di(trimethylsilyl)-(4,5-cyclohexyl)
cyclopenta-3,5-dien-2-one)(1,3-BOC-aminopropyl)ilidene)iron (2-BOC)
In a dried 100 mL Schlenk flask, 1,3-(BOC-aminopropyl) imidazo-
lium bromide (1) (1.370 g, 2.69 mmol) and silver oxide (0.662 g, 2.86
mmol), triscarbonyl(1,3-di(trimethylsilyl)-(4,5-cyclohexyl)cyclopenta-
3,5-dien-2-one) iron (0.996 g, 2.38 mmol) and trimethylamine-N-
oxide (0.268 g, 3.57 mmol) were dissolved in CH3CN (40 mL). The re-
action mixture was stirred at room temperature and protected from light
for 3 h, and then solvent was removed under reduced pressure. The solid
was re-dissolved in toluene (60 mL) and left under reflux for 1 h. At the
end of reaction time, solvent was removed under reduced pressure and
the crude product was purified by column chromatography on neutral
alumina using CH2Cl2/EtOAc (5:5) affording the titled complex as a
yellow powder (55%).
2. Experimental Section
2.1. Materials and methods
All reactions were carried out under a nitrogen atmosphere, using
standard Schlenk techniques. Glassware was oven dried before use.
Dichloromethane (CH2Cl2), diethyl ether (Et2O), toluene and acetoni-
trile (CH3CN) were dried and distilled prior to use. Other solvents such
as ethylacetate (EtOAc), methanol (MeOH), CDCl3, Acetone‑d6 (Sigma
Aldrich) were employed without further purification. Reagents HBF4
(ether complex), KOH, NaOH, Na2SO4, trimethylamine-N-oxide and
Ag2O were used as received by Sigma Aldrich or Alfa Aesar. Compounds
1H NMR (399.9 MHz, 298 K, acetone‑d6): δ(ppm): 6.95 (s, 1H, CH),
6.81 (s, 1H, CH), 3.89 (4H, CH2), 2.96 (4H, CH2), 2.25 (m, 4H, CH2),
1.87 (m, 4H, CH2), 1.56 (4H, CH2), 1.38 (s, 18H, CH3, Boc), 0.27 (s, 18H,
CH3, TMS). 13C NMR (150.8 MHz, 298 K, acetone‑d6): δ(ppm): 217.23
–
(CO), 183.91 (Cq, carbene), 168.86 (C O, CpO), 156.21 (CO, Boc), 124.61
–
triscarbonyl-(
η
4-3,4-bis(4-methoxyphenyl)-2,5-diphenylcyclopenta-2,4-
(CH), 122.3 (CH), 104.21 (C3,4, Cq, CpO), 79.3 (Cq, Boc), 72.11 (C2,5, Cq,
CpO), 67.20 (CH2), 48.57 (CH2), 32.00 (CH2), 32.02 (CH2), 28.63 (CH3,
BOC), 24.71 (–CH2), 22.58 (–CH2), 0.20 (CH3, TMS). FT-IR (CH2Cl2): ν(CO)
dienone)iron and triscarbonyl(1,3-di(trimethylsilyl)-(4,5-cyclohexyl)
cyclopenta-3,5-dien-2-one) iron [20], 1-(3-BOC-aminopropyl)bromide
[22] 1-(3-BOC-aminopropyl)-imidazole [23] were synthesized as pre-
viously reported.
1989, 1926, ν(C O) 1709 cmꢀ 1. ESI-MS: 773 m/z [M+H]+. Anal. Calcd
–
–
(%) for C36H60FeN4O7Si2: C, 55.94; H, 7.82; N, 7.25. Found: C, 56.14; H,
Bisphenol A Diglycidyl Ether (DGEBA) with an epoxy equivalent
weight (EEW) of 170 g/equivalent was purchased from Alfa Aesar. The
commercial epoxy precursor ELANTAS Elan-Tron ® EC 157 was kindly
supplied by Elantas Europe srl.
7.96; N, 7.12.
2.4. Synthesis of di-carbonyl(1,3-di(trimethylsilyl)-(4,5-cyclohexyl)
cyclopenta-3,5-dien-2-one)(1,3-aminopropyl)ilidene)iron (2)
The NMR spectra were recorded using Varian Inova 300 (1H, 300.1;
13C, 75.5 MHz), Varian Mercury Plus VX 400 (1H, 399.9; 13C, 100.6
MHz), Varian Inova 600 (1H, 599.7; 13C, 150.8 MHz) spectrometers at
298 K; chemical shifts were referenced internally to residual solvent
peaks. Infrared spectra were recorded at 298 K on a Perkin-Elmer
Spectrum Two spectrophotometer. ESI-MS spectra were recorded on a
Waters Micromass ZQ 4000 with samples dissolved in MeOH. Elemental
analyses were performed on a Thermo-Quest Flash 1112 Series EA
instrument.
In a dried 100 mL Schlenk flask, di-carbonyl(1,3-di(trimethylsilyl)-
(4,5-cyclohexyl)cyclopenta-3,5-dien-2-one)(1,3-BOC-aminopropyl)
ilidene)iron (2-BOC) (1.025 g, 1.33 mmol) was dissolved in Et2O (20
mL), then 8 equivalent of HBF4⋅Et2O (0,724 mL, 5.32 mmol) were
added. The reaction mixture was stirred at room temperature and pro-
tected from light for 1 h. The solid precipitated was washed with Et2O (3
× 20 mL) and then the residual solvent was removed under reduced
pressure. The solid was re-dissolved in CH2Cl2 (40 mL) and washed with
a basic water solution (KOH 2 M, 2 × 20 mL). The organic phase was
dried over Na2SO4 and then, after filtration, the solvent was removed
under vacuum to yield (85%) di-carbonyl(1,3-di(trimethylsilyl)-(4,5-
cyclohexyl)cyclopenta-3,5-dien-2-one)(1,3-aminopropyl)ilidene)iron
(2) as a yellow solid.
The thermal behavior of the reacting mixtures was evaluated by
Differential Scanning Calorimetry (DSC, Q2000 TA Instruments) and the
measurements were carried out under nitrogen flow, heating the sam-
ples at a heating rate of 1 ◦C/min from –50 to 280. After the first heating,
the samples were quench cooled to 0 ◦C and then heated again with a
heating rate of 20 ◦C/min from 0 to 260 ◦C for Tg evaluation. Ther-
mogravimetric (TGA) measurements were carried out using a TA In-
strument SDT Q600 (heating rate 10 ◦C/min) on 10–20 mg samples
under nitrogen atmosphere (100 mL/min gas flow rate) from room
temperature to 600 ◦C.
1H NMR (399.9 MHz, 298 K, acetone‑d6): δ(ppm): 7.00 (s, 1H, CH),
6.85 (s, 1H, CH), 3.99 (4H, CH2), 2.66 (4H, CH2), 2.28 (m, 4H, CH2),
1.84 (m, 4H, CH2), 1.56 (4H, CH2), 0.31 (s, 18H, CH3, TMS).
13C NMR (150.8 MHz, 298 K, acetone‑d6): δ (ppm) 217.30 (CO),
–
184.7 (Ccarb), 177.01 (C O, CpO), 124.50 (CHNHC), 122.34 (CHNHC),
–
104.35 (C3,4, Cp), 71.22 (C2,5, Cp), 49.62 (CH2), 48.72 (CH2), 32.87
(–NCH2-), 31.72 (CH2), 24.63 (–CH2), 22.61 (–CH2), 0.21 (CH3TMS). FT-
IR (CH2Cl2): ν(CO) 2004, 1944. ESI-MS: 573 m/z [M+H]+. Anal. Calcd
(%) for C26H44FeN4O3Si2: C, 54.53; H, 7.74; N, 9.78. Found: C, 55.01; H,
7.56; N, 9.72.
2.2. Synthesis of 1,3-(BOC-aminopropyl) imidazolium bromide (1)
In a dried 100 mL Schlenk flask, 1-(3-BOC-aminopropyl)bromide
(1.60 g, 6.72 mmol) 1-(3-BOC-aminopropyl)-imidazole (1,17 g, 5.19
mmol) were heated at 110 ◦C for 24 h. At the end of the reaction the
crude is dissolved in EtOAc (30 mL). Then the product was extracted
washing twice with water (2x15 mL). The solvent was removed from the
aqueous phase under vacuum to yield the product 1,3-(BOC-amino-
propyl) imidazolium bromide (1), yield 70%.
2.5. Synthesis of dicarbonyl-(
η
4-3,4-bis(4-methoxyphenyl)-2,5-
diphenylcyclopenta-2,4-dienone)[1,3-BOC-aminopropyl)-ilidene]iron (3-
BOC)
1H NMR (DMSO‑d6): δ(ppm): 9.18 (s, 1H, CH), 7.78 (dd, 1H, CH),
6.97 (dd, 1H, CH), 4.15 (t, 4H, CH2), 2.93 (t, 4H, CH2), 1.90 (quint, 4H,
In a dried 100 mL Schlenk flask, 1,3-(BOC-aminopropyl) imidazolium
bromide (1) (0.500 g, 1.08 mmol), silver oxide (0.300 g, 1.29 mmol),
CH2), 1.37 (s, 18H, CH3, Boc). FT-IR (CH2Cl2): ν(C O) 1705 cmꢀ 1. ESI-MS:
triscarbonyl-(
η
4-3,4-bis(4-methoxyphenyl)-2,5-diphenylcyclopenta-2,4-
–
–
2

A. Cingolani et al.
Inorganica Chimica Acta 519 (2021) 120273
dienone)iron (0.631 g, 1.08 mmol) and trimethylamine-N-oxide (0.122 g,
1.62 mmol) were dissolved in CH3CN (40 mL). The reaction mixture was
stirred at room temperature and protected from light for 3 h, and then
solvent was removed under reduced pressure. The solid was re-dissolved
in toluene (60 mL) and left under reflux for 1 h. At the end of reaction
time, solvent was removed under reduced pressure and the crude product
was purified by column chromatography on neutral alumina using
EtOAc/MeOH (9:1) and affording the titled complex 3-BOC as a yellow
powder (74%).
Table 1
Compositions of the different DGEBA/2, EC157/2, DGEBA/3, and EC157/3
investigated thermoset formulations.
Sample
Epoxy
Hardener
Hardener/epoxy
precursor
Iron (wt/
wt%)
precursor
DGEBA/2-5
DGEBA/2-
7.5
DGEBA
DGEBA
2
2
5
0.98
1.47
7.5
DGEBA/2-
10
DGEBA
2
2
2
2
3
10
5
1.96
0.98
1.47
1.96
0.38
1H NMR (acetone‑d6): δ(ppm): 7.82 (d, 4H, CHaryl), 7.35 (s, 2H, CH,
NHC), 7.15 (m, 10H, CHaryl), 6.75 (d, 4H, CHaryl), 4.20 (t, 2H, CH2), 3.72
(s, 6H, CH3O), 3.70 (t, 4H, CH2), 3.32 (t, 4H, CH2), 1.72 (quint, 4H,
CH2), 1.39 (s, 18H, CH3, Boc). 13C NMR (150.8 MHz, acetone‑d6) δ(ppm):
EC157/2-5
Elan-tron ® EC
157
EC157/2-
7.5
Elan-tron ® EC
157
7.5
10
2.5
EC157/2-
10
Elan-tron ® EC
157
–
218.01 (CO), 181.75 (Cq, carbene), 166.18 (C O, CpO), 159.88 (COCH3),
–
156.44 (CO, Boc), 136.10–123.85 (CHNHC, CHaryl and Caryl), 113.83
(CHaryl), 100.41 (C2,5, Cq, CpO), 80.51 (C3,4, Cq, CpO), 78.51 (Cq, Boc),
67.16 (CH2), 55.38 (OCH3), 48.77 (CH2), 31.86 (CH2), 28.65 (CH3, Boc).
DGEBA/3-
2.5
DGEBA
DGEBA/3-5
DGEBA/3-
7.5
DGEBA
DGEBA
3
3
5
0.76
1.14
7.5
FT-IR (CH2Cl2): ν(CO) 1988, 1933, ν(C O) 1706 cmꢀ 1. ESI-MS: 939 m/z
–
–
[M+H]+, 961 m/z [M+Na]+. Anal. Calcd (%) for C52H58FeN4O9: C,
EC157/3-
2.5
Elan-tron ® EC
157
3
3
3
2.5
5
0.38
0.76
1.14
66.52; H, 6.23; N, 5.97. Found: C, 66.34; H, 6.45; N, 5.86.
EC157/3-5
Elan-tron ® EC
157
2.6. Synthesis of dicarbonyl-(
η
4-3,4-bis(4-methoxyphenyl)-2,5-
EC157/3-
7.5
Elan-tron ® EC
157
7.5
diphenylcyclopenta-2,4-dienone)[1,3-aminopropyl)-ilidene]iron (3)
In a dried 100 mL Schlenk flask, dicarbonyl-(
η
4-3,4-bis(4-methox-
cyclopentadienone as ligands in stable iron(0) complexes [20–21]. This
latter pathway is here exploited in order to prepare bis-amino func-
tionalized N-heterocyclic carbene iron complexes 2 and 3 designed to
work as co-monomer and hardener in the synthesis of epoxy resins.
yphenyl)-2,5-diphenylcyclopenta-2,4-dienone)[1,3-BOC-aminopropyl)-
ilidene]iron (3-BOC) (0.500 g, 0.53 mmol) was dissolved in Et2O (20
mL), then 8 equivalent of HBF4⋅Et2O (0,577 mL, 4,24 mmol). The re-
action mixture was stirred at room temperature and protected from light
for 1 h. The solid precipitated was washed with Et2O (3 × 20 mL) and
then the residual solvent was removed under reduced pressure. The solid
was re-dissolved in CH2Cl2 (40 mL) and washed with a basic water so-
lution (KOH 2 M, 2 × 20 mL). The organic phase was dried over Na2SO4
and then, after filtration the solvent was removed under vacuum to yield
3.1. Synthesis of bis-amino functionalized iron complexes 2 and 3.
At a first stage the BOC protected bis-amino functionalized imida-
zolium salts 1,3-(BOC-aminopropyl) imidazolium bromide (1) was
prepared by reacting 1-(3-BOC-aminopropyl)bromide and 1-(3-BOC-
aminopropyl)-imidazole and heating at 110 ◦C for 24 h in solvent free
conditions (70% yield). NMR and ESI-MS characterization of 1 were
superimposable with those reported for the same salt obtained with a
different synthetic strategy [24]. The bis-amino functionalized iron
complexes 2 and 3 has been then prepared under the conditions
described in Scheme 1.
(88%)
dicarbonyl-(
η
4-3,4-bis(4-methoxyphenyl)-2,5-diphenylcyclo-
penta-2,4-dienone)[1,3-aminopropyl)-ilidene]iron (3) as a yellow solid.
1H NMR (acetone‑d6): δ(ppm): 7.82 (d, 4H, CHaryl), 7.32 (s, 2H, CH,
NHC), 7.12 (m, 10H, CHaryl), 6.75 (d, 4H, CHaryl), 3.73 (s, 6H, CH3O),
3.70 (t, 4H, CH2), 2.74 (t, 4H, CH2), 1.82 (quint, 4H, CH2). 13C NMR
(150.8 MHz, acetone‑d6) δ(ppm): 218.17 (CO), 181.51 (Cq, carbene),
–
166.46 (C O, CpO), 159.83 (COCH3), 136.20–123.99 (CHaryl and Caryl),
–
113.78 (CHaryl), 100.38 (C2,5, Cq, CpO), 80.30 (C3,4, Cq, CpO), 55.38
(OCH3), 49.46 (CH2), 48.68 (CH2), 33.23 (CH2). FT-IR (CH2Cl2): ν(CO)
1995, 1938 cmꢀ 1. ESI-MS: 739 m/z [M+H]+. Anal. Calcd (%) for
The one pot reaction between the two cyclopentadienone iron pre-
cursors and 1 leads to the BOC protected iron complexes 2-BOC (Y =
55%) and 3-BOC (Y = 74%), while the target bis-amino functionalized
complexes 2 (Y = 85%) and 3 (Y = 88%) were obtained by deprotection
of 2-BOC and 3-BOC with HBF4, followed by neutralization of unreacted
acid. Deprotection was monitored with IR spectroscopy by the disap-
C
42H42FeN4O5: C, 68.29; H, 5.73; N, 7.59. Found: C, 68.09; H, 6.01; N,
7.54.
pearance of the C O (BOC) stretching at 1709 cmꢀ 1 (2-BOC) and 1706
–
–
2.7. Formulations preparation
cmꢀ 1 (3-BOC).
All the synthesized compounds were characterized by means of FT-
IR, 1H and 13C NMR spectroscopies, ESI-MS spectrometry and
elemental analyses (see Experimental Section). In general NMR chemi-
cal shifts are in agreement with those reported for congener complexes
[20–21]. In particular, complexes appear as yellow powders that are
stable towards air and moisture while kept in the dark, both in the solid
state and in most common organic solvents. 13C NMR spectra show the
diagnostic signal for the Fe-Ccarbene at 181–185 ppm, while the terminal
CO stretching peaks around 1990–2000 and 1940–1930 cmꢀ 1 can be
observed in the FT-IR spectra in CH2Cl2 solutions as expected [20]. The
structures were further confirmed by ESI-MS analyses: 2-BOC 773 m/z
[M+H+], 2 573 m/z [M+H+], 3-BOC 939 m/z [M+H+], 3 739 m/z
[M+H]+
The epoxy precursor (DGEBA or Elan-Tron ® EC 157) and the bis-
amino functionalized iron complexes (2 or 3) were mixed in a vial at
40 or 50 ◦C (DGEBA and Elan-Tron ® EC 157, respectively) until com-
plete homogenization. Then the mixtures were cooled to RT and used for
the preparation of DSC samples. Different mixtures were produced and
labelled according to the list reported in Table 1.
3. Results and discussion
In order to design curing agents for epoxy resins a bis- or multi-amino
ligand is required. In this frame, N-heterocyclic carbenes (NHCs) are
good candidates for functionalization [25–35] because they are easy to
prepare and tolerant, once coordinated to the metal, to functional
groups [36–40]. Another class of ligands that can provide stability to the
corresponding complexes are variously functionalized cyclo-
pentadienones [41–50]. As above cited, our group has recently devel-
oped a straightforward synthetic procedure that combine NHC and
3.2. Epoxy resin preparation
Commercially available epoxy hardeners are represented, for
3

A. Cingolani et al.
Inorganica Chimica Acta 519 (2021) 120273
Scheme 1. Synthesis of bis-amino NHC iron complexes bearing different cyclopentadienone ligands 2 and 3.
example, by 4,4′-methylene biscyclohexanamine (PACM) and diethyl
toluene diamine (EPIKURE W) that are commonly used to obtain resins
with high thermomechanical properties. By comparison, complexes 2
and 3 are by far more encumbered. In this frame, the investigation of the
ability of complex 3, i.e. the complex with the higher steric hindrance
due to presence of a highly aromatic ligand on the metal, to react with
di-epoxy functionalized moieties, is of particular interest in order to
verify the suitability in the cross linking process for epoxy resin for-
mation. Furthermore, as reported in literature, epoxy monomers and
curing agents with higher aromaticity favor higher char yields [51]
which, in turn, should result in better flame retardancy [52–53].
Moreover, nitrogen containing FR were found to be competitive as
halogen-free flame retardants favoring inert N2 evolution while burning
[5–6]. Finally, the actual incorporation of organometallic units into the
polymer backbone would also avoid all the mixing issues of FRs additive
with polymeric materials, allowing an intrinsic and time-stable
formulation.
A preliminary screening of the reactivity of 3 was thus carried out in
the presence of different amounts diglycidyl ether of bisphenol A
(DGEBA, 5, 7.5 and 10%wt) labelled as DGEBA/3–2.5, DGEBA/3–5 and
DGEBA/3–7.5. With the aim of establishing the occurrence of a curing
process leading to a three-dimensional insoluble network, the analytical
techniques available for the task are limited. Since in the present case
the main purpose of the study is the preliminary assessment of the
crosslinking occurrence and not a full and conclusive scheme of the
reaction mechanism, a simple and reliable tool is required. Hence DSC
measurements were carried out, since calorimetric investigation is able
to provide some crucial information in a single test run for evaluating
the reliability of the hardening system. Indeed, since epoxy opening
reactions are highly exothermic owing to the strain release by the
oxirane cycle’s opening [54], a DSC run can directly measure the overall
rate of exothermic polymerization, as well as the temperature range
required to trigger the process. Moreover, upon re-heating the same
sample, it is also possible to obtain information on the glass transition
temperature reached by the thermoset, a parameter that can be taken as
an index for significant network formation. All the formulations were
thus analyzed by DSC (Fig. 1A) revealing a high-T exotherm accounting
for cross-linking; in the II heating scan (Fig. 1B) a glass transition (Tg) in
the range 135–150 ◦C is observed, as a symptom of resin formation
(Table 2). It is worth noting that, by increasing the amount of hardener,
the crosslinking reaction occurs at lower temperature, while the total
cross-linking enthalpy (ΔH) increases (Table 2).
Fig. 1. DSC thermograms in first (A) and second (B) heating scan of DGEBA/
3–2.5 (
) DGEBA/3–5 (
) and DGEBA/3–7.5 (
).
Table 2
Data obtained by DSC isothermal runs applied at different DGEBA/3
formulations.
Sample
ΔHa (J/g)
Tr max (◦C)
Tg (◦C)
b
DGEBA/3-2.5
DGEBA/3-5
420
437
460
214
200
197
137
144
149
DGEBA/3-7.5
a
ΔH is the heat generated by the cross-linking reaction.
b
Tr max is the temperature corresponding to the maximum rate of the cross-
linking reaction.
with respect to the plain unreacted epoxy precursor, a significant char
production in inert atmosphere (in the range of 20–30% of the initial
weight), that cannot be attributed to the DGEBA degradation pattern
neither solely to the iron-related non-volatile residue. Such a behavior
can be instead ascribed to the presence of the iron-based hardener
An analogous assessment carried out in the presence of 5, 7.5 and
10% of 2 was able to positively cross-link DGEBA though reaching
slightly lower Tg (104 ◦C, 105 ◦C and 110 ◦C respectively) [55]. TGA
curves (Fig. 2) show, in addition to a high thermal stability of the resins
4

A. Cingolani et al.
Inorganica Chimica Acta 519 (2021) 120273
were lower than DGEBA systems (Fig. 3, Table 3), nonetheless con-
firming the thermoset formation.
It is worth noting that by enhancing the amount of hardener, the
reaction enthalpy increases and is also associated to a shift of the exo-
therm peak to lower temperatures (Table 3).
The resins formation was confirmed also by TGA analysis (Fig. 4)
showing a higher thermal stability with respect to the plain Elan-tron ®
EC 157 precursor. The degradation pattern appears similar to those re-
ported for DGEBA/3 systems in Fig. 2, with a residual char content
around 25%wt, and a final solid residue after oxidation that well com-
pares with the expected iron feed.
As reported above, the high char formation confirms the catalytic
effect of iron. It is also worth to point out that the char content of EC157/
3 systems at 600 ◦C is similar to that one of EC157/2 formulations with
Fig. 2. TGA thermograms in nitrogen atmosphere of plain DGEBA (—),
DGEBA/3–2.5 (
), DGEBA/3–5 (
) and DGEBA/3–7.5 (
).
which, as reported in literature [12–14] can catalyze the char formation.
It is also worth to notice that the higher the iron content the higher the
char produced upon resin degradation.
Table 3
Data obtained by DSC isothermal runs applied at different EC157/3
formulations.
Sample
Given these preliminary results, hardener 3 was thus applied to
cross-link a commercial epoxy precursor, Elan-tron® EC 157, intended
for infusion applications. Depending on the hardener system used,
technical datasheet reports the EC 157 epoxy precursor to be able to
reach Tg as high as 80–130 ◦C. In the present case formulations with 2.5,
5, and 7.5%wt of iron-based hardener 3, namely EC157/3–2.5, EC157/
3–5 and EC157/3-0.7.5, were studied. Once again DSC analyses reveal
exotherms at high temperature, while the Tg attained (around 90 ◦C)
b
ΔHa (J/g)
Tr max
Tg (◦C)
EC157/3-2.5
EC157/3-5
420
441
455
198
194
194
90
93
94
EC157/3-7.5
a
ΔH is the heat generated by the cross-linking reaction.
b
Tr max is the temperature corresponding to the maximum rate of the cross-
linking reaction.
Fig. 3. DSC thermograms in first (A) and second (B) heating scan for EC157/3–2.5 (
), EC157/3–5 (
) and EC157/3–7.5 (
).
5

A. Cingolani et al.
Inorganica Chimica Acta 519 (2021) 120273
Fig. 4. TGA thermograms of plain Elan-tron® EC 157 (—), EC157/3–2.5 (
), EC157/3–5 (
) and EC157/3–7.5 (
).
[4] L. Mazzocchetti, T. Benelli, E. Maccaferri, S. Merighi, J. Belcari, A. Zucchelli,
L. Giorgini, Compos.: Part B 145 (2018) 252–260.
the same hardener content [55]. Anyway, considering the higher mo-
lecular weight of 3 with respect of 2, EC157/3 formulations results in an
iron content lower than EC157/2 systems, as shown in Table 1 (See
Experimental Section). Hence, the presence of aromatic moieties seems
to promote the charring ability of the systems which, in turn, should
result in better flame retardancy.
[5] G. Zattini, S. Ballardini, T. Benelli, L. Mazzocchetti, L. Giorgini, Polym. Eng. Sci. 59
(2019) 2488–2497.
[6] T. Benelli, L. Mazzocchetti, E. D’Angelo, M. Lanzi, F. Saraga, L. Sambri, M. Comes
Franchini, L. Giorgini, Polym. Eng. Sci. 57 (2017) 621–630.
[7] S. Molyneux, A.A. Stec, T.R. Hull, Polym. Degrad. Stab. 106 (2014) 36–46.
[8] G.E. Zaikov, S.M. Lomakin, J. Appl. Polym. Sci. 86 (2002) 2449–2462.
[9] E.M. Cooper, A. Covaci, A.L.N. Van Nuijs, T.F. Webster, H.M. Stapleton, Anal.
Bioanal. Chem. 401 (2011) 2123–2132.
4. Conclusions
[10] C.C. Carignan, M.D. McClean, E.M. Cooper, D.J. Watkins, et al., Environ. Int. 55
(2013) 56–61.
[11] a) L. Busetto, M. C. Cassani, R. Mazzoni, Patent (2002) EP 1258487A1; b) L.
Busetto, M.C. Cassani, R. Mazzoni, V.G. Albano, P., Sabatino, P. Frediani, E.
Rivalta, Organometallics, 21 (2002), 4993-4999.
Properly designed bis-amino functionalized iron complexes 2 and 3
were prepared and employed as suitable curing agents for the prepara-
tion of novel hybrid epoxy resins. Cross linking properties have been
demonstrated with both DGEBA and commercial pre-polymer Elan-
tron® EC 157. Different formulations were tested and up to 10%wt of
iron-based hardener with an iron content up to about 2%wt, could be
achieve. All the formulations proved to be able to cross-link epoxy
precursors with good results. Indeed fairly high glass transition tem-
peratures were obtained, confirming that 2 and 3 act as efficient cross-
linkers. Finally, the presence of iron lead to high charring ability which
should result in good flame retardancy. In particular the presence of
aromatic moieties in the functionalized iron complex 3 seems to further
promote the char formation which can act as smoke suppressor and fire
inhibitor via intumescent later formation, thus paving the way to the
application of these materials as potential flame behaviour modifiers.
[12] D.J. Liao, Q.K. Xu, R.W. McCabe, H.V. Babu, X.-P. Hu, N. Pan, D.-Y. Wang,
T. Richard Hull, Ind. Eng. Chem. Res. 56 (2017), 12630–12643 and references
cited therein.
[13] K. Kishore, P. Kannan, K. Iyanar, J. Polym. Sci., Part A: Polym. Chem. 29 (1991)
1039.
[14] N. Najafi-Mohajeri, G.L. Nelson, R. Benrashid, J. Appl. Polym. Sci. 76 (2000)
1847–1856.
[15] E. Scavetta, R. Mazzoni, F. Mariani, R.G. Margutta, A. Bonfiglio, M. Demelas,
S. Fiorilli, M. Marzocchi, B. Fraboni, J. Mater. Chem. B 2 (2014) 2861–2867.
[16] R. Mazzoni, M. Salmi, S. Zacchini, L. Busetto, V. Zanotti, J. Organomet. Chem. 751
(2014) 336–342.
[17] R. Mazzoni, M. Salmi, S. Zacchini, V. Zanotti, Eur. J. Inorg. Chem. 21 (2013)
3710–3718.
[18] L. Busetto, R. Mazzoni, M. Salmi, S. Zacchini, V. Zanotti, Organometallics 31
(2012) 2667–2674.
[19] L. Busetto, F. Marchetti, R. Mazzoni, M. Salmi, S. Zacchini, V. Zanotti,
Organometallics 28 (2009) 3465–3472.
[20] A. Cingolani, C. Cesari, S. Zacchini, V. Zanotti, M.C. Cassani, R. Mazzoni, Dalton
Trans. 44 (2015) 19063–19067.
[21] A. Cingolani, V. Zanotti, S. Zacchini, M. Massi, P.V. Simpson, N. Maheshkumar
Desai, I. Casari, M. Falasca, L. Rigamonti, R. Mazzoni, Appl. Organomet. Chem. 33
(2019), e4779.
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
[22] G.F. Sigler, R. Riaz, Patent WO2000078763, 2000.
[23] I. Lohse, E. Wildermuth, S.P. Brothers, Oncotarget 9 (2018) 35448–35457.
[24] P. Li, Q. Zhao, J.L. Anderson, S. Varanasi, M.R. Coleman, J. Polym. Sci.: Part A
Polym. Chem. 48 (2010) 4036–4046.
[25] W.A. Herrmann, Angew. Chem., Int. Ed. 41 (2002) 1290–1309.
́
[26] S. Díez-Gonzalez, N. Marion, S.P. Nolan, Chem. Rev. 109 (2009) 3612–3676.
Acknowledgments
[27] O. Kühl, Coord. Chem. Rev. 253 (2009) 2481–2492.
[28] L.A. Schaper, S.J. Hock, W.A. Herrmann, F.E. Kühn, Angew. Chem., Int. Ed. 52
(2013) 270–289.
Authors wish to thank Emeritus Professor Luigi Busetto for the
fruitful discussion. Authors wish also to acknowledge the University of
Bologna and Consortium INSTM for financial support.
[29] M.N. Hopkinson, C. Richter, M. Schedler, F. Glorius, Nature 510 (2014) 485–496.
[30] L. Mercs, M. Albrecht, Chem. Soc. Rev. 39 (2010) 1903–1912.
[31] S.P. Nolan, N-Heterocyclic Carbenes: Effective Tools for Organometallic Synthesis,
Wiley VCH, 2014.
[32] E. Peris, Chem. Rev. 118 (2018) 9988–10031.
References
[33] H.V. Huynh, Chem. Rev. 118 (2018) 9457–9492.
[34] C.A. Smith, M.R. Narouz, P.A. Lummis, I. Singh, A. Nazemi, C.-H. Li, C.M. Crudden,
Chem. Rev. 119 (2019) 4986–5056.
[1] S. Qiu, X. Wang, B. Yu, X. Feng, X. Mu, R.K. Yuen, Y. Hu, J. Hazard. Mater. 325
(2017) 327–339.
[35] Q. Liang, D. Song, Chem. Soc. Rev. 49 (2020) 1209–1232.
[36] C. Cesari, R. Mazzoni, E. Matteucci, A. Baschieri, L. Sambri, M. Mella, A. Tagliabue,
F.L. Basile, Carlo Lucarelli, Organometallics 38 (2019) 1041–1051.
[2] N. Saba, M. Jawaid, M.T. Paridah, O.Y. Al-othman, Polym. Adv. Technol. 27 (2016)
577–590.
[3] S. Merighi, L. Mazzocchetti, T. Benelli, E. Maccaferri, A. Zucchelli, A. D’Amore,
L. Giorgini, Polym. Eng. Sci. 59 (2019) 2541–2549.
6

A. Cingolani et al.
Inorganica Chimica Acta 519 (2021) 120273
[37] C. Cesari, R. Conti, A. Cingolani, V. Zanotti, M.C. Cassani, L. Rigamonti,
R. Mazzoni, Catalysts 10 (2020) 264–275.
[45] R. Mazzoni, C. Cesari, V. Zanotti, C. Lucarelli, T. Tabanelli, F. Puzzo, F. Passarini,
̀
E. Neri, G. Marani, R. Prati, F. Vigano, A. Conversano, F. Cavani, ACS Sustainable
[38] C. Cesari, R. Mazzoni, H. Müller-Bunz, M. Albrecht, J. Organomet. Chem. 739
(2015) 256–262.
Chem. Eng. 7 (2019) 224–237.
[46] C. Cesari, A. Cingolani, M. Teti, A. Messori, S. Zacchini, V. Zanotti, R. Mazzoni, Eur.
J. Inorg. Chem. (2020) 1114–1122.
[39] C. Cesari, A. Cingolani, C. Parise, S. Zacchini, V. Zanotti, M.C. Cassani, R. Mazzoni,
RSC Adv. 5 (2015) 94707–94718.
[47] L. Busetto, D. Fabbri, R. Mazzoni, M. Salmi, C. Torri, V. Zanotti, Fuel 90 (2011)
1197–1207.
[40] C. Cesari, S. Conti, S. Zacchini, V. Zanotti, M.C. Cassani, R. Mazzoni, Dalton Trans.
43 (2014) 17240–17243.
[48] T.C. Johnson, G.J. Clarkson, M. Wills, Organometallics 30 (2011) 1859–1868.
[49] A. Quintard, J. Rodriguez, Angew. Chem. Int. Ed. 53 (2014) 4044–4055.
[50] L. Pignataro, C. Gennari, Eur. J. Org. Chem. (2020) 3192–3205.
[51] X. Wang, W. Guo, L. Song, Y. Hu, Compos. Part B Eng. 179 (2019), 107487.
[52] D.W. van Krevelen 16 (1975) 615–620.
¨
[41] M. C. Warner, C. P. Casey, J.-E. Backvall, Topics in Organometallic Chemistry; T.
Ikariya, M. Shibasaki, Eds.; Springer: Berlin/Heidelberg, Germany, 2011; vol. 37,
pp. 85–125. ISBN 978-3-642-20730-3.
[42] B.L. Conley, M.K. Pennington-Boggio, E. Boz, T.J. Williams, Chem. Rev. 110 (2010)
2294–2312.
[53] J. Dai, Y. Peng, N. Teng, Y. Liu, C. Liu, X. Shen, S. Mahmud, J. Zhu, X. Liu, ACS
Sustain. Chem. Eng. 6 (2018) 7589–7599.
[43] T. Pasini, G. Solinas, V. Zanotti, S. Albonetti, F. Cavani, A. Vaccari, A. Mazzanti,
S. Ranieri, R. Mazzoni, Dalton Trans. 43 (2014) 10224–10234.
[44] C. Cesari, L. Sambri, S. Zacchini, V. Zanotti, R. Mazzoni, Organometallics 33 (2014)
2814–2819.
[54] S. Swier, B. Van Mele, J. Polym. Sci. B Polym. Phys. 41 (2003) 594–608.
[55] A. Cingolani, R. Mazzoni, V. Zanotti, M. Sanviti, L. Mazzocchetti, T. Benelli,
L. Giorgini, AIP Conf. Proc. 2196 (2019), 020035.
7