Vanillin-derived epoxy monomer for synthesis of bio-based epoxy thermosets: effect of functionality on thermal, mechanical, chemical and structural properties

Article abstract of DOI:10.1007/s11696-020-01164-8

Abstract: Tri-functional vanillin-derived epoxy monomer was developed through the synthesized di-functional reagent and cured with a series of different types of hardeners (hydroxyl and amine based) to evaluate thermo-mechanical properties of the resultan

Full text of DOI:10.1007/s11696-020-01164-8

Chemical Papers
https://doi.org/10.1007/s11696-020-01164-8
ORIGINAL PAPER
Vanillin‑derived epoxy monomer for synthesis of bio‑based epoxy
thermosets: eꢀect of functionality on thermal, mechanical, chemical
and structural properties
Mohsen Mogheiseh1 · Ramin Karimian1 · Mostafa Khoshsefat2
Received: 28 February 2020 / Accepted: 13 April 2020
©
Institute of Chemistry, Slovak Academy of Sciences 2020
Abstract
Tri-functional vanillin-derived epoxy monomer was developed through the synthesized di-functional reagent and cured
with a series of diꢀerent types of hardeners (hydroxyl and amine based) to evaluate thermo-mechanical properties of the
resultant epoxy thermosets. Through elevating of curing degree by the hardeners, both alpha transition temperature (T ) and
α
−
3
cross-linking density (υ) slightly increased (up to 118 °C and 69.9 mol m ). However, the highest values of storage and loss
modulus (E′ and E″, respectively) were obtained using trimethylolpropane (TMP) in epoxy thermoset systems cooperating
with tri-functional epoxy monomer at 3200 MPa and 359.3 MPa, respectively. The sample produced by TMP exhibited the
highest amount of statistic heat-resistant index (T ) (up to 193 °C). Nevertheless, the longest half-life and the highest T are
s
d
obtained for the sample prepared by incorporation of tris(hydroxymethyl)aminomethane (TRIS) and 1,10-diaminodecane
(
1347 s, 431 °C, respectively) containing amine agents as curing agent. Remarkably, the sample comprising triethanol amine
showed the highest char yield (3.3 mg at 700 °C) among the all samples. However, the sample cured by TRIS exhibited
the highest elasticity, elongation and tensile strength among the obtained samples (up to 0.48 GPa, 6.56% and 22.90 MPa,
respectively).
Electronic supplementary material The online version of this
article (https://doi.org/10.1007/s11696-020-01164-8) contains
supplementary material, which is available to authorized users.
*
Ramin Karimian
karimian.r@gmail.com
1
Chemical Injuries Research Center, Systems Biology
and Poisonings Institute, Baqiyatallah University of Medical
Sciences, Tehran, Iran
2
Key Laboratory of Engineering Plastics, Beijing National
Laboratory for Molecular Sciences, Institute of Chemistry,
Chinese Academy of Sciences, 100080, Beijing, China
Vol.:(0
1
123456789)
3

Chemical Papers
Graphic abstract
Keywords Epoxy thermoset · Bio based · Thermo-mechanical properties · Thermal stability · Vanillin · Oximation
Introduction
as development of bio-based polymer based on vegetable
oil and cyclo-aliphatic compounds (polysaccharide) derived
from cellulose, starch or triglycerides. However, the absence
of aromatic rings makes these moieties less stable (with low
Tg) which is not suitable for demanding thermo-mechanical
performances of this polymer type in their industrial applica-
tions. Therefore, bio-based cross-linked aromatics moieties
are needed to prepare renewable epoxy thermosets (Baron-
cini et al. 2016).
Thermoset polymers are widely used in industrial applica-
tions because of their versatile properties and good durabil-
ity which are provided by their highly cross-linked structure.
Epoxy-based networks are one of such class of materials
aꢀording excellent mechanical properties for structural sta-
bility along with high adhesion and excellent chemical and
heat resistance. As epoxy resins are cured in the presence
of a curing agent or hardener, properties of the cured resins
highly depend on the type of epoxy resin and the curing
agents which, nowadays, derivative of petroleum compounds
are commonly used, as well (Garra et al. 2018; Varshney
et al. 2008). Despite being classified as reprotoxic and
banned for use in food contact applications, diglycidyl ether
of bisphenol acetone (DGEBA) is still widely used for too
many applications (such as coating, bonding and construc-
tion). Therefore, in addition to the scarcity of petro-based
Many reports could be found in the literature on the sub-
ject of bio-based aromatics for high-performance epoxy pol-
ymers. Among them, vanillin as a non-hazardous material is
the only mono-aromatic aldehyde and industrially produced
from lignin, currently (15% of the overall vanillin produc-
tion) (Fache et al. 2014, 2016). Lignin itself could be a prom-
ising candidate as substitutes of BPA for the designing of
epoxy thermosets, but the variety and complex structure of
these materials have made it diꢁcult option for processing.
Fache et al. reported the synthesis and properties of several
diglycidyl compounds by the reactions of methoxyhydro-
quinone, vanillyl alcohol and vanillic acid and vanillylamine
derived from vanillin with epichlorohydrin. Triglycidylether
of vanillylamine could increase the cross-linking density
and Tg of the epoxy network (Feghali et al. 2018; Hollande
et al. 2019; Fache et al. 2015a, b, c). Recently, di-, tri- and
tetra-glycidyl bio-based epoxy monomers are derived from
(
fossil) resources, the use of renewable resources are an eco-
logical and economic necessity, and the polymer industry
has its role for playing a transition (Fache et al. 2015a, b, c;
Bomze et al. 2015; Motahari et al. 2015).
The biosource of polymers has been the subject of much
research worldwide to develop safer epoxy resins and cur-
ing agents, sustainable alternative sources. Great eꢀorts
have been made and were successful in some areas such
1
3

Chemical Papers
divanillyl alcohol which was synthesized with the almost
diꢁcult route (selective enzymatic oxidative coupling and
reduction of aldehyde moieties) in two steps. However, the
resultant epoxy thermosets displayed remarkable properties
with respect to T and T values which range from 140 to
Experimental
Materials and methods
Chemicals
g
α
2
00 °C. This system exhibits similar mechanical properties
and thermal stability than the petro-based epoxy thermoset
Vanillin (99%), 1,1,1-tris(hydroxymethyl)propane (TMP,
98%), triethanolamine (TEA, ≥99%), 1,10-diaminodecane
(DD, 97%), tris(hydroxymethyl)aminomethane (TRIS,
99.8%) cyclopentanone (≥99%), hydrochloric acid (37%),
glacial acetic acid (≥ 99%), epichlorohydrin (99%), tri-
phenylphosphine (99%), benzyl chloride (99%), triethyl-
amine (≥ 99%), hydroxylamine hydrochloride (≥ 96%),
sodium hydroxide (NaOH, ≥ 97%), anhydrous sodium
sulfate (≥ 99%), 1,4-dioxane (≥ 99.5%), chloroform
(≥ 99%), acetonitrile (≥ 99.9%), diethyl ether (≥ 99.7%),
ethanol (≥99.9%) were all purchased from Merck, and tri-
ethyl benzyl ammonium chloride (TEBAC) was prepared
according to the literatures (Nashy et al. 2011). All chemi-
cals were used as received without further puriꢂcation.
(
Feghali et al. 2018; Hollande et al. 2019; Fache et al. 2015a,
b, c; Shibata and Ohkita 2017).
Therefore, the future of vanillin and their derivatives
(
such as di-vanillin and combination with aliphatic bridge)
was grown as bio-based synthons building blocks for renew-
able epoxy thermosets (Shibata and Ohkita 2017; Savonnet
et al. 2018). The properties of cured resins highly depend
on the type of epoxy and curing agent as well. On the other
hand, multi-functional epoxy compounds have high indus-
trial importance and vanillin-based materials can be further
functionalized to prepare compounds to fulꢂll diꢀerent prop-
erties. Indeed, it is inevitable to increase the overall func-
tionality of a given epoxy formulation and cross-link density
aꢀording higher T , which can enhance the thermo-mechan-
g
ical properties of the resultant epoxy network. Hardeners,
undoubtedly, play an eꢀective impact in increasing grid-
ability (network performance) due to their diverse structures
and functionalities. Among all curing agents (such as anhy-
drides, phenols, amines, carboxylic acids and so on), amines
are the most applicable one (Pham and Marks 2002), and
aliphatic amines are highly active even at room temperature
and provide good balance between reactivity and the ꢂnal
properties. However, aromatic samples provide higher ther-
mal and chemical stability (Goulding 2003). On the other
hand, bio-based hardeners (such as amino acids, vegetable
oils, chitosan and so on) are less visible in the literature.
Therefore, the range of bio-based amine hardeners should
be increased which is created to compare the properties of
new hardeners with currently existing systems (Hollande
et al. 2019; Li et al. 2007; Metkar et al. 2014; Stemmelen
et al. 2015).
Synthesis and characterization of epoxy monomers
Synthesis of 2,5‑bis(4‑hydroxy‑3‑methoxybenzylidene)
cyclopentanone (BP) Cyclopentanone (21 g, 0.25 mol)
was slowly added into a round-bottomed ꢃask (250 ml)
which was charged with vanillin (80 g, 0.5 mol) in 100 ml
of 1,4-dioxane and stirred for 30 min at room tempera-
ture. Concentrated hydrochloric acid (3 ml) was slowly (in
30 min) added using a syringe. The resultant solution was
stirred at room temperature (for 5 h) and then left it to stand
(for 3–4 days) at the ambient condition with the occasional
stirring. The mixture was poured into acetic acid/crushed ice
(1:1 w/w). The solid was ꢂltered, washed several times with
distilled water and oven-dried at 80 °C under reduced pres-
sure for overnight. Greenish-yellow powder was obtained.
Yield: 95%. M : 214 °C. Expected formula C H O ,
p
21 20
5
−
1 1
In this study, synthesis and characterization of new tri-
functional bio-based epoxy reagent were reported. The cur-
ing agents were chosen from both aliphatic alcohols and
amines to exhibit the inꢃuence of all sections of reagent’s
structures on thermo-mechanical and physical properties in
the epoxy thermosets.
M =352.4 g mol , H NMR (300 MHz, DMSO-d , δ,
W
6
ppm): 9.69 (s, 2H, alcohol), 7.40 (s, 2H, 1-ethylene), 7.28
(br, 2H), 7.20 (br, J=8 Hz, 2H), 6.93 (d, J=8 Hz, 2H, ben-
zene), 3.88 (s, 6H, methoxy), 3.10 (s, 4H, cyclo-aliphatic).
−
1
FT-IR (KBr, cm ): 3497, 2937, 1680, 1583, 1512, 1447,
1356, 1250, 1211, 1179, 1031, 935, 806. MS (EI, m/z): 352
+
[
M ].
Synthesis of 2,5‑bis(4‑hydroxy‑3‑methoxybenzylidene)
cyclopentanone oxime (BPO) In a round-bottomed ꢃask
(
(
100 ml), NaOH (0.6 g, 0.015 mol) and NH OH·HCl
2
1.042 g, 0.015 mol) were added, respectively, to a solu-
tion of BP in H O/EtOH 1:9 (30 ml). The resultant solu-
2
tion was heated to reꢃux for about 16 h (TLC control).
1
3

Chemical Papers
The solution was poured onto ice/water (2/1) in which
the oxime product immediately precipitates. Finally, the
product was ꢂltered, washed with very cold water, dried
in oven and used as is for the next step. Orange powder
7.15 (d, J=8.4 Hz, 2H), 7.09 (s, 2H), 6.95 (d, J=8.4 Hz, 2H,
benzene), 4.31 (dd, J=11.4, 3.1 Hz, 2H), 4.03 (dd, J=11.4,
5.7 Hz, 2H, phenoxy methylene), 3.90 (s, 6H, methoxy),
3.40 (m, 2H, epoxy methine), 3.04 (s, 4H, cyclo-aliphatic),
2.91 (t, J=4.5 Hz, 2H), 2.76 (dd, J=4.8, 2.6 Hz, 2H, epoxy
was obtained. Yield: 90%. M : 210 °C. Expected formula
p
−
1
1
−1
C H NO , M = 367.4 g mol , H NMR (300 MHz,
methylene). FT-IR (KBr, cm ): 3055, 2920, 1684, 1595,
2
1
21
5
W
DMSO-d , δ, ppm): 10.77 (s, 1H, oxime), 9.67 (s, 2H,
1512, 1451, 1281, 1243, 1141, 1023, 918, 854. MS (EI, m/z):
6
+
alcohol), 7.38 (s, 2H, 1-ethylene), 7.24 (s, 2H), 7.16 (d,
536 [M ].
J = 8.4 Hz, 2H), 6.92 (d, J = 8.0 Hz, 2H, benzene), 3.86 (s,
6
H, methoxy), 3.04 (s, 4H, cyclo-aliphatic). FT-IR (KBr,
General procedure for preparation of epoxy thermosets
−
1
cm ): 3361, 3006, 2957, 2932, 1674, 1592, 1513, 1463,
+
1
259, 1215, 1158, 934, 844, 816. MS (EI, m/z): 365 [M ].
A mixture of GBP/GBPO (1.325 g/1.02 g, epoxy:
0
.0057 mol), TMP (0.255 g, OH: 0.0057 mol) and TPP
Method for preparation of epoxy monomers (glycidylation
process)
(1 wt.% of the total weight of resins) was mixed vigorously
by stirrer bar in glass tube at 150 °C until the molten pre-
polymer got dark (ca. 90 min). The viscous mixture was
poured into a silicon mold (preheated before) and placed
in an oven (vacuumed) for 12 h at 150 °C. The same man-
ner was employed for the TEA (with limited time manner).
However, DD and TRIS (structures are shown in Scheme 1)
were handled diꢀerently. For this purpose, all components
of the polymer are mixed properly at powder form (30 °C),
placed into a silicon mold and heated up to 150 °C for 12 h
(like the other samples). All the components of the thermo-
set structure were mixed in stoichiometric amounts (i.e., two
to one for amines and one to one for hydroxyl groups), which
has been proved as the best ratio to preparation of the ther-
mosets (Fache et al. 2015a, b, c; Shibata and Ohkita 2017).
A round-bottomed ꢃask (50 ml) was charged with BP or
BPO (0.008 mol), TEBAC (0.0008 mol) and epichlorohy-
drin (0.08 mol). The mixture was stirred for 1 h after 80 °C
till a clear dark solution appeared. The solution was cooled
down to room temperature. A prepared aqueous solution of
TEBAC (0.0008 mol) and NaOH (6.4 ml of 5 M solution)
was added (a solid was appeared). The mixture was vigor-
ously stirred for (1 h) at room temperature. After quenching
the reaction mixture with aqueous HCl (0.1 M), chloroform
(
300 ml) and water (200 ml) were added. The mixture was
stirred (for 30 min) until all solid was dissolved. The aque-
ous phase was extracted using chloroform. Organic phases
were combined and dried over anhydrous sodium sulfate.
After ꢂltration, dried organic layer was concentrated via a
rotary evaporator under reduced pressure to remove excess
of the epichlorohydrine. The resultant product was dried (at
Characterization
1
All HNMR experiments were performed using AVANCE
6
9
0 °C for 24 h) to produce GBP and GBPO. Yield (GBP):
III-300 spectrometer in deuterated solvent (such as CDCl
3
7%, pale yellow powder, M : 207 °C. Expected formula
and DMSO-d ). Mass spectrums were recorded by Varian
p
6
−
1
1
C H O , M =464.5 g mol , H NMR (300 MHz, CDCl ,
CH-7A spectrometers. FT-IR spectrums (resolution of 4 at
32 runs) were recorded using Thermo Nicolet AVATAR 370.
The gel content of each sample was tested by the Soxhlet
extraction. 1.5 g of cured sample was extracted for 24 h,
using 200 ml acetone as solvent. For the accuracy, each sam-
ple was tested three times. A diꢀerential scanning calorim-
eter (DSC) instrument (Mettler Toledo DSC822) at a heating
27
28
7
W
3
δ, ppm): 7.50 (s, 2H, 1-ethylene), 7.21 (m, 2H), 7.12 (d,
J=1.7 Hz, 2H), 6.97 (m, 2H, benzene), 4.33 (dd, J=11.4,
3
.2 Hz, 2H), 4.06 (dd, J=11.4, 5.7 Hz, 2H, phenoxy meth-
ylene), 3.92 (s, 6H, methoxy), 3.42 (m, 2H, epoxy methine),
3
2
2
1
.08 (s, 4H, cyclo-aliphatic), 2.93 (m, 2H), 2.78 (dd, J=4.9,
−
1
.6 Hz, 2H, epoxy methylene). FT-IR (KBr, cm ): 3007,
925, 2876, 1686, 1594, 1513, 1461, 1332, 1247, 1136,
−
1
rate of 10 °C min was used to characterize epoxy thermo-
set sample prepared using diꢀerent curing agents. The glass
transition temperature (T ) and heat of fusion (ΔΗ ) data
+
023, 939, 857. MS (EI, m/z): 464 [M ].
Yield (GBPO): 95%, yellow orange powder, M : 132 °C.
p
g
f
−
1 1
Expected formula C H NO , M = 535.6 g mol , H
were determined by integrating the area under corresponding
3
0
33
8
W
NMR (300 MHz, CDCl , δ, ppm): 7.47 (s, 2H, 1-ethylene),
peaks in DSC curves. Thermal gravimetric analysis (TGA)
3
Scheme 1 Multi-functional
hardeners used in this study
1
3

Chemical Papers
was performed on PerkinElmer TGA-7 with a constant rate
The general synthetic pathway for the preparation of all
epoxy network components is shown in Scheme 2. The ꢂrst
step in our synthetic approach was crossed aldol conden-
sation subsequent to dehydration reactions of vanillin and
cyclopentanone according to the reported method (Shibata
and Ohkita 2017). The product was used as monomer to the
synthesis of GBP (Sardjiman et al. 1997). The oximation
reaction was carried out using hydroxyl amine hydrochlo-
−
1
of 20 °C min from 25 to 700 °C on air ꢃow, and dynamic
mechanical thermal analysis (DMTA) measurements were
carried out on Mettler Toledo DMA1, stare system. The
samples had a rectangular geometry (length 30 mm, width
1
0 mm, thickness 15 mm). Uniaxial stretching of samples
−1
was performed while heating at a rate of 5 °C min from 25
to 210 °C, keeping frequency at 1 Hz. In order to evaluate
the mechanical properties of TRIS and DD-based thermo-
sets, curing reactions were then performed on larger quanti-
ties (2 g) in molds enabling tensile tests (details are available
in supplementary information). Tensile tests were performed
on Zwick/Roell Z250 system. The measurements were per-
formed on standardized dog bone sample (width = 2 mm;
thickness=2 mm and length of ꢂxed section=12 mm) using
ride and NaOH in EtOH/H O (9:1, V/V) at reꢃux condi-
2
tion, which is converting ketones to the pre-required oxi-
mes. Oximes are capable of reacting with various sources
of carbon electrophiles, especially with epoxides (Soltani
Rad et al. 2009; Rad et al. 2008). Thus, epichlorohydrin in
the presence of sodium hydroxide and TEBAC (as a base
and a phase transfer catalyst, respectively) reacted with two
diꢀerent nucleophiles (in GBPO) under controlled condi-
tions to aꢀord epoxy monomer as precursor to the synthesis
of a thermoset resin. The details of the synthesis as well
as the spectral data are reported in the experimental sec-
tion. In the FT-IR spectra of BPO, two distinguished peaks
−
1
5
00-N load cell and crosshead speed of 1 mm min .
Results and discussion
−
1
As shown in Scheme 2, the epoxy thermosets prepared in
this study can be divided into two main groups. The ꢂrst
and second groups were created to investigate the inꢃuence
of number and type of functionality of epoxy monomer and
curing agent (hydroxyl and amine based) on properties of the
ꢂnal products, respectively. Hence, all the curing processes
were carried out at same condition to ensure about chain
mobility during gelation and to obtain optimal cross-linking
content. Therefore, eꢀect of these parameters on thermo-
mechanical properties was investigated followed by evalua-
tion the thermal stability of the ꢂnal thermosets.
around 1674 and 3361 cm were observed. These bands
explained conversion of the ketone into the corresponding
−1
oxime by shifting the initial peaks at 1680 and 3497 cm . In
1
H-NMR spectra of BP and BPO, one and two singlet peaks
of –OH protons appeared in the down ꢂeld region of each
spectrum at 9.69 (2H), 9.67 (2H) and 10.77 (1H) ppm. Some
peaks were also observed at 7.38 ppm (and 7.40 ppm for
BP, s-2H) corresponding to the oleꢂnic protons which were
formed by dehydration reaction after crossed aldol conden-
sation. The protons of aromatic ring appeared in the region
of 7.00-7.70 ppm. The aliphatic protons of methoxy group
Scheme 2 Synthesis route of epoxy monomers
1
3

Chemical Papers
−
1
were manifested with a peak at 3.88 ppm. In the case of
GBP and GBPO, two protons of the epoxy methylene group
were separately observed at 2.91 and 2.76 ppm, and the pro-
ton corresponding to the epoxy methane group was seen at
showed higher intensity at 850–950 cm as compared to
GBPO, which is corresponded to lower cross-linking density
at this thermoset. The gel content of prepared epoxy sample
was measured by Soxhlet extraction, which was performed
during 24 h by acetone. As the result indicated, increasing
the degree of functionality in raw material also increases
the amount of gel content (up to 99.7%). Therefore, the
raw reactants were almost completely incorporated into the
thermoset networks. However, the epoxy systems containing
TEA hardener exhibit lowest percentage (down to 95.6%),
which can be explained by unreacted reagent inside of the
epoxy network (Yang et al. 2017).
3
.40 ppm (2H). However, all the data presented with using
1
H-NMR and FT-IR spectrums were conꢂrmed by the elec-
trospray ionization mass spectroscopy (ESI–MS) analysis.
The progress of curing process was followed by evolution
of the most signiꢂcant bands in FT-IR spectroscopy, DSC
thermogram and gel content to evaluate the real chemical
transformation that is carried out. The results are shown in
Fig. 1. The initial spectrum shows two absorption bands
−
1
at range of 850 and 950 cm corresponding to the epoxy
ring. These absorptions have disappeared (completely) in
the spectrum of the cured material, whereas a new broad
As it can be seen in Fig. 2, the use of stoichiomet-
ric amount for mixing led most of the epoxy systems to
fully cure. An optimal T of 113 °C is attained for an
g
−
1
adsorption at 3400–3600 cm appeared due to the forma-
tion of the hydroxyl propyl moieties in the network structure
of the resultant thermosets indicating the incorporation of
reactants into the polymer networks (Guzmán et al. 2019).
However, the absorption pattern of sample prepared by GBP
epoxy system cured by TRIS. It seems with the raising
of amine groups, ΔΗ of the epoxy thermosets showed
f
−
1
much higher value (82.1 J g ) as compared to other sys-
tems containing the hydroxyl group. Furthermore, the
glass transition temperature and heat of fusion of the
Fig. 1 FT-IR spectrum of epoxy thermosets prepared by a same hardener (TMP) and b same epoxy monomer (GBPO) before and after curing
EBP or EBPO)
(
1
3

Chemical Papers
Inꢀuence of functionality
The proꢂle of DMA curves for (both) the epoxy monomers
is depicted in Fig. 3 and summarized in Table 1. It is worth
noting that there is only one structural diꢀerence between
two functionals consistent with Shibata et al. ꢂndings in
some aspects than the current petro-based epoxy monomer
and new tri-functional epoxy monomers. Therefore, the
impact of this diꢀerence on thermo-mechanical properties
was investigated. At a glance, it can be observed that the
thermoset prepared with GBPO showed higher value of T ,
α
υ, E′ content and E″ temperature as compared with analogue
Fig. 2 DSC thermograms of thermosets obtained using EBPO epoxy
sample obtained using GBP (two functionals) (Entry 1 and
system and diꢀerent curing agents
2
—Table 1). This result is explained by the relative high
curing degree of the reagent tri-functionality leading to an
elevation of the cross-linking density and less mobility of the
resultant segments (Fache et al. 2015a, b, c; Thoseby et al.
1986). Actually, with eliminating the steric and topologi-
cal constraint over tri-hydroxyl hardener (in GBPO-TEA),
thermosets increased as the curing degree of reagents
increased. However, it seems that the product prepared
using TEA is not in appropriate amount of composition
for the desired polymer processing. Therefore, among all
T of the thermosets increased (to 110 °C). Thermosets
α
the obtained products, this sample showed the lowest T
obtained by TEA (as hardener), however, diminished the
overall cross-linking (υ) and E′ to the lowest value among
the analogue samples prepared with GBPO. They may
attribute to the unreacted end group or even free monomer
presence inside the epoxy network. Indeed, these materials
act as plasticizers and handle the degree of curing to lower
value (Fache et al. 2015a, b, c; Hu et al. 2014; Guzman et al.
g
−
1
and ΔΗ (70, 92 °C and 6.8 J g ). This observation might
f
be attributed to an excess of either amine or epoxy present
unreacted end groups or even free monomers. Therefore,
these unreacted entities can move and rotate freely, which
diminishes the cross-linking density and T (Fache et al.
g
2
015a, b, c).
2
018).
Compared with the tri-hydroxyl derivatives of hardener,
Thermo‑mechanical properties of epoxy networks
the amine hardeners present a greater functionality which
allows increasing of T of the epoxy thermosets from 97 to
α
The thermo-mechanical characterization of the epoxy
networks was performed by dynamic mechanical analy-
sis (DMA) to investigate storage modulus (E′) and loss
modulus (E″) and tan δ curves as a function of tempera-
118 °C (Entry 5—Table 1). The cross-linking density also
increased with the elevating the functionality of the amine
−
3
groups in the curing agent up to 69.9 mol m ; thus, the
highest cross-linking density was observed for the thermoset
cured with DD. Nevertheless, the EBPO-DD product indi-
cates the lowest E″ of 126.4 MPa, but the TMP-based prod-
ucts exhibit the highest loss modulus values with GBP and
GBPO as epoxy monomer (329.8 and 359.3, respectively).
This result may be attributed to the rigid framework and less
prone to irreversible deformation of the created by EBPO-
DD network (Shibata and Ohkita 2017; Guzman et al. 2017).
ture for each product. The T is defined as the maximum
α
of the tan δ curves corresponding to the transition from
a glassy to a rubbery state. The cross-linking density (υ)
could be calculated from rubber-like elasticity theory
as υ = E′(T
)/φ.R.T
, where T , the maximum of
α+50
α+50
α
the tan δ, E′, the storage modulus at T
, φ the front
α+50K
factor approximately equal to 1 in Flory theory and
−
1
−1
R = 8.314 J mol
K
the gas constant. Also, from the
The molecular weights between cross-links (M ) of
c
rubbery plateau values of the storage modulus, molecular
the cured resins were calculated, and details are given in
weight between cross-links (M ) of the cured resin was
Table 1. The M of tri-functional epoxy resin was near
c
C
calculated as M = 3φρRT/E′, with φ = 1.0 as a front fac-
the half of analogue epoxy monomer (bifunctional) at the
c
−
1
−1
tor and ρ = 1.2 g mol as density of cured resin. R is a
same condition (675.1 and 1043.0 g mol , respectively).
gas constant, and T is the absolute temperature of T
;
With the increasing participation of amine functionality
in network structure of epoxy product, virtually simi-
α+50K
E′ is the elastic response of the material and is related to
mechanical energy stored per cycle upon deformation.
E″ is the viscous response and is related to the dissipated
energy per cycle when the sample is deformed (Iijima
et al. 1992; Fache et al. 2015a, b, c; Kishi et al. 2011).
lar trending was observed, and the lowest M value was
c
observed for the sample contaminated with DD as cur-
−
1
ing agent (50.9 g mol ). This suggests the small distance
between the chains and relatively rigid chemical structure
1
3

Chemical Papers
Fig. 3 Plots of DMA results such as a tan δ, b E′ and c E″ versus temperature for various materials obtained from diꢀerent formulations by cur-
ing agent and epoxy monomer
Table 1 Thermal data of the materials obtained after curing process
T (°C) T (°C) ΔΗ (J g−1) Intensity of υd
a
b
c
GC (%) M (g mol−1) E′ (MPa) E′
e
f
g
E″ (MPa)
i
Sample
α
g
f
c
tan δ peak (×10 mol m−3)
3
(T
)h
α+50K
(
MPa)
EBP-TMP
EBPO-TMP
EBPO-TEA 110
EBPO-DD 109
EBPO-TRIS 118
90
97
–
–
21.1
0.7
0.8
0.6
0.3
0.5
3.4
5.3
1.4
69.9
7.8
98.2
99.1
95.6
99.7
99.4
1043.0
675.1
3467.6
50.9
2400
3200
1500
1700
2100
4.0
6.0
1.6
93.0
9.5
329.8
359.3
211.3
126.4
207.8
94
70, 91 6.8, 6.5
102
113
82.1
25.3
599.5
a
d
b
c
Alpha transition temperature determined by DMTA, glass transition temperature determined by DSC, Heat of fusion measured by DSC,
e
cross-linking density estimated according to rubber-like elasticity theory (Shibata and Ohkita 2017), gel content determined by Soxhlet extrac-
f
g
tion (Yang et al. 2017), molecular weight between cross-links determined by DMTA (Kishi et al. 2011), storage modulus determined by
h
i
DMTA, storage modulus in rubbery state determined at tan δ+50 K. Loss modulus determined by DMTA
in epoxy networks which increase the amount of cross-
At 30 °C, the polymers were all glassy and their storage
−
3
linking as exhibited for this sample (69.9 mol m ). The
modulus was all around 1.5–3.5 GPa, and the prepared
EBPO-TMP had the highest E′ (30 °C) at 3.2 GPa than
the other samples. As shown in Fig. 3a, c, with increasing
of amine groups in curing agent, the intensity of the tan δ
and E″ peak decreases from 0.8 to 0.3 and from 359.3 to
126.4 MPa, respectively. These features could be due to
more cross-linked and less prone to irreversible deforma-
tion (Fache et al. 2015a, b, c).
relationships between the M and the T of the cured resins
C
α
were examined too. As shown, the decrease in M leads to
c
increase in cross-linking density and T in formed epoxy
α
thermosets (up to 118 °C). However, the sample contain-
−
1
ing TEA showed the highest M value (3467.6 g mol
)
c
and relatively high T (110 °C), which is somewhat unex-
α
pected. This suggests the inꢃuence of both M and chemi-
c
cal structure between cross-link points (Kishi et al. 2011).
1
3

Chemical Papers
In addition, tan δ curves of the resultant polymer network
show diꢀerent types of broadenings (Fig. 3a). However, the
materials cured using TEA show broad and bimodal tan
δ curves, which suggests signiꢂcant network heterogene-
ity or coexistence of two diꢀerent structures in the formed
epoxy networks (Gerard et al. 1991; Morancho et al. 2018),
whereas the narrow and unimodal curves were seen for the
sample prepared with TMP (tri-hydroxyl function) as cur-
ing agent. The broadening of the sample cured by TRIS and
DD, however, could be related to the existence of active
functional group of these agents at higher temperature dur-
ing analysis (Morancho et al. 2018). Tensile tests were also
performed on EBPO-TRIS- and EBPO-DD-based thermo-
sets. Results are summarized in Table 2, and tensile traces
are available in supplementary information (Fig. S22). The
TRIS-based thermosets exhibited a Young’s modulus of
value can be calculated according to T = 0.45 (T + 0.6
s d5%
(T
− T ), using two-degradation temperature weight
d5%
d30%
loss, T
and T
. The heat-resistant index shows the
d30%
d5%
temperature of the polymer in the physical heat tolerance
limit (Chiu et al. 2008).
Bimodality can be clearly observed in curves (Fig. 4b)
including ꢂrst and faster degradation process (40 wt% loss at
400 °C) along with a deꢂned shoulder at temperature around
600 °C (with 80 wt% loss). Moreover, more complexity of
degradation pattern was observed for similar materials pre-
pared with both TEA and DD. The result could be explained
with lower cross-linking density of TEA or by degradation
reaction speciꢂc (such as thermo-oxidative degradation) to
the structure, which is common in material containing nitro-
gen atoms or amine groups (Fache et al. 2015a, b, c; Aouf
et al. 2013).
4
80 MPa and an elongation at break of 6.5%. In the case
A vast range of changes were seen at the temperature of
of DD-based product, the value of Young’s modulus and
elongation shows a slightly decrease (down to 290 MPa and
the initial degradation (T ) for all the obtained materials.
d5%
The product cured with TMP shows a higher value of statis-
3.85%, respectively). In fact, the presence of hydroxyl moie-
tic heat-resistant index, T and T
, than the other sam-
d30%
d5%
ties in curing agent allows more ꢃexibility to the network in
comparison with the DD-based epoxy thermoset. Therefore,
the highest value of tensile strength was belonged to EBPO-
TRIS epoxy thermoset (22.90 MPa) (Savonnet et al. 2018).
ples. The thermoset prepared with TEA begins to degrade
at lower temperature (307 °C, 5 wt% loss) as compared to
the other analogue samples (Table 3). Moreover, the lowest
to highest half-life of the epoxy thermosets was attributed
to the product cured with TEA (1221 s) and TRIS (1347 s),
respectively. This fact can be ascribed to increasing of func-
tional group or existence of multiple groups in the hardeners
(Fache et al. 2015a, b, c; Shibata and Ohkita 2017; Savonnet
et al. 2018).
Thermo‑gravimetric analysis (TGA)
Thermal stability of the prepared samples was examined at
high temperature with using TGA, and the results are plot-
ted in Fig. 4. The statistic heat-resistant index (T ) shows a
The diꢀerential curve of thermo-gravimetric analysis
shows that the sample with the highest cross-linked network
s
characteristic of the thermal stability of the cured resin. This
Table 2 Mechanical properties
Sample
Young’s modulus Elongation at
Strain at Fmax (mm)
Ultimate tensile
strength (MPa)
of epoxy networks formed after
curing with DD and TRIS
(
MPa)
break (%)
EBPO-DD
EBPO-TRIS
290± 80
480± 60
3.85± 0.4
6.56± 1.1
0.38± 0.03
0.67± 0.1
10.35± 5.5
22.90± 3.0
Fig. 4 Thermo-gravimetric (a) and diꢀerential thermo-gravimetric (b) curves of the resultant epoxy networks
1
3

Chemical Papers
Table 3 Diꢀerential thermo-gravimetric results for the obtained epoxy networks
a
b
c
d
e
Char yield (%)f
Sample
T (°C)
T
(°C)
T
(°C)
Half life (s)
T (°C)
s
d5%
d30%
d
5
00
600
700
EBP-TMP
191
194
174
181
178
353
360
307
313
312
415
419
388
406
398
1233
1215
1221
1311
1347
419 (65%), 592 (14%)
420 (63%), 562 (20%)
362 (81%), 404 (61%), 606 (21%)
339 (89%), 432 (60%), 627 (19%)
402 (68%), 606 (23%)
41.3
37.6
42.9
43.8
47.4
11.0
7.2
23.4
27.5
25.4
0.7
1.4
3.3
2.5
1.7
EBPO-TMP
EBPO-TEA
EBPO-DD
EBPO-TRIS
a
b
c
Statistic heat-resistant index determined by TGA, temperature of 5% of weight loss in atmosphere, temperature of 30% of weight loss in
d
e
f
atmosphere, determined by TGA, temperature of the maximum rate of degradation in atmosphere, residue after thermal degradation in atmos-
phere
indicating a steady state or equilibrium degradation reac-
tion. The highest hydrolytic stability is observed for the
samples prepared by TMP (1.8 and 2.9%) in the early stage
of degradation; however, the lowest percentage of degrada-
tion belonged to the epoxy thermoset cured by DD at the
end of incubation (7.9% at 144 h). This behavior might
be attributing to high cross-linking density and chemical
stability of the ꢂnal epoxy thermoset (details gathered in
Table S1) (Hollande et al. 2019).
Conclusions
A new bio-based tri-functional epoxy resin was prepared
Fig. 5 Analysis of weight fraction diꢀerence as function of incuba-
with glycidation of tri-hydroxyl precursor which itself
was synthesized with oximation of two-functional mate-
rial derived from vanillin. The parameters such as type and
number of functionality were found to develop the thermo-
mechanical properties and thermal stability of the materi-
als to the higher value. For instance, the epoxy networks
containing amine groups displayed enhanced thermal sta-
tion time at 60 °C in acid aqueous solution
has the lowest weight loss temperature (431.7 °C) (Fig. 4).
However, the fastest and the greatest weight loss just hap-
pens for the material cured with TEA (at 404 °C, 40 wt%
loss) and TMP (at 420 °C, by 37 wt% loss), while this sys-
tem (EBPO-TEA) exhibited the highest residual mass value
bility (such as half-life and T ) and thermo-mechanical
d
(
char yield) at 700 °C, surprisingly. The behavior could be
properties (such as T , υ, E″) comparable to the other
d
explained with the ꢂre-retardant nature of the hardeners
based on nitrogen atoms in the epoxy network. Thus, the
char yield increased as content of the amine group raised in
the epoxy network (from 1.5 to 3.3%) (Rad et al. 2008; Van
Krevelen 1975).
samples, which is attributed to the rigid structure formed
inside the network. As result indicated, the cross-linking
density of EBPO-TMP network is almost two times higher
−
3
−3
(5.3 mol m ) than EBP-TMP network (3.4 mol m );
therefore, higher T was observed in the resulted resins
α
(
97 °C). Various types of hardeners (hydroxyl and amine
Hydrolytic degradation
based) were also tested with GBPO as an enhanced epoxy
monomer. In contrast, the highest residual mass value is
observed for the sample cured with TEA (as a tri-hydroxyl
curing agent). The resultant epoxy thermoset prepared
with DD exhibited the highest hydrolytic stability (down to
7.9%) than the other samples. However, the sample cured
by TRIS exhibited the highest elasticity, elongation and
tensile strength among the obtained samples.
Hydrolytic degradation tests were performed in aqueous
solution of HCl (3 M) at accelerated condition (60 °C).
Degree of degradation was measured in terms of weight
fraction diꢀerence of the remaining sample as a function
of soaking time. Figure 5 shows a rapid increase in deg-
radation, which reaches up to 50.7% in 5 h of incubation
period. However, after 90 h, the values remain unchanged,
1
3

Chemical Papers
Acknowledgements This study was ꢂnancially supported by Grant
No. IR.BMSU.REC.1397.294 at Baqiyatallah University of Medi-
cal Sciences. The authors would also thank Prof. G.H. Zohuri, Dr.
N. Ramezanian, Mr M. Behzadpour, M. Hematian and Sh. Jourabchi
Prog Organ Coat 114:259–267. https://doi.org/10.1016/j.porgc
oat.2017.10.025
Guzmán D, Serra A, Ramis X, Fernández-Francos X, De la Flor S
(2019) Fully renewable thermosets based on bis-eugenol prepared
by thiol-click chemistry. React Funct Polym 136:153–166. https
(
From Ferdowsi University of Mashhad) for all their cooperation.
:
//doi.org/10.1016/j.reactfunctpolym.2018.12.024
Hollande L, Do Marcolino I, Balaguer P, Domenek S, Gross RA, Allais
F (2019) Preparation of renewable epoxy-amine resins with tun-
able thermo-mechanical properties, wettability and degradation
abilities from lignocellulose-and plant oils-derived components.
Front Chem 7:159. https://doi.org/10.3389/fchem.2019.00159
Hu F, La Scala JJ, Sadler JM, Palmese GR (2014) Synthesis and char-
acterization of thermosetting furan-based epoxy systems. Mac-
romolecules 47:3332–3342. https://doi.org/10.1021/ma500687t
Iijima T, Yoshioka N, Tomoi M (1992) Eꢀect of cross-link density
on modiꢂcation of epoxy resins with reactive acrylic elasto-
mers. Eur Polym J 28:573–581. https://doi.org/10.1016/0014-
References
Aouf C, Nouailhas H, Fache M, Caillol S, Boutevin B, Fulcrand H
(
2013) Multi-functionalization of gallic acid. Synthesis of a novel
bio-based epoxy resin. Eur Polym J 49:1185–1195. https://doi.
org/10.1016/j.eurpolymj.2012.11.025
Baroncini EA, Kumar Yadav S, Palmese GR, Stanzione JF III (2016)
Recent advances in bio-based epoxy resins and bio-based
epoxy curing agents. J Appl Polym Sci 133:44103. https://doi.
org/10.1002/app.44103
3
057(92)90025-W
Kishi H, Akamatsu Y, Noguchi M, Fujita A, Matsuda S, Nishida H
2011) Synthesis of epoxy resins from alcohol-liqueꢂed wood and
the mechanical properties of the cured resins. J Appl Polym Sci
(
Bomze D, Knaack P, Liska R (2015) Successful radical induced
cationic frontal polymerization of epoxy-based monomers by
C–C labile compounds. Polym Chem 6:8161–8167. https://doi.
org/10.1039/C5PY01451D
1
20:745–751. https://doi.org/10.1002/app.33199
Li Y, Xiao F, Wong CP (2007) Novel, environmentally friendly
crosslinking system of an epoxy using an amino acid: tryptophan-
cured diglycidyl ether of bisphenol A epoxy. J Polym Sci A Polym
Chem 45:181–190. https://doi.org/10.1002/pola.21742
Chiu YC, Chou IC, Tseng WC, Ma CCM (2008) Preparation and ther-
mal properties of diglycidylether sulfone epoxy. Polym Degrad
Stab 93:668–676. https://doi.org/10.1016/j.polymdegradstab
Metkar PS, Scialdone MA, Moloy KG (2014) Lysinol: a renewably
resourced alternative to petrochemical polyamines and aminoalco-
hols. Green Chem 16:4575–4586. https://doi.org/10.1039/C4GC0
.
2007.12.014
Fache M, Darroman E, Besse V, Auvergne R, Caillol S, Boutevin
B (2014) Vanillin, a promising biobased building-block for
monomer synthesis. Green Chem 16:1987–1998. https://doi.
org/10.1039/C3GC42613K
1
167H
Morancho JM, Ramis X, Fernández-Francos X, Salla JM, Konuray AO,
Serra À (2018) Curing of oꢀ-stoichiometric amine–epoxy thermo-
sets. J Therm Anal Calorim 133:519–527. https://doi.org/10.1007/
s10973-018-7158-2
Fache M, Viola A, Auvergne R, Boutevin B, Caillol S (2015a) Biobased
epoxy thermosets from vanillin-derived oligomers. Eur Polym J
6
8:526–535. https://doi.org/10.1016/j.eurpolymj.2015.03.048
Motahari A, Rostami AA, Omrani A, Ehsani M (2015) On the ther-
mal degradation of a novel epoxy-based nanocomposite cured
with tryptophan as an environment-friendly curing agent. J
Macromol Sci Phys 54:517–532. https://doi.org/10.1080/00222
Fache M, Montérémal C, Boutevin B, Caillol S (2015b) Amine hard-
eners and epoxy cross-linker from aromatic renewable resources.
Eur Polym J 73:344–362. https://doi.org/10.1016/j.eurpo
lymj.2015.10.032
3
48.2015.1019331
Fache M, Auvergne R, Boutevin B, Caillol S (2015c) New vanillin-
derived diepoxy monomers for the synthesis of biobased ther-
mosets. Eur Polym J 67:527–538. https://doi.org/10.1016/j.eurpo
lymj.2014.10.011
Nashy ESH, Megahed MG, El-Ghaꢀar MA (2011) Preparation of
fat-liquor based on jojoba oil under phase transfer catalysis. J
Am Oil Chem Soc 88:1239–1246. https://doi.org/10.1007/s1174
6
-011-1778-z
Fache M, Boutevin B, Caillol S (2016) Epoxy thermosets from model
mixtures of the lignin-to-vanillin process. Green Chem 18:712–
Pham HQ, Marks MJ (2002) Epoxy resins. Encyclopedia of polymer
science and technology. Wiley, Hoboken
7
25. https://doi.org/10.1039/C5GC01070E
Rad MNS, Behrouz S, Dianat M (2008) Aqueous-mediated ring open-
ing of epoxides with oximes: a rapid entry into β-hydroxy oxime
O-ethers as potential β-adrenergic blocking agents. Synthesis
Feghali E, Torr KM, van de Pas DJ, Ortiz P, Vanbroekhoven K, Eevers
W, Vendamme R (2018) Thermosetting polymers from lignin
model compounds and depolymerized lignins. Top Curr Chem
2
008:2055–2064. https://doi.org/10.1055/s-2008-1067122
3
76:32. https://doi.org/10.1007/s41061-018-0211-6
Sardjiman SS, Reksohadiprodjo MS, Hakim L, Van der Goot H,
Timmerman H (1997) 1,5-Diphenyl-1,4-pentadiene-3-ones and
cyclic analogues as antioxidative agents. Synthesis and structure-
activity relationship. Eur J Med Chem 32:625–630. https://doi.
org/10.1016/S0223-5234(97)83288-6
Garra P, Graꢀ B, Schrodj G, Morlet-Savary F, Dietlin C, Fouassier
JP, Lalevée J (2018) Ultrafast epoxy-amine photopolyaddition.
Macromolecules 51:10230–10236. https://doi.org/10.1021/acs.
macromol.8b02340
Gerard JF, Galy J, Pascault JP, Cukierman S, Halary JL (1991) Vis-
coelastic response of model epoxy networks in the glass transi-
tion region. Polym Eng Sci 31:615–621. https://doi.org/10.1002/
pen.760310813
Savonnet E, Grau E, Grelier S, Defoort B, Cramail H (2018) Divanillin-
based epoxy precursors as DGEBA substitutes for biobased epoxy
thermosets. ACS Sustain Chem Eng 6:11008–11017. https://doi.
org/10.1021/acssuschemeng.8b02419
Goulding TM (2003) Epoxy resin adhesives: handbook of adhesive
technology, revised and expanded. CRC Press, Boca Raton
Guzman D, Ramis X, Fernández-Francos X, De la Flor S, Serra A
Shibata M, Ohkita T (2017) Fully biobased epoxy resin systems com-
posed of a vanillin-derived epoxy resin and renewable phenolic
hardeners. Eur Polym J 92:165–173. https://doi.org/10.1016/j.
eurpolymj.2017.05.007
(2017) New bio-based materials obtained by thiol-ene/thiol-epoxy
dual curing click procedures from eugenol derivates. Eur Polym
J 93:530–544. https://doi.org/10.1016/j.eurpolymj.2017.06.026
Guzman D, Ramis X, Fernandez-Francos X, De la Flor S, Serra
A (2018) Preparation of new biobased coatings from a trigly-
cidyl eugenol derivative through thiol-epoxy click reaction.
Soltani Rad MN, Khalaꢂ-Nezhad A, Behrouz S (2009) Design and
synthesis of some novel oxiconazole-like carboacyclic nucleoside
analogues, as potential chemotherapeutic agents. Helv Chim Acta
9
2:1760–1774. https://doi.org/10.1002/hlca.200900051
1
3

Chemical Papers
Stemmelen M, Lapinte V, Habas JP, Robin JJ (2015) Plant oil-based
epoxy resins from fatty diamines and epoxidized vegetable
oil. Eur Polym J 68:536–545. https://doi.org/10.1016/j.eurpo
lymj.2015.03.062
and prediction of molecular and material properties. Macromol-
ecules 41:6837–6842. https://doi.org/10.1021/ma801153e
Yang X, Wang C, Li S, Huang K, Li M, Mao W, Shan C, Xia J (2017)
Study on the synthesis of bio-based epoxy curing agent derived
from myrcene and castor oil and the properties of the cured prod-
ucts. RSC Adv 7:238–247. https://doi.org/10.1039/C6RA24818G
Thoseby MR, Dobinson B, Bull CH (1986) Recent advances with glyci-
dylamine resins. Br Polym J 18:286–291. https://doi.org/10.1002/
pi.4980180503
Van Krevelen DW (1975) Some basic aspects of ꢃame resistance
of polymeric materials. Polymer 16:615–620. https://doi.
org/10.1016/0032-3861(75)90157-3
Publisher’s Note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional aꢁliations.
Varshney V, Patnaik SS, Roy AK, Farmer BL (2008) A molecular
dynamics study of epoxy-based networks: cross-linking procedure
1
3