Catalytic activity of graphite-based nanofillers on cure reaction of epoxy resins

Article abstract of DOI:10.1016/j.polymer.2014.09.019

The strong influence of graphite oxide (GO) nanofiller on the glass transition temperature (T_g) of epoxy resins, generally attributed to restricted molecular mobility of the epoxy matrix by the nanofiller or to the crosslinking of GO layers via the epoxy chains, is investigated. The study confirms that large increases of the glass transition temperature of the nanocomposite can be observed in presence of GO. However, similar T_g increases are observed, when the filler is a high-surface-area graphite (HSAG), lacking oxidized groups. Moreover, these T_g differences tend to disappear as a consequence of aging or thermal annealing. These results suggest that the observed T_g increases are mainly due to a catalytic activity of graphitic layers on the crosslinking reaction between the epoxy resin components (epoxide oligomer and di-amine), rather than to reaction of the epoxide groups with functional groups of GO. This hypothesis is supported by investigating the catalytic activity of graphite-based materials on reactions between analogous monofunctional epoxide and amine compounds.

Full text of DOI:10.1016/j.polymer.2014.09.019

Polymer 55 (2014) 5612e5615
Contents lists available at ScienceDirect
Polymer
journal homepage: www.elsevier.com/locate/polymer
Polymer communication
Catalytic activity of graphite-based nanofillers on cure reaction of
epoxy resins
Marco Mauro a, Maria Rosaria Acocella a, Carola Esposito Corcione b, Alfonso Maffezzoli b,
,
*
Gaetano Guerra a
a
ꢀ
ꢀ
Dipartimento di Chimica e Biologia e Unita di Ricerca INSTM, Universita di Salerno, Fisciano, SA, Italy
Dipartimento di Ingegneria dell'Innovazione, Universita del Salento, Lecce, Italy
b
ꢀ
a r t i c l e i n f o
a b s t r a c t
Article history:
The strong influence of graphite oxide (GO) nanofiller on the glass transition temperature (Tg) of epoxy
resins, generally attributed to restricted molecular mobility of the epoxy matrix by the nanofiller or to
the crosslinking of GO layers via the epoxy chains, is investigated. The study confirms that large increases
of the glass transition temperature of the nanocomposite can be observed in presence of GO. However,
similar Tg increases are observed, when the filler is a high-surface-area graphite (HSAG), lacking oxidized
groups. Moreover, these Tg differences tend to disappear as a consequence of aging or thermal annealing.
These results suggest that the observed Tg increases are mainly due to a catalytic activity of graphitic
layers on the crosslinking reaction between the epoxy resin components (epoxide oligomer and di-
amine), rather than to reaction of the epoxide groups with functional groups of GO. This hypothesis is
supported by investigating the catalytic activity of graphite-based materials on reactions between
analogous monofunctional epoxide and amine compounds.
Received 8 April 2014
Received in revised form
5 September 2014
Accepted 7 September 2014
Available online 16 September 2014
Keywords:
Epoxy resins
Graphite-based nanofillers
Differential Scanning Calorimetry
© 2014 Elsevier Ltd. All rights reserved.
1. Introduction
In particular, although most nanoparticles generally disrupt
crosslinking density and decrease Tg in high Tg thermosets
Polymer nanocomposites have attracted great interest from
researchers and technologists, due to unique properties derived
from the extended interface between the polymer matrix and the
filler [1e4]. Nanocomposites with carbon-based fillers have been
widely investigated and, in particular, many studies have been
focused on nanocomposites with graphite oxide (GO) [5e10], i.e. a
layered material, (generally obtained by chemical oxidation of
graphite with strong acids and oxidants) [11e13], whose layers
mainly contain hydroxyl and epoxide groups at the surface and
carbonyl and carboxyl groups at the edges [14]. The large increase
of the interlayer spacing and the disordered arrangement of the
functional oxidized groups on the layer surfaces lead to weak
interlayer attractive forces, making easy the GO exfoliation by
dispersion in a suitable polymer matrix [15e17]. These studies on
GO reinforced nanocomposites show relevant improvements of
physical properties of epoxy resins, which are mainly due to the
polymer-filler improved adhesion, arising from the oxygenated
groups [5e10].
[18e20], several studies describe significant Tg increases for
polymer matrices, in the presence of GO nanofiller [21e27]. In
fact, according to literature data, this increase is negligible for
apolar polymers like isotactic polypropylene [21], of 4.5 ꢀC for
polystyrene [22], of nearly 8 ꢀC for poly(propylene carbonate)
[23] and definitely higher for polymethylmethacrylate (in the
range 7e17 ꢀC) [24,25]. By far the largest Tg increases, in the
presence of GO nanofiller, have been observed for epoxy resins
(in the range 10e36 ꢀC) [26,27]. These extremely high Tg in-
creases would be due to restricted molecular mobility of the
epoxy matrix by the nanofiller [26] or to the possible crosslinking
of GO layers via the epoxy chains. The latter hypothesis, in
principle, could be supported by recent results of Berry et al. [28]
on the functionalization of GO by ring opening reaction of
epoxidized methyl oleate, as well as of Zheng et al. [29] on the
coupling of epoxy resin onto GO sheets via the ‘‘grafting to’’
method.
In this paper, comparative thermal analyses of epoxy resins fil-
led with different carbon nanofillers (a high-surface-area graphite,
HSAG and a graphite oxide), for different thermal histories, are
conducted. The main aim is to understand the molecular origin of
the influence of GO on the Tg of epoxy resins.
* Corresponding author. Tel.: þ39 089969558.
E-mail address: gguerra@unisa.it (G. Guerra).
http://dx.doi.org/10.1016/j.polymer.2014.09.019
0032-3861/© 2014 Elsevier Ltd. All rights reserved.

M. Mauro et al. / Polymer 55 (2014) 5612e5615
5613
2. Results and discussion
Experimental details are listed in the Supporting Information.
The used high surface area graphite (HSAG, 308 m2/g) exhibits the
X-ray diffraction pattern of Fig. 1A, showing an interlayer distance
of 0.339 nm, a high shape anisotropy (D /Dt ¼ 3.1) and a low de-
k
gree of order in the relative position of parallel graphitic layers (i.e.
a tendency toward the so-called turbostratic graphite) [17,30].
The Hummers' oxidation of HSAG leads to a GO, with an oxygen
content of 32 wt% (excluding the water content) and a decrease of
surface area down to 1.4 m2/g (See Supporting Information). These
surface area values agree well with those reported for GO samples
by Bielawski et al. [31e33], while they are definitely lower than
other literature values [34e36].
The X-ray diffraction pattern of the derived GO, reported in
Fig. 1B, shows an interlayer distance increase from 0.339 nm to
0.84 nm. It is worth noting that the out-of-plane correlation length
decreases from 9.8 nm to 4.2 nm, while the in-plane correlation
length remains almost unchanged (D z 30 nm). Hence, GO pre-
k
sents a high shape anisotropy D /Dt ¼ 7, associated with a strongly
k
reduced order in the direction perpendicular to the graphene oxide
layers.
The WAXD patterns of epoxy-based nanocomposites, filled with
3 wt% of HSAG and GO are shown in Fig.1A0 and B0, respectively. The
WAXD pattern of the nanocomposite with HSAG shows a sharp 002
reflection, clearly showing that the graphite in the epoxy matrix
still consists of multilayer stacks. In the WAXD pattern of the
nanocomposite with GO, on the contrary, reflections of the filler
cannot be identified. This indicates the occurrence of high disorder
in the direction perpendicular to the graphitic layers, possibly
corresponding to a complete exfoliation of the graphene oxide
layers.
Fig. 2. DSC heating scans of the neat epoxy resin (A, A0) and of the epoxy-based
nanocomposites, filled with 3 wt% of HSAG (B, B0) and GO (C, C0): (A, B, C) samples
cured at 150 ꢀC; (A0, B0, C0) the same samples after annealing at 200 ꢀC. The evaluated
Tg are indicated close to the curves.
Moreover, large increases of Tg are observed only for freshly
prepared samples. In fact the Tg of the same samples after aging at
room temperature for 3 months become much closer, due to a large
aging-induced increase for the neat resin (z20 ꢀC) and a negligible
variation for the nanocomposites. After aging, the Tg difference
between the GO nanocomposite and the neat resin becomes close
to 10 ꢀC.
DSC heating scans of the neat epoxy resin and of the corre-
sponding nanocomposites with 3 wt% of HSAG and GO, as cured at
150 ꢀC, are shown in Fig. 2AeC and the derived Tg data are listed in
the second column of Table 1.
An increase of Tg of about 24 ꢀC and 31 ꢀC, with respect to the
neat epoxy resin (Tg ¼ 124 ꢀC, Fig. 2A), is observed in presence of
HSAG and GO (Fig. 2B and C), respectively. The similar Tg increases
for nanocomposites with graphite and graphite oxide clearly in-
dicates that the formation of covalent bonds between the graphene
oxide layers and the crosslinked epoxy is not the main mechanism
inducing these Tg increases.
Also informative are the Tg values of the neat resin (Fig. 2A0) and
of the nanocomposites with G and GO (Fig. 2B0 and C0), after
annealing at 200 ꢀC, which are also listed in the 3rd column of
Table 1. After annealing, the Tg of the neat resin increases of 28 ꢀC
while that one of the GO nanocomposite increases of only 3 ꢀC and
as a consequence the Tg difference between the GO nanocomposite
and the neat resin becomes as low as 6 ꢀC.
The whole set of data relative to Tg (Table 1) clearly suggests that
most of the Tg increase, which is observed for epoxy resins
including GO, has a kinetic rather than a thermodynamic origin.
The simplest hypothesis to rationalize the described behavior is a
possible catalytic effect of graphitic layers on the crosslinking re-
action of the epoxy resin, between the used epoxide oligomer and
di-amine. In fact, it is well known that both graphene and graphite
oxide show catalytic effects on many organic reactions [31,37e41],
being for some reactions more effective and selective than other
catalysts [41]. Moreover, a catalytic activity of carbon nanotubes on
Table 1
Tg of the epoxy resin and its nanocomposites with HSAG and GO, as cured at 150 ꢀ
and subsequently annealed at 200 ꢀC.
C
Sample
Tg, 150 (ꢀC)
Tg, 200 (ꢀC)
Neat epoxy resin
124
148
155
152
156
158
Fig. 1. X-ray diffraction patterns (CuKa) of HSAG (A) and GO (B) and of the epoxy-
Epoxy þ 3 wt% HSAG
based composites with 3 wt% of HSAG (A0) and GO (B0). Miller indexes and d spac-
Epoxy þ 3 wt% GO
ings (nm) are indicated close to the main diffraction peaks.

5614
M. Mauro et al. / Polymer 55 (2014) 5612e5615
Table 2
Ring opening reaction of styrene oxide (SO) by benzylamine (BA).
Entry
Molar ratio SO/BA
Catalyst (3 %wt)
Time (h)
Yield (%)a
1
2
3
4
5
1/1.2 or 3/1
1/1.2
1/1.2
3/1
3/1
e
24
24
24
18
48
e
HSAG
GO
GO
GO
54 (SO-BA)
45 (SO-BA)
45 (SO-BA)
64 (SO-BA-SO)
a
All the yields refer to isolated chromatographically pure compounds, whose structures were confirmed by analytical and spectroscopic data.
epoxy resin curing reactions has been clearly established (mainly
by dynamic and isothermal DSC studies) [41e47], which has been
attributed to the high thermal conductivity of carbon nanotubes
[43] or to the presence of impurities in commercially available
carbon nanofillers [46].
The possible catalytic activity of HSAG and GO, for reactions
between monofunctional low molecular mass epoxides and
amines, such as styrene epoxide and benzylamine, is presently
studied. The choice of monofunctional reactants is of course due to
an easier characterization of uncrosslinked reaction products.
The epoxide ring opening reaction was conducted at room
temperature in presence of 3 wt% of the considered high-surface-
area graphite (HSAG) or of the derived GO, under solvent free
conditions.
As reported in Table 2, already at room temperature and in the
presence of both nanofillers, substantial yields of the addition
product SO-BA are achieved (entries 2 and 3 in Table 2) while the
uncatalyzed reaction does not proceed at all (entry 1 in Table 2).
As for analogy with crosslinking of epoxy resins, particularly
interesting is exploring the possible catalytic activity of the
graphitic nanomaterials on the reaction of the secondary amine SO-
BA with a further epoxide molecule. This study has been conducted
by using the same reaction conditions but a higher epoxide/amine
ratio (3/1, entries 4 and 5 in Table 2): for short and long reaction
times, products SO-BA and SO-BA-SO are prevailingly obtained,
which correspond to addition of one or two epoxide molecules.
It is worth adding that, for both products SO-BA and SO-BA-SO,
differently from reactions in water solutions [48,49], the ring
opening reaction as catalyzed by graphite-based nanofillers is
highly regioselective, following an SN2 type reaction with the
nucleophile attacking the less substituted end (see reaction on the
top of Table 2).
Tg increases in GO nanocomposites. In particular, these Tg increases
observed in presence of GO, for the considered epoxy resin cured at
150 ꢀC, is higher than 30 ꢀC. However, similar Tg increases (24 ꢀC)
are observed, when the filler is a high-surface-area graphite
(HSAG), lacking oxidized groups. The similar results obtained for
GO and graphite nanofillers, exclude the possibility that the
observed large Tg increases could be attributed either to restricted
molecular mobility of the epoxy matrix by interactions with the
dipolar groups of the GO nanofiller or to the possible crosslinking of
GO layers via the epoxy chains.
Additional DSC analyses have shown that these Tg differences
tend to disappear as a consequence of aging or annealing at higher
temperatures (e.g.,
D
Tg,max z 6 ꢀC, for Tannealing ¼ 200 ꢀC). This
clearly indicates that most of the Tg increment, which is observed
for epoxy resins including graphite-based nanofillers, has a kinetic
rather than a thermodynamic origin.
The higher Tg values observed for low curing temperatures, for
epoxy resins with graphite-based nanofillers, can be easily ratio-
nalized by a catalytic activity of graphitic layers on the reaction
between the epoxy and amine groups of the resin, which leads to
higher crosslinking density in milder conditions.
This hypothesis has been clearly supported by experiments
showing catalytic activity of the considered graphite-based nano-
fillers on the epoxide ring opening reaction, for monofunctional
epoxide and amine reactants. In particular, the considered nano-
carbons exert a catalytic activity not only on reactions between
primary amines and epoxide groups but also on reactions between
secondary amines and epoxide groups, like those needed to
crosslink epoxy resins.
Acknowledgments
These experiments clearly show that the considered nano-
carbons exert a catalytic activity not only on reactions between
primary amines and epoxide groups but also on reactions between
secondary amines and epoxide groups, like those needed to
crosslink epoxy-amine resins. Moreover, because our experiments
have been conducted in isothermal conditions, the possibility that
the observed catalytic activity is due to the high thermal conduc-
tivity of the graphitic materials can be ruled out.
We thank Prof. L. Guadagno and Prof. P. Longo of University of
Salerno for useful discussions. Financial support of the “Ministero
ꢀ
dell'Istruzione, dell'Universita e della Ricerca” (PRIN) is gratefully
acknowledged.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.polymer.2014.09.019.
3. Conclusion
References
In summary, comparative thermal analysis of epoxy resins filled
with different carbon nanofillers, confirms the occurrence of large
[1] Messersmith PB, Giannelis EP. Chem Mater 1994;6:1719e25.
[2] Lan T, Kaviratna PD, Pinnavaia TJ. Chem Mater 1995;7:2144e50.

M. Mauro et al. / Polymer 55 (2014) 5612e5615
5615
[3] Leroux F, Besse JP. Chem Mater 2001;13:3507e15.
[4] Coleman JN, Khan U, Gun'ko YK. Adv Mater 2006;18:689e706.
[5] Fang M, Wang K, Lu H, Yang Y, Nutt S. J Mater Chem 2009;19:7098e105.
[6] Kim H, Abdala AA, Macosko CW. Macromolecules 2010;43:6515e30.
[7] Wissert R, Steurer P, Schopp S, Thomann R, Mülhaupt R. Macromol Mater Eng
2010;295:1107e15.
[27] Tang LC, Wan YJ, Yan D, Pei YB, Zhao L, Li YB, et al. Carbon 2013;60:16e27.
[28] Ahn BK, Sung J, Li Y, Kim N, Ikenberry M, Hohn K, et al. Adv Mater 2012;24:
2123e9.
[29] Shen B, Zhai W, Tao M, Lu D, Zheng W. Compos Sci Technol 2013;77:87e94.
[30] Li ZQ, Lu CJ, Xia ZP, Zhou Y, Luo Z. Carbon 2007;45:1686e95.
[31] Dreyer DR, Bielawski CW. Adv Funct Mater 2012;12:3247e53.
[32] Todd AD, Bielawski CW. Catal Sci Technol 2013;3:135e9.
[33] Dreyer DR, Murali S, Zhu Y, Ruoff RS, Bielawski CW. J Mater Chem 2011;21:
3443e7.
[34] Bradder P, Ling SK, Wang S, Liu S. J Chem Eng Data 2011;56:138e41.
[35] Li Y, Du Q, Liu T, Peng X, Wang J, Sun J, et al. Chem Eng Res Des 2013;91:
361e8.
[36] Wang H, Hu YH. Ind Eng Chem Res 2011;50:6132e7.
[37] Sarvari MH, Slarghi H. Helvetica Chim Acta 2005;88:2282e7.
[38] Dreyer DR, Jarvis KA, Ferreira PJ, Bielawski CW. Macromolecules 2011;44:
7659e67.
[8] Lee YR, Kim SC, Lee H, Jeong HM, Raghu AV, Reddy KR, et al. Macromol Res
2011;19:66e71.
[9] Hu H, Chen L, Chen G. Mater Manuf Process 2011;26:618e22.
[10] Guo Y, Bao C, Song L, Yuan B, Hu Y. Ind Eng Chem Res 2011;50:7772e83.
[11] Staudenmaier L. Ber Dtsch Chem Ges 1898;31:1481e7.
[12] Hummers SWR, Offeman E. J Am Chem Soc 1958;80:1339.
[13] Mauro M, Maggio M, Cipolletti V, Galimberti M, Longo P, Guerra G. Carbon
2013;61:395e403.
[14] He H, Klinowski J, Forster M, Lerf A. Chem Phys Lett 1998;287:53e6.
[15] Schniepp HC, Li JL, McAllister MJ, Sai H, Herrera-Alonso M, Adamson DH, et al.
J Phys Chem B 2006;110:8535e9.
[39] Verma S, Mungse HP, Kumar N, Choudhary S, Jain SL, Sain B, et al. Chem
Commun 2011;47:12673e5.
[16] Compton OC, Nguyen ST. Small 2010;6:711e23.
[17] Mauro M, Cipolletti V, Galimberti M, Longo P, Guerra G. J Phys Chem C
2012;116:24809e13.
[18] Zammarano M. Flame retardant polymer nanocomposites. Hoboken, NJ, USA:
John Wiley & Sons; 2006.
[19] Esposito Corcione C, Freuli F, Maffezzoli A. Polym Eng Sci 2013;53:531e9.
[20] Esposito Corcione C, Maffezzoli A. Polym Test 2013;32:880e8.
[21] Song P, Cao Z, Cai Y, Zhao L, Fang Z, Fu S. Polymer 2011;52:4001e10.
[22] Yang J, Huang L, Li L, Zhang Y, Chen F, Zhong M. J Polym Res 2013;20:173.
[23] Bian J, Wei XW, Gong SJ, Zhang H, Guan ZP. J Appl Polym Sci 2012;123:
2743e52.
[24] Li X, McKenna GB. ACS Macro Lett 2012;1:388e91.
[25] Jiang S, Gui Z, Bao C, Dai K, Wang X, Zhou K, et al. Chem Eng J 2013;226:
326e35.
[40] Chu X, Zhu Q, Dai WL, Fan K. RSC Adv 2012;2:7135e9.
[41] Acocella MR, Mauro M, Falivene L, Cavallo L, Guerra G. ACS Catal 2014;4:
492e6.
[42] Valentini L, Armentano I, Puglia D, Kenny JM. Carbon 2004;42:323e9.
[43] Valentini L, Puglia D, Frulloni E, Armentano I, Kenny JM, Santucci S. Compos
Sci Technol 2003;64:23e33.
[44] Wu J, Chung DDL. Carbon 2004;42:3039e42.
[45] Zhou T, Wang X, Liu X, Xiong D. Carbon 2009;47:1112e8.
[46] Chapartegui M, Markaide N, Florez S, Elizetxea C, Fernandez M, Santamaria A.
Polym Eng Sci 2012;52:663e70.
€
[47] Saeb MR, Bakhshandeh E, Khonakdar HA, Mader E, Scheffler C, Heinrich G. Sci
World J 2013;2013.
[48] Bonollo S, Fringuelli F, Pizzo F, Vaccaro L. Green Chem 2006;8:960e4.
[49] Azizi N, Saidi MR. Org Lett 2005;7:3649e51.
[26] Shen XJ, Pei XQ, Fu SY, Friedrich K. Polymer 2013;54:1234e42.