Journal of Adhesion Science and Technology 22 (2008) 101–110

www.brill.nl/jast

Adhesive Characteristics of Epoxy/Dendritic Hyperbranched

Polymer Blends

Debdatta Ratna ∗
NMRL, Shil-Badlapur Road, Anand Nagar P.O., District Thane, Maharashtra - 421 506, India

Received in final form 25 November 2007

Abstract
Adhesive characteristics of blends of a room-temperature-curing epoxy and a dendritic hyperbranched poly-
mer (HBP) were investigated. Significant improvements in both lap shear and T-peel strengths were observed
as a result of blending of HBP. Dynamic mechanical analysis of the cured blend formulations indicates a
two-phase microstructure. The improvement in adhesive bond strength is achieved without significant sac-
rifice in glass-transition temperature of the cured epoxy network. The results were explained in terms of
phase morphology analyzed by scanning electron microscopy. Cavitation and shear yielding are believed to
be responsible for improvement in toughness and adhesive bond strength.
 Koninklijke Brill NV, Leiden, 2008

Keywords
Epoxy, adhesive bond strength, hyperbranched polymers, microstructure

1. Introduction

Epoxy resins are unique among all the thermosetting resins and the same is reflected
in their extreme applications, namely as a matrix for high-temperature composites
[1, 2] and as vibration-damping materials [3]. They are widely used for diverse
applications, namely as adhesives, surface coatings and electrical laminates. How-
ever, when cured with stoichiometric amounts of polyfunctional amines, the high
degree of cross-linking makes the epoxy resin a brittle material having a fracture
energy about two orders of magnitude lower than the engineering thermoplastics
and three orders of magnitude lower than metals [4]. It weakens epoxy’s both peel
and impact strengths and, therefore, limits its applications. Various modifications
strategies (especially liquid rubber toughening) have been adopted to improve the

*Current address: Institut fur Verbundwerkstoffe GmbH (Institute for Composite Technology), Technical
University Kaiserslautern, Erwin-Schrodinger-Str. 58, D-67663 Kaiserslautern, Germany. Tel.: (49-631)
201-7242; Fax: (49-631) 201-7198; e-mail: [email protected]

 Koninklijke Brill NV, Leiden, 2008 DOI:10.1163/156856108X292304

102 D. Ratna / Journal of Adhesion Science and Technology 22 (2008) 101–110

adhesive bond strength as reviewed recently [5]. Liquid rubber toughened epoxies
have been exploited in the field of adhesives [6]; however, modification of epoxy
with a linear liquid rubber is associated with a tremendous increase in viscosity
which is disadvantageous from adhesive application point of view. Because of their
inherent low viscosity, dendritic hyperbranched polymers (HBPs) have drawn con-
siderable attention and have been proposed for application as alternative tougheners
for thermoset resins [7–11].
Dendritic polymers are a new class of three-dimensional, synthetic materials
produced by multiplicative growth from small molecules that incorporate repeti-
tive branching sequences to create an unusual architecture. Over the last few years,
synthesis of a large number of dendritic polymers has been reported in the open
literature [12, 13], as well as in patents [14, 15]. They have potential applications
in a variety of fields such as biomedical, catalysis, coatings, electrical materials and
so on. However, actual exploitation of dendritic polymer technology is still in its
infancy, especially in applications where large amounts are required due to synthe-
sis difficulties such as retaining solubility of the growing polymer and maintaining
regularity and order in the structure, which makes the technology extremely costly.
Recently, dendritic polymers have been produced by a new lower cost hybrid
synthesis process that generates very highly branched, polydisperse molecules.
These materials are called dendritic hyperbranched polymers (HBP) to distinguish
them from the perfect monodisperse materials [16, 17]. They are being produced in
sufficient quantities to allow for investigation of their utility in conventional en-
gineering applications. Due to the compact three-dimensional structure of such
dendritic polymers, these molecules mimic the hydrodynamic volume of spheres
in solution or in melt and flow easily past each other under applied stress. This
results in a low melt viscosity, even at high molecular weights, due to a lack of
restrictive interchain entanglements [18]. The high density of surface functional
groups on dendritic HBP also offers the potential for tailoring their compatibility
either through conversion of dendritic polymer end groups to chemically suitable
moieties, or through in situ reaction to form covalently bonded networks. These two
properties, low viscosity and tailorable compatibility, make them excellent candi-
dates as an additive which could also act simultaneously as a toughening agent [6,
7]. The epoxy/HBP blends have been investigated [11, 19–23] and show consider-
able promise for high performance structural applications.
The reported toughened epoxy adhesive formulations suffer from the problem of
heat cure. There is only a limited literature on ambient temperature curing adhesive
formulations [24, 25]. Heat curing is difficult and impractical for fabrication of
certain structures and requires significant amount of energy. Ambient curing saves
energy and is advantageous for adhesive applications with intricate structures.
The aim of the present work was to develop a room temperature curing epoxy
adhesive toughened with a HBP. The effect of incorporation of a HBP on the bond
strength of epoxy is discussed in this paper.
 D. Ratna / Journal of Adhesion Science and Technology 22 (2008) 101–110 103

2. Experimental
2.1. Materials
The epoxy resin used was a liquid diglycidyl ether of bisphenol A (DGEBA) (Ciba
Geigy, Araldite LY 556) with an equivalent weight of 195 ± 5 per epoxide group.
An ambient temperature amine hardener (Ciba Geigy, HV 953U) with an amine
equivalent of 3.6 equiv/kg was used as a curing agent. The solvents toluene and
methanol used were of analytical grade.
An epoxy functional dendritic hyperbranched polymer (Boltorn E1) with an
epoxy equivalent weight of approx. 875 g/equiv and a molecular weight of approx.
10 500 g/mol, was supplied by Perstorp Speciality Chemicals (Sweden). Boltorn
E1 consists of a highly branched aliphatic polyester backbone with an average 11
reactive epoxy groups per molecule. Chemical structures of epoxy and the HBP are
shown in Fig. 1.
2.2. Preparation of Blends and Curing
The epoxy/HBP blend samples were prepared by mixing the HBP with DGEBA
with thorough stirring. All the formulations were analyzed for their epoxy content
by standard titration [26] with hydrogen bromide in acetic acid. Accordingly, a sto-
ichiometric amount (40 g for one equivalent of epoxy) of curing agent was added
and thoroughly mixed. The resin-hardener mixture was applied onto the substrate

Figure 1. Chemical structures of epoxy resin and HBP.

104 D. Ratna / Journal of Adhesion Science and Technology 22 (2008) 101–110

for adhesion tests and cast into a Teflon mold for dynamic mechanical analysis
(DMA).
2.3. Dynamic Mechanical Analysis
DMA measurements on cured epoxy/HBP blend samples were carried out using a
Dynamic Mechanical Thermal Analyzer (DMTA MK III, Rheometric Scientific) at
a fixed frequency of 1 Hz with a 3◦ C/min heating rate using liquid nitrogen for sub-
ambient region. Storage moduli and loss factors were obtained by a dual cantilever
mode for the sample size 14 mm × 10 mm × 2 mm.
2.4. Evaluation of Epoxy/HBP Blends for Adhesive Bond Strength
Aluminum alloy (B 51 SWP), containing 0.4 wt% Fe, 0.6 wt% Mn, 1 wt% Si
and 0.8 wt% Mg, was used as the substrate for measurement of Al–Al lap-shear
strength. The specimen dimensions used were 150 mm × 25 mm × 4 mm and the
lap length was 10 mm. T-peel tests for steel to rubber bonding were carried out
using mild steel and nitrile rubber as the adherends. The specimen dimensions for
steel and rubbers were 150 mm × 25 mm × 3 mm and 150 mm × 25 mm × 4 mm,
respectively. The lap length was 50 mm.
The aluminum and mild steel panels were grit-blasted. The grit-blasted pan-
els were stored in a desiccator and used within 3 days. The panels were cleaned
with petroleum ether and dried in an oven at 60◦ C prior to use. The nitrile rubber
strips were rubbed with an emery paper and the rubbed surface was cleaned with
dichloroamine.
The adhesive formulation was applied evenly on the surfaces of both adherends,
with a glass rod (thickness of adhesive layer was about 150 µm) and then joined
together using contact pressure. The samples were then allowed to cure at room
temperature for 7 days.
Single lap shear and T-peel strengths were measured for all the samples using
a Universal Testing Machine (UTM) (Hounsfield, Model H-50 KS) according to
ASTM D-1002 and D-1876, respectively. A cross-head speed of 10 mm/min was
used for the measurement of lap-shear strength and 50 mm/min was used for mea-
surement of T-peel strength. The lap shear strength is expressed in MPa, and it was
calculated by dividing the load required to break the adhesive bond (in N) by the
contact area. The T-peel strength (P.S.) was calculated from the following expres-
sion:
P.S. = 2F /w, (1)
where F is the force to separate the joint and w is the width of the specimen. Five
specimens for each formulation were tested and the average values are reported.
2.5. SEM Analysis
A low voltage scanning electron microscope (SEM; Jeol, JSM 840) was used to
examine the fracture surfaces of the toughened epoxy samples. A thin section of
 D. Ratna / Journal of Adhesion Science and Technology 22 (2008) 101–110 105

the fracture surface was cut and mounted on an aluminum stub using a conductive
(silver) paint and was sputter-coated with gold prior to fractographic examination.
SEM micrographs were obtained under conventional secondary electron imaging
conditions with an acceleration voltage of 20 kV.

3. Results and Discussion

The epoxy groups of the epoxy resin are more reactive than those of HBP. When
the homogeneous mixture of epoxy and HBP is cured, the epoxy groups of the
epoxy resin react rapidly. The curing results in the formation of hydroxyl groups
which decreases the compatibility of the epoxy resin and the HBP, leading to the
formation of a two-phase microstructure. The hardener molecules at the interface
are expected to react with both the epoxy groups of the epoxy and unreacted epoxy
groups of HBP to form chemical bonds with the rigid matrix and HBP, which is
necessary for effective toughening [27, 28].
3.1. Dynamic Mechanical Analysis
In order to study the phase morphology, all the samples (cured blend formula-
tions with different concentrations of HBP) were subjected to DMA analysis from
−100◦ C to 150◦ C. The effect of blending of HBP on low temperature and high
temperature relaxation peaks for the blend samples are shown in Figs 2 and 3, re-
spectively, which represent the loss factor (tan δ) plots against temperature. From
the figures it is clear that all the modified networks show two relaxation peaks, one
at −35◦ C and other at 140◦ C. The former peak is attributed to the α-relaxation of
HBP and the latter peak is due to α-relaxation of the unmodified epoxy, as the height

Figure 2. Effect of incorporation of HBP on the low temperature transition of epoxy/HBP blends.
106 D. Ratna / Journal of Adhesion Science and Technology 22 (2008) 101–110

Figure 3. Effect of incorporation of HBP on the high temperature transition of epoxy/HBP blends.

of the low temperature relaxation peak gradually increases with the incorporation of
HBP. The broad peak in the spectrum for pure epoxy is due to β-relaxation of epoxy.
This indicates that there is phase separation resulting in epoxy phase and HBP-rich
epoxy phase as observed in high-temperature curing rubber modified epoxy formu-
lations containing HBP and other liquid rubbers. However, the blends show a lower
Tg compared to pure epoxy, which indicates some phase mixing. Phase mixing to a
certain extent is desirable as it increases the ductility of the matrix leading to better
toughening [29, 30]. This is also reflected in the broadness of the loss peaks which is
generally correlated with dissipation of mechanical energy [5, 18, 19]. As discussed
in the Introduction, the poor peel strength is due to the low fracture toughness of
epoxy and, hence, the toughening effect is expected to contribute significantly in
improvement to peel strength of the epoxy adhesive.
3.2. Adhesive Bond Strength
A pre-polymer-like epoxy undergoes two important processes, namely gelation and
vitrification. Gelation is defined as the condition at which the weight-average mole-
cular weight reaches infinity. Vitrification is defined as the point at which the Tg of
the network becomes the same as the cure temperature. At this point the material is
transformed from a rubbery gel to a gelled glass and the curing reaction becomes
diffusion controlled. Hence, while studying a room temperature curing system it is
important to know the curing time for an optimum functional property.
The test specimens for lap shear test using epoxy/HBP (100/20) blend were
made and tested at various intervals. Figure 4 shows the lap-shear data as a func-
tion of cure time. It was found that lap shear strength increased with cure time
and reached a plateau after 7 days. It is expected that the blend compositions with
 D. Ratna / Journal of Adhesion Science and Technology 22 (2008) 101–110 107

Figure 4. Lap shear strength of joints using epoxy/HBP blend (20 wt% HBP) as adhesive as a function
of cure time.

Figure 5. Lap shear and T-peel strengths of joints using epoxy/HBP blends as adhesive as a function
of HBP content.

lower amount of HBP will reach optimum curing in 7 days. Thus, all the sam-
ples were allowed to cure for 7 days. The effects of blending of HBP on lap shear
and T-peel strengths are shown in Fig. 5. All the blends show better adhesive
bond strength compared to pure epoxy. It may also be noted that the improve-
ment is achieved without significant decrease in the Tg . In our previous work with
CTPEGA-modified epoxy [25] we had found that the improvement in adhesive
bond strength was associated with a significant decrease in the Tg .
108 D. Ratna / Journal of Adhesion Science and Technology 22 (2008) 101–110

3.3. SEM Analysis

In order to explain the adhesive behavior the fracture samples were analyzed by
SEM. The SEM micrographs for pure epoxy and epoxy/HBP blend (100:20) are
shown in Fig. 6. From the micrograph (Fig. 6a) one can see the smooth glassy frac-
tured surface with cracks in different planes in case of unmodified DGEBA. This
indicates brittle fracture of the unmodified DGEBA, which accounts for its poor
impact and peel strengths. The fracture surface of the modified network consists of
two distinct phases: globular HBP particles dispersed in continuous DGEBA ma-
trix. The particles are uniformly distributed throughout the matrix. The micrograph

(a)

(b)

Figure 6. (a) SEM micrograph of the fracture surface of pure epoxy (single-phase morphology).
(b) SEM micrograph of the fracture surface of epoxy/HBP blend (20 wt% HBP) (two-phase mor-
phology).
 D. Ratna / Journal of Adhesion Science and Technology 22 (2008) 101–110 109

(Fig. 6b) shows the broken HBP particles and a stress-whitened zone. Stress whiten-
ing is due to the scattering of visible light from the layer of the scattering centers,
which in this case are voids [31, 32]. The generation of the voids is due to the cav-
itation of rubbery HBP particles, which is the most important energy dissipating
mechanism in the case of rubber-toughened epoxies [33].

4. Conclusion
Epoxy/HBP blends were investigated for their adhesive bond strength. Significant
increase in adhesive bond strength was achieved as a result of blending of HBP.
The improvement in adhesive bond strength is achieved without significant sacri-
fice in thermal properties. Rubber cavitation and shear yielding are believed to be
responsible for increase in toughness and adhesive bond strength.
Acknowledgement
The author is thankful to the Alexander von Humboldt (AVH) Foundation for the
grant of a post-doctoral research fellowship.

References
1. M. Hussain and G. P. Simon, J. Mater. Sci. Lett. 22, 1471 (2003).
2. A. B. Samui, B. C. Chakraborty and D. Ratna, Int. J. Plastic Technol 8, 279 (2004).
3. D. Ratna, N. R. Manoj, L. Chandrasekhar and B. C. Chakraborty, Polym. Adv. Technol. 15, 583
(2004).
4. W. D. Bascom and D. L. Hunston, Adv. Chem. Ser. 222, 193 (1989).
5. D. Ratna, J. Adhesion Sci. Technol. 17, 1655 (2003).
6. A. J. Kinloch, Adhesion and Adhesives: Science and Technology. Chapman and Hall, London
(1987).
7. P. G. Parzuchowski, M. Kiźlińska and G. Rokicki, Polymer 48, 1857 (2007).
8. G. M. Pax, J. Bouchet, X. Bulliard, V. Michaud and J.-A. E. Månson, Int. J. Adhesion Adhesives
27, 593 (2007).
9. H. Dodiuk, Z. Gold and S. Kenig, J. Adhesion Sci. Technol. 18, 301 (2004).
10. H. Dodiuk, K. Lizenboim, I. Epelboum, B. Zalsman and S. Kenig, J. Adhesion Sci. Technol. 18,
1723 (2004).
11. R. Meezzenga and J.-A. E. Månson, J. Mater. Sci. 36, 4883 (2001).
12. F. Vögtle (Ed.), Dendrimers I, Topics in Current Chemistry, Vol. 197. Springer, Berlin (1998).
13. Y. Kim, J. Polym. Sci., Part A Polym. Chem. 36, 1685 (1998).
14. J. F. Jansen, US Patent 5,788,989 (1997).
15. L. A. Milco, US Patent 5,731,095 (1998).
16. M. Jayakannan, J. L. J. Van Dongen, G. C. Behera and S. Ramakrishnan, J. Polym. Sci. Part A:
Polym. Chem. 40, 4463 (2002).
17. B. Voit, J. Polym. Sci. Part A: Polym. Chem. 38, 2505 (2000).
18. F. W. Billmeyer, Textbook of Polymer Science, 3rd edn. Wiley, New York, NY (1984).
19. J. Verrey, Y. Winkler, V. Michaud and J.-A. E. Månson, Composites Sci. Technol. 65, 1527 (2005).
20. R. Mezzenga, L. Boogh and J.-A. E. Månson, Macromolecules 33, 4373 (2000).
110 D. Ratna / Journal of Adhesion Science and Technology 22 (2008) 101–110

21. D. Ratna, R. Varley and G. P. Simon, J. Appl. Polym. Sci. 92 1604 (2004).
22. D. Ratna and G. P. Simon, J. Polym. Mater. 19, 349 (2002).
23. D. Ratna, R. Varley, R. K. S. Raman and G. P. Simon. J. Mater. Sci. 38, 147 (2003).
24. D. Ratna and A. K. Banthia, J. Adhesion Sci. Technol. 14, 15 (2000).
25. D. Ratna, N. R. Manoj, B. C. Chakraborty and P. C. Deb, J. Adhesion Sci. Technol. 17, 623 (2003).
26. A. J. Durbetaki, Anal. Chem. 28, 2000 (1956).
27. M. Murli, D. Ratna, A. B. Samui and B. C. Chakraborty, J. Appl. Polym. Sci. 103, 1723 (2007).
28. D. Ratna and A. K. Banthia, Polym. Eng. Sci. 47, 26 (2007).
29. G. Levita, Adv. Chem. Ser. 222, 93 (1989).
30. D. Ratna and G. P. Simon, Polymer 42, 7739 (2001).
31. A. F. Yee and R. A. Pearson, J. Mater. Sci. 26, 2273 (1991).
32. P. J. Pearce, C. E. M. Morris and B. C. Ennis, Polymer 37, 1137 (1996).
33. B. Russel and R. Chartoff, Polymer 46, 785 (2005).