1-s2.0-S0143749611000194-main

7/28/2019 1-s2.0-S0143749611000194-main

1/15

Nanocomposite adhesives: Mechanical behavior with nanoclay

Erol Sancaktar n, Jason Kuznicki

Department of Polymer Engineering, The University of Akron, Akron, OH 44325, USA

a r t i c l e i n f o

Article history:

Accepted 22 September 2010Available online 21 February 2011

Keywords:

Novel adhesives

Epoxy/epoxides

Microscopy

Mechanical properties of adhesives

Epoxynanoclay composites

a b s t r a c t

The major objective for this research was to examine the role of epoxyclay nanocomposites in the area

of epoxy bonding to porous stone (granite) substrates. Two bisphenol A epoxy systems were selected

based on the prior work that determined optimal adhesive properties from a larger set of epoxy

systems to determine the role of viscosity on the intercalation and exfoliation of the clay tactiods in the

epoxy resin. The systems were characterized and mechanically tested at varying levels of intercalated

and exfoliated organic clay tactiods. In the first stage of the work, epoxyclay systems were

characterized by wide-angle X-ray diffraction (WAXD) to detect inter-laminar distances of clay layers

and to determine if the mixing procedures had indeed dispersed and exfoliated the clay layers

sufficiently. The second stage of the work involved examining mechanical properties of the epoxy

nanoclay systems. Fracture behavior was studied using granite stone substrates in notched double lap

configuration. Compressing a wedge between the cover plates induced the fracture. Fracture toughness

was approximated by the load at fracture. Tensile properties were measured using cast dog bone tensile

samples. The better layered silicate nanocomposite performance was seen with the lower viscosity

resin. The most noticeable improvements in mechanical properties for the lower viscosity resin system

were found to be maximum stress, elastic modulus, and yield stress. Increased toughness and stress

whitening at 1% by weight nanoclay loading revealed that the clay can act as a shear-yielding

toughening agent in this epoxy system.

& 2011 Elsevier Ltd. All rights reserved.

1. Introduction

Nanotechnology provides improvements in the mechanical,

thermal, and permeation properties of epoxy adhesives. In addi-

tion, the use of nanoparticles such as nanofibers, nanotubes and

graphite, various forms of nanosilicates offer promising improve-

ments in the above mentioned adhesive properties with practi-

cality and low cost.

Clay particles of higher aspect ratio with approximately

1 nm 10-100 nm dimensions can be successfully dispersed in

the epoxy matrix to produce nanocomposites containing layered

silicate of the smectic type. The particles are typically montmor-illonite, but can also be saponite and fluorohectorite forms of clay

and have a structure different from the more common clays. They

possess ion exchange properties, and thus, surface activity with a

sodium cationic gallery surface for intercalation, while they react

with organic materials [1,2]. Nanocomposites, which possess high

stiffness and dimensional stability can also be used as flame

retardant materials [3,4].

For structural adhesive applications, the development of

epoxy-based nanocomposites has drawn considerable attention.

Pinnavaia et al. [510] used layered silicate nanolayers as alter-

native inorganic components for the construction of epoxy-based

nanostructured composites. They reported that clay silicate

nanolayers possess high particle aspect ratios comparable to

conventional fibers and they modified the nanocomposite inter-

layer surface by ion-exchange reaction, intercalating the nanoclay

galleries by organic polymer precursors for the formation of

organicinorganic nanocomposites with stable SiO bonds.

They reported that the platy morphology of the silicate layers,

exfoliated to form clay nanocomposites, resulted in drama-tically improved properties such as barrier and mechanical

properties that are not available with conventional composite

materials [510].

The synthesis of amine-cured epoxy nanocomposites through

the polymerization of monomers in the galleries of protonated

onium ion exchanged forms of smectic clays depends on two

crucial factors: firstly, the ability of the onium ion to serve as a

compatibilizing agent, which allows for cointercalation of the

resin and curing agent, and secondly, the ability of the onium ion

to acid-catalyze the intragallery ring opening polymerization

reaction. The catalytic function of the onium ion is important

because, as noted above, it allows the intragallery polymerization

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/ijadhadh

International Journal of Adhesion & Adhesives

0143-7496/$ - see front matter & 2011 Elsevier Ltd. All rights reserved.

doi:10.1016/j.ijadhadh.2010.09.006

n Corresponding author. Tel.: +1 330 972 5508.

E-mail address: [email protected] (E. Sancaktar).

International Journal of Adhesion & Adhesives 31 (2011) 286300
http://-/?-http://www.elsevier.com/locate/ijadhadhhttp://localhost/var/www/apps/conversion/tmp/scratch_5/dx.doi.org/10.1016/j.ijadhadh.2010.09.006mailto:[email protected]://localhost/var/www/apps/conversion/tmp/scratch_5/dx.doi.org/10.1016/j.ijadhadh.2010.09.006http://localhost/var/www/apps/conversion/tmp/scratch_5/dx.doi.org/10.1016/j.ijadhadh.2010.09.006mailto:[email protected]://localhost/var/www/apps/conversion/tmp/scratch_5/dx.doi.org/10.1016/j.ijadhadh.2010.09.006http://www.elsevier.com/locate/ijadhadhhttp://-/?-

7/28/2019 1-s2.0-S0143749611000194-main

2/15

rate to be competitive with the extragallery polymerization rate.

Consequently, the nanolayers can be completely dispersed in the

polymer matrix, where they can fully contribute to the reinforce-

ment mechanism.

The exfoliation and intercalation of montmorillonite nano-

layers in an epoxy matrix is readily achieved with acidic organo-

philic alkylammonium ions such as C18H37NH3+ abbreviated as

C18A. According to Pinnavaia et al., it is necessary for the onium

ion to serve as a compatibilizing agent to allow cointercalation ofthe resin and curing agent, and to acid-catalyze the intragallery

ring opening polymerization reaction. This is necessary for the

synthesis of amine-cured epoxy nanocomposites through the

polymerization of monomers in the galleries of protonated onium

ion exchanged forms of smectic clays [810].

Pinnavaia et al. utilized layered silicate magadiite with a layer

thickness of 1.12 nm, exfoliated in an epoxy matrix for the

formation of polymerinorganic nanolayer composites. Shell Epon

828 elastomeric epoxy nanocomposite reinforced by magadiite in

this fashion exhibited solvent uptake reduced by 2.5-fold in

comparison to the neat resin, as illustrated by an immersion in

propanol for 50 days. Pinnavaia et al. report that after 50 days

immersion in propanol the pristine polymer absorbs 2.5 times

more than the nanocomposite, and at this time the pristine

sample began to crack and break up, whereas the shape and

texture of the nanocomposite sample appeared unchanged [8].

We have to be careful, however, in differentiating between

changes in the diffusion coefficient versus the equilibrium moist-

ure content of nanocomposite resins as a result of nanoclay

additions. For example, Rana [11] and Rana et al. [12] studied

moisture diffusion through neat vinyl ester resin containing

nanoclay by performing transient and steady-state diffusion

experiments and illustrated that montmorillonite clay was

effective in reducing the diffusion coefficient of water through

the vinyl ester resin. Rana went on to report that the solubility

of water in the nanocomposites appeared to increase with the

amount of clay, as reflected in the equilibrium moisture content.

This was due to water adsorption onto the clay. Rana hypothe-

sized that the clay acted as a moisture scavenger by trapping

water and prevented it from reaching any further. His transient

diffusion experiments revealed that the diffusion coefficient, D, of

water at room temperature dropped by approximately 92% after

an incorporation of 5 wt% of Cloisite 10As clay. His results

showed an increase from 0.66% to 2.08% in equilibrium moisture

content during his transient diffusion experiments with nano-

composite films in distilled water at 25 1C, while D went down

from 7.36109 to 0.609109 cm2/s. The advantage in incor-

porating the clay was even more dramatic for diffusion at

elevated temperatures as for the diffusion of distilled water at

42.5 1C, with D going down from 16.43109 to

0.93109 cm2/s. The excess moisture acquired due to the

presence of reinforcing (standard) glass fibers was reduced from

4.360% with no-clay to 3.221% at room temperature, with theaddition of 5% clay [11].

Results by Rana revealed that water uptake with clay rein-

forced epoxy nanocomposite samples should not be measured

using direct (thickness-normalized) water absorption experi-

ments, but rather using diffusion through fully saturated samples,

as well as an excess moisture content in epoxy bonded samples

in-situ. Such measurements should be made by normalizing with

equilibrium (saturation) absorption levels. Since the clay particles

scavenge water, the nanocomposite samples initially absorb

slightly higher amounts of water in comparison to the no-clay

samples, with the water molecules congregating around the clay

particles. On the other hand, the presence of these clay particles

still hinders diffusion of water through the sample, thus protect-

ing the structural interfaces.

Rana also reported increases in the polymer glass transition

temperature and mechanical properties with nanoclay addition

(Cloisite 10A); however, these increases were obtained by only up

to 1% nanoclay addition by weight, except for the tensile modulus.

Addition of higher amounts of nanoclay brought these material

properties back to the levels obtained with the neat resin. The

tensile modulus was found to increase monotonically, up to

$13%, when Cloisite 10A was added up to 5% by weight. With

1 wt% nanoclay addition, the notched impact strength increased$10%, the tensile strength increased $26% and strain at failure

increased $17%. The torsional properties were similarly

enhanced [11].

Since epoxy applications may exist in areas of high moisture

content and under mechanically induced stress, the effect of such

stressing on water uptake by epoxyclay nanocomposites is also

of interest. Sancaktar and Kuznicki [13] used low viscosity liquid

aromatic diglycidyl ether of bisphenol A (DGEBA) epoxy resin

Epon 815C mixed with 0.5% nanoclay (Cloisite 30B) by weight to

produce an exfoliated clayepoxy resin nanocomposite system.

These samples were immersed in water in stressed condition

(flexural stress) to assess the effect of stress on nanocomposite

epoxy system for its water uptake behavior. Application of the

flexural stress affected the water uptake barrier properties for

nanoclay/epoxy nanocomposites, with the flexural stress acting to

decrease the rate of absorption as well as to decrease the

equilibrium moisture content in the nanocomposite. The results

revealed up to 33% reduction in water uptake for the stressed

samples [13].

Park and Jana [1416] reported on conditions for complete

exfoliation of nanoclay particles in epoxynanoclay composite

systems. These authors developed master curves from the values

of storage modulus (G0) of curing epoxy inside the clay galleries

and the complex viscosity (Zn) of curing epoxy outside the

galleries to show that complete exfoliation is produced if the

ratio, G0/Zn is greater than approximately 2 (1/s); otherwise,

intercalated structures are obtained. They argued that the values

of G0 are strong functions of organic treatment of clay, the

chemical structure of curing agents, and the chemical structures

of epoxy resin and provided a comprehensive list of cure pro-

cesses for epoxy/nanoclay systems [14].

Property enhancement in nanoparticle reinforced polymers

depends on the degree of nanoparticle dispersion [1719], which

can be achieved by dispersion of nanoparticles into low viscosity

components (for example liquid hardeners for thermoset resins)

followed by in-situ polymerization [1820]. The stages of nano-

particles dispersion in polymeric matrices involve: (a) intercalation

of particles by polymer chains (a nanoscopic process); (b) flow-

induced exfoliation of individual particles (a mesoscopic process),

and; (c) homogeneous dispersion of exfoliated particles into the

matrix (a macroscopic process). The first of these processes, the

intercalation process is governed solely by diffusion [21], while the

exfoliation and homogenization processes are controlled by mix-ing. Therefore, step (a) involves clay treatment, wet-out and

intercalation. In clay treatment, functionalized onium ions

appropriate for the desired matrix polymer replace the inorganic

cationic counterions adsorbed to the inner surfaces of the clay

galleries to balance the negative charges resulting from an iso-

morphic substitution of lower valence atoms in the clay lattice. The

clay aggregates are infiltrated with polymer or precursor during

the wet-out stage. Intercalation occurs when the polymer or

precursor molecules enter, and swell the galleries. Subsequently,

layers separate and become individual during the exfoliation

step (b). The nano-platelets created in this manner still need to be

dispersed homogeneously in the matrix with step (c) [22].

As mentioned above, low amounts of montmorillonite rein-

forcement typically result in large increases in stiffness,

E. Sancaktar, J. Kuznicki / International Journal of Adhesion & Adhesives 31 (2011) 286300 287

7/28/2019 1-s2.0-S0143749611000194-main

3/15

permeability and heat resistance along with improvements in

tensile and impact strength for the resulting nanocomposites. The

degree of delamination of the clay tactoids to result in large

aspect ratio of the dispersed clay laminas increase permeation

resistance of the nanocomposite. The resulting nanoclay layer

spacing gives an idea about the degree of intercalation towards

exfoliation. This (d001) spacing of the clay crystal lattice is

typically measured by the wide-angle X-ray scattering. Interpre-

tation of the spectra can also provide information on the dis-tribution of stacking sequences [19,23,24].

The results on the mechanical behaviors of clay reinforced

nanocomposite polymers reported above reveal the necessity of

optimization with simultaneous considerations of water diffusion

and strength/strain properties. The important mechanical proper-

ties: tensile strength, impact strength, and failure strain seem to

reach their maximum values at rather lower weight percentages

of clay loading, approximately 1% by weight, and seem to go back

to and, more or less, remain near the no-clay values at nanoclay

loadings above approximately 1% by weight. The diffusion coeffi-

cient, however, decreases monotonically with clay loading. Con-

sequently, if increasing the mechanical properties of the

nanocomposite adhesive is not the main aim, then one can load

58% clay in epoxy adhesives to obtain significant reductions in

the coefficient of diffusion, thus improving the durability and

environmental resistance of structures adhesively bonded by

these nanocomposite adhesives.

Another method of reinforcing epoxy adhesives for improve-

ments in mechanical properties along with the creation of elec-

trical conduction function for them is the use of electrospun

polyacrylonitrile (PAN) fiber precursor-based carbon nanofiber

(CNF) mats incorporated in epoxy resin as shown by Sancaktar

and Aussawasathien [2527]. Mechanical properties of electro-

spun carbon nanofibers (ECNF)/epoxy nanocomposites were

obtained using tensile and flexural tests [25]. The tensile modulus

of the pure epoxy matrix was found to be 5.7 GPa and increased by

46% to 8.4 GPa when reinforced by 6 g wt% ECNF mat. The

maximum flexural tensile stress value at the outer surfaces of

the neat epoxy samples was measured to be approximately

330 MPa, which was approximately twice that of the model epoxy

tensile strength (approximately 150 MPa). When increasing

amounts of the ECNF mat were added to this neat resin, however,

the maximum flexural tensile stress at the outer surface decreased

as a result of the embrittlement effect of the ECNF mat in the

composites. Note that this embrittlement effect, which reduced

the flexural strength of the nanocomposite, was not observed with

the tensile behavior, where the ultimate tensile strength increased

by as much as 27% with 2.06% by weight ECNF mat addition.

Our previous unpublished works using Epon 815C/Epicure

3140 curing agent system revealed the following results regard-

ing the effects of synthetic graphite nanoparticles or Multiwall

Carbon Nanotubes (MWCNT) reinforced nanocomposites: condi-

tions resulting in 1.31% water absorption in the neat resinresulted in 1.22% and 1.09% water absorption along with 20%

and 30% increase in tensile moduli values, respectively, when

synthetic graphite nanoparticles or MWCNT were added in 1% by

weight proportions, respectively.

The primary objectives of this current study were as follows:

Investigate epoxynanoclay composite formation, using two

common epoxy resins with differing viscosities.

Determine the role of viscosity on the intercalation and

exfoliation of the clay tactiods.

Explore the effect of clay loading on mechanical properties of

resins.

Explore effects of mechanical interlocking and crack propaga-

tion with epoxynanoclay composites using fracture tests on

porous granite substrates bonded with the nanocomposite

adhesive and approximate fracture toughness of the nanocom-

posite adhesive systems using the load at fracture.

Thus, we hope to improve the performance of epoxynanoclay

composites in practical applications such as in bonding porous

stones (granite), which are used extensively in construction.

Many such applications will involve epoxy systems under stress,

in humid areas, or under potential fracture. This study attempts tofirst discover basic mechanical benefits of adding different

amounts of nanoclay to adhesive systems. Also, the ability of

the nanoclay phase to work against crack propagation will be

explored, specifically in a mechanical interlocking adhesion situa-

tion, with adhesive having diffused into the porous granite

subsequently failing during fracture testing.

2. Materials and experimental procedures

2.1. Epoxies and curing agents

Two epoxy materials, one type of curing agent, and one type of

adherend material were studied in this work. Epon 815C and Epon830 are DGEBPA resins manufactured by Resolution Performance

Products in Houston, Texas, USA. Epon 815C is a low viscosity

epoxy resin that reacts with a wide range of curing agents. It is

made up of a mixture of 86.4% DGEBPA epoxy combined with 13.6%

monofunctional butyl glycidyl ether called Heloxy Modifier 61

(Resolution Performance Products, Houston, TX, USA). Heloxy

modifiers are reactive diluents that are either monofunctional to

minimize cost, or polyfunctional to maximize properties. Epon

815C is a good choice for adhesion applications as well as

encapsulating and coating processes. Epon 830 is made up of

100% DGEBPA resin, resulting in an above average viscosity, but

low crystallization tendency. This resins chemical resistance makes

it a good choice for industrial floor coatings and grouting and it is

typically the base resin for Heloxy-modified resins. The resin andmodifier properties are given in Table 1. The curing agent used was

Epi-cure 3223, an unmodified aliphatic amine diethylenetriamine

(DETA, Miller-Stephenson Chemical Co., Danbury, CT, USA).

2.2. Organically treated clay

The organically modified Clay was supplied by Southern Clay

Products in Gonzales, Texas. It is Montmorillonite clay modified

with methyl, tallow, bis-2-hydroxyethyl, quaternary ammonium

(Fig. 1). The clay is made up of approximately 1 nm thick layers

electrostatically held together in the form of micron sized

tactiods. Table 2 shows the properties of this clay.

2.3. Intercalation and curing procedures

The procedure for the formation of epoxyclay nanocompo-

sites using Cloisites 30B was based on the previous research [14].

Table 1

Properties of Epon series resin and modifiers.

Resin/

modifier

Epoxide equivalent

weight (g/eq)

Viscosity at

25 1C (P)

Density

(g/cm3)

Number average

molecular weight

815C 180195 57 1.13 700

830 190198 170225 1.16 700

Modifier

61

145155 12 0.91

E. Sancaktar, J. Kuznicki / International Journal of Adhesion & Adhesives 31 (2011) 286300288

7/28/2019 1-s2.0-S0143749611000194-main

4/15

The following procedure was used:

1. The epoxy resin and clay were stirred at 1000 rpm by magnetic

stirrer for 6 h at 90 1C.

2. The mixture was placed in a vacuum oven for 30 min at 80 1C.

This was found to optimize intercalation.

3. The mixture was cooled to room temperature and held for 15 min.

4. A stoichiometric amount of curing agent (see Tables 3 and 4)

was added to the resin and clay mixture and mixed by handfor 5 min at room temperature.

5. The resin-clay-curing agent mixture was degassed in a room

temperature vacuum oven for 510 min to remove air bubbles.

6. The mixture was placed in a dog bone cast mold, film mold, or

applied to granite fracture samples.

7. The system was cured at the temperature and time suggested

by the manufacturer for the application.

The ratios and formulations for a 20 g sample for both epoxy

systems at various clay loadings are given in Tables 3 and 4. Cure

conditions and procedures are shown in Tables 5 and 6. The film

samples were gelled at the desired thickness at room temperature

in a compression mold as a first step. All other samples were

cured in one step.

2.4. X-ray characterization

One-dimensional wide-angle X-ray diffraction measurements

of the epoxyclay nanocomposites were carried out on Rigaku

X-ray diffractometer (Tokyo, Japan) with 1.54 A (Cu) wavelength,

50 kV tube voltage, and 150 mA tube current. The 2-theta scan

range 1.581 at an interval of 0.0501 at a rate of 1.001/minute.

Both transmittance and reflectance spectrums were collected.

Samples of cured epoxyclay nanocomposites at various loadings

were tested to determine the interlayer spacing of clay layers and

the extent of intercalation or exfoliation.

2.5. TEM characterization

The transmission electron microscope (TEM) samples were

prepared by microtoming in a Reichert Ultracut Microtome

(Depew, NY, USA) at room temperature. The samples were then

analyzed in an FEI model TANCNAI-12 TEM (Hillsboro, OR, USA),

using 300 mesh copper grids. TEM images were used to confirm

findings from X-ray diffraction on the degree of intercalation and

exfoliation, as well as the morphology of the dispersed clay phase.

2.6. Tensile testing

Bulk tensile test specimens were prepared at clay loadings of

0%, 0.5%, 1%, 2%, and 4% to determine the mechanical properties.

The epoxy systems were cast in aluminum dog bone tensile

specimen molds and cured at the cure conditions shown

in Tables 5 and 6. The mold was coated with Teflon-based mold

release agent to aid in demolding. The samples were removed and

cooled for 1 h and then sanded. The samples were then tested onan Instron 5567 machine (Canton, MA, USA), using a 1 kN load cell

and a crosshead speed of 5 mm/min. Fig. 2 shows the dimensions

of the dog bone sample. The sample thickness was around

1.5 mm, but some variations occurred in thickness and width

due to sanding of the samples to remove flash and other edge

effects. The sanding procedure started with 150 grit power

sanding, followed by 300 and 400 grit hand sanding.

2.7. Adherend preparation and adhesive bonding

The natural solid granite adherend samples were prepared for

fracture testing by first cutting them with a Work Force CTC550

wet saw (Atlanta, GA, USA) into rectangular specimens of 50 mm

length, 30 mm width, and 10 mm thickness. The samples wererinsed with water, cleaned with compressed air, then rinsed with

acetone and allowed to dry. Double lap joint specimens (Fig. 3)

were prepared using an overlap length of 12.0 mm with an initial

crack length of 7.0 mm. These dimensions provided adhesive

failure rather than failure of the adherend during initial trials.

The samples were clamped under a pressure of 0.550 MPa during

CH2CH2OH

CH3 N+ (tallow)

CH2CH2OH

Fig. 1. Chemical structure of methyl, tallow, bis-2-hydroxyethyl, and quaternary

ammonium.

Table 2

Properties of Cloisites 30B clay.

Clay type Modifier

Concentration

(micro- equivalent/

10 g clay)

Specific gravity X-ray diffraction

results (A)

Typical dry particle sizes (mm)

10% Less than 50% Less than 90% Less than

Cloisites 30B 90 1.98 18.5 2 6 13

Table 3

Formulation of Epon 815C, DETA, and 30B.

Weight percentages (%) 20 g Sample formulation (g)

30B

clay

815C

Resin

DETA curing

agent

30B

clay

815C

Resin

DETA curing

agent

0.0 89.00 11.00 0.00 17.80 2.20

0.5 88.56 10.95 0.10 17.71 2.19

1.0 88.11 10.89 0.20 17.62 2.18

2.0 87.22 10.78 0.40 17.44 2.16

4.0 85.44 10.56 0.80 17.09 2.11

Table 4

Formulation of Epon 830, DETA, and 30B.

Weight percentages (%) 20 g Sample formulation (g)

30B

clay

830

Resin

DETA curing

agent

30B

clay

830

Resin

DETA curing

agent

0.0 90.17 9.83 0.00 18.03 1.97

0.5 89.72 9.78 0.10 17.94 1.961.0 89.27 9.73 0.20 17.85 1.95

2.0 88.37 9.63 0.40 17.67 1.93

4.0 86.56 9.44 0.80 17.31 1.89


7/28/2019 1-s2.0-S0143749611000194-main

5/15

cure to ensure uniform adhesive thickness of around 0.08 mm and

adequate bonding of the substrates. The various formulations of

epoxy and clay were prepared and applied to the areas to be

bonded. The initial crack was formed using a double-edged razor

blade coated with a UV-cured polysiloxane coating, and the

starter blade was removed after curing. The initial crack length

was determined through experimentation and served to initiate

fracture failures.

2.8. Fracture testing

The specimens for fracture tests were prepared as described

in Section 2.7. A wedge type fracture method was developed and

employed using a wedge with a vertex angle of 46.161. The wedge

was placed in between the double side of the specimen and a

compressive load was applied until failure. The use of a 50 kN

load cell was found necessary based on the greater of the crushing

force of the granite substrate and shear strength of the epoxy. The

fracture toughness was approximated by load at fracture. After

considerable testing, the load cell required for dimensions in thiswork was found to be 30 kN. The crosshead speed of 5 mm/min

was used based on the ASTM D3931 method [28]. This test (Fig. 4)

was used to find the load at the catastrophic crack propagation in

an attempt to assess the fracture properties of the epoxy systems

on granite.

2.9. Film sample preparation for TEM analysis

The film samples were cast using Teflon and PET compression

mold fixture, illustrated in Fig. 5. The prepared uncured filled and

neat epoxy systems were poured into the compression mold and

allowed to gel, while under 34.5 MPa pressure to ensure an

average thickness of 0.15 mm. The materials gelled at room

temperature were then removed and cured at 1201

C for 3 h for

Epon 815C, and at 100 1C for 1 h for Epon 830. Since certain

volatiles may escape, especially during elevated temperature

curing, efforts were made to make the film casting process

comparable to the fracture and tensile test specimens. Thus, thecuring was done after the removal of the top Teflon layer, so that

the gelled film was stable enough not to adversely affect the film

thickness uniformity.

Once the film samples were cured they were subjected to

X-ray diffraction and TEM analysis.

3. Results and discussion

3.1. Interpretation of X-ray and TEM findings

Various samples were tested in an attempt to determine the

interlayer distances of the dispersed clay tactiods and subse-

quently the degree of intercalation or exfoliation in the 815C-

DETA system. First, clays that were treated by the manufacturer

and clays treated in our laboratory were compared in their

powdered form. Fig. 6 shows clay 30B to have peak near 4.91representing a d-spacing of 1.80 nm (Fig. 6b). This shows good

agreement with what was reported by the manufacturer

(Table 2). We also treated the clay, which was supplied by

Southern Clay Products Co. in an untreated form, using hexade-

cylamine, and obtained d-spacing of 1.84 nm (Fig. 6a). We expect

little difference in performance of nanocomposites filled with clay

30B or the clay we treated using hexadecylamine, due to small

difference in interlayer distances achieved. Consequently the

pretreated 30B clay was used for the remainder of the work.

Fig. 7 shows the X-ray results for various mixing and curing

procedures tested to confirm their effectiveness. The variousconditions used are outlined in Table 7, with all having 4% weight

Cloisites 30B. Condition 7a is described in [14]. With condition

7b, the mixing of the epoxy with organoclay was doubled to 12 h

with rest of the procedure being the same as that for condition 7a.

Condition 7c represents a system using a two-stage curing

procedure recommended by the manufacturer of the resin and

the curing agent. For condition 7d, the resin and clay were mixed

for the standard time, but extensive hand mixing and subsequent

ultrasonic agitation were used after the addition of the curing

agent. The final condition (7e) depicted is a system using 24%

weight acetone added to the clay as an initial swelling agent prior

to magnetic mixing.

Fig. 8 shows more details of the effect of the acetone on the

clay, showing curves for the clay only (Fig. 8b), clay swelled with

Table 5

Cure conditions and procedures for Epon 815C, DETA, and 30B.

815C/DETA/30B Step 1 Step 2

Specimen type Time (h) Temperature (1C) Pressure (MPa) Time (h) Temperature (1C) Pressure (MPa)

Dog bone 3 120 0.101

Fracture 3 120 0.550

Film 4 26 34.5 3 120 0.101

Table 6

Cure conditions and procedures for Epon 830, DETA, and 30B.

830/DETA/30B Step 1 Step 2

Specimen type Time (h) Temperature (1C) Pressure (MPa) Time (h) Temperature (1C) Pressure (MPa)

Dog bone 1 100 0.101

Fracture 1 100 0.550

12.7 mm

47.0 mm

13.0 mm5.0 mm

Fig. 2. Dimensions of tensile specimen.


7/28/2019 1-s2.0-S0143749611000194-main

6/15

acetone (8c), and the nanocomposite system formed (8a). Exam-

ination of Figs. 7 and 8 leads to the following observations:

The first four curing and mixing conditions (7a, b, c, and d)

showed small differences in the degree of intercalation, as

shown by the interlayer spacings measured.

Blade is removed to form

initial crack of 7.0 mmEpoxy thickness

of 0.08 mm

12.0 mm

50.0 mm

30.0 mm

10.0 mm

5.0 mm

Fig. 3. Diagram of double lap joint.

2 = 46.16

FMode I = (F/2)*tan ()

F

FModeII = F/2

Fig. 4. Force diagram of wedge and double lap sample used in fracture testing

showing the force components acting on the cover plates to induce fracture.

Teflon

Teflon

PET

2.0 mm

0.15 mm

Fig. 5. Diagram of mold used for film casting.

1.5

2 Theta

RelativeIntensity(a.u.)

4.10 nm

4.20 nm

4.20 nm

(a)

(b)

(c)

(d)

(e)

4.41 nm

2.5 3.5 4.5 5.5 6.5 7.5

Fig. 7. X-ray diffraction patterns for 815C-DETA-4% 30B epoxy system at various

curing and mixing conditions: (a) standard procedure; (b) 2 xs epoxy-clay mixing

time; (c) 2-step cure; (d) 7 xs epoxy-clay-curing agent mixing time; and

(e) acetone added.

1.5

2 Theta (Degree)

Rea

ltive

Intens

ity

(a.u.

)1.84 nm

1.80 nm

(a)

(b)

2.5 3.5 4.5 5.5 6.5 7.5

Fig. 6. X-ray diffraction patterns for (a) Cloisites Na+ treated with hexadecyla-

mine and (b) Cloisites 30B.


7/28/2019 1-s2.0-S0143749611000194-main

7/15

The additional 6 h of mixing in condition 7b did not show

substantial benefit, only decreasing the interlayer distance

from 4.41 to 4.20 nm.

The two-stage curing (7c) using a room temperature stage and

an elevated temperature stage resulted in a decrease in an

interlayer distance (4.10 nm) compared to the system cured inone stage at an elevated temperature (4.41 nm). This is

probably because, in the single stage, the temperature

increases early in the curing process, thus reducing viscosity

of the system and lowering the extra-gallery viscous forces

that work against the exfoliation process (see Fig. 9) [14].

The extensive mixing and ultrasonic agitation of the system as it

gelled, also did not provide intercalating benefit (condition 7d). A

similar effect is seen as in condition 7c; the increase of viscosity

as the system gelled actually worked to hinder the effectiveness

of the exfoliating elastic forces inside the clay galleries, despite

the additional mixing. Although condition 7d showed less inter-

layer spacing than the standard system 7a (4.20 compared to

4.41 nm), it had higher gallery height compared to the two-stage

curing process of condition 7c (4.20 compared to 4.10 nm). This

is probably because the system cured (or rather gelled) for

35 min rather than 3 h as in the two-stage cure, and the

additional shear mixing perhaps aided in intercalation.

Due to the increase of viscosity as the system gelled, mixing

past the gel point in condition 7d increased the likelihood ofair bubbles being trapped in the system.

The greatest interlayer spacing for 4% weight clay loading was

found in condition 7e. The addition of acetone as a swelling

agent showed good X-ray results compared to all of the other

conditions in Table 7 and Fig. 7. The reason is best seen

in Fig. 8. The addition of the organic solvent swells the

organoclay from 1.80 to 4.77 nm (8b8c). This mixture, mag-

netically-stirred into the resin, results in a nanocomposite

with a peak shifting far to the left indicating a d-spacing of

greater than 5.88 nm (8a and 7e).

While the addition of acetone to the system proved to increase

the intercalation height of the 4% clay loaded nanocomposites,

subsequent mechanical testing proved that residual acetonemolecules in the epoxy system acted as inhibitors to crosslinking,

resulting in a decrease in crosslink density and consequently

decrease in mechanical properties. This decrease outweighed the

benefits of the increased d-spacing.

The standard mixing procedure in condition 7a was thus

chosen as the optimal procedure among those considered. Next,

X-ray measurements were taken of samples with the clay loading

varying from 0% to 4% as shown in Fig. 10:

The results showed larger d-spacing for the nanocomposites

with lower clay loadings. This is most likely because having

less clay particles increases the chance that resin molecules

will intercalate well into the clay galleries. TEM images

(discussed later) show that the smaller-sized aggregates of

Table 7

Various curing and mixing conditions used on 815C-DETA-4% 30B epoxy system.

Sample Mixing condition (with 4% 30B) Curing condition X-ray result (nm)

Resin/clay Resin/clay/curing

4.1.2a 6 h 90 1C (magnetic stirrer) 5 min 25 1C (hand mix) 3 h, 120 1C 4.41

4.1.2b 12 h 90 1C (magnetic stirrer) 5 min 25 1C (hand mix) 3 h, 120 1C 4.20

4.1.2c 6 h 90 1C (magnetic stirrer) 5 min 25 1C (hand mix) 3 h, 25 1C + 4.10

2 h, 1501C

4.1.2d 6 h 90 1C (magnetic stirrer) 35 min 25 1C (hand mix+ultras onic) 3 h , 12 0 1C 4.01

4.1.2e 6 h 90 1C with acetone (magnetic stirrer) 5 min 25 1C (hand mix) 3 h, 120 1C 45.88

1.5

2 Theta (degree)


1.80 nm

(a)

(b)

(c)

4.77 nm

2.5 3.5 4.5 5.5 6.5 7.5

Fig. 8. X-ray diffraction patterns for (a) 815C-DETA-4% 30B-acetone nanocompo-

site, (b) 30B clay, and (c) 30B clay swelled with acetone.

van der Waals

Electrostatic forceElastic force

Extra-gallery viscous force

Extra-gallery viscous force

(Clay layer)

Fig. 9. Schematic representation of forces acting on a pair of clay layers during

intercalation and exfoliation.

1.5

2 Theta (degree)

Rela

tiveIntensity(a.u.) 4.64 nm

5.19 nm

(a)

(b)

(c)

(d)

(e)

(f)

4.41 nm

2.5 3.5 4.5 5.5 6.5 7.5

Fig. 10. X-ray diffraction patterns for 815C-DETA-30B epoxy system at various

clay loadings: (a) 4%; (b) 2%; (c) 1%; (d) 0.5%; (e) 0%; and (f) air.


7/28/2019 1-s2.0-S0143749611000194-main

8/15

clay in the lower weight percent samples also seem to have

larger d-spacing. The near featureless curve of the 0.5% nanocomposite

(Fig. 10d) is a good indication of the exfoliation of the clay

tactiods to a distance greater than 5.88 nm. Thus, best barrier

and mechanical properties should be expected at this loading.

The curves are also compared to air (10f) and neat epoxy (10e)

as baselines.

Some Epon 815C systems prepared using different curing and

mixing procedures were imaged with TEM to further investigate

the morphology of the clay and epoxy already investigated by the

X-ray diffraction. Figs. 11 and 12 show TEM images of the 2% clay

loaded system at different magnifications. Figs. 13 and 14 show

the system with 0.5% weight clay at the same magnification for

comparison. The TEM samples were collected by microtoming

Fig. 11. Low magnification TEM image of 815C/DETA with 2% weight clay loading,

showing typical size of a clay aggregate (18,500 magnification).

Fig. 12. High magnification TEM image of 815C/DETA with 2% weight clay loading,

showing mostly intercalated nanocomposite structure (195,000 magnification).

Fig. 13. Low magnification TEM image of 815C/DETA with 0.5% weight clay

loading, showing typical size of a clay aggregate (18,500 magnification).

Fig. 14. High magnification TEM image of 815C/DETA with 0.5% weight clay

loading, showing exfoliated nanocomposite structure (195,000 magnification).

1.5

2 Theta (degree)


5.04 nm

5.04 nm

(a)

(b)

(c)

(d)

(e)

5.04 nm

5.04 nm

2.5 3.5 4.5 5.5 6.5 7.5

Fig. 15. X-ray diffraction patterns for 830-DETA-30B epoxy system at various clay

loadings: (a) 4%; (b) 2%; (c) 1%; (d) 0.5%; and (e) air.


7/28/2019 1-s2.0-S0143749611000194-main

9/15

along the cross-section of the film samples created in the manner

stated in Section 2.9. The TEM results showed as follows:

The TEM images seem to agree relatively well with the

interlayer spacing found in X-ray diffraction results. TEM

shows areas of both intercalation and exfoliation, with exfo-liated and intercalated clay aggregate structures similar to

those found by Park and Jana and others (Figs. 11 and

13) [14,29].

As suggested earlier by the X-ray diffraction, it appears that

there is a slightly better exfoliation of the clay layers in 815C

samples with less clay loading. TEM images showed that

typically, 0.5% weight clay content samples had smaller

aggregates compared to the 2% loaded samples (Fig. 11 com-

pared to Fig. 13). This smaller aggregate size results in less

interference of clay tactiods, and consequently better exfolia-

tion as seen in Fig. 14 compared to 12.

For the most part, the clay layers seemed to remain parallel

even for the samples with large d-spacing. This morphology

has been shown to be beneficial for creating zones of shear

Fig. 16. Low magnification TEM image of 830/DETA with 2% weight clay loading,


Fig. 17. High magnification TEM image of 830/DETA with 2% weight clay loading,

showing mostly intercalated nanocomposite structure (195,000 magnification).

Fig. 18. Low magnification TEM image of 830/DETA with 0.5% weight clay loading,


Fig. 19. High magnification TEM image of 830/DETA with 0.5% weight clay

loading, showing mostly an intercalated nanocomposite structure (195,000

magnification).

0

200

400

600

800

1000

1200

1400

1600

0

Clay Loading (wt %)

Young'sModulus(MPa)

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

Fig. 20. Variation of Youngs modulus for 815C-DETA at different clay loadings.


7/28/2019 1-s2.0-S0143749611000194-main

10/15

yielding, and consequently increased rupture strain and

toughness [29,30].

Varying degrees of intercalated and exfoliated nanocompo-

site areas seem to form in both the samples with 0.5% and

2% weight clay. It is reasonable to assume that this

trend occurs for systems at other clay loadings as well, so

various results for mechanical and barrier properties are

expected.

50

55

60

65

70

75

0

Clay Loading (wt %)

Max

Stress

(MPa

)

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

Fig. 21. Variation of maximum stress for 815C-DETA at different clay loadings.

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

Clay Loading (wt %)

MaxStrain(m/m)

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

Fig. 22. Variation of maximum strain for 815C-DETA at different clay loadings.

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0

Clay Loading (wt %)

Strain

atMaximumS

tress(m/m)

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

Fig. 23. Variation of strain at maximum stress for 815C-DETA at different clay

loadings.

0.0

5.0

10.0

15.0

20.0

25.0

30.0

35.0

40.0

45.0

50.0

0

Clay Loading (wt %)

YieldS

tress

(MPa

)

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

Fig. 24. Variation of offset yield stress for 815C-DETA at different clay loadings.

0

1

2

3

4

5

6

7

8

0

Clay Loading (wt %)

Toug

hness

(MPa

)

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

Fig. 25. Variation of toughness, measured as an area under the stressstrain curve

for 815C-DETA, at different clay loadings.

Fig. 26. Digital photograph of 815C-DETA-30B tensile tests specimens at various

clay loadings. The samples were stressed to failure, with the 0% sample showing

some necking and the 1% sample showing stress whitening and some necking, as

observed at upper failure areas in the inset photos.


7/28/2019 1-s2.0-S0143749611000194-main

11/15

The higher viscosity 830-DETA epoxy system was also ana-

lyzed with X-ray diffraction and TEM and the results are shown in

Figs. 1519. Examination of these figures reveals the following

findings:

Compared to the 815C X-ray findings, the 830 system showed

less dramatic of a leftward shift in peaks as the weight

percentage was decreased as seen in Fig. 15. The best results

seem to be from the 0.5% clay loaded system as it has a more

featureless curve (Fig. 15d), but an indication of a peak near

5.04 nm suggests that the degree of exfoliation is not as good

as in the 0.5% loaded 815C system (Fig. 10d). Although the

0

200

400

600

800

1000

1200

1400

1600

0

Clay Loading (wt %)

Young's

Mo

du

lus

(MPa

)

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

Fig. 27. Variation of Youngs modulus for 830-DETA at different clay loadings.

50

55

60

65

70

75

80

85

0

Clay Loading (wt %)

Max

Stress

(MPa

)

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

Fig. 28. Variation of maximum stress for 830-DETA at different clay loadings.

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0

Clay Loading (wt %)

Max

Stra

in(m/m)

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

Fig. 29. Variation of maximum strain for 830-DETA at different clay loadings.

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0

Clay Loading (wt %)

StrainatMa

ximumS

tress(m/m)

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

Fig. 30. Variation of strain at maximum stress for 830-DETA at different clay

loadings.

0.0

5.0

10.0

15.0

20.0

25.0

30.0

35.0

40.0

45.0

0

Clay Loading (wt %)

YieldStress

(MPa

)

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

Fig. 31. Variation of offset yield stress for 830-DETA at different clay loadings.

0

1

2

3

4

5

6

7

8

9

0

Clay Loading (wt %)

Toug

hness

(MPa

)

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

Fig. 32. Variation of toughness, measured as an area under the stressstrain curve

for 830-DETA, at different clay loadings.


7/28/2019 1-s2.0-S0143749611000194-main

12/15

system does appear to intercalate well, increasing the clay

layer distance up to 5.04 nm, curves would also seem to

indicate that exfoliation is less likely or at least more difficult

in the higher viscosity system.

The 830 system followed the same trend as the 815C system,

with the lower loading of clay resulting in a smaller aggregate

size as seen by comparing Figs. 18 and 16.

Higher magnification TEM images showed a lesser degree of

exfoliation and more areas of intercalated nanocomposite

formation at both 0.5% and 2% clay loading levels in the Epon

830 system. This was expected for the much higher viscosity

resin. Magnified images of typical intercalated structures can

be seen in Fig. 17 for the system with 2% weight clay and

in Fig. 19 for the 0.5% loaded system. In the 815C system,

having less clay particles in smaller aggregates worked to an

Displacement (mm)

CrackP

ropagationLoad(kN)

vertical displacement

during crack propagation

Fracture toughness

approximated by the

load at fracture

Fig. 33. Typical load versus displacement plot showing values for fracture load

and vertical displacement during crack propagation.

Failure location

Failure location

Fig. 34. Digital photograph of a bonded granite double lap sample showing typical

shearing failure of the granite substrate (a) and typical failure primarily in the

adhesive layer (b).

Table 8

Load and displacement values for 815C-DETA-30B fracture samples (standard

deviations shown in parentheses).

Clay

loading

(weight

%)

Fracture load

(kN)

FMode I(kN)

FMode

II (kN)

Vertical

displacement

during crack

propagation

(mm)

Mode I

displacement

during crack

propagation

(mm)

0 0.1787 (0.0095) 0.0381 0.0894 0.7833 (0.1259) 0.33380.5 0.2462 (0.0013) 0.0525 0.1231 1.1416 (0.0877) 0.4865

1 0.2009 (0.0026) 0.0428 0.1005 0.9275 (0.0293) 0.3952

2 0.1995 (0.0084) 0.0425 0.0997 0.8721 (0.1550) 0.3716

4 0.1846 (0.0124) 0.0393 0.0923 0.9748 (0.0413) 0.4154

Table 9

Load and displacement values for 830-DETA-30B fracture samples (standard

deviations shown in parentheses).

Clay

loading

(weight

%)

Fracture load

(kN)

FMode I(kN)

FMode

II (kN)

Vertical

displacement

during crack

propagation

(mm)

Mode I

displacement

during crack

propagation

(mm)

0 0.1843 (0.0052) 0.0393 0.0921 0.5915 (0.0506) 0.2521

0.5 0.1926 (0.0229) 0.0410 0.0963 0.7527 (0.0622) 0.3207

1 0.2127 (0.0177) 0.0453 0.1064 0.8474 (0.0375) 0.3611

2 0.2349 (0.0273) 0.0500 0.1174 0.8695 (0.0335) 0.3705

4 0.1256 (0.0131) 0.0268 0.0628 0.6691 (0.1836) 0.2851

0

0.05

0.1

0.15

0.2

0.25

0.3

0

Clay Loading (wt %)

Forcea

tca

tas

trop

hiccrac

kprop

aga

tion

(kN)

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

Fig. 35. Variation of load at catastrophic crack propagation for various clayloadings in the 815C-DETA epoxy system.


7/28/2019 1-s2.0-S0143749611000194-main

13/15

improved exfoliation, but in this case the impact of aggregate

size on exfoliation seemed to be less.

3.2. Analysis of tensile properties

Figs. 2025 show key mechanical properties for the Epon

815C-DETA system loaded with various weight percentages of

30B clay. All were cured at 120 1C for 3 h.X-ray and TEM seemed to indicate the best nanocomposite

formation at 0.5% clay loading. The mechanical properties seem to

agree with this; 0.5% organoclay loading on the 815C-DETA

system had the following effects:

The addition of clay increases Youngs modulus from

1184 MPa for the neat resin to 1317 MPa ( Fig. 20).

The maximum stress also increases from 67.49 to 73.19 MPa

(Fig. 21).

The addition of 0.5% weight clay embrittles the system. This

decreases the maximum strain from 0.1148 to 0.1078 mm/mm

(Fig. 22), but increases strain at maximum stress from 0.0823

to 0.1038 mm/mm (Fig. 23).

As expected, the more brittle system shows a higher yieldstress (45.5 MPa, Fig. 24) but a slightly lower toughness

(5.25 MPa, Fig. 25), measured as area under the stressstrain

curve.

The addition of 1% weight 30B clay seems to cause some

interesting property changes in the epoxy system:

An increase in rupture strain from 0.1148 to 0.1405 mm/mm

and corresponding increase in toughness from 5.38 to

7.30 MPa accompanies this nanocomposite (Fig. 25). The

reason for this is attributed to the shear yielding and toughen-

ing sometimes observed in clay nanocomposites, retaining

some orientation of the clay layers [29,30]. This is further

confirmed by the presence of stress whitening and somenecking in the 1% sample as seen in Fig. 26.

When the clay serves to aid in shear yielding, it cannot act to

embrittle the system as in the 0.5% loaded sample. Further-

more, no significant increase is observed in the modulus or

yield stress for the 1% sample (Figs. 20 and 24, respectively),

and a less dramatic increase in maximum stress from 67.49 to

70.94 MPa (Fig. 21).

The addition of clay greater than 1% seems to have no positive

effect on mechanical properties. In fact, all key properties except

for Youngs modulus seem to steadily decrease with the increased

addition of clay. The Youngs modulus values show an initial

increase at 0.5% clay loading, but they drop down to a level

approximately equal to that for the neat resin at higher clayloadings (Fig. 20). It is possible that adding high amounts of

nanoclay increases the organic content to the level, where it acts

to plasticize the epoxy network and decrease the crosslink

density. Thus, an optimal point of clay loading is observed for

mechanical properties, and in the 815C-DETA-30B system this

appears to be around 0.5% clay loading.

3.2.1. The effect of resin viscosity

The effectiveness of nanocomposite formation was explored

on an epoxy system of much higher viscosity (Epon 830-DETA).

Essentially, the two epoxy systems are chemically very similar,

with the viscosity reduction of 815C achieved by the addition

of reactive diluent of butyl glycidyl ether. This small reactive

molecule effectively reduces the viscosity from around 170225 P

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0

Clay Loading (wt %)

Verticaldisplacementduringcrack

propagation(mm)

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

Fig. 36. Variation of vertical wedge displacement during crack propagation for

various clay loadings in the 815C-DETA epoxy system.

0

0.05

0.1

0.15

0.2

0.25

0.3

0

Clay Loading (wt %)

For

ceatcatastrophiccrackpropagation(kN)

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

Fig. 37. Variation of load at catastrophic crack propagation for various clay

loadings in the 830-DETA epoxy system.

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0

Clay Loading (wt %)

Vert

ica

ldisp

lacemen

tdu

ringcrac

k

propaga

tion

(mm

)

0.5 1 1.5 2 2.5 3 3.5 4 4.5

Fig. 38. Variation of vertical wedge displacement during crack propagation for

various clay loadings in the 830-DETA epoxy system.


7/28/2019 1-s2.0-S0143749611000194-main

14/15

(Epon 830) to 57 P (Epon 815C) and seems to aid in the

intercalation process as the nanocomposites elastic behavior

seems to have been extended with the lower viscosity resin

system. The mechanical properties of the higher viscosity

830-DETA-30B system are plotted in Figs. 2732. Examination

of these figures reveals the following findings:

The higher viscosity 830-DETA system showed less elastic

behavior compared to the 815C-DETA system at every clayloading level with lower yield stress for the Epon 830 system

compared to the Epon 815C.

The higher viscosity system showed higher modulus, higher

maximum stress, and higher rupture strain at nearly every clay

loading level.

The higher viscosity Epon 830 system had the best modulus at

a clay loading of 0.5% and the best yield stress at 2% clay

loading (Figs. 27 and 31, respectively).

In comparison to the neat sample, toughness (measured as an

area under the stressstrain curve) decreased at all clay levels

with only the 2% weight sample providing toughness value

close to that for the neat sample (Fig. 32).

The overall impact of the clay on key mechanical properties of

the 830-DETA-30B systems was less than the impact it had on

the key properties of the 815C-DETA-30B system. This is

attributed to the less effective intercalation and exfoliation

(and thus less effective nanocomposite formation) in the 830

system, due to higher viscosity and a higher molecular weight.

3.3. Assessment of fracture results

Different configurations of the double lap fracture test speci-

men (Fig. 4) were investigated to ensure failure primarily through

the adhesive layer. A crack length of 7 mm was used and the

overlap length was varied from 15 to 12 mm, resulting in load

versus displacement plots of the type shown in Fig. 33. At larger

overlap lengths, shear failure of the granite stone substrate was

observed (see Fig. 34a), so the overlap length was reduced until

failure through epoxy was observed (Fig. 34b). The optimal

overlap length of 12 mm was then used on the two epoxy systems

(815C and 830) at various clay loadings to observe the effect of

clay on the fracture toughness of the epoxy bonded to granite

(approximated by the load at fracture). The fracture test results

are tabulated in Tables 8 and 9. The tables include the fracture

load and vertical wedge displacement during crack propagation

(see Fig. 33) as well as the Mode I displacement perpendicular to

the crack propagation and the force components (FM ode I and

FMode II as seen in Fig. 4) on the granite substrate cover plates. The

load at catastrophic crack propagation and the vertical wedge

displacement during crack propagation for the 815C-DETA system

are plotted in Figs. 35 and 36 and in Figs. 37 and 38 for the 830-

DETA system. Based on these figures the following observationscan be made:

All systems showed brittle failure through the epoxy layer.

The fracture tests seem to reveal results analogous to the

tensile tests. With the higher viscosity 830-DETA resin system,

the advantage of clay addition seems to be less in comparison

to the similarly filled lower viscosity system (Figs. 35 and 37).

The load at failure and total vertical displacement during crack

propagation was highest for the 815C-DETA system (Figs. 35

and 36) at 0.5% 30B clay loading. This clay loading also

provided the largest modulus, maximum stress, and yield

stress in the tensile testing (Figs. 20, 21, and 24), but not the

highest toughness (Fig. 25). This also indicates that although

the clay addition provided maximum toughening properties in

the tensile tests with 1% by weight in 815C-DETA adhesive, it

may not be effective as a crack stopping toughening agent at

this loading level when bonding porous stone substrates.

The addition of clay to the 830 system again seems to show

less of an effect in the higher viscosity resin. With the best

failure load and maximum vertical displacement being

observed around 2% clay loading. This correlates the tensile

test data, with the clay loading levels for high values of

modulus, maximum stress, yield stress and toughness(Figs. 27, 28, 31 and 32) matching the clay loading levels for

maximums in the fracture data (Figs. 37 and 38).

The fracture results reveal the 815C system with larger values

for vertical displacement at nearly every clay loading level

compared to the 830 system.

The difference in viscosity between the two systems seemed to

have minimal effects on the macro scale. Penetration into the

granite pours seem to be similar, probably due to the high

temperature cure causing a reduction in viscosity to a similar

level for both systems at early stages of curing.

4. Conclusions

Between the higher and lower viscosity resins, the betterlayered silicate nanocomposite performance was seen with the

lower viscosity 815C resin. X-ray and TEM images confirm the

levels of exfoliation for each resin and at each clay loading. The

optimal level of nanoclay loading seems to be around 0.5% for

Epon 815C and 2% for Epon 830 for mechanical properties. The

most noticeable improvements in mechanical properties for the

815C system were found to be maximum stress, elastic modulus,

and yield stress. Increased toughness and stress whitening at 1%

loading are evidence that the clay can act as a shear-yielding

toughening agent in this epoxy system. Nanocomposite formation

in the 830 resin appears to be hindered by its higher viscosity

compared to the 815C, showing dramatic improvements in only

yield stress and a slight improvement in maximum stress.

Fracture analysis of the samples seemed to reveal results

similar to those from tensile testing. The addition of clay

improved the fracture load for the samples similar to the way

maximum stress was increased in tensile properties for both Epon

815C and 830 resin-based nanocomposites. One inconsistency

was that for the lower viscosity resin Epon 815C-based nano-

composite, the fracture toughness, which we approximated by

load at fracture, did not show a significant increase at the 1% clay

level as in the case of tensile toughness. This suggests that, when

the lower viscosity resin is used, while the clay may be an

effective toughening agent in regard to shear yielding with 1%

loading, its ability to stop cracks in Mode I situations is less

effective at the same level of clay loading. TEM imaging reinforced

this idea revealing areas of intercalation, which may be sites of

shear yielding. The fracture results showed failure of the

mechanically interlocked adhesive layer. Therefore, we believethat the increase in the bulk adhesive strength provided by the

addition of clay resulted in improved adhesive joints of brittle,

porous (granite) substrates.

References

[1] Vaia RA, Giannelis EP. Lattice model of polymer melt intercalation inorganically-modified layered silicates. Macromolecules 1997;30:79909.

[2] Vaia RA, Giannelis EP. Polymer melt intercalation in organically-modi-fied layered silicates: model predictions and experiment. Macromolecules1997;30:80009.

[3] Lan T, Pinnavaia TJ. Clay-reinforced epoxy nanocomposites. Chem Mater1994;6:22169.

[4] Gilman JW, Kashiwagi T. Nanocomposites: a revolutionary new flame

retardant approach. SAMPE J 1997;33:406.


7/28/2019 1-s2.0-S0143749611000194-main

15/15

[5] Lan T, Kaviratna PD, Pinnavaia TJ. Mechanism of clay tactoid exfoliation inepoxy-clay nanocomposites. Chem Mater 1995:214450.

[6] Lan T, Wang Z, Shi H, Pinnavaia TJ. Clayepoxy nanocomposites: relationshipsbetween reinforcement properties and the extent of clay layer exfoliation.Polym Mater Sci Eng 1995;73:2967.

[7] Shi H, Lan T, Pinnavaia TJ. Interfacial effects on the reinforcement propertiesof polymerorganoclay nanocomposites. Chem Mater 1996;8:15847.

[8] Wang Z, Massam J, Pinnavaia TJ. Epoxyclay nanocomposites. In: PinnavaiaTJ, Beall GW, editors. Polymerclay nanocomposites. Hoboken, New Jersey:

John Wiley & Sons Ltd.; 2000. p. 12749.[9] Wang Z, Pinnavaia TJ. Hybrid organicinorganic nanocomposites: exfoliation

of magadiite nanolayers in an elastomeric epoxy polymer. Chem Mater1998;10:18206.

[10] Wang Z, Lan T, Pinnavaia TJ. Hybrid organicinorganic nanocompositesformed from an epoxy polymer and a layered silicic acid (magadiite). ChemMater 1996;8:22004.

[11] Rana HT. Moisture diffusion through neat and glass-fiber reinforced vinylester resin containing nanoclay. M.S. Thesis Department of Chemical Engi-neering, West Virginia University 2003.

[12] Rana HT, Gupta RK, Gangarao HVS, Sridhar LN. Measurement of moisturediffusivity through layered-silicate nanocomposites. AICHE J 2005;51:324956.

[13] Sancaktar E, Kuznicki J. Stress-induced reduction of water uptake in clay-reinforced epoxy nanocomposites. Curr Nanosci 2006;2:3517.

[14] Park JH, Jana SC. Mechanism of exfoliation of nanoclay particles in epoxyclay nanocomposites. Macromolecules 2003;36:275868.

[15] Park JH, Jana SC. A case study on the effects of plasticization of epoxynetworks by organic treatment on exfoliation of nanoclay. Macromolecules2003;36:83917.

[16] Park JH, Jana SC. The relationship between nano- and micro-structures andmechanical properties in PMMAepoxynanoclay composites. Polymer2003;44:2091100.

[17] Giannelis EP. Polymer layered silicate nanocomposites. Adv Mater 1996;8:2935.

[18] Messersmith PB, Giannelis EP. Synthesis and characterization of layered

silicateepoxy nanocomposites. Chem Mater 1994;6:171925.[19] Messersmith PB, Giannelis EP. Synthesis and barrier properties of poly(e-

caprolactone)-layered silicate nanocomposites. J Polym Sci, Part A: Polym

Chem 1995;33:104757.[20] Messersmith PB, Giannelis EP. Polymer-layered silicate nanocomposites: in-

situ intercalative polymerization of e-caprolactone in layered silicates. Chem

Mater 1993;5:10646.[21] Lee JY, Baljon ARC, Loring RF, Panagiotopoulos AZ. Modeling intercalation

kinetics of polymer silicate nanocomposites. Mater Res Soc Symp Proc

1999;543:13135.[22] Krishnamoorti R, Giannelis EP. Rheology of end-tethered polymer layered

silicate nanocomposites. Macromolecules 1997;30:4097102.[23] Vaia RA, Ishii H, Giannelis EP. Synthesis and properties of two-dimensional

nanostructure by direct intercalation of polymer melts in layered silicates.

Chem Mater 1993;5:16946.[24] Vaia RA, Teukolsky RK, Giannelis EP. Interlayer structure and molecular

environment of alkylammonium layered silicates. Chem Mater 1994;6:

101722.[25] Sancaktar E, Aussawasathien D. Nanocomposites of epoxy with electrospun

carbon nanofibers: mechanical behavior. J Adhes 2009;85:16079.[26] Sancaktar E, Aussawasathien D. Electrospun polyacrylonitrile-based carbon

nanofibers and their silver modifications: surface morphologies and proper-

ties. Curr Nanosci 2008;4:1307.[27] E. Sancaktar, D. Aussawasathien, Effect of Non-Woven Carbon Nanofiber Mat

Presence on Cure Kinetics of Epoxy Nanocomposites, Macromol Symp 264

(2008) 2633.[28] /http://webstore.ansi.org/RecordDetail.aspx?sku=ASTM+ D3931-08S; 2010.[29] Liu T, Tjiu WC, Tong Y, He C, Goh SS. Morphology and fracture behavior of

intercalated epoxy/clay nanocomposites. J Appl Polym Sci 2004;94:123644.[30] Ratna D, Becker O, Krishnamurthy R, Simon GP, Varley RJ. Nanocomposites

based on a combination of epoxy resin, hyperbranched epoxy and a layered

silicate. Polymer 2003;44:744957.

http://webstore.ansi.org/RecordDetail.aspx?sku=ASTM+D3931-08http://webstore.ansi.org/RecordDetail.aspx?sku=ASTM+D3931-08http://webstore.ansi.org/RecordDetail.aspx?sku=ASTM+D3931-08http://webstore.ansi.org/RecordDetail.aspx?sku=ASTM+D3931-08

1-s2.0-S0143749611000194-main

Documents

Transcript of 1-s2.0-S0143749611000194-main