Computational Materials Science 81 (2014) 612–616

Contents lists available at ScienceDirect

Computational Materials Science

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

An analysis of interfacial adhesion in nanocomposites from recycled

polymers
Yasser Zare ⇑, Ali Daraei, Mostafa Vatani, Parya Aghasafari
Department of Mechanics, Tiran Branch, Islamic Azad University, Tiran, Iran

a r t i c l e i n f o a b s t r a c t

Article history: The present paper studies the quantitative analysis of interfacial adhesion in the nanocomposites con-
Received 18 April 2013 taining recycled polymers. The interfacial bonding is evaluated using different models for tensile proper-
Received in revised form 27 July 2013 ties of composites. A good agreement is found between the experimental results of mechanical properties
Accepted 19 August 2013
and the theoretical predictions which validate the current analysis.
Available online 19 October 2013
Many parameters such as (a) in Nicolais–Narkis, (a) in Kunori–Geil, (B) in Pukanszky model and inter-
facial strength (t) show the perfect interfacial adhesion. Moreover, the obtained values of (B) and (t) are
Keywords:
compared with other studies. The current study justifies the recycling of polymers through the incorpo-
Waste polymers
Nanocomposites
ration of nanofillers.
Interfacial adhesion Ó 2013 Elsevier B.V. All rights reserved.
Mechanical properties

1. Introduction
Nowadays, the large amounts of waste polymers threaten the
people life. This problem and the other economic and petroleum
considerations related to polymers have increasingly motivated
the researchers to solve the problem [1,2]. Undoubtedly, the effi-
cient treatment of waste polymeric products is recycling and reus-
ing. However, the major problem in this field involves the
degradation of polymer structure in reprocessing which cause
much poorer properties [3]. In this regards, researchers have tried
to introduce the best modifications that can compensate the loss of
properties.
The addition of other components to the waste polymers seems
to be the simplest and easiest way for reusing the recycled poly-
mers. The nanofillers present more excellent feature for improve-
ment of all mechanical, thermal and barrier properties [4–7].
They increase the interphase surface of the components that devel-
op the performances through a simply processing technology with
low cost. It is interesting to note that the nanofillers can enhance
the melt strength of recycled polymers while the inferior melt
strength causes an inconsistency of material after leaving the ex-
truder which makes the production of sheets or profiles, impracti-
cable [8]. Further, nanofillers increase the intrinsic viscosity of
Poly(ethylene terephthalate) (PET) which is necessary for the
reprocessing of PET wastes [9].
Recently, more studies have been carried out on the recycling of
polymers such as PET [8,10–12], Polypropylene (PP) [13–16], high
⇑ Corresponding author. Tel.: +98 21 66454514; fax: +98 21 66547514.
E-mail address: y.zare@aut.ac.ir (Y. Zare).
density Polyethylene (HDPE) [17,18] and others through the addi-
tion of nanoparticles. In these works, the mechanical properties of
nanocomposites were evaluated by experimental characterization.
The prediction and modeling of behavior have been carried out for
nanocomposites from virgin polymers, while this subject has not
been studied for nanocomposites from recycled polymers.
The analysis of behavior provides more information without
requiring to a large number of experiments. In other words, the
models remove any need to much cost, time and also, difficulties
conducted to examination of properties [19–22]. Furthermore,
the models facilitate the development of most desirable products.
As well known, the best properties of nanocomposites can be ob-
tained when a perfect interfacial adhesion is provided between
the matrix and nanofiller [23,24]. However, whether the nanopar-
ticles can provide a good interfacial adhesion and so, an efficient
reinforcement in waste polymers or not? On the other hand,
whether the mechanical properties of nanocomposites show a
strong interfacial bonding between the recycled matrix and nano-
filler phases?
In this paper, much attempt is made to answer these important
questions through the quantitative analysis of interfacial adhesion
by modeling of mechanical properties.
2. Background
In the condition of poor adhesion, the strength of a composite is
determined by the available effective region of load-bearing matrix
in the absence of filler [25,26]. In this state, the interfacial layer
cannot transfer stress and the tensile strength of composite

0927-0256/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.commatsci.2013.08.041
 Y. Zare et al. / Computational Materials Science 81 (2014) 612–616 613

depends on the effective load bearing cross-section area portion (B) parameter is related to the load carried by the dispersed
(1  W) as: phase depending on the interaction which can be applied as a
quantitative measurement of filler–matrix adhesion. The (B)
rc ¼ rm ð1  wÞ ð1Þ
parameter is shown as:
where (rm) and (rc) are the tensile strength of the matrix and com-  
ri
posite, respectively. If (W) is assumed as a power law function of B ¼ ð1 þ AqlÞ ln ð13Þ
volume fraction of filler (/), Eq. (1) can be presented as: rm
where (A) is the specific surface area of filler (contact surface), (q) is
rc ¼ rm ð1  a/b Þ ð2Þ
density of filler, while (l) and (ri) are the thickness and strength of
where (/) is the volume fraction of the filler and (a) and (b) depend the interphase, respectively. The Pukanszky model can be reformu-
on the filler–matrix interaction and the shape and arrangement of lated as:
particles.
rc 1 þ 2:5/
Based on Eq. (2), Nicolais and Narkis [26,27] suggested a model ln rReduced ¼ ln ¼ B/ ð14Þ
for tensile strength of composites reinforced with spherical parti- rm 1  /
cles as: According to Eq. (14), when a linear correlation is observed be-
2=3 tween the reduced tensile strength (ln rReduced) and the volume
rc ¼ rm ð1  a/ Þ ð3Þ
fraction of filler (/), the model is valid.
In the case of good adhesion, the interfacial layer can transfer a Sato and Furukawa [35,36] also suggested a model for tensile
small portion of stress while the deformation of matrix is very modulus of composites containing an adhesion parameter (f) as:
small. In this case, the tensile strength includes a contribution of " ! #
both matrix and filler properties. Therefore, the value of parameter 0:5/2=3 /2=3 wf
E ¼ Em 1þ ð1  wfÞ  ð15Þ
(a) in Nicolais–Narkis model becomes smaller than 1.21 which 1  /1=3 ð1  /1=3 Þ/
shows the stronger adhesion.
  !
Kunori and Geil [28] related the tensile strength of composites / 1 þ /1=3  /2=3
with (a) parameter, which is a stress concentration factor as: w¼ ð16Þ
3 1  /1=3 þ /2=3
rR ¼ expða/Þ ð4Þ
The (f) parameter of 1 shows the poor adhesion, while f = 0 indi-
where (rR) is the relative tensile strength as (rc/rm). The higher val- cates to the perfect adhesion.
ues of (a) indicates to a greater stress concentration.
In the case of good interfacial adhesion, Piggott and Leidner [29]
3. Results and discussion
introduced an empirical model including a coefficient of particle–
matrix adhesion (a), as:
The tensile strength and modulus of nanocomposites containing
rc ¼ K rm  a/ ð5Þ various nanofillers such as layered silicate, CaCO3, carbon nanotube
The effect of interfacial interaction and filler properties on the (CNT) and different waste matrices such as PET, PP and HDPE were
tensile strength of composites [30] can be presented as: provided from the literature. Table 1 shows the studied nanocom-
  posites and the attributed references in which all experimental
at data and other details were given. Accordingly, I refrain from the
rR ¼ 1 þ 1 / ð6Þ
rm further discussion of the details.
Table 1 illustrates the calculated parameters from models for
where (a) is the aspect ratio of filler and (t) is the interfacial stress
different nanocomposites. The (a) parameter from Nicolais–Narkis
transfer parameter. (a) can be calculated from the developed Hal-
model is obtained from the linear plotting of experimental tensile
pin–Tsai model for randomly 3 dimensional (3D) platelet fillers
strength against (/2/3). As observed in Table 1, the value of (a) for
[31,32] as:
PP/CaCO3 nanocomposite is 0.59. It is lower than 1.21 demonstrat-
E ¼ 0:49 E1 þ 0:51 E2 ð7Þ ing the good interfacial adhesion between recycled PP and CaCO3
  nanofiller.
1 þ gn/ The calculated (a) parameter from Kunori–Geil model (Eq. (4))
E1 ¼ Em ð8Þ
1  g/ is also observed in Table 1. The higher values of (a) which shows
  the greater stress concentration are obtained in samples No. 7
1 þ 2g/ and No. 8. Other (a) values are much smaller than zero presenting
E2 ¼ Em ð9Þ
1  g/ a perfect interfacial adhesion.
Further, the same trend of previous calculations is observed by
g ¼ ðEf =Em  1Þ=ðEf =Em þ nÞ ð10Þ Piggott–Leidner model (Eq. (5)), where the lowest (a) parameter,
indicating to the best interfacial adhesion is obtained for samples
n ¼ 2a ð11Þ
where (E1) and (E2) are the tensile moduli of the composite in the Table 1
longitudinal and transverse directions, respectively. (Em) and (Ef) The calculated interfacial parameters from different models.
are the tensile modulus of matrix and filler, respectively. No. Sample a (Eq. (3)) a (Eq. (4)) a (Eq. (5)) B (Eq. (12)) Refs.
Pukanszky developed a model based on the spontaneous forma-
1 PET/clay – 17.63 23.28 21.45 [10]
tion of interphase in composites assuming the variation of tensile
2 PET/clay – 4.75 6.83 8.48 [11]
strength as a function of composition [33,34]. The Pukanszky mod- 3 PET/clay – 3.27 3.8 6.9 [12]
el can be presented as: 4 PP/clay – 10.77 15.67 15.6 [15]
5 PP/clay – 1.49 1.52 4.91 [13]
1/ 6 PP/CNT – 4.35 7.13 9.3 [14]
rc ¼ rm expðB/Þ ð12Þ
1 þ 2:5/ 7 PP/CaCO3 0.59 1.52 1.48 1.74 [16]
8 HDPE/clay – 1.61 1.61 1.76 [18]
614 Y. Zare et al. / Computational Materials Science 81 (2014) 612–616

No. 1 and No. 4. Also, the specimens No. 7 and No. 8 show the Table 2
worst (a) values. The obtained parameters by models.

The high specific surface area of silicate layers can significantly
increase the interfacial interactions between polymer and filler
phases, attributed to many factors such as constituent types of sys-
tem, degree of dispersion, interactions between particles and poly-
mer and particle–particle interactions [37,38]. The negligible effect
of CaCO3 content on the tensile modulus and strength of PP/CaCO3
sample (No. 7) was reported [16]. Also, the impact strength of sam-
ples decreases by incorporation of CaCO3 nanoparticles, despite the
toughening effect of CaCO3 nanoparticles. The lower level of inter-
facial adhesion can be related to the poor dispersion of CaCO3
nanofiller, even with stearic acid treatment. The weak dispersion
may be induced by the low viscosity of recycled PP together with
the large contents of aggregates in the higher content of CaCO3
(20 wt%).
Moreover, the sample No. 8 does not show any improvement in
the tensile strength of recycled HDPE/clay nanocomposite (except
in 1 wt% of nanoclay content), even in the presence of maleated PE
and the well exfoliated clay platelets [18]. The authors attributed
the slight enforcement of clay to the possible existence of other fill-
ers and additives in the recycled HDPE.
The experimental results of tensile strength are also success-
fully fitted to Eq. (6). From the slope of plotted lines, the values
of (at) are obtained. The aspect ratios of nanofillers (a) have not
been exactly characterized by the authors; therefore, they can be
approximately obtained by the developed Halpin–Tsai model mod-
el (Eqs. (7)–(11)). Fig. 1 illustrates the calculation of (a) from devel-
oped Halpin–Tsai model model for samples that their tensile
modulus was reported and the obtained values of (a) are shown
in Table 2. Consequently, the calculated interfacial strength (t) is
No. Sample a (Eq. (11)) t (Eq. (6)) f (Eq. (15)) Refs.
1 PET/clay 10 58.3 2.5 [10]
2 PET/clay 140 3.6 18 [11]
3 PET/clay 22 8.2 4 [12]
observed in Table 2. The most value of (t) is obtained for sample
No. 1 equivalent to 58.3 MPa, as expected according to the ob-
tained parameters from other models.
Based on the literature results, the composites having only
weak Van der Waals interactions between matrix and filler show
relatively low (t) values [39]. Frankland et al. computed the (t)
for weakly interacted PE–CNT about 2.7 MPa, while the interfacial
strength (t) between 30 and 110 MPa have been found for cross-
linked PE–CNT nanocomposites [40]. Also, Durmus et al. obtained
the (t) value of 5 MPa for linear low density Polyethylene
(LLDPE)/clay nanocomposite [38]. As a result, the interfacial
strength parameter (t) for the present nanocomposites verifies
the strong bonding between recycled polymers and nanofillers.
The values of (B) coefficient from Pukanszky model (Eq. (12))
are obtained through the linear plot of ln (reduced tensile strength)
against nanofiller content (/), as shown in Table 1. The minimum
(B) value of 1.74 and maximum 21.45 are calculated for the current
nanocomposites. They can be considered as good results, compared
with (B) values obtained for other systems, ranging from negative
values up to about 20 as a maximum level reported by Szazdi et al.
[41,42]. The results particularly agree with the further parameters
obtained by other studied models. Demjen et al. [34] found the
maximum (B) of 2.45 for composites containing different amounts

Fig. 1. The calculations of aspect ratio (a) from developed Halpin–Tsai model model for (a) No. 1, (b) No. 2 and (c) No. 3 samples.
 Y. Zare et al. / Computational Materials Science 81 (2014) 612–616
Fig. 2. The prediction of tensile modulus by Sato–Furukawa model at different adhesion parameter (f) for (a) No. 1, (b) No. 2 and (c) No. 3 samples.
of CaCO3 micro-particles treated with different silane coupling
agents and stearic acid. These findings confirm that the nanofillers
introduce a better interfacial interaction to the recycled polymers
compared to the treated micro-particles to the neat polymers.
In addition, Szazdi et al. [42] stated that the usual techniques
such as X-ray diffraction (XRD) and transmission electron micros-
copy (TEM) do not present any reliable information about the de-
gree of exfoliation. They also indicated that the nanocomposites
which do not exhibit any silicate reflection in XRD scans, may have
very poor mechanical properties that shows a small degree of exfo-
liation. They have quantitatively estimated the extent of exfolia-
tion by (B) values assuming the specific surface area of silicate
layers of 750 m2 g1 and (B) value of 195 for possibly perfect exfo-
liation. Their analysis indicated that the extent of exfoliation is ex-
tremely small in most nanocomposites, maximum 10% in the best
case. According to these criteria, the most level of exfoliation in the
present samples is estimated as 11% for sample No. 1 (B = 21.45).
However, the morphological analysis showed the most exfoliated
clay layers in the matrix [10]. First of all, the nanoparticles may
have much lower surface area than 750 m2 g1 (assumed by Szazdi
et al.). Therefore, a (B) value lower than 195 is expected for evalu-
ation of full exfoliation, according to Eq. (13). Also, it is believed
that other factors such as modification of crystalline structure
due to nucleation promotion of nanofillers [5], interaction and
interfacial adhesion between matrix and filler phases [43], chemi-
cal reactions [44] and so on can affect the mechanical properties of
nanocomposites, in spite of Szazdi et al. approach correlating the
less improved mechanical properties to the slighter level of exfoli-
ation of nanoclay [42].
Fig. 2 shows the prediction of tensile modulus by Sato–Furuka-
wa model (Eqs. (15) and (16)) at different adhesion parameter (f)
615
and the most suitable (f) values are presented in Table 2. As indi-
cated, (f) values of 1 and 0 represents the poor and perfect adhe-
sion, respectively. However, all reported values of (f) in Table 2
are lower than zero and the negative (f) parameter has not been
investigated in the previous studies on the nano- and micro-com-
posites. As observed in Fig. 2, the higher level of (f) parameter in-
duces smaller values of tensile modulus. In other word, the best
modulus is achieved by inferior (f) values. As a result, the calcu-
lated (f) data indicates that the little content of nanofillers provide
a significantly improved tensile modulus in the nanocomposites
from recycled polymers. The nanofillers can produce much interfa-
cial interactions with matrices and the best modulus and strength
is introduced by perfect interfacial adhesion between nanoparti-
cles and recycled matrices. In conclusion, the negative values of
(f) show much interfacial bonding between polymer and nanofiller
phases for the studied samples.
4. Conclusion
This paper aimed to study the interfacial adhesion in the vari-
ous nanocomposites containing recycled polymers. The analysis
of interfacial adhesion in the recycled polymer nanocomposites
was carried out using the various models for tensile strength and
modulus of composites. The different types of waste polymers such
as PET, PP, HDPE containing nanoclay, CaCO3 and CNT were chosen
for evaluation. A good agreement was found between the experi-
mental data and the models.
Many interfacial parameters from Nicolais–Narkis, Kunori–Geil,
Piggott–Leidner, Pukanszky and Sato–Furukawa models proved the
perfect interfacial bonding between the nanofillers and recycled
616 Y. Zare et al. / Computational Materials Science 81 (2014) 612–616
polymers in all cases. Also, they showed a similar trend for all
nanocomposites. Furthermore, the values of (B) and (t) were com-
pared with other studies and also, the Szazdi et al. approach about
the correlation between (B) value and exfoliation level of nano-
composites was argued. It was concluded that the nanofillers bring
in a good interfacial adhesion with the waste polymers demon-
strating the good efficiency of nanoparticles in procedure of poly-
mer recycling.
References
[1] A. Ashori, A. Nourbakhsh, Waste Manage. 29 (2009) 1291–1295.
[2] M. Mehrabzadeh, F. Farahmand, J. Appl. Polym. Sci. 80 (2001) 2573–2577.
[3] Y. Zare, Waste Manage. 33 (2013) 598–604.
[4] Y. Zare, H. Garmabi, F. Sharif, J. Appl. Polym. Sci. 122 (2011) 3188–3200.
[5] Y. Zare, H. Garmabi, J. Appl. Polym. Sci. 124 (2012) 1225–1233.
[6] A. Khan, M.H. Shamsi, T.-S. Choi, Comput. Mater. Sci. 45 (2009) 257–265.
[7] C. Pisano, P. Priolo, Ł. Figiel, Comput. Mater. Sci. 55 (2012) 10.
[8] M. Kráčalík, M. Studenovský, J. Mikešová, J. Kovářová, A. Sikora, R. Thomann, C.
Friedrich, J. Appl. Polym. Sci. 106 (2007) 2092–2100.
[9] M. Kráčalík, M. Studenovský, J. Mikešová, A. Sikora, R. Thomann, C. Friedrich, I.
Fortelný, J. Šimoník, J. Appl. Polym. Sci. 106 (2007) 926–937.
[10] M. Bizarria, A.L.F.M. Giraldi, C.M. de Carvalho, J.I. Velasco, M.A. d’Ávila, L.H.I.
Mei, J. Appl. Polym. Sci. 104 (2007) 1839–1844.
[11] B.A. Basilia, H.D. Mendoza, L.G. Cada, Philip. Eng. J. 23 (2011).
[12] S. Hamzehlou, A. Katbab, J. Appl. Polym. Sci. 106 (2007) 1375–1382.
[13] F. Cengiz, Preparation and characterization of recycled polypropylene based
nanocomposites, M.Sc. Dissertation, The Graduate School of Natural and
Applied Science of Middle East Technical University, 2008.
[14] J.L. Gao, Y.H. Liu, D.M. Li, Adv. Mater. Res. 279 (2011) 106–110.
[15] N. Tri Phuong, V. Gilbert, B. Chuong, J. Reinf. Plas. Compos. 27 (2008) 1983–
2000.
[16] A. Elloumi, S. Pimbert, A. Bourmaud, C. Bradai, Polym. Eng. Sci. 50 (2010)
1904–1913.
[17] M.E. Hill, Adding Value to Recycled Polyethylene Through the Addition of
Multi-Scale Reinforcements, The University of Akron, 2005.
[18] Y. Lei, Q. Wu, C.M. Clemons, J. Appl. Polym. Sci. 103 (2007) 3056–3063.
[19] Y. Zare, H. Garmabi, J. Appl. Polym. Sci. 123 (2012) 2309–2319.
[20] S. Ebrahimi, K. Ghafoori-Tabrizi, H. Rafii-Tabar, Comput. Mater. Sci. 53 (2012)
347–353.
[21] A. Alavinasab, R. Jha, G. Ahmadi, C. Cetinkaya, I. Sokolov, Comput. Mater. Sci.
44 (2008) 622–627.
[22] Y. Zare, H. Garmabi, Compos. Part B: Eng. (2013).
[23] B. Lauke, T. Schüller, W. Beckert, Comput. Mater. Sci. 18 (2000) 362–380.
[24] J. Gou, B. Minaie, B. Wang, Z. Liang, C. Zhang, Comput. Mater. Sci. 31 (2004)
225–236.
[25] L. Nicolais, M. Narkis, Polym. Eng. Sci. 11 (1971) 194–199.
[26] L. Nicolais, L. Nicodemo, Int. J. Polym. Mater. 3 (1974) 229–243.
[27] R. Lavengood, L. Nicolais, M. Narkis, J. Appl. Polym. Sci. 17 (1973) 1173–1185.
[28] T. Kunori, P. Geil, J. Macromol. Sci. Part B: Phys. 18 (1980) 135–175.
[29] M. Piggott, J. Leidner, J. Appl. Polym. Sci. 18 (1974) 1619–1623.
[30] W.D. Callister, Materials Science and Engineering: An Introduction, Wiley, New
York, 2003.
[31] E.W. Wong, P.E. Sheehan, C.M. Lieber, Science 277 (1997) 1971.
[32] D. Hull, T. Clyne, An Introduction to Composite Materials, Cambridge Univ.
Press, 1996.
[33] B. Pukanszky, Composites 21 (1990) 255–262.
[34] Z. Demjén, B. Pukanszky, J. Nagy, Compos. Part A: Appl. Sci. Manuf. 29 (1998)
323–329.
[35] L.E. Nielsen, J. Appl. Polym. Sci. 10 (1966) 97–103.
[36] Y. Sato, J. Furukawa, Rubber Chem. Technol. 36 (1963) 1081.
[37] E. Bilotti, R. Zhang, H. Deng, F. Quero, H. Fischer, T. Peijs, Compos. Sci. Technol.
69 (2009) 2587–2595.
[38] A. Durmus, A. Kasßgöz, C.W. Macosko, J. Macromol. Sci. Part B: Phys. 47 (2008)
608–619.
[39] A.H. Barber, S.R. Cohen, H.D. Wagner, Appl. Phys. Lett. 82 (2003) 4140.
[40] S. Frankland, A. Caglar, D. Brenner, M. Griebel, J. Phys. Chem. B 106 (2002)
3046–3048.
[41] L. Szazdi, A. Pozsgay, B. Pukanszky, Eur. Polym. J. 43 (2007) 345–359.
[42] L. Százdi, B. Pukánszky, G.J. Vancso, Polymer 47 (2006) 4638–4648.
[43] L. Racz, B. Pukanszky Jr., A. Pozsgay, B. Pukánszky, Polyamide 6/
montmorillonite nano-composites: effect of interaction on structure and
properties, in: From Colloids to Nanotechnology, Springer, 2004, pp. 96–102.
[44] X. Liu, Q. Wu, Polymer 42 (2001) 10013–10019.