EUROPEAN

POLYMER
European Polymer Journal 42 (2006) 1716–1727
JOURNAL
www.elsevier.com/locate/europolj

Conductive poly(butadiene-co-acrylonitrile)-polyaniline
dodecylbenzenesulfonate [NBR-PAni.DBSA] blends
prepared in solution
a,1 b,*
K.C. Yong , P.J.S. Foot , H. Morgan b, S. Cook c, A.J. Tinker c

a
Materials Research Group, SPC, Faculty of Science, Kingston University, Penrhyn Road, Kingston upon Thames, Surrey KT1 2EE, UK
b
Materials Research Group, Faculty of Science, Kingston University, Penrhyn Road, Kingston upon Thames, Surrey KT1 2EE, UK
c
Tun Abdul Razak Research Centre, Brickendonbury, Hertford, Hertfordshire SG13 8NL, UK

Received 20 January 2006; received in revised form 24 February 2006; accepted 1 March 2006
Available online 19 April 2006

Abstract

Blends of poly(butadiene-co-acrylonitrile) elastomer [NBR] and polyaniline dodecylbenzenesulfonate [PAni.DBSA],

with electrical conductivities up to 102 S cm1, have been prepared by solution mixing and casting. Miscibility was max-
imised for NBR with high acrylonitrile (ACN) content, as predicted on the basis of simple solubility parameter calcula-
tions. Blends prepared using NBR with 48 wt% ACN had the lowest electrical conductivity percolation thresholds, and
were much more conductive than previous thermally mixed blends. Optical and electron micrographs of blends prepared
from NBR 48 wt% ACN also showed the lowest levels of phase separation. The FT-IR spectra of NBR-PAni.DBSA
blends resembled a superposition of the spectra of the pure materials, but with significant peak shifts due to changing inter-
molecular interactions between the polymers. Under DSC analysis, thermal events for blends prepared with NBR 48 wt%
ACN also showed the largest temperature shifts relative to those for the pure polymers, supporting the other evidence for
interaction between the two polymers.
 2006 Elsevier Ltd. All rights reserved.

Keywords: Conductive blend; Polyaniline; Nitrile rubber (NBR)

1. Introduction cables [1]. Conventional electrically conductive elas-

tomers (containing graphite or coarse carbon black
Conventional electrically conductive elastomers or powdered metals) have some very significant dis-
were not of major interest until about 1930, when advantages compared with novel electrically con-
several patents described the use of conductive com- ductive elastomers (which contain intrinsically
pounds for the prevention of corona discharge in conductive polymers). Carbon fillers are relatively
poor conductors, providing useful electromagnetic
*
interference (EMI) shielding effectiveness only with
Corresponding author. Tel.: +44 (0)20 8547 7485; fax: +44
(0)20 8547 7562.
about 40 weight percent (wt%) of filler in elastomers
E-mail address: [email protected] (P.J.S. Foot). [2]. Carbon black-filled elastomers have high perco-
1
Malaysian Rubber Board, Malaysia (on study leave). lation thresholds for electrical conduction, and the

0014-3057/$ - see front matter  2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.eurpolymj.2006.03.007
 K.C. Yong et al. / European Polymer Journal 42 (2006) 1716–1727 1717

conductivity values are not usually much higher Paoli [16] obtained homogeneous films by grafting
than about 108 S cm1 [3]. Metal powder fillers polyaniline with NBR29 through potentiodynamic
have high densities, and weight fractions of 60% or potentiostatic syntheses. The films were insoluble
are often needed to achieve good EMI shielding in toluene and in chloroform.
[2]. The metal surface can also become oxidised, The aim of the present work was to study electri-
and this will reduce the effective conductivity of cally conductive blends produced by combining var-
the composite [4]. ious grades of NBR with different amounts of
Intrinsically conductive polymers may have poor PAni.DBSA. It was decided to produce blends from
stability [5], poor mechanical properties [6] and solution, which required that both components be
higher cost than commodity polymers [2]. An effec- soluble in a common solvent. This method has not
tive method to overcome some of these problems is previously been used to make NBR-based con-
by preparing composite materials containing the ductive blends. Characterisation of the resulting
intrinsically conductive polymers as fillers dispersed NBR-PAni.DBSA blends is reported here, using
in thermoplastic or elastomeric matrices. To intro- Fourier-transform infrared (FT-IR) spectroscopy,
duce a conductive polymer into an insulating elasto- differential scanning calorimetry (DSC), determina-
meric matrix, one can use techniques such as tion of the electrical conductivity of pure polymers
thermomechanical mixing [3,7–9], solution mixing and blends and morphological study by optical
[8,10–15] or electrochemical methods [16,17]. and electron microscopy. The effects of acrylonitrile
In this work, poly(butadiene-co-acrylonitrile) or composition on the morphology and electrical prop-
nitrile rubber [NBR] was selected partly because of erties of the blends were important concerns of these
its high chemical attack resistance [3,9,18]. NBR studies.
should have quite good miscibility with conductive
polymers, especially polyaniline [PAni]. PAni-based 2. Experimental
elastomer blends have been the subject of research
and development activity for a number of years 2.1. Chemicals and raw materials
[3,7–10,12–17]. Special interest has focused on PAni
because the monomer is relatively inexpensive, poly- Aniline monomer (99 wt%, Sigma-Aldrich), ammo-
merisation is straightforward and it provides a high nium persulphate [APS] (98 wt%, Sigma-Aldrich),
yield [19,20]. When PAni base is doped with sulfonic 36.5–38 wt% HCl solution (GPR), anhydrous ferric
acids, it becomes highly electrically conductive, chloride [FeCl3] (98 wt%, Sigma-Aldrich), 33 wt%
reaching 16 S cm1 for p-toluenesulfonic acid ammonia solution (GPR), 70 wt% DBSA solution
(p-TSA) [21]. If the acids contain large n-alkyl in 2-propanol (GPR, Sigma-Aldrich), chloroform
groups, as in dodecylbenzenesulfonic acid (DBSA) (GPR), methanol (GPR) and several commercial
or tetrapropylbenzenesulfonic acid (TPBSA), the grades of NBR [Krynac 27.50, Krynac 34.50
polymer is soluble in organic solvents [21]. Accord- and Krynac 50.75 (Bayer) and Breon N41C45
ing to Poussin et al., undoped PAni (emeraldine (ZEON Chemicals)]. All grades of NBR were
base), hydrochloric acid doped PAni (PAni.HCl), washed with methanol for 24 h using Soxhlet extrac-
p-TSA doped PAni (PAni.TSA) and DBSA doped tion in order to remove chemical additives. By 1H
PAni (PAni.DBSA) are stable to at least 200 C, NMR analysis of solutions in THF-d8 (Bruker
100 C, 260–300 C and 260 C, respectively [21]. AC300), the actual ACN contents for pure NBR
Hence the polyanilines may be suitable thermally sta- were found to be 25.7 ± 1.0 wt% (Krynac 27.50),
ble partners for the production of practical blends. 33.1 ± 1.0 wt% (Krynac 34.50), 41.4 ± 1.0 wt%
Blends of various grades of NBR with PAni have (Breon N41C45) and 48.2 ± 1.0 wt% (Krynac
previously been produced by thermo-mechanical 50.75), respectively.
mixing or electrochemical methods only. Vallim
et al. [3] used a twin-roller mill to blend PAni.HCl, 2.2. Calculation of solubility parameter values
PAni.DBSA, PAni.TPBSA and PAni.TSA with for main materials
NBR of 29 wt% [NBR29] or 45 wt% [NBR45]
ACN content. All the blends showed very low con- In order to estimate the solubility parameters for
ductivities, ranging from about 108 to 1012 S all the grades of pure NBR and PAni.DBSA, Eq. (1)
cm1. It was concluded that NBR45 was more mis- and the values of functional group molar attraction
cible with PAni than with NBR29. Tassi and De constants Fi calculated by Hoy [22] were used.
1718 K.C. Yong et al. / European Polymer Journal 42 (2006) 1716–1727
 P 
q  Fi chloroform were prepared and filtered. Each of
dp ¼ ð1Þ
M0 the pure NBR solutions was added to the PAni.
DBSA solution separately in appropriate amounts
where q is the density of the polymer (kg m3), M0 is to obtain the following compositions (wt%
the formula weight
P of the polymer repeat unit NBR:wt% PAni.DBSA), 50:50, 60:40, 70:30, 80:20,
(kg mol1) and Fi is the sum of all the groups’ mo- 90:10, 95:5, 97.5:2.5 and 99:1, respectively. Each of
lar attraction constants in the polymer repeat unit the above blend solutions was magnetically stirred
[(J m3)0.5 mol1]. The determination was based on for 24 h at room temperature prior to casting.
the smallest repeat unit of the relevant polymer, treat-
ing the co-polymers as being regular, for simplicity. 2.5. Morphological studies (optical microscopy
and TEM)
2.3. Synthesis of PAni.DBSA
NBR-PAni.DBSA blends were studied with a
Conductive emeraldine salt PAni.HCl was syn- Nikon OPTIPHOT-2 optical microscope (magnifi-
thesised by an oxidative chemical polymerisation cation 200·) linked to a computer by a video con-
based on an oxidant (APS)/monomer (aniline) ini- verter (LINKAM VTO 232) for digital image
tial mole ratio of 1.0. Fifty millilitre of aniline and capture. A small drop of each sample blend solution
125.2 g of APS were dissolved in 375 mL and was put onto a microscope cover-glass and allowed
276 mL of 2 M HCl, respectively, which had been to evaporate to form a thin (3.0 lm) transparent
pre-cooled to 1 C. The aniline solution was placed film.
in an ice bath with an overhead stirrer. After the Blends with 5 wt% (generally below the conduc-
addition of a few drops of FeCl3 solution as a cata- tivity threshold) and 10 wt% (generally above the
lyst, the APS solution was added gradually over a conductivity threshold) of PAni.DBSA were also
period of 20 min, with constant stirring to ensure studied by TEM (with a Philips CM12 microscope
thorough mixing. The temperature of the polymeri- at accelerating voltage, 80 kV). An ultra-thin sam-
sation mixture was maintained between 0 and 2 C ple (ca. 150 nm) of each blend was obtained for this
while the pH was adjusted to 0–0.5 by the addition purpose by sectioning with an RMC MT7000
of concentrated HCl. The mixture was left for 6 h to microtome.
complete the polymerisation, and the precipitate
was allowed to settle before decanting and final 2.6. FT-IR spectroscopy
washing with distilled water until the filtrate was
colourless. PAni emeraldine base (EB) was prepared Infra-red spectra of PAni.DBSA, different grades
by de-protonating the wet PAni.HCl in 33% aque- of pure NBR and all NBR-PAni.DBSA blends were
ous ammonia solution (3000–5000 mL) with 24 h recorded on a Perkin–Elmer FT-IR Spectrometer
stirring, followed by washing and re-protonation PARAGON 100. A small amount of each sample
in 1 M aqueous DBSA at a molar ratio EB:DBSA solution was cast onto a KBr window in order to
of 1:1. Filtration was done with a sintered-glass fun- obtain a reasonably thin (3.0 lm), transparent
nel of grade 3 porosity (16–40 lm). All products film.
were washed with large quantities of distilled water
until the filtrates were colourless and no odour of 2.7. Thermal analysis
ammonia could be detected. Final drying of the
powdered filter cake was carried out in a vacuum The PAni.DBSA powder, different grades of
oven at 50 C for 24 h. A sample of PAni.DBSA pure NBR and all NBR-PAni.DBSA blend cast
prepared here was submitted for elemental analysis films were characterised by differential scanning
and was estimated thereby to have a 42% proton- calorimetry. Above ambient temperature (30–
ation level on the basis of the S:N atomic ratio. 400 C), a Mettler Toledo DSC822e (heating rate
20 C/min) was used, and below ambient tempera-
2.4. Blend preparation ture (60 to 0 C), samples were characterised using
a Perkin–Elmer Pyris Diamond DSC (heating rate
Masterbatch solutions of the four different 10 C/min). Each sample was sealed in a 40 lL
grades of pure NBR (50 mg NBR/mL solvent) and aluminium pan and was analysed under N2 atmo-
PAni.DBSA (16.7 mg PAni.DBSA/mL solvent) in sphere. An empty pan was used as the reference.
 K.C. Yong et al. / European Polymer Journal 42 (2006) 1716–1727 1719

Each thermal analysis was repeated twice, and Electrical conductivities were also calculated
the error for all of the results is believed to be from the mean resistance values obtained by the
±1.5 C. 2-probe method using Eq. (3).
The onset temperature of the major exotherm   
1 L
was defined as the intersection of tangents drawn r¼ ð3Þ
R A
to the baseline and major exotherm slope at half-
height in the above-ambient temperature DSC where r is the electrical conductivity, R is the mean
thermograms. The onset temperature of the glass value of apparent resistance (voltage/current ratio)
transition (Tg) was defined as the intersection of tan- for the cast blend, L is the measured electrode spac-
gents drawn to the baseline and major ‘‘endotherm’’ ing distance and A is the cross-sectional area of cast
slope at half-height in the sub-ambient temperature film between the current-carrying electrodes.
DSC thermograms.
3. Results and discussion
2.8. Electrical conductivity measurement
3.1. Calculation of solubility parameters
The dc electrical conductivity values for for materials used
PAni.DBSA, different grades of pure NBR and
NBR-PAni.DBSA blends were calculated from elec- Chloroform has a solubility parameter of 19.0
trical resistance values measured by 2- or 4-probe (MJ m3)0.5 [2], and all grades of NBR were highly
techniques. The equipment employed was a Keith- soluble in it. The PAni.DBSA prepared in this work
ley 617 programmable electrometer and a Keithley was also soluble in CHCl3, giving dark green solu-
224 programmable current source, both under com- tions. The calculated solubility parameters for
puter control. A pressed pellet of pure PAni.DBSA PAni.DBSA and the grades of NBR used here are
(diameter 13 mm, thickness 1.5 mm; prepared in a summarised in Table 1.
KBr press at 10 ton load) was placed in a holder The value for PAni.DBSA agreed with that
with four spring-loaded metallic contacts in order reported in the literature [23]. On the basis of sim-
to measure its electrical resistance. ple solubility parameter considerations, it appears
For the 4-probe van der Pauw technique, samples from Table 1 that miscibility between NBR and
of each NBR-PAni.DBSA blend were cast from solu- PAni.DBSA would be optimised by using the grades
tion onto square microscope slides (625 mm2), and of NBR with high ACN contents, i.e., 41.4 and
the solvent was allowed to evaporate for 24 h. Fine 48.2 wt%.
copper wires 50 mm long were attached to the four
corners of the cast films using small silver paint con- 3.2. Morphological studies (optical microscopy
tacts (Acheson Electrodag 915). Guarded 2-probe and TEM)
electrical resistance measurements were performed
for less conductive cast films (conductivity < All the NBR-PAni.DBSA blends were studied by
107 S cm1). optical microscopy, but only images of blends con-
For the samples measured by the 4-probe tech- taining 5, 20, 30 and 40 wt% of PAni.DBSA pre-
nique, electrical conductivities were calculated from pared with 25.7 and 48.2 ACN wt% rubbers are
the mean resistance values using the van der Pauw shown here (Figs. 1 and 2, respectively). Two visu-
equation (Eq. (2)). ally distinct regions were observed, i.e., the ‘‘white’’

2 ln 2
r¼ ð2Þ
ðR1 þ R2 Þpdf Table 1
Solubility parameter values for PAni.DBSA and all grades of
where r is the electrical conductivity [S cm1], R1 pure NBR
and R2 are the mean values of apparent resistance Material Solubility parameter (MJ m3)0.5
for a cast blend in its two perpendicular contact NBR 25.7 ACN wt% 19.0
configurations, d is the thickness of the sample NBR 33.1 ACN wt% 19.6
and f is a geometric factor (close to unity for sym- NBR 41.4 ACN wt% 20.3
metrical contacts on circular pellets or square NBR 48.2 ACN wt% 20.7
PAni.DBSA 20.8
plaques).
1720 K.C. Yong et al. / European Polymer Journal 42 (2006) 1716–1727

Fig. 1. Optical micrographs (200· magnification) for the blends of NBR 25.7 ACN wt% and PAni.DBSA. Blends contain (a) 60:40,
(b) 70:30, (c) 80:20 and (d) 95:5 wt% NBR:PAni.DBSA.

ones (pale-green in true colour) and the ‘‘black’’ taining NBR with lower ACN contents (25.7 and
ones (actually dark green). 33.1 wt%) showed more obvious phase separation
On increasing the proportion of PAni.DBSA, the than those richer in ACN (41.4 and 48.2 wt%).
area of the dark-coloured region increases, as would The two phases also appear to have a larger area
be expected for a classical particulate composite of contact in the case of the ACN-rich NBR blends.
with a metal or graphite filler rather than a polymer Unfortunately, no comparable micrographs of
blend. There are also some large particles present pure PANI-DBSA films were obtainable. This
even at fairly low loadings (e.g., 5 wt%). However, partly relates to their poorer film-forming ability,
on the basis of the evidence presented below, we but is largely due to their strong optical absorption;
believe that there is significant miscibility between only spin-coated films were thin enough to produce
the polymers. micrographs, and this coating method was found to
The 5 wt% blends appear to contain isolated produce severe effects on the morphology that were
PAni.DBSA particles, and so should be well below not comparable with the blends.
the percolation threshold for conductivity; neverthe- Examples of transmission electron micrographs
less it will be seen below that there is significant of blends (5 and 10 wt% PAni.DBSA, prepared with
residual conductivity in these samples. NBR 41.4 ACN wt%) are shown in Fig. 3. The dar-
The dark regions (rich in PAni.DBSA) are large ker regions of both TEM micrographs are again
conductive particles and their agglomerates, result- related to conductive PAni.DBSA particles in an
ing from some degree of phase separation. The NBR matrix. Relatively large isolated polyaniline
light-coloured regions (rich in NBR) are related to particles are observed in Fig. 3(a), which is consis-
well-blended PAni.DBSA and NBR. Blends con- tent with the fairly low electrical conductivity of
 K.C. Yong et al. / European Polymer Journal 42 (2006) 1716–1727 1721

Fig. 2. Optical micrographs (200· magnification) for the blends of NBR 48.2 ACN wt% and PAni.DBSA. Blends contain (a) 60:40,
(b) 70:30, (c) 80:20 and (d) 95:5 wt% NBR:PAni.DBSA.

Fig. 3. Transmission electron micrographs of NBR 41.4 ACN wt%-PAni.DBSA blends. Blends contain (a) 5 wt% and (b) 10 wt% of
PAni.DBSA.

the 5 wt% PAni.DBSA blend (see Section 3.5). In cluster or to form conductive pathways, which is
Fig. 3(b), the PAni.DBSA particles are starting to consistent with the much higher conductivity for
1722 K.C. Yong et al. / European Polymer Journal 42 (2006) 1716–1727

the 10 wt% PAni.DBSA blend. Numerous very wagging). The spectrum of PAni.DBSA cast film
small particles of PAni.DBSA were revealed by has characteristic absorptions at 3447 cm1 (@NAH
TEM in the NBR-rich regions; most of these are stretching), 1559 cm1 (N@quinoid@N), 1478 cm1
of colloidal dimensions (50–1000 nm) and they (NAbenzoidAN), 1295 cm1 (quinoid@NAbenzoid),
might be expected to make a contribution to electri- 1240 cm1 (C aromaticAN stretching) and 1030 cm1
cal conductivity even below the main percolation (S@O stretching).
threshold. At first sight, the FT-IR spectra of the blend films
are merely superpositions of the spectra for both
3.3. Infrared spectroscopy polymers, but detailed examination reveals signifi-
cant shifts of certain key bands; the mixing of the
Examples of FT-IR spectra for PAni.DBSA, two blended components may be assessed by study-
pure NBR 48.2 ACN wt% and their blends (i.e., ing some of these (Table 2). The bands at 3447 cm1
with 10 wt% and 40 wt% of PAni.DBSA) are shown (NAH stretching) and 1030 cm1 (S@O stretching)
in Fig. 4. are solely due to PAni.DBSA, while that at
FT-IR spectra for cast films of all grades of pure 2237 cm1 derives from NBR. In the spectra of
NBR show the characteristic absorptions at the blends, all these bands have intensities compara-
2237 cm1 (AC„N stretching), 1630 cm1 (C@C ble to those for the pure materials, but two large,
stretch) and 1440–1445 cm1 (out of plane CAH opposing peak shifts can be observed (a decrease

Fig. 4. FT-IR spectra of PAni.DBSA, pure NBR 48.2 ACN wt% and their blends. (a) 1000–1650 cm1, (b) 2150–2300 cm1 and (c) 3400–
3470 cm1.
 K.C. Yong et al. / European Polymer Journal 42 (2006) 1716–1727 1723

Table 2 have much smaller specific surface areas, and there-

Representative IR absorption peaks for PAni.DBSA, all grades fore strong H-bonding characteristic of bulk
of pure NBR and the various NBR-PAni.DBSA blends
PAni.DBSA is again detected in the IR spectra.
Pure material Peak assignment (cm1). Indication of In contrast to the infrared spectra, UV–visible
or blend peak intensity: W, weak; M, medium
(wt% NBR:wt%
transmission spectra were not easy to obtain,
@NAH S@O AC„N because of the large scattering background in many
PAni.DBSA)
stretching stretching stretching
cases. In principle they might have revealed any
PAni.DBSA 3447W 1030M – slight differences in the electronic band structure
NBR 25.7 ACN wt% – – 2237M
NBR 33.1 ACN wt% – – 2237M
of the conducting polymer as a function of blend
NBR 41.4 ACN wt% – – 2237M composition, but the UV–visible band widths were
NBR 48.2 ACN wt% – – 2237M so large that it was impossible to discern any trends.
Blends with NBR 25.7
ACN wt% 3.4. Thermal analysis
90:10 3423W 1030W, 1080W 2237W
80:20 3423W 1030W, 1080W 2237W Figs. 5 and 6 show examples of the sub-ambient
70:30 3419W 1030W, 1080W 2237W
temperature DSC thermograms for two grades of
60:40 3423W 1030W 2236W
50:50 3440W 1030W 2236W pure NBR (with 25.7 and 48.2 ACN wt%) and their
blends with 10–50 wt% of PAni.DBSA. (A DSC
Blends with NBR 33.1
thermogram for pure PAni.DBSA is not shown here
ACN wt%
90:10 3417W 1080W 2237W because there were no obvious thermal processes in
80:20 3417W 1080W 2237W the low-temperature region). Figs. 7 and 8 show
70:30 3413W 1080W 2237W examples of the above-ambient temperature DSC
60:40 3423W 1030W 2236W thermograms for PAni.DBSA, two grades of pure
50:50 3440W 1030W 2236W
NBR and their blends with 10–50 wt% of the con-
Blends with NBR 41.4 ducting polymer.
ACN wt%
90:10 3411W 1080W 2237W
80:20 3411W 1080W 2237W
70:30 3407W 1080W 2236W
60:40 3423W 1029W 2236W
50:50 3437W 1029W 2236W
Blends with NBR 48.2
ACN wt%
90:10 3411W 1081W 2237W
80:20 3411W 1081W 2237W
70:30 3407W 1081W 2236W
60:40 3412W 1029W 2236W
50:50 3437W 1029W 2236W

to 3411–3423 cm1 for the NAH stretch and an

increase to 1080 cm1 for the S@O stretch), attrib-
utable to changes in the intermolecular interactions.
The peak positions for all the dilute blends largely
revert to those of the pure conducting polymer at
higher concentrations (40 wt% PAni.DBSA and
above). In pure PAni.DBSA, the positively charged
NAH hydrogens would have a strong propensity for
hydrogen-bonding with the negative sulfonate oxy-
gens, and the infrared frequency changes suggest
that this bonding is disrupted by polar interactions
with the rubber matrix in the dilute, well-dispersed Fig. 5. Sub-ambient temperature DSC thermograms for NBR-
blends. The larger, more aggregated particles in PAni.DBSA blends of various compositions (wt% NBR:wt%
the concentrated blends (Section 3.2) are likely to PAni.DBSA) prepared with NBR 25.7 ACN wt%.
1724 K.C. Yong et al. / European Polymer Journal 42 (2006) 1716–1727

Fig. 6. Sub-ambient temperature DSC thermograms for NBR- Fig. 7. Above-ambient temperature DSC thermograms for
PAni.DBSA blends of various compositions (wt% NBR:wt% PAni.DBSA, pure NBR 25.7 ACN wt% and their blends of
PAni.DBSA) prepared with NBR 48.2 ACN wt%. different composition (wt% NBR:wt% PAni.DBSA).

The calculated mean Tg values for all the pure cation of the high-temperature miscibility of
NBR grades and NBR-PAni.DBSA blends are PAni.DBSA with NBR in their blends may also
shown in Table 3. The sub-ambient temperature be gained from the onset temperatures shown in
DSC thermograms of the blends have positive Tg Table 4. For high proportions of PAni.DBSA (40–
shifts for the NBR component that increase with 50 wt%), only small shifts in the onset temperature
the proportion of PAni.DBSA up to 30 wt%, and of the major exotherm (w.r.t. that for pure
then decrease when 40 wt% or more is added. This PAni.DBSA) are observed. The blends containing
would be consistent with the existence of effective low to moderate amounts of PAni.DBSA (10–30
dispersion of the conducting polymer at low concen- wt%) showed large shifts in the onset temperature
trations, followed by more complete phase separa- of the major exotherm, echoing the evidence for bet-
tion between the two components at the higher ter mixing in these blends at room temperature and
levels of PAni.DBSA. below.
All the above-ambient temperature DSC thermo-
grams for the blends show thermal processes that 3.5. Electrical conductivity measurement
are combinations of events recorded for pure
NBR and PAni.DBSA, but the events show some The calculated electrical conductivity values for
degree of temperature shift relative to the corre- pure PAni.DBSA, the various grades of pure NBR
sponding processes in the pure polymers. As indi- and blends with different proportions of NBR and
cated by the data in Table 4, all grades of pure PAni.DBSA are shown in Fig. 9. Clearly all grades
NBR are thermally stable up to about 320 C. The of pure NBR are very good electrical insulators,
PAni.DBSA also has good thermal stability up to with conductivities in the region of 1014 S cm1.
about 233 C. The PAni.DBSA used in this project had a high
With the caveat that the polymers are approach- electrical conductivity of 1.2 ± 0.5 S cm1, compa-
ing the onset of thermal decomposition, some indi- rable with the literature values for pressed pellet
 K.C. Yong et al. / European Polymer Journal 42 (2006) 1716–1727 1725

samples [21,23]. The pure conducting polymer did

not form films very reproducibly, but its film con-
ductivity was nevertheless very high (>100 S cm1).
The electrical conductivity of all the blends
increased with the proportion of PAni.DBSA, as
observed in several studies of PAni.DBSA blends
[3,7,9,13,15]; however, the compatibility of the
blended polymers was greater in the present work,
and the properties of the blends were quite
distinctive.
Blends prepared from NBR with higher ACN
contents, i.e., 41.4 and 48.2 ACN wt%, had higher
electrical conductivities (up to about 102 S cm1)
than those prepared with NBR of lower ACN con-
tent, i.e., 25.7 and 33.1 wt% (up to about
103 S cm1). Vallim et al. made a similar observa-
tion for two NBR-PAni.DBSA blends prepared by
mechanical mixing [3]. Hence the blends with less
phase separation were found to have higher electri-
cal conductivity. This suggests either that the more
compatible host material favours the formation of
microscopic conductive pathways, or that the well-
dispersed PAni.DBSA species within the NBR-rich
regions are assisting the process of bulk charge
Fig. 8. Above-ambient temperature DSC thermograms for pure
NBR 48.2 ACN wt% and its blends of various compositions (wt%
transport.
NBR:wt% PAni.DBSA) with PAni.DBSA.

Table 3
The mean glass transition temperature (Tg) values for all grades of pure NBR and for NBR-PAni.DBSA blends
Composition Glass transition temperature (C)
(wt% NBR:wt%
Blend with NBR 25.7 Blend with NBR 33.1 Blend with NBR 41.4 Blend with NBR 48.2
PAni.DBSA)
ACN wt% ACN wt% ACN wt% ACN wt%
Pure NBR 46 36 20 11
90:10 46 35 19 8
80:20 46 34 17 8
70:30 43 32 16 7
60:40 45 34 18 8
50:50 46 35 18 8

Table 4
The mean onset temperature values of major exotherms for the above-ambient temperature DSC thermograms of PAni.DBSA, pure NBR
and NBR-PAni.DBSA blends
Composition Onset temperature of major exotherm (C)
(wt% NBR:wt%
Blend with NBR 25.7 Blend with NBR 33.1 Blend with NBR 41.4 Blend with NBR 48.2
PAni.DBSA)
ACN wt% ACN wt% ACN wt% ACN wt%
Pure NBR 335 330 325 331
90:10 334 324 257 263
80:20 324 323 241 240
70:30 233 232 224 222
60:40 231 239 240 246
50:50 237 237 238 229
PAni.DBSA 233 233 233 233
1726 K.C. Yong et al. / European Polymer Journal 42 (2006) 1716–1727

The conductivity percolation threshold for each achieved by using NBR with high ACN contents,
family of blends (prepared with same grade of i.e., 41.4 and 48.2 wt%.
NBR) was estimated by fitting the data from Fig. 9 However, a simple percolation model for conduc-
to a simple percolation model as defined by Eq. (4) tive particles in an insulating matrix does not fit the
[15,24]. data for the dilute polyaniline blends. The very tiny
t (50–1000 nm) particles of PAni.DBSA revealed by
rf ¼ cðf  fp Þ ð4Þ
TEM microscopy probably provide the small, but
where c is a constant, t is the critical exponent, f is significant conductivity observed in blends having
the volume fraction of the conductive medium and 1–5 wt% loadings of conducting polymer – well
fp is the volume fraction at the percolation thresh- below the main percolation threshold. We propose
old. All the weight fractions referred to above were that these colloidal-sized particles mediate conduc-
converted into volume fractions for this analysis. By tive path formation between the larger particles of
fitting the data to a plot of log electrical conductiv- PAni.DBSA down to very low loadings.
ity (r) versus log (f  fp), it was possible to estimate It is possible that H-bonding has a significant
the values of t and correlation coefficient (R) for effect on the efficiency of intermolecular charge
each case. The estimated electrical conductivity per- transfer. However, from our results it is not possible
colation thresholds (in wt% and vol% of PAni.DB- to comment on the likelihood of slight changes in
SA) are shown in Table 5. Reasonably low electrical the molecular mechanism, as opposed to the large
conductivity percolation thresholds can evidently be changes in morphology and inter-particle contact

-1

-3
Log Conductivity (S.cm-1)

-5

-7

-9

-11

-13

-15
0 10 20 30 40 50
Added PAni.DBSA (wt%)
Blend with NBR 25.7 ACN wt% Blend with NBR 33.1 ACN wt%
Blend with NBR 41.4 ACN wt% Blend with NBR 48.2 ACN wt%

Fig. 9. Electrical conductivities for the different grades of NBR-PAni.DBSA blends as a function of PAni.DBSA content (wt%).

Table 5
Electrical conductivity percolation thresholds for different grades of NBR-PAni.DBSA blends
Grade of NBR Estimated percolation Critical exponent, t Correlation coefficient, R
threshold (wt%/vol%
of PAni.DBSA content)
NBR 25.7 ACN wt% 7.5/6.4 4.2 0.98
NBR 33.1 ACN wt% 7.5/6.4 4.1 0.98
NBR 41.4 ACN wt% 6.6/5.8 3.8 0.99
NBR 48.2 ACN wt% 6.0/5.4 3.6 0.99
 K.C. Yong et al. / European Polymer Journal 42 (2006) 1716–1727 1727

that appear to correlate with the observed changes of this project. The authors also wish to thank
in conductivity. C. Hull, R. Davies and K. Lawrence from Tun
Abdul Razak Research Centre, Hertford for their
4. Conclusion help with the NMR, DSC and TEM analyses.

NBR-PAni.DBSA blends with useful levels of

References
electrical conductivity have been successfully pre-
pared by solution mixing. The conductivities were [1] Norman RH. Conductive rubbers and plastics. England:
much higher than those of comparable materials Elsevier; 1970. p. 3–4.
reported previously [3]. Blends prepared using [2] Jiang G, Gilbert M, Hitt DJ, Wilcox GD, Balasubramanian
NBR with the highest ACN content showed the best K. Compos Part A-Appl Sci 2002;33:745–51.
[3] Vallim MR, Felisberti MI, De Paoli M-A. J Appl Polym Sci
electrical properties (estimated conductivity perco-
2000;75:677–84.
lation threshold 6.0 wt%/5.4 vol% of PAni.DBSA), [4] Bhattacharya SK, Kusy RP. Metal-filled polymers:
owing to its reasonably good miscibility with properties and applications. New York: Marcel Dekker;
PAni.DBSA. 1986.
The FT-IR spectra of pure NBR and PAni. [5] Foot PJS, Kaiser AB. Conducting polymers, Kirk-Othmer
encyclopedia of chemical technology. New York: John
DBSA corresponded well with those reported in
Wiley & Sons Inc.; 2004.
the literature [16,23]. Those for cast films of NBR- [6] Laska J, Zak K, Pron A. Synthetic Met 1997;84:117–8.
PAni.DBSA blends resembled a superposition of [7] Davies SJ, Ryan TG, Wilde CJ, Beyer G. Synthetic Met
the spectra of the pure materials, but some notable 1995;69:209–10.
peak shifts also provided evidence of the bonding [8] Pinho MS, Gorelova MM, Dezzoti M, Soares BG, Pertsin
interaction between the two polymers. From the AJ. J Appl Polym Sci 1998;70:1543–9.
[9] Leyva ME, Barra GMO, Gorelova MM, Soares BG, Sens
low and high temperature DSC thermograms, M. J Appl Polym Sci 2001;80(4):626–33.
NBR-PAni.DBSA blends with the largest tempera- [10] Carone E, D’Ilario L, Martinelli A. J Appl Polym Sci
ture shifts for their thermal events (i.e., those con- 2002;86:1259–64.
taining NBR 48.2 ACN wt%) were also the ones [11] Radhakrishnan S, Saini DR. Polym Int 1994;34:111–7.
[12] Zheng W, Levon K, Taka T, Laasko J, Osterholm J-E.
with the best miscibility as indicated by optical
J Polym Sci Phys 1995;33:1289–306.
and electron microscopy. [13] Gospodinova N, Mokreva P, Tsanov T, Terlemezyan L.
It is likely that miscibility between the various Polymer 1997;38:743–6.
grades of pure NBR (even for the lower ACN con- [14] Cataldo F, Maltese P. Polym Adv Technol 2001;12:293–9.
tents) and the PAni.DBSA could be further [15] Leyva ME, Barra GMO, Soares BG. Synthetic Met
2001;123:443–9.
improved by introducing a coupling agent into the
[16] Tassi EL, De Paoli M-A. Polymer 1992;33(11):2427–30.
system. The coupling agent could be incorporated [17] Zoppi RA, De Paoli M-A. J Electroanal Chem
directly (as one of the system additives) or indirectly 1990;290:275–82.
(by using a chemically modified NBR grade, e.g., [18] Morton M, editor. Rubber technology. 2nd ed. New York:
carboxylated NBR as the elastomeric host). This Van Nostrand Reinhold Company; 1973. p. 16–7.
will be the subject of future work, together with a [19] Anderson MR, Mattes BR, Reiss H, Kaner RB. Science
1991;252:1412–5.
study of the temperature- and frequency-dependent [20] Narkis M, Haba Y, Segal E, Zilberman M, Titleman GI,
conductivity in order to characterise better the Siegmann A. Polym Adv Technol 2000;11:665–73.
mechanism of conduction in these blends. [21] Poussin D, Morgan H, Foot PJS. Polym Int 2003;52:
433–8.
[22] Hoy KL. J Paint Technol 1970;42:76.
Acknowledgements
[23] Morgan H, Foot PJS, Brooks NW. J Mater Sci 2001;36:
5369–77.
The authors would like to acknowledge the [24] Stauffer D. Introduction to percolation theory. Lon-
Malaysian Rubber Board for the financial support don: Taylor and Francis; 1985.