Polymer Testing 89 (2020) 106623

Contents lists available at ScienceDirect

Polymer Testing
journal homepage: http://www.elsevier.com/locate/polytest

Material Properties

Influence of alkaline treatment and acetone extraction of natural rubber

matrix on properties of carbon black filled natural rubber vulcanizates
Phattarawadee Nun-anan a, Suwaluk Wisunthorn a, *, Skulrat Pichaiyut a, Norbert Vennemann b,
€we b, Charoen Nakason a, **
Claudia Kummerlo
a
Faculty of Science and Industrial Technology, Prince of Songkla University, Surat Thani Campus, Surat Thani, 84000, Thailand
b
Faculty of Engineering and Computer Science, University of Applied Sciences Osnabrück, Osnabrück, 49076, Germany

A R T I C L E I N F O A B S T R A C T

Keywords: In this study, air dried sheet (ADS) showed higher molecular weight, proteins, lipids and more gel content than
Alkaline treatment the low protein natural rubber (LPNR) from alkaline treatment or acetone-extracted natural rubber (AENR). After
Acetone extraction removal of proteins and lipids, LPNR and AENR had shorter scorch and cure times among the rubber compounds
Natural rubber vulcanizates
observed. This is due to higher content of free fatty acids, glycerides and sodium salts of fatty acid that might act
Carbon black
Network structures
as cure activators with sulfur curing. Furthermore, a finer dispersion of CB was found in ADS, due to chemical
interactions of CB with proteins and lipids at terminal ends of NR molecules. Also, high molecular weight and gel
in ADS could induce a higher reinforcing index (α) resulting to superior mechanical, dynamic, thermo-
mechanical, hardness and crosslink density. In contrast, CB agglomerates occurred in LPNR and AENR with
low molecular weight and reduced proteins and lipids, causing inferior mechanical, dynamic, thermo-
mechanical, along with loss of stiffness and crosslink density.

1. Introduction [8]. Furthermore, the non-rubber components (especially proteins and

lipids) cause inconsistencies in the properties of natural rubber, partic­
Carbon black (CB) is a reinforcing filler that is widely used in rubber ularly in viscosity and elasticity, which affect NR applications [9]. In
compounds. Addition of CB into a rubber matrix tends to favorably addition, it was claimed that proteins and lipids may be the main
modify various properties, including tensile modulus and strength, components that control various properties of NR including plasticity
hardness, abrasion resistance, and compression resistance [1,2]. Natural retention index (PRI) and the mechanical properties modulus and green
rubber (NR) is the main type of rubber matrix used together with CB in strength in the un-vulcanized state [5,10]. Furthermore, proteins and
various industrial applications. This is because NR has excellent green lipids influence mechanical properties, processability and cure charac­
strength in un-vulcanized state. Also, NR vulcanizates exhibit superior teristics (i.e., cure rate) of vulcanized NR [3,11]. It is noted that the
modulus, tensile strength, tear strength, and low heat build-up [3–5]. It fundamental rubber structures consist of long and branched chains with
has been claimed that the difference in mechanical properties between the proteins and lipids at the two types of terminals: ω-terminal and
synthetic and natural rubber is mainly due to different molecular α-terminal chain ends, respectively [12–14]. The non-rubber compo­
structures, and to the non-rubber components in NR [5]. It has been well nents (i.e., proteins and lipids) at the chain ends are origins of the
established that NR has high molecular weight because it consists mostly branching network (i.e, aggregation) based on interactions among pro­
of long-chain hydrocarbons with numerous repeating units of cis-1, teins and lipids via hydrogen bonding together with ionic linkages to
4-polyisoprene [6]. Also, the non-rubber components including pro­ form microgels [14]. Furthermore, macrogels are also formed by in­
teins, lipids, carbohydrates, and minerals, are other components in NR teractions of the microgels during prolonged storage of NR, which is
[7]. It was also found that properties and molecular characteristics of NR known as storage hardening [14]. This affects various properties of
together with types and levels of non-rubber components are mainly un-filled and filled NR composites including tensile strength, modulus,
dependent upon weather, landscape, soil and natural rubber genotype and other related properties [15].

* Corresponding author.
** Corresponding author.
E-mail addresses: [email protected] (S. Wisunthorn), [email protected] (C. Nakason).

https://doi.org/10.1016/j.polymertesting.2020.106623
Received 17 November 2019; Received in revised form 27 April 2020; Accepted 15 May 2020
Available online 19 May 2020
0142-9418/© 2020 Elsevier Ltd. All rights reserved.
P. Nun-anan et al. Polymer Testing 89 (2020) 106623

A number of studies have investigated effects of non-rubber com­ then formed and washed with clean water until it was free from the acid.
ponents on various properties of filled NR composites [16–18]. It was It was then dried in a hot air oven at 50 � C to obtain ADS. The ADS was
claimed that proteins had strong interactions with filler particles that then cut to small pieces and soaked in 5 %w/v NaOH for 72 h. The
reduced formation of filler-filler networks [17]. On the other hand, the rubber was then separated from the alkaline solution and washed
proteins hindered chemical and physical interactions of silica surfaces thoroughly with clean water, and then dried in a hot air oven at 50 � C for
with silane, reducing reinforcement efficiency and worsening silica 48 h. This alkaline treated sample is called low-protein natural rubber
dispersion [17]. In addition, the gel content in NR caused poorer CB (LPNR). The lipid content in LPNR was then selectively separated by
dispersion, decreased crosslink density, and loss of tensile strength [15]. Soxhlet extraction with acetone for about 48 h. Then the rubber sample
Filler-rubber interactions could also be improved by adding lipid (1–5 was washed several times with acetone and then water, and eventually
phr) into a natural rubber compound [16]. It was found that lipid dried in a vacuum oven at 30 � C for about one week to acetone-extracted
increased the number of entanglement points, which could promote natural rubber (AENR).
stronger interactions between CB and NR molecules. The lipid may have
affected strain-induced crystallization (SIC) in carbon black filled NR
composites by decreasing the crystallinity index [16]. The mechanical 2.3. Gel permeation chromatography
properties of filled deproteinized natural rubber (DPNR) and filled
synthetic cis-1,4-polyisoprene (IR) have been compared [19]. Similar The rubber sample (about 30 � 5 mg) was dissolved in 30 mL THF
tensile and tear strengths were observed for both rubbers at various stabilized with 2,6-di-tert-butyl-4-methylphenol (BHT). The rubber so­
crosslink densities, as the reinforcement by CB overshadowed other lution was then conditioned in dark at 25 � C for about one week. During
parameters affecting properties of NR composites [19]. In addition, ef­ this period, the solution was intermittently stirred for about 1 h/day.
fects of natural rubber type (SMR CV60 or SVR 3L) when filled with 30 The solution was then filtered through a 1.0 μm glass fiber filter
phr carbon black on properties of NR compound and vulcanizate were (Macherey-Nagel GmbH & Co. KG, Düren, Germany). The filtrate was
investigated [20]. It was found that the SVR 3L composites exhibited then injected into a size exclusion chromatography (SEC) system, which
greater tensile strength and recovery from loading force than those with consisted of an HLPC pump 1000, refractive index detector (RI) 2300/
SMR CV60 [20]. This was due to SVR 3L having longer NR chains with 2400, and ultraviolet detector (UV) 2500 (Wissenschaftliche Gera €te
more functionality (number of double bonds), which could form more GmbH, Berlin, Germany). In this work, poly(styrene) solution standards
crosslinks with other units, and hence contribute to higher crosslink were used to calibrate the columns. The measurements were done at 20
density than in the SMR CV60 [20].
�
C at mobile phase flow rate of approximately 1 ml/min. The number-
In the present work, effects of different NR types: air-dried sheet average molecular weight ðMn Þ and weight-average molecular weight
(ADS), low protein natural rubber (LPNR) based on alkaline treatment, ðMw Þ of NR molecules were eventually determined.
and acetone-extracted natural rubber (AENR) of the RRIM600 genotype
on various properties of gum and carbon black filled compounds and
2.4. Nitrogen contents
vulcanizates were investigated. In addition, various properties including
cure characteristic, hardness, crosslink density, mechanical, thermal,
The total nitrogen content of rubber was determined by the Kjeldahl
morphological, stress relaxation and dynamic properties together with
method using a Kjeltec auto 8400 Analyzer (Foss, Hilleroed, Denmark)
dispersion of CB in NR compounds and vulcanizates were also
as described elsewhere [21]. Firstly, a rubber sample of about 0.1 g was
investigated.
mixed with CuSO4: K2SO4 (0.65 g) mixture and then digested with 2.5
ml concentrated H2SO4. Then, the digested material was distilled after
2. Material and methods
adding alkaline solution (i.e., 67% w/v NaOH). The released ammonia
gas was collected in 4 w/v% boric acid in the Kjeltec automatic distilling
2.1. Materials
unit. Finally, the boric acid solution was titrated with 0.01 N H2SO4. The
nitrogen content estimate was multiplied by a factor of 6.25 to the
Three different types of natural rubber: air-dried sheet (ADS), low-
equivalent amount of protein.
protein natural rubber (LPNR), and acetone-extracted natural rubber
(AENR), were used to prepare gum and carbon black filled compounds
and vulcanizates. Acetone, toluene and sodium hydroxide (NaOH) were 2.5. Lipid analysis
manufactured by Fisher Scientific (Loughborough, UK). Tetrahydro­
furan (THF) and 2,6-di-tert-butyl-4-methylphenol (BHT), used to The rubber sample (about 5 g) was Soxhlent reflux extracted with
analyze molecular weight of NR, were manufactured by VWR (Fonte­ acetone (150 mL) as described elsewhere [22]. The solvent was removed
nay-sous-Bois, France) and Panreac (Barcelona, Spain), respectively. in a rotary evaporator at 60 � C, and the residue was dried to constant
The chemicals used to determine the protein content, sulfuric acid weight in a vacuum oven at 104 � C. The content of lipids was eventually
(H2SO4), potassium sulfate (K2SO4) and boric acid were manufactured calculated as follows:
by Merck KGaA (Darmstadt, Germany). Copper (II) sulfate (CuSO4) was
produced by VWR (Fontenay-sous-Bois, France). Carbon black (HAF) Lipids (%wt) ¼ 100 � (W1/W0) (1)
used as a reinforcing filler was manufactured by Thai Carbon Black
where W0 and W1 are are the initial weight of rubber sample and the
Public Co., Ltd. (Ang Thong, Thailand). Activator stearic acid was
final weight of extracted lipids.
manufactured by Imperial Industry Chemical Co., Ltd. (Pathum Thani,
Thailand), and Zinc oxide (ZnO) was manufactured by global Chemical
Co., Ltd. (Samut Prakan, Thailand). The cure accelerator, N-tert-butyl-2 2.6. Gel content
benzothiazole-sulfenamide (TBBS), was manufactured by Fluxys
Belgium (Fluxys, Brussels, Belgium). Finally, the curing agent sulfur was Gel content of gum NR was determined according to ISO 1166. First,
manufactured by Ajax Chemical Co., Ltd. (Samut Prakan, Thailand). gum rubber (about 0.1 g) was dissolved in 30 mL of toluene. It was
shaken for about 1 min and then conditioned at 25 � C for 20 h without
2.2. Preparation of ADS, LPNR and AENR stirring. Centrifugation at 14,000 g was then applied to the rubber so­
lution for about 2 h and the liquid fraction was removed. The remaining
Air dried sheets (ADS) were first prepared by acid coagulation of gel fraction was washed thoroughly with acetone and then dried at 110
natural rubber latex from the RRIM600 genotype. The rubber sheet was �
C for 1 h. Then, the gel content was calculated as follows:

2
P. Nun-anan et al. Polymer Testing 89 (2020) 106623

� �
W0 W1 � �
Gel ​ content ​ ð%wtÞ ¼ *100 (2) dEðTÞ
W0 HðTÞ ¼ T⋅ (4)
dT v¼const

where W1 and W0 are the dry weights after extraction and before
extraction, respectively. 2.9. Morphological properties

Morphological properties of gum and carbon black-filled NR vulca­

2.7. Preparation of NR vulcanizates nizates were determined using three techniques: optical microscopy,
scanning electron microscopy, and a carbon black dispersion tester. In
Various NR types (i.e., ADS, LPNR and AENR) from the RRIM600 optical microscopy the rubber samples were fast cut with a razor blade
genotype were compounded with and without carbon black (N-330), to create smooth surfaces before examining with an optical microscope
with the formulations shown in Table 1. The rubber sample was first (Carl Zeiss Microscopy GmbH, Oberkochen, Germany). In addition,
loaded into an internal mixer, a Brabender Plasticorder (Brabender high-resolution scanning electron microscopy (SEM) (ZeissSupra-40 VP,
GmbH & Co. KG, Duisburg, Germany), and mastication was then per­ Carl Zeiss Microscopy GmbH, Oberkochen, Germany) was also used to
formed for about 2 min at 60 � C and a rotor speed of 60 rpm. For the verify the morphological properties for the various types of rubber
carbon black-filled NR, 40 phr of CB, N-330, was incorporated and samples. For this imaging the sample was immersed in liquid nitrogen
mixing was continued for another 6 min. Then, the zinc oxide and stearic and then cracked to create a fresh surface, which was then sputter-
acid were added in sequence and mixed for 2 min. Finally, the curatives coated with a thin layer of gold under vacuum, before imaging with
(i.e., TBBS accelerator and sulfur) were added and mixing was continued SEM. Moreover, the dispersion of carbon black in rubber vulcanizate
for another 2 min to reach the total mixing time of about 12 min. The was also visualized using a carbon black dispersion tester, 3000 (Mon­
rubber compound was sheeted out on a two-roll mill (YFCR 600, Yong Tech Rubber Testing Solutions, Buchen, Germany), according to ISO
Fong machinery Co, Ltd., Samut Sakorn, Thailand) and then conditioned 11345.
at room temperature for at least 24 h before testing. Curing properties of
the rubber compounds were characterized with a rheometer, MDR 2000
2.10. Crosslink density
(Alpha technologies, Ohio, USA), and then vulcanizate sheets were
fabricated using a compression molding machine (PR1D-W400L450 PM,
Apparent crosslink density of the rubber vulcanizates is typically
Chareon Tut Co., Ltd, Samut Prakarn, Thailand) at 150 � C and 70 bar
assessed with a swelling method. Rubber samples with dimensions 10 �
compression pressure, according to the optimum cure time (tc90). It is
10 � 2 mm3 were immersed in toluene in a dark room at ambient
noted that the gum NR compounds (Table 1) were also prepared for
temperature for one week, to reach equilibrium swelling. The swollen
comparison purposes. In addition, the filler reinforcement index (α) was
samples were then weighed and excess solvent was wiped off from the
estimated from the maximum torque, as follows [23]:
sample surfaces with tissue paper. Each sample was then dried at 60 � C
� �� �
α ¼ ​ ΔLmaxðCBfilled compoundÞ ΔLmaxðgum compoundÞ ΔLmaxðgum compoundÞ (3) to a constant weight. The final dry weight of the rubber sample was
compared to the original dry weight before swelling in toluene. The
where ΔLmaxðCBfilled compoundÞ is the maximum change in torque for the CB- apparent crosslink density was calculated from the Flory-Rehner equa­
filled compound, and the ΔLmaxðgum compoundÞ is the maximum change in tion [27]:
torque of unfilled gum compound. �
lnð1 φρ Þ þ φρ þ χ φ2ρ
V ¼ 1 � (5)
V1 φ3ρ φρ2
2.8. Temperature scanning stress relaxation (TSSR)
where v is the crosslink density (mol/m3), ∅ρ is the volume fraction of
Temperature scanning stress relaxation measurement was carried rubber in a swollen network, V1 refers to the molar volume of toluene,
out with a TSSR meter (Brabender, Duisburg, Germany). It is noted that and χ is the Flory-Huggins interaction parameter between toluene and
TSSR has been used to characterize the relaxation behavior of rubber rubber, for which the value 0.4 was used according to prior literature
vulcanizates. In addition, the non-rubber components and aggregates [28].
(or gel) in NR could be estimated from the stress relaxation spectrum of For the filled NR vulcanizates, the volume fraction of filler was
natural rubber. In the test, the dumbbell-shaped specimen (type 5A, ISO subtracted from the volume fraction of rubber in a swollen network ðφρ Þ.
527) was placed in the sample holder in the TSSR chamber. Then 50% Thus, the volume fraction of rubber in a swollen network was deter­
strain was applied at 23 � C for about 2 h. The non-isothermal testing mined as follows:
over the temperature range from 23 � C to 220 � C had a constant 2 � C/
min heating rate, and the relaxation modulus E(T) for the constant φρ ¼ � �
1
��� � (6)
heating rate (v) was estimated from the stress-temperature curve. m md
1 þ md
ρp

Eventually, the relaxation spectrum H(T) was evaluated by differenti­

ρ s

ating E(T) with respect to temperature T, as follows [24–26].

where m is the mass of the swollen sample, md is the mass after drying
the sample, ρp is density of the polymer, and ρs is density of the solvent
Table 1
Formulation of gum and carbon black-filled NR compounds. [24].

Chemical Quantity (phr)

2.11. Mechanical properties
Gum Carbon black-filled
compound compound
Tensile properties of the rubber vulcanizates were measured with a
Natural rubber(ADS, LPNR and 100.0 100.0
Tinius Olse model 10ST tensile testing machine (Salfords, England). The
AENR)
Stearic acids 0.5 0.5 dumb-bell shaped specimens (ISO 527 die type 5A) were elongated at
ZnO 6.0 6.0 room temperature with an extension speed of 500 mm/min. It is noted
TBBS 0.7 0.7 that the reinforcing efficiency (RE) was evaluated from 100% modulus
Sulfur 3.5 3.5 difference between filled (i.e., M100, filled) and gum (i.e., M100, gum)
N-330 40.0
vulcanizates as related to the total weight of filler, as follows [24].
–

3
P. Nun-anan et al. Polymer Testing 89 (2020) 106623

RE ¼ ðM100 ; filled M100 ; gumÞ=wt%filler (7)

Furthermore, hardness of NR vulcanizates was also measured by
durometer Shore A according to ISO 2781 with a hardness testing ma­
chine (HT 3000, MonTech, Buchen, Germany).

2.12. Dynamic properties

Dynamic properties of the various NR samples were determined by

using Rubber Process Analyzer (RPA 2000, Alpha Technologies, Akron,
USA). The machine was operated at 80 � C and a fixed frequency of 1 Hz
in the strain amplitude range from 0.07% to 500%. In this work, storage
modulus as a function strain amplitude for the various types of NR
vulcanizate was investigated.

2.13. Thermal properties

Thermal properties of the various gum and filled NR vulcanizates

were characterized by dynamic mechanical analysis with DMA 8000
(PerkinElmer Inc., Waltham, USA). The DMA was operated in tension
mode in the temperature range from 100 � C to 100 � C at the fixed 1 Hz
Fig. 1. Stress (σ) as a function of temperature for gum NR (i.e., ADS, LPNR
frequency. Storage modulus and tan δ as functions of temperature were and AENR).
eventually determined, and the glass transition temperature (Tg) was
also estimated from tan δ peaks.
This might be due to the decomposition of network structures in NR,
which originated by phospholipids at α-terminal ends and by proteins at
3. Results and discussion
the ω-terminal ends of NR molecules, so that stress of gum NR decreased
[29,30].
3.1. Molecular characteristics
Moreover, two significant drops of stress are clearly observed at
about 30–40 � C and 40–80 � C, corresponding to two sharp peaks in the
Table 2 shows gel content and average molecular weights (i.e., M n relaxation spectrum, as shown in Fig. 2. The two distinct peaks in the
and Mw) for the three types of NR (i.e., ADS, LPNR and AENR). It is clear temperature ranges 30–40 � C and 40–80 � C are here labeled (A) and (B),
that the average molecular weights (i.e., Mn and Mw) of ADS, LPNR and respectively. It is anticipated that peak (A) may represent proteins at the
AENR had drastically decreased after alkaline treatment and acetone ω-terminals in the form of a loose network structure [29]. On the other
extraction. That is, the Mn was 1425, 461 and 116 kg/mol while Mw was hand, peak (B) at 40–80 � C may be attributed to the decomposition of
2144, 1311 and 353 kg/mol for ADS, LPNR and AENR, in this order. network at lipids at the α-terminals, together with the interaction at
Also, it is clear that higher molecular weight caused more gel or loosely proteins of ω-terminal ends of NR molecules [29,30]. In Fig. 2, it is clear
crosslinked structures. Therefore, treatment of NR by alkaline and that only the gum ADS shows the first peak (A) but the peak had dis­
acetone extractions caused decomposition of rubber molecules together appeared in LPNR and AENR. This implicates the proteins at ω-terminals
with loss of rubber network structures based on interactions at α- and in ADS, with comparatively strong chemical interactions by hydrogen
ω-terminal ends. This affects characteristics of the network structures bonding between the rubber molecules to form networks in ADS. It is
and the molecular weight of NR. noted that the alkaline treatment to form LPNR caused breakdown of
protein molecules and hence reduced active functional groups at the
3.2. TSSR of uncured gum rubbers terminal ends. Moreover, acetone extraction of LPNR to form AENR

Fig. 1 shows stress (σ ) as a function of temperature for the gum NR

samples (i.e., ADS, LPNR and AENR). It can be seen that ADS showed the
highest initial stress (σ 0) at 23 � C (about 0.10 MPa) while the gum LPNR
and AENR gave lower σ0 at about 0.04 and 0.02 MPa, respectively. It is
noted that the σ 0 indicates the stiffness of material. Therefore, the ADS
had higher stiffness than the LPNR and AENR samples. This might be
attributed to more proteins and lipids that increased stiffness and stress
relaxation [29]. In Fig. 1, the stress of gum ADS slightly decreased with
temperatures in the range 30–40 � C. This might be due to interaction by
proteins at the ω-terminal ends of the rubber molecules [29]. However,
constant stress was found for gum LPNR and AENR due to weaker in­
teractions because of having less proteins. In addition, the stress relax­
ation was clear for all types of NR in the temperature range 40–80 � C.

Table 2
Gel content and average molecular weight (i.e., Mn and Mw) for three types of NR
(i.e., ADS, LPNR and AENR).
Sample Gel content (%wt) Mn (kg/mol) Mw (kg/mol)

ADS 40.6 � 1.6 1425 � 53 2144 � 4

LPNR 31.2 � 0.2 461 � 35 1311 � 10
AENR 24.7 � 1.5 116 � 6 353 � 10
Fig. 2. Relaxation spectrum H(T) for gum NR (i.e., ADS, LPNR and AENR).

4
P. Nun-anan et al. Polymer Testing 89 (2020) 106623

further decreased proteins and lipids, leading to smaller networks in

AENR.
In Fig. 2, it is also clearly seen that the gum ADS showed the highest
and broadest peak (B) in the temperature range from 40 � C to 80 � C. The
peak area is about 2.77 MPa*K, compared to 0.97 MPa*K for LPNR and
0.71 MPa*K for AENR. This matches previous studies [30], which
indicated that the strongest chemical interaction were of lipids at the
α-terminals, and there were strong linkages among proteins at ω-ter­
minals. This low protein and lipid contents in LPNR and AENR caused
less network structures. The AENR has the lowest peak intensity due to
having the least proteins and lipids [30]. Also, the peaks are shifted to
lower temperatures upon removal of proteins and lipids in LPNR and
AENR, which indicates poorer thermal resistance. It can be concluded
from the TSSR results that the gum ADS had the most proteins, lipids and
gel content, while gum LPNR and AENR contained less proteins, lipids
and gel than gum ADS. These results for the various types of NR are
shown in Fig. 3, and are consistent with the gel content (Table 2). That
is, the gel content of ADS at about 41 %wt was remarkably decreased
after alkaline treatment and acetone extraction to form LPNR and AENR,
to 31 and 25 %wt, respectively. Fig. 4. Stress (σ) as a function of temperature for gum and carbon black-filled
ADS, LPNR and AENR vulcanizates.
3.3. TSSR of gum and CB-filled NR vulcanizates

Thermo-mechanical behavior in terms of stress relaxation of gum

and CB-filled NR vulcanizates was investigated by the TSSR technique,
with results shown in Figs. 4 and 5. In Fig. 4 it is seen that lower initial
stresses (σ0) are seen for the gum vulcanizate as compared to its CB-filled
counterpart (Table 3). Also, the initial stress slightly increased with
temperature due to entropy effects in the NR [15]. The initial stress may
be related to crosslink density (Table 4) of the rubber network in NR
vulcanizates [31]. In Table 4, among the three types of gum vulcani­
zates, it is confirmed that the ADS vulcanizate had the highest crosslink
density of about 139 mol/mm3 followed by the LPNR (109 mol/mm3)
and AENR (75 mol/mm3) vulcanizates. These results correlate well with
the slope of stress versus temperature curve in the linear region from
initial temperature to 50 � C (Fig. 4). It is clear that the slope of the curve
of the gum ADS vulcanizate is higher than those of the LPNR and AENR
vulcanizates, respectively. In Fig. 4, a higher initial stress is clearly
observed for the filled CB than for the gum vulcanizate counterpart. This
may indicate higher crosslink density (Table 4) and also reinforcement
by filler. The stress markedly decreased in the temperature range 40–80
�
C with peak (a) for the CB-filled ADS vulcanizate in Fig. 5. It is noted
that peak (a) is obvious only for the CB-filled ADS vulcanizate. This may
Fig. 5. Relaxation spectrum H(T) for gum and carbon black-filled of ADS, LPNR
be attributed to physical desorption of bound rubber (or rubber-filler
and AENR vulcanizates.
interactions) related to the dissociation of network structures

Table 3
Initial stress and T90 in the relaxation spectrum H(T) of gum and carbon black-
filled NR vulcanizates.
Sample Initial stress (σ0) T90

Gum ADS 0.41 195.9

Gum LPNR 0.26 186.0
Gum AENR 0.13 150.1
CB-ADS 0.81 193.5
CB-LPNR 0.70 158.2
CB-AENR 0.54 143.8

generated by the end groups (i.e., proteins and lipids). This corresponds
to the relaxation peak of the gum ADS in Fig. 2. However, peak (a) is
missing for the other unfilled vulcanizates and for CB-filled LPNR and
AENR vulcanizates (Fig. 5). This may be attributed to the smaller net­
works formed with very low contents of proteins and lipids. Further­
more, it might be attributed to poor interactions or NR with CB due to
low non-rubber content (i.e., proteins and lipids) in the LPNR and AENR.
Fig. 3. Protein and Lipid contents in ADS, LPNR and AENR made from RRIM In Fig. 5, the broad peak (b) with much greater area underneath than for
600 genotype latex.

5
P. Nun-anan et al. Polymer Testing 89 (2020) 106623

Table 4 followed by LPNR and AENR, in this order.

Crosslink densities of gum NR and CB-filled NR vulcanizates.
Sample Crosslink density (mol/m3)
3.4. Morphological properties
Gum ADS 139.03 � 4.95
Gum LPNR 109.41 � 1.89
Fig. 6 shows light microscope images of gum NR and CB-filled NR
Gum AENR 74.84 � 3.86
CB-ADS 257.89 � 4.11 vulcanziates. The dots in gum NR vulcanizates are ZnO particles
CB-LPNR 202.96 � 0.63 dispersed in the rubber matrix (Fig. 6(a), (b) and 6(c)). It is also seen that
CB-AENR 153.72 � 2.60 the gum ADS vulcanizate showed homogeneous dispersion of ZnO with
smooth surfaces (Fig. 6(a)). In addition, the LPNR vulcanizate (Fig. 6(b))
shows some large particles of ZnO whereas the AENR vulcanizate (Fig. 6
peak (a) was also observed for all vulcanizates. This might be attributed
(c)) showed even larger ZnO particles with more surface and larger
to thermo-oxidative degradation of rubber molecules, cleavage of the
voids. In Fig. 6, it is also seen that the CB-filled ADS vulcanizate showed
network bridges, and related interactions among the rubber chains [32,
more homogeneous dispersion of CB in the matrix (Fig. 6(d)). This can
33]. Clearly for each type of rubber the filled vulcanizate had higher
be explained by the rubber-filler interactions via chemical interaction
peak and larger area underneath it, due to reinforcement by CB filler.
between the active functional groups (i.e., proteins and/or lipids) at the
Among the gum vulcanizates, the ADS vulcanizate showed the tallest
NR molecule terminal ends with a functional groups on CB particle
peak with largest area underneath it, followed by the LPNR and AENR
surface. However, large CB agglomeration is seen in LPNR and AENR
vulcanizates. This is definitely due to the stronger network structures in
matrices (Fig. 6(e) and (f)). This might be attributed to lower molecular
ADS, because of interactions of proteins and lipids at the terminal ends.
weight, and less proteins and lipids in treated NR (i.e., in LPNR and
This correlates well with height and area underneath peak (B), in Fig. 2,
AENR) causing poor dispersion of CB in the rubber matrix. The light
that reflect the thermal stability of these materials [34].
microscope results match well the SEM micrographs in Fig. 7. That is,
In Fig. 5 the peak locations for LPNR and AENR vulcanizates are
small white dots of ZnO particles are seen in gum ADS (Fig. 7(a)) while
shifted toward lower temperatures from that of the gum ADS vulcani­
gum LPNR (Fig. 7(b)) and AENR (Fig. 7(c)) vulcanizates showed larger
zate, suggesting poorer thermal stability [34]. This is due to lower
ZnO clusters. This might be due to higher molecular weight and more
crosslink density and lesser quantity of proteins and lipids that act as
proteins and lipids, and also more gel, in ADS that improved the
antioxidants in NR [35]. Furthermore, the temperature of peak (b) for
dispersion of ZnO.
the CB-filled vulcanizates shifted to lower temperatures from that of the
Furthermore, homogeneous CB dispersion with some small aggre­
gum vulcanizate counterpart. This indicates poorer thermal resistance of
gates was observed in CB fllled ADS vulcanizate (Fig. 7(d)). However,
the CB-filled vulcanizate, which might be due to heat conduction by CB
the LPNR vulcanizate (Fig. 7(e)) had some regions of homogeneous
that could promote thermal degradation of the polymer networks at a
dispersion together with CB agglomerates. In (Fig. 7(f)), it is clear that
lower temperature than in the neat polymer [36]. In Fig. 5, it is also seen
the AENR vulcanizate showed poorer CB dispersion with larger CB ag­
that the CB-filled ADS showed higher peak temperature than CB-filled
glomerates. This is probably due to low viscosity (i.e., low molecular
LPNR and AENR vulcanizates. This better thermal stability of ADS
weight) and poor chemical interactions via proteins and lipids in LPNR
vulcanizate was due to stronger interactions of chain ends of NR mole­
and AENR contributing to CB agglomeration. The light microscope and
cules giving stronger structure and good dispersion of CB in the rubber
SEM results also match well the results from carbon black dispersion
matrix (Figs. 6–8), which enhances the interactions of NR with CB that
tester, in Fig. 8. This confirms the CB-filled ADS vulcanizate (Fig. 8(a))
contribute to crosslink density (Table 4). The thermal stability of NR
has the finest CB dispersion with small aggregates, and dispersion degree
vulcanizates corresponds to T90 (i.e., the temperature at which stress had
of about 94.79%. On the other hand, the CB-filled LPNR vulcanizate
decreased by 90% from the initial stress), shown in Table 3. That is, the
(Fig. 8(b)) had some CB-agglomerates with dispersion degree of about
best thermal stability in terms of T90 was found for the ADS vulcanizate
~93.16%. Also, AENR vulcanizate in Fig. 8(c) had CB dispersion of

Fig. 6. Light microscope images of gum ADS (a), gum LPNR (b), gum AENR (c), CB-ADS (d), CB-LPNR (e), and CB-AENR (f).

6
P. Nun-anan et al. Polymer Testing 89 (2020) 106623

Fig. 7. SEM images of gum ADS (a), gum LPNR (b), gum AENR (c), CB-ADS (d), CB-LPNR (e), and CB-AENR (f) (15000x).

Fig. 8. Carbon black dispersion in ADS (a), LPNR (b), and AENR compounds.

about 88.46% with the largest agglomerates.

3.5. Cure characteristics

Fig. 9 shows cure curves for gum and carbon black-filled ADS, LPNR
and AENR vulcanizates. Also, Table 5 summarizes cure characteristics
and reinforcement index (α) of gum and carbon black-filled NR vulca­
nizates. It can be seen that the gum ADS compound exhibited a long
delay in sulfur vulcanization with the longest scorch and cure times
(Table 5), while the LPNR and AENR gum compounds showed shorter
scorch and cure times than the gum ADS compound. This may be due to
residues of free fatty acids (i.e., stearic acids), glycerides and sodium
salts of fatty acid from the alkaline treatment [37]. In addition, the
hydrolysis of saponifiable lipids could give rise to the free fatty acids
forms (i.e., stearic acids) after the acetone extraction [38]. Moreover,
Heryanto et al. [39] and Tanaka [40] reported low solubility of stearic
acid in acetone and hence the residual lipids (i.e., especially stearic
acids) remained in AENR after acetone extraction. Therefore, the re­
sidual lipids (i.e., especially stearic acids) remained in LPNR and AENR
that might act as cure activators during vulcanization [41–43]. In Fig. 9. Cure curves for gum and carbon black-filled ADS, LPNR and AENR
Table 5, it is also seen that among the gum rubbers, the gum ADS vulcanizates.
compound gave the highest minimum (ML) and maximum torque (MH)
and largest torque difference (MH-ML), followed by the gum LPNR and
AENR compounds in this order. This matches the order of molecular
weights and of gel contents (Table 2) and hence also protein and lipid

7
P. Nun-anan et al. Polymer Testing 89 (2020) 106623

Table 5
Cure characteristics and reinforcement index (α) of gum and carbon black-filled NR vulcanizates.
Sample Minimum torque Maximum torque Torque difference (ΔL, Reinforcement index Scorch time (ts2, Cure time (t90, Cure rate index
(ML) (MH) dNm) (α) min) min) (CRI)

Gum-ADS 1.37 8.00 6.63 – 6.57 11.95 18.59

Gum- 1.16 7.51 6.35 – 2.94 7.11 23.98
LPNR
Gum- 0.76 6.87 6.11 – 2.75 7.10 22.99
AENR
CB-ADS 2.96 17.04 14.08 1.12 2.52 6.12 27.78
CB-LPNR 1.67 14.91 13.24 1.09 2.20 7.46 19.01
CB-AENR 1.63 8.77 7.14 0.17 1.24 12.22 9.11

contents (Fig. 3). It is noted that the torque difference typically indicates 3.6. Mechanical properties
crosslink density of a rubber network. Clearly the crosslink density
(Table 4) is related to the torque difference in Table 5, and these are Fig. 10 shows stress-strain curves of gum and CB-filled NR vulcani­
closely related to the non-rubber content in Fig. 3. By contrast, the AENR zates. Also, Table 6 summarizes tensile properties based on stress-strain
exhibited the lowest MH, ML and MH-ML due to the lowest molecular curves (Fig. 10) together with hardness (Shore A) of the three types of
weight and gel content (Table 2), together with lowest non-rubber gum and CB-filled NR vulcanizates. It is clear that the ADS gave the
contents (Fig. 3). highest tensile strength, moduli and hardness, surpassing the LPNR and
In Fig. 9, it is also seen that CB filler shortened scorch time for all AENR vulcanizates. However, the elongation at break of the gum ADS
types of CB filled NR compounds (Table 5). This might be due to more vulcanizate was below those of LPNR and AENR vulcanizates. This is due
heat being generated during mixing of the filled compounds, increasing to the higher average molecular weight (i.e., Mn and Mw) (Table 2) and
the cure rates in the early stages of vulcanization [44]. Also, chemical higher non-rubber contents (Fig. 3) of the ADS that favor strain-induced
interactions between NR molecules and polar functional groups on CB crystallization (SIC)) of the molecular networks. This increased strength
surfaces have shortened scorch times. However, the CB-filled AENR and but decreased maximum strain [42]. In Fig. 10, a lower SIC of gum LPNR
LPNR had longer cure times than their gum compound counterparts. and AENR vulcanizates is observed from the lesser gradients of their
This might be attributed to low non-rubber content and dispersion of CB stress-strain curves. Therefore, these rubber matrices had poorer ability
in rubber matrix. That is, the low contents of proteins and lipids in gum to crystallize after removal of proteins and lipids (Fig. 3) and with lower
LPNR and AENR caused poor dispersion of CB in the rubber matrices molecular weights (Table 2). SIC of natural rubber might be induced by
(Figs. 6–8) which delayed cure times of their CB-filled vulcanizates. In lipids [5], proteins, and long regular chains [49].
contrast, higher protein and lipid contents in the case of gum ADS In Table 6, it is clear that CB filler reinforced rubber samples, as
improved CB dispersion (Figs. 6–8), which reduced cure time of the indicated by increased moduli, tensile strength, hardness, and crosslink
CB-filled ADS vulcanizates. density (Table 4) from those of the gum NR vulcanizates. Therefore,
In Table 5, it is also seen that the CB-filled NR vulcanizates provided particulate reinforcement by fine particles dispersed as a discontinuous
higher ML, MH and MH-ML than the corresponding gum vulcanizates. phase in the matrix was obvious (Figs. 6–8). The polar functional groups
This is attributed to the reinforcement by CB filler, supported by its on CB interacted with NR via –C– – C- double bonds, and with the polar
physical and chemical interactions with NR [45]. Also, sulfur on the functional groups of protein or lipid moieties at the terminal chain ends.
surface of CB might increase chemical crosslinking with rubber [46]. It is In Table 6, it is also seen that the CB-filled ADS vulcanizate showed the
noted that higher MH-ML indicates a higher crosslink density (Table 4) highest tensile strength, moduli, and hardness, superior to the CB- filled
and increased stiffness from restricted mobility of rubber molecules LPNR and AENR vulcanizates. This relates to higher crosslink density
[47]. The CB-filled ADS gave the highest MH-ML and crosslink density (Table 4), reinforcement index (Table 5), molecular weight, and gel
(Table 4) followed by CB-LPNR and CB-AENR vulcanizates in this order. content (Table 2). Also, finer and smaller CB aggregates were seen in the
Furthermore, MH-ML and crosslink density had the same trend as the ADS vulcanizate (Figs. 6–8).
reinforcement index (α). It is noted that the reinforcement index (α) was
used to indicate dispersion and morphology of the filler in rubber matrix
[48]. That is, a higher α value in the CB-filled ADS vulcanizate may be
due to finer dispersion of CB. On the other hand, CB-filled LPNR and
AENR vulcanizates had lower α values indicating poor dispersion of CB
with strong agglomeration in the rubber matrix. It is clear that α
(Table 5) is in agreement with the degree of carbon black dispersion
(Fig. 8).
In Table 5 the CB-filled ADS compound gave a higher cure rate than
the other types of rubber, with the highest CRI at 27.78 s 1. This is due
to the finer dispersion of carbon black in ADS matrix (Figs. 6–8) giving
the highest thermal conductivity for the ADS compound, which
contributed to cure reactions of rubber molecules to form stronger
networks. However, CB aggregates were observed in CB-filled LPNR
compound and even larger agglomerates were observed in the AENR
compounds. This significantly increased the cure times (t90) (Table 5).
The large CB agglomerates obstructed heat transfer that activates curing
of NR (i.e., LPNR or AENR). Therefore, good dispersion of carbon black
(Figs. 6–8) speeds up curing reactions and improves interactions at
carbon black surfaces, hence improving also the reinforcement index
(Table 5).
Fig. 10. Stress-strain curves for NR gum and carbon black-filled NR vulcani­
zates of ADS, LPNR and AENR.

8
P. Nun-anan et al. Polymer Testing 89 (2020) 106623

Table 6
Mechanical properties of gum vulcanizates and carbon black-filled NR vulcanizates of ADS, LPNR and AENR.
Sample Young’s modulus 100% modulus Reinforcing efficiency 300% modulus Tensile strength Elongation at break Hardness (Shore
(MPa) (MPa) (RE) (MPa) (MPa) (%) A)

Gum ADS 0.23 � 0.06 0.72 � 0.01 – 1.86 � 0.05 19 � 0 764 � 21 36.5 � 0.4
Gum 0.19 � 0.02 0.60 � 0.08 – 1.19 � 0.05 15 � 0 946 � 9 19.1 � 0.8
LPNR
Gum 0.18 � 0.03 0.47 � 0.02 – 0.90 � 1.05 8�1 888 � 10 14.0 � 0.2
AENR
CB-ADS 0.51 � 0.02 2.20 � 0.15 0.056 12.92 � 1.05 30 � 1 491 � 15 57.5 � 0.5
CB-LPNR 0.46 � 0.01 1.89 � 0.19 0.049 11.58 � 1.24 24 � 0 463 � 22 53.3 � 0.6
CB-AENR 0.29 � 0.02 1.37 � 0.03 0.034 7.25 � 0.59 16 � 1 472 � 25 49.6 � 0.6

3.7. Dynamic properties

Table 7
Minimum and maximum moduli, difference between maximum and minimum
Fig. 11 shows storage modulus (G ) as a function of strain amplitude
0
moduli, as well as difference between difference moduli of CB filled and gum
for gum and carbon black-filled NR vulcanizates of ADS, LPNR and vulcanizates (Payne effect) of the gum and CB filled compounds.
AENR. Linear storage moduli are seen for all rubber vulcanizates in the
Sample G’∞ (MPa) G’0 (MPa) Payne effect of filled NR
low strain amplitude regime (i.e., <1.0%). However, the moduli
ΔG’
(MPa) (ΔG00 )/MPa
abruptly declined in the strain amplitude range 3–100% for CB-filled
Gum ADS 0.18 0.11 0.07
compounds and 30–100% for gum compounds. This may be due to –
Gum 0.15 0.10 0.05 –
breakdown of filler-filler networks and disentanglement of rubber net­ LPNR
works, known as the Payne effect [50,51]. It is noted that the Payne Gum 0.09 0.06 0.03 –
effect can be quantified by the difference between maximum and min­ AENR
CB-ADS 0.33 0.09 0.24 0.17
imum moduli (G’0 →G’∞ ) as follows [52,53]:
CB-LPNR 0.30 0.11 0.19 0.14
ΔG’ ¼ ​ G’0 G’∞ (8) CB-AENR 0.24 0.09 0.15 0.12

ΔG’ ¼ ΔG’CB filled vulcanizate ΔG’gum NR vulcanizate (9) modulus. The highest molecular weight (i.e., more long rubber mole­
cules), gel and level of non-rubber components in gum NR which could
where G’0and G’∞
are the maximum and minimum moduli, and ΔG 00
originate a large branching points by the end groups (i.e., proteins and
represents the Payne effect due to filler-filler interactions.
lipids) and hence more entanglement improved dispersion of CB and
Table 7 summarizes the ΔG’ (Eq. (8)) and ΔG’’ (Eq. (9)) of gum NR
gave larger Payne effect, as schematically proposed in Fig. 12(a). This is
and CB-filled NR compounds. It is seen that gum ADS gave high ΔG0 of
reflected by the ΔG00 of about 0.17 MPa of the CB-filled ADS compound
about 0.07 while gum LPNR and AENR compounds had lower values of
(Table 7). Furthermore, the removal of non-rubber components in LPNR
about 0.05 and 0.03 MPa, respectively. This indicates more chain
and AENR matrices caused reduction of ΔG00 to 0.14 and 0.12 MPa,
entanglement and/or interactions among ADS molecules, as a result of
respectively (Table 7). This might be attributed to low chemical in­
higher molecular weight and more gel (Table 2) together with larger
teractions between rubber molecules and less filler-filler interaction due
protein and lipid contents (Fig. 3). It is noted that the Payne effect of CB
to poorer dispersion of CB with larger agglomerates. Also, lower mo­
filled NR calculated based on eq. (9) was 0.17, 0.14 and 0.12 MPa for CB
lecular weight (i.e., more linear molecules), gel, protein and lipid con­
filled ADS, LPNR and AENR compounds, respectively. This reflects the
tent in gum NR after alkaline treatment and acetone extraction were
stronger rubber-filler and filler-filler interactions in ADS with fine
found. This reduction of branching points and chain entanglement
dispersion of CB (Figs. 6–8). This may be attributed to a synergistic effect
lowered ΔG00 in the CB-filled LPNR and AENR vulcanizates. This also
of rubber-rubber, rubber-filler and filler-filler networks on the elastic
allows slippage and disentanglement, as described in Fig. 12(b) and (c).
Thus, it is concluded that the dynamic properties of gum NR and CB
filled NR compounds are closely related to the filler-rubber interactions
and molecular structures (i.e., molecular weight, gel and non-rubber
components). These results are correlated well to molecular weight,
gel content (Table 2) and quantity of non-rubber content (i.e., proteins
and lipids) (Fig. 3).

3.8. Thermal properties

Dynamic mechanical analysis (DMA) was used to investigate the

thermal properties in terms of storage modulus (G0 ) and tan δ as func­
tions of temperature, with results shown in Fig. 13. High relatively
constant G0 in the glassy region at temperatures below 70 � C was
observed. In addition, the CB-filled NR vulcanizates showed higher G0
than the gum vulcanizates. This might be due to the rigid nature of CB
and the mobility restrictions it placed on rubber molecules [54]. G0
abruptly decreased in the temperature range from 50 � C to 40 � C,
indicating glass transition (Fig. 13(a)). The maximum change in
modulus provides the glass transition temperature (Tg) at the peak
location of tan δ (Fig. 13(b)), as summarized in Table 8. Clearly the gum
Fig. 11. Storage modulus (G0 ) as a function of strain amplitude for NR gum and vulcanizates had lower Tg than their CB filled counterparts. This might
carbon black-filled NR vulcanizates of ADS, LPNR and AENR. be attributed to higher flexibility of the gum NR than of the CB-filled

9
P. Nun-anan et al. Polymer Testing 89 (2020) 106623

Fig. 12. A proposed a model of carbon black dispersion in various NR vulcanizates: CB-ADS (a), CB-LPNR (b) and CB-AENR (c).

to stronger filler-rubber interactions. That is, ADS with high molecular

weight, gel content (Table 2), protein and lipid content (Fig. 3), finer
distribution and dispersion of CB (Figs. 6–8) had more restricted mo­
lecular mobility and hence higher Tg. In Table 8, it is also seen that the
largest area underneath the tan δ peak is that for the gum ADS vulcan­
ziate, followed by the LPNR and AENR vulcanziates. This indicates that
more energy was needed to transform the rubber from glassy to rubbery
state, because the ADS has more sulfur bridges and more non-rubber
components that could act as reinforcing fillers. Furthermore, stronger
rubber-filler interactions caused larger area underneath the tan δ peak
for the CB- filled ADS vulcanizate (i.e., 33.6 cm2), whereas the CB-filled
LPNR and AENR vulcanizates had lower values of about 32.5 and 30.4
cm2, respectively. In Table 8, lower area underneath tan δ peak was
obseved with filler that restricted mobility of the rubber molecules [36].

4. Conclusions

NR vulcanizates made of air dried sheet (ADS), low protein natural

rubber (LPNR) from alkaline treatment, or acetone extracted natural
rubber (AENR) were prepared with and without carbon black filler.
Molecular weights, non-rubber contents (i.e., proteins and lipids) and
Fig. 13. Storage modulus (G0 ) (a), and Tan δ peaks (b) as functions of tem­
gel contents were characterized. The results show that ADS had higher
perature for gum and carbon black-filled NR vulcanizates of ADS, LPNR
and AENR. molecular weight, and more proteins, lipids and gel than LPNR and
AENR. Lower proteins and lipids in LPNR and AENR gave shorter scorch
and cure times of the rubber compounds. This may be due to the free
Table 8 fatty acids (i.e., stearic acids), glycerides and sodium salts of fatty acid
Glass transition temperature (Tg) and area underneath the tan δ peaks for gum acting as activators of sulfur vulcanization. Furthermore, well-dispersed
and CB-filled NR vulcanizates. CB was found in ADS. This may be attributed to strong chemical in­
Sample tan (δ) teractions of CB with ends of rubber molecules via proteins and lipids.
Moreover, higher molecular weight and larger quantity of gel in ADS
Tg (oC) Area underneath of the tan (δ) peak (cm2)
could increase the reinforcement index (α). This relates to the excellent
Gum ADS 47.4 42.7
properties of ADS as regards thermo-mechanical, mechanical, hardness,
Gum LPNR 48.9 40.6
Gum AENR 49.9 40.6 dynamic, and dynamic mechanical properties, and crosslink density. In
CB-ADS 45.7 33.6 contrast, large CB agglomerates were found with matrices of lower
CB-LPNR 46.4 32.5 molecular weight, and lesser quantities of proteins and lipids (i.e., LPNR
CB-AENR 48.1 30.4 and AENR vulcanizates) giving inferior mechanical, dynamic, thermo-
mechanical, dynamic mechanical, and stiffness properties, and cross­
vulcanizates. On the other hand, the rubber-filler interactions restricted link density.
mobility of rubber molecules. In addition, the CB-filled ADS had higher
Tg than CB-filled LPNR and AENR vulcanizates. This might be associated

10
P. Nun-anan et al. Polymer Testing 89 (2020) 106623

Data availability vulcanizates: gum and black-filled vulcanizates, J. Appl. Polym. Sci. 97 (3) (2005)
1139–1144.
[20] Y. Phanny, A. Azura, H. Ismail, Effect of different origins of natural rubber on the
The raw/processed data required to reproduce these findings cannot properties of carbon black filled natural rubber composites, ASEAN Eng. J. Part B 2
be shared at this time, as the data also forms part of an ongoing study. (1) (2012) 60–67.
[21] E. Aik-Hwee, S. Kawahara, Y. Tanaka, Determination of low nitrogen content of
purified natural rubber, J. Nat. Rubber Res. 8 (2) (1993), 109-109.
Declaration of competing interest [22] P. Manirakiza, A. Covaci, P. Schepens, Comparative study on total lipid
determination using Soxhlet, Roese-Gottlieb, Bligh & Dyer, and modified Bligh &
Dyer extraction methods, J. Food Compos. Anal. 14 (1) (2001) 93–100.
The authors declare the following financial interests/personal re­ [23] B. Boonstra, Rubber technology and manufacture, in: second ed., in: C.M. Blow,
lationships which may be considered as potential competing interests: C. Hepburn (Eds.), Butterworth, London, for the Plastics and Rubber Institute, vol.
22, 1982, p. 28.
[24] N. Vennemann, K. B€ okamp, D. Br€ oker, Crosslink density of peroxide cured TPV,
Acknowledgement Macromol. Symp. 245 (1) (2006) 641–650.
[25] N. Vennemann, M. Wu, M. Heinz, Thermoelastic properties and relaxation
The authors gratefully acknowledge the financial support from behavior of S-SBR/silica vulcanizates, Rubber World 246 (6) (2012) 18–23.
[26] V. Norbert, Characterization of Thermoplastic Elastomers by Means of
Prince of Songkla University, Contract no. S&T550141S and from Prince Temperature Scanning Stress Relaxation Measurements, Thermoplastic Elastomers,
of Songkla University, Surat Thani Campus, Thailand, Contract no. PSU/ Rijeka, Croatia, 2012, pp. 347–370.
002/2561. Furthermore, we would like to thank Dr. Chatchamon [27] P.J. Flory, J. Rehner Jr., Statistical mechanics of cross-linked polymer networks I.
Rubberlike elasticity, J. Chem. Phys. 11 (11) (1943) 512–520.
Dangkanit Nataworn and Rubber Authority of Thailand, Surat Thani [28] P.J. Flory, Molecular size distribution in three dimensional polymers. II.
province, Thailand for kind support as regards access to facilities and Trifunctional branching units, J. Am. Chem. Soc. 63 (11) (1941) 3091–3096.
equipment. Also, assistance in manuscript preparation by Assoc. Prof. [29] M. Wu, M. Heinz, N. Vennemann, Investigation of un-vulcanized natural rubber by
means of temperature scanning stress relaxation measurements, Adv. Mater. Res.
Dr. Seppo Karrila is acknowledged. 718–720 (2013) 117–123.
[30] P. Nun-anan, S. Wisunthorn, S. Pichaiyut, C.D. Nathaworn, C. Nakason, Influence
References of nonrubber components on properties of unvulcanized natural rubber, Polym.
Adv. Technol. 31 (1) (2020) 44–59.
[31] M. Narathichat, C. Kummerl€ owe, N. Vennemann, C. Nakason, Thermoplastic
[1] A. Mostafa, A. Abouel-Kasem, M. Bayoumi, M. El-Sebaie, Insight into the effect of
natural rubber based on polyamide-12: influence of blending technique and type of
CB loading on tension, compression, hardness and abrasion properties of SBR and
rubber on temperature scanning stress relaxation and other related properties,
NBR filled compounds, Mater. Des. 30 (5) (2009) 1785–1791.
J. Appl. Polym. Sci. 121 (2) (2011) 805–814.
[2] S.S. Choi, B.H. Park, H. Song, Influence of filler type and content on properties of
[32] P. Nun-anan, S. Wisunthorn, S. Pichaiyut, N. Vennemann, C. Nakason, Novel
styrene-butadiene rubber (SBR) compound reinforced with carbon black or silica,
approach to determine non-rubber content in Hevea brasiliensis: influence of clone
Polym. Adv. Technol. 15 (3) (2004) 122–127.
variation on properties of un-vulcanized natural rubber, Ind. Crop. Prod. 118
[3] K. Suchiva, T. Kowitteerawut, L. Srichantamit, Structure properties of purified
(2018) 38–47.
natural rubber, J. Appl. Polym. Sci. 78 (8) (2000) 1495–1504.
[33] N. Vennemann, C. Schwarze, C. Kummerl€ owe, Determination of crosslink density
[4] S. Toki, B.S. Hsiao, S. Amnuaypornsri, J. Sakdapipanich, New insights into the
and network structure of NR vulcanizates by means of TSSR, Adv. Mater. Res. 844
relationship between network structure and strain-induced crystallization in un-
(2014) 482–485.
vulcanized and vulcanized natural rubber by synchrotron X-ray diffraction,
[34] A. Barbe, K. B€okamp, C. Kummerl€ owe, H. Sollmann, N. Vennemann, S. Vinzelberg,
Polymer 50 (9) (2009) 2142–2148.
Investigation of modified SEBS-based thermoplastic elastomers by temperature
[5] N.T. Thuong, O. Yamamoto, P.T. Nghia, K. Cornish, S. Kawahara, Effect of
scanning stress relaxation measurements, Polym. Eng. Sci. 45 (11) (2005)
naturally occurring crosslinking junctions on green strength of natural rubber,
1498–1507.
Polym. Adv. Technol. 28 (3) (2017) 303–311.
[35] M. Morimoto, Effect of nonrubber ingredients in natural rubber on aging
[6] Y. Tanaka, Structural characterization of natural polyisoprenes: solve the mystery
properties. Proceedings in International Rubber Conference, Kuala Lumpur, 1985,
of natural rubber based on structural study, Rubber Chem. Technol. 74 (3) (2001)
pp. 61–72.
355–375.
[36] W. Pongdong, C. Kummerl€ owe, N. Vennemann, A. Thitithammawong, C. Nakason,
[7] E. Hwee, Y. Tanaka, Structure of natural rubber, Trends Polym. Sci. 3 (1993)
A comparative study of rice husk ash and siliceous earth as reinforcing fillers in
493–513.
epoxidized natural rubber composites, Polym. Compos. 39 (2) (2018) 414–426.
[8] P.D. Galiani, M.A. Martins, P. de Souza Gonçalves, C.M. McMahan, L.H. Capparelli
[37] D. Swern, Bailey’s Industrial Oil and Fat Products, Wiley-Blackwell, 1982.
Mattoso, Seasonal and clonal variations in technological and thermal properties of
[38] S.-F. Chen, C. Ng, The Natural Higher Fatty Acid Soaps in Natural Rubber Latex and
raw Hevea natural rubber, J. Appl. Polym. Sci. 122 (4) (2011) 2749–2755.
Their Effect on the Mechanical Stability of the Latex, a Meeting of the Rubber
[9] N. Phupewkeaw, J. Sakdapipanich, Natural rubbers by saponification: physical
Division, American Chemical Society, 1983.
property characterization of uncured and cured low-protein skim, KGK-Kaut.
[39] R. Heryanto, M. Hasan, E.C. Abdullah, A.C. Kumoro, Solubility of stearic acid in
Gummi. Kunst. 64 (3) (2011) 27–32.
various organic solvents and its prediction using non-ideal solution models, Sci.
[10] K. Nawamawat, J.T. Sakdapipanich, C.C. Ho, Effect of deproteinized methods on
Asia 33 (2007) 469–472.
the proteins and properties of natural rubber latex during storage, Macromol.
[40] A. Tanaka, Distribution and origin of abnormal groups in natural rubber, J. Rubber
Symp. 288 (1) (2010) 95–103.
Res. 12 (1) (1997) 11–20.
[11] W. Smitthipong, R. Tantatherdtam, K. Rungsanthien, P. Suwanruji, S. Klanarong,
[41] M. Chukwu, I. Madufor, M. Ayo, L. Ekebafe, Effect of stearic acid level on the
S. Radabutra, S. Thanawan, M.F. Vallat, M. Nardin, K. Mougin, Effect of non-
physical properties of natural rubber vulcanisate, Pac. J. Sci. Technol. 12 (1)
rubber components on properties of sulphur crosslinked natural rubbers, Adv.
(2011) 344–350.
Mater. Res. 844 (2014) 345–348.
[42] B. Poh, E. Tan, Mooney scorch time and cure index of epoxidized natural rubber in
[12] A. Eng, S. Kawahara, Y. Tanaka, Trans-isofrene units in natural rubber, Rubber
presence of sodium carbonate, J. Appl. Polym. Sci. 82 (6) (2001) 1352–1355.
Chem. Technol. 67 (1) (1994) 159–168.
[43] S. Amnuaypornsri, J. Sakdapipanich, Y. Tanaka, Highly purified natural rubber by
[13] L. Tarachiwin, Y. Tanaka, J. Sakdapipanich, Structure and origin of long-chain
saponificaion of latex: analysis of green and cured properties, J. Appl. Polym. 118
branching and gel in natural rubber, KGK-Kaut. Gummi. Kunst. 58 (3) (2005)
(6) (2010) 3524–3531.
115–122.
[44] V. Geethamma, R. Joseph, S. Thomas, Short coir fiber-reinforced natural rubber
[14] L. Tarachiwin, J. Sakdapipanich, K. Ute, T. Kitayama, Y. Tanaka, Structural
composites: effects of fiber length, orientation, and alkali treatment, J. Appl.
characterization of α-terminal group of natural rubber. 2. Decomposition of
Polym. Sci. 55 (4) (1995) 583–594.
branch-points by phospholipase and chemical treatments, Biomacromolecules 6 (4)
[45] J.E. Mark, Physical Properties of Polymers Handbook, Springer, New York, 2007.
(2005) 1858–1863.
[46] Z. Li, J. Zhang, S. Chen, Effects of carbon blacks with various structures on
[15] A. Nimpaiboon, S. Amnuaypornsri, J. Sakdapipanich, Influence of gel content on
vulcanization and reinforcement of filled ethylene-propylene-diene rubber, Express
the physical properties of unfilled and carbon black filled natural rubber
Polym. Lett. 2 (10) (2008) 695–704.
vulcanizates, Polym. Test. 32 (6) (2013) 1135–1144.
[47] N. Hayeemasae, H. Ismail, Improving the tensile properties of natural rubber
[16] T. Xu, Z. Jia, L. Wu, Y. Chen, Y. Luo, D. Jia, Z. Peng, Effect of acetone extract from
compounds containing ground ethylene propylene diene rubber waste by two-stage
natural rubber on the structure and interface interaction in NR/CB composites, RSC
processing, Procedia Chem. 19 (2016) 810–815.
Adv. 7 (42) (2017) 26458–26467.
[48] A. Menon, C. Pillai, G. Nando, Physicomechanical properties of filled natural
[17] S. Sarkawi, W.K. Dierkes, J.W. Noordermeer, The influence of non-rubber
rubber vulcanizates modified with phosphorylated cashew nut shell liquid, J. Appl.
constituents on performance of silica reinforced natural rubber compounds, Eur.
Polym. 68 (8) (1998) 1303–1311.
Polym. J. 49 (10) (2013) 3199–3209.
[49] S. Amnuaypornsri, L. Tarachiwin, J.T. Sakdapipanich, Character of long-chain
[18] N. Rattanasom, A. Poonsuk, T. Makmoon, Effect of curing system on the
branching in highly purified natural rubber, J. Appl. Polym. Sci. 115 (6) (2010)
mechanical properties and heat aging resistance of natural rubber/tire tread
3645–3650.
reclaimed rubber blends, Polym. Test. 24 (6) (2005) 728–732.
[19] N. Rattanasom, U. Thammasiripong, K. Suchiva, Mechanical properties of
deproteinized natural rubber in comparison with synthetic cis-1, 4 polyisoprene

11
P. Nun-anan et al. Polymer Testing 89 (2020) 106623

[50] H. Ismail, J. Nizam, H.A. Khalil, The effect of a compatibilizer on the mechanical [53] A. Krainoi, Y. Nakaramontri, S. Wisunthorn, S. Pichaiyut, C. Nakason,
properties and mass swell of white rice husk ash filled natural rubber/linear low C. Kummerl€ owe, N. Vennemann, S. Kiatkamjornwong, Influence of carbon
density polyethylene blends, Polym. Test. 20 (2) (2001) 125–133. nanotube and ionic liquid on properties of natural rubber nanocomposites, Express
[51] S.A. Baeurle, A. Hotta, A.A. Gusev, A new semi-phenomenological approach to Polym. Lett. 13 (4) (2019).
predict the stress relaxation behavior of thermoplastic elastomers, Polymer 46 (12) [54] J. Ramier, C. Gauthier, L. Chazeau, L. Stelandre, L. Guy, Payne effect in silica-filled
(2005) 4344–4354. styrene–butadiene rubber: influence of surface treatment, J. Polym. Sci., Part B:
[52] A. Payne, R. Whittaker, Low strain dynamic properties of filled rubbers, Rubber Polym. Phys. 45 (3) (2007) 286–298.
Chem. Technol. 44 (2) (1971) 440–478.

12