Effect of non-rubber components on properties of sulphur crosslinked natural rubbers Rungsima Chollakup1,a, Rattana Tantatherdtam1,b, Wirasak Smitthipong1,c, Kanokwan Rungsanthien2,d, Potjanart Suwanruji2,e, Klanarong Sriroth3,f, Siriwat Radabutra4,g, Sombat Thanawan4,h, Marie-France Vallat5,i, Michel Nardin5,j and Karine Mougin5,k 1
Kasetsart Agricultural and Agro-Industrial Product Improvement Institute,
Kasetsart University (KU) 50 Ngam Wong Wan Rd., Bangkok 10900, Thailand. 2
Department of Chemistry Faculty of Science,
KU 50 Ngam Wong Wan Rd., Bangkok 10900, Thailand. 3
Department of Biotechnology, Faculty of Agro-Industry,
KU 50 Ngam Wong Wan Rd., Bangkok 10900, Thailand. 4
Rubber Technology Research Center, Faculty of Science, Mahidol University, 25/25 Phuttamonthon 4 Rd.,Salaya, Nakhon Pathom 73170, Thailand. 5
Institut de Science des Matériaux de Mulhouse, UMR 7361 CNRS UHA, 15 rue Jean Starcky, BP 2488 68057 Mulhouse, France.
a
[email protected],
[email protected],
[email protected],
[email protected], g
[email protected],
[email protected],
[email protected],
[email protected], i
[email protected],
[email protected], k
[email protected]
e
Keywords: natural rubber, non-rubber components, sulphur crosslinking, crosslinking density, adhesion property, AFM
Abstract. Non-rubber components (mainly proteins and lipids) in natural rubber (NR) play important roles for controlling the properties of NR. Crosslinking process creates intermolecular chemical bonds in order to obtain a three-dimensional network, resulting in more elastic rubber. Sulphur crosslinking is the most popular method and is applied in the present study. Two types of NR were prepared, namely, whole natural rubber (WNR) and purified natural rubber (PNR). PNR was deproteinized by centrifugation method and then acetone extraction. These rubbers were crosslinked by an efficient vulcanization (EV) system. They were cured for three curing times (1xt90, 2xt90, 3xt90) at 150°C. WNR presents shorter curing time than PNR because there are some phospholipids and proteins which are natural accelerators for curing reaction. The presence of nonrubber components seems to play a major role on crosslinking density and adhesion phenomenon for rubber/glass system. AFM images of WNR show more heterogeneity and roughness compared to PNR. Introduction Natural rubber (NR) contains 4-6 wt% of proteins, fatty acids, lipids and some inorganic substances in addition to the rubber chains. It has long been known that these non-rubber constituents are responsible for the variability in the properties of NR. Purified natural rubber (PNR) can be obtained by removing proteins, fatty acids and lipids which can be achieved by the deproteinization of the latex and acetone extraction of the resultant deproteinized rubber. PNR, being largely uncontaminated by non-rubber substances, would be suitable for medical applications as a starting material compared with NR which contains non-rubber components (WNR) [1]. Adhesion is one of the important properties for most practical uses of rubber, particularly when NR is used as matrix in composite materials, such as tires, belts and coated fabrics or applications of
NR which are in contact with other solids [2]. The aim of this study is therefore to better understand the effect of non-rubber substances, mainly proteins and lipids, on the properties of sulfur crosslinked natural rubber on the crosslinking density, adhesion behavior and topology characteristic which reflect the influence of the presence and nature of non-rubber constituents in NR vulcanizates. Experimental Two types of NR samples, WNR and PNR, were prepared from Hevea latex with low ammonium content according to a well-known method [3]. Briefly, WNR was prepared by drying of fresh natural rubber latex. PNR was prepared from centrifugation of natural rubber latex and then by acetone extraction. Chemical compositions of these rubber samples were analyzed. Protein contents estimated in terms of the nitrogen contents were determined by the micro-Kjeldahl method. Lipid contents were calculated by lipid extraction yield in chloroform/methanol solutions (2:1 by vol) [4]. Gel content was determined by immersing the rubber samples in toluene and molecular weight was determined by gel permeation chromatography (GPC) [2]. WNR and PNR (200 g each) were separately masticated using a two-roll mill with front rotor speed of 18 rpm and back rotor speed of 20 rpm at 70 ºC for 4 min. Then, the mixtures of ZnO (6 g) and stearic acid (2 g) were added and continuously masticated for 2 min. After TBBS (4 g) was added, the rubber samples were masticated for 1 min. Finally, 3 g of sulfur were added and the rubber samples were masticated for 12 more min. Optimum curing time measurement was taken using an oscillating disk rheometer (ODR) (Monsanto 100S) at 150°C as the time at which the rheometer torque increased to 90% of the total torque change following curing of rubber. Crosslinking density was measured by equilibrium solvent (toluene) swelling method. The crosslink density (ν) was determined from modified Flory-Rehner equation [5]:
1 ln(1−Vr )+Vr + χVr 2 ν =− 1/ 3 V Vs Vr − r 2
(1)
where Vs is the molar volume of solvent (106.9 cm3/mole for toluene), Vr is the volume fraction of rubber in the swollen network, χ is the solubility parameter between polymer and solvent (0.43 + [0.05xVr]). An adhesion probe test [2] was used to measure the level of adhesion between rubber and glass using a tensile testing machine (Lhomargy DY-34). With this device, it was possible to record simultaneously and precisely the force, the displacement and the contact area as a function of time during the experiment. All the parameters were carefully controlled, such as rates of contact (10 mm/min) and separation (10 mm/min), contact time (10 s), contact force (0.2 N) as well as compressive displacement (0.1 mm), and the thickness of the rubber sample (2 mm). The contact area was determined with a high speed camera by means of a mirror and side-views of the interfacial phenomena could be obtained with a second camera. The total energy required for the separation of the two surfaces was obtained by integration of positive values of the force as a function of displacement. However, to compare the different samples, adhesion energy has to be expressed per unit area, according to Eq. 2: G=
1 F ( x) dx A∫
(2)
where G is the adhesion energy (J/m2), F is the measured force, x is the displacement and A is the contact area before separation.
Morphology of vulcanized rubber samples was analyzed by Atomic Force Microscopy (AFM), nanoscope 4 (Santababara, USA) using tapping mode at scan rate of 0.3 Hz and scan size of 10 µm. Results and Discussion We found that protein contents are reduced more than 90% from WNR to PNR. Moreover, lipid contents are also decreased around 50% from WNR to PNR. These results confirm that these NR preparation methods allow us to reduce the non-rubber components in NR sample [3]. The macrostructure of WNR presents more branched chains than that of PNR, certainly due to the higher gel contents and molecular weights of WNR compared to PNR. For the sulphur crosslinked rubbers, WNR presents shorter curing time (1xt90) around 4 than PNR 16 min because there are some phospholipids and proteins which are natural accelerator agents in WNR to accelerate the curing reaction contrary to PNR, in good agreement with the previous work [1]. Concerning to the crosslinking density (ν), we cannot determine the crosslinking density of PNR at 1xt90 because the rubber sample was almost dissolved in toluene. That’s why we have used two additional curing times: 1xt90, 2xt90, 3xt90 at 150°C (Figure 1). The main difference is that WNR presents a higher crosslinking density than PNR. Values of the adhesion energy G of the glass probe onto both uncrosslinked and crosslinked rubbers are presented in Figure 2. Each experimental value represents the average of at least three measurements. For the same contact time at 10 s, a dependence of adhesion energies on the nature of the rubber is observed. The presence of non-rubber components seems to play a major role on adhesion phenomenon, since WNR presents weaker adhesion energy than PNR. Such a result has already been observed for uncrosslinked elastomers [6]. To verify the existence of non-rubber components on the rubber surface, surface analysis by AFM was performed. We can get together two image types: height image and phase contrast image. It should be noted that changes in chemical nature and/or mechanical surface, maybe invisible on the "height image", appear on a "phase contrast image." 10 9
Uncompounded NR Compounded NR
2xt90 1xt90
200
3xt90
WNR
150
3xt90
2xt90
PNR
100
Adhesion energy (J/m2)
Crossinking density (mol/m3)
250
8
Cured NR at 1xt90 Cured NR at 2xt90
7
Cured NR at 3xt90
6 5 4 3
50
2
0 0
5
10
15 20 25 30 35 Curing time (min)
40
45
50
1 0 WNR
Figure 1. Crosslinking density of crosslinked WNR and PNR.
PNR
Figure 2. Adhesion energy G between rubber and glass of WNR and PNR.
Figure 3 and 4 show the AFM images of WNR and PNR, respectively. Irregular: cavities are visible on the height image of WNR (dark areas on the surface). The height image of WNR also shows more roughness compared to that of PNR. The WNR domains are absent in the case of PNR. AFM phase contrast images for both rubbers confirm the differences in topography. The presence at the surface of the WNR areas consists of material with different sizes of domain. These
areas are almost disappeared on the surfaces of PNR. The heterogeneous quantities in the case of WNR surface help to explain the greater decrease in the adhesion energy. The absence or minimal presence of natural different components on the surface (in the case of PNR) leads to stronger adhesion.
Height image
Phase image
Figure 3. AFM images of crosslinked WNR at 1xt90, the images at 2xt90 and 3xt90 are the same as those of 1xt90 (data not show here).
Height
Phase
Figure 4. AFM images of crosslinked PNR at 1xt90, the images at 2xt90 and 3xt90 are the same as those of 1xt90 (data not show here).
Conclusion Uncrosslinked rubbers and crosslinked rubbers by sulfur were characterized by several techniques. These sample preparation methods (both WNR and PNR) allow us to reduce the nonrubber components (essentially proteins and lipids) in NR samples. The curing time of WNR is shorter than that of PNR, certainly due to the effect of non-rubber components. The crosslinking density of WNR is higher than that of PNR at different curing times. A probe test was specially used to determine the level of adhesion between natural rubbers and glass, the non-rubber components decreased the adhesion of rubber on glass. AFM images of WNR showed more heterogeneity and roughness compared to PNR. References [1] K. Suchiva, T. Kowitteerawut, L. Srichantamit. Structure properties of purified natural rubber, J. Appl. Polym. Sci. 78 (2000) 1495-1504. [2] W. Smitthipong, M. Nardin, J. Schultz, T Nipithakul, K. Suchiva. Study of tack properties of uncrosslinked natural rubber, J. Adhes. Sci. Technol. 18 (2004) 1449-1463. [3] K. Rungsanthie, P. Suwanruji, R. Tantatherdtam, R. Chollakup. Effect of non-rubber components on viscosity stabilization of natural rubber, in: The 28th International Conference of Polymer Processing Society, 11-15 December 2012, Pattaya, Thailand. [4] S. Liengprayoon, F. Bonfils, J. Sainte-Beuve, K. Sriroth, E. Dubreucq, L. Vaysse. Development of a new procedure for lipid extraction from Hevea brasiliensis natural rubber, Eur. J. Lipid Sci. Technol. 110 (2008) 563–569. [5] P.J. Flory. Statistical mechanics of swelling of network structures, J. Chem. Phys. 18 (1950) 108 -111. [6] W. Smitthipong, M. Nardin, J. Schultz. Effect of bulk and surface properties on adhesion of rubbers, Kasetsart J. (Nat. Sci.) 42 (2008) 318 - 324.
