- Email: [email protected]
Investigation of aging behavior and mechanism of nitrile-butadiene rubber (NBR) in the accelerated thermal aging environment
Polymer Testing 54 (2016) 59e66
Contents lists available at ScienceDirect
Polymer Testing journal homepage: www.elsevier.com/locate/polytest
Material properties
Investigation of aging behavior and mechanism of nitrile-butadiene rubber (NBR) in the accelerated thermal aging environment Jie Liu a, b, *, Xiangbo Li b, Likun Xu b, Peiqing Zhang a a b
College of Chemistry and Chemical Engineering, Yantai University, Yantai 264005, PR China State Key Laboratory for Marine Corrosion and Protection, Luoyang Ship Material Research Institute (LSMRI), Qingdao 266101, PR China
a r t i c l e i n f o
a b s t r a c t
Article history: Received 2 April 2016 Received in revised form 15 June 2016 Accepted 15 June 2016 Available online 29 June 2016
Nitrile-butadiene rubber (NBR) was exposed to an accelerated thermal aging environment produced by an air-circulating oven for different time periods. NBR aging was evaluated by morphology, crosslink density, mechanical properties, chemical changes and thermal stability. The results showed that the surface damage of NBR turned severe and inhomogeneous, and the aging degree was most serious on the edge region of voids. Crosslinking reactions mainly occurred in the aging process. The tensile strength increased with increase in crosslink density up to a maximum value and thereafter decreased with further increase in crosslink density. X-ray Photoelectron Spectroscopy (XPS) and Pyrolysis Gas Chromatography-Mass Spectrometry (Py-GC/MS) analysis demonstrated that hydroxyl groups were formed and the additives migrated from inner to surface of NBR samples. In addition, the thermogravimetric analysis (TGA) indicated that the thermal stability of NBR did not significantly change in the accelerated thermal aging environment. © 2016 Elsevier Ltd. All rights reserved.
Keywords: Accelerated thermal aging Nitrile-butadiene rubber (NBR) Crosslink density Mechanical properties Chemical changes Thermogravimetric analysis
1. Introduction Nitrile-butadiene rubber (NBR) has been widely used in industry over the years [1e5]. NBR has great potential in seals industry because of its low cost, excellent resistance to oil, fuels and greases, low abrasion rate, and good processability [1,2]. However, the aging resistance of NBR is sensitive to environmental factors because of the presence of unsaturated back-bone of the butadiene part. A degrading environment, such as one characterized by high temperature, high humidity, strong light, low temperature, or a mechanical load, can have a profound effect on durability of NBR [3e9]. For instance, in oxygen containing environments, the hardness, morphology and mechanical properties of NBR rubber can be greatly affected by oxidation, especially at relatively high temperatures [10e14]. In fact, the process of rubber degradation is very slow in an actual working environment, so it takes a long time to obtain rubber with degradation behaviors. Thus, accelerated aging tests are typically used to study the degradation of rubbers in a comparatively short time [15]. Thermal aging of the NBR rubber in air is an extremely complex
* Corresponding author. College of Chemistry and Chemical Engineering, Yantai University, Yantai 264005, PR China. E-mail address: [email protected] (J. Liu). http://dx.doi.org/10.1016/j.polymertesting.2016.06.010 0142-9418/© 2016 Elsevier Ltd. All rights reserved.
process that involving several reactions that proceed in oxygen and without it. Thermal oxidation research of NBR has been carried out for decades. In the previous works [16e20], the effects of thermooxidation on the physical properties of NBR were investigated, and the volatile components and oxidative degradation products changes during thermal aging of NBR were analyzed. During the thermal aging, mainly three changes occurred successively: (1) migration and loss of additives such as plasticizers, stabilizers; (2) extra crosslinking of the rubber network; and (3) oxidation of macromolecular chains [21,22]. Although there are many studies [16e20,23e24] on the accelerated thermal aging behavior of NBR rubber up to now, the thermal aging mechanism of nitrilebutadiene rubber (NBR) still need to be studied further. The objective of the present work was to investigate the aging behavior and mechanism of nitrile-butadiene rubber (NBR) in the accelerated thermal aging environment. The change of morphology was monitored by Scanning Electron Microscope (SEM). Crosslink density of NBR specimens was measured by the solvent swell method. Mechanical properties were monitored by computercontrolled universal tensile testing. The surface chemical changes were detected by Attenuated Total Reflection-Fourier Transform Infrared Spectroscopy (ATR-FTIR), X-ray Photoelectron Spectroscopy (XPS) and Pyrolysis Gas Chromatography-Mass Spectrometry (Py-GC/MS). The thermal stability was evaluated by
60
J. Liu et al. / Polymer Testing 54 (2016) 59e66
700e4000 cm1 with 32 scans at a resolution of 4 cm1. The infrared radiation (IR) penetrates the surfaces of samples to approximately 1 mm. The average of three scans for each sample was taken for peak identification, with diamond as the ATR crystal.
Thermogravimetric Analysis (TGA). 2. Experimental 2.1. Materials
2.7. X-ray photoelectron spectroscopy (XPS) The sulfur vulcanized NBR sheets with thickness of 2 mm were supplied by Changjiang Rubber Manufacturing Limited Company of China. The nitrile content is around 32%. In addition to NBR, carbon black, ZnO, Dibutyl Sebacate plasticizer, DCP curing package and antioxidant D were included in the composition of the sheets. To improve the aging resistance of NBR, a certain percentage of Neoprene was also added into NBR. The specific compounds formulation and technological process were not disclosed because of commercial reasons. The samples were cut into dumbbells and strips according to the standard ASTM D412-2006 for mechanical properties measurements.
XPS analysis was performed using a ESCALAB 250 analyzer (Thermo Inc, USA) equipped with an achromatic Al Ka X-ray source. The pressure during the analysis was 6.7 108 Pa and the take off angle was 45 . The survey spectra in the range of 0e1200 eV were recorded in 1 eV step for each sample. Atomic concentrations of each element were carried out by determining the relevant integral peak intensities. High-resolution scans with step size 0.1 eV for C 1s and O 1s were recorded. The binding energies were corrected by referencing to the hydrocarbon component at 284.8 eV. 2.8. Pyrolysis gas chromatography-mass spectrometry (Py-GC/MS)
2.2. Thermal aging Accelerated thermal aging studies were carried out in an aircirculating oven (Electro-mag with sensitivity of ±1 C) equipped with thermocouples connected to continuous recorders. The test temperature was selected at 65 C for up to 90 days in this work. 2.3. Surface morphology and component analysis The surface morphology of the samples was studied by using Quanta 200 Scanning Electron Microscope (SEM) (Quanta 200, FEI Inc, Netherlands) operating at an accelerating voltage of 20 kV. All the samples for SEM tests were not stained. The surface component analysis was carried out by an INCA Energy Dispersion Spectroscopy (EDS) (Oxford Instrument, UK) accompanied with SEM. 2.4. Crosslink density The volume fraction of rubber in swollen network of the vulcanizates Vr, was determined by means of equilibrium swelling in acetone laboratory grade at 23 C. The equilibrium swelling was used to calculate the crosslink density, which was the number of network chain density by applying the Flory-Rehner equation [25] as follow:
"
1 lnð1 Vr Þ þ Vr þ mVr2 Ve ¼ 1=3 V Vr Vr =2
# (1)
where Ve was crosslink density; m was polymer-solvent interaction parameter; V was molar volume of solvent; Vr was volume fraction of rubber in the swollen gel. 2.5. Evaluation of mechanical properties The mechanical properties were evaluated by tensile tests. The tensile specimens were tested according to ISO 37: 2011 at room temperature (23 C) with a crosshead speed of 500 mm min1 using a computer-controlled universal testing machine (INSTRON 5565, USA) equipped with pneumatic grips. At least five samples were tested in order to get a reliable result. 2.6. Attenuated total reflection-fourier transform infrared spectroscopy (ATR-FTIR) ATR-FTIR spectra of the samples were taken at room temperature using a Nicolet 470 Instrument (Nicolet Inc, USA) in the range
Py-GC/MS analysis were carried out using an EGA/PY-3030D multi-shot pyrolyzer (Frontier Laboratories Ltd., JPN) coupled to a Trace GC Ultra gas chromatograph equipped with a TSQ Quantum XLS mass spectrometer (AGILENT Inc, USA). The sample (about 1.0 mg) was preheated at 300 C for 3 min and pyrolyzed at 450 C for 1 min under a helium (He) atmosphere. The volatiles were carried by carrier gas He through a HP-5MS capillary column (30.0 m, i.d. 0.32 mm 1.0 mm stationary phase thickness) and the mixture was separated. The GC oven temperature program was as follows: 70 C (held for 2 min) to 230 C at 20 C/min (held for 16 min). The interface temperature of GC to MS was 250 C. Electron ionization (70 eV) was used to ionize the pyrolysis products. The MS source temperature was 230 C. 2.9. Thermogravimetric analysis (TGA) The thermogravimetric analysis of NBR was carried out using a TG/DTA Q600 system (TA Inc, USA). Specimens weighing 10e15 mg were heated to 550 C in nitrogen at a feed speed of 60 ml/min, and the heating rates was 10 C/min. Nitrogen flow (80 ml/min) was employed in order to remove all corrosive gases involved in the degraded materials and to avoid thermo-oxidative degradation. 3. Results and discussion 3.1. Surface morphology and component analysis The morphology changes on the surface of samples before and after thermal aging for different times were obtained by Scanning Electron Microscopy (SEM) as shown in Fig. 1. Apparently, the surface of the unaged NBR samples was relatively homogeneous and smooth and with no obvious defects appeared. With increasing aging time, small voids appeared. When the aging time reached 90 days, the surface turned rougher, the size of the voids became greater and more white spots can be observed. The appearance of white spots on the NBR samples was related to precipitation of some additives [26e29], and this phenomenon would be demonstrated below. In order to analyze the degree of thermal aging of NBR samples after 90 days conveniently, the components of oxygen and carbon on the NBR surface were analyzed by Energy Dispersion Spectroscopy (EDS) microanalysis system. Three different regions were selected in Fig. 2, which respectively corresponded to the bottom of one void (Spectrum 1), the edge of voids (Spectrum 2) and the relatively smooth region (Spectrum 3). The atomic weight proportion of oxygen and carbon (O/C) was adopted to determine the
J. Liu et al. / Polymer Testing 54 (2016) 59e66
61
Fig. 1. Micrographs of NBR samples before and after thermal aging: (a) 0 d; (b) 30 d; (c) 90 d.
environment became severe and inhomogeneous with aging time, and the oxidation reactions initiated from the NBR surface, afterwards extended to the inner structures. 3.2. Crosslink density analysis Oxygenated species produced in the accelerated thermal aging environment combined with each other and led to new crosslinking reactions occurrence, which indicated that the crosslink density of NBR samples increased with the thermal aging time [30]. The correlation between crosslink density of NBR samples and the thermal aging time was shown in Fig. 4. The linear relationship was expressed by the equation as follow:
v ¼ 6:84 104 þ 3:87 106 t
Fig. 2. Locations of NBR samples after thermal aging for 90 days for EDS analysis.
order of the degradation reaction. As shown in Fig. 3 and Table 1, there were clear distinctions in the three regions, the proportion of Spectrum 2 was highest, which meant that the degree of thermal aging was most serious for the edge region of voids of NBR samples. And the proportion of Spectrum 1 was lowest, which meant that oxidation reactions were initiated from the surface of NBR samples in the accelerated thermal aging environment. From the SEM and EDS results, it can be demonstrated that the surface damage of NBR in the accelerated thermal aging
(2)
where v was crosslink density; t was the thermal aging time. The analysis results demonstrated that crosslinking reactions mainly occurred in the process of thermal aging. 3.3. Evaluation of mechanical properties The changes of elongation at break and tensile strength with thermal aging time were shown in Fig. 5 and Fig. 6, respectively. With the aging time increase, the elongation at break decreased continuously, and the value decreased 300% in the whole aging process. In contrast, the tensile strength in Fig. 6 had an increasing trend in the first phase of thermal aging and then decreased after 70 days of thermal aging, which was related to the continuous
62
J. Liu et al. / Polymer Testing 54 (2016) 59e66
-3
Crosslink density (mol ·cm )
0.0011
-4
-6
v=6.84*10 +3.87*10 t, R=0.996
0.0010
0.0009
0.0008
0.0007
0.0006
0
20
40
60
80
100
Aging time (d) Fig. 4. Variation of the crosslink density change of the NBR samples with aging time.
500 450
Elongation at break (%)
400 350 300 250 200 150 100
0
20
40
60
80
100
Aging time (d) Fig. 5. Effect of thermal aging time on elongation at break.
10.5
Fig. 3. EDS spectra of NBR samples after thermal aging for 90 days: (a) the bottom of one void; (b) the edge of voids and (c) the relatively smooth region.
Table 1 EDS results of NBR samples after thermal aging for 90 days. Position
11.28 30.25 21.06
1.03 1.14 1.82
0.97 1.10 e
3.48 4.09 3.72
9.5
9.0
8.5
8.0 0
C (wt%) O (wt%) Mg (wt%) S (wt%) Cl (wt%) Zn (wt%) O/C
Spectrum 1 81.70 Spectrum 2 63.42 Spectrum 3 73.40
Tensile strength (MPa)
10.0
1.54 e e
0.14 0.48 0.29
20
40
60
80
100
Aging time (d) Fig. 6. Effect of thermal aging time on tensile strength.
120
J. Liu et al. / Polymer Testing 54 (2016) 59e66
3.4. ATR-FTIR analysis The FTIR spectra of NBR samples before and after thermal aging for 90 days were displayed in Fig. 8. It can be seen that there were nine characteristic peaks in the spectrum of the unaged NBR. The broad peak at 3350 cm1 was due to NeH stretching vibration and OeH symmetric stretching vibration. The peaks at 2918 cm1 and 2848 cm1 corresponded to asymmetric stretching vibration of methylene in the saturated hydrocarbon backbone and symmetric stretching vibration of methylene, respectively. The peak at 2233 cm1 was attributed to eC^N stretching vibration. The peaks at 1730 cm1 and 1660 cm1 were the very typical characteristic of C]O stretching vibration and C]C stretching vibration, separately. The peak at 1580 cm1 should be attributed to additives such as curing package [17]. The peak at 1433 cm1 was assigned to eCH2e deformation vibration, while the peak at 962 cm1 was attributed to CeH deformation vibration, which was characteristic of the trans-1, 4 structure of the butadiene part. It also can be seen from Fig. 8, when the NBR samples were aged for 90 days, the peak at 1730 cm1 decreased obviously in intensity, and the peaks at 3350 cm1, 2918 cm1 and 2848 cm1 increased remarkably in intensity. The peak at 3350 cm1 was mostly due to NeH stretching vibration and OeH symmetric stretching vibration
Fig. 7. Relationship between tensile strength and crosslink density.
before 90 d
Reflectances (%)
increase of crosslink density. The correlation between tensile strength and crosslink density of NBR was shown in Fig. 7. It can be seen that the whole change process was divided into two phases. In phase I, the tensile strength increased continuously and reached a maximum value with the increase of crosslink density, the reason was as followed: when the crosslink density increased during the early aging time, the crosslink points would increase correspondingly, which can make the positions of chain segments relatively stable and decrease the relative movement of network chains. It was beneficial to disperse and transfer the external stress and then led to the increase of tensile strength. In phase II, the tensile strength decreased obviously with the further increase of crosslink density, and the reason was as followed: when the crosslink density reached or exceeded a certain critical value, the average molar mass of the NBR chain between two adjacent crosslink points decreased and the effective mobility of the chain segment was limited, which disturbed the normal orientation of the network chains, thus the stress concentration was caused. This reduced the number of effective network chains and then led to the decrease of tensile strength [24,30,31].
63
3350
2233
1730
2848 1660
2918
1580 1433
4000
3500
3000
2500
2000
1500
964
1000
-1
Wavenumbers (cm ) Fig. 8. ATR-FTIR spectra of NBR samples before and after thermal aging for 90 days.
for unaged NBR samples, however, the NBR samples were badly aged after thermal aging for 90 days, most of antioxidants in NBR samples were consumed, so in the case the peak at 3350 cm1 was mostly due to OeH symmetric stretching vibration. It can be estimated that oxygenated species such as hydroxyl or carboxyl groups were formed for NBR in the process of accelerated thermal aging. Moreover, the decrease of the band at 1730 cm1 in intensity is possibly attributed to the consumption of carbonyl groups, or attributed to the loss of plasticizer during the thermal aging process. From the above analysis results, it indicated that hydroxyl groups were produced when NBR was exposed to the accelerated thermal aging environment.
3.5. XPS analysis Fig. 9 showed the C 1s spectra of NBR surface before and after thermal aging for 90 days. The C 1s spectrum of the unaged NBR can be fitted to two components corresponding to the CeC/CeH (284.8 eV) and CeOeC/CeOH (286.3 eV), respectively. The CeC/ CeH component was quite significant and most likely came from the hydrocarbon backbone of NBR rubber. The presence of CeOeC/ CeOH in the unaged sample could be attributed to some unavoidable side reactions during crosslinking with sulfur and other agents. Since the characteristic peak assigned to the vibrations of CeOH can be observed and changed obviously before and after thermal aging in the FTIR spectra in Fig. 8, however, the characteristic peak assigned to the vibrations of CeOeC cannot be found, it could be deduced that the peak at 286.3 eV was assigned to CeOH rather than CeOeC. After exposure to the thermal aging environment for 90 days, there was a clear chemical change around the carbon atoms, and the area percentage of the two components changed obviously compared with the unaged NBR. It was notable that the area percentage of CeOeC/CeOH decreased 3.11% in Table 2, compared with the unaged NBR, and the possible reason was that Dibutyl Sebacate plasticizer, which was a higher carbon-oxygen ratio (C/O), and antioxidant D (N-phenyl-2-naphthylamine), which was a higher carbon-nitrogen ratio (C/N), migrated from inner to surface of NBR samples. To verify the correctness of the above reasoning, the Py-GC/MS analysis of the unaged and aged NBR samples were carried out.
64
J. Liu et al. / Polymer Testing 54 (2016) 59e66
Fig. 9. High-resolution C 1s spectra of NBR samples: (a) before aging and (b) after aging for 90 days.
Table 2 XPS results from the curve fitting of the high-resolution C1s spectra of NBR before and after thermal aging for 90 days. Aging time (d)
Peak A (eV)
Structure
Peak B (eV)
Structure
0 90
284.8(93.86%) 284.8(96.97%)
CeH/CeC CeH/CeC
286.3 (6.14%) 286.3 (3.03%)
CeOeC/CeOH CeOeC/CeOH
3.6. Py-GC/MS analysis Fig. 10 showed total ion chromatograms of NBR samples. The four characteristic peaks had been marked as shown in Fig. 10, and the structure of typical volatile components is shown in Table 3. Peak 1 was assigned to butadiene monomer and acrylonitrile monomer, which constituted the NBR copolymer, Peak 2 was attributed to the small molecular of Neoprene 1212 modifier, Peak 3 and Peak 4 corresponded to Dibutyl Sebacate plasticizer and antioxidant D, respectively. It can be seen that the abundance of Peak 2, Peak 3 and Peak 4 increased obviously after aging for 90 days. The increase of the abundance of Peak 2 was that cracking reactions occurred on some molecular chains of NBR, small molecular of Neoprene were produced and accumulated continuously on the surface of NBR samples. The increase of the abundance of Peak 3 and Peak 4 showed that Dibutyl Sebacate plasticizer and antioxidant D migrated
Fig. 10. Py-GC/MS chromatograms of NBR samples: (a) control sample before aging; (b) surface of sample after aging for 90 days and (c) inner of sample after aging for 90 days.
continuously from inner to surface of NBR samples in the process of thermal aging, which depended on their own physical properties of rubber additives, such as low molecular weight, low melting point, high volatile etc. Moreover, it was noteworthy that there were no obvious differences in abundance of total ion chromatograms between the sample before degradation and inner of sample after aging for 90 days. The reason was that the migration amount of additives was
J. Liu et al. / Polymer Testing 54 (2016) 59e66
loss, while the second onset temperature was 351.37 C with a 39.27% weight loss. The results demonstrated that the accelerated thermal aging environment did not significantly affect the thermal stability of NBR.
Table 3 The structure of typical volatile components in NBR.
Peak number
65
Structure
4. Conclusions 1
CH2
2
C H
C H
CH2
CH2
C H
CN
Cl Cl O
3
O O O
H N
4
comparatively little from inner to surface of NBR samples, which did not cause the remarkable changes of component content for inner of NBR samples. The surface and inner of NBR samples can be seen as a “micro thin layer” and an “infinite thick layer”, respectively, so migration of little materials from inner to surface of NBR may cause the phenomenon that additives accumulated on the surface appeared. The results showed that additives migrated to the surface of NBR samples in the process of thermal aging, which can powerfully support the conclusions of SEM, ATR-FTIR and XPS analysis.
3.7. Thermogravimetric analysis Fig. 11 showed TGA (differential thermogravimetric) plots of NBR samples before and after thermal aging for 90 days. For the unaged NBR samples, the first onset temperature was 143.37 C with a 8.86% weight loss and the second onset temperature was 355.60 C with a 39.69% weight loss. After thermal aging for 90 days, the first onset temperature was 155.28 C with a 6.46% weight
Fig. 11. TGA plots of NBR samples before and after thermal aging for 90 days.
From the systematic studies on the accelerated thermal aging of NBR, it can be concluded that, in the accelerated thermal aging environment, the surface damage of NBR samples became severe and inhomogeneous, and the oxidation reactions initiated from the NBR surface and extended to the inner structures. Crosslinking reactions mainly occurred in the aging process. With the crosslink density increase, the tensile strength firstly increased and then decreased. In the process of accelerated thermal aging, hydroxyl groups were formed and the additives migrated from inner to surface of NBR samples. The accelerated thermal aging environment did not significantly affect the thermal stability of NBR. Acknowledgements This work was financially supported by the Research Fund of State Key Laboratory for Marine Corrosion (KF160412) and Protection of Luoyang Ship Material Research Institute (LSMRI) and Chong Qing Postdoctoral Science Foundation Grant No. Rc 201301. References [1] T. Yasin, S. Ahmed, F. Yoshii, K. Makuuchi, Radiation vulcanization of acrylonitrile-butadiene rubber with polyfunctional monomers, J. React. Funct. Polym. 53 (2002) 173e181. [2] F.S. Ahmed, M. Shafy, A.A. Abd El-megeed, E.M. Hegazi, The effect of g-irradiation on acrylonitrile-butadiene rubber NBR seal materials with different antioxidants, Mater. Des. 36 (2012) 823e828. rtolas, On the [3] M.J. Martínez-Morlanes, F.J. Medel, M.D. Mariscal, J.A. Pue assessment of oxidative stability of post-irradiation stabilized highly crosslinked UHMWPEs by thermogravimetric, Polym. Test. 29 (2010) 425e432. [4] C. Marazzato, Y. Peneva, E. Lefterova, S. Filippi, L. Minkova, Kinetics of nonisothermal degradation of nanocomposites based on functionalized polyethylenes, Polym. Test. 26 (2007) 526e536. [5] L.H. Nguyen, M. Gu, Decomposition kinetics, life estimation and dielectric study of an acrylate based photopolymer for microfabrication and photonic applications, Macromol. Chem. Phys. 206 (2005) 1659e1684. [6] G.V. Salmoria, P. Klauss, A.T.N. Pires, J. Roeder, V. Soldi, Investigations on cure kinetics and thermal degradation of stereolithography Renshape™ 5260 photosensitive resin, Polym. Test. 27 (2008) 698e704. [7] M. Jaunich, W. Stark, D. Wolff, Comparison of low tem perature properties of different elastomer materials investigated by a new method for compression set measurement, Polym. Test. 31 (2012) 987e992. [8] M. Jaunich, D. Wolff, W. Stark, A new method to evaluate the low temperature function of rubber sealing materials, Polym. Test. 29 (2010) 815e823. [9] R. Fan, Y. Zhang, C. Huang, Effect of crosslink structures on dynamic mechanical properties of natural rubber vulcanizates under different aging conditions, J. Appl. Polym. Sci. 81 (2001) 710e718. [10] K.T. Gillen, R.L. Clough, J. Wise, Prediction of elastomer lifetimes from accelerated thermal-aging experiments, Adv. Chem. Ser. 249 (1996) 557e563. [11] A. Mostafa, A. Abouel-Kasem, M.R. Bayoumi, M.G. El-Sebaie, The influence of CB loading on thermal aging resistance of SBR and NBR rubber compounds under different aging temperature, Mater. Des. 30 (2009) 791e795. € mmling, M. Jaunich, D. Wolff, Effects of heterogeneous aging in com[12] A. Ko pressed HNBR and EPDM O-ring seals, Polym. Degrad. Stab. 126 (2016) 39e46. [13] A.S. Deuri, A.K. Bhowmick, Effect of ageing on critical cut length and morphology of fracture surface in tensile rupture of natural rubber, J. Mater. Sci. 22 (1987) 4299e4306. [14] P. Budrugeac, Accelerated thermal aging of NBR and other materials under air or oxygen pressures and degradation of polymers, Die. Angew. Makromol. Chem. 247 (1997) 19e30. [15] M.M. Abou-Zeid, Radiation effect on properties of carbon black filled NBR/ EPDM rubber blends, Eur. Polym. J. 43 (2007) 4415e4422. [16] J.H. Zhao, R. Yang, R. Iervolino, B. Vorst, S. Barbera, The effect of thermooxidation on the continuous stress relaxation behavior of nitrile rubber, Polym. Degrad. Stab. 115 (2015) 32e37. [17] J.H. Zhao, R. Yang, R. Iervolino, S. Barbera, Changes of chemical structure and mechanical property levels during thermo-oxidative aging of NBR, Rubber Chem. Technol. 86 (2013) 591e603.
66
J. Liu et al. / Polymer Testing 54 (2016) 59e66
[18] R. Yang, J.H. Zhao, Y. Liu, Oxidative degradation products analysis of polymer materials by pyrolysis gas chromatography-mass spectrometry, Polym. Degrad. Stab. 98 (2013) 2466e2472. [19] X. Liu, J.H. Zhao, Y. Liu, R. Yang, Volatile components changes during thermal aging of nitrile rubber by flash evaporation of Py-GC/MS, J. Anal. Appl. Pyrolysis 113 (2015) 93e201. [20] A. Li, The aging phenomena and aging mechanisms of rubbers, Spec. Purp. Rubber Prod. 30 (2009) 56e67. [21] A. Li, Aging and lifetime prediction for rubbers, Rubber Ref. 39 (2009) 2e77. [22] Y. Miyashita, H. Kato, Cross-linking reaction of LDPE and the behavior of decomposition products from cross-linking agents, in: Proceedings of the Twenty-first Symposium on Electrical Insulating Materials, 1988, pp. 259e262. [23] Y. Xiong, G.S. Chen, S.Y. Guo, G.X. Li, Lifetime prediction of NBR composite sheet in aviation kerosene by using nonlinear curve fitting of ATR-FTIR spectra, J. Ind. Eng. Chem. 19 (2013) 1611e1616. [24] F. Delor-Jestin, N. Barrois-Oudin, C. Cardinet, J. Lacoste, J. Lemaire, Thermal ageing of acrylonitrile-butadiene copolymer, Polym. Degrad. Stab. 70 (2000) 1e4. [25] Elena V. Bystritskaya, Tatyana V. Monakhova, Victor B. Ivanov, TGA application for optimising the accelerated aging conditions and predictions of
thermal aging of rubber, Polym. Test. 32 (2013) 197e201. [26] G. Li, J.Z. Tan, J.M. Gong, Chemical aging of the silicone rubber in a simulated and three accelerated proton exchange membrane fuel cell environments, J. Power Sources 217 (2012) 175e183. [27] Alexandre de Souza Rios, Wanderley Ferreira de Amorim Jr., Elineudo Pinho de Moura, Enio Pontes de Deus, Judith Pessoa de Andrade Feitosa, Effects of accelerated aging on mechanic al, thermal and morpholog ical behavior of polyurethane/ep oxy/fi berglass compo sites, Polym. Test. 50 (2016) 152e163. [28] H. Nabil, H. Ismail, A.R. Azura, Comparison of thermo-oxidative ageing and thermal analysis of carbon black-filled NR/Virgin EPDM and NR/Recycled EPDM blends, Polym. Test. 32 (2013) 631e639. [29] A. Choudhury, A.K. Bhowmick, M. Soddemann, Effect of organo-modified clay on accelerated aging resistance of hydrogenated nitrile rubber nanocomposites and their life time prediction, Polym. Degrad. Stab. 95 (2010) 2555e2562. [30] Q.L. Zhao, X.G. Li, J. Gao, Aging of ethylene-propylene-diene monomer (EPDM) in artificial weathering environment, Polym. Degrad. Stab. 92 (2007) 1841e1846. [31] H.M. Dong, M.T. Li, X.D. Liu, C.L. Zhang, Z.F. Li, Effect of crosslink density on thermal property of HNBR, Spec. Purp. Rubber Prod. 31 (2010) 19e21.
Contents lists available at ScienceDirect
Polymer Testing journal homepage: www.elsevier.com/locate/polytest
Material properties
Investigation of aging behavior and mechanism of nitrile-butadiene rubber (NBR) in the accelerated thermal aging environment Jie Liu a, b, *, Xiangbo Li b, Likun Xu b, Peiqing Zhang a a b
College of Chemistry and Chemical Engineering, Yantai University, Yantai 264005, PR China State Key Laboratory for Marine Corrosion and Protection, Luoyang Ship Material Research Institute (LSMRI), Qingdao 266101, PR China
a r t i c l e i n f o
a b s t r a c t
Article history: Received 2 April 2016 Received in revised form 15 June 2016 Accepted 15 June 2016 Available online 29 June 2016
Nitrile-butadiene rubber (NBR) was exposed to an accelerated thermal aging environment produced by an air-circulating oven for different time periods. NBR aging was evaluated by morphology, crosslink density, mechanical properties, chemical changes and thermal stability. The results showed that the surface damage of NBR turned severe and inhomogeneous, and the aging degree was most serious on the edge region of voids. Crosslinking reactions mainly occurred in the aging process. The tensile strength increased with increase in crosslink density up to a maximum value and thereafter decreased with further increase in crosslink density. X-ray Photoelectron Spectroscopy (XPS) and Pyrolysis Gas Chromatography-Mass Spectrometry (Py-GC/MS) analysis demonstrated that hydroxyl groups were formed and the additives migrated from inner to surface of NBR samples. In addition, the thermogravimetric analysis (TGA) indicated that the thermal stability of NBR did not significantly change in the accelerated thermal aging environment. © 2016 Elsevier Ltd. All rights reserved.
Keywords: Accelerated thermal aging Nitrile-butadiene rubber (NBR) Crosslink density Mechanical properties Chemical changes Thermogravimetric analysis
1. Introduction Nitrile-butadiene rubber (NBR) has been widely used in industry over the years [1e5]. NBR has great potential in seals industry because of its low cost, excellent resistance to oil, fuels and greases, low abrasion rate, and good processability [1,2]. However, the aging resistance of NBR is sensitive to environmental factors because of the presence of unsaturated back-bone of the butadiene part. A degrading environment, such as one characterized by high temperature, high humidity, strong light, low temperature, or a mechanical load, can have a profound effect on durability of NBR [3e9]. For instance, in oxygen containing environments, the hardness, morphology and mechanical properties of NBR rubber can be greatly affected by oxidation, especially at relatively high temperatures [10e14]. In fact, the process of rubber degradation is very slow in an actual working environment, so it takes a long time to obtain rubber with degradation behaviors. Thus, accelerated aging tests are typically used to study the degradation of rubbers in a comparatively short time [15]. Thermal aging of the NBR rubber in air is an extremely complex
* Corresponding author. College of Chemistry and Chemical Engineering, Yantai University, Yantai 264005, PR China. E-mail address: [email protected] (J. Liu). http://dx.doi.org/10.1016/j.polymertesting.2016.06.010 0142-9418/© 2016 Elsevier Ltd. All rights reserved.
process that involving several reactions that proceed in oxygen and without it. Thermal oxidation research of NBR has been carried out for decades. In the previous works [16e20], the effects of thermooxidation on the physical properties of NBR were investigated, and the volatile components and oxidative degradation products changes during thermal aging of NBR were analyzed. During the thermal aging, mainly three changes occurred successively: (1) migration and loss of additives such as plasticizers, stabilizers; (2) extra crosslinking of the rubber network; and (3) oxidation of macromolecular chains [21,22]. Although there are many studies [16e20,23e24] on the accelerated thermal aging behavior of NBR rubber up to now, the thermal aging mechanism of nitrilebutadiene rubber (NBR) still need to be studied further. The objective of the present work was to investigate the aging behavior and mechanism of nitrile-butadiene rubber (NBR) in the accelerated thermal aging environment. The change of morphology was monitored by Scanning Electron Microscope (SEM). Crosslink density of NBR specimens was measured by the solvent swell method. Mechanical properties were monitored by computercontrolled universal tensile testing. The surface chemical changes were detected by Attenuated Total Reflection-Fourier Transform Infrared Spectroscopy (ATR-FTIR), X-ray Photoelectron Spectroscopy (XPS) and Pyrolysis Gas Chromatography-Mass Spectrometry (Py-GC/MS). The thermal stability was evaluated by
60
J. Liu et al. / Polymer Testing 54 (2016) 59e66
700e4000 cm1 with 32 scans at a resolution of 4 cm1. The infrared radiation (IR) penetrates the surfaces of samples to approximately 1 mm. The average of three scans for each sample was taken for peak identification, with diamond as the ATR crystal.
Thermogravimetric Analysis (TGA). 2. Experimental 2.1. Materials
2.7. X-ray photoelectron spectroscopy (XPS) The sulfur vulcanized NBR sheets with thickness of 2 mm were supplied by Changjiang Rubber Manufacturing Limited Company of China. The nitrile content is around 32%. In addition to NBR, carbon black, ZnO, Dibutyl Sebacate plasticizer, DCP curing package and antioxidant D were included in the composition of the sheets. To improve the aging resistance of NBR, a certain percentage of Neoprene was also added into NBR. The specific compounds formulation and technological process were not disclosed because of commercial reasons. The samples were cut into dumbbells and strips according to the standard ASTM D412-2006 for mechanical properties measurements.
XPS analysis was performed using a ESCALAB 250 analyzer (Thermo Inc, USA) equipped with an achromatic Al Ka X-ray source. The pressure during the analysis was 6.7 108 Pa and the take off angle was 45 . The survey spectra in the range of 0e1200 eV were recorded in 1 eV step for each sample. Atomic concentrations of each element were carried out by determining the relevant integral peak intensities. High-resolution scans with step size 0.1 eV for C 1s and O 1s were recorded. The binding energies were corrected by referencing to the hydrocarbon component at 284.8 eV. 2.8. Pyrolysis gas chromatography-mass spectrometry (Py-GC/MS)
2.2. Thermal aging Accelerated thermal aging studies were carried out in an aircirculating oven (Electro-mag with sensitivity of ±1 C) equipped with thermocouples connected to continuous recorders. The test temperature was selected at 65 C for up to 90 days in this work. 2.3. Surface morphology and component analysis The surface morphology of the samples was studied by using Quanta 200 Scanning Electron Microscope (SEM) (Quanta 200, FEI Inc, Netherlands) operating at an accelerating voltage of 20 kV. All the samples for SEM tests were not stained. The surface component analysis was carried out by an INCA Energy Dispersion Spectroscopy (EDS) (Oxford Instrument, UK) accompanied with SEM. 2.4. Crosslink density The volume fraction of rubber in swollen network of the vulcanizates Vr, was determined by means of equilibrium swelling in acetone laboratory grade at 23 C. The equilibrium swelling was used to calculate the crosslink density, which was the number of network chain density by applying the Flory-Rehner equation [25] as follow:
"
1 lnð1 Vr Þ þ Vr þ mVr2 Ve ¼ 1=3 V Vr Vr =2
# (1)
where Ve was crosslink density; m was polymer-solvent interaction parameter; V was molar volume of solvent; Vr was volume fraction of rubber in the swollen gel. 2.5. Evaluation of mechanical properties The mechanical properties were evaluated by tensile tests. The tensile specimens were tested according to ISO 37: 2011 at room temperature (23 C) with a crosshead speed of 500 mm min1 using a computer-controlled universal testing machine (INSTRON 5565, USA) equipped with pneumatic grips. At least five samples were tested in order to get a reliable result. 2.6. Attenuated total reflection-fourier transform infrared spectroscopy (ATR-FTIR) ATR-FTIR spectra of the samples were taken at room temperature using a Nicolet 470 Instrument (Nicolet Inc, USA) in the range
Py-GC/MS analysis were carried out using an EGA/PY-3030D multi-shot pyrolyzer (Frontier Laboratories Ltd., JPN) coupled to a Trace GC Ultra gas chromatograph equipped with a TSQ Quantum XLS mass spectrometer (AGILENT Inc, USA). The sample (about 1.0 mg) was preheated at 300 C for 3 min and pyrolyzed at 450 C for 1 min under a helium (He) atmosphere. The volatiles were carried by carrier gas He through a HP-5MS capillary column (30.0 m, i.d. 0.32 mm 1.0 mm stationary phase thickness) and the mixture was separated. The GC oven temperature program was as follows: 70 C (held for 2 min) to 230 C at 20 C/min (held for 16 min). The interface temperature of GC to MS was 250 C. Electron ionization (70 eV) was used to ionize the pyrolysis products. The MS source temperature was 230 C. 2.9. Thermogravimetric analysis (TGA) The thermogravimetric analysis of NBR was carried out using a TG/DTA Q600 system (TA Inc, USA). Specimens weighing 10e15 mg were heated to 550 C in nitrogen at a feed speed of 60 ml/min, and the heating rates was 10 C/min. Nitrogen flow (80 ml/min) was employed in order to remove all corrosive gases involved in the degraded materials and to avoid thermo-oxidative degradation. 3. Results and discussion 3.1. Surface morphology and component analysis The morphology changes on the surface of samples before and after thermal aging for different times were obtained by Scanning Electron Microscopy (SEM) as shown in Fig. 1. Apparently, the surface of the unaged NBR samples was relatively homogeneous and smooth and with no obvious defects appeared. With increasing aging time, small voids appeared. When the aging time reached 90 days, the surface turned rougher, the size of the voids became greater and more white spots can be observed. The appearance of white spots on the NBR samples was related to precipitation of some additives [26e29], and this phenomenon would be demonstrated below. In order to analyze the degree of thermal aging of NBR samples after 90 days conveniently, the components of oxygen and carbon on the NBR surface were analyzed by Energy Dispersion Spectroscopy (EDS) microanalysis system. Three different regions were selected in Fig. 2, which respectively corresponded to the bottom of one void (Spectrum 1), the edge of voids (Spectrum 2) and the relatively smooth region (Spectrum 3). The atomic weight proportion of oxygen and carbon (O/C) was adopted to determine the
J. Liu et al. / Polymer Testing 54 (2016) 59e66
61
Fig. 1. Micrographs of NBR samples before and after thermal aging: (a) 0 d; (b) 30 d; (c) 90 d.
environment became severe and inhomogeneous with aging time, and the oxidation reactions initiated from the NBR surface, afterwards extended to the inner structures. 3.2. Crosslink density analysis Oxygenated species produced in the accelerated thermal aging environment combined with each other and led to new crosslinking reactions occurrence, which indicated that the crosslink density of NBR samples increased with the thermal aging time [30]. The correlation between crosslink density of NBR samples and the thermal aging time was shown in Fig. 4. The linear relationship was expressed by the equation as follow:
v ¼ 6:84 104 þ 3:87 106 t
Fig. 2. Locations of NBR samples after thermal aging for 90 days for EDS analysis.
order of the degradation reaction. As shown in Fig. 3 and Table 1, there were clear distinctions in the three regions, the proportion of Spectrum 2 was highest, which meant that the degree of thermal aging was most serious for the edge region of voids of NBR samples. And the proportion of Spectrum 1 was lowest, which meant that oxidation reactions were initiated from the surface of NBR samples in the accelerated thermal aging environment. From the SEM and EDS results, it can be demonstrated that the surface damage of NBR in the accelerated thermal aging
(2)
where v was crosslink density; t was the thermal aging time. The analysis results demonstrated that crosslinking reactions mainly occurred in the process of thermal aging. 3.3. Evaluation of mechanical properties The changes of elongation at break and tensile strength with thermal aging time were shown in Fig. 5 and Fig. 6, respectively. With the aging time increase, the elongation at break decreased continuously, and the value decreased 300% in the whole aging process. In contrast, the tensile strength in Fig. 6 had an increasing trend in the first phase of thermal aging and then decreased after 70 days of thermal aging, which was related to the continuous
62
J. Liu et al. / Polymer Testing 54 (2016) 59e66
-3
Crosslink density (mol ·cm )
0.0011
-4
-6
v=6.84*10 +3.87*10 t, R=0.996
0.0010
0.0009
0.0008
0.0007
0.0006
0
20
40
60
80
100
Aging time (d) Fig. 4. Variation of the crosslink density change of the NBR samples with aging time.
500 450
Elongation at break (%)
400 350 300 250 200 150 100
0
20
40
60
80
100
Aging time (d) Fig. 5. Effect of thermal aging time on elongation at break.
10.5
Fig. 3. EDS spectra of NBR samples after thermal aging for 90 days: (a) the bottom of one void; (b) the edge of voids and (c) the relatively smooth region.
Table 1 EDS results of NBR samples after thermal aging for 90 days. Position
11.28 30.25 21.06
1.03 1.14 1.82
0.97 1.10 e
3.48 4.09 3.72
9.5
9.0
8.5
8.0 0
C (wt%) O (wt%) Mg (wt%) S (wt%) Cl (wt%) Zn (wt%) O/C
Spectrum 1 81.70 Spectrum 2 63.42 Spectrum 3 73.40
Tensile strength (MPa)
10.0
1.54 e e
0.14 0.48 0.29
20
40
60
80
100
Aging time (d) Fig. 6. Effect of thermal aging time on tensile strength.
120
J. Liu et al. / Polymer Testing 54 (2016) 59e66
3.4. ATR-FTIR analysis The FTIR spectra of NBR samples before and after thermal aging for 90 days were displayed in Fig. 8. It can be seen that there were nine characteristic peaks in the spectrum of the unaged NBR. The broad peak at 3350 cm1 was due to NeH stretching vibration and OeH symmetric stretching vibration. The peaks at 2918 cm1 and 2848 cm1 corresponded to asymmetric stretching vibration of methylene in the saturated hydrocarbon backbone and symmetric stretching vibration of methylene, respectively. The peak at 2233 cm1 was attributed to eC^N stretching vibration. The peaks at 1730 cm1 and 1660 cm1 were the very typical characteristic of C]O stretching vibration and C]C stretching vibration, separately. The peak at 1580 cm1 should be attributed to additives such as curing package [17]. The peak at 1433 cm1 was assigned to eCH2e deformation vibration, while the peak at 962 cm1 was attributed to CeH deformation vibration, which was characteristic of the trans-1, 4 structure of the butadiene part. It also can be seen from Fig. 8, when the NBR samples were aged for 90 days, the peak at 1730 cm1 decreased obviously in intensity, and the peaks at 3350 cm1, 2918 cm1 and 2848 cm1 increased remarkably in intensity. The peak at 3350 cm1 was mostly due to NeH stretching vibration and OeH symmetric stretching vibration
Fig. 7. Relationship between tensile strength and crosslink density.
before 90 d
Reflectances (%)
increase of crosslink density. The correlation between tensile strength and crosslink density of NBR was shown in Fig. 7. It can be seen that the whole change process was divided into two phases. In phase I, the tensile strength increased continuously and reached a maximum value with the increase of crosslink density, the reason was as followed: when the crosslink density increased during the early aging time, the crosslink points would increase correspondingly, which can make the positions of chain segments relatively stable and decrease the relative movement of network chains. It was beneficial to disperse and transfer the external stress and then led to the increase of tensile strength. In phase II, the tensile strength decreased obviously with the further increase of crosslink density, and the reason was as followed: when the crosslink density reached or exceeded a certain critical value, the average molar mass of the NBR chain between two adjacent crosslink points decreased and the effective mobility of the chain segment was limited, which disturbed the normal orientation of the network chains, thus the stress concentration was caused. This reduced the number of effective network chains and then led to the decrease of tensile strength [24,30,31].
63
3350
2233
1730
2848 1660
2918
1580 1433
4000
3500
3000
2500
2000
1500
964
1000
-1
Wavenumbers (cm ) Fig. 8. ATR-FTIR spectra of NBR samples before and after thermal aging for 90 days.
for unaged NBR samples, however, the NBR samples were badly aged after thermal aging for 90 days, most of antioxidants in NBR samples were consumed, so in the case the peak at 3350 cm1 was mostly due to OeH symmetric stretching vibration. It can be estimated that oxygenated species such as hydroxyl or carboxyl groups were formed for NBR in the process of accelerated thermal aging. Moreover, the decrease of the band at 1730 cm1 in intensity is possibly attributed to the consumption of carbonyl groups, or attributed to the loss of plasticizer during the thermal aging process. From the above analysis results, it indicated that hydroxyl groups were produced when NBR was exposed to the accelerated thermal aging environment.
3.5. XPS analysis Fig. 9 showed the C 1s spectra of NBR surface before and after thermal aging for 90 days. The C 1s spectrum of the unaged NBR can be fitted to two components corresponding to the CeC/CeH (284.8 eV) and CeOeC/CeOH (286.3 eV), respectively. The CeC/ CeH component was quite significant and most likely came from the hydrocarbon backbone of NBR rubber. The presence of CeOeC/ CeOH in the unaged sample could be attributed to some unavoidable side reactions during crosslinking with sulfur and other agents. Since the characteristic peak assigned to the vibrations of CeOH can be observed and changed obviously before and after thermal aging in the FTIR spectra in Fig. 8, however, the characteristic peak assigned to the vibrations of CeOeC cannot be found, it could be deduced that the peak at 286.3 eV was assigned to CeOH rather than CeOeC. After exposure to the thermal aging environment for 90 days, there was a clear chemical change around the carbon atoms, and the area percentage of the two components changed obviously compared with the unaged NBR. It was notable that the area percentage of CeOeC/CeOH decreased 3.11% in Table 2, compared with the unaged NBR, and the possible reason was that Dibutyl Sebacate plasticizer, which was a higher carbon-oxygen ratio (C/O), and antioxidant D (N-phenyl-2-naphthylamine), which was a higher carbon-nitrogen ratio (C/N), migrated from inner to surface of NBR samples. To verify the correctness of the above reasoning, the Py-GC/MS analysis of the unaged and aged NBR samples were carried out.
64
J. Liu et al. / Polymer Testing 54 (2016) 59e66
Fig. 9. High-resolution C 1s spectra of NBR samples: (a) before aging and (b) after aging for 90 days.
Table 2 XPS results from the curve fitting of the high-resolution C1s spectra of NBR before and after thermal aging for 90 days. Aging time (d)
Peak A (eV)
Structure
Peak B (eV)
Structure
0 90
284.8(93.86%) 284.8(96.97%)
CeH/CeC CeH/CeC
286.3 (6.14%) 286.3 (3.03%)
CeOeC/CeOH CeOeC/CeOH
3.6. Py-GC/MS analysis Fig. 10 showed total ion chromatograms of NBR samples. The four characteristic peaks had been marked as shown in Fig. 10, and the structure of typical volatile components is shown in Table 3. Peak 1 was assigned to butadiene monomer and acrylonitrile monomer, which constituted the NBR copolymer, Peak 2 was attributed to the small molecular of Neoprene 1212 modifier, Peak 3 and Peak 4 corresponded to Dibutyl Sebacate plasticizer and antioxidant D, respectively. It can be seen that the abundance of Peak 2, Peak 3 and Peak 4 increased obviously after aging for 90 days. The increase of the abundance of Peak 2 was that cracking reactions occurred on some molecular chains of NBR, small molecular of Neoprene were produced and accumulated continuously on the surface of NBR samples. The increase of the abundance of Peak 3 and Peak 4 showed that Dibutyl Sebacate plasticizer and antioxidant D migrated
Fig. 10. Py-GC/MS chromatograms of NBR samples: (a) control sample before aging; (b) surface of sample after aging for 90 days and (c) inner of sample after aging for 90 days.
continuously from inner to surface of NBR samples in the process of thermal aging, which depended on their own physical properties of rubber additives, such as low molecular weight, low melting point, high volatile etc. Moreover, it was noteworthy that there were no obvious differences in abundance of total ion chromatograms between the sample before degradation and inner of sample after aging for 90 days. The reason was that the migration amount of additives was
J. Liu et al. / Polymer Testing 54 (2016) 59e66
loss, while the second onset temperature was 351.37 C with a 39.27% weight loss. The results demonstrated that the accelerated thermal aging environment did not significantly affect the thermal stability of NBR.
Table 3 The structure of typical volatile components in NBR.
Peak number
65
Structure
4. Conclusions 1
CH2
2
C H
C H
CH2
CH2
C H
CN
Cl Cl O
3
O O O
H N
4
comparatively little from inner to surface of NBR samples, which did not cause the remarkable changes of component content for inner of NBR samples. The surface and inner of NBR samples can be seen as a “micro thin layer” and an “infinite thick layer”, respectively, so migration of little materials from inner to surface of NBR may cause the phenomenon that additives accumulated on the surface appeared. The results showed that additives migrated to the surface of NBR samples in the process of thermal aging, which can powerfully support the conclusions of SEM, ATR-FTIR and XPS analysis.
3.7. Thermogravimetric analysis Fig. 11 showed TGA (differential thermogravimetric) plots of NBR samples before and after thermal aging for 90 days. For the unaged NBR samples, the first onset temperature was 143.37 C with a 8.86% weight loss and the second onset temperature was 355.60 C with a 39.69% weight loss. After thermal aging for 90 days, the first onset temperature was 155.28 C with a 6.46% weight
Fig. 11. TGA plots of NBR samples before and after thermal aging for 90 days.
From the systematic studies on the accelerated thermal aging of NBR, it can be concluded that, in the accelerated thermal aging environment, the surface damage of NBR samples became severe and inhomogeneous, and the oxidation reactions initiated from the NBR surface and extended to the inner structures. Crosslinking reactions mainly occurred in the aging process. With the crosslink density increase, the tensile strength firstly increased and then decreased. In the process of accelerated thermal aging, hydroxyl groups were formed and the additives migrated from inner to surface of NBR samples. The accelerated thermal aging environment did not significantly affect the thermal stability of NBR. Acknowledgements This work was financially supported by the Research Fund of State Key Laboratory for Marine Corrosion (KF160412) and Protection of Luoyang Ship Material Research Institute (LSMRI) and Chong Qing Postdoctoral Science Foundation Grant No. Rc 201301. References [1] T. Yasin, S. Ahmed, F. Yoshii, K. Makuuchi, Radiation vulcanization of acrylonitrile-butadiene rubber with polyfunctional monomers, J. React. Funct. Polym. 53 (2002) 173e181. [2] F.S. Ahmed, M. Shafy, A.A. Abd El-megeed, E.M. Hegazi, The effect of g-irradiation on acrylonitrile-butadiene rubber NBR seal materials with different antioxidants, Mater. Des. 36 (2012) 823e828. rtolas, On the [3] M.J. Martínez-Morlanes, F.J. Medel, M.D. Mariscal, J.A. Pue assessment of oxidative stability of post-irradiation stabilized highly crosslinked UHMWPEs by thermogravimetric, Polym. Test. 29 (2010) 425e432. [4] C. Marazzato, Y. Peneva, E. Lefterova, S. Filippi, L. Minkova, Kinetics of nonisothermal degradation of nanocomposites based on functionalized polyethylenes, Polym. Test. 26 (2007) 526e536. [5] L.H. Nguyen, M. Gu, Decomposition kinetics, life estimation and dielectric study of an acrylate based photopolymer for microfabrication and photonic applications, Macromol. Chem. Phys. 206 (2005) 1659e1684. [6] G.V. Salmoria, P. Klauss, A.T.N. Pires, J. Roeder, V. Soldi, Investigations on cure kinetics and thermal degradation of stereolithography Renshape™ 5260 photosensitive resin, Polym. Test. 27 (2008) 698e704. [7] M. Jaunich, W. Stark, D. Wolff, Comparison of low tem perature properties of different elastomer materials investigated by a new method for compression set measurement, Polym. Test. 31 (2012) 987e992. [8] M. Jaunich, D. Wolff, W. Stark, A new method to evaluate the low temperature function of rubber sealing materials, Polym. Test. 29 (2010) 815e823. [9] R. Fan, Y. Zhang, C. Huang, Effect of crosslink structures on dynamic mechanical properties of natural rubber vulcanizates under different aging conditions, J. Appl. Polym. Sci. 81 (2001) 710e718. [10] K.T. Gillen, R.L. Clough, J. Wise, Prediction of elastomer lifetimes from accelerated thermal-aging experiments, Adv. Chem. Ser. 249 (1996) 557e563. [11] A. Mostafa, A. Abouel-Kasem, M.R. Bayoumi, M.G. El-Sebaie, The influence of CB loading on thermal aging resistance of SBR and NBR rubber compounds under different aging temperature, Mater. Des. 30 (2009) 791e795. € mmling, M. Jaunich, D. Wolff, Effects of heterogeneous aging in com[12] A. Ko pressed HNBR and EPDM O-ring seals, Polym. Degrad. Stab. 126 (2016) 39e46. [13] A.S. Deuri, A.K. Bhowmick, Effect of ageing on critical cut length and morphology of fracture surface in tensile rupture of natural rubber, J. Mater. Sci. 22 (1987) 4299e4306. [14] P. Budrugeac, Accelerated thermal aging of NBR and other materials under air or oxygen pressures and degradation of polymers, Die. Angew. Makromol. Chem. 247 (1997) 19e30. [15] M.M. Abou-Zeid, Radiation effect on properties of carbon black filled NBR/ EPDM rubber blends, Eur. Polym. J. 43 (2007) 4415e4422. [16] J.H. Zhao, R. Yang, R. Iervolino, B. Vorst, S. Barbera, The effect of thermooxidation on the continuous stress relaxation behavior of nitrile rubber, Polym. Degrad. Stab. 115 (2015) 32e37. [17] J.H. Zhao, R. Yang, R. Iervolino, S. Barbera, Changes of chemical structure and mechanical property levels during thermo-oxidative aging of NBR, Rubber Chem. Technol. 86 (2013) 591e603.
66
J. Liu et al. / Polymer Testing 54 (2016) 59e66
[18] R. Yang, J.H. Zhao, Y. Liu, Oxidative degradation products analysis of polymer materials by pyrolysis gas chromatography-mass spectrometry, Polym. Degrad. Stab. 98 (2013) 2466e2472. [19] X. Liu, J.H. Zhao, Y. Liu, R. Yang, Volatile components changes during thermal aging of nitrile rubber by flash evaporation of Py-GC/MS, J. Anal. Appl. Pyrolysis 113 (2015) 93e201. [20] A. Li, The aging phenomena and aging mechanisms of rubbers, Spec. Purp. Rubber Prod. 30 (2009) 56e67. [21] A. Li, Aging and lifetime prediction for rubbers, Rubber Ref. 39 (2009) 2e77. [22] Y. Miyashita, H. Kato, Cross-linking reaction of LDPE and the behavior of decomposition products from cross-linking agents, in: Proceedings of the Twenty-first Symposium on Electrical Insulating Materials, 1988, pp. 259e262. [23] Y. Xiong, G.S. Chen, S.Y. Guo, G.X. Li, Lifetime prediction of NBR composite sheet in aviation kerosene by using nonlinear curve fitting of ATR-FTIR spectra, J. Ind. Eng. Chem. 19 (2013) 1611e1616. [24] F. Delor-Jestin, N. Barrois-Oudin, C. Cardinet, J. Lacoste, J. Lemaire, Thermal ageing of acrylonitrile-butadiene copolymer, Polym. Degrad. Stab. 70 (2000) 1e4. [25] Elena V. Bystritskaya, Tatyana V. Monakhova, Victor B. Ivanov, TGA application for optimising the accelerated aging conditions and predictions of
thermal aging of rubber, Polym. Test. 32 (2013) 197e201. [26] G. Li, J.Z. Tan, J.M. Gong, Chemical aging of the silicone rubber in a simulated and three accelerated proton exchange membrane fuel cell environments, J. Power Sources 217 (2012) 175e183. [27] Alexandre de Souza Rios, Wanderley Ferreira de Amorim Jr., Elineudo Pinho de Moura, Enio Pontes de Deus, Judith Pessoa de Andrade Feitosa, Effects of accelerated aging on mechanic al, thermal and morpholog ical behavior of polyurethane/ep oxy/fi berglass compo sites, Polym. Test. 50 (2016) 152e163. [28] H. Nabil, H. Ismail, A.R. Azura, Comparison of thermo-oxidative ageing and thermal analysis of carbon black-filled NR/Virgin EPDM and NR/Recycled EPDM blends, Polym. Test. 32 (2013) 631e639. [29] A. Choudhury, A.K. Bhowmick, M. Soddemann, Effect of organo-modified clay on accelerated aging resistance of hydrogenated nitrile rubber nanocomposites and their life time prediction, Polym. Degrad. Stab. 95 (2010) 2555e2562. [30] Q.L. Zhao, X.G. Li, J. Gao, Aging of ethylene-propylene-diene monomer (EPDM) in artificial weathering environment, Polym. Degrad. Stab. 92 (2007) 1841e1846. [31] H.M. Dong, M.T. Li, X.D. Liu, C.L. Zhang, Z.F. Li, Effect of crosslink density on thermal property of HNBR, Spec. Purp. Rubber Prod. 31 (2010) 19e21.