Application of Peroxide Curing Systems in Cross-Linking of Rubber Magnets Based on NBR and Barium Ferrite
Rubber magnetic composites were prepared by incorporation of barium ferrite in constant amount—50 phr into acrylonitrile-butadiene rubber. Dicumyl peroxide as the curing agent was used for cross-linking of rubber magnets alone, or in combination with four different types of co-agents. The main aim was to examine the influence of curing system composition on magnetic and physical-mechanical properties of composites. The cross-link density and the structure of the formed cross-links were investigated too. The results demonstrated that the type and amount of the co-agent had significant influence on cross-link density, which was reflected in typical change of physical-mechanical properties. The tensile strength increased with increasing amount of co-agents, which can be attributed to the improvement of adhesion and compatibility on the interphase filler-rubber due to the presence of co-agents. Magnetic characteristics were found not to be influenced by the curing system composition. The application of peroxide curing systems consisting of organic peroxide and co-agents leads to the preparation of rubber magnets with not only good magnetic properties but also with improved physical-mechanical properties, which could broaden the sphere of their application uses.
Composites based on rubber matrix and various types of magnetic fillers, the so-called rubber magnetic composites, have been recognized as class of smart materials which are already found to have utilization in microwave and radar technology, motor parts, vibration absorbers, variable impedance surfaces, sensors of magnetic and electromagnetic fields, memory devices, inductor cores, and other technological applications [1, 2].
Vulcanization, often termed as curing, is one of the most important processes in rubber technologies. During this process, plastics rubber compound changes into highly elastic final product—vulcanizate. The fundamental of vulcanization is formation of chemical cross-links between rubber macromolecules. This leads to the creation of three-dimensional network structure within the rubber matrix, by reactions between the functional groups of rubber chains and suitable curing agents. Nowadays, sulfur and peroxide curing systems are still the most widely used for cross-linking of rubber matrices. Vulcanization with sulfur is a complex process leading to the formation of different types of sulfidic cross-links between rubber chains (monosulfidic C-S-C, disulfidic C-S2-C, and polysulfidic cross-links C-Sx-C (x = 3–6)). Sulfur-cured vulcanizates exhibit good tensile properties, high tensile and tear strength, good dynamic characteristics, or good abrasion resistance. The main negatives are poor resistance to thermo-oxidative ageing and weak stability at high temperatures [3, 4].
On the other hand, carbon-carbon linkages are formed between rubber chain segments by application of organic peroxides. As C-C cross-links exhibit higher dissociation energy when compared to sulfidic cross-links, the main features of peroxide-cured elastomers are high-temperature ageing resistance and low compression set at elevated temperatures [5, 6]. Simple formulation of rubber compounding, good electrical properties of vulcanizates, or no staining of the finished parts belong to other advantages of peroxide vulcanization. However, there are also some disadvantages compared to sulfur cured systems, such as low scorch safety and worse dynamic and elastic properties of vulcanizates .
As cross-linking efficiency of peroxide curing is sometimes rather low, multifunctional low molecular weight organic compounds, the so called co-agents, are often added into the rubber formulations cured with peroxides [8–10]. Co-agents boost the cross-linking efficiency by increasing the cross-link density and influencing the structure of the formed cross-links . This subsequently leads to the improvement of physical-mechanical properties of composites.
In this work, composites based on NBR and barium ferrite were cured with dicumyl peroxide alone and its combination with four different types of co-agents. The main aim was to examine the influence of curing system composition on cross-link density and physical-mechanical and magnetic properties of composites.
Acrylonitrile-butadiene rubber, NBR (SKN 3345, content of acrylonitrile 31–35%, Sibur International, Russia), was used as a rubber matrix. Barium ferrite BaFe12O19 (Magnety, Světlá Hora, Czech Republic) was used as a magnetic filler. Dicumyl peroxide, DCP, as a curing agent was supplied from Merck Schuchardt OHG, Germany. Four types of low molecular weight organic compounds: di(ethylene glycol)dimethacrylate (DEGDMA), di(ethylene glycol)diacrylate (DEGDA), trimethylolpropane trimethacrylate (TMPTMA), and trimethylolpropane triacrylate (TMPTA) were tested as co-agents. They were supplied from Sigma-Aldrich, USA.
2.2.1. Preparation and Curing of Rubber Compounds
The compounding of additives was carried out in a laboratory mixer Brabender. The rubber and the filler were mixed in the first step (7 min, 80°C), while the components of the curing system were added in the second step (4 min, 80°C). After that, the blends were homogenized in a two-roll calender.
The prepared rubber compounds were cured by using a hydraulic press Fontijne at 160°C and approximately 15 MPa. The time of vulcanization was identical with their optimum cure time, which was obtained from the corresponding curing isotherms. The rubber compounds were vulcanized in the form of thin sheets (thickness 2 mm, width 15 × 15 cm).
Acrylonitrile-butadiene rubber was filled with anisotropic barium ferrite as a magnetic filler. The content of barium ferrite was kept constant in all rubber compounds—50 phr. The characteristics of the applied magnetic filler are summarized in Table 1.
In the first part of the study, neat dicumyl peroxide was purely used for cross-linking of rubber compounds. The main goal was to investigate the amount of peroxide required for optimal properties of magnetic composites. The composition of rubber compounds is summarized in Table 2.
Then, dicumyl peroxide was used in combination with four types of co-agents. The content of dicumyl peroxide was kept constant—1 phr, while the content of co-agents varied from 5 to 15 phr. The structural formulae of co-agents are illustrated in Figure 1. The aim was to investigate the influence of the type and amount of co-agents on the cross-linking and properties of rubber magnetic composites. The composition of rubber compounds is summarized in Table 3.
The aim of the last part of the study was to examine the amount of dicumyl peroxide on cross-link density and properties of composites cured in the presence of constant amount of co-agents—15 phr. Table 4 summarizes the composition of rubber compounds.
2.2.2. Determination of Cross-Link Density
The cross-link density ν of composites was determined based on swelling of samples in acetone. The samples of composites were swelled in a solvent within time until the equilibrium swelling was reached. The experiments were carried out at laboratory temperature and a swelling time equal to 30 hours. The Krause-modified Flory–Rehner equation for filled vulcanizates  was introduced to calculate the cross-link density based upon the previously obtained equilibrium swelling degree:where is the cross-link density (mol.cm−3), is the volume fraction of rubber in equilibrium swelling sample of vulcanizate in absence of fillers, is the volume fraction of rubber in equilibrium swelling sample of filled vulcanizate, is the molar volume of solvent (for acetone = 73.519 cm3·mol−1), and is the Huggins interaction parameter (for NBR-acetone, χ = 0.54).
2.2.3. Evaluation of Physical-Mechanical Properties
The tensile properties of tested composites were measured by using a Zwick Roell/Z 2.5 appliance at a cross-head speed of 500 mm·min−1 and laboratory temperature in accordance with the valid technical standards, on the double-side dumbbell-shaped test specimens (width 6.4 mm, length 8 cm, and thickness 2 mm). The hardness was measured by using a durometer, and the unit was expressed in Shore A. The results of physical-mechanical properties are the average of five parallel measurements.
2.2.4. Determination of Magnetic Characteristics
Magnetic characteristics of prepared magnetic composites were evaluated at room temperature and a maximum coercivity of Hm = 750 kA m−1. For this purpose, a magnetometer TVM−1 (Vúzort, Praha, Czech Republic) was used. The basic principle of measurement is the induction method of scanning of scattering magnetic flux Φ induced by the magnetic vibrating sample. Magnetic field is generated by means of two cores of a Weiss electromagnet at a minimum distance of poles adapters (7.5 mm). The induced tension proportional to time dependence of magnetic flux in the sample is scanned with the system of four small cores. The cores eliminate the influence of time instability of electromagnet magnetic fields, and the change of magnetic flux is directly proportional to magnetic induction B. The specimens for the magnetic characteristics evaluation were of prism shape (8 × 4 × 2 mm).
2.2.5. Microscopic Analysis of Composites
The microstructure and surface morphology of composites were observed using the scanning electron microscope JEOL JSM-7500F at different accelerating strains. The samples were first cooled down in liquid nitrogen and subsequently fractured into small pieces with a surface area of 3 × 2 mm. Each surface was coated with a thin layer of gold and placed into the SEM.
3. Results and Discussion
3.1. Influence of Dicumyl Peroxide on Cross-Link Density and Properties of Composites
In the first part of the study, 1, 2, 3, and 5 phr of neat DCP were used for cross-linking of rubber composites filled with a constant amount of barium ferrite—50 phr. The main aim was to select a suitable amount of curing agent for achieving the optimal properties of rubber magnets.
It becomes obvious from Table 5 that the higher was the amount of DCP, the higher was the cross-link density. It is generally known that cross-linking of rubber compounds with organic peroxides proceeds via radical mechanism. In the first step, organic peroxide undergoes homolytic cleavage at high temperature to form peroxide-free radicals. The primary radicals can be fragmented into secondary radicals [6, 13, 14]. Peroxide radical species then react with rubber chains leading to the formation of macromolecular radicals, which usually recombine to form carbon-carbon cross-links [13, 15–17]. Thus, it becomes clearly apparent that the higher was the content of DCP, the more free radicals were formed from its thermal decomposition in the rubber matrix. With higher amount of free radicals, the cross-link density increased. The values of physical-mechanical properties of the prepared composites are summarized in Table 5. The hardness of composites followed the increasing trend of cross-link density. On the other hand, the higher is the cross-link density, the higher is restricted the rubber chains mobility, and thus, the elongation at break decreased. The influence of the amount of DCP on tensile strength was not significant. The tensile strength first slightly increased by increasing of DCP from 1 to 2 phr. Then, a slight decrease of the property was observed with the next increasing amount of DCP.
The incorporation of magnetic fillers into rubber matrices leads to the preparation of rubber composites with unique magnetic properties [18, 19]. As in the present study, the content of barium ferrite in composites was kept on a constant level and the possible influence of curing system composition on magnetic properties was examined. The remanent magnetic induction Br and the coercive intensity of magnetic field (coercivity) Hc belong to the most important characteristics of magnetic materials. The first parameter represents the value of residual magnetization retained in the magnetic material, when external magnetic field is removed. The coercivity represents the intensity of external magnetic field, which is needed to abolish the remanent magnetic induction in the material. The higher the values of both characteristics the magnetic materials have, the better the permanent magnets they are. As shown in Table 5, the remanent magnetic induction fluctuated only in the low range of experimental values, independently on the amount of peroxide. The same statement can also be applied on coercivity of composites.
3.2. Influence of the Type and Content of Co-Agents on Cross-Link Density and Properties of Composites Cured with Constant Amount of DCP
The achieved results demonstrated that, with increasing amount of dicumyl peroxide, the cross-link density increased. The increase in the cross-linking degree resulted in the increase of hardness and decrease of elongation at break. However, the influence of DCP content on tensile strength was not significant. Based upon the obtained outputs, it can be stated that the amount of 1 phr DCP is sufficient for reaching optimal tensile properties for rubber magnetic composites. Therefore, in the next step, 1 phr DCP was used in combination with 4 types of commercially available co-agents, which were dosed to the rubber compounds in 5, 10, and 15 phr. The composite cured only with dicumyl peroxide of 1 phr in absence of co-agents is, in graphical illustrations, specified as reference (REF).
As can be observed from Figure 2, the application of 5 phr co-agents leads to the increase of cross-link density of composites. It also becomes apparent that the higher was the amount of co-agents, the higher was the degree of cross-linking. Thus, it can be stated that co-agents take an active part in the curing process. The reaction mechanisms of co-agents in peroxide vulcanization of rubber compounds have been a subject of ongoing research, and several reaction pathways have been proposed for different types of co-agents. Grafting of co-agents among rubber chains and the formation of an interpenetrating network of homopolymerized co-agents and elastomer chains have been suggested as the most common reaction mechanisms and network structure enhancement (Scheme 1) [6, 9, 11, 20–22]. The influence of the applied co-agents on the cross-link density of tested composites can be deduced based on their chemical structure (Figure 1).
With the exception of the composite cured only with peroxide (REF), the lowest cross-link density exhibited composites cured with DCP and di(ethylene glycol)dimethacrylate (DEGDMA). DEGDMA is a linear compound with two active double bonds in vinyl groups (Figure 1(a)). However, vinyl groups are shielded by methyl groups, which can act as steric hindrance against coupling of co-agent molecules onto rubber chains. Therefore, it is expected that DEGDMA can rather undergo homopolymerization reactions than couple onto rubber chains. It must also be noted that, for activation and homopolymerization of co-agent, free radicals formed from peroxide decomposition are required. Then, the concentration of peroxide-free radicals might not be sufficient for cross-links formation. As a consequence, the cross-link density could be lowered. As seen in Figure 2, by application of DEGDMA, the cross-link density of composites slightly increased. Thus, it becomes apparent that, in addition to homopolymerization, coupling of original or homopolymerized molecules of DEGDMA onto rubber chains will proceed concurrently. However, the achieved results suggest that the dominant reaction mechanism of DEGDMA is homopolymerization. Composites cured with DCP and di(ethylene glycol)diacrylate (DEGDA) showed higher cross-link density. The structure of DEGDA is similar to DEGDMA with the absence of methyl groups (Figure 1(b)). Based upon this and achieved results, it is supposed that dominant reaction mechanisms of DEGDA are addition reactions and grafting onto rubber chains. Trimethylolpropane trimethacrylate (TMPTMA) has three vinyl groups with active double bonds, which are sterically hindered by methyl groups (Figure 1(c)). Homopolymerization and addition of TMPTMA onto rubber chains can proceed concurrently; however, the most likely reaction mechanism is expected to be homopolymerization, just as in the case of DEGDMA. The highest degree of cross-linking as well as the highest increase of cross-link density, in dependence on co-agent content, was recorded by application of trimethylolpropane triacrylate (TMPTA). Chemical composition of TMPTA is similar to TMPTMA, again with absence of methyl groups in vinyl positions (Figure 1(d)). Thus, TMPTA exhibits the highest amount of easily accessible double bonds. Based on the outlined facts, it becomes clearly obvious that the prevalent reaction mechanism is grafting onto rubber chains and the formation co-agent bridges between rubber macromolecules.
The hardness (Figure 3) and elongation at break (Figure 4) of composites followed the dependences on the cross-link density. So, the higher was the degree of cross-linking, the higher was hardness, but the lower was the elongation at break. Composites cured with DCP and DEGDMA exhibited the lowest cross-link density, and thus, they also showed the lowest hardness and the highest elongation at break in all co-agent concentration scales. On the other hand, composites cured with DCP and TMPTA showed the highest cross-link density, thus also the highest hardness and the lowest elongation at break.
Application of all tested co-agents in peroxide curing of rubber compounds resulted in the increase of tensile strength of NBR-based magnetic composites. As seen in Figure 5, the tensile strength showed increasing trend with increasing content of co-agents. The tensile strength of the composite cured with DCP and 15 phr TMPTA reached almost 10 MPa, which represents almost threefold increase when compared to the reference. The tensile strength of composites cured with DCP and 15 phr DEGDMA or TMPTMA was almost fourfold higher compared to that of the reference (the tensile strength increased from 3 MPa for the composite cured only with DCP to almost 12 MPa for the composites cured with DCP and 15 phr DEGDMA or TMPTMA). It becomes clearly apparent that presence of co-agents in peroxide curing process has significant influence on tensile characteristics of rubber magnetic composites. Tensile strength of cured rubber compounds is a complex property, dependent not only on the cross-link density but also on the structure of formed cross-links and on the interphase conditions between the rubber and the filler. Magnetic fillers, as also applied barium ferrite, do not provide reinforcing effects to the rubber compounds. As inactive fillers, they are incorporated into rubber compounds in order to impart magnetic characteristics to the composites. As already outlined, co-agents contribute to the formation of complex cross-link structure within the rubber matrix by homopolymerization and grafting onto rubber chains. On the other hand, owing to their polar character, they are supposed to enhance the adhesion to barium ferrite, which also belongs to the polar materials [23, 24]. Thus, they contribute to the increase of adhesion and compatibility on the interphase filler-rubber, by their chemical bonding onto rubber chains and physical couplings onto the ferrite filler. This subsequently leads to the overall reinforcement of rubber magnetic composites. The reinforcement of rubber magnetic composites cured with peroxide curing system and combined sulfur and peroxide systems was also achieved in our previous works [25, 26].
The magnetic properties of composites were also examined. As seen in Figure 6, the remanent magnetic induction fluctuated only in the low range of experimental values, independently on the type and amount of co-agents. The values of coercivity moved from 100 to 102 kA m−1 again, with no dependence on the curing system composition.
3.3. Microscopic Analysis
From SEM images, it is possible to observe that the homogeneity and compatibility between ferrite and the rubber matrix is higher in the case of composite cured with peroxide and of co-agent (DCP and 15 phr of TMPTMA) (Figure 7(b)). In the case of composite cured only with organic peroxide, there is more evidently visible the presence of microcavities and voids on the interphase filler-rubber (Figure 7(a)). The results confirmed the presumption that co-agents improve the adhesion to the applied filler and thus, they contribute to the improvement of adhesion between the rubber and the filler on the interphase.
3.4. Influence of the Content of Dicumyl Peroxide on Cross-Link Density and Properties of Composites Cured with Constant Amount of Co-Agents
The results revealed that co-agents are effectively involved in peroxide curing process of rubber magnetic composites by increasing the cross-link density, influencing the cross-link structure and improving the adhesion on the interphase filler-rubber. Taking into consideration that, for cross-linking of magnetic composites was used 1 phr DCP, it is not clear whether its thermal decomposition leads to the formation of sufficient amount of free peroxide radicals required for chemical reactions of co-agents, mainly at higher co-agent concentrations. Moreover, part of free peroxide radicals is just depleted for carbon-carbon cross-links formation within the rubber matrix. Thus, the concentration of peroxide-free radicals could be lowered. Based upon the outlined facts, the goal of the next research was to investigate the amount of DCP on cross-linking and properties of composite materials with a constant level of co-agents—15 phr. The content of DCP was increased to 2 and 3 phr. Composites cured in the presence of 1, 2, and 3 phr DCP are designated as reference (REF).
It becomes obvious from Figure 8 that, with increasing amount of DCP used for cross-linking, the cross-link density increased. Moreover, peroxide-free radicals initiated chemical reactions of co-agents, which also contributed to the increase of cross-link density by formation of complex cross-link network structure within the rubber matrix. As also seen in Figure 8, the character of cross-link density on the type of applied co-agent remained the same also at higher DCP content. The lowest degree of cross-linking exhibited composites cured with DCP and DEGDMA; on the contrary, the highest cross-link density showed composites cured with DCP and TMPTA.
The increase of hardness with increasing amount of DCP is apparently visible from Figure 9. The lowest hardness showed composites cured in the presence of DEGDMA with the lowest cross-link density. On the other hand, composites cured with DCP and TMPTA with the highest cross-linking degree exhibited also the highest hardness. The elongation at break, by contrast, showed a decreasing tendency with increasing content of DCP (Figure 10). The elongation at break decreased by increasing DCP from 1 to 2 phr. But then, there was only a little change of elongation at break at the maximum amount of DCP. As already outlined, the reason for the decrease of elongation at break can be attributed to the restriction of mobility and flexibility of rubber chains due to a very high cross-link density. Too high degree of cross-linking was also responsible for the decrease of tensile strength (Figure 11). At very high cross-link density, the elasticity of elastomer chains is fairly restricted and the applied tensile and deformation forces cannot be effectively redistributed onto the whole cross-link network structure of rubber matrix uniformly. The regions of rubber matrix with the highest cross-link density can act as stress concentrators causing inception and crack growing. The result is the destruction of the composite by application of deformation forces. However, it should be remarked that the tensile strength of composites cured in the presence of dicumyl peroxide and co-agents was still higher in comparison with reference composites cured only with dicumyl peroxide at all concentrations of DCP.
Finally, it can be stated that magnetic characteristics of composites were not influenced by the composition of the curing system (Figure 12).
Barium ferrite was compounded with NBR in order to prepare rubber magnetic composites. The peroxide curing system, consisting of neat dicumyl peroxide or combination of dicumyl peroxide with co-agents, was used for cross-linking of rubber magnetic composites. The results demonstrated that co-agents take an active part in peroxide curing process, leading to formation of more complex cross-link structure within the rubber matrix. Moreover, co-agents improve the adhesion and compatibility between the rubber and the filler on the interphase. The overall reinforcement of rubber magnetic composites was subsequently achieved. The influence of the curing system composition on magnetic characteristics was not observed.
The data used to support the findings of this study are included within the article.
Conflicts of Interest
The authors confirm that the mentioned received funding in the “Acknowledgment section” did not lead to any conflict of interests regarding the publication of this manuscript. The authors also confirm that there are not any other possible conflicts of interests in the manuscript.
This work was supported by the Slovak Research and Development Agency under the contract nos. APVV-16-0136 and APVV-16-0059 and by the grant agency VEGA (project no. 1/0405/16).
J. Kruželák, R. Sýkora, and I. Hudec, “Sulphur and peroxide vulcanisation of rubber compounds-overview,” Chemical Papers, vol. 70, no. 12, pp. 1533–1555, 2016.View at: Google Scholar
A. Thitithammawong, C. Nakason, K. Sahakaro, and J. W. M. Noordermeer, “Thermoplastic vulcanizates based on epoxidized natural rubber/polypropylene blends: selection of optimal peroxide type and concentration in relation to mixing conditions,” European Polymer Journal, vol. 43, no. 9, pp. 4008–4018, 2007.View at: Publisher Site | Google Scholar
S. K. Henning and R. Costin, “Fundamentals of curing elastomers with peroxides and coagents,” Rubber World, vol. 233, pp. 28–35, 2006.View at: Google Scholar
X. Liu, T. Zhou, Y. Liu, A. Zhang, C. Yuan, and W. Zhang, “Cross-linking process of cis-polybutadiene rubber with peroxides studied by two-dimensional infrared correlation spectroscopy: a detailed tracking,” RSC Advances, vol. 5, no. 14, pp. 10231–10242, 2015.View at: Publisher Site | Google Scholar