1 Introduction Elastomers represent a unique class of materials combining an extremely low elastic modulus and a high elongation at break property. Rubber parts are widely used as rotary shaft seals, O rings or door seals. The rubber type, the amount of additives and the degree of cross-linking can be adjusted in order to cover a broad field of applications, particularly in the sealing industry [1]. The relatively low hardness of the elastomer leads, over the long term, to failure of the sealing part coming along with environmental pollution, energy losses and downtime. Although much progress has been made over the past few decades in improving the sustainability of elastomeric components, satisfying durability issues remain. Aus Wissenschaft und Forschung 33 Tribologie + Schmierungstechnik · 67. Jahrgang · 1/ 2020 DOI 10.30419/ TuS-2020-0004 Abdichtende Elastomerbauteile leiden an hoher Reibung und sind Verschleiß ausgesetzt, was letztendlich zu Fehlstellen und Austritt des abzudichtenden Mediums, anderweitigen Umweltbelastungen sowie Energieverlusten führt. Harte diamantartige amorphe Kohlenstoffschichten (diamond-like carbon, DLC) bewirken eine Verschleiß- und Reibungsreduzierung unter Beibehaltung typischer Elastomereigenschaften. Es wurden Parametervariationen des Beschichtungsprozesses durchgeführt. Die Nanohärte und die Schichtdicke charakterisieren die Widerstandsfähigkeit einer Schicht. Die Untersuchung erfolgt an insgesamt elf Elastomere verschiedener Hersteller aus drei Kautschuktypen (EPDM, FKM und NBR). Die Härte betrug jeweils 75 Shore A. Aus trocken laufenden Reibuntersuchungen ergaben sich Werte für Reibungskoeffizienten und Beurteilung des Verschleißverhaltens der beschichteten und unbeschichteten Elastomerplatten. Es zeigt sich, dass sich durch die Maßnahme der Beschichtung eine Reduzierung des Reibungskoeffizienten von µ unbeschichtet = 1,2 auf µ DLC = 0,2 realisieren lässt. Die Ergebnisse zeigen, dass angepasste Beschichtungen für verschiedene Elastomertypen, die als Dichtmaterialien eingesetzt werden, gegen Reibverschleiß geschützt werden können. Schlüsselwörter Elastomere, Gummireibung, DLC Beschichtung, Amorpher Kohlenstoff, a-C: H, Bias, Dichtung, PECVD, EPDM, NBR, FKM, Verschleiß, Funktionsschichten, Reibungsreduzierung Sealing rubber components suffer from high friction and wear causing leakage of lubricants. Leakages by formation of flaws lead to loss of energy and pollution. Amorphous carbon (diamond-like carbon, DLC) coatings combine their higher hardness with the high flexibility of the elastomeric substrate, also called rubber, to achieve low friction and low wear. Different coating properties were regarded by varying the deposition parameters in the PECVD process. The hardness and thickness were measured in order to carve out the best match to the preassigned requirements for a durable coating. Eleven elastomers from three rubber types (EPDM, FKM and NBR), exhibiting the same value of hardness (75 Shore A), were coated. Dry friction tests with the uncoated and coated elastomers were executed in order to determine the coefficient of friction (COF) and examine the wear resistance of the material. The investigations prove a considerably low coefficient of friction of all three rubber types by using DLC coatings compared to the uncoated variants (µ uncoated = 1.2 to µ DLC = 0.2). The promising results drive the development of customized coatings for various materials used as seals and need to be protected against frictional damages. Keywords Elastomers, Rubber Friction, DLC Coating, amorphous carbon, a-C: H, Bias, Sealing, PECVD, EPDM, NBR, FKM, Wear, Functional Layer Kurzfassung Abstract * Suleyman Bayrak, M. Sc. Dr. Dominik Paulkowski Fraunhofer-Institut für Fertigungstechnik und Angewandte Materialforschung IFAM 28359 Bremen Low Friction and Wear of Elastomers by DLC Coating Suleyman Bayrak, Dominik Paulkowski* T+S_1_2020_ 2.qxp_T+S_2018 04.03.20 15: 03 Seite 33 Material Tester (UMT1) combined with a NanoHead 2 (NH2) module. The evaluation of the roughness was made with an atomic force microscope (AFM). These measurements were performed with the Nanosurf easyScan 2. 2.2 Friction test The friction tests were run in a ball versus oscillating plate contact geometry under dry sliding conditions. Figure 1 illustrates the Universal Material Tester UMT3. Normal and friction force are recorded by a force sensor. Table 2 shows a condensed overview of the test parameters. It is well known, that the temperature and the relative humidity affect the results of the friction tests. In order to exclude thermal fluctuations, every friction test was carried out with a sample holder temperature of 30 °C. 3 Results and discussion 3.1 Cleaning and adhesion test Good adhesion of the coating to the elastomeric substrate was ensured. Therefore a pull-off test with a regular adhesive tape was done. Possible delaminations and other surface damages have been investigated under the light microscope to improve cleaning and deposition process. Figure 2 shows, for an exemplary case, the influence of the detergent on the adhesion strength. The same DLC coating has been coated on the same rubber type (EPDM). Apparently, there is a strong dependence between the adhesion strength and the detergent used in the cleaning process. 3.2 Deposition The negative charged electrode creates a direct bias voltage attracting the positive charged ions in the plasma. The rubber samples are placed on the electrode. By controlling the bias voltage, the acceleration of the ions towards the samples is controlled. As depicted in Figure 3 (a), with increasing (negative) bias voltage an increase in the deposition rate and the temperature can be observed. Aus Wissenschaft und Forschung 34 Tribologie + Schmierungstechnik · 67. Jahrgang · 1/ 2020 DOI 10.30419/ TuS-2020-0004 By modifying the surface of the elastomeric surface, bulk properties (e.g. strong damping, low modulus) can be preserved. Amorphous carbon (a C: H) films are as intended for the tribological functionalization of rubber surfaces. A thin coating thickness of one micrometer enhances the tribological behavior of the substrate drastically. This work deals with a particularly hard variant of the a C: H films, the DLC coatings. The high hardness, low wear rate and chemical inertness are typical properties of those functional layers [2]. Furthermore, the close chemical relationship between the rubber and the a C: H film enables an excellent adhesion to the rubber substrate [3]. 2 Experimental environment 2.1 Samples preparation Several rubber types were taken into account to investigate the influence of the substrate on the coating system. Rubber plates (FKM, NBR and EPDM) with a sample size of 20 x 20 mm are considered in the experiments. Oil contaminations and residual wax limit the adhesive strength of the DLC coating on the rubber surface. The cleaning of the rubber samples is therefore a decisive step before the deposition. The rubber substrates are sonicated in a tempered detergent solution (60 °C) for five minutes and rinsed with demineralized water. 2.1 Coating characterization and deposition process Due to the low deposition temperature and the high ionization energy, plasma-enhanced chemical vapor deposition (PECVD) was applied as coating process. Initially, the sample was treated by Ar plasma to plasma fine clean and activate the surface. An adhesion improving layer is deposited by introducing TMS (Tetramethysilan) into the chamber. For the functional toplayer, Toluene in combination with TMS was used as precursors. As shown in Table 1, the last deposition step is executed with two precursors. Coated silicon samples were used as references to measure the Young’s modulus. The commonly applied method for the determination of the Young’s modulus according to Oliver and Pharr is considered in this work. The nanoindentation was carried out with a Universal Gas Activation Gas flow Bias Time [sccm] [V] [s] Argon 30 400 120 TMS - - - Toluene - - - Table 1: Deposition parameter Adhesion improvement Gas flow Bias Time [sccm] [V] [s] - - - 20 600 120 - - - Functional layer Gas flow Bias Time [sccm] [V] [s] - - - 20 100 - 1200 600 80 - 110 100 - 1200 600 T+S_1_2020_ 2.qxp_T+S_2018 04.03.20 15: 03 Seite 34 Aus Wissenschaft und Forschung 35 Tribologie + Schmierungstechnik · 67. Jahrgang · 1/ 2020 DOI 10.30419/ TuS-2020-0004 Load Stroke Velocity Temperature Time shortterm Time longterm Counterpart Ball diameter Counterpart Material [N] [mm] [mm/ s] [°C] [s] [h] [mm] 4.7 11 200 30 300 24 10 100Cr6 Table 2: Test parameters of the friction test holder for counterpart crank drive force sensor sample holder sample z y x F Load F F Figure 1: Experimental setup for the tribotest with an inserted sample in the sample holder Tenside #1 Tenside #2 Tenside #3 100 μm 100 μm 100 μm Figure 2: Light micrographs of adhesive strength tests from coating to elastomer for 3 detergents, after adhesion test exemplary for EPDM-2 b) a) Figure 3 a: Variation of deposition rate and temperature with bias and b) variation of Young’s modulus with bias; p = 0,024 mbar peratu peratu T+S_1_2020_ 2.qxp_T+S_2018 04.03.20 15: 03 Seite 35 more adhesive bonds and eventually more friction. Whereas samples with a rough topography reduce the adhesive interactions. Surface comparisons are only made for samples with the lowest and highest COF. Exemplary light microscope images of coated and uncoated EPDM emphasize the wear resistance of the DLC coating on rubber materials (Figure 5). A clear wear track can be detected after five minutes sliding the ball against the unprotected substrate (Figure 5 a). The sample, which was coated with 600 V bias voltage, does not reveal any wear signs at all (Figure 5 b). Long-term friction test Additional long-term friction tests are therefore necessary to determine the durability of the DLC coatings. The friction coefficient over the course of 24 hours is shown for two representative rubber types (Figure 6). It can be demonstrated, that all rubber samples show a significantly low COF even after a long period of tribological strain. It is noteworthy that the DLC coating survived the test without any severe damages or delaminations. Figure 7 plots the COF of three rubber types tested with an increased normal load of now 10 N to determine the wear boundaries of the 600 V coatings. The chosen rub- Aus Wissenschaft und Forschung 36 Tribologie + Schmierungstechnik · 67. Jahrgang · 1/ 2020 DOI 10.30419/ TuS-2020-0004 In fact, the Young’s modulus keeps growing linearly and dependently from the bias voltage (Figure 3 b). This is related to the increasing amount of sp3 fraction in the microstructure of the amorphous carbon [2]. 3.3 Friction test Short-term friction test Figure 4 a plots the results of the short-term friction test. The uncoated rubber samples show a high friction coefficient with a relatively high standard deviation. This can be explained by the stick-slip effect typically occurring for uncoated rubbers. A very low coefficient of friction is seen for the DLC coated variants of all rubber types. The coated NBR rubber exhibits on average the lowest coefficient of friction whereas FKM shows a higher COF. That indicates that there is a role of the substrate on the friction since the mechanical properties of the coatings are consistent. One reason might be the surface topography of the coated samples (Figure 4 b). It should be noted that coated FKM shows the lowest roughness for the same coating series. It can be deduced that a low roughness results in a higher contact area between sample and counterpart, a) b) Figure 4: a) Variation of the COF by rubber type and applied DLC coating varied by bias Voltage as well as b) variation of the roughness with rubber type b) coated 600V, 4.7N, 5min uncoated, 4.7N, 5min a) Figure 5: Wear track of a) uncoated rubber and b) 600 V coated rubber after short-term friction test T+S_1_2020_ 2.qxp_T+S_2018 04.03.20 15: 03 Seite 36 ber samples are representative for the particular rubber type. DLC coated NBR and EPDM samples fail prior the end of the test. All FKM rubbers exhibit a constantly low COF. Nomenclature 600 V coated NBR Coated NBR sample; Deposition by 600 V bias voltage a-C: H Amorphous hydrogenated carbon COF Coefficient of friction DLC Diamond-like carbon EPDM Ethylene propylene diene monomer rubber FKM Fluoroelastomer NBR Acrylonitrile butadiene rubber Conclusion Different cleaning detergents have been investigated to carve out most appropriate one for the subsequent coating process. It became evident, that the detergent limits the adhesion strength from coating to elastomer. It was shown that the DLC-coatings significantly reduce the friction of elastomers. By modifying the (negative) bias voltage, it could be deduced that a high bias voltage is beneficial for the mechanical property of the DLC film. Shortand long-term friction test demonstrate the dependency of the substrate on the durability of the DLC coating. Acknowledgement The research project (IGF project 19772 N) is founded by the German Federal Ministry of Economics and Energy due to a decision in the German Bundestag. The responsible research association is the German Rubber Society. References [1] Krumeich, P.: Polymere Dichtungswerkstoffe; Resch Verlag; München; 1988 [2] Robertson, J.: Diamond-like amorphous carbon; Materials Science and Engineering: R 37; p. 129-281; 2002 [3] Martínez-Martínez, D.; Schenkel, M.; Pei, Y.T.; Sánchez- López, J.C.; De Hosson, J. Th.: Microstructure and chemical bonding of DLC films deposited on ACM rubber by PACVD; Surface and Coatings Technology; Vol. 205; p. 75-78; 2011 Aus Wissenschaft und Forschung 37 Tribologie + Schmierungstechnik · 67. Jahrgang · 1/ 2020 DOI 10.30419/ TuS-2020-0004 Figure 6: Coefficient of friction in long-term friction test of a) 600 V coated NBR and b) 600 V coated EPDM; Normal load 4.7 N COF uncoated a) COF uncoated b) Figure 7: Coefficient of friction in long-term friction test of 600 V coated NBR, EPDM, and FKM; Normal load 10 N T+S_1_2020_ 2.qxp_T+S_2018 04.03.20 15: 03 Seite 37