Layered Structure and Wear Mechanism of Concentrated Polymer Brushes Authors: Hikaru Okubo, Daiki Kagiwata, Ken Nakano, Yoshinobu Tsujii Published: 2023 ABSTRACT Concentrated polymer brushes (CPBs), which are significantly denser and thicker than conventional semidilute polymer brushes, have received increasing attention in the field of tribology because of their superlow friction properties. However, despite numerous studies aimed at enhancing CPBs for mechanical applications, the relationship between the specific layered structure and lubrication mechanisms of CPBs is still not completely understood. In this study, to reveal the relationship, simultaneous time-resolved measurements of the interfacial gap, static mechanical response, and dynamic mechanical response of the CPB at the contact interface were conducted using optical interference and precise force measuring methods. Two types of tests (i.e., the “indentation” and “sliding” tests) were alternately performed on a glass substrate coated with the CPB against a steel ball immersed in an ionic liquid. The indentation tests measuring the time-resolved interfacial gap and changes in static and dynamic mechanical responses quantitatively confirmed the presence of dilute, middle, and concentrated layers in the CPB. In the sliding tests, the wear of the CPB was detected by observing a decrease in the interfacial gap at the contact interface. Moreover, the thickness of the dilute layer remained constant with sliding, whereas the thicknesses of the other layers decreased, indicating that the dilute layer was continuously formed due to sliding. Therefore, CPB wear occurs randomly at the friction interface alongside the formation of a dilute layer with low density and stiffness on the surface. Introduction Amidst global environmental concerns, the demand for enhancing the efficiency of mechanical systems is increasing. Consequently, the reduction of friction, accountable for a few tens % of energy losses in machinery such as automotive, has become an important technological issue.1 Furthermore, EU has declared the intention of achieving a climate-neutral society by 20502 and Japan a carbon-neutral society by 2050;3 these targets highlight the importance of developing highly functional sliding machine elements with attributes such as “superlow friction,” “high durability,” and “low environmental impact”.1,4,5 Therefore, the development of sliding machine elements requires technology from fields beyond mechanical engineering. In the field of materials engineering, next-generation research and development has accelerated, and many studies have proposed the development of highly functional polymer materials. Polymer brushes, which comprise elongated conformations of polymer chains grafted onto a substrate, exhibit distinct physical and chemical properties.6−14 Many research groups have explored various applications for artificial synthetic polymer brushes that are chemically or physically grafted onto substrates, including tribological,6−9 biological,10−12 and adhesive applications.13,14 Recently, Tsujii et al. synthesized concentrated polymer brushes (CPBs), which are substantially denser and thicker than conventional semidilute polymer brushes, using advanced surface-initiated controlled radical polymerization.6,15,16 The CPBs have attracted considerable research attention in the tribology field, as they exhibit a superlow friction coefficient (μmin ≈ 10–3) when swollen in suitable solvents under micro- and macroscopic experimental conditions.17−22 In particular, the relationship between specific swollen structures and lubrication mechanisms was recently investigated.17,18,22 Miyazaki et al. suggested that the structure of a swollen poly(methyl methacrylate) (PMMA)-CPB can be classified into two layers: a bulk-concentrated layer and a surface-diluted layer. This classification depends on the molecular weight distribution as a function of film thickness, as determined through its micromechanical response using a nanoindentation method.17 They also reported that such a layered structure affects the durability of the CPB under lubricated conditions.17 Moreover, Vlădescu et al. monitored the stationary and sliding contact interfaces in CPBs by measuring the gap between PMMA-CPB and a counterface steel ball, immersed in an ionic liquid (IL), using an optical interferometry method.18 Using time-resolved gap measurements under stationary conditions and Reynolds equation-based estimations, they defined two layers within the CPB, a bulk-concentrated layer and a surface-diluted layer, based on its macromechanical response.18 They also reported that the film thickness of the surface-diluted layer changes depending on the sliding conditions, indicating that the layer behaves in a “liquid-like” manner.18 However, the relationship between the layered structure and lubrication mechanisms of CPBs has not been fully elucidated. To understand the lubrication mechanism of a tribological system, continuous observation of the contact interface between two solid materials and the development of “time-resolved measurement methods”22,23 are imperative. Simultaneously obtaining multiple pieces of information at the contact interfaces is particularly important to understand the tribological phenomena. Prior studies have developed an apparatus featuring an in situ simultaneous observation system that incorporates “ultra-thin-film optical interferometry” and “friction measuring.” This apparatus measures the interfacial gap and “static” mechanical response, including the contact and friction forces, simultaneously. Using this apparatus, the lubrication mechanisms of various tribological systems have been elucidated by correlating the acquired information.18,23−25 Meanwhile, for polymer materials such as CPBs, measuring “dynamic” mechanical response (i.e., viscoelasticity) is crucial because it strongly reflects the structure and conformation of polymers.26,27 Using surface force and resonance shear measurements, Mizukami et al. revealed that the elastic and damping parameters of PMMA-CPBs at the interfaces strongly depend on the contact and friction forces.26 Therefore, conducting simultaneous in situ measurements of the interfacial gap and dynamic mechanical response of CPBs at the contact interface potentially aids in elucidating the relationship between their layered structure and lubrication mechanisms. In this study, we developed a homemade apparatus to investigate the relationship between the layered structure and lubrication mechanism of the PMMA-CPB. This apparatus conducts simultaneous in situ measurements of the “interfacial gap”, “static mechanical response”, and “dynamic mechanical response” of CPBs at the contact interface. Optical interference and precise force measuring methods were employed to simultaneously analyze the interference spectra at the contact interface and the phase shift between the applied input–output signal and the contact interface. In particular, the apparatus was used to examine the wear mechanism of the CPB from the perspective of the layered CPB structure. Methods Materials The PMMA-CPB was deposited on a sapphire glass disk (diameter: 15 mm, thickness: 2 mm) using surface-initiated atom transfer radical polymerization (SI-ATRP), which precisely immobilizes the polymer chains from the ATRP polymerization initiator. The details of the polymerization procedure have been described in our previous studies21,22 and are briefly described in the Supporting Information. The number-average molecular weight (MN) of the free PMMA was evaluated by using gel permeation chromatography (GPC) (Shodex GPC-101). The dry film thickness H of the PMMA-CPB was determined using ellipsometry (M-2000U, J.A. Woolam Co., NE). The graft density σ of the polymer brushes was estimated using the equation shown in the Supporting Information. The characterized parameters of the PMMA-CPB are as follows: MN = 2.6 × 106, Mw/MN = 1.1, H = 850 ± 50 nm, and σ = 0.22 chains/nm2. A steel ball (AISI 52100 steel ball; Ra: 0.02 μm, diameter: 19.05 mm) was used as a counterface specimen against the PMMA-CPB. The IL, N-(2-methoxyethyl)-N-methylpyrrolidinium bis(trifluoromethanesulfonyl)imide (MEMP-TFSI, viscosity: 0.05 Pa s at 25 °C), was used as the solvent for the PMMA-CPB and test lubricant. Note that the IL is known as a good solvent of the PMMA-CPB: the IL swells the PMMA-CPB and increases its thickness several times. Apparatus Figure 1 shows a schematic of the homemade apparatus used to investigate the relationship between the layered structure and lubrication mechanisms of the swollen PMMA-CPB immersed in the IL. It comprises an “optical unit” and a “mechanical unit”. The optical unit included a high-speed camera (Fastcam Mini AX50, Photron. Co., Japan), an optical camera (Wraycam-Neo-1600, Wraymer Co., Japan), a spectrometer, and a white light source, as shown in Figure 1(a,b). The mechanical unit featured a PMMA-CPB-coated glass disk specimen mounted on a piezostage (P-733.3CL, Physik Instrumente (PI) GmbH & Co., Germany) capable of oscillating in a controlled 3-axis manner and a steel ball mounted on a 3-axis force sensor (9119AA2, Kistler Co., Switzerland). For the optical unit, white light reflected by a half-mirror passes through an objective lens (magnification: 10×) and illuminates the contact area. A portion of this light is reflected by the interface between the glass disk and the swollen CPB layer, whereas another portion passes through the swollen CPB layer and is subsequently reflected by the steel ball, resulting in interference fringes at the contact interface. To determine the interface thickness, the interference spectra were analyzed using thin-film analysis software (FiDiCa, JFE Techno-Research Co., Japan). For the mechanical unit, a sinusoidal displacement with an amplitude of 50 nm and frequency of 10 Hz was generated using a multifunction generator (WF1973, NF Co., Japan) and input into the Z-axis direction of the piezostage. The signal obtained from a strain gauge in the piezostage was then filtered by using a bandpass filter to acquire an AC signal with a frequency of 10 Hz. This AC signal served as a reference signal input to a lock-in amplifier (LI5640, NF Co., Japan). Furthermore, the force signal in the Z-direction was input into the lock-in amplifier, and the phase shift was determined by analyzing the two signals. (a) Appearance and (b) schematic of the homemade apparatus comprising the optical unit and the mechanical unit. Procedures Figure 2 shows a schematic of the test details and procedures. The two types of tests, the indentation test (step-A in Figure 2(a)) and sliding test (step-B in Figure 2(b)), were conducted on the contact immersed in the IL between the PMMA-CPB-coated glass disk and the steel ball. The two types of tests were alternatively conducted: A-0 → B-1 → A-1 → ··· → A-9 → B-10 → A-10, as shown in Figure 2(c). The indentation tests simultaneously monitored the following quantities: the piezostage Z-axis motion (z), three-directional forces (Fx, Fy, and Fz), phase shift (φ), and interfacial gap (hgap) between the steel ball and glass plate. The indentation test conditions were as follows: an applied load of 10 N, loading speed of 100 nm/s, holding time of 30 s, and unloading speed of 100 nm/s. The sliding tests simultaneously monitored the following quantities: the piezostage X-axis motion (x), three-directional forces (Fx, Fy, and Fz), and interfacial gap (hgap), where the interfacial gap during the sliding tests is not discussed in this paper. The sliding test conditions were as follows: an applied load of 10 N, loading speed of 100 nm/s, duration of 60 s, sliding stroke of 50 μm, sliding speed of 100 μm/s, and unloading speed of 100 nm/s. The sampling frequency for measuring the piezostage motion and the three-directional forces was 100 Hz, and that for measuring the interfacial gap was 10 Hz. Schematic of the test details of the (a) indentation test (A-0 to A-10), (b) sliding test (B-1 to B-10), and (c) test procedure of steps A and B. Results and Discussion Indentation Test Results Figure 3 shows the representative raw signals of the piezostage Z-axis motion (z), three-directional forces (Fx, Fy, and Fz), phase shift (φ), and interfacial gap (hgap) between the steel ball and glass plate for steps A-0 and A-10. These steps correspond to the indentation test before friction and after the 10th sliding test, respectively. For both A-0 and A-10, the contact force (Fz) increased and decreased along with an increase and decrease in the Z-axis position during the loading and unloading processes, respectively. Meanwhile, during the loading and unloading processes, the phase shift decreased from 90 to 0° and increased from 0 to 90° as the Z-axis position was increased and decreased, respectively. Notably, the interfacial gap inherently decreases and increases as the Z-axis position increases and decreases during the loading and unloading processes, respectively. The results indicate that the PMMA-CPB interface immersed in the IL exhibited static and dynamic mechanical responses corresponding to compression and decompression. Representative raw signals of the piezostage Z-axis motion (z), three-directional forces (Fx, Fy, and Fz), phase shift (φ), and interfacial gap (hgap) for steps (a) A-0 and (b) A-10. Based on the raw signals obtained (Figure 3), two types of curves (the “force–gap curve” and “phase–gap curve”) were delineated for steps A-0 and A-10, as shown in Figure 4. The curves for steps A-1 to A-9 are shown in Figure S1. It must be mentioned that the force–gap curve is different from typical force–distance curves obtained by a nanoindentation and atomic force microscopy: the force–gap curve is delineated by the gap between the ball and glass providing an absolute value, whereas the force–distance curve is delineated by the drive distance of the scanner providing a relative value, indicating that the gap is directly interpreted as the combined film thickness of the PMMA-CPB and IL for the force–gap curve. Force–gap curves for (a) A-0 and (b) A-10 and phase–gap curves for (c) A-0 and (d) A-10. For the force–gap curves (Figure 4(a,b)), the contact force increased and decreased with decreases and increases in the interfacial gap, respectively. Additionally, several specific interfacial gaps related to the inflection points of the force curves were observed. The specific film thickness of the PMMA-CPB can be determined by identifying the point at which the contact force begins to increase during the loading, defined as the Fz-response start point (hf1), corresponding to 3200 and 2300 nm for steps A-0 and A-10, respectively. Moreover, the minimum specific film thickness of the PMMA-CPB, defined as the compression limit point (hf2), was determined as 2400 and 2000 nm for steps A-0 and A-10, respectively. For the phase–gap curves (Figure 4(c,d)), the phase shift gradually decreased from 90 to 0° during loading and gradually increased from 0 to 90° during the unloading. Several specific interfacial gaps related to the inflection points of the phase curves were also observed. The specific film thickness of the PMMA-CPB can be determined where the phase shift is 90° during the loading, defined as the viscoelastic response starting point (hp1), with values of 3800 and 2900 nm for steps A-0 and A-10, respectively. Notably, if the swollen PMMA-CPB does not come into contact with the steel ball, a phase shift is never detected because the output oscillatory response is not obtained. Therefore, the point with a phase shift of 90° corresponds to the initial contact point between the swollen PMMA-CPB and the steel ball. Moreover, the specific film thickness of the PMMA-CPB where the phase shift is 0° can be determined. This point is defined as the elastic response starting point (hp2), measuring 2900 and 2200 nm for steps A-0 and A-10, respectively. In summary, four specific points can be determined from the curves; the interfacial gap at these points changed gradually depending on the sliding cycles (Figures 4 and S1). This indicates that the wear of the PMMA-CPB occurs at the interface during friction. This point will be discussed later. Note that in Figures 3 and 4, a value of the interfacial gap hgap was plotted every 5 s, although the sampling frequency of hgap was 10 Hz. For determining the four thickness values (hf1, hf2, hp1, and hp2), the complete data set of hgap was used to ensure the thickness resolution. Sliding Test Results Figure 5 shows the raw signals of the piezostage X-axis motion (x) and three-directional forces (Fx, Fy, and Fz) for steps B-1 and B-10, where Fx and Fz are the friction and contact forces, respectively. Figure 6 shows the average friction coefficients μ (= Fx/Fz) obtained in steps from B-1 to B-10 as a function of the sliding time. The raw signals for steps from B-2 to B-9 and a video of the optical observation during friction are shown in Figure S2 and Video S1 in the Supporting Information, respectively. As shown in Figure 6, the friction coefficient gradually decreased from 0.015 to 0.006 with the sliding time. The results suggest that the relatively high-friction interface structure gradually transforms into a relatively low-friction interface structure via sliding. Figure 7 shows optical images of the contact interface after steps B-1 and B-10. The color of the interference fringes considerably changes; this color corresponds to the gap between the steel ball and the glass substrate, that is, the interfacial gap. Actually, the interfacial gap decreased compared to the results of the force–gap curve from A-0 and A-10, as shown in Figure 4. Therefore, the wear of the PMMA-CPB occurred at the interface, and such a wear phenomenon may affect the frictional response of the system. Representative raw signals of the piezostage X-axis motion (x) and three-directional forces (Fx, Fy, and Fz) for steps (a) B-1 and (b) B-10. Average friction coefficient as a function of sliding time. Optical images of the contact interference after steps (a) B-1 and (b) B-10. Layered Structure of the PMMA-CPB In this study, simultaneous measurements of the interfacial gap, static mechanical response, and dynamic mechanical response of PMMA-CPB at the contact interface were conducted and the force–gap and phase–gap curves were obtained, as shown in Figures 4 and S1. From these curves, the specific interfacial thicknesses of the PMMA-CPB were determined as follows: Fz-response start point, hf1, compression limit point, hf2, viscoelastic response start point, hp1, and elasticity response start point, hp2. Based on the definitions of these specific interfacial thicknesses, we characterized the layered structure of the PMMA-CPB, and the dilute, middle, and concentrated layers were defined as follows:123where hd is the thickness of the dilute layer, corresponding to the relatively highly viscous layer; hm is the thickness of the middle layer, corresponding to the transition layer from the dilute layer to the concentrated layer; and hc denotes the thickness of the concentrated layer, corresponding to the relatively highly elastic layer of the PMMA-CPB. Figure 8 and Table 1 show the schematics and characteristics of the layers of the bare PMMA-CPB, as determined from the results of A-0, which represent the conditions before friction. In the previous study, Vlădescu et al. estimated the dilute layer thickness of PMMA-CPB to be several hundred nanometers using time-resolved gap measurements under stationary conditions and Reynolds equation-based estimations.18 The results of this study are consistent with those of the aforementioned study.18 Based on this definition, we will discuss the relationship between the layered structure and wear mechanism of the PMMA-CPB. Schematic of the layered structure: dilute, middle, and concentrated layers of the PMMA-CPB delineated by the simultaneous measurement results. Table 1Characteristics of Each Layer of the PMMA-CPB Obtained from the Indentation Test Resultsnotationrangefilm thicknessclassificationviscoelasticityhdhf1 < hgap < hp1ca. 400 nmdilute layerviscoelastic layerhmhp2 < hgap < hf1ca. 500 nmmiddle layerhchf2 < hgap < hp2ca. 500 nmconcentrated layerelastic layer0 < hgap < hf2ca. 2500 nm Characteristics of Each Layer of the PMMA-CPB Obtained from the Indentation Test Results Layered Structure and Wear Mechanism of the PMMA-CPB From the indentation results shown in Figures 4 and S1, we examined the alterations in the layered structure of PMMA-CPB caused by friction. Figure 9 shows the specific interfacial thickness behavior of each point (hp1, hf1, hp2, and hf2), alongside the thicknesses of the dilute layer hd, middle layer hm, and concentrated layer hc, plotted against the sliding time. As shown in Figure 9(a), the interfacial thickness at each point gradually decreased with sliding time, indicating the occurrence of wear in the PMMA-CPB. Notably, Figure 9(b,c,d) shows that the time-dependent behaviors of hd, hm, and hc were distinct; in other words, the wear behaviors of the dilute, middle, and concentrated layers differed. The dilute layer thickness hd remained constant throughout the sliding test, whereas hm and hc decreased first for several minutes and then remained constant. When the wear mode of the PMMA-CPBs was considered, two wear modes, sequential wear and random wear, were naturally assumed, as shown in Figure 10. The sequential wear mode is characterized by wear proceeding sequentially from the topmost surface of the PMMA-CPBs, as shown in Figure 10(a). Consequently, the polymer chains are gradually and sequentially cleaved by the sliding contact of the counterface material. In contrast, the random wear mode is characterized by wear proceeding randomly at the sliding contacts. Thus, the polymer chains are randomly cleaved by the sliding contact of the counterface material. However, based on the sequential wear mode, the dilute layer is expected to initially disappear, which contradicts the results here, as shown in Figure 9(b). Therefore, the wear of the PMMA-CPB occurs randomly at the sliding contact, indicating that the dilute layer continuously forms with a decrease in the middle and concentrated layers in the PMMA-CPB. Interfacial gap behavior of each point: (a) hp1, hf1, hp2, and hf2, along with the thickness of (b) the dilute layer hd, (c) middle layer hm, and (d) concentrated layer hc as a function of the sliding time. Wear progression mechanisms of the PMMA-CPB for the (a) sequential wear mode and (b) random wear mode. In contrast, the friction coefficient of PMMA-CPB gradually decreased with sliding time, as shown in Figure 6. From the perspective of the layered structure, the friction coefficient decreased with a decrease in the middle and concentrated layer thicknesses of the PMMA-CPB. In Figure 9(a), the ratio of the concentrated layer thickness to the total thickness changed from 75 to 65% during the sliding test. Hence, an optimized layered structure can be employed to achieve superlubricity in the PMMA-CPB. One possible mechanism is that a decrease in the CPB thickness due to the mild wear increases the contact stiffness, causing a reduction of the indentation depth, which decreases the friction force due to the CPB viscoelasticity.28,29 Further studies are necessary to elucidate the details of the relationship between superlubricity and the layered structure of PMMA-CPBs. Conclusions In this study, simultaneous measurements of the interfacial gap, static mechanical response (the three-directional generated forces due to linear relative motions), and dynamic mechanical response (the phase shift due to viscoelasticity) of PMMA-CPBs at the contact interface were conducted using a homemade apparatus to understand the relationship between the layer structure and lubrication mechanisms of PMMA-CPBs. The main conclusions are as follows:1.The static and dynamic mechanical responses of the contact interface between the PMMA-CPB and the steel ball immersed in the IL indicate that the layered structure of the PMMA-CPB is classified into three layers: dilute, middle, and concentrated layers.2.The time-dependent interface thickness behavior of each PMMA-CPB layer was distinct. The dilute layer thickness remained constant throughout the sliding test, whereas the middle and concentrated layer thicknesses decreased with sliding time. This indicates that the wear mechanism of the PMMA-CPB is characterized by random wear at the sliding contacts. Thus, the polymer chains are randomly cleaved by the sliding contact of the counterface material, resulting in the continuous formation of a dilute layer with a decrease in the concentrated layer in the PMMA-CPB.3.The friction coefficient for the tribo-pair of the PMMA-CPB and steel ball lubricated with the IL gradually decreases with a decrease in the ratio of the concentrated layer thickness in the PMMA-CPB. This suggests that an optimized layered structure is employed to achieve superlubricity in PMMA-CPB. 1.The static and dynamic mechanical responses of the contact interface between the PMMA-CPB and the steel ball immersed in the IL indicate that the layered structure of the PMMA-CPB is classified into three layers: dilute, middle, and concentrated layers. The static and dynamic mechanical responses of the contact interface between the PMMA-CPB and the steel ball immersed in the IL indicate that the layered structure of the PMMA-CPB is classified into three layers: dilute, middle, and concentrated layers. 2.The time-dependent interface thickness behavior of each PMMA-CPB layer was distinct. The dilute layer thickness remained constant throughout the sliding test, whereas the middle and concentrated layer thicknesses decreased with sliding time. This indicates that the wear mechanism of the PMMA-CPB is characterized by random wear at the sliding contacts. Thus, the polymer chains are randomly cleaved by the sliding contact of the counterface material, resulting in the continuous formation of a dilute layer with a decrease in the concentrated layer in the PMMA-CPB. The time-dependent interface thickness behavior of each PMMA-CPB layer was distinct. The dilute layer thickness remained constant throughout the sliding test, whereas the middle and concentrated layer thicknesses decreased with sliding time. This indicates that the wear mechanism of the PMMA-CPB is characterized by random wear at the sliding contacts. Thus, the polymer chains are randomly cleaved by the sliding contact of the counterface material, resulting in the continuous formation of a dilute layer with a decrease in the concentrated layer in the PMMA-CPB. 3.The friction coefficient for the tribo-pair of the PMMA-CPB and steel ball lubricated with the IL gradually decreases with a decrease in the ratio of the concentrated layer thickness in the PMMA-CPB. This suggests that an optimized layered structure is employed to achieve superlubricity in PMMA-CPB. The friction coefficient for the tribo-pair of the PMMA-CPB and steel ball lubricated with the IL gradually decreases with a decrease in the ratio of the concentrated layer thickness in the PMMA-CPB. This suggests that an optimized layered structure is employed to achieve superlubricity in PMMA-CPB.