Bridging length and time scales by AFM-based nanotribology: Applications to nanostructured ceramics and polymer surfaces.
The aim of the work described in this Thesis was to develop the necessary platform for quantitative nanotribology by atomic force microscopy at relevant length and time scales and thereby to contribute to bridge the gap between complementary nano- and microtribology. The primary focus was centered on the development of a nanotribology platform that included reliable quantification procedures for friction force measurements and the extension of the range of scanning velocities to 2 mm/s with full environmental control, as well as on the investigation of relevant tribological phenomena on the nanometer scale. The effects of nanostructure, environment and velocity on friction were revealed for different types of samples, ranging from ceramics to polymers, that are important for various types of applications, e.g. advanced coatings. Chapter 2 provided on overview of modern concepts of tribology. The aspects summarized ranged from single asperity to multi asperity friction, including energy dissipation processes, interfacial friction, atomic scale friction, friction anisotropy, velocity dependence, as well as effects of chemistry and environment. Both experimental and computational results were reviewed and instruments for micro- and nanotribology were discussed. The most important issues for quantitative friction force measurements, including friction force calibration, tip apex characterization, and tip stability, as well as critical tests using a variety of LFM probes, were discussed in Chapter 3. In particular, the frequently applied two-step calibration method for friction quantification was experimentally tested and its accuracy was estimated. The crucial parameters that limit this accuracy were identified. As shown, this method may be highly inaccurate for the calibration of Si3N4 cantilevers due to the inherent errors in the calculation of kL originating from poorly defined materials properties and insufficient precision in determination of the value of cantilever thickness. Moreover, the lateral photodiode sensitivity SL was found to be an additional significant source of error in the two-step calibration approach. Therefore, the final friction calibration factors i could not be obtained with high accuracy according to this procedure; relative errors i were 45 - 50% and 35 - 40% for V-shaped and single beam cantilevers, respectively. Finally, the wear-resistance of different tips was investigated and a method for improving the tip stability, by applying wear-resistant Al2O3 coatings by pulsed laser deposition, was developed. In Chapter 4, a new calibration standard that allows one to calibrate all types of LFM probe cantilevers independent of cantilever geometry and tip radius using a direct calibration method (the wedge calibration method, as originally introduced by Ogletree, Carpick, and Salmeron Rev. Sci. Instrum. 1996, 67, 3298 - 3306 and later improved by Varenberg, Etsion, and Halperin Rev. Sci. Instrum. 2003, 74, 3362-3367) was fabricated. The application of this standard specimen enables accurate determination of the calibration factors with an error of ca. 5%. This overcomes the limitations of the two-step friction force calibration procedures discussed in Chapter 3. As shown for oxidized Si(100), thin films of poly(methyl methacrylate) (PMMA), and micropatterned self-assembled monolayers (SAMs) on gold, the calibration of various V-shaped and single beam cantilevers using the wedge method in conjunction with the new universally applicable standard allowed us to perform quantitative nanotribology for a wide range of materials and applications. The development of a high velocity accessory for friction force microscopy measurements in controlled environment (0 - 40% RH and 0 - 40ºC) for a commercial stand-alone atomic force microscope (AFM) was described in Chapter 5. Using the accessory, a broad range of velocities up to several mm/s can be accessed independent of the lateral scan size up to a maximum scan size of 1000 nm with high lateral force signal resolution. The design and calibration of the accessory, as well as validation measurements at high velocities, were discussed. The device was validated in studies of the velocity dependence of friction forces and friction coefficients on organic [PMMA], as well as inorganic [oxidized Si(100)] samples. It was shown that the accessory allows one to bridge the time and length scales from ms to several s and tens of micrometers to nanometers, respectively, in tribological studies on oxidic ceramics systems and amorphous polymers, as also described in Chapters 7 and 8. AFM-based nanotribological measurements on advanced ceramic coatings, which were aimed at unraveling the relation of structural factors and the frictional response, were discussed in Chapter 6. In particular, the nanotribological properties of nanostructured thin films of tetragonal ZrO2 on oxidized Si(100) were investigated as a function of grain size and relative humidity (RH). The nanostructured ZrO2 showed a 50% decrease in friction coefficient Si3N4 compared to oxidized Si(100) in dry nitrogen atmosphere and 40% RH. A maximum of Si3N4 was observed at ca. 40% RH for both samples. No significant difference in friction coefficient was revealed among samples with grain sizes between 12 and 30 nm, which was attributed to insignificant differences in mechanical and nanostructural properties of the samples. In Chapter 7 we focused on performing complementary nano- and microtribology measurements of 3Y-TZP ceramics doped with 8 mol% CuO. The process of soft layer formation as reported by Pasaribu (Ph.D. Thesis, University of Twente, Enschede 2005) was studied at different length scales as a function of sliding distance of specimens previously subjected to pin-on-disc tests against Al2O3 balls. A sharp decrease in friction coefficient measured using AFM with Si3N4 tips (Si3N4nano) for wear tracks with sliding distances above 100 m was observed. For these wear tracks, similar values of Si3N4nano and Si3N4micro were revealed, under the same environmental conditions (40% RH, 26oC). These results are consistent with the formation of a soft layer generated during sliding, which reduces the friction coefficient. In Chapter 8 the first comprehensive AFM study of surface relaxations of PMMA was presented. The broad range of scanning velocities (up to 1 mm/s) accessible using the newly developed high velocity accessory (Chapter 5), temperature control (from -3oC to 26oC), as well as tips with widely different radii (20 nm to 870 nm), allowed us to cover a frequency range from 1 to 107 Hz. Friction data acquired at different temperatures and velocities were corrected for the effect of tip-sample contact pressure and were successfully shifted to yield one mastercurve. The and relaxation processes of PMMA were identified in the Hz and MHz regime, respectively (Tref = 26oC). The activation energies of the surface relaxation processes (Ea ~ 110 kJ/mol and Ea ~ 35 kJ/mol) were found to be significantly lower with respect to reported bulk values and the relaxation frequencies of the processes were noticeably higher compared to the bulk. These results are consistent with the existence of an increased free volume at the polymer surface and indicate a significantly higher mobility of the macromolecules at the film surface. As shown in this Thesis effects of nanostructure, confinement, environment and velocity on nanoscale friction can now be addressed and quantitatively determined for a wide range of materials. Expanding on instrumental, as well as technical advances, the necessary platform for quantitative nanotribology by AFM has been developed and enables one to tackle previously inaccessible phenomena in nanotribology at relevant length and time scales. Ultimately, the advances summarized in this Thesis may contribute to bridge the gap between complementary nano- and microtribology and thus to enable the development of a fundamental understanding of friction based on first principles.