**1. Introduction**

208 Atomic Force Microscopy – Imaging, Measuring and Manipulating Surfaces at the Atomic Scale

Wu X. H., Elsaas C. R., Abare A., Mack M., Keller S., Petroff P. M., DenBaars S. P., Speck J.

Yamada M., Mitani T., Narukawa Y., Shioji S., Niki I., Sonobe S., Deguchi K., Sano M.,

S., & Rosner S. J., (1998). *Appl. Phys. Lett.* Vol.72, pp.692-694.

Mukai T., (2002). *Jpn. J. Appl. Phys.* Vol.41 pp.L1431–L1433.

Worldwide, asphalt concrete is the most commonly used material for the top layer of pavements. The asphalt mixture's ability to provide the necessary stiffness and strength via its strong aggregate skeleton, while at the same time offering a damping and self-restoring ability via its visco-elastic bituminous binder, makes it a uniquely qualified material for increased driving comfort and flexible maintenance and repair actions. Unfortunately, bitumen supply is diminishing as crude sources are depleted and more asphalt refineries install cokers to convert heavy crude components into fuels. It is therefore becoming of imminent urgency to optimize the lifetime of the virgin bitumen from the remaining available crude sources. With 90% of the total European road network having an asphalt surface or incorporating recycled asphalt mixture in one of its base layers, the annual production of asphalt mixtures in Europe is well over 300 million tonnes. It is therefore fair to state that asphalt mixtures play a significant role in the economic viability and international position of the European pavement industry.

The intrinsic self-restoring ability of some bitumen, often referred to as its 'healing potential', could thereby serve as an excellent characteristic that could be capitalized upon. To date, however, there is still very little fundamental insight into what causes some bitumen to be better 'healers' than others. Even less is known about the resulting impact of this healing potential on the overall lifetime of the pavement. Healing potential is therefore very rarely included into pavement lifetime predictions or brought into the planning of maintenance operations, which is a missed opportunity. This will not change until a better understanding is created about the fundamental healing processes, which would allow for tailoring of bitumen during the manufacturing process and could potentially have a significant impact on an increased pavement service lifetime.

Current CE (European Conformity) specifications for bituminous binders do not contribute to advancing the understanding of the healing properties of the bitumen and even in more academic context; researchers often limit themselves to the performance of fatigue tests with and without rest periods from which an overall measure of the stiffness of the sample is calculated. The fatigued samples with rest periods may show a slower decrease of stiffness than the samples that were continuously fatigued. This ratio is then directly used as a quantification of the healing propensity of the bitumen during the rest periods. Yet it could be argued that part of this ratio can be contributed to the visco-elastic unloading behavior of

Atomic Force Microscopy to Characterize the Healing Potential of Asphaltic Materials 211

Fig. 1. Topographic 2D (left) and 3D (right) AFM image of bitumen indicating evidence of

Based on the obtained AFM images from non-contact mode and pulsed-force mode, four different material phases (i.e., hard-bee, soft-bee, hard-matrix and soft-matrix phase) were identified. However, they could not distinguish between soft-bee and soft-matrix phase as the relative stiffness for the two phases were more or less the same. After investigating the bee-shaped structures in the 5μm scale, the authors reported that the distance between the higher parts of the bees was approximately 550nm and appeared to be independent of the

After this, Masson et al. (2006) conducted extensive AFM studies on the microstructure phases in bitumen by using 12 different types of Strategic Highway Research Program (SHRP) bitumen. In this study, bitumen samples were prepared by heat casting films onto glass plates. By using AFM phase detection microscopy (PDM), the detected four phases were defined as catanaphase (bee-shaped), periphase (around catanaphase), paraphrase (solvent regions) and salphase (high phase contrast spots), which were similar to those reported by Jäger et al. (2004). Interestingly, the researchers, for the first time, found very poor correlation between the asphaltene content and the bee-shaped structures. Furthermore, no correlation was found between the PDM results and the SARA fractions (Ssatutares, A-aromatics, R-resins, A-asphaltenes) or acid-base contents of the observed bitumen. In addition, the area ratio of the catanaphase appeared to correlate with the mass parent of Vanadium and Nickel metals for several samples. The authors also reported different microstructures of the same bitumen samples prepared by heat casting and solvent casting. Continuing their study to bitumen stiffness at low temperatures under cryogenic AFM (Masson et al., 2007), at temperatures -10°C to -30°C, they showed that not all the bitumen phases contracted equally. The reason for this different contraction could be that the catana, peri and para phases were related to the existence of domains with different glass transition temperatures. One of the key findings of this low temperature AFM study was that bitumen with multiple phases at room temperature are never entirely rigid, even

De Moraes et al. (2009), studied the high temperature behaviour of bitumen using AFM at different temperatures followed by different rest periods. The authors observed that phase

microstructures

source of bitumen.

when well below the glass transition temperature.

the material and says very little about the chemo-mechanical healing propensity, nor helps in understanding of the controlling parameters.

Bitumen is a complex mixture of molecules of different size and polarity, for which microstructural knowledge is still rather incomplete. As physical properties of bitumen are largely dependent on this microstructure, the prediction of the performance of asphalt pavements is also directly related to this. A detailed knowledge of microstructure is needed to understand the physico-chemistry of bitumen, which can serve as the direct link between the molecular structure and the rheological behaviour (Lesueur et al., 1996; Loeber et al., 1998). Optical microscopy techniques have been employed for more than three centuries to study the microstructure of materials (e.g. Baker, 1742). Researchers (Lu & Redelius 2005; Hesp et al., 2007) used optical microscopy in asphalt field to have a better understanding and visualization of bitumen microstructures. However, because of the opacity and adhesive properties of bitumen, optical microscopy has not received much attention from the asphalt industry. To overcome some of the limitations of optical microscopy, researcher in the asphalt field have chosen to use scanning probe microscopy such as the atomic force microscopy (AFM). AFM is capable of measuring topographic features at atomic and molecular resolutions as compared to the resolution limit of optical microscopy of about 200nm. Moreover, the AFM has the advantage of imaging almost any type of surface which opens the window for investigating microstructures of different polymers and wax modified bitumen.
