The field of atomic force microscopy (AFM), which was invented in the mid 1980’s, has revolutionized our capabilities to explore and understand nanoscale phenomena by allowing unprecedented characterization of surface and interface reactions and molecular and sub-molecular structures. Especially with commercial instruments available for widespread academic and industrial research beginning in the early 1990’s, the atomic force microscope (AFM) has become a main tool in the suite of techniques available for characterization, and is included in most characterization facilities alongside optical and electron microscopes.
AFM now can routinely provide ~10 nm lateral resolution and angstrom vertical resolution on a variety of surfaces and in flexible environments including ambient and in situ fluid imaging and is routinely used to provide a wealth of information including topography, mechanical properties, electrical and magnetic properties on a variety of materials ranging from biological cells to semiconductors to polymers. The heart of the AFM measurements lies in the precisely monitored interaction between a very sharp tip (~10 nm in diameter) mounted on a cantilever (typically 100’s of microns long, tens of microns wide and a few microns thick) and the surface of interest via optical detection. Through this tip-sample interaction, multiple surface properties can be probed on the nanoscale, including nanomechanical properties of polymeric materials. Specifically, an AFM mode called tapping mode or amplitude modulation mode is employed to image polymeric surfaces where the cantilever is oscillated at a resonance frequency and thus gently “taps” along the surface through intermittent contact, resolving features in the material based on its mechanical properties such as stiffness and other viscoelastic properties.
Sample preparation of polymeric samples via ultracryomicrotomy for AFM analysis is critical for two reasons. First, a smooth surface is essential for effective AFM analysis as the maximum vertical range on AFM is typically less than 5µm. This is also means that if there are features that are taller than 5 µm (or whatever the specification on the particular AFM instrument), that surface will not be able to be imaged. Second, many polymer materials come in a processed form where the material has either been injection molded or compression molded and thus forms a rough skin on the surface that is not representative of the bulk material and needs to be removed.
Cryomicrotomy is able to remove the surface skin and provide a smooth surface (all at cold temperatures below Tg of the polymer, which is essential, otherwise the surface features of interest will not be resolved). The Leica EM UC6 system provides convenient AFM attachments where samples can be cryofaced in conjunction with the Leica EM FC6 and directly transferred to the AFM for convenient analysis without removing the specimen, ensuring a smooth flat surface for AFM imaging.
Shown in Figure 1a is an AFM tapping mode phase image with approximately dimensions of 5 µm x 5 µm of a cryotomed impact copolymer. The commercial impact copolymer used for this study is composed of a polypropylene (PP) matrix with micron-sized domains of ethylene-propylene (EP) rubber domains, which further contain ethylene inclusions produced in a serial polymerization reactor. In the image, the PP matrix is observed as the surrounding purple medium, and the EP rubber domain as the large round bright yellow domain in the middle. Within this EP rubber domain there is a further smaller purple inclusion which is composed of ethylene. The color contrast in this phase image is due to a convolution of various mechanical properties where the EP rubber is softer than the surrounding stiffer PP matrix.
In Figure 1b, this impact copolymer was elongated in the direction of the black arrow by 1.7 % (this elongation length is below the yield strain of PP) resulting in the AFM image shown in Figure 1b where the same rubber component is tracked from Figure 1a. Several new features are visible in this new image. First, rips and tears that were present within the rubber in Figure 1a (circled in blue), have now grown and elongated in Figure 1b (also circled in blue). Second, stretch marks (circled in red) between the rubber and the polypropylene matrix have developed at the north and south poles of the rubber domain, indicating a mismatch in Poisson ratios between the EP and PP materials. If the EP domain is stretched mainly along the equatorial line as in the experiment conducted here, then stretch marks would develop mainly at top and bottom of the EP rubber domain as observed in the AFM image in Figure 1b.
Furthermore these marks are asymmetric about the EP rubber domain and appear to be most prominent at the bottom of the domain, though stretch marks are also observed on the top portion of the domain. The sample was allowed to sit overnight at 1.7 % elongation and the next morning revealed a disappearance of the stretch marks as shown in the AFM image of Figure 1c, suggesting the yielding of the PP matrix overnight.
Finally, the effect of the stress within the PP matrix at 2 % elongation is shown in Figure 2. Both topography (a) and phase (b) images of a large-area (15 um) scan size show a number of areas where cracks have formed at the EP-PP interface and propagated into the PP matrix; some of the cracks are highlighted in blue/orange circles. The developments of cracks or shear bands and micro-voids may come from stress amplifi cation in the ICP material due to the presence of EP rubber domains. Maximum stress amplification by a spherical EP rubber domain is inversely proportional to the square root of the crack tip radius and occurs at the poles of the EP rubber domain. All these cracks and shear bands in Figure 2 appear to originate at the polar locations of the EP rubber domains, probably at sharp corners of the rubber domain with extremely small crack tip radii (and therefore maximum stress amplification resulting in a stress singularity at that point). The appearance of these shear bands and micro-voids suggests that the local stresses well exceed the yield stress despite the 2 % global deformation.
The larger cracks propagate several microns within the polypropylene matrix. However, there are also several cracks with significantly smaller dimensions of a couple hundred nm in length and tens of nm in width. Zooming in on the crack circled in orange from Figure 2 is shown in Figure 3 and reveals tiny PP fibrils stretching across the entire width of the track, as shown in the corresponding topography 3(a) and phase 3(b) images, at about 45 degree to the stretching direction suggesting that the cracking is induced by shear deformation. This particular crack is measured to be ~80 nm in depth and ~600 nm in width.
Morphology and interface adhesion of an impact copolymer (ICP) were studied using atomic force microscopy. Effects of deformation were observed within both PP and EP components as well as at the interface between the two materials. A continued stretching of the ICP could lead to delamination of EP from PP matrix. The strain required to separate the EP domains from the PP matrix could be used as a measure of the interfacial adhesion between EP and PP. Most importantly, the corresponding local interfacial stretching extent or void length between EP and PP upon delamination, which can be measured directly by AFM, can be used to calculate the interfacial strength between EP and PP. Presently, there are no direct measurement methods available to determine interfacial adhesive strength of nano- and microscale domains within polymer blends, especially blends generated in situ in polymerization reactors. This AFM examination of micro-domain deformation described qualitatively here could be used for direct determination of interfacial adhesion in complex polymer containing materials such as blends and composites.