Page images
PDF
EPUB

zirconium diboride and elemental silicon specifically to produce a glassy phase at about 1200°C (ref. 55). Carbon-carbon composites cannot usually contain the high amounts of glass-forming materials present in JTA graphite (approximately 40 percent); however, smaller quantities of boron and silicon compounds have been added to the composite in order to provide a source for the continuous generation of new glassy materials (refs. 18 and 56).

One of the problems associated with oxidation-resistant additions to organic precursors is that the addition will either be removed or modified during the process of carbonization/graphitization. Boron added as fine particulates will change to boron carbide during high-temperature treatment, whereas during polymerization of the original resin, it will oxidize and could hydrate (ref. 57). Most of these impurity or inhibition reactions may adversely affect the mechanical properties of the composite.

An alternative method to improve oxidation resistance at lower temperatures involves protecting active sites of the graphite crystal structure with reactive elements such as halogens and phosphorous compounds (refs. 10 and 11). Active sites most prone to oxygen attack are edge sites associated with the planar graphitic lattice, dislocation sites, and vacancies. Reaction of these sites with elements like the halogen gases suggests the production of stable complexes that prevent oxidation.

Microstructural Characterization Techniques

Optical Microscopy

Optical microscopy is used to relate the microstructure of carbon materials to their processing conditions (refs. 58 through 63). Hot-stage microscopy enables the observation of mesophase formation (ref. 64) and changes that occur in the crack morphology of CC as a result of increasing temperature (ref. 3). Image analysis systems have enabled quantification of the microstructure (refs. 64 and 65). Microstructural characterization techniques that have become increasingly popular include the use of polarized light (ref. 66) and differential interference contrast (ref. 67). The popularity of these techniques has largely resulted from the anisotropic properties of certain carbons, a property which leads to the generation of structurally related colors. For example, interference colors are generated when using polarized light and a half-wave retarder plate. These data can be used to assess the orientation of the constituent lamellar planes of anisotropic carbon as they terminate at a polished surface (refs. 66 through 68). The size and shape of these isochromatic regions determine their optical texture (ref. 69). Optical textures vary from isotropic (nongraphitizable) to domain anisotropic (highly graphitizable). The definitions of such textures are shown in table 4 (ref. 28). The use of optical microscopy in characterizing carbonaceous materials is limited to a resolution of about 0.5 μm. Any carbon layer less than 0.5 μm thick will appear as part of

the surrounding material. For example, the highly graphitic sheath (0.1 μm wide) observed at the surface of some PAN fibers (Type I) using the transmission electron microscrope cannot be resolved using optical microscopy (ref. 20).

Table 4. Nomenclature to Describe Optical Texture in Polished Surfaces of Coke (ref. 28*)

[blocks in formation]

Acicular flow domain anisotropy (AFD) >60 μm length; <5 μm width

[blocks in formation]

*Reprinted with permission. Marsh, H.; and Menendez, R.: Mechanisms of Formation of Isotropic and Anisotropic Carbons. Introduction to Carbon Science, H. Marsh, ed., Butterworths (London), 1989, p. 37.

X-Ray Diffraction

X-ray diffraction (XRD) provides useful information on the crystallinity of carbons and other materials. Since the beam size of the X-ray source is large, 1 mm, the results are an average obtained from a large volume of a specimen that may include many different constituents.

The average interplanar spacing d of the analyzed volume can be calculated from the Bragg equation

[blocks in formation]

where is the Bragg angle and d is the interplanar spacing.

The X-ray diffraction patterns reflect the structural order of the crystal. For carbonaceous materials, the widths of the (002) lines are generally used to calculate the crystallite height (Le) using the Scherrer equation (ref. 69),

Lc = [K]/[B002 (20) cos 0]

is the

where K is the Scherrer constant (K = 0.9), λ is the X-ray wavelength, Bragg angle, and B002 is the half-height width of the 002 diffraction line. The crystallite diameter (La) is usually determined from the half-height width, B1/2, of the 10 or 11 patterns for turbostratic materials using the Warren and Bodenstein formula (ref. 70)

La = [KX]/[B1/2(20) cos 0]

where is the Bragg angle, K = 1.77.

Crystallites in turbostratic PAN fibers typically exhibit a spacing between 3.42 Å and 3.45 Å as shown by the width of the (002) diffraction line (ref. 71). Generally, heating graphitizable carbons to temperatures approaching 3000°C increases the degree of graphitization, which can be detected by a decrease in (002) spacing to that of crystal graphite, 3.35 Å. Polycrystalline graphite produces a typical Debye ring pattern in which the width of the appropriate lines can be used to estimate the average crystallite size both in the hexagonal c direction (Le) and in the transverse a direction (La).

Scanning Electron Microscopy

Scanning electron microscopy (SEM) is used to examine the microstructure and surface topography of carbon specimens at magnifications from ~20 to 200 000 times and resolutions more than 100 Å (refs. 72 through 75). The main advantage of this technique is the large depth of focus, which is useful for the examination of fracture surfaces.

Scanning electron microscopy is used to reveal microstructural information of fractured surfaces of carbon composites or of surfaces that have been etched. Etching is often used to distinguish the different carbon components in a composite. However, over-etching should be avoided because undesired artifacts such as pitting and microcracks may be produced.

Transmission Electron Microscopy

Transmission electron microscopy (TEM) is a useful microstructural tool, because it enables direct observation of the phase details, interfaces, and degree of crystallinity within carbonaceous materials (refs. 76 through 82).

To use this technique, it is necessary to produce samples that are thin enough (<2000 Å) for the electron beam to pass through. Grinding and microtome sectioning are the two thinning techniques that have been used almost exclusively in the TEM studies of fibers, mesophase, and other carbonaceous materials. More recently, thin foil sections of CC composites have also been successfully produced using the mechanical dimpling and atom milling techniques (ref. 77).

Three TEM imaging techniques are used in microstructural analysis: bright field, dark field (DF), and lattice image. The bright field is a mode in which the transmitted beam is used to produce an image of the specimen. The DF mode is obtained by tilting the illumination system so that the desired diffracted beam coincides with the optical axis of the microscope. An objective aperture is then inserted to exclude all other beams. Dark-field imaging is used in carbon work because all crystallites of similar orientation can be identified. The diffracting crystallites are generally seen as white regions on a dark background. An illustration of this effect is shown in figure 8, taken from a pitch-based carbon fiber using the (002) reflection (ref. 80). In this figure, individual fibrils are shown highly oriented parallel to the fiber axis. Each fibril exhibits a length of a few micrometers and a thickness of 300-400 Å. Each bright region (fibril) consists of a number of crystallites with similar basal plane orientation.

Lattice imaging is another mode of operation that is well suited for direct observation of the lattice planes in carbonaceous materials. The image obtained in the image plane must be slightly underfocused in order to obtain the lattice image (ref. 83). The lattice image in figure 9 shows the typical (002) basal plane orientation of fiber and matrix regions found in transverse sections. This lattice image was formed using a portion of the (002) ring under a tilted-beam condition. The turbostratic structure of the PAN fiber was observed across the whole fiber thickness, even immediately adjacent to the interface, as seen in figure 9. Compared with the surrounding pitch matrix, the coherent domains in the PAN fiber are much smaller in size and more uniform in structure.

The selected area diffraction (SAD) technique also allows for the determination of crystallite dimensions Lc, La, and d002 in a similar way to that described for X-ray diffraction. The major advantage of SAD over XRD in such measurements is that SAD can be used to obtain crystallite information from very small regions (<1 μm). The degree of perfection of basal plane alignment Z with respect to the

[graphic][subsumed][subsumed]

Figure 8. Dark-field micrograph of fibrils within pitch-based carbon fiber.

fiber axis can be obtained from the half-width measurement of the (002) diffraction arc (ref. 2).

The SAD results of a PAN fiber in a 3-D CC composite indicate that the structure of the PAN fibers is turbostratic. Both SAD and DF show that the degree of perfection for basal plane alignment to the fiber axis increases gradually from fiber core (Z = 38°) to the surface (Z = 22°) (ref. 73).

Microstructure of CC Matrices

Pitch Matrix Composites

The fabrication of CC composites using pitch matrix materials involves first infiltrating the fiber preform with pitch before heating to carbonization temperature. As discussed previously, during carbonization various gases are evolved, and providing the rate of heating is slow, the isotropic pitch is transformed gradually to mesophase.

The molecular structure of mesophase has been described by Brooks and Taylor (ref. 84) to consist of discotic liquid crystals aligned parallel to one another. Although these layers exhibit high in-plane strength and stiffness, they can easily

« PreviousContinue »