Date of Award

12-2014

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Legacy Department

Chemical Engineering

Committee Member

Dr. Amod A. Ogale, Committee Chair

Committee Member

Dr. David P. Anderson

Committee Member

Dr. Christopher L. Cox

Committee Member

Dr. Christopher L. Kitchens

Committee Member

Dr. Mark C. Thies

Abstract

Mesophase pitch-based carbon fibers are known for their excellent thermal and electrical conductivity, high tensile modulus, moderate tensile strength, but poor compressive strength. This collection of properties results from the texture and crystalline structure (together known as microstructure) of the fibers. Fiber microstructure, in turn, develops during processing due to the discotic nature of the mesophase pitch precursor. In prior studies, such important parameters as the size and shape of capillaries in the spinneret, spinning temperature and carbonization temperature have been varied to produce fibers with different microstructures and properties. In this dissertation, the primary research goal was to investigate how the microstructure and resulting transport properties of carbon fibers would be influenced by the incorporation of short aspect ratio multiwalled carbon nanotubes (MWCNTs) or, as a low-cost alternative, carbon black (CB) at ultra-dilute concentrations. Thus, MWCNTs and CB were dispersed into the mesophase pitch precursor at only 0.3 wt%. At this extremely low concentration, rather than acting as traditional fillers, these nanomodifiers served as surface-anchoring agents, which led to changes in the microstructure of the precursor and resulting carbon fibers. These microstructural modifications then impacted fiber and composite properties. In the first part of this study, the effect of nanomodification on fiber microstructure was evaluated. Using light and scanning electron microscopy, it was observed that the cross-section of unmodified (0 wt%) fibers had a well-defined radial texture, with minimal folding of the graphitic layers (average pleat length ~40 nm), especially for the large fraction (~83%) of fibers that exhibited “pac-man” type splitting. The cross-section of fibers modified with CB had a line-centered texture that exhibited increased folding of the graphitic planes (average pleat length ~30 nm) toward the outer surface of the fiber, resulting in ~45% of CB-modified fibers displaying “pac-man” splitting. Fibers modified with MWCNTs were found to have a largely random cross-sectional texture with significant folding of the graphitic planes (average pleat length ~30 nm) across the entire surface, and only ~3% of MWCNT-modified fibers showed “pac-man” splitting. Finally, via x-ray diffraction, it was determined that nanomodification had no adverse impact on crystallite size (Lc ~40 nm and La ~80 nm), orientation (FWHM ~2°), or graphitic perfection (d002 ~0.338 nm). This indicates that nanomodification could be a possible route for producing highly graphitic fibers, which are mechanically toughened by increased folding of the graphitic pleats. The second major component of this work focused on quantifying the density, electrical resistivity, thermal conductivity and mechanical properties of individual carbon fibers (i.e., single filaments). Using a set of calibrated cesium formate aqueous solutions, fiber densities were accurately measured to be 2.20 ≤ ρ0wt% < 2.25 g/cm3, 2.15 ≤ ρMWCNT≤ 2.20 g/cm3, ρCB = 2.20 g/cm3. Thus, it was determined that external incorporation of nanomodifiers led to a small increase in percent void volume (~2%). This is consistent with a majority of literature studies that repeatedly show the undesired introduction of such voids with the incorporation of nanomodifiers. The single-filament electrical resistivity of the MWCNT-modified fibers (2.75±0.13 μΩ∙m) was not found to be significantly different (at a 95% confidence level) from the 0 wt% control (2.52±0.11 μΩ∙m); the CB-modified fibers only showed a slight increase in electrical resistivity (2.75±0.10 μΩ∙m). Similarly, fiber thermal conductivity (~550 W/m∙K) predicted from electrical resistivity values using the Issi-Lavin correlation showed no notable reduction as a result of nanomodification. Both nanomodified fibers showed a decrease in tensile strength (0 wt%: 1.71±0.21 GPa, MWCNT: 1.12±0.11 GPa and CB: 1.23±0.14 GPa) and modulus (0 wt%: 583±26 GPa, MWCNT: 520±26 GPa and CB: 527±30 GPa). Additionally, although a precise compressive strength for MWCNT- and CB-modified fibers could not be obtained (a result of limitations of the current tensile recoil testing method), all experimental fibers were determined to have a compressive strength of at least ~1 GPa. This is an improvement over previous studies. More notably, the difference in fiber structure achieved through nanomodification resulted in fibers with a better balance of compressive-to-tensile strength (σC/σT → 1), which is not observed for most highly conductivity conventional pitch-based carbon fibers. Another novel result from the present study is that the low-cost CB modifier was able to achieve similar changes in microstructure and properties as MWCNTs. In the final phase of this study, using both experimentation and finite element modeling, a method was developed to measure the bulk thermal conductivity of carbon fibers and their unidirectional composites. When applied to experimental fibers, no statistically significant difference in thermal conductivity was observed between MWCNT-modified (468±127 W/m∙K) and 0 wt% (514±179 W/m∙K) fibers. Additionally, these thermal properties were consistent with those predicted from single-filament electrical resistivity values (0 wt%: 569±18 W/m∙K, MWCNT: 533±20 W/m∙K). Thus, these types of composites could be useful as thermal management materials.

Share

COinS