NEW! Exfoliated Graphite Nanofibers (EGNF)
by Prof. A.D. Lueking
Lueking and Penn State co-workers have recently synthesized a new and unique carbon-based material, EGNFs. EGNFs are closely related to exfoliated (macro-scale) graphite and (micro-scale) carbon fibers, but are over an order of magnitude smaller than previously reported with a much more regular and unique structure. The graphite nanofiber (GNF) lattice has been expanded from 3.35 Å (Figure 1A) to form interior pores that terminate along the fiber access, varying from a 4% expansion accompanied by graphitic dislocations (arrows, Figure 1B) to an estimated 15-fold expansion (Figures 1C).1 A subsequent mild oxidation led to structure collapse and the introduction of more graphitic defects (Figure 1D). Despite these structural modifications, the EGNFs retain the slit-pore geometry, nano-scale dimensions, and high aspect ratio of the original GNFs. The width of these new interior pores / expanded lattice ranges from 0.35 nm to 50 nm, depending upon treatment conditions. Further, electron energy loss spectroscopy (EELS) studies of the fibers with 15-fold expansion reveal that the interior pores consist of regions of amorphous carbon (light regions) sandwiched between regions of graphitic carbon (dark regions, Figures 1C).
Exfoliated Graphite. The initial stages of the exfoliation process for the GNFs are based on well-established techniques to exfoliate graphite, with an acid intercalation step followed by a thermal shock to vaporize the acid intercalant and expand the graphitic layers. Although the term exfoliation has been applied previously to carbon fibers and single-wall nanotubes to denote separation of fibrous strands or bundles into individual tube-like structures, this is fundamentally dissimilar to true graphite exfoliation which is a phase transition and expands the space between the graphene sheets (along the c-axis of graphite. “Traditional” carbon fibers are typically ~10 mm or greater in diameter (Figure 2), at least an order of magnitude greater than the new EGNFs. Additionally, exfoliated carbon fibers are generally highly irregular in structure, with the expansion serving to create a flowery bundle of eye-shaped pores within the cross section of the fiber (e.g. Figure 2), a structure that is clearly different from the new EGNFs.
It is generally found that beyond a minimum stacking height, exfoliation is easier for materials that are “thin” with respect to the graphitic c-axis, which can be attributed to the forces needed to overcome the inter-planar van der Waal’s forces at play between adjacent graphite layers. It was also previously reported that exfoliation was limited to graphitic particles with a width along the ab-plane in excess of 75 mm. Particle size limitations along the ab-plane of graphite are attributed to the need for a sizable intercalate “island size” to provide the force needed (upon vaporization of the intercalant) to overcome the van der Waals forces at play between graphitic layers. Thus graphite exfoliation was previously thought to be limited to fibers with micro-, rather than nano-, dimensions. The largest diameter observed in low-resolution images of the GNF precursor was ~300 nm, with most fibers having a diameter less than 200 nm, giving clear indication that exfoliation of graphite layers is not limited to particles with widths of 75 mm. In other words, exfoliation of GNFs, known for the thick dimensions along the c-axis and thin width along the ab-plane, is quite exceptional.
Structural Changes in EGNF. We have noted significant changes in the BET surface area and helium density of the EGNFs.1 EGNF-1000 has a surface area of 555 m2/g, a ten-fold increase over the precursor GNF, which we have attributed to “pitting” within the surface of the EGNF-1000 fiber (apparent in SEM images not shown here) rather than access of the nitrogen molecules to the interior lattice spacing.1 Changes in calculated helium density via buoyancy measurements indicate the 2.6 Å diameter helium atom is not able to penetrate the 3.35 Å pores but is able to penetrate the graphitic dislocations present in the EGNF-700. The increase in helium density for the EGNF-1000 suggests that helium is not able to penetrate the amorphous carbon spacers of the material and/or is inhibited by the outer graphitic layer which is observable in HR-TEM. We are currently developing methods to remove this encapsulating graphitic layer.
Hydrogen Adsorption in EGNF Materials. EGNFs were conceived by Lueking as a means to test a hypothesis that was emerging in the carbon-based hydrogen storage literature in the 2002-2003 timeframe: hydrogen uptake can intercalate and expand the graphite lattice of GNFs. Common graphite exfoliation methods were applied to GNF precursors to test this hypothesis. The increased surface area of EGNF-1000 translates to a three-fold increase in hydrogen uptake at 77 K and 20 bar; its 1.2% hydrogen uptake by weight is comparable to other recent reports of high-surface area nanoporous carbon. Interestingly, the lattice defects and small increases in the graphite lattice spacing of EGNF-700 (i.e. 3.5 Å vs. the original 3.35 Å) result in a fourteen-fold increased hydrogen adsorption at 300 K. The resulting hydrogen uptake of 0.29% at 300 K and 20 bar is far from DOE targets, but this material has not been optimized and may contain residual intercalants that either act as diluents are block access to lattice spacings. Further, we believe the undoped EGNF material may provide a virtual carbon ‘sponge’ that will accept atomic hydrogen in metal-assisted carbon, as outlined in the next section. The results to date, although not optimized, illustrate that the introduction of defects in the graphitic lattice provide a means to control the relative carbon-hydrogen binding energy, and thus the operative hydrogen adsorption temperature. In short, we have outlined evidence that the graphitic lattice may “open” to fully utilize every carbon atom in a graphitic matrix. The possible lattice spacings are smaller than the pores in SWNT bundles, and further, certain theoretical predictions show slit-pores have increased adsorption compared to cylindrical pores.
Current work on this project:
Synthesis, Characterization, and Tuning the Pore Size: Dr. Dania Fonseca
Hydrogen Storage in EGNF (pending)
Probing Hydrogen Permeation through EGNF: Mr. Nageh El-Motalib
For additional information:
1. Lueking, A. D.; Pan, L.; Narayanan, D.; Burgess-Clifford, C. E. J.
Phys. Chem. B. 2005, 109, 12710.
2. Lueking, A.D., et al. In Preparation