WO2007050460A2 - Carbon nanofiber paper and applications - Google Patents

Abstract

A carbon nanofiber paper comprising a web of sulfonated multi-walled nanofibers. In one embodiment, the nanofibers have a stacked cup or cone helix structure, a diameter of about 10 to 500 nm, and carry sulfonic acid groups in an amount of at least about 5.3 x 10-3moles/cm2. The paper can be used in a fuel cell membrane such as a gas diffusion layer or catalyst support.

Description

CARBON NANOFIBER PAPER AND APPLICATIONS

FIELD

[0001] The present invention relates to the preparation and use of a paper or sheet material made from a type of carbon nanofiber with a high concentration of sulfonic acid/sulfonate groups distributed over its nanofiber surface. The sheet described can advantageously be made without a binder, as it derives cohesive strength from the inter- fiber attractions of polar sulfonic-acid-derivatized surfaces. The invention also relates to the use of this highly sulfonated nanofiber sheet material in a fuel cell, and especially in a direct methanol fuel cell or a fuel cell that operates above 100^°C.

BACKGROUND

[0002] Carbon nanotubes and nanofibers are of interest for a host of emerging applications in electronics, composites, energy storage, energy conversion, and bio- related uses. They are generally synthesized from hydrocarbons in the vapor phase at elevated temperature, using processes that involve furnaces, lasers, plasmas, or arc- discharge. Carbon nanofibers have also been produced from a coal feedstock, as described in US. Patent 5,846,509 of Alig et al., issued Dec 8, 1998.

[0003] Single- or multi-wall carbon nanotube/nanofibers are filamental in nature, with diameters from one nanometer to 1000 nanometers. Their structures vary, and include concentric tubes of graphitic sheets, fish-bone structures with angular opposing graphitic platelets, stacks of small graphite sheets, tubular forms with either a 'stacked cup' or 'cone-helix' morphology, and any of these varieties with an added overcoat of pyro lytic or amorphous carbon. These structural types are diagrammed in Figure 1. In this disclosure we will use the term "nanofiber" as opposed to "nanotube" to denote a nanotube with a diameter above IOnm, too large to exhibit quantum effects. It should be clear, however, that a nanofiber can have a hollow core and a gross tubular structure.

[0004] The preparation and properties of carbon nanotubes/nanofibers are described in a number of patents. Tennet U.S. Pat. No 4,663,230, issued May 5, 1987, describes carbon nanofϊbrils that are free of an amorphous carbon coating and consist of multiple graphitic outer layers substantially parallel to the fibril axis. Tubular fibrils with graphitic sheet layers and diameters between 3.5 and 75 nanometers are described in Tennet et al. Pat. No. 5,165,909, issued Nov. 24, 1992, Tennet et al. U.S. Pat. No 5,171,560, issued Dec 15, 1992, and Mandeyille et al U.S. Pat. No 5,500,200, issued Mar. 19, 1996. Nanofibers with a fishbone morphology are described in European Patent Application No. 198,558 to J.W. Geus (published Oct 22, 1986).

[0005] A variety of methods for forming nanotubes into structures have also been disclosed. Snow et al describes interconnected networks of single-wall nanotubes in U.S. Patent 6,918,284, issued July 19, 2005. Cabert et al. discusses a single-wall nanotube catalyst support in U.S. Patent 6,824,755, issued Nov. 30, 2004. The formation of an entangled single wall nanotube solid and method for making the same is disclosed in U.S. Patent 6,899,945 of Smalley et al., issued May 31, 2005. Each of the preceding patents pertain to single-wall nanotubes. U.S. Patent 6,824,689 of Wang et al., issued Nov 30, 2004, describes nanotubes disposed over a porous support such as a foam, felt, mesh, or membrane. Its procedures involve use of a surfactant to facilitate the water dispersion of a hydrophobic nanotube, and subsequently place it on a large-pore support that is preferably metal or ceramic. Rueckes et al. describes nonwoven aggregates of nanotube segments where segments contact one another to define a plurality of conductive paths, and also a nanotube fabric made by depositing a solution of suspended nanotubes on a substrate in U.S. Patents 6,706,402, issued March 16, 2004, and in U.S. Patent 6,835,591, issued Dec. 28, 2004. These two disclosures pertain to making patterned devices for memory arrays, where the nanotubes remain on the substrate. A method of forming a nanotube structure is described in U.S. Patent 6,682,677 of Lubovsky et al., issued Jan 27, 2004. It describes nanotube ribbons made by coagulation spinning. U.S. Patent 6,615,169 of Tennent et al, issued Dec 16, 2003, describes a nanotube network where the nanotubes are surface coated with a network of semiconducting material. U.S. Patent 6,921,575 of Horiuchi et al., issued July 26, 2005, describes forming a network of nanotubes by creating-a^*5igh- viscosity dispersion and then removing the carrier, for purposes of creating wiring, circuits, or planar electrodes. In Patent Application 20040202603, submitted Oct 14, 2004 and incorporated by reference, Fischer et al. disclose an electrode made of functionalized nanotubes, and a porous network of functionalized nanotubes where the functionalized fibrils are cross- linked by a bifunctional or polyfunctional group.

[0006] A sulfonic-acid-substituted carbon nanomaterial for fuel-cell use is described in Patent Application 20040115501, submitted by Hinokuma on June 17, 2004. This application describes carbon clusters with acid functional groups that show proton conductivity even in a dry atmosphere and in a temperature range above 100⁰C.

SUMMARY OF THE INVENTION

[0007] One manifestation of the invention is a sulfonated multiwalled nanofiber. A more particular manifestation of the invention is a sulfonated nanofiber that is prepared from a nanofiber having a structure that includes high surface levels of graphitic sheet edges. These sheet edges are sites where covalently bonded sulfonic acid may be attached, in contrast to the smooth basal-plane surfaces of nanotubular graphite. The high surface coverage of edge sites allows the preparation of a nanofiber with an unusually high level of sulfonic acid groups on its surface. In one embodiment, this makes the nanofibers water dispersible without the use of surfactants, creating solutions that will pass through fine-mesh (90 micron) screens. Such a liquid of well-dispersed nanofibers can be formed into a homogenous freestanding sheet by filtration. In accordance with a further embodiment, it also has the properties of hydrophilicity and proton-conduction along its surface, which is valuable for fuel-cell applications and other electrode uses. In one embodiment, papers made with this process can be as thin as 2 mils (.002 inch) thick.

[0008] This disclosure also pertains particularly to nanofibers that exhibit a high concentration of carbon edge sites on their surfaces, with diameters between IOnm and 500nm. Such nanofiber structures include: 1) the stacked-cup morphology, 2) the cone- helix structure, which is a tight spiral ribbon of graphene sheet that resembles the stacked-cup morphology, 3) nanofibers 1 or 2 coated with turbostratic carbon, and 4) partially graphitized (i.e., heat-treated) versions of the aforementioned nanofiber types. It more particularly applies to the stacked-cup or cone-helix morphology. Nanofibers with a stacked-cup or cone-helix morphology are produced by methods described in U.S. Pat. No. 5,024,818 of Tibbets et al. and U.S. Pat. No. 5,374,415 of Alig et al., which disclose a high-temperature gas phase reaction of hydrocarbons with a catalytic metal particle in a non-oxidizing gas stream.

[0009] Figure 3 provides a closer examination of one of the geometries used in one embodiment of the invention, highlighting the graphitic sheet structure of the edge site and showing the attachment of a surface group. In one embodiment a high concentration of sulfonic acid is distributed over the entire exterior surface of the nanofiber. By contrast, single wall nanotubes (or multi-wall nanotubes that consist of substantially concentric graphitic tubes) have few surface sites where chemical groups may be attached; their structures allow surface bonding only at defect sites or at the tube ends. Thus single-wall nanotubes and multi-wall nanotubes that consist of graphitic layers substantially parallel to the fibril axis and free from an amorphous carbon coating do not pertain to this invention.

[0010] The processing of carbon nanofibers into useful and functional forms is important for exploiting these novel materials. In one embodiment a paper or nonwoven sheet is provided including either pure nanofibers or nanofibers with inclusions of other materials such as fibers or powders of carbon, ceramic, or polymer. Such papers can have a range of conductivities, homogeneities, porosities, wettabilities, surface chemistries and tensile strengths.

[0011] A process of forming a nonwoven paper in accordance with one embodiment of the invention employs a dispersion method to create separated fibrils that can later be filtered to make a homogenous assembly; in another embodiment papers are made using a mechanism of cohesion within the assembly that will hold the paper together; a third embodiment involves matching of a fibril to a filter membrane such that the formed paper cleanly releases from its filter backing.

BRIEF DESCRIPTION OF THE DRAWINGS [0012] FIGURE 1 is a schematic illustration of nanofiber structures: A- Stacked cup structure, B- Cone helix structure, C-Fishbone structure, D- Single wall nanotube, E- Concentric tube structure, F - Stacked platelet structure.

[0013] FIGURE 2 is a schematic comparison of the 'cone-helix' and 'stacked cup' geometries. Images courtesy of Prof. John Dismukes, presentation at the Electrochemical Society Fall Meeting, October 2004 in Hawaii.

[0014] FIGURE 3 is an illustration of a carbon edge site, showing the covalent attachment of a sulfonic acid group.

[0015] FIGURE 4 is a cyclic voltammogram of a sulfonated nanofiber paper sample in 0.5M H₂SO₄. Voltages are referenced to a saturated calomel electrode (SCE). The two peaks correspond to the reversible electrochemical conversion of an attached sulfonic acid group to an attached sulfϊnic acid group (SO₃H / SO₂H).

[0016] FIGURE 5 is a plot of resistance vs. frequency for a sulfonated nanofiber paper: A-Sample tested at room atmosphere, B-The same sample tested after equilibration in a high humidity environment. This data was acquired by electrochemical impedance over the frequency range 20,000 to 0.1 Hz.

[0017] FIGURE 6 is a resistance vs. frequency graph of a nanofiber paper sample after heating it in an inert atmosphere to 650⁰C to strip the sulfonic acid groups off the surface. This paper is a simple electrical conductor with no proton conducting properties.

[0018] FIGURE 7 illustrates comparative performance of different anode types in direct methanol fuel cell testing. The sulfonated nanofiber paper (Curve C) shows about 16 times the current at a comparable catalyst loading compared to the Toray paper standard (Curve A), a 16-fold improvement in catalyst utilization.

[0019] FIGURE 8 is a transmission electron microscope image of an individual PR-24 carbon nanofiber that has been graphitized by heat-treating it to 3000^°C. [0020] FIGURE 9 is a comparison of the commercial ELAT 2500 (Curve A) and the same material modified with a thin layer of sulfonated nanofibers applied with the papermaking technique (Curve B).

[0021] FIGURE 10 illustrates anode configurations used in a comparison example. Configurations A and B used the Toray Carbon Paper (CP) (0.09 mm thickness, De Nora Inc.) as a support, whereas C used the sulfonated nanofiber paper support in accordance with one embodiment of the invention.

DETAILED DESCRIPTION

[0022] The high levels of surface sulfonic acid groups has been demonstrated indirectly with X-ray Photoelectron Spectroscopy (XPS). XPS is a surface analysis technique that has a depth penetration of about 5 to 10 angstroms. This means that when it is applied to a carbon surface, XPS will detect sub-surface atoms (primarily carbon) as well as surface atoms, yielding 'surface' analyses that are overweighted in carbon, and too low in other species. Because XPS detects so many sub-surface carbons, it has been estimated that its reported values for non-carbon surface atoms need to be multiplied by as much as 5.1 to obtain their true surface concentration. (J. Dismukes et al., Electrochemical Society conference presentation, October 2004, Hawaii.)

[0023] An XPS analysis was performed on a PR24 nanofiber made by Applied Sciences of Cedarville Ohio that was sulfonated by attaching azophenyl sulfonate groups to the surface. XPS showed a 1.27% atomic surface concentration of sulfur, and a 2.03% surface concentration of nitrogen (the ratio of S :N in azophenyl sulfonate is 1 :2). At first glance, a measured value of 1.27% atomic surface sulfur appears to be small, but in fact it is close to the theoretical maximum XPS value for an azophenyl sulfonate derivatized carbon surface. The atomic concentration of sulfur within the azophenyl sulfonate group itself is only 9% (hydrogens are not counted). Furthermore, it is possible to attach the azophenyl sulfonate group to less than one half of the nanofiber surface carbons because of valence rules. This drives down the maximum possible atomic sulfur at the surface of a azophenyl-sulfonate derivatized carbon to under 4.5%. Finally, when one considers that XPS will underreport its sulfur analysis because the measurement technique has some depth penetration into the bulk nanofϊber, it is seen that an XPS-measured value of~l% surface sulfur is as high as one might reasonably expect for a monolayer coverage of azophenyl sulfonate.

[0024] Further evidence of high surface coverage by sulfonic acid is supplied by bulk elemental analysis. Ordinarily, a single-layer coverage of groups on a solid represents an infinitesimal fraction of the total solid, so that its measurement by bulk elemental analysis is problematic. However, because the nanofϊbers are so small, surface groups are a non- trivial fraction of the nanofϊber atoms, especially if their surface coverage is high. One can approximately calculate the theoretical sulfur content of a nanotube completely covered with sulfonic acid groups by using the density and diameter of the nanofiber host. The density of individual PR24 nanofibers is known to be ~1.6 g/cc, and their average diameter is ~70 nm. By calculating the area of the nanofiber surface and dividing it by the area occupied by an individual sulfonate group, the maximum number of possible surface sulfonates is determined. The weight of the sulfur in these groups can then be compared to the nanofiber' s weight to give an expected percentage of sulfur content. For PR24, this maximum value is approximately 0.47wt.%.

[0025] The actual bulk elemental analysis of a sulfonated PR24 is complicated by the presence of native sulfur that is introduced during the manufacture of the nanofiber. This condition exists because individual PR-24 nanofϊbers are synthesized in the gas phase by contacting gaseous hydrocarbons with suspended particles of iron sulfide; the nanofϊber emanates from the FeS particle surface. Some of this iron sulfide can be washed from the resulting nanofϊbers with dilute hydrochloric acid, but some of it is retained in a carbon-encapsulated form at one end of the nanofϊber. To distinguish this sulfur content from sulfonic-acid-type sulfur, one can derivatize the carbon surface with azophenyl sulfonate, which contains a 2:1 ratio of nitrogen to sulfur. One can then use the bulk nitrogen content of the nanofϊber as a guide to the amount of sulfonic-acid-type sulfur present.

[0026] When an actual bulk elemental analysis of azophenyl sulfonate-derivatized PR24 fiber is carried out, one obtains a total sulfur content of .64wt.%, and a nitrogen content of .73wt.%. Since we know the ratio of N:S in azophenyl sulfonate is 2:1, it follows that the sulfonic acid type sulfur is V4(.73 wt.%) = .365wt.%. This value signifies a high surface content of sulfonic acid; it approaches the maximum theoretical value calculated above for monolayer surface coverage (.47wt.%). We recognize that this bulk chemical analysis provides no proof of the chemical state of the sulfur or the uniformity of its dispersion on the surface; but it provides evidence that supports the XPS surface analysis data showing a high surface coverage of sulfonic acid.

[0027] Electrochemical data also supports the claim of sulfonic acid groups on the carbon surface. When an electrode consisting of sulfonated PR24 is examined with cyclic voltammetry in an aqueous electrolyte, a pair of current peaks is observed that correspond to the reversible conversion of sulfonic into sulfinic acid (SO3H/SO2H). The voltage of these current peaks corresponds to this specific electrochemical reaction. This cyclic voltammogram is shown in Figure 4.

[0028] An unusually high surface coverage of sulfonic acid groups gives nanofiber in certain embodiments proton conduction properties not found in other carbons. The phenomenon of proton conduction occurs in materials that have sufficient numbers of certain surface groups so that an environment is created where protons can 'hop' from one loosely-bonded association to another. In the most widely-known example of a proton conductor (a highly sulfonated DuPont polymer with the trade name Nafion), protons can 'hop' between nearby sulfonic acid groups that are attached to the polymer. A similar phenomenon can occur on a carbon surface that hosts sulfonic acid groups, if there are enough sulfonic acid or sulfonate groups that are sufficiently close to one another.

[0029] We have collected electrochemical impedance data that demonstrates proton conductivity in a nanofiber paper made according to one embodiment of this disclosure. A 10-mil thick sample of the nanofiber paper was placed between two flat nickel electrodes, and examined at a zero voltage bias with an applied AC perturbation of 5OmV over the frequency range 20,000-0.1 Hz. When the collected data is plotted on a graph comparing real vs. imaginary impedance, it displays one semi-circle at the high- frequency end of the plot. The semi-circle denotes a process involving charged particles, in this case proton conduction. When the collected data is plotted as resistance vs. frequency, a drop in resistance is observed over the same high frequency zone of the spectrum (Figure 5, plot A), reflecting a conductivity enhancement via proton conduction. If the same sample is equilibrated in a high humidity environment and measured again, a much larger resistance drop over the same frequency range is observed (Figure 5, plot B). This is expected, since a humid environment should enhance proton conduction in sulfonated nanofibers just as it does in the sulfonated polymer Nation. The humidified sample again produces a real/imaginary impedance plot with one semicircle at high frequency.

[0030] Finally, if a sample of the nano fiber paper is heated to 650⁰C in an inert atmosphere to strip the sulfonic acid groups from its surface, it exhibits no electrochemical impedance behavior indicative of proton conduction. No semi-circle is observed in a plot of real vs. imaginary impedance, and there is no resistance drop at high frequencies (Figure 7). Equilibrating the sample in a high humidity environment does not affect this result.

[0031] For one embodiment, observations of paper cohesion relative to surface sulfonic acid content (as determined by XPS and elemental analysis) suggest that the disclosed nanofϊber papers may advantageously be made when the surface concentration of sulfonic acid groups is above about 5 x 10^"3 moles/cm². For comparison, the calculated value for a complete monolayer coverage of sulfonic acid groups on a solid surface is approximately 4. x 10^"2 moles/cm²'. Surface coverages in the range of about 5 x 10^~3 to 4 xl 0^~2 moles/cm² fall within the scope of one embodiment of this invention. Measured values above the theoretical maximum (4 x 10^"2 moles/cm²) are also within this invention's scope, as the carbon surface may exhibit such apparent values by virtue of being irregular or porous.

[0032] Carbon nanofibers may be functionalized with a sulfonic acid group by a variety of procedures. U.S. Patent 6,479,030Bl of Firsich, issued Nov 12, 2002 describes a method for sulfonating carbon nanofibrils and the use of a sulfonated nanofibril as an electrode material. Neither the sulfonic acid content nor the surface coverage is specified.

[0033] In the j ournal publication Advanced Functional Materials 13 (5) of May 2003 , p.365, Amma et al present a synthesis for attaching an azobenzene sulfonate to a carbon surface. This article describes a 300 nm "nanowire" (solid filament of carbon) that was treated to functionalize it with azobenzene sulfonate. The bulk elemental analysis of that sample showed a 1.29wt.% sulfur content, well above that calculated for a monolayer surface coverage of sulfonic acid (~0.08wt.%). The authors did not mention this discrepancy; it is possible that their 'nanowire' has porosity such that carbons within the bulk are functionalized.

[0034] In Patent Application 20040202603, submitted Oct 14, 2004, Fischer et al. discloses tubular fullerenes and fibrils comprised of graphene sheets that are substantially cylindrical and parallel to the fiber axis, and procedures for functionalizing them, including sulfonation.

[0035] The aforementioned literature pertains to the direct covalent attachment of a sulfonic acid group (or a group that contains sulfonic acid) to a carbon surface. Such procedures can be desirable, since they introduce no polymers or absorbates that can reduce the conductivity of a nonwoven nano fiber mat, reduce inter-fiber attractions, change its reactivity, or negatively impact fuel cell performance.

[0036] Other methods have been disclosed that add a sulfonated polymer to a nanofibril as a way to incorporate a sulfonic acid group. The polymer wrapping of single-wall nanorubes with polymers that contain a sulfonic acid group (such as polystyrene sulfonate) is disclosed in Patent Application 20040186220 by Smalley et al., filed Sept 23, 2004. The functionalization of a nanotube using a non- wrapping polymer is the subject of U.S. Patent 6,905,667 of Chen et al., issued June 14, 2005; here the disclosed treatment can impart water-solubility to the nanotube, depending on the polymer composition. The preparation of functionalized nanorubes is described in Fisher et al. in U.S. Patent 6,203,814, issued March 20, 2001. Fisher describes substantially cylindrical graphitic nanotubes that are surface-oxidized and then modified by either absorbing cyclic compounds or adding surface groups capable of cross-linking.

[0037] Carbon nanotubes/nanofibers can be useful as components in fuel cells due to their conductive nature, their small dimensions, and their high surface areas. A porous conductive paper consisting of an entangled web of carbon nanofibers can serve as a catalyst support, a gas diffusion membrane, or both. The exact composition and structure of such a paper will impact its suitability for a particular fuel cell application. For example, a carbon paper used as a catalyst support will preferably be hydrophobic or hydrophilic depending on its application. Fuel cells that operate below 100⁰C and use hydrogen as a fuel favor a hydrophobic catalyst support because liquid water is generated during the fuel cell's operation, and hydrophobicity facilitates removal of the water so that the electrode does not "flood". On the other hand, fuel cells that use methanol as a fuel favor a hydrophilic catalyst support, since the fuel enters the cell as a methanol/water mixture that needs to wet the catalyst electrode. And finally, fuel cells that operate above 100⁰C and use hydrogen as a fuel are compatible with a carbon support whose surface imparts proton conductivity in a dry atmosphere at this high temperature range. In general, a fuel cell's performance is sensitive to a catalyst support's surface chemistry, conductivity, and morphology. In one embodiment, the as-produced sulfonic-acid- substituted nanofiber paper is hydrophilic and proton conducting; it may be post-treated by processes such as heating in a non-oxidizing atmosphere or vacuum to render it hydrophobic.

[0038] One embodiment of this invention provides a papermaking method for producing a type of free-standing sheet material composed of multi-walled sulfonated nanofibers. In one embodiment of the invention, nanofibers having structures that support a high degree of surface functionalization with sulfonic acid are used. In one embodiment, the disclosed sheet may be advantageously formed without the use of binders or surfactants. In this embodiment, the sheet derives its cohesive strength from the inter-fiber attractions of polar sulfonic-acid-derivatized surfaces. The sheet is also advantageous because it exhibits proton-conduction properties. Preliminary data included below shows that the sheet is useful in a fuel cell application. [0039] In this disclosure, a 'nanotube' denotes a fύamental form of carbon consisting of either single- or multiple walls of graphene sheets; the filament has a general tubular structure with a hollow core, and a diameter under lOnm. A nano fiber is larger in dimension: it is a multi-wall filamental structure similarly made up of graphene sheets with a hollow core and a diameter between lOnm and lOOOnm.

[0040] In accordance with one embodiment, the nanofibers are multiwall nanofibers made from graphene planes that are oriented at an angle between about 9 and 60 degrees off the axis of the nanofiber, where the nanofibers have a diameter between about 10 and 300 nm and a length between about 10 and 500 microns, and also to nanofibers of this type that have a coating of pyrolytic carbon. Nanofibers with this off-axis geometry preferably consist essentially of a 'stacked cup' geometry or a 'cone helix' geometry with a hollow core. These two nanofiber configurations are depicted in Figure 2; their similar structures may be indistinguishable by Transmission Electron Microscopy, although conceptually the cone helix consists of a single ribbon of graphene sheet wrapped in a tight spiral, while the stacked cone is an assembly of individual graphene cone sections. The invention also includes 'stacked-cup' or 'cone-helix' nanofibers with a finite level of pyrolitic or turbostratic carbon coating.

[0041] The selection of the nanofiber starting material is a key element of one embodiment of the disclosed papermaking process. By using nanofibers whose fiber surface is populated by carbon edge planes, a nanofiber with a relatively high surface concentration (above 5 x 10^"3moles/cm²) of sulfonic acid groups can be prepared.

[0042] The type of sulfonated nanofibers described herein can be highly dispersed in low-ionic strength water with vigorous agitation, stirring, or ultrasonication. The optimum pH for such a dispersion can be determined through a series of "Zeta potential" tests, a common tool used in colloid science or papermaking. One version of this analysis uses electrophoretic light scattering and Laser Doppler Velocimetry (LDV) to determine nanoparticle velocities within a distilled water suspension. These velocities are related to the so-called Zeta potential: the greater the Zeta potential, the greater the repulsion between particulates in the liquid medium, and the higher the likelihood of achieving a stable suspension of well-separated particulates.

[0043] A series of zeta potential tests on sulfonated ~70 nm nanofibers having the "stacked-cup" or "spiral ribbon" morphology have been performed (using a starting material known as Pyrograf III- PR24, made by Applied Sciences of Cedarville Ohio). These tests have determined that a pH within a range of about 6-9 produces a suitable dispersion in distilled water. This result is consistent with our observation that a nanofiber sheet formed by filtering a dispersion of sulfonated PR-24 shows the most tensile strength if the pH values are between 5 and 10.

[0044] We find that pH affects the density of the PR24 paper that is formed. In one embodiment, concentrations are below 1 gram per liter to prevent agglomeration prior to filtration. Increased ionic strength has been observed to reduce the solubility of the dispersed nanofibers (for example, if salt is added to a distilled water suspension of PR- 24, the PR-24 precipitates from solution). Aqueous dispersions with solution conductivities below about 50 microSiemens/cm are preferred in one embodiment of this invention.

[0045] The dispersion made in accordance with one embodiment will pass though a fine-mesh screen (as fine as 90 micron), demonstrating that agglomerates have been broken up. Filter media with much smaller pore sizes (e.g., between 0.5 and 5 micron ) may be used for capturing the dispersed nanotubes/nanofibers and forming them into a sheet. Filter membranes with pore sizes above this range allow an ever-increasing fraction of the dispersion to pass through. Pore sizes below this range may also be used, but the rate of water filtration in such membranes is slower, so that the sheet may take longer to form.

[0046] The origin of the paper's strength can be demonstrated by examining a series of similar nanofibers that have different amounts of surface edge planes, and therefore, different levels of sulfonic acid groups on their surface after sulfonation. . .

[0047] The ~70-nm diameter nanofiber PR24 has a structure that can be interpreted as either a "stacked cup" or a "cone-helix". It is known that its surface has high concentrations of edge planes along the fiber length. The amount of edge sites can be reduced by partially- or completely pre-graphitizing this fiber through heat treatment. Heat treatment gradually eliminates edge surfaces by bonding nearby edges together. With fewer edge sites, fewer sulfonic acid groups may be placed on the fiber surface. The claim that heat treatment causes a reduction in surface bonding sites is supported by data originating from the manufacturer of PR-24 (Applied Sciences, Inc., Cedarville Ohio). Applied Sciences has a process for functionalizing their fiber surface with amines or amides. When this process is applied to PR-24, XPS analysis detects a range of surface nitrogen between 2-3 atom%, but when the same process is applied to a partially- graphitized specimen (treated to about 1400 to 2000°C), the measured surface nitrogen is only~0.6 atom%.

[0048] A Transmission Electron Microscope image of a fully-graphitized (heat treated to 3000 C) PR-24 nanofiber is shown in Figure 8. The 'stacked-cup' or cone-helix' graphite plane geometry is evident, as is the hollow central core. A close examination of the image shows that at the surface the carbon planes have folded over and bonded to one another. This exemplifies the way in which surface bonding sites are reduced through heat-treatment. By contrast, in a partially graphitized fiber, the fiber walls remain crystalline.

[0049] If a partially-graphitized PR24 sample is sulfonated, dispersed, and formed into a paper, the paper has less than one-third the tensile strength of a sheet made from PR24. And if a completely graphitized PR24 is put through the same process, the filtered mat does not hold together at all.

[0050] The above examples suggest that it is not fiber entanglement, but rather the concentration of surface sulfonic acid groups that leads to strength in the disclosed sheet material. Therefore, nanofiber structures that can host high surface concentrations of sulfonic acid groups are used in this embodiment of the invention. [0051] It should be apparent that some of the weaker sheets formed above are within the scope of this invention, because these weak sheets can be made thicker in order to generate enough tensile strength for practical handling. A range of sheet strengths may be obtained depending on the amount of sulfonic acid groups on the nanofiber surface, and on the diameter of the nanofiber. Based on our observations of handling strength and corresponding analyses by XPS and quantitative analysis, in one embodiment, a range of about 5 xlO^"3 to 4 xlO^"2 moles of sulfonic acid per square centimeter of nanofiber surface leads to a useful freestanding sheet when nano fibers between 10-300nm in diameter are used. If other structural factors are equal, the smaller diameter nanofibers in this range will yield a sheet with more tensile strength, as they pack more closely in a nonwoven mat and have greater numbers of inter-fiber contacts. The fiber surface areas (cm²) referred to herein are geometric surface areas calculated using the diameter of the fiber as contrasted with surface areas determined by other methods.

[0052] Another property associated with sulfonated nanofiber in one embodiment is its ability to form a sheet by filtration that will later release cleanly from its filter backing. The source of this advantageous property is the large difference in surface polarity between a non-woven web of sulfonated nanofibers having the preferred structure and a non-polar filter membrane such as polyethylene, polypropylene, or teflon. Nanofibers with polar sulfonated surfaces have more attraction for each other than they do for a non- polar filter surface. This gives a filtered mat of the nanofibers sufficient internal cohesion, and sufficiently low attraction for the filter membrane, that it cleanly releases from the filter membrane after drying.

[0053] To summarize the above points, we disclose a method for forming a carbon nanofiber paper or sheet material. One embodiment of the invention includes the use of multi-wall nanofibers that can achieve a high level of sulfonic acid groups over their surface. Another embodiment of the invention involves the use of low-concentration, low-ionic-strength aqueous dispersions with controlled pH for creating a papermaking feedstock solution of these nanofibers. Another embodiment of the invention involves filtration of this dispersion onto a non-polar membrane to form a sheet that will release from the filter membrane. In accordance with another embodiment of the invention, no binders, surfactants, or cross-linking strategies are required to produce a freestanding sheet material. The invention also includes the nanofϊber papers produced from these multiwall sulfonated nanofibers. Still another embodiment of the invention is a nanofϊber paper with proton conduction properties.

[0054] The processes that may be used to manufacture the sulfonated nanofibers used in this invention consist of syntheses that transform the as-produced nanofibers into highly sulfonated nanofibers. Such syntheses include those specified in U.S. Patent 6,479,030Bl of Firsich, issued Nov 12, 2002, which describes a method for sulfonating carbon nanofibrils, and the synthesis described in the journal publication Advanced Functional Materials 13(5) of May 2003, p.365, where Amma et al. presents a synthesis for attaching an azobenzene sulfonate to a carbon surface. A third option is to use the carbon sulfonation procedure described in Patent Application 20040202603, submitted Oct 14, 2004 by Fischer et al. An applicable sulfonation synthesis will attach a sulfonic acid or sulfonate group to the nano fiber through covalent bonds at a surface level of over 5 x 10^" moles/cm , such that the resulting nanofiber is suitable for nanofϊber papermaking according to the disclosed procedure.

Utilization in Fuel Cells

[0055] The hydrophilicity and proton conduction of the sulfonated nanofϊber paper make it particularly well-suited for a number of applications in fuel cells. In the descriptions below, the tradename 'Nafϊon' refers to a product of the E.I. DuPont company (Nafϊon is a copolymer of tetrafluoroethylene and a perfluorosulfonic acid polymer).

As a Gas Diffusion Layer and Catalyst Support in a Methanol Fuel Cell

[0056] The data shown below is from a methanol fuel cell, where fuel is fed into the fuel cell as a methanol/water mixture.

[0057] Three different anode configurations were tested in order to compare the effect of support material on methanol oxidation current. Each of the tests below used the standard 50:50 PtRu catalyst (Johnson Matthey). In all cases the catalyst was deposited by air spraying a solution of 12wt% Nafion ionomer and 80mg of PtRu suspended in 600μL of isopropanol. The different configurations are illustrated in Fig. 10.

[0058] Type A samples (corresponding to the traditional DMFC Nuvant standard [1-3]) were produced by spraying the Toray CP with a carbon black (Alfa Aesar)/Nafion ionomer solution, followed by drying, and subsequent deposition of the PtRu catalyst (4mg/cm²).

[0059] In Type B samples the Toray CP was sprayed with a solution of single-wall carbon nanotubes (Carbon Nanotechnologies, Inc.) and Nafion ionomer (the nanotubes serving as an alternative to carbon black as a pore-filler and catalyst-support material), followed by drying, and subsequent deposition of the PtRu catalyst (.9mg/cm²).

[0060] For Type C samples the PtRu catalyst was directly deposited onto the nanofiber paper (1.1 mg/cm²).

[0061] Analysis by Scanning Electron Microscopy showed that in each case the surfaces were fairly uniform, and the catalyst well-dispersed. These samples were simultaneously evaluated for electrochemical performance towards MeOH oxidation in the commercial Nuvant test cell; a multi-array instrument that has been shown to yield data that correlate well with real fuel cell performance.

[0062] Samples prepared as described above were analyzed simultaneously in triplicates in order to determine their activity toward methanol oxidation. The test was run under half cell condition in which the anode array had 0.5M Methanol flowing over the backside of each anode at 8mL/min. The cathode had humid hydrogen gas at ~180sccm. Figure 7 shows Linear Sweep Voltammerry (LSV), at scan conditions of 2mV/sec over the voltage range 0 to 1.0V versus Dynamic Hydrogen Electrode (DHE). The data has been normalized for differences in catalyst content.

[0063] It is seen in Figure 7 that the nanofiber electrode used in this example produces 16 times the current compared to the standard catalyzed Toray paper. It also performs four times as well as the single-wall-nanotube-modified Toray paper at about the same catalyst content. This data has been reproduced in triplicate, and its standard deviation falls within 10% of the mean.

In a Fuel Cell Cathode for Oxygen Reduction

[0064] During the operation of a fuel cell there is water generated at the cathode. Fuel cell cathodes are typically treated with a water repellant such as teflon to help wick away liquid water; if the water builds up, gaseous oxygen is prevented from reaching the cathode, dramatically reducing the fuel cell's output. This phenomenon is referred to as

'flooding'.

[0065] Sulfonated nanofibers are very hydrophilic, and so a fuel cell electrode (a gas diffusion layer) made mostly of these nanofibers floods when used as a cathode. But if the electrode is constructed by placing a very thin layer of sulfonated nanofibers on top of a water-repelling gas diffusion layer, the resulting electrode does not flood when used as a cathode. Further, this configuration can produce outstanding enhancements in fuel cell power and catalyst utilization, as shown in the data to follow.

[0066] There are many commercially available water-repelling gas diffusion layer products that are suitable substrates for implementing this strategy. These substrates consist of woven or non- woven carbon fibers covered with a microporous layer of carbon powder. When a dilute papermaking solution of sulfonated nanofibers is filtered through these substrates, the microporous surface layer catches the suspended nanofibers to form a thin nanofiber layer on the surface. If the substrate's surface is irregular, the nanofibers will tend to fill in the cracks or defects in the microporous layer. It is not necessary to have a complete surface coverage of sulfonated nanofibers to obtain enhancement in cathodic fuel cell performance, and too thick a layer will cause flooding. The optimal thickness of the applied sulfonated nanofiber layer will vary depending on the substrate's morphology.

[0067] An example of a suitable substrate for this process is the product ELAT 2500 made by PEMEAS Corp of Somerset, New Jersey. It consists of a woven carbon cloth with a microporous carbon powder coating on one side. The data in Figure 9 compares unmodified ELAT 2500 to the same material modified with a thin layer of sulfonated nanofϊbers using the papermaking process of this disclosure. The nanofiber layer thickness is approximated at 0.0001 inch based on the weight added, although defects in the microporous surface will make the actual thickness lower. The modified material has a visibly different color, a different shade of grayish-black.

[0068] Catalyst was applied to the modified and unmodified ELAT materials in the same way (sprayed-on), and the resulting samples were tested simultaneously in a common cell against a common anode and hydrogen fuel. The testing was done in triplicate, and the data was normalized to account for small differences in catalyst content. It is seen that at 40 degrees C, the nanofiber-modified ELAT provides three times the power output at the same catalyst content, i.e., three times the catalyst utilization. Another comparison at 6O⁰C showed about 2.4 times the power output.

As a Proton Conducting Separator Membrane

[0069] A central component of the fuel cells that burn hydrogen or methanol is a proton conducting membrane. The membrane separates the two halves of the fuel cell (anode and cathode), and its purpose is to be a proton conductor and an electrical insulator. It is generally made of a polymer (such as Nafion) that has sulfonic acid groups attached to a polymer chain. Protons migrate though such a membrane via a "hopping" mechanism, where they "hop" from one sulfonic acid group to another, frequently aided by local water molecules. Nafion performs well in this application below 100⁰C, but performs poorly above this temperature because it dries and cracks.

[0070] An active area of research in fuel cells is to create a fuel cell separator membrane that operates above 100⁰C. Such a membrane will allow the development of a hydrogen-burning fuel cell that operates between 100-180⁰C, a device that has clear advantages over counterparts that operate at around 8O⁰C. They are less susceptible to poisoning by contaminated fuel, but most importantly, they have the potential to operate without humidification. This reduces the fuel cell's complexity, and it eliminates the need to furnish a source of water (something that would add weight and require replenishing in an automobile, for example). [0071] The disclosed nanofiber paper has both proton and electrical conductivity; however, electrical conductivity is not desirable for a separator membrane. The nanofiber paper's electrical conductivity may be sharply reduced by incorporating polymeric material (examples are discussed below) onto or into the nanofiber sheet. This can be done in a number of ways, including melt-pressing or laminating a polymer onto the nanofiber sheet, applying a solution of polymer to the paper followed by solvent drying, electrostatic spraying, or incorporating the polymer during the preparation of the nanofiber paper followed by a heat-treatment to melt the polymer into a solid mass. To produce a useful separator membrane, the incorporated polymer(s) need to fill or coat the porous nanofiber paper such that the paper becomes gas-impermeable, as a fuel cell separator membrane acts to block gas transport between the two halves of a fuel cell.

[0072] Preferentially, the added polymer would either exhibit proton conduction itself or otherwise enhance the proton conduction of the sulfonated nanofiber paper. Examples of polymers that would provide this benefit can include Nafion, sulfonated polyphenylene, polyvinyl alcohol, polybenzimidazole, polyethyleneimine, polyethylene oxide, and related polymers that contain -SO₃H, -OH, -O-, -NH-, or -S- groups. Alternatively, the polymer can serve no function except to make the paper gas- impermeable and electrically non-conductive. Another embodiment of this invention is a sulfonated nanofiber paper that is coated or infiltrated with a polymer. Another embodiment of the invention is a proton-conducting, electrically non-conducting membrane made from a sulfonated nanofiber paper and a polymer. Another embodiment of the invention is a gas-impermeable, electrically non-conductive, proton-conducting membrane. Another is a gas-impermeable, electrically non-conductive, proton conducting membrane that is suitable for use in a fuel cell that operates between 100^° and 180^°C.

As a Cathode in a Hydrogen Fuel Cell that Operates above 100^°C. [0073] In the common type of hydrogen fuel cells that operate at approximately 80^°C, the cathode is preferably relatively hydrophobic, so that the liquid water generated by the fuel cell's operation can be easily swept away from this porous electrode. However, in a fuel cell that operates in the range of 100^°-180^°C, the cathode can be hydrophilic, as there is no concern that it will 'flood' with liquid water at temperatures above water's boiling point. Thus the hydrophilic sulfonated nanofiber paper is suitable for use as a cathode in this temperature range, and it is furthermore anticipated that a sulfonated nanofiber paper will enhance catalyst utilization in^' this application because of its proton conduction properties.

A Bi-layer Nanofiber Paper as an Anode for Tolerance to Reformate Fuel [0074] Reformate fuel is a hydrogen rich fuel derived from hydrocarbons. It can contain some carbon monoxide, which poisons most fuel cell catalysts, especially at operational temperatures below 100°C. One way to mitigate the poisoning problem is to use two different catalysts that are separated from one another within a single anode electrode. The first catalyst that the gas stream encounters is designed to convert CO to CO₂, and the second to serve the normal function of oxidizing hydrogen. Thus the first catalyst acts as a kind of scavenger to purify the fuel. In this scenario the anode is made in layers so that the incoming gas stream can first encounter one catalyst before reaching with the second.

[0075] The papermaking process disclosed is suitable for constructing such a fuel cell anode. Once the main nanofiber paper layer is formed, one catalyst can be applied. Then a second carbon layer can be formed on top before applying the second catalyst. The top carbon layer can consist of the same type of sulfonated nanofibers as the bottom layer, but this is not a requirement. Other types of nanofibers, nanotubes, or carbon powders can comprise the second layer. The main nanofiber paper layer can serve as an effective filter and a support, so that a carbon that lacks the cohesion to be made into a paper by itself (such as a graphitized concentric-wall nanotube) can be formed as the top layer.

As a Novel Type of Fuel Cell Catalyst Support

[0076] The surface coverage of sulfonic acid groups on sulfonated nanofiber paper gives it a high affinity for the sulfonated membranes that often separate the two halves of a fuel cell. An example of such a membrane is Nafion, a trade name of the E.I. DuPont company; this is a copolymer of a perfluorosulfonic acid polymer and tetrafluoroethylene. Thus when the sulfonated nanofiber paper is placed in contact with the Nafϊon membrane, it sticks readily and bonds well. This compatibility and preferential interface is expected to promote good performance in a fuel cell, and also in a fuel cell that is operated in reverse (i.e., creating hydrogen by electrolyzing water).

[0077] Another embodiment of the invention is a laminate composite of the aforesaid nanofiber paper and a sulfonated proton-conducting polymer membrane.

[0078] In accordance with another embodiment of the invention, the sulfonated nanofiber paper is applied to a porous matrix that may include a foam, felt, mesh, or membrane materials.

[0079] Another embodiment is a sulfonated nanofiber paper made with one or more of the polymeric binder discussed above, whose purpose is to enhance paper strength without eliminating the porosity of the paper. This is a distinct purpose from the earlier- discussed embodiment of filling the paper with a polymer to make it impermeable to gas. A polymer which serves as a binder may be present in quantities up to 15wt.%. It should be clear that some added polymers can potentially serve two functions at once, for example, serving as a binder and also an agent to enhance ion conduction.

[0080] Another embodiment is the heat treatment of the formed sulfonated nanofiber sheet to remove sulfonic acid groups and render the sheet hydrophobic. This may be accomplished by heating the material above about 600⁰C for 1 hour in vacuum, inert atmosphere, or a hydrogen-containing atmosphere.

[0081] Another embodiment is a sulfonated nanofiber paper made with incorporated powders or non-carbon fibers. These are made by adding a dispersable powder or fiber to the aqueous papermaking solution prior to forming the sheet. One example of this embodiment is a paper made with incorporated silicon powder (e.g., less than about 200μ); such a paper has an application as a lithium ion battery anode. Another example is a paper made with incorporated nickel nanofibrils; such a paper would be useful for electromagnetic shielding. Depending on their identity and size, such incorporated powders or non-carbon fibers can constitute up to about 70% of the paper weight, although paper strength considerations generally limit this to about 50%. [0082] Another embodiment is the use of the paper as a generic catalyst support for other applications such as water electrolysis, ammonia electrolysis, oxidation of gases, or photocatalysis.

[0083] This invention is illustrated in more detail in the non-limiting examples that follow.

EXAMPLE 1- Nanofiber paper from azophenyl sulfonate derivatized nanofiber

[0084] A sample of PR-24, a multiwall nanofiber with a 'stacked-cup' or cone-helix morphology and an average diameter of 70nm was obtained from Applied Sciences, Cedarville Ohio. A 5 g PR-24 sample was first heated with stirring in 500 ml of dilute aqueous hydrochloric acid (6M) at 60⁰C for 12 hours to leach out impurities. The nanofiber sample was then filtered out and rinsed with distilled water.

[0085] The filtered nanofiber sample was then derivatized with azophenyl sulfonate by using a synthetic procedure similar to that of Amrna et al. in Advanced Functional Materials 13(5) of May 2003, p.365. A solution of 10 g NaNO₂ in 850ml water was prepared, and then 2Og of the sodium salt hydrate of sulfanilic acid was added. Next 20ml of concentrated hydrochloric acid was added, and then finally the nanofiber sample. The solution was stirred for 24 hours, then filtered and washed four times with distilled water, until the conductivity of the filtrate was under 50 microSiemens/cm.

[0086] A 2 g portion of the sample prepared above was dispersed in 3 liters of distilled water with vigorous stirring. The solution's pH was 6.6, and the solution's conductivity was a low 6 microSiemens/cm. After stirring for one half hour, the solution was filtered onto a 11 "xl 1" filter plate covered with a polypropylene membrane having a 0.4 micron pore size (Millipore brand HTTP membrane). After air-drying, the formed nanofiber paper (7 mils thick) was recovered by peeling off the filter membrane, which released cleanly from the nanofiber paper.

EXAMPLE 2 Nanofiber paper made on a carbon mesh reinforcing support [0087] A sample of azophenyl sulfonate derivatized PR-24 was prepared as described in EXAMPLE 1. A 3 gram sample was dispersed in 3 liters of water with stirring. The pH of the solution was initially 5.8; it was adjusted to 7.2 through small additions of a dilute sodium hydroxide solution. After pH adjustment, the conductivity of the solution was 10 microSiemens/cm. The solution was stirred for one-half hour prior to paper preparation.

[0088] A 11 "x 11 " filter plate was first covered with a nonwoven polypropylene/polyethylene filter membrane (SC 18 from Crane Nonwovens, Dalton, MA), and the a 2-mil thick nonwoven web of conventional carbon fibers (Hollingsworth and Vose, West Groton, MA) was placed on top of the filter membrane. The nano fiber suspension was filtered onto this web, forming a nanofiber paper with a carbon mesh reinforcement. After air-drying, the reinforced nanofiber paper was recovered by peeling off the polypropylene membrane. The resulting nanofiber paper was 10 mils thick, and the mesh reinforcement represented about 10% of the total paper weight. This reinforced nanofiber paper exhibits more tensile strength and better conductivity than the plain nanofiber paper.

EXAMPLE 3 Nanofiber paper from sulfonated PR-24 nanofiber made by a different procedure.

[0089] A 5 g PR-24 sample was first heated with stirring in 500 ml of dilute aqueous hydrochloric acid (6M) at 60⁰C for 12 hours to leach out impurities. The nanofiber sample was then filtered out, rinsed with distilled water, and air dried. It was then placed in a quartz tube furnace, and a stream of hydrogen gas (100 cc/min) was passed over the sample as the furnace was ramped to 500⁰C. After holding at 500⁰C for 2 hours, the furnace was allowed to cool to room temperature. The sample was then transferred to a flask, where 150cc of fuming sulfuric acid (20% SO₃) was added, and the mixture heated to 130⁰C for 24 hours under argon. The flask was then transferred to an ice bath to quickly cool it down, and then the sample was filtered. The wet filter cake was allowed to stand in air overnight, during which time it absorbs a good deal of water from the air because it is moist with hygroscopic sulfuric acid. The next morning the filter cake was suctioned further to remove this liquid, and then the filter cake was gradually added to a distilled ice/water mixture with stirring. After filtering the icewater away, the nanofibers were further washed with distilled water to remove traces of acid.

[0090] 2 grams of this sulfonic acid derivatized nano fiber was dispersed in 3 liters of distilled water with vigorous stirring. The solution's pH was found to be 6.4, and its conductivity was low (3 microSiemens/cm). After stirring for one half hour, the solution was filtered onto a 11 "x 11" filter plate covered with a polypropylene membrane having a 0.4 micron pore size (Millipore brand HTTP membrane). After air-drying, the formed nano fiber paper (7 mils thick) was recovered by peeling off the filter membrane, which released cleanly from the nanofiber paper.

EXAMPLE 4 Nanofiber paper made from PR-19, a nanofiber with a thick pyrolytic carbon coating

[0091] A sample of PR-19 nanofiber was obtained from Applied Sciences, Cedarville Ohio. PR-19 consists of a PR-24 nanofiber coated with a thick layer of pyrolytic carbon. Its diameter is in a range around 200 nm, well above the ~70 nm diameter of PR-24. A 5 gram sample of PR-19 was washed with dilute hydrochloric acid and derivatized with azophenyl sulfonate as described in EXAMPLE 1.

[0092] 2.5 grams of the azophenyl sulfonate derivatized PR-19 were dispersed in 4 liters of distilled water with vigorous stirring. The solution pH was 6.7 and the ionic conductivity was 2 micro Siemens/cm. After one half hour of stirring, the dispersion was passed through a 150 micron screen to remove some poorly-dispersed nanofibers, and then the solution was filtered onto a l l"xl 1" filter plate covered with a polypropylene membrane (Celgard 3501, from the Hoechst Celanese Corporation). After the nanofiber sheet was filtered, it was washed with a liter of distilled water before allowing it to air- dry overnight. The formed nanofiber sheet (12 mils thick) was recovered by peeling off the Celgard 3501 filter membrane, which fell away with no adhesion to the nanofiber sheet.

What is claimed is: Having described this invention in detail it will be apparent that numerous variations and modifications are possible without departing from the embodiments and manifestations of the invention as described in the following claims.

Claims

1. A carbon nano fiber paper comprising a web of sulfonated multi-walled nanofibers.

2. The paper of claim 1 wherein the nanofibers have a stacked cup or cone helix structure.

3. The paper of claim 1 wherein the nanofibers have a diameter of about 10 to 500 nm.

4. The paper of claim 3 wherein the nanofibers carry sulfonic acid groups in an amount of at least about 5.3 x 10^"3 moles/cm².

5. The paper of claim 4 wherein the nanofibers carry a coating of pyrolyzed carbon.

6. The paper of claim 1 wherein the paper further comprises a reinforcing porous substrate having a layer of the nanofibers on the surface thereof.

7. The paper of claim 1 wherein the paper is infiltrated with a polymer.

8. The paper of claim 7 wherein the polymer enhances the paper's proton- conducting properties.

9. The paper of claim 7 wherein the polymer is selected from the group consisting of Nafion, sulfonated polyphenylene, polyvinyl alcohol, polybenzimidazole, polyethyleneimine, polyethylene oxide, or the like.

10. The paper of claim 1 wherein the paper is proton-conducting.

11. The paper of claim 1 wherein the paper is gas-permeable.

12. The paper of claim 1 wherein the paper is electrically conductive.

13. The paper of claim 12 wherein the paper is rendered electrically non-conductive through the incorporation of an electrically conductive polymer.

14. The paper of claim 12 wherein the paper is rendered gas-impermeable through the incorporation of a polymer.

15. The paper of claim 1 wherein the nanofibers contain graphene planes oriented at an angle of about 9 to 60° with respect to the longitudinal axis of the nanofiber.

16. The paper of claim 1 wherein the paper is free of surfactants and binders.

17. The paper of claim 1 wherein the paper further includes non-sulfonated nanotubes.

18. The paper of claim 1 wherein the paper further includes organic or inorganic powers or fibers.

19. A fuel cell membrane including the carbon nanofiber paper of claim 1.

20. The membrane of claim 19 wherein the membrane is a fuel cell gas diffusion layer or gas diffusion layer and catalyst support.

21. The membrane of claim 19 wherein the membrane is a fuel cell gas diffusion layer that includes a layer of the sulfonated multi-walled nanofibers having a diameter of about 10 to 500 nm and at least about 5.3 x 10^"3 moles/cm² sulfonic acid groups.

22. The membrane of claim 19 wherein the membrane is a catalyst support that includes a catalyst on a layer of sulfonated multi-wall nanofibers.

23. The membrane of claim 19 wherein the membrane includes a layer of sulfonated multi-walled nanofibers on a water-repelling gas diffusion layer.

24. The membrane of claim 23 wherein the water-repelling gas diffusion layer is a layer of a woven carbon cloth containing microporous carbon powder.

25. The membrane of claim 22 wherein the nanofibers are rendered hydrophobic.

26. A fuel cell including a membrane formed from the carbon nanofiber paper of claim 1.

7. The fuel cell of claim 26 wherein the nanofibers form a gas diffusion layer.