Method for purifying and functionalizing carbon nanomaterials

Abstract

Methods for removing metal nanoparticles from a carbon nanomaterial and methods for increasing nitrogen doping level in a carbon nanomaterial and related carbon nanomaterials are described. In examples, the methods comprise pyrolyzing a mixture comprising a carbon nanomaterial, a fluorine-containing polymer, and a nitrogen-containing compound at a temperature sufficient to remove metal nanoparticles from the carbon nanomaterial to provide a carbon nanomaterial depleted in metal nanoparticles compared to the carbon nanomaterial prior to pyrolysis and to increase nitrogen doping level in the carbon nanomaterial to provide a carbon nanomaterial with increased nitrogen doping level compared to the carbon nanomaterial prior to pyrolysis.

Description

METHOD FOR PURIFYING AND FUNCTIONALIZING CARBONNANOMATERIALSCROSS-REFERENCE TO RELATED APPLICATION

[0001] This application claims the benefit of and priority to U.S. Provisional Patent Application No. 63 / 553,858, filed February 15, 2024, and the contents of which are incorporated by reference herein in their entirety.BACKGROUND

[0002] Carbon nanotubes (CNTs) have garnered considerable attention in both fundamental and applied research since their initial discovery 35 years ago. Both singlewalled (SWCNTs) and multi-walled CNTs (MWCNTs) have found their intrinsic properties exploited in an array of applications including, but not limited to, energy conversion and storage, electronics, and in composite materials. The dependance of these applications on the stark performance improvements made upon introducing CNTs to the host matrix - as a result of their mechano-chemical and thermoelectric properties - created an exponentially increasing market for manufacturing CNTs. Many synthetic approaches have been introduced but suffer from slow processing time, poor yields and controls on the resultant quality of CNTs, feed reagent quality requirements, post-processing costs for removal of impurities, and high energy costs. Three routes, namely, arc-discharge, laser ablation, and chemical vapor deposition (CVD), have been increasingly used to reproducibly synthesize CNTs. Although nanoparticle catalysts are not needed for growth, they are typically used to accelerate the reduction of carbon-containing feed gas and lower the temperature for the growth of CNTs. Metal catalysts such as Ni, Fe, Co, and Mo, or their mixtures are commonly embedded in alumina (AI2O3), titania (TiCh), or magnesia (MgO) inert supports. Depending on the synthetic conditions (environment, catalyst quality, synthesis protocol), the CNT purity of the as-prepared sample can vary between 30-99 wt.% for MW CNT, with metal impurities from the seed catalyst being in the 0.1-10 wt.%. The majority of metal impurities maintain their position within the core of MWCNT as nanoparticles - effectively being embedded atop layers of chemically resilient sp2carbon. Further, it is seldom when these impurity nanoparticles are only located at the base or tips of CNTs but, instead, they are also embedded at varying depths within their walls, making it nearly impossible to remove them without compromising the nanotube structure.This makes their removal extremely difficult and costly, but necessary for many existing and emerging applications in electronics, water treatment, and composite materials, which mandate consistent and reproducible behavior. To that end, a plethora of work has been dedicated to the purification and removal of metal nanoparticle impurities embedded within MWCNT.

[0003] Both gas-phase and liquid-phase methods have been reported to purify CNTs. The conventional methods of removing impurity nanoparticles from CNTs usually damage the graphitic shell that encapsulates the metallic impurity, followed by etching it from within its enclosure. Chemical purification of CNTs is categorized into gas-phase (such as dry oxidation), liquid-phase (like acid etching), and electrochemical oxidation, while physical purification relies on gas-phase techniques such as high-temperature annealing (> 2000 °C in an inert environment in the absence of chemical reagents). While chemical agents can enable one-step purification, this approach is more destructive to the CNT surface and structure compared to the sequential multi-step processes typically employed. Gas-phase purification generally involves oxidizing agents like air, oxygen, ozone, steam, carbon dioxide, or nitrogen dioxide to oxidize carbon, allowing the subsequent removal of metallic impurities via acid leaching. Tradeoffs typically exist between the needed reaction temperature (400 - 1000 °C), the duration of treatment, and the degree of surface damage to the CNT depending on the oxidative strength of the oxidizing gas used. Alternatively, carbon can be gasified into methane using reducing gases like hydrogen (H2; T > 700 °C) or ammonia (NH3; T > 900 °C). Another method involves metal oxidation and volatilization through halogenation (using agents like chlorine or halogenated hydrocarbons, CHCI3 and CCh) without involving carbon gasification. This oxidation occurs because halogens oxidize metals and metalloids more readily than carbon. However, high chlorination temperatures (> 1000 °C), or a subsequent hydrogenation (~ 600 °C), are needed to fully eliminate residual chlorine that is typically bound to carbon surface at lower reaction temperatures. As can be noted from the above-mentioned non- exhaustive summary of purification techniques, both liquid-phase and gas-phase purification strategies tend to require multiple steps and high solvent or energy (temperature) costs - practical limitations due to the corrosive nature of the gas / solvent used - and tend to yield damaged surfaces due to the harsh chemical environment to which the entire CNT sample is exposed.

[0004] One of the primary application fields that have become highly dependent on CNT utilization is electrochemical energy generation, utilization, and storage. Primarily due to the chemical inertness, conductivity, functionalizable potential, and high surface area of CNTs, an array of electrochemical reactions - specifically cathodic reactions - have had their performance metrics improved through the inclusion of CNTs as a support, dopant, or catalytic ingredient within their respective electrocatalysts.

[0005] In particular, the electrochemical CO2 reduction reaction (eCChRR) has witnessed blooming attention in recent years due to its potentially feasible alignment with CO2 capture and utilization (CCU). Of the many reaction products of eCO2RR that have been explored, Cl products - namely, formate / formic acid and carbon monoxide (CO) - have shown technoeconomic viability. While the benchmark CO-producing catalysts are primarily based on noble metals like Au and Ag, various other catalyst design strategies have been investigated. Single atom catalysts (SACs) and dual atom catalysts (DACs), specifically in the form of metal-nitrogen doped carbon (M-N-C), have garnered attention due to their tunable functionality, control over their coordination environment, high atomic utilization and lower cost, and the high charge localization feature. CNTs have been demonstrated as a cost-effective carbon substrate for M-N-C catalysts in eCChRR.

[0006] Atomically precise nanoclusters (NCs, 1-2 nm) combine the tunable activity of single / dual atom catalysts (SACs / DACs, 0.1-0.3 nm) with the stability and anchoring advantages of nanoparticles (>3 nm), making them versatile for electrochemical applications. The main challenge with NCs lies in achieving scalable and uniform synthesis on catalytic substrates, which has primarily relied on wet-chemistry methods that involve the use of ligands. However, the removal of these ligands often disrupts their uniformity which causes aggregation and loss of performance. Further, meaningful stability of ligand- protected NCs under electrochemical operation has not been comparable to SAC / DAC or catalytic nanoparticle analogues, especially at higher current densities.SUMMARY

[0007] To address these and related challenges, the present disclosure provides methods for removing metal nanoparticles from a carbon nanomaterial, methods for increasing nitrogen doping level in a carbon nanomaterial, carbon nanomaterials having less than about 1 percent by weight metal based on the total weight of the carbon nanomaterial and greater than about 6 percent by weight nitrogen based on the total weightof the carbon nanomaterial, articles of manufacture that includes the carbon nanomaterial, catalytic electrodes, and reactors.

[0008] Accordingly, in an aspect, the present disclosure provides a method for removing metal nanoparticles from a carbon nanomaterial, comprising pyrolyzing a mixture of a carbon nanomaterial, a fluorine-containing polymer, and a nitrogen-containing compound at a temperature sufficient to remove metal nanoparticles from the carbon nanomaterial to provide a carbon nanomaterial depleted in metal nanoparticles compared to the carbon nanomaterial prior to pyrolysis.

[0009] The level of metal nanoparticle removal can be verified by TEM, EDX, and XPS.

[0010] In another aspect, the disclosure provides a method for increasing nitrogen doping level in a carbon nanomaterial, comprising pyrolyzing a mixture of a carbon nanomaterial, a fluorine-containing polymer, and a nitrogen-containing compound at a temperature sufficient to increase nitrogen doping level in the carbon nanomaterial to provide a carbon nanomaterial with increased nitrogen doping level compared to the carbon nanomaterial prior to pyrolysis.

[0011] In certain embodiments of the above methods, the nanomaterial is a carbon nanotube (SWCNT or MWCNT), a carbon nanofiber, a graphene, or a graphene nanoplatelet.

[0012] In certain embodiments, the metal nanoparticles are selected from the group consisting of iron, nickel, cobalt, chromium, manganese, and copper nanoparticles, and mixtures thereof.

[0013] In certain embodiments of the methods, the fluorine-containing polymer is a polytetrafluoroethylene (PTFE).

[0014] In certain embodiments of the methods, the nitrogen-containing compound is melamine.

[0015] In certain embodiments, temperatures sufficient to remove metal nanoparticles from the carbon nanomaterial range from greater than about 550°C to less than about 750°C. In certain embodiments, the temperature sufficient to remove metal nanoparticles from the carbon nanomaterial is about 650°C.

[0016] In certain embodiments of the methods, the carbon nanomaterial produced by the method is a non-magnetic carbon nanomaterial.

[0017] In other embodiments of the methods, the carbon nanomaterial produced by the method is doped with nitrogen to provide a nitrogen-functionalized carbon nanomaterial. In certain of these embodiments, the nitrogen-functionalized carbon nanomaterial has improved water dispersion compared to the carbon nanomaterial prior to pyrolysis. In other of these embodiments, the nitrogen-functionalized carbon nanomaterial has a nitrogen doping level at a minimum of about 6 percent by weight based on the total weight of the carbon nanomaterial. In certain embodiment of these methods, the nitrogen- functionalized carbon nanomaterial has a nitrogen doping level from about 6 to 12% by weight based on the total weight of the carbon nanomaterial in the bulk measured by EDX. It will be appreciated that the surface nitrogen doping level measured by XPS is greater than the bulk nitrogen doping level measured by EDX because most doping occurs on the nanomaterial surface.

[0018] In further embodiments of the methods, the carbon nanomaterial produced by the method is not damaged during the process, as measured by transmission electron microscopy (TEM).

[0019] The disclosure provides carbon nanomaterials prepared by the methods described herein.

[0020] In a further aspect, the disclosure provides a carbon nanomaterial having less than about 1 percent by weight metal based on the total weight of the carbon nanomaterial and greater than about 6 percent by weight nitrogen based on the total weight of the carbon nanomaterial.

[0021] In certain embodiments, the nanomaterials produced by the method have about 0.6% by weight of metal content (atomically dispersed form as opposed to nanoparticle form). It will be appreciated that the amount of metal is atomically dispersed on the surface of carbon nanomaterials, likely coordinated with nitrogen doping. For this reason, metal nanoparticles are not observed by TEM. However, atomically dispersed metal species are detected by XPS and EDX. These atomically dispersed metal centers, in coordination with nitrogen dopant, are active sites for catalytic reactions of CO2 reduction and water splitting, among other reactions.

[0022] In another aspect, the disclosure provides articles of manufacture that include the carbon nanomaterials described herein. In certain embodiments, the article is (1) an electrode for CO2 reduction, N2 reduction, water splitting, rechargeable batteries, or fuel cells; (2) used in an application where metal nanoparticles in carbon nanomaterialshave negative impact on their use (e.g., magnetic property and metal toxicity may impair the use of carbon nanomaterials in medical fields; metal nanoparticles may impair the mechanical strength when carbon nanomaterials are used as additives to improve mechanical property); (3) used in an application where it requires carbon nanomaterials to be well dispersed in an aqueous system.

[0023] In an embodiment, the carbon nanomaterial is disposed on a substrate. In an embodiment, the carbon nanomaterial is disposed on a metal foam. In an embodiment, the carbon nanomaterial is disposed on a non-woven carbon paper gas diffusion media comprising a microporous layer.

[0024] In another aspect, the present disclosure provides a catalytic electrode comprising an electrode; and carbon nanomaterial according to any embodiment of the present disclosure disposed on a surface of the electrode.

[0025] In another aspect, the present disclosure provides a reactor comprising a catalytic electrode according to any embodiment of the present disclosure.

[0026] This summary is provided to introduce a selection of concepts in a simplified form that are further described below in the Detailed Description. This summary is not intended to identify key features of the claimed subject matter, nor is it intended to be used as an aid in determining the scope of the claimed subject matter.DESCRIPTION OF THE DRAWINGS

[0027] The foregoing aspects and many of the attendant advantages of the present disclosure will become more readily appreciated as the same become better understood by reference to the following detailed description, when taken in conjunction with the accompanying drawings, wherein:

[0028] FIGURES 1A-1C provide (1A) Bright-field TEM, (IB, 1C) HAADF- STEM and corresponding elemental maps of as-received commercial Raw CNT;

[0029] FIGURES 1D-1F provide (ID) TEM, (IE, IF) HAADF-STEM and corresponding elemental maps of as-prepared carbon nanomaterials according to embodiments of the present disclosure;

[0030] FIGURES 2A and 2B provide (2A) XRD patterns of as-prepared carbon nanomaterials according to embodiments of the present disclosure and (2B) Raman spectra of as-received raw and carbon nanomaterials according to embodiments of the present disclosure;

[0031] FIGURES 2C-2F provide high-resolution XPS spectra for (2C) Ni 2p and (2D) N Is for carbon nanomaterials according to embodiments of the present disclosure; Ni K-edge (2E) XANES and (2F) EXAFS for raw CNT, CNT-Mel-650, and carbon nanomaterials according to embodiments of the present disclosure;

[0032] FIGURES 3A-3F illustrate CO current density and CO Faradaic efficiency as a function of applied potential for CNT-PTFE-Mel samples, where 3A and 3B illustrate carbon nanomaterials according to embodiments of the present disclosure pyrolyzed at varied temperatures, 3C and 3D illustrate CNT co-pyrolyzed with both PTFE and melamine, according to embodiments of the present disclosure, in comparison with those pyrolyzed with PTFE or melamine alone, and 3E and 3F illustrate carbon nanomaterials according to embodiments of the present disclosure prepared by regulargrade CNTs in comparison with that by industrial-grade CNTs, where results are based on H-cell measurements under neutral pH electrolyte (0.5 M KHCO3);

[0033] FIGURES 4A-4D illustrate (4 A) Faradaic efficiency toward CO and applied potentials on CNT-Mel-650 and carbon nanomaterials according to embodiments of the present disclosure at various current densities under pure feed CO2, (4B) 50 vol.% feed CO2 (balance Argon) and (4C) pure CO2 with variations in CNT type during preparation, and (4D) 60-hours stability test at a constant current density of 100 mA cm’2, where all electrochemical results were conducted in a flow-cell using 1.0 M KOH electrolyte;

[0034] FIGURES 5A-5C illustrate (5A) a linear sweep voltammogram (LSV), (5B) scan rate versus change in current density in non-Faradaic potential window for Cai determination, and (5C) electrochemical surface area (ECSA) normalized LSV polarization curves for carbon nanomaterials according to embodiments of the present disclosure and CNT-Mel-650, where all electrochemical results were conducted in a flow-cell using 1.0 M KOH electrolyte without iR correction and the inset in (5A) highlights the onset overpotential taken at 5 mA cm’2;

[0035] FIGURES 6A-6F illustrate (6A) Ni X-edge XANES from DFT-modeled NisN-NC, where vertical lines correspond to peak locations in the Ni X-edge XANES spectra of carbon nanomaterials according to embodiments of the present disclosure; (6B) isometric views of NisN NC atop Configuration 1 of N-doped sp2carbon at the start and after 30 ps under an artificial thermostat of 923 K during AIMD; (6C) the corresponding formation energies (EForm) of the four modeled systems during 30 ps of AIMD; (6D)calculated free energy diagram for eCChRR to CO and (6E) the corresponding difference in limiting potentials for eC02RR to CO and HER on NisN and metallic Ni-NCs and control NiN4 and FeN4 SACs; (6F) charge density difference plots of key intermediate steps on the NisN NC;

[0036] FIGURE 7A schematically illustrates synthesis and growth of Ni-based NCs from initial encapsulated Ni NPs in commercial CNTs according to embodiments of the present disclosure;

[0037] FIGURE 7B is a block diagram of a method for preparing carbon nanomaterials according to embodiments of the present disclosure;

[0038] FIGURES 8 A-8D are images of (8 A) raw CNT, (8B) CNT-Mel-650, (8C) CNT-PTFE-650, and (8D) carbon nanomaterials according to embodiments of the present disclosure, where a neodymium magnet was held above the samples and the magnet is at the top of each image behind a protective sheet of paper;

[0039] FIGURES 9A and 9B illustrate (9A) 90% IR corrected linear sweep voltammetry (LSV) and (9B) Nyquist plots of carbon nanomaterials according to embodiments of the present disclosure and commercial 40 wt.% Pt / C spray coated on nickel foam (NF), and the control bare NF substrate;

[0040] FIGURES 9C and 9D illustrate (9C) cyclic voltammetry versus scan rate for carbon nanomaterials according to embodiments of the present disclosure and (9D) commercial 40 wt.% Pt / C;

[0041] FIGURES 9E-9H illustrate (9E) scan rate versus change in current density in non-Faradaic potential window for Cai determination; (9F) ECSA-normalized current density for carbon nanomaterials according to embodiments of the present disclosure and commercial 40 wt.% Pt / C; (9G) DFT calculated Gibbs binding energies for proton absorption (AGH*) on NisN nanoclusters in carbon nanomaterials according to embodiments of the present disclosure and different geometries of Ni / Fe single (SAC) and dual-atom catalysts (DACs); (9H) 30-minute stability experiment of CNT-Mel-PTFE-650 at an applied overpotential of -233 mV, where electrochemical measurements for FIGURES 9C-9E were conducted in 1.0 M KOH, and all other electrochemical measurements were undertaken in 0.5 M H2SO4 electrolyte in a standard H-cell;

[0042] FIGURE 10 shows raw CNT and carbon nanomaterials according to embodiments of the present disclosure after being added to 3 mL of deionized water and sonicated for 15 min;

[0043] FIGURES 11A-11G provide (11 A) a HAADF image and (11B-11G) corresponding elemental maps of carbon nanomaterials according to embodiments of the present disclosure;

[0044] FIGURES 12A-12G provide (12A) a HAADF image and (12B-12G) corresponding elemental maps of carbon nanomaterials according to embodiments of the present disclosure;

[0045] FIGURES 13A-13G provide (13A) a HAADF image and (13B-13G) corresponding elemental maps of carbon nanomaterials according to embodiments of the present disclosure;

[0046] FIGURES 14A-14G provide (14A) a HAADF image and (14B-14G) corresponding elemental maps of carbon nanomaterials according to embodiments of the present disclosure;

[0047] FIGURES 15A and 15B are TEM and corresponding HAADF images of carbon nanomaterials according to embodiments of the present disclosure; and

[0048] FIGURES 16A and 16B are TEM and corresponding HAADF images of carbon nanomaterials according to embodiments of the present disclosure.DETAILED DESCRIPTION

[0049] In various aspects, the present disclosure provides methods for removing metal nanoparticles from a carbon nanomaterial, methods for increasing nitrogen doping level in a carbon nanomaterial, carbon nanomaterial having less than about 1 percent by weight metal based on the total weight of the carbon nanomaterial and greater than about 6 percent by weight nitrogen based on the total weight of the carbon nanomaterial, articles of manufacture that includes the carbon nanomaterial, catalytic electrodes, and reactors.

[0050] Intrinsic metal impurities in commercial CNTs may preclude their wide application as structural and functional materials in electronics and energy storage and conversion.

[0051] To address these and related challenges, the present disclosure provides a simple method, which, in embodiments, completely removes or greatly removes encapsulated nickel nanoparticles (Ni NPs, 10-50 nm) from CNTs via co-pyrolysis of CNTs with a fluorine-containing polymer, such as PTFE, and a nitrogen-containing compound, such as melamine, at a relatively low temperature (e.g., approximately 650 °C). Further, this approach is adapted or otherwise configured to redistribute the Ni content assurface-anchored Ni-based NCs ranging between 1-1.5 nm in diameter. Thorough characterization provided herein reveals the NCs to be a combination of NisN and metallic Ni, albeit predominantly NisN based. The NCs exhibit exceptional electroactivity for eCChRR to CO, with the carbon nanomaterials according to embodiments of the present disclosure demonstrating an ultra-low onset overpotential, such as of -19 mV, and achieving over 98% CO selectivity across a broad current density range from 100 to 700 mA cm'2. As shown herein, the stability of the nanoclusters was validated through over 60 hours of chronopotentiometry at 100 mA cm'2in an alkaline flow cell. The same synthesis method can be applied to industrial-grade CNTs that have much higher impurity levels, and the electrochemical activity is similar to that made by regular-grade CNTs, suggesting a promising method of utilizing low-cost raw materials as the catalyst feedstock. AIMD simulations on experimentally verified models were performed for 30 ps to shed light on the formation stability of said NCs, revealing that NisN-NCs are the most thermodynamically favorable. Subsequent DFT calculations on the reaction profile for CO production showcases near thermoneutral binding energies and CO selectivity for the NisN- NCs. By contrast, metallic Ni-NCs exhibited less thermodynamically desired adsorption energies with key reaction intermediates and favorable selectivities towards the parasitic HER.METHODS OF MAKING HETERO ATOM-DOPED CARBON NANOTUBES

[0052] Accordingly, in one aspect, the disclosure provides a method for removing metal nanoparticles from a carbon nanomaterial, comprising pyrolyzing a mixture of a carbon nanomaterial, a fluorine-containing polymer, and a nitrogen-containing compound at a temperature sufficient to remove metal nanoparticles from the carbon nanomaterial to provide a carbon nanomaterial depleted in metal nanoparticles compared to the carbon nanomaterial prior to pyrolysis. An example of such a method is illustrated in FIGURE 7B.

[0053] The carbon nanomaterials according to embodiments of the present disclosure, in various embodiments, are referred to according to the components of a precursor mixture and a pyrolyzation temperature. For example, CNT-PTFE-Mel-650 refers to a carbon nanomaterial derived or resulting from a pyrolysis mixture comprising CNT, PTFE, and melamine, which was pyrolyzed at 650 °C.

[0054] In another aspect, the disclosure provides a method for increasing nitrogen doping level in a carbon nanomaterial, comprising pyrolyzing a mixture of a carbonnanomaterial, a fluorine-containing polymer, and a nitrogen-containing compound at a temperature sufficient to increase nitrogen doping level in the carbon nanomaterial to provide a carbon nanomaterial with increased nitrogen doping level compared to the carbon nanomaterial prior to pyrolysis.

[0055] An example of such a method, method 700, is illustrated in FIGURE 7B.

[0056] As shown, in an embodiment, method 700 begins with process block 701, which comprises mixing or providing a mixture comprising a carbon nanomaterial, a fluorine-containing polymer, and a nitrogen-containing compound. In an embodiment, process block 701 is optional.

[0057] In an embodiment, process block 701 is followed by or the method 700 begins with process block 703, which comprises pyrolyzing a mixture comprising a carbon nanomaterial, a fluorine-containing polymer, and a nitrogen-containing compound. As described elsewhere herein, in embodiments, such pyrolyzation of the mixture removes metal nanoparticles from the carbon nanomaterial to provide a carbon nanomaterial depleted in metal nanoparticles compared to the carbon nanomaterial prior to pyrolysis and / or increases nitrogen doping level in the carbon nanomaterial to provide a carbon nanomaterial with increased nitrogen doping level compared to the carbon nanomaterial prior to pyrolysis.

[0058] In an embodiment, process block 703 is followed by process block 705, which comprises washing the resulting carbon nanomaterials under mild acidic conditions (such as with an acidic solution having a pH of less than 1) to remove metal content, such as from surface-bound nanoclusters. In an embodiment, the mild acid-washed carbon nanomaterial contains less than 0.5 at.% residual metal contaminants In an embodiment, process block 705 is optional. As described and demonstrated further herein, by using a combination of precursor materials including a mixture comprising a carbon nanomaterial, a fluorine-containing polymer, and a nitrogen-containing compound, a high N doping level (such as 7-8 at%) of carbon nanomaterials is achieved through a simple, low-temperature method. Because of the high N doping content, the carbon nanomaterial surfaces are more hydrophilic relative to a carbon nanomaterial feedstock. Accordingly, as is shown herein, the carbon nanomaterials according to embodiments of the present disclosure are better dispersed in aqueous solutions than the original raw carbon nanomaterial feedstock. See, for example, FIGURE 10. This increased dispersibility benefits the use of the carbonnanomaterials in applications when they are to be mixed with other ingredients in aqueous solutions.

[0059] Additionally, the present disclosure demonstrates that using certain fluorine-containing polymers (e.g., PTFE) and nitrogen-containing organic compounds (e.g., melamine) have coinciding, relatively low temperature ranges for thermal decomposition. This leads to the formation of certain gas products when pyrolyzing the mixture of the fluorine-containing polymer and nitrogen-containing compound. Without wishing to be bound to any particular theory, it is believed that the gas products penetrate through the carbon layers of the carbon nanomaterial, encapsulating metal nanoparticles therein or thereon and etching metal nanoparticles away from the carbon nanomaterials while doping nitrogen atoms onto the surface of carbon nanomaterials. As shown further herein, when CNTs are mixed with only PTFE or with only melamine in pyrolysis, the metal nanoparticles are not removed. When CNTs are mixed with only PTFE, no nitrogen doping on CNTs is observed. When CNTs are mixed with only melamine, nitrogen doping on CNTs is observed but the N content is low, less than about 2%. Only when carbon nanomaterials are mixed with fluorine-containing polymers and nitrogen-containing compounds altogether before pyrolysis, are metal nanoparticles removed from the carbon nanomaterials. Additionally, the resulting carbon nanomaterials comprise a high N doping content, such as around 7-8 at%.

[0060] In an embodiment, the level of metal nanoparticle removal can be verified by TEM, EDX, and XPS.

[0061] In certain embodiments of the above methods, the nanomaterial is a carbon nanotube (SWCNT or MWCNT), a carbon nanofiber, a graphene, or a graphene nanoplatelet.

[0062] In an embodiment, the carbon nanomaterial used in the pyrolysis mixture comprises metal impurities. Such carbon nanomaterials used in the pyrolysis mixture are distinct from the carbon nanomaterials according to embodiments of the present disclosure, which, in embodiments, have less than 1 wt% metal impurities. In an embodiment, the metal impurities comprise metal impurities chosen from Ni and Fe. In an embodiment, the carbon nanomaterials used for pyrolysis comprise metal impurities in a range of about 1 wt% to about 20 wt%. In an embodiment, the carbon nanomaterials used for pyrolysis comprise metal impurities in a range of about 1 wt% to about 15 wt%. In an embodiment, the carbon nanomaterials used for pyrolysis comprise metal impurities in a range of about2 wt% to about 10 wt%. In an embodiment, the carbon nanomaterials used for pyrolysis comprise metal impurities in a range of about 5 wt% to about 10 wt%. In an embodiment, the carbon nanomaterials used for pyrolysis comprise metal impurities in an amount of about 5 wt%. In an embodiment, the carbon nanomaterials comprise metal impurities in an amount of about 10 wt%.

[0063] In certain embodiments, the metal nanoparticles are selected from the group consisting of iron, nickel, cobalt, chromium, manganese, and copper nanoparticles, and mixtures thereof.

[0064] As above, the methods of the present disclosure comprise pyrolyzing a mixture comprising a fluorine-containing polymer. In certain embodiments of the methods, the fluorine-containing polymer is a polytetrafluoroethylene (PTFE).

[0065] In certain embodiments, the nitrogen-containing compound is an organic nitrogen-containing compound. In certain of these embodiments, the nitrogen-containing compound is selected from the group consisting of urea, melamine, dicyandiamide, or other short-chain nitrogen containing species such as thiourea. In certain embodiments of the methods, the nitrogen-containing compound is melamine.

[0066] In an embodiment, the pyrolysis mixture comprises, consists of, or consists essentially of CNTs, PTFE, and melamine.

[0067] As above, the methods of the present disclosure use a mixture comprising a fluorine-containing polymer, a nitrogen-containing compound, and carbon nanomaterials. The relative ratios of these components can be modified or adjusted, such as based on the composition of the carbon nanomaterial feedstock, such as, in particular, a metal composition of the carbon nanomaterial feedstock. As discussed further herein, a greater metal composition of the carbon nanomaterial feedstock can be addressed by a greater proportion in the mixture of the fluorine-containing polymer and the nitrogencontaining compound. In this regard, an adequate supply of both the nitrogen-containing compound and fluorine-containing polymers as reagents is useful for the effective removal and redistribution of Ni content from encapsulating carbon walls to the top surface of the carbon nanomaterial.

[0068] In an embodiment, a ratio (such as a weight:weight:weight ratio) of the carbon nanomaterial to the nitrogen-containing compound to the fluorine-containing polymer is in a range of about 1 :2:2 to about 1 : 10: 10, such as in a range of about 1 :4:4 to about 1 :8:8.

[0069] In an embodiment, the method includes preparing the mixture of the carbon nanomaterial, the nitrogen-containing compound, and the fluorine-containing polymer. In an embodiment, such preparation comprising pulverizing or mixing the components, such as in a mortar and pestle and the like.

[0070] In an embodiment, the fluorine-containing polymer and the nitrogencontaining compound have evaporation and / or decomposition temperatures in a range of about 10 °C to about 400 °C of each other. In an embodiment, the fluorine-containing polymer and the nitrogen-containing compound have evaporation and / or decomposition temperatures in a range of about 50 °C to about 300 °C. In an embodiment, the fluorine- containing polymer and the nitrogen-containing compound have evaporation and / or decomposition temperatures in a range of about 100 °C to about 250 °C. In an embodiment, the fluorine-containing polymer and the nitrogen-containing compound have evaporation and / or decomposition temperatures in a range of about 100 °C to about 200 °C. As discussed further herein, by having nitrogen-containing compounds and fluorine- containing polymers that evaporate and / or decompose within a relatively narrow temperature range, pyrolyzation mixtures comprising such compounds can generate both evaporated fluorine-containing polymers and nitrogen-containing compounds and / or decomposed components thereof in the presence of carbon nanomaterials. As is also described herein, such conditions can lead to the production of the carbon nanomaterials according to embodiments of the present disclosure.

[0071] In the embodiments, temperatures sufficient to remove metal nanoparticles from the carbon nanomaterial range from greater than about 550°C to less than about 750°C. In certain embodiments, the temperature sufficient to remove metal nanoparticles from the carbon nanomaterial is about 650°C. In certain embodiments, the temperature sufficient to remove metal nanoparticles from the carbon nanomaterial is greater than about 600°C.

[0072] In the embodiments, pyrolyzation temperatures range from greater than about 550°C to less than about 750°C. In certain embodiments, the pyrolyzation temperature is about 650°C. In certain embodiments, the pyrolyzation temperature is greater than about 600°C.

[0073] In certain embodiments of the methods, the carbon nanomaterial produced by the method is a non-magnetic carbon nanomaterial. See, for example, FIGURES 8A-8D and related discussion. Such magnetism, such as ferromagnetism or lack thereof, may betested, for example, by placing a neodymium magnet in proximity to the samples, such as within 1-3 cm. In the absence of such magnetism, the carbon nanomaterials according to embodiments of the present disclosure will not contact the magnet, or be attracted to the magnet sufficiently to overcome gravity.

[0074] Without wishing to be bound by theory, it is understood that the method of the present disclosure removes or substantially reduces the amount or concentration of metal nanoparticles on carbon nanomaterials, thus rendering the resultant carbon nanomaterials according to embodiments of the present disclosure largely non-magnetic.

[0075] In an embodiment, the mixture is pyrolyzed in an inert atmosphere, such as an inert atmosphere comprising or consisting of one or more inert gases. In an embodiment, the one or more inert gases is chosen from argon, nitrogen, xenon, and krypton. In an embodiment, the inert atmosphere does not comprise or comprises substantially no oxygen.

[0076] In an embodiment, the methods of the present disclosure comprise a mild acid washing step after pyrolyzation of the mixture. See process block 705. In an embodiment, the mild acid washing comprises washing the carbon nanomaterial with an acidic solution having a pH of less than 1, such as in a range of about 0.1 to 1. In certain embodiments, this post-pyrolyzation mild acid wash further removes, such as in embodiments totally removes, any residual metal from the carbon nanomaterials. In an embodiment, the resultant mild acid-washed carbon nanomaterials is greater than 99 wt% free of metal contaminants, such as metal contaminants present in feedstock carbon nanomaterials. In an embodiment, removal of metal contaminants can be verified by TEM imaging and elemental mapping. See, for example, FIGURES 11A-16B. In an embodiment, the mild acid-washed carbon nanomaterials comprise less than 0.5 at% metal impurities. As shown, for example, in FIGURES 11 A-16B, which show various images of carbon nanomaterials that have been washed with mild acidic solutions, the carbon nanomaterial structure is not damaged upon acid washing and metal-based nanoclusters are not present on the carbon nanomaterial surface after acid washing.

[0077] In an embodiment, the acid wash does not damage the carbon nanomaterial structure. In an embodiment, the acid wash partially or completely removes nitrogen from the carbon nanomaterial.METHODS FOR USING CARBON NANOTUBES

[0078] In a further aspect, the disclosure provides methods for using the carbon nanomaterials described herein. In certain of these embodiments, the disclosure provides methods for electrochemically reducing a molecule, such as for reducing carbon dioxide to carbon monoxide, comprising contacting the molecule, such as carbon dioxide, with a carbon nanomaterials prepared as described herein and / or according to any embodiments of the present disclosure.

[0079] In an embodiment, the carbon nanomaterials contact the carbon dioxide in a reactor according to an embodiment of the present disclosure.

[0080] In certain embodiments, the method selectively reduces carbon dioxide to carbon monoxide over hydrogen production with above 90% CO selectivity at a current density in the range of 50-500 mA / cm2as measured in a flow cell.

[0081] While carbon dioxide is discussed herein as an example of a reactant, it will be understood that other reactants, such as oxygen (e.g., O2) are possible with the carbon nanomaterials of the present disclosure and within the scope of the present disclosure. In this regard, the carbon nanomaterials of the present disclosure are configured to catalyze several reactions, thus converting several sets of reactants to products.CARBON NANOMATERIALS

[0082] In an aspect, the present disclosure provides carbon nanomaterials, such as CNTs.

[0083] In an embodiment, the disclosure provides carbon nanomaterials, such as may be prepared by the methods described herein.

[0084] In a further aspect, the disclosure provides a carbon nanomaterial having less than about 1 percent by weight metal based on the total weight of the carbon nanomaterial and greater than about 6 percent by weight nitrogen based on the total weight of the carbon nanomaterial.

[0085] In certain embodiments, the nanomaterials produced by the method have about 0.6% by weight of metal content (atomically dispersed form as opposed to nanoparticle form). It will be appreciated that the amount of metal is atomically dispersed on the surface of carbon nanomaterials, likely coordinated with nitrogen doping. For this reason, metal nanoparticles are not observed by TEM. However, atomically dispersed metal species are detected by XPS and EDX. These atomically dispersed metal centers, incoordination with nitrogen dopant, are active sites for catalytic reactions of CO2 reduction and water splitting, among other reactions.

[0086] In other embodiments of the methods, the carbon nanomaterial produced by the method is doped with nitrogen to provide a nitrogen-functionalized carbon nanomaterial. In an embodiment of the present disclosure, the carbon nanomaterials according to embodiments of the present disclosure comprise a nitrogen weight percentage in a range of about 5 wt% to about 8 wt%, such as may be measured by XPS and / or EDX. In an embodiment of the present disclosure, the carbon nanomaterials according to embodiments of the present disclosure comprise a nitrogen weight percentage in a range of about 6 wt% to about 8 wt%. In an embodiment of the present disclosure, the carbon nanomaterials according to embodiments of the present disclosure comprise a nitrogen weight percentage in a range of about 7 wt% to about 8 wt%. In an embodiment of the present disclosure, the carbon nanomaterials according to embodiments of the present disclosure comprise a nitrogen weight percentage of about 6.5 wt% to about 7.5 wt%.

[0087] In an embodiment, the carbon nanomaterials according to embodiments of the present disclosure comprises metal doping, such as Ni and Fe doping, in a range of about 0.5 wt% to about 4.0 wt%, such as may be measured by XPS and / or EDX. In an embodiment, the carbon nanomaterials according to embodiments of the present disclosure comprises metal doping, such as Ni and Fe doping, in a range of about 0.6 wt% to about 3.5 wt%. In an embodiment, the carbon nanomaterials according to embodiments of the present disclosure comprises metal doping, such as Ni and Fe doping, of less than 1 wt%.

[0088] In certain of these embodiments, the nitrogen-functionalized carbon nanomaterial has improved water dispersion compared to the carbon nanomaterial prior to pyrolysis. In other of these embodiments, the nitrogen-functionalized carbon nanomaterial has a nitrogen doping level at a minimum of about 6 percent by weight based on the total weight of the carbon nanomaterial. In certain embodiment of these methods, the nitrogen- functionalized carbon nanomaterial has a nitrogen doping level from about 6 to 12% by weight based on the total weight of the carbon nanomaterial in the bulk measured by EDX. It will be appreciated that the surface nitrogen doping level measured by XPS is greater than the bulk nitrogen doping level measured by EDX because most doping occurs on the nanomaterial surface.

[0089] In further embodiments, the carbon nanomaterial, such as produced by the method of the present disclosure, is not damaged during the process, as measured by transmission electron microscopy (TEM).

[0090] In an embodiment, the carbon nanomaterials of the present disclosure comprise cubic Ni and hexagonal NisN facets, such as may be measured by XRD, which are missing from the X-ray diffractogram of raw carbon nanomaterials (i.e., feedstock materials). As described further herein, such cubic Ni and hexagonal NisN facets can contribute to catalysis, such as in reduction of CO2 and / or H2.CATALYTIC ELECTRODES

[0091] In another aspect, the disclosure provides articles of manufacture that include the carbon nanomaterials according to any embodiments described herein. In certain embodiments, the article is (1) an electrode for CO2 reduction, N2 reduction, water splitting, rechargeable batteries, or fuel cells; (2) used in an application where metal nanoparticles in carbon nanomaterials have negative impact on their use (e.g., magnetic property and metal toxicity may impair the use of carbon nanomaterials in medical fields; metal nanoparticles may impair the mechanical strength when carbon nanomaterials are used as additives to improve mechanical property); (3) used in an application where it requires carbon nanomaterials to be well dispersed in an aqueous system.

[0092] In an aspect, the present disclosure provides a catalytic electrode, such as for reducing a molecule. In an embodiment, the catalytic electrode comprises a carbon nanomaterial according to any embodiment of the present disclosure.

[0093] In an embodiment, the catalytic electrode comprises an electrode, such as an electrode, and a carbon nanomaterial according to any embodiment of the present disclosure disposed on a surface of the electrode.

[0094] In an embodiment, the carbon nanomaterial of the present disclosure is disposed on a substrate. In an embodiment, the carbon nanomaterial is disposed on a metal foam, such as a nickel foam. In an embodiment, the carbon nanomaterial is disposed on a gas diffusion layer. In an embodiment, the gas diffusion layer is a non-woven carbon paper gas diffusion media comprising a microporous layer. In an embodiment, the microporous layer has been PTFE treated, such as to 5 wt%.

[0095] The carbon nanomaterial may be applied to the electrode with any deposition or adherence method, such as by dip coating, spin coating, spray coating, drop casting.

[0096] As discussed further herein with respect to reactors according to the present disclosure, the catalytic electrodes of the present disclosure may be used in reactor, such as for reducing a molecule. Accordingly, in an embodiment, the carbon nanomaterials according to embodiments of the present disclosure have an onset overpotential of -19 mV and 98% or greater CO selectivity in a range of 100-700 mA / cm2.REACTOR

[0097] In another aspect, the present disclosure provides a reactor, such as, in embodiments, to catalytically reduce carbon dioxide to provide carbon monoxide. In other embodiments, the reactor is configured to generate hydrogen gas from water, such as through the application of a voltage.

[0098] In an embodiment, the reactor comprises a catalytic electrode as described with respect to other aspects of the present disclosure. In this regard, the catalytic electrode comprises a carbon nanomaterial according to any embodiment of the present disclosure or made according to any methods of the present disclosure.

[0099] In an embodiment, the reactor is a flow cell configured to flow a gas, such as a gas comprising carbon dioxide, over the catalytic electrode.

[0100] In an embodiment, the reactor comprises a first fluid compartment; a second fluid compartment; an ion exchange membrane fluidically separating the first fluid compartment and the second fluid compartment and configured to allow passage of ions therethrough; a first electrode in electrically conductive communication with an interior portion of the first fluid compartment; and the catalytic electrode according to any embodiment of the present disclosure in electrically conductive communication with an interior portion of the second fluid compartment.

[0101] In an embodiment, the ion exchange membrane is a Nafion membrane. In an embodiment, the ion exchange membrane is an anion exchange membrane.

[0102] In an embodiment, the ion exchange membrane is configured to allow passage of dissolved ions through the membrane, but to block passage of, for example, certain other ions, such as on the basis of charge, or neutral molecules also dissolved in a common solvent.

[0103] In an embodiment, the first fluid compartment and / or the second fluid compartment contain or carry a KHCO3 solution comprising dissolved CO2. In an embodiment, the first fluid compartment and / or the second fluid compartment contain or carry a KOH solution comprising dissolved CO2.EXAMPLESEXAMPLE 1 : SYNTHESIS OF CARBON NANOMATERIALS AND CO2REDUCTION THEREWITH

[0104] The present Example describes synthesis and characterization of carbon nanomaterials according to embodiments of the present disclosure, as well as eCO2RR to CO with such carbon nanomaterialsEXPERIMENTAL SECTION

[0105] Materials. Commercial-grade multiwalled CNTs (>95 wt% purity, Cheap Tubes Inc., referred to as Raw CNT) and industrial-grade multiwalled CNTs (>90 wt% purity, Cheap Tubes Inc., referred to as Raw IndCNT) were used in this study, with the latter containing higher levels of residual metallic impurities. Polytetrafluoroethylene (PTFE, free-flowing, 2.15 g / mL) and melamine (Mel) were both acquired from Sigma- Aldrich. Treatment of the MWCNTs was performed in a 1-inch tubular furnace (Thermal Scientific, Lindberg Blue M) under an argon atmosphere (Airgas, UHP grade).

[0106] Synthesis of CNT-PTFE-Mel and derivatives. In a typical procedure, 100 mg of Raw CNTs, PTFE, and Mel were weighed in a 1 :4:4 mass ratio of CNT:PTFE:Mel and thoroughly mixed by mortar and pestle. The homogenous solid mixture was then loaded into an alumina crucible which was loaded into the tubular furnace. Quartz wool was used downstream of the reactor heating zone to sequester any unreacted and recondensing Mel or PTFE. This is simply a quality control step to prevent downstream blockages of small-diameter connections. Prior to initiating the heating profile, Argon gas was flowed at 100 mL min'1for 10 min to purge ambient O2, which could partially oxidize the Raw CNTs under synthesis conditions. The same flowrate of Argon was allowed for the remainder of the synthesis. The reaction profile entailed a ramp-up rate of 5 °C min'1until 650 °C. This optimized reaction temperature was held for 1 hour followed by natural cooling to ambient temperature.

[0107] Samples are denoted as IndCNT-PTFE-Mel when industrial-grade MWCNT was used instead of commercial-grade CNTs. The synthesis of IndCNT-PTFE- Mel followed the same above-mentioned procedure but with the use of a 1 :8:8 mass ratio of IndCNT:PTFE:Mel to account for the additional metallic impurities of IndCNT.

[0108] Control samples with only Mel, or only PTFE, being mixed with Raw CNTs are denoted as CNT-Mel or CNT-PTFE, respectively. Further, syntheses with mixed PTFE and Mel being placed in an upstream boat with respect to Raw CNT are denoted by the inclusion of PS (physically separated) at the end of their indexed name. Other comparison samples prepared by treating CNTs with PTFE and / or Mel under different maximum temperatures of 400, 550, 750, and 850 °C are denoted by the inclusion of said temperature at the end of their indexed name.MATERIAL CHARACTERIZATIONS:

[0109] Surface and bulk textural and morphological characterization of the samples were performed through scanning electron microscopy (SEM, JEOL JSM7500F) and transmission electron microscopy (TEM, FEI 200kV Titan Themis STEM). Similarly, high-angle annular dark-field scanning TEM / energy dispersive X-ray spectroscopy (HAADF-STEMZEDS) was performed on the same FEI 200kV Titan Themis STEM. Surface elemental composition was measured using an EnviroESCA X-ray photoelectron spectroscopy (XPS) using an Al-Ka radiation source and an energy reference line from the Cis (284.8 eV) of exogenous carbon. Crystallinity measurements and crystallite size approximations were attained from X-ray diffraction analysis (XRD, Bruker Lynxeye detector XTE). Thermogravimetric analysis (TGA, Mettler Toledo TGA / DSC 3+ star system) was used under Argon with a ramping rate of 5 °C min'1from 25 to 650 °C, upon which it was held for 1 hour.

[0110] X-ray Absorption Spectroscopy (XAS) data was collected at the Stanford Synchrotron Radiation Lightsource (SSRL) beamline 11-2. All measurements were performed in fluorescence mode using a Canberra 100-pixel Germanium solid-state monolith detector. Reference foil spectra for Nickel and Iron were acquired alongside each individual scan in transmission mode. Nickel scans were conducted within an energy range of 8103-8805 eV, while iron scans covered 6882-7675 eV. Each sample spectrum was collected at least twice. Data preprocessing was carried out using the SIXPack and Athena software packages to extract XANES and EXAFS data. Ni K-edge and Fe K-edge EXAFS were analyzed using the Hanning window function with k-weights of 2, and fitting wasperformed with the Artemis software package. Scattering paths for model-based EXAFS analysis were generated using the fast Fourier transform extended X-ray absorption fine structure (FEFF) calculation function in Artemis based on the crystal structures of NiPc and FePc. The amplitude reduction factor (So2) for NiPc and FePc fits was applied to calculate the coordination number and bond length. EXAFS fitting was performed within a k-range of ~3-l 1 A’1.ELECTROCHEMICAL MEASUREMENTS:[OHl] H-cell: The conventional H-Cell included of two compartments, separated by an anion exchange membrane (AEM) (Fumasep PK 130, Fuel Cell Store). It includes three electrodes: a working electrode, a Ag / AgCl (3M KC1) reference electrode on the cathode side, and a 1 cm2Pt foil as a counter electrode. A CO2-saturated 0.5 M KHCO3solution was used as both the catholyte and anolyte. The potentials are displayed in their iR corrected form on the reversible hydrogen electrode (RHE) scale. The catalyst ink comprised 3 mg of catalyst in a mixture of 370 pL ethanol, 200 pL water, and 30 pL of PFSA Dispersion (D5, 5%, Fuel Cell Store). Following this for electrode preparation, Toray paper with an active area of 1 cm2was drop cast with 200 pL of the catalyst ink. High-purity CO2(99.999%, Airgas) was introduced into the cathode chamber at a flow rate of 30 mL min'1for 30 minutes to ensure complete saturation of the catholyte, maintaining the flow rate throughout the experiment. The products were analyzed using an online gas chromatograph (GC, GC-2014ATF, Shimadzu), equipped with a thermal conductivity detector (TCD) and a methanizer-assisted flame ionization detector (FID).

[0112] Flow-cell: A three-compartment electrochemical flow-cell was employed for performing current-voltage sweeps on the different as-prepared GDEs. Briefly, the anode and cathode chambers were separated by a Fumasep PK 130 AEM. Nickel foam served as the anode for the oxygen evolution reaction (OER). To prepare the cathode GDE, a catalyst ink was made by combining 10 mg of catalyst, 3 mL of ethanol, and 300 pL of PFSA Dispersion (D5, 5%, Fuel Cell Store). This ink was airbrushed onto a multi-layer gas diffusion layer (GDL, Sigracet 39BB, Fuel Cell Store), achieving an optimized loading of approximately 1 mg cm'2. The cell was then assembled with a 1 cm2working area for both electrodes. Both the anolyte and catholyte, each containing 1 M KOH, were circulated in their respective chambers. An Hg / HgO reference electrode was placed in the catholyte flow. High-purity CO2(Airgas, 99.999%) was continuously fed to the back of the cathode at a flow rate of approximately 30 mL min'1. A DC power supply (Agilent E3633A)provided a constant current to the cell, while a multimeter (AidoTek VC97+) measured the potential between the working and reference electrodes. Measured potentials were converted to the reversible hydrogen electrode (RHE) scale using Eqn. (1). Overpotential at current density i (rji) was taken as the difference between the applied voltage (UaPP(V vs. RHE)) and the equilibrium potential for CO2 reduction to CO (-0.11 (V vs. RHE)). The onset overpotential (qo) is taken at a current density of -5 mA cm'2to minimize scanning artifacts. All potential values for H-cell and flow-cell performance results are reported without iR corrections.

[0113] URHE = Uref + 0.059 x pH + E°ef= Vref + 0.059 x pH + 0.210 VEqn. (1)

[10114] J FE = x 100 jRTEqn. (2)

[0115] An online gas chromatograph (GC, GC2010, Shimadzu) was utilized to measure on-line product concentrations through a thermal conductivity detector (TCD) for H2, CO, and unreacted CO2, as well as a flame ionization detector (FID) for potential hydrocarbons (CH4, C2H4) and high vapor pressure oxygenates. To that end, the selectivity towards eCO2RR to CO or HER to H2 is quantified through Eqn. (2) for Faradaic efficiency (FE). Therein, ‘z’ represents the number of electrons needed to participate in a Faradaic reaction (z = 2 for CO production and HER), ‘ ’ is the Faraday constant (~ 96,485 C mol’J), ‘v’ is the volumetric flowrate of CO2 supplied to the cell,is the applied current (A), ‘P’ is the ambient pressure (101 kPa), ‘ ’ is the volumetric concentration of the target product as measured by the GC, ‘P’ is the universal gas constant (8.314 J mol’1K’1), and ‘F is the ambient temperature used during the performance experiments. For all tested samples, non-ideal selectivity towards CO production entailed the competing HER, without any evidence of other eCO2RR products.COMPUTATIONAL DETAILS:

[0116] Density functional theory (DFT)-based calculations and ab-initio molecular dynamics (AIMD) simulations were conducted with the Vienna ab initio software package (VASP). For the exchange-correlation (XC) functional, the revised Perdew-Burke-Ernzerhof generalized gradient approximation (revPBE-GGA) was utilized,as it has shown improvement in the adsorption phenomena energetics on transition-metal surfaces. Integration of the Brillouin zone was done with the Gamma centered Monkhorst- Pack grid method. The electron-ion interactions were treated using the Projected Augmented Wave (PAW) method. Kohn-Sham orbitals were expanded with a plane wave basis set, and the kinetic energy cutoff was optimized at 500 eV. Structural convergence was achieved when the Hellmann-Feynman forces and total energies were smaller than 10’5eV and 0.02 eV A’1, respectively. To accelerate the convergence of states towards the system-dependent Fermi levels, Gaussian smearing with a finite width of 0.05 eV was applied. For total and projected density of states (T / PDOS) calculations on bulk systems, a 5x5x3 k-point mesh using the MP method was employed to sample the Brillouin zone. Post-processing of data was produced via VASPKIT. A conventional 16-A vacuum region was included to prevent unphysical interactions above the modeled catalyst.

[0117] The computational hydrogen electrode (CHE) model was employed to approximate the effective free energy diagram between intermediate electrochemical steps in the eCCERR, whilst accounting for the energy of proton-electron pairs in an aqueous solution. The free energy of the [H++ e“] reaction under standard conditions is connected to the chemical potential of % H2 gaseous molecular through the standard hydrogen electrode (SHE). The change in free energy with respect to potential (AG(U)) for an elementary step *S + H++ e" (U) —> *SH at a given electrode potential U is calculated using Eqn. (3).

[0118] AG(U) = ESH* - Es* - % E(H2) + AZPE TAS + eUEqn. (3)

[0119] There, ESH* and Es* correspond to the energies of the system with and without the proton (H) interaction, respectively. AZPE and TAS correspond to the change in zero-point vibrational energies and entropy. Both adsorbate-specific AZPE and TAS were calculated using VASP. Moreover, the limiting potential (UL) for a give reaction (eCCERR or HER) is acquired by finding the maximum free energy change AGmax(UL = AGmax / e) at its equilibrium potential. The difference between the UL for eCCERR and HER (AUL = UL(CCCERR) - UL(HER)) offers a probe on selectivity trends, wherein a more positive AUL corresponds to a higher eCCERR selectivity. Further, to identify the minimum energy path and transition state for systems of interest, the nudged elastic band (NEB) method wasapplied. The initial full NEB path contained 5 images between two relaxed intermediates steps. A spring constant of 5 eV A'2was used with the above-mentioned convergence criteria for adsorbate binding energies.

[0120] The AIMD simulations were conducted using the NVT ensemble at both synthesis (923 K) and electrochemical reaction temperatures (300 K) were performed with a time step of 1 femtosecond (fs) with an average of 20-30 picoseconds (ps) for the former. Temperature oscillations were controlled using the Nose-Hoover thermostat. For the simulations, tritium masses were assigned to hydrogen atoms to use a larger timestep. AIMD calculations were carried out at the T point with a plane-wave energy cutoff of 500 eV.METAL AND NITROGEN DOPING LEVEL

[0121] Table 1 shows surface elemental composition from XPS analysis and bulk elemental composition from EDX analysis for the CNT_PTFE_Mel_650 sample. The sample contains a low metal concentration in the bulk (from EDX) at around 0.6 wt.% and a high nitrogen concentration in the bulk (from EDX) at around 6.4 wt.%. In comparison, the raw CNT obtained from the vender contains about 5 wt.% of metal impurities (data from the vender website). This indicates a significant reduction of metal content from the treatment. The XPS data show the surface concentrations of metal and nitrogen are higher than those in the bulk, which is reasonable because most of the doping occurs on the surface of the material. In contrast, the CNT-Mel sample has only around 1 at.% N doping level, much lower than that of CNT_PTFE_Mel_650 (7.3 wt.%); while the CNT-PTFE has no N doping due to the lack of N precursor in synthesis. N doping level found in the literature of N-doped carbon materials is typically in the range of 1 - 3 wt.%. These results indicate that the combination of PTFE and melamine to pyrolyze with commercial CNTs has a synergistic effect that not only removes metal nanoparticles but also increases N doping significantly. The observed surface metal concentration from XPS is likely ascribed to the atomically dispersed metal species on the CNT surface, likely coordinated with the N dopants to form M-Nx sites.

[0122] Table 1 Metal and nitrogen concentrations in the CNT_PTFE_Mel_650 sample as determined by XPS and EDX.DISPERSION IN WATER

[0123] Another application of the materials yielded from this method is their increased dispersibility in water. As shown in FIGURE 10 after a short time mixing CNT_PTFE_Mel_650 is evenly dispersed in water where Raw-CNT shows little to no dispersion.DISCUSSION

[0124] The present disclosure provides an important and innovative approach that harnesses the high loading of metallic impurities typically found in commercial CNTs, converting them into electroactive components integrated within the carbon structure. This method has been validated on two grades of commercially available multi -walled CNTs (MWCNTs) having varied impurity levels: (1) regular-grade CNTs with up to 5 wt% impurities and (2) industrial-grade CNTs with up to 10 wt% impurities, while the industrial-grade is about half of the price of the regular-grade. The present disclosure provides a facile, in embodiment, single-step, low-temperature pyrolysis method — requiring no pre- or post-processing — that effectively removes encapsulated Ni nanoparticles from the tips and tubes of MWCNTs. This process then redistributes the nickel uniformly across the carbon surface, forming 1-1.5 nm NisN nanoclusters centered around nitrogen defect sites, with a very high N content (7—11 at%) on the CNT surfaces. Briefly, in embodiments, polytetrafluoroethylene (PTFE), melamine, and impure MWCNTs were mixed and pyrolyzed under inert atmosphere at a relatively low temperature of 650 °C. FIGURE 7A shows an overview of the hypothesized growth procedure. An array of characterization techniques including TEM, STEM, XAS, XPS, Raman spectroscopy, and XRD have been performed to confirm the removal ofencapsulated Ni NPs and redistribution of Ni-based NCs on MWCNTs. Moreover, first- principles density functional theory (DFT) and ab initio molecular dynamics (AIMD) simulations were performed to gain insights into the stability, catalytic activity and selectivity toward eC02RR versus competing cathodic reactions, as well as formation dynamics of the NCs atop N-doped CNT surface. Computational models were confirmed to mimic the real identity of the NCs through a comparative XANES profile, which revealed NisN to be the predominant stable NC type on the final material. Furthermore, the electrochemical performance of the resultant material, hereby referred to as CNT-PTFE- Mel-650, was tested toward eCChRR. A selectivity greater than 98% was held towards CO production at current densities from 100 to 700 mA cm'2and a continuous operation of 60+ hours in a flow-cell electrolyzer was demonstrated.

[0125] Removal of Encapsulated Ni NPs and their Redistribution as Surface NCs

[0126] The methods of the present disclosure were used on regular-grade MWCNTs (impurities < 5 wt%). In a typical procedure, an initially magnetic Raw CNT sample was thoroughly mixed with melamine and PTFE in a 1 :4:4 mass ratio and pyrolyzed in a tubular furnace at 650 °C for 1 h in an argon environment, resulting in the sample of CNT-PTFE-Mel-650. Briefly, the naming convention is structured as CNT type - PTFE presence - melamine presence - pyrolysis temperature, indicating the CNT type, whether PTFE and melamine were used, and the pyrolysis temperature applied in Celsius. Upon cooling, the resultant black powder was completely non-magnetic, which suggested the removal of impurities. A comparative analysis of HRTEM and HAADF-STEM EDS with elemental mapping in FIGURES 1 A and IB for Raw CNT and FIGURES ID and IE for CNT-PTFE-Mel-650 confirmed the removal of initially encapsulated Ni seed nanoparticles (NPs) ranging from 10-50 nm in diameter. Interestingly, HAADF-STEM imaging in FIGURES IE and IF for CNT-PTFE-Mel-650 demonstrates a thorough and homogenous redistribution of nickel along the surface of the CNTs in the form of 1-2 nm nanoclusters (NCs). Elemental maps in the same figure confirm the presence of Ni-based NCs as well as a homogenous N doping along the carbon backbone. The XPS scan and HAADF-STEM image in FIGURE 1C revealed no surface Ni character, suggesting that the Ni observed in FIGURES 1 A and IB for Raw CNT is encapsulated by graphitic carbon.

[0127] To ascertain the individual effects of PTFE and melamine, control samples of CNT-PTFE-650 and CNT -Mel-650 were prepared and characterized. To that end, TEM, HAADF-STEM, and the corresponding elemental maps of Ni in CNT -Mel-650and CNT-PTFE-650 demonstrate the retention of the metal NPs within the encapsulating graphitic carbon walls. This was also confirmed by the ferromagnetism of both CNT -Mel- 650 and CNT-PTFE-650 samples. This result suggests that the co-existence of PTFE and melamine in the pyrolysis step is advantageous to the removal of encapsulated metal NPs and subsequent formation of NCs on CNT surface. It is important to highlight that studies have shown a direct relationship between the size of a magnetic nanomaterial and its magnetic moment. Specifically, as the diameter decreases to 1 nm, the magnetic moment tends to zero, in accordance with the magnetically dead layer theory. For the sample of CNT -Mel-650, prepared by using 1 :4 mass ratio of CNT:Mel (without PTFE), the material’s properties match those in both of our previous studies and the literature, i.e. CNT treated with melamine alone can lead to a small amount of metallic impurities leaching from CNTs forming single atom sites on the CNT surface. The encapsulated NPs, however, remain largely untouched. Notwithstanding, XPS results for CNT-PTFE-650 show that in the absence of melamine, the effect on N (or F) doping or the leaching of Ni impurities is negligible.

[0128] To investigate the effect of pyrolysis temperature, CNT-PTFE-Mel samples were prepared in the temperature range from 550 to 850 °C. Surprisingly, all samples prepared at 550, 750, and 850 °C were magnetic, in sharp contrast with the sample prepared at 650 °C, suggesting the unique treatment enabled only at around 650 °C. The same pyrolysis method was successfully applied to treating industrial-grade CNT (IndCNT) with a higher level of impurities (up to 10 wt.%). Due to almost double the amount of impurities in Raw IndCNT compared to Raw CNT, a mass ratio of 1 :8:8 of IndCNT:PTFE:Mel was employed. Upon allowing the single physicochemical treatment protocol, magnetism was lost from the initially magnetic Raw IndCNT. Subsequent TEM images of IndCNT -PTFE-Mel-650 show a lack of encapsulated Ni NPs, and HAADF- STEM images confirm the removal of Ni NPs and homogenous redistribution of Ni species from within the graphitic carbon structure to its surface. Therefore, solely based on textural characterization thus far, the requirement of an adequate supply of both melamine and PTFE as reagents is needed for the effective removal and redistribution of Ni content from encapsulating carbon walls to the top surface.

[0129] Characterization of crystallinity was therefore conducted to both scope the inclusion of Ni-based crystalline NCs on the surface of the treated CNTs, as well as to determine the degree of disruption to the crystalline sp2hybridized carbon structure ofCNTs. The indexed XRD patterns in FIGURE 2A for CNT-PTFE-Mel-650 highlights the presence of both cubic Ni and hexagonal NisN facets, which are missing from the X-ray diffractogram of Raw CNT. Briefly, peaks at 44.5° and 51.8° correspond to the (111) and (200) facets of cubic Ni phase (ICDD PDF # 04-0850) and at 43.3° degrees for the (111) facet of the hexagonal NisN phase (ICDD PDF #70-9599). Further, to investigate variations within the crystalline structure of CNTs, Raman spectroscopy analysis was performed in FIGURE 2B. The intensity ratio between the D and G bands allows inference on the degree of graphitization and consequently surface defects present. Comparing Raw CNT with its treated counterpart (CNT-PTFE-Mel-650), a near constant ID / IG ratio of approximately 0.5 is noted. However, a slight decrease from 0.97 to 0.76 is registered for IndCNT and IndCNT-PTFE-Mel-650, respectively. This indicates that although the net structural distortion of IndCNT-PTFE-Mel-650 is negligible, an increase in order (or graphitization) occurs for the initially more defective IndCNT. This is likely a result of M-N-C type bonding around sites with initial carbon vacancies or other pre-existing defects in the Raw CNT and IndCNT.

[0130] For comparison, control samples containing Mel or PTFE were prepared and treated at 650°C. Results of composition distribution from XPS surveys reveal much higher N-doping in the CNT-PTFE-Mel-650 (7.1 at.%) compared to the CNT -Mel-650 (1.3 at.%), although both were treated with the same mass ratio of CNT to melamine. The reported distributions of fitted N Is XPS spectra shows a 7.4x increase in the carbon-bound N distribution for CNT-PTFE-Mel-650 compared to CNT-Mel-650, suggesting that PTFE possibly stabilizes graphitic N sites at 650 °C. Moreover, it can be seen that the rest of nitrogen is largely present in the form of thermodynamically stable NisN NCs (~ 39% of N Is distribution). The Ni 2p XPS spectrum (FIGURE 2C) shows a sharp Ni 2p3 / 2 peak at 855.5 eV, indicating that Ni in the NCs is in a low-valent state (Ni5+, 0 < 5 < 2). Fitting the N Is XPS spectrum reveals peaks for pyridinic N (398.3 eV), Ni-N (399.2 eV), pyrrolic N (400.4 eV), graphitic N (401.3 eV), and oxidized N (405.7 eV). The Nls spectra in FIGURE 2D for CNT-PTFE-Mel-650 indicate a dominant metal nitride character, which closely matches the distribution obtained for IndCNT-PTFE-Mel-650. Notably, IndCNT- PTFE-Mel-650 shows about 52 and 67% higher N and Ni contents, respectively, compared to CNT-PTFE-Mel-650 likely due to a greater availability of Ni impurities for N coupling into stable NisN nanoclusters.

[0131] The results of Ni X-edge XANES and EXAFS from Raw CNT, CNT- PTFE-Mel-650 and CNT-Mel-650 are shown in FIGURE 2E and FIGURE 2F, respectively. Control samples of Ni foil, NiO, and Ni phthalocyanine (NiPc) were used as references for Ni°, Ni2+, and Ni-N valence states, respectively. The absorption edges of Raw CNT and CNT-Mel-650 in FIGURE 2E are quite close to that of Ni foil, albeit a slight shift to higher oxidation state can be noted for the CNT-Mel-650 sample. This suggests that minor Ni5+character is present on the surface as Ni-N-C although the bulk majority remains as encapsulated metallic Ni NPs. The probing depth of XAS is generally considered to be micrometer level, which explains why Ni K-edge spectra are similar for Raw CNT and CNT-Mel-650. In contrast, CNT-PTFE-Mel-650 exhibits an absorption edge closer to NiPc, indicating Ni bonding with other species namely N in the NisN structure. Further, the Fourier transformed (FT) EXAFS spectra in FIGURE 2F shed light on the coordination environments of Ni in the as-prepared samples. Briefly, two primary peaks around 1.5 and 2.2 A in the standard NiPc and Ni foil samples correspond to Ni-N and Ni-Ni coordination shells. A lower Ni-Ni peak around 2.2 A is observed for CNT-Mel- 650 than Raw CNT, suggesting that partial removal of metallic Ni character has been achieved - which supports the surface Ni5+signal registered from XPS in CNT-Mel-650. Similarly, a further reduction in the Ni-Ni peak intensity is observed for CNT-PTFE-Mel- 650, confirming more Ni removal. However, given that no Ni NPs were observed from HAADF-STEM (FIGURES IE and IF) and that the CNT-PTFE-Mel-650 sample is not magnetic, the Ni-Ni peak at 2.2 A is attributed to metallic Ni-NCs rather than Ni-NPs. This is in agreement with NisN NCs, which are confirmed from the presence of a peak around 1.5A in EXAFS. The results show that the coordination number (CN) of Ni-Ni shells falls from 10.5 ± 0.7 in Raw CNT to 7.1 ± 1.7 in CNT-Mel-650 to 1.9 ± 0.8 in CNT-PTFE-Mel- 650. The smallest CN for Ni-Ni shells in CNT-PTFE-Mel-650 suggests the presence of smaller aggregates (ranging from NPs to NCs). When combined with the larger CN of Ni- N shells at 2.5 ± 0.8, this points to the identity of NCs (NisN) rather than an MNC-type SAC moieties.

[0132] TGA analyses and a series of control experiments were conducted to explore the roles of PTFE and melamine as well as pyrolysis temperatures in the etching of Ni NPs and subsequent formation of Ni NCs. TGA results reveal that under argon melamine and PTFE start decomposing at approximately 250 and 475 °C, respectively. This implies an existence of a time period between melamine and PTFE decompositionswithin the CNT-PTFE-Mel mixture under the reaction ramp-up rate of 5 °C min'1, although both are fully decomposed by 650 °C. Control TGA of pure PTFE and melamine samples exhibit complete decomposition at 600 and 350 °C, respectively. These results may explain the retained ferromagnetic behavior of CNT-PTFE-Mel-550. Since PTFE is not fully decomposed at 550 °C, the needed amount of PTFE-Mel gaseous reaction product is limiting at 550 °C and therefore the removal of encapsulated Ni NPs becomes reagentlimited. In addition, a control experiment was performed with a mixture of PTFE and melamine loaded in an upstream crucible while Raw CNT being in a downstream crucible in the same tube furnace. Although the same temperature profile was used for CNT-PTFE- Mel-650, and the resultant material was non-magnetic. This strongly suggests that it is the gas-phase reaction product from the PTFE-melamine mixture at 650 °C that transferred downstream to the CNTs and leached the Ni NPs out of the CNTs. Furthermore, control experiments were done by mixing premade CNT -Mel-650 sample with PTFE which then underwent pyrolysis at 650 and 850 °C, respectively. The two (CNT-Mel-650)+PTFE samples prepared at 650 and 850 °C were magnetic, much like their parent CNT-Mel-650 but unlike CNT-PTFE-Mel-650. This suggests that the unique structure of CNT-PTFE- Mel-650 is ascribed to the synergy between PTFE and melamine during the synthesis, rather than their individual or sequential effects. Most likely, gas phase intermediate products generated from the reaction between melamine and PTFE during the pyrolysis are responsible for the etching of graphitic carbon that encapsulates Ni NPs and the subsequent homogenous surface redistribution of nickel as metal and NisN NCs. Too high of a pyrolysis temperature (greater than 650 °C, e.g., 750 °C) may cause the sintering of the NCs to NPs, manifested by the emergence of ferromagnetic behavior for CNT-PTFE-Mel-750 and CNT-PTFE-Mel-850 samples. In fact, the non-magnetic CNT-PTFE-Mel-650 sample was placed in a second pyrolysis step at 850 °C, and the resultant sample turned out to be magnetic, supporting the sintering hypothesis.

[0133] Electrochemical Performance

[0134] The CNT-PTFE-Mel-650 sample emerges as an outlier in the study, uniquely combining non-magnetic properties, from complete removal of encapsulated Ni NPs, with exceptional CO2 to CO reduction performance. This highlights the pivotal role of pyrolysis temperature and reagents presence in dictating structural and electrocatalytic characteristics. The CNT-PTFE-Mel-650 sample was not only anomalously a nonmagnetic sample in the studied sample set, but it also attained the best activity (FIGURE3A) and average CO selectivity (FFIGURE 3B) across the examined voltage range when using the single-step melamine and PTFE treatment. The performance trends noted for CNT-PTFE-Mel samples prepared at different pyrolysis temperatures (FIGURES 3A and 3B), coupled with the magnetic properties of all but the 650 °C sample, demonstrate that disadvantageous structural and surface chemical changes occur at higher temperatures. For example, the high N-doped character of 7.1 at.% for the 650 °C sample decreases to 1.8 at.% at 850 °C. Interestingly, the Ni content also decreases substantially for the 850 °C sample (from 0.6 to 0.2 at.%) which could be due to an increase in graphitization which reencapsulates the Ni-based NCs. However, it is believed that NC agglomeration or sintering is a result of the decreased defect sites (i.e., N-doped) on the carbon surface at higher temperatures. This is because, as will be discussed in the DFT-AIMD section, the Ni-based NCs bind and get stabilized around the N-doped defect sites. Contrastingly, the 550 °C pyrolysis condition is potentially a reaction-limited regime for the in-situ oxidation of carbon encapsulated Ni NPs in CNTs. This is supported by the maintenance of magnetic properties of Raw CNT.

[0135] Knowing that CNT -Mel-650 retains the majority of encapsulated impurity Ni NPs and generates surface-saturated metal nitride sites (FIGURE 2E) - effectively an M-N-C type SAC - its polarization results and selectivity performance were included for comparison. As can be seen in FIGURE 3C, CNT-PTFE-Mel-650 outperforms the SAC based CNT -Mel-650 in terms of activity and largely in terms of selectivity as well under H-cell testing (FIGURE 3D). CNT-PTFE-650 on the other hand exhibits no electroactivity as previously noted. The electrochemical activity (FIGURE 3E) and selectivity (FIGURE 3F) results for the IndCNT-PTFE-Mel-650 were a very close match to those attained for CNT-PTFE-Mel-650. These findings correlate well with homogenously distributed Ni- based NC character observed for IndCNT-PTFE-Mel-650 in HAADF-STEM. The slight decrease in performance is likely a result of a slight difference in active sites (the metal impurity compositions are more complex), the diameters of nanotubes, and other potential factors that could be explored in future work. Moreover, and to indirectly study the effects Mel and PTFE have on the resultant electrochemical performance, control samples in the absence of melamine (i.e., CNT -PTFE) and of CNT-Mel-650 with a secondary PTFE- assisted pyrolysis ((CNT-Mel-650)+PTFE) were prepared at treatment temperatures of 650 and 850 °C. Although both (CNT-Mel-650)+PTFE samples exhibited comparable performance to CNT-Mel-650 (FIGURE 3C), especially between -0.59 and -0.77 V (vs.RHE), they did not match the electrochemical performance of CNT-PTFE-Mel-650. It suggests that the original structure of CNT -Mel-650 (i.e., SAC) is retained, and that PTFE alone cannot etch out Ni NPs from CNTs. In the absence of an N-source for doping, both control CNT -PTFE samples were not electroactive towards eCChRR.

[0136] Upon optimizing the synthesis procedure to remove encapsulated Ni NPs and in-situ reutilizing them as surface-bound electroactive Ni-based NCs, the eCCERR activity of said NCs was investigated in a more industrial-relevant flow-cell setup. Under flow-cell conditions, both CNT-Mel-650 (SAC) and CNT-PTFE-Mel-650 (NC) exhibited similar eCCERR performance until a CO partial current density of approximately 500 mA cm’2. Notably, the Faradaic efficiency towards CO (FEco) remained above 98% up to a current density of 700 mA cm’2(FIGURE 4A). However, a clear improvement in performance is observed for the NC system at 800 mA cm’2whereby a 2.7-times FEco is retained for the NC. This suggests either more intrinsically active sites or more facile transport of the feed CO2 to the active sites under transport-limited domains. To confirm this, a more transport and kinetically limiting regime was examined by feeding 50 vol.% of CO2 (FIGURE 4B) to the flow-cell instead of the standard 100 vol.% CO2 (FIGURE 4A). More obvious enhancements in both activity and selectivity are noted for the NC based catalyst. As shown in FIGURE 4B, the NC system sustains > 97% FEco at 200 and 400 mA cm’2, respectively, compared with drop to 24% FEco for its SAC counterpart at 400 mA cm’2. Moreover, an average of 22% lower voltage requirement is recorded for the NC catalyst relative to its SAC counterpart under the same transport limiting 50 vol.% feed CO2 concentration. As far as the IndCNT sample performance is concerned under flowcell testing, the IndCNT -PTFE-Mel-650 sample used on average 22% higher overpotential to sustain a given current density (FIGURE 4C) compared to CNT-PTFE-Mel-650. However, the resultant FEco between 200 and 600 mA cm’2were very comparable (> 98%) to that of CNT-PTFE-Mel-650 (FIGURE 4C) suggesting very similar active sites are present in both samples. The reason behind why IndCNT -PTFE-Mel-650 shows slightly lower activity than CNT-PTFE-Mel-650, is elaborated on in the DFT section with theoretical backing from intrinsic site activities and XPS nitridation ratios between the samples.

[0137] The electrochemical stability of the best performing CNT-PTFE-Mel-650 was showcased in its ability to sustain an industrially relevant current density of 100 mA cm’2for over 60 hours (In an embodiment, the carbon nanomaterials according toembodiments of the present disclosure comprises metal doping, such as Ni and Fe doping,). The observed increase in voltage is commonly attributed to the decay in the conductivity and buffering capacity of the electrolyte, as confirmed by restoration of voltage requirement with electrolyte renewal. However, the maintenance of FEco serves as a more reliable indicator of the stability of the active sites. With an average FEco of 95.5% for the entire electrolysis duration, it is safe to ascertain the resilience of the developed electrocatalyst under the testing conditions.

[0138] Electrochemical activity is typically determined based on the overpotential to sustain a given current density. Therefore, the onset overpotential (r|o) was taken as an activity metric to compare with contemporary benchmark electrocatalysts. To that end, potentiostat linear sweep voltammetry (LSV) polarization curves were taken using a Gamry Reference 3000 potentiostat on both the CNT-PTFE-Mel-650 (NC) and the CNT- Mel-650 (SAC). Results in FIGURES 5A demonstrate ultra-low onset overpotentials (r|o) of -19 and -37 mV, respectively, for the NC and SAC type catalysts prepared in this work. This is very competitive with the reported r|0values for NiNx / NCNT and PyPBI@Au / MWNT, which are considered contemporary benchmark electrocatalysts for CO production from eCO2RR, requiring -16 mV and -22 mV, respectively. Further, at a high current density of 500 mA cm'2, it is evident that the NC catalyst requires about 32% lower overpotential compared to the SAC catalyst.

[0139] To probe whether this is due to higher intrinsic activity of the NC compared to the SAC active sites, normalization of the current density with the electrochemical active surface area (ECSA) was performed. The double layer capacitances (Cai) for both CNT -Mel-650 and CNT-PTFE-Mel-650 are shown in FIGURE 5B, calculated from cyclic voltammetry versus scan-rate data. ECSA is obtained as the ratio between Cai and the solution capacitance (Cs), which is taken as 0.04 mF cm'2for the 1.0 M KOH used. Nevertheless, due to the linear proportionality between Cdi and ECSA, it is clear from FIGURE 5B that the CNT-PTFE-Mel-650 has about 78% higher ECSA compared to the SAC character of CNT-Mel-650. Furthermore, upon normalizing the recorded current with the ECSA, the resultant polarization curves in FIGURE 5C demonstrate a subtle improvement in intrinsic activity of the SAC sites relative to the Ni- based NC. However, despite the slightly lower intrinsic activity per active site for the NC catalyst, its significantly higher active site density compared to SACs resulted in a more favorable overall catalytic performance. Additionally, the 5.5x higher nitrogen dopingcontent in the NC catalyst induces polarization of the surrounding carbon atoms, a phenomenon known to facilitate the formation of additional active sites. This synergistic effect underscores the enhanced capability of NCs for eCCERR, making them a promising candidate for scalable and efficient CO2conversion.

[0140] AIMD and DFT Calculations

[0141] After confirming that the structure of the ~1 nm NCs was predominantly NisN (based on a coordination number of 2.5 ± 0.8 from EXAFS Ni X-edge fitting, ab initio molecular dynamics (AIMD) simulations were performed at the synthesis temperature of 923 K (650 °C). Considering the Ni species originate from the 10-50 nm encapsulated NPs, these simulations aimed to investigate whether the NCs would migrate and potentially break down into smaller NCs along the surface of N-doped CNTs with longer synthesis times. To confirm that the modeled structure was analogous to the experimentally designed catalyst, ab initio XANES calculations were initially performed on a fully relaxed NisN moiety to compare the resultant Gaussian broadened spectra (FIGURE 6A) to that experimentally measured (FIGURE 2E). As denoted in FIGURE 6A, the vertical lines at points 3 and 4 correspond to a registered shift in peak location from the experimentally obtained Ni X-edge XANES spectra, which suggested a combination of both NisN and metallic Ni-NCs in the experimental sample. Therefore, both metallic Ni and NisNNCs of 1 nm in diameter were modeled and used for both AIMD and DFT calculations. Considering the distribution of surface N species based on XPS for the experimental CNT- PTFE-Mel-650 sample, two configurational geometries for N-doped sp2carbon surfaces were generated to mimic the same ratio of pyrrolic, pyridinic, and graphitic. Further, both NisN NCs and Ni-NCs were allowed to relax atop these two surface configurations during AIMD calculations at 923 K for 30 ps, with 1 fs per timestep. The evolution of the modeled system for the case of NisN NC under configuration 1 (Configl) is presented in FIGURE 6B. Here, Ni atoms from the NC form anchoring sites with undercoordinated pyrrolic and pyridinic N defect sites. The formation energy (EA™) of all 4 considered systems were obtained and defined as the energy difference between the final structure containing both the NC and the N-doped carbon substrate (Erot) and the sum of the individual free-standing NC (ENC) and N-doped carbon (Esubstrate); EForm = Eibt - (ENC + Esubstrate). The corresponding EForm for the systems considered are presented in FIGURE 6C and show a clear thermodynamic favorability for the aforementioned system (NisN NC under Configl) shown in FIGURE 6B. Interestingly, both Ni-based NC systems showed a very similarformation energy stabilization pathway and reached close final EForm at 30 ps as NisN NC under Config2 of theN-doped carbon substrate. Since experimental characterization cannot distinguish between configurations 1 and 2, collectively identifying both as NisN-NC (or Ni-NC), it was attempted to elucidate computationally, which configuration is more stable. This lack of specificity hinders the accuracy of adsorption calculations, which rely on identifying the more thermodynamically favorable system (i.e. NisN Configl). Since both metallic Ni-NC configurations show similar formation energies, it was concluded that the adsorption energy calculations could be performed on either configuration, as both are equally to form.

[0142] In addition, DFT assessment was utilized to gain an insight into the thermodynamic favorability of different eCCERR to CO reaction intermediates adsorbed atop the presented potential active sites. Herein, four primary structures were considered, namely Ni-NC and NisN-NC (on CNT-PTFE-Mel-650) as well as the control NiN4 and FeN4 SACs (on CNT -Mel-650). For the NCs, all potential active sites atop the nanocluster geometry were considered, including hollow, bridge, and top sites. Additionally, different interfacial locations between the NC and the N-doped substrate were also evaluated. Based on the known reaction pathway of eCCERR to CO, the *CO2 to *COOH steps are generally considered to be the rate-limiting step that affects the experimental overpotential. Therefore, lowering the thermodynamic jump of this step and maintaining a near thermoneutral free energy change is considered ideal. As shown in FIGURE 6D, bothNiN4 and FeN4 SAC show limiting AG*COOH of 1.47 eV for NiN4 and strong adsorption (weak desorption) of *CO on FeN4 (AG*co of -0.774 eV) - both in agreement with prior reports. Calculated results in the same figure for Ni-NCs with active sites along the interface between the anchored NC and substrate, suggest that although the free energy at the *COOH step is smaller than NiN4 (0.47 eV for Ni-NC), the stronger adsorption of CO2 causes the energetic jump between two steps to be even more endothermic (1.87 eV) than NiN4 to *COOH (1.67 eV). Usefully, the Ni top site on the surface of the metallic Ni-NC is more thermodynamically favorable for both the *CCE and *COOH steps than active sites at the interface for this system. However, all plausible active sites for metallic Ni-NCs have very strong *CO adsorption causing the desorption of the product to be the rate-limiting step rather than the *COOH formation step. Interestingly, nitrogen coordination to Ni in both the NiN4 SAC and the NisN NC pathways does not suffer this limitation and has spontaneous desorption energetics. Specifically, the NisN NC system seems to beextremely electroactive towards CO production based on the calculated thermodynamic profile in FIGURE 6D. Near thermoneutral adsorption energies are achieved across the board when considering the reaction to proceed on Ni top sites for both CO2 and *COOH steps and on the neighboring N site for the CO desorption. This means that an ideal system with only NisN NCs should have higher ECSA-normalized activity compared to Ni / Fe SACs. Relating back to the higher ECSA-normalized activity for CNT-Mel-650 (NiN4 / FeN4) than CNT-PTFE-Mel-650 in FIGURE 5C, however suggests that a pronounced negative effect on the apparent ECSA-normalized activity from the metallic Ni-NCs in CNT-PTFE-Mel-650 is evident. This further supports experimental characterization findings that metallic Ni and NisNNCs are present on the surface of CNT- PTFE-Mel-650.

[0143] The Gibbs free energy changes for the competing HER was also calculated on the above mentioned surfaces to compare the resultant difference in limiting potentials (AUL) between eCCERR to CO (UL(CCO2RR)) and HER (UL(HER)). The UL is considered to be a directly proportional thermodynamic indicator of eCO2RR selectivity, without kinetic or transport considerations. As demonstrated in FIGURE 6E, the corresponding AUL for both NiN4 SAC and NisN NC are very selective towards eCO2RR. The latter case is particularly true when neighboring Ni and N top sites on the NisN are considered for the reaction pathway, inferring that the *CO2 and *COOH steps occur on the Ni top site and *CO desorption occurs on the neighboring N site (FIGURE 6D). Interestingly, metallic Ni-NCs are not considered to be selective towards eCCERR. This analysis can be linked to the comparison between the CNT-PTFE-Mel-650 and IndCNT- PTFE-Mel-650 samples. The higher Ni to N content in the IndCNT-PTFE-Mel-650 sample corresponds to a lower ratio of NisN to Ni-NCs, which leads to lower eCCERR selectivity, as confirmed experimentally (FIGURE 4D). Charge density difference plots were calculated and presented in FIGURE 6F to showcase the strong charge accumulation around the vicinity of both Ni and N sites towards *COOH and *CO steps, respectively, explaining the favorable thermodynamic adsorption / desorption behavior of eCCERR intermediates on NisN.EXAMPLE 2: HER PERFORMANCE OF CNT-MEL-PTFE-650

[0144] The present Example shows electrochemical performance of carbon nanomaterials prepared as described in Example 1 towards the cathodic hydrogen evolutionreaction (HER) in a standard H-cell under acidic conditions (0.5 M H2SO4). Hydrogen is produced from the reduction of water molecules (in neutral pH) or protons (in acidic conditions).

[0145] In a typical experiment, the powder catalyst of CNT-Mel-PTFE-650 was spray coated on a 1 x 1 cm2nickel foam (NF) substrate, which was initially cleaned through sonication in isopropyl alcohol, until a catalyst loading of 1 mg cm'2was attained. As a comparison, a commercially available catalyst, 40 wt.% Pt on Vulcan carbon (Pt / C) powder, was spray coated on NF following the same procedure. FIGURE 9A showcases the standard performance of commercial benchmark 40 wt.% Pt / C in acidic conditions towards the HER. All polarization results in FIGURE 9A were 90% iR corrected through the solution resistances (Ru) attained from Nyquist plots (FIGURE 9B). As can be seen, 40 wt.% Pt / C required an overpotential at 10 mA cm'2(rpo) and 100 mA cm'2(rpoo) corresponding to 50 and -160 mV, respectively, as commonly reported. Interestingly, the carbon nanomaterial (CNT-Mel-PTFE-650) sample initially demonstrated a current response poorer than Pt / C. However, after activation (performing 10 cycles of cyclic voltammetry) a notable shift is observed. Namely, upon activation (solid polarization curve in FIGURE 9A) 27 and 178 mV of overpotential were used at 10 and 100 mA cm'2, respectively. At a higher, and commercially relevant, current density of 300 mA cm'2the developed catalyst requires - 13% lower overpotential compared to Pt / C and is able to sustain higher current densities yet. Based on the smaller semi-circles (charge-transfer resistance) from Nyquist plots in FIGURE 9B for Pt / C, it is evident that conductivity is not the governing factor for the enhanced performance of CNT-Mel-PTFE-650. Therefore, cyclic voltammograms versus scan rates were performed in a non-Faradaic potential window (1.1 to 1.2 V vs. RHE) on both samples (FIGURES 9C and 9D), followed by plotting the scan rate versus change in central current densities (FIGURE 9E). The slope of the latter plot (FIGURE 9E) corresponds to the double-layer capacitance (Cai) of the samples. Knowing the solution capacitance of the alkaline electrolyte (1.0 M KOH) these measurements were undertaken in (Cs = 0.04 mF cm'2), the electrochemical active surface area (ECSA) can be determined. Normalizing the attained current in FIGURE 9A by the EC SA determined from FIGURE 9E results in the ECSA-normalized current density polarization curves in FIGURE 9F which are a direct experimental indicator of the intrinsic activity of each sample’s active sites. As can be seen in FIGURE 9F, the more electroactiveresponse of CNT-Mel-PTFE-650 under ECSA normalization suggests higher intrinsic activity of their active sites.

[0146] Moreover, density functional theory (DFT) calculations were performed on geometrically stabilized structures of the CNT-Mel-PTFE-650 to obtain the Gibbs binding energy for proton absorption (AGH*) - a known thermodynamic indicator for HER activity. As disclosed, material characterizations and ab initio molecular dynamics (AIMD) calculations indicate a combination of NisN and metallic Ni nanocluster structures embedded within surface N-defect sites on the CNTs. Furthermore, contemporary single (SAC) and dual-atom (DAC) M-N-C type structures were also calculated for comparison. Water molecules above the modeled catalytic slab were included to factor for water solvation effects. Typically, the closer to thermoneutrality (0 eV) AGH* is the better the activity. As shown in FIGURE 9G, several Ni -based SAC and DAC structures theoretically outperform the activity of the NisN nanoclusters, albeit by a close margin. Moreover, the showcased DFT results do not include kinetic considerations from hidden transition states or transport limitations. Therefore, this suggests that the desired experimentally attained performance of the CNT-Mel-PTFE-650 is largely due to a combination of factors. These include high intrinsic activity for active sites and high active site density. Lastly, the stability of the CNT-Mel-PTFE-650 was investigated under an initial current density of - 180 mA cm’2, corresponding to 233 mV of overpotential. As shown in FIGURE 9H, the brief 30-minutes chronopotentiometry test demonstrates a continued activation in performance, wherein the final current density at the same applied overpotential is 190 mA

[0147] The term “about” means plus or minus 5% of the stated value.

[0148] The principles, representative embodiments, and modes of operation of the present disclosure have been described in the foregoing description. However, aspects of the present disclosure which are intended to be protected are not to be construed as limited to the particular embodiments disclosed. Further, the embodiments described herein are to be regarded as illustrative rather than restrictive. It will be appreciated that variations and changes may be made by others, and equivalents employed, without departing from the spirit of the present disclosure. Accordingly, it is expressly intended that all such variations, changes, and equivalents fall within the spirit and scope of the present disclosure, as claimed.

[0149] While illustrative embodiments have been illustrated and described, it will be appreciated that various changes can be made therein without departing from the spirit and scope of the invention.

Claims

CLAIMSThe embodiments of the invention in which an exclusive property or privilege is claimed are defined as follows:

1. A method for removing metal nanoparticles from a carbon nanomaterial, comprising pyrolyzing a mixture comprising a carbon nanomaterial, a fluorine-containing polymer, and a nitrogen-containing compound at a pyrolyzation temperature sufficient to remove metal nanoparticles from the carbon nanomaterial to provide a carbon nanomaterial depleted in metal nanoparticles compared to the carbon nanomaterial prior to pyrolysis.

2. A method for increasing nitrogen doping level in a carbon nanomaterial, comprising pyrolyzing a mixture comprising a carbon nanomaterial, a fluorine-containing polymer, and a nitrogen-containing compound at a pyrolyzation temperature sufficient to increase nitrogen doping level in the carbon nanomaterial to provide a carbon nanomaterial with increased nitrogen doping level compared to the carbon nanomaterial prior to pyrolysis.

3. The method of Claims 1 or 2, wherein the nanomaterial is a carbon nanotube (SWCNT or MWCNT), a carbon nanofiber, a graphene, or a graphene nanoplatelet.

4. The method of Claim 1, wherein the metal nanoparticles are selected from the group consisting of iron, nickel, cobalt, chromium, manganese, and copper nanoparticles, and mixtures thereof.

5. The method of Claims 1 or 2, wherein the fluorine-containing polymer is a polytetrafluoroethylene (PTFE).

6. The method of Claims 1 or 2, wherein the nitrogen-containing compound is melamine.

7. The method of Claims 1 or 2, wherein the pyrolyzation temperature is about 650°C.

8. The method of Claims 1 or 2, wherein the pyrolyzation temperature is in a range of about 550°C to about 750 °C.

9. The method of any one of Claims 1-8, wherein the carbon nanomaterial produced by the method is a non-magnetic carbon nanomaterial.

10. The method of any one of Claims 1-9, wherein the carbon nanomaterial produced by the method is doped with nitrogen to provide a nitrogen-functionalized carbon nanomaterial.

11. The method of Claim 10, wherein the nitrogen-functionalized carbon nanomaterial has improved water dispersion compared to the carbon nanomaterial prior to pyrolysis.

12. The method of Claim 10, wherein the nitrogen-functionalized carbon nanomaterial has a nitrogen doping level at a minimum of about 6 percent by weight based on the total weight of the carbon nanomaterial.

13. The method of any one of Claims 1-7, wherein the carbon nanomaterial produced by the method is not damaged during the process, as measured by transmission electron microscopy (TEM).

14. The method of any one of Claims 1-13, further comprising washing the carbon nanomaterial under mild acidic conditions to remove metal content from surfacebound nanoclusters, wherein the mild acid-washed carbon nanomaterial contains less than 0.5 at.% residual metal contaminants.

15. A carbon nanomaterial prepared by the method of anyone of Claims 1-14.

16. A carbon nanomaterial having less than about 1 percent by weight metal based on the total weight of the carbon nanomaterial and greater than about 6 percent by weight nitrogen based on the total weight of the carbon nanomaterial.

17. An article of manufacture that includes the carbon nanomaterial of Claim 16.

18. The article of manufacture of Claim 17, wherein the article is an electrode for CO2 reduction, N2 reduction, water splitting, rechargeable batteries, or fuel cells.

19. The article of manufacture of Claim 17, wherein the article is used in an application where metal nanoparticles in carbon nanomaterials have negative impact on their use.

20. The article of manufacture of Claim 17, wherein the article is used in an application where it requires carbon nanomaterials to be well dispersed in an aqueous system.

21. The article of manufacture of Claim 17, wherein the carbon nanomaterial is disposed on a metal foam.

22. The article of manufacture of Claim 17, wherein the carbon nanomaterial is disposed on gas diffusion layer.

Images