Field

The invention relates to nitrogen-containing carbon nanotube catalysts. The invention further relates to fuel cells that comprise carbon nanotube catalysts, and methods of catalysing the oxygen reduction reaction using such catalysts.

Introduction

Carbon nanotubes can be described as sheets of graphene that are rolled so as to form a tubular structure. The resulting tubes can be single-walled, or they can be multi-walled where multiple tubes are grown in concentric circles. Carbon nanotubes generally have a thickness in the range of about 1 - 1000 nm, but can be as long as 10 mm or even longer.

Due to their high aspect ratio and low density, carbon nanotubes have very large surface areas, as high as hundreds of m ² /g. As a consequence, the tubes tend to form aggregates or bundles, but are typically dispersed prior to their use.

A fuel cell is an electrochemical cell that converts the chemical energy from a fuel into electricity through an electrochemical reaction of the fuel (often hydrogen) with oxygen or another oxidizing agent. Fuel cells are different from batteries in that they require a continuous source of fuel and oxygen to sustain the chemical reaction, whereas in a battery the chemical energy usually comes from metals and their ions or oxides that are already present in the battery. Fuel cells can produce electricity continuously for as long as fuel and oxygen are supplied.

A typical fuel cell consists of an anode, a cathode and an electrolyte that allows ions to move between the cathode and the anode of the fuel cell. At the anode, the presence of a catalyst causes the fuel to undergo an oxidation reaction, leading to the production of ions (usually hydrogen ions) and electrons. The ions then move through the electrolyte towards the cathode, and at the same time electrons flow from the anode to the cathode through an external circuit, thereby producing a direct current electricity. At the cathode, a second catalyst causes ions, electrons and oxygen to react to form water and possibly also other products, such as hydrogen peroxide, H ₂ O ₂ . Fuel cells are typically classified by the type of electrolyte they use, and also by their startup time, ranging from about 1 second for proton exchange membrane fuel cells (PEM fuel cells, or PEMFC) to about 10 minutes for solid oxide fuel cells (SOFC). Hydrogen fuel cells convert chemical energy to electrical energy by catalyzing the hydrogen oxidation reaction (HOR) at the anode, where hydrogen is converted into protons and electrons, and the oxygen reduction reaction (ORR) at the cathode, where water is formed. The most common catalyst used at the cathode is platinum (Pt), but its high cost, susceptibility to poisoning and poor durability has hampered large scale commercialization (Cheng, et al., J. Power Sources, 2007, 165:739-56).

Extensive efforts have been put into the development of an alternative ORR catalyst that is not dependent on the use of Pt, not only for fuel cell applications but also for metal-air batteries where the catalysts can be bifunctional and work for both the ORR and the reverse oxygen evolution reaction (OER). For example, nitrogen doped carbon materials, e.g., nitrogen doped carbon nanotubes (N-CNTs) and nitrogen doped graphene (N-graphene) have been shown to be active metal-free ORR catalysts with good durability and resistance to CO poisoning. Other heteroatoms than nitrogen, such as B, O, S and P have also been shown to boost the ORR activity of graphene (Jiao et al., J. Am. Chem. Soc., 2014, 136:4394-4403). Hydrogen fuel cell vehicles use hydrogen gas to power an electric motor. Unlike conventional vehicles which run on gasoline or diesel, fuel cell cars and trucks combine hydrogen and oxygen to produce electricity, which runs a motor. Since they’re powered entirely by electricity, fuel cell vehicles are considered electric vehicles - but unlike other such vehicles, their range and refueling processes are comparable to conventional cars and trucks. Replacing the state- of-the-art fuel cell catalyst platinum for a cheaper and abundant alternative would make the hydrogen economy viable.

In metal-air batteries, oxygen from air works as the cathode of the electrochemical cell, while the anode is typically provided by a metal, such as Lithium. In these electrochemical cells, oxygen is reduced and the metal is oxidized. In lithium-O ₂ systems, catalysts such as Pt/C, Pd/C and Au/C have been reported; these catalysts serve the purpose of reducing the overpotential between the oxygen reduction reaction (ORR) and the oxygen evolution reactions (OER), and promote the recycling of the cathode by facilitating the removal of the solid discharge product when the cell is charged. There is however a need for improved catalysts for use in air batteries, in particular catalysts that do not require use of Pt or other precious metals.

Summary

The present invention has as an objective to address the above-mentioned deficiencies, by providing improved catalysts that are useful in fuel cells and metal-air batteries. The present invention relates to carbon nanotubes with specific constraints on their structure. Thus, CNTs with particular physical parameters have been shown to be especially useful as catalysts, in particular as catalysts for the oxygen reduction reaction (ORR). In this context, the ORR can be either the 2 electron reduction of oxygen to hydrogen peroxide or the four electron reduction of oxygen to water.

In an aspect, the invention relates to an electrochemical catalyst for use in a fuel cell and/or in a metal-air battery, the electrochemical catalyst comprising a plurality of carbon nanotubes (CNTs), wherein the carbon nanotubes are characterized by having a diameter that is on average at least 7 Å.

The carbon nanotubes preferably contain at least one substitution. This can mean that the carbon nanotubes contain a substitution of at least one carbon atom in the carbon nanotubes with an atom selected from the group consisting of N, B, O, S and P. In an embodiment, the carbon nanotubes are nitrogen-containing carbon nanotubes (N-CNTs) that contain substitution of at least one carbon atom in the nanotube with a nitrogen atom. The diameter can be determined as the internal diameter of the nanotube. The diameter can alternatively be determined as the distance between the center of carbon atoms that are diametrically opposite one another, i.e. carbon atoms that are positioned so that a line drawn from the middle of one to the middle of the diametrically opposite carbon atoms passes through the center of the nanotube. Another aspect of the invention relates to a fuel cell, comprising (a) an anode assembly, for converting hydrogen gas to protons and electrons; (b) a cathode assembly, for converting oxygen to water; and (c) at least one proton conductive membrane disposed between the anode and the cathode, wherein the fuel cell is characterized in that at least the cathode assembly is provided with a coating comprising an electrochemical catalyst comprising carbon nanotubes that contain substitutions with at least one atom selected from N, B, O, S and P.

A further aspect of the invention relates to an apparatus, comprising a source of hydrogen and a fuel cell as described herein. The fuel cell can thus comprise (a) an anode assembly, for converting hydrogen gas to protons and electrons; (b) a cathode assembly, for converting oxygen to water; and (c) at least one proton conductive membrane disposed between the anode and the cathode, wherein the fuel cell is characterized in that at least the cathode assembly is provided with a coating comprising a carbon nanotube electrochemical catalyst wherein at least one of the carbon atoms in the carbon nanotubes is substituted by an atom selected from N, B, O, S and P. In yet another aspect the invention relates to a method of generating electricity. The method comprises steps of (i) performing a catalytic oxidation at an anode, where molecular hydrogen is converted to protons and electrons; (ii) performing a catalytic reduction at a cathode, where oxygen is converted to water; (iii) facilitating the passage of protons from the anode to the cathode; and (iv) facilitating the passage of electrons from the anode to the cathode via an external electron circuit, wherein at least the cathode is provided with a coating comprising a carbon nanotube electrochemical catalyst wherein at least one of the carbon atoms in the carbon nanotube is substituted by an atom selected from N, B, O, S and P.

The carbon nanotubes (CNTs) can contain substitutions with a single atomic species, i.e. atoms selected from N, O, B, S and P. The CNTs can also contain mixed substitutions, i.e. the CNTs can be substituted by two or more different atoms selected from N, O, B, S and P.

The above features along with additional details of the invention, are described further in the examples below, which are intended to further illustrate the invention but are not intended to limit its scope in any way.

Brief description of the drawings

The skilled person will understand that the drawings, described below, are for illustration purposes only. The drawings are not intended to limit the scope of the present disclosure in any way. FIG. 1 shows super cells used for N-doped (14,0) zig-zag tube (A), (14,7) chiral tube (B), and armchair (14,14) tube (C). Calculations were performed on the carbon atoms next to the nitrogen. C1 is the carbon atom to the top left of the nitrogen; C11 and C111 are the next atoms in clockwise direction from C1. C1 and C111 are equivalent positions for zig-zag tubes (n,0), and C1 and C11 are equivalent for armchair tubes (n,n). For chiral tubes, all three sites are unique.

FIG. 2 shows free energy diagram for ORR on nitrogen-doped (14,7) CNT at zero potential (U=0 V, upper solid line), equilibrium potential (U=1.23 V, lowest solid line), and at the onset potential (U=0.88 V, middle solid line). The ORR on an optimal catalyst is shown for comparison (dashed upper line). FIG. 3 shows the correlation between DFT calculated binding energies of intermediates. Upper figure: Binding energies of *OOH and *0 are plotted as a function of the binding energy of *OH. The relations are Lower figure: Gibbs free energy of intermediates adsorbed on the N-CNT with respect to clean surface and water as a function of tube diameter. The diameter is the distance from the middle of a carbon atom on one side of the tube to the middle of a carbon atom at the other side of the tube.

FIG. 4 shows (Upper figure): The thermodynamic limiting potential is estimated from the negative of the change in free energy (-DG) for the different reaction steps at the equilibrium potential (U=1.23) as a function of the *OH binding energy. The lines, calculated from the scaling relations, are used to guide the eye, whereas the explicit values are included for each tube. This gives the calculated overpotential for the ORR for each tube as a function of the *OH binding energy. The green -DG ₄ and red -DG ₁ lines are the potential limiting steps. (Lower figure): a zoom of the top of the volcano where N-CNTs indexes are indicated. The (14,7), (12,6) and (8,8) are close to the top of the volcano with OH binding energy of 0.95 eV and estimated overpotential of -0.35 V.

FIG. 5 shows (Upper figure): The limiting potential of each N-CNT against the diameter of the tube. The diameter is the distance from the middle of a carbon atom on one side of the tube to the middle of a carbon atom at the other side of the tube. (Lower figure): The limiting potential of each N-CNT against the chirality vectors (n,m).

Description

In the following, exemplary embodiments of the invention will be described, referring to the figures. These examples are provided to provide further understanding of the invention, without limiting its scope.

In the following description, a series of steps are described. The skilled person will appreciate that unless required by the context, the order of steps is not critical for the resulting configuration and its effect. Further, it will be apparent to the skilled person that irrespective of the order of steps, the presence or absence of time delay between steps, can be present between some or all of the described steps.

In the following description, the description of carbon nanotubes containing substitutions, sometimes referred to as doped carbon nanotubes or, in the case of substitutions with nitrogen, N-CNTs, applies equally to electrochemical catalysts as described and claimed herein, to fuel cells as described and claimed herein, to an apparatus using such doped carbon nanotubes, and to methods of generating electricity using such catalysts, as described and claimed herein. The present invention relates to carbon nanotubes (CNTs) that are useful as catalysts and have certain physical characteristics, in particular their three-dimensional geometry and their diameter. Pure carbon nanotubes (i.e., nanotubes that are not doped, i.e. do not contain substitutions of carbon atoms) have been used as catalyst support in fuel cells, including ORR fuel cells, where the nanotubes serve the role of providing structural support for the platinum catalyst used to catalyse the oxygen reduction reaction (Cha et al., J Power Sources, 2018, 401 :296-302).

Doping of carbon nanotubes, in particular N-doping of carbon nanotubes is known to lead to enhanced electronic properties of the nanotubes, and in particular enhanced electrocatalytic performance (Wei, Catalysts 2015, 5 1574-1602). However, to date, most of the research on ORR activity of N-doped carbon nanotubes has been focused on the effects of doping on the activity of the CNTs and less on their structure-activity relationship.

In the present context, the term“doping” refers to the intentional introduction of one or more foreign atomic species into carbon nanotubes. The doping can be performed using any suitable atoms, including one or more of B, O, S, N and P. This means that the respective atoms are intentionally incorporated in the carbon nanotubes, leading to formation of doped carbon nanotubes (doped CNTs). In some cases, this involves replacing one or more C atom in a carbon nanotube by an atom, e.g. an atom selected from N, O, S, B and P.

Accordingly, when doped, the carbon nanotubes (CNTs) contain a certain fraction of atoms selected from N, O, S, B and P, thus generating carbon nanotubes (CNTs) containing at least one substitution with atoms selected from N, O, S, B and P. In some embodiments, the carbon nanotubes are substituted by one or more nitrogen atom, leading to the formation of nitrogen- containing carbon nanotubes (N-CNTs).

The geometry of a single walled nanotube can be specified by two integers, (n,m), called the chiral vectors, where n ³ m, that describe the diameter and chirality of the tube. Tubes with chiral vector (n,0) are called zig-zag, (n,n) tubes are armchair, and other tubes are chiral. The geometric structure of CNTs heavily influences their electronic properties. CNTs can show metallic or semiconducting behavior as their band gap depends on the diameter and chirality of the tube, thus on (n,m). In general, all armchair tubes (n,n) are metallic and (n,m) tubes, that satisfy n-m = 3j where j is a non-zero integer, only have a small band gap and can be considered metallic at ambient temperature for most practical purposes. Other tubes are semiconductors. However, CNTs with a diameter less than 10 Å have sufficiently strong p* - s* hybridization effects to significantly alter their electronic structure resulting in reduced band gaps for some tubes. This property of small diameter CNTs, combined with nitrogen doping that has been shown to boost ORR activity on CNTs (Maldonade & Stevenson, J. Phys Chem B, 2005, 109:4707-16) makes them an interesting candidate for an ORR catalyst.

A novel method to account for the effect of applied potential in electrochemical reactions has been presented by Norskov et al., referred to the computational hydrogen electrode (CHE), where only the thermochemistry of adsorbed intermediates is taken into account using DFT calculations (Norskov et al, J. Phys Chem B, 2004, 108:17886-92). By using scaling relations of adsorbed intermediates, volcano plots can be constructed that estimates the overpotential needed for a given catalyst in order to eliminate all endergonic steps and make all the proton- electron transfer steps exergonic. The CHE model has been extremely successful to obtain trends in catalytic activity where it has been applied for several electrochemical reactions on various types of catalysts, such as oxygen evolution reactions (OER) on transition metal oxide (TMO) surfaces, ORR on transition metal (TM) surfaces and on N-graphene, N ₂ electroreduction to ammonia on TM surfaces, transition metal nitride (TMN) surfaces and TMO surfaces and CO ₂ reduction reaction (CO ₂ RR) on TM and TMO surfaces. This methodology is able to either predict or confirm the overpotential needed to start these electrochemical reactions in all cases.

For relatively simple reactions such as the hydrogen evolution reaction (HER) or the ORR the overpotential can be used as a good measure of the catalytic activity. It should be noted that the present inventors have seen that in the case of CO ₂ RR that despite that the overpotential can be captured quite accurately within this methodology, the experimental trends in activity and selectivity cannot be reproduced. This is because CO ₂ RR has a complex reaction network, where competition between reaction steps towards different intermediates or products play a crucial role, and there, more detailed model and methods are needed, where proton-transfer barriers are calculated as a function of explicitly applied potential. As described above, this simple methodology (CHE) has been shown to work well for ORR on several catalytic materials, including N-graphene, and can be used to screen for new catalysts.

To date, most of the research on ORR activity of N-CNTs, has been focused on the effect of nitrogen doping of the CNTs and less on their structure. The present invention is a result of using systematic DFT calculations and the CHE model to obtain trends in the ORR activity with changing chirality and diameter of the CNTs while keeping the N-dopant (N content) concentration low and constant to help find the optimal size and shape of tubes to enhance the ORR activity.

As described in more detail in the below Example, DFT calculations have been performed for the ORR on single-walled N-CNTs with different chiralities, ranging from (4,0) to (20,10), having diameters of 3.3 Å - 22 Å. Results show that there is correlation between binding energies of different intermediates with respect to each other, as well as tube diameter. Volcano plot of the estimated overpotential of the N-CNTs with respect to the binding energy of the *OH intermediate based on those correlations demonstrate that the overpotential is generally determined by the first or the last two reaction steps.

Certain N-CNTs have been found to exhibit particularly low overpotential. Thus, a total of 46 N-CNTs were found to have an overpotential that is less than the overpotential for nitrogen doped graphene (0.57 V) for the oxygen reduction reaction (ORR). A set of 22 tested N-CNTs were found to have an overpotential for the ORR less than for Ft (i.e., less than 0.45 V). It was found that three tubes, (8,8), (12,6) and (14,7) have notably low overpotentials of 0.37 V, 0.36 V and 0.35 V, respectively.

The N-CNTs that are the best catalysts were found to have the common feature of having a diameter that is in general at least 8 Å (see FIG. 5).

The overpotential of zig-zag tubes ranging from (4,0) to (20,0), and other tubes with diameter less than 10 Å, was found to be determined by strong binding of *0 and *OH intermediates to the surface. In most cases, the reduction of *OH to H ₂ O was found to be potential limiting, except for a few tubes where the reduction of *0 to *OH was potential limiting. The overpotential of armchair and chiral tubes larger than 16 Å in diameter is a result of week binding of the *00 H intermediate. In the diameter range of 10-16 Å a compromise of these potential limiting steps (PLSs; with either the first or the last protonation step being potential limiting) is found, which results in tubes with intermediate binding of all intermediates and those tubes have the lowest overpotential for ORR. In general, binding energy of the intermediates decreases with increasing tube diameter. The observed trends can in particular be used to design the optimum size and shape of N-CNTs to be used as cathode material for the oxygen reduction reaction in fuel cells.

The electrochemical catalyst in accordance with the invention can in general contain CNTs having any geometric configuration, such as a zig-zag configuration, a chiral configuration or an armchair configuration. In a preferred embodiment, the CNTs have an armchair or a chiral configuration. In some embodiments, the CNTs have an armchair configuration. In some other embodiments, the CNTs have a chiral configuration.

The CNTs can have a diameter that is on average at least 7 Å, at least 8 Å, at least 9 Å, or at least 10 Å. The diameter can be less than 24 Å, less than 23 Å, less than 22 Å, less than 21 Å, less than 20 Å, less than 19 Å, less than 18 Å, less than 17 Å, less than 16 Å or less than 15 Å. The CNTs can have a diameter that is on average less than 24 Å, preferably less than 22 Å, more preferably less than 20 Å, more preferably less than 18 Å, more preferably less than 17 Å, even more preferably less than 16 Å.

The CNTs can have a diameter that is on average in the range of 8 - 22 Å, preferably in the range of 8 - 22 Å, in the range of 8 - 20 Å, in the range of 9 - 20 Å, in the range of 9 - 18 Å, in the range of 9 - 16 Å, in the range of 10 - 20 Å, in the range of 10 - 18 Å, in the range of 10 - 17 Å, or in the range of 10 - 16 Å.

The diameter can be determined as the distance from the middle of a first carbon atom on one side of the tube to the middle of a second carbon atom at the other side of the tube, i.e. the middle of a second carbon atom that is diametrically opposite to the first carbon atom.

The foregoing specified diameter range is intended to mean that for a plurality of tubes, the average diameter of the tubes is within the specified range. Individual tubes may all have the same diameter, or the individual tubes may have different diameters that on average is within the stated range.

The CNTs can have any desired length, as is known in the art. For example, the nanotubes can be as long as 1000 mm in length or even longer. The nanotubes can also be much shorter, for example in the range of about 0.1 to about 100 mm.

The content of substituted atoms (N, O, S, B and/or P) in the CNTs can in general be at least 0.05%, such as at least 0.1 %, at least 0.2%, at least 0.3%, at least 0.4%, at least 0.5%, at least 0.7%, or at least 1.0%. The content in the can be as high as 20%, as high as 19%, as high as 18%, as high as 17%, as high as 16%, as high as 15%, as high as 14%, as high as 13%, as high as 12%, as high as 11 % or as high as 10%.

The carbon nanotubes can accordingly have a content of substituted atoms (e.g., N, O, S, B and/or P) that is on average less than 20%, less than 15 %, less than 10%, less than 5%, less than 3%, less than 1.5 %, or less than 1%.

The nanotubes can in general be single-wall nanotubes or the nanotubes can be multiwalled, i.e. containing two or more nested tubes, wherein an innermost tube is surrounded by one or more outer tubes having a larger diameter so that the inner tube(s) can fit within the outer tube(s). It can however be preferable that the carbon nanotube be a single-wall nanotube. Substituted carbon nanotubes have a certain amount (ratio) of incorporated atoms selected from N, O, S, B and/or P in the nanotubes. In certain embodiments, the atoms are incorporated in the tubes by substituting carbon atoms in the carbon nanotubes by the respective atoms. When the carbon nanotubes contain nitrogen substitutions, the resulting nitrogen-containing carbon nanotubes (N-CNTs) can be selected from the group consisting of (14,7), (12,6), (8,8), (9,6), (15,5), (16,8), (9,9), (10,4), (11 ,2), (14,2), (13,1), (20,0), (7,7), (10,10), (18,0), (12,3), (19,0), (12,4), (10,1), (18,9), (9,3), (11 ,11), (11 ,5), (20,10), (10,7), (15,6), (7,0), (8,5), (17,0), (12,12), (15,0), (16,0), (13,13), (16,4), (14,8), (14,14), (20,8), (10,5), (15,15), (6,6), (16,16),

(7,4), (14,0), (13,0), (20,5) and (12,0).

The N-CNTs can alternatively be selected from the group consisting of (14,7), (12,6), (8,8), (9,6), (15,5), (16,8), (9,9), (10,4), (11 ,2), (14,2), (13,1), (7,7), (10,10), (12,3), (12,4), (10,1), (18,9), (9,3), (11 ,11), (11 ,5), (20,10), (10,7), (15,6), (8,5), (17,0), (12,12), (13,13), (16,4), (14,8), (14,14), (20,8), (10,5), (15,15), (6,6), (16,16), (7,4) and (20,5).

In some embodiments, the N-CNTs are selected from the group consisting of (14,7), (12,6),

(8,8), (9,6), (15,5), (16,8), (9,9), (10,4), (11 ,2), (14,2), (13,1), (20,0), (7,7), (10,10), (18,0), (12,3), (19,0), (12,4), (10,1), (18,9), (9,3) and (11 ,11).

In some embodiments, the N-CNTs are selected from the group consisting of (14,7), (12,6),

(8,8), (9,6), (15,5), (16,8), (9,9), (10,4), (11 ,2), (14,2), (13,1), (7,7), (10,10), (12,3), (12,4), (10,1), (18,9), (9,3) and (11 ,11).

In some embodiments, the N-CNTs are selected from the group consisting of (14,7), (12,6),

(8,8), (9,6), (15,5) and (16,8).

In some embodiments, the N-CNTs are selected from the group consisting of (14,7), (12,6) and (8,8).

The substituted CNTs can have an overpotential for the oxygen reduction reaction that is less than 0.57 V, such as less than 0.55 V, less than 0.50 V, less than 0.45 V, or less than 0.40 V.

Carbon nanotubes can be prepared by methods such as chemical vapor deposition (CVD), thermal annealing under NH ₃ , plasma treatment with N ₂ , wet chemical treatment, laser ablation, arc discharge and molten salt intercalation. Incorporation of a dopant into CNTs can in general be performed by post-treatment (e.g., thermal annealing, plasma treatment or wet chemical treatment) or (more commonly) by incorporation during the synthesis of the CNT (i.e., in situ, such as by arc discharge, CVD, laser ablation).

The introduction of a dopant into CNTs can be achieved by adding the dopant as part of the catalyst or as part of the reactant (e.g., carbon) source. Dopants include heteroatoms such as N, B, O, S and P. When doped with nitrogen, the catalyst or the carbon source can include a source of nitrogen that will be incorporated into the resulting CNT during its synthesis by vapor deposition. Resulting doped CNTs can be characterized by methods known in the art, such as X-ray photoelectron spectroscopy (XPS), X-ray diffraction (XRD), Raman Spectroscopy

Doping of CNTs with heteroatoms results in a change in the CNT geometrical and electronic structure. When a heteroatom is inserted into the backbone of a CNT, the symmetry of the tube is modified and subsequently the structure and properties are altered. Since the heteroatom usually has a different number of valance electrons than the carbon atoms in the CNT, it affects the electronic properties of the tube. The substitutional doping of N has received much attention because major changes in hardness and electrical conductivity have been observed both theoretically and experimentally in N-CNTs (Willalpando-Paez, et al., 2006, Chem Phys Lett 424:345-52). N doping of CNTs has recently been considered as a feasible strategy to fine-tune the electronic properties of CNTs in a well-defined manner. Even small amounts of N incorporation can quite significantly alter the electronic transport properties within a CNT network (Nxumalo & Coville, 2010, Materials 3:2141-71).

Theoretical studies have revealed that doping affects the electronic transport of CNTs. It has been reported that insertion of N into CNTs results in an enhancement in conductivity and an improvement of transport and field emission properties of CNTs. This is believed to be due to the electron donor ability of the N atom that leads to the formation of a n-type semiconductor (Nxumalo & Coville, 2010, Materials 3:2141-71).

In general, incorporation of atoms such as N, B, S, O and/or P into CNTs can result in substitution of individual C atoms in the graphene structure with the respective atom, resulting in so-called graphitic substitution atoms (e.g., N-atoms in the case of N-CNTs). The incorporation can also result in generation of pyridinic substitution atoms, where there are carbon vacancies in the graphene structure surrounding one or more such incorporated pyridinic substitution atoms. Alternatively, the incorporation can lead to the generation of pyrrolic substitution atoms, where an atom such as N, B, S, O and/or P is introduced in a 5-membered ring structure in the graphene backbone. In the case of N-substitution, in the substitution can alternatively lead to the generation of oxidized nitrogen species (NO _x ). The CNTs in accordance with the present invention can have any one of these types of substitutions in the graphene backbone of the CNT. The CNTs can have a single type of substitution incorporated, i.e. a graphitic, pyridinic, pyrrolic (or NO _x in the case of N-CNTs) substitution, or they can have a mixture of two of these types of substitutions, or even a mixture of all these foregoing types. In a preferred embodiment, N-CNTs have predominantly graphitic N-atoms incorporated. A nanotube-containing coating can in principle comprise any mixture of a suitable substrate material and nanotubes. The substrate material can for example be a chemically inert material (i.e. a material that is inert when used in the fuel cell). The support material can preferably be a material that is amenable to being sprayed onto, or otherwise applicable in a thin coating, onto an electrode.

Substrate materials can for example be polymers, ceramic materials, metal alloys, or silicon- based materials (for example silicon wafers). The substrate materials can be dispersed in a suitable solvent prior to being applied to an electrode.

Due to their tendency to form aggregates, carbon nanotubes are usually dispersed prior to use. Their dispersion in one or more suitable substrate material (that can act as a dispersing agent) can therefore be beneficial for their intended use.

Alternatively, the CNTs can be used directly to coat one or more electrodes in a fuel cell. The CNTs can optionally be used in combination with graphene to provide a coating on the one or more electrode, for example a cathode.

The CNTs in accordance with the invention can alternatively, or additionally, be provided as composite materials in which one or more additional material have been provided on the - CNTs. For example, the CNTs can be provided with nanoparticles comprising one or more metallic, semiconducting and/or insulating materials on their surface (Jurn, et al., IEEE Int Conf Control Systems Comput Eng, 2014). The CNTs can also or alternatively be provided with one or more polymer coating, such a Nation coating, a titanium-silicon oxide coating, or other suitable coatings.

It is also possible to provide the CNT catalysts as composites with graphene. The graphene used in such composite materials can be a doped graphene, such as graphene doped with nitrogen (N-graphene). Ideal catalysts for use on the cathode side in fuel cells should ideally have a high electrical conductivity, excellent catalytic activity - in particular for the oxygen reduction reaction - a high surface area, high corrosion resistance and low cost. The CNT electrochemical catalysts described herein fulfil all of these characteristics, and are thus believed to be particularly useful as catalysts in fuel cells and/or in metal - air batteries, such as N-CNT batteries. A further advantage of these catalysts is that their functional properties are a direct consequence of their structural characteristics, i.e. diameter and three-dimensional configuration. No additional modification of the CNTs is needed to enhance their catalytic properties, unlike for example unmodified carbon nanotubes, that lack catalytic activities and have to be chemically modified to allow for incorporation of metal catalysts on their surface (examples include Ft nanotubes, Pt-NTs). Such chemical modification requires harsh conditions, such as treatment with a strong acid to create necessary functional groups on their surface (CO and/or COOH). Acid treatment however has the disadvantage of disturbing the structure of the NTs, which can impact their durability. The catalytic nanotubes of the present invention are free from these disadvantages, since they do not require any further modifications prior to use.

Exemplary embodiments of the invention include the following:

1. An electrochemical catalyst for use in a fuel cell and/or in metal-air battery, the electrochemical catalyst comprising a plurality of carbon nanotubes containing at least one N, B, S, O and/or P substitution, wherein the nanotubes are characterized by having an diameter that is on average at least 7 Å.

2. The electrochemical catalyst of the previous embodiment, wherein the carbon nanotubes have a diameter that is on average at least 8 Å, more preferably at least 9 Å, more preferably at least 10 Å.

3. The electrochemical catalyst of any one of the previous embodiments 1-2, wherein the carbon nanotubes have a diameter that is less than 24 Å, preferably less than 22 Å, more preferably less than 20 Å, more preferably less than 18 Å, more preferably less than 17 Å, even more preferably less than 16 Å.

4. The electrochemical catalyst of any one of the previous embodiments 1-3, wherein the carbon nanotubes have a diameter that is in the range of 7 - 22 Å, preferably in the range of 8 - 20 Å, more preferably in the range of 9 - 20 Å, more preferably in the range of 10 - 20 Å, more preferably in the range of 10 - 18 Å, more preferably in the range of 10 - 17 Å, even more preferably in the range of 10 - 16 Å.

5. The electrochemical catalyst of any one of the previous embodiments 1-4, wherein the nanotubes contain substitution of one or more carbon atoms in the carbon nanotubes by atoms selected from N, B, S, O and/or P.

6. The electrochemical catalyst of the previous embodiment, wherein the incorporated atoms are graphitic, pyridinic or pyrrolic.

7. The electrochemical catalyst of any one of the previous embodiments 1-6, wherein the carbon nanotubes have a content of substituted atoms that is on average at least 0.05%, preferably at least 0.1 %, more preferably at least 0.5%, even more preferably at least 1.0%.

8. The electrochemical catalyst of any one of the previous embodiments 1-7, wherein the carbon nanotubes have a content of substituted atoms that is on average less than 20%.

9. The electrochemical catalyst of any one of the previous embodiments 1-8, wherein the carbon nanotubes have a content of substituted atoms that is on average less than 15%, less than 10%, less than 5%, less than 3%, less than 1.5%, or less than 1%.

10. The electrochemical catalyst of any one of the preceding embodiments 1-9, wherein the carbon nanotubes contain nitrogen substitutions only.

11. The electrochemical catalyst of any one of the previous embodiments 1-10, wherein the electrochemical catalyst is provided in a hydrogen fuel cell.

12. The electrochemical catalyst of any one of the previous embodiments 1-11 , wherein the electrochemical catalyst is provided at the cathode of a fuel cell or an metal-air battery.

13. The electrochemical catalyst of any one of the previous embodiments 1-12, wherein the electrochemical catalyst catalyzes the oxygen reduction reaction (ORR) or the oxygen evolution reaction (OER).

14. The electrochemical catalyst of any one of the previous embodiments 1-13, wherein the catalyst comprises single-wall nitrogen-containing carbon nanotubes.

15. The electrochemical catalyst of any one of the previous embodiments 1-14, wherein the catalyst comprises carbon nanotubes having a chiral and/or armchair configuration.

16. The electrochemical catalyst of any one of the previous embodiments 1-15, wherein the catalyst comprises nitrogen-containing carbon nanotubes (N-CNTs) having a configuration selected from the group consisting of (14,7), (12,6), (8,8), (9,6), (15,5), (16,8), (9,9), (10,4), (11 ,2), (14,2), (13,1), (20,0), (7,7), (10,10), (18,0), (12,3), (19,0), (12,4), (10,1), (18,9), (9,3), (11 ,11), (11 ,5), (20,10), (10,7), (15,6), (7,0), (8,5), (17,0), (12,12), (15,0), (16,0), (13,13), (16,4), (14,8), (14,14), (20,8), (10,5), (15,15), (6,6), (16,16), (7,4), (14,0), (13,0), (20,5) and (12,0). 17. The electrochemical catalyst of any one of the previous embodiments 1-16, wherein the catalyst comprises nitrogen-containing carbon nanotubes having a configuration selected from the group consisting of (14,7), (12,6), (8,8), (9,6), (15,5), (16,8), (9,9), (10,4), (11 ,2), (14,2), (13,1), (20,0), (7,7), (10,10), (18,0), (12,3), (19,0), (12,4), (10,1), (18,9), (9,3) and (11 ,11).

18. The electrochemical catalyst of any one of the previous embodiments 1-17, wherein the catalyst comprises nitrogen-containing carbon nanotubes having a configuration selected from the group consisting of (14,7), (12,6) and (8.8).

19. The electrochemical catalyst of any one of the previous embodiments 1-18, wherein the catalyst comprises nitrogen-containing carbon nanotubes having a configuration selected from the group consisting of (14,7) and (12,6).

20. A fuel cell, comprising: a. an anode assembly, for converting hydrogen gas to protons and electrons; b. a cathode assembly, for converting oxygen to water; c. at least one proton conductive membrane disposed between the anode and the cathode, wherein the fuel cell is characterized in that at least the cathode assembly is provided with a coating comprising an electrochemical catalyst comprising carbon nanotubes containing at least one substitution with at least one atom selected from N, O, B, S and P wherein the carbon nanotubes have a diameter that is on average at least 7 Å.

21. The fuel cell of the previous embodiment, wherein the fuel cell further comprises at least one electrolyte provided between the cathode assembly and the anode assembly.

22. The fuel cell of the previous embodiment, wherein the anode assembly is coupled to at least one source of hydrogen gas.

23. The fuel cell of any one of the previous embodiments 20-22, wherein the cathode assembly comprises at least one electrode comprising a carrier and a coating layer disposed on the surface of the carrier, wherein the plurality of carbon nanotubes are provided in the coating layer. 24. The fuel cell of any one of the previous embodiments 20-23, wherein the fuel cell is adapted to catalyze an oxygen reduction reaction (ORR) wherein oxygen is reduced to produce water.

25. The fuel cell of any one of the previous embodiments 20-24, wherein the electrochemical catalyst is a catalyst as set forth in any one of the embodiments 1 to

19.

26. An apparatus, comprising: a. A source of hydrogen b. A fuel cell, comprising: i. an anode assembly, for converting hydrogen gas to protons and electrons; ii. a cathode assembly, for converting oxygen to water; iii. at least one proton conductive membrane disposed between the

anode and the cathode, wherein the fuel cell is characterized in that at least the cathode assembly is provided with a coating comprising an electrochemical catalyst that comprises carbon nanotubes containing substitutions with at least one atom selected from N, O, B, S and P, wherein the carbon nanotubes have a diameter that is on average at least 7 Å.

27. The apparatus of the previous embodiment, wherein the carbon nanotube electrochemical catalyst is a catalyst as set forth in any one of the embodiments 1 to 19.

28. A method of generating electricity, comprising: a. performing a catalytic oxidation at an anode, where molecular hydrogen is converted to protons and electrons; b. performing a catalytic reduction at a cathode, where oxygen is converted to water; c. facilitating the passage of protons from the anode to the cathode; and d. facilitating the passage of electrons from the anode to the cathode via an external electron circuit; wherein at least the cathode is provided with a coating comprising an electrochemical catalyst that comprises carbon nanotubes containing substitutions with at least one atom selected from N, O, B, S and P, wherein the carbon nanotubes have a diameter that is on average at least 7 Å.

29. The method of the previous embodiment, wherein the electrochemical catalyst is a catalyst as set forth in any one of the embodiments 1 to 19.

30. The method of any one of the previous two embodiments 28-29, wherein the passage of protons from the anode to the cathode is facilitated by providing a proton conductive membrane that separates the anode from the cathode.

As used herein, including in the claims, singular forms of terms are to be construed as also including the plural form and vice versa, unless the context indicates otherwise. Thus, it should be noted that as used herein, the singular forms“a,”“an,” and“the” include plural references unless the context clearly dictates otherwise.

Throughout the description and claims, the terms “comprise”, “including”, “having”, and “contain” and their variations should be understood as meaning“including but not limited to” and are not intended to exclude other components.

The present invention also covers the exact terms, features, values and ranges etc. In case these terms, features, values and ranges etc. are used in conjunction with terms such as about, around, generally, substantially, essentially, at least etc. (i.e.,“about 3” shall also cover exactly 3 or“substantially constant” shall also cover exactly constant).

The term“at least one” should be understood as meaning“one or more”, and therefore includes both embodiments that include one or multiple components. Furthermore, dependent claims that refer to independent claims that describe features with“at least one” have the same meaning, both when the feature is referred to as“the” and“the at least one”.

It will be appreciated that variations to the foregoing embodiments of the invention can be made while still falling within the scope of the invention. Features disclosed in the specification, unless stated otherwise, can be replaced by alternative features serving the same, equivalent or similar purpose. Thus, unless stated otherwise, each feature disclosed represents one example of a generic series of equivalent or similar features.

Use of exemplary language, such as“for instance”,“such as”,“for example” and the like, is merely intended to better illustrate the invention and does not indicate a limitation on the scope of the invention unless so claimed. Any steps described in the specification may be performed in any order or simultaneously, unless the context clearly indicates otherwise.

All of the features and/or steps disclosed in the specification can be combined in any combination, except for combinations where at least some of the features and/or steps are mutually exclusive. In particular, preferred features of the invention are applicable to all aspects of the invention and may be used in any combination.

Examples

In the following, non-limiting experimental examples in accordance with the invention are described.

Example 1

Methodology

The overall redox reaction where O ₂ is reduced and H ₂ is oxidized into water is given by:

and could in principle follow several mechanisms. In this systematic study, we only consider the associative mechanism of the ORR half-cell reaction since that reaction mechanism has been suggested to be the preferred mechanism for ORR on N-graphene ^8,19-22 and since the dissociative pathway has been shown to be unlikely on N-CNTs, due to high energy barriers for O ₂ dissociation. ⁴¹

The associative mechanism:

where * denotes the surface sites where the species adsorb onto.

All DFT calculations were performed using revised Perdew-Burke-Ernzerhof generalized gradient approximation. ⁴² A 1 x 1 x 4 Monkhorst-Pack k-point sampling was used for all calculations and a wave cut-off energy of 400 eV was used to represent the valence electrons with a PAW ⁴³ representation of the core electrons as implemented in the VASP code. ^44-47 The calculations were performed on periodically repeated one-dimensional (1 D) unit cells of single- walled CNTs where one carbon atom was replaced by a nitrogen atom, forming graphitic N- CNTs, and minimum of 16 Å of vacuum was used to avoid inter-tubular interactions. Figure 1 shows 1 D unit cells for zig-zag, armchair and chiral N-CNTs. As the 1 D unit cells for chiral CNTs are usually much larger than for armchair or zig-zag CNTs, ²⁴ the super cells were set up so there was always a minimum of 7 carbon atoms, or 8.5 Å, between nitrogen dopants, see Figure 1. This results in a maximum of 3.125% doping for the smallest unit cell, but most tubes had less than 1 % doping. By keeping the N doping low minimizes possible concentration effects and simulates the effect of a single dopant. Intermediates bind strongest to the carbon atoms next to the nitrogen dopant, which is in accordance to previous calculations on graphitic- N-doped graphene. ¹⁹

The free energy for all intermediate states in eqs. (2)-(5) were calculated for a range of N- CNTs, with intermediates adsorbed on C1 , C11 or C1 11 in Figure 1. The free energy change for elementary reactions in eqs. (2)-(5) was calculated by:

where the reaction energy, DE, was calculated using DFT. DZRE and DS, the difference in zero-point energy and entropy, respectively, were calculated using results from Studt ¹⁹ for the ORR on N-graphene and the stabilization energy from the water layer, included using

results from Yu et al. ²⁰ for the ORR on N-graphene.

The effect of the applied potential, U, was estimated using a simple model (CHE) by Norskov et al., ²⁷ for all reactions involving a proton-electron transfer by shifting the free energy by -neU, where n is the number of electrons in the electrochemical reaction: and DG(O) is calculated with eq. (6), which is the free energy change at U = 0 V (all voltages are given with respect to the reversible hydrogen electrode, RHE). The calculations were done at standard conditions of 298 K, 1 bar Ha and pH=0. With this methodology the potential limiting steps (PLS _s ) can be determined and the overpotential (or the limiting potential) estimated for different catalysts. It should be noted that overpotentials, which we focus on in this work, are independent on pH values. This model has been used to describe the ORR on N-doped graphene ¹⁹ and transition metals, ²⁷ as well as electrolysis of water on transition-metal oxides. ²⁹

Results The free energy diagram for the ORR reaction on a C195N (14,7) N-CNT tube is shown for several applied potentials in Figure 2. Each step of the reduction reaction involves a transfer of a proton-electron pair. The optimal ORR catalyst, with no overpotential, would have all reaction steps equally spaced downhill by 1.23 eV at U = 0 V or thermoneutral at the equilibrium potential of 1.23 V, depicted by a dashed line in Figure 2. The first step, formation of *OOH species on the surface of C195N (14,7) N-CNT, is only downhill by -0.88 eV, so there is under binding of the *00 H species. The reduction from *00 H to *0 and from *0 to *OH is downhill by -2.04 eV and -1.12 eV respectively, resulting in an over binding of those species. The final reduction step from *OH to desorbed H ₂ O is downhill, also by only -0.88 eV. Therefore, the PLSs are the reduction of O ₂ to *00 H and *OH to H ₂ O on (14,7) N-CNTs. Overpotential of only 0.35 V is predicted on this tube which is the difference of the equilibrium potential (1.23 V) and the onset potential (0.88 V), where all reaction steps are downhill in free energy.

To estimate the trends in catalytic activity of N-CNTs between tube sizes of (4,0) and (20,10) the scaling of the adsorption energy of the intermediates was used to construct a volcano plot. Figure 3 shows the scaling of the adsorption energies of *00 H and *0 intermediates with respect to the adsorption energy of *OH (left) and as a function of the diameter of the CNTs (right).

The scaling relations shown in Figure 3 (left) can be used to construct a volcano plot, where the reaction free energy for each elementary step (or the limiting potential) can be expressed as a function the adsorption energy of *OH. This can be used to estimate which adsorption energy of *OH gives the lowest overpotential for the ORR reaction, estimated by the limiting potential. This has been done before to evaluate catalytic trends in e.g. O ₂ and N ₂ reduction on transition metals ^27,30 . In the following, the free energy is expressed as a sum of an energy contribution that scales as the *OH binding energy, a constant coming from zero-point energy and entropy, and the energy shift due to the applied potential. The energy of all the different species is now expressed in terms of the *OH binding energy by using the linear scaling relations from Figure 3 (left):

Figure 4 shows the volcano plot, constructed from equations (8)-(11), where the volcano is outlined by lines representing the first and last protonation steps (elementary steps shown in eqs. (2) and (5) respectively). The line from equation 9 is too far out of frame to be of relevance. The lines are constructed from the scaling relations in Figure 3a and are used to guide the eye. The explicit values for the PLSs for each tube are also included in Figure 4. The right leg is limited by the first protonation step, O ₂ to *OOH, except for the (10,4) and (11 ,2) tubes where the second last step, *0 to *OH, is the PLS. The left leg is limited by the last protonation step, *OH to H ₂ O, in all cases except for a few tubes, where the second last step, *0 to *OH, is the PLS: (4,4), (5,1), (5,2), (6,1), (6,3), (7,1), (8,2), (8,5) and (9,3). This is because there is some scatter in Figure 3a which is also apparent in the volcano plots in Figure 4. The tubes having the second last step being the PLS are all chiral tubes except one armchair tube. All of these tubes have rather small diameter, from 5.5-9.6 Å, which will be discussed further below.

The tubes with the lowest overpotential at the top of the volcano in Figure 4 are (8,8), (12,6) and (14,7) having estimated overpotentials of 0.37 V, 0.36 V and 0.35 V, respectively. Tube (14,7) is exactly at the top of the volcano and has therefore two PLSs instead of only one as on either side of the volcano. This can be seen on the free energy diagram in Figure 2 above.

With similar DFT calculations, the overpotential for ORR on Pt(111) has been estimated to be 0.45 V in good comparison with experiments. ²⁷ As can be seen on Figure 4, 22 N-CNTs are estimated to have lower overpotential than Pt, or down to 0.35 V. Our DFT calculations also estimates the overpotential on N-graphene to be around 0.57 V but over 40 tubes are estimated here to have lower overpotential than N-graphene, which has gotten the most attention experimentally today. Plotting the limiting potential as a function of the diameter of the N-CNTs (see Figure 5) shows the smaller tubes are predicted to have a large overpotential but they also have a large scatter, as e.g. the tubes of around 5 Å of (6,1) and (4,3) where the predicted overpotential is around 0.63 V and 1.46 V, respectively. N-CNTs having diameter greater than 7 Å are predicted to have a lower overpotential than N-graphene and tubes with diameter greater than 10 Å reach a plateau and the scatter is decreased considerably, with predicted overpotential between 0.35 to 0.55 V. The magnitude of the limiting potential gradually increases and approaches slowly the value of N-doped graphene as the diameter of the tubes gets greater. For armchair and chiral N-CNTs with diameter smaller than 10 Å, the PLS is usually the reduction of *OH to H ₂ O (reaction (5)), except in a few cases where reduction of *0 to *OH (reaction (4)) is the PLS, discussed above. The reduction of *OH to H ₂ O is the predicted PLS on all the zig-zag tubes considered in this work, where (20,0) is the largest tube with a diameter of 15.8 Å. For the armchair and chiral tubes larger than 16 Å in diameter the PLS is the reduction of O ₂ to *OOH (reaction in eq. (2)). Between 10-16 Å, the PLS shifts between being reduction of *OH o H ₂ O and the adsorption and reduction of O ₂ to *OOH, leading to a compromise between strong *OH binding and weak *OOH binding. This results in tubes with lower overpotential than on either side of this diameter range.

REFERENCES 1. X. Cheng, Z. Shi, N. Glass, L. Zhang, J. Zhang, D. Song, Z.-S. Liu, H. Wang, and J. Shen.

A review of PEM hydrogen fuel cell contamination: Impacts, mechanisms, and mitigation. Journal of Power Sources, 165:739-756, 2007.

2. Z. Chen, D. Higgins, A. Yu, L. Zhang, and J. Zhang. A review on non-precious metal electrocatalysts for PEM fuel cells. Energy & Environmental Science, 4:3167-3192, 2011. 3. K. Gong, F. Du, Z. Xia, M. Durstock, and L. Dai. Nitrogen-doped carbon nanotube arrays with high electrocatalytic activity fo oxygen reduction. Science, 323:760-764, 2009.

4. M. Xiao,J. Zhu, L. Feng, C. Liu, and W. Xing. Meso/macroporous nitrogen-doped carbon architectures with iron carbide encapsulated in graphitic layers as an efficient and robust catalyst for the oxygen reduction reaction in both acidic and alkaline solutions. Advanced Materials, 27:2521-2527, 2015.

5. Z.-S.Wu, S. Yang, Y. Sun, K. Parvez, X. Feng, and K. Mullen. 3D nitrogen-doped graphene aerogel-supported Fe304 nanoparticles as efficient electrocatalysts for the oxygen reduction reaction. Journal of the American Chemical Society, 134:9082-9085, 2012.

6. Liang, R. F. Zhou, X. M. Chen, Y. H. Tang, and S. Z. Qiao. Fe-N decorated hybrids of CNTs grown on hierarchically porous carbon for high-performance oxygen reduction. Advanced Materials, 26:6074-6079, 2014.

7. Qu, Y. Liu, J.-B. Baek, and L. Dai. Nitrogen-doped graphene as efficient metal-free electrocatalyst for oxygen reduction in fuel cells. ACS Nano, 4:1321-1326, 2010.

8. Jiao, Y. Zheng, M. Jaroniec, and S. Z. Qiao. Origin of the electrocatalytic oxygen reduction activity of graphene-based catalysts: A roadmap to achieve the best performance. J. Am. Chem. Soc., 136:4394-4403, 2014.

9. Zhu, Y.H. Zhu, C.C. Tian, T. Jin, X.J. Yang X.B. Jin, C.Z. Li, H.L. Wang, H.L. Liu, and S.

Dai. Pyrolysis of conjugated nanoporous polycarbazoles to mesoporous N-doped carbon nanotubes as efficient electrocatalysts for the oxygen reduction reaction. J. Mat. Chem. A, 5:4507-4512, 2017.

10. J.C. Li, P.X. Hou, and C. Liu. Heteroatom-doped carbon nanotube and graphene-based electrocatalysts for oxygen reduction reaction. Small, 13:1702002-1702015, 2017.

11. J.C. Li, P.X. Hou, M. Cheng, C. Liu, H.M. Cheng, and M.H. Shao. Carbon nanotube encapsulated in nitrogen and phosphorus co-doped carbon as a bifunctional electrocatalyst for oxygen reduction and evolution reactions. Carbon, 139:156-163, 2018.

12. D. Guo, R. Shibuya, C. Akiba, S. Saji, T. Kondo, and J. Nakamura. Active sites of nitrogen- doped carbon materials for oxygen reduction reaction clarified using model catalysts. Science, 351 :361-365, 2016.

13. H.B. Yang, J. Miao, S.F. Hung, J. Chen, H.B. Tao, X. Wang, L. Zhang, and R. Chen.

Identification of catalytic sites for oxygen reduction and oxygen evolution in N-doped graphene materials: Development of highly efficient metal-free bifunctional electrocatalyst. Sci. Adv., 2:e1501122-e1501133, 2016.

14. J. Shui, M. Wang, F. Du, and L. Dai. N-doped carbon nanomaterials are durable catalysts for oxygen reduction reaction in acidic fuel cells. Sci. Adv., 1 :e1400129, 2015.

15. L. Xue, Y. Li, X. Liu, Q. Liu, J. Shang, H. Duan, L. Dai, and J. Shui. Zigzag carbon as efficient and stable oxygen reduction electrocatalyst for proton exchange membrane fuel cells. Nat. Comm., 9:3819, 2018.

16. Y. Zheng and S.Z. Qiao. N-doping goes sp-hybridized. Nat. Chem., 10:900-902, 2018.

17. S. Sinthika, U.V. Waghmare, and R. Thapa. Structural and electronic descriptors of catalytic activity of graphene-based materials: First-principles theoretical analysis. Small, 14:1703609, 2018. 18. G. Murdachaew and K. Laasonen. Oxygen evolution reaction on nitrogen-doped defective carbon nanotubes and graphene. J. Phys. Chem. C, 122:25882-25892, 2018.

19. F. Studt. The oxygen reduction reaction on nitrogen-doped graphene. Catalysis Letters, 143:58-60, 2013.

20. L. Yu, X. Pan, X. Cao, P. Hu, and X. Bao. Oxygen reduction reaction mechanism on nitrogen-doped graphene: A density functional theory study. Journal of Catalysis, 282: 183- 190, 2011.

21. Y. Okamoto. First-principles molecular dynamics simulation of 02 reduction on nitrogen- doped carbon. Applied Surface Science, 256:335-341 , 2009.

22. T. Ikeda, M. Boero, S.-F. Huang, K. Terakura, M. Oshima, and J. Ozaki. Carbon alloy catalysts: Active sites for oxygen reduction reaction. Journal of Physical Chemistry, 112:14706-14709, 2008.

23. M. S. Dresselhaus, G. Dresselhaus, J. C. Charlier, and E. Hernandrez. Electronic, thermal and mechanical properties of carbon nanotubes. Philosophical Transactions: Mathematical, Physical and Engineering Sciences, 362:2065-2098, 2004.

24. M. S. Dresselhaus, G. Dresselhaus, and R. Saito. Physics of carbon nanotubes. Carbon, 33:883-891 , 1995.

25. X. Blase, Lorin X. Benedict, E. L. Shirley, and S. G. Louie. Hybridization effects and metallicity in small radius carbon nanotubes. Physical Review Letters, 72:1878-1881 , 1994.

26. S. Maldonado and K. J. Stevenson. Influence of nitrogen doping on oxygen reduction electrocatalysis at carbon nanofiber electrodes. The Journal of Physical Chemistry B, 109:4707-4716, 2005.

27. J. K. Norskov, J. Rossmeisl, A. Logadottir, L. Lindqvist, J. R. Kitchin, T. Bligaard, and H.

Jonsson. Origin of the overpotential for oxygen reduction at a fuel-cell cathode. The Journal of Physical Chemistry B, 108:17886-17892, 2004.

28. F. Abild-Pedersen, J. Greeley, F. Studt, J. Rossmeisl, T.R. Munter, P.G. Moses, E.

Skulason, T. Bligaard, and J.K. Norskov. Scaling properties of adsorption energies for hydrogen containing molecules on transition metal surfaces. Physical Review Letters, 99:016105, 2007.

29. J. Rossmeisl, Z.-W. Qu, H. Zhu, G.-J. Kroes, and J. K. Norskov. Electrolysis of water on oxide surfaces. Journal of Electroanalytical Chemistry, 607:83-89, 2007.

30. E. Skulason, T. Bligaard, S. Gudmundsdottir, F. Studt, J. Rossmeisl, F. Abild-Pedersen, T.

Vegge, H. Jonsson, and J. K. Norskov. A theoretical evaluation of possible transition metal electro-catalysts for N2 reduction. Physical Chemistry Chemical Physics, 14:1235-1245, 2012.

31. A.B. Hoskuldsson, Y. Abghoui, A.B. Gunnarsdottir, and E. Skulason. Computational screening of rutile oxides for electrochemical ammonia formation. ACS Sustain. Chem. Eng., 5:10327-10333, 2017.

32. A.A. Peterson and J.K. Norskov. J. Phys. Chem. Lett., 3:251-258, 2012.

33. A. Bhowmik, T. Vegge, and H.A. Hansen. ChemSusChem, 9:3230-3243, 2016. 34. E. Tayyebi, J. Hussain, Y. Abghoui, and E. Skulason. Trends of electrochemical C02 reduction reaction on transition metal oxide catalysts. J. Phys. Chem. C, 122:10078- 10087, 2018.

35. Y. Abghoui, A.L. Garden, V.F. Hlynsson, S. Bjorgvinsdottir, H. Olafsdottir, and E.

Skulason. Enabling electrochemical reduction of nitrogen to ammonia at ambient conditions through rational catalyst design. Phys. Chem. Chem. Phys., 17:4909-4918, 2015.

36. Y. Abghoui, A.L. Garden, J.G. Howalt, T. Vegge, and E. Skulason. Electroreduction of N2 to ammonia at ambient conditions on mononitrides of Zr, Nb, Cr, and V: A DFT guide for experiments. ACS Catal., 6:635-646, 2016.

37. A.A. Peterson, F. Abild-Pedersen, F. Studt, J. Rossmeisl, and J.K. Norskov. Energy Environ. Sci., 3:1311 , 2010.

38. J. Hussain, H. Jonsson, and E. Skulason. Faraday Discuss., 195:619-636, 2016.

39. E. Skulason and H. Jonsson. Atomic scale simulations of heterogeneous electrocatalysis: recent advances. Advances in Physics: X, 2:481 , 2017.

40. J. Hussain, H. Jonsson, and E. Skulason. Calculations of product selectivity in electrochemical C02 reduction. ACS Catal., 8:5240-5249, 2018.

41. D. Srivastava and K. Laasonen. Dissociative adsorption of 02 on negatively charged nitrogen-doped single-walled carbon nanotubes: First-principles calculations. RSC Advances, 6:84155-84163, 2016.

42. B. Hammer, L. B. Hansen, and J. K. Norskov. Improved adsorption energetics within density-functional theory using revised perdew-burke-ernzerhof functionals. Physical Review B, 59:7413-7421 , 1999.

43. P. Blochl. Phys. Rev. B: Condens. Matter Mater. Phys, 50:17953(17979, 1994.

44. G. Kresse and J. Hafner. Phys. Rev. B: Condens. Matter Mater. Phys., 47:558-561 , 1993.

45. G. Kresse and J. Hafner. Phys. Rev. B: Condens. Matter Mater. Phys., 49:14251- 14269,1994.

46. G. Kresse and J. Furthmuller. Comput. Mater. Sci., 6:15-50, 1996.

47. G. Kresse and J. Furthmuller. Phys. Rev. B: Condens. Matter Mater. Phys., 54:11169- 11186, 1996.