Nitrogen-Doped Carbon Nanosheets Decorated With Mn2O3 Nanoparticles for Excellent Oxygen Reduction Reaction

The purpose of this study is to develop an active, low cost, non-precious, stable, and high-performance catalyst for oxygen reduction reaction (ORR). In this regard, Mn2O3-decorated nitrogen-doped carbon nanosheets (Mn2O3/NC) are fabricated by a two-step strategy involving a hydrothermal method and a solid-state method. In the resultant structures, very fine Mn2O3 nanoparticles with an average size of about 5 nm are strongly attached to nitrogen-doped carbon nanosheets. The role of the Mn2O3 nanoparticles is to provide active sites for ORR, while the presence of the nitrogen-doped carbon not only enhances the conductivity of the overall structure but is also helpful for overall stability. The Mn2O3/NC shows good onset potential (0.80 V@−1 mA/cm2), methanol crossover effect, and stability (90%).


INTRODUCTION
Over the past 100 years, global energy intensity has declined continuously due to economic and population growth. World energy consumption is projected to rise by 2.3% per year (Lewis and Nocera, 2006). It is estimated that energy demand will nearly double by 2050 (Du and Eisenberg, 2012). Eighty percent of the current energy resources are from fossil fuels (Ran et al., 2014), but the use of fossil fuels is associated with numerous problems like unsustainably, finite reserves (Dincer, 2000), buildup of greenhouse gases, and the emission of CO 2 (Ran et al., 2014). To reduce the dependency on fossil fuels and to meet global energy demand in a sustainable fashion, renewable energy sources like fuel cells, metal-air batteries, and water splitting are important (Dincer, 2000;Du and Eisenberg, 2012;Ran et al., 2014). These energy systems are governed through electrochemical reactions in which oxygen reduction reaction (ORR) is particularly important (Dincer, 2000;Du and Eisenberg, 2012;Ran et al., 2014).
ORR occurs at the cathode of fuel cells, but its kinetics are six times slower than the anode reaction (Post, 1999;Linda et al., 2000;Pokropivny, 2007;Jingjing et al., 2014;Augustin et al., 2015;Jiao et al., 2015;Wei et al., 2015;Deng et al., 2016;Liu et al., 2016;Lv et al., 2016;Sun et al., 2016;Zhu et al., 2016). Therefore, a catalyst is required to enhance the speed of ORR. In this regard, Platinum (Pt) holds great importance due to its high efficiency and low onset potential (Linda et al., 2000). However, its high cost, rarity, and low stability are major concerns for the abovementioned energy technologies (Hu and Dai, 2016). Thus, the development of a stable, active, and low-cost ORR catalyst is required in order to replace Pt (Lewis and Nocera, 2006).
Transition metals oxides (MnO, CoO, NiO) are considered to be the best alternative to noble metals for ORR. In particular, Mn 2 O 3 , because of the diversity of its oxidation states (2 + , 3 + , 4 + ), shows remarkable catalytic activity toward ORR (Jiao et al., 2015). Mn 2 O 3 has the advantages of low cost, non-toxicity, compatibility with the environment, and abundant reserves (Lv et al., 2016). It also has various structural configurations (Wei et al., 2015). All these aspects of Mn 2 O 3 make it an ideal candidate for ORR. However, there are still some issues that need to be resolved such as its low conductivity (Jiao et al., 2015).
On the other hand, because of availability, low price, and variety of form, carbon-based electro-catalysts have attracted much intention over the last few years. They show excellent ORR activities, comparable and even superior to that of Pt (Hu and Dai, 2016). Carbon-based nanostructures are also used to increase the conductivity for ORR. Carbon nanosheets and nanotubes are mostly used in this regard (Post, 1999;Jingjing et al., 2014). They also increase the surface area, and hence the number of active sites. However, pure carbon has low or moderate electrocatalytic activity (Jingjing et al., 2014;Liu et al., 2016). In order to improve the electrocatalytic activity of carbon nanostructures, different strategies are used, such as doping and combining them with other materials. Doping with heteroatoms like nitrogen (N) in carbon (C) is valuable because the electron affinity of N is greater than that of C; the carbon atoms adjacent to nitrogen dopants counterbalance this electron affinity by creating a net positive charge density. N doping attracts electrons to facilitate the ORR (Deng et al., 2016). It increases not only the electrocatalytic activity  but also the stability.
Keeping the aforementioned aspects in mind, the fabrication of Mn 2 O 3 on NC may be a good approach to enhancing the activity and stability of ORR. The small Mn 2 O 3 particle provides a larger surface area. NC sheets not only provide the support but also increase the conductivity and number of electrons transferred. Herein, we have fabricated Mn 2 O 3decorated nitrogen-doped carbon nanosheets (Mn 2 O 3 /NC) by using hydrothermal and solid-state methods. The resultant Mn 2 O 3 /NC has very small, ∼5-10 nm Mn 2 O 3 nanoparticles, providing a large surface area and number of active sites, while the nitrogen-doped carbon nanosheets increase not only the conductivity but also the stability of ORR. We explored the ORR activity of Mn 2 O 3 supported by NC at different temperatures and concentrations and found that Mn 2 O 3 /NC at 500 • C and with a 20% initial Mn 2 O 3 concentration (Mn 2 O 3 /NC-500-20%) shows the best ORR performance, with low onset potential and excellent stability along with a good methanol crossover effect.

RESULTS AND DISCUSSION
In XRD, the Mn 2 O 3 phase was identified with card number 41-1442. From XRD analysis, it can be seen that Mn 2 O 3 is in the cubic phase with a=b=c=9.04 Å. No other peaks are observed, which confirms the high purity of the material, as stronger and sharper peaks indicate the formation of a well-defined crystalline structure. In order to see the effect of temperature and initial concentration on the XRD pattern, we compare the XRD patterns of three different samples, i.e., Mn 2 O 3 /NC-500-20%, Mn 2 O 3 /NC-500-10%, and Mn 2 O 3 /NC-450-10%. As shown in Figure 1, there is little difference in the intensities of the Mn 2 O 3 nanoparticle XRD peaks for the fabricated three samples, which means that their Mn 2 O 3 particle sizes are similar.
Mn 2 O 3 nanoparticle-decorated nitrogen-doped carbon nanosheets (Mn 2 O 3 /NC) were fabricated through the pyrolysis of Mn(OH) 2 and nitrogen-rich compounds, namely urea, in a muffle furnace. During the pyrolysis process, NH 3 released from urea reduces Mn(OH) 2 into Mn 2 O 3 nanoparticles at the same time as Mn 2 O 3 catalyzes the carbonization of carbon to form N-doped carbon nanosheets. SEM was used to determine the microstructures and morphology of the Mn 2 O 3 /NC. The SEM results for Mn 2 O 3 /NC-450-20%, Mn 2 O 3 /NC-500-20%, Mn 2 O 3 /NC-450-10%, and Mn 2 O 3 /NC-500-10% are shown in Figures S1-S4. SEM images of samples Mn 2 O 3 /NC-500-2.5%, Mn 2 O 3 /NC-500-5%, Mn 2 O 3 /NC-500-10%, and Mn 2 O 3 /NC-500-20%, which have initial concentrations of Mn(OH) 2 of 2.5, 5, 10, and 20% and were fabricated at 500 • C, are shown in Figure S5. We used TEM to image the nanostructures more clearly. The TEM images of a prepared sample in Figure 2 show that Mn 2 O 3 particles are embedded in nitrogen-doped carbon nanosheets. Very thin nanosheets of carbon can be seen in all of the TEM images, while very small ∼5-10 nm nanoparticles of Mn 2 O 3 (black dots) can be observed on the surface of the NC.
Electrochemical measurement was carried out in a threeelectrode electrochemical cell. Rotating disk electrode (RDE) measurements were performed to investigate the ORR activities in 0.1 M KOH solutions. Hg/HgO was used as a reference electrode, and platinum wire was used as a counter electrode. A  Figure 3A. The onset potentials of all the tested samples at a current density of −1 mA/cm 2 are given in Table 1. Figure 3A and Table 1 show that Mn 2 O 3 /NC-500-20% has the best onset potential (0.732 V @-1mA/cm 2 ) of the fabricated samples of Mn 2 O 3 /NC. However, the onset potential of Pt/C (0.93 V @−1mA/cm 2 ) is much better than that of Mn 2 O 3 /NC-500-20%. Furthermore, we conclude from our results that fabrication at 500 • C and an initial concentration of 20% Mn(OH) 2 are the best approximation to the performance of Pt/C. Three samples (Mn 2 O 3 /NC-450-20%, Mn 2 O 3 /NC-550-20%, and Mn 2 O 3 /NC-500-20%) with the same initial concentration of 20% but with different fabrication temperatures are compared in Figure 3B. At high temperature, the organic species are evaporated from the sample (all organic groups such as C will be removed on heating). This loss of organic groups occurs due to the evolution of CO and CO 2 gases and the decomposition of organic ligands by oxidation (Pokropivny, 2007). This is dependent on temperature and the partial pressure of oxygen (Pokropivny, 2007). Therefore, with the decrease in C, Mn 2 O 3 particles become active or the amount of Mn 2 O 3 will increase. When Mn 2 O 3 becomes active, it can lose oxygen. The loss of oxygen will result in the generation of vacancies on the surface. Both the electrical and chemical properties and the performance of the catalyst and its support change because of the presence of such oxygen vacancies (Linda et al., 2000). When Mn 2 O 3 nanoparticles are more accessible to oxygen due to these vacancies, a large triple-phase boundary will result, which further facilitates ORR.
In order to achieve better ORR activity, there should be a balance in the coverage of the surface by oxygenated species, which are specifically anions. These anions are adsorbed on the catalyst surface and have intermediate adsorption energies. Therefore, if the catalyst binds oxygen atoms too strongly, the ORR process will slow down due to the slow rate of removing oxides and anions from the surface. Conversely, if the catalyst binds oxygen atoms too weakly, the ORR process will be limited by the rate of electron and proton transfer to adsorbed O 2 (Augustin et al., 2015). Thus, the catalyst may bind oxygen too strongly at 450 • C and too weakly at 550 • C, so that both conditions produce Mn 2 O 3 /NC that is less active than with a temperature of 500 • C. In addition, through our experiment, we can see that at 550 • C, a large amount of the sample evaporates or vanishes, so with the increase in temperature above 550 • C, the ORR activity will decrease. The activities of samples Mn 2 O 3 /NC-500-10%, Mn 2 O 3 /NC-500-20%, and Mn 2 O 3 /NC-500-30%, all with the same temperature condition of 500 • C, are compared in Figure 3C. It can be concluded that the low number of Mn 2 O 3 particles where the concentration is 10% means that they cannot take part in the reaction properly and that a concentration of 30% means that the amount of Mn 2 O 3 particles is too high, and there is therefore insufficient carbon support. Thus, a concentration of 20% is the best option on the basis of our results, showing a balance between the amounts of carbon and Mn 2 O 3 .
The working electrode was scanned with RDE voltammograms for ORR at different rotation speeds. Linear-sweep voltammograms (LSVs) were obtained of Mn 2 O 3 /NC-500-20% in an O 2 -saturated 0.1 M KOH solution at different rotation speeds of 400-2500 rpm. The current density for the LSV curves increased with increasing rotation speed due to the increase in mass transport on the electrode surface, as shown in Figure 3D.
For practical applications in fuel cells, the methanol crossover effect and stability are both important parameters. The methanol crossover effect was investigated by LSV measurements in 0.1 M KOH with the addition of methanol   Figure 4B, demonstrate that Mn 2 O 3 /NC-500-20% has better long-term durability.
In conclusion, we fabricated nitrogen-doped carbon nanosheets decorated with Mn 2 O 3 nanoparticles by adopting a novel two-step synthesis method. The phase was identified as crystalline Mn 2 O 3 with card number 41-1442. The particle size, calculated from TEM images, was ∼5-10 nm. Mn 2 O 3 /NC-500-20% exhibited good ORR performance with low onset potential (0.732 V@-1 mA/cm 2 ), excellent stability, and a good methanol crossover effect. This low cost and easy fabrication method of Mn 2 O 3 /NC can be used for large-scale fabrication not only of Mn 2 O 3 but also of other transition metal oxide/NCs. Because of its good stability and low onset potential, Mn 2 O 3 -500-20% can be used in fuel cells and metal-air batteries.

Fabrication of Mn 2 O 3
Two-step synthesis, involving a hydrothermal method and a solid-state method, was used.
Briefly, a solution of MnCl 2. 6H 2 O and urea is stirred for 20 min in 20 mL of distilled water. The resulting solution is transferred to a Teflon container and placed in an autoclave. This autoclave process is carried out for 8 h at 120 • C. The solution is then washed several times with distilled water and ethanol in a centrifuge machine; finally, it is dried at 60 • C for 8 h. After drying, this intermediate product is mixed with urea and exposed to different temperatures in a muffle furnace for 2 h. The details of the process are given in Scheme 1.

Characterization of the Materials
According to the elaboration procedure, the Mn 2 O 3 nanoparticles are supported by NC. A scanning electron microscope (SEM) was used for the morphological analysis. We investigated the particle size distribution using transmission electron microscopy (TEM). X-ray Diffraction (XRD) was used in order to investigate the crystal structures of Mn 2 O 3 /NC.

Electrochemical Setup
The electrochemical measurements were carried out in a threeelectrode cell. Rotating disk electrode (RDE) measurements were performed to investigate the ORR activities in 0.1 M KOH. Hg/HgO was used as a reference electrode, and platinum wire was used as a counter electrode. A glassy carbon disc with a diameter of 2.5 mm served as a substrate for the working electrode.

DATA AVAILABILITY STATEMENT
All datasets generated for this study are included in the article/Supplementary Material.

AUTHOR CONTRIBUTIONS
MQas has written the manuscript. MQad has done the experimental work. JH, SB, and MHF helped in characterization. JZ has provided the experimental facilities. MTah has designed and supervised the overall work. FI and MTan have revised the manuscript while MQF helped in experimentation.