Copper- and Nitrogen-Codoped Graphene with Versatile Catalytic Performances for Fenton-Like Reactions and Oxygen Reduction Reaction

: Copper- and nitrogen-codoped reduced graphene oxide material (Cu / N-rGO) was prepared with a hydrothermal method. Its versatile catalytic performances were demonstrated toward the oxidative degradation of rhodamine B (RhB) and oxygen reduction reaction (ORR). The Cu and N codoping of graphene enhanced not only its activation ability toward H 2 O 2 , but also its electrocatalytic ability for ORR. It was observed that the use of 3%Cu / N-rGO together with 40 mmol · L − 1 H 2 O 2 and 4 mmol · L − 1 Na 2 CO 3 could remove more than 94% of the added RhB (30 mg · L − 1 ) in 20 min through a catalytic Fenton-like degradation. Quenching experiments and electron paramagnetic resonance (EPR) measurements indicated that the main reactive species generated in the catalytic oxidation process were surface-bound • OH. The modiﬁed graphene also showed good electrocatalytic activity for ORR reaction in alkaline media through a four-electron mechanism. On the electrode of Cu / N-rGO, the ORR reaction exhibited an onset potential of − 0.1 V and a half-wave potential of − 0.248 V, which were correspondingly close to those on a Pt / C electrode. In comparison with a Pt / C electrode, the 3%Cu / N-rGO electrode showed much greater tolerance to methanol. Such outstanding catalytic properties are attributed to the abundant active sites and the synergism between Cu and N in Cu / N-rGO.


Introduction
Due to serious challenges from environmental pollution and energy shortages, various catalysts have been developed to enhance the related pollutant degradation reactions and energy conversion reactions, especially for Fenton-like reactions, and hydrogen evolution and oxygen reduction reactions [1][2][3][4][5]. Typical investigations are focused on the designs of heterogeneous catalysts with high catalytic activity and selectivity by using noble metals (Pt and Au) [6,7], metal oxides (Fe 3 O 4 and ZnO/Co 3 O 4 ) [8,9], and metal nitrides [10]. These catalysts may have the demerits of either high cost (due to noble metals) or poor durability (due to the less noble metals), which are unfavorable to their large-scale deployment. To overcome this problem, it is interesting and important to develop carbon-based versatile catalysts that have good catalytic performances for both the Fenton-like reactions (in the field of environmental chemistry) and electrochemical oxygen reduction reaction (in the field of energy chemistry).
In fact, carbon-based catalytic materials have recently attracted much attention, such as carbon nanotubes (CNTs) [11,12], graphene [13,14], graphitic carbon nitride (g-C 3 N 4 ) [15][16][17], and carbon The X-ray diffraction (XRD) patterns of rGO, N-rGO, 3%Cu/N-rGO, and 100%Cu/N-rGO are shown in Figure 2a. For all samples, two distinct characteristic peaks of the graphite (002) plane at 2θ = 26.5° and (100) plane at 2θ = 43° were observed, which demonstrated the graphitic crystal structure. However, no copper-based characteristic peaks were detected in the patterns of the 3%Cu/N-rGO sample, due to the low Cu content, poor crystallinity of Cu-based compounds, and/or the formation of Cu-N. For 100%Cu/N-rGO, new reflection peaks at 35.6° and 38.9° can be indexed to the (−111) and (111) planes of CuO (JPCDS card no. , indicating the presence of CuO species. This was consistent with the TEM observation. The X-ray diffraction (XRD) patterns of rGO, N-rGO, 3%Cu/N-rGO, and 100%Cu/N-rGO are shown in Figure 2a. For all samples, two distinct characteristic peaks of the graphite (002) plane at 2θ = 26.5 • and (100) plane at 2θ = 43 • were observed, which demonstrated the graphitic crystal structure. However, no copper-based characteristic peaks were detected in the patterns of the 3%Cu/N-rGO sample, due to the low Cu content, poor crystallinity of Cu-based compounds, and/or the formation of Cu-N. For 100%Cu/N-rGO, new reflection peaks at 35.6 • and 38.9 • can be indexed to the (−111) and (111) planes of CuO (JPCDS card no. , indicating the presence of CuO species. This was consistent with the TEM observation. Figure 2b shows the Raman spectra of rGO, N-rGO, 3%Cu/N-rGO, and 100%Cu/N-rGO. In all the spectra, there were two peaks at around 1350 cm −1 and 1603 cm −1 , corresponding to the D band and G band of graphene, respectively [23]. In comparison with the I D /I G value of rGO (1.55), the values of N-rGO, 3%Cu/N-rGO, and 100%Cu/N-rGO were decreased to 1.11, 1.08, and 1.10, respectively, due to the incorporation of N and Cu atoms. This suggested the formation of the increased number of hopping defects in the graphene layers. The increased number of defect sites can be beneficial to the enhancement of the catalytic activity.
To obtain the specific surface area of rGO, N-rGO, 3%Cu/N-rGO, and 100%Cu/N-rGO, we measured the adsorption−desorption isotherms of N 2 ( Figure S1). It was found that the pore sizes distribute mainly in the range of 2-20 nm. The specific surface area of 3%Cu/N-rGO (425.21 m 2 ·g −1 ) was larger than that of N-rGO (398.05 m 2 ·g −1 ), 100%Cu/N-rGO (186.45 m 2 ·g −1 ), and rGO (177.98 m 2 ·g −1 ) ( Table  S1). The specific surface area declined when the doping context of Cu was 100%, which may be because the CuO species anchored onto the surface of 100%Cu/N-rGO. = 26.5° and (100) plane at 2θ = 43° were observed, which demonstrated the graphitic crystal structure. However, no copper-based characteristic peaks were detected in the patterns of the 3%Cu/N-rGO sample, due to the low Cu content, poor crystallinity of Cu-based compounds, and/or the formation of Cu-N. For 100%Cu/N-rGO, new reflection peaks at 35.6° and 38.9° can be indexed to the (−111) and (111) planes of CuO (JPCDS card no. 45-0937), indicating the presence of CuO species. This was consistent with the TEM observation. The X-ray photoelectron spectroscopy (XPS) was applied to analyze the surface chemical state of rGO, N-rGO, and 3%Cu/N-rGO, as shown in Figure 3a. From the survey scan of 3%Cu/N-rGO, it can be seen that the 3%Cu/N-rGO catalyst contained four kinds of elements, i.e., carbon (70.85%), oxygen (21.84%, Table S2), nitrogen (6.70%), and copper (0.61%). Apparently, the metallic Cu signal was very weak, and the content was only 0.61%. The C 1s spectra (Figure 3b) of rGO, N-rGO, and 3%Cu/N-rGO could be deconvoluted into the following bands: C=C/C−C (284.38 eV), C=N/C−OH (285.6 eV), epoxy C−O−C (286.7 eV), C=O/C−N (288.9 eV), and O−C=O (291.5 eV), respectively [23,34]. Similarly, the N 1s XPS spectra of 3%Cu/N-rGO ( Figure 3c) could be deconvoluted into three peaks around 398.3 eV, 400.0 eV, and 401.2 eV, which were attributed to pyridinic-type N (47.31%, Table S3), pyrrolic-type N (31.54%), and graphitic-type N (6.76%), respectively [23,35]. In addition, there was a particular additional peak at 399.2 eV, which was ascribed to the Cu-coordinated nitrogen (14.39%) [35,36]. The high ratio of pyridinic N (47.31%, Table S3) in 3%Cu/N-rGO could enhance the ORR activity as compared to that of N-rGO (43.10%) [10]; meanwhile, the Cu-coordinated nitrogen could provide new active sites. In the Cu 2p XPS spectrum in Figure 3d, the main peak between 932.1 and 934.5 eV was similar to the values previously reported [36,37], which demonstrated that the valence states of the Cu species in 3%Cu/N-rGO was situated between Cu 0 and Cu 2+ . This result indicated that the Cu δ+ (0 < δ < 2) nature of the Cu-coordinated nitrogen was dominant on the surface of the Cu species, which conforms with previous literature [38,39].

H 2 O 2 Activation Catalyzed by Cu/N-rGO for the Oxidative Degradation of Rhodamine B (RhB)
In our previous work [23], we explored the persulfate activation ability of N-rGO toward the oxidative degradation of organic pollutants. As a green oxidizing agent, H 2 O 2 is superior to persulfate because the activation of the latter disposes of unfavorable sulfates into the natural environment. Therefore, we investigated the H 2 O 2 activation catalyzed by 3%Cu/N-rGO toward the oxidative degradation of RhB. Figure 4a displays the removal of RhB in different systems. It was certain that under the tested conditions the total removal of RhB was contributed to from both the pre-adsorption and the followed degradation. There was no adsorption capacity in the system of H 2 O 2 and H 2 O 2 + Na 2 CO 3 . After adding the catalysts, the adsorption capacity (Figure 4b) of these samples followed the order 3%Cu/N-rGO (345 mg·g −1 ) > N-rGO (232 mg·g −1 ) > rGO (208 mg·g −1 ). Therefore, the doping N and Cu improved the pre-adsorption capacity of RhB on rGO, which may be attributed to more active sites and a larger specific surface area (as confirmed in Figure S1). After the pre-adsorption, the oxidative degradation of RhB was started by adding H 2 O 2 . In the absence of any catalyst, the degradation removal of RhB in 20 min was only 3% by adding 40 mM H 2 O 2 , and it was increased slightly to 13% by Catalysts 2020, 10, 1326 5 of 14 co-adding 4 mM Na 2 CO 3 . The enhancement induced by the introduction of Na 2 CO 3 may be attributed to the greater number of free radicals originating from the interaction between CO 3 2− and • OH [40,41].
This result was consistent with previous reports that carbonates may enhance the degradation of organic pollutants [40,42]. When N-rGO was used as the catalyst system in the presence of H 2 O 2 and Na 2 CO 3 , RhB was partly degraded, giving a corresponding total RhB removal of 47%. Surprisingly, the RhB removal efficiency was increased to be more than 94% when 3%Cu/N-rGO was introduced. This improved activation performance may be ascribed to the increased number of defect sites and the synergism between Cu and N doped in Cu/N-rGO. The transition metal related to Lewis acid can be prone to bind with N (Lewis base) to form a strong M-N coordination coupled with carbon, which contributes to excellent catalytic activity. The nature of the metal-nitrogen species (M-N x ) resulted in an irregular charge distribution and distribution [20,43,44], which ameliorated the RhB adsorption and degradation. Moreover, the transition metal provided new active sites, which was beneficial to the catalytic activation because of its vacant 3D orbital and variable valence state [27].

H2O2 Activation Catalyzed by Cu/N-rGO for the Oxidative Degradation of Rhodamine B (RhB)
In our previous work [23], we explored the persulfate activation ability of N-rGO toward the oxidative degradation of organic pollutants. As a green oxidizing agent, H2O2 is superior to persulfate because the activation of the latter disposes of unfavorable sulfates into the natural environment. Therefore, we investigated the H2O2 activation catalyzed by 3%Cu/N-rGO toward the oxidative degradation of RhB. Figure 4a displays the removal of RhB in different systems. It was certain that under the tested conditions the total removal of RhB was contributed to from both the pre-adsorption and the followed degradation. There was no adsorption capacity in the system of Cu/N-rGO. The transition metal related to Lewis acid can be prone to bind with N (Lewis base) to form a strong M-N coordination coupled with carbon, which contributes to excellent catalytic activity. The nature of the metal-nitrogen species (M-Nx) resulted in an irregular charge distribution and distribution [20,43,44], which ameliorated the RhB adsorption and degradation. Moreover, the transition metal provided new active sites, which was beneficial to the catalytic activation because of its vacant 3D orbital and variable valence state [27].  Figure 5 gives the effects of various reaction parameters (Cu load, H2O2 concentration, Na2CO3 concentration, and temperature) on the adsorption and degradation of RhB over Cu/N-rGO, where the insets are the dependence of the apparent rate constant of RhB degradation on the specified reaction parameter. In order to know the practical Cu load in the different catalysts, we determined that the actual Cu-doping contents of the Cu/N-rGO samples were obtained by the atomic absorption spectroscopy, as listed in Table S4. From Figure 5a, it was found that the total removal of RhB in 20 min was increased from 43% for N-rGO (zero Cu load) to 67%, 79%, and 94% for Cu-loading amounts of 0.5, 1.0, and 3.0 wt.%, respectively. The differences in the total RhB removal observed between the various Cu/N-rGO catalysts were mainly attributed to the effect of the Cu content on the catalytic degradation of RhB. Because the degradation rate in the initial time period (0-1 min) was very fast, we used only the experimental data in the initial one minute to fit with the pseudo first-reaction kinetic model. The calculated rate constant (k) in the oxidation stage was increased from 0.11 for undoping condition to 0.16, 0.25, and 0.27 min −1 for Cu-adding amounts of 0.5, 1.0, and 3.0 wt.%. When the Cu-loading amount was increased to 100 wt.%, the RhB oxidative degradation was not significantly enhanced in relation to 3% Cu/N-rGO, and the apparent rate constant was only about 0.33 min −1 . The above results confirmed that the degradation performance of RhB was best when the Cu-loading amount was 3.0 wt.%. The good performance of 3% Cu/N-rGO was attributed to the moderate coordination of copper and nitrogen atoms, and then the formation of Cu-Nx (2 or 4) active sites that exhibited the strong Lewis acid-base interaction [37,45].
The effect of the initial concentration of H2O2 for RhB degradation in the 3%Cu/N-rGO system was studied, as shown in Figure 5b. With increasing H2O2 concentrations from 5.0 mM to 40.0 mM, the degradation efficiency improved from 73% to 94% and the k value increased from 0.19 to 0.27 min −1 . The enhanced efficiency can be attributed to the greater generation of radical species at higher H2O2 concentrations. The use of 60 mM H2O2 did not further increase the degradation efficiency with a k value of 0.25 min −1 . This was due to insufficient catalytic sites and self-quenching of the  Figure 5 gives the effects of various reaction parameters (Cu load, H 2 O 2 concentration, Na 2 CO 3 concentration, and temperature) on the adsorption and degradation of RhB over Cu/N-rGO, where the insets are the dependence of the apparent rate constant of RhB degradation on the specified reaction parameter. In order to know the practical Cu load in the different catalysts, we determined that the actual Cu-doping contents of the Cu/N-rGO samples were obtained by the atomic absorption spectroscopy, as listed in Table S4. From Figure 5a, it was found that the total removal of RhB in 20 min was increased from 43% for N-rGO (zero Cu load) to 67%, 79%, and 94% for Cu-loading amounts of 0.5, 1.0, and 3.0 wt.%, respectively. The differences in the total RhB removal observed between the various Cu/N-rGO catalysts were mainly attributed to the effect of the Cu content on the catalytic degradation of RhB. Because the degradation rate in the initial time period (0-1 min) was very fast, we used only the experimental data in the initial one minute to fit with the pseudo first-reaction kinetic model. The calculated rate constant (k) in the oxidation stage was increased from 0.11 for undoping condition to 0.16, 0.25, and 0.27 min −1 for Cu-adding amounts of 0.5, 1.0, and 3.0 wt.%. When the Cu-loading amount was increased to 100 wt.%, the RhB oxidative degradation was not significantly enhanced in relation to 3% Cu/N-rGO, and the apparent rate constant was only about 0.33 min −1 . The above results confirmed that the degradation performance of RhB was best when the Cu-loading amount was 3.0 wt.%. The good performance of 3% Cu/N-rGO was attributed to the moderate coordination of copper and nitrogen atoms, and then the formation of Cu-N x (2 or 4) active sites that exhibited the strong Lewis acid-base interaction [37,45].
The effect of the initial concentration of H 2 O 2 for RhB degradation in the 3%Cu/N-rGO system was studied, as shown in Figure 5b. With increasing H 2 O 2 concentrations from 5.0 mM to 40.0 mM, the degradation efficiency improved from 73% to 94% and the k value increased from 0.19 to 0.27 min −1 . The enhanced efficiency can be attributed to the greater generation of radical species at higher H 2 O 2 concentrations. The use of 60 mM H 2 O 2 did not further increase the degradation efficiency with a k value of 0.25 min −1 . This was due to insufficient catalytic sites and self-quenching of the generated superfluous • OH. The performance of the 3% Cu/N-rGO catalyst and the values in the previous articles are shown in Table S5.  Table S5.  Figure 5c displays the effect of Na2CO3 on the catalytic degradation of RhB. It was easily observed that the k value increased from 0.19 to 0.27 min −1 with the increasing Na2CO3 concentrations from 0 to 4 mM, but the addition of 8 mM Na2CO3 induced little further enhancement of the RhB degradation. It seemed that 4 mM Na2CO3 was optimal. This was explained by the fact that the co-existence of the H2O2 molecules and carbonate radicals resulted in the generation of CO3 −• , which triggered a series of reactions to generate other radicals (including the superoxide ion and singlet oxygen) for pollutant degradation as suggested by the equations below [40,41], involving the following reactions: H2O2 + CO3 -• → HCO3 -+ HO2 • (2) O2 •-+ • OH → O2 1 + HO - Figure 5c displays the effect of Na 2 CO 3 on the catalytic degradation of RhB. It was easily observed that the k value increased from 0.19 to 0.27 min −1 with the increasing Na 2 CO 3 concentrations from 0 to 4 mM, but the addition of 8 mM Na 2 CO 3 induced little further enhancement of the RhB degradation. It seemed that 4 mM Na 2 CO 3 was optimal. This was explained by the fact that the co-existence of the H 2 O 2 molecules and carbonate radicals resulted in the generation of CO 3 −• , which triggered a series of reactions to generate other radicals (including the superoxide ion and singlet oxygen) for pollutant degradation as suggested by the equations below [40,41], involving the following reactions: Catalysts 2020, 10, 1326 8 of 14 The degradation of RhB over 3%Cu/N-rGO was investigated at temperatures in the range of 10-55 • C, as shown in Figure 5d. The k value was roughly linearly increased from 0.24 min −1 to 0.36 min −1 with increasing temperatures from 10 • C to 55 • C. The activation energy (E a ) of the degradation of RhB in the 3%Cu/N-rGO-H 2 O 2 -Na 2 CO 3 system was calculated according to the Arrhenius equation of lnk = lnA − E a /RT, where R is the universal gas constant (8.314 J·mol −1 ). The E a value was obtained as 6.93 kJ·mol −1 , indicating the high catalytic activity of 3%Cu/N-rGO.
It was reported • OH, • O 2 − /HO 2 • , and singlet oxygen ( 1 O 2 ) could be generated to the reactive oxygen species for RhB in the H 2 O 2 -Na 2 CO 3 system. To determine the main free radicals for determining the degradation process, electron paramagnetic resonance (EPR) and quenching experiments were carried out. Figure 6a displays the EPR spectra of the tested catalytic system, which were typical for TEMPO (2,2,6,6-tetramethylpiperidine-1-oxyl)- 1 Figure 6b, the RhB degradation was inhibited significantly after adding thiourea, and the removal was decreased from 94% to 53.6%. When benzoquinone or NaN 3 was used as the radical scavenger in the 3%Cu/N-rGO/H 2 O 2 /Na 2 CO 3 system, the degradation efficiency declined slightly as compared with that without adding any radical scavenger. This result illustrated that the dominant free radical was • OH rather than • O 2 − /HO 2 • and 1 O 2 during the catalysis process, which was consistent with the reported literature [48].
HO2 • + HO2 • → O2 1 + H2O2 H2O2 + HCO3 -→ H2O + HCO4 - The degradation of RhB over 3%Cu/N-rGO was investigated at temperatures in the range of 10-55 °C, as shown in Figure 5d. The k value was roughly linearly increased from 0.24 min −1 to 0.36 min −1 with increasing temperatures from 10 °C to 55 °C. The activation energy (Ea) of the degradation of RhB in the 3%Cu/N-rGO-H2O2-Na2CO3 system was calculated according to the Arrhenius equation of lnk = lnA − Ea/RT, where R is the universal gas constant (8.314 J·mol −1 ). The Ea value was obtained as 6.93 kJ·mol −1 , indicating the high catalytic activity of 3%Cu/N-rGO.
It was reported • OH, • O2 − /HO2 • , and singlet oxygen ( 1 O2) could be generated to the reactive oxygen species for RhB in the H2O2-Na2CO3 system. To determine the main free radicals for determining the degradation process, electron paramagnetic resonance (EPR) and quenching experiments were carried out. Figure 6a displays the EPR spectra of the tested catalytic system, which were typical for TEMPO (2,2,6,6-tetramethylpiperidine-1-oxyl)-1 O2, DMPO (5,5-dimethyl-1-pyrroline N-oxide)-• OH, and DMPO-• O2 − /HO2 • . Strong characteristic peaks of DMPO-• OH were detected, but the signals for DMPO-• O2 − /HO2 • were weak and no signals were observed for TEMPO-1 O2. This suggested that • OH was primarily important in the degradation process. Furthermore, thiourea, benzoquinone, and NaN3 were used as the radical scavenger for • OH, • O2 − /HO2 • , and ( 1 O2) [46,47], respectively. According to Figure 6b, the RhB degradation was inhibited significantly after adding thiourea, and the removal was decreased from 94% to 53.6%. When benzoquinone or NaN3 was used as the radical scavenger in the 3%Cu/N-rGO/H2O2/Na2CO3 system, the degradation efficiency declined slightly as compared with that without adding any radical scavenger. This result illustrated that the dominant free radical was • OH rather than • O2 − /HO2 • and 1 O2 during the catalysis process, which was consistent with the reported literature [48].

Electrocatalytic Performance of Cu/N-rGO
The electrocatalytic activity of 3%Cu/N-rGO for the ORR was evaluated with cyclic voltammetry (CV), which was conducted in 0.1 M KOH at a scan rate of 10 mV·s −1 . As shown in Figure 7a,b, the obtained cyclic voltammogram of N-rGO and 3%Cu/N-rGO in the N2-saturated solution was typically capacitive without an observable electrocatalytic active reaction occurring in the potential window from −1.0 to 0.2 V (vs. Ag/AgCl). However, both the electrodes of N-rGO and

Electrocatalytic Performance of Cu/N-rGO
The electrocatalytic activity of 3%Cu/N-rGO for the ORR was evaluated with cyclic voltammetry (CV), which was conducted in 0.1 M KOH at a scan rate of 10 mV·s −1 . As shown in Figure 7a . These results were close to or higher than the values previously reported in the literature (Table S6). It should be mentioned that both the much higher limiting current density and more positive onset potential (and half-wave potential) of 3%Cu/N-rGO proved that the electrocatalytic activity of 3%Cu/N-rGO was much better than 100%Cu/N-rGO, rGO, and N-rGO, and very close to that of Pt/C, indicating that the existence of Cu species played a key role in the catalytic activity. Such outstanding catalytic properties may be attributed to the abundant active sites and the synergism between Cu and N in the Cu/N-rGO.
Catalysts 2020, 10, x FOR PEER REVIEW 9 of 14 3%Cu/N-rGO exhibited an obvious oxygen reduction peak around −0.285 V in the O2-saturated electrolyte, and the peak current (0.7 mA·cm −2 ) on 3%Cu/N-rGO was more than two times (0.26 mA·cm −2 ) that on N-rGO. This was indicative of an excellent ORR electrocatalytic activity of 3%Cu/N-rGO. To further support the result, we compared the linear sweep voltammetry (LSV) of rotating disk electrodes (RDEs) of 3%Cu/N-rGO and other materials including Pt/C at a rotation rate of 2500 rpm in O2-saturated 0.1 M KOH. As shown in Figure  . These results were close to or higher than the values previously reported in the literature (Table S6). It should be mentioned that both the much higher limiting current density and more positive onset potential (and half-wave potential) of 3%Cu/N-rGO proved that the electrocatalytic activity of 3%Cu/N-rGO was much better than 100%Cu/N-rGO, rGO, and N-rGO, and very close to that of Pt/C, indicating that the existence of Cu species played a key role in the catalytic activity. Such outstanding catalytic properties may be attributed to the abundant active sites and the synergism between Cu and N in the Cu/N-rGO. The kinetics of ORR reaction pathways of various catalysts were studied by LSV. The LSV curves (Figure 7d) of the 3%Cu/N-rGO catalyst illustrated that the current densities were increased with the increasing rotation speed, which was due to the increased mass transport, considering a kinetics-controlled process of ORR. As a comparison, the LSV data of rGO, N-rGO, 100%Cu/N-rGO, and Pt/C are plotted in Figure S2. The corresponding Koutecky-Levich (K-L) plots showed nearly parallel fitting lines (inset of Figure 7d), validating that the transferred electron numbers per oxygen molecule in ORR were almost the same. The electron transfer number (n) of 3%Cu/N-rGO was The kinetics of ORR reaction pathways of various catalysts were studied by LSV. The LSV curves (Figure 7d) of the 3%Cu/N-rGO catalyst illustrated that the current densities were increased with the increasing rotation speed, which was due to the increased mass transport, considering a kinetics-controlled process of ORR. As a comparison, the LSV data of rGO, N-rGO, 100%Cu/N-rGO, and Pt/C are plotted in Figure S2. The corresponding Koutecky-Levich (K-L) plots showed nearly parallel fitting lines (inset of Figure 7d), validating that the transferred electron numbers per oxygen molecule in ORR were almost the same. The electron transfer number (n) of 3%Cu/N-rGO was calculated to an average value of 4.04, indicating a better selectivity of the 4e − pathway to directly reduce O 2 to H 2 O. However, the corresponding values (insets of Figure S2a,b) of rGO and N-rGO were 1.51 and 2.16, respectively, suggesting the presence of a less efficient 2e − pathway. For both 100%Cu/N-rGO and Pt/C, the electron transfer number (insets of Figure S2c,d) was close to 4. However, the catalysis ability of 100%Cu/N-rGO decreased greatly compared with 3%Cu/N-rGO, due to the existence of CuO species. Figure 7e presents the Tafel plots of 3%Cu/N-rGO and 20 wt% Pt/C. The Tafel slope value for the 3%Cu/N-rGO catalyst was calculated to be 84.31 mV·dec −1 , which was very close to that of Pt/C (84.1 mV·dec −1 ). This demonstrated that it was easy to drive a larger catalytic current density at a lower overpotential and had a favorable electron transfer kinetics. In a methanol fuel cell, methanol may permeate from the anode to the cathode, leading to the poisoning and deactivation of the cathode catalyst. The durability of 3%Cu/N-rGO was evaluated by chronoamperometric measurements into 0.1 M KOH. As observed in Figure 7f, the 3%Cu/N-rGO catalyst still retained 97.5% of its original current after methanol (AR, 99.5%) injection, whereas the 20%Pt/C displayed a dramatic decline (current loss of 23.9%), illustrating the superior tolerance toward methanol of 3%Cu/N-rGO comparable to that of 20%Pt/C. The result was higher than the values previously reported in the literature (Table S6).

Preparation of the Cu/N-rGO Catalyst
Graphene oxide (GO) was prepared according to a modified Hummers method (see Supplementary Materials for more details) [23,49], and the GO suspension was obtained by dispersing GO in deionized water. Typically, a 12.5 mL portion of GO (4 mg·mL −1 ) suspension was added to 32.5 mL of Cu(NO 3 ) 2 ·3H 2 O aqueous solution. Afterward, the mixture was treated by ultrasonication for 30 min. Then, 5 mL of NH 3 ·H 2 O were added with magnetic stirring. The mixture was then transferred into a 100 mL Teflon-lined autoclave and heated at 180 • C for 6 h. Subsequently, the black precipitates were filtrated and washed with distilled water several times. Finally, the product was dispersed in water as Cu/N-rGO dispersions. N-rGO and rGO samples were similarly prepared in the absence of Cu(NO 3 ) 2 ·3H 2 O and/or NH 3 ·H 2 O, respectively. In the present work, the Cu/N-rGO catalysts were referred to as x%Cu/N-rGO, where x represented the Cu 2+ /GO mass ratio.

Characterizations
The morphology and the element distribution of the catalysts were characterized by transmission electron microscopy (TEM, FEI Tecnai G2 F20, Hillsboro, TX, USA) and energy-dispersive spectroscopy (EDS) (Fei company, Hillsboro, TX, USA). X-ray diffraction (XRD) was applied to obtain the crystal structure by using an X-ray diffractometer (Bruker D8 Advance, Thermo Fisher Technology Co., Ltd, Walsham, MA, USA) under the condition of Cu K α irradiation with operating voltage and current of 40 kV and 40 mA. The surface chemical composition and valence state were analyzed with X-ray photoelectron spectroscopy (XPS) on a VG Multilab 2000 spectrometer (Thermo Fisher Technology Co., Ltd, Walsham, MA, USA) with Al K α radiation (300 W). Raman spectra were obtained on a Thermo Fisher DXR Raman spectrometer (Thermo Fisher Technology Co., Ltd, Walsham, MA, USA) with laser power (532 nm). The free radicals in the DMPO-• OH, DMPO-O 2 −• /HO 2 • , and TEMPO-1 O 2 systems were trapped by electron paramagnetic resonance (EPR) spectra (Bruker, Beijing, China).

Adsorption and Degradation Experiments
The adsorption and catalytic degradation experiments were carried out in a 100 mL beaker at 298 K. The given amount of Cu/N-rGO (40 mg/L) was added to 50 mL solution of RhB (30 mg/L), and the dispersion was magnetically stirred for 16 min to achieve the adsorption/desorption equilibrium between the solution and the Cu/N-rGO solid. Then, a specified amount of H 2 O 2 + Na 2 CO 3 was added to the above suspension to initiate the degradation. Afterward, aliquots (1 mL) were sampled at given time intervals and centrifuged by using an EBA-21 centrifuge (13000 rpm) to remove Cu/N-rGO. The RhB concentration in the supernatant was determined by monitoring its absorbance at λ max = 554 nm on a Cary 60 UV-VIS spectrophotometer (Agilent, Palo Alto, CA, USA).

Electrochemical Measurements
The electrocatalytic activity of Cu/N-rGO for ORR was investigated at room temperature by using a CHI 660A electrochemical station with a standard three-electrode system consisting of a rotating disk electrode with a geometric area of 0.196 cm 2 coated with the catalyst, a Ag/AgCl reference electrode (saturated with 1 M KOH), and a Pt foil counter electrode. The primary catalyst ink was prepared by dispersing 5 mg of Cu/N-rGO into 5 mL of N, N-dimethylformamide (DMF) within 50 µL of 5% naphthol. The as-obtained ink was coated onto the disk electrode, yielding a catalyst loading of 0.1 mg·cm −2 . The ORR reaction of the catalyst was investigated by cyclic voltammetry (CV) in the N 2 /O 2 -saturated 0.1 M KOH electrolyte solution, with a scan rate of 10 mV·s −1 . Linear sweep voltammetry (LSV) curves were measured at various rotating speeds, with a scan rate of 10 mV·s −1 under O 2 condition. The stability of the catalysts was also evaluated by acquiring current−time (I-t) curves at 0.55 V in O 2 -saturated 0.1 M KOH solution.

Conclusions
A versatile catalyst of Cu-and N-codoped graphene was synthesized by a hydrothermal process. The Cu load in the catalyst and several other parameters were optimized in terms of the catalytic activity. It was confirmed that the 3%Cu/N-rGO catalyst showed superior catalytic activity in the H 2 O 2 -Na 2 CO 3 system. Its use achieved a total removal of 94% of the added RhB within 20 min with an apparent degradation rate constant of 0.27 min −1 . Furthermore, the 3%Cu/N-rGO catalyst was applied to oxygen reduction reaction, and displayed a four-electron pathway and excellent tolerance to methanol in ORR. The better catalytic and ORR performances were attributed to the abundant active sites and the synergism between Cu and N in the Cu/N-rGO.