The pressing crises of environmental pollution and fossil fuel energy depletion have propelled humans to explore feasible, reliable, and renewable energy technology solutions [1]. Various fuel cells including the Zn-air battery and H₂-O₂ fuel cells are capable of directly converting chemical energy into electrical energy, and with the absence of heat engine processes and the Carnot cycle, these fuel cells show high energy conversion efficiencies [2, 3]. Recently, the demand for environmentally friendly batteries with high specific energy and excellent reliability has stimulated the investigation of Zn-air batteries, of which the Zn anode is highly recognized for its high capacity, high safety, and low cost. Nonetheless, many conundrums remain unsolved, e.g., sluggish oxygen reduction reactions (ORRs) on the cathode side [4, 5]. The ORR, as the most important cathode reaction in the Zn-air battery, is generally recognized to have a significant impact on the performance of associated devices [6–8]. Although Pt-based catalysts have been regarded as the most efficient cathode catalysts, their prohibitive cost greatly limits their widespread commercial application. Therefore, it is of great importance to develop high-performance non-precious metal ORR electrocatalysts [9, 10].

In recent years, transition metal carbon nitrides (M-N-C, M = Co, Fe, et al.) have emerged as one of the most promising alternative electrocatalysts to Pt-based materials toward the ORR{[13-15]. In these alternative electrocatalysts, the M-N_x moiety normally functions as the major active site [11, 12], and its activity is dependent on the density of the active sites, surface area, and porous structure [13, 14]. Such M-N-C catalysts can be directly synthesized by the pyrolysis of precursors containing transition metal, N, and C species; metal atoms tend to aggregate into large particles due to their high surface energy, and their porous structures cannot be effectively controlled [15–17], which leads to relatively poor catalytic activity. Currently, MOF-derived catalysts have attracted significant research interest due to their potential applications in various fields of energy conversion and storage [18–22]. This has motivated us to explore an efficient method to prepare ideal electrocatalysts with favorable properties, including abundant N species, high surface areas, and adjustable pore sizes for the ORR ^{[5, 23]}.

Herein, we report a facile and reliable route for the controllable synthesis of Fe-incorporated and N-doped carbon dodecahedron nanoarchitectures (Fe-NCDNA), which were exploited as electrocatalysts for the ORR and were applied as cathode electrocatalysts in a Zn-air battery. Both results consistently confirmed the highly attractive catalytic properties of Fe-NCDNA with activity comparable to those of Pt-based materials as well as excellent durability.

Zinc nitrate hexahydrate (Zn(NO₃)₂·6H₂O), 2-methylimidazole (C₄H₆N₂), ferrous sulfate heptahydrate (FeSO₄·7H₂O), absolute methanol (CH₃OH), and absolute ethanol (C₂H₅OH) were purchased from Sinopharm Chemical Reagent Co., Ltd. Nafion solution (5%) and 1, 10-phenanthroline (C₁₂H₈N₂) were purchased from Aladdin Reagent (Shanghai) Co., Ltd. All the reagents were used directly without further purification.

Typically, Zn(NO₃)₂·6H₂O (2.000 g) and 2-methylimidazole (1.813 g) were dissolved in 45 and 30 mL of methanol, respectively. Subsequently, a 2-methylimidazole solution was injected in a Zn(NO₃)₂·6H₂O solution under vigorously stirring for 16 h at room temperature. The precipitates were centrifuged and washed with methanol and deionized water several times, followed by drying in a vacuum oven at 60 ℃ overnight.

The powders of ZIF-8 (0.300 g) and 1, 10-phenanthroline (0.136 g) were added to 20 mL of methanol and sonicated for 30 min at room temperature. Thereafter, 20-mL aliquots of methanol solutions containing 0.012, 0.015, and 0.200 g of FeSO₄·7H₂O were separately injected. After vigorously stirring for 2 h at room temperature, the mixture was evaporated to dryness at 80 ℃. Finally, the Fe-NCDNA-X (X = 1, 2, 3) samples were obtained by calcination at 900 ℃ for 2 h at a heating rate of 5 ℃ min^–1 under Ar atmosphere. Under the same conditions, Fe-NCDNA-0 was synthesized without adding FeSO₄·7H₂O. NC was prepared by the pyrolysis of ZIF-8 under the same annealing condition.

X-ray powder diffraction (XRD) patterns were recorded on an X-ray diffractometer (Miniflex6000, Rigaku) at 40 kV and 15 mA using Cu K_α radiation (λ = 1.54178 Å). The specific surface area and pore size distributions of the samples were analyzed by the Brunauer–Emmett–Teller (BET) method using N₂ adsorption–desorption isotherms on a Micromeritics Instrument Corporation sorption analyzer (Micromeritics TriStar Ⅱ 3020). The Raman spectra of the catalysts were recorded on a LabRAM HR over a range of 300–3000 cm^–1. The Fe concentrations of the catalysts were determined by inductively coupled plasma atomic emission spectroscopy (ICP-MS). The morphologies of the catalysts were determined using Hitachi SU8010 at a high vacuum pressure and an accelerating voltage of 5 kV. A scanning electron microscope (SEM, KYKY-EM3900M) and a transmission electron microscope (TEM, JEM-2010F) were used to observe the morphology and nanostructure of the materials. X-ray photoelectron spectroscopy (XPS) measurements were carried out on an ESCALAB 250Xi XPS system using a monochromatic Al Kα source and a charge neutralizer.

The electrochemical measurements were conducted in a three-electrode system on a CHI760 electrochemical station (CH instrument Co.) at room temperature (25 ℃). A carbon electrode served as the counter electrode. Meanwhile, a Hg/HgO in 1 M KOH solution served as the reference electrode in alkaline medium, and an Ag/AgCl electrode in saturated KCl served as the reference electrode in neutral and acidic media. A rotating disk electrode (RDE) with a glassy carbon (GC) disk of 3 mm in diameter and a rotating ring-disk electrode (RRDE) with a GC disk (4.0 mm diameter) and a Pt ring (5.0 mm inner diameter and 7.0 mm outer diameter) served as the substrate for the working electrode. Experimentally, 5.0 mg of the sample (containing 20 wt% Pt/C) was ultrasonically dispersed in 1 mL of the solution containing 50 μL of Nafion solution, 200 μL of ethanol, and 750 μL of water for 30 min to form homogeneous catalyst ink. The volumes of the catalyst ink were 6 μL and 10 μL for the RDE and RRDE tests, respectively. The O₂-saturated 0.1 M KOH, 0.1 M phosphate buffer solution (PBS, pH = 7), and 0.5 M H₂SO₄ were used as alkaline, neutral, and acidic media, respectively. For the CV measurements, the scan rate was 50 mV s^–1, while for the RDE/RRDE test, it was 5 mV s^–1.

All the potentials in this work were converted to the reversible hydrogen electrode (RHE) potential, and the potential conversion formula is as follows: E_RHE = E_Hg/HgO + 0.059 V × pH (alkaline medium) and E_RHE = E_Ag/AgCl + 0.059 V × pH (neutral/acidic medium).

The electron transfer number (n) was obtained by the Koutecky-Levich (K–L) Eq. (1):

(1)

where J is the measured current density; J_K and J_L are the kinetic and limiting current densities, respectively; w is the linear rotation speed (rpm s^–1) of the disk.

For the RRDE measurements, the polarization curves were recorded at a disk rotation rate of 1600 rpm. The potential of the ring was set at 1.4 V (vs. RHE). The H₂O₂ yield and the electron transfer number (n) were calculated by Eqs. (2) and (3), respectively, as follows:

(2)

(3)

where I_d is the disk current, I_r is the ring current, and N (0.442) is the ring collection efficiency.

Ten milligrams of the sample was ultrasonically dispersed in 1 mL of the solution containing 100 μL of Nafion solution, 450 μL of ethanol, and 450 μL of water for 30 min to form homogeneous catalyst ink. The volume of catalyst ink coated onto carbon paper (1 cm²) was 100 μL, after which the coated carbon paper was dried with a lamp. A Zn plate served as the anode electrode and a 6 M KOH solution was adopted as the electrolyte. The open-circuit voltage (VOC) and galvanostatic discharge data were collected in the LAND testing system. The specific capacitor of the assembled battery was calculated from Eq. (4):

(4)

where i, t, and △m represent the discharge current, discharge time, and the weight of consumed Zn, respectively.

It is a classic reaction, 1, 10-phenanthroline and Fe²⁺ form a red-brown Fe-phenanthroline (Fe-phen) product [26, 27]. And N-rich carbon can be obtained by the direct pyrolysis of ZIF-8. Therefore, the catalysts containing Fe and N were prepared through the carbonization of ZIF-8 with absorbed Fe-phen. During the carbonization process, the N group combined with the Fe species to form FeN_x structure units, which were confined to the porous carbon structure derived from ZIF-8. Fig. 1 schematically illustrates the synthetic process for preparing Fe-NCDNA. The entire synthesis process was implemented by the initial in situ growth of Fe-phen complexes on the surface of ZIF-8 through a reaction between Fe²⁺ and 1, 10-phenanthroline, noting that 1, 10-phenanthroline can be favorably adsorbed on ZIF-8 through π–π interactions [28, 29], which leads to a uniform coating of Fe-phen complexes on ZIF-8. A set of Fe-NCDNA catalysts can be obtained by pyrolyzing the composites of Fe-phen-coated ZIF-8 (ZIF-8@Fe-phen).

According to the XRD pattern (Fig. S1(a)), the as-prepared ZIF-8 catalyst shows diffraction peaks in accordance with the previous result [25], indicating the successful synthesis of ZIF-8. Simultaneously, the prepared ZIF-8 and ZIF-8@Fe-phen are determined by SEM observation to be regular dodecahedrons (Fig. S2). After a high-temperature pyrolysis reaction, ZIF-8 and ZIF-8@Fe-phen were converted to N-rich carbon and Fe, and N co-doped carbon, respectively. The XRD patterns (Fig. 2(a)) of NC and Fe-NCDNA-0 display two broad peaks near 26° and 44°, which are assigned to the (002) and (101) crystal planes of graphitic carbon [30, 31]. Although the peaks of Fe, FeC_x, and FeN_x, located around 44°, are observed for the Fe-NCDNA-3, Fe-NCDNA-2, and Fe-NCDNA-1 materials, respectively, these peaks are not sharp. This may be due to the low loading and the presence of amorphous Fe [32]. ICP-MS results confirm that the final Fe contents are 2.61 wt%, 3.30 wt%, and 4.34 wt% for Fe-NCDNA-1, Fe-NCDNA-2, and Fe-NCDNA-3 catalysts, respectively.

To further unveil the texture of carbon, the Raman spectra of the catalysts were recorded. Fig. 2(b) presents two distinct peaks in the range of 1250‒1750 cm^–1 corresponding to the D (around 1320 cm^–1) and G bands (around 1610 cm^–1). Mostly, the D-band is attributed to the defect sites or grain boundaries, while the G-band corresponds to the band stretching of sp² bonded pairs [33, 34], such as C=C, C=O, and N=C. After normalization, the area ratio of the D-band and G-band (I_D/I_G) can indicate the defect induced by N-doping. Evidently, the I_D/I_G values for NC and Fe-NCDNA-0 are 1.20 and 1.34, respectively, which illustrates that the introduced phenanthroline induced the formation of defects during the pyrolysis process. Conversely, the I_D/I_G value decreases gradually as the Fe content increases, and the value drops from 1.34 to 1.19; this suggests a decrease in the number of defects and an improved degree of graphitization. Since Fe is an excellent catalyst to induce graphitization [33], the inductive effect increases with the Fe content; thus, the I_D/I_G value decreases.

The N₂ adsorption-desorption test was conducted to evaluate the porosity and structural features. Fig. S1(b) depicts the corresponding N₂ adsorption-desorption isotherm of ZIF-8, which is a type-Ⅰ curve. Similarly, it addresses the microporous characteristics (inset in Fig. S1(b)) and typical reversible adsorption. Contrarily, all the catalysts deliver a large specific surface area. As shown in Fig. S3(a), there is a distinct hysteresis loop at P/P₀ of 0.4~1.0, which is a distinctive feature of mesoporous materials. Nevertheless, at a low relative pressure (P/P₀ < 0.40), the adsorption quantity declines sharply, indicating the existence of micropores in all the materials. Meanwhile, we calculate the BET surface area of the samples. The BET surface area of Fe-NCDNA-0 is calculated to be 1240 m² g^–1, which is higher than those of NC (1124 m² g^–1), Fe-NCDNA-1 (809 m² g^–1), Fe-NCDNA-2 (897 m² g^–1), and Fe-NCDNA-3 (649 m² g^–1). A downtrend is observed with the specific surface area when Fe doping is included, and it could be the result of Fe occupying the microporous structure [4, 23]. In addition, Fe-NCDNA-2 has the highest specific surface area among the Fe-doped samples and retains the most abundant pore size distribution compared to those of all the other catalysts (Fig. S3(b)). Combined with the results of the electrochemical test, such a hierarchical pore structure could supply more sites and enhance the accessibility of oxygen and electrolyte [35].

SEM was employed to investigate the morphology of the samples. Fig. S2 shows that the synthetic ZIF-8 is a regular dodecahedron, which is in agreement with previous reports. From the SEM images in Fig. 3, it is clear that all the catalysts retain the similar morphology derived from ZIF-8. Meanwhile, the morphologies of the catalysts changed in varying degrees due to the adjustment of the Fe content. Among them, Fe-NCDNA-2 revealed remarkable distortion and a rough surface. Therefore, according to the N₂ adsorption–desorption test and SEM images, the the amount of Fe not only affect the surface area and pore size distribution but also the morphology.

Furthermore, TEM was adopted to gain insight into the construction of the as-prepared Fe-NCDNA-2 catalyst. Figs. 4(a), 4(b), and Figs. S4(a), S4(b) show that Fe-NCDNA-2 has a carbon nanocage structure with in situ-formed carbon nanosheets. However, the nanoparticles commonly found in Fe, N co-doped carbon materials were not be observed. Meanwhile, the high-resolution TEM (HRTEM) image did not manifest lattice fringes and diffraction rings due to the low content of Fe and the existence of amorphous carbon (Fig. 4(c)) [32, 36, 37]. As displayed in the high-annular dark-field scanning TEM (HAADF-STEM) images (Fig. 4(d)) and corresponding elemental mapping images of the resultant Fe-NCDNA-2 catalyst, a uniform distribution of trace Fe (yellow), N (green), and carbon elements (red) exists in the carbon frameworks.

XPS analysis was performed to identify the elemental composition and surface chemical binding states of the samples. The results clearly reveal that the doping of Fe and N atoms in the carbon matrix was successful (Fig. 5(a)). The C 1s high-resolution spectra (Fig. 5(d)) show that there are two distinct peaks (located at 284.8 and 285.7 eV) and a broad shoulder peak (288.3 eV), which are assigned to C=C, C=N, and C-N, respectively [23]. Furthermore, C=C accounts for more than half of the content; C=N comes second, and C-N has the least content (Fig. S5(b), Table S1). The N 1s high-resolution spectra (Fig. 5(c)) can be subdivided into four different groups: pyridinic N (398.5 eV, 34.77%), pyrrolic N or Fe-N (399.8 eV, 22.54%) [38–41], graphitic N (400.8 eV, 35.50%), and oxygenated N (402.3 eV, 7.19%) [42–44]. According to previous literatures [45, 46], pyridinic N is the main catalytic active site for the ORR in N-doped carbon, and the graphitic N improves the limiting current density. In addition, the content of pyridine N is much higher than that of pyrrolic N (or Fe-N), and the content of graphitic N increases gradually when additional N and Fe are introduced (Fig. S5(a), Table S1). Fig. 3(b) manifests the Fe (Fe 2p) survey spectrum. Based on the reports [47–49], the two distinct peaks at 710.2 and 713.8 eV are attributed to the binding energies of the 2p_3/2 orbitals of the Fe²⁺ and Fe³⁺ species, respectively. For the 2p_1/2 orbitals, the peak at 723.3 eV is assigned to the binding energy of Fe²⁺. This suggests the existence of FeN_x and FeC_x species in Fe-NCDNA-2 [25, 50, 51]. Moreover, a satellite peak is observed at 719.5 eV.

The ORR activity of the electrocatalysts was firstly assessed by cyclic voltammetry (CV) in an Ar or O₂ saturated 0.1 M KOH solution at a scan rate of 50 mV s^–1. As shown in Figs. S6a and S6b, Fe-NCDNA-2 displays a noticeable oxygen reduction peak at 0.857 V vs. RHE, which is superior to those of Fe-NCDNA-1 (0.853 V), Fe-NCDNA-3 (0.835 V), Fe-NCDNA-0 (0.821 V), as well as 20 wt% Pt/C (0.817 V). As a reference, the NC catalyst exhibits a poor oxygen reduction peak of 0.796 V, indicating the vital role of Fe and N atoms in improving the ORR performance. To further evaluate the ORR performance, a linear sweep voltammetry (LSV) test with RDE was executed, and the results were compared to those of the commercial 20 wt% Pt/C catalyst (Fig. 6(a)). According to the plots, Fe-NCDNA-2 delivers the best ORR activity as it has the most positive half-wave potential (E_1/2 = 0.863 V) compared to those of Fe-NCDNA-1 (0.846 V), Fe-NCDNA-3 (0.840 V), Fe-NCDNA-0 (0.860 V), NC (0.807 V), and Pt/C (0.841 V). By comprehensive characterization and electrochemical results, it is apparent that N doping can improve the ORR activity. Moreover, the high-efficiency catalytic unit was formed to further enhance the catalytic activity due to the incorporation of Fe. Meanwhile, we noticed that the Fe content affected the structure of the sample, and Fe-NCDNA-2 could reveal more active sites [52], making its ORR performance the most outstanding. Subsequently, the Tafel slopes of the samples were obtained by analyzing the corresponding RDE polarization curves (Fig. S6(c)). Among them, Fe-NCDNA-2 presented the smallest slope of 80 mV dec^–1 compared to those of Fe-NCDNA-1, Fe-NCDNA-3, Fe-NCDNA-0, NC, and commercial Pt/C. This demonstrates that Fe-NCDNA-2 has a relatively high catalytic current and further corroborates its outstanding ORR activity [53]. For a better understanding of the electrochemical processes, RDE measurements were employed at different rotating rates, and the corresponding kinetic parameters of Fe-NCDNA-2 were calculated using the K−L equation (Figs. 6(b), 6(c)). The computed value is 3.87, which is close to the theoretical value of 4.0 for the ORR. This indicates the efficient reduction of oxygen over the electrocatalyst via a four-electron reduction pathway. Based on the RRDE data, the yield of peroxide species (HO₂^–) and the transferred number of electrons can be determined (Fig. S6(d)). Similarly, the transferred number of electrons is consistent with the calculated value from the K–L plots. Moreover, the HO₂^– yield over Fe-NCDNA-2 is less than 6% compared to 29.6% over the commercial Pt/C in the range of 0.2–0.9 V. This implies that the material has a relatively high four-electron pathway selectivity and avoids the two-electron pathway to peroxide. Chronoamperometric tests (Fig. 6(d)) show that Fe-NCDNA-2 exhibits better long-term stability (80%) than Pt/C (64%), i.e., retention of the initial current, after a 30000 s run in a 0.1 M KOH solution. Methanol tolerance test was employed to assess the poison tolerance of the catalyst. Fig. S7(a) shows that the curves almost coincide, while there is significant oxidation peak in Fig. S7(b), revealing that the Fe-NCDNA-2 catalyst exhibits excellent poison tolerance compared to that of Pt/C.

When in a neutral medium, a set of Fe-NCDNA catalysts also exhibits catalytic properties toward the ORR. For the CV tests (Figs. S8(a), S8(b)), Fe-NCDNA-2 displays an oxygen reduction peak at 0.689 V, which is more positive than that of Pt/C. Similar to the case in alkaline medium, the E_1/2 value of Fe-NCDNA-2 is 0.715 V, suggesting a positive potential compared to those of Fe-NCDNA-1 (0.692 V), Fe-NCDNA-3 (0.628 V), Fe-NCDNA-0 (0.631 V), NC (0.497 V), and Pt/C (0.673 V) (Fig. 7(a)). In addition, the Tafel slopes of the catalysts were obtained from analyzing the corresponding LSV data in RDE at 1600 rpm. Fig. S8(c) demonstrates that Fe-NCDNA-2 has the lowest slope of 105 mV dec^–1. RDE measurements at different rotating rates (Fig. 7(b)) and RRDE measurement at 1600 rpm (Fig. S8(d)) were also conducted to evaluate the electrochemical behavior. From the results, the H₂O^– yield over Fe-NCDNA-2 is less than 8% over the range of 0.1–0.8 V, revealing that the evolution of H₂O^– is dramatically suppressed on the Fe-NCDNA-2 electrocatalyst. Further, the transferred number of electrons determined from the K–L plots (Fig. 7(c)) matches the calculated value obtained by RRDE measurement (close theoretical value of 4.0). This illustrates that the ORR processes over the Fe-NCDNA-2 catalyst occur preferably via a four-electron reduction pathway. As depicted in Fig. 7(d), Fe-NCDNA-2 and Pt/C have similar long-term stabilities with 77% and 76% retention of the initial current, respectively, after catalyzing for 30000 s in 0.1 M PBS solutions. Table 1 lists the electrocatalytic ORR performance parameters of the prepared materials and depicts the great competitiveness over the Pt/C and some Pt-free catalysts in both alkaline and neutral media. [41, 54–58] In addition, Fe-NCDNA-2 has an E_1/2 value of 0.745 V in the O₂-saturated 0.5 M H₂SO₄ solution, which is negative compared with 0.783 V for Pt/C (Fig. S9(a)). Systematic electrochemical studies demonstrate that Fe-NCDNA-2 follows a four-electron reduction pathway and has a high long-term stability (Figs. S9(b)-9(d) and Fig. S10).

A home-made Zn-air battery was constructed to further demonstrate the practical application of Fe-NCDNA-2 as an ORR electrocatalyst. The Zn-air battery was constructed using the Fe-NCDNA-2 catalyst loaded on carbon paper as the air-cathode and Zn plate as the anode in a 6 M KOH electrolyte (Fig. 8(a)). For comparison, the 20 wt% Pt/C catalyst loaded on carbon paper was used as the air-cathode and was also measured under the same condition. The V_OC of the batteries were recorded for 12 h to verify that their voltages are close and that the assembled Zn-air batteries have low self-discharge (Fig. 8(b)). The battery employing the Fe-NCDNA-2 catalyst exhibited a V_OC of 1.46 V; this value is slightly higher than that of Pt/C (1.43 V). The power density is not only one of the important indicators for evaluating battery performance, but also reflects the catalytic activity of the material. It is worth noting that the maximum power density for the Zn-air battery employing the Fe-NCDNA-2 catalyst as the cathode is calculated to be 184 mW cm^–2, which is higher than that of the commercial Pt/C (173 mW cm^–2, Fig. 8(c)). Moreover, from the inset in Fig. 8(c), it is evident that a red LED light (~2.0 V) is lit by two Zn-air batteries connected in series. During the discharge process at a current density 10 mA cm^–2, the batteries with the Fe-NCDNA-2 catalyst presented better performance than that of the commercial Pt/C, and their voltage plateaued at 1.28 V with a specific capacity of 801 mAh g^–1 when normalized to the weight of the consumed Zn. This value is slightly more positive than that of Pt/C (799 mAh g^–1, Fig. 8(d)). The above results are superior to many reported results [52, 53, 58–62] and further confirm a comparable performance and application of Fe-NCDNA-2 toward the ORR. To acquire the rate discharge performance curves, the batteries employing the optimal Fe-NCDNA and Pt/C catalysts as the air cathode were employed for a discharge operation at current densities of 2, 4, 6, 8, 10, and 2 mA cm^–2 (Fig. 8(e)). The voltage decreased steadily as the discharge current density increased. Furthermore, there was no noticeable voltage drop during 30 h of operation. As described in Fig. 8(f), the Zn-air battery employing the Fe-NCDNA-2 catalyst exhibits a high cycling stability when repeatedly subjected to discharge and charge processes at 10 mA cm^-2 for 600 cycles without any distinct augment in the voltage gap of the charge and discharge processes. In addition, the battery with Fe-NCDNA-2 has a smaller voltage polarization, especially after a long cycle, compared to that of the device with Pt/C.

In summary, we reported a facile route for the controllable synthesis of a catalyst with a carbon dodecahedron nanoarchitecture and formed in situ carbon nanosheets loaded uniformly with Fe-N-C species. Systematic electrochemical tests demonstrated that the Fe-NCDNA-2 electrocatalysts exhibited a more positive half-wave potential, a lower Tafel slope, and better long-term stability than the commercial Pt/C in both alkaline and neutral media. The home-made Zn-air batteries employing the optimal Fe-NCDNA catalyst also verified the feasibility of its practical applications as an ORR electrocatalyst. The present work can inspire researchers to explore similar N-doped transition metal porous carbon catalysts for a variety of applications.