Chromium phosphide nanoparticles embedded in porous nitrogen‐/phosphorus‐doped carbon as efficient electrocatalysts for a nitrogen reduction reaction

Abstract

The resource recovery of heavy metals from effluent has significant environmental implications and potential commercial value. Chromium phosphide nanoparticles embedded in a nitrogen‐/phosphorus‐doped porous carbon matrix (CrP/NPC) are synthesized via a consecutive Cr⁶⁺ leachate treatment and resource recovery process. Electrochemical testing shows that CrP/NPC shows excellent nitrogen reduction reaction (NRR) performance, which yields the highest NH₃ production rate of 22.56 μg h⁻¹ mg⁻¹_cat. and Faradaic efficiency (16.37%) at −0.5 V versus the reversible hydrogen electrode in a 0.05 M Na₂SO₄ aqueous solution, as well as robust catalytic stability. The isotopic experiments using ¹⁵N₂ as a nitrogen source confirm that the detected NH₃ is derived from the NRR process. Finally, density functional theory (DFT) calculations show that the electron deficiency environment of the Cr site can significantly reduce the barrier of the NRR process and promote the formation of intermediate species.

INTRODUCTION

Ammonia (NH₃) is one of the largest in-demand chemicals (about 170 Mt year⁻¹) in the modern industry, which is widely used for fertilizers, refrigerants, or chemical feedstocks.^1–3 Recently, the Haber–Bosch technology has become the leading artificial synthesis method for NH₃, but more than 1% of the global annual energy is consumed due to the high-temperature and high-pressure process, which simultaneously generates vast carbon dioxide emissions.^4–6 As a promising alternative process, the electrochemical nitrogen reduction reaction (NRR) to NH₃ driven by renewable energy under normal temperature/pressure conditions has received considerable attention in the past few years.^7–10 However, because of the robust N≡N bonds and the competition of hydrogen evolution reactions (HER), it is essential to develop efficient electrocatalysts with sufficient activity and selectivity.^11–15 Fortunately, inspired by natural nitrogenase, the active center of the electrocatalyst should have the ability to adsorb N₂ molecules and activate the N≡N triple bond.^16–19 In principle, an ideal active center has the ability to combine unoccupied and occupied d-orbitals to synergically accept electron density from and back-donate to N₂.^11,20 To this end, the search for an electrocatalytic NRR active center, mainly focused on Mo, Fe, Rh, Ru, and others, has made great progress in recent years.^21–25 It is worth noting that although the transition metal has unoccupied orbitals and electron donation positions, the d-orbital electrons also facilitate the formation of metal-H bonds,^26–29 which intensifies the competitive HER and inhibits the activity of NRR. In view of this, developing an effective active center with a high affinity for NRR and suppression of the HER competitive reaction is imperative but still challenging.

According to available theoretical calculations toward NRR, one of the promising active centers on the volcano map was the Cr element, which had the ability to activate the N₂ and form weak cohesion to proton.⁶ In recent experimental studies, it was proven that the Cr-based components could serve as efficient electrocatalysts toward NRR.^30–32 For example, Zhang et al.³⁰ reported multishelled hollow Cr₂O₃ microspheres (MHCMs) as high-performance NRR electrocatalysts, which achieved the highest NH₃ production rate (25.3 μg h⁻¹ mg⁻¹_cat.) and Faradaic efficiency (FE; 6.78%) at −0.9 V versus reversible hydrogen electrode (RHE). Nevertheless, most of the reported Cr-based components were oxides; however, the poor conductivity of oxide making large overpotential to trigger the NRR process. Recently, transition metal phosphides (TMPs) as efficient NRR electrocatalysts have attracted considerable attention because of their high conductivity, controllable components, and special physical–chemical properties.^27,33–36 More importantly, the internal electronic structure of the active center can be regulated by the phosphorus binding to the transition-metal lattice, so as to improve the inherent NRR catalytic activity of the active center.^35,36 For instance, Su et al.³⁵ reported small RhP₂ nanoparticles loading on N, P co-doped carbon (RhP₂/NPC) as an efficient NRR electrocatalyst, which demonstrated that the electronic structure and catalytic behavior of Rh were modified by incorporating the P elements, and the NH₃ production rate of RhP₂/NPC was 1.8 times higher than that of the Rh/NPC. Nevertheless, most TMPs were synthesized by the gas–solid reaction of the decomposition of excessive NaH₂PO₂ and red phosphorus, which produced a large amount of toxic and flammable phosphine tail gas. Therefore, it is important to find an environmentally friendly and high-efficiency method to synthesize TMPs toward NRR.

Herein, an efficient NRR electrocatalyst (CrP/NPC) was synthesized by a consecutive bio-assisted Cr⁶⁺ leachate treatment and resource recovery. In this process, the CrP/NPC was obtained via a solid–solid reaction using microorganisms as the biological phosphorus source. The CrP nanoparticles embedded in NPC can endow these composites with more accessible active sites, excellent conductivity, and better stability. Electrochemistry measurements showed that the obtained CrP/NPC had excellent NRR performance, with a high NH₃ yield (22.56 μg h⁻¹ mg⁻¹_cat.) and FE (16.37%) at −0.5 V versus RHE in a 0.05 M Na₂SO₄ aqueous solution. Density functional theory (DFT) calculations demonstrated that the introduction of P affected the intrinsic electronic structure of Cr, resulting in an effective and active strong N≡N bond, and thereby resulting in increased NRR catalytic activity.

RESULTS AND DISCUSSION

The CrP/NPC was synthesized via a consecutive Cr⁶⁺ leachate treatment and resource upgrading process, which is schematically illustrated in Scheme 1. The optical photograph of the overall experimental process is shown in Figure S1. First, the simulated Cr⁶⁺ ion leachate was mixed with a certain quality of microorganism. In this study, saccharomycetes, which were found to be rich in N, P, and C elements (Figure S2) and surface functional groups (Figure S3), were used as model bioadsorbents because they were able to efficiently uptake highly toxic heavy-metal ions,^37,38 and the adsorption capacity of saccharomycetes for Cr⁶⁺ was about 30 mg g⁻¹ (Figure S4). Then, the solution gradually became clear and yellow aggregates (Cr/microorganism composite) were observed at the bottom of the solution, indicating that a large number of Cr⁶⁺ ions were immobilized in the microorganism. Although no significant difference was observed in the X-ray diffraction (XRD) patterns of pristine microorganisms and the Cr/microorganism composite (Figure S5), the morphological characteristics of microorganisms have changed obviously after uptake of Cr⁶⁺. As can be seen in the field emission scanning electron microscopy (FESEM, Figure S6a) images and the biologically sectioned transmission electron microscopy (TEM, Figure S7a) images of the pristine microorganism, the surface and cell wall are smooth without any impurities. As for the Cr/microorganism, the surface and the cell wall become coarse and abundant nanoparticles are observed (Figure S6b and Figure S7b), demonstrating that the Cr⁶⁺ ions are successfully reduced to Cr nanoparticles. Subsequently, we analyzed the possible adsorption/reduction mechanism in this uptake process. The positively charged Cr⁶⁺ ions were incipiently trapped on the surface and the cell wall by electrostatic interactions, and the Zeta potential of the microorganism was about −22.91 mV.²⁷ Then, Cr⁶⁺ ions were fully diffused and penetrated into the cytomembrane, followed by combining with proteins/enzymes.^39,40 Finally, a process of resource utilization was carried out and the Cr/microorganism composites were calcined under an H₂ atmosphere to obtain porous CrP/NPC, which could be used as an efficient electrocatalyst for NRR.

Scheme 1. Schematic illustration of the synthesis process of the chromium phosphide nanoparticles embedded in a nitrogen-/phosphorus-doped porous carbon matrix (CrP/NPC). NRR, nitrogen reduction reaction

The crystal structure of the as-obtained CrP/NPC was first analyzed by XRD. As shown in Figure 1A, the XRD pattern of CrP/NPC show distinct diffraction peaks located at 29.6°, 32.2°, 33.3°, 34.1°, 36.4°, 44.4°, 45.0°, 47.0°, 48.4°, 53.3°, 54.3°, 57.3°, 59.1°, 70.0°, 71.6°, and 77.8°; all of these peaks matched well with the orthorhombic CrP (JCPDS No. 29-0456). In addition, the relatively broad peak and sharp peak located at around 23° and 46° corresponded to the (002) and (100) facets of graphitic carbon, respectively. The morphology and microstructure of the CrP/NPC were then investigated by FESEM and TEM. The typical FESEM images of the obtained CrP/NPC are shown in Figure 1B,C, which show a porous elliptical microsphere shape with a size of around 1–2 µm and a large number of nanoparticles embedded on the surface, which were considered to be CrP. To further demonstrate the structure of CrP/NPC, TEM measurements were performed. As shown in Figure 1D, many dispersed particles 50–100 nm in size were observed on the porous carbon matrix, consistent with the FESEM images. Interestingly, curled graphene can be observed around some of the particles (Figure 1E). The selected area electron diffraction (SAED) pattern of selective nanoparticles (Figure 1F) confirm the monocrystal nature. In Figure 1G, the high-resolution TEM (HRTEM) image reveal well-resolved lattice fringes with an interplanar distance of 0.18 nm, which corresponded to the (211) plane of CrP. The corresponding high-angle annular dark-field scanning TEM (HAADF-STEM) image and energy-dispersive X-ray spectroscopy (EDS) elemental mappings of the CrP/NPC are shown in Figure 1H,I. Significantly, the Cr and P element distributions were found to be strictly superposed to the bright dots in the HAADF-STEM image, confirming that the nanoparticles in the porous carbon matrix are all of CrP. Meanwhile, the P and N elements are evenly distributed in the C matrix. All of the above results confirm the successful synthesis of CrP/NPC.

Figure 1. (A) X-ray diffraction pattern, (B,C) field emission scanning electron microscopy image, (D,E) transmission electron microscopy image (TEM), (F) selected area electron diffraction, (G) high-resolution TEM image, (H) high-angle annular dark-field scanning TEM image, and (I) corresponding energy-dispersive X-ray spectroscopy elemental mapping of CrP/NPC (chromium phosphide nanoparticles embedded in a nitrogen-/phosphorus-doped porous carbon matrix)

The physicochemical characteristics of CrP/NPC were further investigated by N₂ adsorption–desorption isotherms, Raman spectroscopy, and X-ray photoelectron spectroscopy (XPS), as shown in Figure 2. The porous structure of CrP/NPC was determined by the N₂ adsorption–desorption isotherms (Figure 2A). A hysteresis loop was observed, indicating the exist of mesoporous in CrP/NPC. Also, the Brunauer–Emmett–Teller (BET) surface area of the CrP/NPC was measured to be 334.86 m² g⁻¹, which is much larger than that of the pristine microorganisms (1.14 m² g⁻¹) and the Cr/microorganisms (2.14 m² g⁻¹), indicating that the high BET surface of CrP/NPC is derived from its porous structure. In addition, the corresponding pore size distribution confirm that abundant micropores and mesoporous existed in CrP/NPC. The high porosity and specific BET surface area of CrP/NPC was befitted to the electrocatalyst for exposing more active sites for NRR. The graphitic carbon nature of CrP/NPC was further examined by Raman spectroscopy (Figure 2B). The Raman peaks at 1343, 1587, and 2800 cm⁻¹ correspond to the graphite D-band, G-band, and 2D-band, respectively. The lower I_D/I_G (0.86) ratio of CrP/NPC than NPC (1.08) indicate that more crystallized graphitic carbon exist in CrP/NPC, which could improve the conductivity of the electrocatalyst. In addition, the broad peak shape of the 2D band demonstrate the existence of multilayer graphene in the CrP/NPC, which is in accordance with the TEM result in Figure 1E.

Figure 2. (A) N2 adsorption–desorption isotherms. (B) Raman spectroscopy. High-resolution X-ray photoelectron spectroscopy (XPS) spectra for (C) Cr 2p, (D) P 2p, (E) N 1s, and (F) C 1s

The electrocatalytic reactions usually take place on the surface/interface of the electrocatalyst. Therefore, to better understand the chemical composition and electronic structure at the surface of the CrP/NPC, XPS was performed. In Figure S8, the survey spectrum of CrP/NPC show the existence of Cr, P, N, and C elements. In Figure 2C, the high-resolution XPS peaks of the Cr 2p peaks of CrP/NPC reveal that the binding energies (BEs) of 575.65 and 585.05 eV could be assigned to the Cr–P bond in CrP, and the other two peaks (577.2 and 587.1 eV) could be assigned to the oxidized species CrP because of exposure to air. Compared with the BEs of CrP (574.15 and 583.55 eV), this BE shifted by about 1.50 eV in the positive direction, suggesting the electron-deficient state of Cr in CrP/NPC. This result may be attributed to the strong electron transfer between CrP and NPC. The BEs of the P 2p in CrP/NPC (Figure 2D), with the peaks located at 128.4 and 129.3 eV, could be attributed to the P species in phosphide and had a 0.9 eV negative shift compared to CrP, indicating the electron-rich state of P in CrP/NPC. Furthermore, the signal at BE of 129.9 eV was attributed to P–C (P doping). As for the N species, pyrrolic-N (400.1 eV) and pyridinic-N (398.8 eV) could be distinguished from the N 1s spectrum (Figure 2E). The two components fitted from the XPS spectrum of C 1s (Figure 2F) were assigned to sp2 hybridization of C–C bonding (284.5 eV) and nonmetallic doped C (285.4 eV), confirming the N/P doping in graphite carbon.

The electrocatalytic NRR performance for CrP/NPC was evaluated in a 0.05 M Na₂SO₄ aqueous solution using a two-compartment cell separated by a proton-exchanged membrane. To ensure the accuracy of the results, before the electrochemical test, a series of blank tests were conducted to eliminate distractions.⁷ Polarization curves of CrP/NPC in both Ar- and N₂-saturated electrolytes show different current densities from −0.1 to −0.6 V (Figure S9), indicating the occurrence of the NRR process. Figure 3A shows the i–t curves at different potentials for 2 h. After the N₂ electrolytic reaction, Nessler's analysis (Figure S10) show that the NH₃ production rate increased from −0.1 to −0.5 V and reached a maximum of 22.56 μg h⁻¹ mg⁻¹_cat. at −0.5 V (Figure 3B). It is noteworthy that the decrease in the NH₃ yield rates at −0.6 V could be ascribed to the competitive HER. Furthermore, the relationship between the NH₃ yield and the current reveal that the highest FE up to 16.37% was reached at −0.5 V. The highest NH₃ production rate and FE of CrP/NPC were comparable to most of the recently reported non-noble metal-based electrocatalysts, which are summarized in Table S1.

Figure 3. NRR behavior of CrP/NPC. (A) Chronoamperometry curves of CrP/NPC at various potentials for 2 h in N2-saturated 0.05 M Na2SO4. (B) NH3 yield rate and Faradaic efficiency at different potentials. (C) 1H-NMR spectra (600 MHz) of both 14NH4+ and 15NH4+ produced from the NRR reaction (at −0.5 V vs. RHE) using 14N2 or 15N2 as the N2 source. (D) Cycling stability results at −0.5 V versus RHE. CrP/NPC, chromium phosphide nanoparticles embedded in a nitrogen-/phosphorus-doped porous carbon matrix; NMR, nuclear magnetic resonance; NRR, nitrogen reduction reaction; RHE, reversible hydrogen electrode

We also examined the NRR performance of CrP/NPC prepared at different pyrolysis temperatures (700°C to 1000°C), and the optimal pyrolysis temperature could be set at 800°C (Figures S11 and S12). In addition, to further understand the active component of CrP/NPC, control samples (CrP and NPC) were synthesized, and the corresponding XRD patterns and FESEM images are shown in Figures S13–S15, respectively. As shown in Figure S16, the NH₃ production rate of NPC was low (2.71 μg h⁻¹ mg⁻¹_cat.), indicating the inactive NRR performance. By contrast, CrP showed definite activity for NRR (15.34 μg h⁻¹ mg⁻¹_cat.), but still much lower than that of CrP/NPC. This result indicate that the high performance of CrP/NPC may be derived from the synergistic effect of CrP and NPC, and the possible reasons were analyzed as follows. First, although NPC was inactive for NRR, the improved conductivity of the sample could be found in electrochemical impedance spectroscopy (Figure S17), which was beneficial to the electrochemical NRR processes. Second, the strong electron transfer between CrP and NPC (Figure 2C,D) could regulate the electron distribution of the active center, resulting in the higher NRR activity of CrP.

An isotopic experiment was conducted to identify the source of the nitrogen species of the detectable NH₃. As shown in Figure 3C, only ¹⁵NH₄⁺ was detected in the nuclear magnetic resonance test using ¹⁵N₂ as the feed gas, indicating that the NH₃ was generated by the NRR process. The stability testing of CrP/NPC in Figure 3D show that the NH₃ production rate and FE were not substantially changed, demonstrating the good catalytic stability of CrP/NPC for NRR. After stability testing, the morphological differences in FESEM of CrP/NPC were negligible (Figure S18), which confirm the NRR catalytic durability and the structural stability of the electrocatalyst.

To further understand the reaction mechanisms of NRR on CrP/NPC at the atomic level, DFT calculations were performed. From the electrochemical test results, the outstanding NRR performance of the CrP/NPC is derive due to the synergistic effect between CrP nanoparticles and N/P co-doped carbon matrix. To simplify the calculation model, the N₂ adsorption energies in CrP and NPC were first calculated to evaluate the intrinsic N₂ adsorption activity (Figure 4A). In the DFT calculation results, CrP had more negative adsorption energy (−0.42 eV) than NPC (−0.07 eV), indicating that the N₂ molecule had favorable adsorption on CrP but not for NPC. From the charge density difference in CrP after adsorption of N₂ (Figure 4B), we found that the nonbonding electrons in the N₂ molecule were incorporated into the electron deficiency area of CrP, resulting in the effective adsorption and then activation of N₂. In addition, drawing on the previous reports,^41,42 the recommended pathway for N₂ to NH₃ on CrP was calculated by the Gibbs energy (Figure 4C). The reasonable reaction path for N₂-to-NH₃ on CrP was *N₂ → *NNH → *NNH₂ → *N + NH₃ → *NH → *NH₂ → *NH₃ → NH₃. Significantly, the active free energy of N₂ on CrP was as low as 0.09 eV and the rate-determining step of the whole reaction was confirmed as *NH → *NH₂ (1.16 eV).

Figure 4. (A) Nitrogen adsorption energy of chromium phosphide nanoparticles (CrP) and nitrogen-/phosphorus-doped porous carbon (NPC). (B) Charge density difference of CrP after adsorption of N2. (C) Free energy of N2 reduction on CrP; the insets show intermediate configurations. The mazarine, pink, gray, and green balls represent Cr, P, N, and H atoms, respectively

CONCLUSION

In summary, we identified a Cr⁶⁺ wastewater treatment and resource upgrading route to prepare CrP/NPC as a valuable electrocatalyst for NRR. Significantly, the CrP/NPC showed superior NRR performance with an average NH₃ production rate of 22.56 µg h⁻¹ mg⁻¹_cat. and an FE of 16.37% at −0.5 V in a neutral solution. The isotopic experiments confirmed that the synthesized NH₃ was derived from the direct supply of N₂. Theoretical calculations confirmed that the electron-deficient Cr site could accommodate the lone–pair electrons of a N₂ molecule, resulting in efficient adsorption and activation of the strong N≡N bond, and thus facilitating the overall NRR. This study highlights the potential application of TMP for electrocatalytic NRR and provides new insights for preparing novel materials combined with wastewater treatment and upgradation toward applications in energy conversion and storage.

ACKNOWLEDGMENTS

This study was supported by Taishan Scholars Project Special Funds (tsqn201812083), the Natural Science Foundation of Shandong Province (ZR2019YQ20 and 2019JMRH0410), and the National Natural Science Foundation of China (51972147, 52022037 and 52002145).

CONFLICT OF INTEREST

The authors declare no conflict of interest.