Nano Energy 64 (2019) 103879
Contents lists available at ScienceDirect
Nano Energy journal homepage: www.elsevier.com/locate/nanoen
N,P co-doped hollow carbon nanoﬁber membranes with superior mass transfer property for trifunctional metal-free electrocatalysis
Yang Gaoa,b, Zhichang Xiaoc, Debin Kongb,**, Rashid Iqbalb, Quan-Hong Yanga,***, Linjie Zhia,b,* a
School of Chemical Engineering and Technology, Tianjin University, Tianjin, 300072, China Key Laboratory of Nanosystem and Hierarchical Fabrication, National Center for Nanoscience and Technology, Beijing, 100190, China c College of Science, Hebei Agricultural University, Baoding, 071001, China b
A R T I C LE I N FO
A B S T R A C T
Keywords: Hollow carbon nanoﬁber Oxygen reduction reaction Metal-free electrocatalysis Mass transfer
Carbon-based metal-free electrocatalysts have inspired extensive eﬀorts to explore their applications in many nontrivial electrochemical reactions, such as oxygen reduction reaction (ORR), oxygen evolution reaction (OER), and hydrogen evolution reaction (HER), by virtue of the integrated advantages including low cost, sustainability, longevity, and multifunctionality. Herein, N,P co-doped hollow carbon nanoﬁber (N,P–HCNF) membranes were facilely prepared via coaxial electrospinning technology, which are bestowed with a hierarchical porous architecture, turbostratic structures, and abundant catalytically active sites such as doping, defects, and edges. Beneﬁting from structural features of the one-dimensional (1D) carbon hollow nanoarchitecture, which aﬀords plentiful active sites, continuous conducting pathways, and benign mass transfer channels, the resultant catalyst reveals an excellent trifunctional electrocatalytic activity for ORR, OER, and HER. Impressively, it exhibits one of the best metal-free bifunctional electrocatalytic activities in oxygen electrocatalysis as characterized by a low potential deviation (ΔE) of 0.73 V between the half-wave potential (E 1/2) for ORR and the potential reaching 10 mA cm−2 (Ej=10) for OER. Signiﬁcantly, further investigations demonstrate that the eﬀect of mass transfer makes a great diﬀerence to electrocatalytic activity, mainly through enlarged speciﬁc surface area to aﬀect intrinsic catalytic activity and the ionic resistance in pores. This work sheds light on the design, fabrication, and regulation of highly active metal-free electrocatalysts with abundant active sites and tuned pore structures for electrocatalysis and other applications.
1. Introduction With the gradual increase of energy consumption and the burning issues of environmental crisis, there are extensive eﬀorts focusing on renewable energy storage and conversion technologies, including fuel cells , metal-air batteries , and water electrolysis . Oxygen reduction reaction (ORR) and oxygen evolution reaction (OER) play an important role in fuel cells and metal-air batteries, whilst hydrogen evolution reaction (HER) and OER are the two key half-reactions of electrocatalytic water splitting [4–7]. Accordingly, multi-functional electrocatalysts are highly desired and predicated to play critical roles due to their wide and valuable applications in future energy storage and conversion systems. Typically, noble metal catalysts are extensively used, in which Pt-based catalysts are eﬃcient for ORR and HER, whereas IrO2 and RuO2 are highly active towards OER. Nevertheless,
the high cost, low abundance, and poor stability of these noble metal catalysts impede the application at industrial scale. In view of the above-mentioned facts, well-designed transition metal oxides , chalcogenides , nitrides , carbides , and phosphides  have been testiﬁed to be eﬃcient catalysts, but their poor electrical conductivity and inferior catalytic activities compared to noble metal catalysts hinder them from large-scale commercialization. Furthermore, integrating high activities of all three above-mentioned reactions, namely constructing trifunctional electrocatalysts, is also conducive to facilitating device production and to curtailing manufacture cost. Therefore, the fabrication of eﬃcient, low-cost, and trifunctional electrocatalysts for ORR, OER, and HER faces an urgent need. Since the development of metal-free vertically aligned N-doped carbon nanotubes (VA-NCNTs) with an unexceptionable catalytic activity , carbon nanomaterials have received tremendous attention
Corresponding author. School of Chemical Engineering and Technology, Tianjin University, Tianjin, 300072, China. Corresponding author. *** Corresponding author. E-mail addresses: [email protected]
(D. Kong), [email protected]
(Q.-H. Yang), [email protected]
(L. Zhi). **
https://doi.org/10.1016/j.nanoen.2019.103879 Received 28 January 2019; Received in revised form 16 June 2019; Accepted 3 July 2019 Available online 10 July 2019 2211-2855/ © 2019 Elsevier Ltd. All rights reserved.
Nano Energy 64 (2019) 103879
Y. Gao, et al.
(DCDA = dicyandiamide) shell were obtained by coaxial electrospinning technology. After polymerization and pyrolysis, the novel porous N,P-HCNFs in the form of hollow structure were obtained (Fig. S1), with no need of hard templates and without additional procedures to remove the impurities . Herein, PAN and DCDA were deliberately combined to synthesize a catalyst exhibiting a hierarchical pore structure and a remarkable activity, and TPP was selected as the phosphorus source. Interestingly, the pore structures of N,P-HCNFs can be tuned by changing the concentration of PAN (noted as x% m/v) in shell solution, and the corresponding products are marked as N,P–HCNF-x (see the Experimental section for more details). As a typical case of N,P-HCNFs, the morphology of the as-obtained N,P–HCNF-8 was characterized by scanning electron microscopy (SEM) and transmission electron microscopy (TEM). As shown in Fig. 2a–b, the loosely stacked ﬁbrous network of the pristine ﬁlm converted to a compactly intertwined porous structure after polymerization and pyrolysis, which aﬀords continuous conducting pathways. The fractured surfaces of N,P–HCNF-8 clearly elucidate the formation of the hollow tubular structure with an average diameter of ~300 nm (Fig. 2c). With the increase of PAN concentrations, the N,P-HCNFs become wider in diameter along with a much clearer ﬁbrous morphology (Fig. S2). The persistent gray dots shown in the TEM image of the N,P–HCNF-8 ﬁber (Fig. 2d) conﬁrm the structural characteristic of porous walls. High resolution TEM (HRTEM) images shown in Fig. 2e and Fig. S3 reveal that the N,P–HCNF-8 contains a porous network and turbostratic graphitic structures with a large quantity of edges and defects, which could obviously promote electrocatalysis . The HAADF-STEM image and the corresponding element mapping images reﬂect the uniform distribution of C, N, O, and P elements in N,P–HCNF-8 (Fig. 2f–j). The observed porous structure can furnish abundant active sites and continuous 1D paths, both of which are favorable for electrocatalysis through improving mass transfer and electron transport. Further evidences conﬁrm the abundance of edges and defects in N,P-HCNFs. As shown in the Raman spectra in Fig. 3a, a relatively high ID (1348 cm−1)/IG (1586 cm−1) ratio (~0.9) reveals the low graphitization degree of N,P-HCNFs. Meanwhile, the corresponding X-ray diffraction (XRD) patterns (Fig. 3b) show a broad (002) diﬀraction peak at 2θ ≈ 23.5°, which denotes a greatly small graphitic domain size integrated with turbostratic alignment of basal planes within the domains . X-ray photoelectron spectroscopy (XPS) was used to analyze the chemical compositions of N,P-HCNFs. The survey spectra of N,P-HCNFs (Fig. S4) reveal the presence of signiﬁcant amounts of nitrogen (2.62–3.08 at%). Furthermore, the XPS N 1s spectra (Fig. 3c) reveal the presence of four distinguished nitrogen species, including pyridinic N (398.2 eV), pyrrolic N (399.6 eV), quaternary N (400.9 eV), and oxidic N (402.6 eV), with similar contents in diﬀerent samples (Fig. S5) . Remarkably, the dominant pyridinic and quaternary nitrogen species can eﬀectively aﬀect the chemical and electronic environments of the neighboring carbon atoms, thus facilitating electrocatalytic activity . The XPS P 2p spectrum (Fig. S6) of N,P–HCNF-8 is deconvoluted into two diﬀerent bands at 131.0 eV and 133.1 eV, which are assigned to P–C and P–O conﬁgurations, respectively . N2 adsorption measurements were performed to analyze the Brunaue-Emmett-Teller (BET) surface areas and pore structures of N,P-HCNFs. All the samples show a typical type IV isotherm curve with an obvious hysteresis, conﬁrming the existence of mesopores (Fig. 3d). The BET speciﬁc surface area of N,P–HCNF-8 is 760 m2 g−1, which is much larger than the values of N,P–HCNF-7.5 (599 m2 g−1), N,P–HCNF-8.5 (700 m2 g−1), N,P–HCNF9 (443 m2 g−1), and N,P–HCNF-9.5 (427 m2 g−1). Density functional theory (DFT) pore size distribution curves (Fig. 3e) derived from the N2 desorption brunches demonstrate the presence of hierarchical pore structures and markedly improved pore volumes (from 0.19 to 0.38 cm3 g−1) for N,P-HCNFs (Table S1). Therefore, the N,P–HCNF-8 sample with wide pore size distribution possesses a higher surface area and a larger pore volume as compared to other samples, making it highly appropriate for electrocatalysis [34,35].
[14–16]. From a fundamental point of view, the electronic structure of a carbon matrix is of signiﬁcant importance to electrocatalytic reactions, which are initiated with the chemisorption of reactant species, followed by electron and mass transfer. Recent works have demonstrated doping, edges, and defects could eﬀectively optimize charge/ spin distribution and thus modify the chemisorption of intermediates and subsequent electron transfer [17–19]. Speciﬁcally, doping with heteroatoms, especially nitrogen, provides graphitic carbon materials with a good bifunctional activity for ORR and OER due to the diﬀerence of electronegativity between carbon and heteroatoms [20,21]. Sharp edge sites have been conﬁrmed to possess an excellent electrocatalytic HER activity through tuning the electronic structure of the carbon skeleton . In addition, topological defects can give rise to the formation of the preferential conﬁgurations of dopants , thus providing multicomponent active sites, which maybe potentially promising for trifunctional metal-free electrocatalysts. More recently, the integration of multicomponent active sites has demonstrated to be an eﬀective strategy for enhancing trifunctional electrocatalytic activity . However, another crucial factor, mass transfer, is still lack of detailed engineering, which is associated with the speciﬁc surface area (SSA) and porosity of a carbon matrix and also has a great eﬀect on electrocatalytic activity [25,26]. Furthermore, it is still a formidable challenge to construct an appropriate model material with abundant active sites and an excellent mass transfer property and then to explore the eﬀect of mass transfer on electrocatalytic activity. Speciﬁcally, the one-dimensional (1D) carbon hollow nanoarchitecture provides an unique platform to improve the mass transfer property due to their distinct structural features, such as enlarged contact areas, eﬀective ion diﬀusion pathways, and continuous electron transport, and thus improving the accessibility of active sites [27,28]. Therefore, a hierarchical porous architecture, in particular with 1D electrolyte/reactant diﬀusion paths and abundant accessible active sites including doping, defects, and edges, is highly desirable for carbon nanomaterials to realize trifunctional electrocatalysis for ORR, OER, and HER. Herein, we report the ﬁrst example of N,P co-doped hollow carbon nanoﬁber (N,P–HCNF) membranes with a beneﬁcial hierarchical porous architecture and turbostratic structures, containing abundant catalytically active sites such as doping, defects, and edges. By the integration of plentiful accessible active sites and widely distributed pore sizes, the well-designed N,P–HCNF-8 (8 represents the concentration (% w/v) of polyacrylonitrile (PAN)) exhibited a remarkable trifunctional electrocatalytic activity and long-term stability in an alkaline electrolyte (i.e., 0.1 M KOH). It showed an ORR onset and half-wave potential (E1/2) of 0.93 and 0.82 V vs. reversible hydrogen electrode (RHE), a low overpotential of 320 mV vs. RHE at 10 mA cm−2 (Ej=10) for OER, and a comparable HER activity with metal-free catalysts. Impressively, the N,P–HCNF-8 was demonstrated to be one of the best ORR and OER metal-free bifunctional catalysts with a high overall oxygen electrode activity as characterized by a small ΔE (ΔE = Ej=10 - E1/2) of 0.73 V. More intriguingly, since the N,P-HCNFs have identical precursors, preparation conditions, similar structural features, and doping contents, it is reasonable to employ the materials as model catalysts to investigate individually the eﬀect of mass transfer on electrocatalytic performance. We demonstrate here that the mass transfer property of N,P-HCNFs makes a great diﬀerence to electrocatalytic activity, mainly through enlarged SSA to aﬀect intrinsic catalytic activity and the ionic resistance in pores. 2. Results and discussion Inspired by the motivation to prepare a carbon material with ample reactive interfaces and eﬃcient mass transfer pathways, N,P-HCNFs, endowed with multiple active sites and 1D electrolyte/reactant transport paths, are developed as delineated in Fig. 1. Brieﬂy, the pristine core-shell nanoﬁbers composed of a TPP/PVP (TPP = triphenylphosphine, PVP = polyvinylpyrrolidone) core and a DCDA/PAN 2
Nano Energy 64 (2019) 103879
Y. Gao, et al.
Fig. 1. Schematic representation of the fabrication procedures towards N,P-HCNFs.
Fig. 2. SEM images of (a) the pristine ﬁlm of N,P–HCNF-8 before pyrolysis, b) N,P–HCNF-8, and (c) fractured surfaces of N,P–HCNF-8. (d) TEM, (e) HRTEM, and (f) HAADF-STEM images of N,P–HCNF-8. (g–j) EDS mapping images of C, N, O, and P elements of (f).
Fig. 3. (a) Raman spectra, (b) XRD patterns, (c) N 1s XPS spectra, (d) N2 absorption–desorption isotherms, and (e) pore size distribution curves of N,P-HCNFs. 3
Nano Energy 64 (2019) 103879
Y. Gao, et al.
Fig. 4. (a) CVs of N,P–HCNF-8 in O2- and N2-saturated KOH solutions (0.1 M). (b) ORR LSV curves of N,P-HCNFs and Pt/C catalysts at a rotation rate of 1600 rpm in O2-saturated 0.1 M KOH solution with a scan rate of 10 mV s−1. (c) LSV curves for N,P–HCNF-8 at diﬀerent rotating speeds with a scan rate of 10 mV s−1. The inset in (c) shows K–L plots obtained from LSVs at diﬀerent potentials. (d) Tafel slopes derived from (b). (e) Chronoamperometric responses of N,P–HCNF-8 and Pt/C for ORR at 0.7 V versus RHE in 0.1 M KOH with a rotation speed of 900 rpm. (f) Chronoamperometric responses of N,P–HCNF-8 and Pt/C at 0.7 V versus RHE after the introduction of 20 ml of 3 M methanol into 100 ml of 0.1 M KOH solution.
Fig. 5. (a) LSV curves of diﬀerent catalysts for OER in 0.1 M KOH solution at 1600 rpm. (b) Corresponding Tafel plots of these catalysts. (c) LSV curves of various catalysts for both ORR and OER in 0.1 M KOH at 1600 rpm. (d) LSV curves of diﬀerent samples for HER in N2-saturated 0.1 M KOH solution at 1600 rpm. (e) Tafel slopes derived from (d). (f) The values of ΔE for N,P–HCNF-8, N,P–HCNF-8.5 and various catalysts reported previously.
The electrocatalytic performance of N,P-HCNFs was ﬁrst evaluated by cyclic voltammetry (CV) and linear sweep voltammetry (LSV) measurements on a rotating disk electrode (RDE) in O2- or N2-saturated 0.1 M KOH solution with a mass loading of ~0.25 mg cm−2. As shown
in Fig. 4a, a distinct oxygen reduction peak for ORR is observed for N,P–HCNF-8 in O2-saturated KOH solution, conversely no peak is detected in N2-saturated KOH solution. The oxygen reduction peak of N,P–HCNF-8 (0.78 V) is comparable to that of commercial 20 wt% Pt/C 4
Nano Energy 64 (2019) 103879
Y. Gao, et al.
parameters of recently reported bifunctional electrocatalysts. In addition, the HER catalytic performance of N,P-HCNFs was carried out in N2-saturated 0.1 M KOH solution as well. As shown in Fig. 5d, the N,P–HCNF-8 exhibits an onset overpotential of ~400 mV, which is much smaller than those of other reference samples. Furthermore, the potentials needed for the current density of 10 mA cm−2 are −0.55, −0.71, and −0.11 V for N,P–HCNF-8, N,P–HCNF-8.5, and Pt/ C, respectively. While the N,P–HCNF-8 presents a higher overpotential than Pt/C, its HER catalytic activity is still better than those of most of the reported carbon-based materials (Table S3). The corresponding Tafel plots in Fig. 5e demonstrate that the N,P–HCNF-8 has a lower Tafel slope (161 mV dec−1) than N,P–HCNF-7.5 (280 mV dec−1), N,P–HCNF-8.5 (287 mV dec−1), N,P–HCNF-9 (409 mV dec−1), and N,P–HCNF-9.5 (333 mV dec−1). Besides, the chronoamperometric response of N,P–HCNF-8 exhibits a cathodic current attenuation as small as 10%, indicating higher stability than Pt/C (Fig. S9b). Overall, beneﬁting from the hierarchical porous structure and numerous active sites, the N,P–HCNF-8 electrocatalyst shows a prominent trifunctional catalytic activity, in which doping, edges, and defects can regulate the electronic structure of the carbon matrix to facilitate trifunctional electrocatalysis. The unique trifunctional catalytic activity is closely related to the mass transfer behavior of the material as well. Considering ﬂexibility of coaxial electrospinning technology, the obtained N,P-HCNFs with different SSAs were selected to investigate individually the eﬀect of mass transfer on electrocatalytic activity. Particularly, the N,P-HCNFs share identical precursors, preparation conditions, similar structural features, and doping contents, as well as the analogous characteristic of pore size distribution, consequently serving as an ideal model system for the detailed study of the eﬀect of mass transfer on electrocatalytic properties. The LSV curves of N,P-HCNFs indicate the ORR performance gradually decreases with the reduction of SSA (Fig. 4b). To unravel the eﬀect of mass transfer through SSA, two critical factors are deeply deliberated. The ﬁrst is the impact of SSA on intrinsic catalytic activity, which is estimated by the kinetic current density (jk) in the mixed kinetic-diﬀusion controlled region. Since the electrochemical doublelayer capacitance (Cdl) is proportional to the electrochemically active surface area (ECSA) , the larger the SSA of N,P-HCNFs is, the more exposed active sites are (Fig. 6a and Fig. S10). Remarkably, along with the raising of Cdl, jk dramatically increases, indicating enhanced intrinsic catalytic activity. After the normalization of jk by the SSA and ECSA, the activity diﬀerence among N,P-HCNFs curtails distinctly, for example, the diﬀerence of jk between N,P–HCNF-8 and N,P–HCNF-9.5 decreases from 15.6 to 9.1 and 4.6 times, respectively (Fig. S11). These results reﬂect that the increased SSA of N,P-HCNFs contributes at least partially to the enhancement of ORR performance , and the ECSA is also an eﬀective metric to embody the electrocatalytic activity of metalfree electrocatalysts. Another key factor aﬀected by SSA is the ionic resistance in pores. The Nyquist plots of N,P-HCNFs (Fig. 6b) exhibit an approximate 45° slope in the middle frequency region and quasi-
(0.79 V) and more positive than those of other reference samples (Fig. S7). The LSV curves in Fig. 4b reveal that the N,P–HCNF-8 exhibits a positive onset potential of 0.93 V, a half-wave potential (E1/2) of 0.82 V, and a limiting current density of 5.1 mA cm−2, all of which are comparable to those of Pt/C. Interestingly, the N,P–HCNF-8 holds the best ORR performance among all the afore-mentioned metal-free catalysts in terms of onset potential, half-wave potential, and limiting current density, indicating the importance of the tailored hierarchical pore structure for ORR. The electron transfer number per O2 (n) for N,P–HCNF-8 was calculated from the LSV curves (Fig. 4c) according to the Koutechy-Levich (K-L) equation. The K-L plots in the inset of Fig. 4c show a linear relationship between j−1 and ω−1/2, suggesting ﬁrstorder reaction kinetics with regard to the concentration of dissolved oxygen . The n of N,P–HCNF-8 was determined to be 3.89 ± 0.04, which is comparable to that of Pt/C (3.90 ± 0.05, Fig. S8c), conﬁrming a four-electron pathway for ORR. Besides, the excellent activity of N,P–HCNF-8 is also conﬁrmed by its lower Tafel slope (47 mV dec−1) than Pt/C (70 mV dec−1) and other reference samples (Fig. 4d). In addition, the operational stability of the N,P–HCNF-8 and Pt/C catalysts was investigated at 0.7 V for 20 h in O2-saturated 0.1 M KOH solution with a rotation rate of 900 rpm (Fig. 4e). After the tests, about 91% of the initial current density was retained for the N,P–HCNF-8 catalyst, while more than 17% decay in ORR activity occurred for the Pt/C catalyst. Moreover, the N,P–HCNF-8 catalyst also demonstrated excellent tolerance to methanol crossover (Fig. 4f). To further exploit the underlying application of N,P-HCNFs as a bifunctional catalyst for both ORR and OER, LSV curves were implemented to characterize the OER catalytic activities of N,P-HCNFs. As shown in Fig. 5a, the N,P–HCNF-8 exhibits a smaller onset potential and a higher current density than N,P–HCNF-7.5, N,P–HCNF-8.5, N,P–HCNF-9, and N,P–HCNF-9.5. Even compared to the state-of-the-art IrO2 catalyst, the N,P–HCNF-8 also displays a higher catalytic activity across a wide potential range. The N,P–HCNF-8 shows a potential of 1.55 V at the current density of 10 mA cm−2 (Ej=10), which is lower than that of IrO2 (1.60 V). Meanwhile, Tafel plots given in Fig. 5b reveal that the N,P–HCNF-8 displays a smaller Tafel slope of 248 mV dec−1 than N,P–HCNF-7.5 (326 mV dec−1), N,P–HCNF-8.5 (311 mV dec−1), N,P–HCNF-9 (607 mV dec−1), and N,P–HCNF-9.5 (420 mV dec−1), which denotes the more favorable kinetics of N,P–HCNF-8 towards OER than others. Furthermore, the N,P–HCNF-8 shows better stability than IrO2 under continuous operation for 20,000 s (Fig. S9a). Fig. 5c clearly presents the overall oxygen electrocatalytic activity of N,P–HCNF-8 as a bifunctional catalyst, which can be evaluated by the potential deviation (ΔE) between E1/2 for ORR and Ej=10 for OER. The N,P–HCNF-8 electrocatalyst exhibits the smallest value of ΔE (0.73 V), displaying one of the best bifunctional activities among carbon-based metal-free materials (e.g., N-GRW, ΔE = 0.82 V; SHG, ΔE = 0.77 V; NGM, ΔE = 0.90 V, Fig. 5f) [37–39], and even transcending some transition-metals (e.g., [email protected]
–C, ΔE = 0.88 V; Co/NC, ΔE = 0.86 V; NGM-Co, ΔE = 0.95 V) [40–42]. Table S2 shows a further comparison of the main performance
Fig. 6. (a) BET speciﬁc surface areas, Cdl and jk of N,P-HCNFs. (b) Comparison of Nyquist plots obtained from potentiostatic EIS at 0.76 V. The open symbols and solid lines represent the experimental and simulation results, respectively. (c) The ionic resistances (Rion) of various materials. 5
Nano Energy 64 (2019) 103879
Y. Gao, et al.
vertical lines at lower frequency. The transmission line model (TLM) for porous electrodes [45,46] (Fig. S12) was utilized for analyzing the mass transfer features of N,P-HCNFs. The projection of the 45° slope to the real axis, deﬁned as Rion/3, signals the ionic resistance (Rion) for the electrolyte-ﬁlled pores inside the porous catalyst layers, which is an eﬀective parameter to describe the capability of mass transfer during the electrocatalytic process [45–47]. As shown in Fig. 6c, the N,P–HCNF-8 shows a lower Rion of 12.8 Ω than other N,P-HCNFs and even than the commercial Pt/C catalyst with a Rion of 15.2 Ω (Table S4). The schematic calculation of Rion/3 for each catalyst in Fig. 6c is shown in Fig. S13. These studies manifest that the mass transfer process can be expedited by engineering the hierarchical porous structure with a larger SSA. In this regard, much more accessible ECSA arises to increase intrinsic catalytic activity, and a lower ionic resistance occurs to improve ion transport kinetics.
This work was supported by the Ministry of Science and Technology of China [grant number 2012CB933403]; the National Natural Science Foundation of China [grant number 51425302, 51302045]; the Beijing Municipal Science and Technology Commission [grant number Z121100006812003]; Talents Introduction Plan of Hebei Agricultural University [grant number YJ201819]; and the Chinese Academy of Sciences. Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.nanoen.2019.103879. References
3. Conclusions    
In summary, N,P co-doped hollow carbon nanoﬁbers with abundant active sites and hierarchical porous structures were facilely prepared via coaxial electrospinning technology, followed by polymerization and pyrolysis. Beneﬁtting from the architecture features including plentiful active sites, continuous conducting pathways, and benign mass transfer channels, the N,P–HCNF-8 exhibited an excellent trifunctional electrocatalytic activity for ORR, OER, and HER, which is superior to the reference catalysts with lower SSAs as well as numerous other carbonbased metal-free materials. Impressively, the N,P–HCNF-8 with a low potential deviation (ΔE) of 0.73 V between E1/2 for ORR and Ej=10 for OER was demonstrated to be one of the best metal-free bifunctional electrocatalysts in oxygen electrocatalysis. Furthermore, the mass transfer property of N,P-HCNFs was demonstrated to make a great diﬀerence to electrocatalytic activity, mainly through enlarged SSA to aﬀect intrinsic catalytic activity and the ionic resistance in pores. The present work may shed light on the design, fabrication, and tailoring of highly active metal-free electrocatalysts with rich active sites and tuned pore structures for electrocatalysis and other applications.
          
4. Experimental section
   
4.1. Preparation of the pristine ﬁlms of N,P-HCNFs A coaxial electro-spinneret was used in this study. The equipment for the electrospinning process is presented in Fig. 1. The core solution consisted of 0.65 g of polyvinylpyrrolidone (PVP, Mw = 1,300,000, Alfa), 1.0 g of triphenylphosphine (TPP, 99%, Alfa), and 5 mL of N,Ndimethylformamide (DMF, AR, Beijing Chemical Works). The shell solution was prepared by dissolving 0.6 g of polyacrylonitrile (PAN, Mw = 150,000, Sigma) and 0.5 g of dicyandiamide (DCDA, 99%, Aladdin) in certain amounts of DMF to reach diﬀerent concentrations of PAN. For the preparation of N,P–HCNF-8, 7.5 mL of DMF was used to obtain a 8% w/v PAN solution. For other concentrations of PAN, the corresponding products were noted as N,P–HCNF-x (x = 7.5, 8.5, 9 or 9.5). A ﬂat Al foil covered with non-dust cloth was used as a collector and put about 15 cm away from the nozzle tip. A voltage of 16 kV was applied to the solution to start the spinning process with a high voltage source (SL50P60, Spellman High Voltage Electronics Corporation).
           
4.2. Preparation of N,P-HCNFs
The pristine ﬁlms of N,P-HCNFs were placed in a ceramic boat, heated to 550 °C (2 °C min−1), and maintained for 4 h in a tube furnace. The temperature in the furnace was further raised to 1000 °C at a ramp rate of 5 °C min−1 and kept for 1 h under Ar ﬂow. After that, the furnace was cooled down to room temperature naturally. The details of sample characterizations and electrochemical measurements are provided in the Supporting Information.
    
M.K. Debe, Nature 486 (2012) 43–51. Y. Li, H. Dai, Chem. Soc. Rev. 43 (2014) 5257–5275. Y. Xu, M. Kraft, R. Xu, Chem. Soc. Rev. 45 (2016) 3039–3052. H. Wang, H.W. Lee, Y. Deng, Z. Lu, P.C. Hsu, Y. Liu, D. Lin, Y. Cui, Nat. Commun. 6 (2015) 7261. J. Zhang, Z. Zhao, Z. Xia, L. Dai, Nat. Nanotechnol. 10 (2015) 444–452. S. Zhao, Y. Wang, J. Dong, C.-T. He, H. Yin, P. An, K. Zhao, X. Zhang, C. Gao, L. Zhang, J. Lv, J. Wang, J. Zhang, A.M. Khattak, N.A. Khan, Z. Wei, J. Zhang, S. Liu, H. Zhao, Z. Tang, Nat. Energy 1 (2016) 16184. S. Zhao, H. Yin, L. Du, L. He, K. Zhao, L. Chang, G. Yin, H. Zhao, S. Liu, Z. Tang, ACS Nano 8 (2014) 12660–12668. Y. Liang, Y. Li, H. Wang, J. Zhou, J. Wang, T. Regier, H. Dai, Nat. Mater. 10 (2011) 780–786. S.C. Huang, Y.Y. Meng, S.M. He, A. Goswami, Q.L. Wu, J.H. Li, S.F. Tong, T. Asefa, M.M. Wu, Adv. Funct. Mater. 27 (2017) 1606585. X. Jia, Y. Zhao, G. Chen, L. Shang, R. Shi, X. Kang, G.I.N. Waterhouse, L. Wu, C. Tung, T. Zhang, Adv. Energy Mater. 6 (2016) 1502585. H. Fan, H. Yu, Y. Zhang, Y. Zheng, Y. Luo, Z. Dai, B. Li, Y. Zong, Q. Yan, Angew. Chem. Int. Ed. 56 (2017) 12566–12570. Q. Liu, J. Tian, W. Cui, P. Jiang, N. Cheng, A.M. Asiri, X. Sun, Angew. Chem. Int. Ed. 53 (2014) 6710–6714. K. Gong, F. Du, Z. Xia, M. Durstock, L. Dai, Science 323 (2009) 760–764. L. Dai, Y. Xue, L. Qu, H.J. Choi, J.B. Baek, Chem. Rev. 115 (2015) 4823–4892. S. Zhao, D.W. Wang, R. Amal, L. Dai, Adv. Mater. 31 (2019) 1801526. S. Zhao, X. Lu, L. Wang, J. Gale, R. Amal, Adv. Mater. 31 (2019) 1805367. I.Y. Jeon, S. Zhang, L. Zhang, H.J. Choi, J.M. Seo, Z. Xia, L. Dai, J.B. Baek, Adv. Mater. 25 (2013) 6138–6145. Z. Zhao, M. Li, L. Zhang, L. Dai, Z. Xia, Adv. Mater. 27 (2015) 6834–6840. C. Tang, Q. Zhang, Adv. Mater. 29 (2017) 1604103. Y. Zheng, Y. Jiao, M. Jaroniec, Y. Jin, S.Z. Qiao, Small 8 (2012) 3550–3566. G. Wu, A. Santandreu, W. Kellogg, S. Gupta, O. Ogoke, H.G. Zhang, H.L. Wang, L.M. Dai, Nano Energy 29 (2016) 83–110. H. Wang, X.B. Li, L. Gao, H.L. Wu, J. Yang, L. Cai, T.B. Ma, C.H. Tung, L.Z. Wu, G. Yu, Angew. Chem. Int. Ed. 57 (2018) 192–197. Y. Ito, Y. Shen, D. Hojo, Y. Itagaki, T. Fujita, L. Chen, T. Aida, Z. Tang, T. Adschiri, M. Chen, Adv. Mater. 28 (2016) 10644–10651. Z.Y. Lu, J. Wang, S.F. Huang, Y.L. Hou, Y.G. Li, Y.P. Zhao, S.C. Mu, J.J. Zhang, Y.F. Zhao, Nano Energy 42 (2017) 334–340. H.W. Liang, X. Zhuang, S. Bruller, X. Feng, K. Mullen, Nat. Commun. 5 (2014) 4973. M. Wu, K. Wang, M. Yi, Y. Tong, Y. Wang, S. Song, ACS Catal. 7 (2017) 6082–6088. Q. Wei, F. Xiong, S. Tan, L. Huang, E.H. Lan, B. Dunn, L. Mai, Adv. Mater. 29 (2017) 1602300. L.F. Chen, Y. Lu, L. Yu, X.W. Lou, Energy Environ. Sci. 10 (2017) 1777–1783. W. Zhang, Z.Y. Wu, H.L. Jiang, S.H. Yu, J. Am. Chem. Soc. 136 (2014) 14385–14388. Y. Jia, L. Zhang, A. Du, G. Gao, J. Chen, X. Yan, C.L. Brown, X. Yao, Adv. Mater. 28 (2016) 9532–9538. H.T. Chung, D.A. Cullen, D. Higgins, B.T. Sneed, E.F. Holby, K.L. More, P. Zelenay, Science 357 (2017) 479–484. L. Hao, J. Ning, B. Luo, B. Wang, Y. Zhang, Z. Tang, J. Yang, A. Thomas, L. Zhi, J. Am. Chem. Soc. 137 (2015) 219–225. J. Zhang, L. Qu, G. Shi, J. Liu, J. Chen, L. Dai, Angew. Chem. Int. Ed. 55 (2016) 2230–2234. C.G. Hu, X.Y. Chen, Q.B. Dai, M. Wang, L.T. Qu, L.M. Dai, Nano Energy 41 (2017) 367–376. Y. Huang, Y. Wang, C. Tang, J. Wang, Q. Zhang, Y. Wang, J. Zhang, Adv. Mater. 31 (2019) 1803800. B.Y. Xia, Y. Yan, N. Li, H.B. Wu, X.W. Lou, X. Wang, Nat. Energy 1 (2016) 15006. H.B. Yang, J. Miao, S.F. Hung, J. Chen, H.B. Tao, X. Wang, L. Zhang, R. Chen, J. Gao, H.M. Chen, L. Dai, B. Liu, Sci. Adv. 2 (2016) e1501122. C. Hu, L. Dai, Adv. Mater. 29 (2017) 1604942. C. Tang, H.F. Wang, X. Chen, B.Q. Li, T.Z. Hou, B. Zhang, Q. Zhang, M.M. Titirici, F. Wei, Adv. Mater. 28 (2016) 6845–6851.
Nano Energy 64 (2019) 103879
Y. Gao, et al.
 N. Ogihara, Y. Itou, T. Sasaki, Y. Takeuchi, J. Phys. Chem. C 119 (2015) 4612–4619.  N. Ogihara, S. Kawauchi, C. Okuda, Y. Itou, Y. Takeuchi, Y. Ukyo, J. Electrochem. Soc. 159 (2012) A1034–A1039.  H. Sun, L. Mei, J. Liang, Z. Zhao, C. Lee, H. Fei, M. Ding, J. Lau, M. Li, C. Wang, X. Xu, G. Hao, B. Papandrea, I. Shakir, B. Dunn, Y. Huang, X. Duan, Science 356 (2017) 599–604.
 J. Wang, H.H. Wu, D.F. Gao, S. Miao, G.X. Wang, X.H. Bao, Nano Energy 13 (2015) 387–396.  A. Aijaz, J. Masa, C. Rosler, W. Xia, P. Weide, A.J. Botz, R.A. Fischer, W. Schuhmann, M. Muhler, Angew. Chem. Int. Ed. 55 (2016) 4087–4091.  C. Tang, B. Wang, H.F. Wang, Q. Zhang, Adv. Mater. 29 (2017) 1703185.  K. Qu, Y. Zheng, S. Dai, S.Z. Qiao, Nano Energy 19 (2016) 373–381.  J. Snyder, T. Fujita, M.W. Chen, J. Erlebacher, Nat. Mater. 9 (2010) 904–907.