Controllable Anchoring of Graphitic Carbon Nitride on MnO2 Nanoarchitectures for Oxygen Evolution Electrocatalysis

The design and fabrication of eco-friendly and cost-effective (photo)electrocatalysts for the oxygen evolution reaction (OER) is a key research goal for a proper management of water splitting to address the global energy crisis. In this work, we focus on the preparation of supported MnO2/graphitic carbon nitride (g-CN) OER (photo)electrocatalysts by means of a novel preparation strategy. The proposed route consists of the plasma enhanced-chemical vapor deposition (PE-CVD) of MnO2 nanoarchitectures on porous Ni scaffolds, the anchoring of controllable g-CN amounts by an amenable electrophoretic deposition (EPD) process, and the ultimate thermal treatment in air. The inherent method versatility and flexibility afforded defective MnO2/g-CN nanoarchitectures, featuring a g-CN content and nano-organization tunable as a function of EPD duration and the used carbon nitride precursor. Such a modulation had a direct influence on OER functional performances, which, for the best composite system, corresponded to an overpotential of 430 mV at 10 mA/cm2, a Tafel slope of ≈70 mV/dec, and a turnover frequency of 6.52 × 10–3 s–1, accompanied by a very good time stability. The present outcomes, comparing favorably with previous results on analogous systems, were rationalized on the basis of the formation of type-II MnO2/g-CN heterojunctions, and yield valuable insights into this class of green (photo)electrocatalysts for end uses in solar-to-fuel conversion and water treatment.


■ INTRODUCTION
−13 Since large-scale applications of the actual state-of-the-art OER catalysts based on RuO 2 and IrO 2 14−16 are prevented by various issues, encompassing their high cost, insufficient longterm stability, and scarcity, 4,8,10,11 numerous efforts are being dedicated to alternative eco-friendly electrocatalysts, highly required in the framework of improved sustainability. 9,10,12,13,17−39 Among the various routes adopted to improve g-CN performances, [19][20][21][22]27,29,37 the controlled construction of heterojunctions between g-CN and a suitable semiconducting partner can yield a more efficient harvesting of solar light and an improved separation of photoproduced electrons and holes.3 So far, MnO 2 + g-CN composites have been prepared by liquid-phase routes for applications as supercapacitors 33,34,44 and (photo)catalysts for CO 2 reduction, 36,37 toxic product degradation, 18,28,30,31,43 H 2 generation, 3,5,6,26,40 and oxidation reactions of industrial interest.39,45 In a different way, the available studies on MnO 2 /g-CN OER electrocatalysts 8,23,42 (or MnO 2 composites with C tubes/dots/fibers 1, 4,9,11,13,17 ) are more scarce.Furthermore, most of the available studies have f o c u s e d o n p o w d e r e d s y s t e m s , e i t h e r a s such 1, 3,5,18,27,30,32,37,40,41,45 or mixed with binders/additives to yield pastes immobilized on supports, [8][9][10]33,42,44 whereas a primary requirement for real-world applications is the direct growth of active materials onto suitable substrates.15,24,25,46,47 The OER activity of the target systems is directly affected by their composition, structure, morphology, and defect content, 4,5,[11][12][13]16,41,45 as well as by the used support backbone.8,46 In particular, the growth of nanosystems on porous and electrically conductive scaffolds, such as Ni foams (NFs), 16,33,34,44 can favorably yield 3D nanoarchitectures featuring an enriched content of reaction sites, a large active area, and plentiful pathways for ion/charge carrier diffusion.46− 48 As a matter of fact, further research progress in this direction is highly demanded for the development of green (photo)electrocatalysts endowed with attractive service life and performances.
Herein, we report on the fabrication of supported MnO 2 + g-CN nanocomposites by an original preparation strategy, which, to our knowledge, has never been proposed to date for the preparation of such systems.After the growth of MnO 2 nanoarchitectures on Ni foam substrates by means of plasma enhanced-chemical vapor deposition (PE-CVD), the anchoring of graphitic carbon nitride is carried out by electrophoretic deposition (EPD), as sketched in Figure 1.Both PE-CVD and EPD techniques are highly versatile for the production of functional nanomaterials with tailored features.Specifically, PE-CVD benefits from the peculiar characteristics of cold plasmas, enabling MnO 2 growth under conditions far from thermodynamic equilibrium and yielding its efficient dispersion into the used porous substrates. 47,49Furthermore, the continuous plasma bombardment of the growing material yields an efficient MnO 2 /Ni foam interfacial contact and tailored surface reactivity. 47,50The latter feature, along with the high surface-to-volume ratio of the obtained MnO 2 deposits, is a favorable issue for the subsequent g-CN anchoring, taking advantage of the EPD flexibility in the processing of exfoliated micro/nanosheets. 25Thanks to these particular features, the present approach leads to a high density of MnO 2 /g-CN heterojunctions, favorably influencing electrochemical performances. 47,49In this regard, EPD was carried out using two diverse kinds of g-CN with variable defect content and different active areas, 2,7,19,25 to elucidate the interrelations between the characteristics of carbon nitride and the electrocatalytic activity of the resulting MnO 2 + g-CN composites.
The prepared systems contain tetragonal β-MnO 2 , the equilibrium polymorph of manganese(IV) oxide, 46 different from the majority of studies on OER that have been performed on the α phase 8,9,11,26,42 or amorphous MnO 2 . 23The developed nanoarchitectures feature a close MnO 2 /g-CN contact, an open area morphology, and a tunable defect content, resulting in an attractive OER activity.Specifically, the best system yields an overpotential (η) of 430 mV vs the reversible hydrogen electrode (RHE) at a current density of 10 mA/cm 2 and a Tafel slope of ≈70 mV/dec, which, to our knowledge, are among the best values reported to date for MnO 2 + g-CN OER catalysts in alkaline solutions.The attractive modular characteristics of the present green materials open the door to their practical use under real-world conditions.
■ EXPERIMENTAL SECTION Material Synthesis.MnO 2 nanoarchitectures were prepared by PE-CVD from electronic-grade Ar/O 2 plasmas using a custom-built two-electrode apparatus equipped with a 13.56 MHz radio frequency (RF) generator.In each deposition, precleaned 46 Ni foam substrates (lateral size = 1.2 cm × 2.0 cm) were fixed on the grounded electrode.Based on previously reported results, 46 growth processes were carried out using the following experimental settings: RF-power = 20 W; deposition temperature = 300 °C; total Ar flow rate = 65 standard cubic centimeters per minute (sccm); total O 2 flow rate = 5 sccm; duration = 3 h; total pressure = 1.0 mbar.
Anchoring of graphitic carbon nitride on MnO 2 was performed by EPD from two kinds of carbon nitride powders, i.e., g-CN(M) and g- CN(CM) (where M = melamine and CM = melamine + cyanuric acid, indicating the used precursors; see also the Supporting Information, § S-1.1 and Figures S1 and S2).Depositions were performed from freshly prepared suspensions of g-CN and I 2 in acetone, 51 using carbon paper and Ni foam-supported MnO 2 as anode and cathode, respectively, (see also Supporting Information, § S-1.1).Each composite specimen was finally subjected to annealing at 400 °C for 1 h in air (heating rate = 20 °C/min).The content of g-CN in the final composite systems was modulated as a function of EPD duration [from 20 to 600 s for the systems derived from g-CN(CM)].For comparison purposes, g-CN(M)-containing composites were prepared by using an optimized deposition time of 180 s.In the following, the target samples are denoted as MnO 2 −X−Y, where X = M or CM and Y = deposition time (s).Bare manganese dioxide (MnO 2 ) or g-CN specimens (the latter denoted as X−Y) were also fabricated and characterized.
Chemicophysical Investigation.A Zeiss SUPRA 40 VP instrument equipped with an INCA x-act PentaFET Precision spectrometer was used for field emission-scanning electron microscopy (FE-SEM) and energy-dispersive X-ray spectroscopy (EDXS) analyses, at primary beam acceleration voltages between 10 and 20 kV.The average aggregate dimensions were evaluated using the ImageJ software through a statistical image analysis. 52Transmission electron microscopy (TEM), selected area electron diffraction (SAED), high angle annular dark field-scanning TEM (HAADF-STEM) analyses, and STEM-EDXS mapping were carried out using an aberration double-corrected cold FEG JEM ARM200F microscope operated at 200 kV and equipped with a CENTURIO large angle EDX detector, ORIUS Gatan camera, and Quantum GIF.X-ray photoelectron spectroscopy (XPS) analyses were performed using a ThermoFisher ESCALAB 250XI+ apparatus, with a monochromatized Al Kα X-ray source (hν = 1486.6eV).Binding energy (BE) values were corrected for charging by assigning a BE of 284.8 eV to the adventitious C 1s signal (component C 0 in Figure 6a,b below).Atomic percentages (at.%) were computed by peak area integration using ThermoFisher sensitivity factors.Peak fitting was carried out by XPSPEAK software using Gaussian−Lorentzian sum functions.Photoluminescence (PL) spectra for Ni-foam-supported materials were collected in the 300−700 nm wavelength range using an FLS 1000 fluorimeter (Edinburgh Instruments).The following settings were used: excitation wavelength/bandwidth = 280/13 nm; emission bandwidth = 7 nm.
Electrochemical Tests.OER performances were tested in freshly prepared 0.1 mol/L KOH (pH = 12.9) solutions with an electrochemical workstation (Autolab PGSTAT204 potentiostat/ galvanostat) equipped with a three-electrode setup.A Pt coil, Hg/ HgO (MMO), and NF-supported materials were used as counter, reference, and working electrodes, respectively.To prevent wetting of the electrical contact by the electrolyte due to capillary phenomena, a nickel-coated copper wire was carefully welded on the edge of the Ni foam, and the substrate/wire contact was subsequently coated with epoxy resin.
Linear sweep voltammetry (LSV) traces were recorded at a fixed scan rate of 5 mV/s.The onset potential was calculated as the one necessary to reach a current density of 0.02 mA/cm 2 . 53Tafel slopes were determined by analyzing the plots of potential vs log(current density). 12,24,46Applied bias photon-to-current efficiency (ABPE) curves were obtained by the following equation 23,24,31 where j (mA/cm 2 ) and P are the photocurrent density at the potential E RHE and the incident light power density (100 mW/cm 2 ), respectively.Additional details and data on material synthesis and characterization are reported in the Supporting Information.
■ RESULTS AND DISCUSSION Material Characterization.Functionalization of PE-CVD MnO 2 deposits was performed via EPD of carbon nitride for different durations, yielding homogeneous samples (Figure 1, left bottom panel).In this study, efforts were dedicated to investigating the influence of EPD duration on the chemicophysical and functional characteristics of the obtained composite systems.As observed in Figure 1 (center bottom panel), varying solely the EPD duration enabled accurate control of the mass of the anchored g-CN even for a very short deposition time, an important issue in modulating material features depending on this parameter.
The system morphology was analyzed by FE-SEM.As can be observed in Figure S3, bare MnO 2 was dominated by the occurrence of hierarchical 3D f lower-like architectures resulting from the assembly of 2D flake-like structures.The obtained deposits, evenly covering the underlying Ni foam supports, were characterized by an open morphology and a high active area, allowing enhanced contact with the reaction medium, a key feature for OER applications. 8,11,46,48After EPD for moderate times, the oxide matrix showed no remarkable modifications (Figure 2), evidencing that the proposed g-CN anchoring procedure enabled avoiding alterations of the original MnO 2 morphology.Nonetheless, g-CN nano-organization was appreciably influenced by the EPD duration and the nature of carbon nitride precursor powders.In the case of specimen MnO 2 −M-180, g-CN(M) was characterized by aggregates featuring an irregular shape, with dimensions comprised between 0.5 and 10 μm.In a different way, g-CN(CM)-derived specimens presented spherical-like carbon nitride particles with a narrower size distribution (1−3 μm), resulting from the assembly of exfoliated sheets.As shown by an inspection of Figure 2, a higher EPD process duration resulted not only in an increased amount (in agreement with the mass increase observed in Figure 1, center bottom panel) but also in a different spatial distribution of g-CN(CM).Upon going from MnO 2 −CM-20 to MnO 2 −CM-45, more effective coverage of MnO 2 by g-CN aggregates was detected, and g-CN particles were still well dispersed onto the underlying MnO 2 .These morphological features evidenced the occurrence of a close contact between MnO 2 and g-CN(CM), exerting in turn a favorable influence on the ultimate OER activity of composite systems.Nevertheless, a further increase in EPD duration (specimen MnO 2 −CM-180) resulted in an irregular spatial distribution of g-CN, whose globular aggregates completely covered the underlying MnO 2 in some sample regions (see also the EDXS results presented in Figure S4).This phenomenon was even more evident in the sample MnO 2 −CM-600, for which the functionalization with carbon nitride yielded even a partial occlusion of Ni foam pores.These characteristics detrimentally affected the resulting electrochemical activity, as explained below.
Preliminary Raman spectroscopy and X-ray diffraction (XRD) analyses (see Figures S5 and S6 and related observations) evidenced the obtainment of composite systems, in which phase-pure β-MnO 2 and g-CN maintained their chemical identity.To attain a deeper insight into the system nano-organization, with particular regard to carbon nitride spatial distribution and to MnO 2 /g-CN interfaces, an advanced TEM and EDXS investigation was performed on selected samples.Figure 3 25 A thorough analysis revealed that g-CN presented different 3D porous structures: spherical mesoporous, donut-type, and spherical doublelayered (see HAADF-STEM images and STEM-EDXS maps in Figure 4).Regardless of their morphology, all particles were formed by assembled g-CN flakes.The latter were evenly distributed within mesoporous spherical particles, whereas donut-type systems presented a porous shell structure surrounding a hollow internal cavity.Conversely, spherical double-layered systems consisted of a g-CN core region comprised between two denser shells with a thickness of (400  S7) confirmed the 3D morphology of the observed g-CN aggregates.−56 As can be observed in Figure 4d, g-CN particles are well connected to flake-like MnO 2 , yielding an intermixed interfacial region that can promote charge transport processes.
Figure 5 displays the TEM characterization results of sample MnO 2 −M-180.As can be observed, after a critical thickness of 600 nm, the MnO 2 deposit starts switching from a dendritetype to a flake-type growth.According to the collected SAED patterns and HRTEM images, both dendrite-and flake-like MnO 2 particles possess a tetragonal structure.Figure 5c,d shows evidence that dendritic grains present a near-epitaxial relationship with the Ni foam substrate.In a different way, no epitaxial relationship between dendritic and flake-type MnO 2 particles, that showed a random orientation, was detected (see Figure 5e).
Material chemical composition was analyzed by XPS on representative samples.For bare manganese dioxide and composite materials, the O/Mn atomic percentage (at.%)ratio values were very close to 2.0, as expected for MnO 2 .Conversely, N/C atomic percentage ratios (1.0−1.1) were lower than the stoichiometric value for g-C 3 N 4 (1.3), suggesting that the target systems were nitrogen-deficient, a feature favorably impacting their photoactivity. 38C 1s signals resulted from the contribution of three different bands (Table S1 and Figure 6a,b; see Figure 6c for the detailed attribution): C 0 , due to adventitious carbon contamination, at 284.8 eV; 32,45 C 1 , related to C bonded to amino groups (C-NH x ; x = 1, 2) on g-CN heptazine ring edges; 38,59,60 C 2 , attributed to carbon in N−C = N moieties of g-CN aromatic rings. 3,7,34,38,40For composite systems (see also Table S1), the C 1 component could be also related to C−O bonds between g-CN and MnO 2 , 1, 3,31,32,40,43 in line with O 1s peak analysis results (see Figure S9 and Table S5).This C−O bonding suggests the formation of a solid MnO 2 /g-CN linkage, 43 a favorable feature promoting electron−hole separation. 37Upon passing from bare g-CN specimens to the corresponding composites, C 1 and C 2 components underwent a BE increase 6,32,44 (see Table S1), higher upon going from CM-45 vs MnO 2 −CM-45 (0.4 eV) in comparison to the one occurring in MnO 2 −M-180 with respect to M-180 (0.2 eV).These results suggested the occurrence of a g-CN → MnO 2 electron flow, 18 consistent with the formation of g-CN/MnO 2 heterojunctions.N 1s photopeaks were fitted with four components (Figure 7a,b): N 1 , the main one, attributed to C = N−C bicoordinated nitrogen centers; 8,37,61 N 2 , due to tertiary N−(C) 3 atoms in g-CN; 7,35,38,59 N 3 , assigned to uncondensed NH x moieties; 19,32,35,59 N 4 , related to π−π*-electron excitations. 7,30,34,45he contribution of component N 3 to the overall N 1s signal underwent a nearly 2-fold increase upon going from specimen M-180 to CM-45 and from sample MnO 2 −M-180 to MnO 2 − CM-45 (Table S2 and Figure 7c), highlighting that CMderived systems were characterized by a higher content of terminal -NH x groups, i.e., by a lower condensation degree.In line with the C 1s case, 18,23 even the BEs of N 1s components underwent an increase in going from pure nitrides to composite samples (see Table S2), supporting the presence of the aforementioned MnO 2 /g-CN electronic interplay.
Manganese photoelectron peaks are displayed in Figure 8.For bare MnO 2 , the obtainment of manganese(IV) oxide was confirmed by Mn 2p signal shape and position [Table S3; BE(Mn 2p 3/2 ) = 642.6 eV; 15,30,44,62 spin−orbit splitting (SOS) = 11.6 eV 1, 41,43,44 ], as well as by Mn 3s multiplet splitting separation 45,62 (Table S4).For composite materials, Mn 2p and Mn 3s energy positions underwent a red shift in comparison to MnO 2 (−0.2 eV for MnO 2 −M-180 and −0.4 eV for MnO 2 −CM-45; Tables S3 and S4).This phenomenon, in accordance with previously reported data for MnO 2 /g-CN materials, 6,18,23,43,44 confirmed that g-CN and MnO 2 were directly coupled in the fabricated composite systems, 34 and that a built-in electric field was formed at the g-CN/MnO 2 interface. 41The occurrence of the latter promoted a g-CN → MnO 2 electron transfer, a phenomenon which, on the basis of the measured BE shifts, turned out to be more marked for MnO 2 −CM-45 than for MnO 2 −M-180.The g-CN/MnO 2 interaction was further proven by photoluminescence measurements on bare MnO 2 , CM-45, and MnO 2 −CM-45.The recorded spectra (Figure S10) revealed that MnO 2 did not yield any appreciable emission.In the case of g-CN-containing samples, a better charge separation, corresponding to luminescence quenching, was observed for specimen MnO 2 − CM-45 in comparison to CM-45.This phenomenon is accompanied by a red shift of the emission maximum (≈5 nm for MnO 2 −CM-45 vs CM-45), indicating a favorable electronic interaction between MnO 2 and g-CN, suppressing the recombination of photogenerated charge carriers. 63Basing also on FE-SEM and TEM outcomes, this result was related to the exfoliated morphology of the g-CN(CM) complex, enabling shortening of the charge carrier diffusion length and promoting g-CN → MnO 2 electron transfer, 29 in line with XPS results (see above).
Electrochemical Characterization.The catalytic activity of the obtained systems (Figure 9a) was investigated in alkaline media.The LSV curves recorded under irradiation for singlephase and composite materials (Figure 9b) evidenced a net current density increase at potentials >1.50 V vs RHE.The appreciable dark currents at high bias values (Figure S11) indicate that the target materials are reasonable OER catalysts. 14As a general trend, LSV data under illumination indicated that the current densities yielded by composite systems were higher than the ones measured not only for the Ni foam substrate, but also for bare g-CN and MnO 2 samples, provided that an optimal carbon nitride amount was anchored on the underlying manganese oxide (compare Figures 9b and  S12; see also onset potential and Tafel slope values (Figure 9c,d).The extrapolated onset potentials (Figure 9c) were directly influenced by both the used g-CN precursor (as shown by the values for MnO 2 −M-180 and MnO 2 −CM-180, the latter yielding a higher activity) and the EPD duration (as shown by the data for the g-CN(CM)-containing composites).Based on the above-described characterization results, a similar phenomenon was related to the concurrence of (a) the higher exfoliation of CM-derived systems, resulting in an enhanced active area and, in turn, in better contact with the reaction   medium; (b) the improved MnO 2 /g-CN linkage in CMderived systems, and their higher defect content, as evidenced by XPS results.In particular, defects resulting from uncondensed −NH x groups can suppress charge carrier recombination, favorably boosting the system's electrocatalytic activity. 24,25Furthermore, a detailed analysis of Figure 9c revealed, for CM-derived specimens, a marked dependence of onset potentials on EPD process times, with the bestperforming system (MnO 2 −CM-45) yielding a potential drop of ≈0.37 V in comparison to that of bare MnO 2 .Based on FE-SEM outcomes (see Figures 2 and S3 and related observations), these effects can be attributed to the fact that longer EPD durations yield an irregular and excessive accumulation of carbon nitride aggregates on MnO 2 , thus hindering hole transport to the g-CN surface and suppressing the advantages brought about by MnO 2 /g-CN junctions.In a different way, a too low carbon nitride content (as for specimen MnO 2 −CM-20) reduces MnO 2 /g-CN heterocontact density, limiting the benefits yielded by MnO 2 /g-CN coupling and thus compromising the resulting functional performances.
Information about the OER kinetics was gained from the Tafel slopes (Figure 9d).The obtained results indicate that g-CN anchoring on MnO 2 promoted a Tafel slope decrease according to the trend: 45 (the latter sample yielding a reduction of 34 mV/dec with respect to MnO 2 ).As can be observed, the obtained slopes presented the same trend of onset potentials (compare Figure 9c,d), with lower values for CM-derived materials and, in particular, for specimen MnO 2 −CM-45.Since lower Tafel slopes are indicative of an improved reaction kinetics, i.e., of a more efficient O 2 evolution at the anode, 9,11,12 these results further validate the positive influence exerted by g-CN(CM) on OER activity.It is worthwhile to observe that the functional performances of the present systems in terms of current densities at 1.65 V vs RHE, as well as Tafel slope and overpotential values, compared favorably not only with some of the previously reported electrocatalysts with analogous composition but also with various RuO 2 and IrO 2 -based benchmark systems (compare data in Table S6 with the ones in Tables S7 and S8).Taken together, these observations demonstrate that the adopted synthetic procedure yielded MnO 2 + g-CN composites of potential interest for a possible practical implementation.The activity order discussed above was in excellent agreement with the ABPE results.As can be observed in Figure 10a, whereas bare MnO 2 was the worst electrocatalyst, g-CN introduction resulted in an increase of ABPE maximum value (Figure 10b), accompanied also by a favorable curve shaping toward lower bias values (see also Table S6, last column).The best-performing MnO 2 −CM-45 electrocatalyst exhibited a more-than-4-fold ABPE maximum increase in comparison to the pristine MnO 2 and presented the highest turnover frequency (TOF; Table S6).Altogether, these results indicate that the coupling of MnO 2 and g-CN(CM) can advantageously boost electrocatalytic performances, provided that a judicious choice of g-CN loading (i.e., of the EPD process duration) is performed.
In view of eventual real-world end uses, a good system service life is a key requirement. 15,17,24Preliminary CA measurements conducted for 1 h (Figure S13) showed that all of the target materials exhibited good stability.In order to attain additional information, CA experiments were conducted for up to 35 h for the best-performing electrocatalyst (Figure 10c), highlighting a constant current density throughout the whole experiment.To unambiguously discard the occurrence of material alterations upon aging, the target systems were stored for six months under ambient conditions and periodically subjected to electrochemical tests.The obtained data (Figure S14) revealed only modest current variations, allowing us to rule out any significant degradation or dissolution, as confirmed by postoperando XRD and FE-SEM measurements (Figures S15 and S16).These results are of high importance toward an eventual practical utilization, considering also that the present systems are eco-friendly and economically viable.
The electrocatalytic activity enhancement promoted by g-CN anchoring on MnO 2 can be traced back to cooperative morphological, catalytic, and electronic effects.In fact, higharea nanoarchitectures like the present ones feature an enhanced OER activity, 8,46,48 since they can (i) maximize the contact with the working solution and provide enough room for reactant transport even into the inner system regions; (ii) reduce the diffusion distance of charge carriers from catalytic sites to the Ni foam substrate and, hence, to the external circuit.These conclusions are corroborated by Tafel slope values, indicating that the anchoring of g-CN results in a promotional catalytic effect.A last important contribution is provided by the interfacial electronic interplay between MnO 2 and g-CN, and, in particular, by the formation of type-II junctions 63,64 (see Figure 11a).Specifically, upon irradiation and bias application, holes flow from the MnO 2 valence band to the g-CN one, whereas electrons can be transferred from the g-CN conduction band to the MnO 2 one.Electrons are then injected into the Ni foam substrate and, subsequently, transported to the cathode, where they are involved in reduction processes leading to H 2 .The most likely OER active sites are pyridinic nitrogen atoms, on which the highest occupied molecular orbital of g-CN is mainly localized. 65,66he important influence exerted by MnO 2 /g-CN junctions on the ultimate material activity is further highlighted by electrochemical impedance spectroscopy data collected for bare MnO 2 and MnO 2 −CM-45 specimens.In fact, the pertaining results (Figure S17) show, under illumination, a more effective charge transfer resistance decrease for sample MnO 2 −CM-45 with respect to that for bare MnO 2 .
The above observations account for the activity increase brought about by composite systems, but how does one explain the different performances of MnO 2 + g-CN(CM) electrocatalysts obtained using different EPD durations?To attain a deeper insight into this issue, flat band potential values were estimated by the intercept of the square photocurrent density curves with the potential axis (Figure 11b). 14,67The Fermi level potentials reported in Figure 11c are the electrode potentials at which no band bending occurs. 67An inspection of the obtained results reveals that the Fermi level position is directly dependent on the target electrocatalyst, with an increase according to the order MnO 2 −CM-600 < MnO 2 −

■ CONCLUSIONS
In summary, eco-friendly electrocatalysts based on graphitic carbon nitride and MnO 2 were fabricated by an original preparation strategy, consisting of the PE-CVD of manganese dioxide nanoarchitectures on Ni foams followed by g-CN electrophoretic deposition and final annealing in air.Particular attention was devoted to the anchoring on MnO 2 of tunable amounts of exfoliated g-CN with a high active area and defect content.A thorough analytical investigation by complementary techniques highlighted the successful obtainment of nanoarchitectures featuring an even g-CN spatial distribution and intimate MnO 2 /g-CN coupling, which is of key importance to benefit from their synergistic interplay in OER electrocatalysis.The best composite system yielded an overpotential of 430 mV to achieve a current density of 10 mA/cm 2 and a Tafel slope of ≈70 mV/dec, favorably comparing with various previously reported MnO 2 + g-CN OER catalysts.These performances, accompanied by remarkable material stability, could be rationalized on the basis of the formation of type-II MnO 2 / g-CN heterojunctions and on the different Fermi level positions as a function of g-CN content, tunable solely as a function of the EPD duration.
In perspective, a computational investigation of the target systems will be of importance to attain deeper insight into the mechanism governing their electrocatalytic performances.Furthermore, the present nanocomposites stand as promising platforms for water treatment applications, and the fabrication strategy utilized in this study can be exploited to prepare a variety of multicomponent non-noble electrocatalysts for H 2 O splitting end uses.In this regard, it is worthwhile noting that the presently adopted preparation route is potentially viable for the industrial production of large-area, porous electrode architectures since both PE-CVD and EPD allow an easy immobilization of the active catalyst components over a variety of substrate materials.

Figure 1 .
Figure 1.Schematic illustration of the fabrication procedure adopted for MnO 2 -based electrodes functionalized with carbon nitride through: (i) PE-CVD of MnO 2 on Ni foams; (ii) EPD of g-CN on the developed oxide nanoarchitectures.The bottom panel displays the deposited mass of g-CN(CM) (see the text) as a function of EPD duration (center), as well as the images of two representative deposits (left).

Figure 3 .
Figure 3. (a) Cross-sectional HAADF-STEM overview and (b,c) magnified images of MnO 2 −CM-45.The SAED pattern recorded on g-CN aggregates is displayed as an inset in (c).(d) Corresponding STEM-EDXS chemical maps for Ni Kα, Mn Kα, N Kα, O Kα, and C Kα and an overlaid image.Pt and W in parts (a) and (b) are due to sample preparation prior to analysis.

Figure 5 .
Figure 5. TEM analysis for the MnO 2 -M-180 specimen.(a) SAED patterns taken from a [001]-oriented MnO 2 dendritic grain (left) and from the flake-like MnO 2 region (right), marked in the HAADF-STEM image reported in (b).Both patterns can be indexed based on the tetragonal β-MnO 2 structure (P4 2 /mnm space group; a = 4.39 Å, c = 2.87 Å). 57,58 (c) HRTEM image of the MnO 2 /Ni foam interface.(d) Higher magnification micrograph of the area framed by the white box in (c).Inset: Fourier transform pattern, confirming the [100] orientation of epitaxially grown MnO 2 grains.(e) HRTEM image of the dendrite/flake MnO 2 interface.(f) HRTEM image of the upper part of the MnO 2 dendrite region in (b).

Figure 6 .
Figure 6.C 1s photopeaks for representative MnO 2 + g-CN samples: (a) MnO 2 −M-180; (b) MnO 2 −CM-45.(c) Sketch of graphitic carbon nitride structure, 61 in which the various C and N sites are marked in the left and right image, respectively.Color codes identifying nonequivalent carbon atoms are the same ones used for the different fitting components in panels (a,b).

Figure 7 .
Figure 7. N 1s photopeaks for representative MnO 2 + g-CN samples: (a) MnO 2 −M-180; (b) MnO 2 −CM-45.Color codes for the different fitting components identifying nonequivalent nitrogen atoms are defined in Figure 6c.(c) Percentage contribution of the N 3 component to the overall N 1s signal for the analyzed specimens.

Figure 8 .
Figure 8. Mn 2p (left) and Mn 3s (right) photopeaks for the MnO 2 and MnO 2 + g-CN samples.In both cases, the component positions for bare MnO 2 are marked with dashed lines.In the right panel, the two Mn 3s components generated by multiplet splitting are marked by A and B.

Figure 9 .
Figure 9. (a) Photograph of the cell used for electrochemical tests (CE = counter-electrode; WE = working electrode; RE = reference electrode).The anodic photocurrent proves the n-type semiconducting nature of the target materials. 14(b) LSV curves recorded under irradiation, (c) onset potentials, and (d) Tafel slopes for the indicated specimens.

Figure 11 .
Figure 11.(a) Schematic energy band diagram for MnO 2 /g-CN systems. 68Conduction and valence band edges are denoted as CB and VB, respectively.(b) Square photocurrent density (J 2 ) vs applied bias for MnO 2 + g-CN(CM) electrodes.Experimental and fitting curves correspond to dashed and continuous lines.(c) Extrapolated Fermi level potentials for the target specimens.