1 Introduction

Since the last decade, nanomaterials have played a critical role in developing new technologies to generate energy and also in devising novel energy storage devices. However, properties of nanomaterials are directly dependent on polymorphism, morphology, size of particles, size distribution, presence of external coatings, and the type of precursor used in the synthesis [1]. Nano-based materials have been extensively considered to develop new energy systems, specifically for renewable energy storage and conversion. Large-scale energy storage devices [2], such as Li-ion [3], Na–S [4], and redox flow batteries (RFBs) [2] have evolved to promote renewable energy generation and storage. Among the unique storage devices, iron redox flow batteries (IRFBs) are suggested to be most promising for large-scale energy storage due to their long cycle life, charge–discharge capabilities, and flexible design. Similarly, carbon-based materials, such as graphite felts (GF) are preferred as electrodes in IRFBs, due to their excellent strength, high electrical conductivity, porous structure, and acid resistance [5]. However, the poor electrocatalytic activity of IRFBs decreases the capacity and lowers the efficiency at higher current densities, thereby preventing their wider application and large-scale commercialization. Enhancing the wide energetic active sites on the surface of GF electrode used in IRFBs and introducing additional functional groups are considered effective to overcome the limitations [5, 6].

IRFBs are taken into consideration to be efficient for grid-level storage. In IRFBs, Fe2+/Fe3+ and Fe2+/Fe act as positive and negative electrolytes, respectively. A typical reaction occurring in the IRFBs is given in Equations 1–3 listed below:

$${{2\text{Fe}}^{2+}\leftrightarrow 2\text{Fe}}^{3+}+ 2{\text{e}}^{- } { {E}}^{0}=0.77 \text{V} \;\text{vs}. \;\text{NHE}$$
(1)
$${\text{Fe}}^{2+} + 2{\text{e}}^{- }\leftrightarrow {\text{Fe}}^{0} {{E}}^{0}=-0.44 \text{V} \;\text{vs}.\; \text{NHE}$$
(2)
$${3\text{Fe}}^{2+} \leftrightarrow {\text{F}\text{e}}^{0}+ {2\text{Fe}}^{3+} {{E}}^{0}=1.21\; \text{vs}.\; \text{NHE}$$
(3)

Electrodes are the essential components in any RFB systems [7] which in turn are dependent on the composition, structure, and performance of the materials used to develop the electrodes. The performance and efficiency of batteries are directly related to the performance of the electrodes. Hence, considerable attention has been given to develop appropriate electrodes or modify the electrodes to achieve the desired battery performance. Among the different materials used to modify the electrodes metal oxides, rare earth oxides, carbon catalysts, functionalized organic materials, heteroatom-doped catalysts are common. These materials act as electrocatalysts in the modified electrodes and increase the effectiveness of redox reactions by exchanging ions and charges.

The electrochemical activity and cyclic stability of IRFBs may be improved by the usage of many noble metals (Pt, Ir, Ru, Au, Pd, etc.) [8, 9]. Such metals have been deposited on the surfaces of GF by the use of numerous techniques. For instance, an increase in electrochemically actives sites on the surface of GF and reduction in the resistance of the GF lead to improved electrical conductivity in the IRFBs. However, metals on GF are affected by the gases released during the redox reactions occurring in IRFBs. The process of deposition of metals also makes the GF expensive. Several studies have been done to develop low-cost metals and carbon-based catalysts with higher electrocatalytic performance. Instead of noble metals, inexpensive metals and metal oxides (Mn3O4, CeO2, WO3, Ti, W, Ni) [10,11,12,13,14] have additionally been used to enhance the electrochemical reactions, but their catalytic activity largely relies upon the scale and stability of the catalysts on the surface of the GF. Therefore, a combination of metal and metal oxide nanocomposites could enhance the electrochemical activity and stability [14]. It is critical to design nanomaterial-based electrocatalysts that are non-toxic, inexpensive, and provide good electrochemical performance and stability to ensure large-scale energy storage devices like IRFBs are not only efficient but also cost-effective and suitable for large-scale energy storage.

Among the various oxides used, cerium oxide (CeO2) with a bandgap of 2.58 eV has gained significance because of its excellent properties like electrocatalytic activity, high oxygen storage capacity, and low energy barrier among III and IV oxidation states [15]. These unique features help to advance the applications of CeO2 for hydrogen generation [16], solar cells [17], environmental remediation [18], oxidation of CO [19], and in developing supercapacitors with high specific capacity [20] or as cathode/anode component in RFBs [21,22,23]. CeO2 when used for electrode modification shows a significant improvement in the cyclic stability and retention capacity of lithium-ion batteries [24]. It should be noted that integrating a proper co-catalyst with an active catalyst to form a heterojunction is an effective method for enhancing electrocatalytical performance [25, 26]. MnO2 as an electrocatalytically active material shows good electrical conductivity, unique catalytic redox ability, and generation of large amounts of surface oxygen species. Recently, reduced graphene oxide@MnO2 [27] and BiOI/MnO2 [28] composite semiconductors were constructed with an inclusive heterojunction of the MnO2 as an active material. However, the prepared heterojunctions based on MnO2 have limitations in terms of complex processes and high cost. During the development of MnO2-based heterojunctions, the addition of a reducing agent (carbon), which is synthesized using organic matter by carbonization [29, 30] is necessary. Considering the overall aspects, CeO2 and MnO2 justify the synthesis of MnO2@CeO2 heterostructure but a relatively simple and cost-effective process should be developed. The heterostructure was formed by the overlapping of MnO2 particles on the surface of CeO2 particle to provide an efficient electrocatalytic activity for further modification of the GF and to enhance the charge/discharge and cycle stability of the IRFBs.

Herein, novel MnO2@CeO2 composites containing oxygen moieties having porous spherical and dry leaf stick morphology have been developed. MnO2@CeO2 composite was first fabricated on the surface of the GF to increase the electrocatalytic activity. Later, MnO2@CeO2 composite-modified graphite felt electrodes (MGF) were used to increase the performance of IRFBs. The proposed MnO2@CeO2 composite structure and its properties effectively improve the iron redox reactions in the IRFBs. An increase in the electrochemical performance and cyclic stability of the MGF was possible and hence, the proposed strategy has great potential for applications in IRFBs.

2 Materials and methods

2.1 Materials

Raw materials used in the experiments include cerium nitrate hexahydrate (Ce(NO3)3·6H2O), urea (CO (NH2)2), potassium permanganate (KMnO4), manganous sulfate (MnSO4·H2O), ferrous chloride (FeCl2), glycine (C2H5NO2), ammonium chloride (NH4Cl), and Ethanol (C2H5OH) (Analytical reagents, AR) which were procured from Bangalore Scientific and Industrial Supplies, Bangalore, India. All chemicals were used as received, without further purifications. The graphite felt electrode was procured from Rayon Graphite felt (AGFHT), USA. The anionic membrane (FUMASEP FAP-375PP) and Nafion binder used were obtained from Fuel Cell Store, USA. The battery components such as reinforced epoxy end plates of 15 mm thickness, copper plates, and gaskets were fabricated in-house, graphite serpentine flow fields were machined with the help of a local vendor M/s Mersen India Pvt Ltd., Bangalore. A schematic of the battery components and assembled system is shown in Scheme 1.

Scheme 1
scheme 1

A typical assembly of the IRFBs

Full size image

2.2 Preparation of MnO2@CeO2 composite

Precisely 0.26 g of Ce (NO3)3.6H2O and 2.25 g of CO (NH2)2 were dissolved in 60 ml of deionized water under stirring at 200 rpm in a beaker. In separate beakers, 0.448 g of MnSO4·H2O and 1.0 g of KMnO4 were dissolved in 30 ml of deionized water separately. The MnSO4·H2O solution was transferred into KMnO4 solution and the two solutions were mixed thoroughly for about 30 min. Later, the mixed solution was transferred into a stainless steel autoclave, and the temperature was raised to 433 K and maintained for 24 h. After the reaction terminated, the solution was allowed to cool down to room temperature. The brownish precipitate formed was removed from the solution by centrifugation at 8000 rpm and the collected material was washed with deionized water until the water became colorless. The prepared product was dried at 333 K for 12 h, and calcined at 573 K for 3 h in a muffle furnace. The final powder having brownish color was stored in an airtight container [20].

2.3 Modification of bare graphite felt electrode (BGF)

A known quantity of MnO2@CeO2 composite was dispersed in alcohol by continuous stirring at 350 rpm for about 5 h at room temperature. Nafion (5 wt%) was added to the above solution and sonicated for about 15 min. Later, either side of the BGF was coated with MnO2@CeO2 composite using a nozzle spray gun (Aimex H-827) having 1.4 mm internal diameter and operating at an air pressure of 58 PSI. Coated electrodes were vacuum dried for about 24 h at 353 K and 25 PSI.

2.4 Characterization

The prepared material was subject to extensive testing and characterizations. The morphological study of the MnO2@CeO2 composite, BGF and MGF were conducted using SEM (HITACHI, SU3500) equipped with an EDAX system. TEM analysis was done by Philips CM-200. XRD investigations were conducted on BRUKER eco-D8 ADVANCE diffractometer working with CuKα radiation (λ = 1.54 Å). Raman spectra of BGF and MGF were analyzed using (HORIBAJOBINYVON, LABRAM) with 532 nm LASER at an exposure time of 5 s. Charge-coupled device (CCD) was used as detector for the analysis with 1800 lines/mm grating. Determination of specific surface area, pore volume, and pore diameter was done on ASAP 2010 Micrometrics instrument.

2.5 Electrochemical studies

Cyclic voltammetry (CV), Electrochemical Impedance Spectroscopy (EIS, frequency range: 0.01 to 105 Hz), and Tafel analysis were carried out using an electrochemical work station (CHI600E Series) with the conventional three-electrode system. The working electrodes used in the study were BGF, MGFs. An Ag/AgCl electrode was used as the reference electrode and platinum was the counter electrode. FeCl2 solution was prepared using deionized water and used as the electrolyte. All the experiments were performed under ambient temperature.

2.6 Charge–discharge studies

A single-cell having 132 cm2 active area with MGF and BGF as positive and negative electrodes, respectively, were used for the charge–discharge studies (Fig. 1). FUMASEP FAP-375PP was used as a separator.

Fig. 1
figure 1

SEM images of a MnO2@CeO2 composite, b BGF electrode, c 1 mg cm− 2-MGF electrode, d 2 mg cm− 2-MGF electrode, e 3 mg cm− 2-MGF electrode, f EDAX of 2 mg cm− 2-MGF electrode

Full size image

The iron electrolyte was prepared using a mixture of 3.25 m FeCl2, 0.3 m C2H5NO2, and 1.0 m NH4Cl in deionized water. Here, C2H5NO2 acts as ligand and reduces the pH imbalance of the electrolytes. NH4Cl increases the conductivity of the FeCl2 electrolyte. The iron electrolyte which is stored in glass reservoirs and acts as anolyte and catholyte was connected to the redox flow cell using a rubber hose. The flow of the electrolyte into the cell was controlled at 130 ml min− 1 using a peristaltic pump (RH-P100L-200-2 h-1D). The charge–discharge and cycling analysis were performed using a Bitrode life cycle tester (LCV4-100/1–48) by varying the current density from 10 to 50 mA cm− 2 and potential between 0.8 and 1.5 V.

Coulombic efficiency (ηC) was calculated using the following equation:

$$\text{Coulombic}\; \text{efficiency}= \frac{\text{Discharge} \;\text{current} \times \text{Discharge} \;\text{time}}{\text{Charge} \;\text{current} \times \text{Charge} \;\text{time}} \times 100$$
(4)

3 Results and discussion

3.1 Morphological and structural analysis

Figure 1 shows the SEM analysis of the MnO2@CeO2 composite, BGF and MGFs and EDAX of the MGF. The morphology of the composite consists of a combination of dry leaves sticks and spherical dust nanoaggregate-like particles as seen from Fig. 1a. Figure 1b represents the microwire morphology of BGF electrode with clean surface. Figures 1c–e show SEM images of the MGF electrodes with an increasing weight of the MnO2@CeO2 composite particles from 1 to 3 mg cm− 2. The dry leaf sticks and dust nanoaggregate-like particles were uniformly deposited on the surface of the GF electrodes. The deposited MnO2@CeO2 composites are porous, have sharp edges, and are unique in structure. Such morphology helps the composite to have more active sites of oxygenated moieties, which are responsible for the enhancement of electrocatalytic activity in 2 mg cm− 2-MGF electrode. The EDAX results (Fig. 1f) show that the MnO2@CeO2 composite does not contain any impurities.

TEM images of MnO2@CeO2 composite at different magnifications are shown in Fig. 2a–c. The composite particles are in nanoscale and their surface is porous as seen from Fig. 2c. The porous structure may have been formed due to the presence of oxygen in the composite moiety. The images also reveal the presence of actives sites on the surface as observed (Fig. 2b) through the overlapping of particles of MnO2 on the CeO2 particles (Fig. 2a). MnO2@CeO2 composites have more active surface for the catalytic performance of the composite in electrochemical applications [31]. Figure 3 shows the XRD diffractograms of MnO2@CeO2 composite, and BGF and MGF electrodes. MnO2@CeO2 composite exhibits a single crystalline peak which crystallized into a characteristic cubic-type lattice and the lattice constants are consistent with the standard JCPDS (Card No. 00-064-0204) with diffraction peaks at 2θ of 28.75°, 37.44°, 47.13°, 56.43°, 60.07°, 69.17°, and 76.85° corresponding to the diffraction planes of (111), (200), (220), (311), (222), (400), and (331), respectively. MnO2@CeO2 composite crystals had all the characteristic diffraction peaks when compared with the standard, thus indicating that the MnO2@CeO2 composite crystals were grown in all orientations. Upon comparison with the standard file, MnO2@CeO2 composite indicated that Mn ions occupy a position within a cubic lattice of ceria with centrosymmetric crystals belonging to Fm-3 m (225) point group [32]. The crystalline phases of BGF and MGF show diffraction peaks at 25.31°, 42.59°, 52.59°, and 77.98° representing the hkl planes (002), (101), (004), and (110), respectively. The intensity of the peaks broadens and the peaks shift in the MGF which was also observed in BGF electrode and MnO2@CeO2 composite, due to the oxidation of the BGF due to thermal exposure during drying or depositing the MnO2@CeO2 composite. Using Scherrer’s equation, the average crystal size of composite particle was determined to be 7.62 nm.

$${D}=\frac{0.9{\lambda }}{{\beta }\text{cos}{\theta }},$$
(5)

where λ is the wavelength of X-rays used (0.154 nm), θ is the Bragg diffraction angle, and β is the full width at half maximum of the XRD pattern of (111) plane.

Fig. 2
figure 2

TEM images of MnO2@CeO2 composite at different magnifications a 100 nm b 50 nm c 5 nm

Full size image
Fig. 3
figure 3

XRD spectra of pure MnO2@CeO2 composite, and BGF and MGF electrodes

Full size image

3.2 BET surface analysis

The adsorption–desorption isotherms and pore distribution curves for MnO2@CeO2 composite obtained from the BET analysis in N2 atmosphere are given in Fig. 4. The adsorption–desorption isotherm plot of relative pressure scale was from 0.50 to 0.99 representing a type IV monolayer formation together with hysteresis loops of type H3. MnO2@CeO2 composites shapes were porous slit or plate, in the form of non-rigid aggregates of the type H3, confirming the mesoporous structure of MnO2@CeO2 composite particle. The size of the particle was in the range of 2–50 nm which confirms the crystal size of 7.26 nm obtained from the XRD analysis using Scherrer’s equation. The pore size of the mesoporous structure of composite particle was between 1 and 3 nm and porosity was 0.219 cc g− 1. The developed MnO2@CeO2 composite had 99.427 m2 g− 1 of the BET surface area. With the presence of a larger surface area and uniform size of fine nanoparticles, the composites may provide good electrocatalytical performance.

Fig. 4
figure 4

a Nitrogen adsorption–desorption isotherms and b corresponding pore size distribution curves of the MnO2@CeO2 composite

Full size image

3.3 Electrochemical analysis

EIS measurements were carried out for BGF and MGF electrodes in FeCl2 electrolyte and the results are shown in Fig. 5. The spectra obtained show a similar pattern of depressed capacitive semi-circle in the high-frequency region, whereas a sloping straight line in the low-frequency region. Rs is bulk solution resistance and Rp is Faradaic interfacial charge-transfer resistance which are important parameters indicating the capacity of the electrolyte and working electrodes. Rs and Rp of BGF and MGF electrodes are listed in Table 1. The Rs of all the electrodes were around 1.5 Ω, except BGF which has a considerably higher Rs of 2.49 Ω. The Rp value of BGF was 76.23 Ω, which may be the presence of inactive sites on the surface or due to high internal resistance. The 2 mg cm− 2-MGF electrode provides the least Rp value of 2.06 Ω, which may be due to the presence of oxygen moieties within the MnO2@CeO2 composite and on the surface of the electrodes. The oxygen-active sites enhance the redox reaction of the FeCl2 electrolyte at the interface of the electrode. The resistance of MGF was lowest up to a certain extent of modification, beyond which the resistance increases due to the increase in the internal resistance of MnO2@CeO2 composite.

Fig. 5
figure 5

Electrochemical impedance spectra of BGF and MGF electrodes

Full size image
Table 1 Parameters obtained by fitting the Nyquist plots
Full size table

The CVs of BGF and GOMGF electrodes are shown in Fig. 6a. 2 mg cm− 2-MGF electrode showed a significant difference in the redox peaks compared to BGF electrode. Among all modified electrodes, 2 mg cm− 2-MGF electrode exhibits the highest level of electrochemical performance. The relative CV which represents the difference of cathodic and anodic peak potential (∆E) for 2 mg cm− 2-MGF electrode is around 0.203 V at a scan rate of 50 mV s− 1 compared to BGF and MGF electrodes. Similarly, the ratio of anodic (Ipa) and cathodic (Ipc) peak current densities is 1.131 which means considerably higher ∆E than the BGF 0.365 V, the relative Ipa/Ipc is 1.409 which indicates irreversible reactions towards the Fe2+/Fe3+ redox process. The derived electrochemical data are given in Table 2. The lower ∆E and ratio of Ipa and Ipc approaching 1 in 2 mg cm− 2-MGF electrode represent better reversibility and good electrochemical catalytic activity towards Fe2+/Fe3+ redox reactions. Figure 6b shows the CV variations as the scan rate changes from 10 mV s− 1 to 100 mV s− 1. The 2 mg cm− 2-MGF electrode exhibits the best reversibility and electrochemical catalytic activity. The CV studies have good agreement with the EIS analysis of BGF and MGF electrodes (weight cm− 2) based on the ∆E and Ipa/Ipc values.

Fig. 6
figure 6

a Cyclic voltammograms on different electrodes in FeCl2 electrolyte at a scan rate of 50 mV s− 1 and b variation of scan rates at 2 mg cm− 2-MGF electrode

Full size image
Table 2 Electrochemical parameters obtained from CV curves
Full size table

Tafel plot of BGF and MGF electrodes is shown in Fig. 7. The linear polarization resistance values calculated from the Tafel plot are given in Table 3. The calculated linear polarization resistance decreases for the BGF, 1 and 2 mg cm− 2-MGF electrodes, but the 3 mg cm− 2-MGF electrode showed sudden increase in the linear polarization resistance. Among all electrodes, 2 mg cm− 2-MGF electrode provides a very low resistance of 5 ohms, whereas 3 mg cm− 2-MGF electrode shows a much higher resistance of 59 ohms, which may be due to the increase in internal resistance because of high amount of MnO2@CeO2 composite on GF electrode. In 3 mg cm− 2-MGF electrode, there are higher amount of MnO2@CeO2 composite particles as observed from the SEM images (Fig. 1e). The lower linear polarization resistance of MGF may also contribute to the high electrocatalytic activity and leads to the deposition of MnO2@CeO2 composite on the GF electrode.

Fig. 7
figure 7

Tafel plot of BGF and MGF electrodes

Full size image
Table 3 The linear polarization resistance value of the electrodes from the Tafel plot
Full size table

3.4 Performance characterization of single-cell flow battery

The charge–discharge studies were carried out using 132 cm2 cell between current density of 40 and 70 mA cm− 2 with C2 rating. In the cell, BGF and MGF electrodes were used as positive electrodes and BGF as a negative electrode. The efficiencies of the cell using MGF electrodes were calculated and compared at 50 mA cm− 2. However, BGF electrode exhibits poor electrochemical performance and charge–discharge studies were limited to 20 mA cm− 2 as indicated in Fig. 8a. The 2 mg cm− 2-MGF electrode provides the best efficiencies (98.96%) than the BGF and other MGF electrodes as shown in Fig. 8b. The corresponding ηC and PPD are provided in Table 4. The 2 mg cm− 2-MGF electrode shows good reversibility for the redox process for the FeCl2 system (Fe2+/Fe3+), due to the lower charge-transfer resistance and lesser electrochemical polarization as evident in the electrochemical studies.

Fig. 8
figure 8

Charge–discharge studies for a BGF (20 mA cm− 2), b MGF (50 mA cm− 2) electrodes, c 2 mg cm− 2-MGF electrode at 40–70 mA cm− 2, d ηC characteristics against number of cycles

Full size image
Table 4 Battery performance characterization for BGF electrodes (20 mA cm− 2) and MGF electrodes (50 mA cm− 2)
Full size table

Figure 8c represents the performance of 2 mg cm− 2-MGF electrode at current densities between 40 and 70 mA cm− 2. The ηC and PPD were between 88.22 and 99.10% and between 40.77 and 57.32 mW cm− 2, respectively (Table 5).

Table 5 Battery performance characterization using 2 mg cm− 2-MGF electrode
Full size table

In addition, the stability of the cell was studied using 2 mg cm− 2-MGF electrode and the data were plotted for number of cycles Vs ηC as shown in Fig. 8d. The ηC for first cycle was found to be 98.74% which increases to 98.97% at second cycle, but decreases gradually and becomes 85.46% after the 25th cycle. The decrease in the ηC may be due to the loss of stability of electrolyte or saturation of the modified electrode (2 mg cm− 2-MGF). The average ηC and PPD of the cell after 25 cycles was found to be 96.06% and 55.56 mW cm− 2, respectively. At a current density of 70 mA cm− 2, ηC was 88.22%. The performance results provide good evidence of the stability of the MnO2@CeO2 composite-modified electrode.

Most of the reports on IRFBs use non-aqueous iron electrolytes with three-electrode systems and activation area lesser than 25 cm2. Few studies also report aqueous iron electrolytes but mainly focus on ligand optimizations [33,34,35,36,37]. The performance of IRFBs were studied at different conditions and are compared in Table 6. The IRFBs reported have produced better efficiencies of 90–97% at much lower current densities (≤ 10 mA cm− 2). The non-aqueous iron electrolytes used in the IRFBs systems are not economical and also pose a hazard to the environment. In this work, an effort has been made to demonstrate the cell (132 cm2) with higher active area using modified electrode and optimized aqueous electrolyte. The 2 mg cm− 2-MGF electrode provided an efficiency of 98.96% (50 mA cm− 2) ηC as compared with BGF electrode (44.36%@20 mA cm− 2).

Table 6 Performance characterization of various cells in iron electrolytes
Full size table

4 Conclusions

A simple hydrothermal process was used to synthesize the MnO2@CeO2 composite. Synthesized MnO2@CeO2 composites were used to modify GF electrode by electrostatic spraying after ultra-sonication. Performance characterization of BGF and MGF electrodes were studied using FeCl2 as an electrolyte. At 50 mA cm−2, the 2 mg cm−2-MGF electrode exhibits good charge/discharge characteristics with a ηC of 98.96%. The 2 mg cm−2-MGF electrodes had enhanced reversibility of the redox reaction and lower resistance in electrochemical studies, which is beneficial to obtain good results during the cyclic performance of IRFBs. The obtained performance was attributable to the unique morphology and oxygen moieties on surface properties. The modification of the GF electrode results in good efficiency and cyclic performance reveals that IRFBs are potentially feasible for grid-level energy storage.