Development of low-cost nitrogen- and boron-doped carbon black cathode catalysts for the improvement of hydrogen-bromine flow battery cathode kinetics

In this study, carbon blacks, containing nitrogen and boron, are used as cathode catalysts first time in a hydrogen-bromine redox flow cell. Nitrogen and boron doping has been applied to carbon blacks to improve the limited properties of the Vulcan XC72 (VXC). Here, we have examined the effect of nitrogen-/boron-doped Vulcan XC72 carbon black on hydrogen-bromine flow cell performance. Nitrogen gas and boric acid are used as nitrogen and boron sources in the post-doping process, respectively. XRD, SEM, XPS, Raman spectroscopy, and N2 adsorption analyses are used to characterize the structures of cathode catalysts. The electrochemical characterization of the catalysts has been carried out with the cyclic voltammetry technique using the conventional three-electrode system connected to a potentiostat. We have addressed the effect of nitrogen and boron doping into the carbon black onto the flow battery performance by comparing their polarization and power curves. The maximum power densities with the VXC, N-VXC, and B-VXC cathode catalysts have been measured at 0.75 V as 360, 390, and 430 mW cm−2, respectively.


Introduction
Increasing concerns about the environment and changing policies necessitate the use of predominantly renewable energy sources shortly. Government and international organizations support innovative and sustainable solutions aiming to protect the environment [1]. In this context, the use of renewable energy sources has become quite common. Energy systems, such as solar and wind, generate intermittent energy by nature. Electrical energy storage (EES) systems are needed to increase the efficiency of renewable energy and maximize its use [2][3][4]. Flow battery technology is a reversible EES system. Flow batteries store the electrical energy produced by natural resources as chemical energy in intermittent energy generation and distribution systems and convert this chemical energy into electrical energy at the time of demand. The power and energy potential of the flow battery system are decoupled by storing the reactants out of the flow cell [5][6][7][8]. Within the flow battery technologies, the H 2 /Br 2 flow battery (HBFB) is a promising system due to its affordable cost and fast kinetics of bromine pairs (Fig. 1).
In a typical HBFB system, bromine and hydrogen are the reactants used to store electrical energy. Platinum and platinum group catalysts generally catalyze the HOR and HER reactions at the anode. On the other hand, carbon structures or platinum and platinum group catalysts may catalyze bromine oxidation and reduction reactions at the cathode (BOR/ BRR). Between the anode and cathode, there is a solid polymer electrolyte membrane that separates the reactants from each other and provides proton conduction [9][10][11]. The total charge/discharge reaction of HBFB is given below: Although the platinum catalyst used in the anode and cathode is highly active, it is not stable in the bromine environment [12,13]. One of the most critical problems of the HBFB system is that the bromine species (Br − , Br 3− , and Br 5− ) in the cathode reach the anode by crossover through the polymer electrolyte membrane, poison the platinumbased catalyst, and significantly lower the performance of the flow battery. Various studies have been conducted in the literature to solve this problem [14][15][16][17][18]. Another issue affecting the battery performance results from the need for a high-activity and low-cost catalyst that can operate in a corrosive Br 2 /HBr environment for the bromine redox reactions at the cathode.
Carbon materials have plenty of free electrons flowing. In fuel or flow cells, electrochemical reactions are generally catalyzed by carbon-supported catalysts or carbons. However, the activity of free-flowing electrons is considerably low in the carbons. The addition of heteroatoms, such as nitrogen and boron, to the structure, may improve the limited electrical and chemical properties of the carbon-based catalysts [19][20][21]. It is seen in the literature that the contribution of heteroatoms rich in electrons (such as nitrogen and phosphorus) and electron-deficient (such as boron) increases the catalytic activity in reactions using carbon material [22][23][24][25]. Since boron, carbon, and nitrogen are located consecutively in the second row of the periodic table, investigating the catalytic effects of B or N modified carbon is a good practice for comparison. There are various studies on the addition of nitrogen or boron to the carbon structure in the literature. In situ doping and post-doping techniques are the two categories gathering these research studies. In situ doping is the direct addition of heteroatom during the synthesis of carbon material. However, in the post-doping process, the pre-synthesized or existing carbon structures are treated by a nitrogen or boron source, such as urea, NH 3 , N 2 , boric acid, B 2 H 6 . The post-doping technique is an easy and promising process widely used for both nitrogen and boron addition.
In this work, alternative cathode catalysts to platinumbased catalysts, which are unstable in the bromine environment, have been produced. Here, we aimed to develop nitrogen or boron added Vulcan XC72 carbon black cathode catalysts with the post-doping technique for use in hydrogenbromine flow batteries for the first time.

Synthesis of N-doped carbon catalysts
In all experiments, carbon black (Vulcan XC72)was used as the cathode electroactive material.The post-doping method is applied to add nitrogen atoms into the carbon black structure using nitrogen gas as the source. In brief, carbon blacks were pretreated with 6 M HCl for 24 h before nitrogen loading to activate and consequently functionalize the carbon structure. Afterward, carbon blacks were washed in a vacuum filter system with deionized water until the pH was neutral. The carbon blacks were dried overnight at 100 °C. After the pretreatment, the carbon blacks were loaded into a quartz tube and placed in a furnace. The nitrogen loading process was carried out at 900 °C for 100 min with a heating rate of 5 °C min −1 in a tubular furnace. The resultant nitrogen-doped Vulcan XC72 carbon black was named N-VXC.

Synthesis of B-doped carbon catalysts
The boron loading process was achieved with the hydrothermal method using boric acid as the boron source. First, boric acid (99% H 3 BO 3 , Merck®) and carbon black were added respectively to the ethanol-water mixture (v/v = 3:1). The amount of boric acid added was approximately five times the amount of carbon black. After stirring the resultant mixture for an hour, the mixture was transferred into a Tefloncoated stainless steel autoclave. The reaction was carried out at 180 °C for 12 h. After that, the material was filtered and washed with DIW. Finally, the boron-doped Vulcan XC72 carbon black (B-VXC) was vacuum-dried in an oven overnight at 60 °C.

Physical characterization
X-Ray diffraction (XRD, Rigaku Smartlab) spectroscopy was used to investigate boron and nitrogen doped carbon blacks (scanning angle 2-90°, and scanning rate 2 o min −1 ). Scanning electron microscopy (SEM, JEOL JSM-7001F) analyzes were used for surface morphology with an acceleration voltage of 10 kV. N 2 adsorption tests were performed with a surface properties analyzer (Quantachrome Autosorb IQ2) to determine the surface area and pore size of nitrogen and boron-doped carbon black cathode catalysts. Additional physical characterization of the cathode catalytic materials was conducted with X-ray photoelectron spectroscopy (XPS, PHI 5000 VersaProbe) and Raman spectroscopy (Bruker FRA-106/S) measurements.

Electrochemical measurements
Cyclic voltammetry analyzes were performed under nitrogen environment using Ivium workstation and a standard three electrode electrochemical cell. Ag/AgCl (3 M NaCl) and coiled platinum wire with gold-plated connector were used as reference (RE-5B, Basi®) and counter electrodes (MW-1033, Basi®), respectively. Glassy carbon electrode (GCE, d = 3 mm, MF-2066, Basi®) was used as working electrode. GCE was rinsed with large amounts of DI water. For the catalyst ink, 10 µL of 0.05 wt.% Nafion solution was dissolved in 0.5 mL ethanol in an ultrasonic bath. Afterwards

MEA preparation
In this study, solid polymer electrolyte Nafion-212 (DuPont®) was used as a proton exchange membrane in all experiments. Before the membrane electrode assembly (MEA) preparation, Nafion-212 membrane was cut to 4.2 × 4.2 cm −2 in size and pretreated for activation. When preparing the electrode on the anode side, 2.2 × 2.2 cm −2 non-woven Sigracet 25BC carbon paper(SGL Carbon®) with a microporous layer (MPL) that has been 5% PTFE treated was used for all samples. Using the brush-paint method, the commercial Pt/C (67 wt.%, Tanaka TKK®, Japan) catalyst was loaded onto the MPL layer of Sigracet 25BC carbon paper in an amount of 0.75 mg Pt cm −2 .
Pt/C catalyst ink was prepared with the required amount of Pt/C catalyst, 4 mL isopropanol, 2 mL DI water, and required amount of 15 wt.% Nafion solution (30 wt.% dry Nafion based on Pt/C amount, Ion Power®). In all tests, the same electrode was used on the anode side. When preparing the cathode electrode, Sigracet 10AA (SGL Car-bon®) carbon paper was used for all samples. Cathode catalyst (VXC, B-VXC, N-VXC) were loaded onto the 2.2 × 2.2 cm 2 Sigracet 10AA carbon paper using Nafion solution as binder. Loading amount of cathode catalyst was kept at 0.75 mg cm −2 . Finally, by placing the Nafion-212 membrane between the anode and cathode electrodes, the MEA structure was formed in a hot press for 3 min at 130 °C and 2 bars.

Performance of hydrogen-bromine redox flow battery
A 5 cm 2 single cell hydrogen-bromine flow battery with interdigitated flow field architecture was used to record polarization and power curves. All flow battery tests were carried out at 30 °C with a PID controlled heater. MEA, Teflon gaskets (175 µm), graphite bipolar plates with interdigitated flow field, gold-plated electrolytic copper current collector plates and eloxal coated aluminum end plates with heating tapes are components used in a typical single cell hydrogen-bromine flow battery. All cell components are compressed with torque wrench by applying 4 N.m torque per bolt. Performance tests were carried out with the setup shown in Fig. 2. The flow rate of hydrogen gas (anolyte) was set to 0.1 L min −1 (no humidification). 1 M HBr (47.0%, Merck®)/1 M Br 2 (99.0% Merck®) solution (catholyte) was circulated with a peristaltic pump (Watson Marlow 323E) and flow rate was adjusted to 50 mL min −1 . Before each flow battery test, nitrogen gas was circulated from the anode and cathode side of the flow battery for approximately 20 min. Hydrogen-bromine flow battery discharge performance tests were carried out using electronic load (Maynuo®M9714). Performance tests were conducted by recording steady state currents starting from open circuit voltage at 0.05 V intervals by operating electronic load under potentiostatic mode [27].

Physical characterization
Various physical characterization techniques have been used to understand the structure of carbon black cathode catalysts developed with pretreatment and nitrogen and boron addition. The XRD patterns of the samples are given in Fig. 2.
The peaks at 24.8° and 43.7° in all XRD patterns correspond to the (002) and (100) crystal planes of carbon (JCPDS card No. 41-1487), respectively [19]. The graphite plane (002) may refer to the piling of thin graphene layers. Moreover, the sp 2 hybridizing carbon structures are related to the graphite plane (001) [28]. The peak positions of all cathode catalysts are similar without any shift, indicating that doping exerts a negligible effect on the carbon crystal structure. A change in peak intensity occurred at 24.8°. The peak density decreased significantly after nitrogen and boron loading. The decrement in peak density explains the heteroatom (boron or nitrogen) addition into the structure [29,30].
The surface morphology of the nitrogen/boron-loaded Vulcan XC72 cathode catalysts was identified by SEM analysis, and SEM micrographs are presented in Fig. 3. Clustered structures in different forms have been observed. In addition, carbon structures exhibit a partially regular size distribution. The presence of fibrous structures was determined in some parts of the surface shape structures of nitrogen and boron-containing carbon catalysts.
Nitrogen adsorption analyses were conducted to investigate the structural properties of the cathode catalysts. The surface areas of VXC, N-VXC, and B-VXC were calculated using N 2 adsorption/desorption isotherms. N 2 adsorption analysis data of VXC, N-VXC and B-VXC structures are given in Fig. 4.  a and b), N-VXC (c and d) N 2 adsorption-desorption isotherms appear to be Type IV isotherms according to the IUPAC classification, and these features indicate that the catalysts exhibit micro and mesopore structures [31]. Adsorption in the region of P/P 0 partial pressure value close to zero occurs on micropores.
The surface area, macropore size, and micropore size were calculated by BET (Brunauer-Emmett-Teller), BJH, and HK methods, respectively. Results of these structural parameters are tabulated in Table 1.There was no change in the micropore diameter for all samples. N-VXC and B-VXC pores are macropores with pore sizes of 120 and 106 nm, respectively. The smaller pore size of B-VXC compared to VXC (118 nm) indicates the presence of heteroatom contribution in the structure. This result is also evident when examining BET-specific surface areas. Pore activation in heat treatments creates an increase in surface area. As a result of the high-temperature process, only a slight increase in surface area and macropore size occurred for N-VXC. In addition, N-VXC has a more centralized pore size distribution. In other words, it has a higher capacity to provide more  [24,37,38]. N1s peak positions and binding energies have been attributed to functional groups pyridic N (397.5 eV), prolic N (399.4 eV), graphitic N (400.6 eV) [34,39].
Raman spectroscopy results of the nitrogen and boron containing Vulcan XC72 is presented in Fig. 6. The G band indicates the degree of graphitization of the carbon domains, while the D band is refers to with defects in the carbon matrix.When the Raman spectra of VXC were examined, two peaks of 1339 cm −1 and 1598 cm −1 were observed for the D and G bands, respectively (Fig. 5). The D and G bands of the N-VXCindicate a shift of the peaks to 1339 cm −1 and 1585 cm −1 , respectively.
The D and G bands of the B-VXCexhibit a shift of the peaks to 1339 and 1579 cm −1 , respectively. The N-VXC and B-VXC have higher D and G band density ratios (I D /I G ) compared to the Vulcan XC72, indicating that they produce more defects in the carbon frame with nitrogen and boron doping. Thus, defects increase in association with the shift of the G band with the contribution of N or B atoms [40].

Cyclic voltammetry (CV) analysis
The electrochemical activities of nitrogen and boron containing carbon structures and Vulcan XC72 catalysts were researched by cyclic voltammetry analysis. Figure 7 shows cyclic voltagrams in the 0-1.0 V potential region for As the cyclic voltammetry results for VXC, B-VXC, and N-VXC are analyzed, the maximum current values are measured as 0.21, 0.36, 0.44 mA, respectively (Fig. 7a). The bromine reduction reaction peak also increased proportionally with the increase in scanning rate. Depending on the increase in scanning speed, a reverse shift occurred in the bromine reduction peak positions. This indicates that the bromine reduction reaction in the HBr/Br 2 environment is semi-reversible [41].
The stability of nitrogen and boron containing carbon structures and Vulcan XC72 carbon black catalysts was observed by cyclic voltammetry analysis up to 200 cycles at a scan rate of 50 mVs −1 and is presented in Fig. 8.
When the cyclic voltammetry results were examined, no major change was observed in the peaks even at the 200th cycle. This result shows that VXC, B-VXC and N-VXC are stable under these conditions.
The following Randles-Sevcik equation can be used to explain the effect of sweep speed on peak current. Cyclic voltammetry analysis was performed at different scan rates to find the diffusion coefficients of Vulcan XC72, nitrogen and boron containing carbons and bromide ions in HBr solution. In cyclic voltammetry analysis, when the scan rate was changed from 10 to 50 mVs −1 , the oxidation and reduction peak currents increased, showing an almost linear relationship with the square root of the scanning rate [25,26].
Here, i p is the peak current (A), n is the number of electrons transferred, D is the diffusion coefficient (cm 2 s −1 ), A is the area of the electrode surface (cm 2 ), υ (Vs −1 ) is the scanning rate and C is the concentration (mol cm 3 ). Table 1 lists the diffusion coefficients for bromide ions in solution.
When the diffusion coefficients of nitrogen and boroncontaining carbon catalysts in bromine media were examined (Table 2), it was determined that the nitrogen-containing Vulcan XC72 had the highest diffusion coefficient. The faster and shorter transfer of bromine ions in the N-Vulcan XC72 catalyst increased the performance. This result is also compatible with the voltammogramdata.   Linear sweep voltammetry (LSV) was used to evaluate the effect of heteroatom addition on electrochemical performance in bromine medium (Fig. 9).The polarization curves were obtained from linear sweep voltammetry at a constant rotation rate of 400 rpm.
N-VXC has been observed to have the best activity over its entire potential range. As the potential increases, the current also increases, indicating more Br 2 /Br − reduction.

Single cell H 2 /Br 2 flow battery performance
Flow cell performance curves of Vulcan XC72 carbon blacks, containing nitrogen and boron, are presented in Fig. 10. It is seen that B-VXC exhibited the highest performance in the single-cell flow battery tests. The maximum power densities of VXC, N-VXC, and B-VXC were measured as 360, 390, and 430 mW cm −2 , respectively (Fig. 10b).
The most significant difference was observed in the high current density region. The performance degradation due to mass transfer loss in this region is lowest with the B-VXC catalyst. In nitrogen-doped carbon, crossover effects (a return in the polarization curve) due to mass transfer effects were observed at relatively high potentials. Nafion membrane is not resistant enough for the bromine species. This poisons the hydrogen electrode causing a looping behavior in the polarization curve. To avoid performance degradation, it is necessary to develop a crossover resistant membrane and anode catalysts that can perform well in the presence of bromine species [42,43]. The doping of boron atoms into the carbon structure makes the carbon matrix more electropositive, leading to activation in the reduction reaction by increasing bromine adsorption. In addition, the B-VXC alleviated the effects of the bromine species crossover, especially in the high current density region. It is necessary to conduct charge/discharge cycle tests to elucidate this promising effect.

Conclusion
In this study, we investigated the effect of nitrogen-or boron-doped Vulcan XC72 carbon black on the cathode bromine redox reaction kinetics. During discharge mode, the hydrogen-bromine flow cell exhibited promising performance by achieving a peak power of 430 mW cm −2 with the boron-doped cathode electrode.It was confirmed by cyclic voltammetry tests that the catalytic activity of the cathode catalyst increased significantly with heteroatoms added to the carbon black structure. Hydrogen-bromine flow battery performance results showed that boron addition in the cathode catalyst enabled decreasing mass transport limitations at higher current densities.
Nitrogen and boron doping with low-cost methods to the widely used Vulcan XC72 carbon black resulted in remarkable improvements in peak power density of approximately 8.5% and 20%, respectively. Further research is required to boost the performance and stability of hydrogen-bromine flow batteries by developing novel membranes alleviating bromine species crossover and improving anode electrode kinetics with bromine species tolerant HOR/HER catalysts.