MOF-Derived CuPt / NC Electrocatalyst for Oxygen Reduction Reaction

: Metal-organic frameworks (MOFs) have been at the center stage of material science in the recent past because of their structural properties and wide applications in catalysis. MOFs have also been used as hard templates for the preparation of catalysts. In this study, highly active CuPt / NC electrocatalyst was synthesized by pyrolyzing Cu-tpa MOF along with Pt precursor under ﬂowing Ar-H 2 atmosphere. The catalyst was characterized by scanning electron microscopy (SEM), transmission electron microscopy (TEM), X-ray photoelectron spectroscopy (XPS), and X-ray powder di ﬀ raction (XRD). Rotating disk electrode study was performed to determine the oxygen reduction reaction (ORR) activity for CuPt / NC in 0.1 M HClO 4 at di ﬀ erent revolutions per minute (400, 800, 1200, and 1600) and it was also compared with commercial Pt / C catalyst. Further the ORR performance was evaluated by K-L plots and Tafel slope. CuPt / NC shows excellent ORR performance with onset potential of 0.9 V (vs. RHE), which is comparable with commercial Pt / C. The ORR activity of CuPt / NC is demonstrated as an e ﬃ cient electrocatalyst for fuel cell.


Introduction
Fossil fuels are neither a clean nor sustainable source of energy. Therefore, the global energy is going through a transition from fossil fuels to clean and sustainable energy resources [1]. To resolve the rising global energy crisis and the related environmental problems, an immense deal of research attention has been paid for developing substitute green energy conversion and storage technologies, together with water splitting, metal-air batteries, and fuel cells in the past decade [2]. Oxygen reduction reaction electrocatalysis is one of the mainly considered topics because of its significance for electrochemical energy conversion and storage devices. In electrochemical systems it deals with the oxygen electrode reactions. A prevalent interest intended for the electrochemical oxygen reduction reaction has two facets. The following reaction attracts substantial consideration from basic research point of view; in addition to this, it is the most significant reaction for energy conversion application in electrochemical energy devices. It has been focused in theoretical study as 4ereaction, incredibly sensitive to the surface structural and electronic characteristics. It might consist of a number of elementary reactions [3]. Moreover, oxygen reduction reaction (ORR) is the reaction happening at the cathode, which presides over the energy transfer efficiency of fuel cells. phosphorous and nitrogen doping using Cu-MOF, this catalyst exhibited a remarkable performance as a bi-functional electrocatalyst for ORR and hydrogen evolution reaction (HER) [24].
In this study, a novel electrocatalyst for ORR was successfully synthesized by pyrolysis of Cutpa MOF with very low Pt loading. This Cu-tpa MOF-derived electrocatalyst has nano-porous carbon (NC) formed in its structure. The ORR performance of the CuPt/NC is comparable with commercial Pt/C.

Results and Discussion
A detailed literature review was conducted to devise a synthesis technique for Cu-tpa MOF [25][26][27]. Figure 1 shows the FTIR spectra of Cu/NC, Cu Pt/NC, and Cu-tpa MOF. The peaks for C-H stretching vibrations are at 1106 cm −1 and 1174 cm −1 . Whereas, peaks positioned at 1384 cm −1 and 1610 cm −1 are characterized for C-H and >C=C< stretching bonds. Furthermore, peaks at 2373 cm −1 , 3440 cm −1 correspond to the presence of carbonates and O-H bonds absorbed on the surface [4,28,29].  shows SEM images with same magnification for Cu-tpa MOF, Cu/NC, and CuPt/NC. The average Cu-tpa MOF particle size is around 2 µ m, these particles do not possess specific and regular shape but shows a good crystallinity [27]. After pyrolysis, the sample was transformed into a black powder because of the carbonization of organic linker (Figure 2b,c).
During the process of pyrolysis, the Cu-tpa MOF decomposes partially as the unstable organic groups evaporate, leaving only the nano-porous carbon behind. Figure 2b,c illustrate the formation of nano-porous carbon onto the surface of Cu-tpa particles. With Pt loading surface becoming more rough and porous, that can be observed in Figure 2c.  Figure 2 shows SEM images with same magnification for Cu-tpa MOF, Cu/NC, and CuPt/NC. The average Cu-tpa MOF particle size is around 2 µm, these particles do not possess specific and regular shape but shows a good crystallinity [27]. After pyrolysis, the sample was transformed into a black powder because of the carbonization of organic linker (Figure 2b,c).   During the process of pyrolysis, the Cu-tpa MOF decomposes partially as the unstable organic groups evaporate, leaving only the nano-porous carbon behind. Figure 2b,c illustrate the formation of nano-porous carbon onto the surface of Cu-tpa particles. With Pt loading surface becoming more rough and porous, that can be observed in Figure 2c. Figure 3 shows the EDS spectrum of all three samples including Cu-tpa MOF, Cu/NC, and CuPt/NC. Carbon, oxygen, and copper peaks are visible in all three samples, while the Pt peak is observed only in CuPt/NC. Likewise, the elemental percentages from EDS are shown in the Table 1. Cu-MOF has the highest percentage of carbon while the other two pyrolyzed samples have relatively lower percentage of carbon as the unstable organic groups have evaporated after the pyrolysis which also decreases the carbon percentage. The relative percentage of Cu is increased after pyrolysis as seen in Table 1. Because of the very low Pt loading, the EDS only shows 1 wt % of Pt loading in CuPt/NC because of the limited selection of area for EDS, while the actual loading of Pt was 10%.  To study the morphologies and size of Cu Pt/NC, TEM images were recorded as shown in Figure  4. From the TEM images it can be observed that the pyrolyzed sample contains numerous small CuPt nanoparticles distributed uniformly on carbon matrix (Figure 4a). The size of CuPt/NC is nearly 5-7 nm (Figure 4b) as depicted from TEM images.   To study the morphologies and size of Cu Pt/NC, TEM images were recorded as shown in Figure 4. From the TEM images it can be observed that the pyrolyzed sample contains numerous small CuPt nanoparticles distributed uniformly on carbon matrix (Figure 4a). The size of CuPt/NC is nearly 5-7 nm (Figure 4b) as depicted from TEM images.
The crystalline structure of Cu-tpa MOF and CuPt/NC (pyrolyzed) was determined with powder XRD as shown in ( Figure 5). XRD pattern of Cu-tpa MOF has been transformed after pyrolysis because of the removal of unstable organic linker groups and the formation of nano-porous carbon. Cu and Pt  To study the morphologies and size of Cu Pt/NC, TEM images were recorded as shown in Figure  4. From the TEM images it can be observed that the pyrolyzed sample contains numerous small CuPt nanoparticles distributed uniformly on carbon matrix (Figure 4a). The size of CuPt/NC is nearly 5-7 nm (Figure 4b) as depicted from TEM images.   Figure 6 displays the XPS spectra of CuPt/NC which confirms the same surface composition as that of the bulk. XPS further confirms the presence of Pt. Pt (4f) and Cu (2p) peaks were observed at 71.5 eV and 932.9 eV, both the peaks shifted toward higher binding energies as compared to pure Pt (71.2 eV) and pure Cu (932.7 eV). The shift in peaks along with the Pt (4f) peak broadening is attributed to Cu-Pt metal interaction and the formation of CuPt nanoparticles on carbon framework [32,33]. The C1s is aligned at 284.42 eV to compensate for the sample surface charging. Similarly oxygen spectrum deconvoluted into three peaks positioned at 530.15, 531.81, and 532.41 eV represents M-O, C=O, and C-O respectively [34].  Figure 6 displays the XPS spectra of CuPt/NC which confirms the same surface composition as that of the bulk. XPS further confirms the presence of Pt. Pt (4f) and Cu (2p) peaks were observed at 71.5 eV and 932.9 eV, both the peaks shifted toward higher binding energies as compared to pure Pt (71.2 eV) and pure Cu (932.7 eV). The shift in peaks along with the Pt (4f) peak broadening is attributed to Cu-Pt metal interaction and the formation of CuPt nanoparticles on carbon framework [32,33]. The C1s is aligned at 284.42 eV to compensate for the sample surface charging. Similarly oxygen spectrum deconvoluted into three peaks positioned at 530.15, 531.81, and 532.41 eV represents M-O, C=O, and C-O respectively [34].

Electrode Preparation and Electrochemical Evaluation
For evaluation of electrochemical performance, catalyst ink was prepared by dissolving 2 mg of catalyst (Cu/NC, CuPt/NC, and commercial Pt/C catalyst 46.1 wt % Tanaka Kikinzoku Kogyo, TEC10E50E) in 1 mL of stock solution. The suspension was sonicated in ultrasonic bath for 30 min.
Catalysts 2020, 10, 799 6 of 12 Stock solution was prepared by mixing together 7.6 mL of DI water, 2.4 mL of iso-propyl alcohol, and 0.1 mL of 5 wt % NAFION™ dispersion (Ion power LQ−1105-1100 EW) [25]. The catalyst electrode film was prepared on glassy carbon electrode (with diameter of 5 mm from Pine Research AFE5T050GC) by dropping 10 µm of ink on it and dried for 20 min at 700 rpm. The thin films of commercial Pt/C and Cu/NC were also deposited for comparison. The three electrode assembly system was used for electrochemical measurements. Pt wire was used as counter electrode and saturated calomel electrode (SCE) as reference electrode. To test the ORR activity, cyclic voltammetry (CV) and linear sweep voltammetry (LSV) measurements for CuPt/NC and Pt/C catalysts were taken on rotating disk electrode in 0.1 M HClO 4 solution. The solution was purged with N 2 and O 2 gases for 20-30 min before the electrochemical tests [35]. LSV tests were performed at 400, 800, 1200, and 1600 rpm. All the electrochemical tests were conducted at voltage scan rate of 30 mV s −1 .  Figure 6 displays the XPS spectra of CuPt/NC which confirms the same surface composition as that of the bulk. XPS further confirms the presence of Pt. Pt (4f) and Cu (2p) peaks were observed at 71.5 eV and 932.9 eV, both the peaks shifted toward higher binding energies as compared to pure Pt (71.2 eV) and pure Cu (932.7 eV). The shift in peaks along with the Pt (4f) peak broadening is attributed to Cu-Pt metal interaction and the formation of CuPt nanoparticles on carbon framework [32,33]. The C1s is aligned at 284.42 eV to compensate for the sample surface charging. Similarly oxygen spectrum deconvoluted into three peaks positioned at 530.15, 531.81, and 532.41 eV represents M-O, C=O, and C-O respectively [34].

Electrode Preparation and Electrochemical Evaluation
For evaluation of electrochemical performance, catalyst ink was prepared by dissolving 2 mg of catalyst (Cu/NC, CuPt/NC, and commercial Pt/C catalyst 46.1 wt % Tanaka Kikinzoku Kogyo,  For the electrochemical testing of CuPt/NC, first cyclic voltammetry (CV) was performed in 0.1 M HClO 4 with nitrogen saturation. No significant oxygen reduction peak was observed. With oxygen saturation, a distinct peak at a potential of 0.7 V is observed, which corresponds to oxygen reduction reaction (Figure 7a).  Further ORR activity of CuPt/NC is evaluated by linear sweep voltammetry (LSV) measurements at different rotations with RDE, which gives limiting current of 4 mA cm −2 at 1600 rpm for CuPt/NC (Figure 7b). ORR activity of CuPt/NC was compared with commercial Pt/C. CuPt/NC has the highest onset potential of 0.9 V while the value is 0.85 V for the commercial Pt/C catalyst. The Cu/NC has a very low onset potential of 0.26 V. So CuPt/NC showed 50 mV more positive onset potential than the commercial Pt/C catalyst (Figure 7c). The mechanism of the ORR has been confirmed through the Koutecky−Levich (K-L) analysis (Figure 7d). The K-L plots show good linearity, which is assigned to the first order kinetics of ORR, and both catalysts are found to follow similar mechanistic pathways. The number of electrons (n) transferred during the reaction has been calculated from the slope of the K-L plot by following equation: In Equation (2), F is Faraday constant (F = 96,500), n shows number of electrons transferred per O 2 molecules, Do 2 is the diffusion coefficient of oxygen molecule in electrolyte, V is kinetic viscosity, and Co 2 shows bulk quantity/concentration of oxygen and when rotation speed is expressed in rpm then constant 0.2 is adopted [11], transferred number of electrons lies between 3.93 and 3.97 for the voltage range of 0.3-0.6 V [35]. Charge transfer number for commercial Pt/C is 4.00, as it follows 4e pathway for ORR and the n calculated for CuPt/NC is close to 4.00 therefore it also follows 4e pathway [14,36,37].
The mechanism of ORR and quality of the electrocatalyst can be obtained from Tafel analysis. Tafel slope provides an indication of the actual mechanism taking place on the electrode surface and is related to the state of the adsorbed oxygen species and its coverage variation with respect to the potential. The over potential is calculated by the formula (E − E o ) [22]. Furthermore, to identify the kinetic resistance of the reaction, charge transfer coefficients (α) were calculated for the catalysts by following Equation (3): dη The Tafel slopes of Pt/C and CuPt/NC are found to be 190 mV dec −1 and 213 mV dec −1 (Figure 7e). Tafel slopes > 118 mV/dec attributed to either (i) a preceding chemical dissociation or (ii) a consequent chemical combination as well as (iii) an electrode transfer through an oxide layer. In order to avoid misperception, the values chosen up from the Tafel slopes will be cited as "cathodic quantities" in the subsequent sections, as these will directly be used for the investigation of the ORR mechanisms [38]. The results are in accordance with literature, as the low potential region describes first C-H bond breaking in oxygen reduction reaction and it is the rate-determining step because of first electron transfer. Though, at high potential region, the slope values increase because of less exposure of poisonous intermediate species [14,39,40].
Electrochemical impedance spectroscopy (EIS) is also used to determine the electrochemical activity of modified electrode. Charge transfer resistance is measured from Nyquist plots, the lower slope of the plot represents the lower charge transfer resistance. Figure 7f shows that the Nyquist plots of CuPt/NC exhibit lower slope than Pt/C which interprets that the charge transfer resistance for the CuPt/NC is less than that of the commercial Pt/C [41,42]. The overall ORR performance comparison of CuPt/NC and commercial Pt/C along with kinetics parameters is shown in Table 2.  Figure 8 shows the proposed mechanism of oxygen reduction reaction in acidic solution i.e., (HClO 4 ). Molecular oxygen (O 2 ) in solution phase adsorbs first on the surface of electrode. Electrochemical reduction starts from the adsorbed state with K 1 constant either directly to H 2 O or to adsorbed H 2 O 2 as an intermediate that could additionally be reduced to water, by the constant k 3 , where reduction still continues via direct 4e pathway, or desorption of peroxide as the reaction final product, disband into the solution ensuing 2e pathway. In addition, it includes chemical decomposition of two molecules of hydrogen peroxide to water and oxygen.

Catalyst Characterization
SEM images were taken on VEGA3 TESCON at 20 KV at different magnifications. Elemental composition of prepared samples was determined by EDS. Siemens D5000 Powder X-ray diffractometer was used to identify the crystal structure. XRD data were collected between 2θ values of 10°-70° for CuPt/NC and Cu Kα was used as an X-ray source. The FTIR spectrum was acquired by using Agilent Cary 630 spectrophotometer. Transmission electron microscopy (TEM) was performed on CM200-FEG (Philips) at 200 keV. The XPS was performed on VG 220i-XL.

Synthesis of Cu-tpa MOF
In a typical synthesis, 5 g of CuNO3.3H2O was dissolved in 40 mL of deionized (DI) water. Total of 2 g of terephthalic acid was dissolved in a mixture of 40 mL N, N-dimethyl formamide (DMF) and 40 mL of ethanol. Both solutions were mixed by stirring at 80 °C for 24 h on a hot plate. The blue powder obtained was centrifuged and subsequently washed with DMF and DI water. The collected powder was dried at 80 °C in a vacuum oven.

Synthesis of CuPt/NC
For the synthesis of CuPt/NC 100 mg of Cu-tpa, MOF was mixed with 0.1 mL of (96.5 mM) H2PtCl6.H2O (5 wt % solution) and the mixture was dried at 80 °C for 2 h (Figure 9). The dried powder was heated to 350 °C for 1.5 h by placing in ceramic boat in a tube furnace and then heated to 700 °C with a temperature ramp of 2 °C min −1 . It was then kept at 700 °C for 3.5 h under the constant flow of Ar-H2 (90% Ar) mixture in the tube furnace. After cooling down, the black pyrolyzed powder was acid washed with 0.5 M H2SO4 and centrifuged. It was then subjected to multiple washings with DI water and subsequently dried at 80 °C in vacuum oven. Pt loading was calculated on the basis of

Materials
All chemicals and reagents were purchased from Sigma-Aldrich without any additional purification. Chemicals used were copper nitrate trihydrate salt (CuNO 3 ·3H 2 O), terephthalic acid (C 8 H 6 O 4 ), commercial grade chloroplatinic acid (H 2 PtCl 6 ), N, N-dimethylformamide (DMF), ethanol (C 2 H 5 OH), perchloric acid (HCLO 4 ), and 5 wt % Nafion were purchased from Sigma-Aldrich. For preparing aqueous solutions deionized (DI) water was used and distilled water was utilized for washing purpose during catalyst synthesis.

Catalyst Characterization
SEM images were taken on VEGA3 TESCON at 20 KV at different magnifications. Elemental composition of prepared samples was determined by EDS. Siemens D5000 Powder X-ray diffractometer was used to identify the crystal structure. XRD data were collected between 2θ values of 10 • -70 • for CuPt/NC and Cu Kα was used as an X-ray source. The FTIR spectrum was acquired by using Agilent Cary 630 spectrophotometer. Transmission electron microscopy (TEM) was performed on CM200-FEG (Philips) at 200 keV. The XPS was performed on VG 220i-XL.

Synthesis of Cu-tpa MOF
In a typical synthesis, 5 g of CuNO 3 ·3H 2 O was dissolved in 40 mL of deionized (DI) water. Total of 2 g of terephthalic acid was dissolved in a mixture of 40 mL N, N-dimethyl formamide (DMF) and 40 mL of ethanol. Both solutions were mixed by stirring at 80 • C for 24 h on a hot plate. The blue powder obtained was centrifuged and subsequently washed with DMF and DI water. The collected powder was dried at 80 • C in a vacuum oven.

Synthesis of CuPt/NC
For the synthesis of CuPt/NC 100 mg of Cu-tpa, MOF was mixed with 0.1 mL of (96.5 mM) H 2 PtCl 6 ·H 2 O (5 wt % solution) and the mixture was dried at 80 • C for 2 h (Figure 9). The dried powder was heated to 350 • C for 1.5 h by placing in ceramic boat in a tube furnace and then heated to 700 • C with a temperature ramp of 2 • C min −1 . It was then kept at 700 • C for 3.5 h under the constant flow of Ar-H 2 (90% Ar) mixture in the tube furnace. After cooling down, the black pyrolyzed powder was acid washed with 0.5 M H 2 SO 4 and centrifuged. It was then subjected to multiple washings with DI water and subsequently dried at 80 • C in vacuum oven. Pt loading was calculated on the basis of pyrolyzed sample weight and H 2 PtCl 6 loading. Cu/NC was synthesized by the same method without Pt loading.
Catalysts 2020, 10, x FOR PEER REVIEW 11 of 13 pyrolyzed sample weight and H2PtCl6 loading. Cu/NC was synthesized by the same method without Pt loading.

Conclusions
This work demonstrated a highly active and efficient CuPt/NC ORR catalyst derived from Cutpa MOF. Characterization through SEM indicates the Pt loading on nanoporous carbon. TEM images display uniformity of CuPt nanoparticles. Shift in binding energy of Cu2p and Pt4f in XPS is interpreted as Cu-Pt interaction. The onset potential of CuPt/NC is 0.9 V (vs. RHE) and the value of limiting current density is 4 mA/cm 2 . Furthermore, the linearity of K-L plots and the charge transfer number value of 3.93-3.97 indicated single step kinetics. Tafel slope also shows that CuPt/NC and commercial Pt/C follows the same reaction kinetics. From EIS studies it was concluded that CuPt/NC has lower charge transfer resistance than Pt/C. Hence the results emphasize the excellent activity of CuPt/NC as a potential ORR catalyst.