Vat Orange 11—Based Organic Cathode Material for High Rate Rechargeable Magnesium Battery

Materials and electrolyte preparation

Characterization

Electrochemical measurements

The structure of vat orange 11 is confirmed by solid-state ¹³C NMR (Fig. S1 is available online at stacks.iop.org/JES/167/070561/mmedia) as it is not soluble in any solvent and the peaks observed correspond to carbonyl carbons, and carbons due to the aromatic rings. The chemical shift value for carbonyl (–C=O) carbon is ∼183.72 ppm. The chemical shift values for carbons of aromatic rings lies between 110 and 140 ppm. The chemical shift values observed in the ranges, 20−60 ppm and 200−240 ppm are due to spinning bands. The FTIR spectrum (Fig. 1a) reveals peaks around 1667 and 1655 cm⁻¹ that correspond to the carbonyl (–C=O) double bond stretching vibrations. A peak around 1565 cm⁻¹ is due to the –C=C double bond stretching vibrations of the aromatic rings. A small broad peak at 3080 cm⁻¹ corresponds to –N–H stretching vibration. The –C–H bond vibrations are observed between 850 and 750 cm⁻¹. The Raman spectrum as shown in Fig. 1b confirms the structure. The highest intense peak around 1655 cm⁻¹ corresponds to ν (C=O) and the peak around 3070 cm⁻¹ corresponds to ν (N–H) vibrations. The peaks around 1183 and 1433 cm⁻¹ are due to aromatic ring –C=C vibrations. Sharp peaks in the X-ray diffraction (XRD) pattern (Fig. 1c) reveal that the as-extracted, solid material is highly crystalline in nature. The diffraction peaks around 12° and 28° 2θ values are likely to be due to π-π stacking of the aromatic moieties. The morphology (Fig. 1d) shows a granular-like structure with sizes ranging from hundreds of nanometers to a few hundreds of micrometers.

Zoom In Zoom Out Reset image size

Activated carbon shows a capacitive behaviour coupled with good stability, conductivity, wide pore size distribution and it is a preferred current collector material for energy storage devices. It possesses low energy density and high-power density that can take care of slow ion transport and provide good rate performance.³¹ This has been realized in Li-ion battery with LiFePO₄ as the active material. Balducci et al. observed improved rate capability and conductivity along with a good performance at high current rates when activated carbon is used in place of super P carbon.³² So, activated carbon can be used to improve the rate capability of rechargeable Mg-ion batteries. In the present studies, both acetylene black and activated carbon were used as conducting carbon for the electrode preparation.

Electrochemical studies have been carried out in coin cell configuration with Mg foil anode in combination with vat orange 11 in acetylene black or activated carbon separately as cathodes and APC as the electrolyte. Figure S2 shows the cyclic voltammograms of vat orange 11 electrode with acetylene black in the potential range of 1.2−2.4 V for the first five cycles at 0.1 mV s⁻¹ scan rate. In the cathodic scan, the first cycle shows a board peak around 1.4 V which vanishes in subsequent cycles and this peak is possibly related to activation of the electrode material. A broad oxidation peak around 2.05 V along with a small peak at 2.2 V are observed in the anodic scan. In the second cycle, additional small reduction peaks at potentials <2.0 V are observed while the oxidation peak remains almost at the same potential but with lower peak current. The peaks remain unchanged in subsequent cycles. The cyclic voltammograms are very similar to that reported for PAQ¹⁷ wherein redox peaks due to quinone reduction/oxidation are present. Similar, cathodic and anodic peaks are observed when the voltammogram is recorded at different scan rates, from 0.1−1.0 mV s⁻¹ as shown in Figs. 2a and 2b for the electrodes acetylene black and activated carbon respectively. Increase in the peak currents with a slight change in the peak position as a function of scan rate indicates a fairly reversible redox process. The logarithmic plots of peak current vs. scan rates (log i vs. log ν) as shown in Fig. S3 give the slope, b. The slope of log i vs. log ν plot reflects whether the electrochemical process is diffusion-controlled or capacitive in nature. The process is diffusion-controlled when b = 0.5 and capacitive when b = 1. For the electrode with acetylene black, the b values are 0.57 (peak 1) and 0.93 (peak 2) while 0.76 (peak 1) and 1.1 (peak 2) with activated carbon. It indicates a slightly high capacitive contribution of activated carbon as compared to that with acetylene black.

Zoom In Zoom Out Reset image size

The charge-discharge characteristics reveal that the vat orange 11 electrodes can deliver a reversible capacity of ∼35 and 80 mAh g⁻¹ with acetylene black and activated carbon respectively at the current density of 100 mA g⁻¹ for 100 cycles as shown in Figs. 2c and 2d. The capacities offered by acetylene black and activated carbon alone are very low and shown in Fig. S4. The galvanostatic charge-discharge profiles of the cells with acetylene black and activated carbon are shown in Figs. 2e an 2f. A flat plateau observed for the first discharge analogous to the very first cycle in the voltammogram. After the first cycle, a decrease in polarization is observed with the plateau region becoming short. Initial discharge capacities of 172 and 260 mAh g⁻¹ are observed with acetylene black and activated carbon respectively with voltage plateau around 1.5 V. This high capacity in the first cycle may be due to some compositional changes in the material that can be observed from the loss of crystallinity and changes in the IR spectrum as well as red coloration of the separator (Fig. S5) after charge-discharge at 100 mA g⁻¹ current density. The capacities are calculated based on the mass of vat orange 11 used in the electrode. The discharge capacity decreases in the intial cycles and subsequently remains constant. With activated carbon, the capacity stabilizes at ∼80 mAh g⁻¹ for over 100 cycles which is similar to the theoretical capacity (∼83 mAh g⁻¹ considering two-electron process) for vat orange 11. Figures 2g and 2h show the rate performance of the vat orange 11 electrodes at various current densities. The reversible capacity decreases with increasing discharge current density and capacities of 55, 42, 20 and 10 mAh g⁻¹ at current densities of 100, 200, 500 and 1000 mA g⁻¹ respectively are observed for the electrode with acetylene black. However, electrode with activated carbon can deliver a capacity of ∼26 mAh g⁻¹ even at a very high current density of 2000 mA g⁻¹. Hence, the rate performance as well as the capacity values at high current densities are improved for activated carbon electrode as compared to that of acetylene black electrode. The long-term cycling stability test of the electrode was carried out at 500 mA g⁻¹ current density and is shown in Fig. 3. It is stable up to 1000 cycles with ∼98% of Coulombic efficiency and 67% of the capacity retention at the end of 1000 cycles.

Zoom In Zoom Out Reset image size

Salt-controlled dissolution process is employed in which LiCl is used along with APC electrolyte to inhibit the dissolution of vat orange 11. Similar studies have been carried out by Mai et al. using PTCDA as cathode for rechargeable Mg-ion battery where the addition of salts like LiCl, NaCl and KCl to APC electrolyte supressed the dissolution of the organic cathode and high rate capability up to 500 mA g⁻¹ with good cycling stability of 200 cycles at 200 mA g⁻¹ current density are observed.²⁷ The dissolution experiments are carried out in APC and APC-LiCl electrolytes at different time intervals. The optical images (Fig. S6) show a decrease in the dissolution of vat orange 11 in the APC-LiCl electrolyte as compared to that observed in pure APC electrolyte. Further, the dissolution inhibition is confirmed from the FTIR spectra of the electrolytes. As shown in Fig. S7, there is no noticeable change in the transmittance of APC-LiCl electrolyte unlike APC electrolyte after 24 h. of soaking. In the electrolyte without LiCl, the change in intensities of the peaks present at ∼1346, 1420, 1462, 1480 and 1823 cm⁻¹ reflect the dissolution of vat orange 11. The presence of colour in the separator after charge-discharge process is faint (as shown in Fig. S9) in the electrolyte with LiCl as compared to the electrolyte without LiCl. Hence, it is confirmed that the dissolution of vat orange 11 in the electrolyte is relatively suppressed. It is likely that the dissolution of the lithiated dye is decreased when the concentration of the Li⁺ ions in the electrolyte is increased.

Based on the above observations, electrochemical investigations are carried out using 1 M LiCl in the APC electrolyte. Figure 4a shows the voltammograms obtained at various scan rates from 0.1 to 1.0 mV s⁻¹ in the potential range of 1.0−2.5 V vs Mg/Mg²⁺. A set of redox peaks is observed like in the earlier studies but with increased peak currents possibly due to increased conductivity of the electrolyte. The "b" values obtained from the logarithmic plots of peak current vs. scan rate (Fig. S8) indicate a mixed process. The rate performance of vat orange 11/activated carbon electrode as shown in Fig. 4b reveals significant improvement and the capacity vales are also enhanced as compared to the values observed in pure APC electrolyte. An initial discharge capacity of 246 mAh g⁻¹ is observed at the current density of 100 mA g⁻¹ and a stable capacity of ∼80 mAh g⁻¹ is obtained. The long-term cycling stability is carried out at a current density of 500 mA g⁻¹. As shown in Fig. 4c, it is nearly stable up to 1000 cycles with ∼98% of Coulombic efficiency. After the initial decay, the capacity remains almost constant. At the end of 1000 cycles, nearly 95% capacity retention is observed. Still the red coloration of separator (Fig. S9) is observed after charge-discharge cycles (100 cycles at 100 mA g⁻¹ current density) reveals possibilities of some chemical changes during charge-discharge cycles. This is possibly due to the supressed dissolution of the vat orange 11 in the APC-LiCl electrolyte which is similar to the observations made by Mai et al. for PTCDA electrode in APC electrolyte with different salts as mentioned earlier.²⁷

Zoom In Zoom Out Reset image size

The Nyquist plots before and after charge-discharge experiments of vat orange 11/activated carbon electrode in APC-LiCl electrolyte are shown in Fig. 5 and the corresponding equivalent circuit diagram as well as the expanded plots in the high frequency range are shown in the inset of Fig. 5. Similar impedance behaviour has been reported for Li-ion batteries using various vat dyes²⁸ and other inorganic systems as well.³³ Various electrical elements used in the equivalent circuits are, internal resistance (R_in), constant phase element (Q) in parallel with the resistance due to surface passivation (R_sp) and the double layer capacitance (C_dl) in parallel with the Warburg diffusion element (W) due to Mg²⁺ ion diffusion and the charge transfer resistance (R_ct). Initially, the arc of semicircle is large indicating a very low charge transfer rate. After the charge-discharge cycles, the overall cell impedance as well as the arc of the semicircle is reduced that reflects facile Mg²⁺ ion uptake. The R_ct value is decreased from 471 ohm before charge-discharge experiment to 45.5 ohm after the charge-discharge cycles. This indicates that the decreased interfacial resistance after the charge-discharge cycle possibly leads to enhanced cycling and good rate performance. The diffusion constant values are calculated from the EIS data to be 2.528*10⁻¹⁴ cm² s⁻¹ and 4.97*10⁻¹⁴ cm² s⁻¹ before and after the charge-discharge experiments.

Zoom In Zoom Out Reset image size

Although the dissolution problem can be avoided to certain extent, the resulting battery cannot be considered as a pure Mg-ion battery. Hence, a non-nucleophilic electrolyte based on ionic liquid (MMTP₁₃TFSI) is used to study the performance of vat orange 11/acetylene black electrode. Figure 6a shows the galvanostatic charge-discharge profiles of vat orange 11/acetylene black at 100 mA g⁻¹ current density. Unlike APC-based electrolytes, well defined plateaus are observed for both discharge and charge processes in this electrolyte. Discharge and charge capacities of ∼81 and 83 mAh g⁻¹ respectively, are observed for the first cycle and the polarization is found to be very high. After the first cycle, decrease in polarization is observed. Even after 100 cycles, the plateaus are observed clearly, and the curves overlap indicating a highly reversible process. The cycling performance of the electrode at a current density of 100 mA g⁻¹ is shown in Fig. 6b. Initially for few cycles, an activation-type process is observed which could be due to the magnesiation and subsequent oxidation after which the capacity decrease is fairly small and yields a value of ∼80 mAh g⁻¹.

Zoom In Zoom Out Reset image size

The electrode is cycled at various current densities to evaluate the rate performance as shown in Fig. 6c. The MMTP₁₃TFSI electrolyte shows higher capacity and stable performance as compared to APC-based electrolytes. This indicates that although we can limit the dissolution of the electrode material by using LiCl, it is not possible to avoid the nucleophilic nature of APC affecting the performance of the material. There is a change in the composition of the material followed by a decrease in performance.

In the MMTP₁₃TFSI electrolyte, the electrode can deliver a capacity of ∼15 mAh g⁻¹ even at a high current density of 1000 mA g⁻¹ and stable capacity values are observed at all current densities. Further, the capacity value is regained when the cell is cycled back to 100 mA g⁻¹ current density after cycling at high current values, indicating a highly reversible process. The cell is cycled at 500 mA g⁻¹ current density to obtain information on the cycling stability (Fig. 6d). A stable capacity of ∼25 mAh g⁻¹ is obtained upto 500 cycles. Close to the end of 500 cycles, the cell retains ∼63% of the initial capacity which is comparable to that of PAQS where 70% capacity retention has been reported after 100 cycles at a current density of 112 mA g⁻¹.¹⁷ The obtained results with the non-nucleophilic electrolyte is comparable to the performance of anthraquinone-based systems reported in the literature for Mg-ion battery as presented in Table I. The performance of all the organic electrodes reported till date for Mg-based rechargeable batteries is given in Table SI1.^{2,13–27,30,34,35}

Table I.  Comparison of vat orange 11 electrode performance with anthraquinone-based organic compounds for rechargeable Mg-ion battery.

Electrode	Electrolyte	Counter electrode	No. of cycles_capacity @current density	Potential window (V)	References
Poly(anthraquinoyl) sulfides	MHCC	Mg	80 cycles_60 mAh g−1 @50 mA g−1	0.5−2.5	15
	MACC	Mg powder	100 cycles_40 mAh g−1 @50 mA g−1
	MTCC		100 cycles_50 mAh g−1 @ 50 mA g−1
	Mg(HMDS)2-4MgCl2 /THF	Mg	100 cycles_30 mAh g−1 @112 mA g−1	0.5−2.5	17
			100 cycles_100 mAh g−1 @130 mA g−1
			100 cycles_100 mAh g−1 @130 mA g−1
			1000 cycles_60 mAh g−1 @520 mA g−1
	Mg(TFSI)2-MgCl2/glyme	Mg	500 cycles_100 mAh g−1 @260 mA g−1	0.5−2.5	18
	MgCl2-AlCl3/glyme		500 cycles_50 mAh g−1 @260 mA g−1
	Mg(TFSI)2/diglyme	Mg	100 cycles_150 mAh g−1 @130 mA g−1	0.8−2.0	19
	Mg(HMDS)2-MgCl2/THF and PP13TFSI	Mg	100 cycles_80 mAh g−1 @100 mA g−1	1.0−2.5	This work
			500 cycles_26 mAh g−1 @500 mA g−1

(MHCC-[Mg₂Cl₃-6THF] [HMDSAlCl₃] in THF, MACC- MgCl₂–AlCl₃ in THF, MTCC- MgCl₂–Mg(TFSI)₂ in a mixture of THF and glyme).

The mechanistic details on the uptake/release of magnesium ions during charge-discharge is investigated using ex situ FTIR spectra recorded after different charge-discharge cycles at 100 mA g⁻¹ current density. The obtained FTIR spectra are shown in Fig. 7. The pristine electrode shows bands around 1667 and 1655 cm⁻¹ and 1565 cm⁻¹ that correspond to –C=O and –C=C stretching vibrations respectively. In the case of vat 11 orange, the first discharge shows a broad peak centred at ∼1642 cm⁻¹ and the peaks due to carbonyl (–C=O) double bonds have disappeared. A new peak assigned to enolate (–C=O^–) group appears at ∼1388 cm⁻¹. The peak assigned to –C=C double bonds disappears after the first discharge indicating the involvement of –C=C double bonds of the aromatic ring in the magnesium uptake. The above results indicate that the carbonyl (–C=O) double bonds acquire partial single bond character through resonance and get converted to enolate (–C=O⁻) groups during discharge. In the subsequent charging process, the peaks due to –C=O and –C=C double bonds recover to the original positions though they are broadened as compared to the original band shapes. During second discharge, the band corresponding to –C=O double bond is found to be shifted and the band associated with –C=C double bond disappears along with the evolution of –C=O⁻ peak. These results indicate that the magnesiation and demagnesiation process follows a transformation mechanism between carbonyl and enolate groups. Similarly, the IR spectra after 10th, 20th, 50th and 100th charge-discharge cycles show that the material is intact and involves the functional group transformation. The theoretical capacity of the organic molecule is ∼83 mAh g⁻¹ based on two-electron transfer and the values are ∼166 and 249 mAh g⁻¹ for 4 and 6 electron processes respectively. Experimentally observed capacity value (∼80 mAh g⁻¹) indicates that vat 11 orange is involved in a two electron transfer process during charge-discharge process. The possible reaction sequence during discharge-charge process is given in the supporting information (Scheme 1).

Zoom In Zoom Out Reset image size

In summary, vat orange 11 is successfully used as a cathode for secondary non-aqueous Mg-ion battery. Vat orange 11 shows a moderate reversible capacity and rate performance in APC electrolyte. Excellent rate performance, high capacity, and high cycling stability are achieved by using activated carbon instead of acetylene black as conducting carbon along with vat orange 11 in APC-LiCl electrolyte. It delivers a stable capacity of ∼50 mAh g⁻¹ at a high current density of 500 mA g⁻¹ for 1000 cycles. Finally, a Mg-ion battery with non-nucleophilic electrolyte and vat orange 11/acetylene black is realized which shows a high capacity of ∼80 mAh g⁻¹ at 100 mA g⁻¹ current density. The magnesiation and demagnesiation mechanism as probed by ex situ FTIR measurements shows the involvement of functional group transformation between carbonyl (–C=O) and enolate (–C=O⁻) moieties.

We would like to thank V. R. Sai Ganesh, General Manager, Atul Limited, Thane, Maharastra, India for supplying sample of vat orange 11. Debashis Tripathy thanks IISc. for providing the fellowship. DST, New Delhi is acknowledged for research funding.