The Rietveld refinements of XRD patterns of pristine MgCoO2 and 1–5%PANI@MgCoO2 nanocomposites are shown in Fig. 1. The diffraction pattern of MgCoO2 displayed sharp Bragg reflexions (111), (200), (220) (311) and (222) that are indexed to a cubic structure with a well-established Fm3m space group, in agreement with previous reports in the literature [21, 22]. From the XRD peak positions, we have calculated the lattice parameters as a = b = c = 4.2369 (1) Å, and α = β = γ = 90º. Fig. S1 shows the XRD patterns of all the nanocomposites in the 10–80º range, and PANI shows four broad reflections between 10–30º owing to the periodicity parallel and perpendicular to the polymer chains [28, 29]. After PANI coating, the 1–5%PANI@MgCoO2 nanocomposites materials are indexed also using the cubic Fm3m space group. The refined lattices and coherent domain sizes are given Fig. 1. Compared to the pristine MgCoO2, the parameters ‘a’ and ‘V’ of PANI@MgCoO2 nanocomposites remained invariant without affecting the crystal structure.
SEM micrographs revealed that synthesized MgCoO2 exhibits flat and cube-shaped irregular particles with an average diameter of 0.5–0.6 µm. The micrographs of 1–5%PANI@MgCoO2 nanocomposites revealed the presence of small grains of 80–100 nm and rough texture demonstrating that the small PANI grains are embedded on the MgCoO2 particle surface. It is evidenced the structural flexibility of PANI (from 1%, 3%, 4–5%) on coating these sub-micrometric MgCoO2 particles. In addition, the PANI is detected by a progressive appearance of nitrogen atom as recorded from the mapping together with the homogeneously dispersed magnesium, cobalt, and oxygen (Fig. S2). The effect of coating with PANI can facilitate the ion transportation and enhance electron transfer and influence the electrochemical performance [30]. During the synthesis, emeraldine base is transformed into emeraldine salt as shown in Fig. 2F with benzenoid and quinoid rings in the PANI chains. The dependence of physicochemical properties on the synthesis conditions are among the most important PANI properties.
Raman spectra of PANI, MgCoO2 and PANI@MgCoO2 nanocomposites are plotted in Fig. 3. For MgCoO2, three Raman-active bands can be observed (A1g, Eg, and one F2g). The A1g peak at 687 cm− 1 represents the metal-O bond vibrations of the octahedral sublattice [31], noting that it is the Raman vibration peak with the highest intensity out of the three, indicating that the cobalt cations mostly occupy the octahedral site. The F2g peak at roughly 520 cm− 1, shows the stretching vibration of the Co-O bond. The Eg mode at ~ 478 cm− 1 is caused by stretching of Mg-O and Co-O bonds in the tetrahedral sublattice [32]. For the PANI (emeraldine salt), Raman band at 1604 cm− 1 is related to the C–C stretching vibration of benzene ring. The intense band observed at 1170 cm− 1 is assigned to C–H vibrations of aromatic rings [33]. The bands at 874, 812 and 414 cm− 1 correspond to out-of-plane deformations of the rings protonated emeraldine form of PANI [34]. The Raman spectra of PANI@MgCoO2 (1, 3, 4 and 5 wt. %) nanocomposites are quite similar with a higher contribution of carbon bands as PANI content increases. No additional peaks are observed in the spectra, indicating that no chemical reactions between PANI and MgCoO2 occurred.
To verify the advantages of the PANI@MgCoO2 nanocomposite as cathode material for RMBs, the electrochemical discharge/charge behavior and cyclability are investigated (Figure 4A-D) using twofold ion approach (Na+/Mg2+). Previously, the limitations to perform the Mg-cell with pure Mg(TFSI)2+DME electrolyte are confirmed. The galvanostatic curves of pristine MgCoO2, 3%PANI@MgCoO2 and 5%PANI@MgCoO2 during the initial ten cycles at 5 mA g− 1 are selected. During the first discharge a common characteristic in the reactions is observed: i) two plateau at 2.1 and 1V (labeled as 1 and 2 respectively), and ii) a constant decay of the potential in 1–0.05 V (labeled as 3). On the subsequent charge the curves show an extended plateau (between 4-4.5V) and the cell exhibit high polarization, and all the samples recovered around 99% of the capacity. Regarding the capacity, on increasing the PANI content (1, 3, 4 and 5%) the first discharge capacity increaseded from 147.9 (for pristine MgCoO2) to 154.2, 172, 190.1, and 207.8 mA h g− 1, respectively (Fig. 4D). These capacities are lower than the theoretical capacity of MgCoO2. On cycling, the first reversible capacities are respectively 99.3, 103.4, 125, 130.2 and 153 mA h g− 1 which are retained 63% (62.8 mA h g− 1), 77% (80.3 mA h g− 1), 77% (95.7 mA h g− 1), 51% (66.2 mA h g− 1) and 48% (73.2 mA h g− 1) over 40 cycles. Most probably, polyaniline could not tolerate high-voltage oxidation, and the capacity decays constantly while cycling. However, the best synergistic effect is found for 1%PANI@MgCoO2 and 3%PANI@MgCoO2 with 65 and 68% of capacity retention up to 60 cycles, respectively.
According to the ex-situ XRD patterns the cubic phase persist after the complete discharge and discharge/charge and consequently a conversion reaction can be discarded in this system (Fig. 5). The peaks at 36.68, 42.64, 61.91, and 74.23° are assigned to the (111), (200), (220) (311) lattice planes of cubic MgCoO2 (Fig. S3) The aluminium current collector peaks are detected at 38.43, 44.68, 65.05 and 78.17º. The cell parameters at the end of the first discharge and discharge/charge do not show obvious variation for pristine MgCoO2, 3%PANI@MgCoO2 (65%) and 5%PANI@MgCoO2 nanocomposites (4.2377(2)/4.2373(1) Å, 4.2379(2)/ 4.2378(1) Å and 4.2374(3)/ 4.2375(1) Å) as compared to the uncycled electrodes, and the cell volume slightly varied (see table S1). Then, an insertion reaction occurs and could be exemplified for these two samples as follows:
MgCoO2 + 0.64 e− + x Na+ + y Mg2+ ↔ Mg1 + yNaxCoO2 Eq. [1]
3%PANI@MgCoO2 + 0.9 e− + x Na+ + y Mg2+ ↔ 3%PANI@Mg1 + yNaxCoO2 Eq. [2]
To quantify the amount of reacted sodium and/or magnesium, the galvanostatic profile in the Na-cell is performed for reference, which is reflecting that the total discharge capacity is 95 mA h g− 1 with an extension of the reaction at points 1, 2 and 3 of 45, 15 and 35 mA h g− 1, respectively (Fig. S4). That behavior differs in terms of capacity to those observed in Mg-cell. Therefore, at points 1, 2 and 3 (Fig. 4A-C), a sequential insertion of Na/Mg into the cubic structure is occuring through a single-phase mechanism. This outcome differs to previous results where the insertion reaction into 3d transition metal oxides implies a two phase reaction mechanism [35–38]. Also, the insertion of lithium into LiCoO2 cathode (2–4.5 V) is accompained by several phase transitions (hexagonal, monoclinic) leading to internal stress generation [23]. The ICP results confirmed the Mg:Na:Co atomic ratio for pristine electrode is 0.998:0:1.005 and after discharge in Na- and Mg-cells is 0.995:0.397:1.006 and 1.112:0.405:1.015, respectively (Table 1). Then, the extra capacity is due to the participation of Mg2+ ion in the insertion reaction. For 3%PANI@MgCoO2 the atomic ratio in the discharged Mg-cell is 1.154:0.415:0.099 and after the subsequent charge allowed extraction of Mg and Na, as also found in pristine electrode. These results suggest reversible insertion/de-insertion into/from cubic MgCoO2, whose findings show a different mechanism of reaction as compared to the reported conversion reactions (recorded between 0.01–3 V) in LiCoO2 and Co3O4 when cycled versus metallic Li and Na anodes[39–41]. Aditionally, it should be noted the differences of the cubic MgCoO2 compared to the oxyspinel MgM2O4 (M = Mn, Co, Ni) reacting with Mg through a biphasic reaction in which the main problem of Mn3+/Mn+ 4 redox couple is the transformation from cubic to tetragonal spinel diminishing the ionic conductivity due to the smaller ion diffusion chanels in tetrgonal phase [42–44]. Therefore, cubic MgCoO2 exhibit analogous structural characteristics as those in the cubic Mg2MnO4 [6] and MgxNiyCozO2 where the Co and Mg compositions were considered to facilitate the insertion/deinsertion of Mg2+ [45] .
Table 1
ICP results showing the Mg:Na:Co atomic ratio of pristine MgCoO2 and 3%PANI@MgCoO2 after first discharge and discharge/charge (at 5 mA g− 1) in Mg-cells with NaClO4-PC:FEC electrolyte. For the sake of comparison, pristine MgCoO2 is measured at the end of the discharge and discharge/charge in the sodium cell.
Sample
|
Capacity (mA h g− 1)
|
Mg : Na : Co
(ICP)
|
Remarks
|
---|
MgCoO2
|
Pristine
|
--
|
0.998 : 0 : 1.005
|
Pristine
|
MgCoO2
|
Disch.
|
95
|
0.995 : 0.397 : 1.006
|
Na-cell
|
Disch/Ch.
|
95 / 46
|
1.008 : 0.175 : 0.097
|
Na-cell
|
MgCoO2
|
Disch.
|
147.9
|
1.112 : 0.405 : 1.015
|
Mg-cell
|
Disch/Ch.
|
147.9 / 149
|
1.013 : 0.021 : 1.007
|
Mg-cell
|
3%PANI@MgCoO2
|
Disch.
|
172
|
1.154 : 0.415 : 0.099
|
Mg-cell
|
Disch/Ch.
|
172 / 171.9
|
1.017 : 0.011 : 1.012
|
Mg-cell
|
Some properties of the electrode/electrolyte interphase and kinetic response are studied from the impedance spectra for pristine, 3% and 5%PANI@MgCoO2 after 30 cycles. Nyquist plots exhibit two semicircles revealing capacitive phenomena occurring at the surface layer (SL) and charge-transfer (CT) reactions at high and medium frequencies, respectively. The resistances decreased at the end of the discharge at 0.01 V for both coated samples (RSL3%= 0.46 Ω·g RCT3%= 0.92 Ω·g; RSL5%= 0.59 Ω·g, RCT5%= 1.25 Ω·g) as compared to the pristine sample (RSL= 0.61 Ω·g, RCT= 1.65 Ω·g). Additionally, the diffusion coefficients (D+) are calculated at the quasi-equilibrium state using the following equation [46]:
D = \(\frac{1}{2}{\left(\frac{R T}{S {F}^{2}{\sigma }_{w} C}\right)}^{2}\) Eq. [3]
where R is the gas constant (J·K− 1·mol− 1), T is the absolute temperature, S is the area of the electrode surface, F is Faraday’s constant, C is the molar concentration of Mg2+/Na+ ions and σw is the Warburg coefficient. The latter one is determined from the slope of Z’ vs. the reciprocal root square of the lower angular frequencies (ω−1/2) in agreement with the following equation:
\({Z}^{{\prime }}={R}_{SL}+{R}_{CT}+({\sigma }_{W}· {\omega }^{-1/2}\)) Eq. [4]
Benefiting from synergistic effect of both PANI and MgCoO2, the diffusion coefficient adopted these values: 6.92·10− 14 (3%PANI@MgCoO2) > 4.18·10− 14 (5% PANI@MgCoO2) > 3.05·10− 14 (pristine MgCoO2) cm2 s− 1. These results are associated with an improvement of the kinetics of the charge transfer reaction on PANI coated samples.