Rechargeable Lithium-Ion Battery Based on a Cathode of Copper Hexacyanoferrate

The crystal structure and phase of the material was determined by powder X-ray diffraction shown in Fig. 1a. The synthesized CuHCF powders show a high crystallinity and the diffraction peaks can be indexed in the face centered cubic structure of Cu(Fe(CN)₆)_0.667 (86–0513; Fm3m, a = b = c = 10.1 Å). This crystal structure agrees with the results obtained in the EDX analysis where the stoichiometric relationship between Fe and Cu is 0.62 (Fig. S1 (available online at stacks.iop.org/JES/168/080515/mmedia)). The reaction that summarizes the formation of CuHCF can be represented as:

$$\begin{eqnarray}&&\begin{array}{l}{{\rm{K}}}_{3}{\rm{Fe}}{\left({\rm{CN}}\right)}_{6({\rm{aq}})}+{{\rm{CuCl}}}_{2({\rm{aq}})}+{{\rm{nH}}}_{2}{{\rm{O}}}_{({\rm{l}})}\to {\rm{KCu}}\left[{\rm{Fe}}{\left({\rm{CN}}\right)}_{6}\right]\\ \,\times \,{{\rm{nH}}}_{2}{{\rm{O}}}_{({\rm{s}})}+2{{\rm{KCl}}}_{({\rm{aq}})}+{{\rm{H}}}_{2}{{\rm{O}}}_{({\rm{l}})}\end{array}\end{eqnarray} \tag{ 1 }$$

Crystallite size was estimated by Rietveld analysis using the XRD peak indexing software FULLPROF, as shown in Fig. 1b. This information was used to identify the unit cell lattice symmetry cubic belonging the $F\bar{4}3m$ space group (CIF Code 1010359), with the unit cell parameters determined to be as follows: a = b = c = 10.121(4) Å, α = β = γ = 90° and crystal size 67 nm. The R parameters were R_p = 20.0, R_wp = 29.4, R_exp = 35.5 and χ² = 1.54. The CuHCF, whose structure was proposed by Keggin and Miles, ²³ it consists of a simple cubic with chains of –Fe−C−N−Cu−N−C−Fe− along the three crystallographic directions forming an octahedron. Each unit cell consists of eight cells and therefore contains eight insert sites that allow both monovalent and polyvalent ions to be accommodated. Cell size is an important parameter to facilitate insertion of lithium-ion (59 pm), in this case (10.121 Å = 1012 pm), therefore it is possible to house these ions.

Zoom In Zoom Out Reset image size

The co-precipitation of CuHCF in the presence of excess Cu²⁺ proceeds more slowly, allowing for ordered growth of highly crystalline, polydisperse nanoparticles. On the other hand, the CuHCF nanoparticles obtained under these conditions readily precipitate into larger agglomerations, which can be easily filtered and processed for use in battery electrodes. ²⁴ Field emission scanning electron microscopy (FESEM) was used to study the morphology of the CuHCF. The FESEM images shown in Fig. 2a, indicate a particle size ranged between 40 and 70 nm corresponding to usual morphologies of these agglomerates, and consistent with those determined by XRD. Using transmission electron microscopy (TEM, Fig. 2b), it can be seen that the particles are agglomerated and with a cubic morphology that cannot be appreciated by FESEM.

Zoom In Zoom Out Reset image size

Analogs of PB compounds generally have water absorbed in their structure, which can be found in two forms: a) as zeolitic H₂O in the cavities formed by the M'−CN−M network, and b) as coordinated H₂O replacing vacancies of the octahedron cavities formed by Fe(CN)₆ ³⁻. ^25,26 According to the XRD and EDX analysis, the synthesized CuHCF has a vacancy fraction of 0.62. The TGA in Fig. 3 shows that weight loss occurs in two steps.

Zoom In Zoom Out Reset image size

According to Fig. 3, heating from room temperature to about 180 °C, produces 30% of mass loss corresponding to: i) adsorbed humidity, zeolitic water and coordinated water in the structure, i.e., equivalent to 6 molecules per unit of formula. Above 180 °C, a gradual decomposition of the CN group into (CN)₂ occurs, obtaining Cu and Fe as final products. ²⁷ Therefore, the molecular formula of the synthesized compound can be considered to have now six more water molecules, KCu[Fe(CN)₆]_0.667·6H₂O.

The electrochemical performance of the cathode was evaluated using CV and galvanostatic curves (GC) at room temperature, under the experimental conditions described in the experimental section. Figure 4 shows the voltammetric profile of a lithium half-cell with a cathode based on CuHCF, at a scan rate of 0.1 mV s⁻¹ and at a potential ranged of 2.5−4.3 V vs Li⁺/Li. The CV shows a quasi-reversible insertion step (3.6 V) and extraction (3.8 V) of lithium ions, resulting from a separation of peak potentials about 200 mV and a relationship between the oxidation/reduction areas close to 1 These redox reactions correspond to the iron (Fe²⁺/Fe³⁺) carbon-coordinated in the CuHCF structure. This type of compound has only one active center because copper nitrogen-coordinated has such a low solubility product that a reduction of this metal ion shifts to an inaccessible value. ¹⁶ However, a little shoulder observed at 2.6 V vs Li⁺/Li could be attributed to the reduction of copper ions into CuHCF structure. Similar results have been observed by O. Makowski et al. ⁹ in the electrochemical insertion of K⁺ ions into a CuHCF film deposited on a glassy carbon electrode.

Zoom In Zoom Out Reset image size

The lithium-ion insertion process during iron reduction can be represented by Eq. 2:

$$\begin{eqnarray}&&\begin{array}{l}{{\rm{KCu}}}^{\left({\rm{II}}\right)}\left[{{\rm{Fe}}}^{\left({\rm{III}}\right)}{\left({\rm{CN}}\right)}_{6}\right]\cdot 6{{\rm{H}}}_{2}{\rm{O}}+{{\rm{xLi}}}^{+}+{{\rm{xe}}}^{-}\\ \,\leftrightarrow \,{{\rm{Li}}}_{{\rm{x}}}{{\rm{KCu}}}^{\left({\rm{II}}\right)}{\left[{{\rm{Fe}}}^{{\rm{II}}}{\left({\rm{CN}}\right)}_{6}\right]}_{{\rm{x}}}{\left[{{\rm{Fe}}}^{{\rm{III}}}{\left({\rm{CN}}\right)}_{6}\right]}_{\left(1-{\rm{x}}\right)}\cdot 6{{\rm{H}}}_{2}{\rm{O}}\end{array}\end{eqnarray} \tag{ 2 }$$

The GC allow obtaining the reversible specific capacity (C) of both discharge and charge processes. However, despite having the results of both, the subsequent analysis and discussions are made based only in the discharge process. The variation of "C" will allow correlating electrochemistry with structure and morphology. This correlation is possible through the C-rate analysis, reversibility, cyclability, coulombic efficiency, retention, and remaining capacity percentage.

In order to obtain the optimal C-rate condition, the CuHCF material was subjected to ten cycles at different C-rates. Figure 5 shows an increase in storage capacity from 40 mAh g⁻¹ to 60 mAh g⁻¹ when decreasing charging rates from 1 C to C/20, respectively. The average specific capacity and retention capacity for CuHCF obtained in this study are summarized in Table SI. Also, a high discharge/charge reversibility could be observed in all the cases. Comparing the average specific discharge capacities in the initial 1 C state (39 mAh g⁻¹) with the final 1C state (42 mAh g⁻¹), the material is observed to have high mechanical stability.

Zoom In Zoom Out Reset image size

The results summarized in Table SI show an increase in the average specific capacity at longer discharge/charge times, reaching an additional retention of 65.4% at C/20. Therefore, a C/20 value has been chosen to carry out more in-depth cyclability studies. Figure 6 represents the galvanostatic charge/discharge curves as a function of lithium-ion insertion and the variation of the specific capacity and the coulombic efficiency obtained during 300 cycles.

Zoom In Zoom Out Reset image size

In Fig. 6a, charge/discharge profiles have been divided into different zones. In discharge mode (down arrows) can be observed three zones: (I), (II), and (III). In zone (I), from the upper cut-off potential (ca. 4.3 V) until ca. 3.8 V, the potential drops abruptly due to a high cell's polarization. These conditions produce an increase in the internal cell's impedance (ohmic resistance) and an increase in the compound's solubility. Subsequently, in the zone (II) located between 3.8 V to 3.5 V, a slight drop potential is produced (pseudo-plateau). This zone's composition range is around 0.1 < x < 0.4, typical of a topotactic transition described by Eq. 2. Here, lithium ions are inserted into the CuHCF structure occupying the octahedron sites. During this insertion, Fe³⁺ is reduced to Fe²⁺ at a potential close to 3.5 V. In zone (III), at a global composition range between ca. 0.4 < x < 0.73, another pseudo-plateau is observed due to a topotactic transition, where more than one Li⁺ ion per formula unit is inserted during the second and third discharge, showing this small pseudo-plateau. ²⁸ Here, the most lithiated phase is further enriched, while a second phase is impoverished, maintaining a constant global composition. At the end of zone (III), a single-phase reaction occurs again with its abrupt potential drop (lower cut-off potential ca. 2.7 V). Finally, in the zone (III) located in the global composition range between ca. 0.6 < x < 0.73, a capacity increase is observed mainly due to the reduction of Cu²⁺ ions occurring at 2.5 V, which is in agreement with cyclic voltammetry. According to these results, lithium ions' insertion reaches a maximum ca. x = 0.73 during the first discharge cycle and an average ca. x = 0.70, considering the first 150 cycles and a final value x = 0.63 for cycle 300. In charge mode (up arrows) can be observed three zones: (i), (ii), and (iii). In zone (i), a significant potential increase occurs from 2.7 V to 3.3 V, with a remotion of lithium ions around x = 0.25. Subsequently, in zone (ii), a pseudo-plateau between 3.3 V and 3.5 V is observed associated to the oxidation of Fe²⁺ to Fe³⁺ at this potential with the remotion of x = 0.55 (cycle 100) and x = 0.47 (cycle 300) of lithium ions. Finally, in the zone (iii), lithium ions' remotion reaches 80% at cycle 100 and 60% at cycle 300, increasing the voltage from 3.5 V to 4.3 V.

In Fig. 6b, the reversible specific capacity and the coulombic efficiency remain relatively constant during 300 cycles, representing the behavior of the cell subjected to in-depth discharge/charge studies. The average specific capacity achieves a close value to the theoretical specific capacity and a coulombic efficiency of 95%. The stability of the coulombic efficiency is due to the high retention capacity at C/20 (see Table SI). A plot of the percentage of remaining capacity as a function of cycle numbers (cf. Fig. S2) indicates a remaining capacity of 90.5% in the tenth cycle. However, when cycling increases, the remaining available capacity returns to an average of 96.8% for practically 250 cycles. High remaining capacity and high coulombic efficiency suggest that the host material has an excellent reversibility for the charge/discharge processes, rate capability, and life cycle. Additionally, it could be demonstrated excellent structural stability because XRD postmortem (Fig. S3) does not show a phase change, so holding the initial cubic structure.

Another interesting feature in Fig. 6b was the increase in the specific capacity of both processes when the temperature exceeded 40 °C (cycle 50th to 60th), going from ca. 60 mAh g⁻¹ to 80 mAh g⁻¹ for discharge and ca. 90 mAh g⁻¹ for charging. This significant change in the capacity is due to this study was carried out during August in the city of Córdoba (Spain), which during summer reaches temperatures close to 45 °C, being one of the hottest regions in the country (holidays time and the air conditioning in the laboratory was disconnected). The temperature has a significant impact on the performance and lifetime of LiBs. At low temperatures, the ionic conductivity of the electrolyte is significantly reduced, hindering ions transport. ²⁹ In this case, an experiment carried out at 40 °C increased the ionic conductivity of the electrolyte, with the subsequent increase in the storage capacity. The theoretical capacity for KCu[Fe(CN)₆] is 85.1 mAh g⁻¹ in its anhydrous form. Through the ATG analysis, it was possible to determine six more water molecules. Accordingly, it could be expected a decrease in capacity from 85.1 mAh g⁻¹ to value ca. 60 mAh g⁻¹. In practice, the capacities obtained are generally close to 60 mAh g⁻¹ and, in some cases less, since the structure has both zeolitic and adsorbed moisture. ²⁴ Fe(CN)₆ ³⁻ vacancies allow water molecules to be incorporated in two ways in the CuHCF structure: i) the zeolitic water molecules have a strong tendency to reside in or to compete with Li⁺ ions to occupy the interstitial spaces, which may block the transport of Li⁺ ion into the inside of the lattice, thereby lowering the capacity utilization of PB framework; ³⁰ ii), the water is coordinated to the metal ions located in exposed sites of the CuHCF vacancies. The oxygen atoms in the water molecule help protect the positive charge of the transition metal ions that are exposed. In this case, the relationship between vacancies and water is complementary because each additional vacancy exposes an additional interstitial ionic site for coordination and increases the available volume for zeolitic water. Zeolitic water can be removed more easily than coordinated water heating the material, as it is more loosely bound. Water can also influence the thermodynamically favored phase; solvate lithium, reducing its effective ionic potential and mitigating the cubic to the rhombohedral displacement of the crystal structure. ²⁶ Table I shows a comparison between results obtained for different cathodes based on metal hexacyanometalates and other types of compounds that have been studied.

The manganese-based compound's behavior stands out from the table since, at a very high C-rate, it is capable of retaining 67% of theoretical capacity. Concerning our compound, it can be noted that capacities similar to those obtained by other authors are achieved. Additionally, it has been found a decrease in a electrochemical performance of CuHCF in LiBs, compared with NaBs and KBs. ³⁷ One possible interpretation of this behavior involves the size of the inserted species, often expressed in terms of the Stokes radius, which is a parameter experimentally determined by experiments in aqueous systems and can be corrected for non-aqueous solutions. ³⁴ The higher the charge to radius ratio of an ion, the more strongly water molecules will coordinate to it, so a smaller effective ionic radius results in a larger Stokes radius in aqueous solutions. Therefore, batteries based on Li⁺ ions present a worse performance than Na⁺ and K⁺ ions when metalhexacyanometallates are used as active materials.

EIS is a technique that measures resistance (R), capacitance (C), and inductance (L) by monitoring the current response while applying an AC voltage to an electrochemical cell. Impedance is defined as the resistance that interrupts or out of phase current when an AC voltage is applied to the circuit. ³⁵ During the cycles of a LiB, its internal impedance will slowly increase. The higher the impedance of a battery, the more difficult it is to move lithium ions through it. This feature occurs at extreme high or low temperatures, or as a result of the growth of a solid electrolytic interface layer (SEI). ³¹ Figure 7 shows that changes in battery impedance as charge/discharge cycles increase at a rate of 1C, and Table II shows the parameters obtained from the equivalent circuit.

Zoom In Zoom Out Reset image size

EIS consists mainly of three regions, i.e., the low-frequency region, the mid-frequency region, and the high-frequency region. ³⁶ In the low-frequency region, the EIS appears as a straight line with a constant slope associated with a diffusion phenomenon. In this case, for the initial state and first cycle, a constant phase element (CPE2) was adjusted instead of a Warburg impedance because the latter caused a failure in the fit, and a CPE corresponds to a Warburg element of infinite-length. ³⁷ In cycles 5−20−50, the circuit was correctly fit with both elements, showing infinite diffusion in the final process. The mid-frequency region was modeled by a parallel connection of a constant phase element (CPE1) and a resistance (R1) that corresponds to the resistance to charge transfer. The constant phase element was used instead of a capacitor to compensate for the electrode's non-ideal behavior, which corresponds to carbon current-collector (GDL). Finally, R1 corresponds to the electrolyte's resistance that describes the behavior of the battery at high frequencies. The Nyquist plots were analyzed by Zview software and show that an increase in charge/discharge cycles produces a decrease in the charge transfer resistance, from 749 Ω for the initial state to 109 Ω at the end of cycle 50 (smaller semicircles). This decrease in resistance can be attributed to the increase of ionic conductivity with cycles which is related to a positive effect of vacancies in the CuHCF structure. The interconnected cavities offer an alternative route to insertion other than the channels [100], and there is evidence to suggest that the vacancies allow the conduction of larger ions. ²⁶ These changes play a fundamental role in the electrochemical performance and lifecycle of LiBs because it prevents the electrode surface from further reacting with the electrolyte. Furthermore, the double layer's capacity also decreases from 3.16 μF for the initial state to 2.17 μF at the end of the 50 cycles. This is due to a decrease in the insulation layer produced initially when charge/discharge cycles increase as a result of improving conductivity. ³⁸ These results are consistent with those observed for specific capacity in Fig. 6b and demonstrate that the material has a low charge transfer at the beginning. However, as its restructuring begins by incorporating lithium ions, it improves its stability and conductivity.