Nickel-iron layered double hydroxides for an improved Ni/Fe hybrid battery-electrolyser

The transition to renewable electricity sources and green feedstock implies the development of electricity storage and conversion systems to both stabilise the electricity grid and provide electrolytic hydrogen. We have recently introduced the concept of a hybrid Ni/Fe battery-electrolyser (battolyser) for this application1. The hydrogen produced during the Ni/Fe cell charge and continued electrolysis can serve as chemical feedstock and a fuel for long-term storage, while the hybrid battery electrodes provide short term storage. Here, we present Ni–Fe layered double hydroxides (NiFe-LDHs) for enhancing the positive electrode performance. The modified Ni(OH)2 material capacity, high rate performance and stability have been tested over a large range of charge rates (from 0.1C to 20C) over 1000 cycles. The Ni–Fe layered double hydroxides allow the capacity per nickel atom to be multiplied by 1.8 in comparison to the conventional β-Ni(OH)2 material which suggests that the nickel content can be reduced by 45% for the same capacity. This reduction of the nickel content is extremely important as this presents the most costly resource. In addition, Fe doped Ni(OH)2 shows improved ionic and electronic conductivity, OER catalytic activity outperforming the benchmark (Ir/C) catalyst, and long term cycling stability. The implementation of this Fe doped Ni(OH)2 material in the Ni/Fe hybrid battery-electrolyser will bring both electrolysis and battery function forward at reduced material cost and energy loss.

-1 illustrates the Ni-Fe hybrid battery-electrolyser and the function it can provide in several potential application areas. The device has a negative electrode in which Fe(OH) 2 is reduced to Fe upon charge (-0.88 V relative to the standard hydrogen electrode (SHE)): When the positive electrode contains β-Ni(OH) 2 , β-NiOOH is formed upon charge (+0.49 V vs. SHE):

Eq-S 2
For an electrode containing the α-phase x , the charge reaction may be described as

Eq-S 3
During the overcharge, the following reactions take place on the reduced iron and oxidised nickel hydroxide electrode respectively: The extra peak observed at low angle in XRD spectra of the material NiFe7 (2θ=15°) reveals the presence of an extra periodicity (E.P) in the c-axes of the interstratified material which could correspond to the repetition of a sequence of one alpha inter slab distance (6.88 Å), and two beta interslab distances (4.39 Å), adding up and forming a [dα dβ dβ] reflection. To our knowledge, this periodicity has not been mentioned before and reveals a certain regularity and organisation in the interstratification of the alpha and beta phases. In contrast all three iron doped samples show two weight loss steps between 0 and 600 °C. The first one appearing between 0 and 200 °C corresponds to the loss of adsorbed and intercalated water, the second between 200 °C and 400 °C is due to the dissociation of Fe-Ni(OH) 2 into Fe-NiO. The content of adsorbed and intercalated water in the three NiFe-LDH materials corresponds to 18 wt% of the material weight. The beginning of weight loss observed after 700 °C in Figure S-3 a and b is attributed to sulphate anions present in the sample 2 . This confirms that samples NiFe20 and NiFe15 are alpha phase nickel hydroxides with sulphate anions intercalated. The samples Ni-B but also NiFe7 do not show this decrease because Ni-B is a beta nickel hydroxide with no anions intercalated and NiFe7 is a mixed α/β phase interstratified material with probably not enough sulphate intercalated to be noticed on its TGA graph.

Material synthesis
The amount of nickel in the samples, as well as the molar ratio x=Fe/(Ni+Fe) determined by ICP are displayed in Table 2. The molar ratio Fe/(Ni+Fe) is estimated to be 7, 13 and 18 % for NiFe7, NiFe15 and NiFe20 respectively which allows to determine the stoichiometry of Ni and Fe in the material. Sulphate salts being used for the NiFe-LDH synthesis, one can expect SO 4 2to be intercalated in the interlayer space of the crystal of the freshly prepared samples to compensate the excess of positive charges induced by the iron substitution. Thus a proposition of the different materials chemical formula is given in Table S-1.
Several authors suggest that after ageing in KOH, the anions intercalated is being replaced by carbonate anions (in case of use in a battery that is not fully closed to ambient air). Carbonates can indeed also be formed in the KOH solution during its fabrication in ambient air. According to Mendiboure et al. 3  of CO 3 2− with the nickel oxide layers is stronger than for the sulphate anions, the ion-exchange equilibrium constants following the sequence CO 3

the affinity
Thus, Hunter et al. 4 show for NiFe-LDH materials prepared with various anions intercalated that, after ageing in KOH in ambient air, the initial intercalated anions were mostly replaced by carbonates. For their NiFe-SO 4 2material, this anion exchange comes with a reduction of the interlayer distance from 8.6-8.2 Å to 7.6 Å. The same observation was made in the present study, with a reduction of the interlayer space from 8.2 Å to 7.7 Å for material NiFe20 after 1 month of ageing in KOH. Once can make the assumption that after the ageing period, carbonates replace sulphates in all the NiFe-LDH samples. Another possibility in the absence of ambient air CO 2 is replacement of sulphates by smaller OH -. In our experiments we observe the same c-axis reduction for aging in a closed bottle (no CO 2 ingress from air) and aging in a not fully closed battery and electrolysis cell (not fully closed to allow for O 2 and H 2 evolution and occasional water replenishment). Further analysis, such as IR-Spectroscopy, would be necessary to confirm the presence of CO 3and OH -.    Eq-S 13 t bat corresponds to the duration necessary for the charge inserted to equal the discharge capacity, t bat =t d , t el is the duration of the overcharge, t el =t c -t bat (Cf Figure S-6). C d is the discharged capacity, C c is the total charge inserted (overcharge included) and C el is capacity inserted during the overcharge (C el =C c -C bat ). V el and H el are the potential of the electrolysis plateau and the thermoneutral potential respectively.

Energy efficiency calculation
The energy efficiency losses can be disentangled into nickel electrode and iron electrode losses: Eq-S 16 E TN (OER) is the thermoneutral potential for the oxygen evolution reaction, V OER the potential of the OER plateau.
can be also rewritten as a function of its hysteresis and kinetic components (   Fig-S-3) and hysteresis (green area in Figure

 XRD analysis
XRD analysis is performed on the NiFe20 electrode after 1000 cycles to confirm the stability of the alpha phase. The XRD spectrum of the electrode (Red) in Figure S-9 shows some high intensity peaks corresponding to different materials such as carbon and nickel used for the electrode making which are not indexed in the figure being not relevant for the hydroxide. The results show that the material consists mainly of α-Ni(OH) 2 with a d003 peak at 13.8°. The presence of γ-Ni(OH) 2 is also revealed by a peak at 15.7° overlapping partly with the d003 peak of the alpha phase. This reveals that a part of the active material could not be discharged. The small peak at 22.7° can be assigned to the d001 peak of the β-Ni(OH) 2 which suggests that a portion of the nickel hydroxide is converted to β-Ni(OH) 2 during the ageing and the cycling of the sample. The area of the peak suggests that the portion of β-Ni(OH) 2 is low relatively to the alpha/gamma couple but this could explain the reduction of the capacity from 1.57 e -/Ni to 1.4 e -/Ni after the 1000 cycles. A comparison of the NiFe20 electrode with the NiFe20 material aged in KOH highlights a small shift in the d003 and d006 peaks position of the α-Ni(OH) 2 reflecting a reduction of the interlayer distance. As mentioned before, this phenomenon could be explained by the pursuit of the intercalated anions replacement with carbonate anions along the electrode cycling in the not hermetically sealed test cell.