Communication—Layered Double Hydroxide as Intermediate-Temperature Electrolyte for Efficient Water Splitting

Felix P. Lohmann-Richters; Martin Müller; Marcelo Carmo

doi:10.1149/1945-7111/ab8e80

High efficiency is crucial for all electrochemical energy conversion devices. One means to increase efficiency is increasing the operating temperature. The intermediate temperature range of 100 °C to 250 °C is especially attractive, as useful waste heat is generated and challenges related to high temperatures can be avoided. However, electrolytes for intermediate temperature operation remain a challenge. Conventional polymer electrolyte membrane (PEM) or alkaline electrolyzers usually operate at around 80 °C and achieve system efficiencies in the range of 39%–69% of the lower heating value (LHV).^1,2 Higher operating temperatures can increase the efficiency by accelerating electrode kinetics and increasing the electrolyte conductivity.^3–6

The most challenging component for intermediate temperature operation is the electrolyte membrane. Acidic polymer membranes have low mechanical stability at intermediate temperatures and composite membranes have not proved able to overcome this temperature limitation.⁷ Membranes utilizing free acids are not applicable: The acid would be washed out in a water-fed electrolyzer and most of the usually employed materials show very strong corrosion.^8,9 Zirfon, which is often used as the diaphragm in alkaline electrolyzers, has a maximum continuous operating temperature of 110 °C. In addition, its thickness limits the current density in conventional alkaline electrolyzers.¹⁰ While anion exchange membranes are an attractive alternative, their stability still remains a challenge, even at temperatures below 80 °C.¹¹

Layered double hydroxides (LDH) are OH⁻-conducting¹² and a promising electrolyte material or membrane additive in different electrochemical devices, such as batteries, fuel cells, electrochemical ammonia synthesis and oxygen separation.^12–16 LDHs have the general composition $\left[{{\rm{M}}}_{1-x}^{2+}{{\rm{M}}}_{x}^{3+}{\left({\rm{OH}}\right)}_{2}\right]{}^{x+}\left({{\rm{A}}}_{x/m}^{m-}\right).$ Where M are divalent and trivalent metal cations, e.g., Mg²⁺ and Al³⁺, respectively. A is an interlayer anion, e.g., ${{\rm{CO}}}_{3}^{2-}$ or ${{\rm{NO}}}_{3}^{-}.$ LDHs are usually synthesized by means of coprecipitation at high pH and are often aged at up to 180 °C.^17,18 They are thus stable in alkaline solutions. Kim et al. reported that the ionic conductivity of Mg–Al LDH increases from 25 °C to 200 °C under saturated humidity.¹⁹ Meanwhile, Zeng and Zhao employed a pressed layer of Mg–Al LDH as electrolyte in an alkaline electrolyzer at up to 70 °C.²⁰

In this work, we expand the operating temperature range of LDH-based electrolyzers to the intermediate temperature range: We successfully operate an LDH-based electrolyzer at up to 146 °C and observe increasing efficiency with increasing temperature.

Experimental

LDH synthesis

Nitrate Mg–Al LDH ([Mg₂Al(OH)₆]NO₃) was synthesized in a similar way to earlier instances by means of coprecipitation and ion exchange.^18,20,21 Mg(NO₃)₂·6 H₂O (22.5 g, >99%, Merck Suprapur) and Al(NO₃)₃·9 H₂O (16.5 g, >98%, Carl Roth) were dissolved in 196 ml of deionized water. A second solution consisting of Na₂CO₃ (25.9 g, >99.5%, Sigma-Aldrich) and NaOH (19.6 g, >99%, Merck) in 163 ml of deionized water was prepared. The second solution was added dropwise to the first until a pH of 11 was reached. The resulting dispersion was boiled under reflux for 24 h. The white precipitate was separated by centrifugation, washed with water until it reached a neutral pH and then dried at 80 °C under air.

To exchange the carbonate for nitrate, 9 g of the obtained powder were dispersed in methanol (315 ml, technical grade). HNO₃ (15.6 g, 20%, Merck) was diluted in methanol (45 ml) and slowly added to the dispersion, under vigorous stirring. The dispersion was left to stir for one hour. The product was separated by centrifugation, washed with methanol, dried at 80 °C under vacuum and characterized using X-ray diffraction (XRD) and scanning electron microscopy (SEM) (see supporting information available online at stacks.iop.org/JES/167/084512/mmedia).

Cell fabrication

The Mg–Al LDH powder was pressed on nickel foam (42 × 42 mm, 1000 g m⁻², Recemat BV) with 30 kN for 15 min to form a ca. 90 μm-thick, dense layer. Two such foam pieces, with the LDH layers facing to one another, were sandwiched between flow fields with PTFE gaskets, as depicted in Scheme 1. The flow fields were of a single serpentine type. The anode and cathode flow fields were made of titanium, coated with platinum (anode) and platinum and gold (cathode), as employed in earlier investigations.²² Even though platinum is a well-known electrocatalyst, the coating on the flow field should not be catalytically active because there is no ionically conducting connection to the electrolyte layer. The cell was thermally isolated by wrapping it with glass fiber isolation tape.

**Scheme 1.** Schematic representation of the employed electrochemical cell. Blue: Flow-field; shaded: Ni-foam electrode; size not to scale.
Download figure:
Standard image High-resolution image

Electrochemical measurements

A total flow of ca. 6.2 g min⁻¹ of steam from a steam generator (PS100, Stritzel Dampftechnik, Germany; set to 4 barg) was fed through the anode and cathode together. A valve was used to adjust the pressure at the cell outlet to 3.6, 1.7 and 0.2 barg. The temperature resulted from heating by the steam at the set pressure. The measurements were thus performed under a condensing atmosphere. The cell was preconditioned at 146 °C and 2 V for two hours. After each temperature change, the cell was allowed to equilibrate for 20 min at 2 V. The current and the impedance spectrum were measured after holding at the corresponding voltage for three minutes. The impedance was measured between 30.5 kHz and 0.4 Hz using a BioLogic potentiostat HCP-1005. The real part of the high-frequency impedance was taken as an estimate of the ohmic resistance. The ohmic resistance should be dominated by the electrolyte resistance, but also include contact resistances. It can thus be considered an upper boundary for the electrolyte resistance and a lower boundary for the electrolyte conductivity. The difference in the real part of the low- and high-frequency impedance was taken as an estimate of the electrode resistance. A detailed analysis and fitting of the impedance data would be necessary for further interpretation but this is not within the scope of this publication.

Results and Discussion

The synthesized LDH was characterized using XRD and SEM. The diffractogram (Fig. S1) was in good agreement with the one reported in the literature employing similar reaction conditions.¹⁸ The SEM-images revealed platelets of around 100 nm in diameter (Fig. S2).

An electrolysis cell with an electrolyte layer of pressed nitrate Mg–Al LDH ([Mg₂Al(OH)₆]NO₃) was characterized at different temperatures under steam. The cell is schematically shown in Scheme 1. The current density increases with increasing temperature across the entire voltage range (Fig. 1), corroborating the trends reported in the literature.^3–5 The collected data is summarized in Fig. 2, which shows the temperature dependence of the current density measured at 2 V and the electrolyte resistance and electrode resistance estimated using impedance spectroscopy. Both resistances decrease with increasing temperature in accordance with earlier reports^19,23 and illustrate the benefit of increased operating temperature for the electrolyte conductivity and the electrode kinetics.

**Figure 1.** Voltage vs current density plot for different temperatures (dashed line = 9 mA cm⁻²).
Download figure:
Standard image High-resolution image

**Figure 2.** The current density and resistances at 2 V under condensing atmosphere depend on the temperature. The points marked with an X were collected with additional electrical heating, which results in a reduced relative humidity.
Download figure:
Standard image High-resolution image

The order in which the different data points were collected is given in Fig. S3. Even though the long-term stability is not within the scope of this work, the current density is fairly stable throughout the total measurement time. Only a slight initial degradation is observed.

Increasing the flow of steam from ca. 6.2 to 8.3 g min⁻¹ does not significantly change the cell characteristics (Fig. 2). Thus, the flow rate does not limit the current density under these conditions. When the cell is electrically heated so that the relative humidity decreases and no condensation occurs, the current density decreases strongly and the resistances increase (points marked with an X in Fig. 2). This humidity dependence of the LDH's ion conductivity agrees with earlier reports.^20,24 The ionic conductivity at 146 °C is 12 mS cm⁻¹. It has not been corrected for other ohmic resistances and is thus a lower boundary. It is the same order of magnitude as reported by Kim et al. for Mg–Al–LDH with OH⁻ as anion (26 mS cm⁻¹ at 140 °C).¹⁹

The current density at 146 °C is 79 mA cm⁻², which is less than that observed by Allebrod et al. in a steam-fed alkaline electrolysis cell at 150 °C (633 mA cm⁻²).²³ This difference is probably due to the alkaline conditions and micro-structured electrodes employed by Allebrod et al. Zeng and Zhao reported an increase in the current density at 2 V by a factor of 10 when using 0.1 mol l⁻¹ NaOH instead of pure water.²⁰ The electrodes in our electrolysis cell consist of nickel foam only. Microstructured electrodes have been reported to increase the current density by a factor of at least more than three.^25,26 Therefore, a significant increase in current density can be expected if state-of-the-art electrocatalysts and structured electrodes under alkaline conditions are used. However, such investigations are not within the scope of this work, which primarily demonstrates the feasibility of LDHs as intermediate-temperature electrolyte for water electrolysis and the benefit of the temperature increase.

The solid electrolyte employed in this work shows two main advantages over previously reported, intermediate-temperature alkaline diaphragms, which consisted of a liquid electrolyte immobilized in the pores of a ceramic material.³ Firstly, it is solid and does not require careful control of the water partial pressure to avoid evaporation or dilution of the electrolyte when the electrolyzer is fed with steam. The second advantage of the LDH-based membrane presented in this work is its intrinsic conductivity; it can thus be fabricated as dense membrane. This reduces gas crossover, enabling improved gas quality and partial load operation, as well as differential pressure operation. In addition, the conductivity of the liquid is not crucial and a high pH is only needed to facilitate the electrode reactions. Therefore, the concentration of a potential liquid alkaline feed-solution can be strongly reduced compared to classic alkaline electrolyzers—an advantage in terms of corrosion prevention. The conductivity of the LDH might be even further increased by optimizing the LDH composition and controlling the crystallographic orientation.^{15,19,24,27,28} Mechanically stable LDH composites with polymers can be prepared^16,18,29,30 to overcome the mechanical instability of pressed powder layers. Employing high-temperature polymers is crucial to enable a wide operating temperature range.

Designing a system for alkaline electrolysis in the intermediate temperature range, e.g., for up to 200 °C, is certainly challenging but can enable significant gains in efficiency. The efficiency increase becomes obvious when comparing the different temperatures. For 9 mA cm⁻², 1.8 V must be applied at 107 °C, while 1.6 V yield the same current density at 146 °C (Fig. 1). This corresponds to an efficiency increase from 68% to 77% (LHV) and further gains can be expected at higher temperatures. The theoretical efficiency limit is defined by the thermoneutral voltage at the operating temperature, i.e., 84% at 150 °C. Low voltages not only lead to higher efficiency but also enable operation much closer to the thermobalanced voltage, reducing the need for cooling and simplifying the overall system.

Summary

This work shows that layered double hydroxides serving as electrolyte in an electrolyzer enable operation at intermediate temperatures. Remarkable efficiency gains with increasing temperature were observed. This is a first step towards highly efficient intermediate temperature alkaline electrolysis and highlights its potential. The proposed electrolyte shows several advantages over previous intermediate temperature diaphragms. New opportunities are opened for future investigations of optimal electrodes, composite membranes, system design, and operating conditions to bring about further improvements and an improved understanding of the processes and material behavior in the intermediate temperature range. LDHs as electrolytes might also enable intermediate temperature operation in other electrochemical devices such as batteries, fuel cells, electrochemical ammonia synthesis, oxygen separation, and CO₂-reduction.

Acknowledgments

The authors thank Walter Zwaygardt for technical assistance and Andreas Everwand for recording SEM images and X-ray diffractograms.

Communication—Layered Double Hydroxide as Intermediate-Temperature Electrolyte for Efficient Water Splitting

Article metrics

Author e-mails

Author affiliations

Author notes

ORCID iDs

Dates

Abstract