Effect of Deformation on Electrochemical Performance of Aluminum-Air Battery

A schematic describing the process flow to obtain Al specimens with different amounts of dislocations is shown in Fig. 1. The pure Al specimen was fabricated through a conventional melting technique with a tube furnace using Al shot (Part#:00632. Alfa Aesar, USA) with 99.9% purity. As shown in Fig. 1a, the melting process was conducted at 800 °C under an inert 97% Ar + 3% H₂ atmosphere with a flow rate of 400 sccm. The Al shot was melted in a bowl-shape alumina crucible. In order to remove slag during melting, manual stirring using a Fe wire was performed every 15 min before it was poured into a cylindrical copper mold with a diameter of 15 mm. As shown in Fig. 1b, the ingot was initially cold-rolled with thickness reduction of 50% into a specimen with a final thickness of 7.5 mm, followed by heat treatment at 400 °C for 1 h to promote a fully recrystallized Al microstructure (see Fig. 1c). The recrystallized Al specimen was later given another deformation using the cold rolling method to obtain deformed-Al specimens with different thickness reductions of 20% and 50%; specimens with different amount of dislocations were obtained by varying the thickness reduction during cold-rolling. Figures. 1c–1e show the elongation of the grain as more deformation was applied on the specimen. Recrystallized Al, thickness reduction of 20%, and thickness reduction of 50% are referred to as Al-A (Fig. 1c), Al-20 (Fig. 1d), and Al-50 (Fig. 1e), respectively.

Zoom In Zoom Out Reset image size

A scanning electron microscope (SEM, SU-70 Hitachi, Japan) equipped with electron backscatter diffraction (EBSD) was used to examine the microstructures of all specimens. In EBSD analysis, step sizes of 1, 0.7, and 0.2 μm were used for Al-A, Al-20, and Al-50, respectively. The EBSD data were then analyzed, and kernel average misorientation (KAM) calculation was used to determine the amount of deformation by measuring the dislocation density, then construct the subsequent dislocation map for each specimen.

Figures 2a and 2b show the deformed Al anode and set-up used for these experiments, respectively. The Al anode was immersed in the electrolyte, 4 M NaOH (Samchun, 98% NaOH, Korea) with 0.05 M Na₂SnO₃ (Aldrich, 95% Na₂SnO₃, USA) as a corrosion inhibitor. The catalyst at the cathode was coated on Teflon-coated carbon paper using an electrospray method. First, the Pt/C + IrO₂ (50 wt%) was disprsed in ethanol and isopropyl alcohol with 5 wt% Nafion solution (Aldrich 274704, USA). Second, the catalyst solution was placed in a syringe with a positively charged capillary tip. Next, a voltage of 8 kV was applied between the syringe needle tip and the carbon paper using a high-voltage power supply (Korea Switching, Korea). The composite droplets were then electro-sprayed from the needle at 3.0 ml h⁻¹ and collected on the aluminum foil. Finally, the solvent (ethanol and isopropyl alcohol) was dried at room temperature prior to the cell test, and the catalyst loading density was 1 mg cm⁻². Polarization curves were obtained by linear sweep voltage at a scan rate of 10mV s⁻¹ from open circuit voltage (OCV) to 0 V.

Zoom In Zoom Out Reset image size

A comprehensive analysis of the microstructure was conducted using EBSD in order to understand the characteristic behaviors of the deformation and heat treatment on all specimens. As shown in in Fig. 3a, the image quality (IQ) and inverse pole figure (IPF) maps of the Al specimens show coarse grains as well as fully recrystallized microstructures following the initial annealing at 400 °C for 1 h. Figure 3b shows that the elongated grains, which are under the thickness reduction of 20%, are formed by post-processed cold-rolling. Consequently, Fig. 3c shows that not only the elongated grains but also the microstructures are observed under severe plastics deformation, which is under the specimen thickness reduction of 50%. Although the IQ and IPF maps clearly distinguish between the pristine (i.e. recrystallized condition, see Fig. 3a and deformed specimens (see Figs. 3b and 3c), these quantitative analyses show the dependency between the metal deformation and the grain elongation. In particular, KAM distribution is strongly related to dislocation density inside a grain of the deformed or pristine microstructures.^38,39 This distribution is obtained through EBSD analysis and used to quantify the dislocation density. Kubin et al. reported that the KAM value (${\boldsymbol{\vartheta }}$), step size (${\boldsymbol{u}}$), and burger vector (${\boldsymbol{b}}$) of the matrix element could be used to calculate geometrically necessary dislocation (GND) density (${{\boldsymbol{\rho }}}_{{\boldsymbol{GND}}}$) using the following equation⁴⁰:

$$\begin{eqnarray}&&{\rho }_{GND}=\displaystyle \frac{2\vartheta }{ub}\end{eqnarray} \tag{ 4 }$$

The dislocation density is calculated on each point of the KAM map with a misorientation angle below 2°. Subsequently, the dislocation maps of all of the specimens are constructed using MATLAB (The MathWorks, Inc., USA), as shown in Figs. 3d–3f. The dislocation maps clearly show that the dislocation density increases significantly when additional deformation is applied on the specimens. In several regimes shown in red in Figs. 3e and 3f, cell structure with high density of dislocation after certain deformation can be observed. Figure 4a shows the detailed distributions of the dislocation density on each specimen. It is believed that the majority of the dislocation density value increases with a rise in deformation, as indicated by the peak shift of dislocation density shown in Fig. 4a. The Al-A specimen has a dislocation density value of 3.0 × 10¹⁵ m⁻², which gradually increases to 6.5 × 10¹⁵ m⁻² at the specimen thickness reduction of 20% (Al-20). The dislocation density finally reaches approximately 19 × 10¹⁵ m⁻² in the specimen thickness reduction of 50% (Al-50). The fact that overall dislocation density, which is a crucial strengthening mechanism in metallic materials,⁴¹ was increased with thickness reduction indicates that the mean value of hardness linearly esc alates with increasing dislocation density, as shown in Fig. 4b.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Figure 5 shows the I-V-P results, which indicate that the peak power densities of the Al-A, Al-20, and Al-50 specimens are 252, 219, and 174 mWcm⁻², respectively. The OCVs of these specimens are over 2.5, 2.25, and 2.2 V, respectively. The initial voltage drops from 2.4 V to 1.4 V occur in the three cases. It is believed that these common behaviors can be attributed to the partial hydrogen evolution reaction (HER) at the cathode. Previous studies have reported the initial voltage drop as a polarization loss in a typical Al-air system.³³

Zoom In Zoom Out Reset image size

We also suspect that the additional overpotenti.al at the localized cathode area is due to the low oxygen vapor pressure or sluggish oxygen transport, resulting in a hydrogen evolution reaction. However, for our study, these common behaviors at the cathode can be negligible. As discussed above, the Al specimens have different deformation ratios (i.e., dislocation densities). These controlled parameters strongly influence the electrochemical characteristics of the Al anode side and the cathodes are exactly the same. Based on the results of these experiments, two possibilities should be considered: grain size and dislocation density.⁴² Zhu et al. reported that the grain size is strongly correlated with the cycling capability and the grain size.⁴³ Aung and Zhou reported that the grain boundary acts as a physical corrosion barrier.⁴⁴ Small grain size creates more grain boundaries, as a consequence, the rate of corrosion in small-grained microstructure slowed down. Especially, their group exhibited the corrosion rate was found to be increased by 30% with increase in the average grain size from 65 μm to 250 μm. Also, Afshari et al. reported that the corrosion resistance of Fe in alkaline solution considerably increased as the grain size decreased from microcrystalline to nanocrystalline.⁴⁵ In this test, the pre-processing used to control the microstructures is the same, in which the grains are roughly the same. The other factor is the controlled dislocation density of the Al specimen. As shown in Fig. 5, Al-air battery with superior mechanical property (i.e. microhardness) exhibited relatively lower power density. Hamu et al. reported the correlation between the corrosion characteristic and the dislocation density of Al based alloy.⁴² The dislocation density was increased after the severe plastic deformation process (see Fig. 1), and the acceleration in the corrosion rate was correlated with the massive plastic strain imposed. It is therefore believed that the corrosion characteristics of the anode metal strongly correlate with anodic overpotential of metal-air system. As shown in Fig. 5, the dislocation within Al metal at the anode would lead to a sluggish reaction of metal oxidation.

We investigated variations in electrochemical performance with respect to the dislocation density of Al at the anode of the Al-air battery. The pristine Al specimen with minimized deformation has a superior electrochemical performance compared to that of the deformed Al specimen, which we controlled by thickness reduction of 20% and 50%. In addition, we found that a high mechanical property of the processed metal would suppress the electrochemical behaviors of metal oxidation in metal-air battery. Since the metal behaviors at the anode for metal-air system still uncover, we will optimize the metal microstructure to enhance the electrochemical performances of the metal-air battery as a future work.

This work was supported by the Mid-Career Researcher Program (NRF-2018R1A2A1A05077532) through the National Research Foundation of Korea (NRF), funded by the Ministry of Science ICT and Future Planning. This work was also supported by NRF Grant (2017R1C1B501818 and NRF-2018-Global PhD Fellowship Program).