Recent developments in materials for aluminum–air batteries: A review

Journal of Industrial and Engineering Chemistry 32 (2015) 1–20

Contents lists available at ScienceDirect

Journal of Industrial and Engineering Chemistry journal homepage: www.elsevier.com/locate/jiec

Review

Recent developments in materials for aluminum–air batteries: A review Marliyana Mokhtar a, Meor Zainal Meor Talib a,b, Edy Herianto Majlan a,*, Siti Masrinda Tasirin a,b, Wan Muhammad Faris Wan Ramli a, Wan Ramli Wan Daud a,b, Jaafar Sahari a,c a

Fuel Cell Institute, Universiti Kebangsaan Malaysia, 43600 Bangi, Selangor, Malaysia Department of Chemical and Process Engineering, Faculty of Engineering and Built Environment, Universiti Kebangsaan Malaysia, 43600 Bangi, Selangor, Malaysia c Department of Mechanic and Material Engineering, Faculty of Engineering and Built Environment, Universiti Kebangsaan Malaysia, 43600 Bangi, Selangor, Malaysia b

A R T I C L E I N F O

Article history: Received 11 May 2015 Received in revised form 28 July 2015 Accepted 8 August 2015 Available online 17 August 2015 Keywords: Aluminum–air batteries Aluminum anode Air cathode Aqueous electrolyte Corrosion inhibitor

A B S T R A C T

The aluminum–air battery is an attractive candidate as a metal–air battery because of its high theoretical electrochemical equivalent value, 2.98 A h g1, which is higher than those of other active metals, such as magnesium (2.20 A h g1) and zinc (0.82 A h g1). This paper provides an overview of recently developed materials for aluminum–air batteries to be used in various elements, including the anode, air cathode and electrolyte. Aluminum can be alloyed with other active metal elements such as Tin (Sn), Indium (In), Gallium (Ga) and Zinc (Zn). Its binary and tertiary alloys demonstrate improved battery performance. Bifunctional air cathodes fabricated using oxygen reduction reaction (ORR) catalysts, CoO/N-CNT with oxygen evolution reaction (OER) catalysts, Ni–Fe-layered double hydroxide/CNT and MnO2/N-CNT yield excellent results. With regard to electrolytes, several types have been considered: aqueous, nonaqueous, aprotic solvent and solid-state electrolytes. The addition of corrosion inhibitors to an aqueous electrolyte helps to enhance battery performance, whereas non-aqueous and aprotic solvent electrolytes can be used to prevent hydrogen evolution. Polymer electrolytes can overcome the battery leakage problem. As a conclusion, the future research trends related to this type of battery have also been indicated. ß 2015 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Anode material . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pure aluminum . . . . . . . . . . . . . . . . . . . . . . . Aluminum alloys. . . . . . . . . . . . . . . . . . . . . . Selection of aluminum anodes . . . . . . . . . . . Cathode material . . . . . . . . . . . . . . . . . . . . . . . . . . . Nanomaterial and nanotube technology . . . Nitrogen-doped carbon nanotubes, N-CNTs Bifunctional air cathode . . . . . . . . . . . . . . . . Selection of air cathodes. . . . . . . . . . . . . . . . Electrolytes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Aqueous electrolytes. . . . . . . . . . . . . . . . . . . Acidic solutions . . . . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

* Corresponding author. Tel.: +60 3 89118521; fax: +60 3 89118530. E-mail addresses: [email protected], [email protected] (E.H. Majlan). http://dx.doi.org/10.1016/j.jiec.2015.08.004 1226-086X/ß 2015 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

2 2 2 2 3 4 4 4 4 4 8 8 8

2

M. Mokhtar et al. / Journal of Industrial and Engineering Chemistry 32 (2015) 1–20

Neutral salt solutions . . . . . . . . . . . . . Alkaline solutions . . . . . . . . . . . . . . . . Non-aqueous and aprotic electrolytes . . . . . . . . Non-aqueous solutions . . . . . . . . . . . . Aprotic solvents . . . . . . . . . . . . . . . . . . Solid-state electrolytes . . . . . . . . . . . . . . . . . . . . Selection of electrolytes . . . . . . . . . . . . . . . . . . . Corrosion inhibitors . . . . . . . . . . . . . . . . . . . . . . Organic inhibitors . . . . . . . . . . . . . . . . Synthetic inhibitors . . . . . . . . . . . . . . . Future trends in research on aluminum–air batteries . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

Because of the depletion of finite resources and the extensive growth in the demand for alternative energy worldwide, metal–air batteries have been proposed as alternative energy storage devices. Metal–air batteries have received particular attention because of their high energy density and capacity, the lack of dependence of their capacity on load and temperature, their flat discharge voltage and their low fabrication cost (depends on the metal used) [1–6]. Lithium–air (Li–air) batteries have been aggressively studied because of their broad potential for highperformance applications [7–10]. Such batteries can also operate as rechargeable batteries [11,12]. Unfortunately, however, during battery fabrication, the lithium must be handled under inert conditions because it is very sensitive to ambient conditions and poses an explosion hazard [13–15]. This is the greatest challenge for the Li–air battery. As alternatives, other active metal elements such as aluminum have been recommended. Aluminum (Al) is an attractive candidate anode material for metal–air batteries because it has a high theoretical electrochemical equivalent value, 2.98 A h g1, which is the second highest after that of lithium (3.86 A h g1) and higher than those of other active metals, such as magnesium (2.20 A h g1) and zinc (0.82 A h g1) [16–18]. Aluminum is also an inexpensive metal, as it is the second most abundant metallic element after silicon, and is characterized by its environmental friendliness, non-toxicity and high recyclability [19]. The theoretical specific energy of an Al–air battery with an alkaline electrolyte can be as high as 200 W h kg1, and with a neutral salt solution, it is between 300 W h kg1 and 500 W h kg1 [18]. In this paper, we will provide an overview of recent material developments for various elements of aluminum–air batteries, including the anode, air cathode and electrolyte. Each component and material has its own strengths and challenges. This type of battery comprises three main components: an anode, a cathode and an electrolyte. The discharging battery serves as a galvanic cell that drives the electrical current in an external circuit. The electrolyte plays an important role in such a battery because it is the conducting medium through which the two-way charge transfer proceeds between the electrodes [18]. The electrolyte also separates the anode and the cathode to avoid a short circuit and simultaneously provides hydroxide ions to maintain the electrochemical reactions [20]. The oxidation reaction at the anode depends on the type of electrolyte that participates in the reaction:

Cathode : Overall :

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

.. 9 . 11 . 12 . 12 . 12 . 12 . 14 . 15 . 15 . 18 . 19 . 19 . 19 . 19

Another undesired (parasitic) reaction occurs at the anode because of the water reduction reaction. The parasitic hydrogengenerating reaction can be expressed as follows:

Introduction

Anode :

. . . . . . . . . . . . . .

Al ! Al3 þ 3e O2 þ 2H2 O þ 4e ! 4OH 4Al þ 3O2 þ 6H2 O ! 4AlðOHÞ3

(1.1) (1.2) (1.3)

Side reaction : Al þ 3H2 O ! AlðOHÞ3 þ

3 H2 2

(1.4)

One major obstacle that hinders the deployment of the Al–air battery on a commercial scale is the self-corrosion rate of aluminum [21,22]. There are three main processes that occur on the surface of aluminum that hinder further oxidation at the anode in an aqueous-based cell: the formation of an oxide film of Al2O3 or Al(OH)3; the formation of corrosion products, Al(OH)3 and Al(OH)4; and parasitic hydrogen evolution, which lowers the potential of the battery [23,24]. Because of this limitation, further development effort is needed to reduce the corrosion rate. Anode material Pure aluminum Super-pure aluminum with a purity of 99.999% has been used in the past as an anodic material for the Al–air battery system because it has excellent electrochemical properties and operational potential at approximately 0.8 V vs. saturated calomel electrode (SCE) in an aqueous electrolyte [25]. Although it is, in most respects, a very suitable anode material, it suffers from a very high rate of corrosion, which makes it infeasible for application as an alternative power or energy source in place of fossil fuels. Other major difficulties that hinder the operation of Al–air cells are passive hydroxide layer formation and high corrosion currents with parasitic hydrogen evolution on the surface of the aluminum metal. Recently, the anode performance of 2N5 commercial grade aluminum with 99.5% purity has been studied to determine its potential for reducing self-corrosion and simultaneously improving battery performance [26]. Aluminum alloys Modification of the aluminum anode material with low ratios of alloying components in aluminum alloy fabrication is needed to reduce the corrosion rate and increase the operation time for the Al–air battery system. Typical aluminum alloys that are commonly used in batteries at present are Al–Zn, Al–In, Al–Ga and Al–Sn [27–34]. Al alloys that combine multiple alloying components exhibit superior performance by virtue of the beneficial attributes of each individual alloying component. Gallium (Ga) is an attractive alloying element that is known to activate the surface of aluminum in chloride solution. The activation of localized surface sites causes thinning of the passivated oxide film where the gallium is placed. A pure

M. Mokhtar et al. / Journal of Industrial and Engineering Chemistry 32 (2015) 1–20

3

Table 1 Charge, QC, for Al and Al–Sn alloys, determined after the termination of cathodic polarization lasting for 1 s and after the return of the electrode potential to EOCP [36]. t (s)

1

E (V)

LCP

HCP

1.6 1.7 1.8 1.9 2.0

QC (mC cm2) Al

Al-0.02%Sn

Al-0.09%Sn

Al-0.20%Sn

Al-0.40%Sn

17 23 63 390 1407

18 22 60 317 1202

14 22 56 370 1069

10 27 51 112 430

11 32 42 94 151

aluminum anode is known to exhibit enhanced performance with the addition of gallium ions to the chloride electrolyte solution. The alloying of indium and zinc with aluminum metal can produce a low-melting-point alloy that can be used as an anode at high current densities. As an alloying element, zinc has proven to reduce hydrogen evolution at an aluminum anode by increasing the hydrogen evolution potential and thus reducing the anode degradation. Tin (Sn) and indium (In) are alloying elements that, when deposited randomly on the surface of an aluminum anode, can increase the hydrogen evolution overpotential and limit the Al reaction area, thereby causing a positive shift in the anode potential and increased cathodic polarization, which reduces hydrogen growth at the anode. Tin is capable of accelerating aluminum dissolution in an aqueous electrolyte. Cathodic polarization induces the activation of the alloyed aluminum for anodic dissolution. Selection of aluminum anodes Alloyed aluminum (0.1% Sn and 0.5% Mg) from Alcan and 3 unalloyed aluminum anodes – 2N7 commercial grade (99.7% purity), 3N5 grade (99.95% purity) and 5N (99.999% purity) – have been tested in NaOH electrolytes containing corrosion inhibitors such as Na2SnO3 and Al(OH)4. 2N7 grade aluminum exhibits a very high corrosion rate that hinders the electrochemical measurement process, making it infeasible for use as an anode material [35]. The open circuit potential of Alcan alloyed aluminum at 1.87 V vs. Hg/HgO is more negative than that of 3N5 or 5N aluminum by 70 mV. 3N5 aluminum suffers from a higher corrosion rate than those of 5N and Alcan aluminum by 10% and 24%, respectively [35]. Super-pure aluminum with a purity of 99.999% and Al–Sn alloys at various concentrations of 0.02% to 0.40% in 0.5 M NaCl solution exhibit comparable cathodic performances at a high cathodic potential of 1.9 V vs Saturated calomel electrode (SCE); as the potential is decreased, this performance initially decays rapidly to a minimum and then increases again [36]. The high amount of charge spent on reduction, Qc, which is between 320 mC cm2 and 1400 mC cm2, indicates cathodic breakdown and the presence of oxide film hydrates [36]. The low cathodic potential observed at 0.4% Sn rapidly decays with increasing QC throughout this charge range. A higher content of Sn as an alloying element causes a decrease in QC and reduces alkali attack at the oxide layer. Alloyed aluminum with 0.2% or 0.4% Sn exhibits the best performance as anode materials among the investigated Al–Sn alloys [36]. Table 1 summarizes the result of charge, QC, for Al and Al–Sn alloys, determined after the termination of cathodic polarization lasting for 1 s and after the return of the electrode potential to EOCP. Gallium deposited on aluminum metal from gallium ions (Ga3+) through an alkalization process cannot induce complete activation [37]. In a solution of 0.5 M NaCl, the breakdown potential is observed to be 1.2 V, which is more negative than that of the anode without the deposition Ga, namely, 0.83 V [37]. Song et al. [38] has been studied two aluminum foils with different chemical composition, which are Al–1.4  106Sn–8.7  106Fe

named as T3 and Al–22.7  106Sn–10.3  106 Fe named as T24. A comparison of the corrosion behaviors of two aluminum foils, T3 and T24, with different Sn compositions of 1.4  106 and 22.7  106, respectively, has revealed than Sn can be used to enhance the corrosion behavior of aluminum material, but only at low compositions [38]. Water contamination of aluminum anodes with non-aqueous electrolytes such as AlCl3/g-butyrolactone (AlCl3/BLA) and tetraethylammonium chloride/acetonitrile ((C2H5)4NCl/CH3CN) causes large polarization losses of between 100 mV cm2 mA1 and 400 mV cm2 mA1 as a result of the formation of a solid oxidized Al precipitate at the anode and electrolyte degradation leading to the formation of HCl [39]. AlCl3 þ 3H2 O ! AlðOHÞ3 þ HCl

(1.5)

By contrast, water contamination of(C2H5)4NCl/CH3CN gives rise to a completely opposite effect, with the water seeming to facilitate the Al oxidation reaction and to enhance the battery’s performance; as a result, polarization losses are reduced from 20 mV cm2 mA1 to 0.7 mV cm2 mA1 at a low concentration of 0.3 M H2O [39]. Tafel curves of Al–In–Zn anodes with a high concentration of Zn in an alkaline electrolyte reveal a corrosion potential of 1.64 V vs. SCE, which is more negative than the oxidation peak potential observed at 1.4 V because of the formation of a Zn(OH)2 layer on the surface of the aluminum alloy [40]. Because aluminum alloys have more negative open circuit potentials (OCPs)compared with that of pure aluminum (Al–In at 1.27 V and Al–Sn and Al–Sn–In at 1.45 V compared with Al at 0.87 V), corrosion products such as metal oxides that form on the surfaces of aluminum anodes protect them from further corrosion. After corrosion, a large amount of inhomogeneous pitting is observed on the Al–In surface, whereas a rough surface is found on the Al–Sn surface [33]. The polarization curves of Al–Sn and Al–Sn– In exhibit no passive potential region, revealing behavior different from the trend observed in an aluminum anode with an onset potential of 1.47 V caused by the enhanced adsorption of Cl ions. The OCPs of Al–Zn–Ga and Al–In–Ga in a chloride solution are more negative potential value, at 1.38 V and 1.73 V vs. SCE, respectively. The hydrogen evolution rate of Al–In–Ga is higher than that of Al–Zn–Ga for the same Ga content [41]. The current density of the anodic dissolution of Al–Sn–Ga alloys in 2 M NaCl is higher than 200 mA cm2 at an OCP of 1.5 V vs. SCE [42]. The formation of a thick crystalline layer on the surface of the anode causes an increase in the cell voltage [42]. Ferrando introduced the fabrication of a polymer composite or particulate aluminum using a polymer mixture of polyacrylic acid (PAA) and polymethyl methacrylate, which is then bonded by pressing it to an aluminum metal current collector. PAA has the ability to support ionic conduction via H2O molecules along its length, where as polymethyl methacrylate or polystyrene assists and isolates the PAA molecules [43]. A polymer composite aluminum anode with a conventional weight exhibits an average energy output of 154 W h kg1 and a very short operational life of 10.5 h; however,

4

M. Mokhtar et al. / Journal of Industrial and Engineering Chemistry 32 (2015) 1–20

such an anode with a mass of 3 g can produce an energy output of 240 W h kg1 with a longer operational life of 43 h [43]. Elango et al. [44] has been investigated the Al alloys with chemical composition is 97.7 wt.% Al–2.0 wt.% Mn–0.3 wt.% Mg which known as 57S grade aluminium. The anodic polarization of a 57S grade aluminum anode in an alkaline solution of 2 M NaOH containing 0.2 M zinc oxide is higher than the cathodic polarization, indicating a reduction in the corrosion rate caused by the formation of an oxide layer on the aluminum surface [44]. The best aluminium anode chosen to be used in the cell would be Al–Sn alloy with composition of 0.02% to 0.4% [36] which reduce alkali attack on anode surface while open circuit potential (OCP) proven to be more negative with 1.45 V [33] with no passive potential and enhancement of choride ion, Cl. Composite polymer aluminium anode proposed by Ferrando is also an interesting anode material with energy output produce is 240 W h kg1 for 43 h. Cathode material The air cathode is one of the essential components of an aluminum–air battery. Typically, the electrode for a normal battery is prepared by combining conductive additives (carbon) with a binder (polymer) and an active material into a slurry paste, which is then coated onto a metallic foil to form a composite electrode. Because oxygen from the air is reduced at the air cathode via an oxygen reduction reaction (ORR), an oxygen reduction catalyst must be combined with the composite electrode to effectively catalyze the ORR. Conventional metal–air batteries also contain a metal mesh at the cathode to support the composite electrode. The reason why metal–air batteries cannot operate efficiently is related to scientific and technological difficulties associated with air cathodes, such as pore clogging caused by the insolubility of carbonates in the alkaline media, which leads to overvoltage, an increase in the electrode resistance, a low effective surface area for adsorption on the catalyst and a relatively high cathode cost. Studies focusing on materials for the air cathodes of metal–air batteries and fuel cells have been performed for a long time. Thus, the best cathode material can be chosen and optimized based on the findings of aluminum–air battery design research. Most studies of catalyst and air cathode fabrication have been performed using Proton exchange membrane fuel cell (PEMFC), Li– air and zinc–air batteries; in principle, the results of these studies can also be applied to materials for Al–air batteries to enhance the air cathode and overall system efficiency. There is a need to seek out cathode characteristics that contribute to superior performance when incorporated into a metal–air cell. For example, smaller crystallite solids support faster charge transfer kinetics because they have a larger surface-area-to-volume ratio than do larger crystallite solids [45]. The focus of this review is on the technology for the fabrication of air cathodes using nanomaterials and nanostructured designs that have been proven to exhibit high oxygen reduction reaction (ORR) and oxygen evolution reaction (OER) activities, which enhance the potential of a cell and lower its overpotential. To date, however, there is still no low-cost fabrication technique available for this attractive type of cathode. Nanomaterial and nanotube technology Nanomaterial and nanotube technologies have been intensively studied for their superior ORR and OER activities, surface area per unit volume and cell potential when incorporated into fuel cells and metal–air batteries. (a) Single-wall carbon nanotubes (SWCNTs) are composed of single cylindrical tubes of graphite, as shown in Fig. 1(a).

(b) Multi-wall carbon nanotubes (MWCNTs) are formed in 3 types of structures, as shown in Fig. 1(b): (i) Concentric MWCNTs: cylindrical tubes of graphite (SWNTs) aligned on top of each other. (ii) Herringbone MWCNTs: piled-up, truncated graphene cones. (iii) Bamboo MWCNTs: periodic segments of walls blocking off an inner space. (c) Graphenes are single-atom layers of graphite, as shown in Fig. 1(c). (d) Crumpled graphenes, as shown in Fig. 1(d).

Nitrogen-doped carbon nanotubes, N-CNTs In N-CNTs, the inclusion of the nitrogen element in the p cloud of a network of carbon nanotubes disturbs its surface structure, thereby altering the reaction path to the nitrogen surface. Three types of nitrogen-functionalized groups have been identified, namely, pyridic-N, quaternary-N and pyridine-N oxide; however, not all species can be used as cathodic materials. The pyridic-N species is the most suitable nitrogen species for use in N-CNTs because of its location (at the edge of the graphite plane) and its bonding to two carbon atoms. The feasible nitrogen species structures are illustrated in Fig. 2(c) and (d). The ORR on NCNTs has been shown to follow 2 electron pathways of oxygen reduction to peroxide and a possible reduction to water via 4 electron pathways in a PEMFC. Bifunctional air cathode Bifunctional air cathodes are used to facilitate both the oxygen reduction reaction, ORR, and the oxygen evolution reaction, OER. The operating principles for both reactions are applied on the same electrode to achieve reversible and rechargeable metal–air batteries. The ORR requires the use of a porous gas-diffusion electrode with a hydrophobic catalyst, whereas the OER requires a porous electrode sintered on a metallic substrate with a hydrophilic catalyst [46]. The catalyst must be able to endure high-ORR and high-OER conditions to ensure a long operating time and improved performance of the cell. There are three construction principles for bifunctional air cathodes, as proposed by Jorissen; these are illustrated in Fig. 3. Selection of air cathodes The primary challenges associated with air cathode development are insufficient utilization of the anode as a result of severe passivation and corrosion of the electrolyte, sluggish kinetics of the air cathode reaction, poor OER activity and high overpotential. Platinum (Pt) is still the best state-of-the-art catalyst for use in air cathodes because platinum is highly stable and has a high electrocatalytic activity. Although air cathode performance is enhanced at a low Pt loading, it is still not feasible to use a Pt-based air cathode in an Al–air SMFC because of the scarcity and high cost of this material [47]. Air cathodes and catalysts are selected based on their cyclic voltammograms, polarization curves and Koutecky–Levich (K–L) plots. The latter two types of plots can be used to understand the electron transfer number and the pathway taken by the ORR. This review is focused on the performance of ORR and OER electrocatalysts in metal–air batteries and fuel cells. Typical catalysts such as Pt and La0.8Sr0.2MnO3, which are fabricated for use in the air cathode of a lithium–air battery by casting a mixture of a carbon collector, a catalyst and a co-polymer based on polyvinylidenedifluoride (PVDF),exhibit poor discharge capacities that decrease with an increasing number of cycles

M. Mokhtar et al. / Journal of Industrial and Engineering Chemistry 32 (2015) 1–20

5

Fig. 1. (a) SWNTs, (b) MWNTs (adapted and modified from [121,122]), (c) graphene, schematics, (d) crumpled graphene sheet configuration (adapted and modified from [123]) [2].

[48]. Both demonstrate a low capacity retention per cycle of below 1.3%. Notably, Fe3O4 is considered a better catalyst than Fe2O3 or NiO because of its greater capacity retention per cycle, despite its lower first-cycle capacity [48]. By contrast, Fe2O3 exhibits very poor performance, with the lowest capacity retention per cycle, despite having the highest first-cycle capacity [48]. The discharge

Fig. 2. Schematic representation of possible N–C bonding configurations with a nitrogen dopant atom inserted in a graphitic network: (a) substitutional N3sp2CC1N1; (b) pyridine-like N-2sp2CC1N1; (c) pyridine-like N-2sp2CC2N1; (d) bridge-like N-4sp3CC2N1; (e) N-3sp3CC2N1 [124].

and charge potentials are observed to be 2.6 V and 4.6 V, respectively. Although Co3O4 exhibits a good first-cycle capacity of 2000 mA h g1, an extremely high tenth-cycle capacity of 1300 mA h g1 and high capacity retention per cycle, its observed charge potential is 4 V, which is relatively quite low. Table 2 summarizes the discharge capacities of several candidate catalyst materials after various numbers of cycles. The cyclic voltammograms of air cathodes fabricated from highly active Pd-coated silver-nanoparticle-based carbon prepared using the galvanic displacement method in an alkaline solution exhibit a lower reduction peak of Ag2O, which is attributed to the Pd covering and blocking the Ag surface, whereas the reduction peak of Pd oxide is observed at 0.4 V and 0.1 V on a negative scan [49]. However, Pd coated on Ag/C does not yield any improvement in the ORR onset potential because its value is similar to that of Ag/C, at approximately 0 V [49]. In addition, after the coating of Ag with Pd, the half-wave potential shows a positive shift from 0.19 V to 0.12 V [49]. The low ORR can be attributed to the deposition of impurities in the electrolyte on surface of the Pd. Other methods of depositing Pd onto Ag surfaces and the use of anion exchange membranes (AEMs) to isolate the OH species could enhance the air cathode’s performance [49]. The cyclic voltammograms of air cathodes fabricated from multi-walled carbon nanotubes that are prepared using a floating catalyst chemical vapor deposition (CVD) system with ethylene as a carbon collector, melamine as a nitrogen source and ferrocene as a catalyst precursor exhibit a distinct ORR current peak at 0.13 V vs. Hg/HgO [50]. The ORR at the cathode follows 4 electron transfer reaction pathways, as indicated by the fact that the electron

6

M. Mokhtar et al. / Journal of Industrial and Engineering Chemistry 32 (2015) 1–20

Fig. 3. Construction principles for bifunctional oxygen/air electrodes [46].

transfer number is found to be between 3.70 and 2.99 in 0.1 M and 12 M KOH, respectively [50]. The poor ORR and OER performances of conventional electrodes using an alkaline medium, KOH, in which the potentials drop by as much as 0.86 V and 1.65 V, respectively, as the current decreases, have been attributed to trace CO2 leaking from the air and forming insoluble carbonates at the interface between the air cathode and the alkaline medium. The infiltration of CO2 gas into an aqueous alkaline medium can be prevented by using AEMs as separators and interfaces between the air cathode and the alkaline medium; such separators are formed by hot pressing the air cathode and the membrane together [51]. The ORR and OER potentials of such an AEM-air cathode are only slightly decreased, by 0.02 V and 0.03 V, respectively. The effect of CO2 is therefore reduced, with a low overvoltage and delayed deterioration of the electrolyte [51]. The ORR and OER potentials of an AEM-Pt/Ir-based carbon electrode decrease and increase by 0.16 V, respectively, upon an increase in the KOH concentration from 2 M to 4 M at 10 mA cm2 [51]. The Koutecky–Levich plots for air cathodes in zinc–air SMFCs fabricated from N-CNTs synthesized from ethylenediamine as a precursor via chemical vapor deposition (CVD) at 800 8C reveal that the ORR at the N-CNTs follows 4 electron pathways, similar to those of Pt, and that the first exchange transfer is the ratedetermining step. The measured Tafel slopes are 60 mV dec1 at low current density and 260 mV dec1 to 500 mV dec1 at high current density [52]. The oxygen reduction capability of a composite air cathode prepared by coating a conductive polymer and silver nanoparticle catalyst onto a carbon current collector (C-cp-Ag) via electrodeposition is double that of an uncoated glassy carbon electrode and compares favorably with those of Pt- and Au-based electrodes [45]. The positioning of the Ag catalyst nanoparticles on the surface of the composite electrode allows the oxygen gas molecules ready access to the catalyst surface. Composite air cathodes that are manufactured using this multiple coating method are also low in cost [45]. Air cathodes made of a nanoscale manganese oxide (MnOx) catalyst on carbon nanofoam paper (CNFP) prepared by infiltrating the CNFP with aqueous resorcinol-formaldehyde(RF) exploit the

high electrical conductivity, high specific gravity and threedimensional porosity of CNFP [53]. The onset potential for the ORR activity of an MnOx-CNFP cathode with macropores (>50 nm) is greater than 60 mV, representing superior performance to that achieved with mesopores (<50 nm) [53]. Hydrophobic MnOx carbon nanofoam can also be prepared via dipping in a PVDF solution in NMP to minimize electrolyte flooding and enhance electrocatalytic performance [53]. Both MnO2 and N-CNTs are already known to exhibit superior ORR activity, but N-CNTs alone have very poor OER activity. The OER activity of an air cathode can be increased if both MnO2/CNT and N-CNTs are mixed in ethanol and dried in an oven at a 1:1 ratio to form a composite bifunctional catalyst, MnO2/N-CNT [54]. The half-wave potential and OER activity of such an electrode are enhanced compared with those of individual reference electrodes fabricated from either MnO2/CNT or N-CNTs. The OER activity of the bifunctional catalyst cathode is higher than that of the N-CNTs, with a value of 1 V after 50 cycles, and the current density produced is also 7.5 times higher than that produced by an MnO2based catalyst, with a value of 25 mA cm2 at 0.55 V. The OER peak of another bifunctional catalyst, Ni/H3Mo12O40P, fabricated by integrating molybdophosphoric acid (H3Mo12O40P) with a Nafion ink solution at a ratio of 1:6 or with a nickelnanopowder-based carbon electrode at a ratio of 1:1 in an alkaline medium, is less intense and more negative than that of a Ni-based electrode at 0.27 V or 0.32 V, respectively [55]. The ORR cathodic peak observed at 0.21 V vs. SCE indicates a relatively low overpotential of the electrode. The improvement in the ORR and OER performance may be attributable to the presence of the PMo12O40 species in the electrode. The ORR and OER potentials of the electrode tested in 0.4 M KOH yielded inferior results, which may be attributable to hydrogen evolution at 0.53 V and 0.52 V, respectively, making the electrode infeasible for use in alkaline media because of the large potential difference [55]. An alternative bifunctional catalyst consisting of an ORR catalyst of CoO/N-CNT and an OER catalyst of Ni–Fe-layered double hydroxide/CNT for use in Zn–air batteries has been prepared by dispersing a transition metal oxide and oxidized MWCNT powder in a solvent, which then underwent a solvothermal reaction. The

M. Mokhtar et al. / Journal of Industrial and Engineering Chemistry 32 (2015) 1–20

7

Table 2 Discharge voltage and discharge capacities at cycles 1, 5, and 10 [48]. Catalyst

Discharge voltage (V)

Capacity of cycle 1 (mA hg1)

Capacity of cycle 5 (mA hg1)

Capacity of cycle 10 (mA hg1)

Capacity retention per cycle (%)

Pt La0.8Sr0.2MnO3 Fe2O3 Fe2O3–carbon loaded NiO Fe3O4 Co3O4 CuO CoFe2O4

2.55 2.6 2.6 2.6 2.6 2.6 2.6 2.6 2.6

470 750 2700 2500 1600 1200 2000 900 1200

60 75 500 280 900 1200 1900 900 900

60 40 75 75 600 800 1300 600 800

1.28 0.53 0.28 0.3 3.75 6.67 6.5 6.67 6.67

ORR potential of the CoO/N-CNT catalyst is 20 mV more negative than that of Pt/C, whereas the onset potential is greater than 20 mV at 100 mA cm2, representing superior performance to that of Pt/C [56]. However, although the Ni–Fe-LDH-based electrode exhibits a higher OER activity compared with the latter, the onset potential of Ni–Fe -LDH/CNT is 20 mV more negative than that of Ir/C at a current density of 50 mA cm2 [56]. This bifunctional catalyst is attractive as an air electrode material because of its long-term stability, with charge-discharge cycles of up to 20 h. A novel bifunctional catalyst consisting of the solid-state electrolyte Li1.35T1.75Al0.25P2.7Si0.3O2 (LTAP) and a neoteric crosslinked gel (CNG) can produce a charge density of 19,050 mA h g1 when used as an air cathode in a rechargeable lithium–air battery [57]. The high number of active sites in the SWNTs and the ionic liquid in the CNG provide facile electron passage. The discharge of a Li–air cell is known to be facilitated by the highly reversible O2 species process, and the long discharge time in an ambient air environment causes most of the O2 to form CO3 [57]. All-solidstate materials of this type can be readily used in an Al–air cell [57]. A novel nitrogen-doped graphene air cathode catalyst, N–Fe– CNT/CNP, fabricated using iron acetate (Fe(CH3COO)2) as an iron precursor and cyanamide (NCNH2) as a nitrogen precursor, has been found to outperform a Pt catalyst in alkaline media because of the more favorable O2 adsorption sites created by the nitrogen disrupting the graphene structure and the CNPs hindering linear CNT growth [58]. The half-potential of the resulting air cathode at a loading of 0.2 mg cm2 is 0.87 V, which is slightly lower than that of a Pt/C catalyst air cathode at a loading of 60 mg cm2 (0.97 V) [58]. Because manganese oxide catalysts are known to exhibit high ORR activity, are cost-effective compared with noble metal catalysts and have superior physicochemical properties, a- and d-manganese oxide catalysts have been prepared via a simple reaction with KMnO4 and used in the air cathodes of zinc–air SMFCs; the resulting air cathodes were found to produce higher ORR currents of 66.08 A and 62.21 A, respectively, compared with those fabricated using commercial g-manganese oxide [59]. Both catalysts demonstrate superior performance to that of g-manganese oxide because the crystal structure of the latter does not contribute to the ORR activity. The results of this study also

reveal that the ORR activity of a-MnO2 is greater than that of dMnO2 [59]. The different pore structures and sizes of the carbon materials used as cathode materials for lithium–air batteries, such as carbon aerogels ranging in size between 10 mm and 100 mm and activated carbon particles ranging in size between 5 mm and 20 mm, affect the battery performance [60]. The discharge capacity that can be achieved using carbon aerogels is 4155 mA h g1, greater than that for activated carbon (264 m A h g1). Despite having a smaller surface area of 688 m2 g1 compared with the 1908 m2 g1 surface area of activated carbon, carbon aerogel has a larger pore volume (2.53 cm3 g1) compared with that of activated carbon (0.55 cm3 g1) and can therefore accommodate a larger volume of discharge product [60]. The discharge capacity of a cathode fabricated using a dual cast (DC) soaked electrolyte, when operated with oxygen, increases with increasing current density compared with the discharge capacity of a cathode fabricated using a separate cast (SC) soaked electrolyte [61]. When operated without oxygen, cathodes fabricated using embedded DC electrolytes exhibit higher capacity retention compared with soaked DC cathodes because of the improved distribution of the electrolyte within the cathode and the decreased pore flooding of the electrolyte [61]. However, the discharge capacity of a calendered PTFE cathode in a dimethyl ether (DME)based electrolyte has been shown to be superior to that of an embedded DC cathode at a discharge current greater than 1 mA cm2 [61]. Moreover, RTIL DC cathodes exhibit a very poor discharge capacity of 500 mA h g1 at 0.01 mA cm2. Tafel polarization analyses of a calendered PTFE cathode and an embedded DC cathode reveal quite similar performances, despite the fact that no irreversible product is produced by the latter in carbonate media but not in a lithium–air battery system [61]. The low-current region observed at approximately 119 mV indicates that both the main ORRs have values near 1.0, thus favoring a one-electron exchange reaction [61]. Table 3 summarizes the discharge current (mA h g1 carbon) of cathodes using various operating gases and discharge currents. The onset potential of FeAgMo2O8, an OER catalyst for alkalibased metal–air batteries prepared via the spray drying and calcination of mixtures of Fe(NO3)3, AgNO3 and (NH4)6Mo7O24

Table 3 Discharge capacities (mA h g1 carbon) of cathodes using various operating gas and discharge currents [61]. Cathode

Operating gas

Soaked SC Soaked DC Embedded DC Soaked DC Embedded DC Embedded DC

O2 O2 O2 Air, no flow Air, no flow Air, flow

Calendered PTFE

O2

Discharge current (mA cm2) 0.1 1680 2160 2330 1620 1530 – 0.1 1100

0.4 770 1330 1340 370 940 1380 1 1300

0.7 570 880 740 170 450 690 2 600

1 430 630 570 60 260 530 4 200

8

M. Mokhtar et al. / Journal of Industrial and Engineering Chemistry 32 (2015) 1–20

aqueous solutions, is 1.54, which is lower than the potential at the reference catalyst (FeWO4), namely, 1.65 V [62]. The discharge potential increases from 1.05 V to 1.165 V after 10 cycles, whereas the charging potential decreases from 2.1 V to 2.05 V [62]. The half-wave potential for the ORR in acidic electrolytes at Pt/ TiO2-functionalized graphene sheet (Pt/TiO2–FGS) electrocatalysts prepared by impregnating TiO2-nanoparticle-functionalized graphene sheets into Pt catalysts is 0.83 V; this is higher than the potential at the reference catalyst, a Pt/C-modified GC electrode, which is 0.80 V [63]. By contrast, the half-wave potential of this catalyst in alkaline media, 0.72 V, is higher than the potential of bulk Pt by 60 mV [63]. The Pt/TiO2-FGS cathode also exhibits a higher specific activity than does the Pt/C cathode [63]. The discharge potential of a bifunctional catalyst, Pt/IrO2, in which an ORR catalyst (Pt) is combined with an OER catalyst (IrO2)through thermal oxidation, is similar to that of a reference Pt/ C air cathode at 3.5 V, indicating that the IrO2 does not participate in ORR activity. However, the charging potential of this catalyst is improved to 4.5–4.6 V compared with the 4.25 V observed for the Pt/C cathode, leading to the formation of a thinner oxide layer on the surface of the Ptas a result of reduced oxidation and a lower OER overpotential [64]. The onset potential for a Pt catalyst supported on 60% tungsten carbide (WC) is 0.06 V, which is slightly higher than those for 20% WC–Pt and Pt/C catalysts (0.04 V and 0.01 V, respectively) [65]. Both the 60% and 20% WC catalysts show a positive shift compared with Pt/C, indicating a high level of ORR activity in alkaline media. The ORR current densities for Pt/WC and Pt/C catalysts are very similar [65]. After the first charge-discharge cycle, the ORR activity of a threedimensional composite air cathode (3D Cfelt-cp-Ag), consisting of a carbon felt collector, a conductive polymer and a low-loading Ag coating of 0.08 mg cm2, in a non-aqueous medium is four-fold higher compared with that of a planar C-cp-Ag electrode [66]. However, the peak Coulomb flux for the 3D electrode decreases from 0.59 C in cycle 1 to 0.36 C in cycle 2 because of the faster ORR rate in the 3D electrode compared with that in the planar electrode [66]. The O2 reduction peak of a silver-supported Co3O4-based carbon electrode, Ag–Co3O4/C, prepared by dispersing hydrothermally produced Co3O4/C into a colloid mixture of KBH4, trisodium citrate, water and AgNO3, is positively shifted compared with that of Ag/C, indicating high ORR catalytic activity. The K–L plot for this catalyst shows that the electron transfer number is approximately 4.0 at potentials between 0.05 V and 0.2 V, following a 4-electron transfer pathway, whereas the Ag/C result indicates a 2-electron transfer pathway to the production of peroxide [67]. The ORR and OER peaks of an alternative bifunctional catalyst, Sm0.5Sr0.5CoO3(SSC), are similar to those of a perovskite catalyst, LCC, at 0.65 V and 0.86 V, respectively [68]. Although the SSC catalyst exhibits a similar electrocatalytic performance to that of the LCC catalyst, a cathode containing SSC has been found to have a longer life of 106 cycles compared with 68 cycles for an LCC cathode at 0.3 V [68]. Hence, SSC is a superior bifunctional catalyst compared with LCC [68]. Furthermore, a new fabrication of bifunctional air cathode using ORR catalyst, CoO/N-CNT and OER catalyst, Ni–Fe-layered double hydroxide/CNT with onset potential 20 mV more positive than Pt/C and 20 mV more negative than Ir/C during charging chosen as the best air cathode [69]. MnO2/N-CNT also shows excellent results as bifunctional catalyst [54]. Electrolytes Aluminum metal used as the anode in the Al–air batteries is known to react violently with acidic electrolyte producing

hydrogen and huge amount of heat. This causes rapid anode corrosion and complex thermal management. The performance of electrolyte is observed using the Electrochemical Impedance Spectroscopy, EIS used to investigate cathode and electrolyte interface behavior. An interception with real axis at highest frequency shows resistance value that indicates bulk resistance and cathode and collector resistances. Cyclic voltammetry also observed in the electrolyte. Aqueous electrolytes Alkaline electrolyte materials are commonly used in aqueous– electrolyte-based metal–air batteries because the ORR is more favorable in an alkaline electrolyte, with faster reaction kinetics and a lower overpotential compared with acidic electrolytes. Aqueous electrolytes are used because of their low cost, widespread availability and high ionic conductivity. The aqueous electrolytes that are most commonly used in aluminum–air batteries are sodium chloride, sodium hydroxide, potassium hydroxide, brine solution and saline solution. The tendency for hydrogen evolution to occur at the cathode hinders the further oxidation of the aluminum. A major drawback of alkaline aqueous electrolytes is carbonate precipitation due to the reaction between carbon dioxide and hydroxide. This causes the clogging of the electrode pores because of the insolubility of carbonates in alkaline aqueous electrolytes. Sidereaction :

CO2 þ OH ! CO3 2 þ H2 O

(1.6)

Aqueous electrolytes can be categorized based on their pH scale as alkaline solutions (7 < pH  13), neutral salt solutions (pH = 7) or acidic solutions (2  pH < 7). Acidic solutions Aqueous electrolytes with a pH below 7, in the acidic range, have been used since the 1990s. Alloying pure Al with other active metal elements enhances the performance of an aluminum–air battery with an acidic electrolyte. Saidman and Bessone [27] deposited In onto a pure Al surface to form an Al–In alloy and used 0.5 M chloride acid as the aqueous electrolyte solution. The reaction proceeded at room temperature, and the results revealed that the activation process depends on the concentration of In3+ ions. At higher In3+ concentrations, for potentials more positive than 1.50 V, the electroreduction of In3+ can occur, whereas the electroreduction of InO2 ions on quasi-bare Al occurs at potentials more negative than 1.50 V. In addition to being alloyed with indium, Al also can be alloyed with silicon (Si). Mazhar et al. [70] reported that the results of a study using a polarization technique indicated that Al alloys with higher Si contents yield more negative potential values, and upon increasing the acid concentration, the potential values were found to shift even more negative, to between 0.8 V and 1.0 V. Recently, Flamini and Saidman [41] alloyed Al with Zn and investigated the resulting performance using various types of acids at the same concentration, pH and operating temperature. Tafel plots showed that the potential value for the Al–Zn alloy in 0.5 M chloride acid was more negative than that in 0.5 M acetic acid: 1.02 V and 0.80 V, respectively. Furthermore, the development of ternary alloys of Al for use in acidic electrolytes has also been studied. Munoz et al. [71] and Ma et al. [72] studied Al–Zn–In alloys, Flamini and Saidman [41] studied Al–Zn–Ga and Al–In–Ga alloys, and Ma et al. [73] studied Al–Zn–Mg alloys. Al–Zn–In alloys exhibit potential values between 0.9 V and 1.1 V in 0.5 M chloride acid, whereas in 1 M chloride acid, their potential is approximately 0.7 V [71,72]. Based on the obtained

M. Mokhtar et al. / Journal of Industrial and Engineering Chemistry 32 (2015) 1–20

9

Table 4 Summary of acidic electrolytes for aluminum–air batteries. Electrolytes

Concentration (M)

Operation temperature (K)

Anode

Technique

Result/performance

References

Chloride acid pH 3

0.5

298

Al–In

Potentiostatic and potentiodynamic techniques, SEM

[27]

Chloride acid pH 2

0.5

303

Pure Al, Al–7%Si, Al–11%Si, Al–22%Si

Chemical, polarization and EIS measurements, SEM

Chloride acid pH 3

0.5

Room

Al–Zn–Ga and Al– In–Ga

EDX analysis, SEM

Acetic acid pH 3

0.5

Room

Al–Zn–Ga and Al– In–Ga

EDX analysis, SEM

Chloride acid pH 5

0.5

298

Ternary alloy Al– 5%Zn–0.02%In Al–4.1%Zn– 0.02%In–0.09%Si

–

Chloride acid pH 4

0.6

Room

Al–(Zn–5.02% Mg– 0.9%% Mn–0.44% Si–0.06% Fe– 0.001%Ti–0.04% Cu–0.03% In–0.02% C–0.05%)

Measurements of selfcorrosion, potentiodynamic polarization, cyclic polarization experiment combined with OCP technique and SEM

HCl

1

Room

Al–(Zn–5.02% Mg– 0.9%% Mn–0.44% Si–0.06% Fe– 0.001%Ti–0.04% Cu–0.03% In–0.02% C–0.05%)

HCl

1

298

Al–(Zn–5.02% Mg– 1.01% In–0.02% Si– 0.09% Fe–0.001%Ti– 0.05% Cu–0.02%)

Measurements of selfcorrosion, potentiodynamic polarization, cyclic polarization experiment combined with OCP technique and SEM Cyclic polarization, selfcorrosion and SEM

Activation process depends on the amount of In deposited at the bare Al surface, the actual In3 + concentration, the Cl-concentration and the local pH Capacitive behavior of the oxide covered surface is replaced by resistive behavior as immersion time increases in HCl solutions and the pitting by chloride ions initiates more readily in acidic media potential value: shift to the negative side (between 0.8 to 1.0 V) Saturated calomel electrode (SCE) Al– Zn–Ga alloys in chloride solution toward more negative values Al–4.0 wt.%Zn: 1.02 V Al–4.0 wt.%Zn–0.5 wt.%Ga: 0.97 V Al–4.0 wt.%Zn–2.5 wt.%Ga: 1.26 V Al–4.0 wt.%Zn–5.0 wt.%Ga: 1.27 V SCE Al–In–Ga alloys in chloride solution varies between 1.62 and 1.73 V. Al–0.2 wt.%In–0.5 wt.%Ga: 1.64 V Al–0.2 wt.%In–2.5 wt.%Ga: 1.78 V Al–0.2 wt.%In–5.0 wt.%Ga: 1.78 V Saturated calomel electrode, SCE (V): Al–4.0 wt.%Zn: 0.80 Al–4.0 wt.%Zn–2.5 wt.%Ga: 0.92 Al–4.0 wt.%Zn–5.0 wt.%Ga: 1 The presence of In in true electric contact with Al and Zn promotes Cl adsorption at potentials more positive than 1.1 V, considered as the pzc of such interface Alloy undergoes two types of localized corrosion process: hemispherical and crystallographic pitting Hemispherical pits: on the surface of this material under a simply exposure in chloride solution Crystallographic pits: to polarize the alloy The open circuit potential = 1.0 V The open circuit potential = 0.57 V

results, it can be concluded that in acidic solutions, these alloys suffer a low degree of corrosion. For Al–Zn–Ga and Al–In–Ga alloys, the potential values in 0.5 M chloride acid become more negative as the Ga content in the alloys is increased. The reported potentials are between 0.97 V and 1.27 V for Al–Zn–Ga alloys and between 1.64 V and 1.78 V for Al–In–Ga alloys [41]. These authors also tested the same alloys in acetic acid and found the potential values to be below 1.0 V [41]. Ma et al. [73] recently investigated the potentials of Al–Zn–Mg alloys. The reported open circuit potential values are between 0.6 V and 1.0 V for a chloride acid solution with the reaction proceeding at room temperature and a pH range of 1 to 4. The cited report also states that this alloy is subject to a fairly low degree of pitting corrosion. The performances of acidic electrolytes for aluminum–air batteries are summarized in Table 4.

The alloy undergoes pitting in acidic solutions and the open circuit potential = 0.70 V

[70]

[41]

[71]

[73]

[72]

Neutral salt solutions The effectiveness of neutral salt solution electrolytes has been studied since the early 1970s. The majority of researchers have reported that the potential values for pure Al are between 0.65 V and 1.1 V when it reacts in sodium chloride (NaCl) solutions [25,29,31,33,36,74]. The results indicate that the potential values of Al depend on the concentration of the NaCl solution and the operating temperature. The potential increases as the temperature and the concentration of the NaCl solution are increased [25,33,74]. Similar to their performance in acidic solutions, binary and ternary alloys of Al are also beneficial for enhancing the performance in aluminum–air batteries with neutral salt solution electrolytes [25,28,29,31–33,36,72,73,75–80]. Despic´, Drazˇic´, Purenovic´ and Cikovic´ [28] have reported potential values of Al–In alloys in neutral salt solutions of between

M. Mokhtar et al. / Journal of Industrial and Engineering Chemistry 32 (2015) 1–20

10

Table 5 Summary of neutral salt electrolytes for aluminum–air batteries. Electrolytes

Concentration (M)

Operation temperature (K)

Anode

Result/performance

References

NaCl

2

298

Pure Al (99.999%)

[125]

NaCl

–

–

Al–In and Al–Tl (up to 0.2%)

Neutral salt

–

–

Ternary alloy, Al–0.01 In–0.01 Ga

NaCl

0.6

–

Al, Al–In and Al–Ga–In alloys

NaCl

0.5

298

Pure Al, Al–0.02% Sn, Al–0.09% Sn, Al–0.20% Sn, Al–0.40% Sn

NaCl

2

293

Al (98%): Al–0.1% In, Al–0.2% Sn and Al–0.1% In–0.2% Sn

NaCl

2

–

Al (purity 99.8%)

Average discharge voltage: 0.399 V Capacity density: 2752 mA h g1 Energy density: 1096 W h kg1 The potential values of Al–In alloys in neutral salt solution were between 1.4 V to 1.7 V Corrosion rate of the alloys in neutral salt solutions is also decreased compared to that of pure aluminium The result exhibited a more negative rest potential than the Al–In alloy and a corrosion stability superior to that of the Al–Ga alloy. The negative difference effect was found to depend on the cation of the neutral salt in solution and the lowest effect was obtained in ammonium chloride solutions TH result obtained showed the Al–In alloy exhibits the highest negative open circuit potential in 0.6 m NaCl and the corrosion resistance of the tested electrodes decreases in the following order: Al > Al– Ga–In > Al–In Al SCE:  0.8 V Al–Ga–In SCE: 1.05 V Al–In SCE: 1.20 V The charge value decreases with the increase of the Sn content in the alloy. Based on the anodic current– time responses shape and the charge values indicate that the longer stay at the cathodic potential activates alloys with 0.20% an 0.40% Sn for anodic dissolution The Al–In alloy exhibiting the most remarkable characteristics and the addition of In as alloying component to aluminum reduces electrode polarization, decreases hydrogen evolution rate and increases the anode efficiency SCE: 0.87 V Smooth surface without any damages SCE: 1.45 V Rough morphology that is consistent with strong dissolution SCE: 1.27 V Smooth surface with large number of randomly distributed pits SCE: 1.46 V Rough surface with a metallic luster OCP (mV) vs SCE = 800 OCP (mV) vs SCE = 820 OCP (mV) vs SCE = 890 OCP (mV) vs SCE = 1530 OCP (mV) vs SCE = 850 OCP (mV) vs SCE = 1510 OCP (mV) vs SCE = 800 OCP (mV) vs SCE = 1500 OCP (mV) vs SCE = 1530 OCP (mV) vs SCE = 1530 The limitation on the extent of constant current dissolution did not arise from changes to the electrolyte composition or to the formation of a protective surface anodic film. The cell can discharge at 0.29 A for 140 h with the working voltage keeping over 1.1 V, the utilization ratio of aluminum anode is over 44%, and the life of battery is longer than 2400 h The potential of corrosion of the sample, 0.760 V, is close to that of the pitting nucleation, 0.720 V Localized corrosion: hemispherical and crystallographic pits Both alloys suffer from pitting corrosion and the corrosion of 7A09 is much more serious than 3A21. The EIS tests revealed that the high corrosion rate for 7A09 alloy and good corrosion resistance for 3A21 alloy

Al–0.2% Sn

Al–0.1% In

Al–0.2% Sn–0.1% In NaCl

2

293

NaCl

2

293

NaCl

3.5

293

Aluminum alloy doped with Ga, In, Sn, Bi, Pb and Mn

NaCl

3.5

303

Alloy AA5083 (Mg–4.9% Mn– 0.5% Si–0.13% Fe–0.3%Ti–0.03% Cu–0.08% Cr–0.13%)

NaCl

3.5

–

NaCl

0.6

Room

3A21 (Mg–0.05%% Mn–1.3% Si– 0.6% Fe–0.7%Ti–0.15% Cu–0.2% Zn–0.1%) 7A09 (Mg–2.5% Mn–0.15% Si– 0.5% Fe–0.5%Ti–0.1% Cu–1.6% Cr–0.23% Zn–5.5%) Al–(Zn–5.02% Mg–0.9%% Mn– 0.44% Si–0.06% Fe–0.001%Ti– 0.04% Cu–0.03% In–0.02% C– 0.05%)

99.99% Al Al–3% Mg Al–5% Mg Al–0.5% Mg–0.1% Sn–0.05% Ga Al–0.4% Mg–0.1% Sn Al–0.4% Mg–0.4% Sn–0.03% Ga Al–0.4% Mg–0.03% Ga Al–0.1% Sn–0.03% Ga Al–0.6% Mg–0.1% Sn–0.05% Ga Al–0.4% Mg–0.07% Sn–0.05% Ga Al + (0.4 wt.% Mg, 0.07 wt.% Sn and 0.05 wt.% Ga)

The open circuit potential = around 1.0 V Localized corrosion: hemispherical and crystallographic pits

[28]

[31]

[36]

[32]

[33]

[25]

[78]

[77]

[76]

[126]

[73]

M. Mokhtar et al. / Journal of Industrial and Engineering Chemistry 32 (2015) 1–20

11

Table 5 (Continued ) Electrolytes

Concentration (M)

Operation temperature (K)

Anode

Result/performance

References

NaCl

0.6

298

Al–(Zn-5.02% Mg–1.01% In– 0.02% Si–0.09% Fe–0.001%Ti– 0.05% Cu–0.02%)

[72]

NaCl

2

298

NaCl

0.5 1 2 3 4 5 6 7 0.5

303

Al–(Mg–0.5% Ga–0.02% Sn–0.1%) Al–(Mg–0.5% Ga–0.02% Sn–0.1% Mn–0.5%) Al Al Al Al Al Al Al Al Al Al (99.999%; 5N) + 0.1%Tl + 0.1%In

This alloy undergoes relative light pitting in neutral solutions and the alloy exhibits a notable corrosion resistance in neutral chloride solutions (validated by Rp and Icorr measurement) The open circuit potential = 0.9 V OCP (V) = 1.185 OCP (V) = 1.236

Seawater Seawater

Seawater

293

Salinity of about 35 g L1

296

333 296

AA1100 (Mg–0.002%% Mn– 0.004% Si–0.15% Fe–0.51%Ti– 0.02% Cu–0.06% Zn–0.002% Cr– 0.001%) AA5083 (Mg–4.25%% Mn–0.61% Si–0.14% Fe–0.37%Ti–0.02% Cu– 0.01% Zn–0.01% Cr–0.17%)

333

1.4 V and 1.7 V, and these alloys exhibit decreased corrosion rates compared with that of pure Al. El Abedin and Endres [31] also investigated such an alloy in 0.6 M NaCl solution and observed a potentials of approximately 1.2 V. Gudic´, Smoljko and Klisˇkic´ [33] and Smoljko et al. [32] have tested Al–In alloys in 2 M NaCl solution and found that these alloys exhibit a random distribution of pitting corrosion and a potential value of approximately 1.3 V [33]. These authors also reported that these alloys not only offer increased anode efficiency but also help to reduce the rate of hydrogen evolution [32]. Another excellent alloy combination is Al–Sn. Gudic´, Radosˇevic´, Smoljko and Klisˇkic´ [36] investigated the efficiency of Al–Sn alloys in 0.5 M NaCl solution for varying Sn compositions and concluded that the most effective alloys in terms of anodic dissolution were those containing 0.20% and 0.40% Sn [36]. El Shayeb, Abd El Wahab and Zein El Abedin [29] also studied the performance of Al–Sn alloys and observed potential values of approximately 0.98 V in 0.6 M NaCl solution. In 2 M NaCl solution, the potential value of this alloy shifts more negative to approximately 1.45 V, and strong anode dissolution occurs, which reduces the electrode polarization [32,33]. Al–Mg and Al–Zn alloys have also has been studied. The potentials of Al–Mg alloys in 2 M NaCl solution are between 0.8 V and 0.9 V, with higher potentials corresponding to higher Mg contents [25]. Al–Zn alloys also exhibit an elevated potential value of 1.14 V when tested in 0.6 M NaCl solution [29]. Gudic´, Smoljko and Klisˇkic´ [33] and Smoljko, Gudic´, Kuzmanic´ and Kliskic´ [32] also investigated the performance of ternary Al alloys (Al–In–Sn) in 2 M NaCl solution, obtaining a potential value of approximately 1.46 V and observing rough surface corrosion. Subsequently, Al–In–Ga alloys were studied by Despic´, Drazˇic´, Purenovic´ and Cikovic´ [28] and by El Abedin and Endres [31]. These authors revealed that these ternary alloys exhibit potentials that

OCP (V) = 0.66 OCP (V) = 0.67 OCP (V) = 0.65 OCP (V) = 0.75 OCP (V) = 1.1 OCP (V) = 0.92 OCP (V) = 0.81 OCP (V) = 0.74 OCP (V) = 0.68 The ternary alloy Al(T)–0.1% In–0.1% Tl is uniformly dissolved in sea water and has a potential of 0.9V (SCE) in the current density region of 1–10 mA cm2. SCE corrosion potential (V) 920

[127]

[74]

[81]

[82]

1160 950

1185 The breakdown potential of the two alloys decreased with an increase in test temperature with better corrosion resistance for alloy 1100

are more negative than those of pure Al, although slightly less so than those of Al–In alloys. Among ternary alloys that include Al–Zn, El Shayeb et al. [29] investigated the performance of Al–Zn–Sn alloys and Ma et al. [72,73] investigated the performance of Al–Zn–Mg and Al–Zn–In alloys. Based on the results thus obtained, it can be concluded that these alloys can be ranked in order of increasing potential as follows: Al–Zn–In < Al–Zn–Mg < Al–Zn–Sn. Nestoridi et al. [25] studied the performances of Al–Mg–Sn and Al–Mg–Ga alloys in 2 M NaCl solution, and the results show that the potential increases with increasing Mg content. These authors also reported that the potential of the Al–Mg–Sn alloy was slightly higher than that of the Al–Mg–Ga alloy. They also studied an Al– Sn–Ga alloy and found that this alloy exhibited a higher potential than that of either the Al–Mg–Sn or Al–Mg–Ga alloy. Another neutral salt solution that is a promising electrolyte candidate is seawater. Akmal et al. [74] have reported that the open circuit potential of pure Al in seawater at 303 K is 0.68 V. For the aluminum–thallium–indium (Al–Tl–In) alloy studied by Mance, Cerovic´ and Mihajlovic´ [81], a potential value of approximately 0.9 V was observed. Furthermore, this potential was found to shift to a more negative value as the temperature of the seawater solution was increased. Using a weight loss technique, Ezuber et al. [82] revealed that Al alloys undergo pitting corrosion and that the corrosion intensity increases with increasing operating temperature. Neutral salt electrolytes have been widely used in aluminum– air batteries, and their performances are summarized in Table 5. Alkaline solutions Two types of alkaline solutions are typically used in aluminum– air batteries: potassium hydroxide (KOH) and sodium hydroxide (NaOH). The proton conductance in alkaline solutions has been widely studied for many years.

12

M. Mokhtar et al. / Journal of Industrial and Engineering Chemistry 32 (2015) 1–20

Wilhelmsen et al. [34] used a weight loss technique to study the rate of anode corrosion in 4 M KOH and claimed that the addition of other active metal elements helped to improve the corrosion resistance of the anode. Chu and Savinell [83] also investigated KOH solutions at the same operating temperature, 333 K, and reported that there was no mass transfer effect, although the KOH concentration and temperature had a significant effect on the polarization characteristics. Doche et al. [84], Tang et al. [40] and Choi et al. [85] investigated the rate of anode corrosion in 4 M NaOH. Doche, Rameau, Durand and Novel-Cattin [84] used a steady-state technique to identify the corrosion dissolution of pure Al, and the delineated polarization curves indicated that pure Al exhibits a passive state. Using polarization techniques, Tang et al. [40] and Choi et al. [85] studied the corrosion potential of pure Al and observed a potential value of approximately 1.5 V. The effectiveness of alkaline solutions as electrolytes for Al alloy anodes, in addition to pure Al anodes, has also been studied since the late 1980s [86]. Choi et al. [85] investigated the corrosion potential of Al alloys in 4 M NaOH and claimed that the presence of In, Sn, Ga and/or Mg helped to decrease the corrosion rate. However, the use of Fe was not recommended by the authors because only a slight reduction in the corrosion rate was observed. Wilhelmsen et al. [34] used a weight loss method to study the corrosion rate of Al–In alloys and reported a similar conclusion. Upon the addition of 0.1% by weight of In, the anode corrosion resistance greatly improved. Al–Zn–In and Al–Zn–Mg alloys have also been studied by Ma et al. [72,73] in 4 M NaOH. The reaction proceeded at room temperature, and the potentials were observed to be approximately 1.65 V and 1.67 V, respectively. Al–Zn–Mg alloys undergo hemispherical and crystallographic pitting corrosion, whereas Al–Zn–In alloys suffer only crystallographic pitting corrosion. As indicated by the results obtained, these alloys undergo an intensive corrosion process in alkaline solutions as a result of their chemical dissolution by OH. Recently, Gao et al. [87] investigated the effectiveness of bismuth (Bi) when alloyed with an Al–Mg–Sn–Ga alloy. Using a gas-collecting method, the authors investigated the self-corrosion of this alloy in 4 M NaOH and found that the potential shifted more negative when the Bi content was increased. Furthermore, Wang et al. [80] have implemented a promising hybrid concept to solve the problem of parasitic hydrogen evolution simply by using ordinary kitchen aluminum foil, achieving an open circuit potential of approximately 1.45 V. Analysis of this system also revealed that the power density could be increased by increasing the concentration of the alkaline solution. Table 6 summarizes the alkaline electrolytes whose use in aluminum–air batteries has previously been reported. Non-aqueous and aprotic electrolytes Non-aqueous solutions Non-aqueous and aprotic electrolytes are used in Li–air batteries because aqueous electrolytes cannot be used in batteries of this type. This technology can also be used to manufacture compact portable aluminum air batteries by exploiting its high energy density. There are 4 major types of aprotic electrolytes that are commonly used in lithium–air batteries: organic carbonates, ethers, ionic liquids and solid-state electrolytes. Table 7 summarizes the electrolyte materials used in lithium–air batteries. Aprotic electrolytes in lithium–air batteries typically exhibit one-electron ORRs that yield superoxide ions, O2, which eventually form Li2O2. This cell reaction produces a very high energy density that tends to cause the electrolyte materials to decompose and clog the air cathode, which can limit the discharge

capacity [88]. Moreover, the discharge products of aprotic electrolytes in lithium–air batteries have a low conductivity and high electrical resistance. Additives such as Li, K and tetrabutylammonium (TBA) salt can be mixed with the electrolyte to help dissolve the oxygen reduction products and increase the discharge capacity of the battery. However, there was no clear indication regarding whether aprotic electrolytes could be used in aluminum–air batteries until it was recently reported that an EMImCl/ AlCl3 acidic solution had been developed as an innovative electrolyte for aluminum–air batteries, exhibiting current densities of up to 0.6 mA cm2 [89]. In another study by Gelman et al. [90], it was reported current densities of up to 1.5 mA cm2 could be achieved by using EMIm(HF)2.3F as the electrolyte. Despite these findings, there is still a great deal of uninvestigated potential in the field of non-aqueous electrolyte research for aluminum–air batteries. Aprotic solvents Considering that aprotic–solvent-based electrolytes are widely used in Li–air batteries, similar electrolytes could be used in the proposed Al–air system until further studies have been performed. This type of electrolyte is used to prevent hydrogen evolution due to parasitic reactions from degrading the electrode. The conventional aprotic–solvent-based electrolytes that are used in these cells are carbonates, ethers and esters. Typically, when this type of electrolyte is used, the ORR mechanism favors the one-electron reduction of oxygen to the superoxide ion, which then undergoes another one-electron reduction to the peroxide ion. This ORR can only proceed if an alkali metal such as Li is used as the anode material. Aprotic-solvent-based electrolytes face the same challenge as do aqueous electrolytes, namely, the formation of carbonates that consume the electrolyte and plug the electrode pores. The ORR onset potential of an aprotic ether-based electrolyte, PYR14TFSI–TEGDME–LiCF3SO3, prepared by dissolving LiCF3SO3 in tetraethylene glycol dimethyl ether (TEGDME) and mixing with pure PYR14TFSI, at a Super P-based carbon air cathode is 2.8 V vs. Li+/Li, whereas its OER onset potential is 3.5 V vs. Li+/Li [91]. The lifetime of such a battery is only 6 charge-discharge cycles, as the electrolyte decomposes. Electrochemical impedance spectroscopy (EIS) testsindicate that the PYR14TFSI-TEGDME-LiCF3SO3 electrolyte has a low bulk resistance and a small charge transfer resistance [91]. Solid-state electrolytes Solid-state electrolytes are an attractive type of electrolyte that can be used to fabricate compact Al–air batteries with mechanically interchangeable anodes for cell phones and laptop computers. The current capacity of one type of alkaline solid-state electrolyte, a PEO–PVA–KOH polymer membrane prepared by solvent casting PEO, PVA and KOH, in a Zn–air battery is 1305 mA h, which is close to the theoretical capacity of 1560 mA h [92]. The bulk resistance of this electrolyte is on the order of 1–5 V, and its ionic conductivity can be as high as 102 S cm1 [92]. The reduction and oxidation current densities in Zn–air or Al– air batteries with composite hydrophilic solid polymer electrolyte membranes (SPEMs), prepared via the impregnation of a nonwoven polypropylene/polyethylene (PP/PE) membrane with a highly ion-conducting solution of cast polyvinyl alcohol/polyacrylic acid (PVA/PAA), increase with increasing PAA content. The membrane’s compatibility with a Zn–air cell lasts only up to 12 cycles, and its compatibility with an Al–air cell steadily decreases over 12 cycles. The bulk resistance of the electrolyte is in the range of 0.1–0.4 V, and its highest ionic conductivity of 0.301 S cm1 is observed at a PVA/PAA

M. Mokhtar et al. / Journal of Industrial and Engineering Chemistry 32 (2015) 1–20

13

Table 6 Summary of alkaline electrolytes for aluminum–air batteries. Electrolytes

Concentration (M)

Operation temperature (K)

Anode

Technique

Result/performance

References

KOH

4

253 and 333

Pure Al and Al–In alloy

Al

Weight loss experiments reveal that addition of 0.1 weight percent In greatly improves the corrosion resistance of the anode This study indicated that the effects of KOH concentration, aluminates concentration and temperature on the polarization characteristics are significant, but that there is no effect due to mass transfer Average discharge voltage: 1.37 V Capacity density: 2439 mA h g1 Energy density: 3364 W h kg1 Average discharge voltage: 1.398 V Capacity density: 1648 mA h g1 Energy density: 2429 W h kg1 The steady state polarization curves showed that aluminum exhibits a passive state at both 298 K and 333 K The potential value gained was around 1.5 V Average discharge voltage: 1.531 V Capacity density: 2308 mA h g1 Energy density: 3525 W h kg1 In, Mn, Sn, and Mg decreased the corrosion rate of the Al alloys, while Ga enhanced corrosion significantly and accelerated consumption of the anode. Fe was not beneficial to improve the electrochemical properties of the Al anode in that it caused a decrease in the cell voltage and reduced corrosion rate slightly

[34]

333

Weight loss experiments and polarization measurements, SEM and energy dispersive X-ray spectrometry of the anode surfaces Rotating aluminum cylinder electrode

The addition of Bi increases segregative phases which further activate the alloys to get more negative open circuit potentials, at the same time self-corrosion increases The open circuit potential = 1.67 V

[87]

KOH

KOH

4

298

Pure Al (99.999%)

EIS electrochemical test, polarization measurements, battery performance test and SEM

NaOH

4

298

Pure Al (99.999%)

NaOH

4

298 and 333

Pure Al

NaOH

4

Room

Pure Al

NaOH

4

Room

Pure Al (99.999%)

EIS electrochemical test, polarization measurements, battery performance test and SEM-EBSD

NaOH

4

Pure Al

Potentiodynamic polarization tests and electrochemical impedance spectroscopy

NaOH

4

Room

Al-based alloys + Fe, Ga, In, Sn, Mg, and Mn Al–1Mg–0.1Sn–0.1Ga–xBi alloys; xBi: 0, 0.05, 0.1, 0.2%

NaOH

4

Room

Al–(Zn–5.02% Mg–0.9%% Mn–0.44% Si–0.06% Fe– 0.001%Ti–0.04% Cu–0.03% In–0.02% C–0.05%)

NaOH

4

298

Al–(Zn–5.02% Mg–1.01% In–0.02% Si–0.09% Fe– 0.001%Ti–0.05% Cu– 0.02%)

NaOH

1 to 5

Room

Ordinary kitchen aluminum foil = Al purity of 97.6 wt.% (impurities: O 1.13, Fe 0.68, and Ag 0.59 wt.%)

Corrosion–dissolution characteristics of pure aluminum, steady state techniques and voltammetric analysis Anodic polarizing curve, SEM, EDAX

SEM/EDAX and electrochemical measurements, the self-corrosion was tested by using the gascollecting method

Measurements of self-corrosion, potentiodynamic polarization, cyclic polarization experiment combined with OCP technique and SEM. Cyclic polarization, self-corrosion and SEM.

Volume and evolution rate of hydrogen collected in the measuring cylinder

weight ratio of 10:7.5. The cathodic and anodic potential peaks are located at 0.091 V and 0.091 V, respectively. However, the power density of the Al–air battery system is 1.2 mW cm2, which is lower than that observed for the Zn–air battery [47]. Notably, the attempt to apply a gel–polymer-based alkaline electrolyte, hydroponic gel potassium hydroxide (HPG–KOH), in Al–air batteries ended in catastrophic failure because the aluminum anode was found to corrode rapidly into Al(OH)3, producing large amounts of hydrogen gas and a high concentration of OH [21].

The open circuit potential = 1.65 V and the alloy undergoes an intense corrosion process in alkaline solutions due to chemical dissolution by OHThe aluminum/air sub-cell has an open circuit voltage of 1.45 V

[83]

[125]

[84]

[40] [128]

[85]

[73]

[72]

[80]

The recent discovery of a new aluminum-ion-conducting polymer membrane has opened up new research opportunities regarding Al–air SMFCs without catalysts, similar to Li–ion batteries [93]. Moreover, the application of the polymer ceramic composite, non-oxide-based organic and oxide-based organic electrolytes that are presently used in Li–air batteries can be extended to Al–air batteries [94]. The anion-exchange membranes (AEMs) and cation-exchange membranes (CEMs) that have been used as interfaces between the electrode and the electrolyte can be used directly as electrolytes

M. Mokhtar et al. / Journal of Industrial and Engineering Chemistry 32 (2015) 1–20

14

Table 7 Aprotic electrolytes used in lithium–air batteries [88]. Types

Typical examples

Advantages

Disadvantages

References

Organic carbonates

Propylene Ethylene Diethyl Dimethyl Dimethoxyethane (DME) tetraethylene glycol dimethyl ether crown ethers

A high oxidation potential (HOMO) at ca. 4.7 V relatively low viscosity

Electrolytes decomposition

[129,130]

Highly stable with Li metal, high oxidation potential over 4.5 V. Low volatility, relatively high stability to some Li2O2 A high oxidation potential (5.3 V vs Li+/Li), non-flammability, a low vapour pressure, thermal stability, low toxicity, high boiling points, and a high Li-salt solubility Act as separator and good chemical stability

Electrolytes decomposition

[131,132]

A high viscosity and low ion diffusion

[133,134]

Low Li–ion conductivity

[135,136]

Ethers

Ionic liquids

1-Ethyl-3-methylimidazolium bis(triflouromethanesulfonyl)imide (EMITFSI)

Solid-state electrolyte

Lithium aluminium germanium phosphate glass-ceramic Li3x PO4y Ny (LiPON) Lithium aluminium titanium phosphate ceramics, 1-ethyl-3-methylimidazolium bis(trifluoromethanesulfonyl) imide

Table 8 Summary of polymer and solid-state electrolytes for aluminum–air batteries. Polymer host

Electrolyte

Analysis method

Performance/result

Reference

PAA

H2SO4 KI-additive

Electrochemical measurements—EIS Fourier transform infrared (FTIR) spectra (KBr pellet) AFM analyses

[137,138]

PAA

KOH ZnO-additive

PVA/PAA

KOH

Hydroponics gelling agent (HPG)

KOH

–

Sb(V)-doped SnP2O7

Measure the conductivity using a DDSJ-318 conductivity testing system from Shanghai Leici Co., Ltd. Discharge characteristics using a NEWARE BTS5V3A battery testing system. DSC and TGA: analyze the thermal properties of this solid state PVA/PAA polymer membrane XRD: investigate the crystalline structure SEM: examine the surface morphology and the cross-sectional view of the membrane EIS: measured ionic conductivity of PVA/PAA polymer electrolyte membrane with KOH electrolyte OCV: measured by storing the cell in an opencircuit condition for 24 h at a temperature of 25 8C characterized using X-ray diffraction (XRD) FESEM: observation of Al anode failure XRD: identify the crystalline structure of the aluminum electrode before charge–discharge tests SEM-EDX: observe the Al, O and Si elements Polarization measurements

EIS: an increase in the concentration of PAA results in an increase in the size of the semicircle and impedance of the interface Inhibition efficiency increases with increasing inhibitor concentration high ionic conductivity: 0.460 S cm1. optimal composition of 36 wt.% KOH and 6 wt.% AA is confirmed to obtain the balance between mechanism strength and conductivity peak capacity: up to 1166 mA h g1 energy density: 1230 mW h g1 Advantage:obtain the higher ionic conductivity, the higher mechanical strength and very good thermal for SPE ionic conductivity: 0.301 S cm1 polymer chain becomes more flexible for ionic transport due to the amorphous structure PVA/PAA polymer membrane at the composition ratio of 10:5 showed a good balance between enhanced ionic conductivity and reasonable mechanical properties for the solid polymer electrolyte applications Advantage: its rapid and super-absorbent solution properties At 0.6 M KOH concentration: highest capacity:105.0 mA h g1 power density: 5.5 mW cm2 High discharge capacity: 800 mA h g1 Open circuit voltage: up to 1.6 V

[51]. The application of AEMs and CEMs as electrolytes in Al–air batteries as replacements for the widely used aqueous electrolytes offers opportunities for future study [17]. Table 8 presents a summary of the polymer and solid-state electrolytes used in aluminum–air battery applications. Selection of electrolytes In this section, we will discuss and analyze the best potential values and performance of aluminium–air battery for aqueous solution, non-aqueous solution, aprotic and solid-state electrolytes. In acidic solution, when the concentration of electrolyte increase, the potential values are also increase. By adding some active metal elements as Al alloys, we can realize that there are some increment of the potential values involved. For binary Al alloys, we can summarized that the addition of In element gives the highest potential value compared to Zn and Si elements at the same concentration of chloride acid.

[20]

[139,140]

[21]

[141]

While for ternary Al alloys, combination of Al–In–Ga gives the highest potential values compared to Al–Zn–Ga and Al–Zn–Mg which the difference of the value up to 0.5 V. Based on this finding, we can state that In element gives a big impact to the aluminium– air battery performance in acidic solution. This ternary Al alloys also has been tested in other type of acid solution. From the observation, chloride acid solution much more suitable to be used as acidic electrolyte in aluminium–air battery compared to acetic acid solution due to the potential values gained. Neutral salt that commonly used in aluminium–air battery as electrolyte is NaCl solution. Generally, when the concentration of NaCl increase, the potential values also increase. Same goes to the temperature of aluminium–air battery. Most of the previous researches reported that the potential values for pure Al anode in NaCl solution was around 0.8 V and the concentration of NaCl solution that usually studied was 2 M. But Akmal [74] reported that the highest potential value for pure Al anode is 1.1 V in 4 M NaCl solution.

M. Mokhtar et al. / Journal of Industrial and Engineering Chemistry 32 (2015) 1–20

For binary Al alloys, combination of Al–Sn gives the highest potential value compared to Al–In and Al–Mg in 2 M NaCl solution. The potential values also increase when the amount of Sn element in the alloy increase. The effect of Sn element also shown in the ternary Al alloys. The combination of Al–Sn–Ga provide the highest potential value compared to Al–In–Sn, Al–Mg–Sn and Al–Mg–Ga in 2 M NaCl solution. Based on the results thus obtained, it can be concluded that ternary Al alloys can be ranked in the order of increasing potential as follows: Al–Mg–Ga > Al–Mg–Sn > Al–In– Sn > Al–Sn–Ga. Beside NaCl solution, seawater also has been studied as the neutral salt electrolyte. But unfortunately, eventhough they were used ternary Al alloys as anode, the highest potential values obtained from the study was only around 1.1 V [82]. For alkaline electrolyte, there are two solution that commonly used which are KOH and NaOH. KOH solution shown better aluminium–air battery performance compared to NaOH solution but it also shorten the discharge voltage values due to the high corrosion rate. Because of this reason, most of the researches prefer to use NaOH solution as alkaline electrolyte. The common concentration of NaOH solution that has been used in the previous studies was 4 M at room temperature. The potential value gained for pure Al anode in 4 M NaOH solution was up to 1.5 V. For Al alloys, we can summarized that the addition of Zn element gives the highest potential value compared other active metal elements in the same NaOH concentration. The potential value obtained was around 1.65 V [72,73]. For non-aqueous and aprotic electrolytes, the usage of these types of electrolytes in aluminium–air battery is still at the early stage. There are a few studies that has been reported using aprotic electrolytes. Based on the results obtained, the EMIm(HF)2.3F electrolyte gives the highest current densities compared to EMImCl/AlCl3 electrolyte in aluminium–air battery. Besides, recommendation for using aprotic ether based electrolytes for future aluminium–air battery is also has been discussed because this electrolyte has low bulk resistance that can help to increase the conductivity value. Similar with non-aqueous and aprotic electrolytes, there are a few studies that has been carried out using solid state electrolytes in aluminium–air battery. The usage of solid state electrolyte is a new approaches to prevent leakage challenge faced at the air cathode of this battery. Based on the previous reports, polymer host membranes that typically used in aluminium–air battery are PAA and PVA doped with alkaline electrolyte. The PAA/KOH with ZnO as corrosion inhibitor gives the highest ionic conductivity value compared to other polymer host membranes that has been tested in this battery. Further studies focus on the conductivity, mechanical and thermal stabilities of the membrane need to be improve to increase the aluminium–air battery performance and durability. Corrosion inhibitors Corrosion inhibitors are used to prevent the uncontrolled corrosion of the Al electrode in the electrolyte during current discharge. The primary mechanism of corrosion inhibition is through the adsorption of inhibitor molecules on the corroding metal surface, thus effectively lowering the corrosion reaction to a controlled level. Because the use of chemical inhibitors has been limited by environmental regulations, plant extracts are currently receiving increasing attention for their potential use as alternative corrosion-inhibiting materials. Organic inhibitors In acidic media. Recently, Fares et al. [95] investigated the use of polyethylene glycol (PEG) in the presence of the antibiotic

15

ciprofloxacin as an inhibitor additive for acidic media. The results revealed that the efficiency of the PEG increased from 61% to 91% with the addition of ciprofloxacin. Ciprofloxacin also increased the kinetic thermodynamic parameters, such as the activation enthalpy and entropy of the Al corrosion reaction. These authors also added a natural pectin polymer to an acidic medium and reported that the activation energy, enthalpy of activation and entropy increased with increasing pectin concentration [96]. For both inhibitors, the authors used the Langmuir adsorption isotherm to study the adsorption process on the Al surface [95,96]. Deng and Li [97] investigated an extract of Jasminum nudiflorum Lindl. leaves (JNLLE) as a green corrosion inhibitor in hydrochloric acid (HCl) solution using a weight loss method, polarization curves and EIS. The polarization curves reveal that JNLLE acts as a cathodic inhibitor, and the adsorption follows the Langmuir adsorption isotherm. Ezeokonkwo et al. [98] studied the Al surface adsorption process based on the Temkin adsorption isotherm and tested it by using Eucalyptus citriodora extract as the inhibitor in HCl solution. In this experiment, the authors also replaced the Al with mild steel to compare the inhibitor efficiency for both metals. The results revealed superior corrosion-inhibiting performance for Al compared with that for mild steel. Using a weight loss method, Yadav et al. [99] studied an extract of the Ziziphus mauritiana fruit as a green corrosion inhibitor for Al in HCl solution. It was tested at room temperature and for various extract concentrations between 0.0644 g L1 and 1.288 g L1. From these tests, it can be concluded that the efficiency of the inhibitor increases with increasing inhibitor concentration. The corrosion inhibition of aluminum alloy AA3001 in 0.1 M HCl solution induced by an extract of Commiphora pedunculata (CP) gum has been evaluated by Ameh and Eddy [100]. Using gravimetric and thermometric methods of monitoring corrosion, it was shown that as the concentration of CP gum increased, the inhibition efficiency also increased; however, the inhibition efficiency also decreased with increasing temperature. Halambek et al. [101] investigated an ethanol solution of Ocimum basilicum L. oil as an inhibitor for Al in 0.5 M HCl solution using a weight loss method, potentiodynamic polarization and EIS. The results revealed that the presence of basil oil in the HCl solution caused the current density to decrease; the EIS result indicated that this compound induced the formation of a protective layer on the Al surface. The use of a silicate-based extract from rice husk ash as a corrosion inhibitor for aluminum alloy Al 6061 in 0.5 M HCl solution has been studied by Othman et al. [102]. Using a weight loss method, it was shown that the addition of a small amount of this extract led to a decrease in the Al 6061 weight loss. In addition to HCl solution, another acidic medium that has also been studied is sulfuric acid (H2SO4). Obi-Egbedi et al. [103] investigated the natural Al corrosion inhibition effect of Spondias mombin L. extract in 0.5 M H2SO4. Using a standard gravimetric technique, it was found that the inhibition efficiency increased as the extract concentration increased. In neutral salt media. Using weight loss and polarization methods, Halambek et al. [104,105] studied the effectiveness of the natural oil extracted from Lavandula angustifolia L. and an ethanolic solution of Laurus nobilis L. oil as inhibitors for the corrosion of Al alloys in 3% NaCl solution. The results indicate that both inhibitors provide good protection and help to prevent pitting corrosion on the surfaces of Al alloys. In alkaline media. The inhibiting effect of an extract of Phyllanthus amarus leaves on the corrosion of Al surfaces in 2 M NaOH solution has been studied by Abiola and Otaigbe [106] using a chemical technique. The results indicated that at the highest concentration

16

Table 9 Summary of natural inhibitors of the corrosion of aluminum in various types of media. Metal

Medium

Inhibitor

Technique

Adsorption isotherm

Performance

References

Al

2 M NaOH solution

Extract of Phyllanthus amarus leaves

Chemical technique

Langmuir adsorption isotherm

[106]

Al

2 M sodium hydroxide (NaOH) solutions

Gossypium hirsutum L. leave extracts (GLE) and seed extracts (GSE)

Chemical technique

Al

2 M sodium hydroxide solution

Cationic surfactant cetyltrimethyl ammonium bromide (CTAB) and lupine seed extract

Electrochemical techniques and chemical gasometry measurements

Kineticthermodynamic model and Flory-Huggins isotherm

Al

1 M NaOH solution

Solanum trilobatum leaves extract

Mixed type inhibitor

Al

1 M NaOH solution

Extract of Lupinus varius L.

Weight loss, hydrogen evolution, polarization and electrochemical impedance spectroscopy methods Weight loss technique

76% efficiency at the highest concentration in the alkaline environment and the inhibition efficiency increased with increasing concentration of the extract The inhibition efficiency increased with increasing concentration of the extracts. The GLE gave 97% inhibition efficiency while the GSE gave 94% at the highest concentration Potentiodynamic polarization curve measurements showed that lupine seed extract controls both the anodic dissolution of aluminum and the hydrogen gas evolved at the cathodic sites of aluminum surface 94% efficiency at the highest concentration in the alkaline environment and the inhibition efficiency increased with increasing concentration of the extract

Al

Basic medium

Naphthol compound, 4(4-nitrophenylazo)-1naphthol (44NIN)

Al

0.5 M NaOH solution

Stem extract of Bacopa monnieri

Al

HCl and NaOH media

Phenol

Al

Acidic medium, 2.0. M HCl

i-Carrageenan a natural

2 M HCl solution

polymer

Pefloxacinmesylate, acting as zwitterionic mediator

Aningeria robusta extract

Potassium iodide (KI)

Classical chemical (gravimetric) and spectrophotometric (UV–vis and FTIR) methods Potentiodynamic polarization, electrochemical impedance spectroscopy (EIS) methods, and weight loss measurements Quantum electrochemical approaches based on density functional theory and cluster/ polarized continuum model Scanning electron microscope (SEM)

Hydrogen evolution method

Langmuir and Temkin adsorption isotherms

The inhibition efficiency increased with increasing the concentration of the extract and decreased with increasing temperature The inhibition efficiency was found to increase with the azo dye concentration but not with temperature

[107]

[108]

[109]

[110]

[142]

Mixed-type inhibitor; Langmuir adsorption isotherm model

The inhibition efficiencies obtained were in good agreement

[143]

Mixed-type inhibition mechanism

The inhibitor adsorption in acid is more likely to have a physical nature and this studies confirmed that the corrosion rate in alkaline solution is substantially greater than in HCl media

[144]

Langmuir adsorption isotherm

Activation energy of corrosion and other thermodynamic parameters such as standard free energy, standard enthalpy, and standard entropy of the adsorption process revealed better and well-ordered physical adsorption layers in presence of pefloxacin Inhibition efficiency synergistically increased on addition of potassium iodide but decreased with increase in temperature

[145]

Langmuir adsorption isotherm

[146]

M. Mokhtar et al. / Journal of Industrial and Engineering Chemistry 32 (2015) 1–20

Al

Additive

1 M HCl

Extract of Ipomoea invulcrata (IP)

Al

1 M HCl

Coconut coir dust extract (CCDE)

Al

1 M hydrochloric acid solution

Anethum graveolens L. essential oil

Al-SiC composites

1 N HCl

Sulfaguanidine drug (SGD)

Pure Al

1 N HCl

Cumin extract

Al and stainless steel

1 M HCl acidic solution

Juglans regia L.

Al

1 M HCl solution

Mentha pulegium extract

Al

0.5 M HCl

Ethanol solution of Ocimum basilicum L. oil

Al 6061

0.5 M hydrochloric acid (HCl)

Silicate-based extracted from rice husk ash

Al and copper

HCl solution

0.0644–1.288 g L1 Ziziphus mauritiana Fruit Extract

Aluminum alloy AA 3001

0.1 M HCl

Commiphora pedunculata (CP) gum

Al

HCl solution

Jasminum nudiflorum Lindl. leaves extract (JNLLE)

Al

Acidic medium

Polyethylene glycol (PEG)

Al

Acidic media

Pectin natural polymer

With/without KI and Potassium thiocyanate (KSCN)

Antibiotic ciprofloxacin

conventional weight loss technique

Langmuir adsorption isotherm

Weight loss and hydrogen evolution techniques Weight loss, potentiodynamic polarization, and EIS methods Mass loss studies, gasometric measurements, potentiodynamic polarization and impedance analysis Weight loss, Galvanostatic polarization and EIS techniques Gravimetric (weight loss), potentiodynamic polarization and EIS methods Gravimetric, gasometric and EIS techniques

Langmuir adsorption isotherm Langmuir adsorption isotherm

Inhibition efficiency increases with concentration but decreases with increase in temperature and immersion time The inhibition efficiency increases with increasing temperature and concentration of the extract Inhibition efficiency increased with increasing oil concentration, but decreased with temperature

[147]

[148]

[149]

Temkin’s adsorption isotherm

Quantum mechanical calculations validated the inhibition efficiency of the compound studied by electrochemical methods

[150]

Langmuir Adsorption isotherm; Mixed type inhibitor

Reveal that inhibition efficiency increases with increase in concentration of inhibitors but decreases with increase in temperature An increase in inhibition efficiency was identified with the increased concentration of the inhibitor

[151]

Langmuir isotherm

Temkin adsorption model

Weight loss measurements, potentiodynamic polarization and EIS methods Weight loss, potentiodynamic polarization and optical or SEM Weight loss method

Thermodynamic adsorption

Gravimetric and thermometric methods of monitoring corrosion Weight loss, polarization curves, EIS and SEM methods

Langmuir adsorption mode

Langmuir adsorption isotherm

Langmuir adsorption isotherm

SEM

Langmuir isotherm

–

Langmuir isotherm fit

The inhibition efficiency depended on the concentration of the plant extract as well as on the time of exposure of the aluminum samples in HCl solutions containing the extract EIS results confirmed that investigated compound formed protective layer on aluminum surface

[153]

[101]

[102]

[99]

[100]

[97]

[95]

[96]

17

Addition of silicate-based corrosion inhibitor was exhibited the decreasing of the weight loss of Al 6061 in acidic medium Inhibition efficiency increases with increase in concentration of the inhibitor and reached maximum of 76.8% at room temperature for aluminum at 1.288 g L1 concentration of extract The inhibition efficiency of CP gum was found to increase with an increase in concentration but to decrease with increasing temperature Polarization curves reveal that JNLLE acts as the cathodic inhibitor. EIS exhibits a large capacitive loop at high frequencies followed by a large inductive one at low frequency values Ciprofloxacin increase the inhibition efficiency of PEG from 61 to 91% and it also increases the kinetic thermodynamic parameters The maximum inhibition efficiency obtained at 10 8C using pectin concentration was 91% and at 40 8C it severely declined to 31%

[152]

M. Mokhtar et al. / Journal of Industrial and Engineering Chemistry 32 (2015) 1–20

Al

[104]

[105]

[103]

Weight loss, polarization measurements and SEM Weight loss method, potentiodynamic polarization, and linear polarization 3% NaCl solution Pure Al and AA5754 aluminum alloy

(30%, v/v) ethanolic solution of Laurus nobilis L. oil

3% NaCl solution Al-3. Mg alloy

Natural oil extracted from Lavandula angustifolia L.

0.5 M H2SO4 Al

extracts of Spondias mombin L.

Potassium iodide (KI)

Standard gravimetric technique

Langmuir’s adsorption isotherm

Equilibrium constants of the adsorption processes predict better corrosion inhibition of aluminum than mild steel Inhibition efficiency of the extract increased with an increase in concentration of the S. mombin L. extract but decreased with temperature and synergistically increased on addition of potassium iodide Found that the L. angustifolia L. oil provides a good protection to Al-3. Mg alloy against pitting corrosion in sodium chloride solution The results confirmed that AA5754 alloy has better corrosion resistance than pure aluminum, while the oil investigated has better inhibition action on corrosion process of pure aluminum Langmuir adsorption isotherm

Performance

Temkin adsorption isotherm Weight loss technique HClsoln

Eucalyptus citriodora

Medium Metal

Al and Mild steel

Table 9 (Continued )

Inhibitor

Additive

Technique

Adsorption isotherm

[98]

M. Mokhtar et al. / Journal of Industrial and Engineering Chemistry 32 (2015) 1–20 References

18

tested, the inhibition efficiency was approximately 76%. Using the same technique and alkaline medium, the authors also tested the effects of extracts of Gossypium hirsutum L. leaves (GLE) and seeds (GSE) as Al corrosion inhibitors [107]. It was found that GLE was more effective than GSE, with inhibition efficiencies of 97% and 94%, respectively. Abdel-Gaber et al. [108] evaluated the effects of the cationic surfactant cetyltrimethylammonium bromide (CTAB) and lupine seed extract as Al surface corrosion inhibitors in 2 M NaOH solution using an electrochemical technique, and the result revealed that the lupine seed extract controlled both the Al anodic dissolution and the hydrogen gas evolution on the cathodic side. Recently, the effect of an extract of Solanum trilobatum leaves as an Al corrosion inhibitor in 1 M NaOH solution was investigated by Geetha et al. [109]. It was found that at the highest concentration tested, the inhibition efficiency was approximately 94%. Irshedat et al. [110] studied Al corrosion inhibition of Lupinus varius I. extract in 1 M NaOH solution. Using a weight loss method, it was concluded that the inhibition efficiency increased with increasing extract concentration and decreased with increasing temperature. Table 9 summarizes all natural inhibitors for the corrosion of aluminum in various types of media that have been studied over the last few years. Synthetic inhibitors The ion additives that have been studied in neutral salt media are In3+, Sn3+ and Zn2+ ions. El Shayeb et al. [30] have investigated the effect of adding In3+ ions on the Al anode potential in 0.6 M NaCl solution, and the results revealed that the activation of pure Al and Al alloys increased as the In3+ concentration was increased. These authors also investigated the use of Sn3+ ions as an additive to 0.6 M NaCl solutions [29], and the potential values with respect to SCE for pure Al and Al alloys were found to be between 0.95 V and 1.1 V. Pure Al and Al alloys have also been tested in 0.6 M NaCl solutions with Zn3+ ions as an additive by El Abedin and Endres [31]. The results revealed that the Al–In alloy exhibited a higher potential than did pure Al or Al–Ga–In, with values of 1.2 V, 0.8 V and 1.05 V, respectively. Instead of using ions as an additive, Tang et al. [40] studied the effect on the Al potential of adding zinc chloride (ZnCl 2 ) to a 0.5 M NaCl solution. The results indicated that the addition of ZnCl 2 improved the activation potentials of pure Al and Al alloys. Later, cerium(III) chloride (CeCl 3 ) was evaluated for use as an additive in a 3.5 M NaCl solution by Zhou et al. [111], and the results revealed an improved discharge performance of the Al anode with the addition of CeCl 3 to the NaCl solution. In alkaline media, the chemical additives that have been most commonly used are zinc oxide (ZnO) and sodium stannate (Na2SnO3). Wang et al. [112] and Martin and Zhu [113] evaluated the use of ZnO as an additive in KOH solutions, where as Rashvand Avei et al. [114] evaluated its use in NaOH solutions. All authors claimed that the addition of ZnO helped to prevented Al corrosion and improved the performance of the Al anode. Meanwhile, Doche et al. [35] have studied the use of Na2SnO3 in NaOH solutions, whereas Chang et al. [115] and Zeng et al. [24] have studied its use in KOH solutions. The results revealed that the addition of sodium stannate shifted the potential values more strongly negative. Other chemical additives that have been explored include BMIMBF4, polyethylene glycol, NaSnO3, and alkaline citrate [57,112,116–119]. For further information, the reader may refer to several previous reviews that have discussed in detail the additives used in alkaline media for Al corrosion inhibition [16,22,120].

M. Mokhtar et al. / Journal of Industrial and Engineering Chemistry 32 (2015) 1–20

Future trends in research on aluminum–air batteries Polymer and gel electrolytes have recently been widely studied for use in various electrochemical cells, including metal–air batteries. Because few studies have been performed regarding the use of polymer and gel electrolytes in aluminum–air batteries, there is a clear need for further research in this area. Polymer and gel electrolytes are very attractive for application in this type of battery because they are lightweight and flexible, with high electrochemical and thermal stability. The modification of polymer electrolytes also can increase their ionic conductivity and help to improve battery performance. Such modifications can be achieved by blending two or more polymers together, adding plasticizers or inorganic inert fillers to polymer electrolytes, or cross-linking polymers. Furthermore, the addition of corrosion inhibitors to polymer electrolytes can help to reduce the self-corrosion of the aluminum anode and enhance the durability of aluminum–air batteries. Even when high-purity aluminum is not used as the anode, corrosion inhibitors are still beneficial for maintaining battery performance. From the perspective of battery design, polymer electrolytes can be used to minimize the leakage problem that commonly arises in the case of liquid electrolytes. Such leakage can affect battery performance and the durability of the battery stack. At present, many promising innovations have been achieved in the aspects of both materials and design to overcome the challenges of leakage and corrosion faced in aluminum–air batteries.

Conclusion Although aluminum–air batteries are well known as a mechanically rechargeable type of battery with features that can satisfy the basic requirements of many electrochemical device applications, from small portable devices to large power systems, further development of these batteries is necessary to facilitate their commercialization. Each component of this type of battery offers its own advantages and challenges. In general, the battery performance increases when Al alloys, especially ternary alloys, are used as anodes. Studies of alloying Al with active metal elements such as Sn, In and Ga have proven that these elements beneficially influence the battery performance. At present, one attractive possibility for use as the air cathode is a newly fabricated bifunctional air cathode based on an ORR catalyst, CoO/N-CNT, and an OER catalyst, Ni–Fe-layered double hydroxide/CNT, with an onset potential that is 20 mV more positive than that of a Pt/C cathode and 20 mV more negative than that of an Ir/C cathode during charging. With regard to electrolytes, all aqueous electrolytes must face the twin challenges of Al anode corrosion and hydrogen evolution. The alkaline solution offers superior performance compared with the acidic and neutral salt solutions. The addition of corrosion inhibitors to the aqueous electrolyte has also been found to improve battery performance. By contrast, nonaqueous and aprotic solvent electrolytes can be used to prevent hydrogen evolution due to parasitic reactions from degrading the electrode, whereas polymer electrolytes can overcome the leakage problems faced by batteries with liquid electrolytes and minimize the thickness of the battery. Acknowledgments The authors gratefully acknowledge financial support for this work by the Malaysian Ministry of Higher Education under Research Grant Nos. LRGS/2013/UKM-UKM/TP-01, FRGS/2/2013/ TK04/UKM/02/3 and MyBrain15.

19

References [1] K.F. Blurton, A.F. Sammells, J. Power Sources 4 (1979) 263. [2] A. Kraytsberg, Y. Ein-Eli, Nano Energy 2 (2013) 468. [3] M. Kar, T.J. Simons, M. Forsyth, D.R. MacFarlane, Phys. Chem. Chem. Phys. 16 (2014) 18658. [4] Z.-L. Wang, D. Xu, J.-J. Xu, X.-B. Zhang, Chem. Soc. Rev. 43 (2014) 7746. [5] J.-S. Lee, S. Tai Kim, R. Cao, N.-S. Choi, M. Liu, K.T. Lee, J. Cho, Adv. Energy Mater. 1 (2011) 34. [6] M.A. Rahman, X. Wang, C. Wen, J. Electrochem. Soc. 160 (2013) A1759. [7] J.G. Kim, B. Son, S. Mukherjee, N. Schuppert, A. Bates, O. Kwon, M.J. Choi, H.Y. Chung, S. Park, J. Power Sources 282 (2015) 299. [8] M.-K. Song, S. Park, F.M. Alamgir, J. Cho, M. Liu, Mater. Sci. Eng.: R: Rep. 72 (2011) 203. [9] Y. Sun, Nano Energy 2 (2013) 801. [10] J. Yuan, J.-S. Yu, B. Sunde´n, J. Power Sources 278 (2015) 352. [11] N. Imanishi, O. Yamamoto, Mater. Today 17 (2014) 24. [12] J. Wang, Y. Li, X. Sun, Nano Energy 2 (2013) 443. [13] R. Padbury, X. Zhang, J. Power Sources 196 (2011) 4436. [14] R. Mori, RSC Adv. 4 (2014) 30346. [15] H. Zhang, X. Zhong, J.C. Shaw, L. Liu, Y. Huang, X. Duan, Energy Environ. Sci. 6 (2013) 2621. [16] Q. Li, N.J. Bjerrum, J. Power Sources 110 (2002) 1. [17] G. Agladze, P. Nikoleishvili, V. Kveselava, G. Tsurtsumia, G. Gorelishvili, D. Gogoli, I. Kakhniashvili, J. Power Sources 218 (2012) 46. [18] D.R. Linden, Linden’s Handbook of Batteries, 4th ed., 2011, pp. 33.31–33.58 (T.B.). [19] H.G. Schwarz, Encycl. Energy 1 (2004) 81. [20] Z. Zhang, C. Zuo, Z. Liu, Y. Yu, Y. Zuo, Y. Song, J. Power Sources 251 (2014) 470. [21] A.A. Mohamad, Corros. Sci. 50 (2008) 3475. [22] D.R. Egan, C. Ponce De Leo´n, R.J.K. Wood, R.L. Jones, K.R. Stokes, F.C. Walsh, J. Power Sources 236 (2013) 293. [23] A.Z. Zhuk, A.E. Sheindlin, B.V. Kleymenov, E.I. Shkolnikov, M.Y. Lopatin, J. Power Sources 157 (2006) 921. [24] X.X. Zeng, J.M. Wang, Q.L. Wang, D.S. Kong, H.B. Shao, J.Q. Zhang, C.N. Cao, Mater. Chem. Phys. 121 (2010) 459. [25] M. Nestoridi, D. Pletcher, R.J.K. Wood, S. Wang, R.L. Jones, K.R. Stokes, I. Wilcock, J. Power Sources 178 (2008) 445. [26] Y.-J. Cho, I.-J. Park, H.-J. Lee, J.-G. Kim, J. Power Sources 277 (2015) 370. [27] S.B. Saidman, J.B. Bessone, Electrochim. Acta 42 (1997) 413. [28] A.R. Despic´, D.M. Drazˇic´, M.M. Purenovic´, N. Cikovic´, J. Appl. Electrochem. 6 (1976) 527. [29] H.A. El Shayeb, F.M. Abd El Wahab, S. Zein El Abedin, Corros. Sci. 43 (2001) 655. [30] H.A. El Shayeb, F.M. Abd El Wahab, S. Zein El Abedin, J. Appl. Electrochem. 29 (1999) 473. [31] S.Z. El Abedin, F. Endres, J. Appl. Electrochem. 34 (2004) 1071. [32] I. Smoljko, S. Gudic´, N. Kuzmanic´, M. Kliskic´, J. Appl. Electrochem. 42 (2012) 969. [33] S. Gudic´, I. Smoljko, M. Klisˇkic´, Mater. Chem. Phys. 121 (2010) 561. [34] W. Wilhelmsen, T. Arnesen, Ø. Hasvold, N.J. Størkersen, Electrochim. Acta 36 (1991) 79. [35] M.L. Doche, F. Novel-Cattin, R. Durand, J.J. Rameau, J. Power Sources 65 (1997) 197. [36] S. Gudic´, J. Radosˇevic´, I. Smoljko, M. Klisˇkic´, Electrochim. Acta 50 (2005) 5624. [37] D.O. Flamini, S.B. Saidman, J.B. Bessone, Corros. Sci. 48 (2006) 1413. [38] J.-b. Song, W.-m. Mao, H. Yang, H.-p. Feng, Trans. Nonferrous Metals Soc. China 18 (2008) 879. [39] S. Licht, G. Levitin, R. Tel-Vered, C. Yarnitzky, Electrochem. Commun. 2 (2000) 329. [40] Y. Tang, L. Lu, H.W. Roesky, L. Wang, B. Huang, J. Power Sources 138 (2004) 313. [41] D.O. Flamini, S.B. Saidman, Mater. Chem. Phys. 136 (2012) 103. [42] P.R. Birkin, M. Nestoridi, D. Pletcher, Electrochim. Acta 54 (2009) 6668. [43] W.A. Ferrando, J. Power Sources 130 (2004) 309. [44] A. Elango, V.M. Periasamy, M. Paramasiv, J. Metall. Mater. Sci. 46 (2004) 247. [45] E.S. Takeuchi, A.C. Marschilok, K.J. Takeuchi, Electrochemistry 80 (2012) 700. [46] L. Jo¨rissen, J. Power Sources 155 (2006) 23. [47] F. Cheng, J. Chen, Chem. Soc. Rev. 41 (2012) 2172. [48] A. De´bart, J. Bao, G. Armstrong, P.G. Bruce, J. Power Sources 174 (2007) 1177. [49] L. Jiang, A. Hsu, D. Chu, R. Chen, Electrochim. Acta 55 (2010) 4506. [50] H. Li, H. Liu, Z. Jong, W. Qu, D. Geng, X. Sun, H. Wang, Int. J. Hydrogen Energy 36 (2011) 2258. [51] N. Fujiwara, M. Yao, Z. Siroma, H. Senoh, T. Ioroi, K. Yasuda, J. Power Sources 196 (2011) 808. [52] S. Zhu, Z. Chen, B. Li, D. Higgins, H. Wang, H. Li, Z. Chen, Electrochim. Acta 56 (2011) 5080. [53] C.N. Chervin, J.W. Long, N.L. Brandell, J.M. Wallace, N.W. Kucko, D.R. Rolison, J. Power Sources 207 (2012) 191. [54] Z. Chen, A. Yu, R. Ahmed, H. Wang, H. Li, Z. Chen, Electrochim. Acta 69 (2012) 295. [55] R.G.A. Wills, M. Kourasi, A.A. Shah, F.C.Walsh, Electrochem. Commun. 22 (2012) 174. [56] Y. Li, M. Gong, Y. Liang, J. Feng, J.E. Kim, H. Wang, G. Hong, B. Zhang, H. Dai, Nat. Commun. 4 (2013) 1805. [57] Y. Zhang, S. Liang, A. Javed, D. Guan, Met. Mater. Int. 19 (2013) 555. [58] H.T. Chung, J.H. Won, P. Zelenay, Nat. Commun. 4 (2013). [59] Y. Huang, Y. Lin, W. Li, Electrochim. Acta 99 (2013) 161. [60] S.B. Ma, D.J. Lee, V. Roev, D. Im, S.-G. Doo, J. Power Sources. [61] J. Adams, M. Karulkar, V. Anandan, J. Power Sources 239 (2013) 132.

20

M. Mokhtar et al. / Journal of Industrial and Engineering Chemistry 32 (2015) 1–20

[62] A. Garsuch, A. Panchenko, C. Querner, A. Karpov, S. Huber, R. Oesten, Electrochem. Commun. 12 (2010) 1642. [63] K. Tiido, N. Alexeyeva, M. Couillard, C. Bock, B.R. MacDougall, K. Tammeveski, Electrochim. Acta 107 (2013) 509. [64] K. Huang, Y. Li, Y. Xing, Electrochim. Acta 103 (2013) 44. [65] A.C. Garcia, E.A. Ticianelli, Electrochim. Acta 106 (2013) 453. [66] A.C. Marschilok, S.H. Lee, C.C. Milleville, P. Chen, E.S. Takeuchi, K.J. Takeuchi, J. Compos. Mater. 47 (2013) 33. [67] Y. Wang, X. Lu, Y. Liu, Y. Deng, Electrochem. Commun. 31 (2013) 108. [68] S. Velraj, J.H. Zhu, J. Power Sources 227 (2013) 48. [69] B. Li, X. Lu, J. Yuan, Y. Zhu, L. Li, Ionics 21 (2014) 141. [70] A.A. Mazhar, S.T. Arab, E.A. Noor, J. Appl. Electrochem. 31 (2001) 1131. [71] A.G. Mun˜oz, S.B. Saidman, J.B. Bessone, Corros. Sci. 44 (2002) 2171. [72] J. Ma, J. Wen, Q. Li, Q. Zhang, Int. J. Hydrogen Energy 38 (2013) 14896. [73] J. Ma, J. Wen, Q. Li, Q. Zhang, J. Power Sources 226 (2013) 156. [74] M. Akmal, R. Othman, M.H. Ani, Adv. Mater. Res. 701 (2013) 314. [75] A. Aballe, M. Bethencourt, F.J. Botana, J. Cano, M. Marcos, Corros. Rev. 18 (2000) 1. [76] A. Aballe, M. Bethencourt, F.J. Botana, M.J. Cano, M. Marcos, Corros. Sci. 43 (2001) 1657. [77] B. Han, G. Liang, Rare Metals 25 (2006) 360. [78] M. Nestoridi, D. Pletcher, J.A. Wharton, R.J.K. Wood, J. Power Sources 193 (2009) 895. [79] H.J. Fan, H.Y. Sun, L.J. Sun, W. Wang, B.N. Zang, Corros. Sci. Prot. Technol. 24 (2012) 149. [80] L. Wang, W. Wang, G. Yang, D. Liu, J. Xuan, H. Wang, M.K.H. Leung, F. Liu, Int. J. Hydrogen Energy 38 (2013) 14801. [81] A. Mance, D. Cerovic´, A. Mihajlovic´, J. Appl. Electrochem. 14 (1984) 459. [82] H. Ezuber, A. El-Houd, F. El-Shawesh, Mater. Des. 29 (2008) 801. [83] D. Chu, R.F. Savinell, Electrochim. Acta 36 (1991) 1631. [84] M.L. Doche, J.J. Rameau, R. Durand, F. Novel-Cattin, Corros. Sci. 41 (1999) 805. [85] Y.I. Choi, R.S. Kalubarme, H. Jang, C.J. Park, J. Korean Inst. Metals Mater. 49 (2011) 839. [86] D.D. Macdonald, K.H. Lee, A. Moccari, D. Harrington, Corrosion 44 (1988) 652. [87] J.W. Gao, J.B. Wen, J.G. He, Adv. Mater. Res. 800 (2013) 488. [88] J. Wang, Y. Li, X. Sun, Nano Energy 2 (2013) 443. [89] R. Revel, T. Audichon, S. Gonzalez, J. Power Sources 272 (2014) 415. [90] D. Gelman, B. Shvartsev, Y. Ein-Eli, J. Mater. Chem., A 2 (2014) 20237. [91] L. Cecchetto, M. Salomon, B. Scrosati, F. Croce, J. Power Sources 213 (2012) 233. [92] C.-C. Yang, S.-J. Lin, J. Power Sources 112 (2002) 497. [93] R. Mori, RSC Adv. 3 (2013) 11547. [94] Y. Sun, Nano Energy 2 (2013) 801. [95] M.M. Fares, A.K. Maayta, J.A. Al-Mustafa, J. Adhes. Sci. Technol. 27 (2013) 2495. [96] M.M. Fares, A.K. Maayta, M.M. Al-Qudah, Corros. Sci. 60 (2012) 112. [97] S. Deng, X. Li, Corros. Sci. 64 (2012) 253. [98] M.A. Ezeokonkwo, P.O. Ukoha, N.J.N. Nnaji, Int. J. Chem. Sci. 10 (2012) 1365. [99] S. Yadav, G. Choudhary, A. Sharma, Int. J. ChemTech Res. 5 (2013) 1815. [100] P.O. Ameh, N.O. Eddy, Res. Chem. Intermed. (2013) 1. [101] J. Halambek, A. Zˇutinic´, K. Berkovic´, Int. J. Electrochem. Sci. 8 (2013) 11201. [102] N.K. Othman, N. Mohamad, R. Zulkafli, A. Jalar, AIP Conf. Proc. 1528 (2013) 243. [103] N.O. Obi-Egbedi, I.B. Obot, S.A. Umoren, Arabian J. Chem. 5 (2012) 361. [104] J. Halambek, K. Berkovic´, J. Vorkapic´-Furacˇ, Mater. Chem. Phys. 137 (2013) 788. [105] J. Halambek, K. Berkovic´, J. Vorkapic´-Furacˇ, Corros. Sci. 52 (2010) 3978. [106] O.K. Abiola, J.O.E. Otaigbe, Corros. Sci. 51 (2009) 2790. [107] O.K. Abiola, J.O.E. Otaigbe, O.J. Kio, Corros. Sci. 51 (2009) 1879. [108] A.M. Abdel-Gaber, E. Khamis, H. Abo-Eldahab, S. Adeel, Mater. Chem. Phys. 124 (2010) 773. [109] S. Geetha, S. Lakshmi, K. Bharathi, J. Chem. Pharm. Res. 5 (2013) 195. [110] M.K. Irshedat, E.M. Nawafleh, T.T. Bataineh, R. Muhaidat, M.A. Al-Qudah, A.A. Alomary, Portug. Electrochim. Acta 31 (2013) 1. [111] Y.H. Zhou, X.H. Guo, S.D. Zhang, D.R. Zhou, Cailiao Kexue yu Gongyi (Mater. Sci. Technol.), 15 (2007) 192.

[112] X.Y. Wang, J.M. Wang, Q.L. Wang, H.B. Shao, J.Q. Zhang, Mater. Corros. 62 (2011) 1149. [113] A.D. Martin, J.H. Zhu, ECS Electrochem. Lett. 1 (2012) A13. [114] M. Rashvand Avei, M. Jafarian, H. Moghanni Bavil Olyaei, F. Gobal, S.M. Hosseini, M.G. Mahjani, Mater. Chem. Phys. 143 (2013) 133. [115] X. Chang, J. Wang, H. Shao, J. Wang, X. Zeng, J. Zhang, C. Cao, Acta PhysicoChim. Sin. 24 (2008) 1620. [116] A.A. Al-Suhybani, Y.H. Sultan, W.A. Hamid, Materialwiss. Werkstofftech. 22 (1991) 301. [117] V. Kapali, S. Venkatakrishna Iyer, V. Balaramachandran, K.B. Sarangapani, M. Ganesan, M. Anbu Kulandainathan, A. Sheik Mideen, J. Power Sources 39 (1992) 263. [118] J.M. Wang, J.B. Wang, H.B. Shao, X.X. Zeng, J.Q. Zhang, C.N. Cao, Mater. Corros. 60 (2009) 977. [119] D.D. Macdonald, C. English, J. Appl. Electrochem. 20 (1990) 405. [120] J. Zhang, M. Klasky, B.C. Letellier, J. Nucl. Mater. 384 (2009) 175. [121] T. Laino, A. Curioni, Chemistry—A Eur. J. 18 (2012) 3510. [122] Y. Mo, S.P. Ong, G. Ceder, Phys. Rev. B: Condens. Matter 84 (2011) 205446. [123] L. Zhang, X. Zhang, Z. Wang, J. Xu, D. Xu, L. Wang, Chem. Commun. 48 (2012) 7598. [124] W.Y. Wong, W.R.W. Daud, A.B. Mohamad, A.A.H. Kadhum, E.H. Majlan, K.S. Loh, Diamond Relat. Mater. 22 (2012) 12. [125] L. Fan, H. Lu, J. Leng, Electrochim. Acta 165 (2015) 22. [126] X. Wang, Z. Liu, Q. Kong, W. Jiang, J. Yao, C. Zhang, G. Cui, Solid State Ionics 262 (2014) 747. [127] J. Ma, J. Wen, J. Gao, Q. Li, J. Power Sources 253 (2014) 419. [128] L. Fan, H. Lu, J. Power Sources 284 (2015) 409. [129] Z. Hong, M. Wei, T. Lan, G. Cao, Nano Energy 1 (2012) 466. [130] L. Ji, H. Zheng, A. Ismach, Z. Tan, S. Xun, E. Lin, V. Battaglia, V. Srinivasan, Y. Zhang, Nano Energy 1 (2012) 164. [131] C. Laoire, S. Mukerjee, E.J. Plichta, M.A. Hendrickson, K.M. Abraham, J. Electrochem. Soc. 158 (2011) A302. [132] W. Xu, K. Xu, V.V. Viswanathan, S.A. Towne, J.S. Hardy, J. Xiao, Z. Nie, D. Hu, D. Wang, J.-G. Zhang, J. Power Sources 196 (2011) 9631. [133] D. Zhang, R. Li, T. Huang, A. Yu, J. Power Sources 195 (2010) 1202. [134] T. Kuboki, T. Okuyama, T. Ohsaki, N. Takami, J. Power Sources 146 (2005) 766. [135] B. Kumar, J. Kumar, R. Leese, J.P. Fellner, S.J. Rodrigues, K.M. Abraham, J. Electrochem. Soc. 157 (2010) A50. [136] B. Kumar, J. Kumar, J. Electrochem. Soc. 157 (2010) A611. [137] S. Umoren, Y. Li, F. Wang, J. Solid State Electrochem. 14 (2010) 2293. [138] S.A. Umoren, Y. Li, F.H. Wang, J. Appl. Electrochem. 41 (2011) 307. [139] G.M. Wu, S.J. Lin, C.C. Yang, J. Membr. Sci. 275 (2006) 127. [140] G.M. Wu, S.J. Lin, C.C. Yang, J. Membr. Sci. 280 (2006) 802. [141] T. Hibino, K. Kobayashi, M. Nagao, J. Mater. Chem., A 1 (2013) 14844. [142] U. Eduok, E. Inam, S.A. Umoren, I.A. Akpan, Geosyst. Eng. 16 (2013) 146. [143] A. Singh, E.E. Ebenso, M.A. Quraishi, Int. J. Electrochem. Sci. 7 (2012) 3409. [144] M. Lashgari, A.M. Malek, Electrochim. Acta 55 (2010) 5253. [145] M.M. Fares, A.K. Maayta, J.A. Al-Mustafa, Corros. Sci. 65 (2012) 223. [146] I.B. Obot, S.A. Umoren, N.O. Obi-Egbedi, J. Mater. Environ. Sci. 2 (2011) 49. [147] I.B. Obot, N.O. Obi-Egbedi, S.A. Umoren, E.E. Ebenso, Int. J. Electrochem. Sci. 5 (2010) 994. [148] S.A. Umoren, U.M. Eduok, A.U. Israel, I.B. Obot, M.M. Solomon, Green Chem. Lett. Rev. 5 (2012) 303. [149] J. Halambek, K. Berkovic, Int. J. Electrochem. Sci. 7 (2012) 8356. [150] V. Umasankar, S. Karthikeyan, M. Anthony Xavier, J. Corros. Sci. Eng. 16 (2013). [151] D.G. Ladha, U.J. Naik, N.K. Shah, J. Mater. Environ. Sci. 4 (2013) 701. [152] M. Akin, S. Nalbantoglu, O. Cuhadar, D. Uzun, N. Saki, Res. Chem. Intermed. 41 (2015) 899. [153] A. Khadraoui, A. Khelifa, L. Touafri, H. Hamitouche, R. Mehdaoui, J. Mater. Environ. Sci. 4 (2013) 663.