1 Introduction

Isaac Newton and Michael Faraday have given us the intellectual tools to wheel the battery of innovations for clean energy and efficient techno-development of global community. Extensive usage of energy coupled with global climate change has necessitated the growth of renewable energy production and storage systems. World-wide interest in electric vehicles (EVs) has focused on miniaturization of light weight rechargeable batteries. Li-ion batteries with high energy efficiency and more cyclic capabilities has become the most promising candidate for EV applications. However, its low storage capacity (100–200 Wh kg−1), safety issues and high cost has led to extensive research activities to find alternative energy storage technology. Metal–air battery systems such as Li–air, Mg–air, Zn–air, Al–air and Si–air are highly promising for powering automobiles, industrial equipment, computers, electronic device, hearing aids and a large number of utility items [1,2,3,4,5,6,7,8]. Apart from possessing high energy density (400–1700 Wh kg−1), metal–air batteries are compact, low cost, lighter and environment friendly [9,10,11,12,13,14,15]. The reason for high energy density possessed by metal–air batteries is its cathode which operates on oxygen present in the air during discharge, replacing costly chemicals used in present Li-ion batteries. Early work on metal–air batteries encountered the problems associated with the material at the air cathode surface, thermal management system along with various technical problems associated with anodes. Theoretical specific energy densities and open circuit voltages of metal–air battery systems are shown in Fig. 1.

Fig. 1
figure 1

Comparison of theoretical specific energy densities and open circuit voltages (OCVs) of different metal–air batteries [10]

Full size image

Lithium metal is considered to be the best candidate as anode material because of its high theoretical specific energy (3860 Ah kg−1), low density (0.59 g cm−3) and the lowest electrode potential (3.045 V vs. SCE). Si–air battery has comparable specific energy density (5360 Wh kg−1) with Li–air battery (5200 Wh kg−1). Also, silicon (28.2%) is more abundant in earth’s crust when compared with lithium (0.002%) and is also less expensive than lithium metal. The practical performances of some battery systems including metal air batteries have been represented in the form of Ragone plots [16] (Fig. 2). It may be noted from the Ragone plots that metal–air batteries deliver higher specific energy while Li–ion battery provides the best power capability. Also, Li–air and Si–air batteries provide specific energies of the same order but there is large difference in their specific power capabilities. Moreover, rechargeability of Si–air batteries is a challenging task.

Fig. 2
figure 2

a Ragone plot [16] for the important batteries systems; b the comparison of different metal air battery systems. The typical logarithmic axes of Ragone plot a is changed to logarithmic y and linear x in b in order to represent the differences between the metal–air batteries

Full size image

Si–air battery research was initiated by Prof. Yair-Ein-Eli and team at the Technion–Israel Institute of Technology in 2009 [17]. This novel type of primary battery had an ideal specific energy of E = 8470 Wh kg−1 and a long-term stable voltage of 1.0–1.2 V. It consisted of doped silicon wafers as fuel, an air cathode, and a specially synthesized ionic liquid (EMI·(HF)2·3F) as electrolyte.

2 Silicon–air cell construction

Yair-Ein-Eli et al. [17] used a silicon single-crystal anode, air cathode and specially synthesized room temperature ionic liquid (EMI·(HF)2·3F) as electrolyte. Graphic representation of cell assembly is shown in Fig. 3. The cell is made up of three plastic plates. The first plate is made of multiple pieces of silicon wafers of size 1 cm x1cm which were pressed into a viton O-ring. The exposed area of silicon wafers is 0.5 cm2. The second sheet consists of circular sheets of graphitic air electrodes having 0.5 cm2 exposed area. The third sheet is employed to make an outer circle periphery which is used to make electrical contact. High grade copper wires are used for making terminal contacts. Si wafer anodes were pre-treated before cell construction; the treatment involved immersion of Si-wafers in HF solution (1HF:5H2O) for 10 s to remove surface oxide layer followed by cleaning with de-ionized (DI) water and finally drying in a nitrogen stream. The anode-separator-cathode is sandwiched and 0.5 ml of the ionic liquid was added to complete the cell assembly. For adequate wetting of all the components, the cell assembly is maintained at OCV for 4 h during which no current was drawn from the cell assembly. The cell was later used to draw current to initiate the discharge process.

Fig. 3
figure 3

Silicon–air cell assembly [17]

Full size image

3 Anode for Si–air battery

3.1 Silicon wafers

Doped silicon wafers are widely used as anode material for Si–air batteries. Silicon <100> crystals, both moderately doped (n-type) and profoundly doped (n++-type), are held with a screwed back stainless steel contact plate electroplated with thin layer of gold [18]. The corrosion rates for various types of silicon material as anode is shown in Fig. 4.

Fig. 4
figure 4

Silicon electrode corrosion rates in EMI.(HF)2.3F [19]

Full size image

Silicon corrosion rates and polarization voltammograms form a basis for anode selection. On one hand, p++ silicon has lower corrosion rates and better cell voltage at high currents whereas on the other hand, n++ silicon has higher OCV and better cell voltages at low currents. Even though n++doped Si exhibited greater corrosion rates, it has been a subject of interest due to its greater cell voltage at low values of current.

3.2 Nanostructured silicon

Nanostructured silicon/arrays of silicon nanowires or nanorods have been used as anode material in lithium-ion batteries to overcome cracking and pulverization [20, 21]. A robust electrical contact is maintained when silicon nanowires are used as anode material since each nanowire is directly connected to the current collecting substrate. Mingyuan Ge et al. [22] reported that porous silicon nanowires synthesized by direct etching of boron-doped silicon wafers possesses high porosity and electron conductivity. As anode material, these nanowires exhibit high structure stability, higher electrochemical performance and long cycle life. Even after 250 cycles, the capacity remains stable above 2000, 1600, and 1100 mAh g−1 at current rates of 2, 4, and 18 A g−1 respectively. Zhong et al. [23] used nanostructured silicon as the anode material in the design of Si–air battery with alkaline electrolyte. Nano-engineered silicon anodes have been found to increase the reversible charge capacity and longer cycle life for new generation batteries. The surface of the silicon wafer is reformed by metal-assisted electro less etching method to facilitate formation of micro-porous layer on the silicon wafers [24,25,26,27]. The native oxide layer is removed by immersing the clean silicon pieces into a buffered oxide etching (BOE) agent.

3.3 Doped silicon

Silicon doped with Arsenic(As), Antimony(Sb) and Boron(B) with different orientations, i.e. <100> and <111> , was used as anode by Durmus et al. [28, 29]. The following six types of silicon anodes were used.

  1. i.

    As-doped [Crystal family type <100>] (Resistivity: 0.001–0.007 Ω cm)

  2. ii.

    As-doped [Crystal family type <111>] (Resistivity: 0.001–0.010 Ω cm)

  3. iii.

    Sb-doped [Crystal family type <100>] (Resistivity: 0.007–0.020 Ω cm)

  4. iv.

    Sb-doped [Crystal family type <111>] (Resistivity: 0.022–0.028 Ω cm)

  5. v.

    B-doped [Crystal family type <100>] (Resistivity: 0.001–0.005 Ω cm)

  6. vi.

    B-doped [Crystal family type <111>] (Resistivity: 0.002–0.016 Ω cm)

Two stage surface treatment of the doped wafer is carried out with argon/oxygen plasma to remove organic contaminants. The native oxide layers are removed from the wafer surfaces with the help of argon/sulfur hexafluoride plasma. The cell assembly is fabricated and kept for 4 h in a climate chamber which is maintained at 25 °C and 50% relative humidity. During this period no current is drawn and the silicon–air cells are held at their open-circuit voltage (OCV) before carrying out discharge process.

The 3D imagesFootnote 1 (Fig. 5) revealed that the surface characteristics of <100> vs. <111> oriented Si anodes vary significantly at low discharge currents. With both, arsenic-doped and antimony-doped Si anodes having orientation <100>, homogeneous 3D images are observed while with <111> oriented Si anode, 3D images consists of polygons. When cells are discharged at 0.1 mA cm−2 for 20 h, it is observed that the surface characteristics for corresponding orientation at mesoscale are almost similar for As- and Sb-doped Si anodes. 3D images of B-doped (p-type) and B-doped (n-type) Si anodes, under similar conditions of discharge, exhibited large differences in their surface characteristics. B-doped Si anodes with <100> orientation presented a homogeneous surface with a few individual small pores, while <111> oriented anodes showed a grain like structure.

Fig. 5
figure 5

3D images of As-doped Si electrode with <100> orientation. a optical microscopy image from the surface; b green square in a as imaged by AFM; c a line scan of green square position, as indicated in b; d 3D reconstruction of b

Full size image

Thus, the surface characteristics of B-doped Si anodes differ significantly from As- and Sb- doped Si anodes, indicating possibly two types of discharge mechanism at the anode surfaces. The studies on different anodes for Si–air batteries recommend that n++ doped Si anode may be a better choice for higher cell voltage and low current applications while nano-engineered Si anode can be used in applications where higher charge capacity and longer cycle life is needed predominantly.

4 Cathode for Si–air battery

4.1 Air cathode

Air cathode composition and structure play an important role in Si–air battery performance. Porous carbon imbibed with a catalyst, and a polymer binder is used as the air cathode. The specific capacity of an air electrode depends on several factors. Cheng et al. [30] and Xiao et al. [31] studied the effect of surface areas of different graphitic materials on specific capacity of the battery and it was found to be proportional to the surface area. Yang et al. [32] investigated the effect of pore size on specific capacity of Si–air cell and reported that air cathode with meso-porous meso-cellular carbon foam have higher specific capacity. Instead of too large or too small, optimum pore size is recommended. The thickness of the air cathode also plays an important role. Zhang et al. [33] found that the cells with thicker electrodes have lower specific capacity, which is endorsed to slow oxygen diffusion through thicker air cathodes.

Beattie et al. [34] reported that specific capacity of the cells is inversely proportional to the amount of carbon loaded in the cathode. Another factor that affects the electrochemical performance of an air electrode is oxygen diffusion through the air electrode, which can be facilitated by increasing oxygen pressure. Read et al. [35], Yang and Xia [36] and Tran et al. [37] reported that the specific capacity linearly increases with the oxygen pressure. The porosities, wettability and electrical contact of electrodes are determined by the ratio of carbon, catalyst, and polymer, which ultimately affects the electrochemical performance of Si–air cells.

4.2 Catalysts for air cathode

4.2.1 Noble metals

Noble metals such as platinum and Pt-based alloys have been used as oxygen reduction reaction electro catalysts for many decades. The nanoparticles of PtAu exhibited high stability and exemplary catalytic activity for carbon air cathodes of fuel cells and battery systems. Several studies [38,39,40,41] reported that ORR properties of Pt-based catalysts can be further improved by changing the size of the particles to nano scale and changing the surface characteristics of Pt-based nanostructures. When alloyed with other noble or transition metals, the performance of Pt based catalysts can be improved along with a reduction in cost [42,43,44,45,46,47].

4.2.2 Carbonaceous materials

Graphene and carbonaceous materials have been preferred as electro catalysts for air cathode of Si–air battery systems. These have found application as metal free catalyst or catalyst support because of their superior properties which include low cost, widespread availability, low electrical resistivity, very big surface areas and noble stability under severe conditions [48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64]. However, it has been reported that the electrochemical performance of these types of electro catalysts is highly dependent on their structure, morphology, pore size, and wall thickness. Appropriate pre-treatment of the surface of carbonaceous materials improves its functionality and enhance the durability of ORR catalysts.

4.2.3 Transition-metal oxides

High spin transition-metal complexes having spinels and perovskites morphology represents another class of cathode electro catalysts, highly useful in Si–air battery system [65,66,67]. These have shown excellent stability in alkaline medium whereas the stability in acidic electrolyte is relatively poor. Oxides of titanium as co-catalysts have further enhanced the activity of noble metals and its alloys with non-precious metals. Chen et al. [68] examined titanium nitrides and reported that these are highly efficient cathode electro catalysts. Transition metal carbides act as promising supports to promote the stability and activity of Pt- based ORR electro catalysts [69, 70]. The variable valence of transition metals such as Mn, Co, Fe and Ni gives rise to multiple structures to the oxides [71]. Mn-based oxides have been extensively studied as electro catalysts for air-electrodes. Electrochemically active manganese oxides such as MnO2, Mn2O3, Mn3O4, and Mn5O8 exhibits high catalytic activity [72,73,74,75].

4.2.4 Inorganic–organic composites

The integration of organic matrix with nano-sized metal constitutes inorganic–organic composites. It is another class of materials that are used as noble metal-free electro catalysts for oxygen electrochemistry [76,77,78,79,80,81,82,83,84]. The non-precious Ni/Fe metals easily form composite with polyoxometalate (POM) and o-Anisidine (oA) and have exhibited high catalytic activity and stability towards oxygen evolution reaction (OER). A competition between metal cation takes place to make complex with composite. Nanostructured NixFey(OH)2 powders, coated onto various substrate with the aid of polymeric binders, have shown high catalytic activity in alkaline electrolytes. Such composites can be prepared by pyrolysis of metal, nitrogen and carbon as these is not stable in strong acidic or alkali media. The catalytic activities of these composites depend on several parameters which include the metal, the precursors, the substrate and the synthesis method [85]. Nano engineered metal composites with heterocyclic conjugated polymers (polyaniline, polypyrrole and poly(3,4-ethylenedioxithio-phene) have also been used as electro catalysts [78, 86, 87]. These composites have advantage over other types of catalysts as the incorporation of metal ions in matrices creates large number of active sites [86]. Further, the bridges between matrix and metal promote physical and electrical contact [87]. Inorganic–organic composites such as FeCo-EDA catalysts have shown outstanding ORR activity in alkaline electrolytes and exhibits nearly 3 times higher mass activity than commercial Pt/C catalyst [88].

4.2.5 Hetero-atom doped carbons

Heteroatom (N & S) doped carbon nano-materials [89,90,91,92] are the most promising ORR catalyst as it markedly reduces the cost and increase the efficiency of metal–air battery systems/fuel cells. Vertically aligned nitrogen-doped carbon nanotube (VA-NCNT) arrays can act as a metal-free electrode to catalyze a 4e− ORR process with a three times higher electro catalytic activity and better long-term stability than commercially available Pt/C electrodes in an alkaline cell [89]. These carbon-based metal-free ORR catalysts are also free from the CO poisoning and methanol crossover effects. Choi et al. [93, 94] synthesized nitrogen and sulfur-doped carbons and studied their electro catalytic activity towards oxygen reduction reactions (ORR). The carbon catalyst synthesized from cysteine and dual-doped with nitrogen and sulfur, exhibited the highest onset potential (0.55 V vs. Ag/AgCl) and electrochemical activity in acidic media, − 0.2 mA (at 0.2 V vs. Ag/AgCl), which is about 43% of that of commercial Pt/C (40 wt%). From XPS studies they showed that in the carbon catalyst, sulfur was doped as sulfate or sulfonate. It was concluded that not only nitrogen doping but also sulfur doping of carbon improve its electro catalytic activity towards ORR. Zhang et al. [95, 96]. synthesized graphene-based B/N co-doped carbon nano sheets (G-CBP). It was argued that high surface area, unique 2D sheet nanostructure, and high heteroatom-doping contents (5.4% B and 5.3% N) of G-CBP results into an excellent electrochemical performance for the oxygen reduction reaction under alkaline conditions (0.1 M KOH).

4.3 Oxygen reduction reaction (ORR) in air electrode

Polytetrafluoroethylene (PTFE) with activated carbon black (0.45–0.5 g cm−2 loaded) is used as air cathode [17]. The mixture is encapsulated onto a nickel 200 mesh under high pressure and manganese dioxide is added as cathode catalyst. The reduction of oxygen occurring at the air electrode is said to proceed either as 4 electrons or 2 electrons as shown:

$${\text{O}}_{2} + \, 4{\text{e}}^{ - } \leftrightharpoons 2{\text{O}}^{2 - }$$
(1)
$${\text{O}}_{2} + \, 2{\text{e}}^{ - } \leftrightharpoons {\text{O}}_{2}^{2 - }$$
(2)

The scanning electron microscopy (SEM) micrographs of air cathode revealed that activated carbon on cathode is covered with sedimentation during the battery discharge.

Peter Jakes et al. [18] studied the role of manganese oxide (MnO2), which acts as an oxygen- reduction reaction (ORR) catalyst on Si–air battery performance. The chemical state of Mn ions in the MnO2 catalysts was examined by XP and EPR spectroscopy. It was inferred from the results that during the discharge process, MnO2 also undergo changes during an electrochemical reaction when the electrolyte has fluoride ions, commonly present in the RTIL.

$${\text{MnO}}_{2} + \, 8\left( {\text{HF}} \right)_{3} {\text{F }} + \, 2{\text{e}}^{ - } \leftrightharpoons {\text{MnF}}_{2} + \, 10\left( {\text{HF}} \right)_{2} {\text{F }} + \, 2{\text{H}}_{2} {\text{O}}$$
(3)

The formation of non-ORR catalyst MnF2 results in the loss of available catalytic sites and also lowers the ionic conductivity of the electrolyte. The change of MnO2 to MnF2 greatly affects the Si–air battery performance by limiting the discharge capacity and inhibiting the cell discharge of Si–air battery. The electrochemical changes at air cathode are quite complex and complete understanding will require extensive studies with different catalyst and electrode/electrolyte systems. However, studies on new catalyst and new ionic solvent may lead to a suitable system to better understand the phenomenon.

5 Electrolytes for Si–air battery

5.1 Room temperature ionic liquid (RTIL)

A room temperature ionic liquid (RTIL), EMI.(HF)2·3F is the most common type of electrolyte used in a Si–air battery. The EMI.(HF)2·3F is synthesised by a reaction of 1-ethyl-3-methylimidazolium chloride and hydrogen fluoride by Hagiwara et al. [97]. The electrolyte has low viscosity and high conductivity amongst all RTILs with 100 mS cm−1, besides chemical stability in air and high tolerance at extreme relative humidity conditions [98, 99]. The schematic of first experimental Si–Air Battery that utilised RTIL as the electrolyte is shown in Fig. 6.

Fig. 6
figure 6

Schematic diagram of a silicon–air battery [28, 29]; b A refill-type Si–air cell [28, 29]

Full size image

Potentiodynamic polarization studies of the RTIL-silicon wafer couples (anodic half-cell) and of the IL–air electrode (cathodic half-cell) is used to evaluate EMI.(HF)2.3F as a viable candidate for electrolyte in a Si–air battery is shown in Fig. 7.

Fig. 7
figure 7

Polarization voltammograms [19] recorded with RTIL [EMI·(HF)2·3F] at 5 mV s−1 for various electrodes: silicon single-crystal anodes and air cathodes

Full size image

The half-cell reactions, crystallization reaction and overall reaction when two anions of RTIL contribute to the oxidation of silicon at anode and reduction at air cathode are:

$${\text{Anode: }}\;{\text{Si }} + \, 12\left( {\text{HF}} \right)_{2} {\text{F}}^{ - } \leftrightharpoons 8\left( {\text{HF}} \right)_{3} {\text{F}}^{ - } {\text{SiF}}_{4} + \, 4{\text{e}}^{ - }$$
(4)
$${\text{Cathode: }}\;{\text{O}}_{2} + 12\left( {\text{HF}} \right)_{3} {\text{F}}^{ - } + 4{\text{e}}^{ - } \leftrightharpoons 16 \, \left( {\text{HF}} \right)_{2} {\text{F}}^{ - } + \, 2{\text{H}}_{2}$$
(5)
$${\text{Crystallization: }}\;{\text{SiF}}_{4} + 2{\text{H}}_{2} {\text{O}} + 4\left( {\text{HF}} \right)_{2} {\text{F}}^{ - } \leftrightharpoons {\text{SiO}}_{2} + 4\left( {\text{HF}} \right)_{3} {\text{F}}^{ - }$$
(6)
$${\text{Overall }}\,{\text{Cell }}\,{\text{Reaction:}} \;{\text{Si}} + {\text{O}}_{2 } \leftrightharpoons {\text{SiO}}_{2}$$
(7)

5.1.1 Effect of water in the RTIL

Cohn et al. [100] examined the effect of addition of water in the RTIL electrolyte. It has been concluded that discharge performance of Si–air battery system is greatly obstructed since during cell discharge, the SiO2 precipitates on air cathode’s porous carbon. The precipitation of SiO2 layer at a certain point of cell discharge prevents diffusion of oxygen which results into a short supply of oxygen at the electrolyte–electrode interface of air cathode. This results into lowering in discharge capacities of cathode than expected theoretically. Since water plays an important role in the formation of SiO2, it was argued that the formation of SiO2 may shift towards separator or anode by adding water into the electrolyte. If reaction is shifted to take place at the bulk electrolyte rather than at the electrolyte–electrode interface, an alternative location for the SiO2 formation can be introduced causing suffocation to the air–cathode. This is due to the loss of available catalytic sites on the porous surface of the air cathode for oxygen reduction.

The Si–air cells under study are discharged at a constant current density of 0.3 mAh cm−2 and the cut-off potential of 0.5 V. Variation in water content in the electrolyte did not have any affect in open circuit potential (OCP) for the Si–air cell. The dependence of discharge capacity of the Si–air cell is studied as a function of water content of RTIL at a fixed discharge current of 0.3 mAh cm−2. Initially, the discharge capacity increases with water content, reaching a maximum value of 72.5 mAh cm−2 at 15 vol% of water. On increasing water content beyond 15%, a sharp decrease (up to 40%) is observed in the cell discharge capacity up to 80% water content. Beyond 80% water content of RTIL, the cell could be operated as it did not display any discharge. From these results, it has been concluded that addition of water in RTIL may be having two opposing effect. One, it facilitates the formation of SiO2 in the bulk electrolyte and not at the air electrode due to which porosity of air electrode remains more or less intact. Second, the activity of RTIL in the electrolyte decreases on increasing the content of water resulting into loss of its distinctive properties of facilitating cathode reaction. With increment in the percentage of water in the electrolyte, the fraction of RTIL in the mixture decreases, and therefore, the unique properties of the RTIL are reduced.

5.1.2 Gel polymer electrolytes

Gel polymer electrolytes (GPEs) are formed by integrating a liquid electrolyte into a polymer matrix. Although GPEs are in solid state but the conduction in them is similar to a liquid electrolyte. The use of a gel electrolyte as a moisture barrier eliminates the need to handle liquid electrolyte and simplifies technical issues in cell architecture. Shape flexibility, mechanical stability, high safety and a modest loss in ionic conductivity during discharge process have made GPEs come into focus as electrolytes. They are commonly used in a variety of applications including Li-Ion batteries, solar cells and chemical sensors [101,102,103,104].

Yair-Ein-Eli et al. [105] explored the possibility of utilizing a composite polymer electrolyte in Si–air battery system. The gelled polymer electrolyte was derived using specially synthesized EMI.(HF)2.3F ionic liquid and 2-hydroxyethyl methacrylate (HEMA) as the matrix. The combination of the RTIL and polymer composites are also studied extensively by Tsuda et al. [106].

When the battery using GPE of 50–70 mol% ionic liquid is discharged at a current density of 0.1 mA cm−2 in an ambient atmosphere, it’s operating potential increases with increasing percentage of RTIL. It is observed that at 50 mol% RTIL, it is 0.4 V; at 60 mol% it is 0.5 V and at 70 mol% it is 0.6 V (discharge current was fixed at 0.1 mA cm−2). When compared with discharge potentials of the cells using pure RTIL, the operating potential of GPEs is ~ 0.5 V lower. It has been argued that difference in ionic conductivity of the GPEs is responsible for this variation in operating potential. For instance, it is observed that the ionic conductivity of 60 mol% RTIL is 23 mS cm−1 which is lower than the ionic conductivity of pure RTIL (100 mS cm−1).

Gelled electrolytes comprising of 40–70 mol% of EMI.(HF)2·3F RTIL and HEMA polymer are found to be free standing and mechanically stable. These GPE can also be well integrated with both Si anode and air cathode. The cell comprising of these electrolytes exhibited a long discharge time of 850 h at 0.1 mA cm−2 operative discharge current density. However, these cells have lower cell voltage when compared with the cells using pure RTIL, and it has been attributed to higher ionic conductivity of the later. Thus, use of GPEs as electrolyte in Si–air battery is feasible for low power devices.

5.1.3 Corrosion analysis of silicon anode with RTIL electrolyte

Jakobi et al. [28, 29] studied corrosion mass losses of As-, Sb-, and B-doped Si anodes at different discharge rates and varying discharge time (up to 20 h) (Fig. 8). It was concluded that the electrochemical etching increases surface area significantly along with formation of oxide over layer. The corrosion mass of all types of Si anodes increases with the increase in current density. The electrochemical etching is capable of producing a strong non-porous Si electrode incorporating thick porous layers and tiny pores.

Fig. 8
figure 8

Durmus et al. [28, 29]

Influence of dopant type and orientation of silicon anodes on corrosion of Si–air cells for a <100> oriented As-doped Si anode; b <111> oriented As-doped Si anode; c <100> oriented Sb-doped Si anode; d <111> oriented Sb-doped Si anode; e <100> oriented B-doped Si anode; f <111> oriented B-doped Si anode.

Full size image

Such Si anodes shall be favoured in alkaline Si–air batteries due to their enriched discharge behaviour in comparison to flat Si anodes. Further these high surface area nonporous Si anodes in room temperature ionic liquids are capable of minimizing strong passivation and self-discharge of Si anodes, enhancing its extended utilization, resulting into development of Si–air batteries having stable voltage profile. The increase in corrosion mass loss for current densities (up to 0.3 mA cm−2) is found to be linear for n-type Si anodes. The corrosion mass is significantly lower for n-type Si anodes with <111> orientation when compared with <100> orientation. Similar trend is found in the As- and Sb-doped Si anodes as well as in p-type Si anodes to n-type Si anodes. At higher current densities, the corrosion rate is significantly high. Discharge potentials for various dopants at varying current densities are found as detailed below.

  1. (a)

    For low current densities: As100> As111≈ Sb100> Sb111 ≫ B100> B111

  2. (b)

    For high current densities: As100> As111 ≫ Sb100> B100> B111> Sb111

The anode efficiency, which is dependent on corrosion of an anode, is an essential factor for screening of materials for Si–air batteries. Anode efficiency cannot be 100%; it is a factor that is dependent on the corrosion of the doped anode in the electrolyte. Anode mass conversion efficiency of differently doped Si-anodes for varying current densities are analysed and ranked as: Anodic mass conversion efficiency rankings: B100 ≈ B111 ≈ As111 > Sb111 ≫ Sb100 ≈ As100

On the basis of these findings, it has been concluded that <111> oriented As-doped silicon is the first-rate choice as the anode in Si–air batteries, employing EMI·(HF)2·3F as the electrolyte.

5.1.4 Mechanism of discharge termination in silicon–air batteries

The cell impedance during the discharge process of a Si–Air battery employing RTIL electrolyte has also been investigated [107]. Electrochemical Impedance Spectroscopy (EIS) is utilized to study the generation and development of Si anode and air cathode interfaces in the battery during storage and operation. The effect of exchanging the anode and cathode during discharge and at the end of the discharge is also studied. In order to determine which electrode is contributing to the cell impedance as the most dominant source, it is needed to obtain and analyse the impedance spectra of each electrode. The study of anode impedance exhibited two well-defined semicircles which correspond to high frequency and low frequency regions. The high frequency zone is attributed to space charge layer capacitance and a charge transfer resistance whereas the low frequency zone is attributed to the formation of thin oxide layer on the sub-micron pores during anodic reaction that adds to resistance and capacitance of the cells [108, 109]. Since the total impedance at the air cathode is considerably lower than the impedance at the anode, it can be neglected. Therefore, the overall cell impedance is originated from the silicon anode impedance.

The role of silicon in the potential drop of the battery during discharge is identified by employing a three-electrode configuration to the Si–air system. The initial drop of the air cathode is larger than the Si anode which can be attributed to dissolved oxygen in the electrolyte that is easily accessible and is consumed in the beginning of the operation but once more and more oxygen is consumed, the over potential of the air cathode increases and the voltage of the air cathode remains unmodified at 0.4 V. On the other hand, the over potential of the Si anode behaves differently. The over potential at electrodes increases with continuous discharge of the battery. The voltage profile of Si–air battery mimics the anode profile which clearly confirms the assumption that Si anode is the dominant electrode which controls and defines the discharge behaviour of a Si–air system. Hence, the capacity of the battery is limited by the Si anode rather than the air cathode.

The importance of the Si-anode with respect to discharge of Si–air battery is investigated by replacing the anode electrode during the discharge as well as at the end of the discharge. When the anode is replaced during the middle of the discharge, there is no effect on the discharge voltage. However, the discharge capacity almost doubles at the end of the operation of the second silicon anode. With each changing period, the capacity decreases to such an extent that after the fifth replacement, the increase in the discharge capacity is negligible as compared to previous exchanges. The reason why the fifth replacement does not give a significant increase on discharge capacity is because of the degradation of the air cathode and the electrolyte during the discharge process.

5.2 Alkaline electrolyte

A primary Si–Air battery employing nanostructured silicon as anode material with an alkaline solution based electrolyte was investigated by Zhong et al. [23]. Typical galvanostatic discharge of the cell exhibited that the cell can be used to draw current at 0.05 mA cm−2 over a period of 30 h at an operating potential of 1.2 V. In contrast, in a control experiment using an unmodified silicon wafer, the cell is completely discharged in less than 10 min at a potential of 1.1 V i.e. the potential quickly drops to zero. The electrode reactions are:

$${\text{Anode: Si}} + 4{\text{OH}}^{ - } \leftrightharpoons {\text{Si}}\left( {\text{OH}} \right)_{4} + \, 4{\text{e}}^{ - } {\text{E}}_{\text{o}} = 1.69\,{\text{V}}$$
(8)
$${\text{Cathode: }}\,{\text{O}}_{2} + 2{\text{H}}_{2} {\text{O}} + 4{\text{e}}^{ - } \leftrightharpoons 4{\text{OH}}^{ - } {\text{E}}_{\text{o}} = 0.40\,{\text{V}}$$
(9)

The main benefit of employing an alkaline solution is its ability to dissolve Si(OH)4. The surface can be covered by silicon oxide if the dissolution rate of Si(OH)4 is not fast enough. The formation of a porous surface structure increases the dissolution rate of Si(OH)4 in the KOH electrolyte. This facilitates the effective removal of the oxide and reactivates the Si-surface. On the other hand, the unmodified silicon surface become passive very quickly and results in a very short discharge time.

Various concentrations of the alkaline solution as electrolyte are studied. It is reported that the self-discharge rate is highly dependent on the electrolyte composition. It is observed that anodic dissolution potential is directly proportional to KOH concentrations (Table 1).

Table 1 Specific capacities of silicon–air batteries
Full size table

5.2.1 Corrosion analysis of silicon anode with alkaline electrolyte

Durmus et al. [110] studied the corrosion of As-doped (0.001–0.007 µcm) crystalline Si wafers in 0.5–12 M KOH solution. It is concluded that the chemical corrosion of Si is a surface reaction limited process. While in 0.3–5 M KOH, anisotropic etching leads to very rough surfaces and pyramidal hillocks, smoother surfaces at > 5 M KOH is attributed to polishing process. Hence, the anodic oxidation of As-doped Si does not depend on strength of KOH. Durmus et al. proposed that the electron transfer is the rate determining step of corrosion of As-doped Si using galvanostatic discharge data. They further suggested that the focus on the corrosion inhibition will help in improving the shelf-life and discharge performance of the primary Si–air batteries. It can be concluded that weak KOH enhances the shelf life while strong KOH will ensure continuous discharge at high current densities. The effect of KOH concentration on the corrosion rates of As-doped Si anodes can be understood by considering the following reaction mechanism:

$${\text{Si}} + 2{\text{OH}}^{ - } + 2{\text{H}}_{2} {\text{O}} \leftrightharpoons {\text{SiO}}_{2} \left( {\text{OH}} \right)_{2}^{2 - } + \, 2{\text{H}}_{2}$$
(10)

At 15 °C, Si undergoes corrosion @ 0.66 µm h−1, which increases to 1.46 µm h−1 at 25 °C. Further increase in temperature showed exponential increase of the corrosion rates, observed to be 13.62 µm h−1 at 60 °C. This behaviour is in accordance with Arrhenius equation which states that etching rate increases with temperature.

The variation of KOH changes the relative concentration of OH− and H2O in the solutions, which results in a change in their activities. In case of hydroxide ions, the activity increases almost linearly beginning from 1 M of KOH whereas KOH activity decreases [111,112,113].

6 Conclusion and outlook

This mini review is an effort to summarise recent developments of Si-anode, air cathode, different electrolytes and their effect on the discharge mechanisms and working potentials in a Si–air battery system. Despite the intense research, there is still much scope for further advancement of this novel type of Si–air technology. Prospective directions for future research in Silicon–air batteries could focus on the following:

  1. 1.

    Pure silicon wafers are the most common material for Si anode but the drawback is their high corrosion rates. It is required to find a Si anode that is capable to sustain its capacity over hundreds of discharge and charge cycles. Therefore, nano-structured silicon and doped silicon can be used to improve the performance of the Si anode. Structural modifications such as electro deposition or chemical modification by the means of additives has proven to be an ideal solution to find such an anode material. The additives must be effective in small quantities since a large amount of additive can reduce the overall capacity of the silicon material. More studies are needed to properly understand the charge and discharge mechanisms for the optimisation of the discharge time. A more effective methodology is needed to find solution for corrosion or passivation in Si–air battery.

  2. 2.

    The air cathode is an essential part of any type of metal–air battery but its kinetics of the oxygen reduction reaction hinders its performance. Therefore, electro catalysts play an important role in the cathode. Noble metals such as Pt and Pt-alloy based catalysts have been extensively used as electro catalysts in air electrode. To further reduce the cost, different type of catalysts such as high spin transition metal complexes and inorganic–organic composites can be used as a highly efficient alternative. It has been observed that the specific capacity of an air electrode depends on various factors. Thus, a suitable choice for an air electrode would be to choose a very thin and highly porous graphite nano-sheet with optimal pore size which is impregnated with an appropriate electro catalyst. Further development of low cost catalysts with improved activity and durability is necessary to improve the sluggish kinetics of air electrode. One such example of an electro catalyst could be inexpensive bi-functional catalysts which is active to both oxygen evolution reaction and oxygen reduction reaction.

  3. 3.

    New highly stable electrolyte technologies that allow long-lasting operation of Si–air batteries are required for further advancement. At present, Ionic liquids are the most promising electrolytes for Si–Air batteries. Thermally and mechanically stable gel polymer electrolyte appears to be a viable candidate as electrolyte for future research. Solidification of the electrolyte system is important to maintain the safe operation & easy handling of the battery and structural integrity under various deformations. It is also highly important to study the properties at the electrode–electrolyte interface of the battery.

  4. 4.

    Simulations are needed in order to provide an optimal design of Si–air batteries in terms of structural design of anode and cathode, as well as dispersion of electrolyte, electro catalyst, current collector and depolarizers in order to provide a corrosion-free contact between the electrodes and electrolyte. Moreover, the understanding of silicon dendritic growth in Si–air batteries is still in its infancy. Although there are many challenges faced, it is worth applying efforts to develop Si–air batteries for next-generation energy storage technology.