Water Electrolysis Anode Based on 430 Stainless Steel Coated with Cobalt Recycled from Li-Ion Batteries

Abstract

In this paper, a new environmentally-friendly anode for hydrogen production was developed based on 430 stainless steel with an electrodeposited cobalt layer. The novelty of this work is the cobalt source once the electrodeposition bath was obtained from acid dissolution of a spent Li-ion battery cathode. The oxygen evolution reaction on electrodeposited cobalt in 1 M KOH is compatible with the E. Kobussen mechanism. The water discharge is related with reaction determinant step in low overpotential. The cobalt electrodeposition (3 Ccm⁻²) promotes a significant improvement of 430 stainless steel anodic properties for oxygen evolution reaction. When the overpotential reaches 370 mV, the density current for 430 stainless steel with electrodeposit cobalt is 19 mA·cm⁻² against 0.80 mA·cm⁻² for 430 stainless steel without cobalt. Thus, the anode construction described in this paper is an excellent option for Li-ion battery recycling.

1. Introduction

Renewable energy source such as hydrogen is considered crucial for a more sustainable future [1]. In fact, since daily processes even space exploration will depend, largely of hydrogen production within few years [2]. The hydrogen is considered environmentally friendly once the reaction between H₂ with O₂ molecules (forward direction in Equation (1)) has only water molecules as a sub-product, beyond high molar enthalpy value (ΔH = 285.8 kJ/mol at 1013 mbar and 298 K) [2]. In this context, the water electrolysis is the most attractive route among the existed processes for very pure hydrogen production (and also O₂) in large scale (inverse direction in Equation (1)) [1].

H_2(g) +1/2O_2(g) → H₂O_(g) ΔH = 285.8 kJ/mol

(1)

In water electrolysis, the overpotential for oxygen evolution reaction (OER) is a limiting factor and responsible for the greatest energy loss source when the electrolysis is performed in alkaline or acid media [3,4]. However, alkaline media is preferred because in this condition the catalysts used in water electrolysis are more active and stable [4]. The OER also is present in others strategic power sources such as aqueous lithium air batteries [5], NiMH [3], and Zn-air [6]. Many papers are dedicated to investigate the alternative anode for OER with high efficiency [7]. The RuO₂ and IrO₂ are suggested as the benchmark for OER in alkaline solution [8]. However, these oxides have a prohibitive cost for large scale application [8]. Commercial electrodes for water electrolysis at an industrial scale must have important characteristics such as: high conductivity, high corrosion resistance, high catalytic effect for oxygen evolution reaction, and moreover low cost [9]. For Ni foam, the current density for OER is 10 mA·cm⁻² in η = 390 mV (vs. NHE) in 0.5 M KOH. When the Ni foam receives a Co layer by electroless deposition (~2.3 μm) the overpotential reduces for 270 mV [10]. This shows that a cobalt layer around 2.3 μm is enough to decrease the overvoltage for OER approximately in 100 mV. Although the Ni anode has a good anodic behavior in OER, its cost is still high compared to the other substrates such as stainless steels. The Ni price is 8.66 US$/Kg against 1.64 US$/Kg and 0.450 US$/Kg for 316SS and 430SS, respectively [11]. In this context, a special interest is devoted to stainless steel as anode for OER in alkaline solutions [9]. The literature revels that the 316 stainless steel (316SS) shows a density current of 10 mA·cm⁻² in η = 370 mV in 1.0 mol·L⁻¹ KOH [12]. This value is similar to that found for Ni Foam in the same conditions [12]. This is expected since the 316 has 12% of Ni in its composition [13]. The 430SS has little more than 0.75% of Ni in its composition and approximately the same percentage of others alloys elements [13]. The low Ni content makes the 430SS cheaper, however, it is also less active for OER. [13]. Nevertheless, is possible improve the 430SS anodic properties with cobalt electrodeposition. The great advantage of this method is the possibility of cobalt recovery from Li-ion battery recycling [14].

Many papers describe the cobalt recycling from spent Li-ion batteries as a valuable and environmentally-friendly method [14,15,16]. Moreover, very pure cobalt can be obtained by acidic dissolution of lithium cobalt oxide (LiCoO₂) present in spent Li-ion battery cathode [16]. The recycling of LiCoO₂ is important, both economically and environmentally [16]. In this paper, we studied the application of 430SS, coated with recycled cobalt, as anode for OER in 1 M KOH solution. The cobalt used was obtained through dissolution of spent cathode of cell phones Li-ion batteries. The electrochemical characterization of this anode was accomplished through cyclic and linear voltammetry and electrochemical impedance spectroscopy (EIS).

2. Material and Methods

2.1. Preparation of Cobalt Electrodeposition Bath

The Li-ion battery used in this paper was chosen based on the LiCoO₂ presence as cathodic material [16]. The Li-battery used was of SAMSUNG^® model SM-G530 (SANSUNG^®, China). The Li-ion battery was manually dismantled and physically separated into their different parts: anode, cathode, steel, separators, and current collectors. The cathode powder was washed with distilled water under agitation to facilitate the detachment of the active material from current collector. The Li-ion battery electrolyte is a nonaqueous solution composed of lithium salts, such as 1 M LiPF₆, in alkyl organic carbonates (ethylene, propylene, and dimethyl carbonates) [17]. Thus, the spent cathode was heated at 200 °C for 5 h for eliminate the organic carbonates [17]. The active material was filtered and washed with distilled water to remove possible lithium salts such as LiPF₆ and LiClO₄. The temperature of 40 °C degrees was maintained through a thermostatic bath. The recycled powder was dried in air for 24 h. A mass of 250.10 g of positive electrodes was dissolved in an aqueous solution containing 470.00 mL of 3.00 mol·L⁻¹ H₂SO₄ and 30.00 mL of 30% (v/v) H₂O₂. The LiCoO₂ acidic dissolution efficiency increases with acid concentration [17,18,19]. The H₂O₂ reduces the Co³⁺, insoluble in water, to the Co²⁺, soluble in aqueous solution. This behavior is represented by the chemical equation of the dissolution process (Equation (2)) [18].

2LiCoO_2(s) + 2H₂O_2(aq) + 3H₂SO_4(aq) → 2CoSO_4(aq) + 3/2O_2(g) + Li₂SO_4(aq) + 5H₂O_(l)

(2)

In order to obtain the cobalt electrodeposition bath, the system was maintained under constant magnetic agitation at 80 °C for 2 h. The filtration was performed for separation of carbonaceous material and for obtention of cobalt electrodeposition bath.

The concentrations of Mn, Ni, and Cu were analyzed with atomic absorption spectrophotometry (AAS) on a Hitachi-Z 8200 and the values were below of detection limit (10⁻⁵ mol·L⁻¹). The concentration of Co²⁺ was adjusted for 1 mol·L⁻¹ and the solid NaOH was used for adjust the pH = 3 [20]. The pH of electrodeposition bath was maintained equal to 3.0 and the cobalt concentration was 1.00 mol·L⁻¹. This high concentration was chosen to increase the electrodeposition efficiency [16,17,18,19,20].

2.2. Electrochemical Measurements

Electrochemical measurements were made using an IVIUM power supply (Ivium Technologies BV, De Zaale, The Netherlands). The working electrode was made of commercial ferritic stainless steel 430 (430SS). The steel samples were prepared as rectangular foils with a geometric area of 1.00 cm². The auxiliary electrode, with an area of 3.75 cm², was made of platinum. A saturated calomel electrode (SCE) reference electrode was used. The AC-impedance measurements were carried out over the frequency range of 10⁴ to 10⁻² Hz at the rest potentials. The Nyqusit diagrams were adjusted with the IVIUMSOFT. The conversion for reversible hydrogen electrode (RHE) was made using the Equation (3) and the overpotential (η) was calculated using Equation (4) [9].

$${\mathrm{E}}_{\left(\mathrm{RHE}\right)}={\mathrm{E}}_{\left(\mathrm{SCE}\right)}+\left(0.059\mathrm{pH}+0.042\right)$$

(3)

$$\mathsf{\eta }={\mathrm{E}}_{\left(\mathrm{SCE}\right)}+\left(0.059\mathrm{pH}+0.042\right)-1.23\mathrm{V}$$

(4)

The working electrodes were sanded with 600-grit sandpaper before each measurement and washed with distilled water. All the electrochemical measurements were performed without solution agitation, at 25 °C. The bath used in cobalt electrodeposition was obtained by acidic dissolution of spent cathode of Li-ion batteries. The potential used for electrodeposition of cobalt onto 430SS was −1.0 V. The charge density was controlled in 3 Ccm⁻² [9]. The scanning electron microscope JEOL JXAmodel 8900 RL (JEOL, Peabody, MA, USA) was used for surface morphology observations. The cobalt electrodeposit was characterized by X-ray diffraction on a 200 B Rotaflex-Rigaku (Rigaku, Tokyo, Japan) with Cu Kα irradiation, a Cu filter, and scanning speed of 2°·min⁻¹.

3. Results and Discussion

3.1. Characterization of Electrodeposited Cobalt onto 430SS

Figure 1 shows the cyclic voltammetry obtained by electrodeposition of cobalt on 430SS. The starting potential was −0.5 V (rest potential). Sweeping in the direct direction is observed an increase of the cathodic current from −1 V. Some studies of cobalt electrodeposition in $\mathrm{pH}\ge 4$, proposes that the mechanism involves the formation of Co(OH)₂ and the subsequent reduction of cobalt hydroxide to metallic cobalt [17,18,19,20]. Considering the pH used in this study (pH = 3), the presence of other phases, such as Co(OH)₂, in the cobalt electrodeposit was not expected [20].

Figure 2 shows the scanning electron microscopy (SEM) of electrodeposited cobalt. The morphology of electrodeposited cobalt is commonly observed in other papers [15,16].

The Figure 3 shows the X-ray diffraction of 430SS with (black line) and without (red line) electrodeposited cobalt. The XDR was compared with the JCPDS card for cobalt in hexagonal compact phase (hcp) (no. 05-0727) and face centered cubic phase (fcc) (no. 15-0806). The presence of both hcp and fcc phases in electrodeposited cobalt also described in the literature [16,17,18,19,20]. No peaks related to the Co(OH)₂ phase were observed, which may be an indication that the cobalt electrodeposition mechanism does not involves this intermediate [20].

Figure 4 shows the cyclic voltammetry of 430SS with cobalt coating (430SS/Co) in 1.0 mol·L⁻¹ KOH. The first peak (A₁) is due the formation of Co(OH)₂ layer onto cobalt previously electrodeposited [14,21]. The peak A₂ is related to oxidation of Co²⁺ to Co³⁺ due formation of monometallic layered double hydroxides (LDH) onto electrodeposited cobalt [21]. The A₂ can also is related with Co₃O₄ formation [22,23]. The peak A₃ is relative to electrochemical reaction related to conversion of Co³⁺ to Co⁴⁺ [14]. The formation of Co⁴⁺ is important because in alkaline media the OER generally involves the OH⁻ adsorption as first step (Equation (5)). Thus, a transition metal ion with multiple valences and strong bonding power is necessary [21]. Equation (5) represents the adsorption of OH^- by charged transition metal (Mⁿ⁺) in the hydroxide/oxide crystalline net.

Mⁿ⁺_(net) + OH⁻_(aq) → MOHⁿ⁺_(ads) + e⁻

(5)

3.2. Mechanism of Oxygen Evolution

Before study the electrochemical behavior of 430SS coated with recycled cobalt as an anode in water electrolysis, some theoretical considerations need to be performed [24,25]. The most commonly proposed OER pathway for Co oxides is due to the E. Kobussen mechanism [24]. For this mechanism, only the first two steps have significant importance [24,25]. Initially, the hydroxyl ions are absorbed in the active sites, where ‘S’ is the non-occupied active sites (Equation (6)). The reaction determining step (rds) is the water discharge, where ‘SOH’ is the occupied active sites (Equation (7)). If water discharge (Equation (7)) is the rds, the first step (Equation (6)) is in quasi-equilibrium conditions. The anodic and cathodic current are approximately equals, thus the Nerstian equilibrium can be considered in terms of occupied (SOH = θ) and non-occupied (S = 1 − θ) sites. The total current (${\mathrm{i}}_1$) for first step can be represented by Equation (8) where:

F: The Faraday constant.

η: the overpotential.

k₁ and k₋₁: The direct and inverse velocity constant rate, respectively, for the first step.

R: The ideal gas constant (8.314 J·mol⁻¹).

T: The absolute temperature (K).

α: The symmetry coefficient of charge transfer barrier.

$$\mathrm{S}+{\mathrm{OH}}^{-}\begin{array}{c}\to \\ \leftarrow \end{array}\mathrm{SOH}+{\mathrm{e}}^{-}$$

(6)

$$\mathrm{SOH}+{\mathrm{OH}}^{-}\begin{array}{c}\to \\ \leftarrow \end{array}\mathrm{SO}+{\mathrm{e}}^{-}+{\mathrm{H}}_2\mathrm{O}$$

(7)

$${\mathrm{i}}_1={\mathrm{Fk}}_1\mathsf{\theta }\left[{\mathrm{OH}}^{-}\right]{\mathrm{e}}^{\left[\frac{\left(1-\mathsf{\alpha }\right)\mathrm{F}\mathsf{\eta }}{\mathrm{RT}}\right]}-{\mathrm{Fk}}_{-1}\left(1-\mathsf{\theta }\right){\mathrm{e}}^{-\left[\frac{\mathsf{\alpha }\mathrm{F}\mathsf{\eta }}{\mathrm{RT}}\right]}$$

(8)

Since the current represented by the Equation (8) is approximately zero, one can obtain the relation of non-occupied and occupied sites on the electrode surface (Equation (9)).

$$\frac{\mathsf{\theta }}{\left(1-\mathsf{\theta }\right)}=\frac{{\mathrm{k}}_1}{{\mathrm{k}}_{-1}}\left[{\mathrm{OH}}^{-}\right]{\mathrm{e}}^{\left[\frac{\mathrm{F}\mathsf{\eta }}{\mathrm{RT}}\right]}$$

(9)

When the ΔH of adsorption varies with surface covered, the Temkin isotherm more appropriately describes the electrochemical interface. To represent coverage-dependent deviations in adsorption free energy from Langmuir conditions (Equation (9)), the covering dependent function rθ is added where r is a coefficient of increase in adsorption energy (Equation (10)) [24,25].

$$\frac{\mathsf{\theta }}{\left(1-\mathsf{\theta }\right)}={\mathrm{K}}_2\left[{\mathrm{OH}}^{-}\right]{\mathrm{e}}^{\left[\frac{\mathrm{F}\mathsf{\eta }}{\mathrm{RT}}\right]}{\mathrm{e}}^{\left[\frac{-\Delta {\mathrm{G}}^0+\mathrm{r}}{\mathrm{RT}}\right]}$$

(10)

If one considers a region in which 0.2 < θ < 0.8 the variation of $\ln \left(\frac{\mathsf{\theta }}{1-\mathsf{\theta }}\right)$ is much less than that of the term ${\mathrm{e}}^{\mathrm{r}\mathsf{\theta }/\mathrm{RT}}$, thus the term in parenthesis in the Equation (10) is practically constant. The function rθ can be given as a function of the overpotential and the concentration of hydroxyl ions (Equations (11) and (12)).

$$\mathrm{r}\mathsf{\theta }=\left[\ln \left(\frac{\mathsf{\theta }}{\left(1-\mathsf{\theta }\right)}\right)\mathrm{RT}+\mathrm{RTln}-\Delta {\mathrm{G}}^0+\mathrm{RTln}\left[{\mathrm{OH}}^{-}\right]\right]+\mathrm{F}\mathsf{\eta }$$

(11)

$$\mathrm{r}\mathsf{\theta }={\mathrm{K}}_3+\mathrm{F}\mathsf{\eta }$$

(12)

$${\mathrm{K}}_3=\left[\ln \left(\frac{\mathsf{\theta }}{\left(1-\mathsf{\theta }\right)}\right)\mathrm{RT}+\mathrm{RTln}-\Delta {\mathrm{G}}^0+\mathrm{RTln}\left[{\mathrm{OH}}^{-}\right]\right]$$

The equation for the rate determining step (rds) can be written as shown in the Equation (13) where δ is a symmetry factor (0 < δ < 1) for the adsorption process, which depends in magnitude and form of the activation energy barrier. Replacing Equation (12) in Equation (13), the Tafel coefficient is obtained for rds considering a possible situation where $\mathsf{\alpha }=0.5$ and $\mathsf{\delta }=0.5$ (Equation (14)).

$${\mathrm{i}}_2{\mathrm{Fk}}_2\mathsf{\theta }\left[{\mathrm{OH}}^{-}\right]{\mathrm{e}}^{\left[\frac{\left(1-\mathsf{\alpha }\right)\mathrm{F}\mathsf{\eta }}{\mathrm{RT}}\right]}{\mathrm{e}}^{\left[\frac{\left(1-\mathsf{\delta }\right)\mathrm{r}}{\mathrm{RT}}\right]}$$

(13)

$$\left(\frac{{\mathrm{dlni}}_2}{\mathrm{d}\mathsf{\eta }}\right)=\frac{\mathrm{F}\mathsf{\eta }}{\mathrm{RT}}$$

(14)

Figure 5 compares the linear voltammetry of 430SS and 430SS/Co in a 1.0 mol·L⁻¹ KOH. The cobalt electrodeposition significantly reduces the onset potential for oxygen evolution reaction (OER). The onset potential for OER onto 430SS occurs around η = 380 mV. By other hand, for 430SS/co the onset potential is around η = 280 mV. The onset potential was compatible with that obtained in the literature for metallic cobalt [25]. The Tafel coefficient $\left(\frac{\mathrm{dlni}}{\mathrm{d}\mathsf{\eta }}\right)$ was calculed for both electrode and for 430SS. Only one distinct region was found with $\frac{\mathrm{dlni}}{\mathrm{d}\mathsf{\eta }}=$ 38.94 V⁻¹ (~59 mV·dec⁻¹). Considering T = 298 K, this Tafel coefficient agree with rds related with water discharge (Equation (14)). For 430SS/Co, two distinct regions are found in linear voltammetry. The first region agrees with water discharge as rds (59 mV·dec⁻¹). The second region has a Tafel coefficient around 118 mV·dec⁻¹. In this region there is a high degree of coverage of adsorbed species [25]. Thus, the rds for the OER is mainly due to the step described by Equation (6). One can observe that the cobalt electrodeposition (3 Ccm⁻²) promotes a significant improvement of 430SS anodic properties for OER. When the overpotential reaches 370 mV, the density current for 430SS/Co is 19 mA·cm⁻² against 0.80 mA·cm⁻² for 430SS. The 430SS/Co exhibits the double of density current to OER if compared with 316SS [9] or Ni foam [12]. Considering the linear voltammetry of metallic cobalt in 1 M KOH, the literature describes a value around 10 mA·cm⁻² of density current of OER when the overpotential is around 370 mV [25]. Probably due to a surface area effect, the density current found in this work is approximately double. This shows the advantage of working with cobalt electrodeposition.

3.3. Electrochemical Performance

The electrochemical interfacial process is better investigated using electrochemical impedance spectroscopy (EIS). Figure 6 and Figure 7 compares the EIS measures of 430 stainless steel (430SS) and 430 stainless steel with recycled metallic cobalt coating (430SS/Co) in a 1.0 mol·L⁻¹ KOH after cyclic voltammetry study. The R₁ is related with the resistance of ionic motion in the solution bulk, R₂ and R₃ are the polarization resistances and C_dl is a non-ideal capacitance of double layer. For EIS of 430SS (Figure 6) the interface is compatible with Randles circuit [24]. However, for 430SS/Co (Figure 7) the LDH formed onto electrodeposited cobalt behaves as a supercapacitor, thus two constant-phase elements must be taken into account as shown in Figure 7 [21].

4. Conclusions

In this paper, a new anode for hydrogen production was developed based on 430 stainless steel with an electrodeposited cobalt layer. The novelty of this work is the cobalt source once the electrodeposition bath is obtained of acid dissolution of Li-ion batteries spent cathode. The electrodeposited cobalt in 1 M KOH behaves as supercapacitor. Considering the linear voltammetry in 1 M KOH, when the overpotential reaches 370 mV, the anodic density current for 430SS/Co is 19 mA·cm⁻² against 0.80 mA·cm⁻² for 430SS. Thus, the anode construction described in this paper is an excellent option for Li-ion battery recycling. Future work must be performed to study the reproducibility of the OER catalytic performance once the materials collected from spent batteries vary in composition.