New strategy toward enhanced air electrode for Li–air batteries: apply a polydopamine coating and dissolved catalyst

Abstract

In this article, we report a new strategy for enhancing the electrochemical performance of Li–air batteries through the combined use of polydopamine-coated carbon nanotubes (CNTs) and a dissolved LiI catalyst. The polydopamine layer modified the CNT surface from hydrophobic to hydrophilic, improving the wettability of the CNT-based electrode. This, in turn, increased the area of the electrode that actively participates in the reaction between the lithium ions and the dissolved oxygen in the electrolyte. LiI contained in the electrolyte effectively decreases the overpotential of the Li–air cell, by freely accessing the whole surface of the wetted electrode. We introduced the polydopamine-coated CNTs and LiI-doped electrolyte as an alternative to electrodes composed of carbon and solid catalysts. The electrochemical properties of this new system, including its capacity, overpotential, and cyclic performance, were characterized.

---

1. Introduction

Li–O₂ batteries comprise one of the most promising new technologies among the various next-generation energy storage systems because their energy density is several-fold higher as compared to that of commercial Li-ion batteries.^1–10 Several types of Li–O₂ batteries have been developed according to the electrolytes employed, such as aqueous, non-aqueous, and hybrid systems. Because of its relatively simple configuration and large real capacity, a non-aqueous system is the most active among the known types of Li–air batteries.^11–18 Significant research has been devoted towards the performance enhancement of non-aqueous Li–air cells.^18–23 However, they still face fundamental and practical challenges such as low rate capability, limited cycle life resulting from the instability of the electrode and electrolyte, and significant overpotential on charging due to the slow kinetics of dissociation for the reaction products (such as Li₂O₂).^24–29 These critical issues are highly attributable to the air electrode, the reaction place between lithium ions and oxygen, and the direct contact of the electrolyte with the electrode. With the aim of enhancing the properties of non-aqueous Li–air batteries, numerous studies have focused on modifications of the air electrode and the electrolyte.^18–29

In this study, the combination of polydopamine-coated carbon and a dissolved catalyst (LiI) was postulated as a new strategy toward an enhanced air electrode for Li–air batteries. A polydopamine layer was introduced to enhance the wettability of the air electrode (composed of carbon) with the electrolyte. In practice, the capacity of the air electrode is not proportional to the mass of the catalyst employed in the electrode, but rather, the area of the two-phase boundary between the electrolyte (containing Li ions and dissolved oxygen) and the electrode.¹⁵ A second point to consider is that an electrode immersed deeply in the electrolyte cannot actively participate in the catalytic reaction, because it cannot be supplied with enough oxygen for efficient reaction with lithium ions due to the low solubility and mobility of oxygen in the electrolyte.^11,12 So, a wetted electrode surrounded by a thin electrolyte layer is ideal because it can maintain the appropriate oxygen diffusion length and concentration through the surface of the electrode, and thus facilitate an efficient reaction between oxygen and lithium ions.¹⁵ To approach the ideal wetted electrode by increasing the wettability between the carbon-based air electrode and the electrolyte, we proposed a polydopamine coating on the surface of the carbon. A polydopamine coating is notable for being very effective in modifying the properties of various organic and inorganic surfaces.^30–33 In particular, a polydopamine coating can enhance the wettability of a coated surface with organic electrolytes, which have been used in Li battery systems.³³ The enhanced wettability of the air electrode with the electrolyte is expected to increase the homogeneously wetted area, which acts as the active reaction area (reaction sites), and improves the catalytic activity of the air electrode.

Polydopamine coating may be an effective method to enhance the wettability of the carbon-based electrode. However, the polydopamine layer may deteriorate the catalytic activity of the carbon surface because it is essentially a non-conducting material. Moreover, carbon can function well as an oxygen reduction reaction (ORR) catalyst, but it is not effective for the oxygen evolution reaction (OER). This may increase the overpotential on charging. To support the electrode based on carbon and strengthen the advantage of the polydopamine-coated carbon with good wettability, LiI was used as a dissolved catalyst. LiI can act as an electron–hole transfer agent between reaction products such as Li₂O₂ and the electrode, such that LiI can be considered a redox mediator.^34–36 Therefore, the use of LiI is expected to facilitate the oxidation of the solid reaction products, and reduce the overpotential of the air electrode on charging. Moreover, the dissolved LiI can freely access the whole surface of the electrode wetted by the electrolyte, and can efficiently take advantage of the polydopamine-coated carbon.

In this study, carbon nanotubes (CNTs) were used as the carbon source in the air electrode and were coated by a polydopamine layer. An oxide catalyst was not used for the air electrode in order to characterize the effect of the surface-modified carbon on the electrochemical performance of the cell. Instead, LiI was introduced as a dissolved catalyst to enhance the catalytic reaction of the air electrode. The combination of polydopamine-coated CNTs and LiI is expected to improve the electrochemical performance due to enhanced wettability by the electrolyte and the effective catalytic activity of the LiI.

2. Experimental

The polydopamine-coated CNTs were prepared by the simple immersion of commercial CNTs (CNT-90, Hanwha Chemical Corp.) into a solution of dopamine hydrochloride (Sigma Aldrich) containing Tris buffer (10 mM, pH 8.5, Sigma Aldrich) and methanol as co-solvents (CH₃OH–buffer = 1 : 1 v/v). The solution was mechanically stirred for 30 min at room temperature until well dispersed. Then, the suspension was centrifuged and the solid was subsequently washed with distilled water and ethanol. The microstructures of the polydopamine-coated CNTs were observed by transmission electron microscopy (TEM, AP Tech Tecnai G2 F30 S-Twin). X-ray photoelectron spectroscopy (XPS, PHI 5000 VersaProbe™ (Ulvac-PHI)) analysis was used to confirm the presence of the dopamine coating on the CNTs.

The air electrode was prepared by mixing 90 wt% polydopamine-coated CNTs and 10 wt% poly(vinylidene fluoride) (PVDF) binder. The loading weight of the electrode was 0.5 ± 0.05 mg. To check the wettability of the electrode, contact angles between the electrode and water droplets were investigated. The electrochemical performance of the electrodes was examined using a modified Swagelok cell consisting of an air electrode, a metallic Li anode, a Whatman glass filter/separator, and an electrolyte comprising 1 M bis(trifluoromethane)sulfonimide lithium salt (LiTFSi) in tetra(ethylene glycol)dimethyl ether (TEGDME). When LiI was used as the catalyst, 0.05 M LiI was dissolved into the electrolyte. The cells were assembled in an Ar-filled glove box and were subjected to galvanostatic cycling using a charge–discharge system. The experiments were carried out under O₂ (1 atm) in an air system. X-ray diffraction (XRD, Philips) was used to analyze the reaction products on the surface of the electrode during the initial cycles, and scanning electron microscopy (SEM, AP Tech Tecnai G2 F30 S-Twin) was employed to observe the surface morphology of the electrode during cycling. Fourier transform infrared spectroscopy (FTIR) spectra of the electrodes were also collected on an FTIR-4200 (Jasco) to assess the reaction products that accumulated during cycling.

3. Results and discussion

The morphologies of the pristine and polydopamine-coated CNTs were observed and compared by TEM. Fig. 1a and c shows typical pristine multiwall-CNTs having smooth side walls. In contrast, the polydopamine-coated CNTs are covered by a thin-film shell (the expected polydopamine layer), which has a lighter colour than the side walls of the regular CNTs due to its lower density (Fig. 1b and d). In the higher-resolution image (Fig. 1f), the polydopamine layer is clearly observed. The thickness of the polydopamine layer was several nanometers, and could be controlled by the time allowed for self-polymerization in the buffer solution. In this study, the prepared polydopamine layer was kept as thin as possible because of its non-conducting nature which could deteriorate the catalytic activity of the carbon and hinder the movement of electrons from the carbon. The composition of the surface of the CNTs (highlighted by the red rectangles in Fig. 1g and h) was investigated by energy-dispersive X-ray spectroscopy (EDS). The insets in Fig. 1g and h show that the polydopamine-coated CNT surfaces contain a relatively higher content of N than the surface of a pristine CNT, which confirms the presence of the polydopamine layer. The formation of the coating was also confirmed by XPS analysis. Fig. 2 shows the XPS spectra of the CNTs before and after polydopamine treatment. The pristine CNTs exhibit only peaks corresponding to C, whereas the polydopamine-coated CNTs exhibit new peaks attributed to N and O. These newly detected peaks are coincident with the reported XPS peaks of polydopamine,³⁷ thus verifying the formation of the polydopamine layer on the surface of the CNTs.

Fig. 1 TEM images of (a and c) pristine CNTs and (b and d) polydopamine-coated CNTs. High resolution TEM images of (e) pristine CNTs and (f) polydopamine-coated CNTs. EDS results for (g) pristine CNTs and (h) polydopamine-coated CNTs.

Fig. 2 XPS spectra of pristine CNTs and polydopamine-coated CNTs (coated for 30 min and 4 h).

To characterize electrochemical performance, air electrodes employing pristine and polydopamine-coated CNTs were prepared. For convenience, hereafter, we refer to the electrode employing polydopamine-coated CNTs as the “dopa electrode,” and the electrode employing untreated CNTs as the “pristine electrode.” Before the electrochemical tests, the wettabilities of the electrodes were observed to examine the effect of the polydopamine layer. Fig. 3a and b shows images of the contact angle of a water droplet on the air electrode surface. As expected, the wetting angle of the pristine electrode is relatively high (∼83°) due to the hydrophobic surfaces of the CNTs. However, the angle is significantly decreased to ∼42° by applying the polydopamine coating on the CNT surfaces, indicating the modification of the air electrode surface from hydrophobic to hydrophilic. Our efforts to observe the contact angle of a droplet of the TEGDME-based electrolyte were thwarted by the easy infiltration of the electrolyte into the porous electrode: the TEGDME-based electrolyte also is hydrophilic in nature. However, TEGDME has lower polarity than water, so it shows better wettability on the surface of carbon having low polarity.¹⁸ As a method to show the effect of the enhanced wettability of the polydopamine-coated CNTs, we observed how far the electrolyte spread on a flat electrode. If the electrolyte droplet spread more widely on the surface of a flat electrode, it can be inferred that a wider area of the porous electrode (prepared on Ni mesh) could be wetted by a thin electrolyte layer. Flat electrodes employing pristine and polydopamine-coated CNTs were prepared on the surface of flat copper foil and pressed smooth. A small amount of electrolyte (40 μL) was dropped on the surface of the flat electrode and its spread was observed. As shown in Fig. 3c and d, the wetted area of the flat dopa electrode was ∼32 cm², whereas that of the flat pristine electrode was just ∼18 cm². This indicates that a wider area of the dopa electrode can be wetted by a thin electrolyte layer, which may increase the active reaction area between the lithium ions and oxygen dissolved in the electrolyte.

Fig. 3 Contact angle of a water droplet on the (a) pristine and (b) dopa electrodes. Electrolyte spread on the surface of the (c) flat pristine and (d) flat dopa electrodes.

Fig. 4 shows the initial discharge–charge profiles of the pristine and dopa electrodes at current densities of 400, 600, and 1000 mA g⁻¹ in the voltage range 2.35–4.35 V. The capacity values reported in this article are based on the total electrode mass (CNT + binder), which is reasonable for the presentation of energy storage ability. As shown in Fig. 4a–c, the discharge and charge capacities of the dopa electrode were higher than those of the pristine electrode at all current densities. Moreover, the overpotential of the dopa electrode was slightly lower than that of the pristine electrode (Fig. 4d). This result may be attributed to the enhanced wettability of the dopa electrode. The electrolyte can be wetted by a thin layer over a wider surface area of the dopa electrode than the pristine electrode, which leads to a wider area of participation in the reaction between the lithium ions and dissolved oxygen for the dopa electrode. This wider active reaction area may increase the capacity and decrease the overpotential. As mentioned before, the possibility that the polydopamine layer acts as an obstacle for the catalytic reaction cannot be excluded because it is essentially a non-conducting material. However, the initial discharge–charge profiles show that the positive effects of the polydopamine layer (such as enhanced wettability) can compensate for the negative effect of non-conduction.

Fig. 4 Initial discharge–charge profiles of the pristine and dopa electrodes at current densities of (a) 400, (b) 600, and (c) 1000 mA g⁻¹. (d) Comparison of overpotentials measured at 400 mA g⁻¹ (capacity was limited to 1000 mA h g_electrode⁻¹).

To assess the phase structure of the reaction products, an XRD analysis of the pristine and dopa electrodes was performed during the initial cycling process, as shown in Fig. 5. The electrodes were collected at the measuring points indicated at the bottom of the figure. For both electrodes, Li₂O₂ is clearly formed during the initial discharge process and was dissociated during the initial charge process. Broad Li₂O₂ peaks indicate that the reaction products contain a quasi-amorphous phase as well as a crystalline phase. The diffraction peaks for the dopa electrode in the fully discharged state seem to be more intense than those of the pristine electrode, which is likely due to the higher discharge capacity that results in a larger quantity of reaction products (Li₂O₂). It is clear that the polydopamine layer on the surface of the dopa electrode does not disturb the formation and dissociation of reaction products such as Li₂O₂.

Fig. 5 XRD patterns of the electrodes during initial cycles: (a) pristine and (b) dopa electrodes. The lower portion of the figure indicates the points of measurement.

From these studies, we can confirm that the polydopamine coating on the surface of the CNTs effectively increases the capacity of the electrode. However, carbon is not effective for the dissociation of reaction products during the charging process (OER), which leads to a high overpotential and limited cycling performance. To support the electrode based on carbon and facilitate the dissociation of reaction products, LiI was introduced as a catalyst. Li dissolves in the electrolyte and can freely access the reaction products through the electrolyte. These LiI properties are expected to create a synergic effect with the good wettability of the polydopamine coating. Fig. 6 shows the initial discharge–charge profile of the electrode at a current density of 1000 mA g⁻¹ in the voltage range 2.35–4.35 V. As expected, the overpotential during the charging process is significantly decreased due to the effect of the LiI catalyst. In particular, when the dopa electrode was tested using the LiI-containing electrolyte, both a high capacity and low potential could be obtained. The average voltage difference between the discharge and charge processes seems to be less than 1 V. Considering that the cell was measured at a high current density (1000 mA g⁻¹), this result constitutes a very meaningful value.

Fig. 6 Comparison of the initial discharge–charge profiles of the pristine and dopa electrodes, measured using plain and LiI-containing electrolytes (current density = 1000 mA g⁻¹).

Fig. 7 presents the initial discharge–charge profiles of the pristine and dopa electrodes cycled using the LiI-containing electrolyte at current densities of 400, 600, and 1000 mA g⁻¹ in the voltage range 2.35–4.35 V. When the profiles are compared with those of Fig. 4, the overpotential is clearly decreased at all current densities because of the catalytic effect of LiI. As shown in Fig. 7b, the polydopamine layer still effectively increases the capacity of the cell tested using the LiI-containing electrolyte. The average voltage difference between the discharge and charge profiles measured at 400 mA g⁻¹ was just 0.7–0.8 V. Considering the capacity and overpotential, it is clear that the adoption of the polydopamine-coated carbon and dissolved catalyst (LiI) system constitutes a new approach for an enhanced air electrode with high catalytic activity.

Fig. 7 Initial discharge–charge profiles of the electrodes measured using LiI-containing electrolyte at current densities of 400, 600, and 1000 mA g⁻¹: (a) pristine and (b) dopa electrodes.

Typically, the air electrodes for Li–air batteries have been prepared based on porous carbon, which acts as the reaction site between lithium ions and oxygen. One thing we should consider is that the oxygen is dissolved in the electrolyte solution before its reaction with lithium ion. Therefore, the electrolyte filling of the air electrode can be an important factor affecting the overall performance of Li–air batteries. Fig. 8a illustrates a typical electrode based on carbon material. In practice, some part of the electrode is immersed in the electrolyte and some part is exposed to the air. Generally, the electrolyte filling of the electrode is insufficient because the electrolyte is essentially hydrophilic whereas the carbon presents a hydrophobic surface. The exposed part of the electrode cannot participate in the reaction. Furthermore, the reaction in the deeply immersed part of the electrode is also negligible, because the solubility of oxygen in the electrolyte is too low and the mobility of the dissolved oxygen is also poor, leading to the lack of dissolved oxygen in the deeply immersed part. Therefore, the active reaction area (indicated by the red rectangle in Fig. 8a) will be limited to just a portion of the electrode area. Another point to consider is the contact problem between the reaction products and the air electrode during the charging process. Most of the air electrodes contain solid catalysts such as MnO₂, Co₃O₄, Pt, Ru, and RuO₂,^16–22,38 which enhance the catalytic activity and cyclic performance of the cell. However, a portion of the reaction products could be isolated from the surface of the air electrode containing a solid catalyst during the charging process, as illustrated in Fig. 8a (bottom). This can increase the overpotential and limit the effective dissociation of the reaction products.

Fig. 8 Schematic diagrams of the reaction between the electrodes and reaction products: (a) a general electrode; (and b) the electrode employing polydopamine-coated CNTs and the dissolved LiI catalyst.

For the typical air electrode, then, the combination of the polydopamine-coated carbon and the dissolved LiI catalyst is expected to effectively suppress the problems and facilitate the reaction between lithium ions and oxygen. Fig. 8b presents the advantages of our new approach. The polydopamine coating can improve the wettability of the electrolyte on the air electrode, which can increase the active reaction area of the electrode, as shown in Fig. 8b (top). Moreover, the dissolved LiI catalyst in the electrolyte can freely access the reaction products and act as an electron–hole transfer agent between the electrode and the reaction products. This solves the contact problem between the reaction products and the air electrode, as shown at the bottom of Fig. 8b. The enhanced wettability of the air electrode due to the polydopamine layer facilitates the access of the dissolved catalyst in the electrolyte.

Fig. 9 shows the cyclic performance of the pristine and dopa electrodes. The discharge capacities of the electrodes were limited to 1000 mA h g_electrode⁻¹ to avoid a large depth-of-discharge.³⁹ The voltage range was 2.0–4.35 V and the current density was 400 mA g⁻¹. While the pristine electrode maintained its capacity for 61 cycles, the dopa electrode showed enhanced cyclic performance (73 cycles). Furthermore, when the LiI-containing electrolyte was used, the limited capacity (1000 mA h g_electrode⁻¹) of the dopa electrode was maintained for over 100 cycles. The effect of the dopa electrode and LiI-containing electrolyte was also observed the cycling test measured at higher current density of 1000 mA g⁻¹ (Fig. S1†), and using full depth of discharge (Fig. S2†). These results confirm that the combination of the polydopamine layer and dissolved LiI catalyst effectively improves the cyclic performance of the Li–air cell.

Fig. 9 Cyclic performance of the pristine and dopa electrodes at a current density of 400 mA g⁻¹, measured using plain and LiI-containing electrolytes.

The cyclic performance of Li–air batteries is highly dependent on undesired parasitic reactions such as the decomposition of the electrolyte and the formation of Li₂CO₃ due to the reaction of carbon and Li₂O₂.^25,26,40 The reaction products formed from parasitic reactions are not dissociated on charging and can easily accumulate on the surface of the air electrode. These accumulated reaction products prevent the access of the electrolyte (containing oxygen and Li ions) to the electrode, thus limiting cycle performance. Thus, the enhanced cyclic performance of the dopa electrode may imply the efficient dissociation of reaction products or the suppression of parasitic reactions during cycling. To analyze these possibilities, SEM images and FTIR spectra of the electrode were examined.

Fig. 10 shows the SEM images of the electrode before testing and after the initial discharge, initial charge, and 50^th charge processes. The cycling conditions were the same as those used in Fig. 9 (400 mA g⁻¹ current density and 1000 mA h g_electrode⁻¹ capacity limit). As shown in Fig. 10c–e, the surfaces of the electrodes after the first discharge are covered with reaction products. These seem to be dissociated after the initial charge, as presented in Fig. 10f–h. However, some of the CNT fibres seem to be covered with a film-like layer, although it is difficult to discern clearly. This suggests the possibility of imperfect dissociation of the reaction products. The morphologies of the electrodes after 50 cycles are noticeably different. In spite of its charged state, the pristine electrode after 50 cycles is fully covered with reaction products (Fig. 10i). The CNTs are nearly fully obscured by such products, so the fibrous shapes cannot be observed in the surface image. Thus, reaction products are constantly accumulated during cycling, which can explain the relatively poor cycle life of the pristine electrode as shown in Fig. 9. In Fig. 10j, the dopa electrode (without LiI) is also covered with reaction products after 50 cycles (charged state). However, several holes and fibre-like shapes can be observed on the surface of the electrode, which may allow the electrolyte surface access, albeit at limited levels. This result implies that the polydopamine layer can suppress the accumulation of the reaction products during cycling.

Fig. 10 SEM images of the (a) pristine electrode before testing; (b) dopa electrode before testing; (c) pristine electrode after initial discharge; (d) dopa electrode after initial discharge; (e) dopa electrode after initial discharge using LiI-containing electrolyte; (f) pristine electrode after initial charge; (g) dopa electrode after initial charge; (h) dopa electrode after initial charge using LiI-containing electrolyte; (i) pristine electrode after 50^th charge; (j) dopa electrode after 50^th charge; and (k) dopa electrode after 50^th charge using LiI-containing electrolyte.

The accumulated reaction products seem to be reduced considerably when the dopa electrode was cycled using the LiI-containing electrolyte. As shown in Fig. 10k, the electrode surface has many vacant spaces (or holes) and exhibits a fibre-like shape. This indicates that the dissolved LiI also plays an important role in suppressing the accumulation of reaction products, which can explain the enhanced cyclic performance of the dopa electrode in this electrolyte (Fig. 9).

To investigate the reaction products in detail, FTIR spectra were collected from the pristine and dopa electrodes after the initial discharge, initial charge, and 50^th charge processes. Fig. 11a shows the spectra for the pristine electrode. In the spectrum of the first discharged electrode, broad peaks between 400 and 600 cm⁻¹ show the formation of Li₂O₂. The broad peaks between 1400 and 1600 cm⁻¹ and the sharp peak at ∼870 cm⁻¹ may be attributed to the Li₂O₂ exposed to the air. The discharged electrode may contain a small amount of Li₂CO₃. However, it is difficult to confirm because the Li₂O₂ exposed to the air has an FTIR profile very similar to that of Li₂CO₃. The peaks at 400–500, 600–700, 1350–1500, and 1500–1700 cm⁻¹ (marked with ◆) may indicate the formation of organic compounds such as CH₃CO₂Li and HCO₂Li (which have similar FTIR spectra), due to parasitic reactions involving the electrolyte.^26,40 Some peaks due to organic components at 400–500 and 1350–1500 cm⁻¹ overlap with Li₂O₂ peaks and complicate the analysis of the residual materials.

Fig. 11 FTIR spectra of the electrodes collected after initial discharge, initial charge, and 50^th charge: (a) pristine electrode; (b) dopa electrode; (c) dopa electrode using LiI-containing electrolyte.

After the first charge process, the intensities of the peaks at 400–600 cm⁻¹ were significantly reduced due to the dissociation of Li₂O₂. However, it is clear that a small amount of Li₂O₂ or organic materials (detected at 400–500 cm⁻¹) still remained on the electrode. The peaks attributed to the organic materials at 1350–1500 and 1500–1700 cm⁻¹ were either not decreased or somewhat increased during the charge process, because they were not dissociated. Instead, they are easily formed during the charging process due to decomposition of the electrolyte at high voltage. When the pristine electrode was cycled 50 times, not only the Li₂O₂ peak (at 400–600 cm⁻¹) but also the organic peaks (at 600–700 and 1500–1700 cm⁻¹) were considerably increased. This result suggests that the accumulation of reaction products during cycling is mainly the result of electrolyte decomposition, as well as the incomplete dissociation of Li₂O₂ and Li₂CO₃.

Fig. 11b presents the FTIR spectra of the dopa electrode cycled in the absence of LiI. The spectrum of the first discharged dopa electrode is very similar to that of the first discharged pristine electrode. However, the first charged electrode reveals a significantly different spectrum. Fig. 12a compares the FTIR spectra of the first charged electrodes. Notably, the intensities of the peaks related to Li₂O₂ (at 400–600 and 1400–1600 cm⁻¹) are significantly reduced when compared to those of the first charged pristine electrode. Thus, the dopa electrode is more effective in the dissociation of Li₂O₂. Fig. 12b compares the FTIR spectra of the electrodes charged fifty times. The accumulation of Li₂O₂ or Li₂CO₃ on the dopa electrode during 50 cycles seemed similar to that of the pristine electrode. However, the low intensities of the organic peaks at 400–500 and 600–700 cm⁻¹ of the dopa electrode after fifty cycles imply that the polydopamine coating can suppress the formation of organic materials during cycling. However, the peaks at 1500–1700 cm⁻¹ were still large after 50 cycles.

Fig. 12 FTIR spectra of the electrodes collected (a) after 1^st charge and (b) after 50^th charge.

The FTIR spectra of the dopa electrode cycled using the LiI-containing electrolyte are shown in Fig. 11c. The spectrum of the first discharged electrode was nearly the same as the other first discharged electrodes (Fig. 11a and b). However, the first and 50^th charged electrodes produced significantly different spectra. In the spectrum of the first charged dopa electrode using the LiI-containing electrolyte, the intensities of the peaks at 400–600 and 1400–1600 cm⁻¹, which are attributed to Li₂O₂ (or Li₂CO₃) residues, are negligible. This result is noteworthy because a considerable amount of Li₂O₂ (or Li₂CO₃) was detected in the FTIR spectrum of the first charged pristine electrode (Fig. 11a and 12a). Moreover, the organic-based peaks at 400–500, 1350–1500, and 1500–1700 cm⁻¹ were significantly reduced compared to the spectra of the other electrodes. In the spectrum of the 50^th cycled dopa electrode using the LiI-containing electrolyte, all the organic peaks related to the parasitic reactions were present at low intensities. Only the peaks at 400–600 cm⁻¹ related to Li₂O₂, an ideal reaction product in a Li–air cell, were significantly present in the spectrum. This clearly shows that the accumulation of reaction products during cycling was significantly decreased. Specifically, the formation of organic materials was considerably reduced compared to the other electrodes (Fig. 12b), which confirms that the dopa electrode combined with LiI as a dissolved catalyst is very effective in suppressing parasitic reactions related to the decomposition of the electrolyte. In previous reports,^25,41 defects in the carbon facilitate unwanted electrolyte decomposition on cycling. The polydopamine layer on the surface of the CNTs may restrict direct contact between the carbon and the electrolyte, and subsequently suppress parasitic reactions. Furthermore, the dissolved LiI can reduce the charging voltage (lower the overpotential), which can prevent the active decomposition of the electrolyte at a high voltage range. The enhanced cyclic performance of the dopa electrode combined with LiI can be explained by the effects of the reduced accumulation of reaction products during cycling.

4. Conclusions

In this study, we successfully introduced polydopamine-coated CNTs and a dissolved catalyst as a new approach toward enhanced air electrodes for Li–air batteries. A thin polydopamine layer (several nanometers in thickness) coated on the surface of the CNTs effectively improved the wettability of the electrode. This was expected to increase the area of the electrode that actively participates in the reaction between the lithium ions and dissolved oxygen in the electrolyte. The dopa electrode showed a higher capacity and lower overpotential than the pristine electrode, which can be explained by the effect of an increased active reaction area. The polydopamine coating does not disturb the formation and dissociation of reaction products such as Li₂O₂. The addition of the dissolved LiI catalyst significantly decreases the overpotential of the Li–air cell. When the polydopamine-coated CNTs were tested using the LiI-containing electrolyte, the electrochemical performance of the electrode, such as its discharge capacity, overpotential, and cyclic performance, was remarkably enhanced by the synergic effect of the two changes. From the SEM and FTIR analyses, we confirmed that the polydopamine coating and the dissolved catalyst effectively dissociate the reaction products and suppress the parasitic reactions of the electrolyte, leading to a reduction in the accumulated reaction products and enhanced cyclic performance.

Acknowledgements

This work was supported by the National Research Foundation of Korea Grant funded by the Korean government (MEST) (NRF-2009-C1AAA001-0094219), and by an Energy Efficiency & Resources of the Korea Institute of Energy Technology Evaluation and Planning (20112010100110) grant funded by the Korea government Ministry of Knowledge Economy.

Notes and references

1. F. Li, H. Kitaura and H. Zhou, Energy Environ. Sci., 2013, 6, 2302.
2. J. Jung, K. Song, D. R. Bae, S. W. Lee, G. Lee and Y. M. Kang, Nanoscale, 2013, 5, 11845.
3. H. Wang, K. Xie, L. Wang and Y. Hana, RSC Adv., 2013, 3, 8236.
4. F. Li, T. Zhang and H. Zhou, Energy Environ. Sci., 2013, 6, 1125.
5. D. Zhu, L. Zhang, M. Song, X. Wang and Y. Chen, Chem. Commun., 2013, 49, 9573.
6. H. Kim, H.-D. Lim, J. Kim and K. Kang, J. Mater. Chem. A, 2014, 2, 33.
7. J. Gomez, E. E. Kalu, R. Nelson, M. H. Weatherspoon and J. P. Zheng, J. Mater. Chem. A, 2013, 1, 3287.
8. Y. Yu, B. Zhang, Y.-B. He, Z.-D. Huang, S.-W. Oh and J.-K. Kim, J. Mater. Chem. A, 2013, 1, 1163.
9. W.-M. Liu, T.-T. Gao, Y. Yang, Q. Suna and Z.-W. Fu, Phys. Chem. Chem. Phys., 2013, 15, 15806.
10. G. Zhao, Z. Xu and K. Sun, J. Mater. Chem. A, 2013, 1, 12862.
11. R. Padbury and X. Zhang, J. Power Sources, 2011, 196, 4436.
12. A. Kraytsberg and Y. Ein-Eli, J. Power Sources, 2011, 196, 886.
13. R. Black, B. Adams and L. F. Nazar, Adv. Energy Mater., 2012, 2, 801.
14. P. G. Bruce, S. A. Freunberger, L. J. Hardwick and J. M. Tarascon, Nat. Mater., 2012, 11, 19.
15. F. Li, T. Zhang and H. Zhou, Energy Environ. Sci., 2013, 6, 1125.
16. J. Christensen, P. Albertus, R. S. Sanchez-Carrera, T. Lohmann, B. Kozinsky, R. Liedtke, J. Ahmed and A. Kojic, J. Electrochem. Soc., 2012, 159, R1.
17. C. S. Park, K. S. Kim and Y. J. Park, J. Power Sources, 2013, 244, 72.
18. J.-S. Lee, S. T. Kim, R. Cao, N.-S. Choi, M. Liu, K. T. Lee and J. Cho, Adv. Energy Mater., 2011, 1, 34.
19. A. Zahoor, M. Christy, Y. J. Hwang and K. S. Nahm, J. Electrochem. Sci. Technol., 2012, 3, 14.
20. T. H. Yoon and Y. J. Park, J. Power Sources, 2013, 244, 344.
21. J. G. Zhang, D. Wang, W. Xu, J. Xiao and R. E. Williford, J. Power Sources, 2010, 195, 4332.
22. W.-H. Ryu, T. H. Yoon, S. H. Song, S. Jeon, Y. J. Park and I. D. Kim, Nano Lett., 2013, 9, 4190.
23. H.-D. Lim, K.-Y. Park, H. Song, E. Y. Jang, H. Gwon, J. Kim, Y. H. Kim, M. D. Lima, R. V. Robles, X. Lepro, R. H. Baughman and K. Kang, Adv. Mater., 2013, 9, 1348.
24. Y.-C. Lu and Y. Shao-Horn, J. Phys. Chem. Lett., 2013, 4, 93.
25. B. D. McCloskey, A. Speidel, R. Scheffler, D. C. Miller, V. Viswanathan, J. S. Hummelshøj, J. K. Nørskov and A. C. Luntz, J. Phys. Chem. Lett., 2012, 3, 997.
26. B. M. Gallant, R. R. Mitchell, D. G. Kwabi, J. Zhou, L. Zuin, C. V. Thompson and Y. Shao-Horn, J. Phys. Chem. C, 2012, 116, 20800.
27. M. M. O. Thotiyl, S. A. Freunberger, Z. Peng and P. G. Bruce, J. Am. Chem. Soc., 2013, 135, 494.
28. B. M. Gallant, D. G. Kwabi, R. R. Mitchell, J. Zhou, C. V. Thompson and Y. Shao-Horn, Energy Environ. Sci., 2013, 6, 2518.
29. B. D. Adams, C. Radtke, R. Black, M. L. Trudeau, K. Zaghib and L. F. Nazar, Energy Environ. Sci., 2013, 6, 1772.
30. H. Lee, B. P. Lee and P. B. Messersmith, Nature, 2007, 448, 338.
31. H. Lee, S. M. Dellatore, W. M. Miller and P. B. Messersmith, Science, 2007, 318, 426.
32. J. K. Ryu, S. H. Ku, H. S. Lee and C. B. Park, Adv. Funct. Mater., 2010, 20, 2132.
33. S. M. Kang, I. You, W. K. Cho, H. K. Shon, T. G. Lee, I. S. Choi, J. M. Karp and H. Lee, Angew. Chem., Int. Ed., 2010, 49, 9401.
34. V. Giordani, W. Walker, V. Bryantsev, J. Uddin, S. Zecevic, D. Addison and G. Chase, 16th IMLB Abstract, 2012, p. S6-3.
35. V. Chase, Z. Strahinja, W. Wesley, U. Jasim, S. Kenji, G. Vincent, B. Vyacheslav, B. Mario and A. Dan, US Pat., WO2011/133982 A1, 2011.
36. Y. Chen, S. A. Freunberger, Z. Peng, O. Fontainer and P. G. Bruce, Nat. Chem., 2013, 5, 489.
37. M.-H. Ryou, Y. M. Lee, J.-K. Park and J. W. Choi, Adv. Mater., 2011, 23, 3066.
38. C. S. Park, J. H. Kim and Y. J. Park, J. Electroceram., 2013, 31, 224.
39. A. Débart, A. J. Paterson, J. Bao and P. G. Bruce, Angew. Chem., 2008, 47, 4521.
40. M. O. Thotiyl, S. A. Freunberger, Z. Peng, Y. Chen, Z. Liu and P. G. Bruce, Nat. Mater., 2013, 12, 1050.
41. J. Lu, Y. Lei, K. C. Lau, X. Luo, P. Du, J. Wen, R. S. Assary, U. Das, D. J. Miller, J. W. Elam, H. M. Albishri, D. A. El-hady, Y. K. Sun, L. A. Curtiss and K. Amine, Nat. Commun., 2013, 4, 2383.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra01015a

---