Performance Enhancement and Side Reactions in Rechargeable

Dec 31, 2015 - The use of a highly conductive MWCNT network allows for high-capacity utilization because of rapid and efficient electron transport to ...
0 downloads 17 Views 8MB Size
Research Article www.acsami.org

Performance Enhancement and Side Reactions in Rechargeable Nickel−Iron Batteries with Nanostructured Electrodes Danni Lei,†,‡ Dong-Chan Lee,‡ Alexandre Magasinski,‡ Enbo Zhao,‡,§ Daniel Steingart,∥ and Gleb Yushin*,‡ †

School of Materials Science and Engineering, Hunan University, Changsha 410082, P. R. China School of Materials Science and Engineering and §School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, Georgia 30332, United States ∥ Andlinger Center for Energy and the Environment, Department of Mechanical and Aerospace Engineering, Princeton University, Princeton, New Jersey 08544, United States ‡

S Supporting Information *

ABSTRACT: We report for the first time a solution-based synthesis of strongly coupled nanoFe/multiwalled carbon nanotube (MWCNT) and nanoNiO/MWCNT nanocomposite materials for use as anodes and cathodes in rechargeable alkaline Ni−Fe batteries. The produced aqueous batteries demonstrate very high discharge capacities (800 mAh gFe−1 at 200 mA g−1 current density), which exceed that of commercial Ni−Fe cells by nearly 1 order of magnitude at comparable current densities. These cells also showed the lack of any “activation”, typical in commercial batteries, where low initial capacity slowly increases during the initial 20−50 cycles. The use of a highly conductive MWCNT network allows for high-capacity utilization because of rapid and efficient electron transport to active metal nanoparticles in oxidized [such as Fe(OH)2 or Fe3O4] states. The flexible nature of MWCNTs accommodates significant volume changes taking place during phase transformation accompanying reduction− oxidation reactions in metal electrodes. At the same time, we report and discuss that high surface areas of active nanoparticles lead to multiple side reactions. Dissolution of Fe anodes leads to reprecipitation of significantly larger anode particles. Dissolution of Ni cathodes leads to precipitation of Ni metal on the anode, thus blocking transport of OH− anions. The electrolyte molarity and composition have a significant impact on the capacity utilization and cycling stability. KEYWORDS: aqueous, alkaline, batteries, nanocomposite, carbon nanotube, iron, nickel

1. INTRODUCTION

importantly, greatly reduced capital and operating cost, inclusive of recycling.9−13 Although the nickel−iron (Ni−Fe) battery with NiO(OH) as a cathode and Fe as an anode was invented by Waldemar Jungner and Thomas Edison over 100 years ago,14,15 this chemistry was recently reconsidered for various applications because (1) both Fe and Ni are available elements in nature (Fe more so than Ni), (2) the Fe anode is not toxic (particularly when compared to lead and cadmium anodes), and (3) the recent emergence of hybrid nanoFe-based electrodes showed the possibility of enhancing rates and improving capacity utilization (and thus achieving higher

Efficient, robust, scalable, safe, and cheap power storage solutions are needed to accommodate the increasing demands of renewable energy harvesting (such as solar) for the off-grid operations and newly emerging and rapidly growing market of electric vehicles. Rechargeable batteries are particularly suitable for this large-scale storage of electrical energy because of their scalability and energy efficiency.1−3 Lithium-ion batteries (LIBs) play an important role in the storage of electrical energy because of their long cycling capability and high energy density.4−6 However, the flammability and cost of LIBs limit their market adoption rate in price-sensitive applications.7,8 Recently, “classic” rechargeable aqueous batteries have attracted renewed attention because of their environmental friendliness, intrinsic flame resistance of aqueous electrolytes, and, most © 2015 American Chemical Society

Received: November 2, 2015 Accepted: December 31, 2015 Published: December 31, 2015 2088

DOI: 10.1021/acsami.5b10547 ACS Appl. Mater. Interfaces 2016, 8, 2088−2096

Research Article

ACS Applied Materials & Interfaces

Figure 1. Synthesis processes of Fe/MWCNT and NiO/MWCNT. necked round-bottomed flask and refluxed at 180 °C in an oil bath for 1 h under magnetic stirring. Subsequently, the flask was taken out and allowed to cool to room temperature naturally. The mixtures were separated through filter paper (pore size = 2.5 μm; Whatman, USA) and washed with pure ethanol several times. The composites were dried in a vacuum oven overnight at 80 °C. Finally, the as-prepared powders were annealed in a reducing atmosphere (we used 4% H2 in argon) at 600 °C for 2 h. A NiO/MWCNT nanocomposite cathode material was also synthesized by a two-step process,27 similar to that of the anodes (Figure 1). A total of 200 mg of MWCNT was mixed with 100 mL of NMP and sonicated for 1 h. Then, 1.160 g of Ni(NO3)2·6H2O (99.999%; Sigma-Aldrich, USA) was poured into the above suspension and magnetically stirred for 1 h. The suspensions were transferred into a 150 mL three-necked round-bottomed flask and refluxed at 180 °C in an oil bath for 1 h with magnetic stirring. Subsequently, the flask was taken out and allowed to cool to room temperature naturally. The mixtures were separated through filter paper (pore size = 2.5 μm; Whatman, USA) and washed with absolute ethanol several times. The composites were dried in a vacuum at 80 °C overnight. Finally, the asprepared powders were annealed in an air atmosphere at 300 °C for 1 h and kept for further use. Figure 1 shows the schematic of the synthesis processes of both Fe/MWCNT and NiO/MWCNT nanostructured composites. 2.2. Material Characterization. Scanning electron microscopy (SEM) micrographs have been collected on a field-emission LEO 1530 microscope (Zeiss, Germany) at a beam voltage of 5 kV and a working distance of 3 mm for imaging and 8.5 mm for energy-dispersive spectroscopy (EDS) studies. High-resolution transmission electron microscopy (TEM) was performed on a TecnaiG2F30 microscope (FEI, The Netherlands) operating at 100 kV. Powder X-ray diffraction (XRD; X’Pert PRO Alpha-1, USA) using Cu Kα radiation was employed to identify the crystalline phase of the Fe/MWCNT and NiO/MWCNT composites. The isotherms of N2 gas adsorption on the surfaces of MWCNT, Fe/MWCNT, and NiO/MWCNT were collected at 77 K using an ASAP 2020 surface area and porosity measurement system (Micromeritics Inc., USA). Positive time-of-flight (ToF) secondary ion mass spectrometry (SIMS) spectra were obtained with a ToF.SIMS5 spectrometer (ION-TOF, Germany). Electrochemical impedance spectroscopy (EIS) tests were carried out with a Gamry Reference 600 potentiostat/galvanostat/ZRA (Gamry Instruments, Inc., USA) in the frequency range of 10 mHz to 106 Hz. 2.3. Electrochemical Characterization. The anode electrode paste film was prepared by mixing the active material, conducting agent, binder, and additive together in ethanol: 77 wt % Fe/MWCNT nanocomposites, 10 wt % purified exfoliated graphite (PEG; Superior Graphite, USA), 10 wt % poly(tetrafluoroethylene) (60 wt % suspension; Sigma-Aldrich, USA), and 3 wt % Bi2S3 (90%; SigmaAldrich, USA), respectively. A porous Ni foam (1.4 mm thickness; 590 μm pore size; Novamet, USA) was used as a current collector. The electrodes were dried at 80 °C in a vacuum overnight and compressed to ∼0.5 mm prior to use. The area of the electrodes is ∼1 cm2 with a

energy and potentially lower cost per unit energy) compared to electrodes produced using a century-old cell design.16−18 Although the Ni−Fe battery has many compelling advantages, the high self-discharge and poor charging efficiency, poor capacity utilization, and discharge rate capability are the principal drawbacks that limit large-scale adoption of Fe electrodes given the cost of current options.19−21 The low charging efficiency is due to the hydrogen evolution (HE) on the Fe electrode that competes for the current with the desired reaction between the anode and electrolyte and may prevent reversible battery charging. Previous studies on Fe electrodes have reported that bismuth additives increase the hydrogen overpotential, which noticeably reduces HE at comparable cell charge voltage.22−24 The formation of an insulating layer of iron hydroxide [Fe(OH)2] at high rates of discharge results in passivation of the electrode.22 Various researchers suggested that Fe−carbon (C) nanocomposites with improved electrode conductivity may mitigate such a passivation and improve the rate performance.18,24,25 Recently, Liu et al. proposed a hybrid iron active material based on the graphene foam/carbon nanotubes (CNTs) as the anode for a Ni−Fe battery and showed high specific energy/power characteristics with an impressive anode capacity of up to 270 mAh gFe−1 and a long cycle life.26 In this study, we report on an alternative synthesis route, where strongly coupled nanoFe/multiwalled carbon nanotube (MWCNT) and nanoNiO/MWCNT hybrid electrodes are produced via a solution-based deposition of the metal (metal oxide) precursor salts on the MWCNT surface from solutions. The direct attachment of Fe and NiO nanoparticles on the oxidized MWCNT surface provides robust contact between the respective nanoparticles and conductive CNTs, realizing efficient conduction of charge carriers and enhancing the structural stability. With the addition of LiOH into the electrolyte, these Ni−Fe batteries demonstrate a remarkably high discharge capacity of up to 800 mAh gFe−1. Equally importantly, our postmortem analysis revealed key mechanisms of the degradation of cells with nanostructured electrodes, which we discuss here for the first time.

2. EXPERIMENTAL SECTION 2.1. Synthesis. An Fe/MWCNT nanocomposite anode material was synthesized by a two-step process (Figure 1). In a typical procedure, 170 mg of MWCNT (>90%; Research Nanomaterials, Inc., USA) was mixed with 150 mL of N-methyl-2-pyrrolidone (NMP) and sonicated for 1 h. Then, 1.039 g of Fe(NO3)2·9H2O (99.99%; SigmaAldrich, USA) was poured into the above suspension and magnetically stirred for 1 h. The suspensions were transferred into a 300 mL three2089

DOI: 10.1021/acsami.5b10547 ACS Appl. Mater. Interfaces 2016, 8, 2088−2096

Research Article

ACS Applied Materials & Interfaces

Figure 2. Characterization of the Fe/MWCNT material: (a and b) SEM micrographs of the as-prepared samples before and after annealing; (c) TEM micrograph of Fe/MWCNT and HRTEM micrograph of an individual Fe nanoparticle; (d) XRD pattern of Fe/MWCNT.

Figure 3. Characterization of the NiO/MWCNT material: (a and b) SEM micrographs of the as-prepared samples before and after annealing; (c) TEM micrograph of NiO/MWCNT and HRTEM micrograph of an individual NiO nanoparticle; (d) XRD pattern of NiO/MWCNT. loading of ∼5 mg cm−2. The “commercial-like” anodes were made with Fe particles (200 nm, MTI) using the same methods. The NiO/ MWCNT electrode slurry was prepared by mixing the active material, conducting agent, and binder together in NMP: 80 wt % NiO/ MWCNT hybrid material, 10 wt % PureBlack conductive carbon additive (Superior Graphite, USA), and 10 wt % poly(vinylidene fluoride). The suspension was impregnated into a 2 cm × 2 cm Ni foam current collector (the loading is ∼10 mg cm−2), dried at 80 °C, and then annealed in argon at 200 °C for 30 min. The annealed electrodes were compressed to ∼0.5 mm before measurement. The “commercial-like” cathodes were made with Ni(OH)2 particles (Sigma-Aldrich, USA) as active materials. In the process of fabricating the beaker cells, the above Fe/MWCNT electrodes were used as anodes and NiO/MWCNT as cathodes. We used a porous polypropylene membrane Celgard 3401 (Celgard, USA) as a separator and KOH + LiOH aqueous solutions as electrolytes. The charge/ discharge tests were carried out using a multichannel charge/discharge battery testing system (BT2000; Arbin Instruments, USA). The

current densities of 150 mA g−1 for charge and 200 mA g−1 for discharge were selected. In all electrochemical measurements, fresh electrodes without any precycling were used. Notably, we used a voltage holding at 1.8 V for 7 min at the charge step, which can avoid generating excessive HE while allowing a nearly full reduction of Fe(OH)2 to Fe.

3. RESULTS AND DISCUSSION SEM studies of the produced Fe/MWCNT composites show 30−100 nm Fe particles directly connected with MWCNT and uniformly coating their surface (Figure 2). Evidently, a relatively uniform Fe(NO3)2·9H2O salt coating around the MWCNTs (Figure 2a) transform into individual Fe nanoparticles during annealing in a reducing atmosphere (Figure 2b). Fe constitutes ∼50 wt % of the Fe/MWCNT composite, as measured using thermogravimetric analysis (TGA) in air and additionally confirmed by EDS. TEM studies verified the strong 2090

DOI: 10.1021/acsami.5b10547 ACS Appl. Mater. Interfaces 2016, 8, 2088−2096

Research Article

ACS Applied Materials & Interfaces

Figure 4. N2 sorption isotherms and specific surface areas of (a) MWCNT, (b) Fe/MWCNT, (c) NiO/MWCNT, (d) commercial Fe, and (e) commercial Ni(OH)2.

Figure 5. Cycling stability of the Fe/MWCNT nanocomposite anode against the (a) commercial Ni(OH)2 and (b) NiO/MWCNT nanocomposite cathodes in comparison with benchmark materials. Typical discharge curves of the Ni−Fe batteries made with the Fe/MWCNT nanocomposite anode and (c) commercial Ni(OH)2 and (d) NiO/MWCNT nanocomposite cathodes.

coupling between Fe nanoparticles and MWCNTs (Figure 2c). Well-resolved lattice fringes of 0.21 nm observed in highresolution TEM (HRTEM) of Figure 2d correspond to the (110) plane of Fe (JCPDS no. 06-0696). XRD analysis further confirmed the formation of pure Fe without impurities (Figure 2d). All of the peaks are indexed to Fe except for the small peak at 26.4°, which results from the diffraction of MWCNTs. The limited resolution of SEM made it difficult to observe the smaller NiO nanoparticles on the surface of MWCNT (Figure 3b). However, TEM studies identified the shape and size of the NiO particles more clearly. Figure 3c clearly shows that ∼10 nm NiO particles are uniformly coating on the MWCNT surface. The well-defined lattice fringes with distances of 0.20 nm correspond to the d spacings of the (012) planes. As shown in Figure 3d, the XRD peaks of this

sample correspond to NiO (JCPDS no. 44-1159). The mass loading of NiO is around 50 wt % according to our TGA experiments in air. Figure 4 shows N2 sorption isotherms and specific surface areas of the original MWCNT samples as well as the Fe/ MWCNT (Figure 4a,b) and NiO/MWCNT (Figure 4c) nanocomposites. The deposition of Fe particles slightly reduced the specific surface area from 96 to 77 m2 g−1 because of a lower Fe surface-to-volume ratio and a higher Fe density. However, the formation of significantly smaller nanoparticles of NiO resulted in a higher surface area of 133 m2 g−1 due to the higher NiO surface-to-volume ratio. The isotherms and surface areas of the larger commercial Fe and Ni(OH)2 nanopowders used in some of the benchmark electrodes are shown for comparison (Figure 4d,e). 2091

DOI: 10.1021/acsami.5b10547 ACS Appl. Mater. Interfaces 2016, 8, 2088−2096

Research Article

ACS Applied Materials & Interfaces

Figure 6. Cycling stability (a) and typical discharge curves (b) of the Ni−Fe batteries made of Fe/MWCNT and NiO/MWCNT nanostructured electrodes cycled in three different electrolytes. discharge

Figure 5 shows the results of electrochemical characterization of Ni−Fe batteries with and without MWCNT-based electrodes. When anodes are tested against a commercial Ni(OH)2 powder-based cathode in full cells (Figure 5a), a significantly better performance of the Fe/MWCNT nanocomposite compared to the commercial Fe nanopowder mixed with MWCNT (with approximately the same carbon weight fractions) and the commercial Fe nanopowder-based anode is clearly seen (Figure 5a). While the initial capacity of the commercial Fe nanopowder mixed with MWCNT is comparable (only slightly smaller) to that of the Fe/ MWCNT nanocomposite (∼700 mAh gFe−1 vs ∼800 mAh gFe−1), it degrades to 50% of the initial capacity in less than 8 cycles and decreases to 100 mAh gFe−1 at 40th cycle. The performance of the commercial Fe nanopowder-based anode is even worse: the initial capacity of 350 mAh gFe−1 fades to just 50 mAh gFe−1 at 5th cycle. In contrast, the Fe/MWCNT nanocomposite exhibits a combination of very high capacity and significantly better stability than the other two samples, which we contribute to strong coupling between the Fe nanoparticles and MWCNT needed for rapid and efficient electron transport and good mechanical stability. We clarify that commercial Ni−Fe batteries (which utilize sintered Fe powder electrodes or Fe powders filled into porous metal tubes) commonly exhibit 1 order of magnitude smaller capacity (only ∼100 mAh gFe−1). However, they are stable for thousands of cycles. The high-capacity Fe/MWCNT nanocomposites still lose nearly 50% of the initial capacity after 100 cycles. The origin of such degradation will be discussed further. Figure 5b shows similar plots for the same anodes but tested against the NiO/MWCNT nanocomposite cathode. Qualitatively, the results and the observed trends are similar: full cells with the Fe/MWCNT nanocomposite anodes show a combination of better capacity and significantly better stability. After 100 cycles, the cell made of the Fe/MWCNT nanocomposite anode and the NiO/MWCNT nanocomposite cathode showed a discharge capacity of 425 mAh gFe−1, which is almost the same as that of the cell made of Fe/MWCNT and commercial Ni(OH)2 powder-based cathode (423 mAh gFe−1). As expected, both the kinetics and stability of the cells are mostly limited by the Fe anode performance because of an anode-limited cell design. The discharge curves corresponding to cells with these two types of cathodes (Figure 5c,d) show two main plateaus, which are attributed to the following reaction.28 Their corresponding charge curves are also shown in Figure S3.

Fe + 2OH− HooooooooI Fe(OH)2 + 2e− charge

(1)

discharge

3Fe(OH)2 + 2OH− HooooooooI Fe3O4 + 4H 2O + 2e− charge

(2)

We can see that the second (lower) voltage plateau of the cell with the commercial Ni(OH)2 powder-based cathode is higher than that of the cell with the NiO/MWCNT nanocomposite cathode. This suggests higher polarization in the cell with the NiO/MWCNT cathode, contrary to our initial expectations based on the higher electrical conductivity and smaller diffusion pathways in the nanostructured cathode. As we will later demonstrate, the higher surface area of nanostructured NiO may increase the dissolution of Ni. The dissolved Nicomprising ions (such as [Ni(OH)3]−) are reduced on the anode, thus depositing Ni in a metallic state. Because of a relatively low potential of the anode, Ni stays in this reduced metallic state, which has a very low permeability for OH− anions. As a result, nanostructured NiO/MWCNT-based cathodes increase the full cell polarization (compare Figure 5c showing discharge profiles for cells with regular cathodes with Figure 5d showing profiles for cells with nanostructured cathodes). The cyclic voltammograms in Figure 2Sa,b show that NiO/MWCNT exhibits slightly higher oxidation potential (0.50 V vs Hg/HgO) than commercial Ni(OH)2 (0.46 V vs Hg/HgO), which leads to higher polarization. This is related to the different form and size of these cathode materials. The corresponding EIS studies of full cells (Figure S2c,d and Table S1) will further support this observation. A typical electrolyte used in all commercial Fe−Ni cells is an aqueous solution of a 7.0 M KOH and 1.0 M LiOH mixture. However, this electrolyte has been optimized for conventional, microscale electrodes. Because we observed that the behavior of nanostructured anodes and cathodes is markedly different from that of conventional electrodes (nanostructured electrodes exhibit much higher surface area available for various side reactions), we were interested in further investigating the impact of the electrolyte molarity and composition on the capacity utilization and cell stability. Figure 6 first compares the nanostructured electrode-based cell performance in 7.0 M KOH + 1.0 M LiOH (425 mAh gFe−1 after 100 cycles) and 3.5 M KOH + 0.5 M LiOH (353 mAh gFe−1 after 100 cycles). Clearly, higher electrolyte molarity and correspondingly higher pH values result in better cell stability and overall performance. Such results might be somewhat expected because HE should be suppressed in more basic electrolytes according to Nernst’s law.29 This suppression not only prevents the formation of H2 2092

DOI: 10.1021/acsami.5b10547 ACS Appl. Mater. Interfaces 2016, 8, 2088−2096

Research Article

ACS Applied Materials & Interfaces

Figure 7. (a) Impact of LiOH on the concentration of Fe found in electrolytes after 10 cycles. (b) XRD of the Fe/MWCNT nanocomposite anodes at the discharged state after 10 cycles.

Figure 8. SEM micrographs of Fe/MWCNT nanocomposite anodes at discharged states after 10 cycles in (a) 7.0 M KOH + 1.0 M LiOH and (b) 8.0 M KOH electrolytes.

The formation of porous Fe(OH)2 and Fe3O4 may be related to the known incorporation of Li+ into the oxide lattice.30 Another research group similarly proposed that Li+ could be reduced within the oxide lattice of Fe3O4 to produce LixFeyOz intercalation-compound intermediates, which can be reduced to Fe more easily than Fe3O4, thereby improving the rate capability of the Fe anode.31 Other previous publications have also reported that the addition of Li+ changes kinetic parameters, such as exchange current densities and transfer coefficients.32−35 However, the origin of these phenomena and the mechanistic role of Li on both the observed improvements and the metal dissolution have not been thoroughly investigated. In order to gain a better understanding of the origins of the degradation of cells with nanostructured electrodes, we conducted postmortem analysis of the anodes (in the discharged state) after cycling. The anodes from both cells (cycled in the electrolyte with and without LiOH) showed the presence of Fe3O4 (Figure 7b). This results confirm that Li+ only affects the kinetics of Fe0 → Fe2+ and Fe2+ → Fe3+ oxidation but does not bring a qualitative change in the phase composition. Except the peaks of Fe3O4, there are peaks corresponding to Ni (JCPDS no. 44-1159) and C (JCPDS no. 41-1487), which come from Ni foam and PEG, respectively. Postmortem SEM (Figure 8) and TEM (Figure 9) studies revealed a striking impact of cycling on the morphology of Fe anodes. Electron microscopy techniques clearly demonstrated the formation of large particles of Fe3O4 on the anode surface (Figures 8 and 9). Interestingly though, in the case of LiOH present in the electrolyte, a large portion of the reprecipitated Fe3O4 was in the form of nanoporous, nanocrystalline powder (Figure 9a). However, in the case of the pure KOH electrolyte, only large crystalline Fe3O4 particles with clear crystalline facets were produced (grown during cycling). It is difficult to quantify this phenomenon numerically, but qualitatively the degree of

bubbles (which reduce the electrode−electrolyte contact area, leading to lower capacity utilization) but also assists in a more complete charging of the cell (because HE consumes part of the current during the charging process). The anodes cycled in a lower concentration (and thus lower pH) electrolyte slowly accumulate some amounts of Fe(OH)2 and Fe3O4 (not fully reduced to Fe) that are still present at the end of charging and thus become unavailable for the subsequent cycling. Another interesting observation is the impact of LiOH. By comparing the performance of cells in 8.0 M KOH and 7.0 M KOH + 1.0 M LiOH, we observe significantly higher discharge capacities in the LiOH-comprising cells (Figure 6a). However, cells with pure KOH showed better stability. From the corresponding discharge curves (Figure 6b), we see that the systems without LiOH both reduce the capacity of the first anodic oxidation step (eq 1), as well as the potential for the second anodic oxidation step (eq 2). This suggests significantly higher resistance. We propose that LiOH induces the formation of a more porous Fe(OH)2 layer and thus enables faster OH− anion diffusion across the electrolyte/solid interface and correspondingly lower charge-transfer resistance, reduced cell polarization, and accelerated kinetics of Fe2+ → Fe3+ oxidation (eq 2) and higher capacity utilization. Unfortunately, we found that LiOH additionally enhances the dissolution of Fe and its reprecipitation in the oxidized state. In order to demonstrate this hypothesis on the anode dissolution, we collected and dried cycled electrolytes and conducted chemical analysis studies on thus-produced samples. Even just after 10 cycles, as much as 37 mg of Fe was already dissolved in the 7.0 M KOH + 1.0 M LiOH electrolyte (Figure 7a). In cells tested in pure KOH, Fe dissolution also took place, but the degree of dissolution was 3−4 times lower (presumably because of the higher density and thus lower effective surface area of Fe(OH)2). 2093

DOI: 10.1021/acsami.5b10547 ACS Appl. Mater. Interfaces 2016, 8, 2088−2096

Research Article

ACS Applied Materials & Interfaces

our cell configuration. Instead, we assembled new cells with commercial Fe nanopowders and used a Ti current collector. We used ToF SIMS for these measurements. While the electrode based on commercial Fe powders still contained some amount of Ni impurities, we could still detect a noticeably increasing concentration of Ni after only 10 cycles (Table 1). Table 1. Results of ToF SIMS Analysis Showing Differences in the Amount of Ni Present within Fe Anodes at Discharged States after 10 Cycles in 7.0 M KOH + 1.0 M LiOH and 8.0 M KOH Electrolytesa anode

cathode

electrolyte

Fe powder on Ti mesh NiO/ MWCNT

7 M KOH +1M LiOH 8 M KOH

Fe+ peak area

Ni+ peak area

Ni+/Fe+ ratio

655649

3776

5.8 × 10−3

156011

1740

11.2 × 10−3

509703

4197

8.2 × 10−3

The collection area in all tests was 150 μm × 150 μm. The depth of the profile was 100 nm for all of the electrodes. a

Figure 9. TEM micrographs of Fe/MWCNT nanocomposite anodes at discharged states after 10 cycles in (a and b) 7.0 M KOH + 1.0 M LiOH and (c−e) 8.0 M KOH electrolytes.

Furthermore, a higher concentration of Ni was detected in LiOH-comprising electrolyte. Higher degree of Ni dissolution and precipitation on the anode with LiOH addition should reduce the kinetics of LiOH-comprising cells, in contrast to our electrochemical observations (Figure 6). We propose that the difference in the morphological changes in the anode induced by LiOH (such as the formation of porous Fe3O4) had a stronger positive impact on the kinetics than the negative impact induced by the precipitation of Ni. Still, faster Ni precipitation should enhance the degradation rate, as we indeed observed in our tests (Figure 6a). We found a recent publication that comprehensively reviews the properties of different phases of Ni(OH)2 and discusses the possibility of a very slow dissolution of α-Ni(OH)2 in concentrated alkaline media under certain conditions.36 However, the impact of Ni(OH)2 cathode dissolution on the stability of an Fe anode and the impact and role of LiOH on the degree of the dissolution have never been covered in the literature. EIS analyses provided complementary insight into the influence of Li+ (Figure 10) and cathode materials (Figure S2b−d) on anodic processes. The experimental Nyquist and Bode plots of full cells were fitted using the equivalent circuit with a constant-phase-angle element (CPE). Tables 2 and S1 show the calculated values for the ohmic resistance (Re) and total charge-transfer resistance (Rct) of full cells. Rct depends on the conductivity of the interface between the active material and electrolyte, on the electrode surface area, and on the porous morphology of the electrode.37,38 The cell cycled in the LiOH-containing electrolyte exhibits significantly lower Rct (7.6 Ω) than the cells built with pure KOH (21.5 Ω), which suggests faster reaction kinetics in LiOH-comprising cells and correspondingly smaller polarization and higher voltage plateaus, as previously observed and discussed (Figure 6). The higher Rct (7.6 Ω) for the cell based on the nanostructured NiO/MWCNT than that of the commercial Ni(OH)2 (4.2 Ω) is consistent with our previous observation about the difference of voltage plateaus (Figure 5c,d). The fitted Bode plots of all of the full cells (Figures 10b and S2d) matched with our experimental data very well, which proves the validity of our fitting in all frequency regions.

Fe3O4 reprecipitation was more pronounced in the LiOHcomprising electrolyte according to our SEM and TEM results, supporting the previously discussed results (Figure 7a). The formation of larger crystalline Fe3O4 in KOH is certainly undesirable because it limits the rate performance and capacity utilization of these electrodes (and somewhat defeats the purpose of producing nanostructured Fe−C composite anodes). However, once formed, these larger particles shall exhibit slower further dissolution. At the same time, comparing the morphologies of cycled Fe/MWCNT (Figure 8a), commercial Fe (Figure S1a), and Fe/MWCNT (Figure S1b), we found that while the commercial samples did not form such large particles, the active particles became significantly agglomerated and Fe 3 O 4 lost electrical contact with MWCNT, leading to degradation. These results are consistent with the previous discussion about the discharge capacity. According to the potential-pH equilibrium (Pourbaix) diagram for the Fe−H2O system at 25 °C, Fe-comprising soluble species mostly form at low pH values.28 However, at pH above 14.2, HFeO2 − ions may form in the potential range from about −0.7 to −1 V vs SHE, leading to Fe anode dissolution (corrosion) and reprecipitation. To our surprise, this phenomenon was never discussed in any of the recent papers on alkaline batteries with nanostructured Fe anodes. The high surface area of a nanostructured Fe anode, unfortunately, greatly enhances the dissolution kinetics. The development of new electrolytes or favorable surface termination technologies (or OH−-permeable protective surface coatings) might be necessary in order to prevent the dissolution and degradation of nanostructured electrodes. We also hypothesized that some Ni dissolution may also take place during cycling and that the presence of LiOH in the electrolyte may impact Ni cathode dissolution, similarly to its apparent impact on the dissolution of Fe anodes. However, because of the presence of Ni catalysts in MWCNT and the use of a Ni current collector, this hypothesis was difficult to test in 2094

DOI: 10.1021/acsami.5b10547 ACS Appl. Mater. Interfaces 2016, 8, 2088−2096

Research Article

ACS Applied Materials & Interfaces

Figure 10. (a) Nyquist and (b) Bode plots of Ni−Fe batteries made with Fe/MWCNT and NiO/MWCNT nanostructured electrodes at discharged states after 10 cycles in different electrolytes. Solid lines in the Nyquist and Bode plots were fitted to the equivalent circuit with CPEs.

10 cycles, CV of commercial Ni(OH)2 and NiO/ MWCNT, typical discharge and charge curves of Fe/ MWCNT//Ni(OH) 2 and Fe/MWCNT//NiO/ MWCNT, impedance parameters derived using the equivalent circuit model of Fe/MWCNT//Ni(OH)2 and Fe/MWCNT//NiO/MWCNT at discharged states after 10 cycles (PDF)

Table 2. Impedance Parameters Derived Using an Equivalent Circuit Model for Ni−Fe Batteries Made with Fe/MWCNT and NiO/MWCNT Nanostructured Electrodes at Discharged States after 10 Cycles in Different Electrolytes electrolyte composition

Re (Ω)

Rct (Ω)

7 M KOH + 1 M LiOH 8 M KOH

0.36 0.64

7.6 21.5



4. CONCLUSIONS We have successfully utilized a solution-based method to deposit and anchor Fe and NiO nanoparticles onto the surface of MWCNT for use in Ni−Fe batteries. On the positive side, the produced composites exhibited high uniformity and their applications in rechargeable alkaline batteries resulted in remarkably high capacity utilization and moderately good stability. On the negative side, the high specific surface areas of nanostructured materials in both the Ni cathode and Fe anode were found to enhance multiple side reactions, which were discovered to deteriorate the cell performance. On the basis of our postmortem analyses (EIS, TEM, SEM, EDS, XRD, and ToF SIMS), we came to the following conclusions: (i) at high alkaline electrolyte concentration and thus high pH values, Fe dissolution and reprecipitation takes place, which reduces the rate performance and capacity utilization of the nanostructured Fe anodes; (ii) at lower pH values, Fe dissolution could be mitigated, but HE takes place, which becomes particularly significant if high-surface-area nanostructured Fe anodes are used; (iii) the addition of LiOH to KOH electrolyte enhances Fe dissolution but reduces anode polarization and increases capacity utilization; these findings correlate well with the formation of porous oxidized Fe in LiOH-comprising electrolytes; (iv) the dissolution of Ni cathodes (particularly significant in nanoparticle-based electrodes with higher surface area due to their small dimensions and higher surface energy due to the higher curvature) leads to the deposition of Ni metal onto the anode, which slows the cell rate performance characteristics by blocking OH− anions. This new knowledge should assist in the further advancement of rechargeable Ni−Fe battery technology.



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work was partially supported by the Advanced Research Projects Agency-Energy (Grant DE-AR0000400). A fellowship to D.L. was provided by the China Scholarship Council (Grant 201306130006).



REFERENCES

(1) Armand, M.; Tarascon, J. M. Building Better Batteries. Nature 2008, 451, 652−657. (2) Yang, Z.; Zhang, J.; Kintner-Meyer, M. C.; Lu, X.; Choi, D.; Lemmon, J. P.; Liu, J. Electrochemical Energy Storage for Green Grid. Chem. Rev. 2011, 111, 3577−3613. (3) Narayanan, S. R.; Prakash, G. K. S.; Manohar, A.; Yang, B.; Malkhandi, S.; Kindler, A. Materials Challenges and Technical Approaches for Realizing Inexpensive and Robust Iron−Air Batteries for Large-scale Energy Storage. Solid State Ionics 2012, 216, 105−109. (4) Tarascon, J. M.; Armand, M. Issues and Challenges Facing Rechargeable Lithium Batteries. Nature 2001, 414, 359−367. (5) Lee, Y. J.; Yi, H.; Kim, W. J.; Kang, K.; Yun, D. S.; Strano, M. S.; Ceder, G.; Belcher, A. M. Fabricating Genetically Engineered HighPower Lithium-Ion Batteries Using Multiple Virus Genes. Science 2009, 324, 1051−1055. (6) Choi, N. S.; Chen, Z.; Freunberger, S. A.; Ji, X.; Sun, Y. K.; Amine, K.; Yushin, G.; Nazar, L. F.; Cho, J.; Bruce, P. G. Challenges Facing Lithium Batteries and Electrical Double-Layer Capacitors. Angew. Chem., Int. Ed. 2012, 51, 9994−10024. (7) Goodenough, J. B.; Kim, Y. Challenges for Rechargeable Li Batteries. Chem. Mater. 2010, 22, 587−603. (8) Etacheri, V.; Marom, R.; Elazari, R.; Salitra, G.; Aurbach, D. Challenges in the Development of Advanced Li-Ion Batteries: a review. Energy Environ. Sci. 2011, 4, 3243−3262. (9) Gao, X. P.; Yao, S. M.; Yan, T. Y.; Zhou, Z. Alkaline Rechargeable Ni/Co Batteries: Cobalt Hydroxides as Negative Electrode Materials. Energy Environ. Sci. 2009, 2, 502−505. (10) Luo, J. Y.; Cui, W. J.; He, P.; Xia, Y. Y. Raising the Cycling Stability of Aqueous Lithium-Ion Batteries by Eliminating Oxygen in the Electrolyte. Nat. Chem. 2010, 2, 760−765.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsami.5b10547. SEM micrographs of commercial Fe and commercial Fe/ MWCNT anodes at discharged states after 10 cycles, Nyquist and Bode plots of Fe/MWCNT//Ni(OH)2 and Fe/MWCNT//NiO/MWCNT at discharged states after 2095

DOI: 10.1021/acsami.5b10547 ACS Appl. Mater. Interfaces 2016, 8, 2088−2096

Research Article

ACS Applied Materials & Interfaces

(32) Juanto, S.; Schrebler, R. S.; Zerbino, J. O.; Vilche, J. R.; Arvia, A. J. The Influence of Different Cations on the Eletrochemical and Ellipsometric Behaviour of Iron in Alkaline Media. Electrochim. Acta 1991, 36, 1143−1150. (33) Geronov, Y.; Tomov, T.; Georgiev, S. Mössbauer Spectroscopy Investigation of the Iron Electrode During Cycling in Alkaline Solution. J. Appl. Electrochem. 1975, 5, 351−358. (34) Hampson, N. A.; Latham, R. J.; Marshall, A.; Giles, R. D. Some Aspects of the Electrochemical Behaviour of the Iron Electrode in Alkaline Solutions. Electrochim. Acta 1974, 19, 397−401. (35) Hills, S. Beneficial Effect of Lithiated Electrolyte on Iron Battery Electrodes. J. Electrochem. Soc. 1965, 112, 1048−1049. (36) Hall, D. S.; Lockwood, D. J.; Bock, C.; MacDougall, B. R. Nickel Hydroxides and Related Materials: A Review of Their Structures, Synthesis and Properties. Proc. R. Soc. London, Ser. A 2015, 471, 20140792. (37) Li, X.; Rong, J.; Wei, B. Electrochemical Behavior of SingleWalled Carbon Nanotube Supercapacitors under Compressive Stress. ACS Nano 2010, 4, 6039−6049. (38) Qi, J. L.; Wang, X.; Lin, J. H.; Zhang, F.; Feng, J. C.; Fei, W. D. Vertically Oriented Few-Layer Graphene-Nanocup Hybrid Structured Electrodes for High-Performance Supercapacitors. J. Mater. Chem. A 2015, 3, 12396−12403.

(11) Liu, S.; Pan, G. L.; Yan, N. F.; Gao, X. P. Aqueous TiO2/ Ni(OH)2 Rechargeable Battery with a High Voltage Based on Proton and Lithium Insertion/Extraction Reactions. Energy Environ. Sci. 2010, 3, 1732−1735. (12) Xu, C.; Li, B.; Du, H.; Kang, F. Energetic Zinc Ion Chemistry: the Rechargeable Zinc Ion Battery. Angew. Chem., Int. Ed. 2012, 51, 933−935. (13) Liu, Z.; Tay, S. W.; Li, X. Rechargeable Battery Using a Novel Iron Oxide Nanorods Anode and a Nickel Hydroxide Cathode in an Aqueous Electrolyte. Chem. Commun. 2011, 47, 12473−12475. (14) Chakkaravarthy, C.; Periasamy, P.; Jegannathan, S.; Vasu, K. I. The Nickel/Iron Battery. J. Power Sources 1991, 35, 21−35. (15) Shukla, A. K.; Ravikumar, M. K.; Balasubramanian, T. S. Nickel/ Iron Batteries. J. Power Sources 1994, 51, 29−36. (16) Periasamy, P.; Ramesh Babu, B.; Venkatakrishna Iyer, S. Performance Characterization of Sintered Iron Electrodes in Nickel/ Iron Alkaline Batteries. J. Power Sources 1996, 62, 9−14. (17) Halpert, G. Past Developments and the Future of Nickel Electrode Cell Technology. J. Power Sources 1984, 12, 177−192. (18) Jiang, W.; Liang, F.; Wang, J.; Su, L.; Wu, Y.; Wang, L. Enhanced Electrochemical Performances of FeOx-Graphene Nanocomposites as Anode Materials for Alkaline Nickel-Iron Batteries. RSC Adv. 2014, 4, 15394−15399. (19) Souza, C. A. C.; Carlos, I. A.; Lopes, M.; Finazzi, G. A.; de Almeida, M. R. H. Self-Discharge of Fe−Ni Alkaline Batteries. J. Power Sources 2004, 132, 288−290. (20) Manohar, A. K.; Malkhandi, S.; Yang, B.; Yang, C.; Surya Prakash, G. K.; Narayanan, S. R. A High-Performance Rechargeable Iron Electrode for Large-Scale Battery-Based Energy Storage. J. Electrochem. Soc. 2012, 159, A1209−A1214. (21) Malkhandi, S.; Yang, B.; Manohar, A. K.; Prakash, G. K. S.; Narayanan, S. R. Self-Assembled Monolayers of n-Alkanethiols Suppress Hydrogen Evolution and Increase the Efficiency of Rechargeable Iron Battery Electrodes. J. Am. Chem. Soc. 2013, 135, 347−353. (22) Manohar, A. K.; Yang, C.; Malkhandi, S.; Prakash, G. K. S.; Narayanan, S. R. Enhancing the Performance of the Rechargeable Iron Electrode in Alkaline Batteries with Bismuth Oxide and Iron Sulfide Additives. J. Electrochem. Soc. 2013, 160, A2078−A2084. (23) Balasubramanian, T. S.; Shukla, A. K. Effect of Metal-Sulfide Additives on Charge/Discharge Reactions of the Alkaline Iron Electrode. J. Power Sources 1993, 41, 99−105. (24) Rajan, A. S.; Sampath, S.; Shukla, A. K. An in Situ CarbonGrafted Alkaline Iron Electrode for Iron-Based Accumulators. Energy Environ. Sci. 2014, 7, 1110−1116. (25) Wang, H.; Liang, Y.; Gong, M.; Li, Y.; Chang, W.; Mefford, T.; Zhou, J.; Wang, J.; Regier, T.; Wei, F.; Dai, H. An Ultrafast Nickel− Iron Battery from Strongly Coupled Inorganic Nanoparticle/Nanocarbon Hybrid Materials. Nat. Commun. 2012, 3, 917. (26) Liu, J.; Chen, M.; Zhang, L.; Jiang, J.; Yan, J.; Huang, Y.; Lin, J.; Fan, H. J.; Shen, Z. X. A Flexible Alkaline Rechargeable Ni/Fe Battery Based on Graphene Foam/Carbon Nanotubes Hybrid Film. Nano Lett. 2014, 14, 7180−7187. (27) Xu, C.; Sun, J.; Gao, L. Large Scale Synthesis of Nickel Oxide/ Multiwalled Carbon Nanotube Composites by Direct Thermal Decomposition and Their Lithium Storage Properties. J. Power Sources 2011, 196, 5138−5142. (28) Vijayamohanan, K.; Balasubramanian, T. S.; Shukla, A. K. Rechargeable Alkaline Iron Electrodes. J. Power Sources 1991, 34, 269− 285. (29) Kozin, L. F.; Mashkova, N. V.; Manilevich, F. D. Hydrogen Overvoltage at Manganese in Ammonium-Bromide-Perchloric Solution. Prot. Met. Phys. Chem. Surf. 2009, 45, 25−30. (30) Kalaignan, G. P.; Muralidharan, V. S.; Vasu, K. I. Triangular Potential Sweep Voltammetric Study of Porous Iron Electrodes in Alkali Solutions. J. Appl. Electrochem. 1987, 17, 1083−1092. (31) Casellato, U.; Comisso, N.; Mengoli, G. Effect of Li Ions on Reduction of Fe Oxides in Aqueous Alkaline Medium. Electrochim. Acta 2006, 51, 5669−5681. 2096

DOI: 10.1021/acsami.5b10547 ACS Appl. Mater. Interfaces 2016, 8, 2088−2096