Performance of Thixotropic Gel Electrolytes in the Rechargeable

Dec 14, 2016 - Lambert , D. W. H.; Greenwood , P. H. J.; Reed , M. C. Advances in Gelled-electrolyte Technology for Valve-regulated Lead-acid Batterie...
0 downloads 0 Views 6MB Size
Research Article pubs.acs.org/journal/ascecg

Performance of Thixotropic Gel Electrolytes in the Rechargeable Aqueous Zn/LiMn2O4 Battery Tuan K. A. Hoang,†,§ The Nam Long Doan,†,§ Changyu Lu,†,§ Mahmoudreza Ghaznavi,† Hongbin Zhao,†,‡ and P. Chen*,† †

Downloaded via NEW MEXICO STATE UNIV on July 1, 2018 at 13:52:48 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

Department of Chemical Engineering and Waterloo Institute for Nanotechnology, University of Waterloo, 200 University Avenue West, Waterloo, Ontario N2L3G1, Canada ‡ Department of Chemistry, College of Science, Shanghai University, Shanghai 200444, P. R. China ABSTRACT: Novel aqueous gel electrolytes have been prepared using a combination of thixotropic and nonthixotropic gelling agents. These gels are used in rechargeable hybrid aqueous batteries for the first time. Thermogravimetric analysis proves that the retentions of the gel electrolytes in the absorbed glass mat (AGM) separator are significantly higher than that of the conventional liquid electrolyte. Batteries with the gel electrolytes exhibit up to 10% improvement in specific reversible capacity and higher rate capability, and the capacity retention is up to 8% higher than that of the reference battery after 1000 cycles at 4 C. Furthermore, the electrodes of the gel batteries are well-preserved, and dendrite formation on the Zn anode is suppressed as confirmed by scanning electron microscopy and chronoamperometry techniques. These significant improvements are due to the new gel electrolytes and the synergistic effect of a thixotropic and a nonthixotropic gelling agent. KEYWORDS: Aqueous battery, Gel electrolyte, Thixotropy, Dendrite, Thermal tolerance



INTRODUCTION The use of gel electrolytes has been implemented in lithium-ion batteries.1−10 Early types of aqueous rechargeable lithium batteries (ARLBs) have been reported since 1994; however, there are not many ARLBs employing gel or solid state electrolytes reported.11,12 This is unusual since gel electrolytes have been used commercially in other rechargeable aqueous batteries. For example, commercial lead-acid batteries use silicacontaining aqueous gel electrolytes,13 and there are aqueous gel zinc-air batteries reported.14 The move from conventional aqueous electrolytes to gel electrolytes is logical because the gel can prevent water evaporation thanks to hydrogen bonding between the gelling agent and water. As a result, many problems, such as increases in electrolyte concentrations and salt precipitation, may be alleviated with the use of gel electrolytes.13 The rechargeable hybrid aqueous battery (ReHAB) contains a lithium intercalation compound in the cathode, zinc as the anode, and an aqueous solution containing Li+ and Zn2+ as the electrolyte.15,16 The ReHAB was introduced in 2012, and it can store up to 80 W h kg−1 energy density. This value is more than double that of a lead-acid battery (ca. 30−40 W h kg−1).13 More importantly, the ReHAB is constructed mostly from less expensive and environmentally benign materials. This would provide great cost and safety advantages over lead-acid batteries and the lithium-ion analogues. Current research activities in our group and others aim to develop the ReHAB system into commercial products and compete against lead-acid batteries, which constitute about 52% of the current battery market.17 In © 2016 American Chemical Society

order to improve the performance and service life, a simple silica-containing liquid electrolyte has been implemented in the absorbed glass mat (AGM) separator.18 The introduction of nanosilica in the aqueous electrolyte is considered a major success since several physical properties of the ReHAB have been improved, including cyclability, electrode protection against the corrosive electrolyte, and significantly lower selfdischarge.18 However, use of gel electrolyte prepared with nanosilica is not practical because of the high silica loading (up to 10%), and the short gelling time (order of seconds). The short gelling time creates a technical problem when introducing the electrolyte into the AGM separator because the size of the separator in large-scale batteries is usually 2 orders of magnitude higher than its lab-scale analogue (about 100 cm2 vs 1 cm2). In addition, high silica loading increases the electrolyte cost and reduces the ion concentration, which are not good for scaling up processes. In this work, we propose the use of a hybrid gelling system, namely, the mixture of β-cyclodextrin (CD, a 7-membered sugar ring molecule) and fumed silica (FS). β-Cyclodextrin has multiple OH functional groups, which allows the formation of innate hydrogen bonding with water molecules. Through the optimization of the concentration ratio between the thixotropic (FS) and the nonthixotropic gelling agent (CD), new gel electrolytes are obtained, and better battery performance is Received: October 22, 2016 Revised: November 28, 2016 Published: December 14, 2016 1804

DOI: 10.1021/acssuschemeng.6b02553 ACS Sustainable Chem. Eng. 2017, 5, 1804−1811

Research Article

ACS Sustainable Chemistry & Engineering

Chronoamperometry was implemented to shed light on the zinc deposition onto the anode when an overpotential was applied. The test was conducted in a three-electrode system consisting of a working electrode of ca. 2.5 cm2 of polished zinc, a nickel foam counter electrode, and a saturated calomel reference electrode. The zinc electrode size was redetermined again after finishing the experiment, and it was used to determine the current density (mA·cm−1). The electrodes are connected to a multichannel potentiostat (VMP3, Biologic). Then, an overpotential of −120 mV versus Zn2+/Zn was maintained for 1 h. The generated currents were recorded and plotted as mA·cm−1.

observed. Specific reversible capacity and cyclability of the cells containing such a gel electrolyte are both improved as a result of the suppression of dendrite formation on the metal anode, which is supported by the surprisingly low absolute value of Zn2+ ion deposition current in chronoamperometry studies. In addition, the gelling time and gel strength can be controlled upon adjusting the concentrations of gelling agents. Longer gelling time is needed to allow gel penetration into the AGM separator of large commercial cells.





EXPERIMENTAL SECTION

RESULTS AND DISCUSSION Figure 1 shows the nitrogen adsorption and desorption isotherms of fumed silica, which evaluates its physical character.

Preparation of Batteries. Cathode and anode preparations were following previously reported procedures.18 The gel electrolytes were prepared from gelling additives and liquid aqueous electrolyte. The liquid aqueous electrolyte was prepared from Li2SO4·H2O (SigmaAldrich, 98%), ZnSO4·7H2O (Sigma-Aldrich, 98%), and deionized water so that the aqueous electrolyte contains 2 M Li2SO4 and 1 M ZnSO4; pH was adjusted to 4.00 ± 0.05. Batteries prepared from this electrolyte were named “00”. By mixing conventional liquid electrolyte with as-received fumed silica (Sigma-Aldrich) and β-cyclodextrin (Sigma-Aldrich, 98%), gel electrolytes containing 5% fumed silica (FS), 4% FS + 1% β-cyclodextrin (CD), and 3% FS + 2% CD were prepared. Batteries prepared from such electrolytes will be named “50”, “41”, and “32”, respectively. Instrumentation. Fresh cathode and anode and post-battery-run cathodes and anodes are examined by a powder X-ray diffraction (XRD, D8 Discover, Brüker) instrument equipped with a Cu Kα radiation source, with a scan speed of 1° per minute, ranging from 10° to 90°. They are also characterized by field emission scanning electron microscopy (FE-SEM, Leo-1550, Zeiss) using extra high tension (EHT) of 8 kV. Fumed silica was characterized by XRD and the accelerated surface area and porosimetry analyzer (ASAP2020, Micromeritics) using a 20-point nitrogen adsorption and desorption program. Functional groups of CD and FS were analyzed by Fourier transform infrared spectroscopy (FT-IR, Brüker). For thermogravimetric analysis, the AGM separator was cut into small pieces, and each piece has the average weight of ca. 20 mg. The pristine electrolyte (00) and gel electrolytes (50, 41, 32) were added into the AGM before TGA studies (SDT-Q600, TA Instruments). Temperature was ramped from room condition up to 120 °C with a ramp rate of 1 °C min−1. Gel strength is studied by the ball penetration experiment. In a typical run, 16 g of electrolyte was stored in a 20 mL transparent glass vial. For every hour, a 5 mm-in-diameter stainless steel ball was dropped into the electrolyte from 2 cm above the surface. Each experiment was conducted at least 3 times. The penetration depth was measured as the distance from the gel surface to the position where the ball stops. The maximum measured penetration depth in this study is 2.6 cm, meaning that the ball has completely penetrated the gel volume and reached the bottom of the vial.13,19,20 Electrochemical Measurements. The electrochemical tests were implemented on Swagelok-type cells. Each cell was composed of a LiMn2O4/KS-6/PFDF composite positive electrode and a polished zinc metal negative electrode, separated by an AGM separator. The loading of cathode is 4−6 mg of LiMn2O4 per cm2. This loading allows sufficiently high flows of the working ions within the batteries so that the differences in post-run anodes and cathodes are confidently distinguished. Gel electrolyte was injected into the separator (ca. 0.4 g). The charge−discharge performance studies were conducted on a multichannel battery tester (BTS-5 V5 mA and BTS-5 V10 mA, Neware) between the voltage of 1.4 and 2.1 V (1 C is defined as 115 mA h g−1). Cyclic voltammetry (CV) tests were conducted on a multichannel potentiostat (VMP3, Biologic) between 1.4 and 2.1 V versus Zn2+/Zn0 at a scan rate of 0.1 mV s−1. In addition, ac impedance measurements were performed with amplitude of 10 mV at the applied frequency range from 0.1 Hz to 1 MHz after the batteries were cycled for 10 cycles at 4 C rate. All electrochemical data were obtained at ∼20 °C.

Figure 1. Nitrogen adsorption and desorption on fumed silica.

The adsorption and desorption isotherms are generally classified as type-II isotherms, which represent nonporous materials.21 The BET surface area is 331.11 m2 g−1, and the Barrett−Joyner−Halenda (BJH) desorption cumulative surface area of pore is analyzed at 238.80 m2 g−1, which is equivalent to 72% of the surface area. The total BJH desorption pore volume is calculated at ca. 0.74 cm3 g−1. The huge specific surface area of fumed silica and functional groups offer tremendous benefits to its surface chemistry. Scheme 1 shows the representative structure of β-cyclodextrin and amorphous silica. Both materials contain multiple hydroxyl groups on the surface. Each CD molecule contains seven glucopyranoside units, which are linked together via an oxygen bridge to form a closed ring.22 Such a ring exhibits an inner diameter of 0.78 nm and a depth of 0.78 nm, and this Scheme 1. Structure of β-Cyclodextrin and a Typical Surface of Amorphous Silica

1805

DOI: 10.1021/acssuschemeng.6b02553 ACS Sustainable Chem. Eng. 2017, 5, 1804−1811

Research Article

ACS Sustainable Chemistry & Engineering provides fundamentals for rich host−guest and inclusion chemistries. The primary and secondary hydroxyl groups of CD are polar and may initiate hydrogen bonding. The pKa,1 of CD is about 12.2, and the pKa1,2 is about 13.5.23,24 FS has a pKa value of ∼6.8.25 The hydroxyl groups are dipoles and are clearly detected by infrared spectroscopy. Such functional groups from both fumed silica and CD may initiate gelation with aqueous electrolytes by the Coulombic interaction and hydrogenbonding formation with the water component of the electrolyte. The FT-IR results of CD, FS, and gel electrolytes containing CD and FS are reported in Figure 2. The OH and CO/

Figure 3. TGA of the conventional (00), the 5% FS (50), the 4% FS + 1% CD (41), and the 3% FS + 2% CD (32) electrolytes loaded in AGM.

electrolyte in AGM evaporates fully at about 70 °C, it goes up to ca. 120 °C to remove most of the water in the 10 wt % silicacontaining electrolyte. In the current system, only 3−5 wt % of fumed silica is used, but the gel can retain water at up to 120 °C. When increasing CD up to 2 wt %, the temperature required for fully dehydration of the electrolyte increases accordingly. The (32) electrolyte in AGM requires a temperature of up to 110 °C to fully release the water. It may be because there is a higher number of hydroxyl groups on the surface of β-cyclodextrin than on the surface of silica (Figure 2). Furthermore, the CD possesses a 0.67 nm “micropore”, which may exhibit a capillary filling effect. It may be concluded that the gel electrolytes are generally more stable at room temperature, and the water component of the electrolyte can effectively be kept in the gel since the AGM piece containing pristine aqueous electrolyte is dried off after exposure to the environment for about 30 min. Furthermore, the gel electrolytes may be more tolerant to temperature change than the pristine aqueous electrolyte since the gels are more resistant to heating compared with the aqueous electrolyte. Literature regarding silica-containing gel electrolytes for lithium-ion batteries confirms that adding silica in the electrolyte brings in higher stability and synergy between electrode and electrolyte, which leads to enhanced performance of the batteries.31,32 Table 1 shows the gelling time of the as-prepared gel electrolytes. After the gels transfer to the solid state, shear

Figure 2. FT-IR of CD, FS, 5% FS, 4% FS + 1% CD, 3% FS + 2% CD.

CC vibration modes of CD are observed at ca. 3289 cm−1 and ca. 1070−1077 cm−1.26 The OH stretching vibration band (at ca. 3200−3400 cm−1), the in-plane stretching vibration band of the SiO bond of the SiOH group (at ca. 960 cm−1), and the deformation vibration of surface H OH (at ca. 1640 cm−1) are really weak.27−29 The asymmetric stretching vibration of SiOSi is detected at 1070 cm−1, and the symmetric stretching vibration of SiOSi is observed at 800 cm−1. These bands are characteristics of amorphous silica. The FT-IR patterns of gel electrolytes exhibit a very strong and broad overlapped signal at about 3250−3261 cm−1, which is responsible for the vibrational modes of the hydrogen-bonded hydroxyl groups in the gel electrolytes, and the HOH bending vibration at 1640 cm−1 is significantly enhanced. The irregular and defect structure of the gel is characterized by the peak at 604−607 cm−1.30 The water retention capability of gel electrolytes in the AGM separator is confirmed by thermal gravimetric analysis, with temperature ramp rate of 1 °C min−1 up to 120 °C (Figure 3). Water in the AGM piece containing aqueous electrolyte evaporates almost completely around the temperature of 64 °C, while AGM pieces containing gel electrolytes complete dehydration only when the temperature reaches ca. 80, 100, 110 °C for samples 5 wt % of FS in aqueous electrolyte (the 50 electrolyte), 4 wt % FS + 1 wt % CD (the 41 electrolyte), and 3 wt % FS + 2 wt % CD (the 32 electrolyte), respectively (Figure 3). The TGA results illustrate that the CD additive increases the retention of water in the gel electrolytes. When in the AGM separator, the (50) electrolyte exhibits higher water retention than the pristine aqueous electrolyte (00) does. Our previous work about silica nanoparticle doped electrolytes in AGM shows good water retention upon heating.18 While the aqueous

Table 1. Gelling Time of the Gels When Prepared and When inside the AGM Separator gel types

5% FS

4% FS + 1% CD

3% FS + 2% CD

as-prepared as-prepared, after disturbeda

4h 1.5 h

17 h 3.7 h

33 h 7.9 h

a

This is the second gelling time. It means the gel was disturbed to become liquid, which was left undisturbed for gelling to occur. The gelling time was then collected.

stresses can be applied to liquefy the gel, and the gelling times of disturbed gel electrolytes were recorded. Sample (50) requires about 4 h to completely change to the gel state while samples (41) and (32) demand longer times. It must be mentioned that, by adding less than 5% of FS only, the electrolyte does not form a gel. Upon addition of higher quantities of CD in the gel, longer times are required for the gel 1806

DOI: 10.1021/acssuschemeng.6b02553 ACS Sustainable Chem. Eng. 2017, 5, 1804−1811

Research Article

ACS Sustainable Chemistry & Engineering to form, and the corresponding gels are softer than the sample (50) gel. The gel electrolyte cannot be obtained if more than 2% CD or less than 3% FS is used. The first gelling time, right after mixing, is always higher than consecutive gelling times after the application of shear stress. The second and successive gelling times are not changed significantly. After introduction of the as-prepared gel electrolyte into the AGM separator, the gel forms on the magnitude of minutes by visual inspection. This gelling duration is much shorter than the gelling time of the independent gels when they are outside of the separator. The results are supported by the observation from ball penetration experiments (Table 2). Increasing CD contents lead to the Table 2. Physical Characteristics of Gel Electrolytes gel types ball penetration depth (mm) conductivity of electrolyte (mS cm−1)a

5% FS

4% FS + 1% CD

3% FS + 2% CD

7.3

8.9

9.7

60.10 ± 0.10

61.73 ± 0.15

58.47 ± 0.31

Conductivity of the conventional aqueous electrolyte: 63.16 ± 1.30 (mS cm−1).

a

deeper penetrations of the ball into the gel electrolytes, and this represents the decreasing trend of the mechanical strength of the gel. The conductivity of the gel electrolytes is in the range 58.47−61.73 mS cm−1. In a comparison to the ionic conductivity of the pristine aqueous electrolyte, the average conductivity of the gels is ca. 5% smaller, which is reasonable because of the higher viscosity of gel electrolytes and thus slightly lower mobility of ions compared to pristine aqueous electrolyte. The conductivity does not change significantly upon changing the CD concentration from 0 to 2 wt %. Furthermore, the effects of the gel electrolytes on the electrochemical performance of the ReHAB are evaluated in details. After the batteries were assembled, tests were conducted including cyclic voltammetry, potentio-electrochemical impedance spectroscopy, and charge−discharge at various C-rates. The cyclic voltammetry results are presented on Figure 4a. All currents are normalized to the electrode loadings. In the CV results, there are two distinctive pairs of reduction/oxidation peaks, which correspond to the two steps of intercalation/deintercalation of Li+ into/out of the LiMn2O4 lattice. It is observed that there is only a slight increase in polarization in gelled battery systems compared to that of the conventional battery, suggesting that the changes in internal resistivity and kinetics of electrochemical reactions in the batteries are not significantly altered. This suggests that different types of electrolytes exert similar electrochemistry impacts on the intercalation and deintercalation of lithium ions in and out of the cathode. This is expected because both types of gelling agents are electrochemically inert under the testing conditions and in the voltage window of 1.4−2.1 V versus Zn2+/Zn. The CV curves of the batteries using different electrolytes show no significant changes, in contrast to the significant decrease in polarization observed when thiourea, an electrochemically active additive, is used as a dopant in the electrolyte.33 Potential static impedance spectroscopy (PEIS) results of batteries using the conventional liquid electrolyte and gelled electrolytes are presented in Figure 4b. The Ohmic resistance in (00) is slightly lower than those of gel electrolytes in agreement with slightly higher conductivity of (00) (see Table

Figure 4. (a) CV of the battery using conventional liquid electrolyte (the 00), batteries using 5% FS electrolyte (the 50), 4% FS + 1% CD electrolyte (the 41), and 3% FS + 2% CD electrolyte (the 32). (b) PEIS of batteries using conventional liquid electrolyte (the 00), 5% FS electrolyte (the 50), 4% FS + 1% CD electrolyte (the 41), and 3% FS + 2% CD electrolyte (the 32).

2). Furthermore, the charge transfer resistance of the gel electrolytes with CD is slightly higher than the electrolytes without CD, the (00) and the (50). This may be due to the slower diffusion of Li+ from the electrolytes through the interface between the electrolyte and the cathode to the inside of the cathode.34−38 Specific discharge capacity at different C-rates is summarized in Table 3. The (41) and (32) batteries deliver slightly higher specific discharge capacities than that of the (50) battery, whose performance is the same compared with that of the (00) cell. Increasing the C-rate leads to the decrease of the specific discharge capacity. The (00) battery retains ca. 80% of its capacity when increasing the C-rate from 0.2 to 4 C while it is 83% observed on the (41) cell. The specific discharge capacity of the (41) and the (32) batteries is ca. 10% higher than those of (50) and (00) at the 4 C rate. The results on the variation in rate capability imply that the batteries using gel electrolytes may perform better than that using the liquid electrolyte. The cousage of FS, a thixotropic gelling agent, and CD, a nonthixotropic gelling agent, may be a better choice compared with using the silica gelling agent alone since sample (41) offers a highest specific discharge capacity and highest capacity retention. In fact, current aqueous gels of lead-acid batteries are often constructed from thixotropic and nonthixotropic gelling agents.13 Figure 5a represents cyclability results of (00), (50), (41), and (32) batteries at 4 C for up to 1000 cycles. Figure 5b represents constant current (CC)−constant voltage (CV) 1807

DOI: 10.1021/acssuschemeng.6b02553 ACS Sustainable Chem. Eng. 2017, 5, 1804−1811

Research Article

ACS Sustainable Chemistry & Engineering Table 3. Discharge Capacities of Conventional and Doped Batteries (mAh g−1)

a

sample

conventional battery

5% FS

4% FS + 1% CD

3% FS + 2% CD

at 0.2 C at 4 C at 0.2 Ca

120.0 ± 3.8 95.6 ± 7.0 118.4 ± 5.1

121.1 ± 3.0 97.5 ± 2.0 119.7 ± 1.9

132.3 ± 5.7 109.6 ± 7.2 130.0 ± 4.8

130.9 ± 4.6 107.0 ± 9.0 128.5 ± 4.9

After a total 25 cycles at 0.2, 0.5, 1, 2, and 4 C.

The XRD patterns of the cathodes and anodes of the batteries after 1000 cycles in CC mode are obtained and shown in Figure 6a,b. All patterns shown in Figure 6a are normalized

Figure 5. (a) Constant current mode cyclability of batteries at 4 C up to 1000 cycles. (b) Constant current (CC)−constant voltage (CV) cycling of batteries at 4 C up to 1000 cycles, the current cutoff in CV period is set at 10% comparing with current at CC period. Figure 6. (a) XRD of cathode before and after battery run. (b) XRD of anode before and after battery run.

cycling of batteries at 4 C for up to 1000 cycles. The current cutoff in the CV period is set at 10% compared with the current in the CC period. Capacity fading is observed on all batteries because of unwanted side reactions, including the manganese dissolution, water evaporation, water decomposition, hydrogen evolution, and Zn corrosion.3,15,39 A typical (00) battery exhibits ∼53% capacity retention after 1000 cycles at 4 C, while (50), (41), and (32) batteries retain ∼60%, ∼61%, and ∼57%, respectively. With charge−discharge cycling of the batteries under a constant current−constant voltage protocol, all batteries show improved cycle life. The conventional (00) battery retains ∼62% while gelled batteries (50), (41), and (32) exhibit ∼64%, ∼68%, and ∼65% capacity retention. In general, the gel batteries show better cyclability than that of the conventional one, especially after a few hundred cycles. When cycling the batteries under both constant current and constant current−constant voltage modes, the gelled batteries outperform the conventional by a difference of ca. 6−8% (errors are estimated at ca. ±2.5%). In nonaqueous systems, lithium-ion batteries using silica-containing gel electrolytes also give a better performance compared with LiBs using original polymer gel electrolytes.40,41

by the normalization of peak (111) of the LiMn2O4 phase. The post-run cathodes of (00), (50), and (32) seem clean, but there is LiMn2O4·H2O detected in the post-run cathode of (41). The LiMn2O4 spinel structure remains intact on all post-run cathodes, and no significant peak shift is detected. In the fresh cathode, only one peak of graphite appears at ∼26°, while in the cycled cathodes not only is this peak significantly increased compared with peaks of other phases, but also new peaks of graphite appear. This changes in the peaks’ intensity exhibit the morphology and composition changes of different phases in the cathode. This might be due to the partial amorphization of graphitic carbon and the gradual dissolution of LiMn2O4 into the electrolyte.15,42,43 However, a comparison of the ratio of the graphite peaks to LiMn2O4 peaks indicates that the morphology changes are pronounced most in the postrun cathode of the reference (00) battery and least in the postrun cathode of the gelled (41) battery. The existence of a small quantity of LiMn2O4·H2O on the surface of the (41) anode may be due to incomplete drying of the cathode prior to the XRD experiment. 1808

DOI: 10.1021/acssuschemeng.6b02553 ACS Sustainable Chem. Eng. 2017, 5, 1804−1811

Research Article

ACS Sustainable Chemistry & Engineering It is observed on the XRD patterns of the post-run anode that new phases of ZnSO4·7H2O, Zn(OH)2, and Zn4SO4(OH)6 are detected in the XRD patterns of the post-run anodes of (00), (41), and (32). All patterns in Figure 6b are normalized by using the Zn (101) peak as standard. Interestingly, the postrun anode of the (50) battery shows the least changes in the anode after cycling. Structural changes on the anode surface are more pronounced in samples with the CD additive. From the literature, dendrites tend to growth vertically to the sample surface from the (100) and (110) crystal facets.44−46 After 1000 cycles of deposition−dissolution, all of the post-run anodes exhibit slightly higher (100) and (110) peaks. This mean dendrite may be a problem at higher cycle number, and it is not a problem just after 1000 cycles. This represents the dendrite resistance of the Zn when in contact with the gel electrolytes. The intensities of peaks (110) and (100) are slightly lower for the (41) post-run anode than for other samples. The uneven zinc deposition and dendrite growth phenomena must be mitigated so that the batteries can achieve long calendar life. Otherwise, the dendrites develop and ultimately reach the cathode to create a short circuit. Chronoamperometry is the method to identify the deposition current of zinc on the working electrode, e.g., the anode of the ReHAB, under a specific condition. The use of chronoamperometry as a tool to evaluate dendrite formation on the zinc surface has been studied widely.47−49 Smaller deposition currents when a polished zinc foil are in contact with the gel electrolyte, suggesting that the growth of dendrite on the zinc surface is suppressed. Previous studies suggested that highly uniform zinc deposition is expected at low current density.50−53 Furthermore, it has been shown that high mechanical modulus of electrolyte can suppress the dendrites growing on metal electrodes.54 On the other hand, immobilization of anions leads to a larger transference number of cations that are shown to suppress dendrite growth.55 Both silica particles and CD with OH groups can reduce the mobility of SO42− anions in the gel electrolytes through electrical interactions and hydrogen bonding. In this work, we apply a stepping voltage of −120 mV; the responding deposition current on the zinc working electrode is recorded and presented on Figure 7. The gel electrolytes offer 3.8 times the smaller absolute values in current density compared with that offered by the pristine aqueous electrolyte. Thus, it can be expected that the zinc deposition on the anode

of the gelled batteries is highly uniform. The results are supported by the scanning electron microscopy results (Figure 8). There is no visible trace of silica on the cathode, suggesting

Figure 8. SEM of anode and cathode of batteries after 1000 charge− discharge cycles. (a, b) Anode and cathode of the conventional (00). (c, d) Anode and cathode of the (50). (e, f) Anode and cathode of the (41). (g, h) Anode and cathode of the (32).

that this type of silica-containing electrolyte does not adhere well to the cathode, unlike what was observed on the post-run cathode of the battery using a silica nanoparticle doped electrolyte.18 However, upon addition of CD, the zinc surface is flatter compared with the zinc surface of post-run anodes of the (00) and the (50) batteries. This is a ground-breaking improvement since dendrites are clearly visible on the anode of the post-run battery using the pristine aqueous electrolyte since some of the AGM fragments are buried in the surface of the sample. All batteries are cycled for 1000 cycles before they were deassembled for microscopy examination. The post-run cathode of the (50) battery is similar to that of the conventional (00). The post-run anode of (00) shows dendrites growing on the zinc surface, while there is no sign of dendrite on the surface of the zinc in the post-run anode of the gelled batteries, suggesting that zinc dendrite formation is suppressed.



CONCLUSIONS The gel electrolytes containing thixotropic fumed silica and nonthixotropic β-cyclodextrin as gelling agents have been prepared and implemented in the rechargeable hybrid aqueous batteries (ReHABs). In the case where the fumed silica is the only gelling agent, a minimum of 5 wt % of fumed silica is required to form a gel electrolyte. The batteries assembled from

Figure 7. Chronoamperometry current density profiles of the Zn in different electrolytes during 1 h at the 120 mV overpotential vs Zn2+/ Zn. 1809

DOI: 10.1021/acssuschemeng.6b02553 ACS Sustainable Chem. Eng. 2017, 5, 1804−1811

Research Article

ACS Sustainable Chemistry & Engineering

(4) Wang, Q. J.; Song, W.-L.; Fan, L.-Z.; Shi, Q. Effect of Polyacrylonitrile on Triethylene Glycol Diacetate-2-propenoic Acid Butyl Ester Gel Polymer Electrolytes with Interpenetrating Crosslinked Network for Flexible Lithium Ion Batteries. J. Power Sources 2015, 295, 139−148. (5) Hazama, T.; Fujii, K.; Sakai, T.; Aoki, M.; Mimura, H.; Eguchi, H.; Todorov, Y.; Yoshimoto, N.; Morita, M. High-Performance Gel Electrolytes with Tetra-armed Polymer Network for Li Ion Batteries. J. Power Sources 2015, 286, 470−474. (6) Li, X. W.; Zhang, Z. X.; Yang, L.; Tachibana, K.; Hirano, S.-I. TiO2-based Ionogel Electrolytes for Lithium Metal Batteries. J. Power Sources 2015, 293, 831−834. (7) Fan, H. H.; Li, H. X.; Fan, L.-Z.; Shi, Q. Preparation and Electrochemical Properties of Gel Polymer Electrolytes Using Triethylene Glycol Diacetate-2-propenoic Acid Butyl Ester Copolymer for High Energy Density Lithium-ion Batteries. J. Power Sources 2014, 249, 392−396. (8) Fasciani, C.; Panero, S.; Hassoun, J.; Scrosati, B. Novel Configuration of Poly(vinylidenedifluoride)-based Gel Polymer Electrolyte for Application in Lithium-ion Batteries. J. Power Sources 2015, 294, 180−186. (9) Quinzeni, I.; Ferrari, S.; Quartarone, E.; Tomasi, C.; Fagnoni, M.; Mustarelli, P. Li-doped Mixtures of Alkoxy-N-methylpyrrolidinium Bis(trifluoromethanesulfonyl)-imide and Organic Carbonates as Safe Liquid Electrolytes for Lithium Batteries. J. Power Sources 2013, 237, 204−209. (10) Ferrari, S.; Quartarone, E.; Tomasi, C.; Ravelli, D.; Protti, S.; Fagnoni, M.; Mustarelli, P. Alkoxy Substituted Imidazolium-based Ionic Liquids as Electrolytes for Lithium Batteries. J. Power Sources 2013, 235, 142−147. (11) Li, W.; Dahn, J. R.; Wainwright, D. S. Rechargeable Lithium Batteries with Aqueous Electrolytes. Science 1994, 264, 1115−1118. (12) Kim, H.; Hong, J.; Park, K.-Y.; Kim, H.; Kim, S.-W.; Kang, K. Aqueous Rechargeable Li and Na Ion Batteries. Chem. Rev. 2014, 114, 11788−11827. (13) Lambert, D. W. H.; Greenwood, P. H. J.; Reed, M. C. Advances in Gelled-electrolyte Technology for Valve-regulated Lead-acid Batteries. J. Power Sources 2002, 107, 173−179. (14) Mohamad, A. A. Zn/gelled 6 M KOH/O2 Zinc−air Battery. J. Power Sources 2006, 159, 752−757. (15) Yan, J.; Wang, J.; Liu, H.; Bakenov, Z.; Gosselink, D.; Chen, P. Rechargeable Hybrid Aqueous Batteries. J. Power Sources 2012, 216, 222−226. (16) Han, Z.; Askhatova, D.; Doan, T. N. L.; Hoang, T. K. A.; Chen, P. Experimental and Mathematical Studies on Cycle Life of Rechargeable Hybrid Aqueous Batteries. J. Power Sources 2015, 279, 238−245. (17) Pillot, C. Avicenne Market Review; 2014. http://www.avicenne. com/pdf/The_rechargeable_battery_market_C_Pillot_ BATTERIES_2014_Nice_France_September2014.pdf (retrieved July 18th, 2016). (18) Lu, C. Y.; Hoang, T. K. A.; Doan, T. N. L.; Zhao, H. B.; Pan, R.; Yang, L.; Guan, W. S.; Chen, P. Rechargeable Hybrid Aqueous Batteries Using Silica Nanoparticle Doped Aqueous Electrolytes. Appl. Energy 2016, 170, 58−64. (19) Tantichanakul, T.; Chailapakul, O.; Tantavichet, N. Influence of Fumed Silica and Additives on the Gel Formation and Performance of Gel Valve-regulated Lead-acid Batteries. J. Ind. Eng. Chem. 2013, 19, 2085−2091. (20) Pan, K.; Shi, G.; Li, A.; Li, H.; Zhao, R.; Wang, F.; Zhang, W.; Chen, Q.; Chen, H.; Xiong, Z.; Finlow, D. The Performance of a Silica-based Mixed Gel Electrolyte in Lead Acid Batteries. J. Power Sources 2012, 209, 262−268. (21) Gregg, S. J.; Sing, K. S. W. Adsorption, Surface Area and Porosity; Academic Press: London, U.K., 1982. (22) Davis, M. E.; Brewster, M. E. Cyclodextrin-based Pharmaceutics: Past, Present and Future. Nat. Rev. Drug Discovery 2004, 3, 1023− 1035.

this type of gel exhibit higher cyclability than batteries using the conventional liquid electrolyte. Chronoamperometric studies reveal that the absolute values of deposition current densities are 3.5−4.0 times lower when gel electrolytes are used. The zinc deposition on the anode of the gelled batteries is uniform and flat, having no trace of dendrites, while dendrites are observed on the post-run anode of the reference battery. With the use of gel electrolytes containing both fumed silica and βcyclodextrin, the thermal tolerance improves, and gelling time and gelling strength are controllable upon controlling the concentration of thixotropic and nonthixotropic gelling agents. Compared with that of the reference battery, the specific discharge capacity is enhanced by 10% and the cyclability by up to 8% after 1000 cycles.



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Tuan K. A. Hoang: 0000-0002-5632-1976 Author Contributions §

T.K.A.H., T.N.L.D., and C.L. contributed equally to this research. T.K.A.H. conveyed the idea and prepared the gel electrolytes. T.N.L.D. conducted the chronoamperometry and scanning electron microscopy. C.L. prepared and tested batteries. T.N.L.D., M.G., and H.Z. helped with in-depth analyses. T.K.A.H. drafted and improved the paper under the guidance of P.C., who is the principal investigator and leads the research. All authors have given approval to the final version of this manuscript. Funding

This research was financially supported by Positec, Natural Sciences and Engineering Research Council of Canada (NSERC), Canadian Foundation for Innovation (CFI), Mitacs (IT06144, IT06460), Doctor Postgraduate Technical Project of Chang’an University (2014G5290009), the Fundamental Research Funds for the Central Universities (0009− 2014G2290017 and 2014G3292007), China Scholarship Council (.201406560003, 201406895017), and Shanghai University International Cooperation and Exchange Fund. Notes

The authors declare no competing financial interest.



ABBREVIATIONS Conventional, or the “00”, battery containing LiMn2O4 cathode, Zn anode, and aqueous electrolyte; the “50”, battery using gel electrolyte containing 5 wt % fumed silica; the “41”, battery using gel electrolyte containing 4 wt % fumed silica and 1 wt % β-cyclodextrin; the “32”, battery using gel electrolyte containing 3 wt % fumed silica and 2 wt % β-cyclodextrin



REFERENCES

(1) Wang, Y.; Zhong, W.-H. Development of Electrolytes towards Achieving Safe and High-Performance Energy-Storage Devices: A Review. ChemElectroChem 2015, 2, 22−36. (2) Hassoun, J.; Scrosati, B. ReviewAdvances in Anode and Electrolyte Materials for the Progress of Lithium-Ion and beyond Lithium-Ion Batteries. J. Electrochem. Soc. 2015, 162, A2582−A2588. (3) Xu, G. J.; Liu, Z. H.; Zhang, C. J.; Cui, G. L.; Chen, L. Q. Strategies for Improving the Cyclability and Thermostability of LiMn2O4-based Batteries at Elevated Temperatures. J. Mater. Chem. A 2015, 3, 4092−4123. 1810

DOI: 10.1021/acssuschemeng.6b02553 ACS Sustainable Chem. Eng. 2017, 5, 1804−1811

Research Article

ACS Sustainable Chemistry & Engineering (23) Norkus, E.; Vaškelis, A.; Vaitkus, R.; Reklaitis, J. On Cu(II) Complex Formation with Saccharose and Glycerol in Alkaline Solutions. J. Inorg. Biochem. 1995, 60, 299−302. (24) Gaidamauskas, E.; Norkus, E.; Butkus, E.; Crans, D. C.; Grinciene, G. Deprotonation of β-cyclodextrin in Alkaline Solutions. Carbohydr. Res. 2009, 344, 250−254. (25) Scherer, G. W.; Luong, J. C. Glasses from Colloids. J. Non-Cryst. Solids 1984, 63, 163−172. (26) Fu, L.; Lai, G. S.; Yu, A. M. Preparation of β-cyclodextrin Functionalized Reduced Graphene Oxide: Application for Electrochemical Determination of Paracetamol. RSC Adv. 2015, 5, 76973− 76978. (27) Zhdanov, S. P.; Kosheleva, L. S.; Titova, T. I. IR study of hydroxylated silica. Langmuir 1987, 3, 960−967. (28) Rezakazemi, M.; Vatani, A.; Mohammadi, T. Synergistic interactions between POSS and fumed silica and their effect on the properties of crosslinked PDMS nanocomposite membranes. RSC Adv. 2015, 5, 82460−82470. (29) Losq, C. L.; Cody, G. D.; Mysen, B. O. Complex IR Spectra of OH− Groups in Silicate Glasses: Implications for the Use of the 4500 cm−1 IR Peak as a Marker of OH− Groups Concentration. Am. Mineral. 2015, 100, 945−950. (30) Bertoluzza, A.; Fagnano, C.; Morrelli, M. A.; Gottardi, V.; Guglielmi, M. Raman and Infrared Spectra on Silica Gel Evolving Toward Glass. J. Non-Cryst. Solids 1982, 48, 117−128. (31) Raghavan, P.; Choi, J.-H.; Cheruvally, G.; Chauhan, G. S.; Ahn, H.-J.; Nah, C. W.; Ahn, J.-H. J. Power Sources 2008, 184, 437−443. (32) Shimano, S.; Zhou, H. S.; Honma, I. Preparation of Nanohybrid Solid-State Electrolytes with Liquidlike Mobilities by Solidifying Ionic Liquids with Silica Particles. Chem. Mater. 2007, 19, 5216−5221. (33) Wu, X. W.; Li, Y. H.; Li, C. C.; He, Z. X.; Xiang, Y. H.; Xiong, L.; Chen, D.; Yu, Y.; Sun, K.; He, Z. Q.; Chen, P. The Electrochemical Performance Improvement of LiMn2O4/Zn Based on Zinc Foil as the Current Collector and Thiourea as an Electrolyte Additive. J. Power Sources 2015, 300, 453−459. (34) Chen, S. Y.; Mi, C. H.; Su, L. H.; Gao, B.; Fu, Q. B.; Zhang, X. G. Improved Performances of Mechanical-activated LiMn2O4/ MWNTs Cathode for Aqueous Rechargeable Lithium Batteries. J. Appl. Electrochem. 2009, 39, 1943−1948. (35) Bisquert, J.; Garcia-Belmonte, G.; Bueno, P.; Longo, E.; Bulhões, L. O. S. Impedance of Constant Phase Element (CPE)blocked Diffusion in Film Electrodes. J. Electroanal. Chem. 1998, 452, 229−234. (36) Lu, D. S.; Li, W. S.; Zuo, X. X.; Yuan, Z. Z.; Huang, Q. M. Study on Electrode Kinetics of Li+ Insertion in LixMn2O4 (0 ≤ x ≤ 1) by Electrochemical Impedance Spectroscopy. J. Phys. Chem. C 2007, 111, 12067−12074. (37) Levi, M. D.; Gamolsky, K.; Aurbach, D.; Heider, U.; Oesten, R. On Electrochemical Impedance Measurements of LixCo0.2Ni0.8O2 and LixNiO2 intercalation electrodes. Electrochim. Acta 2000, 45, 1781− 1789. (38) Aurbach, D.; Markovsky, B.; Levi, M. D.; Levi, E.; Schechter, A.; Moshkovich, M.; Cohen, Y. New Insights into the Interactions Between Electrode Materials and Electrolyte Solutions for Advanced Nonaqueous Batteries. J. Power Sources 1999, 81-82, 95−111. (39) Hoang, T. K. A.; Doan, T. N. L.; Sun, E. K. K.; Chen, P. Corrosion Chemistry and Protection of Zinc & Zinc Alloys by Polymer-containing Materials for Potential Use in Rechargeable Aqueous Batteries. RSC Adv. 2015, 5, 41677−41691. (40) Aricò, A. S.; Bruce, P.; Scrosati, B.; Tarascon, J.-M.; van Schalkwijk, W. Nanostructured Materials for Advanced Energy Conversion and Storage Devices. Nat. Mater. 2005, 4, 366−377. (41) Walls, H. J.; Zhou, J.; Yerian, J. A.; Fedkiw, P. S.; Khan, S. A.; Stowe, M. K.; Baker, G. L. Fumed Silica-based Composite Polymer Electrolytes: Synthesis, Rheology, and Electrochemistry. J. Power Sources 2000, 89, 156−162. (42) Esbenshade, J. L.; Fox, M. D.; Gewirth, A. A. LiMn2O4@Au Particles as Cathodes for Li-Ion Batteries. J. Electrochem. Soc. 2015, 162, A26−A29.

(43) Wang, H. B.; Zeng, Y. Q.; Huang, K. L.; Liu, S. Q.; Chen, L. Q. Improvement of Cycle Performance of Lithium Ion Cell LiMn2O4/ LixV2O5 with Aqueous Solution Electrolyte by Polypyrrole Coating on Anode. Electrochim. Acta 2007, 52, 5102−5107. (44) Sawada, Y. Transition of Growth Form from Dendrite to Aggregate. Phys. A 1986, 140A, 134−141. (45) Mackinnon, D. J.; Brannen, J. M.; Fenn, P. L. Characterization of Impurity Effects in Zinc Electrowinning from Industrial Acid Sulphate Electrolyte. J. Appl. Electrochem. 1987, 17, 1129−1143. (46) Mackinnon, D. J.; Fenn, P. L. The Effect of Tin on Zinc Electrowinning from Industrial Acid Sulphate Electrolyte. J. Appl. Electrochem. 1984, 14, 701−707. (47) Xu, M.; Ivey, D. G.; Qu, W.; Xie, Z. Study of the Mechanism for Electrodeposition of Dendrite-free Zinc in an Alkaline Electrolyte Modified with 1-ethyl-3-methylimidazolium Dicyanamide. J. Power Sources 2015, 274, 1249−1253. (48) Xu, M.; Ivey, D. G.; Xie, Z.; Qu, W. Electrochemical Behavior of Zn/Zn(II) Couples in Aprotic Ionic Liquids Based on Pyrrolidinium and Imidazolium Cations and Bis(trifluoromethanesulfonyl)imide and Dicyanamide anions. Electrochim. Acta 2013, 89, 756−762. (49) Xu, M.; Ivey, D. G.; Qu, W.; Xie, Z.; Dy, E.; Yuan, X. Z. Zn/ Zn(II) Redox Kinetics and Zn Deposit Morphology in Water Added Ionic Liquids with Bis(trifluoromethanesulfonyl)imide Anions. J. Electrochem. Soc. 2014, 161, A128−A136. (50) Wang, K. L.; Pei, P. C.; Ma, Z.; Xu, H. C.; Li, P. C.; Wang, X. Z. Morphology Control of Zinc Regeneration for Zinc−air Fuel Cell and Battery. J. Power Sources 2014, 271, 65−75. (51) Banik, S. J.; Akolkar, R. Suppressing Dendrite Growth during Zinc Electrodeposition by PEG-200 Additive. J. Electrochem. Soc. 2013, 160, D519−D523. (52) Yuan, Y. F.; Tu, J. P.; Wu, H. M.; Wang, S. F.; Zhang, W. K.; Huang, H. Effects of stannous ions on the electrochemical performance of the alkaline zinc electrode. J. Appl. Electrochem. 2007, 37, 249− 253. (53) Banik, S. J.; Akolkar, R. Suppressing Dendritic Growth during Alkaline Zinc Electrodeposition using Polyethyleneimine Additive. Electrochim. Acta 2015, 179, 475−481. (54) Monroe, C.; Newman, J. The Impact of Elastic Deformation on Deposition Kinetics at Lithium/Polymer Interfaces. J. Electrochem. Soc. 2005, 152 (2), A396−A404. (55) Tikekar, M. D.; Archer, L. A.; Koch, D. L. Stability Analysis of Electrodeposition across a Structured Electrolyte with Immobilized Anions. J. Electrochem. Soc. 2014, 161 (6), A847−A855.

1811

DOI: 10.1021/acssuschemeng.6b02553 ACS Sustainable Chem. Eng. 2017, 5, 1804−1811