Realizing the Long Lifespan of Molybdenum Trioxide in Aqueous Aluminum Ion Batteries Through Potential Regulation
Haodong Fan, Xuejin Li, Jie Zhou, Xiaoning Wang, Xiuli Gao, Haoyu Hu, Li Zhou, Tonghui Cai, Yongpeng Cui, Pengyun Liu, Qingzhong Xue, Zifeng Yan, Wei Xing
Abstract
Open AccessRenewablesRESEARCH ARTICLES14 Aug 2023Realizing the Long Lifespan of Molybdenum Trioxide in Aqueous Aluminum Ion Batteries Through Potential Regulation Haodong Fan†, Xuejin Li†, Jie Zhou, Xiaoning Wang, Xiuli Gao, Haoyu Hu, Li Zhou, Tonghui Cai, Yongpeng Cui, Pengyun Liu, Qingzhong Xue, Zifeng Yan and Wei Xing Haodong Fan† State Key Laboratory of Heavy Oil Processing, School of Materials Science and Engineering, China University of Petroleum, Qingdao 266580 , Xuejin Li† *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] State Key Laboratory of Heavy Oil Processing, School of Materials Science and Engineering, China University of Petroleum, Qingdao 266580 , Jie Zhou State Key Laboratory of Heavy Oil Processing, School of Materials Science and Engineering, China University of Petroleum, Qingdao 266580 , Xiaoning Wang State Key Laboratory of Heavy Oil Processing, School of Materials Science and Engineering, China University of Petroleum, Qingdao 266580 , Xiuli Gao State Key Laboratory of Heavy Oil Processing, School of Materials Science and Engineering, China University of Petroleum, Qingdao 266580 , Haoyu Hu State Key Laboratory of Heavy Oil Processing, School of Materials Science and Engineering, China University of Petroleum, Qingdao 266580 , Li Zhou State Key Laboratory of Heavy Oil Processing, School of Materials Science and Engineering, China University of Petroleum, Qingdao 266580 , Tonghui Cai State Key Laboratory of Heavy Oil Processing, School of Materials Science and Engineering, China University of Petroleum, Qingdao 266580 , Yongpeng Cui State Key Laboratory of Heavy Oil Processing, School of Materials Science and Engineering, China University of Petroleum, Qingdao 266580 , Pengyun Liu State Key Laboratory of Heavy Oil Processing, School of Materials Science and Engineering, China University of Petroleum, Qingdao 266580 , Qingzhong Xue State Key Laboratory of Heavy Oil Processing, School of Materials Science and Engineering, China University of Petroleum, Qingdao 266580 , Zifeng Yan State Key Laboratory of Heavy Oil Processing, School of Materials Science and Engineering, China University of Petroleum, Qingdao 266580 and Wei Xing *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] State Key Laboratory of Heavy Oil Processing, School of Materials Science and Engineering, China University of Petroleum, Qingdao 266580 https://doi.org/10.31635/renewables.023.202200019 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail MoO3 is one of the most promising anode materials for aqueous aluminum batteries due to its high theoretical capacity and suitable aluminum insertion/de-insertion potential. However, the inferior cycling stability limits its further application, and the failure mechanism is still unclear. In this article, we provide a straightforward potential regulation technique to manage phase evolution during the charge/discharge process, which ultimately results in a markedly enhanced MoO3 electrode cycling stability. The failure mechanism study reveals that the excessive oxidation of the electrode during charge/discharge generates the H0.34MoO3 phase, which has high solubility and is the primary cause of MoO3 deactivation. Although the dissolved Mo species will be deposited onto the electrode sheet again, the deposition is not electrochemically active and cannot contribute to the capacitance. Controlling the cutoff potential prevented the production of H0.34MoO3, resulting in excellent cycling performance (80.1% capacity retention after 4000 cycles). The as-assembled α-MoO3//MnO2 full battery exhibits high discharge plateaus (1.4 and 0.9 V), large specific capacity (200 mAhg−1 at 2 Ag−1), and ultra-high coulombic efficiency (99%). The research presented here may contribute to the development of highly stable electrode materials for aqueous batteries. Download figure Download PowerPoint Introduction Aluminum (Al)-ion batteries hold great promise in grid-level energy storage by virtue of the exceptional theoretical capacity, abundant reserves, and low cost of Al.1–4 Specifically, aqueous aluminum-ion batteries (AAIBs) have attracted great attention because of the fast kinetics, environmental friendliness, and inherent safe nature of aqueous electrolytes.5,6 However, the performance of AAIBs is seriously limited by the unsatisfactory plating/stripping of metallic Al anode. This obstacle mainly arises from the much lower equilibrium redox potential of Al (−1.66 V vs standard hydrogen electrolyte), which is a competitive disadvantage for Al relative to the hydrogen evolution reaction (HER).7–10 Another reason is that an unwelcome passivating Al2O3 film forms due to the Al corrosion, and hydrogen is simultaneously evolved. Various tactics, such as protective layer coating, and selecting highly concentrated electrolytes,8,11,12 have been devised to address the problems of metallic Al anodes caused by hydrogen evolution and the Al passivating coating. Inspired by the solid electrolyte interface (SEI) in lithium-ion batteries, researchers first created an artificial SEI on Al anode, either "in situ" by 5 M aluminum triflate electrolyte13 or "ex situ" by ionic liquid (IL) electrolyte,8 to increase the cyclic stability of AAIBs. According to their findings, the SEI not only prevents the growth of passivating Al2O3 films but also facilitates quick Al3+ ion movement, enabling reversible Al plating and striping. The addition of alloying elements can reduce aluminum corrosion and passivation, which is another efficient method used to improve the stability of the aluminum anode. To significantly improve the anode's corrosion resistance, a variety of alloy anodes, such as eutectic Al-Ce alloys, Zn-Al alloys, Al-Cu alloys, and Al-Fe alloys, have been developed.9,14–18 In addition, an electrolyte method using hydrogel or "water-in-salt" electrolytes was created to produce rechargeable AAIBs.19–24 These types of electrolytes effectively minimize free water molecules, which results in less HER or corrosion of aluminum. The cyclability of AAIBs has been enhanced by the aforementioned tactics, but the anode side's reversible Al plating/stripping process has caused some questions. Research from groups of Azimi7 and Lin10 have demonstrated that the previously observed SEI on an Al anode is unstable and that HER continues to be the anodic response in the absence of Al deposition. To achieve reversible Al plating/striping in aqueous electrolytes, further work needs be completed. Another strategy is substituting metallic Al anode with alternative host materials, which has been widely applied in aqueous lithium-ion batteries25–27 and sodium-ion batteries.28–31 To prevent hydrogen evolution, the redox potential of such anode materials must be higher than the HER potential. In addition, the redox potential must be low enough to deliver excellent output voltage for a fully assembled battery. Taking both into account, a number of materials have been reported as anodes in AAIBs, including MoO3,32,33 TiO2,34,35 WO3,36 and anthraquinone.37 Due to its greater theoretical capacity and lower Al3+ intercalation/de-intercalation potential (−1.0∼0.5 V vs Ag/AgCl), MoO3 is thought to be the most promising of this group. The weak cycle performance of MoO3, however, prevents further advancement in AAIB. Some studies state that the capacity degradation was caused by the repeated intercalation and de-intercalation of Al3+, which caused MoO3 to collapse structurally. Despite the development of a few techniques, such as protective layer coating and the hydrogel electrolyte approach,23,38–41 the cycling performance of MoO3 is still unsatisfactory. Moreover, the capacity deterioration mechanism of MoO3 is yet unknown. Fully comprehending the energy storage mechanism and structural/phase evolution of MoO3 during the charge/discharge process is of vital importance to ameliorate MoO3. Herein, α-MoO3 was selected as the anode of AAIB and its phase structure evolution during the charge/discharge process, as well as the capacity deterioration mechanism, were analyzed in detail. We discovered that co-intercalation of Al3+ and H+ contributed to the capacity storage. During the charging process, an unstable phase of H0.34MoO3 forms, which is prone to dissolution in the electrolyte and seriously reduces the capacity of the battery. A voltage control strategy is recommended to stop the production of unstable H0.34MoO3 and thus successfully improve the cycling stability. The capacity remains as high as 80.1% after 4000 cycles in 0.5 M Al2(SO4)3 electrolyte after charge/discharge potential regulation, which is a major improvement over the capacity without regulation (only 17% remained after 50 cycles). Furthermore, we selected α-MnO2 as the cathode electrode material to assemble the full battery with the MoO3 anode. The MnO2//MoO3 battery shows high specific capacity of 180 mAh g−1 at 5 A g−1, and the capacity remains at 80% after 500 cycles. Results and Discussion α-MoO3 with an orthorhombic crystal structure (Figure 1a) was prepared by a hydrothermal method. As illustrated in Figure 1b, the X-ray diffraction (XRD) pattern of MoO3 powder manifests obvious peaks at 12.76°, 23.32°, 25.70°, 27.33°, and 38.97°, corresponding to (020), (110), (040), (021), and (060) crystal planes,42 respectively, which correlates exactly to the PDF card (JCPDS 05-0508). Figure 1c shows the Raman spectrum of MoO3 and the peaks at 289.7, 664, 819, and 994.6 cm−1 correspond to O=Mo=O, Mo–O shared by edges, Mo–O shared by corners, and the unshared Mo=O bond,43 respectively. Figure 1d,e is the scanning electron microscopy (SEM) and transmission electron microscopy (TEM) images of MoO3, which both demonstrate uniform rod-shaped morphology, with a width of about 250 nm and a length of up to micrometers. The mapping diagram shows that the elements consist of Mo and O (Figure 1f), and are distributed uniformly. The high-resolution TEM (HRTEM) image clearly depicts the lattice fringes of (004) and (110) planes with spacings of 0.35 and 0.38 nm, respectively (Figure 1g). The corresponding electron diffraction pattern confirms the single crystal structure of α-MoO3 (and Figure 1h). An enlarged X-ray photoelectron spectroscopy (XPS) spectrum of Mo 3d is shown in Figure 1i. Peaks at 236.05 and 232.90 eV correlate to Mo6+ 3d orbital binding energy, indicating that the synthesized product is pure phase MoO3. Figure 1 | Physical characterizations of α-MoO3. (a) The crystal structure of MoO3, where the orange spheres represent oxygen atoms and the green spheres represent Mo atoms. (b) XRD pattern. (c) Raman pattern, (d) SEM image, (e) and (f) TEM image and element mapping images, (g) HRTEM image, (h) selected area electron diffraction pattern, and (i) Mo 3d XPS spectra. Download figure Download PowerPoint The electrochemical performance of α-MoO3 was first evaluated with a three-electrode system to explore the possible failure mechanism of α-MoO3 in aqueous electrolytes. Figure 2a exhibits the cyclic voltammetry (CV) curve of α-MoO3 anode at various scan rates, and all curves display obvious redox peaks. Galvanostatic charge-discharge (GCD) profiles at various current densities also present flat charge/discharge plateaus and high coulombic efficiencies ( Supporting Information Figure S1). The low redox potential of MoO3 in 0.5 M Al2(SO4)3 electrolyte (–0.43 V vs Hg/Hg2SO4), high specific capacity (216 mAh g−1 at 1 A g−1), and impressive coulombic efficiency (more than 95%) imply that α-MoO3 is a reliable candidate anode for AAIBs. The cycle performance was examined at a high current density of 5 A g−1, as shown in Figure 2b. However, after 50 cycles, the capacity significantly decreases and is just 17% of the original capacity (31 mAh g−1). After cycling, the redox peaks nearly vanish from the CV curves (Figure 2c), whereas the charge/discharge plateaus progressively dissipate (Figure 2d). The reason behind this phenomenon is either the deactivation of Mo active sites or the gradual dissolution of MoO3. The structural evolution of MoO3 during cycling was then detected using XRD and Raman spectra of the MoO3 electrode after various charge/discharge cycles. The XRD data have been normalized and all electrodes were prepared simultaneously with the same area and mass loading. Apart from the peaks of MoO3, several new peaks corresponding to hydrogen molybdenum bronzes (H0.34MoO3, JCPDS 34-1230, denoting as phase I)44 emerge after the first cycle, suggesting that MoO3 underwent a partially irreversible conversion during the initial charge/discharge cycle. Different redox peaks in the CV curve at the second cycle confirm the irreversible conversion ( Supporting Information Figure S2). In the subsequent continuous cycling process, the characteristic peaks of α-MoO3 in the XRD patterns gradually reduce in intensity (Figure 2e) and the redox peaks in CV curves gradually disappear. Regarding the Raman spectra (Figure 2f), the peaks corresponding to α-MoO3 did not decrease much in intensity, while two broad peaks in the range of 200–500 cm−1 and 600–900 cm−1 are more obvious. Raman and XRD results suggest that a majority of the α-MoO3 has been converted into other Mo-based species and that the residual α-MoO3 is too few for XRD to distinguish. We also studied the morphology variation of the electrode during cycling (Figure 2g). An obvious rod-shaped feature is observed before cycling, while a dense surface with some pits is characterized after the cycling process. Energy-dispersive X-ray spectrometry (EDS) mapping reveals the uniform distribution of Al, Mo, and O ( Supporting Information Figure S3). We speculate that partial α-MoO3 dissolves and then a new species containing Al, Mo, and O is generated. The dissolution of active α-MoO3 should be the main reason for capacity deterioration. Figure 2 | Electrochemical performance of α-MoO3 and characterization of the electrode after cycling: (a) CV curves at different scan rates between 1 and 20 mV s−1, (b) cyclic stability at 5A g−1, (c) CV curves of α-MoO3 at 5mV s−1after the first cycle and the 200th cycle, (d) the charge/discharge curves after different cycles from 1 to 200 cycles, (e) XRD patterns, and (f) Raman spectra after different cycles, (g) SEM images of α-MoO3 electrode before and after cycling, (h) and (i) SEM images of the electrode that obtained by electrochemical deposition. Download figure Download PowerPoint After charge/discharge cycling ( Supporting Information Figure S4), the electrolyte becomes distinctly yellow as a result of the dissolution of Mo ions. The UV spectrum of the electrolyte presents apparent peaks at 209 and 239 nm corresponding to Mo6+ ions ( Supporting Information Figure S4), which is similar to that of 0.2 mM (NH4)6Mo7O24·4H2O/0.5 M Al2(SO4)3 solution. The Mo6+ ions in the electrolyte may be electrodeposited back to the electrode. After an electrochemical deposition procedure in the cycled electrolyte with graphite paper as the current collector, apparent nanoparticles composed of Al, Mo, and O were deposited (Figure 2h,i and Supporting Information Figure S5). The electrodeposited species contributes to two broad peaks in the Raman spectrum ( Supporting Information Figure S6a), which is consistent with the findings in Figure 2f. But this electrodeposited species is not electrochemically active and contributes little capacity as illustrated by the CV curve ( Supporting Information Figure S6b). Therefore, we suggest that the substantial capacity degradation should be attributed to the active Mo species dissolving during the charge/discharge process. In light of the fact that α-MoO3 is inherently insoluble in the electrolyte ( Supporting Information Figure S7), the dissolution primarily takes place during the charge–discharge operation. To boost the cycling stability of α-MoO3, the dissolution must be suppressed. To examine the possible reason for the Mo dissolution, the electrode was cycled under varying potential regions (region I: −0.45∼0 V and region II: −0.9∼−0.45 V). Figure 3a,b shows that in region I, the capacity decreases considerably (14.6%) after 100 charge/discharge cycles, whereas in region II, capacity remains intact even after 100 repeated cycles, suggesting that the process of dissolution mainly takes place at high potential. The difference in discharge profile in the first cycle is caused by the irreversible phase conversion, which will be discussed later. Cycling performance under various upper cutoff potentials (−0.45, −0.3, and 0 V) reveals that the α-MoO3 electrode cycled in the potential range of −0.9 to −0.3 V exhibits the best cycling stability (capacity maintains at 80.1% after 4000 cycles) with a high capacity (137 mAh g−1 at 5 A g−1) (Figure 3c). When the lower cutoff potential is set to −1.1 V, serious hydrogen evolution occurs as illustrated by the H2 bubbles on the electrode surface ( Supporting Information Figure S8), and results in low coulombic efficiency ( Supporting Information Figure S9). All the aforementioned findings imply that the active Mo species experiences runoff within the potential range of −0.3 to 0 V. XRD of α-MoO3 electrodes charged to various potentials has been carried out to identify the possible soluble intermediate species. XRD patterns demonstrate the generation of H0.9MoO3 (JCPDS 53-1024, phase III), H0.6MoO3 (JCPDS 01-070-4477, phase II),45 and a mixture of H0.34MoO3 and MoO3 at −0.45, −0.3, and 0 V, respectively (Figure 3d). Lower hydrogen concentrations in HxMoO3 make it more acid soluble and less acid resistant.46 Thus, we propose that the generation of H0.34MoO3 upon charging to 0 V should be the primary reason for Mo dissolution and capacity deterioration. Figure 3 | Electrochemical performance of the different potential ranges and characterizations of the failure mechanism: changes in capacity before and after 100 cycles in specific regions (a) region I: −0.45 to 0 V and (b) region II: −0.9 to −0.45 V; (c) the long-term cycling performance of MoO3 in different potential ranges; (d) XRD patterns of electrodes charged to different potentials; (e) XRD patterns of the electrodes after 100 cycles in the range of −0.9∼0 V and −0.9∼−0.3V, respectively; (f) Raman patterns of the electrodes after 100 cycles in the range of −0.9∼0 V and −0.9∼−0.3V, respectively; (g) SEM image of MoO3 electrode after cycling in the range of −0.9∼−0.3V. Download figure Download PowerPoint The electrode exhibits stable charge/discharge behavior in the potential range of −0.9 to −0.3 V as illustrated by the unchanged GCD profiles after various cycles ( Supporting Information Figure S10). After 100 charge/discharge cycles, the electrode XRD pattern displays identical peaks (Figure 3e). The Raman spectrum of the electrode also indicates that no deposited Mo species was formed when the electrode is charged/discharged in the potential range from −0.9 to −0.3 V (Figure 3f). The rod-like structure is well maintained and the dense layer of Mo-based deposition is not observed (Figure 3g). These results demonstrate that the potential regulation strategy is effective in enhancing the α-MoO3 electrode stability in aqueous electrolyte. After that, a comprehensive evaluation of the electrochemistry was conducted in a half-cell within the potential range of −0.9 to −0.3 V. All CV curves at various scan rates exhibit three reduction peaks and one oxidation peak (Figure 4a). The CV curves show no obvious polarization as the scan rate is increased from 1 to 30 mV s−1, demonstrating good reaction kinetics. The peak current i can be related to the scan rate ν by the following equation47: i = a v b (1)where a and b are adjustable parameters. When b is close to 0.5, electrochemical reactions are predominantly governed by ion diffusion, while b values close to 1 indicate surface-dependent capacitive behavior. According to Figure 4b, Peak1, Peak2, and Peak4 all have b values near 0.5, whereas Peak3 has a b value near 1, indicating that redox reactions are mainly controlled by diffusion.47–49 The electrode nevertheless displays fast kinetics even though the redox reaction is mostly controlled by the ion-diffusion process. With a tiny polarization of only 50 mV, the GCD profiles display three discharge platforms and one (Figure With scan rate and current the reduction reaction contributes to more capacity due to its behavior. a current density of 1 A g−1, a large capacity of mAh g−1 is and mAh g−1 is maintained even at a high current density of 20 A g−1, its excellent rate performance (Figure Moreover, a high coulombic efficiency of is indicating the high of the redox We have also the CV of pure graphite paper in the same electrolytes ( Supporting Information Figure and is no obvious redox peak in the CV demonstrating the capacity of graphite Figure | The electrochemical performance of α-MoO3 and structural evolution during charging and within the potential range of (a) CV curves at different scan rates from 1 to 20 mV s−1, (b) (i) at peak (c) charge/discharge curves at different current densities from 1 to 20 A g−1, (d) rate (e) and discharge (f) XRD patterns at different charge/discharge (g) Al XPS spectra of MoO3 electrodes at fully charged and fully and (h) of the intercalation/de-intercalation of H+ and Al3+ in the crystal structure of α-MoO3. Download figure Download PowerPoint XRD (Figure were then carried out to the charge/discharge behavior of MoO3 electrode. After to V in the first cycle the diffraction peaks corresponding to α-MoO3 at 12.76°, and to and respectively, and a new peak at suggesting the gradual ion intercalation in α-MoO3 and the of The peaks at and vanish after to V while two new peaks at and the diffraction peak at to two peaks at and which that the H0.34MoO3 phase has into a phase of (JCPDS and (JCPDS With further to −0.9 V, only characteristic peaks corresponding to are In the following charging process, the electrode gradually to H0.9MoO3 charging to V at and H0.6MoO3 charging to −0.3 V at XRD results indicate that a gradual H+ occurs during the process, but the H+ is not fully reversible in the first cycle, which can also be by the irreversible changes of Mo ( Supporting Information Figure The irreversible reaction is either caused by the electrochemical from H0.6MoO3 to H0.34MoO3 or H0.34MoO3 to In the following cycles, a reversible conversion between and occurs in Figure and the of H+ intercalation can be by the XRD is that the phase changes from H0.6MoO3 to H0.9MoO3 after to V, which that the first discharge between −0.3 to V is to H+ should be that a discharge can be between and yet the phase of electrode remains unchanged Therefore, Al3+ is also to provide a major to the charge/discharge capacity and the discharge V should be to the Al3+ The of Al3+ is by the XPS which shows that the Al arises after the discharge process and then after the process (Figure mapping results of the electrode of the intercalation of Al3+ ( Supporting Information Figure The phase further changes from H0.9MoO3 to after to V, suggesting that the discharge V should be to the H+ should be that with the current the discharge of H+ intercalation gradually due to the kinetics of During the charging process, the first between −0.9 and V should from the de-intercalation as by the XRD while the V is attributed to the Al3+ We also evaluated the electrochemical performance of the MoO3 electrode in M electrolyte to the H+ in 0.5 M which a much lower specific capacity, as shown from CV and GCD results ( Supporting Information Figure Therefore, we propose that the energy storage in the MoO3 electrode is by the of H+ and As illustrated in Figure during the discharge process, H+ and Al3+ into MoO3 In the following charging process, H+ and Al3+ The reaction in the first cycle is not fully and a of in the lattice of MoO3 after fully In the following cycles, a reversible between and H0.6MoO3 takes place that results in a stable charge/discharge To the of α-MoO3, an α-MoO3//MnO2 full battery was assembled as illustrated in Figure The α-MoO3 anode energy by while the cathode with a The reaction potential of the α-MoO3 anode can be within the range from −0.9 to −0.3 V by the mass of MoO3 to (Figure As a two discharge plateaus were observed at V and V, respectively. As shown in Figure the assembled full battery presents two of apparent redox peaks at and 0.9 V at all scan rates from 1 to mV The GCD profiles at various current densities are shown in Figure which all exhibit obvious charge/discharge A high specific capacity of 200 mAh g−1 is at a current density of 2 A The full battery also presents excellent rate with a high capacity retention of even at a high current density of A g−1 and a high coulombic efficiency of over (Figure Moreover, the battery exhibits cycling which presents a high capacity of mAh g−1 after 500 repeated charge/discharge cycles (Figure 80% of the original As illustrated in Figure the battery can a large composed of for over 5 suggesting the promising of the full Figure 5 | Electrochemical performance of full (a) of the charge/discharge mechanism of the assembled full (b) the charge/discharge curve and the corresponding voltage curve of the cathode and anode at 2 A (c) CV curves at different scan rates from 1 to mV (d) charge/discharge curves at different current densities from 2 to A (e) rate capacity of the full (f) cyclic stability at 5 A and (g) three successfully light up the Download figure Download PowerPoint a metallic Al anode with alternative MoO3 anodes can the corrosion and hydrogen evolution of metallic Al in AAIB. However, the inferior stability of MoO3 in aqueous electrolyte still limits its In this the failure mechanism of MoO3 is discovered and reveals that the generation of serious Mo With a potential regulation the generation of is an enhanced cycling stability of MoO3 in aqueous