Ting
Yang‡
,
Diheng
Xin‡
,
Nan
Zhang
,
Jing
Li
,
Xianchi
Zhang
,
Liqin
Dang
,
Qi
Li
,
Jie
Sun
,
Xuexia
He
,
Ruibin
Jiang
,
Zonghuai
Liu
and
Zhibin
Lei§
*
Key Laboratory of Applied Surface and Colloid Chemistry, MOE, Shaanxi Engineering Lab for Advanced Energy Technology, Shaanxi Key Laboratory for Advanced Energy Devices, School of Materials Science and Engineering, Shaanxi Normal University, 620 West Chang'an Street, Xi'an, Shaanxi 710119, China. E-mail: zblei@snnu.edu.cn; Fax: +86-29-81530702; Tel: +86-29-81530810
First published on 20th March 2024
V2O5 is one of the promising cathodes for aqueous zinc ion batteries. However, its performance is largely hindered by its low conductivity and poor cycling stability due to the electrode dissolution. In this work, an interfacial polymerization strategy is developed to prepare the V2O5@poly(3,4-ethylenedioxythiophene) (PEDOT) core-sheath nanowire electrode. The interfacial redox reactions between the vapor 3,4-ethylenedioxythiophene molecules and V2O5 nanowires initiate the polymerization reaction to yield uniform PEDOT sheaths with varying thickness controlled by the reaction duration. The PEDOT sheaths are found to improve the electrode conductivity, suppress the V2O5 nanowire dissolution, and improve the specific capacity. Theoretical simulation further shows that the PEDOT sheath weakens the interactions between Zn2+ and the V2O5 host, thus benefiting the extraction of Zn2+ from the host electrode and boosting the cycling stability. Consequently, V2O5@PEDOT-50m delivers a specific capacity of 293 mA h g−1 at 0.1 A g−1 and 225 mA h g−1 at 1 A g−1, which are superior to 205 mA h g−1 and 142 mA h g−1 of the pristine V2O5 nanowire electrode, respectively. Moreover, V2O5@PEDOT-50m maintains 97.8% and 99% capacity retention after 100 and 2000 cycles, respectively. The significantly enhanced performances with respect to the corresponding V2O5 nanowire counterpart demonstrate that the PEDOT sheaths developed by the interfacial polymerization could become an effective method to stabilize the vanadate-based cathodes for zinc ion storage.
Rechargeable aqueous zinc ion batteries (ZIBs) hold great potential for grid-scale energy storage by virtue of the large theoretical capacity (820 mA h g−1), suitable redox potential (−0.763 V vs. SHE), and high volume capacity (5855 mA h cm−1) of the metal zinc.4–6 The use of an aqueous electrolyte can significantly reduce the safety risk and simultaneously improve the rate capability of aqueous ZIBs due to the fact that their ion conductivity is several orders of magnitude higher than that of an organic electrolyte.7 Moreover, the abundant zinc resources, simple manufacturing technique, and low-toxicity aqueous electrolyte make the ZIBs a highly competitive energy storage device. However, the practical application of ZIBs is primarily hindered by the dissolution and poor stability of cathodes, dendrite growth, and side reactions of the Zn anode sides.8,9 In particular, the exploration of stable cathodes with high specific capacity and long cycling life remains a challenge for the future practical application of ZIBs.
So far, manganese-based oxides,10,11 Prussian blue analogues12,13 and vanadium based compounds14–18 have been extensively investigated as potential cathodes for ZIBs. Among them, the vanadium oxides have attracted much interest due to their multiple redox reactions and layered or tunnel structure.19–21 V2O5 is a layered structure in which each layer is built by the edge and corner-sharing VO5 pyramid, and can offer a theoretical capacity of 589 mA h g−1 based on the two-electron redox reaction.5 Despite these structural advantages, the V2O5 cathodes face capacity decay due to electrode dissolution and structural collapse during the repeated charging and discharging process. To resolve these issues, pre-insertion of cations,14,22,23 metal ion doping,24 intercalation of H2O molecules25 or two-dimensional materials18,26 have been reported to stabilize the V2O5 cathodes. In spite of the improved performances, some issues still exist, including decreased valence state of V, reduced active sites and increased interactions between the pre-inserted cations and oxide layers of the host cathode.26 On the other hand, the intrinsic low conductivity of V2O5 (10−3–10−4 S cm−1)7,27 is inadequate for high-rate applications. Consequently, the conductive Ti3C2Tx MXene layer or SWCNT are assembled with vanadium oxides to enhance the electrode conductivity and suppress the vanadium dissolution.28,29 Alternatively, the conductive polymers poly(3,4-ethylenedioxythiophene) (PEDOT)30,31 are also attractive for stabilizing V2O5-based cathodes.32,33 For instance, the small molecules PEDOT were inserted into V2O5 nanoflakes to improve the electrode conductivity and facilitate the Zn2+ diffusion kinetics.30 In another work by Cao et al.,32 polymerization of PEDOT layers on the sodium vanadate nanobelts (Na0.76V6O15) was found to produce oxygen vacancies, which offers a large interplanar distance, enhances electron transfer and benefits Zn2+ diffusion. However, the PEDOT layers with controllable thickness still remain a great challenge for high-efficiency Zn2+ storage. Moreover, the effects of the PEDOT layer thickness on the reaction kinetics have rarely been investigated.
In this work, a facile interfacial polymerization reaction between vapor 3, 4-vinyl dioxthiophene (EDOT) and V2O5 nanowires is reported to prepare the V2O5@PEDOT core-sheath electrode. The redox reactions at the EDOT/V2O5 interfaces yield uniform PEDOT sheaths with thickness varying from 23 to 43 nm by extending the polymerization durations. The PEDOT sheaths are proven to enhance the nanowire conductivity, weaken the interactions between the Zn2+ ions and V2O5 host layers, and suppress the electrode dissolution. As a consequence, the V2O5@PEDOT nanowires exhibit dramatically enhanced performances in term of the specific capacity, long-term cyclability and rate performance, making the interfacial polymerization a facile yet effective solution to stabilize vanadium-based cathodes for future grid-scale zinc ion storage.
Fig. 2 Digital photographs (a), SEM images (b–g), XRD patterns (h), Raman spectra in the low (i) and high frequency (j) regions of the aerogels composed of V2O5 nanowires and V2O5@PEDOT nanowires. |
The phase structures of the V2O5@PEDOT products were identified by XRD and Raman spectroscopy (Fig. 2h–j). The V2O5 nanowires exhibit sharp diffraction peaks at 15.3°, 20.2°, 21.7° and 31.0°, which can be indexed to the (200), (001), (101) and (301) diffraction planes of orthorhombic V2O5 (JCPDS no 41-1426),34,35 respectively, and suggests high purity and good crystallinity of the nanowires. In contrast, the V2O5@PEDOT samples exhibit XRD patterns similar to those of the pristine V2O5, except that all of the intensities are slightly reduced. A plausible reason for such changes is due to the scarified V2O5 nanowires, which react with EDOT molecules and yield the low crystallinity of V2O5@PEDOT. The Raman spectroscopy of pristine V2O5 and V2O5@PEDOT nanowires are illustrated in Fig. 2i and j, respectively. In the low-frequency region, the intense peak at around 139 cm−1 is attributed to the skeleton bending vibration of the V–O–V bond. The peaks at 286 and 408 cm−1 correspond to the bending vibrations of O3–VO and V–O3–V, respectively.18 The distinct peak at 993 cm−1 is ascribed to the in-phase stretching vibration mode of the apical VO bond.36,37 According to Fig. 2i, the Raman peaks of V2O5@PEDOT do not show noticeable changes in both position and intensity after PEDOT coating. In the high-frequency region (Fig. 2j), Raman bands at around 1359, 1437, 1492 and 1573 cm−1 are attributed to the symmetric and asymmetric tensile vibrations of Cα and Cβ in PEDOT.38,39 However, unlike the profiles of V2O5@PEDOT in the low-frequency region, these peaks in the high-frequency region gradually increase with the polymerization duration, which are consistent with the optical images, and thus indicate that more PEDOT have been coated on the V2O5 nanowires at longer reaction times.
The uniform coating of PEDOT on the V2O5 nanowires are further confirmed by TEM characterization. Fig. 3a–f shows the TEM images of a single V2O5 and V2O5@PEDOT nanowire at different reaction times. Clearly, both V2O5 and V2O5@PEDOT nanowire display smooth surfaces without any PEDOT aggregates, which coincides with the SEM images and demonstrates the homogeneous PEDOT coating on each V2O5 nanowire. The uniform coating is further elucidated by the elemental mapping images, as displayed in the overlay (Fig. 3b1–3f1) and individual elemental mapping images (Fig. 3b2–3f2). In contrast to the V2O5 nanowire (Fig. 3a1 and 3a2), all V2O5@PEDOT samples display additional S and C elements, which are exclusively from the PEDOT molecules. The uniform distribution of both S and C elements outside the V2O5 nanowires reveals the effective interfacial polymerization reactions. Besides the uniform PEDOT sheaths, their thicknesses could also be facilely adjusted by simply controlling the interfacial polymerization durations. On the basis of the overlay elemental mapping (Fig. 3b1–3f1), the PEDOT sheath thickness can be roughly determined, which are plotted in Fig. 4a. The results showed that extending the polymerization time from 20 to 60 min leads to the thickness increase of the PEDOT sheath from 23 to 43 nm. Importantly, such PEDOT sheaths could enhance the overall conductivity of the V2O5@PEDOT electrodes. As depicted in Fig. 4b, the V2O5@PEDOT-30 m exhibits a conductivity that is 4-fold higher (4.2 × 10−3 S cm−1) as compared with pristine V2O5 nanowires (1.09 × 10−3 S cm−1). When the reaction time is over 40 min, the conductivity of V2O5@PEDOT reaches an approximate plateau of ∼6 × 10−3 S cm−1.
Fig. 3 TEM images (a–f), the corresponding overlay elemental mapping images (a1–f1), and the V, O, S, C elemental mapping images (a2–f2) of the V2O5 nanowires and V2O5@PEDOT nanowires. |
Fig. 4 PEDOT sheath thickness (a), the conductivity (b), survey XPS spectra (c), V 2p XPS spectra (d) and S 2p XPS spectra (e) of the V2O5 and V2O5@PEDOT nanowires. |
The XPS characterization was carried out to determine the valence states of the V2O5 nanowire at different polymerization durations (Fig. 4c–e). Besides the peaks at 536 and 516 eV corresponding to the O 1s and V 2p (Fig. 4c), respectively, the survey XPS spectra of the V2O5@PEDOT nanowires exhibit additional S weak signals at ∼164 eV, confirming the co-existence of V2O5 and PEDOT in the V2O5@PEDOT nanowires. The high-resolution XPS spectrum of the V2O5 nanowires exhibits a strong peak at 517.3 eV and a weak peak at 524.5 eV (Fig. 4d). These peaks correspond to the spin–orbit splitting of V 2p3/2 and V 2p1/2, respectively, and suggest a V5+ chemical valence in the V2O5 nanowires.14 As the interfacial polymerization reactions proceed, these two peaks shift toward low binding energy. More importantly, accompanied with the down-shift of these peaks, two additional peaks at 516.2 and 523.0 eV are observed for all of the V2O5@PEDOT samples. These weak peaks are ascribed to V4+.40 The reduction of partial V5+ into V4+ by EDOT molecules induces the uniform PEDOT sheaths, which are further confirmed by the S 2p XPS spectra. As displayed in Fig. 4e, two peaks of S 2p3/2 (163.7 eV) and S 2p1/2 (165.1 eV) stem from the thiophene in PEDOT molecules.38 By integrating the peak areas of V4+ to the total V peak areas, the amount of V4+ in the V2O5@PEDOT was calculated in Table S1.† The ratio of the V4+/V increases with the polymerization time, and reaches a maximal value of 16.5% for V2O5@PEDOT-50m. Such changes decrease the average valence state of V from +5.0 of the initial V2O5 nanowires to +4.85 of V2O5@PEDOT-60m (Table S1†). Moreover, due to the formation of the PEDOT sheath, the contents of the surface sulfur increase, while that of the surface vanadium decreases accordingly. The above observations coincide well with the thickness of PEDOT sheaths, and thus demonstrate that the structure of V2O5@PEDOT nanowires can be facilely controlled by adjusting the interfacial oxidation polymerization time.
The electrochemical properties of the V2O5@PEDOT nanowires as cathodes for zinc ion storage were assessed by assembling a series of coin-type cells with zinc foil as the anode and 2.0 M Zn(CF3SO3)2 as the aqueous electrolyte. Fig. 5a and Fig. S3† show the galvanostatic charge discharge profiles of the V2O5@PEDOT nanowires at 0.1 and 0.5 A g−1, respectively. Clearly, two discharge potential plateaus in the range of 1.01–0.81 V and 0.64–0.40 V (vs. Zn2+/Zn) are evidenced, which are attributed to the insertion of Zn2+ into the V2O5@PEDOT nanowires. Fig. 5b compares the discharge capacity of V2O5@PEDOT at a current density of 0.1 and 1.0 A g−1. Specifically, the V2O5@PEDOT-50m electrode exhibits discharge capacities of 293 and 225 mA h g−1 at 0.1 and 1.0 A g−1, respectively, which are much higher than that of the V2O5 nanowires (205 and 142 mA h g−1) and PEDOT nanowires electrode (68 and 58 mA h g−1) at the same current density (Fig. S4†). It is noted that the capacity of the V2O5@PEDOT-50m is much higher than those of the cationic intercalated vanadium oxide cathodes, which are usually below 200 mA h g−1.17 The Nyquist plots of the V2O5 and V2O5@PEDOT-50m nanowires electrodes are compared in Fig. S5.† Clearly, the V2O5@PEDOT-50m electrode exhibits significantly reduced charge transfer resistance, which is closely related to the conductive PEDOT sheath coating. The enhanced conductivity affords V2O5@PEDOT-50m with high rate capability. The average discharge capacities decrease from 293 to 224, 169, 136 and 110 mA h g−1 upon the increase of the current density from 0.1 to 0.5, 1.0, 2.0 and 5.0 A g−1, respectively (Fig. 5c). Importantly, even with the 50-fold current increase from 0.1 to 5.0 A g−1, the electrode is still capable of retaining 37.5% capacity (110 mA h g−1) with respect to the initial capacity of 293 mA h g−1 at 0.1 A g−1 (Fig. 5d). In contrast, only 29.0% capacity can be maintained for the V2O5 nanowires electrode under the identical testing condition (Fig. 5d and S6†). Moreover, the V2O5@PEDOT-50m electrode could resume 62% of its initial capacity when returning the current density to initial 0.1 A g−1 after 90 cycles. On the basis of these results, the larger capacity and enhanced rate capacity of the V2O5@PEDOT-50m electrode are attributed to the uniform PEDOT sheaths, which not only improve the electrode conductivity, but also boost the electrode electrochemical durability.
Apart from the enhanced capacity and rate capability, PEDOT sheaths also significantly boost the cycling performances of the V2O5@PEDOT electrode. In spite of different PEDOT sheath thicknesses, the V2O5@PEDOT electrodes exhibit overall enhanced cycling stability with respect to the V2O5 nanowire electrodes at 1.0 A g−1 (Fig. S7†). Taking V2O5@PEDOT-30m as example, a capacity retention of 82.6% relative to the 32.1% of the V2O5 nanowires has been achieved. Fig. 5e and S8† show the long-term durability of the V2O5@PEDOT-50m electrode at 0.1 A g−1 and 10 A g−1, respectively. After continuous cycling at 0.1 A g−1 for 100 cycles, the V2O5@PEDOT-50m electrode exhibits a discharge capacity of 265 mA h g−1, and a capacity retention of 97.8% with respect to its first discharge capacity of 271 mA h g−1 (Fig. 5e). Conversely, although the V2O5 nanowires electrode has an initial discharge capacity of 206 mA h g−1, it decays rapidly to 52.5 mA h g−1 after 100 cycles at 0.1 A g−1 (25.5% capacity retention). In the meantime, at a high current density of 10 A g−1, an initial discharge capacity of 89 mA h g−1 and a capacity maintenance of 99% (88.2 mA h g−1) have been achieved for the V2O5@PEDOT-50m electrode after 2000 cycles (Fig. S8†). These performances are superior to the counterpart V2O5 nanowires with a low initial capacity of 51.8 mA h g−1 and a poor capacity maintenance of 41.8%. Table S2† summarizes the performance of vanadium and manganese-based cathodes. While the V2O5@PEDOT-50m displays relatively low specific capacity as compared with the V2O5·nH2O/graphene,25 porous V2O5 microspheres41 or 3D@V2O5,42 its cycling stability is remarkably improved as compared with V2O5, cationic intercalated vanadium oxides,43,44 and some MnO2-based cathodes10,45 (Table S2†). In addition to the above outstanding electrochemical stability, the V2O5@PEDOT-50m electrode exhibits nearly 100% coulombic efficiency at both low and high current densities, suggesting the superior reversibility of the Zn2+ insertion/extraction within this electrode. The elemental mapping images of the cycled V2O5@PEDOT-50m are shown in Fig. 5f. The core-sheath structure is well maintained even after continuous charging and discharging at 0.1 A g−1 for 100 cycles, further confirming the vital role of the PEDOT sheaths in protecting the structural integrity of the V2O5@PEDOT-50m electrode.
It is known that the strong electrostatic interactions between the Zn2+ and V2O5 host could trap a certain amount of Zn2+,9 leading to the unsatisfying electrochemical reversibility and poor cycling performances. In order to explore the impact of the PEDOT sheaths on this effect, density functional theory (DFT) was used to calculate the binding energy of the inserted Zn2+ with the V2O5 host. Our calculation model was established using the V2O5·1.5H2O molecular formula. According to the aforementioned XPS quantitative analysis, the bulk phase of V2O5@PEDODT was composed of 83.3% V5+, while the surface V4+ coated with PEDOT molecules was 16.7% (Fig. 6a). It is noteworthy that V4+ in the host V2O5 are more likely to combine with O from PEDOT to form the most stable V–O bonding (Fig. 6b). In this case, the binding energy of Zn2+ with V2O5@PEDOT is markedly reduced to −0.07 eV as compared with −2.29 eV of the pristine V2O5 nanowires electrode (Fig. 6c). Such weakened electrostatic interactions between Zn2+ and the V2O5 host suggest that most of the inserted Zn2+ can be reversibly extracted from V2O5@PEDOT. On the other hand, the impact of the PEDOT coating on the vanadium dissolution was also investigated by placing the active electrodes in 2 M aqueous Zn(CF3SO3)2 electrolyte for different days. The dissolved vanadium in the electrolyte after 12 days is determined to be 9.2 mg L−1 for the V2O5@PEDOT-50m electrode, which is one-half lower than 17.7 mg L−1 of the V2O5 nanowires electrode (Fig. 6d). The remarkably reduced binding energy and the suppressed vanadium dissolution contribute to the enhanced cycling stability of V2O5@PEDOT-50m, as discussed in Fig. 5e and S8.†
The CV curves of V2O5@PEDOT-50m recorded at 0.1 mV s−1 in the potential range of 0.2–1.8 V is presented in Fig. 7a. For the V2O5@PEDOT-50m electrode, three prominent oxidation peaks at around 0.75, 1.05, 1.40 V and the corresponding reduction peaks at 0.46 and 0.78 V are evident, which can be ascribed to the redox couples of V5+/V4+ and V4+/V3+, respectively.46 As compared with the V2O5 nanowires electrode, the oxidation peaks of the V2O5@PEDOT-50 m electrode slightly shift towards high potential and their reduction peaks shift to low potential. Such polarization phenomenon is induced by the slow ion diffusion (discussed below). This phenomenon becomes more evident at higher scan rates (Fig. 7b), where an apparent ion concentration gradient would limit the redox reactions near the electrode.7
To understand the Zn2+ storage behaviors, the electrochemical kinetics were analyzed using the method developed by Dunn et al.47
i = avb |
In this equation, the adjustable coefficient parameter b can be utilized to distinguish the pseudocapacitive or diffusion-controlled electrochemical behavior.23Fig. 7c plots the CV curves of the V2O5@PEDOT-50m electrode of various scan rates. The linear fitting of log(i) vs. log(v) yields the b values of 0.53, 0.52, 0.59 and 0.55 for the four peaks, respectively. These values are close to 0.5 and suggest that the diffusion-controlled behaviors dominate the insertion/extraction of the Zn2+ ion into/from the V2O5@PEDOT-50m nanowires. Whereas, the V2O5 nanowires electrode presents a higher b value (Fig. S9†) under the identical condition. This comparison means that the protective PEDOT sheaths would decrease the Zn2+ ion kinetics by slowing the ion diffusion into the V2O5@PEDOT-50m electrode.
In order to quantitatively analyze the current contribution from the pseudocapacitive and diffusion-controlled electrochemical process, the following equation is adopted:48,49
i = k1v + k2v1/2 | (1) |
The structure evolution of V2O5@PEDOT-50m at different charge and discharge stages was investigated by ex situ XRD (Fig. 8a–c). In the discharging process (Fig. 8b), the diffraction peaks of V2O5@PEDOT-50m gradually weaken, while new diffraction peaks at 6.6°, 12.2°, 20.5°, 28.5° and 33.6° gradually occur, which indicate the formation of ZnxV2O5·nH2O due to Zn2+ and H2O co-insertion.28 This observation suggests that V2O5 in the V2O5@PEDOT-50m undergoes structural evolution and transforms into the layered ZnxV2O5·nH2O with the increase of the interlayer spacing from 4.4 to 13.4 Å upon discharging to 0.2 V.52 In the following charging process (Fig. 8c), Zn2+ ions are extracted from the ZnxV2O5·nH2O and the characteristic peaks of V2O5@PEDOT-50m re-appear at the full charge state of 1.8 V. It is noteworthy that from the second cycles on, the insertion and extraction of Zn2+ and H2O proceed concurrently in the ZnxV2O5·nH2O and V2O5@PEDOT-50m electrode because the two phases co-exist when discharged to 1.0 V or charged to 1.8 V (Fig. S10†). Moreover, the similar XRD patterns in the discharge and charge processes of the first and second cycles also confirm the highly reversible Zn2+ insertion and extraction. The excellent electrochemical reversibility was also probed by ex situ XPS. The survey XPS spectrum of the V2O5@PEDOT-50m in the first cycle at a discharge state of 0.2 V displays weak Zn 3s (139.9 eV) and sharp Zn Auger (476.2 and 499.8 eV) signals (Fig. 8d). However, these peaks become very weak when fully charged to 1.8 V. Such change means that the markedly reduced binding energy of Zn2+ with the V2O5@PEDOT host (Fig. 6c) greatly facilitates the extraction of Zn2+ ions from the cathode. This contrast is more evident by comparing the Zn 2p XPS spectra of the V2O5@PEDOT-50m electrode (Fig. 8e). As compared with the pristine electrode, the Zn 2p peaks at the first cycle are very strong when discharged to 0.2 V and become very weak when charged to 1.8 V, suggesting the reversible Zn2+ insertion and extraction process. The insertion/extraction of Zn2+ results in the valence change of V in the cathode (Fig. 8f). When discharged to 0.2 V, the intensity of V4+ (516.4 eV of V 2p3/2) significantly increases and V3+ (515.4 eV of V 2p3/2) apparently appears. A quantitative XPS analysis reveals that the proportion of V4+ increases up to 57.7% and V3+ increases to 12.8%, in contrast to the 16.5% value of V4+ in the pristine electrode, and confirms that the reduction of V5+ is due to the insertion of Zn2+. In the full charge state of 1.8 V, the XPS results of V return to that of the pristine sample. The above results demonstrate the highly reversible electrochemical Zn2+/H2O insertion and extraction process into and from the electrode.
Footnotes |
† Electronic supplementary information (ESI) available: Schematic diagram showing the preparation of V2O5@PEDOT nanowires, SEM images of V2O5 and the V2O5@PEDOT aerogel, elemental composition, GCD curves of the V2O5 nanowire and PEDOT nanowire at 0.1 and 0.5 A g−1, EIS spectra, GCD curves of the V2O5 nanowire at various current densities, cycling performance of V2O5@PEDOT, performance comparisons of the V2O5@PEDOT-50m nanowires electrode with previous cathodes, CV curves of the V2O5 nanowires and the corresponding curves of log(i) vs. log(ν), ex situ XRD patterns of the V2O5@PEDOT-50m electrode in the second cycle. See DOI: https://doi.org/10.1039/d4ta01136h |
‡ Ting Yang and Diheng Xin contributed equally to this work. |
§ Prof. Zhibin Lei, School of Materials Science and Engineering, Shaanxi Normal University, 199 South Chang'an Road, Xi'an, Shaanxi, 710062, China. |
This journal is © The Royal Society of Chemistry 2024 |