Research Article

www.advmat.de

An Ultrafast, Durable, and High-Loading Polymer Anode

for Aqueous Zinc-Ion Batteries and Supercapacitors
Zhixiao Xu, Matthew Li, Wenyuan Sun, Tian Tang, Jun Lu,* and Xiaolei Wang*

activated carbon for ZICs with seem-

Zn metal has shown promise as an anode material for grid-level energy ingly high energy densities (based on the
storage, yet is challenged by dendritic growth and low Coulombic efficiency. cathodes).[1,3,4] However, in most of these
Herein, an ultrafast, stable, and high-loading polymer anode for aqueous cases, excessive Zn anodes were used to
Zn-ion batteries and capacitors (ZIBs and ZICs) is developed by engineering compensate for the uncontrollable den-
dritic growth, corrosion behavior, and
both the electrode and electrolyte. The anode polymer is rationally prepared low Coulombic efficiency (CE) of the Zn
to have a suitable electronic structure and a large π-conjugated structure, metal.[2,4–6] As such, most of the ZIBs
whereas the electrolyte is manufactured based on the superiority of triflate and ZICs developed using these electrode
anions over sulfate anions, as analyzed and confirmed via experiments and materials exhibited actual device energy
simulations. This dual engineering results in an optimal polymer anode and power densities that were signifi-
cantly lower than the corresponding theo-
with a low discharge potential, near-theoretical capacity, ultrahigh-loading
retical values.[7,8] Another key performance
capability (≈50 mg cm−2), ultrafast rate (100 A g−1), and ultralong lifespan (one metric that needs to be improved in ZIBs
million cycles). Its mechanism involves reversible Zn2+/proton co-storage at is their high-rate charge and discharge. In
the carbonyl site. When the polymer anode is coupled with cathodes for both addition to the dendrite-derived problems
ZIB and ZIC applications, the devices demonstrate ultrahigh power densities commonly experienced in batteries based
and ultralong lifespans, far surpassing those of corresponding Zn-metal- on other metal anodes, ZIBs are disad-
vantaged by hydrogen gas evolution from
based devices. water decomposition especially under
high-rate charge/discharge, leading to
poor cycle life.[2,4]
1. Introduction Based on the history of lithium-ion batteries,[9] it is of great
value to develop Zn-metal-free anodes for rocking-chair ZIBs,
With high theoretical capacities (820 mAh g−1 and which, however, remains a challenging objective.[10] An ideal
5855 mAh cm−3), suitable electrochemical potential (−0.763 V Zn-free anode should possess a low potential vs Zn2+/Zn, high
vs standard hydrogen electrode), low cost, and rich abun- theoretical capacity, good electronic conductivity, robust struc-
dance of component materials, aqueous zinc-ion batteries ture, and electrochemical stability.[10] One material that was
(ZIBs) and capacitors (ZICs) have exhibited great promise for explored for its possible use as an anode for ZIBs is Chevrel-
grid-level energy storage and other stationary applications.[1–3] phase Mo6S8, which has demonstrated a discharge voltage of
The past years have witnessed rapid progress in the develop- 0.36 V vs Zn2+/Zn; however, its low capacity and short life
ment of cathode materials, including Mn- and V-based oxides, have limited its application.[11,12] Since then, better ZIB anode
Prussian blue analogs, and organic materials for ZIBs and materials, such as copper selenide (Cu2−xSe),[13] sodium-inter-
calated TiS2 (Na0.14TiS2),[14] and hexagonal molybdenum oxide
Z. Xu, X. Wang
(h-MoO3),[15] with either lower discharge voltages (e.g., 0.3 V vs
Department of Chemical and Materials Engineering Zn2+/Zn) or longer cycle life (e.g., 30 000 cycles), have been dis-
University of Alberta covered. Possible anode materials are also not limited to inor-
9211-116 Street NW., Edmonton, Alberta T6G 1H9, Canada ganic substances; organic materials, particularly carbonyl-rich
E-mail: [email protected] molecules and polymers, have also demonstrated great poten-
M. Li, J. Lu tial as anodes[16,17] because of their low water solubility, high
Chemical Sciences and Engineering Division
Argonne National Laboratory electrochemical stability, oxygen-elimination capability,[18] pas-
9700 Cass Ave, Lemont, IL 60439, USA sivation toward hydrogen evolution reactions,[19] and low redox
E-mail: [email protected] potentials relative to the Zn redox potential.[17,20–22] For example,
W. Sun, T. Tang Lu et al. devised a carbonyl-rich dianhydride molecule with fast
Department of Mechanical Engineering kinetics by enhancing its π–π stacking interaction; however, the
University of Alberta
9211-116 Street NW., Edmonton, Alberta T6G 1H9, Canada
molecule exhibited poor cycling stability.[23] On the other hand,
Feng et al. constructed a 2D-imide-based covalent organic
The ORCID identification number(s) for the author(s) of this article
can be found under https://doi.org/10.1002/adma.202200077. framework as a rapid Zn2+storage anode and revealed a two-step
Zn2+storage mechanism involving a reversible carbonyl-enolate
DOI: 10.1002/adma.202200077 group transition.[24] However, although using a Zn-free anode

Adv. Mater. 2022, 34, 2200077 2200077 (1 of 11) © 2022 Wiley-VCH GmbH
 15214095, 2022, 23, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adma.202200077 by Tohoku University, Wiley Online Library on [07/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
www.advancedsciencenews.com www.advmat.de

is helpful at raising the potential and inhibiting hydrogen evo- among these substances, PI-1 and PI-2 exhibit the greatest
lution, there remains a large gap in achieving fast and stable promise as anode materials for ZIBs.
charge/discharge in Zn-based energy-storage systems. To verify the DFT results, the aforementioned organic mate-
In this study, we searched for an anode/electrolyte configu- rials were synthesized and characterized. To increase the elec-
ration tailored specifically toward achieving fast and stable trical conductivity, CNTs were added during the polymerization
cycling in ZIBs and ZICs. With the guidance of theoretical cal- process to produce PI-1/CNT, PI-2/CNT, and PI-3/CNT. The
culations, a series of molecules and polymers with and without successful preparation of these organic materials was con-
carbon nanotubes (CNTs) were synthesized and used as ZIB firmed by the Fourier-transform infrared (FTIR) spectra, in
anodes. Among these materials, a polymer/CNT hybrid based which peaks corresponding to carbonyl and imide groups were
on perylene 3,4,9,10-tetracarboxylic dianhydride demonstrated clearly detected (Figure 1C and Figure S1, Supporting Informa-
the best performance. Electrolytes containing different zinc tion). The X-ray diffraction (XRD) patterns reveal highly crys-
salts in a variety of concentrations were then studied experi- talline structures with negligible changes after compositing
mentally and via simulation, revealing that the incorporation with CNTs (Figure 1D and Figures S2, Supporting Information)
of triflate anions, mild acidity, high ionic conductivity, and low due to the ultralow content of the CNTs (2 wt%). This was fur-
viscosity can produce sufficiently high transport properties ther confirmed by thermogravimetric analysis (TGA) curves
that enable fast and stable ions storage. The optimized combi- (Figure 1E), which also revealed the excellent thermal stability
nation will enable the polymer anode to have a low discharge of the PI, as evidenced by negligible weight loss at tempera-
voltage (0.2–0.4 V vs Zn2+/Zn), approaching-theoretical capacity tures lower than 500 °C in air. Scanning electron microscopy
(128 mAh g−1 at 50 mA g−1), high CE (≈100%), ultrahigh loading (SEM) and transmission electron microscopy (TEM) images
(≈50 mg cm−2), fast charging capability (100 A g−1), and extraor- show the different morphologies of these organic materials.
dinary stability for up to one million cycles. The storage mecha- Whereas the molecules exhibit nanorod-like microstructures,
nism was uncovered via electrochemical measurements and the polymers are clearly composed of nanosheets (Figure S3,
in/ex situ characterizations, revealing a mechanism involving Supporting Information). For example, PI-1 exists in the form
reversible Zn2+ (de-)coordination and proton co-interaction at of ultrathin nanosheets, and when CNTs are introduced to
the carbonyl group in the polymer. To demonstrate its practical create PI-1/CNT, the combination generates 3D architectures
use, the polymer/CNT anode was coupled not only with bat- characterized by tiny nanosheets wrapped around the CNTs
tery cathodes for ZIBs, which manifested high power densities (Figure 1F–H). By contrast, in PI-2 and PI-3, the nanosheets are
(9.1 kW kg−1) and long life (50 000 cycles), but also with super- assembled into flower-like structures (Figure S3, Supporting
capacitor cathodes for ZICs, resulting in record-high stabilities Information). These micro-/nanostructures benefit the expo-
for over sub-million cycles at 10 and 200 A g−1, which are far sure of redox-active sites and the transportation of electrons
better than those of corresponding devices based on Zn-metal and ions during battery charge/discharge.[25] The polymer/CNT
anodes, even in excessive amounts. hybrid (herein, PI-1/CNT is used as an example) was further
observed via high-resolution TEM (HRTEM), which revealed
that the interface between the PI-1 and CNTs is characterized by
2. Results and Discussion crystalline CNT cores coated by PI-1 shells with thicknesses of
20–100 nm (Figure S4, Supporting Information). Furthermore,
2.1. Anode Material Design and Selection high-angle annular dark-field (HAADF) scanning TEM (STEM)
and the corresponding elemental mapping showed a homoge-
The electronic structures of organic materials are associated nous distribution of C, N, and O elements throughout the PI-1/
with their electrochemical behaviors in batteries. As such, den- CNT hybrid (Figure 1I).
sity functional theory (DFT) calculations were first conducted Subsequently, these organic materials were used as electrodes
to investigate the highest occupied molecular orbital (HOMO) for ZIB tests, with Zn as the counter/reference electrode and
and lowest unoccupied molecular orbital (LUMO) levels and 2.0 m ZnSO4 as the electrolyte. As shown in Figure 1B, although
bandgaps of several carbonyl-rich molecules (perylenetetra- the molecules demonstrated high initial capacities (155 mAh g−1
carboxylic dianhydride or PTCDA, and perylenetetracarboxylic for PTCDA and 122 mAh g−1 for PTCDI), these capacities
diimide or PTCDI) and polymers (polyimides: PI-1, PI-2, and decayed rapidly to 30 mAh g−1 for PTCDA and 60 mAh g−1 for
PI-3), which are either commercially available or easy to manu- PTCDI after 40 cycles. By contrast, the PTCDA-derived polymer,
facture. As shown in Figure 1A, the LUMO levels and bandgaps PI-1, exhibited a much higher stability (>90 mAh g−1 for over
for PTCDA, PTCDI, PI-1, PI-2, and PI-3 are (−4.29 eV, 2.51 eV), 100 cycles), which can be explained by the lower solubility of
(−3.97 eV, 2.49 eV), (−3.88 eV, 2.48 eV), (−3.86, 3.57 eV), and the polymer than those of the molecules (Figures S5–S7, Sup-
(−3.58 eV, 4.43 eV), respectively. In general, a higher LUMO porting Information). A comparison between the different
level leads to a lower discharge potential, whereas a narrower polymers clearly showed that the performance trend follows
bandgap favors higher electronic conductivity. Thus, according PI-1 > PI-2 > PI-3, indicating that a larger π-conjugation struc-
to this principle, the discharge potentials for these organic ture benefits the stability of the ZIB. Although PI-2 and PI-3
materials should follow this order: PI-3 < PI-2 < PI-1 < PTCDI have lower discharge voltages, their wider bandgaps and lower
< PTCDA. Meanwhile, their conductivities should obey this π-conjugations suggest lower electrical conductivities, leading
order: PI-1 > PTCDI > PTCDA > PI-2 > PI-3. Although PI-3 has to fast capacity decay and large voltage hysteresis (Figures S8
the lowest discharge potential, its very wide bandgap and low and S9, Supporting Information). Evidently, among these mole-
conductivity make it unsuitable for Zn-ion storage. Therefore, cules and polymers, PI-1 demonstrated the highest intrinsic

Adv. Mater. 2022, 34, 2200077 2200077 (2 of 11) © 2022 Wiley-VCH GmbH
 15214095, 2022, 23, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adma.202200077 by Tohoku University, Wiley Online Library on [07/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
www.advancedsciencenews.com www.advmat.de

Figure 1. Electrode engineering and structural characterization of PI-1 and PI-1/CNT. A) Electronic structure of different molecules and polymers, B)
cycling performance of different organic electrodes at 0.1 A g−1, C) FTIR spectra, D) XRD patterns, and E) TGA profiles of PI-1/CNT and PI-1. F) SEM
image, G,H) TEM images, and I) HAADF image and elemental mapping images of PI-1/CNT.

redox activity, which matched well with the DFT results. More- to optimize the PI-1/CNT-based ZIBs by investigating the influ-
over, the performance of PI-1 could be further enhanced after ence of zinc salt and concentration on the battery performance.
CNT coupling into PI-1/CNT; by comparison, the pristine CNT For this experiment, we formulated 0.5 m ZnSO4, 1.0 m ZnSO4,
exhibited negligible capacity and poor reversibility (Figure 1B 2.0 m ZnSO4, 1.0 m Zn(OTf)2, 1.5 m Zn(OTf)2, and 2.0 m
and Figure S10, Supporting Information). Therefore, the Zn(OTf)2 in water.
resultant PI-1/CNT composite material is an excellent candi- First, the rate performance of PI-1/CNT in different elec-
date as an anode for fast and stable storage of Zn2+ ions. trolytes was examined. As shown in Figure 2A, as the cur-
rent density was increased from 0.2 to 5 A g−1, the capacity
in aqueous Zn(OTf)2 decreased more slowly compared with
2.2. Electrolyte Engineering and Molecular Dynamics Simulation that in ZnSO4, suggesting the superiority of Zn(OTf)2 over
ZnSO4 as an electrolyte salt for fast charging (Figure S11a–c,
The electrolyte composition has a crucial impact on Zn-based Supporting Information). To demonstrate this better, a 1.0 m
energy-storage devices.[26] Of critical importance is the specific electrolyte is used for the following example (Figure 2B). At
pairing between the electrode and electrolyte. Here, we aimed the low rate, for example, 0.2 A g−1, the PI-1/CNTs in the two

Adv. Mater. 2022, 34, 2200077 2200077 (3 of 11) © 2022 Wiley-VCH GmbH
 15214095, 2022, 23, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adma.202200077 by Tohoku University, Wiley Online Library on [07/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
www.advancedsciencenews.com www.advmat.de

Figure 2. Electrolyte engineering and MD simulation. A) Capacity values of PI-1/CNT in different electrolyte at current densities from 0.2 to 5 A g−1.
B) Voltage–capacity profiles of PI-1/CNT in 1.0 m Zn(OTf)2 and 1.0 m ZnSO4 at low (0.2 A g−1) and high (5.0 A g−1) current rates. C) EIS of PI-1/CNT in
1.0 m Zn(OTf)2 and 1.0 m ZnSO4. D) Capacity and capacity retention of PI-1/CNT in different electrolytes after ultralong cycling. E) Cycling performance
and CE profiles of PI-1/CNT in 1.0 m Zn(OTf)2 and 1.0 m ZnSO4 tested at 5.0 A g−1. F) Ionic conductivities, viscosities, and pH values of different
electrolytes. G) MD-simulated distribution of ions and water near the cathode surface. H) Number density profiles of water and Zn2+ along the normal
direction of the cathode surface.

electrolytes exhibited comparable capacities (115 mAh g−1 for respectively. When Zn(OTf)2 was used, a high CE of ≈100%
ZnSO4 and 112 mAh g−1 for Zn(OTf)2). By contrast, at the high was maintained during long cycling, whereas when ZnSO4
rate, for example, 5.0 A g−1, a high capacity of 100 mAh g−1 was was used, later cycles exhibited unstable CE values, indicating
attained using Zn(OTf)2, surpassing that attained using ZnSO4 the relatively poor reversibility of the electrolyte (Figure 2E).
(91 mAh g−1). Furthermore, Figure 2B shows that the voltage Among all electrolytes examined in this study (Figure 2F), 1.5 m
drops for Zn(OTf)2 were smaller than those for the ZnSO4 Zn(OTf)2 affords the highest ionic conductivity (68.5 mS cm−1),
electrolyte, suggesting that the ionic conductivity of Zn(OTf)2 a low viscosity (2.32 mPa s), and a mild acidity (pH = 4.8). These
is higher. This inference was supported by the lower charge properties contribute to this electrolyte exhibiting the highest
transfer resistance of Zn(OTf)2 (50 Ω) than that of ZnSO4 performance, with the fastest rate capability (101 mAh g−1
(160 Ω), as revealed by the electrochemical impedance spectra at 5.0 A g−1) and highest stability (88% capacity retention and
(EIS, Figure 2C). 89 mAh g−1 capacity after 50k cycles), among all electrolytes
Afterward, the cycling performance of PI-1/CNT in dif- examined herein.
ferent electrolytes was evaluated. As shown in Figure 2D and To gain insight into the superiority of triflate (OTf−) over
Figure S10, Supporting Information, Zn(OTf)2 electrolytes ena- sulfate (SO42−) anions, molecular dynamics (MD) simula-
bled high capacity retentions of more than 80% after 50 000 tions were performed. First, an electrical double-layer model
cycles at 5.0 A g−1, exceeding that produced using ZnSO4, which was built to simulate the dynamics of ions and water in two
enabled retentions of only 53–58% in the same conditions, electrolytes (1.0 m Zn(OTf)2 vs 1.0 m ZnSO4) and at different
again demonstrating that Zn(OTf)2 is the better electrolyte salt. potentials (Figure S12, Supporting Information).[27] Because
Figure 2E shows that the resulting capacities for 1.0 m Zn(OTf)2 of the presence of the hydrophobic CF3 group in OTf−, the
and 1.0 m ZnSO4 after 50k cycles were 85.5 and 51.7 mAh g−1, hydrophobicity of the triflate anion is exactly opposite to that

Adv. Mater. 2022, 34, 2200077 2200077 (4 of 11) © 2022 Wiley-VCH GmbH
 15214095, 2022, 23, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adma.202200077 by Tohoku University, Wiley Online Library on [07/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
www.advancedsciencenews.com www.advmat.de

of the sulfate anion (Figure S13, Supporting Information). electrode materials, including Mn-based[32,33] and V-based[34,35]
This hydrophobicity disparity between SO42− and OTf− led to oxides, Prussian blue analogs,[36,37] and organics,[38–40] among
distinctly different electrochemical double-layer structures at others (Figure 3D and Table S1, Supporting Information).[41,42]
the electrode surfaces (defined as the layer within 0–0.75 nm High-loading electrodes are of great importance for lowering
to the electrode surface), which determined the distribution of production costs and increasing energy densities at the cell
ions and water near the electrode surface (Figure 2G). At an level.[43–45] Accordingly, we further tested PI-1/CNT electrodes
applied potential of U = 0.5 V, the hydrophobic triflate anions under high loadings (4.08–50 mg cm−2) using a constant amount
created an H2O-scarce environment at the inner Helmholtz of electrolyte (100 µL). As shown in Figure 3E, at a loading of
layer (IHL, defined as the layer within 0–0.25 nm to the elec- 4.08 mg cm−2, the specific capacities were 104, 102, 99, 97, 94,
trode surface, as shown in Figure S14, Supporting Informa- and 85 mAh g−1 at current densities of 0.2, 0.5, 1.0, 2.0, and
tion), in contrast to the H2O-rich environment generated by 5.0 A g−1, respectively. Furthermore, the capacity at 5.0 A g−1
the hydrophilic sulfate anions. At an increased potential of was maintained at ≈80 mAh g−1 (94.1% retention) after 30 000
U = 1.1 V, the triflate anions were further enriched in the IHL cycles, corresponding to a low capacity loss of 0.0167% per cycle.
of the electrode, which also decreased the amount of H2O and More impressively, without the binder and current collector,
reduced any water-related reactions (Figures S14 and S15, Sup- free-standing electrodes with loadings of ≈12.5–50.0 mg cm−2
porting Information). By contrast, when the ZnSO4 electrolyte can be easily manufactured via direct pressing of PI-1/CNT
was used, the amount of H2O in the IHL exhibited only slight powders into tablets under high pressure (Figures S18–S22,
changes with respect to the increase in potential. Whereas most and Videos S1 and S2, Supporting Information). As shown in
of the Zn2+ ions were expelled from the double layers in the the inset of Figure 3F and Figure S20, Supporting Information,
ZnSO4 electrolyte, an abundant amount of Zn2+ ions remained a free-standing PI-1/CNT electrode of 19.7 mg cm−2 loading
in the Zn(OTf)2 electrolyte within a distance of 0.75 nm from has an area ≈1.27 cm2 and a thickness of 150 µm. As shown in
the cathode (Figure 2H). The density profiles of Zn2+ and H2O Figure 3F and Figures S18 and S19, Supporting Information,
indicated that more Zn2+ and fewer H2O molecules accumu- electrodes of 12.5 and 19.7 mg cm−2 loadings delivered excel-
lated near the electrode surface in the Zn(OTf)2 electrolyte than lent rate and cycling performance with ≈90% capacity reten-
in the ZnSO4 electrolyte (Figure 2G–H), leading to easier access tion after 500 cycles even under the lean electrolyte condition
to Zn2+ for redox reactions and fewer water-related reactions in (6.3 or 4.0 µL of electrolyte per milligram of PI-1/CNT, denoted
the Zn(OTf)2 electrolyte compared to those in ZnSO4. as E/P). More excitingly, under ultrahigh loadings of 30.1 and
50.0 mg cm−2 with lean electrolytes of 2.6 and 1.6 µL mg−1,
respectively, PI-1/CNT-based ZIBs can still run for over
2.3. Fast and Stable Cycling, High Loadings, 100 cycles with high capacity retentions (Figures S21 and S22,
and Kinetics Analysis Supporting Information). The capacity loss can be attributed to
the Zn-metal side with dendrite growth rather than to the PI-1/
Taken together, the results presented thus far indicate that the CNT side with an intact structure, as shown in the disassem-
ideal pairing of electrode material and electrolyte composi- bled cell after long cycling (Figures S23–S25, Supporting Infor-
tion should employ a PI-1/CNT electrode and 1.5 m Zn(OTf)2 mation). The areal capacities against mass loading at different
electrolyte. To demonstrate the efficacy of this combination, its rates were plotted in Figure 3G, which shows that, at current
rate performance is presented in Figure 3A, showing that at densities between 0.1 and 0.5 A g−1, the areal capacity increased
a low current density of 0.05 A g−1, a near-theoretical capacity linearly with respect to the mass loading up until 30.1 mg cm−2,
of 128 mAh g−1 can be delivered. When the current density indicating that, at these rates, no limitation was reached in
was increased to 0.2, 1.0, 5.0, 10, 30, and 50 A g−1, the specific the mass loading. Moreover, high areal capacities of 5.0 and
capacities were 113, 107, 100, 80, 65, and 59 mAh g−1, respec- 3.0 mAh cm−2 can be obtained using the 50.0 mg cm−2 electrode
tively. Surprisingly, even at an ultrahigh rate of 100 A g−1, corre- at areal current densities of 1.0 and 25 mA cm−2, respectively.
sponding to a charging time of 1.4 s, the capacity could still be By contrast, PI-1 electrodes with high loadings cannot achieve
maintained at 40 mAh g−1. At all current rates, the average dis- high rates and long cycles (Figures S26–S28, Supporting Infor-
charge voltages were maintained at ≈0.35 V vs Zn2+/Zn, which mation), suggesting the essential role of CNT in making high-
is promising for the application of PI-1/CNT as a Zn-metal-free loading electrodes with application potentials.[45] These results
anode. In addition to this exceptional rate performance, the sta- demonstrate the fast rate, long life, and high loading capability
bility was also extraordinary, with negligible capacity losses of of the PI-1/CNT electrode, showing its promise as a future
0.1 A g−1 over 250 cycles (Figure S16, Supporting Information), anode for practical use in full power cells.
of 10 A g−1 over 30 000 cycles (Figure 3B), and of 50 A g−1 over To comprehend the fast kinetics in the PI-1/CNT electrode,
60 000 cycles (Figure S17, Supporting Information), whereas a cyclic voltammetry (CV) tests were conducted, with rates ranging
75% capacity was retained after one million cycles at 100 A g−1 from 0.3 to 2.0 mV s−1 (Figure 3F). In all the curves, there are
(Figure 3C). The performance presented here is one of the two pairs of redox peaks, indicating a two-step ion storage pro-
longest cycle life observed in aqueous batteries and is compa- cess.[24] As the scan rate was increased, there were no signifi-
rable to that of state-of-the-art fast-charging batteries, posing a cant changes in the peak shapes and polarization potentials
serious technological contender.[28–31] To the best of our knowl- between redox peaks, implying high reversibility in the redox
edge, this simultaneous achievement of ultrafast and ultralong reactions. To quantitatively study the ion storage kinetics, the
cycling has hardly been reported for Zn-based energy devices equation i = aνb, in which i and ν represent the current density
and is better than those reported thus far for almost all ZIB and scan rate, respectively (a and b are constants, where b = 0.5

Adv. Mater. 2022, 34, 2200077 2200077 (5 of 11) © 2022 Wiley-VCH GmbH
 15214095, 2022, 23, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adma.202200077 by Tohoku University, Wiley Online Library on [07/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
www.advancedsciencenews.com www.advmat.de

Figure 3. Optimized electrochemical performance and kinetics analysis of PI-1/CNT electrode. A) Voltage–capacity profiles of PI-1/CNT at different
current densities from 0.2 to 100 A g−1. B) Ultralong cycle life of PI-1/CNT at different current densities (10 or 50 A g−1). C) Million-cycle life of PI-1/
CNT at 100 A g−1. D) Comparison of rate and cycling performance between different ZIB electrode materials. E–G) High-loading PI-1/CNT electrodes:
E) rate and cycling performance of PI-1/CNT with high loading of 4.08 mg cm−2; F) cycling performance of ultrahigh-loading PI-1/CNT electrodes; the
inset of (F) shows the optical image of the 19.7 mg cm−2 electrode; and G) plots of areal capacity against mass loadings. H) CV curves of PI-1/CNT at
different scan rates from 0.3 to 2 mV s−1 and I) corresponding kinetics profiles of the different redox peaks A, B, and C.

indicates a diffusion-controlled process, while b = 1.0 suggests scan rates. As shown in Figure S29, Supporting Information,
a pseudocapacitive procedure), was employed. To determine the at rates of 0.3, 0.4, 0.5, 0.75, 1.0, and 2.0 mV s−1, the capaci-
b values, the curves of log(i) vs log(ν) were plotted and linearly tive contributions were 80%, 82%, 84%, 86%, 90%, and 94%,
fitted to the redox peaks. The fitting slopes, which were equal respectively, further verifying the dominance of the pseudoca-
to the b value, for the A, B, and C peaks were 0.84, 0.95, and pacitive process in the PI-1/CNT electrode, which could explain
0.88, respectively. These approaching-1.0 values indicate that its exceptional rate and cycling performance.
the PI-1/CNT electrode mainly follows a pseudocapacitive pro-
cess instead of solid-state diffusion for the Zn2+ ion storage.
To further quantify the contributions from the pseudocapaci- 2.4. Storage Mechanism Analysis
tive and diffusive processes, the current was divided into two
parts in accordance with i = k1ν + k2ν0.5. When the k1 and k2 To understand the charge-storage mechanism of PI-1/CNT in
values are determined, the percentages of current controlled by the ZIB charge/discharge process, CV tests were performed in
pseudocapacitance and diffusion can be calculated for different different electrolytes. Figure 4A shows the CV profiles of the

Adv. Mater. 2022, 34, 2200077 2200077 (6 of 11) © 2022 Wiley-VCH GmbH
 15214095, 2022, 23, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adma.202200077 by Tohoku University, Wiley Online Library on [07/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
www.advancedsciencenews.com www.advmat.de

Figure 4. Storage mechanism of PI-1/CNT electrode. A) CV profiles of PI-1/CNT in different Zn/Na-based electrolytes. B) SEM mapping images of
PI-1/CNT electrode discharged to 0.1 V in 0.5 m Zn(OTf)2–DMF. C) CV profiles of PI-1/CNT in different aqueous electrolytes with controlled pH values.
D) SEM images of PI-1/CNT electrode discharged to 0.1 V (upper part) and charged back to 1.1 V (lower part). E) Ex situ XRD patterns of PI-1/CNT
electrode at pristine and discharged status. F) SEM image and corresponding elemental mapping images of PI-1/CNT electrode discharged to 0.1 V
in 1.5 m Zn(OTf)2. G,H) In situ ATR-FTIR spectra of PI-1/CNT at different discharge/charge states. I,J) Ex situ high-resolution C1s (I) and O1s (J) XPS
spectra of PI-1/CNT electrode at pristine, DC0.1V, and C1.1V statuses.

PI-1/CNT electrode in the separate electrolytes 0.5 m Zn(OTf)2– 0.5 m Zn(OTf)2–DMF and 0.5 m Zn(OTf)2–H2O electrolytes,
DMF, 0.5 m Zn(OTf)2–H2O, and 1.0 m NaOTf–H2O. The CV suggesting the insertion of Zn2+ ions in the PI-1/CNT elec-
profiles of 1.0 m NaOTf–H2O and 0.5 m Zn(OTf)2–H2O show trode. This inference was further confirmed by SEM mapping
apparently different redox peaks, implying the insertion of dif- images, which revealed a uniform distribution of Zn elements
ferent cations in the PI-1/CNT electrode (Zn2+ or Na+). The in the PI-1/CNT electrode after discharge to 0.1 V in the 0.5 m
positive shift in the reduction peak for 0.5 m Zn(OTf)2–H2O Zn(OTf)2–DMF electrolyte (Figure 4B). On the other hand, rela-
relative to that for 1.0 m NaOTf–H2O could be explained by the tive to the profile of the 0.5 m Zn(OTf)2–DMF electrolyte, there
larger charge density of Zn2+ than that of Na+.[46] By contrast, exists another oxidation peak at 0.4 V for the 0.5 m Zn(OTf)2–H2O
identical redox peaks at 0.7 and 0.25 V were detected for the electrolyte, which may have resulted from proton insertion.

Adv. Mater. 2022, 34, 2200077 2200077 (7 of 11) © 2022 Wiley-VCH GmbH
 15214095, 2022, 23, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adma.202200077 by Tohoku University, Wiley Online Library on [07/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
www.advancedsciencenews.com www.advmat.de

To confirm whether protons do interact with the PI-1/CNT was charged back to 1.1 V, stronger CO but weaker CO
electrode, CV tests were conducted in three electrolytes with peaks were detected for C1.1V compared with those for DC0.1V,
controlled pH values: ZnSO4 (pH = 5), H2SO4 (pH = 5), and suggesting the de-coordination process from Zn2+(CO−)2
H2SO4 (pH = 1). It was determined that there exists an oxida- (or COH) to Zn2+/H+ and CO. This was further confirmed
tion platform at ≈0.45 V for all electrolytes, implying its assign- by the high-resolution O1s XPS spectra (Figure 4J), in which
ment to proton insertion. Relative to the profile of H2SO4 the CO/CO bonding increased after discharge but declined
(pH = 5), that of H2SO4 (pH = 1) shows one more oxidation peak after charging. These results imply that the carbonyl group in
at ≈0.8 V, which can be ascribed to Zn2+ insertion due to Zn- PI-1 is the active center for the reversible storage of Zn2+ ions
metal dissolution in the strongly acidic electrolyte (Figure S30, and protons.
Supporting Information). The insertion of protons into PI-1/
CNT was further confirmed by ex situ SEM and XRD char-
acterizations. The SEM images of the discharged electrode 2.5. Application of PI-1/CNT as Zn-Free Anode for ZIBs and ZICs
show numerous nanoflakes covering the surface of PI-1/CNT
(Figure 4D). Based on XRD patterns (Figure 4E) and SEM map- Now that the exceptional performance of our designed anode
ping images (Figure 4F), these nanoflakes were determined to system has been confirmed, we further validated its practical
be zinc sulfate hydroxide. Because zinc sulfate hydroxide pre- use as a Zn-free anode in aqueous Zn-based energy-storage
cipitates in solutions with pH > 5.5, its formation in an elec- devices, including batteries and supercapacitors. To construct a
trolyte with a pH of ≈5 should be due to proton insertion into rocking-chair-type ZIB, commercial MnO2 was first discharged
the polymer electrode, leaving OH− behind and causing a local to obtain ZnxMnO2, which was then used as the cathode.
pH increase.[47] After the polymer electrode was charged back Thereafter, this cathode was combined with PI-1/CNT anode,
to 1.1 V, these nanoflakes disappeared, indicating a revers- and the mass ratio of negative to positive electrode (N/P) was
ible proton insertion/de-insertion mechanism (Figure 4D).[48] maintained at 1 (both 1.32 mg). For comparison, a Zn//MnO2
Overall, these results indicate that both Zn2+ and protons act battery was also assembled, where an electrodeposited Zn
as active cations for energy storage in ZIBs based on PI-1/CNT metal with a capacity of 23 mAh (mass of ≈28 mg) was used
electrodes. as the anode, corresponding to an N/P mass ratio of ≈23.3.
To further elucidate the active sites for storing Zn2+/pro- The voltage windows for discharging/charging were set to
tons in the PI-1/CNT electrode, in situ FTIR and ex situ X-ray 0–1.65 V for the PI-1/CNT//ZnxMnO2 battery and to 0.5–1.8 V
photoelectron spectroscopy (XPS) analyses were performed for the Zn//MnO2 battery. The voltage versus capacity profiles
on the PI-1/CNT electrode at different charge/discharge sta- of the two batteries at different current densities are shown in
tuses (Figure 4G). Holey coin cells and attenuated total reflec- Figure 5A and Figure S32, Supporting Information, where the
tion (ATR)-FTIR microscopy were employed to investigate the current densities and capacities are based on MnO2. At a low
structural changes of PI-1/CNT with the uptake/removal of current density of 0.1 A g−1, the PI-1/CNT-based battery was able
Zn ions (Figure S31, Supporting Information). PI-1/CNT was to exhibit a high capacity of 110 mAh g−1 and an average dis-
loaded onto a stainless-steel mesh with a 125-mesh size to charge potential of 0.91 V (Figure 5A). At current rates of 1, 2, 5,
create an electrode in which infrared light could pass through 10, and 20 A g−1, capacities of 97.6, 95, 88, 75, and 58 mAh g−1,
the hole and enable the IR spectra of PI-1/CNT to be obtained respectively, were achieved (Figure 5A,B). Even when the rate
during battery (dis)charging. The characteristic peaks for car- was increased 350 times to 35 A g−1 (charging time: 4 s), the PI-1/
bonyl (CO), imide (CNC), and CO groups were located CNT//ZnxMnO2 full cell still delivered a capacity of 39 mAh g−1,
at 1650–1700, ≈1350, and ≈1150 cm−1, respectively. For clear corresponding to a 35.5% retention of capacity at 0.1 A g−1,
comparison, the carbonyl group (CO) was normalized to 1. which indicated an excellent rate capability. By contrast, the
During discharge, the intensity ratio of the carbonyl group to Zn//MnO2 battery with the Zn-metal anode delivered only 120,
the imide group declined, whereas the corresponding ratio of 95.6, 64.4, and 39.8 mAh g−1 at rates of 1, 2, 5, and 10 A g−1,
CO to CO increased, suggesting a coordination reaction respectively (Figure S33, Supporting Information), suggesting
between the Zn ions/proton and carbonyl groups to generate that PI-1/CNT is an efficient anode for achieving high-rate and
ZnOC or COH groups (Figure 4H). During charging, high-power batteries. The capacity can be further increased by
the intensities of the aforementioned groups reversed back using a N/P mass ratio of ≈2.5 (Figure S34, Supporting Infor-
to their original states, with a decreased CO/CO ratio and mation). To further illustrate the exceptional performance of
increased carbonyl-to-imide ratio, indicating the removal of the our material, Ragone curves were plotted in Figure 5C based on
Zn ion/proton from PI-1 (Figure 4H). XPS was also conducted both anode and cathode materials. According to these data, the
to monitor the bonding changes in PI-1/CNT during the bat- PI-1/CNT//ZnxMnO2 battery exhibited a high energy density of
tery (dis)charge. As shown in the C1s XPS spectra (Figure 4I) 50.5 Wh kg−1 at 45.5 W kg−1 and was able to maintain 10.1 Wh kg−1
for the pristine, DC0.1V, and C1.1V electrodes, five peaks can be even at a high power density of 9100 W kg−1 (Figure 5C). The
deconvoluted at ≈284.8, ≈285.3, 285.7, 286.9, and 289.3 eV, cor- high power and energy densities of the PI-1/CNT//ZnxMnO2
responding to sp2 CC, sp3 CC, CN, CO, and CO bonds, battery are among the best documented for rocking-chair-type
respectively. Compared with the pristine electrode, DC0.1V ZIBs (Figure 5C and Tables S2 and S3, Supporting Informa-
exhibited less CO but more CO bands, indicating the reduc- tion).[13,14,49] Figure 5D shows the cycling performance and CE
tion of the carbonyl group to the CO group and its subse- values of the PI-1/CNT//ZnxMnO2 and Zn//MnO2 cells at a
quent coordination with Zn2+/H+, forming Zn2+(CO−)2 or high current density of 2.0 A g−1. Although a large excess of
COH during discharge. On the other hand, after the electrode Zn metal was used as the anode, the Zn//MnO2 battery could

Adv. Mater. 2022, 34, 2200077 2200077 (8 of 11) © 2022 Wiley-VCH GmbH
 15214095, 2022, 23, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adma.202200077 by Tohoku University, Wiley Online Library on [07/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
www.advancedsciencenews.com www.advmat.de

Figure 5. Application of PI-1/CNT as Zn-metal-free anode for ZIBs and ZICs. A) Galvanostatic charge and discharge potential profiles of PI-1/CNT//
ZnxMnO2 at different current rates. B) Rate performance of PI-1/CNT//ZnxMnO2 and corresponding CE values. C) Ragone plots of PI-1/CNT//ZnxMnO2
and documented rocking-chair-type ZIBs (energy and power densities are based on total active material mass of cathode and anode). D) Cycling
comparison between PI-1/CNT//ZnxMnO2 and Zn//MnO2 with equal MnO2 mass at current density of 2.0 A g−1. E) Long cycle performance of PI-1/
CNT//AC and Zn//AC at high current density of 10 A g−1. F) Ultrafast and stable cycling of PI-1/CNT//AC capacitor (N/P mass ratio = 1.15). G) Capacity
retention against cycle numbers for different materials in reported papers.

deliver only a 58% capacity retention after 6000 cycles, whereas Information). Whereas the photo of the former shows a color-
the PI-1/CNT anode enabled a high capacity retention of 83% less electrolyte, the photo of the latter shows that it was full
after 6000 cycles and 43% retention after 50 000 cycles with of yellow product, likely caused by Mn-ion dissolution. The
100% CE (Figure S35, Supporting Information). This stability post-cycling PI-1/CNT anode well maintained its crystal struc-
is expected to be further enhanced by the use of a better MnO2 ture and showed no sign of dendrites (Figure S37, Supporting
cathode. To comprehend the highly differing levels of perfor- Information), whereas the Zn anode was full of “dead” and
mance between the PI-1/CNT//ZnxMnO2 and Zn//MnO2 cells, dendritic Zn metal (Figure S38, Supporting Information). In
they were disassembled, and digital photos were obtained to consideration of practical use, Zn-rich ZnMn2O4 cathode[50]
show their contrasting conditions (Figure S36, Supporting was also prepared (Figure S40, Supporting Information) and

Adv. Mater. 2022, 34, 2200077 2200077 (9 of 11) © 2022 Wiley-VCH GmbH
 15214095, 2022, 23, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adma.202200077 by Tohoku University, Wiley Online Library on [07/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
www.advancedsciencenews.com www.advmat.de

coupled with the PI-1/CNT anode for battery tests in which the water solubility, high LUMO level, narrow bandgap, and large
PI-1/CNT anode also presents higher rate capability and longer π-conjugation, the anode material PI-1 exhibits high redox activity
cycling stability than Zn anode-based full cells (Figure S41, Sup- and stability towards Zn2+ storage. On the other hand, the electro-
porting Information). Impressively, a PI-1/CNT//ZnMn2O4/C lyte solution 1.5 m Zn(OTf)2 exhibits high ionic conductivity, mild
battery using free-standing and high-loading electrodes can acidity, low viscosity, and rich Zn2+ access. These sets of charac-
stably run for more than 4000 cycles (Figure S42, Supporting teristics of the anode material and electrolyte solution benefit
Information). These results clearly demonstrate the immense the fast charging and stable cycling of the cell. This combination
capability of this polymer anode at enabling high-power and enables a low discharge voltage, near-theoretical capacity, ultra-
long-life ZIBs. high loadings up to 50 mg cm−2, rapid kinetics of up to 100 A g−1,
To further demonstrate its application in zinc-ion hybrid and long lifespans of up to 1 000 000 cycles. The study revealed
supercapacitors, PI-1/CNT was first discharged to 0.1 V in a a reversible Zn2+/H+ co-interaction mechanism at the carbonyl
half-cell and extracted to serve as an anode and paired with com- site of PI-1. Furthermore, full cells based on the PI-1/CNT anode
mercial activated carbon (AC) as the cathode,[51] with an N/P delivered the highest power density in rocking-chair-type ZIBs
mass ratio of 0.67. For comparison, Zn was also electrodepos- and the longest cycle life in ZICs, surpassing those of corre-
ited onto a Cu foil to form a Zn anode with a capacity of 5 mAh sponding devices based on excessive Zn-metal anodes. Given
(≈6.1 mg). The galvanostatic charge/discharge was conducted in the excellent thermal properties of polyimide, this research will
voltage windows of 0.15–1.8 V for the Zn//AC supercapacitor and open a new avenue for the preparation of cheap, sustainable,
0–1.65 V for the PI-1/CNT//AC supercapacitor at different cur- fast, stable, and high-loading organic anodes for aqueous energy-
rent densities (Figures S43 and S44, Supporting Information), storage devices over a wider range of temperatures.[54]
where the current densities and capacitances were based on the
mass of the AC electrode. The discharge capacitances of PI-1/
CNT//AC were 124.8, 102.4, 97.6, 94.4, and 92.8 F g−1 at current Supporting Information
rates of 0.2, 1, 2, 5, and 10 A g−1, respectively, indicating excel-
Supporting Information is available from the Wiley Online Library or
lent rate performance. These capacitance values were higher
from the author.
than those of Zn//AC at high current densities (Figure S44,
Supporting Information). Figure 4E compares the cycling per-
formance of the two ZICs; at 10.0 A g−1, the PI-1/CNT//AC
capacitor exhibited remarkable stability, with 66% capacitance
Acknowledgements
retention after 100 000 cycles and 50% retention after half This work was financially supported by the Natural Sciences and
million cycles, and ≈100% CE. By stark contrast, the Zn//AC Engineering Research Council of Canada (NSERC), through the
capacitor first exhibited a dramatic decrease in capacitance after Discovery Grant Program (RGPIN-2018-06725) and the Discovery
Accelerator Supplement Grant program (RGPAS-2018-522651), and
only 60 000 cycles with a mere 32% retention and failed to work
by the New Frontiers in Research Fund-Exploration program (NFRFE-
after 107 000 cycles. Both capacitors were also disassembled to 2019-00488). Xiaolei Wang also acknowledges financial support from
investigate their different electrochemical behaviors (Figure S45, the University of Alberta and Future Energy Systems (FES-T06-Q03).
Supporting Information). Photographs show the light-yellow- The work from Argonne National Laboratory was supported by the U.S.
colored PI-1/CNT//AC capacitor in contrast to the dark-brown Department of Energy (DOE), Office of Energy Efficiency and Renewable
colored Zn//AC capacitor, indicating the occurrence of severe side Energy, Vehicle Technologies Office. Argonne National Laboratory is
operated for the U.S. DOE, Office of Science, by UChicago Argonne, LLC,
reactions, which caused the failure of the Zn//AC capacitor. This
under contract No. DE-AC02-06CH11357. The authors also acknowledge
failure was caused by dendrites formed in the Zn-metal anode the computing resources and technical support from Western Canada
(Figure S46, Supporting Information). By contrast, no dendrites Research Grid (WestGrid).
were detected in the PI-1/CNT anode even after 0.5 million cycles
(Figure S47, Supporting Information). Furthermore, another
PI-1/CNT//AC capacitor was assembled with an N/P ratio of Conflict of Interest
≈1.15, which enabled ultrafast charging/discharging (200 A g−1)
and ultrahigh stability (93% capacitance retention) over 0.3 mil- The authors declare no conflict of interest.
lion cycles (Figure 5F). Remarkably, the 200 A g−1 charging rate
and half-million cycles achieved using PI-1/CNT//AC are some
of the fastest and longest life cycles ever reported for Zn-based Data Availability Statement
supercapacitors (Figure 5G and Table S4, Supporting Informa- The data that support the findings of this study are available from the
tion).[3,52,53] These results confirm the advantages of metal-free corresponding author upon reasonable request.
PI-1/CNT as a practical anode material for ZICs.

Keywords
3. Conclusions fast charging/discharging, high-loading electrodes, polyimide, polymer
anodes, zinc-ion batteries/capacitors
We report a polymer anode with an unprecedented high rate and
cycling performance for use in aqueous Zn-ion batteries and Received: January 4, 2022
supercapacitors derived via the careful tailoring of the electrode Revised: March 1, 2022
material and its paired electrolyte composition. Because of its low Published online: April 29, 2022

Adv. Mater. 2022, 34, 2200077 2200077 (10 of 11) © 2022 Wiley-VCH GmbH
 15214095, 2022, 23, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adma.202200077 by Tohoku University, Wiley Online Library on [07/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
www.advancedsciencenews.com www.advmat.de

[1] X. Jia, C. Liu, Z. G. Neale, J. Yang, G. Cao, Chem. Rev. 2020, 120, [29] X. Wu, J. J. Hong, W. Shin, L. Ma, T. Liu, X. Bi, Y. Yuan, Y. Qi,
7795. T. W. Surta, W. Huang, J. Neuefeind, T. Wu, P. A. Greaney, J. Lu,
[2] F. Wan, X. Zhou, Y. Lu, Z. Niu, J. Chen, ACS Energy Lett. 2020, 5, X. Ji, Nat. Energy 2019, 4, 123.
3569. [30] X. Shen, T. Sun, L. Yang, A. Krasnoslobodtsev, R. Sabirianov,
[3] Z. Li, Y. An, S. Dong, C. Chen, L. Wu, Y. Sun, X. Zhang, Energy M. Sealy, W.-N. Mei, Z. Wu, L. Tan, Nat. Commun. 2021, 12, 820.
Storage Mater. 2020, 31, 252. [31] X. Wang, G. Li, M. H. Seo, F. M. Hassan, M. A. Hoque, Z. Chen,
[4] L. E. Blanc, D. Kundu, L. F. Nazar, Joule 2020, 4, 771. Adv. Energy Mater. 2015, 5, 1501106.
[5] M. Song, H. Tan, D. L. Chao, H. J. Fan, Adv. Funct. Mater. 2018, 28, [32] Y. Zhao, Y. Zhu, X. Zhang, InfoMat 2020, 2, 237.
1802564. [33] T. Xiong, Z. G. Yu, H. Wu, Y. Du, Q. Xie, J. Chen, Y.-W. Zhang,
[6] V. Yufit, F. Tariq, D. S. Eastwood, M. Biton, B. Wu, P. D. Lee, S. J. Pennycook, W. S. V. Lee, J. Xue, Adv. Energy Mater. 2019, 9,
N. P. Brandon, Joule 2019, 3, 485. 1803815.
[7] J. F. Parker, J. S. Ko, D. R. Rolison, J. W. Long, Joule 2018, 2, 2519. [34] L. Wang, K.-W. Huang, J. Chen, J. Zheng, Sci. Adv. 2019, 5, eaax4279.
[8] Z. Xu, S. Jin, N. Zhang, W. Deng, M. H. Seo, X. Wang, Nano Lett. [35] J. W. Ding, Z. G. Du, L. Q. Gu, B. Li, L. Z. Wang, S. W. Wang,
2022, 22, 1350. Y. J. Gong, S. B. Yang, Adv. Mater. 2018, 30, 1800762.
[9] J. Mei, J. Wang, H. Gu, Y. Du, H. Wang, Y. Yamauchi, T. Liao, Z. Sun, [36] Q. Yang, F. Mo, Z. Liu, L. Ma, X. Li, D. Fang, S. Chen, S. Zhang,
Z. Yin, Adv. Mater. 2021, 33, 2004920. C. Zhi, Adv. Mater. 2019, 31, 1901521.
[10] Y. Tian, Y. An, C. Wei, B. Xi, S. Xiong, J. Feng, Y. Qian, Adv. Energy [37] L. Ma, S. Chen, C. Long, X. Li, Y. Zhao, Z. Liu, Z. Huang, B. Dong,
Mater. 2020, 11, 2002529. J. A. Zapien, C. Zhi, Adv. Energy Mater. 2019, 9, 1902446.
[11] M. S. Chae, J. W. Heo, S.-C. Lim, S.-T. Hong, Inorg. Chem. 2016, 55, [38] Y. Wang, C. Wang, Z. Ni, Y. Gu, B. Wang, Z. Guo, Z. Wang, D. Bin,
3294. J. Ma, Y. Wang, Adv. Mater. 2020, 32, 2000338.
[12] Y. Cheng, L. Luo, L. Zhong, J. Chen, B. Li, W. Wang, S. X. Mao, [39] H.-Y. Shi, Y.-J. Ye, K. Liu, Y. Song, X. Sun, Angew. Chem., Int. Ed.
C. Wang, V. L. Sprenkle, G. Li, J. Liu, ACS Appl. Mater. Interfaces 2018, 57, 16359.
2016, 8, 13673. [40] W. Yang, X. Du, J. Zhao, Z. Chen, J. Li, J. Xie, Y. Zhang, Z. Cui,
[13] Y. Yang, J. Xiao, J. Cai, G. Wang, W. Du, Y. Zhang, X. Lu, C. C. Li, Q. Kong, Z. Zhao, C. Wang, Q. Zhang, G. Cui, Joule 2020, 4, 1557.
Adv. Funct. Mater. 2020, 31, 2005092. [41] G. Wang, B. Kohn, U. Scheler, F. Wang, S. Oswald, M. Löffler,
[14] W. Li, K. Wang, S. Cheng, K. Jiang, Adv. Energy Mater. 2019, 9, D. Tan, P. Zhang, J. Zhang, X. Feng, Adv. Mater. 2020, 32, 1905681.
1900993. [42] D. Kundu, B. D. Adams, V. D. Ort, S. H. Vajargah, L. F. Nazar, Nat.
[15] T. Xiong, Y. Zhang, Y. Wang, W. S. V. Lee, J. Xue, J. Mater. Chem. A Energy 2016, 1, 16119.
2020, 8, 9006. [43] M. Li, Y. Zhang, Z. Bai, W. W. Liu, T. Liu, J. Gim, G. Jiang, Y. Yuan,
[16] Y. Lu, Q. Zhang, L. Li, Z. Niu, J. Chen, Chem 2018, 4, 2786. D. Luo, K. Feng, R. S. Yassar, X. Wang, Z. Chen, J. Lu, Adv. Mater.
[17] P. Poizot, J. Gaubicher, S. Renault, L. Dubois, Y. Liang, Y. Yao, 2018, 30, 1804271.
Chem. Rev. 2020, 120, 6490. [44] F. Wu, M. Liu, Y. Li, X. Feng, K. Zhang, Y. Bai, X. Wang, C. Wu, Elec-
[18] L. Chen, W. Li, Z. Guo, Y. Wang, C. Wang, Y. Che, Y. Xia, J. Electro- trochem. Energy Rev. 2021, 4, 382.
chem. Soc. 2015, 162, A1972. [45] S.-H. Park, P. J. King, R. Tian, C. S. Boland, J. Coelho, C. Zhang,
[19] Y. Wang, X. Cui, Y. Zhang, L. Zhang, X. Gong, G. Zheng, Adv. Mater. P. McBean, N. McEvoy, M. P. Kremer, D. Daly, J. N. Coleman,
2016, 28, 7626. V. Nicolosi, Nat. Energy 2019, 4, 560.
[20] Y. Lu, J. Chen, Nat. Rev. Chem. 2020, 4, 127. [46] Y. Lu, Y. Cai, Q. Zhang, J. Chen, Adv. Mater. 2021, https://doi.
[21] H. Peng, Q. Yu, S. Wang, J. Kim, A. E. Rowan, A. K. Nanjundan, org/10.1002/adma.202104150.
Y. Yamauchi, J. Yu, Adv. Sci. 2019, 6, 1900431. [47] Z. Lin, H.-Y. Shi, L. Lin, X. Yang, W. Wu, X. Sun, Nat. Commun.
[22] D. Wu, G. Zhang, D. Lu, L. Ma, Z. Xu, X. Xi, R. Liu, P. Liu, Y. Su, 2021, 12, 4424.
J. Mater. Chem. A 2018, 6, 13613. [48] W. Wang, V. S. Kale, Z. Cao, S. Kandambeth, W. Zhang, J. Ming,
[23] H. Zhang, Y. Fang, F. Yang, X. Liu, X. Lu, Energy Environ. Sci. 2020, P. T. Parvatkar, E. Abou-Hamad, O. Shekhah, L. Cavallo,
13, 2515. M. Eddaoudi, H. N. Alshareef, ACS Energy Lett. 2020, 5, 2256.
[24] M. Yu, N. Chandrasekhar, R. K. M. Raghupathy, K. H. Ly, H. Zhang, [49] N. Liu, X. Wu, Y. Zhang, Y. Yin, C. Sun, Y. Mao, L. Fan, N. Zhang,
E. Dmitrieva, C. Liang, X. Lu, T. D. Kühne, H. Mirhosseini, Adv. Sci. 2020, 7, 2000146.
I. M. Weidinger, X. Feng, J. Am. Chem. Soc. 2020, 142, 19570. [50] N. Zhang, F. Cheng, Y. Liu, Q. Zhao, K. Lei, C. Chen, X. Liu, J. Chen,
[25] Z. Xu, W. Deng, X. Wang, Electrochem. Energy Rev. 2021, 4, 269. J. Am. Chem. Soc. 2016, 138, 12894.
[26] Z. Ye, Z. Cao, M. O. Lam Chee, P. Dong, P. M. Ajayan, J. Shen, [51] Z. Xu, R. Ma, X. Wang, Energy Storage Mater. 2022, 46, 233.
M. Ye, Energy Storage Mater. 2020, 32, 290. [52] H. Zhang, Q. Liu, Y. Fang, C. Teng, X. Liu, P. Fang, Y. Tong, X. Lu,
[27] W. Sun, F. Wang, B. Zhang, M. Zhang, V. Küpers, X. Ji, C. Theile, Adv. Mater. 2019, 31, 1904948.
P. Bieker, K. Xu, C. Wang, M. Winter, Science 2021, 371, 46. [53] H. Wang, W. Ye, Y. Yang, Y. Zhong, Y. Hu, Nano Energy 2021, 85,
[28] H. Chen, H. Xu, S. Wang, T. Huang, J. Xi, S. Cai, F. Guo, Z. Xu, 105942.
W. Gao, C. Gao, Sci. Adv. 2017, 3, eaao7233. [54] X. Dong, Z. Guo, Z. Guo, Y. Wang, Y. Xia, Joule 2018, 2, 902.

Adv. Mater. 2022, 34, 2200077 2200077 (11 of 11) © 2022 Wiley-VCH GmbH