Unveiling the Interlayer Spacing Dependence of Zn-Ion Storage Performance in Layered Vanadium Phosphates

Layered vanadium phosphate (VOPO 4 ·2H 2 O) is reported as a promising cathode material for rechargeable aqueous Zn 2+ batteries (ZIBs) owing to its unique layered framework and high discharge plateau. However, its sluggish Zn 2+ diffusion kinetics, the low specific capacity and poor electrochemcial stability remains a major issue in battery application. In this work, a group of phenylamine (PA)-intercalated VOPO 4 ·2H 2 O with varied interlayer spacing (14.8, 15.6 and 16.5 Å) is synthesized respectively via a solvothermal method for cathode of aqueous ZIBs. The specific capacity is quite dependent on d -spacing in PA-VOPO 4 ·2H 2 O system followed by an approximate linear tendency, and the maximum interlayer spacing (16.5 Å phase) results in a discharge capacity of 268.2 mAh·g -1 at 0.1 A·g -1 with a high discharge plateau of ~ 1.3 V and an energy density of 328.5 Wh·Kg -1 . Both of the experimental data and DFT calculation identify that the optimal 16.5 Å spacing can boost fast Zinc-ion diffusion with an ultrahigh diffusion coefficient of ~ 5.7 × 10 -8 cm -2 ·s -1 . The intercalation of PA molecules also significantly increases the hydrophobility in the aqueous electrolyte, resulting in the inhibiting the decomposition / dissolution of VOPO 4 ·2H 2 O and remarkably improved cycling stability over 2000 cycles at 5.0 A·g -1 with a capacity retention of ~200 mAh · g -1 . Our study provides a feasible solution on the sluggish Zn 2+ diffusion kinetics and poor cyclstability, and also shows a clear understanding on the interlayer chemistry of layered phosphates towards aqueous Zinc-ion storage.

3 Nowadays, although rechargeable lithium-ion batteries (LIBs) are widely used for various portable electronics and electric vehicles, limited Li sources and severe safety issue are two main drawbacks of LIBs. With the rapid development of modern electronics and advanced equipments, beyond-lithium-ion (Na + , K + , Mg 2+ , Ca 2+ , Zn 2+ , Al 3+ ) batteries have recently attracted growing attention for energy storage systems. 1 Among them, Zinc is abundant metallic elements in the earth crust far exceeding Li. More importantly, rechargeable aqueous Zinc-ion batteries (ZIBs) are based on aqueous electrolyte, resulting in much superior safety than that of alkali metal (Li + , Na + , K + ) based secondary batteries. Consequently, ZIBs have been considered as very promising candidates for next-generation battery systems. 2,3 Nevertheless, the revolution of aqueous ZIBs is still at the early stage and far from the prospective applications, which is mainly limited by the absence of suitable cathode materials especially with high energy density and sufficient long lifespan. 4 The origin of this issue should be ascribed to the intrinsic difference between univalent lithium ion and divalent zinc ion. Since the similar radius of Li + (0.076 nm) and Zn 2+ (0.074 nm), the surface charge density of divalent Zn 2+ is more than the twice of Li + , and the much stronger electrostatic interaction between Zn 2+ and lattice framework of the electrode materials thus leads to sluggish diffusion kinetics and even structure deterioration. 5,6 Inorganic layered compounds, based on a typical layered structure and rich intercalation chemistry, have been particularly studied in metal-ion batteries. Especially, layered vanadium phosphate, VOPO 4 · 2H 2 O, has been reported as an important cathode material for ZIBs accompanied with a Zn 2+ intercalation / de-intercalation mechanism in its interlayer galleries. [7][8][9] Remarkably, taking advantage of the enhanced ionicity of V-O bonds with the existence of PO 4 3-, layered VOPO 4 · 2H 2 O exhibites a much higher discharge plateau (1.1 -1.2 V) than other V-based cathode materials. [10][11][12] It should be emphasized that energy density is quite dependent on the discharge plateau of cathode, and such a merit of layered VOPO 4 · 2H 2 O is highly desirable to achieve high energy density of ZIBs. 10 Unfortunately, the specific capacity, rate capability of VOPO 4  Recent progress on lithium-/ sodium-/ magnesium-ion batteries unveils the interlayer spacing of layered materials as natural two-dimensional (2D) ion transport / diffusion channels. [13][14][15][16] In principle, increasing the interlayer spacing of layered structure facilitates the ion transport by creating a lower energy barrier. 6   7 CR2032 coin cells were then assembled employing these PA-VOP samples as cathode material, Zn foil as anode and 2M Zn(CF 3 SO 3 ) 2 solution as aqueous electrolyte, respectively. Typical cyclic voltammetry (CV) curves of Zn // PA-VOP (16.5 Å phase) battery reveals a general active process in the first scan cycle, while the curves of the stabilized second and third cycle gives a pair of oxidation / reduction peak located at around 1.6 V and 1.3 V, respectively (Supplementary Fig. 6). It is should be noteworthy that PA molecules is generally inactive in mild aqueous electrolyte, and the electrochemical activity and capacity contribution should be mostly provided from VOPO 4 framework. [29][30][31] The galvanostatic charge discharge (GCD) curves also exhibit steady charge and discharge plateaus with well agreement with the voltage range in CV curves ( Supplementary Fig. 6, 7).
It has been reported that non-metal proton (H + ) has also been recognized as charge carrier ions for some aqueous batteries. 32 To clarify this, we compare the electrochemical performance on both aqueous and non-aqueous system (0.5 M ZnSO 4 / acetonitrile, Supplementary Fig. 8). No apparent difference has been observed between the CV peaks, thereby ruling out the possible contribution from proton intercalation in our system. Ex-situ XRD at different voltages in the second cycle reveals a revisable Zn 2+ intercalation / de-intercalation mechanism of the PA-VOP cathode ( Supplementary Fig. 9). We found the specific capacity of the PA-VOP cathodes is quite dependent on its interlayer spacing. As shown in Fig. 2g interlayer spacing at a current density of 5 A·g -1 , respectively.
We reveal the capacity-interlayer spacing dependence could be followed by an approximate linear trend as summarized in Fig. 3a. The specific capacity of 16.5 Å phase (268.2 mAh· g -1 ) at the current density of 0.1 A· g -1 is about two times higher than that of pristine VOP (138.4 mAh· g -1 ) and also well above that of layered PPy-VOPO 4 cathode (~ 80 mAh· g -1 ) reported before. 23 (Fig. 3d). 56 This excellent energy and powder densities promote the application in flexible and wearable electronic devices (Insert in Fig. 3d).  Reaction/diffusion kinetics mechanism. In order to clarify the aforementioned d-spacing dependent battery-performance, corresponding study on electronic conductivity and Zinc-ion diffusion kinetics is further conducted. Four-probe conductivity test shows a 2 ~ 7 times higher electronic conductivity of PA-VOP (16.5 Å) in contrast to that of pristine VOP with pressure increased, which is mainly attributed to the intrinsic high conductivity of phenylamine (Fig. 4a). CV curves at varied scan rates are employed to distinguish the capacitive / ion diffusion contribution by analyzing the dependence of the logarithm of peak current density and the scan rate ( Supplementary Fig. 10). Theoretically, the peak current (i) and sweep rate (v) in CV curve follows the rule as: 57,58 Where  (Fig. 4b, c).
On the other hand, Zinc-ion diffusion kinetics is further identified by galvanostatic intermittent titration technique (GITT) (Fig. 4d). Strikingly, an ultrahigh Zinc-ion diffusion coefficient of ~ 5.7 × 10 -8 cm -2 ·s -1 has been detected in 16.5 Å phase, nearly 5 orders of magnitude higher than that of pristine VOP cathode (6.2 × 10 -13 cm -2 ·s -1 ) and also much higher than that of conventional cathode materials  (Fig. 4e). This drastic increase demonstrates a great improvement in terms of Zn 2+ intercalation / de-intercalation kinetics by an enlarged 2D interlayer channel as illustrated in Fig. 4f. In principle, large Zinc-ion diffusion coefficient is generally observed in ultrathin samples due to the shortened effective diffusion path and larger activated surface, which drastically facilitate Zn 2+ diffusion and charge transfer. 28,60,61 Our previous study also confirms fast Zn-ion diffusion ability in 6.0-nanometer ultrafine spinel oxide 11 nanodots. 61 However, in the present study, the as-obtained 16.5 Å phase show a significant increase on thickness compared with the pristine VOP sample (Fig 2d-f). Therefore, the thick samples display much faster kinetics than that of thin VOP nanoplates unusually. 60 octahedron gives an optimal absorption energy of 0.07 eV, which can be determined as the potential adsorption site (called as C site) for zinc-ion diffusion. The as-optimized diffusion path denotes hopping process through the right above site of [VO 6 ] octahedron and [PO 4 ] tetrahedron between the adjacent C sites (Fig. 4g, h). Subsequently, an interlayer-dependent activation energy decrease tendency is derived   (Supplementary Fig. 11). One can see the with rapid decay in the first 300 cycles (Fig. 5a). To clarify the actual mechanism, ex-situ XRD characterizations of VOP and PA-VOP (16.5 Å phase) cathode before and after 300 cycles are carried out as shown in Fig. 5b. Apparently, the characteristic diffraction peaks of VOP cathode completely disappear after 300 cycles, indicating the phase conversion of VOP during the long-term cycling. In contrast, the PA-VOP (16.5 Å phase) cathode well maintained the initial phase without any impurity diffraction peaks. SEM observation also reveals PA-VOP (16.5 Å phase) sample well maintain the initial plate-like morphology compared to the drastic aggregation of VOP cathode after 300 cycles (Fig.   5c). All results demonstate the greatly enhanced structural reversibility of our PA-VOP sample after PA intercalation.
To further probe the origin of the remarkable difference on long-lifespan, we compare the chemical stability of both VOP and PA-VOP samples in aqueous 2M ZnSO 4 electrolyte. As shown in Fig. 5d, it is clear that VOP sample generally disappears in the bottom of solution followed by a typical color change into deep red after 5 days due to the decomposition /dissolution of VOPO 4 · nH 2 O into VO x (s) and PO 4 3-. 9 However, the PA-VOP (16.5 Å phase) sample still floats on the liquid surface with much better stability after 5 days (Fig. 5d). Such a result inspires our consideration on the difference on surface-wettability of these two samples. Accordingly, Fig. 5e compares the water-based contact angle on the surface of pressed powder of VOP and PA-VOP sample. The PA-VOP and VOP smaple shows a contanct angle of 56.7 ° and 9.8 ° respectively, demonstring the surface is much more hydrophobic after PA intercalation.
Rationally, this hydrophobic surface plays a decisive role in the inhibiting the decomposition/dissolution of VOPO 4 in aqueous electrolyte, thus leading to much improved long lifespan of the as-constructed ZIBs. Most recently, Sun et al. reported that the decomposition/dissolution of VOPO 4 · 2H 2 O can be prevented by PO 4 3addition and high salt concentration of the aqueous electrolyte 9 . Such a strategy promotes a long-term cycling over 500 cycles with a stable capacity of only 90 mAh·g -1 at 2.0 A·g -1 but also increases the battery cost. In contrast, our strategy by PA-intercalation requires no addition of any reagent/salt addition, and realizes much superior cycling stability over 2000 cycles with a capacity of ~200 mAh·g -1 .

Discussion
In summary, we developed a controllable phenylamine-intercalated strategy for layered vanadium Electronic conductivity test is carried out on a four-probe conductivity tester (ST2253y).

17
The slurry-coated foil is cut into Φ15 mm electrode as cathode, while the zinc foil washed with ethanol and glass fiber membrane is used as the anode and separator, respectively, and 2M Zn(CF 3 SO 3 ) 2 is prepared as the electrolyte. The CR-2032 cell was assembled in air using the beforehand electrodes and other relevant components. LAND battery test system (CT2001A) was employed to evaluate the electrochemical performance of the battery: galvanostatic charge-discharge (GCD), rate capability and long-term cycle performance. Cyclic voltammetry (CV) test at different scan rate is performed on electrochemical workstation (CHI660E). Galvanostatic intermittent titration technique (GITT) is performed under a modified GCD mode, in which an operation period includes two parts: a charge / discharge procedure last for 10 min at 0.05 A·g -1 and a followed pause time for 10 min.
where I p (A·g -1 ) is the peak current density at different scan rate, (mV·s -1 ) is the specific scan rate, C1 and C2 are the corresponding constant factors of the capacity contribution of surface pseudocapacitive effect and battery-type effect, respectively.
With a deformation of the above equation, the specific contribution rate of different internal mechanisims can be solved according to the following equation: The specific energy density (Wh·kg -1 ) and average specific power density (W·kg -1 ) of the batteries is calculated in terms of the following equations: Where E s is the calculated specific energy density (Wh·kg -1 ), P s is the average specific power density (W·kg -1 ); C s (mAh·g -1 ) is the specific capacity of the battery, V 0 and V 1 is the voltage lower limit and voltage upper limit of discharge procedure, respectively, and t is the discharge time (h). All the parameters calculated are based on the mass loading of the active materials (VOP or PA-VOP).
Diffusion coefficient (D Zn 2+ ) of Zinc-ion can be experimentally calculated by GITT method in terms of the following equation: Where D is the diffusion coefficient of Zinc-ion, τ is the relaxation time of current pulse, L is diffusion length which is approximate to the thickness of coated slurry, ΔE s and ΔE t is the voltage change produced by current pulse and the galvanostatic charge / discharge, respectively.
Simulation details. The modeling in this study was performed in the framework of the DFT as implemented in the Vienna Ab initio Simulation Package (VASP). The functional of Perdew-Burke-Ernzerhof based on generalized gradient approximation (GGA) was applied to describe the exchange-correlation energy. In addition, the zero damping DFT-D3 dispersion correction method of Grimme was accounted for VdW interaction in the system. A vacuum space of 15 Å was adopted.
The plane-wave cutoff energy was set to be 480 eV, and the k-mesh was determined to be 7 × 7 × 1 according to the convergence test, which makes the energy accuracy within 1.0 × 10 −3 eV atom −1 .
Finally, a double-layered VOPO 4 model was constructed, with the corresponding interlayer spacing obtained from our previous XRD and HRTEM analysis, thereby the diffusion of zinc ion between VOPO 4 layers was simulated using the climbing-image nudged elastic band method.