Introduction

The boosting application of aqueous zinc metal batteries mainly hinges on the stable solid-electrolyte interface (SEI) and inhibited side reactions1,2,3,4. On the one hand, water-induced parasitic reaction in aqueous electrolyte systems leads to hydrogen evolution reactions (HER) and Zn dendrites formation at Zn anode surface, which will result in a structure intolerant and chemically unstable solid-electrolyte interphase (SEI)5,6. On the other hand, the water-induced loose SEI is prone to detach from Zn metal, which leads to dead Zn and electrolyte exhaustion7,8. In aqueous Zn-ion batteries (AZIBs), the SEI formation is cardinally derived from electrolyte decomposition which mainly form into zinc-hydroxide passivation layer. Generally, the electrolyte (cations, anions and solvent) derived SEI is essentially constituted of both inorganic components (such as ZnF2, ZnCO3, etc.) and organic compounds. Benefiting from its good electronic insulation and compact structure, the inorganic passivation layer can prevent continuous electrolyte reduction and solvent intrusion into SEI layer to suppress further electrolyte decomposition9,10,11. However, the high Young’s modulus of the inorganic layer tends to crack, when it grows into a limiting size during the Zn-ion deposition electrochemical reaction. Meanwhile, the flexibility of the organic compounds (such as C-O-C chains) can efficiently alleviate the influence of Zn dendrites growth triggered volume expansion, which is critical to build stable and homogeneous SEI within the large size anode, such as pouch cell configuration12,13,14. Thus, constructing compatible organic-inorganic hybrid SEI is pivotal for the long-term stable operation of practical AZIBs, in which electrolyte engineering is a feasible strategy to build the appropriate SEI15,16.

In general, the electrolyte solvation environment, especially the inner Helmholtz plane, is the crucial factor to determine the SEI structure at the interface17. The dilute electrolyte ( < 2 M) is the mainstream electrolyte system for practical AZIBs due to the economic feasibility, in which the primary solvation sheath of Zn-ions is mainly surrounded by water molecules or hydrated ion. In such case, water induced side reactions can severely deteriorate the interfacial electrochemical stability and corresponding charge storage capability18. In contrast, anion dominated interfacial chemistry could be conducive to rationally manipulate the SEI structure while suppressing the HER side reaction19,20. To realize anion-dominated interfacial chemistry, the most common strategy is to directly promote electrolyte salt concentration to an extreme high extent21. The anions are unavoidably forced into the Zn-ion solvation sheath, forming contact ion pairs or even ion aggregates for compelling the anion decomposition at the anode surface. For instance, the high-concentrated electrolyte (HCE) consisting of 1 M Zn(TFSI)2 + 20 M Li(TFSI)2 has been demonstrated to effectively stabilize Zn metal anode because of the ion pairs (Zn-TFSI)+ enabled inorganic-abundant SEI22. Unfortunately, the practical implementation of high concentrated electrolyte is inevitably encountered with issues of high viscosity, sluggish ion kinetics and high cost.

Supramolecular materials, such as crown ethers and cyclodextrins (CD), have been introduced as electrolyte additives to stabilize Li or Zn anode, based on their superamolecularly driven interactions, macrocyclic architectures, molecular recognition, and the capture ability of “host-guest” structure, which render its promising features to regulate the solvation sheath of electrolyte ions on the surface of Zn anode23,24. The truncated cone like cavity structure in combination with a relatively hydrophobic center and a hydrophilic outer surface could contribute to molecular recognition and capture capability25. Thus, the hydrophobic central could be expected to accommodate decomposable OTf- or TFSI- anions to form CD@anion complex, subsequently achieving anion-enriched interfacial chemistry through CD@anion complex preferentially absorbing on Zn/electrolyte interface. In addition, due to the direct interaction between anion receptors and anions, stability of anions could be effectively altered, rather than indirectly regulating the stability between anions and cations by reducing the number of water molecules in traditional high-concentration22 (HC) or ionic liquid (IL) electrolytes7,13,21. Thus, the proposed strategy of anion receptors avoids the problems of high viscosity, high cost, and slow ion kinetics in HC and IL electrolyte systems, providing the possibility for the regulation of anion solvation structure under small addition amounts.

In this work, we introduce β-CD additive as anion-receptors into 2 M Zn(OTf)2 aqueous electrolyte to form β-CD@OTf- complex, in which successfully constructed an anion-enrichment interface in dilute electrolyte system. The inner Helmholtz layer could be manipulated via the formation of the β-CD@OTf- complex near Zn anode surface, which in situ converts into organic/inorganic (ZnF2/ZnCO3/ZnS‒(C-O-C/*CF/*CF3)) hybrid SEI during the initial electrochemical plating/stripping reaction. The hybrid SEI simultaneously enables the homogenous Zn deposition and restrained dendrite formation. The thus-derived Zn | |Zn pouch cell demonstrates a predominant electrochemical stability with a cumulative capacity of 6450 mAh cm−2 even at extreme conditions (high areal capacity of 10 mAh cm−2, Zn utilization rate (ZUR) of 34.2%). Such anion receptors additives strategy furnishes a new avenue to promote anion decomposition SEI in dilute electrolyte for large-size homogeneously stable Zn deposition of AZIBs. In addition, the Zn | |ZnxV2O5 full cell also demonstrates ~ 90% areal capacity retention even after 600 cycles even at a low N/P ratio of approximate 2.1 (47.6% ZUR).

Results

Anion-enrichment IHP design

To clarify the ion distribution mechanism at electrode/electrolyte interface, we should primarily familiarize the concept of inner Helmholtz plane (IHP). IHP is defined as the inner layer of electric double layer (EDL) and electrical center of specific adsorbed ions, which is closely correlated with SEI formation26. In conventional dilute Zn(OTf)2 electrolyte, the primary solvation sheath of Zn2+ consists of abundant H2O molecule and slight OTf- anions, resulting in H2O-enrichment and OTf- anion-deficient IHP accompanied by solvated Zn2+ clusters transport from bulk electrolyte to Zn anode surface (Supplementary Fig. 1). H2O molecule dominated IHP is prone to form solvent-derived loose and porous SEI due to the continuous H2 evolution (Supplementary Fig. 2a), which fails to isolate H2O molecules from Zn electrodes and hinders the Zn ions transport. In contrast, in the proposed anion receptors regulated electrolyte, β-CD accompanied with OTf- anions adsorbs on IHP owing to the “host to guest” interaction and preferential adsorption behavior of β-CD (Fig. 1b). With the continuous electric field stimulation, the β-CD@OTf--rich IHP tends to decompose to organic-inorganic hybrid SEI that facilitate Zn ion transport and block the contact between H2O and Zn anode (Fig. 1a and Supplementary Fig. 2b).

Fig. 1: Design strategy of anion-enrichment interface.
figure 1

a Schematic illustration of anion receptors β-CD enabled β-CD@OTf--rich IHP and thus-derived SEI. b Electrode/electrolyte interface consist of inner Helmholtz plane (IHP), outer Helmholtz plane (OHP), and diffusion layer. c Binding energy comparison of OTf- anions with varied CD molecules. d Raman spectra of pure Zn(OTf)2, β-CD and β-CD@OTf- complex. e Schematic snapshots of simulated interfacial structures in β-CD-Zn(OTf)2 electrolyte without applied charge (0 µC cm−2). f LSV curves on Ti electrode at the scan rate of 5 mV s−1. The inset is DFT calculations on the HOMO and LUMO level of bare OTf- anion and β-CD@OTf-. g The corresponding number density profiles along the vertical graphene surface under condition of 0 µC cm−2 in Zn(OTf)2 and β-CD-Zn(OTf)2, respectively. h The partial enlarged snapshot of Fig. 1e suggesting β-CD accommodate OTf- anions.

Full size image

What is the reason for choosing β-CD as an anion receptor to build anion-enrichment interface? Primarily, anion receptors should possess sufficient electron withdrawing ability to chelate with anions27. As aforementioned, supermolecular macrocycle tends to accommodate anions through “host to guest” interaction, based on the macrocyclic confinement effect. CD is a macrocyclic oligomer constructed of six to eight glucopyranose units linked by α−1,4 glycosidic bonds in shape of slightly cone-shaped ring, which is a fascinating candidate molecule to capture anions due to the appropriate macrocyclic size, hydrophobic inner cavity and hydrophilic outer surface. Owing to the varied amount of glucose residues, CD usually presents three typical configurations, including α-CD, β-CD, γ-CD with diverse cavity size (Supplementary Fig. 3). As displayed in Supplementary Figs. 4, 7, we simulate the electrostatic potential mapping (ESP) for all three molecular van der Waals surface to intuitively reflect the best active sites for ion storage. Obviously, the inner cavity of CD molecule assumes more positive ESP relative to the outer cavity with weak electron withdrawing ability, which indicating that the non-polar environment inside inner cavity is conducive to accommodate OTf- anions. In addition, we also perform independent gradient model (IGM) analysis to analyze the interactions between CD molecule and OTf- anion, which demonstrates that the “host-guest” interactions are cardinally composed of H-bond and Van der Waals (vdW) force (Supplementary Fig. 4b). However, anion receptors are not only required to accommodate anions, but also need to be able to adsorb on IHP near the Zn metal surface to construct anion-enrichment IHP interface. As summarized in Supplementary Fig. 4c, the anion adsorption behavior has been also systematically studied through differential charge density distribution between Zn slab and CD molecule, in which the electrons are prone to transfer from CD to Zn surface for constructing a strong electrostatic adsorption.

To screen the appropriate CD molecule for accommodating OTf- anions, the binding energy between CD molecules and OTf- anions has been also simulated, in which β-CD presents lower binding energy with OTf- of −8.38 eV than that of α-CD and γ-CD (Fig. 1c), suggesting that OTf- anions is more inclined to be encapsulated into the inner cavity of β-CD. Thus, the β-CD has been selected as the target anion-receptor to accommodate Zn(OTf)2 electrolyte. To verify the capability of β-CD serving as electrolyte additive, the solubility of β-CD in pure H2O solutions and Zn(OTf)2 electrolyte has been first studied. The soluability of β-CD was significantly enhanced from 10 mM to 150 mM after the introduction of 2 M Zn(OTf)2 (Supplementary Fig. 8), which significantly interpreting that there is indeed a special interaction between β-CD and OTf- anion. Meanwhile, as shown in the Raman spectra of Fig. 1d, the characteristic peak of OTf- anion located at 756 cm−1 also demonstrates a slight red shift compared with β-CD@ OTf- complex, which also indicates the interaction between OTf- and β-CD. Similarly, as depicted in FT-IR results of Supplementary Fig. 9a, the peak intensity of the β-CD related C-O-C groups almost disappears when introduced into Zn(OTf)2 solution, which possibly demonstrates the successful entry of OTf- into the inner cavity of β-CD. We also ruled out the negligible possibility of interaction between Zn2+ ions and β-CD molecules (Supplementary Fig. 10–12 and Note 1). In addition, the encapsulation of OTf- anions in β-CD also results in the Zn2+ transference number increase from 0.32 in Zn(OTf)2 electrolyte to 0.56 in β-CD/Zn(OTf)2 electrolyte (Supplementary Fig. 13). The promoted Zn2+ migration behavior is cardinally affected by Zn2+ solvation environment accompanying by the restricted anions transport, which is coincided with the function of anion receptor. Subsequently, the adsorption behavior of β-CD was studied through experiments and theoretical calculations. The adsorption energy of α-CD, β-CD, γ-CD molecules on Zn (002) slab is calculated to be −2.92, −3.32 and −2.74 eV, respectively (Supplementary Fig. 14), revealing the strong interaction of β-CD with Zn metal. As shown in FT-IR spectra of supplementary Fig. 15, after immersed in β-CD aqueous solution, the stretching vibration of C-O-C polar functional groups belonging to β-CD appeared on the surface of Zn metal, which directly proving the adsorption of β-CD on Zn metal surface. Therefore, according to the accommodation ability between OTf- ions and CD molecules with proper adsorption behavior on Zn surface, we conceived that β-CD may be the appropriate anion receptor for constructing anion-enrichment IHP.

The molecular dynamics (MD) simulations are employed to dissect ion distribution circumstances at electrode/electrolyte interface (Fig. 1e, g, h, Supplementary Fig. 16, 17 and Table 2). Figure 1e, g and Supplementary Fig. 16 demonstrate that OTf- anions tends to adsorb on Zn/electrolyte interface in β-CD-Zn(OTf)2 system with minor water molecules surround at Zn/electrolyte interface with 0 µC cm−2 applied charge, which is mainly attributed to the β-CD@OTf- complexes intensive adsorption. Conversely, in pure Zn(OTf)2 electrolytes, H2O molecules inevitably adsorb on Zn/electrolyte interface due to the absence of any shields (Supplementary Fig. 16). The analogous phenomenon take place even with an electric field applied of −6 µC/cm2 (Supplementary Fig. 17), in which the distribution of OTf- anions and β-CD at Zn/electrolyte interface are highly correlated. It suggests that in the thus-designed β-CD regulated electrolyte tends to form anion-enrichment interface accompanied by β-CD triggered OTf- anions movement onto the Zn anode surface. Furthermore, we also checked the Zn2+ solvation structure (Supplementary Fig. 18, 19) and anlayzed the occupation ratio of various state of water molecules (Supplementary Fig. 20, 21) to clarify the reasons for decrease in the number of water molecules (Supplementary Note 2). In addition, in the β-CD-containing electrolyte, the reduction potential of OTf- anions has been also manipulated via the strong interaction with β-CD. As shown in the inset of Fig. 1f, the lowest unoccupied molecular orbital (LUMO) energy level which represents the ability to gain electrons of OTf- anions was reduced from 5.38 eV to 2.87 eV owing to the interaction between β-CD and OTf-, indicating the increased reducibility of OTf- in β-CD-Zn(OTf)2 electrolyte, which is favorable for the decomposition of OTf- anions. This prediction is also experimentally authenticated by the Linear sweep voltammetry (LSV) tests in Fig. 1f. Obviously, OTf- anions decompose approximately at 0.07 V, when the potential reaches at −25 mV (vs. Zn/Zn2+), a sharper current decrease at around −0.025 V is observed, which is ascribed to the accelerated β-CD@OTf- reduction. For pure 2 M Zn(OTf)2 system, inconspicuous current decrease is also observed at the same potential of −0.025 V, which is identical to DFT prediction. Therefore, the anion-enrichment interface and the thus-advanced reduction potential of β-CD@OTf- complexes synergistically contribute to the emergence of organic-inorganic hybrid SEI, escorting the stable Zn metal anode.

Anion receptor@anion complexes derived SEI chemistry

The SEI structure of different electrolytes was systematically investigated through X-ray photoelectron spectroscopy (XPS) (Fig. 2a and Supplementary Fig. 22, 23). For pure Zn(OTf)2 electrolyte, the C 1 s spectra were cardinally composed of four species, the C-H/C-C (284.8 eV), the C-O-C (285.7 eV), the *CF (289.2 eV) and the *CF3 (293.2 eV), arising from the incomplete decomposition of Zn2+-OTf- clusters (Supplementary Fig. 23). However, inorganic CO32- ( ~ 287.6 eV) was observed in C 1 s spectra on Zn anode harvested from β-CD-Zn(OTf)2 electrolyte (Fig. 2a), which could be ascribed to the sufficient reduction of β-CD@OTf- complex. Note that dissolved CO2 from air also could be another reason for the formation of ZnCO3 species. Accompanied with prolonged sputtering time (5 min), the peak intensity of ZnCO3 almost remains invariable. The existence of ZnCO3 after such a powerful resistance against Ar+ etching indicates that the underlying ZnCO3 interphase penetrates into a certain thickness. A similar situation was also observed in the S 2p spectra, the detected ZnSO3 (167.8 eV) in β-CD-Zn(OTf)2 system also disappeared on Zn anode recovered from pure Zn(OTf)2 system (Fig. 2a), with ZnS (164.6 eV) species existing in both systems. Besides, ZnF2 species was observed in the F 1 s spectra for both electrolyte systems. The ratio for inorganic fluorine to organic fluorine progressively aggrandized accompanied with Ar+ etching, which also suggests the formation of ZnF2-rich interphase. Based on previous studies, the inorganic interphase, such as ZnCO3 and ZnF2, could generally endow the Zn anode surface with high interface energy, thereby inhibiting Zn dendrite formation via promoting the Zn2+ lateral migration instead of vertical growth1,28.

Fig. 2: The structure characterization of SEI in Zn(OTf)2 electrolyte with or without β-CD.
figure 2

a C 1 s, F 1 s, S 2p, and O 1 s XPS spectra of Zn anode extracted from β-CD-Zn(OTf)2 electrolyte with Ar+ sputtering time for 0, 2, and 5 min. b The high-resolution transmission electron microscopy (HRTEM) image of layered-structure SEI on Zn surface in β-CD-Zn(OTf)2 electrolyte system. The normalized strength of varied species as a function of sputtering time for SEI derived in (c) Zn(OTf)2 and (g) β-CD-Zn(OTf)2 electrolyte, respectively. 3D TOF-SIMS mappings of the various species on the cycled Zn metal surface in (d) Zn(OTf)2 and (h) β-CD-Zn(OTf)2 electrolytes. e, f, i, j HRTEM images of marked regions 1–4 in Fig. 2b with lattice fringes, respectively.

Full size image

In addition, although C-O-C, ZnF2 and ZnS components were detected on the Zn electrodes recovered from above two electrolyte systems, there were significant differences in their relative peak intensities and content ratio. Specifically, as summarized in Fig. 2c, the SEI constructed in pure Zn(OTf)2 electrolyte exhibits the feature with inadequate anion-derived SEI with ZnF2 and ZnS only increasing to the normalized strength of 2.5 and 0.41, while the initial normalized strength of C-O-C species drops from ~3.2 to 1.5 (Fig. 2c). In contrast, with respect to β-CD-containing electrolytes, the formed SEI shows an enhanced organic-inorganic hybrid feature, in which C-O-C species generates the normalized strength of 4.2 while the ZnF2 and ZnS occurs at a higher level from 2.5 to 3.6 and 0.41 to 0.46, respectively (Fig. 2g). The promoted organic outer layer is mainly attributed to the co-decomposition of β-CD@OTf- complexes, whilst the enhanced derived inorganic interphase is originated from promoted OTf- anion reduction. The above results indicates that the SEI formed in β-CD-containing system is essentially dominated by organic species of C-O-C/*CF/*CF3 species and inorganic species of ZnF2/ZnCO3/ZnSO3/ZnS, which is presented in the layered structure with ZnF2/ZnCO3/ZnSO3/ZnS in inner layer and organic C-O-C/*CF/*CF3 in outer layer. In such a scenario, inorganic inner layer of SEI holds certain mechanical properties to prevent Zn dendrite formation, and pushes H2O molecules away from Zn surface, while the organic outer layer of SEI alleviates the volume strain during Zn growth due to its porosity and flexibility, with simultaneously promoting the electrolyte penetration29.

To further clarify the composition distribution of interfacial SEI, the concentration of organic CH2O2-, inorganic ZnF- and ZnCO3- species on the Zn anode cycled in Zn(OTf)2 and β-CD-Zn(OTf)2 electrolytes was resolved through time-of-flight secondary ion mass spectroscopy (TOF-SIMS). Notably, organic CH2O2- species in β-CD-Zn(OTf)2 system presents a much stronger intensity than that of Zn(OTf)2, which is benefited from the adsorption of β-CD on IHP near Zn anode (Fig. 2d, h). The two-dimensional and three-dimensional strength distribution of specific C- and O- fragments also exhibits the consistent trend (Supplementary Figs. 24, 25). The roughness of organic species endows the SEI with enough flexibility to relieve volume expansion brought by dendrite growth30. Similarly, the intensity of inorganic ZnF- and ZnCO3- fragments with the support of β-CD is significantly higher than pure Zn(OTf)2 system, which demonstrates uniform intensity distribution with an enhanced organic-inorganic SEI. Meanwhile, the intensity of inorganic compounds maintains the stable high value whereas the organic species intensity gradually declines when Ar+ sputtering etching into the inner layer of SEI. In combination with in-depth XPS analysis and TOF-SIMS results, a layered structure SEI with flexible organic outer layer (C-O-C/*CF/*CF3) and rigid inorganic inner layer (ZnF2/ZnCO3/ZnSO3/ZnS) have been also intuitively demonstrated. This SEI can prevent electron tunneling effect to prohibit the reduction of H2O molecules while promoting the Zn2+ migration22.

The intuitive dissection of layer-structured SEI at the Zn/electrolyte interface was eventually implemented with high-resolution transmission electron microscopy (HRTEM). The interfacial SEI exhibits evident dual-layer characteristic, with constituting inner inorganic (dark domains) and outer organic layers (bright domains) which conforms with species distribution from TOF-SIMS via distinguishing from imaging contrast (Fig. 2b). Inorganic components are recognized as ZnF2 and ZnCO3 mixtures deduced from lattice fringes (0.26 nm corresponds for (101) of ZnF2 and 0.28 nm corresponds for (104) of ZnCO3) (Fig. 2e, i). Such inorganic SEI components mainly contributes the occurrence of semicircles at high frequencies in the electrochemical impedance spectra (EIS) for β-CD-Zn(OTf)2 electrolyte (red lines in Supplementary Fig. 26)31, whereas outer layer is essentially in the amorphous state, implying the organic-rich species (Fig. 2f, j). The inorganic and organic SEI components distribution recognition is well coordinated with XPS and TOP-SIMS analysis. In addition, we could also recognized that the thickness of the organic outer layer and inorganic inner layer are ~25 nm and ~15 nm, respectively. Therefore, a legible layer-structured SEI with ZnF2-rich inorganic inner layer and C-O-C/*CF/CF3-rich organic outer layer has been generally confirmed based on a series of XPS, TOP-SIMS, HRTEM and corresponding energy-dispersive X-ray (EDX) spectroscopy mapping characterizations (Supplementary Fig. 27). Furthermore, as shown in supplementary Fig. 28, we conducted element mapping analysis on the SEI formed in β-CD-Zn(OTf)2 electrolyte based on HRTEM image. The white and pink boxed areas represent the organic layer domain and the inorganic layer domain, respectively. From the elemental spectra in supplementary Fig. 28b, c, we can see that both organic and inorganic domains contain a certain amount of F element except for C and O element. The specific proportion of element content is given in Supplementary Tables 4, 5. Plainly, the organic layer contains 1% F element compared with 10% in inorganic layer, suggesting that the decomposition of OTf- anions directly participates in the formation of the organic part of SEI, the organic part of SEI is not only entirely derived from the decomposition of β-CD molecules, but also from OTf- anions.

Zn plating/stripping behavior

β-CD additive endows Zn(OTf)2 electrolyte with anion-rich IHP through host-guest interaction and preferential adsorption mechanism, which could generate a uniform organic-inorganic hybrid SEI. It is expected to exert a significant impact on the deposition and stripping behavior of Zn ions. Scanning electron microscopy (SEM) was first carried out to observe Zn deposits morphology after 20 cycles in form of Zn | |Zn symmetric cell with 2 M Zn(OTf)2 electrolyte. As shown in Fig. 3e, f, obvious mossy Zn deposits and massive by-products are detected on Zn surface, with a thickness even up to 40 µm from the cross-sectional SEM images (Supplementary Fig. 29a, b). It is also demonstrating highly rough and inhomogeneous Zn deposition. As schematically illustrated in Fig. 3a, b, the anion-deficient and H2O-rich IHP leads to an inhomogeneous and loose solvent-derived SEI containing nauseous by-products (such as ZnO), which fails to expel H2O from Zn anode surface or suppress uncontrollable Zn2+ diffusion. On the contrary, Zn deposits from β-CD-containing Zn(OTf)2 electrolyte (Fig. 3c, d) generally demonstrate the uniform and flat morphology with a dense deposition layer (~16 µm, Fig. 3g, h and Supplementary Fig. 29c, d). Such promoted Zn deposition behavior is benefited from β-CD@OTf--enrich IHP and thus-derived organic-inorganic hybrid layered SEI (Fig. 3c. d). X-ray diffraction (XRD) has been also further performed to investigate the Zn deposition microstructure. As shown in Fig. 3i, the Zn deposition layer harvested from β-CD-containing electrolyte exhibits a strength increased optimal deposition surface (002). Quantitatively, the ratio of peak (002) intensity to peak (100) intensity increases from 0.45 to 1.89, which indicates that the β-CD additive remarkably strengthen Zn (002) crystalline plane deposition. We attribute this phenomenon to two reasons: 1. Adsorbed β-CD molecules shield the protrusions on Zn surface, enhancing the orientation of Zn2+ deposition. 2. The inorganic ZnF2 component of SEI possesses high interfacial energy1, which is conducive to promoting the Zn2+ lateral growth rather than vertical growth.

Fig. 3: Impact of β-CD-Zn(OTf)2 and thus-derived SEI on Zn plating/stripping.
figure 3

Scheme of IHP interfacial chemistry in (a) Zn(OTf)2 and (c) β-CD-Zn(OTf)2 electrolyte, respectively. Adapted with permission from (ACS Nano, 2022, 16, 9667–9678). Copyright (2022) American Chemical Society. Schematic illustration of the effect of SEI formation on Zn deposition in (b) Zn(OTf)2 and (d) β-CD-Zn(OTf)2 electrolyte. Planar and cross-sectional SEM images of Zn anode after 20 cycles in (e, f) Zn(OTf)2 electrolyte and (g, h) β-CD-Zn(OTf)2 electrolyte. i XRD patterns of Zn deposits in Zn | |Zn symmetric cell. j CV curves of Zn | |Ti cells at the scan rate of 10 mV s−1 in varied systems. k The differential capacitance-potential curves for the copper (Cu) electrode at 0.66 V-0.46 V (vs. Zn/Zn2+). l Chronoamperometry curves of Zn | |Zn symmetric cells at overpotential of −150 mV.

Full size image

To systematically study the Zn-ion deposition behavior difference before and after addition of β-CD, a series of electrochemical characterization has been performed to study their interfacial electrochemical behavior. Firstly, the nucleation behavior of Zn2+ was studied through performing cyclic voltammograms (CV) characterization on Zn | |Ti asymmetric cell at a scan rate of 10 mV s−1. The Zn nucleation within β-CD-containing electrolyte exhibits a slightly extended overpotential than that of Zn(OTf)2 electrolyte ( | A’A | = 24 mV, Fig. 3j), which contributes to a refining Zinc nucleus with a promoted uniform Zn deposition. Note that the nucleation overpotential increase can also be ascribed to the competitive adsorption of β-CD on IHP, which can be served as a shielding layer to induce Zn homogeneous deposition to exclude H2O from IHP.

The alternating current (AC) voltammetry measurement at a range of 0.66–0.46 V was further put forward to characterize the adsorption behavior of β-CD. As displayed in Fig. 3k, the value of potential of zero charge (PZC) shifts from 0.624 V to 0.583 V when introducing β-CD additive, which indicates the EDL capacitance variation on Zn anode surface. It attributes to that the specific adsorption of β-CD substitutes part of Zn ion sites. In addition, chronoamperometry (CA) measurement was further employed to study the differences of Zn2+ diffusion behavior on Zn anode surface at an overpotential of −150 mV. As depicted in Fig. 3l, the current of the pure Zn(OTf)2 electrolyte demonstrates a continuous decrease within 400 s, indicating their uncontrollable 2D Zn2+ diffusion. In this regard, Zn2+ is prone to preferentially deposit at sites with a low surface energy, subsequent inducing an initial protrusion which is named as ‘tip effect’32,33. Supplementary Fig. 30b presents the deposited Zn recovered from the symmetric cells in pure Zn(OTf)2 electrolyte after failure, in which a large amount of dead Zn has been detected, which even grows in to a large size with intertwining glassy fiber from separator. Therefore, Zn deposits tends to penetrate into the separator with the formation of dead Zn due to the insufficient contact between the protrusions and the substrate. However, after forming stable anion dominated IHP with β-CD regulation, the Zn-ion diffusion behavior experiences a transition from 2D uncontrollable Zn2+ diffusion to 3D diffusion. Simultaneously, the responsive current of β-CD-Zn(OTf)2 electrolyte system tends to stabilize from ~100 s, with response current switching from 29.2 mA cm−2 to 6.4 mA cm−2.

Zn plating/stripping reversibility under high areal capacity

To explore the anion dominated IHP derived uniform Zn deposition effects on AZIBs, systematic electrochemical studies on both Zn | |Zn symmetric cells and Zn | |Cu asymmetric cells have been put forward to evaluate the Zn plating/stripping reversibility in 2 M Zn(OTf)2 with and without β-CD addition. First, the influence of addition amount on the Zn plating/stripping stability was estimated in Zn | |Zn cells. As depicted in Supplementary Fig. 31, the 2 M Zn(OTf)2 with 60 mM β-CD addition delivers the best cycling performance. When the addition amount surpassed 60 mM, the symmetric cells experienced a sharp electrochemical performance deterioration, with a significant voltage hysteresis increase. It could be attributed to the crowded effect of β-CD-OTf complex at Zn/electrolyte interface to severely increase the charge transfer resistance. As shown in Fig. 4a, the symmetric cell with 60 mM β-CD-Zn(OTf)2 electrolyte can sustain for 940 h cycles with the cumulative capacity of 4685 mAh cm−2 at a low current density of 5 mA cm−2 with an areal density of 2 mAh cm−2. In contrast, the cell with pure Zn(OTf)2 electrolyte only survived for 64 h before battery short circuit. To detect the affordability of the β-CD@OTf- derived SEI enabled AZIBs for extreme conditions, the long cycling performances of symmetric cells has been also characterized at an high current density of 30 mA cm−2 with a high areal density of 10 mAh cm−2, which corresponds to a Zn utilization rate (ZUR) of 34.2%. As shown in Fig. 4b, the symmetric cell with β-CD-containing electrolyte can sustainably cycle for 226 h with an astonishing cumulative capacity of 6800 mAh cm−2, which is more than 130 times of the pure Zn(OTf)2 systems (Supplementary Fig. 32). These excellent performances achieved under extreme areal capacity and current density are mainly attributed to the anion-dominated IHP derived organic-inorganic hybrid SEI and corresponding promoted Zn plating/stripping stability without obvious noxious volume expansion or detrimental water molecules triggered side-reactions. To rationally compare the Zn deposition stability, we introduced the concept of “life factor” to systematically evaluate the cyclic performance under a high areal capacity. The “life factor” is defined as cumulative capacity divided by Zn total capacity, in which the Zn total capacity equals to areal capacity/ZUR. As summarized in Fig. 4c and Supplementary Table 6, the current β-CD regulated electrolyte system demonstrates a high life factor of 242, which prominently outperforms the previously reported Zn metal reversibility under harsh conditions of high areal capacities5,13,14,30,34,35,36,37,38,39,40, the calculation formula of life factor are as follows: \({{\rm{Life}}}\,{{\rm{factor}}}=\frac{{{\rm{Cumulative}}}\,{{\rm{capacity}}}}{{{\rm{Total}}}\,{{{\rm{Zn}}}}^{0}},Total\,Z{n}^{0}=\frac{Areal\,capacity}{ZUR}\). The surface morphology of Zn anode has been also characterized for both pure Zn(OTf)2 and β-CD regulated electrolytes. As shown in Supplementary Fig. 34, obvious Zn dendrites and by-products have been observed on the cycled Zn electrode harvested from Zn(OTf)2 electrolyte, while the corresponding parasitic reactions were not discernable in β-CD-containing electrolyte. Even after Zn stripping, there are still no visible by-products remained on the electrode surface, mainly benefited from the effective inhibition of enhanced SEI on hydrogen evolution reaction (HER). As demonstrated in Supplementary Fig. 35a, the larger overpotential (43 mV) of HER has been detected with the addition of β-CD at the current density of 5 mV s−1, suggesting the prominent inhibition effect on HER. In addition, we examined the differences in H2 evolution among various electrolyte systems through differential electrochemical mass spectrometry (DEMS). As shown in supplementary Fig. 36, the flux of H2 evolution of the Zn | |Zn symmetric cell in pure Zn(OTf)2 electrolyte reaches 0.044 μmol min−1 after cycling for 390 min, which is obviously higher than the value of 0.03 μmol min−1 in the β-CD-Zn(OTf)2 system. This result intuitively demonstrated the inhibitory effect of β-CD on the HER. The cycled products with 50 cycles at a high areal density has been detected through X-ray diffraction (XRD). The peaks at 6.5°, 10.1° and 11.2° which are assigned to by-product of Znx(OTf)y(OH)2x-y·nH2O, which are easily to be found on Zn anodes in pure Zn(OTf)2 electrolyte (Supplementary Fig. 35b). Conversely, the Zn anodes recovered from β-CD-containing electrolyte presents the consistent peaks with original Zn foil, which verifies the prominent effect of formed organic-inorganic SEI on suppressing by-products during Zn plating/stripping.

Fig. 4: Reversibility of Zn plating/stripping under extreme conditions.
figure 4

Galvanostatic Zn plating/ stripping in a Zn | |Zn symmetric cell at (a) 5 mA cm−2 and 2 mAh cm−2 and (b) 30 mA cm−2 and 10 mAh cm−2. c The comparison of “life factor” between this work and previous reported works5,13,14,30,34,35,36,37,38,39,40. d Zn plating/ stripping coulombic efficiency (CE) at 15 mA cm−2 and 14 mAh cm−2 with 47.8% ZUR (50 µm Zn foil) in different electrolytes. e Zn plating/ stripping CE at a low rate of 5 mA cm−2 and 14 mAh cm−2 in Zn | |Cu asymmetric cells. f Schematic and digital photo of Zn | |Zn symmetric pouch cell. g Long cycling performance of corresponding Zn | |Zn pouch cells in different electrolytes at 20 mA cm−2 and 10 mAh cm−2 with the ZUR of 34.2%.

Full size image

To further verify the feasibility of the thus-designed organic-inorganic hybrid SEI, the Zn | |Cu cells assembled with 50 µm thickness Zn foils were cycled at the high current density of 15 mA cm−2 and high areal capacity of 14 mAh cm−2 (ZUR = 47.8%). The slight overpotential increase of Zn | |Cu cell with β-CD-Zn(OTf)2 electrolyte compared with that of Zn(OTf)2 electrolyte is ascribed to the insolated organic-rich outer layer, which aggrandize the interfacial resistance for electrochemical Zn deposition and stripping (Supplementary Fig. 37). As depicted in Fig. 4d, the cell with β-CD-Zn(OTf)2 electrolyte maintained 278 cycles until short circuit, which is equivalent to a large cumulative capacity of 7280 mAh cm−2. Conversely, the ZIB with pure Zn(OTf)2 electrolyte can only survive 2 cycles with a cumulative capacity of 56 mAh cm−2. Furthermore, as areal capacity increased to 28 mAh cm−2 with a 60 µm Zn foil and 80% ZUR = 80%, β-CD-Zn(OTf)2 electrolyte based ZIB could still sustain 24 cycles with delivering a cumulative capacity of 1300 mAh cm−2, while the cell with pure Zn(OTf)2 electrolyte is almost incapable to operate (Supplementary Fig. 38a, b).

Supplementary Table 7 summarized the latest progress in Zn asymmetric cells with respect to electrolyte additive and SEI regulation strategy. The current β-CD additive regulated SEI strategy tends to demonstrate the highest cumulative capacity even at a high ZUR of 80% and an areal capacity of 28 mAh cm−2, which is far more exceeding the recently reported works for AZIBs with electrolyte additive and SEI design5,13,31,41,42,43,44. The surface morphology in Supplementary Fig. 38c, d of Zn electrodes from cycled Zn | |Cu cells also indicated the discrepancies of electrochemical performances. The SEM images (Supplementary Fig. 38c, d) indicates that the regular crystalline and smooth surface morphology for deposited Zn of β-CD regulated system is completely different from the microstructure of the loose-nest like dead Zn mixed with massive separator glass fibers. Impurities and side-reaction products tends to accumulate at low current density with the influence of side reactions such as HER and parasitic reaction corrosion. As depicted in Fig. 4e, the relatively low current density of 5 mA cm−2 was applied to further investigate the Zn reversibility, which presents a 99% coulombic efficiency (CE) within 2 cycles, while stabilized at 98.5% for subsequent cycles, indicating that organic-inorganic SEI protection layer could form in a short time to effectively suppress side reactions at the Zn/electrolyte interface.

To verify its sustainability in practical application scenarios, the calendar life is indispensable. The calendar life mainly measures the cycling performance of batteries under intermittent usage conditions. Hence, we tested the lifespan of Zn | |Zn symmetric cells with the intermittent rest of 40 h (Supplementary Fig. 39). Distinctly, the symmetric Zn | |Zn cells with β-CD-Zn(OTf)2 electrolyte exhibit a long cycling performances over 920 h during intermittent charge/discharge process, which is approximately 51 times that of pure Zn(OTf)2 electrolyte.

The another way to examine practicality is assembling pouch-type cells. As shown in Fig. 4f, the electrochemical performance has been studied in a Zn | |Zn pouch cell with an area of 21.32 cm2 (5.2 cm \(\times\) 4.1 cm), even at an extreme condition of 20 mA cm−2 current density and areal capacity of 10 mAh cm−2. Being beneficial from the organic-inorganic hybrid SEI protection and inhibited H2O activity, the thus-derived Zn | |Zn pouch cell can stably operate for 320 h with delivering a cumulative capacity of 6430 mAh cm−2 at a relatively high ZUR of 34.2%, which is far surpassing the 90 mAh cm−2 of the pure Zn(OTf)2 electrolyte counterpart (Fig. 4g). This splendid performance exhibits the remarkable competitiveness in numerous reporting works about Zn metal anode6,9,43,45 (Supplementary Table 8). Besides, we also have examined the cycling ability of Zn | |Zn pouch cell at the relative low current density and areal density (Supplementary Fig. 40-41). Even at a low current density of 1 mA cm−2 and 1 mAh cm−2, the Zn | |Zn symmetric pouch cell with β-CD-Zn(OTf)2 electrolyte still can sustain 680 h cycles, while the Zn | |Zn pouch cell with pure Zn(OTf)2 electrolyte only survives for 3 h (Supplementary Fig. 40). For CE test, the Zn | |Cu pouch cell in β-CD-Zn(OTf)2 electrolyte can maintain 213 cycles with average CE up to 99.17%, which is far exceeding that of Zn(OTf)2 system (Supplementary Fig. 41), demonstrating the distinguishing Zn anode electrochemical reversibility with large size configuration according to the β-CD triggered robust SEI enhanced mechanical and electrochemical stability.

Electrochemical performance of Zn | |ZnxV2O5 full cell with Low N/P ratio

Ultimately, to further verify its practical application as a full AZIB, the Zn | |ZnxV2O5 full cells with thin Zn anode (10 µm) as anode with a high mass loading (6.7 and 11 mg cm−2) ZnxV2O5 as cathode has been assembled to evaluate the β-CD regulated SEI effects on the electrochemical performance under extreme conditions such as low N/P ratio (Supplementary Figs. 43–45). As displayed in Fig. 5a, the Zn | |ZnxV2O5 full cell with limited Zn reservation (N/P = 3.4) in β-CD-containing electrolyte can sustain 850 cycles with 90.4% capacity retention, while the full cells with pure Zn(OTf)2 electrolyte experienced a sharp capacity decline and CE fluctuation (Supplementary Fig. 43–44) due to the aggravated parasitic reactions. To further investigate the full cell in a close to real operating condition, the full cell was assembled with a high ZnxV2O5 cathode mass loading of 11 mg cm−2 and an ultra-thin Zn anode of 10 µm, which corresponds to the N/P ratio of 2.1 (47.6% ZUR) (Fig. 5b and Supplementary Fig. 44–45). Apparently, as shown in Fig. 5b, the Zn | |ZnxV2O5 full cell with β-CD-Zn(OTf)2 electrolyte exhibits a high initial areal capacity of 1.39 mAh cm−2, which still retains a capacity of 1.26 mAh cm−2 after 600 cycles. As summarized in the comparison curve of Fig. 5c and Supplementary Table 91,41,42,46,47,48,49,50,51,52,53, compared with previous reported works with a high N/P ratio and low ZUR, the thus-assembled full cell demonstrates the highest cumulative capacity of 740 mAh cm−2 even at the lowest N/P ratio of 2.1. Such excellent performance of the Zn | |ZnxV2O5 full cell with β-CD-Zn(OTf)2 electrolyte under extreme conditions unambiguously demonstrates the positive impacts of the β-CD triggered organic-inorganic hybrid SEI for inhibiting nauseous Zn dendrites and side reaction products even under harsh operating conditions. The EIS plot and XPS spectra in supplementary Fig. 46 was further to verify the hybrid SEI formation in the full cell. As displayed in supplementary Fig. 46a, the evident semicircle in high-frequency region in EIS tests indicated that hybrid SEI was constructed successfully in full cells with β-CD-containing electrolytes, while the semicircle is imperceptible in pure Zn(OTf)2 electrolyte system. Besides, the abundant organic (C-O-C/*CF*/CF3 species) and inorganic components (ZnCO3 and ZnF2) was detected on the C1s and F1s XPS spectra (Supplementary Fig. 46b, c) of Zn anodes cycled in β-CD-containing electrolytes, further elucidating the formation of enhanced organic-inorganic hybrid SEI.

Fig. 5: Electrochemical performances of Zn | |ZnxV2O5 full cells under extreme conditions.
figure 5

a Long-term cycling performances at 5 A g−1 with N/P ratio of 3.4. b Long-term cycling performances at 5 A g−1 with N/P ratio of 2.1. c Comparison of cumulative capacity and N/P ratio1,41,42,46,47,48,49,50,51,52,53. d Structure diagram of assembled pouch full cell. e Cycling performance of the pouch cell Zn | |ZnxV2O5 with β-CD-Zn(OTf)2 electrolytes at the current density of 0.5 A g−1.

Full size image

To verify the practical application potential of β-CD-Zn(OTf)2 electrolyte, a pouch cell was assembled with Zn foil anode (50 µm, 29.2 mAh cm−2) and a high mass loading cathode (8 mg cm−2, 2.2 mAh cm−2), as schematically illustrated in Fig. 5d. As depicted in Fig. 5e, the pouch cell of Zn | |ZnxV2O5 with β-CD-containing electrolytes presents a high initial capacity of 42 mAh at 0.5 A g−1 and the prominent capacity retention of 90.6% after 100 cycles (Supplementary Fig. 50), validating the stable Zn reversibility in a practical pouch configuration.

In summary, a β-CD electrolyte additive triggered organic-inorganic hybrid layer-SEI structure regulation strategy has been developed to reach long stability of ZIB under extreme conditions, such as high current density, tremendous accumulative capacity and even low N/P ratio. The decomposition of β-CD@OTf- complexes at IHP derived inorganic inner layer (ZnF2/ZnCO3) and organic outer layer (C-O-C/*CF/*CF3 amorphous species) composed SEI has been systematically characterized through MD/DFT calculation and a series of experimental physicochemical/electrochemical characterizations, which verify its effects on stable reversibility via suppressing HER and Zn dendrite formation. As a result, both symmetric cell and full cell even at pouch cell configuration displayed excellent stability and accumulative capacity even at a high current density and low N/P ratio. This work shed lights on the importance of electrolyte additive effects on the IHP structure and corresponding SEI formation for reaching practical ZIB.

Methods

Preparation of electrolyte and electrodes

The 2 M Zn(OTf)2 electrolyte was prepared by dissolving 2 mol Zn(OTf)2 powder in 1 kg deionized (DI) water. The β-CD-Zn(OTf)2 electrolytes were formed by adding varied amounts of β-CD (30 mM, 60 mM, 100 mM, 150 mM) powder into 2 M Zn(OTf)2 solutions. The β-CD aqueous solution was made through dissolving certain β-CD powder into DI water. The final optimized addition amount is 60 mM which is named as “With β-CD”. The synthesis of ZnxV2O5·H2O was depended on hydrothermal method. Specifically, 0.364 g V2O5 and 0.0745 g Zn(NO3)2·6H2O were primarily dissolved in 60 mL DI water under continuous stirring, subsequent adding 2 mL H2O2 dropwise into the as-mixed solution until the bubbles disappear. Ultimately, the obtained orange solution was packed into 100 mL Teflon-lined autoclave and reacted at 120 °C for 12 h. The blackish green product was obtained after washing and freeze-drying. The ZnxV2O5·H2O cathodes utilized in this work were consisted of 70 wt% ZnxV2O5·H2O, 20 wt% Super P and 10 wt% polyvinylidene fluoride (PVDF). The current collector was carbon paper. The active mass loading of ZnxV2O5·H2O electrodes for normal test is ~3 mg cm−2, and the high mass loading of ZnxV2O5·H2O cathodes is 6.7 mg cm−2 (1.7 mAh cm−2) and 11 mg cm−2 (2.8 mAh cm−2), respectively. The purchased Zn foil was polished to remove extra oxides to obtain bare Zn. Subsequently, polished Zn foils were stamped into small circular plates, which were directly used as Zn electrodes.

Cell assembling

The Zn | |Zn symmetric cell was assembled with Zn foil (50 µm, 29.28 mAh cm−2) as anode and cathode simultaneously and glass fiber (GF) separator. The diameter of Zn electrode is 13 mm, the diameter and thickness of GF separator is 19 mm and 260 μm, respectively. The Zn | |Cu asymmetric cell was fabricated with Zn foil as anode and copper (Cu) foil as cathode. For Zn | |Zn symmetric pouch cell, the Zn foil (50 µm, 29.28 mAh cm−2) was directly served as anode and cathode with an area of 21.32 cm2 (5.2 cm × 4.1 cm), a piece of glass fiber immersed in 3 mL electrolyte as separator. The Zn | |ZnxV2O5 H2O pouch cell was assembled with Zn foil (50 µm, 29.28 mAh cm−2) and ZnxV2O5·H2O cathode (8 mg cm−2), the pole ear is Ti foil. The mass percentage of active materials is 70% wt.

Characterization

The properties and structure of electrolytes were analyzed through Raman (LabRAM HR800 (JobinYvon)) with laser of 532 nm, Fourier-transform infrared spectroscopy (FT-IR) (Bruker Tensor 27), and nuclear magnetic resonance (NMR) spectroscopy (400 SB Bruker Avance III spectrometer). The morphology of Zn electrodes was detected on SEM (FEI Scios) and TEM (JEOL JEM-F200) equipped with an energy-dispersive X-ray spectroscopy detector. The X-ray photoelectron spectroscopy (XPS) was performed on Thermo Scientific ESCA Lab 250Xi to characterize interfacial structure, etch rate is ~0.24 nm/s. X-ray diffraction (XRD) patterns was collected on Bruker D8 advanced diffractometer with Cu-Kα radiation (λ = 1.5406 Å). TOF-SIMS (Germany, TOF-SIMS5) was utilized to explore the depth distribution of SEI.

Electrochemical tests

The basic electrochemical tests were performed on workstation (CHI 660e), and the long-cycling tests were recorded on Neware battery testing system (CT-4008T). Specifically, electrochemical impedance spectroscopy (EIS) was employed with amplitude of 5 mV, the range of frequency was 0.01 Hz to 10 0000 Hz. The cyclic voltammetry (CV) of Zn | |Ti cells was tested at a scan rate of 10 mV s−1 from potential of -0.4 V to 1.2 V. The differential capacitance-potential curve of Zn | |Cu cell was obtained from current voltammetry (AC) test at a scan rate of 10 mV s−1 with frequency of 6 Hz and phase angles of 0°and 90°, the voltage range is from 0.46 V to 0.66 V. The capacitance was calculated following the equation: \(C={\left(2\pi f{Z}_{{im}}\right)}^{-1}\), where \({Z}_{{im}}=Z\sin \varnothing\), \(\varnothing={ar}\tan \frac{{i}_{90}}{{i}_{0}}\), \(Z=\frac{A}{\sqrt{{{i}_{90}}^{2}+{{i}_{0}}^{2}}}\), \({i}_{90}\) and \({i}_{0}\) are the currents at the phase angles of 0 and 90, respectively54. The cut-off voltage of coulombic efficiency (CE) test is set at 0.5 V.

Model building

To calculate adsorption energy of α-cyclodextrin, β-cyclodextrin and γ-cyclodextrin on Zn(002), Zn(002) slab was modeled with 9×9×4 unit cell, and a vacuum thickness was set 20 Å to remove any interaction between periodic slab, the two bottom layers were fixed when calculated. And α-cyclodextrin, β-cyclodextrin and γ-cyclodextrin were obtained from CCSD.

In order to build 2 M Zn(OTf)2 electrolyte consistent with experiment, the system includes 301 Zn2+, 602 OTf- and 4818 H2O in a 5.0 \(\times\) 5.0 \(\times\) 10.0 nm3 cubic box. The initial structure is created by GROMACS tool55,56,57 and PACKMOL58,59.

In order to build 2 M Zn(OTf)2 (β-cyclodextrin) electrolyte consistent with experiment, the system includes 301 Zn2+, 602 OTf-, 9 β-cyclodextrin and 4459 H2O in a 5.0 \(\times\) 5.0 \(\times\) 10.0 nm3 cubic box conformed to experimental concentration of adding additive. The initial structure is also created by GROMACS tool and PACKMOL.

Simulation methods

DFT calculations

Calculations on the absorbed energy of various cyclodextrin are performed using Vienna Ab initio Simulation Package (VASP) at a version of 5.4.4 with the projector augmented wave (PAW) method and a plane wave basis set. The method is density functional theory (DFT) with generalized gradient approximations (GGA) of Perdew-Burke-Ernzerhof (PBE) functional. A dispersion correction, DFT-D3 method with Becke-Jonson damping, was included in the calculations. The energy cut-off is set to 400 eV. Larger energy cut-off does not produce more accurate prediction basing on our benchmark calculation. Reciprocal space was sampled by Γ-centered Monkhorst-Pack scheme with a grid of 1 × 1 × 1. Spin polarization does not have an appreciable effect on the overall energies and is not included in the calculations to reduce computational demands. The partial occupancies for each orbital are set with the first order Methfessel-Paxton scheme in the smearing width of 0.2 eV. The dipole moment corrections for the total energy are considered in the direction normal to the surface. The self-consistent electronic step is considered converged when the change of total energy and eigenvalues change between two steps are both smaller than 1e−5 eV.

A conjugate-gradient algorithm is used to relax the ions in energy minimization. The minimization was considered converged when all the atomic force are smaller than 0.05 eV/Å, where convergence to an energy minimum.

The cluster density functional theory (DFT) calculation is at the level of B3lyp/6-31 G (d, p) as implemented in Gaussian 16. A geometry optimization is applied to fully relax the molecule structure. The universal continuum solvation (SMD) is used to consider the solvation effect implicitly. Electrostatic potential (ESP) is derived from electronic density from DFT calculation fitted using Multiwfn60 (Version 3.7). All the atomic structures and ESP were visualized and rendered using VMD61 (Version 1.9.3). Multiwfn is also used to provide the analysis of independent gradient model (IGM)62,63 to quantify covalent and noncovalent interactions.

The periodic DFT calculation is at the level of Perdew-Burke-Ernzerhof (PBE) with a plane wave basis set as implemented in VASP (Version 5.4.4). The energy cut-off is set to 400 eV. The smearing width of 0.2 eV is utilized to set the partial occupancies for each orbital via the first-order Methfessel-Paxton (MP) scheme. Only gamma point is considered in the calculation. The DFT-D3 method with Becke-Jonson damping was included to consider for dispersion correction. The corrections for dipole moment in z-axis (the direction normal to the surface) are considered when calculating the total energy. The self-consistent electronic step is considered converged when the change of total energy and eigenvalues change between the two steps are both smaller than 1e−5 eV. For energy minimization, the ions are relaxed utilizing a conjugate-gradient (CG) algorithm. The convergence of the minimization was deemed complete when the magnitude of each atomic force was less than 0.02 eV/Å. The solvation effect is considered with a linearized Poisson-Boltzmann implicit solvation model as implemented in VASPsol. Charge density difference was analyzed using VASPKIT64,65 (version 1.4.0). The simulation model of periodic DFT calculation consists of a 9×9×4 Zn (002) slab with one β-CD molecule placed on the surface.

MM-MD simulation

The force field of water molecules is modeled by SPC/E66. And the force field of β-cyclodextrin was OPLS-AA and optimized. In the initial force field, the charge of Zn cation is +2, and the van der Waals (van der Waals, vdW) parameter is from UFF (Universal Force Field, UFF). The charge of OTf- anion is −1, and the vdW parameter is also from UFF force field. The intramolecular parameters of OTf- and β-cyclodextrin were obtained by fitting the results of DFT (Density Functional Theory, DFT) calculations. The final FF parameters of Zn2+ and H2O are used in ref. 67.

GROMACS was selected as simulated tools55,56, simulated process includes energy minimization, 2 ns NPT and 1 ns NVT, and timestep and temperature was 1 fs and 298 K respectively, and the cutoff radius for coulomb interaction and vdW is 0.9 nm. The long-range coulomb interaction is calculated by PME method68 and long-range correction was taken into vdW energy and pressure. The single layer graphene was employed position restraint for keeping the electrode stable.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.