Supramolecular Polymer Intertwined Free-Standing Bifunctional Membrane Catalysts for All-Temperature Flexible Zn–Air Batteries

Highlights Rational design of 3D flexible free-standing membrane catalysts with protonation between supramolecular polymers and nitrogen-deficient carbon nitride nanotubes frameworks via facile bottom-up self-conversion approach. PEMAC@NDCN demonstrates the lowest reversible oxygen bifunctional activity of 0.61 V with exceptional long-lasting durability, surpassing the commercial and reported champion oxygen catalysts. Engineering catalytically active cathode materials enabled all-temperature flexible Zn–air batteries with high electrochemical and mechanical performances (temperature of − 40 to 70 °C) under harsh operations. Supplementary Information The online version contains supplementary material available at 10.1007/s40820-022-00927-0.

Performed current densities were also referenced according to the measured geometric surface areas. Electrochemical impedance spectra were obtained for the frequency range of 100 Hz to 1 MHz with a constant bias of 0.2 V. The long-life durability of the catalysts was measured by continuous potentiodynamic sweeps for a scan rate of 100 mV s -1 . The mass loading of membrane catalysts and reference Pt/C/RuO2 has been placed identically unless otherwise stated (0.15 mg cm -2 ).

S1.4 Oxygen Reduction Reaction Kinetics
The ORR kinetics were evaluated by using different rotational speed LSV profiles. The measured total current density is the sum of inverse of kinetic current (JK) and diffusion current (Jd). Every atom or ion on the electrode reacts immediately as the applied overpotential is sufficiently high. The number of oxygen molecules at the electrode surface is almost zero, facilitating diffusion-limiting plateau. Therefore, the diffusion current is related only to the RDE rotational speeds.
The transferred electron number (n) in oxygen reduction was determined according to the Koutecky-Levich (K-L) equation [S1-S3]: Nano-Micro Letters S3/S32 B = 0.2nF(D O 2 ) 2/3 ν −1/6 C O 2 (S3) where B represents the Levich slope, JK represents the kinetic current, J represents the measured total current, represents the electrode rotation rate, n represents the number of electrons transferred for each oxygen molecule, F represents the Faraday constant (F = 96485 C mol -1 ), DO2 represents the O2 diffusion coefficient in 0.1 M KOH (DO2 = 1.9×10 -5 cm 2 s -1 ), ν represents the kinetic viscosity (0.01 cm 2 s -1 ) and CO2 represents the concentration of O2 (CO2 = 1.2 ×10 -6 mol -1 cm -3 ). The considered rotation speeds are in rpm, and therefore, the constant factor 0.2 is multiplied. The peroxide species during ORR reactions were determined by measuring the RRDE polarization profiles for the ring potential of 1.3 V vs. RHE. Based on the following expressions, the transferred electron number (n) and peroxide (H2O2) yield were evaluated as [S4]: n = 4 I d I d + I r /N (S4) 2 − (%) = 100 2I r /N I d + I r /N (S5) where Ir represents the ring current, Id represents the disk current, and N represents the current collection efficiency of the Pt ring. N was determined to be 0.42.

Alkaline Zn-air batteries
The rechargeable Zn-air battery performance was analyzed on home-built electrochemical cells. Air cathodes were constructed by uniform punching of PEMAC@NDCN membrane (1 cm × 1 cm) and carbon paper was utilized as current collectors. Further, for comparison, Pt/C + RuO2 catalyst slurry was fabricated by mixing carbon black, polytetrafluoroethylene, and the catalysts (1:1:8 w/w) in ethanol/Nafion solution. The mass loadings of the PEMAC@NDCN and reference Pt/C+RuO2 catalysts for rechargeable ZABs was 1 mg cm -2 . Mass ratio of reference Pt/C and RuO2 was of 1:1. Here, 6 M KOH and 0.2 M zinc acetate was used as the electrolyte for the reversible electrochemical reactions. Catalyst-loaded (Pt/C+RuO2) carbon paper was used as an air cathode and polished Zn plate (0.5 mm thickness) as anode.

Flexible solid-state Zn-air batteries
Flexible Zn-air batteries were fabricated by including PEMAC@NDCN membrane catalysts as the air cathodes, the chitosan biocellulosics (CBCs fabricated as per our previous reference 1 ) as the solid electrolyte (50 μm in thickness), and Zn foil (0.3 mm) as the anode. Then, membrane catalysts and zinc foils were positioned on opposite sides of the bio-cellulose membrane electrolyte, without any external current collector. Finally, the assembled devices were pressed cautiously and encapsulated with sustainable latex. For comparison, Pt/C + RuO2 catalyst slurry was also fabricated by mixing carbon black, polytetrafluoroethylene, and the catalysts (1:1:8 w/w) in ethanol/Nafion solution. The mass loadings of the PEMAC@NDCN and reference Pt/C+RuO2 catalysts for rechargeable ZABs was 2 mg cm -2 and electrode area of 2 cm 2 .

Battery testing
All the fabricated ZABs were evaluated under atmospheric conditions. The galvanostatic discharge and charge voltage profiles were conducted on a LAND CT2001A multichannel battery testing system. The cycling for alkaline and flexible solid ZABs was performed for 10 min per cycle (discharge: 5 min; charge: 5 min) with 20 and 50 mA cm −2 current density, respectively. The specific capacities were determined using the galvanostatic discharge profiles standardized to the consumed mass of Zn. The energy efficiency was calculated from the ratio of discharge to charge voltages. The power densities of both ZABs were calculated by expression as P = V × I. Figures   Fig. S1 Spectroscopic characterizations. Fourier transforms IR spectra for PEMAC@NDCN compared to those of PVA@NDCN, PAA@NDCN, NDCN, and bulk CN The molecular binding structures for PEMAC@NDCN compared to those of PVA@NDCN, PAA@NDCN, NDCN, and bulk CN are evaluated by FTIR spectra. The characteristic striazine ring structure breathing mode was observed for 800-891 cm -1 for bulk CN, which illustrates carbon nitride frameworks [S5]. Furthermore, the decreasing intensity of s-triazine peaks for NDCN, PEMAC@NDCN, PVA@NDCN, and PAA@NDCN confirms the formation of nitrogen deficiency and loading of polymeric layers. The peaks from 1100-1600 cm -1 are attributed to the stretching bindings of C-N, C=C, and C=N, which demonstrates the existence of pyridinic, pyrrolic, and graphitic nitrogen species, as evident with XPS N 1s spectra [S6]. Characteristic broad bindings over 3000-3500 cm -1 corresponds to the C=C-H, C-H, stretch, amine or water bindings [S7]. Presence of electron-donating species such as ortho, para and meta phases in polymer-based NDCN structures, which suggests the enhancement for charge transfer during electrochemical reactions. The stretching vibrations for 1418 and 1327 cm -1 were observed conforming to the anti-symmetric and symmetric carboxylic anions and C=C, C=O, C-H bond structures, which indicates the formation of PEMAC@NDCN.S8 PEMAC@NDCN, PVA@NDCN, and PAA@NDCN illustrates the weak binding for ~2181 cm -1 , which clarifies the stretching vibrations for N=C=N and N=C=O. This illustrates the presence of pyridinic N + Ospecies as evident in XPS results [S9]. The variation in the C-N or C=C stretching bindings compared to those bulk CN illustrates the construction of polymeric CN structures. The presence of peaks at 1632 and 1689 cm -1 corresponds to the formation of C-O or C-O-H groups from the polymeric chain. The existence of C=C, C=O, OH, C-H illustrates the negatively charged polymeric moieties attached to the surface of NDCN, which is critical for enhancing electrochemical kinetics. FTIR characterization revealed the formation of polymer-derived metal-free NDCN frameworks.       We performed a detailed analysis of the OER/ORR overpotentials (η OER/ORR ) by constructing free energy diagrams (FEDs) under different electrode potentials (U=0 V, 0.402 V and η V) (Supporting Information Equation S6-S21). From the FEDs for bulk CN, NDCN and polymer-assisted NDCN structures in Figs. S14 and S15, we can deduce the η OER/ORR corresponding to the largest endothermic reaction Gibbs free energy change (ΔG) under the equilibrium potential (U=0.402 V in alkaline media), denoted by green lines in Figs. S14-S15. Applying an additional η OER/ORR to overcome the potential-determining step (G OER/ORR ), we can finally get an overall downhill reaction represented by blue lines in FEDs. The calculated results reveal that the PEMAC@NDCN has the lowest overpotentials (η OER/ORR ) of 0.28 and 0.38 V, respectively. In addition, these values represent much better catalytic activities for conventional Pt (0.45 V) and RuO2 (0.42 V).        Note the "a" and "b" represents the XPS and EDS results Nano-Micro Letters S17/S32    Co-FPOH 600 s/cycle for 1200 cycles; 450 h@5 mA cm -2 [S16] Co@IC/MoC@PC 1200 s/cycle for 300 cycles; 100 h@1 mA cm -2 [S17] Nitride/N-Ti 3 C 2 3600 s/cycle for 120 cycles; 120 h@20 mA cm -2 [S18]

S4 Note S2: Computational Details and Modeling
All ab initio calculations were performed with the Vienna Ab initio Simulation Package (VASP 5.4.4) [S81-S84]. We used the BEEF-vdW [S85] exchange-correlation functional using the projector augmented wave (PAW) method [S86, S87] with a generalized gradient to accurately describe the chemisorption as well as physisorption on the catalyst surface. Integration in the Brillouin zone was performed based on the Monkhorst-Pack scheme using a Γ-centered 3 × 2 × 1 k-point mesh in each primitive lattice vector of the reciprocal space for geometric optimization. Plane-wave cutoff energy of 500 eV was used. Lattice constants and internal atomic positions were fully optimized until the residual forces were less than 0.04 eV Å -1 . The vacuum slab space of a unit cell in the z-direction was set to 16 Å to avoid interactions between layers. The schematics of our models are shown as Figs. S12-S13. For the convenience of identifying the active site positions, we name them by element name and Arabic number, and the details of the active site naming are included in Figs. 7a and S12.

S4.1 OER and ORR Reaction Pathways
In this work, we used the theoretically well-defined free energy diagram (FED) approach proposed by Norskov group. It has been a generally accepted approach for electrochemical studies based on the standard density functional theory (DFT) in combination with the computational standard hydrogen electrode (SHE) model. For the investigation of various catalytic reactions on specific surface structures, we mainly consider the thermodynamic stabilities of the intermediates as the main descriptor, which determines the catalytic performance [S88, S89].
In the case of oxygen evolution reaction (OER), the four-step four-electron reaction has been generally acceptable, where this process in an alkaline environment can be described as follows.

S4.2 Derivation of the Free Energy Relations
We calculated the reaction Gibbs free energies of the intermediates of O2*, OOH*, O* and OH* on the bulk CN, NDCN, and polymer-assisted NDCN to determine the potentialdetermining step of OER and ORR, considering all possible active sites. For each step, the reaction Gibbs free energy ΔGads can be expressed by ΔGads = ΔEads + ΔZPE -TΔS (S15) where ZPE is the zero-point energy, T is the temperature, and ΔS is the entropy change. Using this equation (S15), we can construct a free energy diagram (FED) considering the following four-step four-electron reaction with equations at standard conditions [S90].

S4.3 Free Energy Diagram (FED) and Overpotential (η)
From the calculated ΔG OER/ORR values, we can deduce a critical parameter of electrocatalytic activity, which is the magnitude of the OER and ORR potential-determining step (G OER/ORR ) in a four-step four-electron reaction. It is the specific reaction point with the largest ΔG in the OER and ORR elementary reaction steps; i.e., the concluding step to achieve an overall downhill reaction in the free energy diagram (FED) with increasing potential (Figs. S14 and S15): After finding the largest ΔG value meaning the bottleneck point of ORR, we can finally get the theoretical overpotential in the alkaline condition in the following equation (