Materials. 2,5-Dimethylpyrrole (DMP, 98%), dipyrromethane (DPM, 98%), tetraphenylporphyrin (TPP, 99%), anhydrous AlCl3 (99.99%), 1-ethyl-3-methylimidazolium chloride (EMImCl, 98%) tetraphenylporphinesulfonate (TPP-SO3−, 95%), polyaniline (PANI, 98%) were purchased from Shanghai Macklin Biochemical Co., Ltd. ZnSO4·7H2O and phenyl magnesium chloride (PhMgCl) solution were purchased from Shanghai Aladdin Biochemical Technology Co., Ltd. Tetrahydrofuran (THF) was purchased from DodoChem Co., Ltd. Aluminum foil, zinc foil, magnesium foil, titanium foil and tantalum foil were bought from Runde Metal Products Co., Ltd. All metal foils were polished by fine-grained sandpaper, and the thickness of metal foils were controlled by a roller press (MTI Corporation). The flake graphite powder (99.9%) was purchased from Energy Chemical Co., Ltd. The aqueous CNT suspension and PTFE binder for high mass loading cathodes was purchased from Canrd New Energy Technology Co., Ltd.
Preparation of electrolytes. Preparation of [EMIm]Cl/AlCl3 electrolyte: anhydrous AlCl3 was slowly added into [EMIm]Cl slowly with continuous stirring. The molar ratio [EMIm]Cl: AlCl3 = 1:2. The ionic liquid was treated with Al foil at 90°C for 6 h to remove hydrochloric acid in the electrolyte.
Preparation of (PhMgCl)2/AlCl3 electrolyte: AlCl3 powder was slowly added into PhMgCl THF solution in an Ar atmosphere, and its concentration was 0.25 M.
Preparation of 2 M ZnSO4/CuSO4 electrolyte: ZnSO4·7H2O or CuSO4·5H2O was dissolved in deionized water to prepare 2 M electrolyte. Deionized water was obtained from Milli-Q water purification system (18.2 MΩ cm− 1).
Material characterizations.
Field-emission SEM measurements were conducted on Hitachi SU8000 equipped with an EDX detector. Pole figures were obtained at room temperature on a D8 ADVANCE Da Vinci with CuKα radiation. TEM sample was prepared by a dual-beam focused ion beam (FIB)-SEM (FEI Quanta 3D FEG) with a nanomanipulator using a standard lift-out procedure. Platinum was deposited for surface protection from the Ga ion beam. The initial lamella was milled, cleaned, lifted out and mounted on a TEM grid. Thinning was first conducted with a beam current of 500 pA, and then final polishing with a beam current of tens of pA to remove amorphization and Ga implantation. TEM measurements were performed on JEOL 2100F with Gatan double tilt analytical holder to identify the crystallography of electrodeposits.
Afm-ir Spectroscopy Measurements
The local nanoscale IR spectra and absorption mappings were carried out on a NanoIR2-fs instrument (Anasys Instruments Inc.) under contact/tapping mode. These AFM-IR measurements are based on the photothermal-induced resonance of the AFM probe (PR-EX-Nir2-10). The resonant amplitudes of the AFM cantilever are positively correlated with the absorbed IR radiation on samples. The AFM-IR spectra and 3D intensity mappings were analyzed using Analysis Studio software for the NanoIR2 system. The equipped AFM tip for KPFM measurements was Pt/Ir coated (25 nm thick) with ~ 60 kHz resonance frequency (PR-EX-KPFM-5). Surface potential images were acquired at a scan rate of 0.8 Hz and tip lift height was 20 nm under tapping mode.
Preparation Of High Mass Loading Cathodes
High mass loading graphite cathodes were fabricated by a tape-casting method. The aqueous mixture of CNT suspension and flake graphite powder (2:8 by weight) was sonicated for 1 h. The slurry was then cast onto polyethylene terephthalate (PET) foil using a doctor blade, then slowly dried at 40°C for 8 h, followed by vacuum drying at 100°C for 8 h. and then vacuum infiltrated. The freestanding cathodes were dried at 80°C for 12 h. The freestanding cathodes can be peeled off form the PET substrate after drying. High mass loading PANI cathodes were fabricated by a dry electrode method. The mixture of PANI, Ketjenblack and PTFE (8.5:1:0.5 by weight) was grinded, placed between two PET substrates and rolled out. The freestanding cathodes were dried at 100°C for 8 h.
Electrochemical measurements.
All Al metal batteries (symmetric Al batteries, Al/Ti batteries and Ti/C full batteries) were assembled in an argon-filled glovebox with H2O and O2 content below 0.3 ppm using pouch-type batteries. Glass fiber membrane (934-AH, Whatman) was used as separator. LAND battery cycler and VMP3 potentiostat/galvanostat (Bio-Logic) were used for the electrochemical measurements. Coulombic efficiency was measured in multivalent metal/Ti cells. A fixed amount multivalent metal (e.g., Zn. Mg Al) was plated on Ti electrode and then stripped back under different current densities until the cutoff voltage increased to 1 V. The symmetrical cells using 20 µm Al anode were cycled at current densities of 1 mA cm− 2 under fixed charge/discharge capacities of 3 mAh cm− 2. Long-term cycling lifespan tests of Ti/C full cells were cycled at 2 mA cm− 2 within a voltage range of 0.4 V-2.4 V versus Al3+/Al. Long-term cycle lifespan tests of Zn/PANI full cells were charged/discharged at 3 mA cm− 2 within a voltage range of 0.5 V-1.5 V versus Zn2+/Zn. Tafel plots were obtained from the symmetric cells between − 0.15 V and 0.15 V at a scan rate of 0.1mV s− 1.
Theoretical Calculations
All DFT calculations were performed with the Gaussian 16 software package 37. Geometry optimizations of all the minima and transition states were carried out at the B3LYP level 38, 39 of theory with additional Grimme’s D3 dispersion correction (Becke-Johnson damping)40 with the def2-SVP41 basis set. Vibrational frequencies were computed at the same level to evaluate its zero-point vibrational energy (ZPVE) and thermal corrections at 298 K, and to check whether the optimized structures are real intermediates. Single point calculations were carried out at the same level of theory with a larger basis set def2-TZVPP41, 42. The ESP surfaces were drawn with isovalue of 0.0004 a.u. based on the single point calculation at B3LYP-D3(BJ)/def2-TZVPP with VMD43.
Spin-polarized density functional theory (DFT) calculations were performed using the DMol3 code. The electron exchange-correlation potential was conducted by the Perdew-Burke-Ernzerhof (PBE) functional of generalized gradient approximation (GGA). In all calculations, a global orbital cutoff with the value of 4.4 Å was used for the numerical atomic orbital basis, and the convergence tolerances were set to be 1 × 10− 5 Ha for energy, 5 × 10− 3 Å for maximum displacement, 0.02 Ha/Å for maximum force, and 1.0 × 10− 6 eV/atom for SCF tolerance. The Al (200) surface is modeled by a three-layers surface supercell (75 Al atoms per layer) with only considering the γ point for saving the computational resources. The long-range dispersion correction for the van der Waals interaction was implemented through the DFT-D method in all calculations.