Enhancing the Cycle Life of Lithium‐Anode‐Free Batteries through Polydopamine‐Coated Substrates

Anode‐free lithium metal batteries (AFLMBs) show promise as a means of further enhancing the energy density of current lithium‐ion batteries, as they do not require conventional graphite anodes. The anode‐free configuration, however, suffers from inferior chemical stability of the solid electrolyte interphase (SEI) layer and experiences inhomogeneous lithium deposition during charge/discharge processes, resulting in rapid capacity fading. To address these issues, a carbonized polydopamine (CPD) coating is applied to the copper current collector. The CPD‐coated copper current collector promotes highly efficient and reversible lithium plating and stripping processes, resulting in a densely packed lithium deposition that significantly improves cycling stability. The anode‐free full cell, consisting of CPD‐coated copper current collector and a LiFePO4 cathode, demonstrates significantly improved electrochemical performance, with a capacity retention of more than 63% after 100 cycles at a current rate of 0.3C. The stability of the SEI layer and the presence of lithiophilic sites are verified through a range of techniques, including optical microscopy, Raman spectroscopy, X‐ray photoelectron spectroscopy, chronoamperometry, and electrochemical impedance spectroscopy. Based on these collective findings, it can be inferred that the use of CPD coating provides a simple way to enhance the electrochemical performance of AFLMBs.


Introduction
As energy demands continue to grow and the need to reduce CO 2 emissions from fossil fuels becomes more urgent, advanced energy-storage systems are becoming increasingly essential.
In recent years, rechargeable lithium-ion batteries (LIBs) have become the most prevalent power source for various electronic devices and electric vehicles.Generally, LIBs incorporate a graphite anode and inorganic cathode materials and have proven to be highly successful in energy-storage applications.[3][4][5] At present, a variety of approaches are being employed to achieve an even greater energy density, with many studies focusing on the use of multivalent ions, conversion-type electrodes, or the design of redox-flow batteries. [6]ince the first demonstration of LIBs in the 1970s, Li metal anodes have remained an attractive candidate due to their high theoretical capacity of 3860 mAh g À1 and low electrochemical potential of À3.04 V versus standard hydrogen electrode. [4,5]owever, lithium metal batteries are more susceptible to inherent safety risks and limited cycle life owing to the highly reactive nature of metallic lithium.During the charging and discharging processes, the inhomogeneous deposition of lithium potentially results in the formation of dendritic structures and dead lithium.9][10][11][12][13][14][15] These side reactions could lead to critical safety hazards.Thus, several approaches have been proposed to address the challenges posed by lithium metal anodes, including the addition of electrolyte additives, the application of artificial coatings on the metal surface, the incorporation of lithium host structures, and the introduction of lithiophilic alloy metals [7,8,10,11,16] More recently, anode-less or anode-free lithium metal batteries (AFLMBs), which do not utilize any form of anode active material, have received significant attention.The underlying idea of anode-free cell is to eliminate the graphite anode and deposit lithium metal directly onto the copper current collector.As a result, the volume previously occupied by the graphite can be repurposed for additional cathode active materials, thereby increasing the cell energy density.Furthermore, this anode-free architecture is expected to reduce costs and simplify cell assembly procedures. [8,9]Up to this point, the electrochemical performance of AFLMBs has been hampered by several factors.Many issues are common across all lithium metal anode cases, such as DOI: 10.1002/aesr.202300051Anode-free lithium metal batteries (AFLMBs) show promise as a means of further enhancing the energy density of current lithium-ion batteries, as they do not require conventional graphite anodes.The anode-free configuration, however, suffers from inferior chemical stability of the solid electrolyte interphase (SEI) layer and experiences inhomogeneous lithium deposition during charge/discharge processes, resulting in rapid capacity fading.To address these issues, a carbonized polydopamine (CPD) coating is applied to the copper current collector.The CPD-coated copper current collector promotes highly efficient and reversible lithium plating and stripping processes, resulting in a densely packed lithium deposition that significantly improves cycling stability.The anode-free full cell, consisting of CPD-coated copper current collector and a LiFePO 4 cathode, demonstrates significantly improved electrochemical performance, with a capacity retention of more than 63% after 100 cycles at a current rate of 0.3C.The stability of the SEI layer and the presence of lithiophilic sites are verified through a range of techniques, including optical microscopy, Raman spectroscopy, X-ray photoelectron spectroscopy, chronoamperometry, and electrochemical impedance spectroscopy.Based on these collective findings, it can be inferred that the use of CPD coating provides a simple way to enhance the electrochemical performance of AFLMBs.][19][20][21] Moreover, irreversible consumption of deposited lithium on copper during the initial cycles was a significant challenge for anode-free cells. [9,12,22][26] Herein, we demonstrate a simple method to enhance the cycle life of AFLMBs by modifying the surface of a conventional copper current collector.Our approach involves coating the substrate with a carbonized polydopamine (CPD) layer.The studies have revealed that the adhesion arises from the 3,4-dihydroxy-Lphenylalanine molecule, which can bind to a substrate via either or both of its amino or hydroxyl groups. [27,28]Based on these findings, Lee et al. developed a coating technique called polydopamine (PDA) coating. [29]This method has been widely adopted for hydrophilic surface modifications. [30,31]In this study, we employed a facile two-step process to prepare a CPD-Cu current collector.Electrochemical analysis revealed that the nucleation energy barrier in CPD-Cu was significantly reduced, possibly due to the presence of nitrogen-and oxygen-containing functional groups that created defective sites.In situ and ex situ analyses further confirmed the formation of a stable SEI layer and uniform lithium plating.Additionally, the surface modification of the copper foil yielded lithiophilic characteristics and stable electrochemical performance when combined with a LiFePO 4 cathode.The assembled full cell exhibited an initial specific capacity of 147 mAh g À1 and maintained 63.2% of its initial capacity after 100 cycles.Our analysis confirmed that the catechol functional group in PDA adhered well to the copper substrate, and the nitrogen-functional groups that remained after carbonization served as a reservoir for Li ions.Overall, our results suggest that mussel-inspired PDA coating provides a simple and effective approach for reversible lithium plating and stripping in AFLMBs.

Synthesis and Characterization of CPD-Cu Substrate
The CPD-Cu current collector was prepared in a facile two-step process.The fabrication route for synthesizing the CPD-Cu substrate is illustrated in Figure 1a and detailed procedures are described in Experimental Section.Briefly, the pristine copper foil (Pri-Cu) underwent PDA coating followed by calcination at 800 °C under a nitrogen environment.The morphology of Pri-Cu foil and CPD-Cu were investigated by field-emission scanning electron microscopy (FE-SEM) and the images are presented in Figure S1a-c, Supporting Information, respectively.The polymerization using dopamine hydrochloride resulted in a smooth and uniform morphology of PDA-coated copper substrate (PDA-Cu).During the heat-treatment process in a nitrogen atmosphere, PDA underwent thermal reduction and rearranged into 2D morphology sheets consisting of ordered sub-nanometer structures, which is similar to the previous report. [32]Furthermore, it appears that copper did not undergo significant oxidation during the calcination process at 800 °C.Although native oxide layer may be present on the surface of Pri-Cu prior to the calcination process, a previous reference leads us to suspect that CuO formation is unlikely to have contributed to reduced lithiophilicity. [33]The X-ray diffraction (XRD) spectra of Pri-Cu, PDA-Cu, and CPD-Cu (Figure S2, Supporting Information) displayed similar profiles for against each sample.In addition, the atomic force microscopy (AFM) topographical image (Figure 1b) confirmed the smooth coating morphology, while the line scans taken across the scratches made on CPD-Cu (Figure 1c) verified a thin conformal coating thickness of approximately 16-17 nm.
Raman spectroscopy was performed (Figure 2) to identify the structural variation of CPD-Cu and PDA-Cu.The spectra of CPD-Cu showed characteristic D (1339 cm À1 ) and G (1597 cm À1 ) bands corresponding to graphitic compounds. [34]urthermore, the intensity ratio I D /I G of CPD-Cu increased from 0.91 (PDA-Cu) to 1.02 upon thermal reduction, indicating a decrease in the average size of the sp 2 domains. [35]The I D /I G ratio can also signify the in-plane crystallite size (L a ) of the graphene structure, which can be calculated from the Tuinstra-Koenig relationship. [36]a ðnmÞ ¼ ð2.4 Â 10 À10 Þλ 4 where λ is the excitation wavelength.The L a values obtained from Equation ( 1) provided earlier for PDA-Cu and CPD-Cu were 18.4 and 16.4 nm, respectively.Thus, the decrease in L a value suggested the generation of defects during thermal reduction.
To gain further insights into the nature of defects, X-ray photoelectron spectroscopy (XPS) was employed to conduct chemical structural analysis of PDA-Cu and CPD-Cu (Figure 2b-e).The observed peaks at approximately 284.5, 400.2, and 533.3 eV corresponded to the C 1s, N 1s, and O 1s groups, respectively.The survey scan of CPD-Cu indicated the presence of 6.32% nitrogen and 13.64% oxygen atoms.The deconvolution of the C 1s spectrum revealed three distinct peaks located at 284.5, 287.8, and 288.9 eV, which corresponded to C-C/C = C, C-N/C-OH, and C = O, respectively. [34]Similarly, the N 1s spectra exhibited three binding energies at 397.8 eV for pyridinic, 399.15 eV for pyrrolic, and 400.15 eV for graphitic nitrogen. [37]Raman and XPS measurements have confirmed a higher density of lithiophilic defective sites on the CPD-Cu surface.These sites are expected to promote a more uniform lithiation-deposition process. [37,38]

Comparison of Li-Deposition Behavior on Different Anode-Free Substrate
The Li-deposition behavior on different substrates was investigated using half-cell configuration with Pri-Cu, PDA-Cu, or CPD-Cu as the working electrode and Li metal as both the counter and reference electrodes.At an areal capacity of 1 mAh cm À2 with a current density of 1 mA cm À2 , CPD-Cu electrode exhibited an exceptionally low nucleation overpotential of 10 mV (Figure 3a).In contrast, both the Pri-Cu and PDA-Cu electrodes experienced significantly higher nucleation overpotentials of 240 and 80 mV, respectively, under the same conditions.The significant difference in overpotential indicates that the lithiophilic CPD layer resulted in decreasing the nucleation barrier by promoting a more uniform distribution of charge on the electrode surface. [37,38]Indeed, we observed a larger irreversible capacity during the early stages of lithium plating on the CPD-Cu.Two potential explanations for this phenomenon are electrolyte decomposition and the consumption of lithium by multiple defective sites. [39]e utilized ultraviolet photoelectron spectroscopy (UPS) (Figure S3a   (2.29 eV), and lithium bis(trifluoromethanesulfonyl)imide (À1.03 eV), [40] the relatively large energy gap of CPD-Cu compared to Pri-Cu may minimize the chances of an electrolyte reduction on the electrode surface. [41]To put it differently, the CPD layer exhibited reduced the Fermi level, which increased the stability of the electrode surface in contact with the electrolyte.Therefore, lithium consumption by multiple defective sites in CPD layer is a more plausible explanation for the initial irreversible capacity than electrolyte decomposition.
Upon repeated Li plating and stripping cycles (Figure 3b), CE of Pri-Cu cells became increasingly erratic, even prior to reaching 50 cycles.This low CE is possibly on account of the reaction between lithium and the electrolyte, resulting in the irreversible loss of active lithium.Meanwhile, PDA-Cu demonstrated improved cycling performance, extending the cycle life up to 164 cycles.Most notably, CPD-Cu cells recorded significantly improved cycling performance, with stable CE observed for up to 300 cycles, surpassing the CE of both Pri-Cu and PDA-Cu cells.
Ensuring the long-term reversibility of lithium deposition and dissolution requires the formation of a stable SEI layer.In this study, we employed EIS measurements and ex situ XPS to analyze the SEI layer.Prior to this, the cells were subjected to precycling for five cycles ranging from 1.5 to 0.05 V at a scan rate of 1 mV s À1 .The EIS measurements showed a semicircle followed by Warburg impedance for both Pri-Cu and CPD-Cu, as presented in Figure 3c.Notably, the magnitude of the semicircle in Pri-Cu was twice as large as that in CPD-Cu.To gain a better understanding of these findings, we performed ex situ XPS analyses of the SEI layer on pre-cycled samples, as shown in Figure S4a,b, Supporting Information.The results revealed peaks at 398.5, 400, and 402 eV corresponding to Li 3 N, LiN x O y , and NO 2 À species composition in the SEI layer on the Pri-Cu electrode at 0 V. [42,43] The Li 3 N and LiN x O y are formed during SEI layer formation with LiNO 3 additive in the electrolyte.Interestingly, the SEI composition for the CPD-Cu sample showed a higher concentration of Li 3 N, which is more ionically conductive than LiN x O y and would form a denser and more compact layer, resulting in a lower R SEI for the CPD-Cu sample than the Pri-Cu sample.After subjecting the electrodes to 60 cycles at a current density of 1 mA cm À2 and a capacity of 1 mAh cm À2 , the semicircle magnitude for CPD-Cu remained relatively constant at 114 Ω, while Pri-Cu showed a fourfold increase in the semicircle magnitude (750.8Ω), suggesting that Pri-Cu induced the formation of a relatively thicker SEI layer.While it is unclear which factor, SEI composition or SEI thickness, is dominant, our ex situ EIS and XPS results provide compelling evidence of the stable SEI layer formation in the CPD-Cu.This stable SEI layer is critical for passivating the electrode surface and preventing any unwanted side reactions with the electrolyte.
To obtain a more comprehensive evaluation of the SEI layer, chronoamperometry (shown in Figure S3b,c, Supporting Information) was performed.CPD-Cu displayed lower current density than Pri-Cu during initial chronoamperometry measurement at 0 V (Figure S3b, Supporting Information), indicating lower electrolyte decomposition on CPD-Cu.Moreover, CPD-Cu sample exhibited lower current density and shorter duration than Pri-Cu after the SEI layer formation (Figure S3c, Supporting Information), confirming stable SEI layer formation in CPD-Cu.These results suggest that the defective sites present in CPD coating layer aided in uniform lithium ionic flux on the electrode surface, resulting in uniform Li deposition.
After depositing 1 mAh cm À2 of lithium on Pri-Cu, PDA-Cu, and CPD-Cu, we conducted SEM analysis to examine the morphology of the plated lithium.As shown in Figure 3e, uneven and dendritic Li was deposited on the Pri-Cu, likely due to an inhomogeneous lithium ionic flux distribution on the surface of the copper, especially during the initial plating induced by surface heterogeneities.This dendritic lithium deposition caused side reactions with Li metal and electrolyte and results in the generation of dead lithium, which affects the CE in subsequent cycles. [19,44]In contrast, PDA-Cu exhibited relatively smoother but nonuniform lithium islands (Figure 3f ), while CPD-Cu showed dendrite-free uniform lithium metal deposition (Figure 3g).We also analyzed the morphology of Li deposition on the edge regions (i.e., areas with high charge concentration) and observed uniform deposition of Li (Figure S5a,b, Supporting Information).It is presumed that the slow heating ramp up to 300 °C during thermal annealing resulted in the melting of PDA and subsequent uniform surface coverage, which explains the absence of lithium islands in CPD-Cu case.
To investigate the enhanced cycling efficiency of CPD-Cu, ex situ SEM studies were performed after subjecting the lithium plating/stripping processes to 20 cycles, at a current density of 1 mA cm À2 and an areal capacity of 1 mAh cm À2 .Figure S6a, Supporting Information presents an image of the lithiated state, which shows a uniform lithium-coated surface morphology with no dendrite formation.In Figure S6b, Supporting Information, the surface morphology after delithiation appears corrugated, indicating the complete removal of deposited lithium and no formation of dead lithium.The uniform and dendrite-free Li deposition observed on the CPD-Cu suggests that an interconnected defective structure potentially enabled homogeneous Li þ flux distribution on the entire surface of CPD-Cu during the plating process.

In Situ Raman Analysis of SEI Layer Stability on Pri-Cu and CPD-Cu
In situ Raman spectra of Pri-Cu and CPD-Cu were measured at different voltages (1.5, 1.0, 0.5, 0 V) and after 15 and 30 min of lithium deposition to analyze the stability of the formed SEI layer (Figure 4a,b).For the Pri-Cu sample, Raman bands around 250-380 cm À1 were diminished as the voltage approached 0 V. Since these bands are attributed to the SEI layer (LiF, Li 2 O, and LiOH), [45,46] the decrement suggests that the initially formed SEI layer deteriorated upon the Li plating process. [44]Raman bands at 746 cm À1 , assigned to the S-N-S vibration in TFSI À , [47] also deteriorated upon Li plating, suggesting the decomposition of electrolyte.In contrast, CPD-Cu displayed consistent intensities even after plating for 30 min, indicating improved stability of the SEI layer.Furthermore, the intensity of the Raman band at 1350 cm À1 , corresponding to the sp 3 distorted graphitic carbon, was highly reduced in the voltage region below 0 V, implying potential lithium deposition on the defective carbon sites.In addition, it is worth noting that defects induced the band at 1350 cm À1 corresponding to the sp 3 distorted graphitic carbon and the band at 1580 cm À1 was attributed to sp 2 -hybridized graphitic carbon. [48]We confirmed that the intensity of band at 1350 cm À1 was highly reduced in the voltage region below 0 V, which implies potential lithium deposition possibly occurred on the defective carbon sites. [39]o analyze the spatial homogeneity of the formed SEI layer, we conducted in situ Raman mapping at wavelengths of 280 and 1350 cm À1 , which corresponded to the SEI layer and effective carbon sites, respectively. [45,46,48]To normalize the Raman spectra, we utilized the reference peak at 940 cm À1 , which corresponds to the DOL solvent.The Raman mapping at the 280 cm À1 band of Pri-Cu exhibited highly nonuniform intensity at 0.05 V and such inhomogeneity was maintained even after plating for 30 min at the current density of 0.5 mA cm À2 (Figure 4c), implying highly inhomogeneous SEI layer on Pri-Cu.In contrast, CPD-Cu displayed uniform intensity throughout the mapped area at 0.05 V and maintained a homogeneous distribution even after plating for 30 min (Figure 4d), indicating a spatially homogeneous and stable SEI layer on the CPD-Cu surface.Additionally, Raman mapping of CPD-Cu at 1350 cm À1 showed a uniform distribution of defective carbon sites throughout the CPD-Cu surface (Figure 4e).Upon Li plating for 30 min, the intensity at 1350 cm À1 was highly reduced, indicating that Li metal was preferentially deposited on the defective sites of the CPD coating layer.

Electrochemical Performance of Anode-Free Full Cells
In the final step of our experiment, we assembled either Pri-Cu or CPD-Cu into an anode-free full cell configuration and characterized the electrochemical performance.We used commercially available LiFePO 4 (LFP) coated on aluminum foil as the cathode and conditioned the assembled cells at a current rate of 0.1C for three cycles before measuring their cycling performance at a rate of 0.3C.From this point on, Pri-Cu and CPD-Cu full cells are referred to as Pri-Cu-LFP and CPD-Cu-LFP, respectively.During the initial charge, both full cells exhibited a single voltage plateau at 3.47 V, but CPD-Cu-LFP displayed a larger initial discharge capacity of 147 mAh g À1 compared to 132 mAh g À1 of Pri-Cu-LFP (Figure 5a).In terms of cycling capability, Pri-Cu-LFP and CPD-Cu-LFP retained 37.0% and 63.2% of their initial capacity after 100 cycles, respectively (Figure 5b).Other reports suggest that electrochemical instability often arises from localized current density and inconsistent current distribution, which hinders the reversible deposition and stripping of lithium on the copper surface. [9,17]The improved cycling stability of CPD-Cu-LFP in this study most likely arose from the presence of heteroatoms (N-and O-) and defects, which assisted to form a stable SEI layer and promote uniform Li nucleation.
We conducted in situ Li plating/stripping experiment to further characterize the dendrite-free Li metal surface in CPD-Cu.As illustrated in Figure 5c, the working electrode consisted of either Pri-Cu or CPD-Cu, while a Li-metal-attached copper foil was used as the counter electrode.The working electrode surface was examined with an optical microscope and continuously monitored at a current density of 0.5 mA cm À2 .During the lithium-deposition process (Figure 5d), we clearly identified numerous dendrites forming in various locations.In contrast, we did not observe any noticeable dendrites for CPD-Cu (Figure 5e), which is consistent with the previous cycling data.These observations provide compelling evidence and suggest that the CPD layer plays a crucial role in promoting smooth Li metal deposition in a dynamic Li plating/stripping environment.

Conclusion
In summary, we applied a simple CPD coating on conventional copper current collector to demonstrate uniform Li deposition without forming dendrites.The CPD-coated film effectively facilitated the homogeneous deposition of Li ions and the formation of a stable SEI layer.We configured a full-cell system by coupling the CPD-Cu film with LiFePO 4 cathode to measure its performance.The results of our experiments indicated that the CPD-Cu configured cell exhibited noticeably improved results, maintaining a capacity of over 63% after 100 cycles at a rate of 0.3C, far exceeding the performance of a standard copper current collector.In situ optical microscopy and Raman mapping results indicate that the CPD coating leads to uniform deposition of lithium across the current collector, resulting in a densely packed lithium layer that significantly enhances cycling stability.This study demonstrates that a simple bio-inspired coating could be an effective method for improving the interfacial stability of anode-free configuration and presents a step toward the practical design and realization of AFLMBs.

Experimental Section
Synthesis: The PDA was synthesized in accordance with the procedures described in an earlier literature, [33] with few modifications.The substrate for coating the thin PDA film in this investigation was a commercially available copper foil (20 μm).The substrate was sequentially cleaned with acetone, isopropanol, and deionized water and was immersed in 100 mL of Tris buffer (trishydroxymethyl aminomethane solution, 10 mM with pH 8.5) containing 0.2 g of dopamine hydrochloride.The polymerization of the coating was conducted at 25 °C for 24 h.The coated substrates were extracted from the solution, washed multiple times with distilled water, and dried overnight at 70 °C.
PDA-coated substrates were heat-treated in a tube furnace in N 2 atmosphere to form CPD coating on copper foil.Using a two-step heat-treatment procedure, the substrate was heated at a rate of 2 °C min À1 until reaching 300 °C, followed by a rate of 5 °C min À1 until reaching 800 °C.The temperature was then held at 800 °C for 1 h.After heat treatment, the sample was naturally cooled down to room temperature.
Materials and Electrochemical Characterizations: The XRD measurements of the as-synthesized substrate were performed using powder-XRD (RIKAGU, D/Max-2500 with Cu Kα [λ = 1.5418Å]).Raman spectra were obtained using a Lab RAM HR Evolution (HORIBA, Japan) with laser excitation at a wavelength of 514 nm.In situ Raman spectra and mapping analysis were carried out using an in-house-built in operando cell with Pri-Cu/Li and CPD-Cu/Li configuration with 1 M LiTFSI þ 0.2 M LiNO 3 electrolyte, assembled in an argon-filled glove box maintained at <1 ppm oxygen and H 2 O levels.Raman measurements were carried out using a high-resolution Raman system (LabRAM HR Evolution Visible_NIR) equipped with an excitation wavelength of 514 nm.The cell was firmly mounted to an automatic scanning stage and the laser was focused on the Pri-Cu or CPD-Cu electrode.Raman spectrum measurements were carried out at different voltage steps during the plating process (1.5, 1.0, 0.5, 0 V and after 15 and 30 min of Li deposition) with 0.5 mA cm À2 rate.Raman mapping was constructed along 10 Â 10 μm area of Pri-Cu or CPD-Cu electrode before (holding at 0.05 V) and after 0.25 mAh cm À2 lithium deposition on respective electrodes.The morphological analysis was carried out using an FE-SEM (Magellan400, FEI) equipped with energy-dispersive spectroscopy.UPS analysis to calculate the work function was conducted using in situ XPS (Axis-Supra by Kratos) with a He-I X-ray source (hν = 21.2 eV).The electrochemical performances were evaluated by assembling CR2032-type coin cells assembled in an argon-filled glove box maintained at <1 ppm oxygen and H 2 O levels.The 1 M LiTFSI in DME/DOL (1:1 w/w) with 0.2 M LiNO 3 was used as the electrolyte (90 μL cell À1 ) and Celgard 2400 was used as the separator.For the CE test, half-cells were made using as-prepared CPD-Cu or Pri-Cu (1.1 cm À2 ) as the working electrode and the lithium metal foil as the counter electrode.The galvanostatic measurements were carried out using a battery cycler (Won-A-Tech WBS3000).Cycling tests were performed by first plating lithium on the working electrode at desired current densities and capacities, followed by stripping up to 500 mV.EIS was measured after the SEI layer formation (5 cycles in the voltage range of 1.5-0.05V at 1 mV s À1 ) as well as after 60 cycles at a current density of 1 mA cm À2 to a capacity of 1 mAh cm À2 .Chronoamperometry was measured at the initial state at 0 V and after the SEI layer formation.Both EIS and chronoamperometry were measured using Bio-Logic VSP potentiostat.Commercially available LiFePO 4 cathode from Welcos Co., Korea, with binder and super P of 80:10:10 composition coated in an Al foil was used as the cathode for anode-free full cell.The CPD-Cu/LFP or Pri-Cu/LFP full cells were cycled between 4.0 and 2.0 V at 0.1C rate for the first 3 cycles and 0.3C rate for the following cycles.All electrochemical measurements were operated at 25 °C.

Figure 1 .
Figure 1.a) Schematic representation of the synthesis of carbonized polydopamine (CPD)-Cu via polydopamine coating.b) AFM topographical image of CPD-Cu foil shows smooth coating morphology and c) the line scans across the scratch made on the surface of CPD-Cu.
, Supporting Information) to obtain a more detailed evaluation of the voltage profile.The work function was calculated by subtracting the energy difference between the secondary edge and the Fermi edge using He-I (21.2 eV) as the source energy.The obtained values were 4.20 and 5.08 eV for Pri-Cu and CPD-Cu, respectively.The Fermi levels of Pri-Cu and CPD-Cu correspond to the negative value of their work function, that is, À4.20 and À5.08 eV.Since these values are lower than the lowest unoccupied molecular orbital of the electrolytes, such as 1,2 dimethoxyethane (DME) (2.24 eV), 1,3-dioxolane (DOL)

Figure 3 .
Figure 3. a) Voltage profiles and b) Coulombic efficiency of Li plating/stripping on Pri-Cu, PDA-Cu and CPD-Cu at current density from 1 to 1 mAh cm À2 areal capacity.c) Nyquist plots of Pri-Cu and CPD-Cu measured at 25 °C before and d) after 60 cycles in delithiated state.Scanning electron microscopy (SEM) images of e) Pri-Cu, f ) PDA-Cu, and g) CPD-Cu after lithium deposition at 1 mA cm À2 to an areal capacity of 1 mAh cm À2 .

Figure 4 .
Figure 4.In situ Raman spectra of a) Pri-Cu and b) CPD-Cu at different cutoff voltages during lithium plating at a current density of 0.5 mA cm À2 .c) In situ Raman mappings at 280 cm À1 of Pri-Cu at 0.05 V and after plating for 30 min at a current density of 0.5 mA cm À2 , and d) CPD-Cu at 0.05 V and after plating for 30 min.e) In situ Raman mappings at 1350 cm À1 of CPD-Cu at 0.05 V and after plating for 30 min.

Figure 5 .
Figure 5. a) Galvanostatic voltage profiles of Pri-Cu-LFP and CPD-Cu-LFP at 0.1C rate and b) corresponding cycling performance at 0.3C.c) Schematic illustrating the half-cell used for in situ optical microscopy measurement.d) Optical images of Pri-Cu and e) CPD-Cu to observe the Li-deposition behavior.