All-polymer particulate slurry batteries

Redox flow batteries are promising for large-scale energy storage, but some long-standing problems such as safety issues, system cost and cycling stability must be resolved. Here we demonstrate a type of redox flow battery that is based on all-polymer particulate slurry electrolytes. Micro-sized and uniformly dispersed all-polymer particulate suspensions are utilized as redox-active materials in redox flow batteries, breaking through the solubility limit and facilitating the application of insoluble redox-active materials. Expensive ion-exchange membranes are replaced by commercial dialysis membranes, which can simultaneously realize the rapid shuttling of H+ ions and cut off the migration of redox-active particulates across the separator via size exclusion. In result, the all-polymer particulate slurry redox flow batteries exhibit a highly reversible multi-electron redox process, rapid electrochemical kinetics and ultra-stable long-term cycling capability.

other key components in large-scale energy storage systems.
Also, we have adjusted the descriptions of the significance and novelty of our work as you suggested, as follows: ------------------------------"A new class of RFBs based on combining redox-active polymers or colloids with size-exclusion separator could inhibit the crossover of active species, improve the cycling stability and reduce the battery cost 23,49 ." "To alleviate these issues, herein, we propose the design of a new type of RFBs based on aqueousdispersed all-polymer particulate slurry electrolytes with multi-electron redox capability and fast charge transfer." "The application of particulate slurry electrolytes in RFBs makes it possible to replace expensive ion exchange membranes with much cheaper commercial dialysis membranes through the mechanism of size exclusion 23,49 ." "Different from incorporating redox molecules as pendants into inactive polymer colloids, we utilized polymers with redox-active sites on their main-chain aromatic rings." "In summary, we developed a new type of RFBs by utilizing all-polymer particulate slurries as aqueous catholyte and anolyte." i. The manuscript claims that the suspension breaks the solubility limit by increasing the concentration 1.6× (from ~ 0.6 mol L-1 to 1 mol L-1, based on quinone-unit concentration), which theoretically enables higher Ah L-1. However, it appears as though having the active material as particles may limit the capacity. Although the authors mention that the capacity may not come from only the surface of the particles, there is no experiment proving that. In addition, the manuscript reports a combined gravimetric capacity of 74 mAh per gram of catholyte+anolyte, which is substantially lower (~1/3×) than the theoretical combined capacity of catholyte + anolyte (~200 mAh g-1). The authors should discuss and indicate with calculations what are the performance advantages and limitations of the particle system with the current state-of-the-art organic RFB (i.e. best results in the field). Please include solubility comparisons as well.
Response: Thanks for your helpful suggestion. The CV curves of PHQ particulates and benzoquinone monomer (BQ) were compared to demonstrate the site-hopping mechanism of charge transferring during redox processes. PHQ with average diameters of 1 μm and 50 μm are termed as PHQ-1 and PHQ-50 (Supplementary Figure 7). As shown in Supplementary Figure 8a, with the decrease of particulate size, the electrochemical polarization (ΔE) of PHQ particulates was reduced from 618 mV  to 200 mV (PHQ-1), indicating the sluggish charge transfer within and among PHQ particulates was improved. At the scan rate of 0.025 V/s, the CV oxidation peak currents of BQ, PHQ-1 and PHQ-50 are 0.74, 0.38 and 0.25 mA/cm 2 , respectively, indicating the large particle size has negative effect to the utilization ratio of redox active species 33 . When the scan rate is decreased to 0.006 V/s, the CV oxidation peak currents of BQ, PHQ-1 and PHQ-50 are 0.10, 0.08, 0.06 mA/cm 2 , respectively (Supplementary Figure 8b). These results indicate that the smaller particle size may lead to more charge transfer inside the particulates and more polymer units involved in the redox reaction.
Based on the theoretical model proposed in the literatures 33,49,53 , the site-hopping mechanism is proposed to elucidate the charge transfer of particulates during redox processes. As shown in the new Fig. 1b (see below), the redox-active sites on particulate surface are firstly reduced when approaching to the electrode. Then, the charges transport across the polymer chains by electron hopping between the highly populated redox-active groups 49,52,53 . In addition to the above site-hopping mechanism, the π-conjugated structures of PHQ and PI also enhance the charge transfer during redox processes. Besides, the strong acidic environment and high H + concentration could also enhance the proton conductivity of polymer particulates, thus accelerating the transfer of electrolyte ions through polymer and increasing the Faradaic response 54 .
We also supplemented the battery performance comparison with the state-of-the-art organic RFBs and discussed the performance advantages and limitations of the APPSBs system. As shown in Supplementary Table 1, the mole concentration of PHQ and PI units for battery testing (1.0 mol/L) in our APPSBs is larger than most of other batteries (usually less than 1.0 mol/L). Benefit from the robust polymer frameworks and the absence of side reactions, the APPSBs in this work exhibits good longterm cycling stability, which is superior to many organic RFBs. Although the rate and capacity utilization of APPSBs are relatively lower than some other aqueous systems, they're still higher than nonaqueous systems. We suggest that the utilization ratio of active materials could be further improved by tuning the size, microstructure and compositions of electrochemical-active particulates, such as constructing conductive agent composites, as well as adding proper electrolyte stabilizer without the compromise of electrochemical performances.
We have added new "To investigate the charge transfer mechanism of particulates, the electrochemical properties of benzoquinone monomer (BQ) and PHQ particulates with different sizes were compared. PHQ with average diameters of 1 μm and 50 μm are termed as PHQ-1 and PHQ-50 (Supplementary Figure 7). As shown in Supplementary Figure 8a, with the decrease of particulate size, the electrochemical polarization (ΔE) of PHQ particulates was reduced from 618 mV (PHQ-50) to 200 mV (PHQ-1), indicating the sluggish charge transfer within and among PHQ particulates was improved. At the scan rate of 0.025 V/s, the CV oxidation peak currents of BQ, PHQ-1 and PHQ-50 are 0.74, 0.38 and 0.25 mA/cm 2 , respectively, indicating the large particle size has negative effect to the utilization ratio of redox active species 33 . When the scan rate is decreased to 0.006 V/s, the CV oxidation peak currents of BQ, PHQ-1 and PHQ-50 are 0.10, 0.08, 0.06 mA/cm 2 , respectively (Supplementary Figure 8b). These results indicate that the smaller particle size may lead to more charge transfer inside the particulates and more units involved in the redox reaction. Based on the theoretical model proposed in the literatures 33,49,53 , a site-hopping mechanism is proposed to elucidate the charge transfer of particulates during redox processes. As shown in Fig. 1b, the redox-active sites on particulate surface are firstly reduced when approaching to the electrode. Then, the charges transport across the polymer chains by electron hopping between the highly populated redox-active groups 49,52,53 ." "Promoted by the good redox kinetics and reversibility of PHQ and PI, the PHQ/PI APPSBs can be stably operated for long-term cycling, delivering capacity of 4.95 Ah/L (3.1 Wh/L) at the current density of 20 mA/cm 2 with a capacity retention of 70% after 300 cycles, which is comparable to other current state-of-art RFBs (Supplementary Table 1   ii. Are the PHQ and PI particles porous? What is the stability of these suspensions at low temperature? The authors should also discuss the effect of the particle size on the suspension's electrochemical properties, such as the diffusion coefficients. Response: Thanks very much for your valuable questions. We have investigated the porosity characteristics of PHQ and PI particulates by N2 adsorption−desorption isotherms. The Brunauer−Emmer−Teller (BET) surface areas of PHQ, PI1 and PI2 particulates are 5.5, 19.9, 2.0 m 2 /g, respectively (Supplementary Figure 19). The pore size distribution in Supplementary Figure 19b demonstrates the presence of some mesopores in PI1 particulates, although the pore volume is relatively low.
At low temperature (4 °C), visual observations of the dispersion stability of polymer particulate slurries only show slight sedimentation after three days (see the new Supplementary Figure 1b, as below). Besides, the electrochemical tests of APPSBs were normally performed with the rapid circulation of electrolytes driven by peristaltic pumps, so the precipitation of polymer particulates was very minimal under the flow mode, even at low temperature.
Following your valuable suggestions, we have compared the dispersibility, viscosity, and electrochemical properties of polymer particulates with different sizes. The particulate sizes of PI were decreased to <1 μm via ball-milling method. CV tests and RDE measurements show the smaller PI particulates possess better charge transfer and higher capacity utilization (see the new Supplementary Figure 24-26 below). Detailed comparison is listed in the new Supplementary Table 2. We have added the following discussion and figures in the revised Manuscript and Supplementary Data, as below: ------------------------------"Nitrogen adsorption/desorption isotherms were recorded at 77 K using a Micrometrics ASAP 2020 analyzer. The surface area and pore size distribution were determined by Brunauer-Emmett-Teller (BET) and Barrett-Joyner-Halenda method, respectively." "Significant difference of redox kinetics of PI1 and PI2 particulates may be partially originated from their different pore structures. Brunauer−Emmer−Teller (BET) analysis shows the specific surface areas of 19.9 and 2.0 m 2 /g for PI1 and PI2 particulates, respectively (Supplementary Figure 20). The pore size distribution demonstrates the presence of some mesopores of PI1, although the pore volume is relatively low. Compared to the compact stacking structure of PI2 resulted from the rigid benzene rings, the loosely-stacking structure of PI1 attributed by the flexible alkyl segments on the main chains may accelerate the transfer of abundant electrolyte ions into the particulates." "…at room temperature (25 °C) and low temperature (4 °C) …" "The influences of particle size on the electrochemical and physicochemical properties of polymer particulates were investigated. Briefly, ballmilling processes were performed to further decrease the size of PI1 and PI2 particulates, and the control samples after ballmilling for 48 h were termed as PI1ballmilled, PI2-ballmilled, respectively. As shown in the SEM images and DLS curves (Supplementary Figure 22 & 23), the average particle size of PI1 decreased from 2.7 μm (PI1) to 0.8 μm (PI1ballmilled), and the average particle size of PI2 decreased from 5.6 μm (PI2) to 0.9 μm (PI2-ballmilled). The Zeta potentials of PI1-ballmilled and PI2-ballmilled were measured to be 47.8 mV and 34.3 mV, indicating the dispersibility and stability were improved after ball-milling. CV analysisi revealed the increased Faradaic response of the smaller PI1 and PI2 (Supplementary Figure 24). Diffusion coefficient, including the physical transport of particulates to the electrode and the charge transport of the particulates, are calculated to be 1.7×10 -7 cm 2 /s (PI1), 3.4×10 -7 cm 2 /s (PI1-ballmilled), 0.7×10 -7 cm 2 /s (PI2), and 1.3×10 -7 cm 2 /s (PI2-ballmilled) (Supplementary Figure 10, 18, 25, 26). The smaller particle size accelerates the particulate diffusion and charge transport in the redox process (Supplementary Table 2). Constant-current charge/discharge tests of PHQ/PI2-ballmilled APPSB shows discharging capacity of 8.40 Ah/L (6.05 Wh/L) at the current density of 5 mA/cm 2 , larger than that of PHQ/PI2 APPSB (4.30 Ah/L, 3.02 Wh/L), demonstrating higher capacity utilization of smaller polymer particulates (Supplementary Figure 27). We suggest that the utilization ratio of active materials could be further improved by tuning the size, microstructure and compositions of electrochemical-active particulates, such as constructing conductive agent composites, as well as adding proper electrolyte stabilizer without the compromise of electrochemical performances."  iii. What is the reason for the low coulombic efficiency at 5 mA cm-2? Assuming there is no crossover of material, is there any parasitic reaction happening in this battery?

Supplementary
Response: Thanks very much for the good questions. As shown in the new Supplementary Figure 4, we noticed that the CV curve of PHQ particulate suspension at the 50 th cycle was nearly overlapped with that at the 1 st and 2 nd cycles, while the ratio of oxidation capacity to reduction capacity (Q1/Q2) were changed at different scan rates. The Q1/Q2 is 1.08 at 0.1 V/s and 1.34 at 0.025 V/s, respectively. We attributed the small oxidation peak at 1.2 V (vs. SHE) to the electro-polymerization of PHQ 55,56 . As shown in Supplementary Figure 5, when the terminal hydroquinone was oxidized to the protonated benzoquinone in the electro-oxidation process, it might react with the non-protonated hydroquinone, leading to the electro-polymerization of polymer chains. The electro-polymerization provides additional oxidation capacity, but it doesn't affect the reversibility of redox-active groups. The parasitic reaction of electro-polymerization can be suppressed by increasing the charge rate, as indicated by the Tafel plots (Supplementary Figure 6). We have added the related discussion and figures in the revised Manuscript and Supplementary data, as below: ------------------------------"As shown in Supplementary Figure 4, the CV curve of PHQ particulate suspension at the 50th cycle was nearly overlapped with that at the 1st and 2nd cycles, while the ratio of oxidation capacity to reduction capacity (Q1/Q2) were changed at different scan rates. The Q1/Q2 is 1.08 at 0.1 V/s and 1.34 at 0.025 V/s, respectively. The small oxidation peak at 1.2 V (vs. SHE) is proposed to be originated from the electro-polymerization of PHQ 55,56 . As illustrated in Supplementary Figure 5, when the terminal hydroquinone was oxidized to the protonated benzoquinone in the electro-oxidation process, it might react with the non-protonated hydroquinone, leading to the electro-polymerization of polymer chains. The electro-polymerization can provide additional oxidation capacity, but it doesn't affect the reversibility of redox-active groups. The parasitic reaction of electro-polymerization can be suppressed by increasing the current rate, as indicated by the Tafel plots (Supplementary Figure 6)." iv. What is the mechanism for the capacity fading? Although the dialysis membrane keeps the particles in different compartments, the polymers could degrade and release low molecular weight portions during the battery cycles. A careful analysis of the suspensions after battery cycling must be provided to indicate that the particles keep approximately the original dimensions.
Response: Thanks very much for the good questions. The capacity fading may be attributed to the aggregation and sedimentation of active particulates after long-time cycling. DLS and Zeta potential measurements reveal the similar dispersibility and stability of the diluted polymer particulate suspensions after cycling test (Supplementary Figure 13). As shown in the SEM images (Supplementary Figure 14), most of the PHQ and PI1 particulates keep approximately the original dimensions, but some particulates aggregate into larger particles. For PI2, the aggregation is much worse. The agglomerates may precipitate in the flow grooves and reservoirs, resulting in the capacity fading. Therefore, to further improve the cycling stability of APPSBs, we suggest that proper particulate stabilizer without the compromise of electrochemical performance could be introduced.
------------------------------"DLS and Zeta potential measurements reveal the similar dispersibility and stability of the diluted polymer particulate suspensions after cycling test (Supplementary Figure 13). As shown in the SEM images ( Supplementary Figure 14), most of the PHQ and PI1 particulates keep approximately the original dimensions, but some particulates aggregate into larger particles. For PI2, the aggregation is much worse. The agglomerates may precipitate in the flow grooves and reservoirs, resulting in the capacity fading. Therefore, to further improve the cycling stability of APPSBs, we suggest that proper particulate stabilizer without the compromise of electrochemical performance could be introduced."
v. In Fig. S9, the wavelength of the peaks should be included in the graph. I do not observe any 320 nm peak redshift. The oxidized and reduced species spectra for PHQ and PI seems to have several transition peaks in common. This could indicate partial oxidation/reduction in the particles. "After the electro-oxidation of PHQ, a new absorption peak at 246 nm was emerged ( Supplementary  Figure 15a), and the absorption peak at 326 nm was enhanced and red-shifted, indicating the conjugation effect between the carbonyl groups and the backbone of benzene rings 36 . On the other hand, as shown in Supplementary Figure 15b, the electro-reduction of PI1 leads to new absorption peaks at 304 nm, 539 nm and 645 nm 37 , owing to the spatial charge distribution variation of π−conjugation system."

vi. Please clarify the metrics calculation. Is the capacity per L (Ah L-1) ¬based on the total volume of suspension (~ 20 mL)? For the current density, what is the area based on?
Response: Thanks. The capacity per L (Ah L -1 ) is based on the total volume of suspension (~ 20 mL). The area is based on the area of carbon paper electrode (1×1 cm 2 ).

Reviewer #2 (Remarks to the Author):
This manuscript demonstrated the idea of All-polymer Particulate Slurry Batteries(APPSB), which is similar to the idea of inorganic material particle slurry batteries, e.g. using lithium ion battery materials. The authors claimed the APPSB can break though the solubility limits of active materials and take advantage of the fast kinetic of the redox reactions between the polymer particles and protons.  33,49,52 , the CV curves of PHQ particulates and benzoquinone monomer (BQ) were compared to demonstrate the site-hopping mechanism of charge transferring during redox processes. As shown in the new Fig. 1b (see below), the redox-active sites on the surface of PHQ and PI particulates are firstly reduced when approaching to the electrode. Then, the charges transport across the polymer chains by electron hopping between the highly populated redox-active groups 49,52,53 . In addition to the above site-hopping mechanism, the π-conjugated structures of PHQ and PI also enhance the charge transfer during redox processes. Besides, the strong acidic environment and high H + concentration could also enhance the proton conductivity of polymer particulates, thus accelerating the transfer of electrolyte ions through polymer and increasing the Faradaic response 54 . We also added the functions of each component of the cell in the caption of Supplementary Figure  11. The titanium plate acts as a highly corrosion-resistant current collector. The graphite plate with narrow grooves provides the flow channel of the circulating electrolyte, and the carbon paper electrode provides the electrochemical reaction sites for particulates.
The PI is normally regarded as the electron-insulator, but it can transport charges by site hopping mechanism during redox processes 49,52,53 , whereby the electrons can transfer from one redox-active imide group to another on a neighboring unit.
We have added the related discussion and figures in the revised Manuscript and Supplementary Data, as follows: ------------------------------"According to the Randles-Sevcik equation, a linear increase of the peak current (i) against the square root of the scan rate (v 1/2 ) is observed for PHQ (Fig. 4a & Supplementary Figure 10a) and PI1 (Fig. 4d  & Supplementary Figure 10d), respectively, indicating the occurrence of charge diffusion inside the particulates rather than only on the surface. "To investigate the charge transfer mechanism of particulates, the electrochemical properties of benzoquinone monomer (BQ) and PHQ particulates with different sizes were compared. PHQ with average diameters of 1 μm and 50 μm are termed as PHQ-1 and PHQ-50 (Supplementary Figure 7). As shown in Supplementary Figure 8a, with the decrease of particulate size, the electrochemical polarization (ΔE) of PHQ particulates was reduced from 618 mV (PHQ-50) to 200 mV (PHQ-1), indicating the sluggish charge transfer within and among PHQ particulates was improved. At the scan rate of 0.025 V/s, the CV oxidation peak currents of BQ, PHQ-1 and PHQ-50 are 0.74, 0.38 and 0.25 mA/cm 2 , respectively, indicating the large particle size has negative effect to the utilization ratio of redox active species 33 . When the scan rate is decreased to 0.006 V/s, the CV oxidation peak currents of BQ, PHQ-1 and PHQ-50 are 0.10, 0.08, 0.06 mA/cm 2 , respectively (Supplementary Figure 8b). These results indicate that the smaller particle size may lead to more charge transfer inside the particulates and more units involved in the redox reaction. Based on the theoretical model proposed in the literatures 33,49,53 , a site-hopping mechanism is proposed to elucidate the charge transfer of particulates during redox processes. As shown in Fig. 1b, the redox-active sites on particulate surface are firstly reduced when approaching to the electrode. Then, the charges transport across the polymer chains by electron hopping between the highly populated redox-active groups 49,52,53 ."

Supplementary Figure 11 | Structure configuration and components of APPSB cell.
The stainless-steel plate fixed with screws works as the frame of the cell. The PTFE plate prevents the short-circuit of the cell. The titanium plate acts as a highly corrosion-resistant current collector. The graphite plate with narrow grooves provides the flow channel of the circulating electrolyte, and the carbon paper electrode provides the electrochemical reaction sites for particulates. The silicone rubber seal prevents the leakage of electrolyte. The PTFE frame connecting with silicone rubber tubes is the gateway for electrolytes pumping into and out of the cell. The separator keeps the two electrodes apart to prevent electrical short-circuit while allowing the transport of ions during charge/discharge processes.

2.
Despite of the evidence showed by the authors that the redox reactions are very reversible, the electrochemical performance of this slurry battery is in my opinion much worse compared with all vanadium flow battery. Although being a proof-of-concept paper, the authors need to demonstrate the room for further improvement. The authors need to demonstrate the electrochemcial performance with a higher particle concentration in the suspensions.
Response: Thanks very much for your good suggestion. To demonstrate the improved volumetric specific capacity at higher concentration of redox-active units, the particulate sizes of PI1 and PI2 were further decreased to <1 μm via ballmilling method, and these control samples were denoted as PI1ballmilled and PI2-ballmilled, respectively. The CV tests and RDE measurements of PI1-balmilled and PI2-ballmilled samples show that the smaller polymer particulates possess better charge transfer and higher capacity utilization (Supplementary Figure 23-26 and Supplementary Table 2). Briefly, constant-current charge/discharge tests of PHQ/PI2-ballmilled APPSB shows discharging capacity of 8.40 Ah/L at 5 mA/cm 2 , which is larger than that of original PHQ/PI2 APPSB (4.30 Ah/L). This result confirms that it is possible to further improve the volumetric energy density of APPSBs with a higher concentration of redox-active species. We suggest that the utilization ratio of active materials could be further improved by tuning the size, microstructure and compositions of electrochemical-active particulates, such as constructing conductive agent composites, as well as adding proper electrolyte stabilizer without the compromise of electrochemical performances. We have added the related discussion and figures in the revised Manuscript and Supplementary Data, as follows: ------------------------------"The influences of particle size on the electrochemical and physicochemical properties of polymer particulates were investigated. Briefly, ballmilling processes were performed to further decrease the size of PI1 and PI2 particulates, and the control samples after ballmilling for 48 h were termed as PI1ballmilled, PI2-ballmilled, respectively. As shown in the SEM images and DLS curves (Supplementary Figure 22 & 23), the average particle size of PI1 decreased from 2.7 μm (PI1) to 0.8 μm (PI1ballmilled), and the average particle size of PI2 decreased from 5.6 μm (PI2) to 0.9 μm (PI2-ballmilled). The Zeta potentials of PI1-ballmilled and PI2-ballmilled were measured to be 47.8 mV and 34.3 mV, indicating the dispersibility and stability were improved after ball-milling. CV analysisi revealed the increased Faradaic response of the smaller PI1 and PI2 (Supplementary Figure 24). Diffusion coefficient, including the physical transport of particulates to the electrode and the charge transport of the particulates, are calculated to be 1.7×10 -7 cm 2 /s (PI1), 3.4×10 -7 cm 2 /s (PI1-ballmilled), 0.7×10 -7 cm 2 /s (PI2), and 1.3×10 -7 cm 2 /s (PI2-ballmilled) (Supplementary Figure 10, 18, 25, 26). The smaller particle size accelerates the particulate diffusion and charge transport in the redox process (Supplementary Table 2). Constant-current charge/discharge tests of PHQ/PI2-ballmilled APPSB shows discharging capacity of 8.40 Ah/L (6.05 Wh/L) at the current density of 5 mA/cm 2 , larger than that of PHQ/PI2 APPSB (4.30 Ah/L, 3.02 Wh/L), demonstrating higher capacity utilization of smaller polymer particulates (Supplementary Figure 27). We suggest that the utilization ratio of active materials could be further improved by tuning the size, microstructure and compositions of electrochemical-active particulates, such as constructing conductive agent composites, as well as adding proper electrolyte stabilizer without the compromise of electrochemical performances."  3. The author also need to calculate the energy density beside the specific capacity to give the readers a more comprehensive picture. Also in the manuscript, the authors should compare the performance of this polymer slurry battery with state-of-art flow battery.

Supplementary
Response: Thanks for your helpful suggestion. We have added the energy density beside the specific capacity and the performance comparison with the current state-of-the art organic RFBs. Aqueous and organic electrolytes in the current state-of-art organic RFB are both included in the comparison. As shown in Supplementary Table 1, the mole concentration of PHQ and PI units for battery testing (1.0 mol/L) in our APPSBs is larger than most of other batteries (usually less than 1.0 mol/L). Benefit from the robust polymer frameworks and the absence of side reactions, the APPSBs in this work exhibits good long-term cycling stability, which is superior to many organic RFBs. Although the rate and capacity utilization of APPSBs are relatively lower than some other aqueous systems, they're still higher than nonaqueous systems. We suggest that the utilization ratio of active materials could be further improved by tuning the size, microstructure and compositions of electrochemical-active particulates, such as constructing conductive agent composites, as well as adding proper electrolyte stabilizer without the compromise of electrochemical performances. 4. In page 9 line 271, the coulombic efficiency is 87%. Why is it so low even after the authors claimed there was no side reactions? Also in page 10 line 289, why the coulombic efficiency increases with increasing current densities?

------------------------------
Response: Thanks very much for the good question. As shown in the new Supplementary Figure 4, we noticed that the CV curve of PHQ particulate suspension at the 50 th cycle was nearly overlapped with that at the 1 st and 2 nd cycles, while the ratio of oxidation capacity to reduction capacity (Q1/Q2) were changed at different scan rates. The Q1/Q2 is 1.08 at 0.1 V/s and 1.34 at 0.025 V/s, respectively. We attributed the small oxidation peak at 1.2 V (vs. SHE) to the electro-polymerization of PHQ 55,56 . As shown in Supplementary Figure 5, when the terminal hydroquinone was oxidized to the protonated benzoquinone in the electro-oxidation process, it might react with the non-protonated hydroquinone, leading to the electro-polymerization of polymer chains. The electro-polymerization provides additional oxidation capacity, but it doesn't affect the reversibility of redox-active groups. The parasitic reaction of electro-polymerization can be suppressed by increasing the charge rate, as indicated by the Tafel plots (Supplementary Figure 6).
We have added the related discussion and figures in the revised Manuscript and Supplementary data, as below: ------------------------------"As shown in Supplementary Figure 4, the CV curve of PHQ particulate suspension at the 50th cycle was nearly overlapped with that at the 1st and 2nd cycles, while the ratio of oxidation capacity to reduction capacity (Q1/Q2) were changed at different scan rates. The Q1/Q2 is 1.08 at 0.1 V/s and 1.34 at 0.025 V/s, respectively. The small oxidation peak at 1.2 V (vs. SHE) is proposed to be originated from the electro-polymerization of PHQ 55,56 . As illustrated in Supplementary Figure 5, when the terminal hydroquinone was oxidized to the protonated benzoquinone in the electro-oxidation process, it might react with the non-protonated hydroquinone, leading to the electro-polymerization of polymer chains. The electro-polymerization can provide additional oxidation capacity, but it doesn't affect the reversibility of redox-active groups. The parasitic reaction of electro-polymerization can be suppressed by increasing the current rate, as indicated by the Tafel plots (Supplementary Figure 6)."