Introduction

Electrochemistry involving conversion reactions is a crucial factor and a key challenge in the advancement of battery chemistries1,2,3. A major obstacle during these processes is the kinetic bottleneck caused by interfacial charge transfer among electrolytes, electrodes, and the insulating or incompatible phase materials, leading to high overpotentials and irreversible conversion4,5,6,7,8. Catalytic conversion has emerged as an effective strategy for reducing the reaction energy barrier and thus enhancing redox reaction kinetics in static batteries5,6,7,9,10. Solid-state catalysts have been mainly adopted in the catalytic conversion reaction, which are normally integrated with electrodes and exhibit fixed active sites11,12,13,14.

Aqueous flow batteries (AFBs) are among the most promising technologies for safe, long-duration, and large-scale energy storage (Fig. 1a)4,5,7,15,16,17,18,19,20,21. In particular, aqueous solid-hybrid zinc-bromide (Zn–Br) FBs have been extensively studied as low-cost systems on account of their earth abundance and high electrochemical redox potential (Zn2+/Zn and Br–/Br2 redox at − 0.763 V and − 1.08 V vs. standard hydrogen electrode (SHE), respectively)4,5,7. However, the intrinsic Br–/Br2 reactions and high electrochemical polarization from the oil phase Br-complexes lead to the sluggish conversion kinetics in Zn–Br FBs, which further limit the operating power density and lifespan of the batteries. Especially in cold/frozen environments, the challenges related to electrolyte freezing and salt precipitation undeniably exacerbate issues with Br–/Br2 reactions in Zn–Br FBs. Unlike conventional static batteries, AFBs are equipped with large volumes of electrolytes containing the active material. Unfortunately, most of the progress is in developing conventional non-flowable, solid-state catalysts, which provide only limited active sites for the redox reaction in aqueous Zn–Br FBs (Fig. 1b-left)9,22. Considering the “flowable” feature and large amount of electrolyte of the AFBs, conventional anchored solid catalyst particles, which function only at the stationary interface, cannot affect in electrolyte and may not be the ultimate catalyst design for Zn–Br FBs (Fig. 1b-right)11,23,24,25,26. A more suitable design is considered to introduce the catalytic functions to electrolytes, resulting in flowable catalytic electrolytes, which can function both at the interface and in the electrolyte.

Fig. 1: Designs and functions of conventional static catalyst particles and our flowable catalytic electrolyte.
Fig. 1: Designs and functions of conventional static catalyst particles and our flowable catalytic electrolyte.
Full size image

a Schematic diagram of AFBs with large amounts of flowable electrolytes. b Common catalytic design: conventional static solid-state catalyst particles anchored on the electrode, which only function at the interface. c Proposed catalytic electrolytes strategy: flowable catalytic active sites in electrolytes to enhance interaction both in-electrolyte and at-interface.

Quantum dots (QDs), recently awarded the 2023 Nobel Prize in Chemistry, have potential as high-performance catalysts owing to their quantum size effect, which provides a high specific surface area and numerous active sites that enhance catalytic efficiency27,28,29,30. QDs can be chemically modified to control different surface ligands, thereby enhancing catalytic activity and selectivity31,32,33. More uniquely, QDs with multifunctional ligands can form stable colloidal solutions in liquid media, facilitating direct in-electrolyte interactions with other electrolyte components to strongly regulate electrolyte properties34,35. Inspired by these merits, we hypothesized that QDs could be utilized to develop colloidal catalytic electrolytes, which are distinct from conventional electrode-anchored catalyst particles, to synergistically regulate both interface conversion kinetics and electrolyte chemistry in AFB systems by enhancing in-electrolyte and at-interface contact between redox pairs and catalytic sites (Fig. 1c). Notably, QDs with specific functional groups may interact with electrolytes to disrupt hydrogen-bonding networks, thus preventing electrolyte freezing36.

Herein, we developed colloidal catalytic electrolytes by using low-cost carbon QDs (CQDs) and used them in Zn–Br FBs as a model system. High-catalytic CQDs, as stable and highly dispersed active sites in electrolytes, enhance both in-electrolyte and at-interface interactions between the electrode and Br-based active species, thereby accelerating the Br−/Br2 reaction kinetics. The high-specific-surface CQDs with abundant functional groups (–COOH) derived through surface ligand engineering exhibit strong in-electrolyte interactions with other electrolyte components, disrupting the hydrogen-bond network prone to freezing and thereby achieving anti-freezing electrolytes. Consequently, CQD-COOH catalytic electrolytes enable Zn–Br FBs to have a long lifespan at a high current density at room temperature, outperforming conventional Zn–Br FBs. The Zn–Br FBs developed in the present study exhibited high-rate performance and extended cycling at low temperatures ( − 20 °C).

Results

Design of CQDs catalytic electrolytes

The CQDs with hydroxyl and carboxyl functional groups (CQD-OH, CQD-COOH) are synthesized by a facile pyrolysis and solvothermal method34,35. Transmission electron microscopy (TEM) images of CQD-OH and CQD-COOH revealed uniformly sized QDs with diameters of approximately ~ 11.21 nm and ~ 9.29 nm (Fig. 2a, b). High-resolution TEM (HRTEM) images further indicated lattice fringes of approximately 0.2 nm, which corresponded well with the (002) crystal facets of carbon (Supplementary Fig. 1)37. Meanwhile, the small sizes of CQD-OH and CQD-COOH were confirmed by measuring the zeta potential for the size distribution (Supplementary Fig. 2), suggesting the good dispersion of CQDs (–OH, –COOH) in electrolytes.

Fig. 2: Structure and catalytic activity characterization of CQD catalytic electrolytes.
Fig. 2: Structure and catalytic activity characterization of CQD catalytic electrolytes.
Full size image

a, b TEM and HRTEM images of CQD-OH and CQD-COOH. c FTIR spectra of CQD-OH and CQD-COOH. d NH3-TPD and CO2-TPD results of CQD-OH and CQD-COOH. The optimized structure and ELF model: (e, f) pristine carbon felt, (g, h) CQD-OH, and (i, j) CQD-COOH (Atoms: light pink/white, dark gray, and yellow/red are represented as H, C, and O, respectively). The details of the optimized structure can be seen in Supplementary Data 1. k The Gibbs free-energy profiles of Br-based reaction with pristine carbon felt, CQD-OH, and CQD-COOH (Iso-surfaces: lilac color and black blue are represented as positive and negative zones, respectively; Atoms: light pink, dark gray, and yellow are represented as H, C, and O, respectively). l Digital images of stable colloidal catalytic electrolytes. m Viscosity of BE, CQD-OH, and CQD-COOH catalytic electrolytes. n CV curves of Br-redox reaction (0.05 M ZnBr2 + 0.5 M H2SO4) based on BE, CQD-OH, and CQD-COOH catalytic electrolytes. o Arrhenius plot with corresponding Eg for the Br-redox reaction process among BE, CQD-OH, and CQD-COOH catalytic electrolytes at the onset potential of 1.8 V. The concentration of CQDs added was consistently maintained at 0.1 mg·mL–1 in panel (l–o). All error bars represent standard deviations from three independent tests.

As shown in Fig. 2c, Fourier transform infrared spectroscopy (FTIR) unveiled –OH and –COOH vibration peaks, indicating the presence of hydrophilic ligands in the CQDs38,39. Two characteristic peaks around 1350 and 1550 cm–1 in the Raman spectra can be attributed to the carbon vibrations of the D-band and G-band in the CQDs (–OH, –COOH) (Supplementary Fig. 3)40. In addition, we noted that the characteristic peaks of C–OH and C = O in the C 1 s and O 1 s high-resolution X-ray photoelectron spectroscopy (XPS) can be ascribed to hydroxyl and carboxyl functional groups in CQDs (Supplementary Fig. 4), respectively34,41. Moreover, –OH and –COOH functional groups were covalently anchored to the CQD surface through C–O and O–C = O bonds, respectively, supported by the distinct sp2/sp3 carbon framework (Supplementary Fig. 4). Collectively, these results confirm the successful synthesis of CQD-OH and CQD-COOH and the formation of colloidal electrolytes (Supplementary Fig. 5). Furthermore, temperature-programmed desorption (TPD) analysis was employed to confirm further the anchoring of diverse surface-active functional groups on CQDs42,43,44. CQD-COOH exhibited an NH3 desorption peak at approximately 380 °C (Fig. 2d-top), indicating abundant, strong acidic sites on the CQD-COOH surface. By contrast, CQD-OH displayed strong CO2 desorption peaks (Fig. 2d-bottom), revealing that it has more basic sites than CQD-COOH. These results confirmed successful functionalization, wherein CQD-COOH and CQD-OH possess plentiful COOH-derived acid sites and OH-derived base sites.

We choose Zn–Br FBs as a model system to demonstrate the effects of catalytic electrolytes. First, the active sites in Br-species reactions were systematically investigated through density functional theory (DFT) simulations. The adsorption and desorption capacities of catalysts significantly influence the reaction kinetics of active materials. Electron localization function (ELF) analysis indicated that the pristine carbon felt (CF) structure exhibited an indistinct charge distribution (Fig. 2e, f), leading to suboptimal activation of Br species45. In contrast, CQD-OH (Fig. 2g, h) exhibited enhanced electron localization around the –OH groups, which induced a localized positive charge on the adjacent carbon atoms, indicating that these positively charged sites easily anchor Br–. Similarly, for CQD-COOH (Fig. 2i, j), the delocalized electron structure of the carboxyl group led to a more uniform charge distribution, indicating that the adjacent carbon atom exhibited a relatively lower charge transfer compared to that of the –OH functional group. This result may reflect the influence of sp3-hybridized carbon arising from the incorporation of the –COOH moiety into the carbon framework (Supplementary Fig. 3)46,47. Owing to the unique architecture of CQD-COOH, its pronounced surface dipole and dipole-induced-dipole (halogen-bond-like) interactions with the highly polarizable Br2 molecule could synergistically enhance Br2 adsorption48,49. These electronic structure differences in CQDs suggested that –OH and –COOH groups serve as complementary active centers, facilitating the selective adsorption of different Br species.

According to the main mechanisms underlying the halogen evolution reaction, specifically the Volmer–Tafel and Volmer–Heyrovsky mechanisms, the Br-based reduction reaction pathway can be described by three states: *Br−, *Br2, and *Br3−. The *Br3− state is influenced by the complexing agent 1-ethyl-1-methylpyrrolidinium bromide; * denotes an adsorption site9,45,50. Gibbs free energy (ΔG) was calculated using the structural model illustrated in Fig. 2e–j, k. The initial state (*) was set at 0 eV (Fig. 2k). Ideally, the optimization criterion for ΔG values should be step-specific: while thermodynamically spontaneous steps benefit from strongly negative ΔG, non-spontaneous processes require minimized energy barriers (near-zero ΔG) to achieve balanced adsorption-desorption equilibrium in bromine electrochemistry. The ΔG value of the *Br⁻ state for the CQD-OH is more spontaneously negative than that for the CQD-COOH and pristine CFs, indicating that the site with –OH functionalization strongly stabilizes adsorbed Br⁻ due to the enhanced electron localization of positive partial charge. Furthermore, the CQD-COOH active sites require the smallest additional ΔG value for the *Br2 state, which plays a crucial role in the Br-based reduction reaction, likely because of the –COOH-functionalized surface with a strong dipole that stabilizes neutral Br2. Therefore, given the adsorption–desorption abilities of Br-based species, the CQDs as active centers exhibit optimal catalytic activity for optimizing Br redox kinetics in different steps. Given that the rate-determining step resides in the Br–-to-Br2 conversion, the synergistic interaction between interfacial charge distributions and functional groups enhances Br-species adsorption-desorption dynamics, with the catalytic performance following the order: CQD-COOH > CQD-OH > pristine CF. Moreover, to quantify how these interactions translate into reaction barriers (Supplementary Fig. 6), we carried out climbing-image nudged elastic band (CI-NEB) calculations for the Br2-to-Br⁻ reduction step. The calculated activation energies on pristine CF, CQD-OH, and CQD-COOH surfaces were 1.64 eV, 1.05 eV, and 0.86 eV, respectively. The lower barrier on CQD-COOH and CQD-OH confirmed that carboxyl groups facilitate optimized interfacial charge transfer and accelerate the conversion of Br species, which can be further confirmed by the Raman results (Supplementary Fig. 7).

Because of their small size and abundance of hydrophilic functional groups, CQD-OH and CQD-COOH can form stable colloidal catalytic electrolytes (Fig. 2l), improving interactions between electrodes and electrolyte components. Furthermore, as displayed in Fig. 2m, the viscosity of the CQD-OH and CQD-COOH catalytic electrolytes is essentially the same as that of blank electrolytes (BE), indicating that the small CQDs are well-dispersed in the electrolytes.

Building on the theoretical calculation of the distinct CQD features, we evaluated the performance of the CQD catalytic electrolytes for Br-based reactions. The cyclic voltammetry (CV) curves for all electrolytes exhibited pronounced reduction/oxidation peaks corresponding to the electrochemical reactions of Br−/Br2 (Fig. 2n)9. Firstly, a small reduction peak observed at 1.1 V (vs. Ag/AgCl) can be attributed to the Br2-to-BrO3– inevitable parasitic reaction pathway in acidic media, yet it is not the primary reaction. Specifically, the CQD-COOH catalytic electrolytes had high peak currents with the lowest redox peak potential, indicating the effective acceleration of the Br-based reaction. Similarly, CV curves for the CQD-OH catalytic electrolytes showed high peak currents, although their redox peak potentials were slightly higher than those of the CQD-COOH catalytic electrolytes. Notably, these advantages in the CV curves were apparent when compared with pristine CFs with BE. Moreover, the activation energy (Ea) is a fundamental parameter for evaluating Br−/Br2 reaction kinetics in conversion chemistry2. According to the Arrhenius equation, the Ea values (Fig. 2o) derived from the charge-transfer resistance fitting at various temperatures (Supplementary Fig. 8) are 7.48, 8.23, and 11.8 kJ·mol−1 for CQD-COOH, CQD-OH catalytic electrolytes, and BE, respectively. Electrochemical linear sweep voltammetry (LSV) of the Br redox couple (Supplementary Fig. 9) further corroborated that both oxidation and reduction currents are higher in CQD-based electrolytes than in the BE system, and the derived Tafel slopes were consistently smaller for CQD-COOH and CQD-OH than for BE. These results demonstrate that reversible Br−/Br2 reactions are thermodynamically favorable, with CQD catalytic electrolytes and CQD-COOH catalytic electrolytes exhibiting the highest catalytic activity.

Electrochemical performance of CQD catalytic electrolytes

Due to the high voltage and cost-effectiveness of Zn–Br FBs (Supplementary Fig. 10), this system is used as a model system to demonstrate the advantages of CQD catalytic electrolytes. We evaluated the electrochemical performance (at room temperature) of Zn–Br FBs fabricated using CQD-COOH catalytic electrolytes at room temperature. First, we found that Zn–Br FBs with electrolytes containing 0.1 mg·mL−1 of CQD-COOH exhibited the optimal energy efficiency (EE) of 88.65% (Supplementary Fig. 11). An excess of catalysts led to the aggregation of CQDs, thereby reducing the number of catalytic sites and impairing the ionic conductivity of the electrolytes (Supplementary Fig. 12). Thus, 0.1 mg·mL–1 CQD-COOH was identified to be the optimal concentration for subsequent investigations. The as-fabricated Zn–Br FBs with CQD-COOH catalytic electrolytes exhibited high-rate performance across a current density range of 40–180 mA·cm−2 (Fig. 3a). At a current density of 50 mA·cm−2, the Zn–Br FBs with CQD-COOH catalytic electrolytes achieved an EE of 83.95% and a Coulombic efficiency (CE) of > 98.5%. By contrast, Zn–Br FBs with BEs had an EE of 60.12% under the same testing conditions, which was further confirmed by corresponding galvanostatic charge-discharge (GCD) profiles (Fig. 3b). The results indicated a lower voltage gap in Zn–Br FBs with CQD-COOH catalytic electrolytes (0.18 V) than in those with BEs (0.36 V) at a current density of 50 mA·cm−2. When the current density was increased to 90 mA·cm−2, the EE of BE-based Zn–Br FBs declined drastically, attributable to cell failure due to the slow kinetics of the cathode and Zn-based side reactions (Supplementary Fig. 13). The BE-based Zn–Br FBs, which had a peak power density of 162.82 mW·cm−2 at a current density of 200 mA·cm−2, and the Zn–Br FBs with CQD-COOH catalytic electrolytes had a higher power density of 389.88 mW·cm−2 at a current density of 400 mA·cm−2 (Fig. 3c).

Fig. 3: Electrochemical performance of Zn–Br FBs at room temperature.
Fig. 3: Electrochemical performance of Zn–Br FBs at room temperature.
Full size image

a CE and EE of Zn–Br FBs with BE and CQD-COOH catalytic electrolytes under 40 to 180 mA·cm−2 (Each of the negolyte and posolyte contains 10 ml of 3 M KCl + 2 M ZnBr2 + 0.4 M MEP, polyolefin porous-membrane, 4 cm2 membrane area). b GCD profiles at 50 mA·cm−2 of Zn–Br FBs with BE and CQD-COOH catalytic electrolytes. c Discharge polarization curves with corresponding power density curves of Zn–Br FBs with BE and CQD-COOH catalytic electrolytes. d Cycling performance of Zn–Br FBs with BE and CQD-COOH catalytic electrolytes at 100 mA·cm−2 with an areal capacity of 20 mAh·cm−2. e The representative voltage profiles in the region of 0 − 70 h and 1900 − 1970 h of Zn–Br FBs with CQD-COOH catalytic electrolytes. f Cycling performance of Zn–Br FBs with BE and CQD-COOH catalytic electrolytes at a current density of 80 mA·cm−2 and an areal capacity of 40 mAh·cm−2. g Comparison of Zn–Br FBs in terms of current density, EE, and lifespan, where the symbol size shows the corresponding working time. The concentration of CQDs added was consistently maintained at 0.1 mg·mL−1. The source of the literature data shown in this figure can be found in the Supplementary Information, Table 1.

On the other hand, the Zn–Br FBs with CQD-COOH catalytic electrolytes maintained an EE of 79.5% and a CE of approximately 98.7% (Fig. 3d), with a capacity decay rate of 0.00107%/cycle over 5000 cycles (areal capacity: 20 mAh·cm–2; current density: 100 mA·cm−2). Conversely, the Zn–Br FBs with BEs exhibited poor cycling performance, lasting only approximately 150 cycles, with low EE (73.2%) and rapid capacity decay (Fig. 3d and Supplementary Fig. 14). The problems with BE-based FBs are primarily attributable to sluggish Br-based reaction kinetics and dendrite growth, which ultimately lead to battery failure. The GCD profiles presented in Fig. 3e and Supplementary Fig. 14 further confirmed the reduced voltage loss and enhanced overall stability of the Zn–Br FBs with CQD-COOH catalytic electrolytes during cycling. To evaluate the potential of high-energy Zn–Br FBs at higher areal capacities and current densities, we tested the Zn–Br FBs with CQD-COOH catalytic electrolytes. Long-cycling performance, along with corresponding GCD curves at an areal capacity of 40 mAh·cm−2 and a current density of 80 mA·cm−2, as displayed in Fig. 3f and Supplementary Fig. 15. The Zn–Br FBs with CQD-COOH catalytic electrolytes exhibited prolonged operation over 2000 cycles under high-areal-capacity conditions, with a CE of approximately 98.7% and an EE of approximately 82.4% (Fig. 3f), outperforming the BE-based FBs, which had an EE of ~ 75.7%. These results may be attributable to the colloidal nature of CQD-COOH catalytic electrolytes and the enhanced Br-based interactions facilitated by the abundant carboxyl active sites, which improved the interfacial Br−/Br2 catalytic reaction efficiency. This result is consistent with the characterization and theoretical calculations presented in Fig. 2.

To demonstrate the universal applicability of CQD catalytic electrolytes in improving the performance of Zn–Br FBs, we performed a systematic comparison by using CQD-OH at the same concentration (0.1 mg·mL−1) to form a homogeneous colloidal solution. The Zn–Br FBs with CQD-OH catalytic electrolytes exhibited higher rate performance and peak power density than did the pristine Zn–Br FBs with BEs (Supplementary Figs. 16–19). Furthermore, the Zn–Br FBs with CQD-OH catalytic electrolytes exhibited improved EE and stability over 5000 cycles (Supplementary Fig. 18), with a decay rate of 0.00201%/cycle during long-term cycling tests. These Zn–Br FB cells exhibited durable operation over 2000 cycles with a CE of approximately 98.3% and an EE of approximately 79.7% at an areal capacity of 40 mAh·cm−2 (Supplementary Fig. 19).

Subsequently, we compared various performances (e.g., EE, current density, and lifespan) between our CQD catalytic electrolytes and previously reported strategies, particularly conventional electrode-anchored catalysts for Zn–Br FB systems (Fig. 3g and Supplementary Table 1). The Zn–Br FBs with CQD catalytic electrolytes had an EE of 79.5% at high current density and a lifespan of 1970 h, which were higher than most relevant reported value. Thus, CQD catalytic electrolytes effectively enhance Br-based reaction kinetics and maintain highly active catalytic sites even under conditions of high current density or high areal capacity.

Anti-freezing effect of CQD catalytic electrolytes

Large-scale energy storage systems are particularly vulnerable to seasonal variations. The inefficiency of batteries under low-temperature conditions poses a persistent challenge. To address this problem, we systematically analyzed the effect of electrolytes in FBs on their efficiency under low-temperature conditions. The blank solution used in this study exhibited a noticeable freezing phenomenon when the temperature decreased to − 30 °C (Fig. 4a). By contrast, the CQD catalytic electrolytes exhibited anti-freezing properties, maintaining a flowable state even at the same low temperature (Fig. 4a and Supplementary Fig. 21). Notably, although the addition of large amounts of salt in the BE can improve its low-temperature performance compared to DI water, the electrolyte froze immediately at – 22 °C (Supplementary Fig. 22), indicating that ice crystals had already formed around – 20 °C. This suggested a reduction in low-temperature performance below – 20 °C.

Fig. 4: The electrolyte structures, including H-bonds, the surface functional groups of CQDs, and ion interactions.
Fig. 4: The electrolyte structures, including H-bonds, the surface functional groups of CQDs, and ion interactions.
Full size image

a The digital photographs of the blank solution, CQD-OH, and CQD-COOH catalytic electrolytes with 0.1 mg·mL−1 concentration at − 30 °C. b The fitted O − H stretching vibration represents water molecules with υStrong H-bond, υMedium H-bond, and υWeak H-bond. c The proportion of υStrong H-bond and υWeak H-bond as a function of different CQDs ( − COOH, −OH) concentration-based catalytic electrolytes. d DSC test of CQD-OH and CQD-COOH catalytic electrolytes with 0.1 mg mL–1 concentration from – 50 to 5 °C. e The ionic conductivities of different CQDs ( − COOH, −OH) concentrations at the temperature of – 30 °C. f The ionic conductivities of the CQD-OH and CQD-COOH with 0.1 mg·mL–1 concentration in the temperature range of – 30–25 °C. g The schematic of the structure evolution of CQDs catalytic electrolytes with abundant anti-freezing functional groups at low temperatures. All error bars represent standard deviations from three independent tests.

Regulating the number of hydrogen bonds and reducing the proportion of highly hydrogen-bound water molecules can help prevent water from freezing. To elucidate the mechanisms underlying the anti-freezing properties of CQDs, we investigated (through FTIR spectroscopy) the effect of CQD concentration on water molecule bonding51,52. As shown in Fig. 4b, the broad peak corresponding to the –OH stretching vibration of water could be deconvoluted into three subpeaks representing υStrong H-bond, υMedium H-bond, and υWeak H-bond stretch. The first two stretches were associated with fully hydrogen-bonded water molecules. The υStrong H-bond corresponded to –OH stretch for tetrahedrally coordinated water ( ~ 3200 cm–1), whereas υMedium H-bond corresponded to –OH stretch for not fully coordinated water ( ~ 3400 cm−1)37. Furthermore, υWeak H-bond stretch ( ~ 3600 cm−1) corresponded to water molecules where one –OH oscillator was hydrogen-bonded to another molecule, whereas the other –OH oscillator remained free, forming a weakly hydrogen-bonded structure37,53.

With increasing CQD concentration in the catalytic electrolytes, the intensity of υStrong H-bond stretch decreased, whereas that of the υWeak H-bond stretch increased, indicating that the CQD catalytic electrolytes effectively disrupted the hydrogen-bonding network of water molecules (Fig. 4c and Supplementary Figs. 23 and 24). This phenomenon may be attributable to the strong interactions between the oxygen-containing functional groups on the surface of CQDs and water molecules31,32. Notably, the CQD-COOH catalytic electrolytes exhibit a higher content of υWeak H-bond stretch compared with CQD-OH catalytic electrolytes, demonstrating a stronger interaction between the highly polar carboxyl functional groups and water molecules. To further investigate the effect of the developed electrolytes at low temperatures, we measured their freezing point through differential scanning calorimetry (DSC). The freezing points of CQD-OH-0.1 and CQD-COOH-0.1 catalytic electrolytes were approximately − 24.2 °C and − 26.1 °C, respectively (Fig. 4d and Supplementary Fig. 25). Therefore, CQD-COOH catalytic electrolytes exhibit the greatest anti-freezing effect among the Zn–Br FB-based electrolytes, which is consistent with the FTIR results.

The ionic conductivities of the CQD catalytic electrolytes increased progressively with increasing concentrations at − 30 °C (Fig. 4e and Supplementary Fig. 26). This trend underscores the anti-freezing properties of CQD catalytic electrolytes, suggesting high ionic conductivity under low-temperature conditions. Notably, the ionic conductivity of the CQD-COOH system was twice as high as that of the CQD-OH system when the same concentration was used at − 30 °C (Fig. 4e). This difference highlights stronger interactions between the highly polar –COOH groups in CQD-COOH and water molecules. This phenomenon was further supported by the comparison of ionic conductivity across different electrolytes at varying temperatures, ranging from 25 °C to − 30 °C (Fig. 4f and Supplementary Fig. 27). The CQD catalytic electrolytes consistently exhibited high ionic conductivities, with values of 15.62 mS·cm−1 for CQD-COOH-0.1 and 6.96 mS·cm−1 for CQD-OH-0.1 at − 30 °C. Therefore, the original hydrogen-bond structure of water was disrupted by the strong interactions of CQD-COOH and CQD-OH with their abundant functional groups and water molecules (Fig. 4g and Supplementary Figs. 28 and 29). This disruption inhibited the formation of salt-ice crystals, which typically occur at low temperatures. The CQD catalytic electrolytes promoted the formation of weak hydrogen-bond configurations, effectively reducing the freezing point of water. The enhanced ion interactions help maintain the ionic conductivity of the electrolyte, ensuring that Zn–Br FB systems can operate efficiently under low-temperature conditions.

Low-temperature electrochemical performance of CQD catalytic electrolytes

To assess the potential application of the CQD catalytic electrolytes in FB systems in cold environments, we evaluated the electrochemical performance of Zn–Br FBs at low temperatures (both the reservoirs and cells were maintained at − 20 °C) in Supplementary Fig. 30. The CV curve for CQD-COOH catalytic electrolytes exhibited distinct peak currents with low redox peak potentials (Fig. 5a), indicating that multifunctional electrolytes retain high Br-based reaction kinetics at low temperatures. Moreover, the Zn–Br FBs with CQD-COOH catalytic electrolytes (Fig. 5b, c) exhibited stable cycling during the variable temperature test (0 °C to − 25 °C), with EEs of 84.7%, 82.6%, 80.6%, and 76.1% at 0 °C, − 10 °C, − 20 °C, and − 25 °C, respectively. The results suggest that CQD-COOH catalytic electrolytes possess anti-freezing properties and maintain high-efficiency catalytic activity at low temperatures. The Zn–Br FBs with CQD-COOH catalytic electrolytes delivered a high power of 200.18 mW·cm−2 (Fig. 5d). At low temperatures, these FBs exhibited high rate performance ranging from 30 to 120 mA·cm−2 (Fig. 5e, f). Furthermore, they maintained stable cycling operation over 1900 cycles, with an EE of 74.2% at a current density of 40 mA·cm−2 at low temperatures (Fig. 5g). This finding further confirms the advantages of our electrolytes with high-efficiency catalytic active sites and anti-freezing properties.

Fig. 5: Electrochemical performance of Zn–Br FBs at low temperature.
Fig. 5: Electrochemical performance of Zn–Br FBs at low temperature.
Full size image

a CV curves of CQD-COOH catalytic electrolytes at a low temperature of − 20 °C. b CE and EE of Zn–Br FBs with/without CQD-COOH catalytic electrolytes at a low temperature ranging from 0 to − 25 °C (Each of the negolyte and posolyte contains 10 ml of 3 M KCl + 2 M ZnBr2, polyolefin porous-membrane, 4 cm2 membrane area) with (c) corresponding GCD profiles. d Discharge polarization curves with corresponding power density curves of CQD-COOH catalytic electrolytes in Zn–Br FBs at − 20 °C. e CE and EE of Zn–Br FBs with/without CQD-COOH catalytic electrolytes under 30 to 120 mA·cm−2 at − 20 °C with (f) corresponding GCD profiles. g Long cycling Zn–Br FBs using CQD-COOH catalytic electrolytes and BE at − 20 °C with corresponding GCD profiles in the inset. The concentration of CQDs added was consistently maintained at 0.1 mg·mL−1.

Similarly, the CV curve, anti-freezing performance, rate performance, voltage profiles, power density, and cycling stability of the Zn–Br FBs with CQD-OH catalytic electrolytes at − 20 °C are presented in Supplementary Figs. 31–35). These Zn–Br FBs also exhibited a strong anti-freezing ability at − 25 °C, a rate performance value of 30–120 mA·cm−2, and a power density of 170.96 mW·cm−2 at a current density of 180 mA·cm−2 at − 20 °C. Notably, they further exhibited decent cycling stability over 1400 cycles (over 1400 h) at a current density of 40 mA·cm−2, with an EE of 72.2% at a temperature of − 20 °C. Compared to previously reported strategies at low temperature conditions, Zn–Br FBs incorporating CQD catalytic electrolytes demonstrated enhanced energy efficiency and prolonged cycling durations (Supplementary Fig. 36 and Table 2). These results indicate that our catalytic electrolytes can enhance the electrochemical performance of electrolytes and confer an anti-freezing benefit, thereby improving the environmental suitability of Zn–Br FBs for potential large-scale energy storage systems.

Discussion

In summary, we successfully designed multifunctional CQD–based flowable catalytic electrolytes, which are distinct from conventional electrode-anchored solid catalyst particles. These highly active catalytic electrolytes achieve catalysis both in-electrolyte and at the interface between Br-redox pairs and the catalytic active sites, improving Br−/Br2 reaction kinetics at both room and low temperatures for FBs. We used the Zn–Br FBs as the model systems to integrate CQD-COOH catalytic electrolytes, achieving a power density of 389.88 mW·cm−2, stable cycling > 5000 cycles at 100 mA·cm−2, and a long lifespan of > 1966 h with an EE of 82.4% at room temperature. Even at − 20 °C, our Zn–Br FBs using CQD-COOH catalytic electrolytes also exhibited stable cycling of > 1900 cycles, with an EE of approximately 74.2% at 40 mA·cm−2. Furthermore, Zn–Br FBs with CQD-OH catalytic electrolytes had been demonstrated to exhibit similar battery performance, which achieved high EE at 100 mA·cm−2 with stable cycling of > 5000 cycles at ambient temperatures (25 ± 3 °C). These FBs also exhibited anti-freezing performance for long cycles ( > 1400 cycles) at 40 mA·cm−2 and a low temperature of − 20 °C. This work can be a suitable catalyst design for flow batteries, that is, flowable catalytic electrolytes for flowable batteries.

Methods

Materials

All chemicals were used as received. Zinc bromide (ZnBr2, ≥ 98%), potassium chloride (KCl, ≥ 99%), 1-ethyl-1-methylpyrrolidinium bromide (C7H16BrN, ≥ 99%), citric acid (C6H8O7, ≥ 99.5%), pyrene (C16H10, ≥ 98%), nitric acid (HNO3) and sodium hydroxide (NaOH, ≥ 98%) were purchased from Sigma-Aldrich. Carbon felt (CF, 4.6 mm, carbon ≥ 99 %, bulk density 0.12–0.14 g·cm−2) was purchased from Yi Deshang Carbon Technology. Polyolefin-based porous membrane was purchased from Daramic® HP (Thickness: ~ 0.2 mm; Porosity: 60%).

Synthesis of carbon quantum dots with different functional groups (CQD-OH/COOH)34,41

CQD-OH: The synthesis of CQD-OH began with the nitration of pyrene (0.2 g) in hot HNO3 (16 mL) at 80 °C under reflux with continuous stirring for 12 h. After cooling, the mixture was quenched with deionized water (100 mL) and filtered using a 0.22 μm membrane to eliminate residual acid. The obtained yellow 1,3,6-trinitropyrene (0.3 g, 90% yield) was dispersed in NaOH aqueous solution (60 mL, 0.2 M) via ultrasonication for 2 h. This suspension was transferred to a Teflon-lined autoclave (100 mL) and heated at 200 °C for 10 h. After cooling, the product was filtered through a 0.22 μm membrane to remove insoluble carbon materials and further dialyzed (MWCO: 3500 Da) for 48 h to exclude sodium salts and small molecular impurities. The purified black CQD-OH were dried at 80 °C for subsequent characterization.

CQD-COOH: In a standard preparation, citric acid (100 g) was placed in a glass beaker covered with a glass slide and heated at 180 °C for 150 mins in air. The resulting yellow solid was collected and dispersed in water under stirring for 10 h. The suspension was then filtered through a 0.2 μm syringe filter and subjected to freeze-drying, yielding CQD-COOH as a yellow-orange powder.

Electrochemical characterization

Cyclic voltammetry (CV) measurements were carried out using a CHI 760E electrochemical workstation equipped with a three-electrode configuration, employing a glassy carbon electrode (6 mm diameter) as the working electrode, a graphite plate (2 × 2 cm2) as the counter electrode, and an Ag/AgCl electrode as the reference electrode. The CV stability was assessed within a potential window of 0 to 1.5 V in an electrolyte consisting of 0.05 M ZnBr2 and 0.5 M H2SO4, at a scan rate of 10 mV·s−1, to examine the reversibility of Br−/Br3− redox reaction on electrodes.

The activation energy for the electrode reaction in a Zn–Br flow cell was determined through variable-temperature electrochemical impedance spectroscopy (EIS). Each half-cell contained 10 mL of electrolyte comprising 3 M KCl, 2 M ZnBr2, 0.4 M MEP, and 0.1 mg·mL−1 CQDs, separated by a polyolefin porous membrane with an active area of 4 cm2. Impedance spectra were collected over a temperature range of 253 K to 313 K (− 20 °C–40 °C), allowing 10 minutes of thermal equilibration at each step. Measurements were conducted with an applied DC potential of 1.8 V, across frequencies from 100,000 Hz to 0.01 Hz, and an AC amplitude of 5 mV.

Ionic conductivity of the liquid electrolyte (100 μL) was evaluated via EIS using a symmetric coin-type cell (CR2032) with titanium spacers (purity ≥ 99.9%, thickness 1.0 mm, area ≈ 1.96 cm2) as blocking electrodes and a glass fiber separator (Whatman GF/D, thickness ~ 675 μm). as blocking electrodes and a glass fiber separator (Whatman GF/D, ~ 675 μm thick). The EIS was performed with an initial potential of 0 V, frequency spanning from 100,000 Hz to 0.1 Hz, and a 5 mV perturbation amplitude. Each test was repeated three times. The ionic conductivity (σ, in mS·cm−1) was derived from the bulk resistance (R, in Ω) taken at the minimum of −Z″ in the Nyquist plot, using the relation σ = (1/R) · (L/A), where L and A represent the thickness of the separator and the area of the spacer, respectively.

For ambient/room temperature tests (25 ± 3 °C), both posolyte and negolyte solutions were prepared in deionized water with 3 M KCl, 2 M ZnBr2, and 0.4 M MEP, with or without CQDs at specified concentrations. Low-temperature electrolytes omitted MEP and consisted of 3 M KCl and 2 M ZnBr2, with or without CQDs. CQD-enhanced electrolytes were ultrasonicated for 15 min to improve dispersion. The Zn–Br flow battery was constructed using stainless-steel frames to secure pre-treated 3D carbon felt electrodes (4 cm2 geometric area, 3 mm thickness) and form flow channels. Electrolytes were circulated at a rate of 60 mL·min−1 using a peristaltic pump (Chuang Rui Precision Pump). All cell tests were performed in a temperature-controlled environment.

Galvanostatic testing of the Zn–Br flow batteries was conducted on a LAND CT2001A system at both room temperature (23–25 °C) and low temperature (− 30–0 °C). Current densities ranged from 40 to 120 mAh·cm−2. Charging was controlled by a fixed areal capacity of either 20 mAh·cm−2 or 40 mAh·cm−2, while discharging ended at a cut-off voltage of 0.8 V. The theoretical capacity was based on the posolyte (bromine side), which served as the capacity-limiting electrode.

Characterizations

The crystal structure was studied by X-ray diffraction (XRD, X’Pert Pro MPD, Philips, Holland) using Cu Kα as the radiation source under 40 kV and 40 mA. Morphologies were probed by Transmission electron microscopy (TEM, FEI Talos F200x). X-ray photoelectron spectroscopy (XPS) spectra were recorded on a photoelectron spectrometer (ESCALAB 250, Thermo Scientific, America), where the binding energy (BE) of the elements was calibrated by the BE of C 1 s (284.60 eV). The differential scanning calorimeter (DSC) was carried out in Netzsch DSC 200 F3 in the procedure of − 80–200 °C with a cooling rate of 10 °C·min−1, constant temperature for 2 mins, and − 80–200 °C with a heating rate of 10 °C·min−1. The Fourier-transform infrared (FT-IR) spectroscopy was performed using a PerkinElmer ATR-FTIR spectrometer. The CO2 temperature-programmed desorption (TPD) and NH3-TPD analysis were conducted on Microtrac BELCat II. The samples were pretreated in He flow at 60 °C with a rate of 30 mL·min−1 for 0.5 h, cooled to room temperature, and then swept in CO2 (NH3) flow with a rate of 30 mL·min−1 for 0.5 h. After treatment in He flows for 1 h at room temperature to remove physical adsorption, the sample was raised at a heating rate of 15 °C·min−1 to 500 °C. The signals were monitored by a TCD detector. The size distribution of CQDs was probed by the nano sizer/Zeta-potential tester (Malvern Zetasizer Nano ZS90). Ex-situ Raman characterization: the electrolyte was carefully extracted from the cycled cell immediately after charging for 10 mins. The sample was transferred using a syringe and loaded into a sealed quartz cell with a Teflon cap to prevent corrosion. Raman spectra were acquired within 5 mins of sampling using a WITec RAMAN alpha 300 R spectrometer equipped with a 532 nm laser. All measurements were performed at room temperature under ambient conditions without intentional heating or cooling.

Density functional theory (DFT) calculation

All calculations were performed within the framework of Density Functional Theory (DFT) using the Vienna Ab-initio Simulation Package (VASP) in conjunction with the projector augmented wave (PAW) method54,55,56,57. The exchange–correlation functional was treated using the Perdew–Burke–Ernzerhof (PBE) formulation under the generalized gradient approximation (GGA)58. A vacuum layer of 10 Å was applied normal to the surface to minimize interlayer interactions. The plane-wave basis set was expanded with a kinetic energy cutoff of 500 eV, and Brillouin zone integration was carried out using a 4 × 4 × 1 Monkhorst–Pack k-point grid59. The electronic and ionic relaxation convergence criteria were set to 10−4 eV and 0.02 eV·Å−1, respectively. The Gibbs free energy change (ΔG) for each reaction step was computed according to the equation:

$$\Delta {{{\rm{G}}}}=\Delta {{{\rm{E}}}}+\Delta {{{\rm{ZPE}}}}{{{-}}}{{{\rm{T}}}}\Delta {{{\rm{S}}}}\,({{{\rm{T}}}}=298.15{{{\rm{K}}}})$$
(1)

where ΔE represents the change in total energy, ΔZPE is the difference in zero-point energy, and ΔS denotes the change in entropy.

The zero-point energy and entropy corrections were derived from the NIST-JANAF thermochemical tables for gas-phase species and from vibrational frequency calculations for adsorbed intermediates. The reaction energy barriers for Br2 reduction on CF, CQD-OH, and CQD-COOH were determined via the climbing-image nudged elastic band (CI-NEB) method to assess the kinetics of the delithiation process.