Robust high-temperature potassium-ion batteries enabled by carboxyl functional group energy storage

Significance

Distinctively different from the popularly reported works, an energy storage mechanism is proposed for exploring robust high-temperature potassium-ion batteries (PIBs) with high cycle stability. This is based on an example of p-phthalic acid with two carboxyl functional groups as the redox centers. The cycle stabilities achieved under both low and high current densities are the best among those of high-temperature PIBs reported previously.

Abstract

The popularly reported energy storage mechanisms of potassium-ion batteries (PIBs) are based on alloy-, de-intercalation-, and conversion-type processes, which inevitably lead to structural damage of the electrodes caused by intercalation/de-intercalation of K⁺ with a relatively large radius, which is accompanied by poor cycle stabilities. Here, we report the exploration of robust high-temperature PIBs enabled by a carboxyl functional group energy storage mechanism, which is based on an example of p-phthalic acid (PTA) with two carboxyl functional groups as the redox centers. In such a case, the intercalation/de-intercalation of K⁺ can be performed via surface reactions with relieved volume change, thus favoring excellent cycle stability for PIBs against high temperatures. As proof of concept, at the fixed working temperature of 62.5 °C, the initial discharge and charge specific capacities of the PTA electrode are ∼660 and 165 mA⋅h⋅g⁻¹, respectively, at a current density of 100 mA⋅g⁻¹, with 86% specific capacity retention after 160 cycles. Meanwhile, it delivers 81.5% specific capacity retention after 390 cycles under a high current density of 500 mA⋅g⁻¹. The cycle stabilities achieved under both low and high current densities are the best among those of high-temperature PIBs reported previously.

With the consumption of energy, advanced green energy systems with high specific capacity, long-term cycle stability, and high power/energy density are highly desired (1–3). In terms of the abundant reserves of sodium with similar chemical properties to Li, sodium ion batteries (SIBs) are expected to replace the currently popular lithium ion batteries (LIBs) (4, 5). Unfortunately, with respect to the higher negative redox potential of Na⁺ than that of Li⁺, the energy densities of SIBs are still relatively low (6). As a promising alternative, potassium-ion batteries (PIBs) have lower negative redox potential than SIBs (−2.93 vs. −2.71 V), implying their higher energy/power densities. Moreover, K⁺ has weaker Lewis acidity compared with both Li⁺ and Na⁺, indicating that the ionic conductivity of PIBs is better than that of LIBs and SIBs (7). What is more, PIBs can use aluminum instead of copper foil as the current collector to reduce costs (8, 9), representing their more exciting commercial applications, especially because K⁺ resources are much richer than Li⁺ (2.09 vs. 0.0017 wt%) (10, 11).

Unfortunately, the intrinsically large K⁺ radius makes it highly difficult to intercalate/de-intercalate in electrode materials with sluggish kinetics (12, 13), which causes PIBs to often deliver low-rate performances and low specific capacities with poor cycle stabilities, especially under high-temperature and harsh working conditions with a significantly accelerated K⁺ intercalation/de-intercalation process. For instance, Zhang and coworkers (14) reported that KVPO₄ could enable PIBs to work up to 55 °C. Wang and coworkers (15) reported that an azo group as the redox center favored the stable charge/discharge of PIBs up to 60 °C. Nevertheless, the cycling stability was still poor for PIBs. The reported works suggested that PIBs could run for 50 cycles at 50 °C (14) and 80 (15) and 60 (16) cycles at 60 °C, respectively. This might be mainly attributed to conventional K⁺ storage mechanisms, namely, alloy-, deintercalation-, and conversion-type processes (7, 17–19), which can facilitate structural damage with rapid capacity degradation induced by the large volume change under high temperatures (15, 20, 21).

So, to overcome the poor cycle stability of PIBs, it might fundamentally rely on the progress of K⁺ storage mechanisms. Here, we report a K⁺ storage mechanism toward the exploration of robust PIBs against high temperatures, which is based on a carboxyl functional group with two carboxyl functional groups as the redox centers. In such cases, the intercalation/de-intercalation of K⁺ can be performed via surface reactions between carboxyl functional groups and K⁺ with relieved volume change, thus favoring the specific capacity of PIBs against high temperatures. Correspondingly, as an example for PIB electrodes based on p-phthalic acid (PTA) materials, when operated at 62.5 °C, their initial discharge and charge specific capacities are ∼660 and 165 mA⋅h⋅g⁻¹, respectively, at a current density of 100 mA⋅g⁻¹, with 86% specific capacity retention after 160 cycles. Meanwhile, the specific capacity holds 81.5% retention after 390 cycles under a high current density of 500 mA⋅g⁻¹. Such achieved cycle stabilities under both low and high current densities are state of the art among those of PIBs reported previously.

Results and Discussion

Fig. 1 A–C presents typical scanning electron microscope (SEM) images of commercial PTA under different magnifications, demonstrating that they are composed of numerous small blocks (also see SI Appendix, Fig. S1). The recorded X-ray diffraction (XRD) pattern (Fig. 1D) indicates that PTA is highly crystalline. Fig. 1E shows the typical thermogravimetric analysis (TGA) curve of PTA, demonstrating that it can be stable against high temperatures up to 300 °C. SI Appendix, Fig. S2 and Fig. 1F show the representative transmission electron microscope (TEM) images under different magnifications, and the corresponding high-resolution TEM (HRTEM) image is given in SI Appendix, Fig. S3A. SI Appendix, Fig. S3 B–D presents the energy-dispersive X-ray spectroscopy (EDX) characterizations, suggesting that PTA consists mainly of C and O elements with uniform spatial distributions.

Fig. 1.

(A–C) Typical SEM images of PTA under different magnifications. (D) Typical XRD pattern of PTA. (E) TGA analysis of PTA. (F) Typical TEM image of PTA.

Fig. 2 demonstrates the electrochemical performance of the PTA electrode at a high temperature of 62.5 °C, which is a record among high-temperature PIBs reported previously. Fig. 2A gives the cyclic voltammetry (CV) of PTA at a scan rate of 0.5 mV⋅s⁻¹ with voltages ranging from 0.15 to 3.0 V. In the first scan, there are three irreversible cathodic peaks at ∼2.0, ∼1.5, and ∼1.0 V, respectively. These irreversible cathodic peaks indicate that an irreversible electrode side reaction has occurred, accompanied by the formation of solid electrolyte interphase (SEI) films (22). Over the subsequent cycles, the cathodic peaks are shifted by 0.3 V, clarifying that K⁺ has been embedded into PTA with the formation of K-intercalated compounds after discharge. In the anodic scan, one spike-like peak centered at ∼0.75 V is observed, which can be attributed to the depotassiation of K-PTA. The nearly overlapped cathodic/anodic peaks witness its excellent cyclic stability and tiny polarization. Fig. 2B shows the charge/discharge slopes of the PTA electrode over 0.15 to 3.0 V at a current density of 100 mA⋅g⁻¹. The charge and discharge plateaus are ∼0.75 and ∼0.3 V, respectively, both of which are lower than those of anode materials working under high temperatures reported previously (SI Appendix, Table S1). The low voltage platform could guarantee a high output potential, leading to a desired high energy density (23). As shown in SI Appendix, Fig. S4, the initial discharge and charge specific capacities are ca. ∼660 and 165 mA⋅h⋅g⁻¹, respectively, which are fundamentally enhanced in terms of those under room temperature (RT) (SI Appendix, Fig. S5A; i.e., ∼320 and 35 mA⋅h⋅g⁻¹). The nearly overlapped charge/discharge slopes further verify its excellent cyclic stability with tiny polarization, which is consistent with the CV results (Fig. 2A). Fig. 2C shows the rate performance of the PTA electrode, representing specific capacities of 200, 180, 150, and 100 mA⋅h⋅g⁻¹ at current densities of 100, 200, 300, and 500 mA⋅g⁻¹, respectively. Even at a high current density of 1,000 mA⋅g⁻¹, the electrode still holds a specific capacity of 70 mA⋅h⋅g⁻¹, demonstrating excellent high-rate performance. Moreover, once the current density returns to 100 mA⋅g⁻¹, the electrode still maintains 190 mA⋅h⋅g⁻¹, further underscoring its excellent rate performance. Fig. 2 D and E gives the cycling performances and coulombic efficiencies of the PTA electrode. It discloses that the as-constructed PIBs have a specific capacity of 142 mA⋅h⋅g⁻¹ with 86% retention after 160 cycles at a low current density of 100 mA⋅g⁻¹. Furthermore, the specific capacity delivers 81.5% retention after 390 cycles at a high current density of 500 mA⋅g⁻¹. What is more, regardless of the applied low and high current densities, the coulombic efficiencies are almost 100% over every cycle. The recorded cycling stabilities under both low and high current densities are state of the art among high-temperature K⁺ batteries reported previously (14, 15).

Fig. 2.

Electrochemical performance of the PTA electrode at 62.5 °C. (A) CV curves at a scan rate of 0.5 mV⋅s⁻¹. (B) Charge/discharge curves at a current density of 100 mA⋅g⁻¹ at the first, second, fifth, and 10th cycles. (C) Rate performance at different current densities. (D and E) Cycling performances and coulombic efficiencies at low and high current densities of 100 and 500 mA⋅g⁻¹, respectively.

To demonstrate the K⁺ storage mechanism of PIBs based on PTA electrodes, Fourier transform infrared (FTIR) spectra before and after full discharge are investigated, as shown in Fig. 3A. It seems that the C=O bond becomes weaker after full discharge, implying that the C=O bonds should act as the redox centers. That is to say, K⁺ is stored in the carboxyl functional group and will take part in the electrochemical redox reaction. From the FTIR spectra, the red shift of the wavenumber along with the weakened C=O in the discharging process implies that electronic density around oxygen should be higher, thus leading to a higher frequency (or faster infrared vibration) and increased bonding energies. Accordingly, electronic density around oxygen would be improved, which allows the high electronegative oxygen to capture electrons easily during discharge, resulting in the gradual increase of bonding energy between C−O⁻ and K⁺ with decreased C=O (24–27). Fig. 3B shows the typical X-ray photoelectron spectroscopy (XPS) survey scan of PTA after full discharge, confirming that the electrode contains 16.34, 40.53, and 43.12 at% of K, O, and C in terms of those before full discharge (SI Appendix, Fig. S6A), respectively. There is a new peak of K2p within the XPS survey scan after full discharge, representing the reaction of K⁺ with PTA during the electrochemical redox reaction. The other two peaks could be assigned to S2s and N1s, owing to the decomposition of bisfluorosulfonylimide potassium (KFSI) salt (28). As shown in Fig. 3 C and D, the high-resolution XPS spectrum of C1s could be fitted by two main peaks at 284.6 and 289.2 eV, corresponding to the C=O and COOH– bonds, respectively (29, 30). After full discharge, it could be deconvoluted into three types of C species, which peaked at 284.6, 285.5, and 286.8 eV, corresponding to the C=O, C–C, and C–O bonds, respectively (29, 31). This shows that the intensity of the C=O bond is reduced after full discharge, indicating that the electrochemical redox reaction occurs in the carboxyl functional group. Fig. 3 E and F provides the O1s spectrum of PTA before and after full discharge. Before discharge, the O1s spectrum could be divided into two main peaks at 531.7 and 532.5 eV, attributed to the C=O and C–OH/C–O–C bonds, respectively (22). After full discharge, regardless of the fact that they are still located at 531.7 and 532.5 eV, but the intensity of the C=O bond is weakened after full discharge, similar to that observed for the C1s spectrum, further confirming that K⁺ storage should occur in the surface carboxyl functional groups. Compared to that before full discharge (SI Appendix, Fig. S6A), the spectrum of K2p of the PTA electrode after full discharge (SI Appendix, Fig. S6B) could be divided into two peaks at 292.8 and 295.7 eV, assigned to K2p3/2 and K2p1/2, respectively, clarifying that K⁺ is inserted and extracted within PTA during the electrochemical redox reaction (32). Notably, such K⁺ storage is distinctively different from the typically reported mechanism, namely, alloy-, de-intercalation-, and conversion-type processes (7, 17–19, 33), which is based on the carboxyl functional group with two carboxyl functional groups as the redox centers. In such case, the intercalation/de-intercalation of K⁺ would be performed via surface reactions between carboxyl functional groups and K⁺ with relieved volume change, thus endowing the specific capacity of PIBs against high temperatures.

Fig. 3.

(A) Typical FTIR patterns of PTA before and after full discharge. (B) Typical XPS survey spectrum of PTA after full discharge. (C) High-resolution XPS spectrum of C1s before full discharge. (D) High-resolution XPS spectrum of C1s after full discharge. (E) High-resolution XPS spectrum of O1s before full discharge. (F) High-resolution XPS spectrum of O1s after full discharge.

To further clarify the mechanism of energy storage, the reaction kinetics of PTA is investigated according to the recorded CV profiles at various scan rates, as shown in Fig. 4A. The value of b (used to determine whether there is surface-controlled electrochemical behavior during the charge/discharge process) can be calculated based on the following (Eq. 1) (34, 35):

$$i=a\times {\text{v}}^b.$$ [1]

Accordingly, as shown in Fig. 4B, the anodic and cathodic peaks display b values of 0.89 and 0.884, both of which are close to 1. This implies that the charge/discharge processes are mainly driven by a surface-controlled electrochemical reaction, further reflecting that K⁺ storage happens in the surface carboxyl groups over the electrochemical reaction. Such a surface-controlled K⁺ storage mechanism could be responsible for the excellent high-temperature electrochemical performance, since the surface reaction could stabilize the SEI with relieved volume change. Fig. 4C is a schematic illustration of the K⁺ storage mechanism in the current case. It is known that PTA holds two carboxyl groups per formula unit, which could act as the redox centers to react with two K⁺. During the discharge process, electronic density would be improved around oxygen. This suggests that oxygen with high electronegativity is facile to capture two electrons, which breaks the double bonds of two carboxyl functional groups by bonding two K⁺ for energy storage. During the charging process, electronic density around oxygen atoms would be gradually decreased after PTA loses two electrons, which accounts for the weakened infrared vibration and decreased bonding energies between C−O⁻ and K⁺, thus facilitating two K⁺ extracted from PTA, accompanied by the formation of C=O bonds. Fig. 4D presents Nyquist plots of the PTA electrode at RT and 62.5 °C. It shows that the interphase resistance (consisting of charge transfer and SEI resistances) is ∼20,000 Ω at RT. However, when operated at a high temperature of 62.5 °C, the resistance decreases remarkably to ∼5,000 Ω. This could be attributed to the fact that, at RT, the intrinsic block structure of PTA (Fig. 1 B and C) is poor for the penetration of electrolytes with hindered transfer of electrons/ions. However, at a high temperature of 62.5 °C, the reaction kinetics as well as the transfer of electrons and K⁺ would be fundamentally enhanced, responsible for its much better performance than that at RT. Fig. 4E shows the electrochemical impedance spectroscopies (EISs) of the PTA electrode before and after 100 cycles at 1,000 mA⋅g⁻¹, verifying its smaller semicircle after cycling with reduced charge transfer and solid electrolyte interphase resistances. This suggests that the conductivity of PTA is increased after reacting with K⁺, endowing the good cycling stability of the electrode (36).

Fig. 4.

(A) CV curves of PTA at various scan rates from 0.2 to 1.0 mV⋅s⁻¹. (B) The log(i) vs. log(v) plots of PTA at specific peak currents. (C) Schematic diagram of the electrochemical redox reaction mechanism of PTA. (D) Nyquist plots of PTA at RT and 62.5 °C before cycling. (E) Nyquist plots of PTA before and after 100 cycles at 1,000 mA⋅g⁻¹ under 62.5 °C.

Conclusions

In summary, a potassium storage mechanism is proposed for exploring robust high-temperature PIBs with high cycle stability, which is based on an example of PTA with two carboxyl functional groups as the redox centers. In this case, intercalation/de-intercalation of K⁺ would be performed via surface reactions with relieved volume change. At the fixed working temperature of 62.5 °C, the initial discharge and charge specific capacities of the PTA electrode are ∼660 and 165 mA⋅h⋅g⁻¹, respectively, at a current density of 100 mA⋅g⁻¹, with 86% specific capacity retention after 160 cycles. Meanwhile, it delivers 81.5% specific capacity retention after 390 cycles under a high current density of 500 mA⋅g⁻¹. Such achieved cycle stabilities under both low and high current densities are the best among those of high-temperature PIBs reported previously. The current work might give some meaningful insight on the exploration of advanced PIBs with high stability against high-temperature working environments.

Materials and Methods

Material Synthesis.

To investigate the K⁺ storage mechanism based on carboxyl functional groups with promising practical applications, commercial PTA (C₈H₆O₄; Aladdin) was preferred as the electrode material without further treatment.

Microstructural Characterization.

XRD with Cu–Kα radiation (λ = 0.15406 Å) (X’pertpro; Philips) was used to determine the phase compositions. Field emission SEM (S4800; Hitachi) and HRTEM (JEM-2100F; JEOL) equipped with EDX (Quantax-STEM; Bruker) were used to observe the microstructures. The FTIR spectra were recorded on an ALPHA-T system with KBr pellets from 400 to 4,000 cm⁻¹. The surface element and valence states were characterized by XPS with monochromatic Al Kα (ES-CALAB 250Xi; Thermo Fisher Scientific) as the reference. TGA was performed by a TG/DTA machine (PerkinElmer; Diamond) under Ar atmosphere with a heating rate of 10 K/min.

Electrochemical Measurements.

To prepare the working electrodes, PTA was mixed with carbon blacks and polyvinylidene fluoride (A.R., 99%; Aladdin) with a typical weight ratio of 70:20:10 by grinding assisted with a mortar, followed by dispersal into N-methyl pyrrolidone (NMP) (A.R., 99%; Aladdin) with a typical weight ratio of NMP:PTA that was 10:1. After stirring for ∼20 h, the resultant slurries were casted onto a copper foil, and subjected to drying at 80 °C for ∼12 h. The active materials of the circular electrodes with fixed mass loadings between 1.0 and 1.8 mg⋅cm⁻². The obtained electrodes were assembled into 2,032 button batteries in an argon-filled glove box with water and oxygen contents kept lower than 0.1 ppm. After that, 3 M KFSI in ethylene glycol dimethyl ether (DME) was selected as the electrolyte, and potassium plates cut into pieces acted as the counter electrode, respectively. A commercial glass fiber membrane was used as the separator, which could work well under high temperatures (<500 °C). The CVs were obtained by a Chenhua CHI660E electrochemical workstation at a scan rate of 0.5 mV⋅s⁻¹ under potentials that ranged from 0.15 to 3.0 V. The EIS was measured also by a Chenhua CHI660E electrochemical workstation with frequencies that ranged from 100 kHz to 0.1 Hz. To disclose the high-temperature electrochemical performance, a heating pot was applied to maintain the as-constructed batteries under a 62.5 °C environment during the tests.

Data Availability

All study data are included in the article and/or SI Appendix.