A novel high-efficiency integrated system combining a thermally regenerative electrochemical cycle and a flow battery†

Bo Wang^a, Li Zhao^a, Kun Ge^b, Weicong Xu*^a, Ruihua Chen*^a and Shuai Deng^a
aState Key Laboratory of Engines, Tianjin University, Tianjin 300350, China. E-mail: xuweicong@tju.edu.cn
^bCollege of Power and Energy Engineering, Harbin Engineering University, Harbin 150001, China

First published on 7th August 2024

---

---

1. Introduction

Low-temperature heat energy is abundant and widely distributed in geothermal energy, solar energy and factory waste heat.^1,2 According to statistics, the utilization rate of global primary energy is only 28%, and the vast majority of wasted energy is low-temperature heat energy with temperature lower than 100 °C.³ The low-temperature heat recovery is not only conducive to alleviating the energy crisis but also can save costs and reduce greenhouse gas emissions and has huge application potential.^4,5 However, because of its characteristics of low energy density and small temperature difference with the environment, the recovery of low-temperature heat energy has been a great challenge.

In recent years, liquid-based electrochemical heat to electricity conversion systems such as thermally regenerative batteries (TRBs), thermo-electrochemical cells (TECs) and thermally regenerative electrochemical cycles (TRECs) have caught the attention of scholars around the world.⁶ These systems have the advantages of low material cost, environmental friendliness and ideal thermoelectric conversion performance, which make them effective candidates for the recovery and utilization of medium- and low-temperature heat energy.⁷ TRBs use the complexation of metals and ligands to generate electricity and then regenerate the electrolyte through thermal separation technologies such as distillation, so as to achieve thermoelectric conversion.^8,9 TECs operate based on Seebeck coefficient under non-isothermal conditions, converting heat into electric energy with the electrochemical battery.^10–12 TRECs utilize the temperature dependence of the electrode potential to achieve the conversion of heat to electric energy through the cycle operation of heating, charging, cooling and discharging.¹³ Relative to other systems, the TREC stands out for its excellent heat to electricity conversion efficiency and relative Carnot efficiency and has become a promising heat to electricity conversion system (Fig. S1a and b†).

The heat to electricity conversion performance of a TREC is usually described by temperature coefficient that represents the relationship between the electrode potential and temperature.¹⁴ As shown in Fig. 1a, the electrode potential decreases from E₁(T₁) to E₂(T₂) with the increase of temperature for systems with a negative temperature coefficient. For a half-cell reaction, A + ne⁻ → B, the temperature coefficient α is defined as:

	(1)

where V is the electrode potential, T is the temperature, n is the number of electrons transferred in the reaction, F is the Faraday constant, ΔS is the change in partial molar entropy of the half-cell reaction under isothermal conditions, and S_A and S_B are the partial molar entropy of species A and B, respectively. The temperature coefficient α_cell of the full-cell reaction is the difference between the temperature coefficients of the positive and negative reactions:

αcell = α+ − α−	(2)

where α₊ is the temperature coefficient of the positive electrode and α₋ is the temperature coefficient of the negative electrode. Redox pairs with opposite positive and negative electrode temperature coefficients are usually selected to make the temperature coefficient of the full-cell reaction as large as possible.

Taking a system with negative temperature coefficient as an example, the operation principle of a TREC is shown in Fig. 1b. The cycle consists of four processes: heating up in the open circuit state, isothermal charging, cooling down in the open circuit state and isothermal discharging. In the first step, the system is heated from low temperature T_L to high temperature T_H, the electrode potential decreases, and the entropy increases. In the second step, the system is charged at high temperature. In the third step, the system is cooled from high temperature T_H to low temperature T_L, the electrode potential increases, and the entropy decreases. In the fourth step, the system is discharged at low temperature and returns to the initial state. And when the temperature coefficient is positive, the system consists of four processes: cooling in the open circuit state, isothermal charging, heating in the open circuit state and isothermal discharging. The heat released by the cooling process can preheat the heating process, thus achieving the thermal regeneration process.

The thermally regenerative electrochemical cycle for efficiently harvesting low-temperature heat energy was first reported by Lee et al. in 2014, which used a CuHCF (copper hexacyanoferrate) cathode and a Cu/Cu²⁺ anode to convert low-temperature heat energy into electric energy. When the system operated between 10 °C and 60 °C, the heat to electricity conversion efficiency reached 3.7% without any heat recuperation, and the relative Carnot efficiency reached 24.65%.¹⁴ Then, scholars studied the TREC from aspects such as the ion exchange membrane, charging-free TREC system, electrode and electrolyte. A membrane-free battery was reported in which the heat to electricity conversion efficiency was 1.6% without heat recuperation when the battery was discharged at 15 °C and charged at 55 °C.¹⁵ A charging-free TREC system was proposed that used electrodes consisting of solid Prussian blue particles and achieved a heat to electricity conversion efficiency of 2.0% when the system was operated between 20 °C and 60 °C.¹⁶ Notably, a charging-free continuous TREC system with lithium ferrocyanide and lithium iron phosphate could achieve a thermoelectric efficiency of 19.91% relative to the Carnot efficiency, providing a sustainable power supply throughout the day.¹⁷ The TREC system with the NiHCF cathode and Zn anode achieved a heat to electricity conversion efficiency of 2.41% at the temperature difference of 30 °C.¹⁸ Using the carbon–copper composite negative electrode to increase the electrode reaction contact area, the heat to electricity conversion efficiency increased from 2.97% to 3.86% compared with the system with a copper sheet negative electrode.¹⁹ Besides, the heat to electricity conversion efficiency of the system was 5.17% when manganese hexacyanoferrate (MnHCF) was used as a cathode material.²⁰ The effects of electrolyte on the performance of CuHCF electrodes were investigated through intercalating different monovalent cations (Na⁺, K⁺, Rb⁺, NH₄⁺ and Cs⁺) into electrolyte. Using electrolyte with intercalation of Rb⁺, the system achieved a heat to electricity conversion efficiency of 4.34% when the TREC operates between 10 and 50 °C.^21,22 And using Na/K mixed electrolyte, the heat to electricity conversion efficiency and the relative Carnot efficiency can reach 3.9% and 36.4% when the system operated between 12 °C and 46 °C.²³ The influence of anion properties had also been studied, the system with ClO⁴⁻ electrolyte achieved a heat to electric energy conversion efficiency of 4.1% and a relative Carnot efficiency of 27%, when the cell operated between 10 °C and 60 °C without heat recuperation.²⁴ In addition, it is also possible to combine thermally responsive ionic liquids (TRIL) with the TREC, which endowed the electrolyte with temperature-driven phase change behavior. The heat to electricity conversion efficiency of the special system reached 1.32% operating between 10 °C and 30 °C, and the relative Carnot efficiency was 20.0%.²⁵ To reduce costs, the system using the neutralization flow battery (NFB) was proposed, which could achieve a power of 10–24 μW cm⁻² and an efficiency of 1.4–2.9% at the temperature difference of 25 °C.²⁶ Moreover, the important parameters of existing studies are summarized in Table S1.†

TREC systems achieve high heat to electricity conversion efficiency and relative Carnot efficiency, but they suffer from low power density compared to other liquid-based electrochemical heat to electricity conversion systems such as TECs and TRBs (Fig. S1a†). Actually, paying attention to an individual reactor of the TREC system, the operations are very similar to that of the flow battery. The flow battery energy storage system can directly interconvert chemical energy and electrical energy and achieves an anticipated high power density due to its unique ability to decouple the energy capacity and power density.^27–29 Compared to other flow batteries such as the iron–chromium redox flow battery, zinc–bromine redox flow battery and polysulfide polybromide redox flow battery, the all-vanadium flow battery (VRFB) uses the same substance on both the negative and positive electrodes which allows it to avoid the effects of cross-contamination by redox substances.^29,30 A variety of studies on vanadium electrolytes, such as additives, are also very beneficial for subsequent electrolyte optimization studies. Moreover, due to the benefit of low cost, the VRFB is expected to become a long-term energy storage system for commercial applications.³¹ However, affected by electrode characteristics and ion exchange film materials,^32–34 the energy efficiency of the system is still lower than 87%, with great potential for improvement.³⁵ It can be considered that the TREC is a specific system operation mode that converts heat to power, realized with the help of the flow battery, by charging and discharging at different temperatures. Therefore, the combination of a TREC and flow battery is expected to improve the power density of a TREC and the energy efficiency of a flow battery simultaneously.

Herein, a new energy storage and power generation system integrated with the TREC and flow battery energy is proposed, which has both the low-temperature heat energy recovery power generation of the TREC system and the energy storage function of the flow battery. While simplifying the system structure, the integrated system can compensate for the low power density of the TREC and the low energy efficiency of flow battery energy storage systems. The integrated system, as shown in Fig. 2a, consists of a flow battery unit, two electrolyte reservoirs, two circulating pumps, two heat exchangers, and some connecting fittings. Compared to the flow battery energy storage system, the integrated system only adds two heat exchangers mounted on the delivery pipes. Its operation strategy is shown in Fig. 2b. Taking a system with negative temperature coefficient as an example, since the reaction equilibrium potential is higher at low temperature, the discharging process is conducted at a low temperature, while the charging process is conducted at a high temperature. By passing cold/hot fluids with different temperatures into the heat exchangers, the system is cooled/heated to meet the temperature requirements of different processes and then achieve low-temperature heat recovery and heat to electric energy conversion. In the first step, the system is heated to a high temperature T_H, and the charging process is carried out at the high temperature. Electrochemical reactions take place at the positive and negative electrodes to convert electric energy and heat energy into chemical energy and store it in the electrolyte. In the second step, the system is cooled to a low temperature T_L, and the discharging process is carried out at the low temperature. The electrochemical reactions opposite to the first stage take place at the positive and negative electrodes, converting the chemical energy stored in the electrolyte into electrical energy and output. Through the above process, the integrated system realizes the dual functions of heat to electricity conversion and energy storage while improving the energy efficiency of the flow battery energy storage system.

An all-vanadium flow battery was selected as the flow battery unit in this study, and the integrated system was experimentally studied from the aspects of operating temperature, current density and cycle stability.

2. Experimental

2.1 Experimental system

The schematic diagram of the experimental system is shown in Fig. 3. The whole experimental system is composed of a battery test system, a flow battery unit, a transmission system, and a temperature detection and control system, and specific devices are shown in Fig. S2.†

The 4008T-5V6A-S1 battery test system is provided by NEWARE Technology Limited Co., Ltd., which is utilized to apply and record the voltage and current to the battery unit. The flow battery unit is a laboratory-scale redox flow battery manufactured by Wuhan Zhisheng New Energy Co., Ltd. It consists of metal end plates, a polytetrafluoroethylene insulation sheet, a gold-plated copper collector plate, a serpentine hard graphite plate, a neoprene electrode frame, electrodes, a selective ion exchange membrane and a seal. These components are stacked one by one and are fastened with hexagonal screws to obtain a flow battery unit.

The transmission system includes two liquid storage reservoirs, two peristaltic pumps and some connecting fittings. A DIPump550 intelligent peristaltic pump is provided by Kamoer (Shanghai) Fluid Technology Co., Ltd. Connecting pipes are anti-corrosion hoses with an inner diameter of 3 mm and an outer diameter of 5 mm. The peristaltic pump adjusts the electrolyte flow by changing the speed, and the actual flow is related to the diameter of the pipe, the viscosity of the liquid, and the resistance of the system. The volume of electrolytic liquid pumped is measured by setting different speed values and controlling the working time. The results show that, consistent with the reference data given by the manufacturer, the flow rate of the peristaltic pump in the experimental system is linearly related to its speed; specifically, flow rate (mL min⁻¹) = speed (rpm) × 0.8, with the error less than 2%.

The temperature detection and control system consists of an Agilent 34980A data acquisition instrument, a platinum thermal resistance, two heater bands and a JOANLAB BHS-2 thermostatic water bath procured from Ningbo Qunan Experimental Instrument Co., Ltd. A tee connector made of PTFE material is used at the inlet and outlet of the electrolyte, and the platinum thermal resistance probe is placed in the inlet and outlet electrolyte to measure the temperature. The resistance information collected by the platinum thermal resistance is converted into a 4–20 mA current signal and then input into the data acquisition instrument to obtain the temperature data. Temperature control is achieved by heating the electrolyte in the thermostatic water bath. The anolyte and catholyte storage tanks are placed in the thermostatic water bath and the temperature is set. To compensate for the heat dissipation of the electrolyte flowing through the connecting pipe, the heater bands are installed outside the pipes for heat preservation.

3. Materials and methods

The electrolyte used in this experiment is commercial all-vanadium electrolyte (provided by Wuhan Zhisheng New Energy Co., Ltd.); the concentration of vanadium ions is 1.7 M, and the concentration of SO₄²⁻ is 4.7 M. The anolyte and catholyte are evenly distributed in the two storage tanks, and the total volume of electrolyte is 80 mL. The initial state of vanadium ions in the anolyte and catholyte is V^3.5+ (a mixture of V³⁺ and V⁴⁺). In addition, each liquid storage tank is liquid-sealed with 5 mL liquid paraffin covering the electrolyte. A protective layer composed of inert oils such as paraffin can prevent the air in the catholyte tank from oxidizing V²⁺ to minimize the risk of the charged electrolyte reacting with oxygen in the air and can also protect the electrolyte from evaporation. The graphite felt electrode with an area of 8 × 8 cm² is used in the experiment, the thickness of the electrode before compression is 4.35 mm, and other technical parameters are shown in Table S2.† The ion exchange membrane is a perfluorosulfonic acid proton membrane, and its technical parameters are shown in Table S3.†

After the experimental device is assembled and connected, the speed of the peristaltic pump is set to 30 rpm, and the work step is set in the NEWARE BTSClient80 software. Since the initial state of vanadium ions in electrolyte is V^3.5+ and the positive and negative electrode potential difference is 0 V, the electrolyte has to be activated. When activated, the charging and discharging cutoff voltages are set to 1.65 V and 0.8 V, the charging and discharging current densities are set to 80 mA cm⁻², and the number of charge and discharge cycles are 2. After activation, the anolyte and catholyte are V^2+/5+ after fully charged and V^3+/4+ after fully discharged, and this was followed by other work steps.

Before performing battery performance tests at different temperatures, it is necessary to change the temperature. The reservoirs should be immersed in a thermostat water bath, with both the water bath and the heater band being activated. The desired temperature is set, and the data acquisition instrument is used to collect and observe the temperature data. When the temperature is stable at the preset high temperature, the charging and discharging tests are performed. The charging and discharging cutoff voltages are set to 1.65 V and 0.8 V, and the charging and discharging current densities are set to 40 mA cm⁻². After 2–3 cycles, information on the capacitance, voltage, current and energy during the charging and discharging processes is obtained. Subsequently, the temperature or current density is adjusted, and the procedure is reiterated to ascertain performance variability.

4. Results and discussion

4.1 Measurement of temperature coefficient

Temperature coefficient is a pivotal parameter affecting the performance of the integrated system. In order to ascertain the temperature coefficient of commercial all-vanadium electrolyte, a series of pulse charging experiments were conducted at different temperatures. The charging and discharging current densities were set to 40 mA cm⁻², the flow rate was set to 30 rpm (24 mL min⁻¹), and the charging and discharging cutoff voltages were 1.65 V and 0.8 V, respectively. The experiments were systematically conducted at distinct temperatures: 10 °C, 20 °C, 30 °C, and 40 °C. Following each charging at 0.1 A h, the capacitors were allowed to stand for 3 minutes to facilitate the observation and documentation of the charging voltage. The state of charge (SOC) was determined by the charged capacity and the total capacity. This procedure enabled the derivation of open circuit voltages at equivalent SOCs across the tested temperatures, which are essential for the computation of the temperature coefficient. The experimental outcomes are graphically depicted in Fig. 4. Fig. 4a shows the change in the charging voltage at the aforementioned temperatures, revealing a regular decrease in the charging voltage with increasing temperature. The open circuit voltages of the battery were measured at various temperatures (10 °C, 20 °C, 30 °C and 40 °C) for different states of charge (SOC = 0.3, 0.5, 0.7), and the resultant data were subjected to linear regression analysis. The temperature coefficients for different states of charge (SOC = 0.3, 0.5, 0.7) were −1.38 mV K⁻¹, −1.46 mV K⁻¹ and −1.43 mV K⁻¹, respectively, as shown in Fig. 4b. According to the Nernst equation, the temperature coefficient has concentration dependence, and the difference in temperature coefficient at different SOCs is reasonable.³⁶ In order to verify the accuracy of the temperature coefficient, an experiment was conducted on the temperature coefficient of the half battery, referencing the experimental method of Rajan et al.³⁷ The detailed experimental method and operations are shown in ESI Note 5.† The findings conclusively validate the efficacy of the full temperature coefficient.

4.2 Effect of temperature on the system performance

Operating temperature is a critical parameter for integrated systems, so it is necessary to investigate the effect of temperature on the system performance. Given the objective of the integrated system's operation is to harness and utilize medium- and low-grade thermal energy, it is imperative to prioritize the high-temperature charging process. Therefore, a comprehensive battery performance test was conducted at different operating temperatures (10 °C, 20 °C, 30 °C and 40 °C) to obtain the charging and discharging voltage curves and associated efficiency parameters of the system. Efficiency is an important criterion for evaluating the VRFB performance and can be expressed in a variety of ways. When evaluating the performance of the flow battery, three metrics are commonly used: coulombic efficiency (CE), voltage efficiency (VE) and energy efficiency (EE).

Fig. 5a shows the charging curve at high temperature and discharging curve at low temperature when the system is operated between 10 °C and 40 °C. The horizontal axis is the battery capacity and the vertical axis is the battery voltage. The charging capacity and discharging capacity of the system are 1.6038 A h and 1.550 A h, and the charging energy and discharging energy are 2.3170 W h and 2.0650 W h, respectively. The integrated system achieves the coulombic efficiency, voltage efficiency, and energy efficiency of 96.65%, 92.22%, and 89.12%, respectively. Additionally, the power density and the energy density reach 523.96 W m⁻² and 25.81 W h L⁻¹.

In addition to CE, VE and EE, it is also necessary to pay attention to the conversion efficiency of heat energy to electric energy. The calculation formula η_TREC for the efficiency of the TREC system is given by:¹⁴

	(3)

where ΔTΔS is the maximum output work of the system and E_loss is the energy lost by the system due to factors such as battery resistance. The energy input to complete the cycle consists of two parts: the heat absorbed at the heat source (Q_H = T_HΔS) and the external heat required to heat up the system (Q_HR = C_pΔT). However, since the energy lost by the all-vanadium flow battery itself may be greater than the energy gained by the TREC, the η_TREC is negative possibly. Therefore, the normalized thermal efficiency of the integrated system is given by:³⁸

	(4)

where W_dis (J) is the discharging energy, W_cha (J) is the charging energy, U_D,TREC (J) is the discharging energy of the TREC system, U_C,TREC (J) is the charging energy of the TREC system, U_D,TRef (J) is the discharging energy at the reference temperature, U_C,TRef (J) is the charging energy at the reference temperature, and T_Ref (K) is the reference temperature. T_H (K) is the operating high temperature. α_cell (mV K⁻¹) is the full temperature coefficient of the cell obtained by experiment. q_c (C) is the charge capacity of the cell. C_p (J K⁻¹) is the heat capacity of the electrolyte that is obtained by experiment; the specific experimental system is shown in Fig. S4,† and the results are shown in Table S4.† ΔT (K) is the temperature difference during charging and discharging. η_sys considers the energy gained from the thermodynamic cycle relative to the isothermal cycle at the reference temperature. By calculation, the normalized thermal efficiency of the integrated system is 2.54%.

Fig. 5b presents the charging voltage curve at different temperatures (10 °C, 20 °C, 30 °C and 40 °C) and the discharging voltage curve at 10 °C. An increase of the operating high temperature leads to a reduction of the electrode potential, consequently lowering the charging voltage at a certain SOC. Specifically, elevating the temperature of the charging process diminishes the electrical energy required when the charging capacity is fixed. The system is charged at low potential and discharged at high potential, which not only enhances the energy efficiency of the system but also achieves the conversion of heat into electricity.

Furthermore, the performance of the integrated system across various operating temperatures has been scrutinized. Keeping the operating low temperature at 10 °C, the operating high temperature is changed from 20 °C to 40 °C. The findings of these experiments are depicted in Fig. 5c. As the experimental operating temperature increases from 20 °C to 40 °C, there is an increase in the energy efficiency of the integrated system. This is attributed to the reduction in charging voltage and the consequent decrease in the energy input required for charging. However, the normalized efficiency of the integrated system exhibits a contrary trend, declining with increasing temperature. According to eqn (4), the input heat energy of the system comprises the heat energy absorbed by the reaction and the heat energy required to heat up the system. When the temperature increases, the increase of the heat energy required to heat up the system is greater than the decrease of the charging energy, so the normalization efficiency of the system shows a downward trend.

For the all-vanadium flow battery system, an increment in the operating temperature results in a decrease in coulombic efficiency, an increase in voltage efficiency, and a moderate decrease in energy efficiency, as illustrated in Fig. 5d. The increase of temperature augments the diffusive activity of vanadium ions, thereby facilitating the disproportionation reaction of anolyte and catholyte ions through the ion exchange membrane. This reduces the concentration of active ions in the anolyte and catholyte, culminating in a decrement in the coulombic efficiency of the battery.³⁹ Since the charge transfer rate and mass transfer rate are positively correlated with the temperature of the electrolyte, increasing the temperature of the electrolyte will enhance the reaction kinetics and reduce the viscosity of the electrolyte, thus increasing the voltage efficiency of the battery.⁴⁰ The energy efficiency influenced by both the coulombic efficiency and voltage efficiency exhibits a gradual downward trend.

Fig. 5d also shows CE, VE and EE of the integrated system operating between 10 °C and 40 °C. Comparing the integrated system with the VRFB system, it is found that the coulombic efficiency of the integrated system is close to that of the VRFB at 10 °C and exceeds that of the VRFB at 40 °C. Furthermore, while the VRFB attains the highest voltage efficiency at 40 °C, the integrated system exhibits a higher voltage efficiency than that of VRFB at 10 °C. The integrated system also claims the highest energy efficiency, with improvements of 3.5% and 8.2% over the VRFB systems operating at 10 °C and 40 °C, respectively. This shows that the integrated system combined with the TREC delivers superior performance.

4.3 Effect of current density on the system performance

To elucidate the impact of current density on the system performance, a series of charging and discharging tests were conducted at varying current densities, specifically 20 mA cm⁻², 40 mA cm⁻², and 60 mA cm⁻². Experiments record the charging and discharging voltages and other information and then the changes in the capacity and the system efficiencies.

As shown in Fig. 6a, with the increase of current density, the high-temperature charging voltage curve of the battery increases, while the low-temperature discharging voltage curve shows a declining trend. It is noteworthy that the low-temperature discharging process is more susceptible to the influence of current density compared to the high-temperature charging process. The charging and discharging processes take place at different temperatures and are therefore affected by both temperature and current density. An elevation in temperature will diminish the overpotential of the battery. An increase in current density will accelerate the electrochemical reaction process and also lead to increased polarization inside the electrode.⁴¹ Under the combined action of temperature and current density, the influence of current density on the high-temperature charging process is weakened, while the influence on the low temperature discharge process is comparatively pronounced. Fig. 6b illustrates the influence of current density variation on battery charging and discharging capacity. An increase in current density leads to a decrement in both charging and discharging capacities, with the charging capacity experiencing a more pronounced reduction. The high charging current density reduces the charging time, as shown in Fig. 6c, leaving insufficient time for the vanadium ions to achieve complete diffusion. At the same time, the high current density accelerates electron transfer, creating a scenario where the V³⁺ are inadequate to react with the surplus electrons.⁴² When the battery voltage reaches the predetermined cutoff voltage, a substantial amount of V³⁺ at the negative electrode and V⁴⁺ at the positive electrode remains unconverted to V²⁺ and V⁵⁺, respectively, resulting in reduced charging and discharging capacities.

As indicated in Fig. 6d and e, for VRFB systems at 10 °C and 40 °C, the coulombic efficiency increases with the increase of current density, while the voltage efficiency decreases. The high current density improves the electrode activity and inhibit the crossover of vanadium ions, so the coulombic efficiency gradually increases with the increase of current density. In the charging and discharging process of battery, the battery voltage is significantly affected by polarization. Increased polarization leads to a pronounced deviation of the voltage from its equilibrium potential and the voltage loss, resulting in the reduction of voltage efficiency. With the increase of current density, both ohmic and concentration polarization intensify, further contributing to the decline in the voltage efficiency.

For VRFB systems at 10 °C and 40 °C, the relationship between the energy efficiency and current density shows an opposite tendency. At low temperature, the energy efficiency decreases with the increase of current density. In contrast, the energy efficiency increases with the increase of current density at high temperature. The energy efficiency is affected by both the coulombic efficiency and voltage efficiency, necessitating a vigilant examination of their respective changes. At low temperature, when the current density increases from 20 mA cm⁻² to 60 mA cm⁻², the coulombic efficiency increases by 2.63% and the voltage efficiency decreases by 9.26%. At high temperature, the coulombic efficiency increases by 11.42% and the voltage efficiency decreases by 4.88%. The disparate relationships between the energy efficiency and current density at low and high temperatures can be ascribed to different impacts of current density on the coulombic efficiency and voltage efficiency. At the lower temperature, the current density exerts a more pronounced adverse effect on the voltage efficiency, resulting in a declining trend in energy efficiency. In contrast, at the higher temperature, the current density confers a more significant enhancement to coulombic efficiency, leading to an ascending trend in energy efficiency.

4.4 Cycling stability study

Stable operation is an important prerequisite for the practical application of the integrated system. Hence, the long-term stability of the system has been studied comprehensively. Due to the phenomenon of incomplete charge and discharge, transmembrane transport of vanadium ions and oxidation in contact with air, the battery capacity will inevitably decay. Initially, capacity retention ability has been investigated as shown in Fig. 7a and b. It is found that the charging and discharging voltage curves remain remarkably stable for 50 cycles. The experimental results show that the battery charging capacity still maintains 92% of its initial value after 50 cycles, which proves that the system can run stably for a long time. A slight increment in charging capacity during the initial cycles may be attributed to the electrolyte's progressive activation. The decrease of battery capacity is influenced by many factors. The main reason is the different diffusion of vanadium ions across the membrane, which leads to the accumulation of vanadium ions on one side and the dilution of vanadium ions on the other side.⁴³ The concentration imbalance caused by vanadium ion crossover creates an osmotic pressure gradient between the two sides, and then water molecule migration occurs.⁴⁴ These factors further cause volume imbalance, valence imbalance, and concentration imbalance of active substances in positive and negative electrolytes.⁴⁵ These imbalances ultimately lead to a gradual decline in the battery capacity.

Moreover, charging and discharging tests are carried out at low temperature and high temperature, respectively, in order to obtain the changes in its coulombic efficiency, voltage efficiency and energy efficiency for 10 cycles. The coulombic efficiency, voltage efficiency and energy efficiency stabilized at 96.6%, 89.1% and 86.1% at 10 °C, respectively. At 40 °C, they stabilized at 87.2%, 94.4% and 82.3%, respectively, as shown in Fig. 7c. The above results indicate that the system can operate stably at both low and high temperatures.

In addition, it is imperative to assess the thermal stability of vanadium ions at different temperatures. The thermal stabilities of V⁵⁺ at high temperature and V²⁺, V³⁺, and V⁴⁺ at low temperature were investigated, respectively. The 20 mL fully charged catholyte was heated to high temperatures (40 °C, 50 °C, and 60 °C) and then maintained at high temperatures for 2 hours. The precipitation of V⁵⁺ was observed and recorded at 1 h interval, as shown in Fig. S5.† At 40 °C, the electrolyte remains clear, and V⁵⁺ did not precipitate after 2 h. At 50 °C, no precipitation was observed after 1 h, and precipitation was observed after 2 h. At 60 °C, precipitation was observed after 1 h and increased significantly after 2 h. Therefore, it is concluded that for the commercial vanadium electrolyte with 1.7 M vanadium ions and 4.7 M SO₄²⁻, the maximum operating temperature should not exceed 40 °C. It is advisable to reach 40 °C for a long running time and 50 °C for a short running time. Similarly, the fully charged anolyte and fully discharged anolyte and catholyte are cooled to 0 °C and left for two hours, and the change is observed every hour. The results are shown in Fig. S6.† It is noteworthy that after the two-hour interval, no precipitation was observed, thereby confirming the electrolyte's stability at low temperatures. Therefore, the applicable temperature range for the system could be considered 0–45 °C.

5. Conclusions

To address the limitations of low power density of TRECs and the low energy efficiency of flow battery energy storage systems, this study proposed a novel integrated system combining the heat to electricity conversion of the TREC and energy storage of the flow battery. Further experimental research studies are conducted selecting an all-vanadium flow battery as the flow battery unit. The following conclusions were reached:

(1) The experimental results show that the temperature coefficient of commercial electrolyte (the concentration of vanadium ions is 1.7 M and the concentration of SO₄²⁻ is 4.7 M) is −1.46 mV K⁻¹ when the SOC is 0.5. When the system operates in the temperature range of 10 °C to 40 °C, the coulombic efficiency, voltage efficiency and energy efficiency are 96.65%, 92.22% and 89.12%, respectively. The power density and the energy density reach 523.96 W m⁻² and 25.81 W h L⁻¹, and the normalized thermal efficiency is calculated to be 2.54%.

(2) The study delves into the effects of temperature variations on the performance of the integrated system and the VRFB system. It is observed that the energy efficiency of the integrated system exhibits an increasing trend with increasing operating temperatures, while the normalized thermal efficiency exhibits a decline. For the VRFB system, as the temperature increases from 10 °C to 40 °C, the coulombic efficiency of the battery gradually decreases, the voltage efficiency increases, and the energy efficiency decreases. Operating between 10 °C and 40 °C, the energy efficiency of the integrated system is 3.5% and 8.2% higher than that of the VRFB system operating at 10 °C and 40 °C, respectively.

(3) The study also investigates the influence of current density on the performance of both the integrated system and the VRFB system. The charging and discharging processes are affected by the current density differently, and the low temperature discharging process is more sensitive to the increase of current density. For the VRFB system, with the increase of current density, the coulombic efficiency increases at both high and low temperatures, while the voltage efficiency decreases. Notably, the energy efficiency of the VRFB system exhibits contrasting trends at different temperatures, increasing at high temperatures and decreasing at low temperatures. This divergence can be attributed to the varying impacts that current density exerts on the coulombic efficiency and voltage efficiency of the system at various temperatures.

Data availability

All relevant data are within the manuscript and its additional files.

Author contributions

Bo Wang: conceptualization, writing – original draft, writing – review & editing. Li Zhao: methodology, funding acquisition, project administration. Kun Ge: experimental assistance. Weicong Xu: writing – review & editing, funding acquisition, project administration. Ruihua Chen: writing – review & editing. Shuai Deng: writing – review & editing.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the China Postdoctoral Science Foundation (2021TQ0237 and 2022M722354), National Natural Science Foundation of China (52176017), and the Postdoctoral Fellowship Program of CPSF under Grant Number GZB20240520.

References

1. S. Chu and A. Majumdar, Nature, 2012, 488, 294–303 CrossRef CAS PubMed .
2. Z. Yuan, L. Wei, J. D. Afroze, K. Goh, Y. Chen, Y. Yu, Q. She and Y. Chen, J. Membr. Sci., 2019, 579, 90–101 CrossRef CAS .
3. C. Forman, I. K. Muritala, R. Pardemann and B. Meyer, Renew. Sustain. Energy Rev., 2016, 57, 1568–1579 CrossRef .
4. A. P. Straub, N. Y. Yip, S. Lin, J. Lee and M. Elimelech, Nat. Energy, 2016, 1, 16090 CrossRef CAS .
5. S. I. Garcia, R. F. Garcia, J. C. Carril and D. I. Garcia, Renew. Sustain. Energy Rev., 2018, 81, 760–767 CrossRef .
6. M. Rahimi, A. P. Straub, F. Zhang, X. Zhu, M. Elimelech, C. A. Gorski and B. E. Logan, Energy Environ. Sci., 2018, 11, 276–285 RSC .
7. D. Huo, H. Tian, G. Shu and W. Wang, Sustain. Energy Technol. Assessments, 2022, 49, 101802 CrossRef .
8. W. Wang, G. Shu, H. Tian, D. Huo and X. Zhu, J. Power Sources, 2019, 424, 184–192 CrossRef CAS .
9. W. Wang, G. Shu, X. Zhu and H. Tian, J. Mater. Chem. A, 2020, 8, 12351–12360 RSC .
10. H. Gerischer, C. W. Tobias, R. C. Alkire and D. M. Kolb, Advances in Electrochemical Science and Engineering, VCH Publishers Inc, New York, 2008 Search PubMed .
11. T. Migita, N. Tachikawa, Y. Katayama and T. Miura, Electrochemistry, 2009, 77, 639–641 CrossRef CAS .
12. J. Duan, G. Feng, B. Yu, J. Li, M. Chen, P. Yang, J. Feng, K. Liu and J. Zhou, Nat. Commun., 2018, 9, 5146 CrossRef PubMed .
13. C. Gao, S. W. Lee and Y. Yang, ACS Energy Lett., 2017, 2, 2326–2334 CrossRef CAS .
14. S. W. Lee, Y. Yang, H. Lee, H. Ghasemi, D. Kraemer, G. Chen and Y. Cui, Nat. Commun., 2014, 5, 3942 CrossRef CAS PubMed .
15. Y. Yang, J. Loomis, H. Ghasemi, S. W. Lee, Y. J. Wang, Y. Cui and G. Chen, Nano Lett., 2014, 14, 6578–6583 CrossRef CAS PubMed .
16. Y. Yang, S. W. Lee, H. Ghasemi, J. Loomisa, X. Li, D. Kraemer, G. Zheng, Y. Cui, and G. Chen, Proceedings of the National Academy of Sciences, 2014, vol. 111, pp. 17011–17016 Search PubMed .
17. H. Zhang, Z. Wang, H. Li, M. Salla, Y. Song, S. Huang, S. Huang, X. Wang, K. Liu, G. Xu, J. Huang, C. Qiu and Q. Wang, Joule, 2024, 7, 1515–1528 CrossRef .
18. C. Cheng, S. Wang, P. Tan, Y. Dai, J. Yu, R. Cheng, S. Feng and M. Ni, ACS Energy Lett., 2021, 6, 329–336 CrossRef CAS .
19. D. Huo, H. Tian, W. Wang and G. Shu, Sustain. Energy Technol. Assessments, 2022, 53, 102793 CrossRef .
20. S. Jung, Y. J. Lim, J. Kwon, K. Wei, J. Lee, K. Kim, S. G. Kang and Y. Kim, J. Power Sources, 2022, 524, 231080 CrossRef CAS .
21. C. Gao, Y. Liu, B. Chen, J. Yun, E. Feng, Y. Kim, M. Kim, A. Choi, H. W. Lee and S. W. Lee, Adv. Mater., 2021, 33, 2004717 CrossRef CAS PubMed .
22. D. Huo, H. Tian, W. Wang and G. Shu, Energy Convers. Manage., 2022, 255, 115306 CrossRef CAS .
23. D. Huo, H. Tian, W. Wang and G. Shu, Mater. Today Nano, 2022, 18, 100206 CrossRef CAS .
24. X. Li, A. Wu, J. Li, Z. Li, D. Lee and S. W. Lee, ACS Energy Lett., 2023, 8, 4061–4068 CrossRef CAS .
25. A. Wu, X. Li, D. Lee, J. Li, J. Yun, C. Jiang, Z. Li and S. W. Lee, Nano Energy, 2023, 105, 108022 CrossRef CAS .
26. P. Loktionov, D. Konev, R. Pichugov and A. Antipo, Energy Convers. Manage., 2024, 299, 117830 CrossRef CAS .
27. O. C. Esan, X. Shi, Z. Pan, X. Huo, L. An and T. S. Zhao, Adv. Energy Mater., 2020, 10, 2000758 CrossRef CAS .
28. M. Skyllas-Kazacos, M. H. Chakrabarti, S. A. Hajimolana, F. S. Mjalli and M. Saleem, J. Electrochem. Soc., 2011, 158, 55–79 CrossRef .
29. Z. Huang and A. Mu, Int. J. Energy Res., 2021, 45, 14170–14193 CrossRef .
30. Y. Guo, J. Huang and J. Feng, J. Ind. Eng. Chem., 2023, 118, 33–43 CrossRef CAS .
31. A. Lucas and S. Chondrogiannis, Electr. Power Energy Syst., 2016, 80, 26–36 CrossRef .
32. J. Sun, L. Zenga, H. R. Jiang, C. Y. H. Chao and T. S. Zhao, J. Power Sources, 2018, 405, 106–113 CrossRef CAS .
33. M. H. Hossain, N. Abdullah, K. H. n. Tan, R. Saidur, M. A. M. Radzi and S. Shafie, Chemical Record, 2024, 24, 108321 CrossRef PubMed .
34. N. Zhao, A. Platt, H. Riley, R. Qiao, R. Neagu and Z. Shi, J. Energy Storage, 2023, 72, 108321 CrossRef .
35. Z. Yu, X. Jia, Y. Cai, R. Su, Q. Zhu, T. Zhao and H. Jiang, Energy Storage Mater., 2024, 69, 103404 CrossRef .
36. J. Bleeker, S. Reichert, J. Veerman and D. A. Vermaas, Sci. Rep., 2022, 12, 7993 CrossRef CAS PubMed .
37. A. Rajan, I. S. McKay and S. K. Yee, Nat. Energy, 2022, 7, 320–328 CrossRef CAS .
38. D. Reynard, C. R. Dennison, A. Battistel and H. H. Girault, J. Power Sources, 2018, 390, 30–37 CrossRef CAS .
39. T. Zou, X. Shi and L. Yu, J. Power Sources, 2021, 490, 229514 CrossRef CAS .
40. J. Ren, Y. Li, Z. Wang, J. Sun, Q. Yu, X. Fan and T. Zhao, Int. J. Heat Mass Transfer, 2023, 203, 123818 CrossRef CAS .
41. Z. Huang, Y. Liu, X. Xie, Q. Huang and C. Huang, J. Electrochem. Soc., 2023, 170, 100502 CrossRef .
42. M. Zarei-Jelyani, M. M. Loghavi, M. Babaiee and R. Eqra, Electrochim. Acta, 2023, 443, 141922 CrossRef CAS .
43. N. Wang, W. Zhou and F. Zhang, Ionics, 2019, 25, 221–229 CrossRef CAS .
44. D. K. Kim, E. J. Choi, H. H. Song and M. S. Kim, J. Membr. Sci., 2016, 497, 194–208 CrossRef CAS .
45. X. Yan, J. Sun, L. Gao, W. Zheng, Y. Dai, X. Ruan and G. He, Int. J. Hydrogen Energy, 2018, 43, 301–310 CrossRef CAS .

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4ta03715d

This journal is © The Royal Society of Chemistry 2024