Preparation and characterization of SPES/PVA (double-layer) membrane for vanadium redox flow battery

Abstract

The double-layer membrane consisting of sulfonated poly(ether sulfone) (SPES) sub-layer and polyvinyl alcohol (PVA) sub-layer (denoted as SPES/PVA membrane) was prepared and employed as the separator for vanadium redox flow battery (VRB) system to evaluate the vanadium ions permeability and cell performance. The SPES/PVA membrane is a double-layer structure and exhibits dramatically lower vanadium ions permeability and better cell performance compared to the pristine SPES membrane, PVA membrane, and Nafion117 membrane. The vanadium ion permeability of SPES/PVA membrane is one order of magnitude lower than that of Nafion117 membrane. In further work, the single cell with SPES/PVA membrane showed significantly lower capacity loss, higher coulombic efficiency (>92.5%), and higher energy efficiency (>83.9%) than Nafion117 membrane. In the self-discharge test, SPES/PVA membrane showed 1.8 times longer duration in the open circuit decay than Nafion117 membrane. With all the good properties and low cost, this new kind of double-layer membrane is suggested to have excellent commercial prospects as an ion exchange membrane for VRB systems.

Keywords

Sulfonated poly(ether sulfone)polyvinyl alcohol vanadium redox flow battery double-layer ion exchange membrane

Introduction

New eco-friendly and renewable energy sources, such as solar, tidal, and wind, are among the central topics of our times with the concerns of energy shortage and environmental protection. However, these renewable energy sources are commonly unstable, and how to store them safely and efficiently has become a huge challenge for their large-scale application. Vanadium redox flow battery (VRB) system, firstly proposed by Skyllas-Kazacos’ team¹ in 1985, has received considerable attention in the last few years due to its technical benefits: faster response time, longer cycle life, flexible design, deeper discharge capability, and lower pollution emitting in energy storage. As shown in Figure 1,² the VRB single cell consists of two electrolyte tanks with electrolytes of V(II)/V(III) and V(IV)/V(V) in sulfuric acid (H₂SO₄) solution, two pumps, and a battery stack section where the redox electrode reaction takes place. As the pivotal component of VRB, ion exchange membranes are used to separate the two kinds of electrolytes and complete the current circuit by transferring ions.

Figure 1.

All-VRB. VRB: vanadium redox flow battery.

The perfect ion exchange membrane should possess low vanadium ions permeation, good proton conductivity, and high chemical and mechanical stability. At present, perfluorosulfonic polymers, such as DuPont Nafion, are the most commonly used ion exchange membrane materials. Although they show both better proton conductivity and chemical stability, the extremely high cost and high crossover of these membranes have limited their commercialization.³ Consequently, new alternative ion exchange membrane materials are being sought.⁴

The multilayer membranes have been synthesized and studied for many years. Compared to blend membranes, the multilayer membranes contain more than one block phase, which present self-governed properties. Hence, the multilayer membranes can involve all the characteristics of each sub-layer. In this article, one new kind of double-layer membrane consisting of sulfonated poly(ether sulfone) (SPES) sub-layer and polyvinyl alcohol (PVA) sub-layer was developed by a two-step casting method. The sub-layer near anode rich in PVA is expected to provide blockage of V ion permeation, and the SPES sub-layer near cathode maintains the high mechanical stability and low swelling ratio, which may be beneficial in providing the tight contact between electrolyte membrane and catalyst layer.

Properties of the double-layer membrane as an ion exchange membrane and its preliminary application in VRB were investigated in detail.

Experimental

Preparation of SPES membrane

For preparing the SPES, 10 g of Poly ether sulfones (PES) powder was slowly added to H₂SO₄ (98 wt%, 100 ml) and 20 ml of CSA was added drop by drop at room temperature for 5 h. Then, the reaction was terminated by pouring into ice-cold water under mechanical agitation. The fibers were then washed with deionized water until the neutral pH was achieved and stirred overnight to remove the residual H₂SO₄. The polymer fibers were finally dried at 60°C for 12 h and 110°C for 3 h in a vacuum box.

Dry SPES was weighed and dissolved in N,N-dimethylacetamide (DMAC) to form a 10 wt% solution, which was cast onto a glass plate by using a casting knife and heated to evaporate most of the solvent. The cast membranes were dried at 80°C for 10 h and then at 100°C for 10 h. After cooling to room temperature, the resulting cast membranes were peeled off from the glass plate. The membranes were kept in deionized water after washing several times in deionized water.

Preparation of PVA membrane

Similar to the manufacturing procedure of SPES membrane, dry PVA was dissolved in DMAC to form a 10 wt% solution. Suitable amount of phosphoric acid was added and stirred to obtain a homogeneous solution, which was then cast onto a casting plate. The solution was dried under vacuum at 55°C for 12 h, and then the temperature was increased to 115°C for 3 h. The formed membrane was designated as PVA membrane.

Preparation of double-layer membrane

The SPES sub-layer was firstly prepared as described in the “Preparation of SPES membrane” section. The mixed solution of PVA and phosphoric acid in DMAC was then cast onto the preformed SPES sub-layer. The solution and SPES sub-layer were subsequently both dried under vacuum at 60°C for 12 h and at 120°C for 3 h. The prepared double-layer membranes were finally peeled off from the plates by immersing into deionized water.

Membrane characterization

Thickness of membrane

The thickness is an important parameter for the membrane. In this experiment, the thickness of membranes was measured using a micrometer. The dry sample membrane with dimensions of 2 cm × 2 cm was cut and used for measurement. We measured the thickness of four corners and the center of the membrane and then averaged the results.

Water uptake

The water uptake (W_u) is one important property of ion exchange membrane. In this work, W_u was calculated according to the following equation:⁵

W_{u} = \frac{W_{S} - W_{dry}}{W_{dry}} \times 100 %

where W_s and W_dry are the weight of the saturated and dry membranes, respectively.

Ion exchange capacity

The ion exchange capacity (IEC) of the membranes is calculated by the conventional titration method.⁶ The acidic form of the membrane was converted to the sodium form by immersing in 1 M NaCl solution for 24 h so that the proton ions could be exchanged for sodium ions. The exchanged proton ions in the solution were titrated with 0.05 M NaOH solution:

IEC = \frac{V_{NaOH} C_{NaOH}}{W_{d}}

where V_NaOH is the volume of the consumed NaOH solution, C_NaOH is the concentration of the NaOH solution, and W_d refers to the dried weight of membrane.

The degree of sulfonation (DS) can be calculated by the following equation described by literature⁶

IEC = \frac{1000 DS}{28 + 102 DS}

Proton conductivity

The proton conductivity was measured by electrochemical impedance spectroscopy using a CHI760C electrochemical workstation with an AC perturbation of 5.0 mV over frequency range from 1.0 Hz to 100 kHz. The proton conductivity (σ) was calculated by the following equation⁷

σ = L / A R

where L is the distance between the two electrodes, A is the actual contact area of sample membrane, and R is the resistance of membrane.

Vanadium permeability

The V(IV) permeability measurement was conducted in a membrane separated diffusion cell.⁸ Measurement device is shown in Figure 2: The right reservoir was filled with 1.0 M VOSO₄ in 2 M H₂SO₄ solution and the left one was filled with 1.0 M MgSO₄ in 2 M H₂SO_4. Here, MgSO₄ was used to equalize the ionic strengths and to minimize the osmotic pressure effects. Magnetic stirrers were also used in both reservoirs to avoid the concentration polarization. The effective membrane area exposed to the solution was 3.14 cm² and the solution volume of each half-cell was 40 ml. Samples from the left reservoir were collected at regular time interval.

Figure 2.

Device diagram for measurement of vanadium permeability.

The concentration of V(IV) in the sample solution was measured by UV spectrometer (λ_max = 765 nm).⁹ Given that the change of V(IV) concentration in the right reservoir is so low that the relationship between V(IV) concentration in the left reservoir and time could be described by Fick’s diffusion law as:

V_{L} \frac{d C_{L}^{(t)}}{d t} = \frac{A \times P}{L} (C_{R} - C_{L} (t))

where V_L is the volume (cm³) of the solution in the left reservoir, A is the effective area (cm²), L is the thickness (μm) of the membrane, P is the permeability of V(IV), C_R is the V(IV) concentration in the right reservoir, and C_L is the V(IV) concentration (M) in the left reservoir according to time t (min). The ion selectivity of the membrane is determined by the ratio of conductivity and permeability.¹⁰

VRB single cell performance

The single cell VRB used for charge–discharge test was assembled by sandwiching a membrane with two carbon felt electrodes, clamped by two graphite polar plates. All these components were fixed between two stainless plates; 2.0 M V(II)/V(III) in 3.0 M H₂SO₄ solution and 2.0 M V(IV)/V(V) in 3.0 M H₂SO₄ solution, serving as negative and positive electrolytes, respectively, were cyclically pumped into the corresponding half-cell. The volume of electrolyte solution was 75 ml in each half-cell and the actual area of the electrode was 30 cm². The VRB single cells were charged and discharged at current densities of 36 mA·cm⁻². The cell was charged and discharged by a battery test system. To avoid the corrosion of the graphite plates, the upper limit of charge voltage was set as 1.5 V and the lower limit of discharge voltage was set as 0.8 V.

Results and discussion

Primary properties of the membranes

Thickness, IEC, DS, tensile strength, and water uptake of the SPES membrane, PVA membrane, SPES/PVA membrane, and Nafion117 membrane are summarized in Table 1. It can be seen from the table that the IEC of the double-layer membranes decreases due to the decreasing numbers of –SO₃H groups per gram in the double-layer membrane and water uptake of the double-layer membrane increases upon the PVA sub-layer. Considering the high crystallinity of PVA, it is reasonable that the addition of PVA would extend the crystalline region of the polymer and suppress its water adsorption. Therefore, the blend membrane is less likely to swell in water compared to pristine SPES and Nafion117 membranes.

Table 1.

Primary properties of the membranes.

Samples	Thickness (μm)	ICE (mmol g⁻¹)	DS (%)	Tensile strength (MPa)	Water uptake (%)
Nafion117	110	0.96	—	15	19
SPES	160	1.72	60	11	50
PVA	160	0.05	—	31	76
SPES/PVA	170	1.67	58	27	58

IEC: ion exchange capacity; DS: degree of sulfonation; SPES: sulfonated poly(ether sulfone); PVA: polyvinyl alcohol.

It is well-known that the swelling behavior of membranes is closely related to the mechanical stability. Compared to the Nafion117 and pure SPES membranes, the double-layer membrane exhibits much higher tensile strength, which indicates that the PVA sub-layer can really reinforce SPES membrane with better mechanical stability. It is considered that the adsorption and diffusion of water molecules in SPES mainly take place in the hydrophilic region, induced by sulfonic acid groups, while the crystalline region formed from hydrophobic main chain provides the membrane mechanical stability. The addition of PVA might improve the crystallinity of the polymer blends and further improve their mechanical stability.

Vanadium permeability

Vanadium ion transfer through the ion exchange membrane is disadvantageous to the VRB because it will lead to self-discharge of the battery and lower coulombic efficiency (CE). The vanadium ion permeability (P) of membranes is listed in Table 2. An approximately linear relation between time and concentration of V(IV) in the right cell is clearly demonstrated in Figure 3. It is found that the pure SPES membrane has higher P due to the well-known fact that permeability rises with DS: The increment and enrichment of –SO₃H groups in the membrane can provide more and larger continuous transport channels for cations.

Table 2.

Conductivity, vanadium permeability, and selectivity of the membranes.

Samples	Conductivity (S cm⁻¹)	Permeability of V(IV) (cm² min⁻¹)	Selectivity (S min cm⁻³)
Nafion117	0.100	1.30 × 10⁻⁶	7.7 × 10⁴
SPES	0.054	1.75 × 10⁻⁶	3.1 × 10⁴
PVA	0.12	9.14 × 10⁻⁶	1.3 × 10⁴
SPES/PVA	0.076	2.80 × 10⁻⁷	2.5 × 10⁵

SPES: sulfonated poly(ether sulfone); PVA: polyvinyl alcohol.

Figure 3.

V(IV) concentration in the left reservoir of the permeation measuring device with Nafion117, SPES, and SPES/PVA membranes. SPES: sulfonated poly(ether sulfone); PVA: polyvinyl alcohol.

Comparison of SPES/PVA membrane and SPES membrane reveals that the PVA sub-layer has a big effect on the permeation rate of vanadium ions. As discussed in literature,¹¹ the crystallization of PVA can confine the swelling behavior of the membrane, which is closely related to its ion selectivity. With the addition of PVA, swelling is decreased in a certain extent. Table 2 shows that the proton conductivity of SPES/PVA double-layer membranes is higher due to the good proton conductivity of PVA layer. The potential performance of VRB membranes is often evaluated in terms of the ratio of proton conductivity to vanadium ions permeability defined as selectivity. Higher selectivity is necessary for achieving better performance. Table 2 and Figure 4 show that the ionic selectivity of the SPES/PVA membrane is dramatically higher than that of Nafion117 membrane.

Figure 4.

Selectivity of Nafion117, SPES, PVA, and SPES/PVA membranes. SPES: sulfonated poly(ether sulfone); PVA: polyvinyl alcohol.

VRB single cell performance

The charge–discharge performance of CE and energy efficiency (EE) of the VRB single cells assembled with pure SPES, SPES/PVA, and Nafion117 membranes at a current density of 36 mA cm⁻² for 100 cycles is listed in Table 3. The results clearly indicate that the PVA layer helps to improve the CE in single cell performance.

Table 3.

VRB single cell performances at 36 mA·cm⁻² after 100 cycles.

Samples	CE (%)	VE (%)	EE (%)
Nafion117	85.7	92.5	79.3
SPES	84.8	89.2	75.6
PVA	80.3	93.4	75.0
SPES/PVA	92.6	90.7	83.9

VRB: vanadium redox flow battery; CE: coulombic efficiency; VE: voltage efficiency; EE: energy efficiency; SPES: sulfonated poly(ether sulfone); PVA: polyvinyl alcohol.

Discharge voltage and charge voltage are both determined by the thermodynamic reduction potential of the redox couples in each half-cell and the overpotential of the cell. In VRB system, the ohmic overpotential is partly attributed to the membrane resistance. Therefore, higher membrane resistance will lead to higher charge voltage and lower discharge voltage, which is clearly shown in Figure 5. The single cell with SPES/PVA membrane exhibits apparently higher voltage efficiency compared with that of Nafion117. This result is well-matched with the lower area resistance of SPES/PVA membrane.

Figure 5.

Charge–discharge curves of VRB single cell with Nafion117 membrane and SPES/PVA membrane, respectively, (36 mA·cm⁻²). VRB: vanadium redox flow battery; SPES: sulfonated poly(ether sulfone); PVA: polyvinyl alcohol.

EE is an important indicator of energy loss during the charge–discharge process and often used in the energy storage system. Owing to the good balance of conductivity and ion permeability, the SPES/PVA membrane exhibits the highest EE in this work, indicating that SPES/PVA membrane has a great potential for VRB application from a comprehensive view.

Open circuit voltage (OCV) is usually used to evaluate the degree of self-discharge arising from ion permeability across the membrane. In this test, 2.0 M V(II)/V(III) in 3.0 M H₂SO₄ solution and 2.0 M V(IV)/V(V) in 3.0 M H₂SO₄ solution were cyclically pumped into negative and positive half-cell, respectively. As demonstrated in Figure 6, the OCV decay rate of SPES/PVA membrane is apparently lower than that of SPES and Nafion117 membranes. The entire self-discharge procedure of the VRB cell with SPES/PVA membrane lasts for more than 158 h, while the OCV of the cell with Nafion117 comes down to 0.8 V in less than 90 h. This result reveals that SPES/PVA membrane significantly lowers the permeation rate of vanadium ions which is well-matched with the result of permeability test.

Figure 6.

OCV decay of the VRB single cells with PVA membrane, SPES membrane, Nafion117 membrane, and SPES/PVA membrane, respectively. OCV: open circuit voltage; VRB: vanadium redox flow battery; SPES: sulfonated poly(ether sulfone); PVA: polyvinyl alcohol.

Conclusions

To exert the virtue of PVA and SPES fully, double-layer membrane was fabricated in this work. The vanadium ion permeability of the SPES/PVA double-layer membranes is significantly lower than that of Nafion117 membrane. VRB single cell with SPES/PVA membrane has higher CE, EE, and a much lower self-discharge rate than the cell with Nafion117 membrane. Therefore, due to the good performances and low cost, the SPES/PVA membrane is expected to have excellent commercial prospects as an ion exchange membrane for VRB systems.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Guanlin Li

References

Sukkar

Skyllas-Kazacos

. Modification of membranes using polyelectrolytes to improve water transfer properties in the vanadium redox battery. J Membrane Sci 2003; 222: 249–264.

Fabjan

Garche

Harrer

. The vanadium redox-battery: an efficient storage unit for photovoltaic system. Electrochim Acta 2001; 47: 825–831.

Mohammadi

Chieng

Skyllas-Kazacos

. Water transport study across commercial ion exchange membranes in the vanadium redox flow battery. J Membrane Sci 1997; 133: 151–159.

Ling

Jia

Liu

. Preparation and characterization of sulfonated poly (ether sulfone)/sulfonated poly(ether ether ketone) blend membrane for vanadium redox flow battery. J Membrane Sci 2012; 415: 306–312.

Gao

Sun

Gao

. Novel acid-base hybrid membrane based on amine-functionalized reduced graphene oxide and sulfonated polyimide for vanadium redox flow battery. Electrochimica Acta 2015; 158: 24–34.

Huang

RYM

Shao

Burns

. Sulfonation of poly (ether ether ketone) (PEEK): kinetic study and characterization. J Appl Polym Sci 2011; 82: 2651–2660.

Zhang

. Sulfonated polyimide/s-MoS2 double-layer membrane with high proton selectivity and good stability for vanadium redox flow battery. J Membrane Sci 2015; 49: 179–189.

Cao

. Morphology and performance of poly(ether sulfone)/sulfonated poly(ether ether ketone) blend porous membranes for vanadium flow battery application. RSC ADV 2014; 4: 400–406.

Wiedemann

Heintz

Lichtenthaler

. Transport properties of vanadium ions in cation exchange membranes: determination of diffusion coefficients using a dialysis cell. J Membr Sci 1998; 141: 215–221.

10.

Pivovar

Wang

Cussler

. Pervaporation membranes in direct methanol fuel cells. J Membrane Sci 1999; 154: 155–162.

11.

Mai

Zhang

. Nafion/polyvinylidene fluoride blend membranes with improved ion selectivity for vanadium redox flow battery application. J Power Sources 2011; 196: 5737–5741.