The ionic conductivity, mechanical performance and morphology of two-phase structural electrolytes based on polyethylene glycol, epoxy resin and nano-silica

Materials Science and Engineering B 219 (2017) 37–44

Contents lists available at ScienceDirect

Materials Science and Engineering B journal homepage: www.elsevier.com/locate/mseb

The ionic conductivity, mechanical performance and morphology of twophase structural electrolytes based on polyethylene glycol, epoxy resin and nano-silica Qihang Feng, Jiping Yang ⇑, Yalin Yu, Fangyu Tian, Boming Zhang, Mengjie Feng, Shubin Wang School of Materials Science and Engineering, Beihang University, Beijing 100191, China

a r t i c l e

i n f o

Article history: Received 6 December 2016 Received in revised form 2 March 2017 Accepted 3 March 2017

Keywords: Structural electrolytes Epoxy resin Nano-silica Mechanical performance Ionic conductivity

a b s t r a c t As one of significant parts of structural power composites, structural electrolytes have desirable mechanical properties like structural resins while integrating enough ionic conductivity to work as electrolytes. Here, a series of polyethylene glycol (PEG)-epoxy-based electrolytes filled with nano-silica were prepared. The ionic conductivity and mechanical performance were studied as functions of PEG content, lithium salt concentration, nano-silica content and different curing agents. It was found that, PEG-600 and PEG-2000 content in the epoxy electrolyte system had a significant effect on their ionic conductivity. Furthermore, increasing the nano-silica content in the system induced increased ionic conductivity, decreased glass transition temperature and mechanical properties, and more interconnected irregular network in the cured systems. The introduction of rigid m-xylylenediamine resulted in enhanced mechanical properties and reasonably decreased ionic conductivity. As a result, these two-phase epoxy structural electrolytes have great potential to be used in the multifunctional energy storage devices. Ó 2017 Elsevier B.V. All rights reserved.

1. Introduction Lightweight has been more and more attractive in material design and fabrication for the promotion of energy efficiency in many applications [1]. The structural power composites (SPCs), a series of multifunctional materials simultaneously realizing the goal of load bearing and energy storage, are believed to be one of the most promising alternatives for conventional structural materials in manufacturing energy storage devices [2–4]. As vital components in SPCs, structural electrolytes play a critical role in providing high ionic conductivity and transferring mechanical loads [5–7]. In order for the optimization of structural electrolytes, not only mechanical properties, but also the electrochemical performance, i.e., the ionic conductivity, was considered to realize a better multifunctional performance [8–11]. The preparation of structural electrolytes mainly focused on two major strategies: one is the polymerization of certain kinds of monomers to form the capabilities of ionic-conductivity and load-bearing simultaneously, the other is blending ionicconducting components with structural components to combine these two properties together. Snyder and co-workers [12,13] synthesized homopolymers and copolymers of the vinyl ester deriva⇑ Corresponding author. E-mail address: [email protected] (J. Yang). http://dx.doi.org/10.1016/j.mseb.2017.03.001 0921-5107/Ó 2017 Elsevier B.V. All rights reserved.

tives containing polyethylene glycol (PEG), with a structural scaffold generated through crosslinking and the PEG side chains to enhance their ion transport capability. Willgert et al. [14,15] synthesized polyethylene oxide (PEO)-dimethacrylate/PEO-mono methacrylate lithium salt mixtures by photo-initiated curing, and the relationship between these polymers’ ionic conductivity and mechanical properties was investigated. However, despite their manufacture process is simple and convenient, the trade-off between mechanical and conductive properties makes it difficult to realize an acceptable multifunctional performance for the SPC [12–15]. For another strategy, epoxy resin, a class of structural resin with high mechanical modulus and strength, have been widely used as structural polymer matrix. And the addition of the immiscible electrolyte into epoxy/amine system can form the two-phase phase structure, with the epoxy networks bearing loads and the electrolyte phase achieving ionic conductivity [16–18]. Recently, ionic liquid, propylene carbonate (PC) were used as additives to generate a bi-component structural electrolyte [17]. Their mechanical performance was relatively low because of plasticization of the structural network. In addition, PEO or PEG was blended with epoxy and curing agent to study their mechanical properties, miscibility, and morphology [19–22]. Lu et al. [19] investigated the PEO/epoxy/amine system after curing and found that the molecular weight of PEO

38

Q. Feng et al. / Materials Science and Engineering B 219 (2017) 37–44

2.2. Preparation of samples

had a crucial influence on the components’ miscibility. Hu et al. [20] found that for PEO-20000/epoxy blends cured by 1,3,5trihydroxybenzene, the crystallization of PEO affected the forming kinetics and morphology of crosslinking network. The morphology and kinetics of the PEO crystallization were in turn influenced by the content and chemical structural of epoxy/amine system [21]. Feng et al. [22] found that the addition of PEG simultaneously enhanced the cryogenic tensile strength, ductility of epoxy resins. However, to our best knowledge, few literatures available have ever discussed the relationship between ionic conductivity and mechanical properties of PEG/epoxy systems though PEG has very high ionic conductivity, which merits further study to optimize the multifunctional performance of structural electrolytes. Because bisphenol A epoxy resin contains the aromatic rings which present good mechanical performance after curing, and is miscible with PEG [23,24], bisphenol A epoxy resin was blended with PEG, lithium salt, nano-silica and curing agent to form the two-phase microstructures in this paper. Two different curing agents, polyetheramine D-400 and m-xylylenediamine, were used to form the crosslinking network for load-bearing, and PEG mixed with lithium salt acted as the ionic conductive pathways. Nanosilica was added into the system and the influence of lithium salt concentration and nano-silica content on ionic conductivity, mechanical performance and morphology of the structural electrolytes were discussed. Our results suggest that these two-phase structural electrolytes containing PEG, epoxy resin and nanosilica have great potential to be used as structural electrolytes in the multifunctional energy storage devices.

First, epoxy systems with different PEG content were prepared, the molar ratio of lithium and etheric oxygen in PEG ([Li]/[EO]) and the content of nano-silica were kept for 0.1 and 2.5 wt%, respectively. Then, the composition of PEG was changed to the mixture of PEG-2000 and PEG-600, the ratio of [Li]/[EO] and nano-silica content were still kept for 0.1 and 2.5 wt%, respectively. Furthermore, epoxy systems with varied [Li]/[EO] and nano-silica content were prepared. The general preparation procedure was as follows: First, PEG and LiTF were mixed and ultra-sonicated until homogeneous liquid was formed and no precipitate was left. Then nano-silica and E51 were added into the mixture, followed by transferring them into the planetary mixer (ZYMC-200, Shenzhen ZYE Technology Limited). And they were mixed with a speed of 2000 rpm until a homogeneous mixture was formed. Finally, curing agents were added into the mixture and evacuated to degas until a uniform mixture was formed. Two different curing agents, polyetheramine D-400 and MXDA, were used, respectively. And the molar ratio between epoxy group in E51 and active hydrogen in curing agent for all systems was kept constant for 1. Three different shapes of samples, disc samples for electrochemical impedance spectroscopy and morphology, bars for dynamic mechanical analysis (DMA) and dumbbell samples for tensile testing, were made by different Teflon molds. The mixture was poured into different preheated molds and degassed in vacuum oven at 60 °C for 20 min. The curing cycle was 80 °C for 2 h, 100 °C for 2 h and 120 °C for 2 h, the rate at which all the heating and cooling cycle was 2 °C/min. Macroscopic phase separation cannot be observed during the curing process. After demolding, the samples were machined to be smooth for characterizations. All samples should be kept in an ambient temperature vacuum drier prior to characterization to remove absorbing moisture.

2. Experimental 2.1. Materials

2.3. Characterizations

The epoxy resin used in this study was liquid diglycidyl ether of bisphenol A (DGEBA) type epoxy resin E51 with an epoxy value of 0.51 purchased from Jinan Sunny Chemical Technology Co., Ltd. Nano-silica (purity: 99.5%, particle size: 30 nm) was from Shanghai Maikun Chemical Co. Ltd. PEG-2000 (molecular weight: 2000 Da) and PEG-600 (molecular weight: 600 Da) were purchased from TCI Shanghai. Lithium trifluoromethanesulfonate (LiTF) was from China Ship building Industry Corporation. Polyetheramine D-400 and m-Xylylenediamine (MXDA) were bought from Aladdin. The chemical structures of DGEBA, Polyetheramine D-400 and MXDA were shown in Fig. 1. All materials were dried in an ambient vacuum oven prior to use to remove moisture. All other chemicals were purchased from Beijing Chemical Works and used as received.

OH

CH3

O

2.3.1. Electrochemical impedance spectroscopy The ionic conductivity of the cured disc samples (2–3 mm thickness, 20 mm diameter) was tested with CHI660E (Shanghai Chenhua Device Company, China) electrochemical workstation by electrochemical impedance analysis, which was carried out at ambient temperature (23–25 °C) using an amplitude of 10 mV in the frequency range of 1 Hz to 1 MHz. Samples were placed into a fixer for clamping with two stainless steel electrodes, and the whole fixer was sealed into an incubator. The high frequency arc was associated with the ionic conduction process in the bulk of the electrolyte system [25]. The low frequency part was attributed to blocking double layer capacitance near the electrode–electrolyte

O

O CH3 H2N

CH3 O

n

H2N

DGEBA

CH3 NH2

O CH3

O O

X

Polyetheramine D-400

CH3 NH2

MXDA

Fig. 1. Chemical structures of DGEBA, polyetheramine D-400 and MXDA.

Q. Feng et al. / Materials Science and Engineering B 219 (2017) 37–44

interface caused by the ion migration [26,27]. Bulk resistance (Rb) of the tested sample was obtained from the high-frequency intercept on the real axis [28]. The resulting resistance was used to calculate ionic conductivity, r, using the following Eq. (1) [28]:

r¼

l Rb  A

ð1Þ

where l is the disc’s thickness and A is the cross-sectional area which is calculated from the disc’s diameter. Each formulation was tested for five samples. The thickness and diameter of each sample were measured by a digital caliper. 2.3.2. Differential scanning calorimetry The thermal behavior of cured epoxy systems was studied by differential scanning calorimetry (DSC, Shimadzu DSC-60Plus) in a nitrogen atmosphere. The temperature was raised from room temperature to 150 °C using 3–5 mg of sample, at a heating rate of 10 °C/min. To ensure the reproducibility, for each electrolyte system, three parallel tests were conducted, respectively [29]. 2.3.3. Dynamical mechanical analysis Dynamic mechanical analysis (DMA, METTLER DMA-1) was conducted in a method of 3-point bending at a frequency of 1 Hz with an amplitude of 5 lm. And the epoxy systems were heated from 100 °C to 150 °C at a heating rate of 3 °C/min. Glass transition temperature of the cured electrolytes was determined from the peak of the tan d-temperature curve. 2.3.4. Tensile testing The mechanical properties of the cured dumbbell samples were evaluated by tensile testing according to ASTM D790 using an electrical universal testing machine (Instron 5565) at a crosshead speed of 5 mm/min. The tensile strength and tensile modulus were determined with a 1 kN force transducer. Gage section dimensions were as follows: length: 60–65 mm; width: 5–6 mm; thickness: 3– 4 mm. For each formulation, eight specimens were tested in order to guarantee at least 5 effective data of tensile properties. 2.3.5. Scanning electron microscope Scanning electron microscopy (SEM, CamScan JEOL 6010) studies were performed after washing and drying samples extensively. Samples with different content of nano-silica were soaked for over a week in ethanol and the ethanol was changed every 12 h to extract out the ionic conductive components. After soaking, all samples were placed in vacuum at 70 °C until the mass of each sample was constant. 3. Results and discussion 3.1. Effect of PEG content In these PEG-epoxy resin electrolyte systems, lithium salt was dissolved in PEG due to the interaction between lithium ion and the etheric oxygen (EO). In order to investigate the influence of PEG content on the ionic conductivity of the cured resin electrolyte systems, different PEG-2000 content systems were prepared with a nano-silica content of 2.5 wt% and [Li]/[EO] ratio of 0.1. The Nyquist impedance plots and the equivalent circuit of different PEG-2000 content systems were shown in Supporting information (Fig. S1(a) and (b)). Their ionic conductivity was calculated according to Eq. (1) and presented in Fig. 2(a). When the content of PEG-2000 varied from 30 wt% to 60 wt%, the ionic conductivity increased with the increase of the PEG content in the epoxy electrolyte systems. It is well known that the ionic conductivity is greatly affected by charge carrier concentra-

39

tion of the systems [30–33]. Since the molar ratio of [Li]/[EO] was fixed, and the lithium ion is the charge carrier in ionic conductive phase, when the content of PEG-2000 was increased, the concentration of charge carrier was increased [33]. On the other hand, the content of epoxy/amine decreased, leading to a lower crosslinking density in the system, so the restriction onto the movement of PEG segment from polymer network was weaken. As a result, with the increase of ionic conductive component, there was a nearly one order of magnitude’s increase in ionic conductivity. When PEG-2000 content increased from 50 wt% to 60 wt%, the improvement in ionic conductivity was relatively less obvious than the improvement caused by the increase of PEG-2000 content from 40 wt% to 50 wt%. Meanwhile, the system with PEG-2000 content of 60 wt% was too soft for mold-releasing and its storage modulus (83.1 MPa) was too low (Fig. S2). When the content of PEG-2000 reached 70 wt%, the system was still liquid after curing due to the polymer network was too weak to bear mechanical loads. To increase the ionic conductivity further, liquid PEG-600 at different content was added to the epoxy system, and the mass fraction of PEG (PEG 600 + PEG 2000) was kept unchanged at 50 wt%. The Nyquist impedance plots and the equivalent circuit of the systems with different PEG-600 content were shown in Supporting information (Fig. S3 (a) and (b)). As shown in Fig. 2(b), the ionic conductivity of epoxy electrolyte system increased with the increase of PEG-600 mass fraction. Because the segment movement plays a critical role in the ionic conductive behavior [34], it is advantageous for lithium ion to interact with PEG segment when low molecular weight PEG-600 was blended into the epoxy electrolyte system. On the other hand, PEG-600 is liquid at room temperature, and the increase of PEG-600 content inhibits the crystallization ability of PEG 2000 (the DSC curves are shown in Fig. S4), prompting a more effective dissociation of lithium ion in lithium salt [35], therefore resulting in increased ionic conductivity in epoxy systems. However, higher PEG-600 content (for example, when the content of PEG-600 reached 30 wt%) resulted in a soft electrolyte system after curing (Fig. S5), which is not good for a higher mechanical load. Therefore the PEG-600 content of 25 wt% and PEG 2000 content of 25 wt% in the electrolyte systems was chosen in following research. 3.2. Effect of [Li]/[EO] and nano-silica content 3.2.1. Ionic conductivity The relationship between ionic conductivity and the concentration of lithium salt concentration was depicted in Fig. 3(a). The Nyquist plots of different [Li]/[EO] ratio (Fig. S6(a)–(c)) and the equivalent circuit (Fig. S6(d)) were presented in Supporting information. When the molar ratio of lithium salt to etheric oxygen ([Li]/[EO]) increased from 0.10 to 0.125, the ionic conductivity of epoxy systems at same content of nano-silica was increased. But higher lithium salt concentration did not improve the system’s ionic conductivity when the nano-silica content was above 5 wt %. The slight decrease in ionic conductivity was due to the limitation of segment’s movement and the formation of ion pairs which did not contribute to ionic transport [36], because the epoxy electrolyte system with 0.150 molar ratio of [Li]/[EO] was more viscous than the systems with 0.100 and 0.125 molar ratio of [Li]/[EO] before curing. On the other hand, for electrolyte systems with the same lithium salt concentration, ionic conductivity of epoxy electrolyte systems increased with the increase of the nano-silica content from 0 to 10 wt%, as presented in Fig. 3(b). To explain these investigations, DSC was used to trace the thermal change of the cured epoxy electrolyte systems with different lithium concentration and nano-silica content. In these DSC curves,

Q. Feng et al. / Materials Science and Engineering B 219 (2017) 37–44

(a)

10

10

(b)

Ionic conductivity/(mS/cm)

Ionic conductivity/(mS/cm)

40

-3

-4

30

40

50

PEG-2000 content/wt %

10

-1

10

-2

10

-3

60

0

10

20

PEG-600 content/wt %

30

-1

Ionic conductivity/(mS/cm)

10

10

Nano-silica content:

(a)

10% 7.5% 5.0% 2.5%

-2

0

0.100

0.125

0.150

molar ratio of [Li]/EO]

Ionic conductivity/(mS/cm)

Fig. 2. The influence of PEG-2000 content (a), and PEG-600 content (b) on the ionic conductivity of epoxy electrolyte system with a nano-silica content of 2.5 wt% and [Li]/ [EO] ratio of 0.1.

-1

10

(b)

[Li]/[EO]: 0.125 0.15 0.10

-2

10

0.0

2.5

5.0

7.5

10.0

Nano-silica content/wt %

Fig. 3. The influence of molar ratio of [Li]/[EO] (a), nano-silica content (b) on the ionic conductivity of epoxy electrolyte systems.

‘exo’ meant exothermic, so if the direction of the peak was the same as the arrow, the peak represented an exothermic process, otherwise it represented an endothermic process. As presented in Fig. 4(a), all of the filler-free systems (nano-silica content = 0) had an obvious endothermic peak, which was corresponded to the melting of PEG 2000 in the system. However, the endothermic peaks gradually decreased with the addition of nano-silica, and nearly disappeared when the nano-silica content was more than 5%, as shown in Fig. 4(b)–(d). It was observed that the added nano-fillers in the polymer could increase the amorphous content in the polymer, because they could act as intercalating agent into polymer chain domain so that chain segments arranged less regularly, and the motion of chain segments was enhanced, resulting in the disappearance of the melting endothermic peaks in the epoxy electrolyte systems [37,38]. Considering the influence of the nano-silica content on their ionic conductivity of epoxy electrolyte systems presented in Fig. 2(b), the addition of nano-silica in epoxy electrolyte systems is a promising method to increase their ionic conductivity.

3.2.2. Mechanical properties Furthermore, the effect of the nano-silica content on the dynamical mechanical properties of the epoxy electrolyte systems with the molar ratio of [Li]/[EO] = 0.1, 0.125 and 0.15 were investigated and presented in Fig. 5. It was found that, the glass transition temperature (Tg), determined by the tan d peak in the dynamical mechanical property curves, decreased with the increase of nano-silica content. For the system with molar ratio of [Li]/[EO] = 0.125, when the nanosilica content in the epoxy system increased from 0 to 10 wt%, Tg

of resulting epoxy system dropped nearly 20 °C. The storage modulus at ambient temperature (23 °C) showed the same tendency, decreased from 450 MPa to nearly 50 MPa with an increased nano-silica content in the epoxy system. This is because that, on one hand, nano-silica with a large specific surface area has a strong interaction with the surrounding environment, thus the Tg and storage modulus will be increased [40]. On the other hand, increasing the nano-silica content can enhance the chain flexibility of PEG as discussed above, leading to a decreased Tg and storage modulus. For the same nano-silica content, the system with molar ratio of [Li]/[EO] = 0.15 were softer than that with [Li]/[EO] = 0.125, and the storage modulus of samples was much smaller. It was mainly caused by lithium salt which hindered the movement of polymer chains, resulting in inhibition of the crystallization process [39]. Therefore the static tensile test was conducted for the epoxy systems at [Li]/[EO] ratio = 0.125 (Fig. 6). The tensile strength and tensile modulus was decreased with an increased nano-silica content in the epoxy systems, which was in accordance with the Tg and storage modulus change tendency. With the addition of nanosilica in epoxy systems, the tensile modulus was decreased substantially from 320 MPa to 132 MPa, which was because the addition of the nano-silica’s decreased the crystallinity of the ionic conductive PEG phase, as confirmed in Fig. 4(c).

3.2.3. The morphology As the ionic conductivity and mechanical performance were affected by the system’s microstructure [16,18], the morphology of epoxy systems with different content of nano-silica was investigated using SEM, of which the ratio of [Li]/[EO] was 0.125 and cured by polyetheramine D-400, the ionic conductive components

Q. Feng et al. / Materials Science and Engineering B 219 (2017) 37–44

41

Fig. 4. DSC traces for the cured epoxy systems with different lithium concentration and without nano-silica (a), different nano-silica content at [Li]/[EO] = 0.1 (b), 0.125 (c) and 0.15 (d), respectively.

Fig. 5. The influence of [Li]/[EO] and nano-silica content on the Tg and storage modulus of the epoxy systems at ambient temperature.

Fig. 6. Tensile strength and modulus of epoxy systems with different nano-silica content. Samples were all cured by D-400 and the ratio of [Li]/[EO] was 0.125.

PEG and lithium salt were removed prior to SEM by soaking all samples into ethanol and drying in the oven until the weight of the sample was constant [16]. As presented in Fig. 7, all samples exhibited a two-phase microstructure after removing ionic conductive components, with epoxy forming a porous network. In the cross section of epoxy system without nano-silica (Fig. 7(a)), the two-phase phase was composed of long parallel epoxy phase, with a width of approximately 5 lm, and irregular-shaped holes with a non-uniform diameter varying from 1 to 3 lm scattered in the epoxy phase. After adding 2.5 wt% of nano-silica (Fig. 7(b)), the shape of the epoxy phase was transformed to continuous sheet penetrated with holes that had a

diameter of 1–5 lm. The holes were mostly isolated, and the amount was higher than the system without nano-silica for the same area. When the content of nano-silica reached 5.0 wt% (Fig. 7(c)), the holes were relatively uniformly distributed, some of which became connected. And the space between holes was narrower. When the content of nano-silica was higher as 7.5% and 10% (Fig. 7(d) and (e)), more and more holes formed, which is beneficial to a more continuous ionic-conductive phase. On the other hand, the crosslinking network was weaker and had more defects in the continuous phase containing 7.5 wt% and 10 wt% nano-silica. Therefore, the weakness of polymer network’s restriction onto the ion mobility [41], as well as the formation of interconnected

42

Q. Feng et al. / Materials Science and Engineering B 219 (2017) 37–44

Fig. 7. SEM images of epoxy systems with different nano-silica content after the extraction of the ionic conductive components: (a) The system of nano-silica-free; (b–e) systems with nano-silica content of 2.5 wt%, 5.0% wt%, 7.5% wt% and 10% wt%, respectively.

ion-transporting pathways [42], resulted in the enhanced ionic conductivity in these epoxy systems.

3.3. Effect of different curing agents Furthermore, the effect of different curing agents on the ionic conductivity and mechanical performance of epoxy systems was studied. When m-xylylenediamine (MXDA) was used, the content of ionic conductive and mechanical-bearing component were kept constant compared to D-400 cured systems and the ratio of [Li]/ [EO] was still 0.125. It was clear that the ionic conductivity of the epoxy systems cured by MXDA was lower than epoxy systems cured by D-400, as shown in Fig. 8. This is reasonable, because the aromatic MXDA improves the rigidity of the crosslinking epoxy network after curing, and the restriction onto the chain segment is stronger compared to the aliphatic D-400, so the segment movement is weaker [43,44]. On the other hand, the interaction between etheric oxygen in D-400 chains and lithium ion also might contribute to

Fig. 8. The ionic conductivity of epoxy systems using different curing agents: D-400 (solid triangles) and MXDA (open triangles). The ratio of [Li]/[EO] was 0.125 for all epoxy systems.

Q. Feng et al. / Materials Science and Engineering B 219 (2017) 37–44

the ion-transporting, which could enhance the ionic conductivity in the D-400 cured epoxy systems. Furthermore, DSC curves showed that the increase of nanosilica content decreased the crystallinity of the electrolyte system (Fig. S7). In addition, the utilization of MXDA as curing agent did not change the tendency of mechanical properties (storage modulus, tensile strength and tensile modulus) of the epoxy systems with the different content of nano-silica, but these properties were increased compared to the epoxy systems cured by D-400 at the same nano-silica content, as shown in Fig. 9(a) and (b). The Tg also decreased with the increase of nano-silica content in the epoxy systems, but was higher than those epoxy systems at the same nano-silica content cured by D-400. It is reasonable because the structure of benzene ring in MXDA hindered the movement of adjacent segments in the crosslinking network [45–47], the rigidity of the polymer network was improved, inducing increased Tg, storage modulus, tensile strength and tensile modulus for the epoxy systems cured by MXDA at the same nano-silica content. 3.4. Multifunctional performance of different structural electrolyte systems It is well-known that there has been a trade-off relationship between the mechanical performance and ionic conductivity in the structural electrolyte systems, because higher mechanical performance lies on higher stiffness of the polymer chains, and higher ionic conductivity needs a more flexible chain segments and interpenetrated pathways throughout the matrix [15]. The effective way to overcome such a conflict is to change the microstructure of the structural electrolyte systems by controlling the composition and process parameters of preparation, thereby achieving a controllable phase separation [3]. Multifunctionality plots are effective to consider the trade-off relationship between ionic conductivity and tensile modulus for structural electrolytes.[16] To evaluate the effect of nano-silica’s addition on the systems’ multifunctional performance, the ionic conductivity of the system were plotted against tensile modulus (Fig. 10). Some of structural electrolytes’ ionic conductivity and mechanical properties in the literatures (points 1–4) were collected in Fig. 10, in which point 5 and point 6 represented some of our optimized epoxy electrolyte systems in this work. When the content of nano-silica was 10 wt%, the epoxy system cured by D-400 (point 5) had a room temperature ionic conductivity of 0.086 mS/cm and a tensile modulus of 135 MPa, showing a good combination between the ionic conductivity and tensile modulus. When MXDA was used as curing agent (point 6), the ionic conductivity could be 0.055 mS/cm while tensile modulus reached to 157 MPa. Curing agent containing a structure of benzene ring,

43

Fig. 10. Summary of ionic conductivity and tensile modulus data on structural electrolytes in some literatures: 1-(Weight ratio between modified poly(methyl methacrylate) and ionic liquid: 50:50) [48]; 2-(Weight ratio between epoxy and ionic liquid: 60:40) [49]; 3-(Weight ratio between epoxy-PEO and ionic liquid: 60:40) [50]; 4-(Electrolytes based on epoxy system VTM266) [16]; Solid square point 5 and point 6 represented some of epoxy electrolyte systems in our study, 5([Li]/[EO] = 0.125, 10 wt% nano-silica, curing agent: D-400), 6-([Li]/[EO] = 0.125, 10 wt% nano-silica, curing agent: MXDA).

i.e., MXDA, can bring an enhancement of rigidity of the crosslinking epoxy network, result in increased tensile modulus and decreased ionic conductivity in the cured epoxy systems due to stronger restriction onto chain segment of PEG. These optimized epoxy system had better multifunctional property than epoxy resin and PEG. Compared to the results in literatures (points 1–4 in Fig. 10), our epoxy systems have acceptable ionic conductivity and tensile modulus simultaneously, which are more approaching to the threshold [16,17,40] as shown in Fig. 10, suggesting that the two-phase epoxy electrolyte systems based on PEG, epoxy resin and nanosilica is a promising candidate for structural power composites. 4. Conclusions Structural electrolytes based on PEG-epoxy resins were prepared using PEG, LiTF, nano-silica and epoxy/curing agent together. The ionic conductivity increased with the increase of PEG-600 and PEG-2000 content in the epoxy system. Furthermore, the content of nano-silica in the epoxy systems had great influence on the ionic conductivity and mechanical properties of epoxy systems. The ionic conductivity increased with the increase of the nano-silica content in the epoxy systems. With the increase of nano-silica content in the epoxy systems at [Li]/[EO] ratio of 0.125, the homogeneity of the cured epoxy systems increased presented using DSC

Fig. 9. (a) The Tg and storage modulus at ambient temperature, (b) tensile strength and tensile modulus of epoxy systems cured by MXDA at different content nano-silica. The ratio of [Li]/[EO] was 0.125 for all epoxy systems.

44

Q. Feng et al. / Materials Science and Engineering B 219 (2017) 37–44

results, while the Tg, tensile strength and tensile modulus decreased, and more interconnected irregular network morphology was obtained. Both ionic conductivity and mechanical properties of epoxy systems with certain [Li]/[EO] ratio and nano-silica content achieved a certain degree of balance compared to the initial PEG and epoxy resin systems. When the ratio of [Li]/[EO] was 0.125 and the nano-silica content was 10 wt%, epoxy systems cured by D-400 and MXDA had better multifunctional performance in our study. Acknowledgements This work was financially supported by the National Natural Science Foundation of China (Grant No. 21174009, 21574003). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.mseb.2017.03. 001. References [1] R.F. Gibson, Compos. Struct. 92 (2010) 2793–2810. [2] N. Shirshova, H. Qian, M.S.P. Shaffer, J.H.G. Steinke, E.S. Greenhalgh, P.T. Curtis, A. Kucernak, A. Bismarck, Compos. Part A 46 (2013) 96–107. [3] L.E. Asp, E.S. Greenhalgh, Compos. Sci. Technol. 101 (2014) 41–61. [4] L.E. Asp, Plast., Rubber Compos. 42 (2013) 144–149. [5] D.J. O’Brien, D.M. Baechle, E.D. Wetzel, J. Compos. Mater. 45 (2011) 2797–2809. [6] N. Shirshova, H. Qian, M. Houlle, J.H.G. Steinke, A.R.J. Kucernak, Q.P.V. Fontana, E.S. Greenhalgh, A. Bismarck, M.S.P. Shaffer, Faraday Discuss. 172 (2014) 81– 103. [7] E.S. Greenhalgh, J. Ankersen, L.E. Asp, A. Bismarck, Q.P.V. Fontana, M. Houlle, G. Kalinka, A. Kucernak, M. Mistry, S. Nguyen, H. Qian, M.S.P. Shaffer, N. Shirshova, J.H.G. Steinke, M. Wienrich, J. Compos. Mater. 49 (2015) 1823–1834. [8] Y.G. Andreev, P.G. Bruce, Electrochim. Acta 45 (2000) 1417–1423. [9] A.M. Stephan, K.S. Nahm, Polymer 47 (2006) 5952–5964. [10] E. Jacques, M.H. Kjell, D. Zenkert, G. Lindbergh, M. Behm, M. Willgert, Compos. Sci. Technol. 72 (2012) 792–798. [11] P. Liu, E. Sherman, A. Jacobsen, J. Power Sources 189 (2009) 646–650. [12] J.F. Snyder, R.H. Carter, E.D. Wetzel, Chem. Mater. 19 (2007) 3793–3801. [13] J.F. Snyder, E.D. Wetzel, C.M. Watson, Polymer 50 (2009) 4906–4916. [14] M. Willgert, M.H. Kjell, G. Lindbergh, M. Johansson, Solid State Ionics 236 (2013) 22–29. [15] M. Wingert, M.H. Kjell, E. Jacques, M. Behm, G. Lindbergh, M. Johansson, Eur. Polym. J. 47 (2011) 2372–2378. [16] N. Shirshova, A. Bismarck, S. Carreyette, Q.P.V. Fontana, E.S. Greenhalgh, P. Jacobsson, P. Johansson, M.J. Marczewski, G. Kalinka, A.R.J. Kucernak, J. Scheers, M.S.P. Shaffer, J.H.G. Steinke, M. Wienrich, J. Mater. Chem. A 1 (2013) 15300– 15309.

[17] N. Shirshova, P. Johansson, M.J. Marczewski, E. Kot, D. Ensling, A. Bismarck, J.H. G. Steinke, J. Mater. Chem. A 1 (2013) 9612–9619. [18] N. Shirshova, A. Bismarck, E.S. Greenhalgh, P. Johansson, G. Kalinka, M.J. Marczewski, M.S.P. Shaffer, M. Wienrich, J. Phys. Chem. C 118 (2014) 28377– 28387. [19] S. Hu, R. Zhang, X. Wang, P. Sun, W. Lv, Q. Liu, N. Jia, Eur. Phys. J. E 38 (2015) 118–126. [20] L. Hu, H. Lv, S. Zheng, J. Polym. Sci. B 42 (2004) 2567–2575. [21] Q. Guo, C. Harrats, G. Groeninckx, M.H.J. Koch, Polymer 42 (2001) 4127–4140. [22] Q. Feng, J. Yang, Y. Liu, H. Xiao, S. Fu, J. Mater. Sci. Technol. 30 (2014) 90–96. [23] A.M. Rocco, C.P. Fonseca, F.A.M. Loureiro, R.P. Pereira, Solid State Ionics 166 (2004) 115–126. [24] Y. Kang, K. Cheong, K.-A. Noh, C. Lee, D.-Y. Seung, J. Power Sources 119 (2003) 432–437. [25] S. Lanfredi, P.S. Saia, R. Lebullenger, A.C. Hernandes, Solid State Ionics 146 (2002) 329–339. [26] N.K. Karan, D.K. Pradhan, R. Thomas, B. Natesan, R.S. Katiyar, Solid State Ionics 179 (2008) 689–696. [27] R. Epur, M. Ramanathan, F.R. Beck, A. Manivannan, P.N. Kumta, Mater. Sci. Eng., B 177 (2012) 1151–1156. [28] A.R. Polu, D.K. Kim, H.-W. Rhee, Ionics 21 (2015) 2771–2780. [29] K. Karuppasamy, S. Thanikaikarasan, R. Antony, S. Balakumar, X.S. Shajan, Ionics 18 (2012) 737–745. [30] Y.T. Chen, Y.C. Chuang, J.H. Su, H.C. Yu, Y.W. Chen-Yang, J. Power Sources 196 (2011) 2802–2809. [31] R. Kuruba, M.K. Datta, K. Damodaran, P.H. Jampani, B. Gattu, P.P. Patel, P.M. Shanthi, S. Damle, P.N. Kumta, J. Power Sources 298 (2015) 331–340. [32] S.-J. Kwon, D.-G. Kim, J. Shim, J.H. Lee, J.-H. Baik, J.-C. Lee, Polymer 55 (2014) 2799–2808. [33] A. Sun, F.R. Beck, D. Haynes, J.A. Poston Jr., S.R. Narayanan, P.N. Kumta, A. Manivannan, Mater. Sci. Eng., B 177 (2012) 1729–1733. [34] P. Johansson, Polymer 42 (2001) 4367–4373. [35] D. Devaux, R. Bouchet, D. Gle, R. Denoyel, Solid State Ionics 227 (2012) 119– 127. [36] P. Pradeepa, S.E. Raj, G. Sowmya, J. Kalaiselvimary, M.R. Prabhu, Mater. Sci. Eng., B 205 (2016) 6–17. [37] L. Fan, Z. Dang, G. Wei, C. Nan, M. Li, Mater. Sci. Eng., B 99 (2003) 340–343. [38] T. Kuila, H. Acharya, S.K. Srivastava, B.K. Samantaray, S. Kureti, Mater. Sci. Eng., B 137 (2007) 217–224. [39] S. Klongkan, J. Pumchusak, Electrochim. Acta 161 (2015) 171–176. [40] Y. Yu, B. Zhang, Y. Wang, G. Qi, F. Tian, J. Yang, S. Wang, Mater. Des. 104 (2016) 126–133. [41] A. Dey, S. Karan, A. Dey, S.K. De, Mater. Res. Bull. 46 (2011) 2009–2015. [42] A. Dey, T. Ghoshal, S. Karan, S.K. De, J. Appl. Phys. 110 (2011) 043707. [43] W.J. Liang, C.L. Kuo, C.L. Lin, P.L. Kuo, J. Polym. Sci. A 40 (2002) 1226–1235. [44] P.-L. Kuo, W.-J. Liang, T.-Y. Chen, Polymer 44 (2003) 2957–2964. [45] K. Huang, Y. Zhang, M. Li, J. Lian, X. Yang, J. Xia, Prog. Org. Coat. 74 (2012) 240– 247. [46] T. Li, X. Liu, Y. Jiang, S. Ma, J. Zhu, J. Appl. Polym. Sci. 133 (2016), http://dx.doi. org/10.1002/APP.44219. [47] G. Wang, G. Jiang, J. Zhang, Thermochim. Acta 589 (2014) 197–206. [48] F. Gayet, L. Viau, F. Leroux, S. Monge, J.-J. Robin, A. Vioux, J. Mater. Chem. 20 (2010) 9456–9462. [49] K. Matsumoto, T. Endo, J. Polym. Sci. A 49 (2011) 1874–1880. [50] K. Matsumoto, T. Endo, Macromolecules 41 (2008) 6981–6986.