Effects on the Properties after Addition of Lithium Salt in Poly(ethylene oxide)/Poly(methyl acrylate) Blends

Abstract

The studies of phase behavior, dielectric relaxation, and other properties of poly(ethylene oxide) (PEO)/poly(methyl acrylate) (PMA) blends with the addition of lithium perchlorate (LiClO₄) were done for different blend compositions. Samples were prepared by a solution casting technique. The binary PEO/PMA blends exhibit a single and compositional-dependent glass transition temperature (T_g), which is also true for ternary mixtures of PEO/PMA/LiClO₄ when PEO was in excess with low content of salt. These may indicate miscibility of the constituents for the molten systems and amorphous domains of the systems at room temperature from the macroscopic point of view. Subsequently, the morphology of PEO/PMA blends with or without salt are correlated to the phase behavior of the systems. Phase morphology and molecular interaction of polymer chains by salt ions of the systems may rule the dielectric or electric relaxation at room temperature, which was estimated using electrochemical impedance spectroscopy (EIS). The frequency-dependent impedance spectra are of interest for the elucidation of polarization and relaxation of the charged entities for the systems. Relaxation can be noted only when a sufficient amount of salt is added into the systems.

1. Introduction

Lithium (Li)-ion batteries that exhibit appreciably high energy and power density are the choice of the electrochemical energy storage for portable electronics/devices, hybrid/full electric vehicles, etc. [1,2,3]. Li-ion batteries are rechargeable devices, where both electrodes are intercalation materials, and the commercial electrolyte is commonly Li salt dissolved in a mixture of organic solvents [3,4,5]. Extensive studies on organic solvent-free electrolytes, for example on solid polymer electrolytes (SPEs), have been carried out [6,7] since 1980. One of the popular polymer hosts for SPEs is poly(ethylene oxide) (PEO), which is a semi-crystalline polymer with crystallinity amount up to 70% [8,9,10,11,12]. It is widely accepted that ionic percolation mainly takes place in the amorphous regions of polymer [6,11].

One of the strategies to enhance the ionic conductivity (σ_DC) of binary mixtures of PEO and Li salt is to add a second polymer that is miscible or homogenous with PEO (in the melt or amorphous phase), for example poly(methyl acrylate) [9,13], poly(methyl methacrylate) [14], polyacrylate (PAc) [15], poly(propylene oxide) (PPO) [16], poly(ε-caprolactone) (PCL) [17,18], etc., with the aim to suppress the crystallinity of PEO. However, suppression of crystallinity in PEO alone does not always lead to enhancement of conductivity for the systems as compared to the PEO–salt systems [19]. This is due to highly complex systems that develop when Li salt is added to the miscible or homogenous PEO (binary) blends at room temperature, let say at 25 °C, where the conductivity is measured for useful applications. Therefore, the salt content, blend compositions, the glass transition temperature (T_g) of the second polymer, and homogenous or heterogeneous amorphous phase of the ternary systems will affect the conductivity of the materials.

In this study, high molar mass PEO is blended with an amorphous polymer, i.e., PMA, with the T_g roughly at 10 °C, and with addition of lithium perchlorate (LiClO₄). PEO/PMA blends are known to be homogenous above melting point of PEO (~65 °C) and in the amorphous phase for all blend compositions under the experimental condition [13,20]. Upon cooling from the melt, the PEO starts to crystallize and phase separates from the mixture. Hence, there is co-existence of PEO crystalline phase and the amorphous mixture (amorphous PEO and PMA) at 25 °C (room temperature).

The addition of the ternary component, the Li salt to PEO/PMA blends may trigger phase separation of PEO and PMA in the amorphous phase of the system depending on the composition of the blends and the salt content [13]. Hence, the amorphous phase of the ternary systems can be homogenous or heterogeneous (c.f. text below) depending on the composition of the blends as well as the salt content, which in the end may lead to enhancement or reduction on conductivity as compared to the binary PEO–salt system at a constant mass fraction of salt (W_S) or at constant mass fraction of PEO (W_PEO). It is noteworthy that these polymer electrolyte systems are becoming more complex and complicated in the ternary mixture at 25 °C.

As a result, these composition-dependent ternary systems of PEO/PMA/LiClO₄ will have different morphologies, which lead to different conductivities at constant salt content or constant blend composition. The composition-dependent morphologies of the ternary mixtures in the amorphous region are listed in

Table 1. Brief interpretations on the morphology of the ternary mixtures in the amorphous phase.

Situation	Possible Morphology	System
(I)	Amorphous polyethylene oxide (PEO) incorporated with Li-salt/amorphous polymethyl acrylate (PMA)	PEO/PMA blends with salt content more than ~5 wt.% when PMA is in excess
(II)	Amorphous mixture of PEO and PMA incorporated with Li–salt	PEO/PMA blends with salt content less than ~10 wt.% when PEO is in excess
(III)	Amorphous PEO incorporated with Li–salt/pure Li–salt	PEO with salt content more than ~10 wt.%
(IV)	Amorphous PEO incorporated with Li–salt/amorphous PMA/pure Li–salt	All PEO/PMA blends with salt content more than ~10 wt.%

as an overview.

Situation (I) is based on the existence of two distinct phases (liquid–liquid separation) in the system and situation (II) assume two amorphous phases of PEO and PMA incorporated with Li–salt exist as a single phase. However, another possibility may emerge in the binary and ternary systems with much higher concentration of salt, where the precipitation of Li–salt may be observed out of the polymer matrix when the mixtures are up to saturation, i.e., situations (III) and (IV) [21]. Situation (IV) considers the two amorphous phases of PEO and PMA are no longer homogeneous (miscible) as in (II), along with the presence of pure salt (salt precipitation) after the addition of sufficiently high salt content or at blend composition with minor PEO. Morphology with situation (I) might be effective for conductivity enhancement, as it may increase the distribution of Li–salt into the preferred polymer phase, consequently increasing the polymer–salt interaction. In short, all the morphologies discussed above may play a vital part to the electrical phenomenology of PEO/PMA SPEs. Thus, we note here that frequency-dependent impedance studies are deemed important to elucidate the dielectric and electric relaxation of dipolar entities in the polymer electrolytes as impedance (Z) is one of the central quantities in impedance spectroscopy.

This work is an extension of the previous contribution to the solid polymer electrolytes (SPEs). The scope of the discussion of these ternary PEO/PMA/LiClO₄ systems were expanded with in-depth theoretical analyses and additional experiments as compared to the previous study [13]. Differential scanning calorimetry (DSC) was employed to estimate the properties of the amorphous phase (on the T_g) as well as the crystalline phase (on the melting behavior and its crystallinity) of the ternary systems. Morphology and dielectric response of the systems were investigated by optical microscopy (OM) and electrochemical impedance spectroscopy (EIS), respectively. Besides, polarization and relaxation of dipoles in composition-dependent PEO/PMA blends after addition of Li–salt was studied by EIS and will be discussed.

2. Experimental

2.1. Materials

Characteristics of the polymers and salt are given in

Table 2. Characteristics of materials.

Characteristics	PEO	PMA	LiClO4
Mw(a)/g mol−1	-	-	106.4
Mη(b)/g mol−1	300,000	40,000	-
Tm(c)/°C	65	-	236 (f)
Tg(d)/°C	−53	13	-
ρ(e)/g cm−3	1.21	1.22	2.42
Molecular Structure
Supplier	Sigma Aldrich Chemical Co. (St. Louis, MI, USA)		Acrõs Organics Co. (Geel, Antwerp, Belgium)

^(a) Mass-average molar mass as determined by the supplier. ^(b) Viscosity-average molar mass as determined by the supplier. ^(c) Melting temperature of PEO estimated in this study. ^(d) Glass transition temperature estimated in this study. ^(e) Density as determined by the supplier. ^(f) Melting temperature from reference [22].

. The polymers were purified prior to further preparation. PEO was purified by dissolution in chloroform (Merck, Darmstadt, Germany) following with precipitation in n-hexane (Merck, Darmstadt, Germany). The as-received PMA in toluene was precipitated in n-hexane before blending. The LiClO₄ was dried at 120 °C for at least 24 h.

2.2. Preparation of Samples

Binary blends of PEO/PMA were prepared by using a solution casting technique. Solid film PEO₈₀ denotes PEO/PMA 80/20 blend (m/m) and analogue sample coding for other compositions. Quantity m_i represents mass of component i. The solid solution comprising of PEO, PMA, and LiClO₄ were also prepared by solution casting technique. The mass fraction of polymer (PEO) and mass fraction of salt were estimated as below:

$$W_{\mathrm{PEO}}=\frac{m_{\mathrm{PEO}}}{m_{\mathrm{PEO}}+m_{\mathrm{PMA}}+m_{\mathrm{S}}} \mathrm{and} W_{\mathrm{S}}\frac{m_{\mathrm{S}}}{m_{\mathrm{PEO}}+m_{\mathrm{PMA}}+m_{\mathrm{S}}}$$

(1)

Corresponding W_S to mole fraction of salt (X_S) and salt content (Y_S) is shown in

Table A1. Salt concentration of the systems.

${\mathit{W}}_{\mathbf{S}}=\frac{{\mathit{m}}_{\mathbf{S}}}{{\mathit{m}}_{\mathbf{S}}+{\mathit{m}}_{\mathbf{P}\mathbf{E}\mathbf{O}}+{\mathit{m}}_{\mathbf{P}\mathbf{M}\mathbf{A}}}$	${\mathit{X}}_{\mathbf{S}}=\frac{{\mathit{n}}_{\mathbf{S}}}{{\mathit{n}}_{\mathbf{S}}+{\mathit{n}}_{\mathbf{P}\mathbf{E}\mathbf{O}}+{\mathit{n}}_{\mathbf{P}\mathbf{M}\mathbf{A}}}{ }^{\left(\mathbf{a}\right)}$	${\mathit{Y}}_{\mathbf{S}}=\frac{{\mathit{m}}_{\mathbf{S}}}{{\mathit{m}}_{\mathbf{P}\mathbf{E}\mathbf{O}}+{\mathit{m}}_{\mathbf{P}\mathbf{M}\mathbf{A}}}$
0	0	0
0.0196	0.0082	0.02
0.0476	0.0203	0.05
0.0654	0.0282	0.07
0.0909	0.0398	0.10
0.107	0.0473	0.12
0.130	0.0585	0.15

^(a)n = number of moles.

in Appendix A. All the components were dissolved in acetonitrile (ACN) and stirred for 24 h at 50 °C. The solution was casted onto Teflon^® dish and left to dry at room temperature until the solvent evaporated. The samples were oven dried at 50 °C for no less than 24 h for removal of residual solvent. This was followed by heating the samples at 80 °C for ½ h under nitrogen atmosphere. Subsequently, the samples were isothermally crystallized at 25 °C for 24 h in the convection oven before vacuum drying at 25 °C for 24 h. All samples were then kept in desiccators at 25 °C. Then, the samples were again vacuum dried at 25 °C for 24 h prior to any characterization.

The solution-cast samples were thermally treated under inert atmosphere above the melting temperature (T_m) of PEO for a certain period of time to erase the thermal history of the sample during the sample preparation pathway as well as for complete mixing of all components in the melt. The consistency in sample preparation is crucial for the reproducible properties reported here. Hence, the electrolyte systems discussed herein are close towards equilibrium condition.

3. Characterization

3.1. Differential Scanning Calorimetry (DSC)

The quantities T_g, T_m, and melting enthalpy (ΔH_m) of samples were analyzed from the heating cycle of DSC. The samples were studied using DSC TA Q200 (TA Instrument, New Castle, DE, USA) equipped with RCS90 cooling system (TA Instrument, New Castle, DE, USA). Nitrogen gas was purged during analysis at a rate of 50 mL min⁻¹ to avoid thermo-oxidative degradation of samples. Roughly 10 to 15 mg of thin film samples were encapsulated in aluminum DSC sample pans for analysis. Calibration of DSC using high-purity indium standard was done prior to analysis. Samples were heated up from −90 °C to 80 °C at a heating rate of 10 °C min⁻¹. The T_g was estimated at half extrapolated change in heat capacity (ΔC_p) or adopted from Moynihan’s approach (if relaxation endotherm overlaid the glass transition) for estimation of T_g in a more precise manner [21]. T_m was extracted as the maximum of the endothermal peak in the DSC trace. ΔH_m was estimated from the area underneath the melting endotherm. The crystallinity (X^*) of the PEO phase in the blends was estimated by Equation (2).

$$X^{*}=\left(\frac{∆H_{\mathrm{m}}}{∆H_{\mathrm{ref}} · W_{\mathrm{PEO}}}\right) \times 100\%$$

(2)

where ΔH_ref = 188.3 J g⁻¹ is the melting enthalpy of 100% crystalline PEO, and W_PEO is the mass fraction of PEO in the blends.

3.2. Optical Microscopy (OM)

The morphology of each sample was captured using AxioVision Control software (Zeiss, Oberkochen, Germany) connected to the Axioplan 2 imaging polarizing optical microscope (Zeiss, Oberkochen, Germany) equipped with a Linkam TM600/s hotstage (Linkam, Surrey, UK). Three percent (m/v) of the sample was dissolved in acetonitrile and was heated at 50 °C for 24 h. After dissolution, the polymer solution was casted drop-by-drop on top of a glass cover slip and allowed to dry at room temperature for at least 24 h. The sample was heated up at 80 °C and was annealed for ½ h, followed by quenched cooling and isothermal treatment at T_c = 25 °C until complete crystallization. Minimum 5 micrographs were captured with 10× magnification at two different temperatures T = 80 °C (with non-polarized mode) with annealing time of ½ h to give sufficient time for complete mixing in the melt and at T = 25 °C (with polarized mode) with annealing time of 24 h for complete crystallization of PEO.

3.3. Impedance Spectroscopy (IS)

Impedance measurement of each sample was done at 25 °C using a Hioki 3532-50 Hi Tester impedance analyzer (Hioki, Chubu, Japan) equipped with a computer for data collection over the frequency range from 50 Hz to 2 MHz. Two stainless steel electrodes with a diameter of 20 mm were used as the current collector and the blocking electrode for the ions. The sample was placed in between the two blocking electrodes for measurement. The σ_DC value was estimated from the bulk resistance (R_b) following equation of σ_DC = L/(A·R_b), where quantities L and A denote thickness of the sample and surface area in touch with the two stainless steel disc electrodes, respectively. The quantity L was measured with Digimatic Caliper (Mitutoyo, Kanagawa, Japan) at three different spots that were in contact to the electrodes. The quantity L was averaged from the three measurements and the thickness of the dried samples is maintained within the range of 0.25–0.35 mm.

σ_DC was either estimated from both Nyquist plot and from the real (Z’) or imaginary parts (Z″) values of impedance at frequency for fully stabilized network ($f_{\max }^{Z^{″}}$) [Z’($f_{\max }^{Z^{″}}$) and Z″($f_{\max }^{Z^{″}}$)] at maximum of Z″, and each result obtained by both methods is maintained with an error of less than or equal to 5%. Values of σ_DC reported here were the averages of three impedance analyses from three different spots of the thin sample with errors of σ_DC approximately at 10%.

4. Results and Discussion

4.1. Glass Transition Temperature

Glass transition temperature (T_g) may be seen as one criterion of the miscibility of a polymer blend from a macroscopic point of view. The blend is seen as a single-phase polymer blend (miscible blend) when it exhibits a single and composition-dependent T_g. Immiscibility can be concluded when two T_gs that are approximately close to the T_g of parent polymers are observed. The Fox equation as in Equation (3) is commonly referred to for the evaluation or prediction of the T_g values of miscible binary blends.

$$\frac{1}{T_{\mathrm{g}}}=\frac{W_{\mathrm{A}}}{T_{\mathrm{g}, \mathrm{A}}}+\frac{W_{\mathrm{B}}}{T_{\mathrm{g}, \mathrm{B}}}$$

(3)

where W_A, T_g,_A and W_B, T_g,_B represent the mass fraction and T_g of the respective polymer. The dashed curve in Figure 1b is calculated after Equation (3). The entire composition of semi-crystalline/amorphous PEO/PMA blends in this study exhibit only a single and composition-dependent T_g. The heating cycle of DSC thermograms are displayed in Figure 1a. Moreover, the experimental T_g values are in good agreement with the T_gs predicted after the Fox equation. This may suggest that the polymer pair is miscible and homogeneous in the amorphous phase as well as in the molten state under this experimental condition for all compositions.

Figure 2a displays that the T_g of PMA decreases with elevating salt content (W_S). Unlike PMA, T_g of neat PEO elevates with increasing W_S. This phenomenon implies that the salt may be more soluble in PEO than PMA. It is interesting to note that the miscible PEO/PMA blends act differently with the addition of salt as shown in Figure 2. The blends with W_PEO ≥ 0.7 after the addition of salt W_S still show single-composition-dependent T_g for all studied salt concentrations (W_S = 0–0.17), which may indicate miscibility of the ternary mixtures in the amorphous phase and in the molten state. PEO₈₀ displays the highest T_g values at W_S = 0–0.09 as compared to other systems W_PEO ≥ 0.7 with the same amount of salt. It implies PEO₈₀ may be an effective host for LiClO₄ for the enhancement of the ionic conductivity. Normally, the increase in T_g indicates the stiffness of the polymer chains at an increasing temperature. In this case, the stiffness of the chain segments comes from the interaction of salt molecules and polymer chains. However, the increase in T_g is only limited in the range of low salt content, as it increases up to saturation of the mixture due to the solubility limit of salt in the respective polymers [23,24]. When W_PEO ≤ 0.6 at W_S ≥ 0.05, the heterogeneity in the melt is inferred (i.e., liquid–liquid phase separation). It is deduced by the presence of two T_gs that correspond to the parent polymers. From Figure 2b, another observation on the salt localization in the PEO phase more than PMA can be seen. In the immiscible systems, the T_gs of PEO are constantly observed at slightly higher values than the neat PEO, whereas the T_gs of PMA lie closely to that of the neat PMA.

These T_g findings are in good agreement with the descriptions in the Introduction. Macroscopically, we observe situation (I) in the system with W_PEO ≤ 0.6 at W_S ≥ 0.07, and situation (II) in the system of W_PEO ≥ 0.7 at W_S ≤ 0.10. One may also observe the morphology of situation (III) in binary PEO–salt systems (W_PEO = 1) at W_S ≥ 0.10 or beyond the salt saturation of the mixture. This is confirmed by the T_g of the respective mixture, which is closed to T_g of neat PEO. Furthermore, the precipitation of Li–salt was observed from optical inspection. The mixture with morphology (III) can be differentiated into salt-rich or salt-poor phases [21]. The salt-rich phase (phase″) is mainly the pure salt phase, and salt-poor phase (phase’) is the phase of highly diluted salt solution in the polymer phase. The salt content of salt-poor phase (W_S’) can be elucidated instantly from the plot of T_g against W_S. Hence, the preferential localization of salt in the different polymer phases of the immiscible ternary mixture can be detected based on their respective T_g values. For instance, the PEO₅₀ system added with W_S at 0.091 (black arrow in Figure 2b) is seen to contain approximately 1 wt.% of Li–salt dispersed in the PEO phase as the T_g of PEO in PEO₅₀ is equivalent to T_g of PEO at W_S = 0.01 (black arrow in Figure 2a).

These ternary mixtures are miscible in the molten state and in the amorphous state under this experimental condition at sufficiently low salt content for all blend compositions. The mixtures are also miscible in molten state and in the amorphous phase at high salt content only when the PEO content is in excess, W_PEO ≥ 0.7. The liquid–liquid phase separation in the molten state may be deduced when the PEO content is W_PEO ≤ 0.6 at high salt content, W_S ≥ 0.05. The mixtures under discussion are comprised semi-crystalline/amorphous PEO/PMA SPEs. Hence, at 25 °C, the SPEs (which is true either for miscible or liquid–liquid phase-separated systems in the melt) are seen as liquid–solid phase separated systems due to the crystallization of PEO, which takes place upon cooling from the melt. The morphological studies of these SPEs that will be discussed in the subsequent section were carried out at temperatures T = 80 °C and T = 25 °C using optical microscopy (OM) with the aim of elucidating the phase behavior of the mixture in the melt and in the liquid–solid form from the microscopic level aside from distinguishing the correlation between T_g and morphology. However, the correlation between the electrical properties and miscibility will not be attempted because only liquid–solid phase-separated systems at 25 °C were studied in all cases. Instead, preferential localization of salt in different polymer phases and the percolation path of the systems will be correlated to the relaxation and conductance of the systems.

From Figure 2a, we may also discuss the variation of T_g of the polymer blends that are still miscible in the molten and amorphous state after the addition of salt, for instance, W_PEO ≥ 0.7, in terms of thermodynamics based on the empirical observation [21]. The quantity of T_g, and its ΔC_P can be correlated as follows

$$\ln \left(\frac{T_{\mathrm{g}}}{T_{\mathrm{g}}^{\mathrm{o}}}\right)= \frac{-∆S^{ \mathrm{glass}}}{∆C_{\mathrm{P}}}$$

(4)

ΔS^glass denotes the change in entropy of the glassy state after addition of salts that indirectly describes the deviation of the glass from the equilibrium. Equation (4a) illustrates the ΔC_P serves as the factor of proportionality of –ΔS^glass and T_g/T_g^o. It is also true in the small range of T_g^o to T_g that ΔC_P is assumed to be independent of temperature (this observation is only applicable for thermodynamically miscible system, i.e., PEO–PEO₇₀ and PMA). Approximately, the quantity of T_g/T_g^o may also be expressed as below

$$\frac{T_{\mathrm{g}}}{T_{\mathrm{g}}^{\mathrm{o}}}=1+\frac{1}{T_{\mathrm{g}}^{\mathrm{o}}}\left(\frac{\partial T_{\mathrm{g}}}{\partial W_{\mathrm{S}}}\right)W_{\mathrm{S}}$$

(5)

where the slope of plot T_g against W_S reads, $\frac{1}{T_{\mathrm{g}}^{\mathrm{o}}}\left(\frac{\partial T_{\mathrm{g}}}{\partial W_{\mathrm{S}}}\right)$ ≡ Γ. The relationship of Equations (4) and (5) explains that the slope Γ is interconnected to the partial molar entropy of the polymer pair in the ternary mixture that can be defined as $-∆\tilde{S}\equiv \frac{\partial ∆S^{\mathrm{glass}}}{\partial W_{\mathrm{S}}}\equiv \Gamma \overline{∆C_{\mathrm{P}}}$. The mean value of ΔC_P of the polymer pair with different salt content is employed for estimation of $\mathsf{\Delta }\tilde{S}$ on condition that the ternary mixtures lies in the linearity. Furthermore, the glass transition is not accompanied by the change in energy; hence, we may formulate the related chemical potential or molar partial Gibbs free energy by

$$∆G= -T_{\mathrm{g}}^{\mathrm{o}}∆\tilde{S}$$

(6)

The relationship of (4) and (5) shows when the slope Γ is positive; the reduced T_g/T_g^o of the mixture should be larger than unity and ΔS^glass would be negative. The positive gradient, Γ > 0 reflects the freezing in of degrees of freedom of the polymer chains, whereas a negative gradient, Γ < 0 shows the increase in the chain’s degrees of freedom of the polymer. Experimentally, we observe that the salt affects the respective binary polymer-salt mixtures differently. PEO has a positive slope, Γ > 0 with increasing W_S, whereas PMA has a negative slope, Γ < 0 with increasing W_S. It suggests that the addition of salt in PEO leads to a higher stiffness of polymer chains whereas salt in PMA increases the flexibility of polymer chains. This discussion is also relevant for miscible ternary mixtures of PEO/PMA/LiClO₄. The gradient Γ of W_PEO ≥ 0.7 (i.e., PEO₉₀ and PEO₇₀) is reduced gradually with the increasing content of PMA. This indicates that addition of PMA in the PEO–salt systems increases the degree of freedom of the polymer chains in the miscible polymer blends (increasing the chain flexibility). Besides, it is worth noting that addition of certain amount of PMA into the PEO–salt system also drives the systems closer to equilibrium, as reflected in the ΔG. The corresponding parameters Γ of miscible PEO/PMA/LiClO₄ mixtures are summarized in Table 2.

4.2. Crystallinity

The degree of crystallinity (X*) of PEO in the ternary mixtures was estimated from Equation (2). The melting enthalpy (ΔH_m) of PEO was obtained from the area under the melting endotherm of the thermograms. ΔH_m is an essential quantity for the estimation of the degree of crystallinity (X^*) for semi-crystalline polymer as the extent of the liquid–solid phase separation can be evaluated from crystallinity. The degree of crystallinity allows for rough estimation of the amorphous content. Hence, in principle, the reduction in crystallinity of the crystalline phase leads to an increase in the amorphous phase. It is widely accepted that ion percolation (conductivity) takes place only in the amorphous phase, thus, a rough estimation of the composition of the amorphous phase might be useful to understand electric conductivity. Equation (2) describes the degree of X^* of PEO, which is directly proportional to ΔH_m of PEO, if the enthalpy of 100% crystallinity of PEO serves as the factor of proportionality

X^* ∝ ΔH_m

(7)

Figure 3 depicts the X^* of PEO in the ternary mixtures as a function of W_PEO. The dashed curve in the plot of X^* against W_PEO is the constancy curve of crystallinity of PEO in the ternary mixtures. The suppression of X^* from the constancy curve may be caused by the entrapment of the other component in the crystalline phase (in this case, PEO crystalline). Figure 3 illustrates that the X^* of salt-free high molar mass PEO is 70% and starts to gradually level off to ~20% with increasing of salt content up to W_S = 0.167 for binary PEO–salt system. The deviation from the PEO crystallinity constancy curve is observed for all systems with different blend compositions and salt content.

When PEO crystallinity in both salt-free and salt-added systems is at low salt content, W_S ≤ 0.05 displays no significant difference. The further suppression of X^* of PEO is observed with higher content of salt, W_S ≥ 0.07, which is true for all blends under discussion. This indicates that with small amounts of salt in the mixture, the PEO crystallinity of the semi-crystalline/amorphous mixture is not really affected by the presence of salt and the further suppression of X^* only starts with increasing amounts of salt in the mixture. These observations are true for both miscible and immiscible systems of this study. The suppression of crystallinity is normally true for miscible systems, as the crystallization behavior of the crystalline component in the immiscible systems (or liquid–liquid phase separated systems in the melt) is expected to have close approximation to the neat crystalline polymer, and it is normally unaffected by the presence of other components. Nevertheless, from this study, we note that salt and PMA are fond of PEO. Hence, we may see the exclusion of salt and PMA from PEO spherulites is not completely perfect, and the entrapment of amorphous phase of PEO, PMA, and salt in PEO spherulites lead to suppression of PEO crystallinity as well as depression of apparent melting temperature of PEO in the systems, which will be discussed in the next section.

4.3. Melting Behavior

The apparent melting temperature (T_m) was obtained from the maximum of the melting endotherm of the DSC curve. Figure 4 depicts the apparent T_m of PEO in the ternary mixtures as a function of the mole fraction of salt (X_S). As mentioned before, the decrease in the crystallinity of the crystalline phase is generally influenced by the miscibility of the mixture, which is also true for the apparent T_m. The depression of the apparent T_m of PEO is expected in miscible systems. We note here, both melting point and enthalpy are greatly depending on the amount of salt content. The apparent T_ms of PEO of the as-prepared samples are depressed significantly from the constancy curve with increasing salt content for all compositions. This implies that the dissolution of Li–salt in the polymer matrix (in PEO) increases, which is in good approximation to the T_g as well as the crystallinity findings.

For binary polymer–salt mixtures with a crystalline constituent, we may employ the melting point depression of the crystalline polymer for estimation of deviations from perfect behavior [25,26]. In this case, when binary PEO/Li mixtures are completely miscible in the molten state, we employ the melting point depression of PEO for evaluation of the solid solution from perfect behavior (i.e., PEO–PEO₇₀ at W_S ≤ 0.10). The estimation is expressed by

$$T_{\mathrm{m}}=T_{\mathrm{m}}^{\mathrm{o}}+\frac{R{(T_{\mathrm{m}}^{\mathrm{o}})}^2}{∆H_{\mathrm{ref}}}\left(\frac{\ln {\gamma }_{\mathrm{P}}}{{\ln X}_{\mathrm{P}}}+1\right){\ln X}_{\mathrm{P}}$$

(8)

where X_P denotes the mole fraction of PEO in the molten state, and γ_P is the corresponding activity coefficient for the solution, whereas the quantities T_m^o and ∆H_ref symbolize the melting temperature of neat PEO and the melting enthalpy of 100% crystallinity of PEO, respectively. For systems with low salt content (or the completely miscible systems, PEO–PEO₇₀ at W_S ≤ 0.10), we may assume ln X_P = −X_S. Hence, we see from Equation (8), the information about the activity coefficient γ_P may be determined instantly from the linear plot of T_m against X_S. This relationship is valid as long as the PEO crystalline phase crystallizes out from the molten mixture upon cooling from the melt. We apply Equation (8) to the as-prepared PEO/Li mixture for estimation of deviations from the ideal behavior (c.f. red-dashed curve in Figure 4). From Figure 4, we obtained ∆T/K = 147.3 X_S for the PEO/Li mixture. The deviation of melting point depression from perfect behavior [∆T(X_S)] for the PEO/Li mixture, specifically at higher salt content (X_S > 0.05), is observed. This indicates at low salt contents that the melting point depression is not influenced by the salt, and it shows to a good approximation perfect behavior, displaying a very small γ_P = 1 + 0.30 X_S. This observation is also in good agreement with the findings discussed in references [26,27], where the estimation of the melting point depression of the high molar mass PEO (M_η = 3 × 10⁵ g mol⁻¹) against X_S under similar experimental conditions yields γ_P close to unity (1 + 0.38 X_S) as reported in reference [27]. This indicates that the PEO/Li system under these experimental conditions behaves nearly perfectly. Similar trend observations are found for other miscible mixtures, such as PEO₉₀, PEO₈₀, and PEO₇₀, where drastic deviations only occur for systems with higher content of salt (X_S > 0.05). This implies that all systems that are still miscible with low salt contents under these experimental conditions behave nearly perfect as in binary PEO/Li systems. The results retrieved from Equation (8) are summarized in

Table 3. The characteristic quantities obtained from Equations (5) and (8) for PEO/PMA/LiClO₄ systems.

Systems	Tgo/K	Γ Retrieved from Equation (5)	${\overline{{∆\mathit{C}}_{\mathbf{P}} }}^{\left(\mathbf{a}\right)}/$ J mol−1 K−1	ΔG/ kJ mol−1	Tmo/K	Melting Point Depression Retrieved from Equation (8) [Correlation]	Activity Coefficient (γp)
PEO	220	1.20	8.9 ± 2.6	2.4	338	∆T/K = 147.3 XS [0.999]	1 + 0.30 XS
PEO90	223	1.15	8.2 ± 2.5	2.1	337	∆T/K = 123.7 XS [0.999]	1 + 0.37 XS
PEO80	231	1.27	9.2 ± 2.4	2.7	336	∆T/K = 163.7 XS [0.999]	1 + 0.45 XS
PEO70	233	0.21	8.1 ± 1.7	0.4	335	∆T/K = 91.8 XS [0.999]	1 + 0.19 XS
PEO60	237	-	-	-	335	-	-
PEO50	246	-	-	-	334	-	-
PEO40	258	-	-	-	333	-	-
PEO20	273	-	-	-	331	-	-
PMA	286	−0.72	43.4 ± 1.1	−8.9	-	-	-

^(a) Mean value of ΔC_P of the polymer pair in the linear range.

.

4.4. Optical Microscopy

From T_g measurements, we note the existence of single T_gs for binary PEO/PMA blends, which is also true for ternary PEO/PMA/LiClO₄ mixtures at sufficiently small salt concentration under described experimental conditions. The heterogeneity (immiscibility) is observed only when the mass fraction of PEO is minor, W_PEO ≤ 0.6, and the mass fraction of salt is high, W_S ≥ 0.05. The optical micrographs of selected ternary PEO/PMA/LiClO₄ systems taken at 10× magnification is shown in

Table 4. Micrographs of PEO/PMA/LiClO₄ systems at temperature of 80 °C after 30 min and 25 °C after 24 h of isothermal annealing. Micrographs were taken with 10× magnification. The scale bar at each micrograph corresponds to 100 μm.

System		80 °C	25 °C
PEO	WS = 0
	WS = 0.0196
	WS = 0.091
PEO80	WS = 0
	WS = 0.0196
	WS = 0.091
PEO50	WS = 0
	WS = 0.0654
	WS = 0.091

. The micrographs were captured at two different temperatures, i.e., 80 °C after 30 min (in the molten state) and 25 °C after 24 h (in the solid state). As expected, the binary PEO/PMA blends in the molten state (T = 80 °C) display only a single-phase without any visible boundary with polymer interfaces. This observation is also true for ternary mixtures with small concentration of salt (i.e., systems of PEO, PEO₈₀, and PEO₅₀ without the presence of salt). The two-phase structure occurs when the mass fraction of PEO is minor, W_PEO ≤ 0.6, and mass fraction salt is high, W_S ≥ 0.05. It is supported by the micrographs of PEO₅₀ after addition of W_S ≥ 0.05, where a two-phase structure with clear boundaries at the polymer interface is observed. PEO₅₀ with W_S = 0.065 displays a matrix-droplet morphology, and it starts to transform into co-continuous morphology with addition of higher salt content at W_S = 0.091. This observation on PEO₅₀ with salt may be associated to the situation (I) and (IV) as described in the introduction, where the liquid–liquid phase separation is inferred in the melt and amorphous phase of the semi-crystalline/amorphous PEO/PMA blends after addition of higher content of salt (this is in good agreement to the T_g results; system with W_PEO ≤ 0.6 at W_S ≥ 0.05, which possess two T_gs that correspond to PEO and PMA).

At 25 °C, where PEO is crystalline below the melting point T_m = ~65 °C, the samples are placed between crossed polarizers of the optical microscope. Using polarized light, the neat PEO displays large spherulites with a fine fibrillar texture and a clear Maltese cross. As expected, the number of nucleation sites of PEO spherulites for PEO, PEO₈₀, and PEO₅₀ increase with and without salt due to the exclusion of PMA and Li–salt from the PEO crystalline phase during crystallization This is also reflected by the T_m results, where the melting point depression of PEO is observed with increasing salt content. This is due the fact that PMA and Li–salt most likely have attractive interactions with PEO phases, which eventually lead to the incomplete exclusion. Coarsening and irregularity of PEO spherulite fibrils with blurred grain boundary are clearly noted for all systems at high salt content. There are several spherical dark spots of PMA within the continuous phase of PEO spherulites for PEO₅₀ at W_S = 0.065 and 0.091.

The proposed schematic diagrams of PEO₅₀ systems at W_S = 0 and 0.091 in the molten and solid state based on the T_g results and polarized optical microscopy are shown in Figure 5. We conclude that both systems possess different morphologies molten at 80 °C or isothermally crystallized at 25 °C. Immiscibility caused by liquid–liquid phase separation (at 80 °C) (c.f. Figure 5c) and liquid–solid phase separation in PEO₅₀ are observed while cooling down from 80 °C to 25 °C (c.f. Figure 5d) after addition of salt W_S = 0.091. Figure 5 was assigned with the possible morphologies from Table 1.

4.5. Impedance Spectroscopy

The impedance spectra of all systems were measured at 25 °C. Selected binary systems of PEO, PMA, and its blends, such as PEO₉₀, PEO₈₀, PEO₇₀, PEO₆₀ and PEO₅₀, added with different salt content will be discussed under this section, where the morphologies of these systems are correlated to situation (I), (II), (III), or (IV). It is widely accepted that the percolation network only bounds in the amorphous domain of salt-added systems. This phase is considered as not in thermodynamic equilibrium [28,29,30,31]. It may be adequate to discuss the dielectric relaxation process in this domain as a fluctuation–dissipation process [32]. Hence, in the following, the development of relaxation or polarization of dipoles are discussed from the phenomenological point of view.

We observe here that the real part (Z’) and imaginary part (Z″) of impedance reflect to Ohmic resistance and non-Ohmic resistance, respectively. The quantity Z″ displays the characteristic frequencies mainly for dielectric (or dipole) relaxation resulting from local motions of the charged entities and electrode polarization developing from the accumulation of charged entities at the electrode–electrolyte interfaces, after being subjected to an external electric field. These characteristic frequencies are noted as $f_{\max }^{Z^{″}}$ and $f_{\min }^{Z^{″}}$, respectively. Besides, the intersection between Z’ and Z″ that is noted as $f_{\mathrm{cross}}^{Z^{\prime }-Z^{″}}$ indicates to the development of the percolation charge entities. We recognize the characteristic frequencies as the average of relaxation time constant (ω⁻¹). A system with one relaxation time (ω⁻¹) is recognized as Debye relaxation [33]. In general, these phenomena can be observed in a system with highly resistive behavior. A system with high capacitive behavior normally behaves contrarily. Apparently, the impedance plot Z″ = Z″ (Z’) in the Z’-Z″ plane results in semicircle with a radius of R_b/2. Hence, the plot of Z″ vs. frequency (f) as displayed in Figure 6 allows for the determination of the bulk resistance (R_b), with |Z^″_max| = R_b/2. This R_b quantity is defined as inversely proportional to the DC conductivity (R_b ∝ 1/σ_DC). Hence, investigation of the relaxation process in terms of impedance from phenomenological perspective may implicitly elucidate the dielectric and electrical behavior of the system.

Figure 6 represents the impedance spectra, Z’ (f) and Z″ (f), of PEO at low salt content W_S = 0.0196. We observe the existence of the three characteristic frequencies as mentioned before in the impedance spectra of the system. The $f_{\max }^{Z^{″}}$ is noted as the main dipole relaxation, which originates from the alignment of dipoles resulting from the interplay of the sample’s resistance and capacitance in the fully stabilized network. The $f_{\min }^{Z^{″}}$ is noted as the electrode polarization (or called a double-layer capacitance due to the accumulation of charges at the interface of electrode and electrolyte) especially at low frequency. This observation frequently leads to dispersion of relaxation times (i.e., $f_{\max }^{Z^{″}}$ < $f_{\mathrm{cross}}^{Z^{\prime }-Z^{″}}$) and normally becomes severe in the system with high salt content because of the inhomogeneity in the system (c.f. red triangle markers for PEO–salt in Figure 7). However, the distance of $f_{\max }^{Z^{″}}$ and $f_{\mathrm{cross}}^{Z^{\prime }-Z^{″}}$ for PEO at W_S = 0.0196 is small which may indicate that the system is close to Debye relaxation. We suggest that the orientation of dipolar entities in PEO–salt is dominantly restricted to short-range motion (local motion) and insignificant long-range motion. The PMA–salt has no $f_{\max }^{Z^{″}}$ at low salt content and one broad $f_{\max }^{Z^{″}}$ is observed in Z″ spectrum of system with high salt content at frequency. This indicates no dipolar entities in PMA with low salt content and only appears after adding higher content of salt under the experimental condition.

Figure 7 displays the impedance spectra of PEO/PMA mixtures at different W_S. PEO, PMA, and its blends with salt are above the glassy state of parent polymers (c.f. glass transition temperature). R_b values clearly decrease with increasing salt content for all systems and this indirectly infers the increasing DC conductivity. For system with PEO in excess, under W_S = const., either for low or high content of salt, the R_b decreases slightly as compared to the binary PEO/Li salt system. This may indicate that the system with PEO in excess behaves similar to PEO especially at low salt content, for instance, PEO₈₀ at W_S = 0.0196 (low content of salt). A large distribution of relaxation times is observed for PEO and PEO₈₀ with high salt content W_S = 0.0909 as well as for PEO₅₀ at W_S = 0.0196–0.0909 and PMA at W_S = 0.0909, based on the observation of the characteristic frequencies $f_{\max }^{Z^{″}}$ < $f_{\mathrm{cross}}^{Z^{\prime }-Z^{″}}$ at the respective compositions. Additionally, the presence of $f_{\min }^{Z^{″}}$, which points towards electrode polarization, may also contributes to the distribution of relaxation times of a system (i.e., PEO–PEO₅₀ at W_S ≥ 0.0196 in Figure 7). This deviation may indicate to non-Debye response that normally occurs because of the increase in inhomogeneity in the system. We may summarize that, at high salt content, the distribution of characteristic frequencies is always denoted as

$$f_{\max }^{Z^{″}}<f_{\mathrm{cross}}^{Z^{\prime }-Z″ }$$

(9)

whereas at low salt content, it is seen as

$$f_{\max }^{Z″}\approx f_{\mathrm{cross}}^{Z^{\prime }-Z″}$$

(10)

At W_S = const. especially at high salt content, when PMA is in excess or in symmetric systems (i.e., W_PEO ≤ 0.5) R_b and characteristic frequencies shift to higher and lower values, respectively, as compared to binary PEO/Li salt blends or system with PEO in excess. It is noted that these systems are seen as liquid–liquid phase separated as reflected in the T_g results of the same system, where situations (I) and (IV) are correlated. This behavior may indicate preferential localization of salt in the different polymer phases of the immiscible ternary mixture. For instance, based on the T_g results, we may assume that the PEO phase in PEO₅₀ at W_S = 0.091 contains less salt as compared to the binary PEO/Li salt system at the same composition. This phenomenon might be one of the reasons for the enhancement of the bulk resistance as well as large deviation from Debye response in the liquid–liquid phase separated systems.

Further insight, Figure 7 illustrates that impedance spectra of PEO/PMA blends comprise low and high concentration of salt, respectively. Interesting points of the impedance spectra are suggested as follows. Blends with PEO in excess show similar behavior of neat PEO at low salt concentration. The amorphous network-like phase of PEO in the blends is almost not influenced by the PMA phase. The PEO amorphous phase contains slightly more salt than in the neat PEO; as a consequence, bulk resistance is slightly reduced. Additionally, we note that $f_{\min }^{Z^{″}},$ $f_{\max }^{Z^{″}} \left(\mathrm{PEO}\right) < f_{\min }^{Z^{″}},$ $f_{\max }^{Z^{″}} \left({\mathrm{PEO}}_{80}\right)$ and $f_{\mathrm{cross}}^{Z^{\prime }-Z^{″}}\left(\mathrm{PEO}\right)<f_{\mathrm{cross}}^{Z^{\prime }-Z^{″}}\left({\mathrm{PEO}}_{80}\right)$. Enhanced content of PMA in blends leads increasingly to higher R_b at Y_S = const. On top of that, $f_{\min }^{Z^{″}},$ $f_{\max }^{Z^{″}}, \mathrm{and}$ $f_{\mathrm{cross}}^{Z^{\prime }-Z^{″}}$ shift to lower values. It indicates “effective″ salt content of the amorphous PEO phase lessens more and more. Again, one observes more dispersion of relaxation times, $f_{\min }^{Z^{″}}<f_{\max }^{Z^{″}}<f_{\mathrm{cross}}^{Z^{\prime }-Z″}$. In short, dielectric behavior of PEO and of the blends with PMA is governed by amorphous network-like phases. Addition of Li–salt leads to dissolution of low “effective″ amounts of salt in these phases. Therefore, corresponding polymer–salt mixture can be seen to good approximation as a perfect solution. This PEO-like behavior is maintained when low content PMA is added to PEO at room temperature.

Figure 8 depicts the double-logarithmic plots of conductivity (σ_DC) as a function of salt concentration (Y_S). The salt concentration Y_S is defined as the ratio of mass of salt to the mass of polymer. The double-logarithmic plots of σ_DC vs. Y_S for PEO/PMA/LiClO₄ systems denote a power law dependence resulting from the functional relationship between the two quantities [27,34]. The linear relationship of the double-logarithmic plot of σ_DC vs. Y_S can be read as

$${\sigma }_{\mathrm{DC}}={\sigma }_o{Y}_{\mathrm{S}}^x$$

(11)

which yields the quantity of σ_o and allows for the estimation of the exponent x [27,34]. The exponent x in Equation (11) is the extent of association between salt molecules and the polymer segments. The quantity σ_o in Equation (11) is the simplified form of charge mobility in the system. The full form equation can be read as

$${\sigma }_{\mathrm{DC}}=N_{\mathrm{A}}e\mu \alpha \frac{{\rho }_{\mathrm{P}}}{M_{\mathrm{S}}}{Y}_{\mathrm{S}}^x$$

(12)

where μα is the mobility, N_A is the Avogadro’s constant, e is the elementary charge, ρ_p is the polymer’s density, and M_S represents the molar mass of LiClO₄, where the molecular characteristics for determination of quantity μα are listed in Table 1. The density of each blend was calculated using the additivity rule. All characteristic quantities of the linear regression function for all systems adopted from Equation (12) are summarized in

Table 5. The characteristic quantities obtained from Equation (12) for PEO/PMA/LiClO₄ systems.

Systems	Regression Function/S cm−1	Correlation	X	μα /cm2 V−1 s−1
PEO	σDC = 2.82 × 10−4 YS1.96	0.950	1.96	2.57 × 10−7
PEO90	σDC = 1.10 × 10−4 YS1.87	0.978	1.87	1.01 × 10−7
PEO80	σDC = 4.19 × 10−3 YS2.28	0.944	2.28	3.81 × 10−6
PEO70	σDC = 3.16 × 10−5 YS1.68	0.901	1.68	2.87 × 10−8
PEO60	σDC = 2.79 × 10−6 YS1.58	0.944	1.58	2.54 × 10−9
PEO50	σDC = 1.95 × 10−6 YS1.59	0.957	1.59	1.77 × 10−10
PMA	σDC = 1.57 × 10−6 YS1.59	0.906	1.59	1.43 × 10−9

.

From Figure 8, binary PMA/Li salt systems display the lowest σ_DC as compared to the binary PEO/Li salt blends and in the blends of PEO/PMA systems due to the fact that the μα mobility in PMA is by three orders of magnitude lower than in PEO. Moreover, PEO has stronger association between salt molecules and polymer segments than salt in PMA, as reflected in exponent x = 1.96 for PEO as compared to x = 1.59 for PMA, which might be the reason for lower σ_DC of PMA. This indicates that the salts interact with PMA very weakly, which is in agreement with the T_g findings of the systems. PEO₈₀ at W_S = 0.107 has the highest value of ion conductivity at 1.43 × 10⁻⁵ S cm⁻¹ as compared to the other blend compositions with W_S = const. This is in good agreement where PEO₈₀ also has the highest value of exponent x = 2.28 and μα = 3.81 × 10⁻⁶ cm² V⁻¹ s⁻¹ as compared to the other blend compositions. The blends W_PEO = 0.5 exhibit lower values of μα mobility than PEO but higher than μα mobility of PMA. Hence, we may conclude the ionic conductivities of PEO/PMA/Li blends are dominated by the percolation of PEO in the amorphous phase. Blending small contents of PMA with PEO, when PEO in excess with addition of salt enhances the σ_DC of systems as compared to binary systems of PEO and salt with W_S = const. However, the conductivities of the symmetric system, i.e., PEO₅₀, decrease by two orders of magnitude as compared to PEO–salt systems. This might be caused by the fact that the Li cation is not coordinating well with the chains of both polymers for the phase-separated systems at room temperature. Therefore, we may assume that when PEO with addition of salt is dispersed in PMA domains such as for systems W_PEO ≤ 0.5 where the Li–salt in the PEO phase may not have sufficient percolating paths unlike the systems with PEO in excess.

5. Conclusions

The influence of salt in binary PEO/PMA blends was elucidated using DSC, OM, and EIS. The changes in phase behavior indicated by T_g and morphologies are correlated to the dielectric relaxation using impedance spectroscopy. Based on T_gs, PEO/PMA remains miscible when PEO is in excess after the addition of low contents of salt. However, this behavior is not seen in systems of W_PEO < 0.6 at salt content W_S > 0.05 due to the fact of preferential solubility of salt in the PEO phase, which leads to liquid–liquid phase separation. These findings are well reflected with the results from optical microscopy, where the morphologies of PEO/PMA blends with or without addition of salt can be distinguished in agreement with T_g findings. Subsequently, the evaluation of polarization and dielectric relaxation of dipoles in the composition dependent PEO/PMA blends with salts at room temperature are discussed from the phenomenological point of view. Dielectric relaxation for systems with PEO in excess behaves similar as in PEO after the addition of salt. Main dipole relaxation and bulk resistance of the salt-added systems are shifted to higher frequency and lower impedance, respectively, with increasing salt content. This phenomenon is also true for all systems under discussion. Moreover, the double-logarithmic plot of σ_DC vs. salt concentration manifests the presence of power-law distribution between the quantities. It indicates that the conductivity of the compositional-dependent PEO/PMA blends with salts are governed by the conductivity of PEO, as the power-law distribution of compositions with PEO in excess are higher than in PMA.