Thermal Analysis of Battery Separator Film

Introduction

Lithium ion batteries (LIB) are rapidly becoming the most common source of stored energy for everything from personal electronic devices to electric vehicles and long-term energy storage. A diagram of a battery is shown in Figure 1.

One of the key components of the battery is the porous separator which prevents contact between the anode and cathode and allows transport of the lithium ions during charging and discharging cycles. Some of the requirements for a battery separator include: good electronic insulator, minimal electrolyte resistance, mechanical and dimensional stability, chemical resistance to the electrolyte, ability to prevent migration of colloidal or soluble species between the electrodes, readily wetted by electrolyte, and uniformity in thickness and properties (2). Polyolefin separators made from polypropylene (PP), polyethylene (PE), or laminations of PE and PP are often used for lithium ion batteries with organic electrolytes.

Polyolefin separators are made by wet or dry processes all of which result in forming micropores in the film and in the case of uniaxially stretched films, imparting high tensile strength in the machine direction (MD) and relatively weak properties in the transverse direction (TD). Biaxially stretched films made from β-nucleated isotactic PP and the wet process result in films with comparable properties in both directions. The advantages and disadvantages of the processes are extensively discussed in the literature (2) (3) (4).

The purpose for this note is to detail the basic thermal analysis and mechanical techniques used to characterize a typical separator made from PP.

Experimental

Sample – Celgard 2400 polypropylene separator, 60 mm x 10 mm x 25 μm

Results and Discussions

Thermogravimetric Analysis

TGA results are shown in Figure 2. Main mass loss is 98.31% with a residue of 1.68% which is the filler. The decomposition point is often taken as the temperature at some arbitrary extent of mass loss – typically less than or equal to 5%. It is important to determine the decomposition temperature for the DSC experiment as thermal transitions above that temperature are indiscernible from decomposition. The temperatures corresponding to various extents of mass loss are shown in Table 6.

Differential Scanning Calorimetry

The DSC transitions are summarized in Table 7.

The relatively high heat of fusion in the first heat Figure 3 may be due to a combination of high isotacticity and change to the polymer chain structure due to machine direction stretching. The heat of fusion drops in the second heat (Figure 5) to a value more typical in a commercial grade polypropylene. The difference in the heat of fusion between the 1st and 2nd heats is 36.0 J/g.

The second heat also shows evidence of β-spherulites which melt at ~ 149 °C. The relative fractions of the melting endotherms were estimated using numerical analysis software (5). The relatively low fraction of β-spherulites (Figure 6, Table 8) likely indicates the resin is sensitive to beta formation dependent on cooling rate and in this case does not contain a beta nucleator. Formation of pores by biaxially stretching beta nucleated isotactic polypropylene is well known (4). The observed glass transitions are typical for polypropylene homopolymer.

Thermomechanical Analysis

Figure 7 shows the TMA analysis of the battery separator in the machine direction.

The parameters in Table 9 were determined using the guidelines proposed in a NASA document on procedures for evaluating lithium ion battery separators (1). The subjective nature of the guidelines may be mitigated to some degree by examining the derivative of the strain data with respect to temperature (Figure 8). An alternative method to choose the shrinkage onset temperature may be determined by plotting the derivative of strain with respect to temperature and choosing the temperature corresponding to 0% strain in the vicinity where contraction begins using the TRIOS software. The deformation temperature may be determined by using the onset tool in the derivative by extrapolating a tangent line drawn from the lower temperature range to one drawn from the deflection as deformation accelerates as demonstrated in Figure 10. This results in slightly different values obtained in Figure 7 but may improve laboratory precision (Table 10). The rupture temperature is determined as the minimum from the strain percent as function of temperature data shown in Figure 7.

One of the most common uses of DMA is to determine viscoelastic properties of materials. This is accomplished by applying an oscillatory force or stress (σ) and measuring the displacement or strain (ε). For purely elastic solids (Hookean solid), the strain is in perfect phase or the phase angle delta (δ) is zero. For a purely viscous fluid (Newtonian liquid) δ = 90°. Viscoelastic polymers will have phase angles somewhere in between. In simplistic terms, modulus (E*) describes a materials resistance to deformation, and can be resolved into storage modulus (E’) which is the elastic component or stored energy, and loss modulus (E”) which is the fluid component and dissipated as heat. In mathematical terms they expressed as:

E*=σ/ε (Complex Modulus)
E’=E* cos⁡δ (Storage Modulus)
E”=E* sin⁡δ (Loss Modulus) tan⁡δ=E”/E’

An important parameter determined in a DMA viscoelastic experiment is the glass transition temperature (TG) above which the material will be less rigid and behave more rubbery and below which will be more rigid. The glass transition is often reported as the peak of the loss modulus or the peak of tan delta and can vary based on the technique used to measure it, so the method of determining the TG should be reported. The glass transition in the separator film is 8.9 °C (peak of tan delta) as shown in Figure 13. The number of and temperature of the glass transition(s) can also yield valuable information regarding polymer type for example, homopolymer or copolymer (random or block) or physical blend particularly in the case of polypropylene.