A phase diagram was generated for di-alkyl phthalates, which are liquid–liquid phase-separation systems, using compounds in which the chain length of the alkyl group was changed. These results are shown in Figure 2.
Figure 2Phase diagram of PE-dialkyl phthalate.2 Lines indicate the following dialkyl phthalates: —, dibutyl phthalate (DBP); - - - - , dioctyl phthalate (DOP); . . . . , diisodecyl phthalate (DIDP); — · —, ditridecyl phthalate (DTDP).
Full size image
The chain length of the alkyl group of the di-alkyl phthalate increased from butyl (DBP) to octyl (DOP), isodecyl (diisodecyl phthalate) and tridecyl (ditridecyl phthalate) as the compatibility with PE increased. With increasing chain length, the binodal line of phase separation decreased. This decrease was about 80 °C for tridecyl relative to butyl. Moreover, the monotectic point shifted to the side of the lower PE concentration. The PE concentration of tridecyl was about 30% less than that of butyl.
PE liquid–liquid phase separation and solid–liquid phase separation membranes were fabricated, and the structure of the membranes was observed. PE with a molecular weight of 280 000 was dissolved at 250 °C in liquid paraffin or in DOP as a solvent at a PE concentration of 35 wt%. This solution was rapidly cooled by immersion in ice water and allowed to stand at room temperature for slow cooling. Cooling the solution produced a phase-separation membrane. A cross-section of the phase-separation membrane from which the solvent was extracted was observed using an electron microscope. The results at 6000 × are shown in the upper part of Figure 3.
Figure 3Scanning electron micrograph of a cross-section of the PE phase-separation membrane.2 Slow cooling rate: 15 °C min−1. (230 °C → 100 °C); rapid cooling rate: 50 °C min−1 (230 °C → 100 °C); Mv of PE: 2.8 × 105; concentration of PE: 35 wt%.
Full size image
After solid–liquid phase separation, cell structures indicating pore formation were rarely observed, but these cell structures were abundant after liquid–liquid phase separation. These results were similar to the results of the polypropylene-solvent systems reported by Lloyd et al.15, 16 These solvents were evaluated to create the phase diagram in Figure 2. Phase-separation membranes were fabricated by rapid cooling and slow cooling, and cross-sections were observed by an electron microscope. The results are shown in the lower part of Figure 3. In all of the slow cooling systems and in the systems that contain solvents with a low compatibility with PE, only cell structures were present. In the systems that contained solvents with a good compatibility with PE, structures other than cell structures were present in the rapidly cooled samples. These structures are similar to the surface of the solid–liquid phase-separation membrane. When cell structures were present, the diameter of the cell structures was measured from the electron micrographs in Figure 3. These results are shown in Table 1 .
Table 1 Effects of molecular weight of PE, cooling rate and solvent on cell diameter2Full size table
The following observations were made: (1) As the compatibility of the solvent with PE decreased, the cell diameter increased. (2) The cell diameter was larger after slow cooling than after fast cooling. (3) As the PE molecular weight decreased, the cell diameter increased.
In liquid–liquid phase-separation membranes, cell structures were recognized (Figure 3). Many nonthrough holes may be present. As the compatibility between PE and the solvent increased and the PE molecular weight increased, the cell structures decreased (Figure 3 and Table 1).
No cell structures were recognized in the solid–liquid phase-separation membrane. Instead, very minute pores were seen (Figure 3). The liquid–liquid phase-separation membrane in which the nonthrough holes are dominant is unsuitable for use as a separator, and the pores of the solid–liquid phase-separation membrane were too small to function as the pores of the separator. The phase-separation membrane also has insufficient strength for a separator (described later in Table 2 ). Therefore, extension was performed to change the pore structures and increase the strength.
Table 2 Effects on the preparation process on the properties of membranes made by the phase- separation or extension process.2Full size table
Pressed sheets (about 550 μm) were fabricated from the liquid–liquid phase and solid–liquid phase-separation systems containing PE and a different predetermined solvent. The phase-separation membranes were subjected to two cycles of extension/extraction. Extension was performed without solvent extraction, thereafter solvent extraction was performed. Next, extraction/extension was performed to extract the solvent from the phase-separation membrane, thereafter extension was performed again. The properties of the resulting films were measured. These results are shown in Table 2.
Compared with unextended films (phase-separation membranes), both extension/extraction films and extraction/extension films have high puncture strength per unit thickness, and pore structures through which air could pass were present in the extended films. The air permeability of phase-separation membranes was 10 000 s or more, and they have a much lower air permeability than extended films. The systems with extension/extraction were stronger than the systems with extraction/extension but had a smaller pore diameter. In the systems with extraction/extension, when 5 × 5 times extensions were performed, a break occurred during extension. A stable extension was performed only up to a 3 × 3 times extension. Results for the 3 × 3 times extensions are shown in Table 2. Therefore, systems with extraction/extension (particularly in the liquid–liquid phase-separation system) contain large pores because of the extraction of the solvent-rich phase before extension. Extension was not sufficiently performed, and pore formation based on the pores progressed. A high-ratio extension led to film break because of extreme pore formation. In the other systems with extension/extraction, the extended film was dense, and the solvent was not extracted. Extension was efficiently performed in these systems, and the strength was improved. No pores formed as a result of extension, and the pore diameter was small. The differences between extraction/extension and extension/extraction processes are greater than the differences in the structure of the phase-separation membrane (that is, the difference due to the solvent used or the difference between liquid–liquid phase separation and solid–liquid phase separation) (Table 2).
The relationship between the extension ratio and mechanical strength (puncture strength) of microporous films fabricated using the extension/extraction process was examined. PE with different molecular weights was used, and extension was performed by changing the extension ratio. The relationship between the extension ratio and puncture strength is shown in Figure 4.
Figure 4Effects of molecular weight on the relationship between extension ratio and puncture strength.2 Symbols indicate the molecular weight of PE. •: 2.8 × 105; ▪: 4.5 × 105; ▴: 2.3 × 106.
Full size image
In PE (Mv: 280 000 and 450 000), puncture strength increased proportionally to the extension ratio in the range of 3 × 3 times to 10 × 10 times. A high extension ratio increased the strength. In the extension ratio range from 3 × 3 (nine times) to 10 × 10 (100 times), strength increased with an increase in the molecular weight up to a PE molecular weight of 450 000. The increase in strength was small compared with the increase in the extension ratio in the range of UHMWPE (molecular weight: 2 300 000). For the 100 times extension, the puncture strength decreased in the UHMWPE range compared with that for the PE molecular weights of 280 000 and 450 000. The entanglement of UHMWPE might not be sufficient for the PE/liquid paraffin systems, and alignment by extension was not successful. Thus, no alignment contributed to the strength. Had a method been found in which UHMWPE and liquid paraffin were mixed well without decreasing the molecular weight of the PE, the strength would have improved further.
Seven major conclusions were reached about this manufacturing process. (1) Phase diagrams for PE-solvent systems were generated, and phase-separation temperature data can be arranged by compatibility between the PE and solvents. (2) During liquid–liquid phase separation, cell structures were mainly produced. (3) In solid–liquid phase separation, no cell structures were found. (4) For the extension of phase-separation membranes, the extension effect by extension/extraction was greater than that for extraction/extension. (5) In both extraction/extension and extension/extraction, the pore diameter for the extension of the solid–liquid phase-separation membrane was smaller than that for the extension of the liquid–liquid phase-separation membrane. (6) The difference in the extension process (extraction/extension or extension/extraction) was greater than the difference in the structure of the phase-separation membrane before extension, and the study of the extension was important. (7) As a result of studying the extension/extraction of the solid–liquid phase-separation membrane, there was an improvement in the puncture strength (extension effect) up to an extension ratio of about 100 times.
A phase diagram of a typical PE-solvent system was generated by checking the optical cloud point, but this system contains a large amount of an inorganic powder. Therefore, analysis by this method was impossible. A method for measuring the phase-separation state of a three-component system was devised in this study for the first time. A predetermined amount of PE, an inorganic powder and a solvent were introduced into a Plastomill. The temperature was increased while kneading until the inorganic powder was dispersed, and the PE and the solvent were uniform. Next, the temperature was decreased with continued kneading, and torque was measured. With a decrease in temperature, the uniform phase changes to a liquid–liquid two-phase mixture, and an inflection point in the torque was seen. Usually, in a single polymer system, only a monotonous increase in torque occurs with a decrease in temperature. The temperature at which a decrease in torque starts was defined as the starting phase-separation temperature, and the temperature at which an increase in torque begins to be seen again was defined as the phase-separation ending temperature. Using the method described in the Experimental procedure section, composition was measured (Figure 5) and the sample was heated to a temperature of 250 °C while stirring. After stirring for 5 min, the heating was stopped, and the temperature was decreased with constant stirring. The torque at this time was measured. The measured torque–temperature curves of the three-component and two-component systems are shown in Figure 5.
Figure 5Torque vs temperature of two- and three-component systems.1 (a) A three-component system containing PE (30 wt%), DOP (64 wt%) and silica (6 wt%; surface area of silica: 200 m2 g−1) and (b) a two-component system containing PE (30 wt%) and DOP (70 wt%). Arrows show the starting temperature of phase separation.
Full size image
The starting phase-separation temperature of the three-component system was about 12 °C lower than that of the two-component system. An analysis was performed using the binodal line of the PE concentration and the phase-separation temperature in the phase diagram for the PE-DOP two-component phase separation.2 A decrease of about 12 °C in phase-separation temperature is assumed to correspond to an increase of about 5% in PE concentration (that is, a decrease of about 5% in DOP).
Changes in the starting phase-separation temperature by this torque method depended on the composition ratio of the three-component system. The following seven types of three-component system were examined: (1) a standard composition with a PE/DOP/silica (VN3) weight ratio of 44:41:15, (2 and 3) two types of composition in which the weight ratio of PE/DOP was fixed and the weight of silica was changed by about ±20%, (4 and 5) two types of composition in which the weight ratio of the PE/silica was fixed and the weight of DOP was changed by about ±15%, and (6 and 7) two types of composition in which the weight ratio of the DOP/silica was fixed and the weight of PE was changed by about ±15%. These systems were fabricated, and the starting phase-separation temperature was measured. The relationship between this starting phase-separation temperature and PE/(PE+DOP) in PE/DOP/silica is shown in Figure 6.
Figure 6Starting temperature of phase separation for three-component systems.1 Symbols indicate the composition. •: standard composition; ○: low silica wt%; ▴: high silica wt%; ▪: high PE wt%; *: high or low DOP wt%.
Full size image
Three major results were observed: (1) The starting phase-separation temperature decreased with an increase in silica. (2) The starting phase-separation temperature increased with a decrease in silica. (3) An increase and decrease in the amount of PE and the amount of DOP (without an increase and decrease in the amount of silica) were similar to the typical changes in the PE concentration of PE and DOP. The starting phase-separation temperatures of these cases (a change in the amount of PE and a change in the amount of DOP) were slightly different, and the curves gradually decline to the right, as in the two-component phase-separation curve (Figure 3).
In the phase separation of these three-component systems, the apparent solvent amount is assumed to decrease because the silica absorbs the solvent. The following three results were determined from Figure 6. (1) The phase-separation temperature decreases when the silica concentration increases because the amount of absorbed solvent increases because of increased silica. Therefore, the amount of solvent that substantially involved with PE during the phase separation decreased. As a result, the apparent PE concentration increased, and the phase-separation temperature decreased. (2) A decrease in silica leads to a decrease in the apparent PE concentration and an increase in phase-separation temperature. (3) Changes in the amount of PE and the amount of DOP were similar to the usual change in PE concentration in PE and DOP.
To examine the effect of silica on phase-separation temperature and phase-separation rate, silica samples with different surface areas (130, 200 and 300 m2 g−1; AEROSIL) were used, and the phase-separation state was measured and compared with that of the two-component system. These results are shown in Figure 7.
Figure 7Effects of silica surface area on the phase-separation rate and temperature of a two-component system.1 The heavy line shows the rate of phase separation and the light line shows the phase separation starting temperature. #130, #200 and #300 exhibit a three-component system, and 130, 200 and 300 show the surface area of the silica.
Full size image
Four major observations were made: (1) When the surface area of the silica was changed from 130 to 300 m2 g−1, a slight change in the starting phase-separation temperature (about 175 °C) was seen. In addition, little change was seen in the phase-separation ending temperature (about 145 °C). This measurement was not as accurate as the measurement of the starting temperature. (2) Torque increased with an increase in the surface area of silica over the entire temperature range. (3) The rate of phase separation for silica with a surface area of 300 m2 g−1 was slower than that for silica with a surface area of 130 m2 g−1. (4) The phase-separation rate of the three-component system that contains silica with a surface area of 300 m2 g−1 (#300 torque curve in Figure 7) has a smaller slope and was slower than the phase-separation rate of the two-component system (REF torque curve in Figure 7).
Both the starting phase-separation temperature and the end temperature were the same, regardless of the surface area of the silica, because the solvent was absorbed by the silica rather than adsorbed on the silica surface. The amount of solvent absorbed depended on the weight of silica. Further, the silica absorbed a certain amount of DOP before the start of phase separation. Our assumption from the results in (2) and (3) was as follows: As the surface area of silica increased, the viscosity (torque) of the three-component system increased, and as the viscosity of the system increased, the rate of phase separation became slower, the decrease in torque was slow and the slope was small. However, no significant difference in the phase-separation rate was observed between 130 and 200 m2 g−1 silica, and no clear conclusion was obtained.
When the three-component system (the system with a surface area of 300 m2 g−1) was compared with the two-component system with the same viscosity (the PE/DOP composition ratio of 34.6/65.4), the two-component system had nearly the same viscosity as the three-component system at 230 °C, and the three-component system had a much slower phase-separation rate. The effect of decreasing the phase-separation rate was recognized in silica. The extent of the effect depended on the viscosity of the system. Three-component systems had a slower rate even in the low-viscosity systems that contained additional silica (that is, the systems with a surface area of 130 and 200 m2 g−1).
For the first time, the addition of silica was observed to slow the phase-separation rate.
To examine the morphology of the three-component and two-component films, cross-sections of both phase-separation membranes were compared. The electron microscopy results are shown in Figure 8.
Figure 8Morphology of the PE phase-separation membrane prepared from two- and three-component systems.1
Full size image
In two-component films, large (1–3 μm) cell structures were observed after slow cooling, and small (about 1 μm) cell structures were seen after rapid cooling. In three-component films, no cell structures were seen. The size of the cell structures of the two-component liquid–liquid phase-separation system (DOP) has already been described.2
Few cell structures were observed in three-component films. The growth of these cell structures is represented by the product of the phase-separation rate, and the time in the viscosity range (temperature range) in which growth is possible.17 The phase-separation rate was fast in the two-component systems, but the phase-separation rate was slow in the three-component systems. Therefore, the temperature of the system decreased to the crystallization temperature of the PE solution without the fusion of nuclei (the growth of cell structures), and crystallization occurred without the growth of cell structures. The resulting film structure does not contain cell structures.
To examine the dispersed state of silica in the three-component phase-separation membrane, silica was extracted with NaOH, and the membrane was examined. Little change was seen in the structure of the film before and after the extraction of silica (electron micrograph not shown). The primary particle diameter of the silica used was 0.01–0.02 μm. No change occurred in the structure before or after extraction, indicating that the silica is dispersed as primary particles and does not aggregate.
A film obtained by three-component phase separation was processed by the following two methods: (1) an extension/extraction process in which extension was performed without the extraction of the solvent and the silica; and (2) the solvent and the silica were extracted from the phase-separation membrane and then extension was performed. The effect of these processes on the film properties was examined. The results are shown in Table 3 .
Table 3 Physical properties of the membranes made by different processes1 (extraction/extension and extension/extraction)Full size table
Two important observations were made: (1) In the case of the extension/extraction process, the pore diameter was small, and the strength was high. (2) In the case of extraction/extension, the pore diameter was at least four times larger than the pore diameter in the first process, and the strength was not as high. These differences between processes were similar to the results of processing the two-component phase-separation membrane.2
The extension rate and extension ratio of phase-separation membranes with the standard three-component composition were measured. Phase-separation membranes with a thickness of about 100, 200 and 300 μm and films obtained by extracting the solvent and the silica from the phase-separation membranes were prepared and subjected to a 3 × 3 times extension at three rates (10, 100 and 300% per second). The relationship between film thickness after extension (after extraction was performed when the phase-separation membrane was directly extended) and puncture strength is shown in Figure 9.
Figure 9Effect of extension rate on the puncture strength of the films prepared by the phase separation/extension/extraction process.1 Symbols indicate the extension rate. •: 10% per second; ▪: 100% per second and ▴: 300% per second.
Full size image
Using Figure 9, interpolation and extrapolation were performed at 25 μm to obtain puncture strength at 25 μm. The relationship between puncture strength at 25 μm and extension rate is shown in Figure 10.
Figure 10Effect of extension ratio on the puncture strength of the films prepared by extension/extraction (▪) and extraction/extension (▴).1
Full size image
When the extension rate was increased, an improvement in strength was seen for both processes. This effect was slightly greater for the extension/extraction process. These results were caused during the increase in the extension rate; the molecule slip was reduced. The effect of the extension was clear. To examine the effect of the extension rate on the morphology of films; films obtained by the extension/extraction process were observed using an electron microscope. These results are shown in Figure 11.
Figure 11Scanning electron micrograph of the PE phase-separation membrane showing the effect of extension rate on the morphology of the membrane.1
Full size image
With an increase in the extension rate (10–600% per second), the density of the film surface increased. The films that extended at a high rate were expected to have reduced molecule slip, increased strength and a very uniform structure.
The effect of the extension ratio on the strength of the films (including the molecular-weight effect) was also studied. A method similar to that of the experiment determining the effect of the extension rate was used. The results for the microporous films made by extraction/extension with two molecular weights are shown in Figure 12. The results for the microporous films made by extension/extraction with two molecular weights are shown in Figure 13.
Figure 12Effect of extension ratio on the puncture strength of the film prepared by the phase separation/extraction/extension process.1 Symbols indicate the molecular weight of PE. ▴: 6.0 × 104 and ▪: 8.0 × 104.
Full size image
Figure 13Effect of extension ratio on the puncture strength of the films prepared by the phase separation/extension/extraction process.1 Symbols indicate the molecular weight of PE. ▴: 6.0 × 104 and ▪: 8.0 × 104.
Full size image
As a result of this experiment, three observations were made: (1) For both processes, the puncture strength increased with an increase in the extension ratio. (2) For the extraction/extension process, when the PE molecular weight increased from 600 000 to 800 000, the puncture strength also increased. (3) In the extraction/extension process, the range in which extension was possible was slightly narrower than the range in the extension/extraction process. When the extension ratio was more than about 5 × 5 (25 times) in extraction/extension and more than about 7 × 7 (49 times) in extension/extraction, a break during extension occurred.
The range of the extension ratio suggests that during extraction/extension, extension was performed with voids present, and part of the extension force was used to expand the voids (as in the two-component systems).2 During the high-ratio extension, void expansion is assumed to have led to the break. During extension/extraction, no voids were present, and therefore, the extension effect was improved. Extension was assumed to be possible without break up to the higher ratio. Electron micrographs of the films obtained by both processes are shown in Figure 14. Films with a large pore diameter and a high void ratio (porosity) were obtained, and for extension/extraction, films with a small pore diameter and high strength were obtained (Table 3).
Figure 14Differences in the morphology of PE membranes prepared by the extraction/extension process and the extension/extraction process.1
Full size image
The estimation of the phase-separation state, the structure of phase-separation membranes and the processing (extension) of phase-separation membranes and the film structure were studied for three-component systems containing PE, inorganic powder and a solvent.
Five conclusions were reached from these studies: (1) The phase-separation state of the three-component system, which was not previously studied because the optical analysis of the phase-separation state was conventionally impossible, was successfully estimated from the change in viscosity (torque). (2) On addition of silica, silica is assumed to absorb the solvent, and the apparent PE concentration increased. As a result, the phase-separation temperature decreased. (3) With the addition of the inorganic powder, the viscosity of the system increased. Further, when silica with a large surface area was used, the viscosity of the system was further increased. (4) The phase-separation rate of the three-component system was slower than that of the two-component system. Therefore, the fusion of nuclei was slow, and the obtained film was a dense film without cell structures. Little change was observed in this system, even when silica was extracted. (5) When the extension of the phase-separation membrane was compared for the two processes, the extension/extraction film had a high strength and a small pore diameter, and the extraction/extension film had a low strength and a large pore diameter.
A method can be designed for producing a material that exhibits target film properties by selecting the extraction/extension, extension/extraction and extension ratio based on the target values for the important properties of a separator (the strength and pore diameter).
The first use of a microporous film separator was in a Li primary battery. The polypropylene film had too high a melting point as an LIB separator, and therefore, when the temperature inside the battery is high, the polypropylene film cannot perform the fuse function. The Asahi Kasei Corporation first demonstrated that a microporous PE film is the optimum LIB separator with a fuse function.26
The physical properties of the films obtained by the wet process (the two-component and three-component methods in the thermally induced phase-separation films developed by Asahi Kasei Corporation) and the dry process (the conventional method that produces porous films through extension-induced pore forming) were compared and studied.
The wet process involves three steps: (1) A polymer, a solvent and an inorganic powder are uniformly mixed at a high temperature. (2) The temperature is decreased to thermally induce the phase separation of the mixture into the polymer and the solvent+inorganic powder. (3) The solvent+inorganic powder phase is extracted. Before and after this step, extension is performed in some cases.
The dry process also involves three steps: (1) A polymer is extruded into a film at a high draft ratio. (2) The film is annealed for crystallization. (3) The film is extended at low temperature and then at high temperature to form pores at the crystal interface by extension stress.
Electron micrographs of the films obtained by these processes are shown in Figure 15. A microporous film with a large pore diameter was made for the first time by the three-component method. The mean pore diameter of the three-component film measured by the mercury intrusion method was 0.51 μm, which was about three times the mean pore diameter of the two-component film (0.15 μm). The dry process (the extension-induced pore-forming method) produced films with a small mean pore diameter of 0.09 μm. The Garley value, which represents air permeability, was low for the three-component film, indicating the ease of air permeation for this film and a film (pore) structure through which ions easily permeate.
Figure 15Electron micrograph of microporous membranes made by different processes.3
Full size image
The physical properties of the various films are shown in Table 4 . The initial capacity retention rate after 500 cycles of charge and discharge was 85% for the three-component film, 80% for the two-component film and 70% for the porous film made by the dry process. The mechanism was not clear, but these results may be due to the large pore diameter of the three-component film. An oligomer may have been produced by the side reaction of the electrolytic solution during charge and discharge and been deposited, but no clogging occurred. Therefore, the initial capacity retention rate was high.
Table 4 Physical properties of the microporous membranes3Full size table
The fuse temperature could be decreased for the three-component film by mixing in linear low-density PE (LLDPE). An example of the fuse temperature measurement is shown in Figure 16. Resistance value was measured, and the temperature was increased. The resistance value sharply increased around 135 °C. Without the addition of LLDPE, the resistance value sharply increased at about 140 °C. This increase in resistance value was caused by the clogging of pores in the microporous film. For one film, the resistance value sharply decreased around 150 °C. For another film, the resistance value did not decrease until 170 °C at the last measurement, and these films contained UHMWPE. The film in which the resistance value decreased indicates that conduction occurred because of PE film break. The film in which the resistance value did not decrease until 170 °C was incorporated into a battery, and a safety test was performed.
Figure 16Temperature dependency of impedance showing the thermal stability of the separator.
Full size image
Measurements were made of the pore structures of the films obtained by the wet process (the two-component method and the three-component method). The pore diameter for the three-component method was greater than the pore diameter for the two-component method. For the three-component method, the cycling properties were improved under certain battery-making conditions because the three-component film has a large pore diameter, and the deteriorated electrolytic solution did not easily clog the pores. In the separator with a large pore diameter, the discharge properties at low temperatures were good. The addition of LLDPE to the polymer composition decreased the fuse temperature. Owing to the decrease in fuse temperature, this polymer has a role as a safety device. When the temperature inside the battery increases because of overcharge, safety improves.
Four conclusions were reached by measuring the pore structures of the films obtained by the wet process (the two-component method and the three-component method). (1) The pore diameter produced by the three-component method was greater than the pore diameter produced by the two-component method. (2) For the three-component method, the cycle properties were improved under certain battery-manufacturing conditions because the three-component film has a large pore diameter, and the deteriorated electrolytic solution did not easily clog the pores. (3) For the separator with a large pore diameter, the discharge properties at low temperatures were good. (4) The addition of LLDPE to the polymer composition decreased the fuse temperature. Owing to this decrease in fuse temperature, this material is useful as a safety device. When the temperature inside the battery increases (for example, because the overcharge increases), safety improves.
For more information pe separator, please get in touch with us!