GB1592238A - Process for manufacturing ultra-thin sintered pvc battery separators - Google Patents

Description

(54) PROCESS FOR MANUFACTURING ULTRA-THIN SINTERED PVC BATTERY SEPARATORS (71) We, GENERAL MOTORS COR PORATION, a Company incorporated under the laws of the State of Delaware, in the United States of America, of Grand Boulevard, in the City of Detroit, State of Michigan, in the United States of America (Assignees of VAN VECHTAN RIES BERG Jr) do hereby declare the invention for which we pray that a patent may be granted to us, and the method by which it is to be performed to be particularly described in and by the following statement: This invention relates to a dry-sintering process (ie. without solvents or plasticizers) for the manufacture of PVC (polyvinyl chloride) battery separators in thinner (ie.

less than .3048 mm) strips than it was practically possible to do heretofore without materially reducing the separator's strength, its resistance to interelectrode dendrite growth and its electrical conductivity in the battery.

Battery separators function essentially to electrically isolate the positive and negative plates from each other. They prevent direct contact and suppress interelectrode dendrite growth, commonly referred to as "treeing", which causes shorting of the respective plates. An ideal separator would isolate the plates without inhibiting electrolyte mobility, and without increasing the battery's internal resistance. The separator manufacturer's ability to achieve the ideal, however, is thwarted by practical manufacturing limitations. Processes are sought which will yield maximum total porosity and thinness (ie. for achieving low electrical resistance), and minimum pore size (ie. for achieving maximum "treeing" resistance). The relationship that exists between the total porosity and size of the pores defining that porosity for a separator of given thickness can be quantitatively characterized in terms of the separator's air permeability according to the Gurley porosimeter method and is referred to herein as the separator's "porosity profile".

Heretofore, high-speed, dry-sintering processes have been able to produce separators having thicknesses as low as .3556 mm and total porosities of about 50%, but with average pore sizes no less than about 14 microns. With pores this large, the .3556 mm thickness is necessary to provide adequate strength and "treeing" resistance. Prior to the present invention, such processes have not been able to produce acceptable separators less than 0.3048 mm inch thick at commercially practical rates. Moreover, separators that have been made have proven unacceptable for applications such as Pb-Ca maintenance-free batteries which have a higher "treeing" resistance requirement not met by the larger pores in the thinner separators. In this respect, acceptable separators are herein intended to mean those which have an electrical resistance (ie.

at 26.7"C) which does not exceed about 0.00019 Ohms/cm2 for each 25.4 mm of web thickness, and have a porosity yielding an air permeability of not less than 24 secs for passing 300 ccs of air in a Model 4100 Gurley Densometer with a 16.13 mm2 orifice and a 141.7g weight (ie. 24 Gurley).

Separators with Gurley ratings above about 60 secs, on the other hand, tend to have too high a resistance for most applications.

The method of the present invention comprehends: sintering PVC powder mixed with, preferably 3% to 15% by volume, leachable, pore-forming particles which are less than 10 microns in diameter (average); warm deforming or calendaring of the sintered mix at a temperature above 1200C but less than the sintering temperature to reduce the sintered pore size without collapsing the pores, and to stabilize the strip against in-service growth; the particles are subsequently leached out to leave only the smaller pores. Earlier attempts to reduce the pore size by simply calendaring the sintered sheet but without the pore-formers or with the pore-formers but at too high a temperature only increased the electrical resistance to an unacceptable level.

Moreover, calendaring at too low a temperature following sintering would not fix the separator's thickness against in-service expansion as will be pointed out hereinafter.

The pore-forming particles used in combination with the warm deformation step have an average particle size of less than 10 microns and preferably range from 1 to 7 microns in diameter and have an average particle size of less than 4 microns. Preferably, the pore-forming particles are comprised of materials which evolve gas under PVC sintering conditions (ie. evolve a gas in the sintering oven) yet leave soluble (ie in acid or water) residue in the PVC. One such preferred gas-evolving material is sodium bicarbonate which gives off about 20% of its weight as CO2 at 210 C and leaves somewhat smaller (ie about 10%) sodium carbonate particles in their stead. A particular advantage of sodium bicarbonate over other gasifiable pore-formers is that its bulk density (ie. ca 0.47 g/cc) is near that of the PVC (ie. ca 0.53 g/cc) which greatly simplifies mixing and fluidization of the PVC-bicarb mixes.

In carrying out the process of this invention, the mix is spread on to a moving metal belt as a layer less than 0.3048 mm thick (usually about 0.254 mm). The particle layer is heated (e.g. about 213"C) as it passes through an elongated oven to sinter the PVC particles into a continuous strip. Following sintering, the strip is cooled to a temperature of 1200C to 1500C. (preferably about 135"C), and at this temperature is compressed, preferably between calendar rollers, to a thickness which is no greater than one-half (preferably about one-third) its sintered thickness. This warm compression deforms the warm PVC particles, improves their bond strength to each other and shrinks the pores between them. In this step, the leachable particles serve to prevent collapse of the pores and, like a core in moulding, to some extent generally define the pores themselves.

Following compression, the still warm strip elastically recovers much, but not all, of its lost thickness and hence remains somewhat permanently deformed. More specifically, it generally recovers 75% to 90% (ie. preferably about 90%) of its sintered thickness. The precise amount of recovery in each instance will vary with the degree of compression and the compression temperature used, and it has generally been observed that greater compression is required at the lower compression temperatures (ie. nearer 1200C to achieve the desired pore size and recovery than is needed at the higher compression temperatures (ie. nearer 1500C).

Following recovery, the strip is cooled to fix the post-compression thickness achieved at the exit of the calendaring rolls. The leachable pore-forming particles remain with the PVC throughout the foregoing, but are ultimately removed by the time the battery is in service: they may be immediately removed as by a distinct leaching step, but preferably are left in situ and are ultimately removed in the battery by the action of the acid therein. The particular combination of process parameters (eg.

composition, layer thickness, sintering time/ temperature, and degree and temperature of compression) is chosen to achieve a particular design thickness after the calendaring rolls. Following cooling and fixing of the separator thickness, the separator strip is ready for cutting and forming into individual separators or separator-envelopes according to the many techniques known to those skilled in the art. Conventional spacer ribs may be formed on the separator at the time the powder layer is spread on to the belt during calendaring or at any other time, as is well known to those skilled in the art.

Warm compression in the 1200C to 1500C range following sintering has been found essential to fix the post-compression thickness against further growth during the service life of the battery. It has been observed that when the PVC is compressed at temperatures less than about 1200C, an initial partial elastic recovery occurs immediately after compression, but that this thickness is not permanent and a secondary elastic recovery later occurs in the battery in service which unduly internally stresses its elements. This problem has been particularly noticed in automobile batteries located in engine compartments which see as much as 110 C. temperatures. On the other hand, strips compressed at temperatures above 1500C do not recover as much after compression and tend to yield separators with unnecessarily high electrical resistance.

As indicated, the pore-forming particles preferably gasify in the sintering oven and yield a pore-forming residue which is then leached out after the warm compression step. Most preferably, the gasifiable poreforming particles are sodium bicarbonate in the 1 to 7 micron particle range which evolve harmless CO2 and leave sodium carbonate as the residue which does not upset the battery chemistry when removed by the electrolyte in the completely assembled battery.

The Figures generally illustrate, in side elevation, apparatus for carrying out the process of the present invention. Figure 2 is an enlargement of portions of Figure 1.

Fixed thickness PVC battery separators can be made by the process of this invention which are less than about 0.254 mm thick, have greater than 50% porosity, have pores which are, for the most part, less than about 10 microns in diameter and have Gurley air permeabilities greater than 24 secs. The high porosity helps to keep the electrical resistance low by ensuring adequate electrolyte volume and mobility within the cell while the small pore size inhibits the "treeing" through of these thin separators.

Separators have been made by this invention as low as 0.2032 mm thick and with an average pore size of about 7.5 microns (as determined by a mercury porosimeter Aminico Model 7-7118).

Separator-grade PVC particles useful with this invention comprise for the most part particle mixes in which the particles range in diameter from about 13 microns to about 67 microns with an average particle size of less than 36 microns. Thinner separators are made with preferred PVC particles which vary for the most part from about 15 microns to about 48 microns and have an average particle diameter of less than 30 microns. Particle sizes and distributions herein for both the PVC and pore-forming agents are as determined by a Coulter Electronics Counter Model PA11.

The pore-forming particles have an average particle size which is no greater than the 10 micron pore size sought to be obtained in the finished separator. Particular success has been obtained with sodium bicarbonate particles ranging from about 1 micron to about 7 microns in diameter and an average particle size of about 3.2 microns. The sodium bicarbonate content of the PVCbicarbonate mix can vary from as low as 3% to as high as 15% by volume, but 5% to 10% yields consistently acceptable results.

The 5% sodium bicarbonate-PVC mixes seem to achieve about the best balance between acceptable electrical resistance, "treeing" resistance and handling strength.

Otherwise, when the bicarbonate content falls below 3%, the resistance of the compressed separator is unacceptably high. On the other hand, when the bicarbonate content exceeds about 15%, the separator has a lower resistance to "treeing" and is generally too weak and fragile to sustain the normal handling in the plant.

The gasifiable, pore-forming particles which leave leachable resides (ie.

NaHCO3) are preferred over particles which are leachable but do not gas in the oven. The gasifiable pore-formers yield sintered strips whose porosity (ie. before compression) is higher than that predictable based solely on the volume of pore former alone. Just why this is so is not clearly understood though it is believed that the gassing in the oven has a lofting effect on the PVC which lowers the density of the sintered strip prior to compressing. It is also noted that the pore-forming particles themselves grow somewhat smaller during gassing which contributes to the small pore formation achieved during the warm compression step of the process. In one example of this apparent lofting phenomena, a control sample of PVC powder (ie. without a pore-former) was sintered and yielded an uncompressed separator with a porosity of about 50%. The same PVC powder, but with 5% by volume sodium bicarbonate added, had an uncompressed porosity of about 62% (ie. with the carbonate residue still present). When the residue was leached out, the uncompressed porosity of the separator rose to about 65%. It has further been observed that 50% porous PVC control samples (ie. without gassing poreformers) have a porosity approaching only about 40% after the warm compression step whereas those containing soda, as above, are about 48% porous after warm compression (ie. before removal of the salt), and in excess of 50% (ie. 51%-52%) after the carbonate is leached out.

Separator strip material made in accordance with this invention may be processed in substantially the same manner as described in our prior United States Patent 3,551,210.

The drawings of this application depict apparatus like that of Figure 2 of US Patent 3,551,210 but with the addition of means for the warm compression of the separator strip following sintering. In carrying out the present process, the PVC particles are conditioned as necessary for moisture and agglomeration control followed by homogeneous mixing with the pore-forming particles. The specific means for accomplishing this is not part of the present invention but both conditioning and mixing may be conveniently achieved by known fluidization techniques. Figure 1 depicts a conditioning and mixing means 2 for providing the PVC-pore-forming mix to a feed hopper 4 (see Figure 2 for enlargement).

The hopper 4 dispenses the mix on to a continuous stainless steel belt 6 (ie. about 0.813 mm thick) behind a comb like scraper blade 8 which is profiled to form conventional spacing ribs on the strip while spreading the powders. The spacer ribs are preferably combed into the powder layer while it is being spread on to the belt, and the compression means merely compresses the webs between the ribs without appreciably acting on the ribs themselves. It is recognized, however, that the powder may be spread flat and the ribs put thereon after compression and recovery as by hot melt beading, corrugating, or embossing as is well known in the art.

The belt 6 moves at a rate of about 60.96m/min. under the feeding hopper 4 and thereunder receives a layer of mix having a thickness equal to about the height of the dam 10 above the belt 6. The dam 10 is positioned about 0.025 inch above the belt 6 and the comb 8 adjusted (ie. to about 0.02 inch above the belt) to produce a 0.012 inch thick powder layer 11 downstream thereof.

The height of the dam 10 and comb 8 can be varied by appropriate dam and comb adjusting means 12 and 14, respectively. Excess powders form behind the comb 8 a mound 16 which is kept in a constant rolling or eddy-like motion by means of a vacuum skimming device 18 which is so located as to prevent excess powders upstream of the comb 8 from raising the head of the mound 16 to the point that it becomes stagnant.

The powder layer 11 flowing from under the comb 8 is then heated and sintered in a long oven 20. Preferably it is rapidly preheated (ie. to about 1900C) to just below its sintering temperature, and then more slowly heated to sintering of the PVC at about 210 C - 213"C. In the particular embodiment shown, the initial rapid heating of the particles to the 1900C. presintering temperature is accomplished in the first two stages of four-stage oven 20 having gas burners 22 heating the separators through the stainless steel belt 6 which tends to form a thin skin on the bottom of the strip where the PVC is hottest. This skin has a somewhat higher density than the rest of the separator, but even here the pore-forming particles serve to keep the skin from completely sealing off that surface of the separator. The first two burners are located approximately 2 inches below the stainless steel belt 6. The first oven stage is approximately 14.6 m long and the oven temperature is maintained at about 315"C. The second oven stage is about 8.5 m long and is maintained at an oven temperature of about 205"C. The third and fourth oven stages finish the heating and sintering and are 8.5 m and 9.7 m long, respectively, and maintained at oven temperatures of about 320"C and 245"C, respectively. It is to be appreciated that these temperature readings will vary depending on the location of the temperature sensor in each oven, but they do serve to indicate the nature of the preheating and sintering steps used to manufacture separators by the process of this invention.

After sintering, the strip is cooled to a temperature of about 1200C. to 1500C as determined by a temperature probe 13 (see Figure 2 enlargement) contacting the underside of the belt 6 just before the compression means. While forced cooling would be acceptable, it appears that merely extending the length of the line between the oven exit and the compression rollers (discussed hereafter) is sufficient to permit adequate cooling before compression. At the aforesaid 1200C to 1500C temperature, the sintered strip enters the nip of compression rollers 24 which compress the strip between the upper roller and the belt 6. As indicated above, the compression rollers may have flat surfaces if the strip is flat or may have annular grooves for accommodating the ribs if they are already formed on the strip. In this latter case, only the portions of the rollers that are between the annular recesses compress the web portions (ie. between the ribs) of the separator strip. Upon exiting the compressing rollers 24, the strip recovers to about 80%-95% (ie. depending on the temperatures of the PVC and degree of compression) of its as-sintered thickness before compression, which is the design thickness of the separator. Air cooling to room temperature after the warm compression fixes the thickness of the strip against further elastic recovery and swelling while in service. Finally, the strip is peeled from the belt 6 as by a stripper means 26 and cut into desired lengths as by blade 28.

0.254 mm thick PVC separators compressed (ie. at about 135"C) to about onethird their as-sintered thickness using the preferred 5% NaHCO3 mix have demonstrated resistances of about 0.00155 ohms/ cm2 and Gurley air permeabilities of about 30 secs. With the same material, 0.3 mm thick PVC separators made this way have demonstrated 0.002 ohms/cm2 and Gurley air permeabilities of about 42 secs. With the same material, 0.203 mm thick separators made this way have demonstrated 0.0013 ohms/cm2 resistance and Gurley air permeabilities of about 33 secs. These resistance measurements were determined in a typical battery separator test cell 27"C using 1.280 specific gravity H2SO4.

WHAT WE CLAIM IS: 1. A method of making micro-porous battery separators by a dry-sintering process including spreading PVC (polyvinyl chloride) particles into a thin layer on a moving belt, sintering the particles to form a continuous electrolyte-permeable strip having therein a multiplicity of interconnected pores, and cutting the strip into individual separators, said method including mixing the PVC with leachable pore-forming particles of less than 10 microns average size; spreading the mix on to said belt to a layer thickness of less than 0.3048 mm prior to said sintering; following sintering compressing and elastically deforming said strip at a temperature greater than 1200C but less than the sintering temperature so as to reduce the thickness of the sintered strip by at least one half and to deform the PVC substantially about said pore-forming particles; permitting the strip while at substan

**WARNING** end of DESC field may overlap start of CLMS **.

Claims (9)

**WARNING** start of CLMS field may overlap end of DESC **. The belt 6 moves at a rate of about 60.96m/min. under the feeding hopper 4 and thereunder receives a layer of mix having a thickness equal to about the height of the dam 10 above the belt 6. The dam 10 is positioned about 0.025 inch above the belt 6 and the comb 8 adjusted (ie. to about 0.02 inch above the belt) to produce a 0.012 inch thick powder layer 11 downstream thereof. The height of the dam 10 and comb 8 can be varied by appropriate dam and comb adjusting means 12 and 14, respectively. Excess powders form behind the comb 8 a mound 16 which is kept in a constant rolling or eddy-like motion by means of a vacuum skimming device 18 which is so located as to prevent excess powders upstream of the comb 8 from raising the head of the mound 16 to the point that it becomes stagnant. The powder layer 11 flowing from under the comb 8 is then heated and sintered in a long oven 20. Preferably it is rapidly preheated (ie. to about 1900C) to just below its sintering temperature, and then more slowly heated to sintering of the PVC at about 210 C - 213"C. In the particular embodiment shown, the initial rapid heating of the particles to the 1900C. presintering temperature is accomplished in the first two stages of four-stage oven 20 having gas burners 22 heating the separators through the stainless steel belt 6 which tends to form a thin skin on the bottom of the strip where the PVC is hottest. This skin has a somewhat higher density than the rest of the separator, but even here the pore-forming particles serve to keep the skin from completely sealing off that surface of the separator. The first two burners are located approximately 2 inches below the stainless steel belt 6. The first oven stage is approximately 14.6 m long and the oven temperature is maintained at about 315"C. The second oven stage is about 8.5 m long and is maintained at an oven temperature of about 205"C. The third and fourth oven stages finish the heating and sintering and are 8.5 m and 9.7 m long, respectively, and maintained at oven temperatures of about 320"C and 245"C, respectively. It is to be appreciated that these temperature readings will vary depending on the location of the temperature sensor in each oven, but they do serve to indicate the nature of the preheating and sintering steps used to manufacture separators by the process of this invention. After sintering, the strip is cooled to a temperature of about 1200C. to 1500C as determined by a temperature probe 13 (see Figure 2 enlargement) contacting the underside of the belt 6 just before the compression means. While forced cooling would be acceptable, it appears that merely extending the length of the line between the oven exit and the compression rollers (discussed hereafter) is sufficient to permit adequate cooling before compression. At the aforesaid 1200C to 1500C temperature, the sintered strip enters the nip of compression rollers 24 which compress the strip between the upper roller and the belt 6. As indicated above, the compression rollers may have flat surfaces if the strip is flat or may have annular grooves for accommodating the ribs if they are already formed on the strip. In this latter case, only the portions of the rollers that are between the annular recesses compress the web portions (ie. between the ribs) of the separator strip. Upon exiting the compressing rollers 24, the strip recovers to about 80%-95% (ie. depending on the temperatures of the PVC and degree of compression) of its as-sintered thickness before compression, which is the design thickness of the separator. Air cooling to room temperature after the warm compression fixes the thickness of the strip against further elastic recovery and swelling while in service. Finally, the strip is peeled from the belt 6 as by a stripper means 26 and cut into desired lengths as by blade 28. 0.254 mm thick PVC separators compressed (ie. at about 135"C) to about onethird their as-sintered thickness using the preferred 5% NaHCO3 mix have demonstrated resistances of about 0.00155 ohms/ cm2 and Gurley air permeabilities of about 30 secs. With the same material, 0.3 mm thick PVC separators made this way have demonstrated 0.002 ohms/cm2 and Gurley air permeabilities of about 42 secs. With the same material, 0.203 mm thick separators made this way have demonstrated 0.0013 ohms/cm2 resistance and Gurley air permeabilities of about 33 secs. These resistance measurements were determined in a typical battery separator test cell 27"C using 1.280 specific gravity H2SO4. WHAT WE CLAIM IS:

1. A method of making micro-porous battery separators by a dry-sintering process including spreading PVC (polyvinyl chloride) particles into a thin layer on a moving belt, sintering the particles to form a continuous electrolyte-permeable strip having therein a multiplicity of interconnected pores, and cutting the strip into individual separators, said method including mixing the PVC with leachable pore-forming particles of less than 10 microns average size; spreading the mix on to said belt to a layer thickness of less than 0.3048 mm prior to said sintering; following sintering compressing and elastically deforming said strip at a temperature greater than 1200C but less than the sintering temperature so as to reduce the thickness of the sintered strip by at least one half and to deform the PVC substantially about said pore-forming particles; permitting the strip while at substan

tially said compression temperature, to elastically recover much, but not all, of its thickness as sintered; and cooling said strip so as substantially to fix the thickness of said strip at said recovered thickness thereafter and throughout its service life, the amount of said pore-forming particles and the extent and temperature of said compression being such that, upon removal of said poreforming particles from the strip, the strip will have a porosity of at least 24 seconds Gurley air permeability (measured as indicated herein).

2. A method according to claim 1, in which the mix comprises 3% to 15% by volume of the pore-forming particles.

3. A method according to claim 1 or 2, in which said particles are present throughout the steps of the method but are ultimately leached out of the sintered strip.

4. A method according to any of claims 1 to 3, in which said particles are gasifiable during sintering to yield a leachable particulate residue the average particle size of which is less than 10 microns.

5. A method according to claim 3, in which the strip is compressed to reduce the size of the pores to approximately that of the particles comprising said residue.

6. A method according to any of claims 1 to 5, in which said particles are sodium bicarbonate, with an average particle size of 1 to 7 microns, the strip is cooled at 135"C, and is permitted, while still warm to recover at least 80% of its thickness when sintered.

7. A method according to any of claims 1 to 6, in which said strip is compressed to approximately one third of its thickness as sintered.

8. Battery separators made by the method according to claim 6 or 7, in which said strip material has a porosity of more than 50%.

9. Battery separators made by the method according to any of claims 1 to 7.