CA2260005C - Glass fiber separators for batteries - Google Patents

Abstract

A glass fiber separator material is disclosed. The separator is composed of a mass of intermeshed glass fibers substantially all of which have a fiber diameter not greater than about 20 .mu.m, and at least 5 percent w/w of which have a fiber diameter less than 1 .mu.m, and, distributed through the glass fibers, and from 0.2 percent w/w to 20 percent w/w of cellulose fibrils. The fibrils are from a slurry having a Canadian freeness sufficiently low that the separator material has a tensi le strength greater than on otherwise identical separator where glass fibers having an average diameter greater than 1 .mu.m replace the cellulose fibrils.

Description

TITLE
GLASS FIBER SEPARATORS FOR BATTERIES
EACKGROUND OF THE INVENTION
FIELD OF THE INVENTION
The present invention relates generally to the field of batteries and, more specifically, to separators containing glass fibers which are positioned between the positive and negative plates of batteries and to a method for producing such separators. As is subsequently discussed in more detail, separators containing glass fibers are well known.
Long before glass fiber separators, however, cedar veneers were used as a separator material, and were replaced by microporous, hard rubbery separators and cellulose separators impregnated with resins.
DESCRIPTION OF THE PRIOR ART
Valve regulated ("sealed" - "recombinant") lead acid (VRLA) batteries are known;
they usually comprise a plurality of positive and negative plates, as in a prismatic cell, or layers of separator and positive and negative electrodes wound together, as in a "jelly roll"
cell. The plates are arranged so that they alternate, negative - positive -negative, etc., with separator material and paste separating each plate from adjacent plates.
The separator, which, typically, is a mat of glass fibers, is an inert material;
it stores battery acid, applies a force to paste-grid interfaces, and provides low electric resistance. In addition, in VRLA batteries, there arc innumerable gas channels in the separator material through which oxygen can migrate from the positive electrode, when generated there, to the negative electrode where it can be recombined with hydrogen, according to the oxygen cycle. One of the most important functions of a separator in a VRLA battery is to force the paste into contact with the plates, and cause a pressure between the plates.
Glass fiber separator material, typically, is produced commercially on paper making equipment including fourdrinier machines and rotoformers, inclined fourdrinier machines and extended wire rotoformers. In the production of separator made of glass fibers for VRLA batteries, it is preferred that no organic material be added to a furnish from which separator sheets are made; the entanglement of individual fibers serves to maintain the sheet in a cohesive structure, and water glass, which sometimes forms on the fiber surfaces, serves as a binder. Organic binders, however, tend to decrease the ability of a separator to wick acid, and to decrease the amount of acid a separator can hold. A
great deal of work has been directed to modifying the glass fiber furnish from which separators are produced to improve battery performance and/or lower the cost of the separator. Some of the work has entailed the addition of synthetic fibers for various S_ reasons, such as the use of thermoformable plastic fibers so that the separator can be heat sealed on its edges to envelop a plate. Other work, which pertains to the field of this invention, has been directed to the use of filler, e.g., silica, to provide separators which are comparable to all glass fiber separators, at a lower cost. Separators made from glass fibers to which cellulose has been added and polyolefin fibers to which cellulose has been added have also been suggested. Prior art patents are discussed below.
US Patent No. 4,465,748 (Harris) discloses glass fiber sheet material for use as a separator in an electrochemical cell, and made from 5 to 35 percent w/w of glass fibers less than 1 pm in diameter; the patent also discloses a glass fiber sheet for such use wherein there are fibers of a continuous range of fiber diameters and lengths, and most of the fibers are not over 5 mm in length.
US patent No. 4,216,280, (Kono et a1.), discloses glass fiber sheet material for use as a plate separator in a battery, and made from 50 to 95 percent w/w of glass fibers less than 1 p.m in diameter and 50 to 5 percent w/w of coarser glass fibers. The coarser glass fibers, the reference says, have a fiber diameter larger than 5 pm, preferably larger than 10 Vim, and it is advantageous for some of the coarser fibers to have diameters of 10- pm to 30 wm.
US Patent No. 4,205,122 (Minra et al) discloses a battery separator of reduced electric resistance comprising a self supporting, non woven mat consisting essentially of a mixture of olefinic resin fibers having a coarseness of from 4 to 13 decigrex and olefinic resin fibers having a coarseness of less than 4 decigrex, the latter fibers being present in an amount of not less than 3 parts by weight per 100 parts by weight of fibers;
up to about 600 parts by weight of inert filler materials per 100 parts by of fibers can also be used. The battery separator is produced by subjecting a suitable aqueous dispersion to a sheet-forming operation, drying the resulting wet, non-woven mat, and heat treating the dried mat at a temperature ranging from a point- 'Z0° lower than the melting point of the aforementioned fibers to a point about 50° higher than the melting point.

US Patent No. 4,216,281 (O'Rell et al.) discloses a separator material produced from a furnish containing 30 to 70 percent w/w of polyolefin synthetic pulp, 15 to 65 percent w/w of a siliceous filler and 1 to 35 percent w/w of "long" fibers which can be polyester fibers, glass fibers, or a mixture of the two. Cellulose in an amount up to about percent w/w is disclosed as an optional ingredient of the furnish.
US Patent No. 4,363,856 (Waterhouse) discloses a separator material made from a furnish composed of polyolefin pulp fibers and glass fibers, and names polyester staple fibers, polyolefin staple fibers and cellulose pulp fibers as alternative constituents of the furnish.
10 US Patent No. 4,387,144 (McCallum) discloses a battery separator having a low electrical resistance after extended use which is made by thermal consolidation and thermal embossing of a paper web formed from a furnish containing a synthetic pulp the fibrils of which are filled with an inorganic filler, the web incorporating a wetting agent which is preferably an organic sulphonate, and organic succinate, or phenol ethoxylate.
US patent No. 4,373,015 (Peters et al.), discloses sheet material for use as a separator in a battery, and "comprising organic polymeric fibers"; both of the examples of the reference describe the sheet material as "short staple fiber polyester matting about 0.3 mm thick", and indicate that the polyester fibers range from about 1 p.m to about 6 p,m in diameter.
~0_ Sheet separators for use in conventional (not valve regulated) batteries and comprising both glass fibers and organic fibers are disclosed in all of the following US
patents: No. 4,529,677 (Bodendorf); No. 4,363,856 (Waterhousc); and No.

4,359,511 (Strzempko).
US patent No. 4,367,271, Hasegawa, discloses storage battery separators composed of acrylic fibrils in an amount of up to about 10 percent w/w, balance glass fibers.
Japanese patent document 55/146,872 discloses a separator material comprising glass fibers (50-85 percent w/w) and organic fibers (50-15 percent w/w).
US patent No. :1,245,013, Clegg et al., discloses a separator made by overlaying a first sheet of fibrous material including polyethylene fibers with a second sheet of 30 fibrous material including polyethylene and having a synthetic pulp content higher than the first sheet.
US Patent No. 4,908,282, Badger, discloses a separator comprising a sheet made from first fibers which impart to the sheet an absorbency greater than 90°~o and second 26815-55(S) fibers which impart to the sheet an absorbency less than 8(1% wherein the first and second fibers are present in such proportions that the sheet has an absorbency of from 75 to 95%.
This patent discloses that fine glass fibers have a high absorbency. that coarse glass fibers have a low atuorbency, and that hydrophobic organic fibers have an cxtrcmet)~
low absvrbeney, and that, when this separator is saturated with clcxtrolyte, unfilled voids remain so that gas can transfer from plate to plate for recombination.
US Patent No. 5,091.275 (F3rccht ct at.) discloses a glass fiber separator which expands when exposed to electrolyte. Tile separator comprises eiass fibers which arc LQ impregnated with an aqueous solution of colloidal silica particles and a sulfate sail. The separator is produced by forming a paper making web of glass filxrs, impregnating the web with the aqueous mixture of silica and the salt, lightly comprcwing the impregnated web to remove some of the aqueous solution, partially drying tttc web, compressing the web to a final thickness and completing the drying of the web. Tltc web is preferably compressed to a thickness which is less than the distance het~~ecn plates in a given cell, so that insertion of an assembled cell stack into a case is facilitated. When electcol~~te is added to the case. the salt dissolves in the electrolyte and the sepuator expands to provide good contact between the plates and the aepantors. According to the patent, the silica contributes to the recombination performance of cells incorporating the pre-compressed ,~,~ separator. The silica also contributes a great deal of stiffness to Ihc separator, so much so that the separator may be characterized as rigid.
It has been determined that the production of battery separator by paper-making techniques fmm a furnish of glass fibers and silica powder leads to problems wttich arc caused by variations in the concentration of the silica powder in the furnish.
Typical glass fiber furnishes have a liquid content exceeding 9R percent wIN~. In the course of making separator sheets, most of the w:ttcr is removed from the furnish in the first few feet of a screen on which the furnish is cast. The water, known as white water, is recycled and winds up back in the hcadhox of the machine. If the furnish is composed exclusively of glass fibers, virtuall~~ none of the fibers pass through the wire and wind up in the white ~Q water. However, furnishes comprising glass fibers and silica powder do not fare so well.
In the absence of a retention aid, significant amounts of silica powder from such furnishes do pass through the paper making wire :uuf wind up in the white mater. Left unchecked, 26815-55(S) this phenomenon causes the concentration of silica powder in the furnish to increase, undesirably changing the properties of the furnish. Heretofore, the problem of silica powder and the like passing through a paper making wire has hecn avoided through the use of binders as retention aids.
US patent No. ?,477,0(H) discloses a syntltctic fiber papa prcxluccd from fibrillae and fibers made by methods wherein a solution of the fiber is extruded through very small orifices (spinnerets) and then the extruded solution is altowcd to congeal either in a precipitating bath or by evaporation of the soUcnt or by temperature changes (see column 2, lines 25 and foitowing). The patent says that fibers of cellulose acetate, LQ cellulose nitrate, regenerated cellulose from viscose, *Vinylitc (a synthetic resin made by polymerization of vim~l compounds).~'Aralac (a fibrous product made from skim milk casein), and spun glass" which range in length from '/~ inch to I inch and in diameter from 1?.-80 microns and fibriilac prcfcrahly derived from flat, Manila hemp, caroa or hemp can be used to make the paper. At Icwt ~)t) pcrccn t of the fibrillac should he from O.U015 to O.t~25 inch in length and from tf.tHHHH)37 to O.iHI()pU4d inch in width.
BRIEF DESCRIPTION Q~'LEf>t1_N_SMAN.~IN~NT10N
The instant invention is based upcm the discovery that comparatively small additions of wood pulp, if beaten or refined to a sufficient degree to product a highly fibrillated cellulose fiber, to a glass fitxr furnish witable for use in making battery ,?~Q separator material, (l) cause surprisingly high incrca~es in some of the strength pmpctties of separator made from the furnish.
(3) impmvc the cut through resistance of , separator made from the furnish.
(3) and have a unique characteristic in that they hold a greater proportion of acid introduced thereunto when the separator is subsequently compressed.
In addition, the separator is repulpahle, in the sense that it can he used as a constituent of a glass fiber which is used to prc~ducc ~new" separator; furthermore, batteries made from glass fiber separator material which contaitte comparatively small amounts of wood pulp which has been beaten or refined so a sufficient degree, have remarkably long service lives, as indicated by their performance in cycling tests. In general, the pulp slurry should * Trade mark 26815-55(S) be beaten or rE:fined to a Canadian freeness not greater than about 650 cc, or to an equivalent freeness by other measurement te<:hniques, :~.nd a remarkablEv increase in tensile strength is achieved wren the pulp is beaten or refined to a Canadian freenE:ss not c~rf;~:~ter than about 120 cc, or to an equivalent freE:ness by otaer measurement techniques.
Thus, in one aspect the invention provides a glass fiber separator material. comprising a mass of. intermeshed glass fibers substantia~lvy all of which have a fiber diameter not greater than ak~>out 20 um, and at least 5 percent w/w of which havre a fiber diameter less than 1 um, and, distributed through the glass fibers, from 0.2 percent w/w to 20 percent w/w ct~~_:Lulose pulp beaten to a Canadian standard freeness not g=r_eater than 120 <~c~..
In a second aspect, there is provided a sealed lead/sulfuric acid recombinant storage battery comprising a plurality of lead plates in a closed care, a glass fiber separator material as df-.:~cribed i:n the first aspect between adjacent ones cf said p:Lates, and a bady of a sulfuric acid electrolyte absorbed by said glass fiber separator material and maintained in contacts: with each of tre adjacent ones of said plates.

' I' . , 26815-55 {S) 6a BRIEF Dl~ CRI ON OF THE DRAWINGS
Fig. 1 is a plot of the percent w/w of added cellulose in glass fiber separator material according to the invention vs. the liters per second of air flowing through the separator material under test conditions that arc subsequently described herein.
Fig. ? is a plot of tensile strength, lx~th machine direction ("Tensile, MD") and cross direction ("Tensile, CD"), vs. percent w/w of added cellulose in glass fiber battery separator according to the invention.
Fig. 3 is a plot of percent of initial capacity vs. number of test cycles for a battery according to the invention and for a control battery.
Figs. ~ through 9 arc plots of thickness (the values plotted are 1000 times the thickness of the separator in mm) vs. load and rebound thickness vs. load for fide glass fiber separator materials according to the in4~entinn and a control, where rebound thickness is 1000 times the thickness of a separator material in mm after that separator has been subjected to a load and the load has been reduced to 0.55 pounds per square inch (3.79 ICPa); the data in Figs. 4 through 9 are for dry separator material.
2S2 Figs. 10 through 15 are plots similar to those of Figs. 4 through 9, showing thickness vs. load and rebound thickness ~s. load for the five glass fiber separator materials according to the invention and for the control, but arc based cm data where, before testing. each of the separator materials had been wet with seven times its weight of sulfuric acid hating a specific gravity of 1?86.
Figs. 1G and 17 are plots similar to Figs. ~1 and S, differing in that interpolated points are plotted in the former, so that successive points along the X axis represent equal ;
increments of cellulose content; while experimental values arc plotted in the latter and, as a consequence, as is subsequently explained herein, successive points along the X axis do not always represent equal increments of cellulose content.
~f DEFINITIONS
. Subsequently herein, the.term "percent v/v" means percent by volume; the term "percent w/w" and the symbol 'fin mean percent by weight: the term "wire", as applied to, 26815-55(S) a paper making machine, means the surface of the machine cm which a furnish is cast in producing paper, amd can be, for example, the screen of a Fourdrinier machine or the vacuum drum of a rotoformcr machine; p()rC S1ZCS rCp(1('ICd llCTe111, unless otherwise indicated, arc in microns, and arc determined by the first bubble method or by liquid porosimetry, Coultcr: alt temperatures arc in °C.; and the followinb ahbrcviations have the meanings indicated: Erm = micron or microns; mg=milligram or milligrams:
g=gram or grams; kg=kilogram or kilograms; 1=liter or liters; mi=milliiitcr or milliliters; cc=cubic centimeter or cubic centimeters; pef=pound per cubic foot or pounds per cubic foot;
m=meter or meters; cm=centimeter or centimeter; mm=millimeter or millimeters;
LQ m=meter or meters; mil=inch x t()'' or inches x 10'' (multiply times '_'S.:1 to convert to mm); KPa=pressure in thousands of Ncwtons per square meter: psi=pounds per square inch (multiply time. C~.R~> to convert to KPa); and KN=force in thousands of 'Jcwtons.

Glass fiber separator hand sheets were produced in a lalxlrator~~ apparatus by y~ depositing a furnish on a wire or screen, and draining the furnish. The apparatus comprised a tank with a screen in the bouom, a drain below the screen, a ualvc which opened and closed tilt drain, and a hand paddle which was moved back and forth to simulate the movement of a furnish in commercial papcrmaking apparatus and establish a "machine direGion" parallel to the direction of paddle mcncment. The furnish was 2~ produced by charging to the tank acidified water, pH 2.7, and solids composed of 74.5 percent w/w $chuflcr 206 glass fihcrs, average filxr diameter (1.76 llm. 1'.R
percent w/w '~ Evanite 6 i0 glass fibers, nominal fiber diameter 2.6 t,lm, and I?.$
percent w/w A2()-BC
'h inch glass fibers, nominal fiber diarnetcr 13 t.tm, stirring for about a minute, charging to the tank a kraft pulp slurry which had a Canadian frecness of 57cc and a consistency "~ of 1.235 percent, and stirring for an additional ? minutes. The composition in the mixer, after the pulp addition, contained 73 percent wlw Schullcr 206 glass fibers, 12.5 percent w/w Evanite 610 glass fibers, 12.~ percent w/w A20-BC-'h inch glees fibers and percent w/w pulp fibrils. The furnish and the pulp were stirred for about two minutes, after which the valve was opened so that the water drained through the screen while the separator was retained on the screen. The furnish contained enough glass fibers to produce a separator having a gramnlage of 30 glm~ at a thickness of 0.1~ mm.
The separator hand sheet was heated in a drying oven to about 1 SO° for 30 minutes. Two * Trade mark 26815-55(S) s separator sheets produced as described above were tested and various data, summarized below, were collected (the data arc avcragGS of the determinations on the two sheets).
Frazicr permeability, in the following data and elsewhere herein, is in Uscc/m' C~ ''t) mm HZO. The tests, instruments and apparatus used to determine v arious properties in F.xampie 1 and elsewhere herein arc described in a publication entitled BCllItl3SM
Standard Test Methcxls, Battery Council Intcrnational~
Grammagc (g/m~) 3~,,7 Thickness, mm (undo a IQ load of 1t).3d KPa): ().IS

Tensile, MD

(Ncwtons/m): 3f,z Tensile, CD

(Ncwtons/m): ?75 Elongation. MD

(percent of total): 1.3 Elongation. CD

(percent c~f total):

Port Sizc-first Z~Q bu(~(~Ic mctluxl, E~m 3() Frazicr Permeability 9R

Port size-liquid porosimctrv, Coultcr. ~xm minimum 5.1 maximum IR.S

mean S 5 "Frazier Permeability" values reported herein were determined using Frazicr Permeability tester 91A (TnPP! T'rSlOM-RS).
"Wicking", as reported almcc and subscyucntly herein, was determined by the ~Q procedure described in U.S. patent No. 5,'"S,~9R, column 7, lines '?t) and following, using water instead of sulfuric acid as there described; the test is known as t)tc Japanese Industrial Standard rncthod.
The composition of the Schullcr 2t)6 glass fibers used in Example 1 and in subsequent Examples var~~ slightly from time to time. Mean values, in percent w/w, calculated from data furnished by Schuller for the period when the examples were carried out are given below:
SiO, 65.40 Na,O 16.11 A1.,0~ 2.99 K~O 0.69 Ca0 5.88 B,03 5.31 Mg0 2.79 F~ 1.02 Schuller also indicates that the glass contains Fe~O,, TiO~, ZrO~, Cr,Oz, SrO, BaO, MnO, ZnO, Li~O, SO, and Pb in amounts less than 0.1%.
The nominal composition of the Evanite 610 glass fibers used in Example 1 and in subsequent Examples varies, in percent w/w, within the following ranges:
SiO, 60.0 - 69.0 AI,Oz 3.0 - 6.0 Ca0 5.0 - 7.0 Mg0 2.5 - 4.5 Na,O 8.0 - 12.0 K,O 0.5 - 3.0 B~O~ < 0.02 F, 0.0 - 1.0 Zn0 <0.04 Fe~O~ <0.02 The A20-BC-'/a inch glass fibers used in the procedure described above and in other procedures described herein are commercially available from Schuller under the indicated designation.
Glass fiber separator sheets according to the invention were produced on a pilot plant paper making machine by depositing a furnish on an advancing wire, through which water from the furnish drained. The furnish was produced in a mixer from acidified water, pH 2.7, and solids composed of Schulier 206 glass fibers, Schuller 210X glass fibers, nominal fiber diameter 3.0 um and the same composition as the 206 fibers, and 1/z inch glass fibers. The furnish was stirred in the mixer for about one minute, after which time a kraft pulp slurry which had a Canadian freeness of 57cc and a consistency of 1.235 percent was added to the furnish in the mixer. The composition in the mixer, after the pulp addition, contained about 7 parts by weight of Schuller 206 glass fibers, about 1 part by weight of each of Schuller 210 glass fibers, A20-BC-'h inch glass fibers, and about 0.6 part by weight of pulp fibrils. The furnish and the pulp were stirred for about two minutes, after which time the pulp-containing furnish was charged to the headbox of the pilot plant machine. An addition of 0.6 part by weight of pulp fibrils from red wood pulp 5 that had been beaten to a Canadian_ freeness less than 100 cc was then made to the material in the headbox, and the furnish which resulted was flowed onto the advancing wire to produce a separator having a grammage of 30 g/m' at a thickness of 0.15 mm.
The separator was ultimately heated in a drying oven to about 150° for 30 minutes. The separator had a loss on ignition a little over 12 percent w/w, indicating a total pulp 10 content of about 12 percent w/w. The procedure described in this paragraph constitutes the best mode presently contemplated by the inventors with respect to the production of battery separator material according to the invention.
Cells according to the invention were made using the separator material produced in the pilot plant paper machine as described above, and were subjected to life testing in comparison with batteries made using conventional, all glass separators, but otherwise identical. Batter's capacity after each cycle, as a percentage of initial capacity, is set forth in Table I, below (the control battery test was terminated after 7 cycles):

Table I
Number of cycles Capacity, percent of initial According to invention Control 1 113.5 103.6 2 115.6 93.6 3 111.9 76.0 4 109.3 53.4 5 107.4 34.0 G 105.3 25.1 7 103.6 20.9 8 101.7 9 100.0 * *

10 98.6 11 97.2 ~5 12 95.5 13 93.7 14 90.1 87.6 1G 86.1 ~0 17 80.0 18 74.9 19 74.0 67.3 The data in Table I are presented graphically in Fig. 3, which was computer generated by entering the foregoing data for the battery of the invention and for the control after cycles 1 through 7, but entering zero for the percent of initial capacity after cycles 8 through 20.

Glass fiber separator hand sheets were also produced from other furnishes which contained varying amounts of kraft pulp that had been beaten to a consistency of 0.9906 percent and a Canadian freeness of 57cc. The furnishes also contained the previously identified Schuller 206, 210X and A20-BC-'/z inch glass fibers. The hand sheets were produced in a laboratory apparatus by depositing a furnish on a wire or screen, and draining the furnish. The apparatus comprised a tank with a screen in the bottom, a drain below the screen, a valve which opened and closed the drain, and paddles which were moved back and forth to simulate the movement of a furnish in commercial papermaking apparatus and establish a "machine direction" parallel to the direction of paddle movement. The furnish and the pulp were stirred for about two minutes, after which the valve was opened so that the water drained through the screen while the separator was retained on the screen. The furnish that was charged contained enough glass fibers to produce a separator having a grammage of 30 g/m'- at a thickness of 0.15 mm.
The separator hand sheet was heated in a drying oven to about 150° for 30 minutes. The final compositions, in percent w/w, of representative ones of the furnishes and the properties of the hand sheets that were produced are set forth in Table II, below, where, as in other tables herein, unless otherwise indicated, tensile strength is in pounds per inch of width of the separator (multiply times 0.175 to convert to kilonewtons pcr meter), elongation is in percent, stiffness is "Gurley Stiffness" in mg, pore sizes arc in um, electrical resistance is in ohms per square inch of the separator, and LOSS 011 lgllltl011 1S 117 percent w/w. The compositions of the furnishes are given in the following table:
Composition Example Example Example Example Example of 2 3 ~ 5 6 furnish A20-BC 'h 10 10 10 10 10 inch fibers Cellulose 1 3 7 10 15 Table II
Property Example Example Example Example Example 2 3 :1 grammage, 119.9 121.7 119.3 119.9 119.4 5_ g/m2 Thickness, mm (10.34 KPA) 0.765 0.850 0.653 0.620 0.591 (20 KPa) 0.726 0.753 0.644 0.590 0.570 Tensile, 1-00Newtons/m MD 71.7 135.0 135.7 139.2 149.5 CD 84.7 117.8 108.9 125.4 130.2 Elongation Percent MD 1.37 2.00 1.96 2.08 2.13 1-55CD 1.83 1.67 1.61 1.70 1.92 Frazier Permeability65.7 50.2 13.:1 5.9 n.d.

Wicking seconds/10 83 89 104 153 247 mm ~0 Stiffness, mg Pore Size-first Bubble 25 Method, ,um 16.5 16.0 20.1 21.6 24.0 Electrical 0.002 0.003 0.009 0.011 0.014 Resistance LOI,% 3.3 52 9.0 12.5 18.1 Pore size-30 liquid Porosimetry Coulter, ,um Min 5.570 5.386 3.734 2.628 1.697 Max 42.24 42.24 26.07 17.80 12.43 35 Mean 8.875 8.507 5.753 4.425 3.497 In the foregoing table and in subsequent tables the entry "n.d." means not determined, in the cases of Examples 6 and 11, because the porosity was too low for a determination of Frazier Permeability.
40 Control glass fiber separator hand sheets were produced by the same method from a furnish which was composed of 80 percent w/w of Schuller 210X glass fibers, percent w/w of A20-BC-lh inch glass fibers and 10 percent w/w Schuller 206 glass fibers. The average test results for two control sheets are set forth in Table III, below:
Table III
Grammage, g/m' 117.1 Thickness, mm (10.34 KPA) 0.857 g/m' (?0 KPa) 0.717 g/m' Tensile, Newtons per M

MD 10.8 CD 11.0 Elongation, %

MD 0.70 CD 1.21 Frazier Permeability178.4 Wicking 62 seconds/10 mm Stiffness, mg ~0 CD 655 Pore Size-first bubble Method, ~m 11.0 Pore size-liquid porosimetry, ~5 Coulter, q.m Min 6.86 Max 65.97 Mean 12.98 Electrical n.d.

30 Resistance LOI, % 0.31 Thickness in mm x 1000 of samples of the hand sheets produced as described in Examples 2 through 6 and of the control sheets was also determined under various loads, 35 both in an as produced condition and after having been wet with 7 times its dry weight of sulfuric acid, specific gravit}~ 1.286. All thicknesses reported herein were determined by the method described in U.S. patent No.5,336,27>. The example numbers are column headings in Table IV, below, and thicknesses (the values reported are measured thicknesses in mm x 1000) when the samples were in the as produced condition, at applied loads in KPa indicated in the left column, are set forth under the identifying headings:
Table IV
Applied Load,ControlEx. 2 Ex. Ex. 4 Ex. Ex. 6 KPa 3 5 3.79 38 36.5 31 28.5 26 6.06 35 30.5 26 25.5 23 9.51 29.5 27.5 23 23.5 ? 1 19.5 13.71 25.5 25.5 21 22.5 20 18.5 ~_0 17.57 22 23.5 20 21.5 19 17.5 23.98 20 22.5 18.5 20 19 17 28.87 19 21.5 17.5 19.5 18 16.5 42.65 16.5 19 16.5 18.5 17 15.5 15 "Rebound" thicknesses in mm x 1000 (after the excess of the load above 3.79-KPa was removed from each "as produced" sample) are given in Table V, under headings which give the load that was applied, and from which each sample "rebounded";
the values reported are 1000 x thicknesses in mm at the loads indicated in the left column of the table:
Table V
Applied Control Ex.2 Ex.3 Ex.4 Ex.S Ex.6 Load, KPa 6.06 36 33.5 28.5 27.5 24.5 26.5 9.51 33.5 30.5 29 26.5 23.5 25.5 13.71 31.5 29.5 27 25.5 22.5 26 17.57 29.5 28.5 25.5 25.5 22.5 26 23.98 29 27.5 25 25.5 22.5 25 28.87 28 27.5 25 24.5 22 23.5 42.65 27 27 24 24.5 22 23 _30 The data in Tables and V
are presented graphically in computer generated Figs.

through of the drawings, where the loads are shown in psi, and successive points along the X
axis, which are equally spaced from one another, represent 0.55 psi (3.79KPa}, 0.88 psi (6.06 KPa), 1.38 psi (9.51 KPa), 1.99 psi (I3.71 KPa), 2.55 psi (17.57 KPa), 3.48 psi (23.98 KPa), 4.19 psi (28.87 KPa), and 6.19 psi (42.65 KPa}.
Accordingly, Figs. 4 through 9 are skewed in the sense that, for example, a given distance between the first and second points represents a change from 0.55 psi (3.79 KPa) to 0.88 psi (G.OG
KPa), while the same distance between the last two points represents a change from 4.19 psi (28.87 KPa) to 6.19 psi (42.65 KPa). In order to represent the data from the control 5_ sheets and from Example 2 in a more nearly conventional plot, thickness and rebound thickness (in mm x 1000) were calculated by interpolation from the experimental data for loads of O.G9 psi (4.75 KPa), 1.19 psi (8.20 KPa), 1.G9 psi (11.64 KPa), 2.19 psi (15.09 KPa), 2.69 psi (18.53 KPa), 3.19 (21.98 KPa), 3.G9 psi (25.4? KPa). 4.G9 psi (32.31 KPa), 5.19 psi (35.76 KPa), and 5.69 psi (39.20 KPa). These and the experimental values (in mm x 1000) at 4.19 psi (28.86 KPa) and at G.19 psi (42.65 KPa) are set forth in Tables VI and VII, respectively:
Table VI
Applied Load,Control,Example Control,Example KPa thickness2, Rebound 2, thickness Rebound 4.75 3 G.7 34 8.20 31.G 28.G 34.8 32 11.64 28.0 2G.7 32.3 30 15.09 24.3 24.8 30.5 29.G

18.53 22.8 23.8 29.5 28.4 ?0 21.98 20.G ~~.8 29.2 X8.4 25.42 20.3 22.7 28.7 X7.5 28.86 30 22.5 28 27.5 32.31 19.2 21.7 27.8 27.4 35.76 18.3 20.8 27.5 27.3 39.20 17.4 20.2 27.3 27.2 42.65 1G.5 19 27 27 The data from Table VI are presented graphically in Figs. 1(i and 17, which are computer generated plots using loads in KPa. It will be noted that the curves of Figs.
16 and 17 are similar in shape to those of the corresponding curves of Figs. 4 and 5, which is deemed to indicate that valid conclusions can be reached from the skewed curves.
Thickness and rebound thickness measurements were also made on the separator materials of Examples 2 through G and the controls after the materials had been wet with sulfuric acid having a specific gravity of 1.286. The applied loads in KPa are given in the left hand column of Table VII, below, and thicknesses are set forth under the headings which identify the samples; the reported thicknesses are 1000 times the measured thicknesses of the separator in mm:
5_ Table VII
Applied ControlEx.2 Ex.3 Ex.4 Ex.S Ex.6 load, KPa 3.79 36 20.5 28 29 27.5 27.5 6.06 31.5 27 26 26 25 24,5 9.51 28.5 24 23 24 22 22.5 13.91 26.5 22.5 21 22.5 20.5 20,5 17.57 24 21.5 20 21.7 19.5 19 23.98 20.5 20.5 19 20 19 17.5 28.87 19 19.5 18 19 18 16.5 42.65 17.5 17.5 16.5 17.5 16.5 15.5 "Rebound" thicknesses (after the excess of the load above 3.79 KPa was removed from each sample that had been wet with sulfuric acid) are given in Table VIII, below, adjacent entries in the left hand column which give the load that was applied, and from which each sample "rebounded"; the values reported are 1000 x measured thicknesses in mm):
Table VIII
Applied ControlEx. Ex. Ex. Ex. Ex.
Ioad,KPa 2 3 ~ 5 6 6.06 32.5 27.5 26.5 27.5 27 25.5 9.51 31 25.5 25.5 26.5 25 24.5 13.91 29 25.5 25 25 25 23.5 17.57 27.5 25.5 25 25 25 23.5 23.98 24.5 24.5 24 25 24.5 23.5 28.87 24 24.5 24 25 24 22.5 42.65 23.5 24.5 24 24.5 24.5 ~~.5 The data from Tables VII and VIII are plotted in Figs. 10 through 15, where loads are in KPa. The data of Tables IV, V, VII and VIII and Figs. 4-15 indicate that the separator materials of Examples 2 through 6, above, all have sufficient resiliency that they can be compressed between the plates of a lead acid battery, and that their major surfaces will be urged against the adjacent plates with sufficient force for the battery to perform satisfactorily.

Glass fiber separator hand sheets were also produced by the method described in Example 1 from other furnishes which contained varying amounts of kraft pulp that had been beaten to a consistency of U.990G percent and a Canadian freeness of 57cc, and were then dipped in a latex, 3 percent w/w solids. The final compositions, in percent w/w, of representative ones of the furnishes are set forth in Table IX, below, and the properties of separators produced from the furnishes are set forth in Table X, below, where thickness of the separator material is in mm:
Table IX
Composition Example Example Example Example Example of Furnish 7 8 9 10 11 ''0 1/z inch fibers Cellulose 1 3 7 10 1 S

Table X
Property Example Example Example Example Example grammage 121.6 121.9 127.5 123.1 122.7 S g/mz Thickness, mm (10.34 KPa) 0.792 0.778 0.750 0.742 0.603 (20 KPa) 0.760 0.745 0.720 0.698 0.585 Tensile ~_0 Newtons/m MD 93.0 120.6 139.2 152.3 168.8 CD 80.6 102.0 122.0 139.2 158.5 Elongation, Percent MD 1.8 2.3 1.9 ?.3 1.9 15 CD 1.5 2.1 2.0 2.1 .0 ?

Frazier 8.97 5.08 1.39 0.918 n.d.

Permeabilit~~

Wicking seconds/10 225 184 253 261 391 mm 0 Stiffness, mg Pore Size- 16.8 16.1 19.4 20.5 25,4 First Bubble ?5 Method, ~.m Pore size- ' liquid Porosimetrv Coulter, p,m 30 Min 5.283 4.726 3.427 2.285 1.09?

Max 46.54 40.89 27.52 21.73 11.88 Mean 9.550 7.881 5.839 4.902 2.920 LOI, % 6.7 8.4 12.7 17.1 21.3 Still other glass fiber separator hand sheets were produced by the method described in Example 1 from substantially the furnish of Examples 7-lI which contained various small amounts of kraft pulp that had been beaten to a consistency of 1.235 percent and a Canadian freeness of 57cc. The final compositions, in percent w/w, of representative ones of the furnishes are set forth in Table XI, below, and their properties are set forth in Table XII, below, where thickness is in mm:
Table XI
CompositionExample Example Example Example Example 5 of Furnish 12 13 14 15 6 210 X 77 79 79'/a 79'/z 793/a A20-BC 10 10 10 1.0 10 1/z inch fibers 10 Cellulose 3 1 /q 1/2 1/4 Table XII
Property Example Example Example Example Example 15 grammage 118.4 115.6 117.2 116.4 116.3 g/mz Thickness, mm (10.34 KPa) 0.757 0.751 0.778 0.774 0.797 (20 KPa) 0.662 0.694 0.716 0.703 0.722 20 Tensile Newtons/m MD 49.5 25.3 23.8 20.0 18.5 CD 43.8 20.2 20.7 20.0 2.54 Percent Elongation 8.41 5.75 6.58 6.68 7.82 MD

CD 8.23 6.48 6.06 6.13 8.89 Frazier 129.6 175.2 175.2 186.4 200.8 Permeability Wicking seconds/IO 74 76 72 G7 62 mm Surface area 0.6874 0.6114 0.6603 0.6513 0.7030 Corr. 9.9970 9.9962 9.9991 9.996? 9.9970 Pore size-liquid Porosimetry, Coulter, pm Min 6.050 5.941 7.050 6.496 7.589 Max 44.71 50.49 62.08 70.13 78.26 Mean 10.65 12.04 12.32 12.59 12.17 LOI 0.46 1.56 1.28 0.89 0.75 Control glass fiber separator hand sheets were produced by the same method from a furnish which was composed of 80 percent w/w of Schuller 21UX glass fibers, percent w/w of A-2,0-BC 1/z inch glass fibers and 10 percent w/w Schuller 206 glass fibers. The test results, average of two, are set forth in Table XIII, below, where 5 thickness is in mm:
Table XIII
grammage, g/m' 113.7 Thickness, mm 10 (10.34 KPa) 0.742 (2.0 KPa) 0.600 Tensile Newtons/m MD 10.1 CD 11.0 Elongation, MD 0.96 CD 1.27 Frazier 22.4 ~0_ Permeability Wicking seconds/10 mm 62 The data concerning Frazier permeability from Table X (Examples 12 through 16) and from Table XI (for the corresponding controls) are presented graphically in Fig.
l, which is a computer generated plot of Frazier permeability (called CFM on the drawing) vs. cellulose content. It will be noted that Fig. 1 has points on the X axis for 1.25, 1.5, 1.75. ?.0, 2.25, ?.5 and 2.75 percent pulp. To cause the plot to show these points, for which there was no experimental data, Frazier permeability was calculated for each, of these pulp contents by interpolation between the experimental values at 1.0 percent and at 3.0 percent. The experimental and calculated data input to generate Fig.
2 are set forth below:

Percent w/w cellulose ; Frazier Permeability 0.0 ;

o.~s ; ~s.os 0.5 ?3.2,5 5_ 0.75 ; 21.9 1.0 ; ?1.85 1.25 (Calc) ; ?1.14 1.5 (Caic) ; ?0.44 1.75 (Calc) ; 19.73 2.0 (Calc) 19.03 2.25 (Calc) ; 18.32 ?.5 (Calc) ; 17.61 2.75 (Calc) ; 16.91 3.0 ; 16.2 The data concerning tensile strength from Table XII and from Table XIII are presented graphically in Fig. ?, which is composed of two computer generated plots of tensile strength in pounds per inch (machine direction, in one case, and cross direction in the other) vs. cellulose content. It will be noted that Fig. 2 has points on the X
axis for 20 1.25, 1.5, 1.75. 2Ø ?.25, 2.5 and 2.75 percent pulp. To cause the plot to show these ordinate points, for which there was no experimental data, tensile strength in both directions was calculated for each of these pulp contents by interpolation between the experimental values at 1.0 percent and at 3.0 percent. The experimental and calculated data input to generate Fig. 2 are set forth below:

Percent w/w cellulose ; Tensile, MD (Pounds per inch) 0.0 ; 1.46 0.25 ; 2.685 0.5 ; 2.90 5_ 0.75 ; 2.45 5 1.0 ; 3.63 1.25 (Calc) ; 4.07 1.5 (Calc) ; 4.52 1.75 (Calc) ; 4.96 ~0 2.0 (Calc) ; 5.41 ?.25 (Calc) ; 5.85 ''.5 (Calc) ; 6.29 2.75 (Calc) ; 6.74 3.0 ; 7.18 Percent w/w cellulose ; Tensile, CD (Pounds per inch) 0.0 1.55 0.25 ; 2.54 0.5 ; 2.72 0.75 3.005 1.0 ; 2.93 1.25 (Calc) ; 3.36 1.5 (Calc) ; 3.79 1.75 (Calc) ; 4.22 ~5 ?.0 (Calc) ' 4.65 2.25 (Calc) ; 5.07 2.5 (Calc) ; 5.50 ?.75 (Calc) ; 5.93 3.0 ; 6.36 30 If the calculated data were not plotted, the computer generated plot would move the point representing 3.0 percent w/w pulp to the left to the point which represents 1.25 percent w/w pulp in Fig. 2, so that the curves would rise sharply from tensile strengths of 1.93 and 3.63 at 1.0 percent w/w pulp to tensile strengths of 6.36 and 7.10 at 3.0 percent w/w pulp, but the distance along the X axis from 1.0 to 3.0 would be the same 35 as the distance from 0.75 to 1Ø

Still other glass fiber separator hand sheets were produced by the method described in Example 1 from furnishes containing 35 parts by weight of 206 glass fibers, 65 parts by weight of 210 glass fibers and about 1-2 parts by weight of kraft pulp that had been beaten to various Canadian freenesses. The Canadian freeness of representative ones of the furnishes and various properties of the separators produced therefrom are set forth in Table XIV, below, where thickness is in mm. Because of the _5 small size of the samples and lack of uniformity of the furnishes, the loss on ignition ("LOI") of the hand sheets is the best indication of the cellulose content of the furnish from which it was produced. A hand sheet containing no cellulose can be expected to have a loss on ignition of about '/2%.
Table XIV
Property Example Example Example Example Canadian Freeness 660 548 4?0 225 Grammage 147 143 141 143 g/m' Thickness, mm 10 KPa 0.96 0.9? 0.88 0.g9 KPa 0.84 0.81 0.82 0.88 50 KPa 0.79 0.70 0.70 0.68 Average total 1.8 2.3 ~.3 1.9 tensile, pounds per inch 20 Average 2.2 2.4 2.8 2.1 elongation, %

Loss on ignition, 1.6 1.3 ?.0 1.7 ~l0 Average Tensile 0.01?? 0.0161 0.()163 0.0133 g/m' ' 'S

Table XIV (continued) Property Example Example Example Example 21 ~~ 23 ~4 Canadian Freeness 120 40 30 20 Grammage 143 142 137 146 5_ g/m2 Thickness, mm 10 KPa 0.97 0.91 0.94 0.92 20 KPa 0.84 0.80 0.82 0.82 50 KPa 0.73 0.70 0.70 0.72 10 Average total 2.4 2.5 3.0 4,5 tensile, pounds per inch Average 2.2 2.3 2.3 ?.>

elongation, %

Loss on ignition, 1.8 1.5 1.8 ~,6 %n 15 Average Tensile 0.0133 0.0176 0.0219 0.0308 g/m' Still other glass fiber separator hand sheets were produced by the method described in Example 1 from furnishes containing 35 parts by weight of 206 glass fibers, G5 parts by weight of 210 glass fibers and 3-5 parts by weight of kraft pulp that had been beaten to various Canadian freenesses. The Canac~;a" fr~PnPCe of representative ones of the furnishes and various properties of the separators produced therefrom are set forth in Table XV, below, where thickness is in mm:
25 Table XV
Property Example Example Example Example 25 ~6 27 28 Canadian Freeness 660 548 420 ~~5 ~Grammage 148 144 138 141 g/m' Av erage total 2.6 3.0 2.7 .8 ?

tensile, pounds per inch Average 1.9 2.5 3.1 2,2 elongation, ~o Loss on ignition, 3.5 3.7 3.8 4.0 ~n Average Tensile 0.0176 0.0208 0.0196 0.0199 g/m' Table XV (continued) Property Example Example Example Example Canadian Freeness 120 40 30 Grammage 141 140 141 141 m Average total 3.5 3.5 5.1 7,0 tensile, pounds per inch Average 1.9 2.0 2.1 2,0 elongation, %

Loss on ignition, 4.5 3.6 3.6 4.1 %

Average Tensile 0.0248 0.0260 0.036'_' 0.0496 g/m' ~5_ EXAMPLES 33-4U
Still other glass fiber separator hand sheets were produced by the method described in Example 1 from furnishes containing 35 parts by weight of 206 glass fibers, 65 parts by weight of 210 glass fibers and 9 to 11 parts by weight of kraft pulp that had been beaten to various Canadian freeness. The Canadian freeness of representative ones of the furnishes and various properties of the separators produced therefrom are set forth in Table XIV, below, where thickness is in mm:
Table XVI
Property Example Example Example Example Canadian Freeness 660 548 420 225 Grammage 148 146 140 145 g/m~

Average total 2.5 3.8 4.5 5.1 tensile, pounds per inch Average 2.1 2.1 2.1 2,0 ~0_ elongation, Loss on ignition, 11.3 11.5 8.7 10.0 %

Average Tensile 0.0169 0.0261 0.0319 0.0364 g/m' Table XVI (continued) Property Example Example Example Example Canadian Freeness 120 40 3p Grammage 138 144 140 150 5_ g/m' Average total 6.9 7.8 9.0 13.3 tensile, pounds per inch Average 2.0 2.3 1.8 ?.?

elongation, %

Loss on ignition, 12.0 10.~ 11.5 11.0 %

Average Tensile 0.0500 0.0542 0.0643 0.0887 g/m' As has been indicated above, a remarkable increase in tensile strength is achieved when separator material according to the invention is produced using pulp that has been beaten or refined to a Canadian freeness not greater than about 120 cc. This increase is illustrated by the data of Examples 17 through 40 concerning tensile strength 5_ of separator materials according to the instant invention produced from furnishes containing varying amounts of wood pulp which had been refined to several different Canadian freenesses. The data concerning average tensile strength in g/m' vs.
Canadian freeness are presented graphically in charts A, B and C, below. Chart A is a plot of the indicated data from Examples 17 through 24; Chart B is a plot of the indicated data from Examples 25 through 32; and Chart C is a plot of the indicated data from Examples 33 through 40.

Chart A
Chart B
Chart C

It has been found that the separator material produced as described in each of the foregoing Examples can be charged to conventional papermaking apparatus, and "repulped", either as the sole source for glass fibers and cellulose fibrils or supplemented with additional glass fibers and cellulose fibrils to produce a furnish 5 which can be deposited on the moving wire of paper making apparatus as described above to produce separator material. As a consequence, there is no need for any of the separator material according to the instant invention to be scrapped; instead, it can be recycled. Further, separator material according to the instant invention has improved puncture strength than otherwise identical separator material which does not contain 10 cellulose fibrils; as a consequence, increased yields of acceptable lead acid batteries having expanded metal or continuous cast grids can be achieved.
As has been explained above, separator material made from first fibers which impart to the sheet an absorbency greater than 90~~n and second fibers which impart to the sheet an absorbency less than 80% wherein the first and second fibers are present 15 in such proportions that the sheet has an absorbency of from 75 to 95%, when saturated with electrolyte, still has unfilled voids so that gas can transfer from plate to plate for recombination. Such separator material can be produced according to the instant invention by adding to a slurry COlltallllllg, in suitable proportions, first fibers which impart to the sheet an absorbency greater than 90% and second fibers which impart to 20 the sheet an absorbency less than 80%, from 0.2 percent w/w to 20 percent w/w of a slurry of cellulose fibrils having a Canadian frccness sufficiently low that a separator material produced from the resulting slurry has a tensile strength greater than an otherwise identical separator where glass fibers having an average diameter greater than 1 ~.m replace the cellulose fibrils. Preferably, the fibers which impart to the sheet an 25 absorbency less than 80% include both relatively coarse glass fibers and hydrophobic organic fibers. Polyethylene, polypropylene, acrylic and polyester fibers are examples of preferred hydrophobic organic fibers.
A preferred separator according to the invention having an absorbency (as defined in the above identified Badger patent, of from 75 to 95% which, when saturated with electrolyte, still has unfilled voids so that gas can transfer from plate to plate for recombination contains 33.6 parts by weight Schuller 206 glass fibers or an equivalent, 50.4 parts by weight Schuller 210X fibers or an equivalent, 11 parts by weight Schuller A20-BC'/a inch glass fibers or equivalent, and 5 parts by weight of polyethylene fibers, and, in addition, from 0.~ percent w/w to ?0 percent w/w of cellulose fibrils from a slurry having a Canadian freeness sufficiently low that the separator material has a tensile strength greater than an otherwise identical separator where glass fibers having an average diameter greater than 1 p.m replace the cellulose fibrils.
it will be appreciated that various changes and modifications can be made from the specific details of the invention as described above without departing from the spirit and scope thereof as defined in the attached claims.

Claims (6)

CLAIMS:

1. A glass fiber separator material comprising a mass of intermeshed glass fibers substantially all of which have a fiber diameter not greater than about 20 µm, and at least percent w/w of which have a fiber diameter less than 1 µm, and, distributed through the glass fibers, from 0.2 percent w/w to 20 percent w/w cellulose pulp beaten to a Canadian standard freeness not greater than 120 cc.

2. A glass fiber separator material as claimed in claim 1 wherein cellulose fibrils in the cellulose pulp are impregnated with a solidified, synthetic resin.

3. A glass fiber separator material as claimed in claim 1 wherein cellulose fibrils in the cellulose pulp are from a slurry which has a Canadian freeness not greater than 100 cc.

4. A glass fiber separator material as claimed in claim 1 wherein glass fiber separator material has two opposed major surfaces and cellulose fibrils from the cellulose pulp which are adjacent one of the two opposed major surfaces are impregnated with a solidified, synthetic resin, while cellulose fibrils from the cellulose pulp which are adjacent the other of the two opposed major surfaces are not so impregnated.

5. A glass fiber separator material as claimed in claim 1 wherein there are also hydrophobic synthetic fibers in the mass of glass fibers, the hydrophobic synthetic fibers are intermeshed with the grass fibers and with one another, and the size distribution of the glass fibers and the proportions of glass and hydrophobic synthetic fibers are such that the separator has an absorbency for a sulfuric acid electrolyte of from 75 percent v/v to 95 percent v/v.

6. A sealed lead/sulfuric acid recombinant storage battery comprising a plurality of lead plates in a closed case, a glass fiber separator material as claimed in claim 1 between adjacent ones of said plates, and a body of a sulfuric acid electrolyte absorbed by said glass fiber separator material and maintained in contact with each of the adjacent ones of said plates.