US3141790A

US3141790A - Burned basic refractory shapes

Info

Publication number: US3141790A
Authority: US
Publication date: 1964-07-21
Anticipated expiration: 1981-07-21

Description

y 21, 1964 B. DAVIES ETAL BURNED BASIC REFRACTORY SHAPES Filed Jan. 25, 1963 [IF/VEST P404 4V5 PETER HARRY HAVRl/VEK W f M ATTORNEY United States Patent poration of Pennsylvania Filed Jan. 25, 1963, Ser. No. 254,013 9 (liaims. (Cl. 117-123) This invention relates to furnace structure for the oxygen steelmaking process. More particularly, this invention relates to the working lining of oxygen steelmaking vessels.

In the oxygen steelmaking processes, which have been variously designated as the LD process, basic oxygen furnace process, oxygen converter process, etc., the furnace structure normally consists of a metal shell having a refractory lining therein. The lining for oxygen vessels has heretofore generally consisted of an inner or working lining of tar bonded, chemically bonded, or burned basic brick, an exterior or shell lining of burned magne site brick, and usually an intermediate layer of a tar bonded ramming mix of composition similar to that of the working lining. This invention is primarily concerned with the working lining, and a composition suitable for use therein. In a particular aspect, the invention relates to improved lining material for the cone section and a slag splash zone of an oxygen converter lining.

Refractory shapes may be termed chemically bonded or ceramically bonded. The chemically bonded shapes are green or unfired shapes, and are put into service in this form. ceramically bonded shapes are shapes which have been subjected to elevated burning or firing temperatures, to obtain a ceramic bond throughout the particulate material from which the shape is fabricated. This invention relates particularly to the latter.

In both chemically and ceramically bonded shapes, the bond itself is probably one of the most important properties. In order to obtain a shape which is easily handled and which stands up in service without spalling, peeling or breaking away, the bond must be strong and relatively stable. For certain uses, high density and low porosity are equally important properties. These latter properties are desirable in a shape, to resist penetration by corrosive metallurgical slags and fumes which tend to destroy a shape in service.

Typically, the lining of an oxygen converter has been made of basic refractories, because the slag of the oxygen steelmaking process is, itself, basic. Prior refractories most commonly used are those consisting essentially of tar impregnated or tar bonded, dead burned magnesite, dead burned dolomite, and sometimes lime mixed with the foregoing. Since this invention relates particularly to burned or ceramically bonded shapes, it is most closely related to the art of tar impregnated basic refractories. An example of a good tar impregnated dead burned magnesite refractory is one made according to the teachings of United States application, Serial No. 113,094, Davies et al., filed May 29, 1961, entitled Burned Refractory Product, now United States Patent No. 3,106,475, and owned by the same assignee as the instant invention. The shapes of the invention of the co-pending application are particularly adapted to manufacture by conventional techniques using commercially "ice available materials. One of their outstanding properties is resistance to impact, abrasion and spalling. These brick are characterized by good strength at room temperature and by very good compressive strength at elevated temperatures, as indicated by resistance to subsidence in a load test at elevated temperatures. However, in service, ceramically bonded brick are subjected to tensile stress due to expansion and the like. And, while the refractories of application Serial No. 113,094 have high compressive strength, they do not resist tensile stresses as much as might be desired.

Accordingly, it is an object of this invention to provide ceramically bonded shapes having improved tensile strength, particularly suited for use in fabrication of the lining of oxygen converter vessels.

It is another object of this invention to provide refractory shapes having improved resistance to tensile stresses in service installations.

In one aspect, this invention is predicated on the discovery that controlled adulteration of very high purity dead burned magnesia grain (i.e., 97+% MgO, by weight, and on the basis of an oxide analysis) and use of this grain in the manufacture of a ceramically bonded shape with subsequent tar impregnation thereof, provides shapes giving superior service in an oxygen converter vessel. The R 0 content of the grain and the resulting shapes is critical. We have found it must be less than about 1%, by weight, of the grain. As known to those in the art, R 0 refers to Cr O Fe O and A1 0 Our so-calied adulterating ingredients are lime and silica. The silica must constitute less than about 1% of the grain and shapes, with at least 3 but not more than 4 times as much lime as silica expressed as CaO and SiO A better understanding and other features and advantages of our invention will become obvious to those skilled in the metallurgical and refractory arts, by a reading of the following detailed description with reference to specific examples and to the drawings. In these drawings:

FIG. 1 is a photornicograph of a portion of a ceramically bonded shape according to this invention; and

FIG. 2 is a photomicrograph of a previous ceramically bonded shape.

Workers in the art of refractories and related ceramic fields have long recognized that lime and silica can be balanced in a basic refractory, i.e. one fabricated of a magnesite or a dolomite, to good advantage. The United States Patent to Pitt, No. 2,245,297, the Patent to Mc- Caughey et al., No. 1,965,605, the Patent to Lee, No. 2,089,970, the Patent to Lee, No. 2,229,297, and others, are all indicative of prior work in this area. The basic goal of these workers appears to have been stabilization of the materials with which they worked (especially the lime). Also, they reported desirable improvement in various physical properties, including strength.

The ceramically bonded shapes according to the copending application, Serial No. 113,094, mentioned above, on infrequent occasion exhibit a distressing loss of strength at temperatures above 2000 F. This loss of hot strength is believed to be caused by the weakening of the very thin silicate films that appear to coat and bond the separate magnesia or periclase crystals which constitute the shapes. FIG. 2 is a photomicrograph (1000x magnification) of a shape according to the co-pending application, in which periclase crystals 10 are shown cemented in place and largely surrounded by monticellite 11. and merwinite 12 (both silicate minerals) films. The numeral 13 indicates voids.

More recently, we have come to realize that to develop the best tensile strength in a magnesite brick, there must be a direct attachment of magnesite or periclase crystal to periclase crystal without an intervening film of silicate.

In order to achieve this direct attachment, we made brick from the most pure commercially available refractory magnesite of which we know, having the analysis of grain AA shown in Table I, using conventional brickmaking and burning temperatures. Upon testing these brick for modulus of rupture at 2300 F., we found they were no stronger than those made of lower purity magnesite. Petrographic examination revealed that a silicate film still separated most of the separate magnesia or periclase crystals (similar to the showing of FIG. 2). In attempting to overcome this problem, we burned similar refractory shapes to a temperature of 3400 F. which far exceeds any commercial burning practice but this did not remove the silicate film nor improve the strength at 2300 F.

It became apparent that no physical processing was likely to remove this silicate film, because the lowest surface energy for the system was still a magnesium silicatemagnesia bond, rather than the desired magnesia-magnesia bond.

The studies of prior workers, such as those of the United States patents, above noted, and our own Work,

suggested that conversion of these silicate films to a dicalcium silicate would be useful.

In exploratory studies, we prepared a size graded batch of dead burned magnesite, to which we added a small amount of lime hydrate and a small amount of chromic oxide to stabilize any dicalciurn silicate against mineral inversion, if this silicate should form in burning. The magnesite used had the following chemical analysis.

Table 1 Grain AA:

SiO percent 0.9 A1 0 do 0.4 Fe O do 0.3 CaO do 1.1

MgO do 97.3 Ratio CaO/SiO 1.22/1

All parts by Weight, and on the basis of an oxide analysis.

The size grading of the magnesite grain was substantially as follows:

Table 11 Percent 4 on mesh 40 10 on 28 mesh -28 on 65 mesh 15 65 mesh to ball mill fines 30 To about 98.5 parts of the magnesite, size graded as indicated, we added 1.3 parts of lime hydrate (technical grade) and about 0.2 part chromic oxide (also technical grade). These materials were dry mixed for about five minutes and then for an additional five minutes while adding 2 parts lignin liquor and 2 parts water, based on the total weight of the mixture, as a tempering fluid. Brick were made from the batch by pressing on a power press at 8000 psi. and burned 4 high on edge to cone 23 (about 2820" F.). These brick were subjected to physical testing, the results of Which are set forth in Table Ill.

Table III Linear change in burning percent -0.1 Bulk density, p.c.f. (av. 20) do 174 Modulus of rupture, p.s.i. (av. 3)

At room temperature 1360 At 1500 F. (5 hr. hold) 930 At 2000 F. (5 hr. hold) 870 At 2300 F. (5 hr. hold) 400 Apparent porosity (av. 5) "percent" 20.1

Comparative batches of the same magnesite without the lime hydrate or chromic oxide additions are mixes A through D of Table V. The lime hydrate addition of the Table III mix obviously improved the tensile strength at elevated temperatures. This is readily apparent from comparing the modulus of rupture at 2300 F. for Table III with mixes A through D of Table V, which contained no added lime hydrate. However, the increased strength was still less than that desired. Microscopic studies of brick made from the mix to which the lime hydrate had been added still indicated a silicate filming, somewhat similar to that previously observed with brick of the type shown in FIG. 2 of the drawings.

We therefore concluded, that to change the characteristics of this system some additive had to be made to the brick mix. Thus, to brick batch mixtures of the type described in Tables I and II, 1% of the following oxides 'were added: A1 0 Cr O Fe O NiO, CaO, ZnO, CuO,

TiO (all technical grade and 325 mesh). Brick were pressed from these batches and burned at cone 23 (some of this work is reported in more detail below). None of these additions had any appreciable effect on the strength at 2300 F.

It was then decided to more fully explore the field of controlled lime additions to the magnesite grain-making batch, with attention to the relationship of lime content to the other oxides. We prepared an extended series of these magnesites, and one which proved to have extremely interesting properties Was the grain BB of Table IV. It was made by utilizing the same magnesium hydroxide slurry, which was the initial ingredient of grain AA. Additions of lime hydrate were made which, after caustic burning, gave an analysis typically as follows.

Table IV Grain BB:

MgO percent 96.3 SiO do 0.8 CaO do 2.9 F6203 .d0 0.3 A1 0 do 0.4 Ratio CaO/SiO 3.63/1

This caustic calcined mixture, with all free and substantially all chemically combined water removed, was formed into small briquettes at a pressure of about 20,000 psi, which briquettes were fed to a shaft kiln where they were dead burned at a temperature above about 3000 F. The resulting dead burned briquettes were crushed to formulate brickmaking grain. This grain, Whose analysis is typified by Table IV, was then used for fabrication of various brickmaking batches, as reported in Table V, mixes E, F, G and H. These batches were subjected to brick fabricating and burning techniques, substantially the same as described above, with reference to Table I. Table V sets forth in detail the composition of the various batches and the results of physical testing of brick made from the batches. The abbreviation BMF refers to ball milled fines. The brick were burned to cone 23 (2820 F.) with the top temperature being held for 10 hours.

Table V Grain Grain 100% Grain BB AAbut BB but Magncsite Grain 100% Grain AA (Co-burned with 020) with with BB in AAin BMF BMF Mix Letter A B G D E F G H I J Batch:

3+10 mesh 20 30 40 20 30 40 -4+10 mesh. 20 34 10 34 34 10+28mesh 20 20 20 28 20 20 20 2s 28 BMF (5060%325mesh) 3s 40 40 40 3s 40 40 40 38 38 Ligniu Liquor and water, about 50/50 weight ratio percent 5 5 5 5 5 5 5 5 5 5 Bulk Density, p.c.f. (Av. 10) 180 181 181 182 180 181 181 182 178 180 Modulus of Rupture, p.s .i.:

At Room Temperature, about 72 F.

(Av.3 2,450 2,370 2,190 2,200 2,150 2,570 2,740 2,760 2,150 2,300 At 2300F.(I-I0ld Time 5Hours Av. 3)-. 180 140 130 120 1040 0 3 500 860 Chemical Analysis of Batch:

Silica (SiOz) percent 0.8 0.8 0.8 0.8 0.8 0.8 0.3 0.8 0.8 0.8 Lime 03o do 0.9 0.0 0.9 0.9 2.9 2.0 2.9 2.0 1.7 2.1 Lime/Silica Ratio do 1.13 1 1.13 1 1.13/1 1.13 1 3.62/1 3. 02 1 3. 02 1 3.62/1 2.13 1 2.63/1

The results of the testing reported in Table V Were of the co-burned grain, having a balanced lime/silica completely unexpected. The brick fabricated entirely ratio of 3.62:1. The batch was substantially the same from the special co-burned grain (mixes E through H) as mix E of Table V. We added 1%, by weight, of with a lime/silica ratio of about 3.62 to 1 had a modulus technical grade chromic oxide (all 325 mesh) to the of rupture at 2300 F. over 4 and almost 5 times as batch, and manufactured it into brick according to the great as the comparative samples (A through D) not same techniques discussed with reference to Table Ill having the benefit of the lime addition before dead burnabove. The resulting burned brick had a modulus of ing of the magnesia. Mixes I and J indicated beneficial 3O rupture at 2300 of only 390 p.s.i. This indicated that characteristics were imparted when even a portion of the R 0 type materials might generally have a detrimental batch was fabricated of the co-burned grain with a careeffect on the brick of our invention. We, therefore, prefully balanced lime/silica ratio, with best results obtained pared two additional batches of the same grain, and when the special grain was included as the coarse fraction having the same size grading, to the first of which we (mix I added A1 0 (technical grade alumina) to total about 1% Microscopic examination of the brick fabricated entirefor the batch, and to the second of which we added ly from the special co-burned grain revealed a dramatic Fe O (pigment grade iron oxide) to tatal about 1% of change in appearance as compared to previous ceramthe batch. Brick fabricated of the batch having the ically bonded brick of this type, as typified by FIG. 2 alumina addition had a modulus of rupture of 280 (IOGOX magnification). FIG. 1 is a photomicrograph 40 psi. at 2300 F. Brick made from the batch with the Of a brick fabricated entirely Of the special co-burned iron oxide addition had a somewhat higher modulus of grain X magnificatiofll These l r brlck are rupture at 2300 F.; namely, 410 p.s.i. We, therefore, chflfac'ieflled y direct P6116135e t0 peflclas? crystal concluded that the total amount of R 0 materials in gram attachment 1, large Wh areas 15 the grain and brick must be maintained at a level below are the periclase grains. The lime and silica appear to 1%, by weight In small generally k i Pools or 5 In our extensive work in the field of this invention, we posits -n l iif g i to sometimes find that if all of the other factors are suitably i i iii v iidg a tnca mum S1 mate 6 mebu M are s confined, control of the ratio of (1210 to $10 is an adapt- The results of Table V are even more surprising when 2:3; 222 if i i i i g g zi gj ig compared to the results of Table III above, in which lime t b 3/1 6 Zia/opera e a ff a i 2 hydrate was added to a size graded dead burned magra etwger} an m makPlg nesite batch to provide an overall lime/silica ratio of that f f g non'magl}esia crystefnmemgledlent about 2 in the batch. In further studies, co-burned grain 1S Lncalclum slhijace 2 y at first was manufactured, in which the lime/silica ratio was appeaf PuZZlmg i a 1'fl01BC 1l1ar 8, t s mineral about 4.16 to 1. Under identical testing conditions, it 01 5 C210 and S10 111 the ratio of only 2.8/1. was found that a lime/silica ratio of this magnitude and In working in this field, we find it convenient to apslightly higher is satisfactory, but such ratios undesirably praise our mixes in terms of the ratio of CaO to Si0 of complicate manufacturing. Also greater amounts of lim the various compositions. If a magncsite should be made inC i116 hydfationtend'ency Of p made from 5119b available, in which all of the Cat) and Si0 were present grain. The entire series of tests established to our satrsas dicalcium Silicate (29195502), m C310 /Si02 ratio faction that the Optimum range for thellme/slhca f would be 1.87/1, which is calculated from the molecular fehould be 3 4 t0 4 to 1 i weights; if present soley as tricalcium' silicate is somewhat flexible, with satisfactory product resulting with a ratio as high as 4.5 to 1. (3CaO-SiO Since the pooling of the silicate was so unexpected and I different from the previous silicate filming we had ob- 0210/3102. ratio be 2-8/1; P served, it appeared likely the presence of other oxides y as WOHfIIStOF-UTB 2) the F 2 Iatlo Would in the brick mix might affect the silicate surface tension be if a magfifislte Should Contain more C210 than the and give Still better Properties to the brick 7O 2.8/1 ratio of tricalcium silicate, the excess CaO would For example, We expected that a ll amount f necessarily be present as some mineral other than calcium chromic oxide, such as that added to the batch Which insilicates, since there is no calcium silicate more calcareous cluded the lime hydrate reported in Table III, would than tricalcium silicate. Such excess CaO is thought of as provide even greater strength for shapes according to existing as ferrites, or aluminates (if the requisite Fe O this invention. We, therefore, prepared another batch and A1 0 are present) or as free lime.

The co-burned grain BB of Table IV is illustrative. Its CaO/SiO ratio was 3.63/1, which is considerably above the 2.8/1 of tricalcium silicate. Nevertheless, in the microscopic examinations no silicate other than tricalcium silicate and free (3210 was unobserved. In our experience, as we proceed upwards to the neighborhood of 4/1 ratios, free lime can be expected to make its appearance. A small amount can be tolerated but, for reasons mentioned elsewhere, free lime in the grain or finished brick introduces difficulties, which can become prohibitive in nature.

The foregoing discussion of experimental data, demonstrates the superiority in physical properties and, in particular, the superiority in hot strength as measured by' modulus of rupture at 2300 F, for brick made in accordance with our invention. We also demonstrated that they exhibited adequate resistance to thermal shock, as evidenced by 0.0% spalling loss when subjected to ASTM C-122 thermal shock tests. The combination of thermal shock resistance and hot strength renders them particularly suitable in locations in an oxygen process vessel, where mechanical wear and rapid temperature changes are present; namely, in the cone and slag splash zones of oxygen converter vessels. Of course, they are tar impregnated for use in these vessels. We suggest between 4 and 10%, by weight, of tar in the brick as being most suitable. Between 6 and 8% tar is best.

A group of burned brick, fabricated according to this invention, are giving outstanding service in oxygen converter vessels. The brick are impregnated with cokable, non-aquesous, carbonaceous materials such as tar and pitch. They were impregnated by immersion in the carbonaceous material, generally heated to about 400 F.

The heated and liquified carbonaceous material easily penetrates throughout the brick and, in cross section, the brick show remarkably uniform distribution of the carbonaceous material, including a thin layer completely about exterior surfaces. Commercially available tar or pitch of petroleum or coal base can be used for impregnation. Generally, we suggest a coal based pitch, having a softening point on the order of 150 F.

Unless otherwise stated, all chemical analyses are on the basis of an oxide analysis, in conformity with the common practice of reporting the chemical analyses of refractory materials. All size grading is according to the Tyler mesh series. All parts and percentages are by weight. The chemical analyses of raw materials, which are reported, should be considered typical. However, the lime, silica and R content, reported in the various analyses, should be considered accurate to the first decimal place.

Having thus described the invention in detail and with suflicient particularity as to enable those skilled in the art to practice it, what is desired to have protected by Letters Patent is set forth in the following claims.

We claim:

1. Tar impregnated, ceramically bonded basic refractory shapes suitable for use in metallurgical vessels, said shapes, by weight and on the basis of an oxide analysis, consisting essentially of from about 96 to 97% MgO, there being Ca() and SiO in a weight ratio between 3 to 1 and about 4 to 1, and materials of the group C1203, A1 0 and Fe O in an amount not exceeding 1%, said shapes microscopically characterized by periclase to periclase crystal attachment and with the Ca() and SiO content largely present in spaced disconnected pockets between the periclase crystals and being mineralogically characterized largely as tricalcium silicate.

2. Tar impregnated ceramically bonded basic refractory shapes suitable for use in metallurgical vessels, said shapes made from a batch of size graded dead burned magnesia grain, at least a portion of the grain by Weight and on the basis of an oxide analysis, consisting essentially 8 of fromrabout 96 to 97% MgO, there being CaO and SiO in a weight ratio between 3 to 1 and about 4 to 1, and materials of the group Cr O A1 0 and Fe O in an amount not exceeding 1%, and being made by coburning the MgO, Cat) and Si0 ingredients thereof.

3. Ceramically bonded basic refractory shapes suitable for use in metallurgical vessels, said shapes, by weight and on the basis of an oxide analysis, consisting essentially of from about 96 to 97% MgO, there being CaO and SiO in a weight ratio between 3 to 1 and about 4 to l, and materials of the group Cr O A1 0 and Fe O in an amount not exceeding 1%, said shapes microscopically characterized by periclase to periclase crystal attachment and with the CaO and SiO content largely present in spaced disconnected pockets between the periclase crystals and being mineralogically characterized largely as tricalcium silicate.

4. Ceramically bonded basic refractory shapes suitable for use in metallurgical vessels, said shapes made from a batch of size graded dead burned magnesia grain, at least a portion of the grain by weight and on the basis of an oxide analysis, consisting essentially of from about 96 to about 97% MgO, there being CaO and SiO in a weight ratio between 3 to 1 and about 4 to 1, and materials of the group Cr O A1 0 and Fe O in an amount not exceeding 1%, and being made by co-burning the MgO, CaO and SiO ingredients thereof.

5. Tar impregnated, ceramically bonded basic refractory shapes suitable for use in metallurgical vessels, said shapes, by weight anl on the basis of an oxide analysis, consisting essentially of from about 0.8 to no more than about 1% SiO CaO in an amount sufiicient to obtain a CaO to SiO weight ratio between about 3 to 1 and about 4 to 1, the remainder consisting essentially of MgO and materials of the group Fe O Cr O and A1 0 said latter materials constituting less than about 1% of the total weight of the batch.

6. Tar impregnated ceramically bonded basic refractory shapes suitable for use in metallurgical vessels, said shapes made from a batch of size graded dead burned -magnesia grain, at least a portion of the grain, by weight and on the basis of an oxide analysis, consisting essentially of from about 0.8 to no more than about 1% SiO CaO in an amount sufiicient to obtain 9. C210 to SiO weight ratio between about 3 to 1 and about 4 to 1, the remainder consisting essentially of MgO and materials of the group F3203, Cr O and A1 0 said latter materials constituting less than about 1% of the total weight of the grain which constitutes at least a portion of the batch.

7. Ceramically bonded basic refractory shapes suitable for use in metallurgical vessels, said shapes, by weight and on the basis of an oxide analysis, consisting essentially of from about 0.8 to no more than about 1% SiO CaO in an amount suflicient to obtain a CaO to SiO weight ratio between about 3 to 1 and about 4 to 1, the remainder consisting essentially of MgO and materials of the group Fe O Cr O and A1 0 said latter materials constituting less than about 1% of the total weight of the batch.

8. Ceramically bonded basic refractory shapes made rom a batch of size graded refractory grain consisting essentially of from about 0.8 to no more than about 1% SiO CaO in an amount suificient to obtain a Cat) to SiO weight ratio between about 3 to 1 and about 4 to 1, the remainder consisting essentially of MgO and materials of the group Fe O Cr O and A1 0 said latter materials constituting less than about 1% of the total weight of the batch.

9. Ceramically bonded basic refractory shapes suitable for use in metallurgical vessels, said shapes made from a batch of size graded dead burned magnesia grain, at least a portion of the grain, by weight and on the basis of an l0 oxide analysis, consisting essentially of from about 0.8 to References Cited in the file of this patent no more than about SiO C30 in an amount Sulfit t clen to obtaln a CaO to S10 weight who between abou 3,106,475 Davies et a1 Oct 8, 1963 3 to l and about 4 to 1, the remainder consisting essentially of MgO and materials of the group Fe O Cr O 5 OTHER REFERENCES and A1 0 said latter materials constituting less than Holt 1: Tar Bonds Oxygen Vessel Bricks, in Steel,

about 1% of the total Weight of the batch vol. 143, pp. 7478, July 7, 1953, Ts 300 174, 266-43.

Patent No, 141,790 July 21,, 1964 Ben Davies et all.

It is hereby certified that err ent requiring correction and that th corrected below or appears in the above numbered pate said Letters Patent should read as Column 9, line 7,

after "weight" inser which constitutes at 16 t of the grain ast a portion Signed and sealed this 12th day of January 1965 (SEAL) Attest:

ERNEST W. SWIDER' EDWARD J. BRENNER Attesting Officer Commissioner of Patents

Claims

1. TAR IMPREGNATED CERAMICALLY BONDED BASIC REFRACTORY SHAPES SUITABLE FOR USE IN METALLURGICAL VESSELS, SAID SHAPES, BY WEIGHT AND ON THE BASIS OF AN OXIDE ANALYSIS, CONSISTING ESSENTIALLY OF FROM ABOUT 96 TO 97% MGO, THERE BEING CAO AND SIO2 IN A WEIGHT RATIO BETWEEN 3 TO 1 AND ABOUT 4 TO 1, AND MATERIAL OF THE GROUP CR2O3, AL2O3, AND FE2O3, IN AN AMOUNT NOT EXCEEDING 1%, SAID SHAPES MICROSCOPICALLY CHARACTERIZED BY PERICLASE TO PERICLASE CRYSTAL ATTACHMENT AND WITH THE CAO AND SIO2 CONTENT LARGELY PRESENT IN SPACED DISCONNECTED POCKETS BE-