Hollow Microspheres of Silica Glass and Method of Manufacture

The present invention relates to manufacture of hollow glass microspheres, and more particularly to manufacture of "high-silica" glass microspheres from a gel and to the resulting product.

It has been found desirable in the art of manufacturing hollow glass microspheres or shells to provide microspheres constructed essentially of a single oxide, preferably silica. Such single oxide shells avoid many potential problems associated with multiple component glass shells, such as inhomogeneities and phase separation. Essentially pure silica shells are particularly desirable for their high strength and resistance to devitrification. Moreover, essentially pure silica microspheres are considered desirable for use as inertial confinement fusion targets because of controllable and predictable gas permeation characteristics, and because of the high softening temperatures characteristic of silica. However, the high softening temperature of pure silica (1667.degree. C.) presents a major difficulty in attempts to manufacture pure or essentially pure silica microspheres.

In the manufacture of silica microspheres from a glass gel, boron is often added to the gel for lowering the viscosity and, therefore, the softening temperature of the dried and crushed gel. However, the resulting glass microspheres possess a substantial boron concentration. Concerning the manufacture of hollow glass microspheres from gels and a gel powder, reference may be had to the following art: the U.S. patents to Veatch et al Nos. 3,030,215, Beck et al 3,365,315, and Budrick et al 4,017,290 and 4,021,253; Souers et al, "Fabrication of the Glass Microballoon Laser Target", UCRL-51609, Sept. 26, 1974; and 1977 Annual Report on Laser Fusion Research, KMS Fusion, Inc., pages 1-12 to 1-15.

A general object of the present invention is to provide a method of manufacturing hollow glass microspheres of essentially pure silica glass. More specifically, an object of the invention is to provide a method of manufacturing hollow glass microspheres composed of at least 99% silica.

Another object of the invention is to provide a gel powder suitable for manufacture of essentially pure silica glass microspheres. A further and related object of the invention is to provide a method of gel formation suitable for manufacture of such frit.

Yet another object of the invention is to provide hollow glass microspheres composed of at least 99% silica.

In accordance with the invention it has been found that introduction of alkali metals along with boron in the glass gel and powder, specifically potassium and sodium, improves the viscosity/temperature forming properties of the frit, but yields glass microspheres having high concentrations of silica. Indeed, it has been found that, whereas addition of either alkali metals or boron alone to the gel and gel powder yields microspheres with substantial alkali or boron concentrations, addition of both results in substantially complete volatilization of the additives in the forming process to yield a microsphere of more than 99% silica glass. It has also been discovered that formation of the shells in the presence of water vapor materially assists and enhances such additive volatilization.

In the following description and claims, the terms "microspheres" and "shells" are used synonomously and refer to hollow discrete substantially spherical structures having a finite wall thickness. Unless otherwise indicated, all concentrations are expressed in mole percent and all concentration ratios are mole ratios. Concentrations of glass-forming components of the glass gels are expressed in oxide equivalent mole percent. The terms "gel" and "gel powder" are employed in their usual meanings, the term "gel powder" referring specifically to a particle formed by crushing and drying a metal organic gel containing oxidizable metallic glass-forming components and a blowing agent. (The above-referenced Budrick et al Pat. No. 4,021,253 employs the term "frit" for what is termed "gel powder" herein).

The invention, together with additional objects, features and advantages thereof, will be best understood from the following description, the appended claims and the accompanying drawings in which:

FIG. 1 is a semi-schematic elevational view of an oven for forming glass microspheres from glass frit and is not to scale;

FIG. 2 is a visual photomicrograph at 39X of one example of glass microspheres discussed hereinafter;

FIG. 3 is an interference photomicrograph of the shells of FIG. 2 at 39X; and

FIGS. 4-13 are photomicrographs at 39X of further shell samples to be discussed and alternately correspond to FIGS. 2 and 3.

A number of different glass gel powder and shell samples have been prepared and tested to demonstrate the invention. Two methods of gel preparation were employed:

Gel Method I: Alkaline metals were combined in ethanol under argon, with the amount of ethanol being sufficient to dissolve the alkali metals. The quantities of metals depended upon the desired concentrations of each in the resulting powder. Tetraethoxysilane in an amount sufficient to yield desired powder silica concentration was then added after the metals were fully dissolved. A mixture of 50% ethanol and 50% water (by volume) was then added dropwise to the solution until the result approached gel viscosity. Triethoxyborane was then added in an amount calculated to yield a desired boron concentration in the powder and the solution was allowed to gel. The gel was heated to 55.degree. C. and held for several hours (overnight), and then vacuum dried at 100.degree. C. The gel was then crushed and sieved to yield a gel powder of desired particle size.

Gel Method II: Tetraethoxysilane, ethanol, water and hydrochloric acid were combined and heated to 70.degree. C. The mass ratio of tetraethoxysilane to ethanol was about 2.0 g/ml, with the amount of tetraethoxysilane being such as to yield desired powder silica concentration. The water was in a 1/1 mole ratio to the tetraethoxysilane. The hydrochloric acid was to act as a catalyst and the amount is not critical. A second mixture of potassium and sodium acetates, water and acetic acid was combined to form a clear solution, with the ratio of water to acetic acid being 2/1 by volume. Again, the amounts of potassium and sodium acetates depend upon desired powder concentrations. The two solutions were then combined while stirring. Triethoxyborane was then added and the solution spontaneously gelled. The gel was then dried, crushed and sieved as in Gel Method I.

It will be appreciated that the above-described gels and powders contain oxidizable glass-forming components (silicon, boron, potassium and sodium), and solvents and materials (which may include one or more of the solvents) which will operate as blowing agents. The following table illustrates concentrations of the glass-forming components in terms of oxide-equivalence for five powder specimens tested to demonstrate the invention. (The use of oxide-equivalence when expressing concentrations of the glass-forming components permits expression of such concentrations relative to each other without regard to the other materials in the powder.)

Glass microspheres were formed from the above frit by dropping the powder particles into a tower furnace of the type illustrated in FIG. 1. The furnace 10 of FIG. 1 is similar to that shown in the above-identified Budrick et al U.S. Pat. No. 4,017,290 and comprises a vertical three-section ceramic pipe 12 having a 7.6 cm diameter and an overall high of 3.9 meters. Pipe 12 is heated uniformly by ninety-six silicon carbide heating elements 14. Furnace 10 is divided into six separately controlled and monitored sections 18-28 each having three thermocouples 30 connected to suitable temperature control and recording circuitry (not shown). The overall furnace is encased in a layer of high temperature insulation 32.

The lower end of pipe 12 is submerged in a container 16 of distilled water which serves to seal the bottom of tube 12 and also to provide water vapor in the furnace atmosphere. Some water vapor is considered essential to facilitate the alkali and boron losses required to achieve high silica glass shells in accordance with the invention. The mechanism of the relationship between humidity and shell formation is not fully understood at the present time. However, it has been found that a water temperature of about 60.degree. C. yields adequate humidity. Higher water temperatures and consequently higher water vapor pressures may be employed. Gel powder is fed into the furnace tube at the top either manually, as by a spatula, or by a motorized powder feeder. The powder and resultant shells fall by gravity through the air/water vapor atmosphere of the furnace pipe and are collected in a container of water.

In formation of each of the shell samples set forth in the following table, chamber temperature was held at 1500.degree. C. and a powder particle size in the range of 106-125 .mu.m was utilized:

Observation of FIGS. 2-13 demonstrates that shell quality increases with powder oxide-equivalent potassium concentration at least in the range of 3.3% to 11.98%. Shell samples A and B, as shown in visual photomicrograph FIGS. 2 and 4, comprised poorer quality microspheres and, for this reason, were not analyzed for final composition. It will be observed, however, in interference photomicrograph FIGS. 3 and 5 that wall uniformity, shown by the degree of concentricity between the interference rings, increased from shell sample A to shell sample B with increasing potassium and sodium concentrations in the gel powders (Table 1).

Powders C produced good quality shells with reasonable wall uniformity (FIGS. 6 and 7) but resulted in lower than desired silica concentration (Table 2). Powders D and E also produced good quality shells with reasonable wall uniformity (FIGS. 8-13) and essentially pure silica shells--i.e. having a silica concentration of more than 99%. Shell sample D' (FIGS. 12 and 13) was run to check repeatability using powder D, which is presently preferred for overall quality and composition. Shell sample D was additionally subjected to inspection by contact microradiography as described in Henderson et al., "Microradiographic Characterization of Laser Fusion Pellets", Rev. Sci. Instrum., 48, 835-840 (July 1977) which showed an average wall non-uniformity of the batch as about 26%.

Shell sample C having the highest powder sodium concentration possesses a significantly lower silica concentration than do shell samples D, D' and E (although such silica concentration in shell sample C is higher than now commercially available in glass microspheres). Thus, it appears from a comparison of shell samples C-E that sodium concentration in the starting powder should vary as a direct function of potassium concentration to obtain essentially pure silica shells of good quality. A potassium/sodium oxide equivalent mole percent ratio in the powder of at least 1.5 appears desirable. A ratio of between 4 and 5 is preferred, as with the ratio of 4.4 in powders D and E. The total alkali oxide concentration (sodium oxide plus potassium oxide) should exceed about 8%, as with powders C, D and E. In particular, total alkali oxide concentration of between 14% and 15% (powders D and E) is preferred at the present time. It is believed that a minimum of about 1% boron in the gel powder is required to produce high silica shells of acceptable quality. Powder compositions having more than 20% boron concentration would probably not produce quality shells having more than 99% silica under any reasonable conditions. Rubidium and cesium are believed to be suitable substituted for potassium.

The exact mechanism by means of which the alkali metals not only themselves volatilize, particularly in the presence of water vapor, but also promote volatilization of the boron is not fully understood. Thus, although trials with gel D at substantially larger powder sizes (330-355 .mu.m sieve cut) at an oven temperature of 1500.degree. C. produced shells with only 93% silica concentration, it is believed that variation of other conditions, such as increasing oven temperature, may permit manufacture of larger essentially pure silica shells. Lowering of oven temperature to 1250.degree. C. using gel D and gel particles in the range of 106-125 .mu.m was also found to decrease ultimate silica concentration to 95%. Again, however, it may be possible to compensate for lower oven temperature, such as by decreasing particle size.

Thus, the invention in its broadest envisions a boron concentration (oxide equivalent) in the starting gel powder in the range of about 1% to 20%, with an intermediate range of about 4% to 10% and particularly about 6% to 7% being preferred; a potassium concentration in the range of about 5% to 25%, with the intermediate range of about 8% to 15% and particularly about 12% being preferred; a sodium concentration of less than about 10%, with an intermediate range of about 1% to 6% and particularly about 2.5% to 3% being preferred; a potassium-to-sodium ratio greater than 1.5, preferably between 4 and 5, and particularly about 4.4; and the remainder consisting essentially of silicon.

Although the invention has been disclosed with particular reference to manufacture of glass microspheres for inertial confinement fusion targets, other potential applications are envisioned. For example, glass microspheres may be used for building insulation or as a lightweight filler in many molded articles of manufacture. For a further discussion of the application of glass microsphere technology, see Wehrenberg et al., "Shedding Pounds in Plastics: Microspheres are Moving", Mechanical Engineering, October 1978, pages 58-63.