Summer Undergraduate Research Fellowships 2003

9/31/03

Summary:

This project was designed to study the effect of additives on the crystallization behavior of E-glass under a specific heat treatment. This was intended to determine the feasibility of recycling of the millions of tons of waste E-glass produced every year into useable products, with a minimum of cost and effort. E-glass samples were obtained and crushed, the additions were made, and the samples were melted, annealed, and cut into standardized samples. The samples were heat treated, allowed to cool, and then mounted in epoxy and polished. The thickness of the crystallization layer was measured using an optical microscope. It was determined that this material is very difficult to crystallize using this heat treatment program, and that the only effective way to do so is to shift the composition of the material using additions of Al₂O₃ and CaO.

Introduction:

In 1999, approximately 3.5 billion pounds of glass fiber were produced worldwide. E-glass was the most commonof all the types of fiber produced⁽¹⁾. There has been growing interest in researching new ways of recycling this material. This project investigated the feasibility of recycling waste E-glass by transforming it into a glass-ceramic material, which could then be used in other applications. The process developed must minimize the cost and time required for the waste material to undergo the transformation. This means that the amount of additives used, and the cost of those additives must be kept as low as possible, and the equipment used should be equipment that would already be found in glass-fiber manufacturing plants.

Glass-ceramic materials are solid materials with a crystalline structure obtained by the controlled devitrification of glasses. To make a glass-ceramic product, a glass is melted, formed into the desired shape, and heat treated to allow the molecules in the glass to connect and order themselves into a crystalline structure. When done properly, the end result is a product with a crystalline microstructure with little to no voids, microcracking, or porosity⁽²⁾ (reducing these flaws is often the most difficult part of processing ceramic materials).

Glasses are amorphous materials, or materials with no ordered molecular structure. Materials crystallize by nucleation and crystal growth. In the nucleation stage, many small nuclei form, but do not grow. In the growth stage, the previously formed nuclei grow larger and larger, until the material can no longer form crystals of the growing phase.

It would be desirable to transform the waste E-glass into glass ceramic for several reasons. Interest in glass-ceramic materials has been increasing as manufacturers realize the beneficial properties of glass-ceramics. The main benefit of glass-ceramics is that they are easily made and shaped like traditional glass products, as well as having many of the desirable properties of ceramic materials (high strength, high resistance to thermal shock, hardness, etc.). Other properties of glass-ceramics, such as thermal expansion, can be varied over a large range (from negative to positive)⁽³⁾, making them ideal for many applications. With these properties, it is now possible for manufacturers to produce products which before were previously not possible, as complex shapes are difficult to form from ceramics, but can be formed from glasses, crystallization produces properties outside the limitations of traditional glass products, with values in the range of many ceramic products.

Owens-Corning Fiberglass (OCF) E-glass was selected since it makes up a significant portion of the glass fiber produced in the U.S. today ⁽¹⁾. OCF E-glass is a calcium-alumino-silicate (CAS) glass, with small amounts of magnesium, iron, boron, titanium, and sodium oxides. OCF E-glass is known for having a high tensile strength even at high temperatures (hence its primary application in fiber-reinforced composites) and a high chemical resistance, especially to H₂O.

Very little research has been done on the crystallization of CAS glasses. R.G Duan and K.M. Liang ^{(4, 5)}studied the crystallization of the entire system in much more depth at the University of Beijing in 1997. They did not study any commercial glasses.

The effect of additives on the crystal layer thickness was studied in this project. Additions included TiO₂ which was used as a nucleating agent, Li₂CO₃ which was added in order to de-stabilize the glass, and CaO and Al₂O₃ which were added to shift the composition of the glass into different crystal-forming regions of the phase diagram.

Methods:

In order to crystallize the original glass into the desired glass-ceramic product, the samples were heat treated at 1000 ºC for times ranging from 1 to 2 hours. The thickness of the crystal layer was then measured.

OCF E-glass was obtained from OCF in the form of 25 – 30g ‘marbles’, which were placed in a Pt crucible for melting. The glass and crucible were placed into an electrically heated furnace and held at 1500 ºC for 30 minutes. After 30 minutes, the crucible was removed and air cooled to room temperature. If additions were necessary, the glass was mechanically crushed into a fine powder, mixed with the desired amount of additives, remelted for 30 minutes and allowed to air cool again. The glass transition temperature (T_g) of the glass was determined in a differential scanning calorimeter (DSC). Once the T_g was known, the sample was annealed in a programmable furnace. Samples were heated to within 5 ºC of their respective T_g’s at a rate of 5 K/minute, held for 30 minutes, and cooled at a rate of 3 K/ minute to room temperature. Once cooled, the glass was cut into samples approximately 1.5 mm thick. The samples were placed on a piece of Pt foil, and heat treated at 1000ºC in the melting furnace for a specified amount of time (typically 15 – 180 minutes). Once heat treated, the sample was removed and allowed to air cool to room temperature. The sample was then vertically mounted in fast-setting 2-part epoxy and polished to 1200 grit. Once polished, the thickness of the crystal layer was measured using an optical microscope with an eyepiece scale. The thickness of the sample layer was recorded.

Results:

The results of the heat treatments with different additives are listed in Table I.

Table I. Crystal Layer Thickness

As is shown in the table, the base glass does not crystallize on its own, using this heat treatment. It can also be seen that the only additives with any promise are 20% CaO and 10% Al₂O₃.

Discussion:

It was found that E-glass is very difficult to crystallize. The glass itself is very resistant to crystallization and will not nucleate from the bulk of the material with any of the additives or heat treatments used in this study. The addition a of nucleating agent (TiO₂) did not improve crystallization noticeably. Addition of Li₂O₃ to destabilize the glass⁽⁶⁾ had an only a moderate effect on the growth of crystals in the sample.

In order to determine the stabilizing effect of the B₂O₃ on the material, a simplified E-glass composition was made, with the formula 26.7 CaO- 13.3 Al₂O₃ – 60SiO₂. This glass did not crystallize, demonstrating that the stabilizing effect of the B₂O₃ is negligible.

Addition of Al₂O₃ or CaO causes the glass to crystallize. Addition of these oxides shifted the composition of the material on the phase diagram, causing a shift in the composition of the crystals formed. The CaO additions were the most successful. Increasing the temperature of the heat treatments to 1050 ºC had a more significant effect on increasing the amount of crystallization than increasing the length of the heat treatment (these results are qualitative). This result agrees with the previous studies of the E-glass crystallization ^{(4, 5, 6)}.

Conclusions:

E-glass is very difficult to crystallize. The effects of nucleating agents such as TiO2 are not significant. The only way to get the material to crystallize in an acceptable manner is to shift the composition of the crystalline phase by adding either Al₂O₃ or CaO (the latter being preferred). In future studies, the compositions which produced more desirable results could be studied further, with heat treatments at different temperatures and for different times.

References:

1. http://www.energy.ca.gov/process/pubs/composites.pdf

2. Boyd, D.C., and MacDowell, J.F. (eds.). Advances in Ceramics, Vol. 18: Commercial Glasses. The American Ceramic Society, Inc., Columbus, OH, pp. 157-176, 1986.

3. Prindle, W. R., Danielson, P. S., and Malmendier, J. W. “Glass Processing”, in

Engineered Materials Handbook: Ceramics and Glasses, Vol.4. ASM

International, Materials Park, OH, pp. 377-394, 1991.

4. Duan, R.G. and Liang, K.M. “A Study on the crystallization of CaO-Al₂O₃-SiO₂

system glasses”. Journal of Materials Processing Technology, vol. 75, iss. 1-3,

pp. 235-239, March, 1998.

5. Duan, R.G., Liang, K. M., and Gu, S. R. “A Study on the Mechanism of Crystal

Growth in the Process of Crystallization of Glasses.” Materials Research

Bulletin, vol. 33, number 8, pp. 1143-1149, 1998.

6. Barbieri, L., Corradi, A., Leonelli, C., Siligardi, C., Manfredini, T., and Pellacani, G.

Effect of TiO₂ Addition of the Properties of Complex Aluminosilicate Glasses

and Glass-Ceramics. Materials Research Bulletin, vol. 32, number 6, pp. 637-

648, 1997.