Adaptation of Existing Raman Spectrometer for In Situ, Real-time Composition Analysis of Gaseous Bubbles in Glass Melts, 20-9144

Printer Friendly Version

Principal Investigators
Vijay Jain (Todd Mintz)
Darrell S. Dunn

Inclusive Dates: 6/25/99 - 10/25/00

Background - Fewer defects are tolerated in glass products because of increasingly stringent quality standards. Many defects in glass products are generated in the form of bubbles. The most dominant processes for the formation of bubbles in glass melts can be attributed to one or more processes such as gases originally introduced physically in the batch (porosity), gases released during decomposition of batch materials (carbonate salts), gases released due to reactions at the glass-refractory interface, and gases released due to redox conditions in the glass (reboil). The determination of gas components within bubbles is a useful approach for diagnosing a fining problem or identifying the origin of a specific defect. In addition, verification and validation of mathematical models for the identification of defects require experimental data on gas bubble composition and size. Compositional analyses of gas bubbles have been performed using gas chromatography and mass spectrometry. Both techniques require extensive set up, sample preparation time, and are not applicable for analyzing all gases. In this project, an existing fiber optic Raman spectrometer with 532-nm laser was used for collecting high-temperature spectra and analyzing gas constituents in the bubbles.

Approach - The main advantages of Raman spectroscopy as an in situ analytical tool are its ability to be used remotely with long lead optical fibers and its ability to probe complex, corrosive liquid and gaseous environments to obtain quantitative information regarding molecular or ionic composition. While Raman spectroscopy cannot distinguish all gaseous species, the separation in Raman peaks for O2, N2, CO, CO2, SO2, and H2 are sufficient for differentiation. Initially, gas bubbles were introduced in glass and were analyzed at room temperature. Following the initial determination of maximum temperature at which Raman spectra can be collected without black body radiation, gas bubbles were introduced into glass melts and analyzed.

Accomplishment - Study demonstrated that spectra can be collected using Raman spectroscopy from room temperature to 700°C. Black body radiation, which interferes with Raman spectra, increases with temperature. Current study demonstrates, as shown in the illustration, black body radiation becomes significant at 800°C. This temperature is the maximum at which the existing 532-nm Raman system can be used without significant interference from black body radiation. Study also indicated that CO2 introduced in the melt can be detected by Raman spectroscopy at room temperature but required precise focusing on the bubble. Analyses of gas bubbles in glass had some additional limitations, including limited optical efficiency and a large depth of focus that did not allow precise focusing on the bubbles in the glass. As a result of these limitations, obtaining spectra of actual bubbles at elevated temperatures was not possible using the 532-nm system with a 17-inch focal length objective. To perform this analysis, the geometry of the system would have to be revised such that a shorter focal length objective could be used. This modification would require a small, high-temperature furnace that would allow the fiber optic probe head to be located in close proximity to the glass and appropriate shielding to prevent heat damage to the system optics. In addition, study of bubble formation at temperature higher than 800°C will require use of a continuous wave 425-nm laser.

      Measurement & Nondestructive Evaluation of Materials & Structures Program
2000 IR&D Home SwRI Home