46
47
maintain morphology upon heating until the sintering temperature is reached. The data shows that regardless of variations of particle morphology, the compositionally different glass frits demonstrated viscous flow and sintered within the
ɳ
s range. The insensitivity of viscous sintering to particle size and geometry has been shown in a number of studies.35Figure 28.Plot showing sintering viscosity as a function of SSA, Group I (circles) and Group II (triangles) data are shown.
Figure 29. Sintering temperature as a function of specific surface area.
0 2 4 6 8 10 12 14
0.010 0.100 1.000 10.000
Log [Sintering Viscosity (Pa∙s)]
Specific Surface Area (m2/g)
600 620 640 660 680 700 720 740
0.010 0.100 1.000 10.000
Sintering Temperature (°C)
Specific Surface Area (m2/g)
48
Figure 30. Scalped container glass fired at (a) 675°C and (b) 680°C for two hours. (Size bar = 100µm)
Figure 31. Scalped float glass fired at (a) 665°C and (b) 670°C for two hours. (Size bar = 100µm)
Figure 32. Particles maintain morphology during sintering, (a) container glass fired at 675°C and (b) float glass fired at 650°C for two hours. (Size bar = 200µm)
(a) 675°C
(b) 670°C (b) 680°C
(a) 665°C
(a) 675°C (b) 650°C
49
CONCLUSIONS
The primary hypothesis of this work was that glass frit sinters within a narrow and predictable viscosity window,
ɳ
s. Sintering viscosities were determined for six different glass chemistries and eight different particle sizes. Each glass frit was found to densify, make the transition from continuous to closed porosity, within the viscosity range,ɳ
s = 109.6 ± 1.9 Pa∙s. The variance in sintering viscosity is likely due to the susceptibility of certain glass compositions to water adsorption and other LOI specimens, such impurities have been shown to decrease the glass transition temperature (Tg). Discrepancies between HSM glass transition and sintering temperatures and those measured experimentally, also alludes to the susceptibility and impact of water on glass. Further, calculated glass transition temperatures were lower than those calculated in SciGlass, lending towards the idea that water absorption or some other impurity did decreases the glass transition temperatures.Densification of each analyzed glass was localized to temperatures just above the glass transition temperature even over a wide range of specific surface areas. Viscous flow, which is not dominated by surface energy reduction, is the primary mechanism for glass densification. Viscous flow or viscous phase sintering is evident in two main ways: (1) glass frit sintered at the same sintering viscosity before and after scalping the particle size and distribution, and (2) glass frit maintained unique particle morphology until the sintering temperature was reached.
This work concluded that glass frits, regardless of chemistry and particle size, sintered via viscous flow within a narrow viscosity window. The results and discussion touched on some questions that arose that were outside the scope of this work, but if these questions were to be answered could potentially support the notion of a universal sintering viscosity for glass frits.
50
SUGGESTIONS FOR FUTURE WORKS
While this work found that six different glass chemistries densified within a narrow viscosity window, future work could be done to further narrow the sintering viscosity range. Areas of future work could include analysis of frit water content via Fourier- transform infrared spectroscopy (FTIR) and thermogravimetric analysis (TGA), electron diffraction spectroscopy (EDS) for suspected phase separation, and structural analysis via x-ray diffraction (XRD) for suspected crystallization such as what was seen in heat treated fused silica frit.
It would be especially valuable to understand the amount of water present in the glass frits that were used for experiments regarding the primary hypothesis. Extended glass-water interactions have been shown to change viscosity-temperature characteristics and can depress the glass transition temperature (Tg). TGA would be valuable to deduce the amount of physically and chemically bound water present. For these purposes, the sensitivity of TGA may be enhanced by a lower ramp rate. Hot stage microscopy (HSM) and differential calorimetry (DSC) would also be useful, in addition to TGA, for determining the temperatures at which chemically bound water is released. Determination of a relationship between water content and glass composition would be invaluable in determining water content impact on sintering viscosity.
Regarding the secondary hypothesis, it would be useful to explore sintering temperatures of glasses with extremely large and extremely small particle sizes, e.g.
fusing two solid sheets of glass and sintering nano-frits/powders. Determination of sintering viscosities for extreme specific surface areas would give insight into factors effecting viscous flow.
51
REFERENCES
1. R. M. German, Sintering Theory and Practice; Ch. 1. John Wiley & Sons, New York, 1996.
2. M. N. Rahaman, Ceramic Processing and Sintering, 2nd ed.; Ch. 1. Marcel Dekker, New York, 1995.
3. M. N. Rahaman, Ceramic Processing and Sintering, 2nd ed.; Ch. 7. Marcel Dekker, New York, 1995.
4. M. N. Rahaman, Ceramic Processing and Sintering, 2nd ed.; Ch. 8. Marcel Dekker, New York, 1995.
5. H. Ravash, L. Vanherpe, J. Vleugels, and N. Moelans, "Three-Dimensional Phase- Field Study of Grain Coarsening and Grain Shape Accomodation in the Final Stage of Liquid-Phase Sintering," J. Euro. Ceram. Soc., 37, 2265-75 (2017).
6. M. N. Rahaman, Ceramic Processing and Sintering, 2nd ed.; Ch. 10. Marcel Dekker, New York, 1995.
7. R. M. German, Liquid Phase Sintering; Ch. 2. Plenum Press, New York, 1985.
8. G. W. Scherer, "Sintering of Low-Density Glasses: I, Theory," J. Am. Ceram. Soc., 60 [5-6] 236-9 (1977).
9. R. M. German, Sintering Theory and Practice; Ch. 3. John Wiley & Sons, New York, 1996.
10. L. Zagar, O. E. Klinger, W. Stumpfe, K. Loewer, "Surface Energy of Silicate Melts and Its Importance in Sintering Processes," Int. Round Table Conf. Sintering 14, 75-84 (1982). As cited in R. M. German, Sintering Theory and Practice; p. 82. John Wiley & Sons, New York, 1996.
11. G. W. Scherer, S. Calas, and R. Sempere, "Densification Kinetics and Structural Evolution During Sintering of Silica Aerogel," J. Non-Cryst. Solids, 240, 118-30 (1998).
52
12. W. D. Kingery, H. K. Bowen, and D. R. Uhlmann, Introduction to Ceramics, 2nd ed.; Ch. 14. John Wiley & Sons, New York, 1976.
13. J. E. Shelby, Introduction to Glass Science and Technology, 2nd ed.; Ch. 6. TJ International, Padstow, Cornwall, UK, 2005.
14. R. J. Loucks and J. C. Mauro, The Physics of Glass: An Introduction to Fundamental Concepts; Ch. 5. John Wiley & Sons, (in press).
15. J. E. Shelby, Introduction to Glass Science and Technology, 2nd ed.; Ch. 2. TJ International, Padstow, Cornwall, UK, 2005.
16. W. D. Kingery, H. K. Bowen, and D. R. Uhlmann, Introduction to Ceramics, 2nd ed.; Ch. 3. John Wiley & Sons, New York, 1976.
17. R. J. Loucks and J. C. Mauro, The Physics of Glass: An Introduction to Fundamental Concepts; Ch. 1. John Wiley & Sons, (in press).
18. J. C. Mauro, D. C. Allan, and M. Potuzak, "Nonequlibrium Viscosity of Glass,"
Phys. Rev. B: Condens. Matter Mater. Phys., 80 [9] 094204-18 (2009).
19. C. A. Angell, "Formation of Glasses from Liquids and Biopolymers," Science, 267 [5206] 1924-35 (1995).
20. J. C. Mauro, Y. Yue, A. J. Ellison, P. K. Gupta, and D. C. Allan, "Viscosity of Glass-Forming Liquids," Proc. Natl. Acad. Sci. U. S. A., 106 [47] 19780-4 (2009).
21. A. K. Varshneya, Fundamentals of Inorganic Glasses; Ch. 9. Alden Group, Sheffield, UK, 2006.
22. B. Oistad, "The Comparison of Model Viscosity Calculations to Measured Values";
B.S. Thesis. Alfred University, Alfred, NY, 2017.
23. A. Fluegel, A. K. Varshneya, D. A. Earl, T. P. Seward, and D. Oksoy, "Improved Composition-Property Relations in Silicate Glasses, Part I: Viscosity," Ceramic Trans., 170, 129-43, 2004.
53
24. T. P. Seward III, "Modeling of Glass Making Processes for Efficiency" (2003) Center for Glass Research. Accessed on: April 2017. Available at
<http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.579.8057&rep=rep1&
type=pdf>
25. "Measurement of Viscosity of Glass Between Softening Point and Annealing Range (Approximately 108 Pa∙s to Approximately 1013 Pa∙s) by Beam Bending (Metric)," ASTM Designation C 1350M-96. American Society for Testing and Materials, West Conshohocken, PA.
26. H. E. Hagy, "Experimental Evaluation of Beam-Bending Method of Determing Glass Viscosities in the Range 108 to 1015 Poises," J. Am. Ceram. Soc., 46 [2] 93-7 (1963).
27. C. A. Angell and M. Hemmati, "Glass Transitions and Critical Points in Orientationally Disordered Crystals and Structural Glassformers: ("Strong" Liquids Are More Interesting Than We Thought)" (2013) Accessed on: Nov. 2016.
Available at <http://www.public.asu.edu/~caangell/PL501%20Sendai%20CAA- Hemmati%20Sendai%20(printed).pdf>
28. J. Otminski, "Predicting Glass Formation from Melt Viscosity and Fragility"; B.S.
Thesis. Alfred University, Alfred, NY, 2013.
29. M. Ahmed, "Characterization of Glaze Melting Behavior with Hot Stage Microscopy"; M.S. Thesis. Alfred University, 2002.
30. A. M. Buchtel, "Pyroplastic Deformation of Whitewares"; M.S. Thesis. Alfred University, 2003.
31. "Standard Reference Material 717a," NIST Designation SRM 717a. National Institute of Standards and Technology, Gaithersburg, MD.
32. A. K. Varshneya, Fundamentals of Inorganic Glasses; Ch. 4. Alden Group, Sheffield, UK, 2006.
33. C. H. Yoon, "High-Temperature Study of Defects and Homogeneity in Glass";
Ph.D. Thesis. Alfred University, Alfred, NY, 1998.
34. W. C. LaCourse, "Determination of Skin Viscosity of Glass Via Fiber Elongation"
(1988) ResearchGate. Accessed on: September 2017. Available at:
https://www.researchgate.net/publication/277813983
54
35. R. K. Bordia, S.-J. L. Kang, and E. A. Olevsky, "Current Understanding and Future Research Directions at the Onset of the Next Century of Sintering Science and Technology," J. Am. Ceram. Soc., 100 [12] 2314-52 (2017).
36. J. K. Mackenzie and R. Shuttleworth, "A Phenomenological Theory of Sintering,"
Proc. Phys. Soc. B, 62 [12] 833-52 (1949).
55
APPENDIX A
Introduction
Throughout this work, discussion of the main factor influencing densification of glass is of the characteristic viscosity. Still, it is important to mention that while time is of lesser importance, heat distribution, a function of time and composition, is vital to achieving a sintering viscosity. In addition to initial sintering trials of varying hold times, a design of experiments (DOE) was conducted to determine a necessary or appropriate amount of time to carry out successful isothermal sintering of glass frit in a customized tube furnace. The DOE looked at three factors influencing heat distribution: depth, position and equilibration time (see Figures 33 & 34.) The position factor was found to be insignificant - the temperature that a specimen would experience, whether near the top of the D-tube (a short specimen) or near the top of the tube furnace (a tall specimen), were relatively the same. The depth factor was found to be significant - the temperature of the furnace varied nonlinearly with distance from the front mouth of the furnace. The hottest point, or hot zone, was determined to be approximately 50cm from the front. Therefore, the compact frit specimen(s) were placed at this distance between two, stationary thermocouples. The factor of furnace equilibration time was also found to be significant.
As a result, the D-tube and substrates were heated in the furnace during the heating ramp to a chosen setpoint. And to further avoid thermal lag, samples were pressed and quartered to obtain small and repeatable dimensions. The results of the DOE are plotted and given in Figure 35. Overall, time was not found to be a significant factor during the viscous sintering of glass frit, as has been shown by others.36 However, this DOE showed that the furnace setup required a minimum of 45-minutes at temperature to stabilize. Further, initial sintering trials showed that 60-minutes at temperature was not enough to overcome thermal lag within a crucible packed sample. Ultimately, the two-hour hold time (longer than necessary for thermal distribution through the system) was chosen for comparison to initial densification trials and to ensure complete thermal distribution. It is interesting to note that there were no observable microstructural differences between glass frit that sintered at a given temperature for 60-minutes compared to 2-hours, the only difference was the volume of glass frit that sintered in each specimen, partial and whole, respectively.
56 Experimental Procedures
Figure 33. Fishbone diagram for DOE.
Figure 34. DOE three-factor factorial.
57 Results and Discussion
Figure 35. (a) Pareto chart showing that depth (distance from front of furnace) and furnace equilibration time are significant factors, and (b) a box plot showing that at higher temperatures, in the hot zone, temperature uncertainties decrease drastically.
(a)
(b)