Open Access
Add to BibSonomy
Abstract
The key-feature of ZERODUR as a glass-ceramic material which enables various optical technologies is the nearest-zero expansion. This unique property is based on the thermal expansion characteristic of a negative expansion phase consisting of nano-sized crystals, which turns the class of lithium aluminosilicate (LAS)-glass-ceramics to a state-of-the-art transparent nanomaterial. Some insight is given in the origin and design of this key-property as well as the resulting fit to the requests of various optical technology devices ranging back from 1960s up today without limits insight.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
In the course of 2022 as the International Year of Glass, the glass-ceramic product ZERODUR is one example of an optical material that already has and will continue to play an essential role in Society and its megatrends, e.g. in the fields of microelectronics and communication. The success story started several decades ago from mirror substrates of telescopes supporting scientists to understand the Universe and continues, e.g, within more recent technologies in micro and extreme ultraviolet lithography needed to produce microchips for various devices of modern technologies.
In every case described here, the basic reason for the use of ZERODUR is not found in its “classical” optical properties, but upon its outstanding feature of nearest-zero expansion which can only be realized by profound knowledge and control of composition and related process technology.
2. LAS-glass-ceramic basics
2.1 Glass-ceramics introduction
The unusual and yet outstanding property of near-zero expansion is a key characteristic of a compositional range characterized by the main components Li2O, Al2O3 and SiO2 - in the starting glass as well as in the evolving crystalline phase upon heat treatment, therefore setting the abbreviated name of “LAS-glass-ceramics”.
The material class “glass-ceramics” is described in several standard books, e.g. by McMillan [1], Höland and Beall [2] and others, whereas Bach and Krause [3] edited a monograph on low-expansion glass-ceramics only.
As a material class of its own, glass-ceramics - also called “vitroceramics” or “sitalls” - were traditionally defined as e.g. stated by McMillan [1]: “Glass-Ceramics are polycrystalline solids prepared by the controlled nucleation of glasses” which was recently updated to a more general view by the group “Technical Committee 7 of ICG” et al. [4]: “Glass-ceramics are inorganic, non-metallic materials prepared by controlled crystallization of glasses via different processing methods. They contain at least one type of functional crystalline phase and a residual glass. The volume fraction crystallized may vary from ppm to almost 100%”.
In the family of “LAS-glass-ceramics” the major functional crystalline phase is based on a solid solution of β-eucryptite which differs from the vast majority of other crystals by its negative thermal expansion behavior, detected by Hummel in 1951 [5]. This negative expansion of the crystal phase may compensate the positive expansion of the surrounding residual glass, resulting to a near zero expansion of the resulting glass-ceramic material.
2.2 LAS-glass-ceramics: introduction to thermal expansion
The characteristic feature of thermal expansion distinguishes LAS-glass-ceramics from other type of materials significantly, as e.g. illustrated in the following graph (Fig. 1) describing the relative length change (dl/lo) as a function of temperature of commercially available Ti-doped fused silica and sintered cordierite ceramics:
Fig. 1. Thermal expansion ZERODUR TAILORED, Ti-doped SiO2 and Sintered Cordierite, modified from [6]
Within these measurements, the relative length change (dl/lo) at 0°C generally is set as reference value zero. All samples depicted here were analyzed in the same dilatometer [6].
A unique feature of ZERODUR can already be seen here: The dilatation of ZERODUR is generally extremely small in the range of “extended” room temperature, describing the range of 19°C-25°C with extended variation up to 20K, marking an essential advantage for multiple technical applications.
In addition, it can be figured out that the comparison of thermal expansion data between different materials should take into account the temperature range of data collection, as e.g. in the example given above the material of sintered cordierite shrinks up to the range of room temperature followed by a pronounced material expansion upon elevated temperatures.
The slope of the measured length change is known as “coefficient of thermal expansion” – abbreviated “CTE” in the following – and given in the next graph (Fig. 2) for the above mentioned materials:
Fig. 2. Coefficient of Thermal Expansion ZERODUR TAILORED 22°C, Ti-doped SiO2 and Sintered Cordierite, modified from [6]
2.3 LAS-glass-ceramics: Introduction to high-quartz solid solution
The fundamental results by Hummel [5] were achieved upon sintering experiments of batch materials of Li2CO3, Al2O3 and SiO2 of natural flint. This basic research of crystallization effects within broad compositional ranges resulted to solid solution samples of Li2O•Al2O3•nSiO2 with very low to near –zero expansion in the range of 3 ≤ n ≤ 10.
As β-eucryptite has a distinct stoichiometry of Li2O•Al2O3•2SiO2, other scientists approached by the concept of “stuffed derivate” for SiO2 polymorphs, quoted e.g. in [2,3]. Meanwhile, the related functional phase of solid solutions of β-eucryptite in various LAS-glass-ceramic products mainly is called “high-quartz solid solution” in industry [3]. This high–quartz solid solution crystal with hexagonal symmetry is a metastable phase which transforms at elevated temperatures to a solid solution phase with tetragonal symmetry based on keatite, also often referred to as solid solution of β-spodumene (with a stoichiometry of Li2O•Al2O3•4SiO2), respectively [2,3].
The solid solution crystal phase in LAS-glass-ceramics is also regarded as a “stuffed derivate” [5] where SiO2 is substituted by LiAlO2: Hereby, Al3+ is substituting Si4+ in the lattice while the charge is compensated by Li+ which is interstitially accommodated in channels of this structure. Beside LiAlO2 incorporated in this “stuffed” quartz structure, MgAl2O4 [7] and ZnAl2O4 [8] can substitute SiO2 in this specific type of quartz structure, too. Petzoldt investigated the stability range of this pseudo-quaternary system LiAlO2 - MgAl2O4 - ZnAl2O4 – SiO2 intensively [9], especially in respect to the thermal expansion of the lattice parameters in a- and c-axis directions. In addition, he investigated the substitution of SiO2 by AlPO4 [10] discovering a broad range of solubility. His results on mutual solubility of these 6 oxides forming the high-quartz solid solution in LAS-glass-ceramics were summarized in a general manner [10]:
These finding describe the broad compositional range of high–quartz solid solution - and related properties - opening up multiple opportunities for material development activities.
The following graph (Fig. 3) illustrates the variation of thermal expansion behavior of ZERODUR TAILORED 22°C and 2 LAS-glass-ceramics of other vendors, all of them containing high-quartz solid solution as functional crystal phase:
Fig. 3. Coefficient of thermal expansion of ZERODUR TAILORED 22°C and 2 LAS-glass-ceramics of competitors, modified from [6]
The thermal expansion of many glass-ceramic materials show a non-linear temperature dependence. Here, the different slopes of thermal expansion coefficients within one family of glass-ceramic material clearly demonstrate that the desired nearest zero expansion is not given per se, but required and to this day still requires intensive developmental work and efforts.
2.4 LAS-glass-ceramic products require nucleation agents
Basic optical principles have to be fulfilled to achieve a transparent glass-ceramic material, e.g. described by Beall and Duke [11], 1969, who point out 3 main topics:
- 1.) First of all, the crystals have to be smaller than the wavelength of the visible light.
- 2.) The refractive index difference between crystals and the residual glass phase have to be very small.
- 3.) The optical anisotropy within the crystallites have to be very small, too.
Subsequently, the catalytic capacities to provoke the crystallization of the desired functional crystal phases has to evolve upon consecutive processing only.
Instead of “nucleation agents” the term of “seeds” can be found as well as “mineralizer”: Singer [12] described in 1929 the use e.g. of oxides of Ti, Zr, Ce, Mn, Cr, Fe and V as “real” mineralizer of glasses and enamels, as they increase the number of crystal nuclei without taking part in the process of (functional phase) crystallization themselves.
The high efficiency of TiO2 as nucleating agent for LAS-glass-ceramic was described by Stookey of Corning in the late 1950’s [13], marking the decisive breakthrough for innovation in the field of LAS glass-ceramic products. According to the filing dates of patent applications, more or less in the same period of time up to the early 1960’s inventors in several companies and countries conducted crystallization experiments with relevance for LAS-glass-ceramic developments up to date, e.g. including the work of Baum at Siemens in 1960 [14]. Hayami, Tanaka and Ogura at AIST (Japan) [15] already claimed the use of MoO3 and WO3 as nucleating agent merely parallel to the work of McMillan of English Electric Company [16]. In 1959 [17], he describes the addition of P2O5 serving as a nucleating agent in LAS-glass-ceramics shortly before Tashiro and Takagi of Nippon Electric Glass [18], whereas Babcock et al. from Owens-Illinois Inc. introduce SnO2 as nucleating agent in transparent and colored LAS-glass-ceramics in 1964 [19]. The combination of TiO2 und ZrO2 turned out to be more favorable to achieve transparent LAS-glass-ceramics compared to the use of TiO2 as nucleating agent only, as reported by Sack and Scheidler of SCHOTT [20]. Their colleague Müller [21] investigated the effect of SnO2, Ta2O5, ThO2, Nb2O5 and ZrO2 especially as mixtures thereof and pointed out that their ability to form mutual solid solutions appears crucial for nucleation efficiency in high-quartz solid solution glass-ceramics, too.
It has to be mentioned that the role of P2O5 in LAS-glass-ceramics is still discussed ambiguously today [22,23].
2.5 Microstructural development in LAS-glass-ceramics
The process of crystallization in heterogeneous nucleated glass-ceramic materials can be divided into several stages in the course of thermal treatment, e.g. [24,25,26]:
- 1.) an amorphous phase separation where one phase contains nucleation agents
- 2.) nucleation agents start to crystallize
- 3.) subsequent crystallization of the main functional phase – see e.g. Figure 4
With the means of high resolution analytics, e.g. transmission electron microscopy (TEM), the scale of the starting phase separation as well as the crystallization of the nucleating agents has been determined to be only single digit nanometer scale of about 3–5 nm up to 5–8 nm, respectively, see e.g. Maier and Müller [24], 1987. The observation of phase separation by TEM images was underlined by short-range structural investigations of crystallizing LAS-materials [25]: During the course of heat treatment, the environment of Ti4+ ions changes from tetrahedral to octahedral coordination which is attributed to the change from the amorphous to the crystalline status, respectively.
Additionally, Maier and Müller [24] concluded that high-quartz solid solution crystals grow from a central ZrTiO4-nucleus due to epitaxial relation between translation vectors of ZrTiO4 and β-quartz; this remains a topic of current high-resolution TEM-analysis and interpretation today, see e.g. [27].
Progress in analytical instrumentation now allows the more detailed investigation of this relevant stage of crystallization: Phase separated, amorphous droplets of nucleation agents not only start to crystallize, but are additionally surrounded by an alumina enriched layer [28], enriching the scientific discussion on the mechanism of heterogeneous nucleation in LAS-glass-ceramics with further detailed observation.
Upon carefully selected thermal procedure, the finally resulting microstructure may result in a nano-sized microstructure with a high amount of high-quartz solid solution crystals characterized by an average grain size below 100 nm – see e.g. grey grains in Fig. 4 –, therefore ensuring transparency in the visual range.
The amount of high-quartz solid solution of the LAS-glass-ceramic laboratory sample in Fig. 4 was determined to 75% with an average grain size of 63 nm via XRD (X-ray diffraction).
2.6 Laboratory techniques in material development of LAS-glass-ceramics
The preparation of LAS-glass-ceramic products via controlled crystallization of the initial LAS-glasses requires the choice of appropriate nucleation agents in combination with a suitable nucleation process, generally induced via thermal treatment.
In the case of transparent glass-ceramics, an optimum nucleation process aiming at a high number of nano-sized nuclei is a major factor to achieve a high degree of transparency.
In return, this correlation to visual transparency can be used to identify the suitable type and amount of nucleation agents as well as appropriate process parameters.
McMillan [1] suggests to heat-treat several samples of the glassy material at various temperatures in the nucleation range for a fixed period of time and to transfer them to a 2nd furnace for crystal growth treatment. The subsequent analysis of crystal number and growth, e.g. by optical measurement, leads to the determination of the optimum nucleation parameters at the chosen set of temperature/time conditions. More efficiently, this procedure can be undertaken in a laboratory furnace with gradient zones, as a standard method of “gradient heating” today, e.g. described by Scheidler [29]. After determination of an appropriate nucleation temperature, this technique also can serve to identify a suitable temperature/time set of parameters for crystal growth, as shown in the following schematic graph (Fig. 5): A pre-nucleated bar of parent LAS-glass is placed in the gradient furnace for a given time. In correlation to the maximum temperature, increases in the average crystal size are accompanied by a gradual transformation from transparent to translucent and finally opaque appearance. Within the transparent rage, the final amount of high-quartz solid solution determines the CTE, whereas in the translucent and finally opaque ranges, keatite solid solution starts to evolve and later dominates the thermal expansion behavior.
Fig. 5. Heat treatment of a parent LAS glass in a gradient furnace [Scheidler, 29] Visual appearance: 1,2: Transparent 3,4: Opaque
Simple techniques of Differential Thermal Analysis (DTA) and/or Differential Scanning Calorimetry (DSC) techniques can be used alternatively or in addition to determine the optimum nucleation temperature: Sack and Scheidler [30], 1970, recommend the pre-heating of DTA-samples at different temperatures in the nucleation range for a fixed period of time, too, and compare the thereby induced shift of DTA-peak with the untreated sample. The difference ΔTpeak between the untreated and the pre-heated sample is depicted in relation to the temperature of pre-heating. A positive value of ΔTpeak points at an increase of nuclei during the pre-heating whereas a negative value of ΔTpeak is adverse to nucleation efficiency. Marotta et al. [31] proposed the same principle of basic DTA-technique in 1981, simplifying it even more by performing the above described sample pre-heating in the DTA furnace as isothermal hold during the DTA run and cross-check their results with crystalizing material of Li2O-2SiO2 glasses. The refinement of this technique [32] also results in data on nucleation rates, using reciprocal temperature with “1/Tp” for the samples with isothermal hold and “1/Tp°” for the untreated sample. Meanwhile, this DTA-technique is often called the “Marotta-method”.
Fig. 6. Marotta-method results for ZERODUR. Quantity on the y-axis is proportional to logarithm of the nucleation rate, taken from [33].
As previewed by Marotta et al. who set up their experiments with Li2O-2SiO2 material, this technique has also been verified with ZERODUR as an example of LAS-glass-ceramic (Fig. 6), see Davis & Mitra [33].
3. Industrial application of LAS-glass-ceramics
Retrospectively, it did not take a long time from promising laboratory results to the first industrial products based on LAS-glass-ceramics. Main applications were and are still found in household applications, beginning with cookware, Corning product Pyroceram [2], followed by Cer-Vit of Owens-Illinois and Hercuvit by PPG as well as Neoceram N-11 by Nippon Electric Glass [2,3]. At first, the majority of these LAS-glass-ceramic products were based on keatite solid solution as the functional phase with low expansion characteristic [2,3]. Beside cookware, the use as cook top panels started with small formats [3], setting the pace to a broad household consumer market in the 1960s and 1970s preceding today’s modern transparent cooktop panels or fireplace windows [2,3], e.g. product ranges CERAN and ROBAX; by SCHOTT, KeraBlack and KeraLite by EUROKERA and Neoceram N-0 by Nippon Electric Glass as well as fire-rated doors, see e.g. FireLite by Nippon Electric Glass and varieties of PYRAN by SCHOTT.
With high-quartz solid solution as the functional crystal phase, these products exhibit not only low, but near zero expansion in the range of room temperature to 700°C ensuring high resistance to thermal shock [2,3].
The transparency of high-quartz solid solution glass-ceramics originates from the nano-size microstructure, and is an important feature for the customer e.g. using fireplace windows and (mostly tinted) cooktop panels. In contrast to application based features, in the case of ZERODUR, the transparency serves to inspect the inner quality of thick and large blocks of material to ensure quality demands for high precision optical components (Fig. 7). Classical optical inspections methods such as visual analysis, refractive index (Fig. 8) and stress birefringence come into play to inspect the LAS parent glass thoroughly.
Fig. 8. ZERODUR block of 200 mm diameter demonstrating H4 refractive index homogeneity, taken from [34].
3.1 LAS-glass-ceramics: astronomy
Stable shape and dimension is a prerequisite of components in precision optical devices, such as equipment for astronomical applications. The composition of ZERODUR providing nearest zero expansion was especially designed by Petzoldt and coworkers [10,35] to meet the requirements for telescope mirrors in observatories starting in the 1960s when the Max-Planck-Institute of Astronomy in Heidelberg (Germany) requested mirror telescopes of 4 meters in diameter [3,34]. As one of the leading companies for optical glasses, decades of manufacturing high quality products enabled SCHOTT to set up the technology for melting, casting, and cooling for mirror substrates of the 4 meter class starting in 1973 [34].
The subsequent process of ceramization to achieve glass-ceramic blanks of several tons requires several month, with the first 4-m-class glass-ceramic substrate completed in 1975 [34] that finally started operation in the observatory at Calar Alto (Spain).
The world heaviest monolithically manufactured glass in the 1990s [36] mark a milestone in glass production as yet unsurpassed: The European Southern Observatory (ESO) ordered monolithic mirror substrates for the 8.2-meter Very Large Telescope (VLT) to operate in the area of Atacama (Chile). Therefore, an upscaled spin-casting technique to process about 44 tons of molten glass (Fig. 9) was set up [36].
For even larger telescope mirrors, smaller segments are combined out of mainly hexagonal shaped substrates that require extremely high reproducibility.
The Extremely Large Telescope (ELT) of ESO will be the largest optical telescope in the world with a primary mirror of 39 m aperture (Fig. 10).
This primary mirror will consist of 798 hexagonal mirror segments with slightly different shape to cope with the aspherical surface. The production of these mirror blanks by SCHOTT is ongoing until 2024 [37], while the secondary and tertiary ZERODUR mirror blanks of 4.25 m and 4 m have already been delivered [37].
The “technical first light” for the ELT is targeted for end of 2027 and is expected to collect up to 20 times more light out of space compared to existing large telescopes [38].
Besides earthbound telescopes, ZERODUR is widely used for optical and structural components in satellites, too [34]. Some of these devices have been space-born for many years under the impact of cosmic radiation, but no issue had been detected so far [34,37]. Nevertheless, the effects of radiation loads were investigated recently by Carré et al. [39].
Ring laser gyroscopes [3] are part of the navigation system in airplanes. Here, the helium permeability has to be low in order to avoid undesired He gas diffusion and ZERODUR is distinct from other low–expansion materials material by more than 2 orders of magnitude [3].
3.2 ZERODUR: adaptation to EUV-lithography
The production of electronic elements via microlithography requires highest precision in alignment and positioning of movable stages where ZERODUR is an established material for this application for the past ∼2 decades [3,34].
The demand of semiconductor integrated circuit chips will continue to lead to the demand of feature structures becoming finer than 20 nm. Therefore, the next step beyond optical microlithography is marked by Extreme Ultraviolet (EUV) lithography technology aiming to realize structures below 10 nm. In order to ensure image stability with tighter tolerances, even existing materials with extreme low thermal expansion have to be improved.
In addition to the qualification of ZERODUR as material for mechanical support, the requirements for CTE in respect to substrate materials in EUV lithography were crosschecked for ZERODUR.
In the case of ZERODUR, the time and temperature conditions of ceramization can be varied to a considerable degree in order to tailor the final CTE value.
This is illustrated in Fig. 11 where all samples were taken from the same block of parent glass [40]. Prescribed variations of process parameters lead to rotation of the Δl/l0 curve. The given 5 examples of slopes all represent some process parameters and therefore demonstrate the variety of slopes generally achievable.
This special feature of ZERODUR opens up the possibility to tailor the CTE to be minimum within a desired temperature range and even to influence the position of zero crossing.
Having a closer look at the temperature range of interest for the EUV lithography community, one example in Fig. 12 reveals a nearest zero-expansion of 0 ± 5 ppb/K (0 ± 0.005 ppm/K) in the range of about 10 to 33°C and a zero-crossing of the CTE slope at 22.5°C.
As a result of carefully adjusted ceramization parameters, samples of ZERODUR exhibiting CTE of 0 ± 5 ppb/K in the range of 19 to 25°C corresponding to the lowest expansion class of the standard SEMI P37 [41] were obtained.
3.3 ZERODUR tailored
ZERODUR has been designed originally for applications in the temperature interval from –30°C to + 70°C [3]. Usually, ZERODUR is characterized by its CTE (0-50°C) which serves to define expansion classes, e.g. “Class 0” is given by a CTE (0-50°C) value range of 0 ± 0.02 × 10−6/K, equivalent to ± 20 ppb/K (0-50°C).
Decades of production experience lead to more precise control of the ceramization process on one hand and in combination with improvements of dilatometry accuracy, SCHOTT introduced new CTE grades with even tighter tolerances, see Table 1 [42].
Table 1. ZERODUR CTE grades available at SCHOTT
As the thermal expansion of LAS-glass-ceramics is non-linear to the temperature – see e.g., Figs. 1 & 2 – SCHOTT introduced ZERODUR TAILORED in 2012 [42] in order to provide material with expansion characteristic specific adjusted (=“tailored”) to the application temperature profile.
Cleanrooms for the production of integrated circuits (IC) are one example of the hereby addressed application as they are operated at 22°C by well controlled temperature conditioning.
In the temperature range around 22°C, various low expanding materials – see Fig. 1 & 2 – exhibit a steady slope, either increasing as seen for sintered cordierite and Ti-doped silica or decreasing in the case of other commercially available LAS-glass-ceramics [6]. Taking a closer look at the temperature range between 0°C and 40°C clearly point out that only ZERODUR TAILORED 22°C meets the tight tolerance of 0 ± 0.010 ppm/K, or 0 ± 10 ppb/K, respectively, in case of temperature changes of several K. (see Fig. 13)
4. One outlook–adaptation to potential future demands
ZERODUR TAILORED 22°C has the smallest changes in CTE within the tolerance of 0 ± 10 ppb/K, (0 ± 0.010 ppm/K), over the widest temperature range of 32 K.
Nevertheless, during manifold processing steps of ZERODUR at customers, the range of near room temperature might be exceeded for various reasons either for (repeated) short periods of time or as optimum point of operation.
In order to be prepared for this potential future demand, a R&D project aimed to develop a LAS glass-ceramic with enhanced 10 ppb/K-range of thermal expansion. The resulting compositions were melted and ceramized with process parameters close to actual conditions in case for upcoming scale-up. Figure 14 shows examples of this development where the small band of 0 ± 10 ppb/K CTE-tolerance was nearly doubled to a temperature range of 57 K [43].
The composition and processes to attain the unique material ZERODUR were developed, established and improved for more than half a century at SCHOTT. Uncountable tons of material have been produced meeting demands for numerous application with different focused product specifications. Customers - and scientists - came up with a wide perspective of performance, and consequently, a broad spectrum of material and product data was gained and collected throughout these years, as summarized e.g. by Hartmann et al. [37,44].
With new technologies and fields of applications emerging, this steady process to (at the same time) deepen and widen insights in the material class of LAS-glass-ceramics continues…
Acknowledgments
The author likes to thank many colleagues of the ZERODUR team in development and production who gave insight and offer the opportunity to co-work in this exciting world of ZERODUR. A special thank is expressed to my colleagues Thomas Westerhoff, Ralf Jedamzik and Tedi-Marie Usher-Ditzian for internal reviewing this paper.
In memoriam of our friend and colleague Mark J. Davis, Phd (1960-2021).
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. P.W. McMillan, Glass-Ceramics, 2nd ed. (Academic Press Inc., 1979).
2. W. Höland and G.H. Beall, Glass-Ceramic Technology, 3rd ed. (Wiley, 2020).
3. H. Bach and D. Krause (editors), Low Thermal Expansion Glass-Ceramics, 2nd ed. (Springer, 2005).
4. J. Deubener, M. Allix, M. J. Davis, A. Duran, T. Höche, T. Honma, T. Komatsu, S. Krüger, I. Mitra, R. Müller, S. Nakane, M. J. Pascual, J. W. P. Schmelzer, E. D. Zanotto, and S. Zhou, “Updated definition of glass-ceramics,” J. Non-Cryst. Solids 501, 3–10 (2018). [CrossRef]
5. F. A. Hummel, “Thermal expansion properties of some synthetic lithia materials,” J.Am.Cer.Soc. 34(8), 235–239 (1951). [CrossRef]
6. R. Jedamzik, C. Kunisch, J. Nieder, and T. Westerhoff, “Glass ceramic ZERODUR enabling nanometer precision,” Proc. SPIE 9052, 90522I (2014). [CrossRef]
7. W. Schreyer and J. F. Schairer, “Metastable solid solutions with quartz type structure on the join MgAl2O4 –SiO2“,” Zeitschrift für Kristallographie 116(1-2), 60–82 (1961). [CrossRef]
8. G.H. Beall, G.H. Karstetter, and H.L. Rittler, “Crystallization and chemical strengthening of stuffed β-quartz glass-ceramics,” J American Ceramic Society 50(4), 181–190 (1967). [CrossRef]
9. J. Petzoldt, “Metastabile Mischkristalle mit Quarzstruktur im Oxidsystem Li2O-MgO-ZnO-Al2O3-SiO2,” Glastechnische Berichte 40, 385–396 (1967).
10. J. Petzoldt, “Der Einbau von P2O5 in metastabile Mischkristalle mit Quarzstruktur des Grundsystems Li2O-MgO-ZnO-Al2O3-SiO2,” Glastechnische Berichte 41(H.5), 181–189 (1968).
11. G. H. Beall and D. A. Duke, “Transparent glass-ceramics,” J. Mater. Sci. 4(4), 340–352 (1969). [CrossRef]
12. F. Singer, “Verfahren zur Regelung der Entglasung von Schmelzen,” DE570148, filed 1929
13. D. Stookey, “Method of making ceramics and product thereof,” US2,920,971, filed 1956
14. W. Baum, “Ferromag. Körper hoher Temperaturwechselbeständigkeit…,” DE1130349, filed 1960
15. R. Hayami, H. Tanaka, and T. Ogura, “Process for manufacturing ceramic-like products from glass by microscopic crystallization,” US3238085, first filed in Japan 1960
16. P.W. McMillan, “Keramikgegenstand und Verfahren zu seiner Herstellung,” DE1421866, filed in GB 1961
17. P.W. McMillan, “Improvements in or relating to Devitrified Glass Ceramics,” GB924996, filed 1958
18. M. Tashiro and K. Takagy, “Method of preventing scum formation in glass melt and glass-ceramic products,” US3282712, first filed in Japan 1962
19. C.L. Babcock, R.A. Busdiecker, and E.C. Hagedorn, “Glass,” GB1124001, first filed in US 1964
20. W. Sack and H. Scheidler, “Einfluß der Keimbildner TiO2 und ZrO2 auf die sich ausscheidenden Kristallphasen bei der Bildung von Glaskeramik,” Glastechn. Berichte 39(H.3), S.126–130 (1965).
21. G. Müller, “Zur Wirkungsweise von Gemischen oxidischer Keimbildner in Glaskeramiken des Hochquarz-Mischkristalltyps“, Glastechnische Berichte,” Zeitschrift für Glaskunde 45(H.5), S.189–194 (1972).
22. P. Glatz, M. Comte, L. Cormier, M. Montagne, B. Doumert, and G.G. Moore, “Different role of phosphorous in the nucleation of lithium aluminosilicate glases,” J. Non-Cryst. Solids 493, 48–56 (2018). [CrossRef]
23. C. Venkateswaran, H. Sreemoolanadhan, and R. Vaish, “Lithium aluminosilicate (LAS) glass-ceramics: A review of recent progress,” Int. Mat. Rev. 67, 620 (2021). [CrossRef]
24. V. Maier and G. Müller, “Mechanism of oxide nucleation in lithium aluminosilicate glass-ceramics,” J. Am. Ceram. Soc. 70(8), C-176–C-178 (1987). [CrossRef]
25. V. Maier and R. Frahm, “X-rax absorption studies of the short-range order in lithium aluminosilicate glass-ceramics,” Glastech. Ber. 62 (1989).
26. L. R. Pinckney and G. H. Beall, “Microstructural evolution in some silicate glass-ceramics: a review,” J. Am. Cer. Soc. 91(3), 773–779 (2008). [CrossRef]
27. E. Kleebusch, C. Rüssel, C. Patzig, and T. Höche, “Evidence of epitaxial growth of high-quartz solid solution on ZrTiO4 nuclei in a Li2O-Al2O3-SiO2 glass,” J. Alloy Compounds 748, 73–79 (2018). [CrossRef]
28. S. Bhattacharyya and T. Höche, “Direct evidence of Al-Rich Layers around Nanosized ZrTiO4 in Glass: Putting the Role of Nucleation Agents in Perspective,” Crystal Growth Design 10(1), 379–385 (2010). [CrossRef]
29. H. Scheidler, “Herstellung und Eigenschaften von Glaskeramik-Werkstoffen,” Silikat-Journal 11, 6 (1972).
30. W. Sack and H. Scheidler, “Einfluß von Viskosität, Phasenstabilität und Kristallisationsverhalten auf die Herstellung von temperaturwechselbeständigem Natrium-Boro-Silikat- und Lithium-Aluminium-Silicat-Wirtschaftsglas“,” Glastechn. Berichte 43(H.9), 359–368 (1970).
31. A. Marotta, A. Buri, and F. Branda, “Nucleation in glass and thermal analysis,” J Mater Sci 16(2), 341–344 (1981). [CrossRef]
32. A. Marotta, A. Buri, F. Branda, and S. Saiello, “Nucleation and Crystallization of Li2O-2SiO2 glass – a DTA study,” in Nucleation and Crystallization in Glasses, Advances in Ceramics Vol 4, (Am. Cer. Soc., 1982).
33. M. J. Davis and I. Mitra, “Crystallization Measurements Using DTA Methods: Applications to ZERODUR,” J. Am. Cer. Soc. 86(9), 1540–1546 (2003). [CrossRef]
34. T. Döhring, R. Jedamzik, T. Westerhoff, and P. Hartmann, “Four decades of ZERODUR mirror substrates for astronomy,” Proc. SPIE 7281, 728103 (2009). [CrossRef]
35. J. Petzoldt and H. Scheidler, “Verfahren zur Herstellung transparenter Glaskeramiken,” DE1596860, filed 1967
36. R. Müller, H. Höneß, H. Morian, and H. Loch, “Manufacture of the first primary mirror blank for the Very Large Telescope (VLT),” Proc. SPIE 2199, 164 (1994). [CrossRef]
37. P. Hartmann, R. Jedamzik, A. Carré, J. Krieg, and T. Westerhoff, “Glass-ceramic ZERODUR: Even closer to zero thermal expansion: a review, part 1,” J. Astron. Telesc. Instrum. Syst. 7(02), 020901 (2021). [CrossRef]
38. Homepage ESO, https://elt.eso.org
39. A. Carré, M. Valente, T. Hull, J. Krieg, and T. Westerhoff, “An empirical approach for space-based mirrors to evaluate the extent of radiation compaction under expected environmental dose,” Proc. SPIE 11852 (2021).
40. I. Mitra, M.J. Davis, J. Alkemper, R. Müller, H. Kohlmann, L. Aschke, E. Mörsen, S. Ritter, H. Hack, and W. Pannhorst, “Thermal Expansion Behavior of Proposed EUVL Substrate Materials,” Proc. SPIE 4688 (2002).
41. SEMI P37-1109 “Specification for extreme ultraviolet lithography substrates and blanks,” origin. publ. 2001.
42. P. Hartmann, R. Jedamzik, and T. Westerhoff, “Zero-expansion glass ceramic ZERODUR - Recent developments reveal high potential,” Proc. SPIE 8450, 845022 (2012). [CrossRef]
43. I. Mitra and C. Kunisch, “Precision component,” US10730787, first filed in Germany 2017.
44. P. Hartmann, R. Jedamzik, A. Carré, J. Krieg, and T. Westerhoff, “Glass-ceramic ZERODUR: Even closer to zero thermal expansion: a review, part 2,” J. Astron. Telesc. Instrum. Syst. 7(2), 020902 (2021). [CrossRef]
