# Fileset

[134_26024.pdf](https://mdr.nims.go.jp/filesets/3f2bb43a-3518-49a6-b1e2-a62e48a257bf/download)

## Creator

Richard Weber, Stephen K. Wilke, Jared Rafferty, Abdulrahman Al-Rubkhi, Chris Benmore, Benjamin Moulton, Alan Kastengren, [Shinji Kohara](https://orcid.org/0000-0001-9596-2680), Rina Shimonishi, Chihiro Koyama, Takehiko Ishikawa

## Rights

[Creative Commons BY Attribution 4.0 International](https://creativecommons.org/licenses/by/4.0/)

## Other metadata

[Structure, properties and microgravity processing of liquids and glasses](https://mdr.nims.go.jp/datasets/65ab1df4-4127-497f-a19b-cc216d6f5c7a)

## Fulltext

Structure, properties and microgravity processing of liquids and glassesFULL PAPERStructure, properties and microgravity processing of liquids and glassesRichard Weber1,2,³, Stephen K. Wilke1,2, Jared Rafferty1, Abdulrahman Al-Rubkhi1,Chris Benmore2, Benjamin Moulton3, Alan Kastengren2, Shinji Kohara4,Rina Shimonishi5, Chihiro Koyama5 and Takehiko Ishikawa51Materials Development, Inc., Arlington Heights, IL 60004, USA2X-ray Science Division, Advanced Photon Source, Argonne National Laboratory, Argonne, IL 60439, USA3Alfred University, Alfred, NY 14802, USA4National Institute for Materials Science, Tsukuba, Ibaraki 305–0047, Japan5Japan Aerospace Exploration Agency, Tsukuba, Ibaraki 305–8505, JapanContainerless processing was used to access and study supercooled liquids and glasses that cannot be made usingconventional melting approaches. In-situ measurements of melt atomic structure and density provided insightinto how the glass forms. Experiments included making measurements in microgravity where buoyancy drivenconvection and sedimentation are suppressed. Here we examine the structure and properties of rare earth-aluminate composition liquids and some glasses made from them. The structures show a fundamentally differentnetwork behavior from the classical Zachariasen model. The network comprises four and about 40% fivecoordinated aluminum ions that share corners or edges and often form triply bonded species with an oxygen ion.The concept of Kn is used to evaluate the glass forming behavior in terms of network connectivity and bonding.The temperature dependence of density of the liquid is reported and discussed in the context of processingmolten materials in reduced gravity where bubbles can be trapped in the liquid due to lack of buoyancy. Theinterior structure in samples was investigated using X-ray tomography to investigate how bubbles can clusterinside a liquid drop.Key-words : Glass, Liquid, Supercooling, Containerless processing, Atomic structure, Density[Received March 15, 2026; Accepted March 19, 2026; Published online April 23, 2026]1. IntroductionMuch industrial value-add materials processing involveshigh temperature liquids. This is particularly the case forproduction of glass where mixtures of metal oxides aretypically melted and then cooled in a way the preventsnucleation of crystals.1,2) Containerless processing canenhance glass formation by avoiding both heterogeneousnucleation of crystals and chemical contamination ofmelts.3–5) In this work, containerless processing was usedto access and study supercooled liquids and glasses thatcannot be made using conventional melting approaches. Inaddition, the use of in-situ measurements of melt proper-ties and atomic structure were used to gain insight intohow the glass forms. The research includes the use ofmicrogravity experiments to investigate uncontained meltsin conditions where buoyancy driven convection andsedimentation are suppressed.6,7)These containerless techniques expand the compositionrange that can be explored, enabling processing of veryfragile liquids in conditions that can result in glasses with-out the traditional network forming components. Theseglasses can offer fundamental insights in how the structurechanges to create a non-crystalline product. It also pro-vides samples of materials that can be used to determineproperties and benchmark research directions to developnew glass materials.2. MethodsTwo containerless methods were used. Experiments inthe ground-based laboratory and at synchrotron beamlinesused aerodynamic levitation in combination with carbondioxide laser beam heating.8,9) This technique is useful forwork with molten metal oxides ranging from borates10) andcarbonates11) to highly refractory materials such as zirco-nium dioxide.12) Forces derived from gas flow through aconverging-diverging nozzle can trap liquid drops fromabout 1mm to several mm in diameter. Sample temperatureis controlled by adjusting the laser beam heating power.Selected compositions were investigated in micrograv-ity using the JAXA Electrostatic Levitation Furnace(ELF).13) The operation and methods used with the in-strument is described in detail in the literature13–15) andbriefly summarized here. Samples develop a small chargedue to loss or gain of electrons (typically loss of electronsby thermionic emission). Three orthogonal pairs of elec-³ Corresponding author: R. Weber; E-mail: rweber@anl.gov‡ Preface for this article: DOI https://doi.org/10.2109/jcersj2.134.P4-1Journal of the Ceramic Society of Japan 134 [5] 355-360 2026DOI https://doi.org/10.2109/jcersj2.26024 JCS-Japan©2026 The Ceramic Society of Japan 355This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/),which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.https://doi.org/10.2109/jcersj2.134.P4-1https://doi.org/10.2109/jcersj2.134.P4-1https://doi.org/10.2109/jcersj2.26024https://creativecommons.org/licenses/by/4.0/trodes are used to repel the sample and provide stablepositioning. The sample position is monitored using animaging system and its position is actively stabilized bycontrolling the voltage on the electrodes. Samples are heat-ed using four 40W diode lasers arranged in a tetrahedralarray. As in the aerodynamic levitator, sample temperatureis controlled by adjusting the laser beam heating power.Samples levitated in the ELF are imaged by using abacklight to project a silhouette of the sample onto a videocamera sensor. Analysis of video images enables calcu-lation of the volume of the sample and hence measurementof its density.15) Samples can also be excited into an oscil-lation mode by pulsing the voltage on the positioningelectrodes at selected frequencies. This enables investiga-tion of fluid properties such as surface tension and vis-cosity that can be derived from the resonance oscillationfrequency and the damping of motion when the stimula-tion is stopped.13,14)Various glass compositions were synthesized using con-tainerless methods. Here we will focus on rare earth alu-minates as a system that exhibits some of the behavior thatcan occur in fragile liquids. Compositions based onytterbium and lanthanum were of particular interest sincethey span a range of rare earth ion radius. The large lan-thanum ion enables glass formations. Compositions basedon the smaller ytterbium ion do not form glass.16)2.1 Atomic structureHigh-energy X-ray diffraction measurements were madeat Sector 6-ID-D of the Advanced Photon Source, ArgonneNational Laboratory (Lemont, IL, USA). An aerodynamiclevitator was used as the sample environment.9) Sampleswere levitated on a stream of air and heated with a carbondioxide laser beam until molten. Melt temperature wascontrolled by adjusting the incident laser beam power(measurement uncertainty of «30K17)). X-ray diffractionmeasurements were collected at several different temper-atures. The diffraction of 99.96 keV X-rays was measuredin transmission geometry using an area detector (Varex4343CT) arranged to provide a range of momentum trans-fers 0.3 < Q < 24.5¡¹1. Data were corrected and reducedfollowing previously described procedures18–20) resultingin the X-ray total structure factors and pair distributionfunctions (PDFs).21) The PDFs were calculated with amaximum Q value of 20.5¡¹1 and no modification (e.g.,Lorch) function was applied.Structural models were obtained from Empirical Poten-tial Structure Refinement (EPSR) of the X-ray structurefactors measured at selected temperatures. EPSR is areverse Monte Carlo based simulation technique, in whichsimple, pre-defined interatomic potentials are refined bycomparing the simulated and experimental scattering tobring the former into agreement with the latter.22) The out-put of the EPSR model enables interpretation of the struc-ture to provide an understanding of the coordination andbonding of cations in the sample. Details of the EPSRsimulations including the initial reference potentials havebeen published previously – see Refs. 26) and 27).2.2 DensityIn the ELF instrument, a video image of the samplesilhouette is recorded using a camera and ultravioletbacklight positioned on opposite sides of the levitationposition. The camera is regularly calibrated with precisionstainless steel spheres of similar diameter to the samples.For density analysis, individual frames were extractedfrom the video record of each sample during cooling (i.e.,after the heating lasers were turned off). The perimeter ofthe silhouette was fitted with a sixth order Legendre poly-nomial, from which the spheroid sample’s volume wascalculated.21) Density and thermal expansion were thencalculated based on the sample volume and mass, whichwas measured on Earth before and after the experiments inELF. This technique has been applied for several moltenmaterials including lanthanoid sesquioxides23) Ga2O3,24)and Au.25) Two sets of experiments were run in ELF. Thefirst set were run in air at a pressure of 2 bar. The secondwere run in high purity argon at a pressure of 2 bar.2.3 MicrostructureX-ray microtomography measurements were performedat Sector 7-BM-B of the Advanced Photon Source, follow-ing the same setup and reconstruction procedures as de-scribed previously.17) To measure internal porosity, thethree-dimensional reconstructions were analyzed as se-quences of individual cross-sections. In the image se-quence, each frame represents a two-dimensional cross-section of the sample, with a thickness of 1 pixel. Eachframe contained a bright area in the middle (the sampleportion) and the surrounding background, which wasdarker. Macro scripts were implemented in ImageJ to run athreshold that distinguished the sample from background.Only pixels in the sample area were counted. Then, thehole filling operation was used to fill-in any internal pores,and pixels within the sample were again counted. Porositywas calculated as the ratio of filled-in pixels to the totalnumber of sample pixels.3. Results3.1 Atomic structureThe atomic structure of several rare earth aluminum gar-net composition melts has already been reported else-where.16,26,27) One important finding from the structuralinvestigation arose from a comparison of the rare earthatomic environments in aluminate melts and glasses. Inbinary rare earth aluminates, glass forming ability de-creases as the rare earth cation size decreases,16) and vitri-fication is not possible for rare earths smaller than thulium.This trend correlates with changes in the Al–O coordina-tion number distributions, network linkedness and con-nectivity. In these systems, the atomic network comprisesa mixture of Al–O polyhedra of varying coordinationnumber linked predominantly via corner-sharing, thoughwith up to 17% edge-sharing in liquid above the meltingpoint. Because the Al–O units include substantial fractionsof five- and six-coordinate polyhedra (e.g., 36 and 6% inYbAG, ytterbium aluminum garnet Yb3Al5O12 composi-Weber et al.: Structure, properties and microgravity processing of liquids and glassesJCS-Japan356tion, at 2630K) in addition to the typical tetrahedra,26) theconventional Qn connectivity analysis cannot be used todirectly compare structure in different rare earth aluminatemelts. The Qn calculation assumes that the network isentirely tetrahedral with only corner-sharing and no triplybonded oxygen. In YbAG, all three of these assumptionsare invalid.In order to directly compare the connectivity of differ-ent melt structures, we have defined a new connectivitymetric, Kn. Like Qn, the value of n in Kn corresponds to thenumber of neighboring network units to a given unit in thenetwork, according to:n ¼ BO� ESþ 2� TBO ð1ÞIn Eq. (1), BO are the number of bridging oxygen, whichlink any two neighboring network units. ES are the num-ber of edge-shares between a given network unit and itsneighboring units. TBO are the number of triply-bondedoxygen, since these correspond to connections to twoneighboring units. A full description of the Kn calculationsand connectivity results are given in Ref. 26). A few illus-trations are provided in Fig. 1 for structural environmentsfound in molten YbAG at 2460K. In the first example, a[4]Al tetrahedron has three BO and one TBO. Each BOconnects to one neighboring network unit, and the TBOconnects to two neighboring network unit. All linkages arecorner-sharing (no edge-sharing), so Eq. (1) gives n =3 ¹ 0 + 2*1 = 5. If a traditional Qn analysis was usedinstead, the TBO would be ignored and only 3 neighboringunits would be counted, which illustrates why the assump-tions of Qn do not apply for this type of network structure.3.2 Density and microstructureThe density of molten YbAG was measured in twoseparate campaigns of microgravity experiments, whicheach required flying samples to the International SpaceStation for measurements in the ELF. In the first campaign,three replicate samples of YbAG were used, which wereinitially polycrystalline spheroids approximately 2mm indiameter. All three replicates were successfully levitatedand melted, and their measured densities are shown inFig. 2(a). The densities were consistent within ³1.5%among the replicates, which is close to the typical uncer-tainty for ELF density measurement.13) Intriguingly, allthree replicates exhibited a density maximum ca. 2000K,slightly below the equilibrium melting point of Tm =2283K.28) Anomalous density trends such as this maxi-mum have been correlated with polyamorphic transitionsin some liquids,29) suggesting that polyamorphism mayoccur in supercooled liquid YbAG. To explore this pos-sibility, a thorough investigation was completed of theatomic structure in molten YbAG, combining X-ray dif-fraction, neutron diffraction with isotope substitution, andstructural modeling. However, the structure measurementsand simulations showed no indication of polyamorphism.The explanation for the anomalous density trend withtemperature was found using X-ray tomography to probethe internal microstructure of the YbAG samples. BecauseYbAG does not form a glass even in containerless condi-tions (for ³2mm free cooled spheres), the melt quenchedproducts were polycrystalline. As shown in Figs. 3(a)–3(c), the three replicates from the first flight campaign thatwas run with air as the process gas contained large internalbubbles, which were likely responsible for the apparentdensity maximum. Each sample contained two large voidsseparated by a thin, 10–15¯m solid layer that presumablywas a thin liquid layer held in place by surface tensionFig. 1. Structural arrangements in molten YbAG. The con-nectivity metric Kn is used quantify the number of neighboringnetwork Al–O units to a given Al–O polyhedron. In each exam-ple, the polyhedron being analyzed is shown in opaque green,and its neighboring Al–O polyhedra are in translucent blue. Forthese three examples, the value of n is 5. Oxygen atoms are onlyshown for the central unit, for clarity. Ytterbium atoms are notshown.Fig. 2. Density of liquid YbAG. (a) Measurements during thefirst flight campaign, using three replicate samples that were laterfound to contain large internal bubbles. (b) Measurements of thesecond flight campaign, using two replicates that contained nointernal bubbles. Replicates are marked in different colors.Journal of the Ceramic Society of Japan 134 [5] 355-360 2026 JCS-Japan357during the melt processing. Surprisingly, the thin featuresremained intact even after solidification. It is also surpris-ing that all three samples contained similarly large voids.Other features evident in the tomography cross-sectionsinclude smaller voids and networks of cracks that likelyformed during the rapid solidification.During a second flight campaign, additional samples ofYbAG samples were run in high purity argon and the den-sity measurements were repeated. This time, the densityshowed a typical, monotonic thermal expansion with tem-perature. The densities of two replicate samples are shownin Fig. 2(b). The density of YbAG increases from 5.50 to5.70 g cm¹3 during cooling from 2400 to 1000K. Bothreplicates spontaneously crystallized ca. 1000K below theestimated glass transition of Tg = 1159K.These two samples were also probed with X-ray tomog-raphy and found to contain no internal voids. Representa-tive cross-sections are shown in Figs. 3(d) and 3(e). Simi-lar to the first campaign, the second campaign samplescontain cracks throughout the volume that likely formeddue to the sudden volume change during solidification.The absence of bubbles validates the density data collectedin the second campaign. Additionally, this study illustrateswhy ground-based characterization is essential to supportmicrogravity experiments: the thermophysical propertymeasurements could only be interpreted correctly by com-bining several ground-based probes for atomic structureand microstructure. In cases where small, roughly spher-ical bubbles are formed, the density can be corrected byadjusting the volume to account for the voids. However,for the large voids like those in Figs. 3(a)–3(c), the voidslikely change size during cooling due to the temperature ofany entrapped gas and surface tension effects, so densitycorrection based on tomography is not reliable in thesecases.4. DiscussionThe ability to synthesize novel glasses and to accessdeeply supercooled liquids provides a useful way to inves-tigate the structure and properties of unconventional glassformers.The structure analysis for supercooled liquids showsthat linkedness in the Al–O networks is mostly corner-sharing with about 40% edge-sharing among five- and six-coordinated Al units in YbAG. If network connectivity iscompared using the Kn metric, which accounts for theeffects of five-coordinate Al, edge-sharing, and triply bond-ed oxygen, and the Qn approach, which only considerscorner-shared four-coordinate species, significant differ-ences are apparent. Molten Yb and La based RE3Al5O12composition materials exhibit similar polyhedra linkednessin a Qn analysis (i.e., when five-coordinate Al is excludedfrom network analysis). When five-coordinate aluminum isincluded in the analysis, the lower glass forming ability ofYbAG compared to LAG correlates with larger fractions offive- and six-coordinate aluminum ions, lower connectivityamong four-coordinated Al units, increased edge-sharinginvolving five-coordinated Al units, and a higher fractionof triply bonded oxygen.26) The analysis provides a tool forevaluation of glass forming tendency in fragile liquids.30,31)Fig. 3. Cross-sections of YbAG polycrystalline beads recovered from melt processing in microgravity.Replicates from (a–c) processed in air first and (d–e) processed in high purity argon second flight campaigns. Thedark areas are voids, the light areas are oxide sample.Weber et al.: Structure, properties and microgravity processing of liquids and glassesJCS-Japan358The Kn connectivity analysis approach is applicable acrossnetwork-forming oxide materials, and its applicability willbe discussed in future publications for other network form-ing oxides based on silicates, aluminates, borates, andtitanates. No evidence of proposed polyamorphic phasetransitions was found in the atomic structure data. It isnotable that the structure found in the glasses differs sig-nificantly from those of the equilibrium crystalline phases.This suggests that the energetics of glass formation will besignificantly different from some of the classical glassformers such as silicates where the non-equilibrium struc-tures are similar to the fully relaxed crystal but with freevolume due to bonding of polyhedra.The observation of multiple internal bubbles in the firstcampaign samples highlights an important opportunity andchallenge of space-based melt processing. The delicateinternal structures present in the recovered samples wouldnot have been possible to form in a terrestrial environmentbecause the buoyancy of the bubbles generally causesthem to rise to the top of a molten sample. Bubbles the sizeof the observed voids typically burst when they reach thesample surface during levitation melting on Earth. Thenear absence of gravity eliminates such buoyancy and itsassociated convection in the melt, thus stabilizing thepresence of bubbles. While this is potentially problematicfor thermophysical property measurements, it provides aunique processing capability. For example, it could beexploited to create glass forms with intentionally patternedbubbles or domains of a second phase of different den-sity.6,32) Such patterned materials may have useful appli-cations in optical devices that exploit cavity waveguidesand patterned structures with different refractive indexmaterials to control transmission of light.33) It is notablethat the large bubbles were only present in samples pro-cessed in air and not those processed in high purity argon.All flight samples were X-rayed before launch to check forlarge internal voids or cracks. None of the samples mea-sured exhibited large pores before they were processed, sothe formation of the large voids is likely an effect of theoxygen-containing atmosphere in the first flight campaign.In terms of materials processing in reduced gravity, thepresence of bubbles may not always be desirable. In caseswhere bubbles need to be eliminated, materials may needto be processed using artificial gravity (e.g. in a centrifuge)to separate phases with different densities. These findingshave practical implications for in-situ resource utilizationprocessing that involves liquid phase processing.5. ConclusionsContainerless liquid phase processing provides access tosupercooled melts and in some cases glasses that cannot bemade by conventional methods. In this work, we inves-tigated the structure and thermophysical properties ofmolten rare earth aluminates that have potential applica-tions in device materials.The structure of non-traditional network former glassesand liquid was found to exhibit significant differences incoordination and bonding from conventional glasses.Formation of edge shared four- and five-coordinatedcations and the presence of significant amounts of three-coordinated oxygen characterized the highly disorderedstructures.Experiments in microgravity revealed novel and repro-duceable behavior of gas bubbles in liquids. These behav-iors have potential for synthesis of new and useful non-equilibrium materials in microgravity.Acknowledgements This work was supported by theNational Aeronautics and Space Administration (NASA)through grant numbers 80NSSC19K1288 and80NSSC18K0059 and US Department of Energy grantnumber DE-SC0018601. This research was performed onAPS beam time award (DOI: https://doi.org/10.46936/APS-183682/60011543) from the Advanced Photon Source, a U.S.Department of Energy (DOE) Office of Science user facilityoperated for the DOE Office of Science by Argonne NationalLaboratory under Contract No. DE-AC02-06CH11357.References1) H. G. Pfaender, “Schott Guide to Glass”, Chapman &Hall, London (1996).2) W. Vogel, “Chemistry of Glass”, translated by N.Kreidel, Am. Ceram. Soc., Columbus, OH (1985).3) J. K. R. Weber, Int. J. Appl. Glass Sci. 1, 248 (2010).4) M. Kaneko, J. Yu, A. Masuno, H. Inoue, M. S. VijayaKumar, O. Odawara and S. Yoda, J. Am. Ceram. Soc.95, 79 (2021).5) P. Kidkhunthod, L. B. Skinner, A. C. Barnes, W.Klysubun and H. E. Fischer, Phys. Rev. B 90, 094206(2014).6) National Academies of Sciences, Engineering, andMedicine. 2023. Thriving in Space: Ensuring the Futureof Biological and Physical Sciences Research: ADecadal Survey for 2023–2032. Washington, DC: TheNational Academies Press.7) 2019 ISS R&D Conference Materials Science in SpaceWorkshop Report, July 29, 2019, Hyatt Regency,Atlanta, Georgia.8) L. B. Skinner, A. C. Barnes and W. Crichton, J. Phys.-Condens. Mat. 18, L407 (2006).9) J. K. R. Weber, C. J. Benmore, G. Jennings, M. C.Wilding and J. B. Parise, Nucl. Instrum. Meth. A 624,728 (2010).10) O. L. G. Alderman, C. J. Benmore, A. Lin, A.Tamalonis and J. K. R. Weber, J. Am. Ceram. Soc.101, 3357 (2018).11) M. C. Wilding, M. Wilson, O. L. G. Alderman, C.Benmore, J. K. R. Weber, J. B. Parise, A. Tamalonis andL. B. Skinner, Sci. Rep.-UK 6, 24415 (2016).12) S. Kohara, J. Akola, L. Patrikeev, M. Ropo, K. Ohara,M. Itou, A. Fujiwara, J. Yahiro, J. T. Okada, T.Ishikawa, A. Mizuno, A. Masuno, Y. Watanabe and T.Usuki, Nat. Commun. 5, 5892 (2014).13) T. Ishikawa, C. Koyama, H. Oda and H. Saruwatari, Int.J. Microgravity Sci. Appl. 39, 390101 (2022).14) W.-K. Rhim, K. Ohsaka, P. F. Paradis and R. E. Spjut,Rev. Sci. Instrum. 70, 2796 (1999).15) S. K. Chung, D. B. Thiessen and W.-K. Rhim, Rev. Sci.Instrum. 67, 3175 (1996).16) Y. Watanabe, A. Masuno and H. Inoue, J. Non-Cryst.Journal of the Ceramic Society of Japan 134 [5] 355-360 2026 JCS-Japan359https://doi.org/10.46936/APS-183682/60011543https://doi.org/10.46936/APS-183682/60011543Solids 358, 3563 (2012).17) S. K. Wilke, A. Al-Rubkhi, C. Koyama, T. Ishikawa, H.Oda, B. Topper, E. M. Tsekrekas, D. Möncke, O. L. G.Alderman, V. Menon, J. Rafferty, E. Clark, A. L.Kastengren, C. J. Benmore, J. Ilavsky, J. Neuefeind, S.Kohara, M. SanSoucie, B. Phillips and R. Weber, npjMicrogravity 10, 26 (2024).18) S. K. Wilke, O. L. G. Alderman, C. J. Benmore, J.Neuefeind and R. Weber, Sci. Rep.-UK 12, 8258 (2022).19) D. A. Keen, J. Appl. Crystallogr. 34, 172 (2001).20) C. J. Benmore, ISRN Mater. Sci. 2012, 1 (2012).21) C. J. Benmore, “X-ray and Neutron Diffraction fromGlasses and Liquids”, 3rd ed., Elsevier Inc. (2023).22) A. K. Soper, Chem. Phys. 202, 295 (1996).23) C. Koyama, T. Ishikawa, H. Oda, H. Saruwatari, S.Ueno, M. Oshio, Y. Watanabe and Y. Nakata, J. Am.Ceram. Soc. 104, 2913 (2020).24) K. Yoshida, H. Kumagai, T. Yamane, A. Hayashi, C.Koyama, H. Oda, T. Ito and T. Ishikawa, Appl. Phys.Express 15, 085503 (2022).25) J. Nawer, T. Ishikawa, H. Oda, H. Saruwatari, C.Koyama, X. Xiao, S. Schneider, M. Kolbe and D. M.Matson, npj Microgravity 9, 38 (2023).26) S. K. Wilke, C. J. Benmore, R. E. Youngman, B. J. A.Moulton, A. Al-Rubkhi and R. Weber, Phys. Chem.Chem. Phys. 27, 13464 (2025).27) S. K. Wilke, A. Al-Rubkhi, C. J. Benmore, J.Neuefeind, C. Koyama, T. Ishikawa, R. Shimonishiand R. Weber, J. Chem. Phys. 162, 124501 (2025).28) P. Wu and A. D. Pelton, J. Alloy. Compd. 179, 259(1992).29) S. Aasland and P. F. McMillan, Nature 369, 633 (1994).30) C. A. Angell, Science 267, 1924 (1995).31) H. Tanaka, J. Non-Cryst. Solids 351, 678 (2005).32) https://ntrs.nasa.gov/api/citations/20240000706/downloads/white%20paper%20-%20Non-equilibrium%20and%20Glassy%20Oxides.pdf.33) M. Kumar, T. Miura, Y. Sakurai and F. Koyama, Tun-able hollow optical waveguide and its applications, in“Frontiers in Guided Wave Optics and Optoelectronics”,Ed. by B. P. Pal, IntechOpen Limited, London, UK(2010) Chap. 2.Weber et al.: Structure, properties and microgravity processing of liquids and glassesJCS-Japan360https://ntrs.nasa.gov/api/citations/20240000706/downloads/white%20paper%20-%20Non-equilibrium%20and%20Glassy%20Oxides.pdfhttps://ntrs.nasa.gov/api/citations/20240000706/downloads/white%20paper%20-%20Non-equilibrium%20and%20Glassy%20Oxides.pdfhttps://ntrs.nasa.gov/api/citations/20240000706/downloads/white%20paper%20-%20Non-equilibrium%20and%20Glassy%20Oxides.pdf