# Fileset

[Ono_et_al-2025-NPG_Asia_Materials.pdf](https://mdr.nims.go.jp/filesets/b531327f-8aad-4180-8250-075bf2a2be4d/download)

## Creator

[Madoka Ono](https://orcid.org/0000-0002-1216-2731), Yasuhito Tanabe, Masaya Fujioka, Hiroki Yamada, [Koji Ohara](https://orcid.org/0000-0002-3134-512X), [Shinji Kohara](https://orcid.org/0000-0001-9596-2680), Masanori Fujinami, Junji Nishii

## Rights

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

## Other metadata

[Direct observation of the topological pruning in silica glass network; the key for realizing extreme transparency](https://mdr.nims.go.jp/datasets/1923fba1-9d66-4e43-a240-a19ce3bb3797)

## Fulltext

Direct observation of the topological pruning in silica glass network; the key for realizing extreme transparencyOno et al. NPG Asia Materials            (2025) 17:5 https://doi.org/10.1038/s41427-025-00589-5 NPG Asia MaterialsART ICLE Open Ac ce s sDirect observation of the topological pruning insilica glass network; the key for realizing extremetransparencyMadoka Ono 1,2, Yasuhito Tanabe2, Masaya Fujioka3, Hiroki Yamada4, Koji Ohara 4,5, Shinji Kohara 6,Masanori Fujinami7 and Junji Nishii2AbstractThe optical transparency of silica glass significantly improves when subjected to compression at its meltingtemperature. Using a rare hydrostatic iso-pressure apparatus capable of reaching 0.98 GPa at 1800 °C with Ar gas as thepressure medium, we obtained centimeter-sized glass samples, allowing us to measure various properties. Both thedensity and refractive index increased with pressure, while the refractive index dispersion decreased monotonically.However, Rayleigh scattering intensity, and small ring structures show a minimum around 0.8 GPa. High-energy X-rayscattering analysis indicates that the short-range structure, around 4 Å, governs the monotonic trends in the averagedphysical properties, such as density and refractive index. In contrast, non-monotonic changes are observed with thedisappearance of intermediate-range order at around 8 Å. This simplification of structural ordering is crucial forachieving extreme transparency in silica glass. The effect of suppression of the 8 Å order is well explained by thepredicted topological pruning phenomenon, where large voids and small unstable ring structures vanish, leading tothe minimal light scattering under high pressure. Our experimental findings also reveal that the optimal pressure forachieving this transparency is much lower than previously predicted, which makes the process more feasible for mass-production applications.IntroductionSiO2 is abundant on the earth’s crust, and its vitreousstate, silica glass, is an indispensable material for modernsociety as optical fiber1–3 and optics for lithography4,5.Optical fibers are undoubtedly the most importantinfrastructure in constructing the worldwide opticalcommunication network. Ever since silica glass fiber wasfirst invented in 1970, its transparency improved largelyfor the first ten years. However, it seems almost saturatedsince then, not because its specification reached a satis-factory level; the optical loss of fiber should be muchlower considering the next-generation communication,quantum cryptographic, and/or quantum information are,in principle, not amplifiable6,7. More than 80% of theremaining optical loss is attributed to Rayleigh scattering;the scattering caused by structural fluctuations in silicaglass8. Thus, it had long been believed that stabilizing the-Si–O- structure by thermal annealing, and/or chemicaldoping to accelerate the relaxation of the network is theonly solution9–11. In 2018, M. Ono et al. proposed a newmethod to suppress Rayleigh scattering using pressure12.The scheme is based on the shrinkage of the empty spacesin silica glass, called voids that can be considered as theparticles that scatter light as Rayleigh scattering13. Theapplication of pressure significantly suppressed the lightscattering loss. Molecular dynamics (MD) simulationspredicted the further reduction of Rayleigh scatteringcontinues until the pressure goes up to 4 GPa14. But it hadnever been examined experimentally due to the difficultyin isostatic hot compression of large-sized glass by using© The Author(s) 2025OpenAccessThis article is licensedunder aCreativeCommonsAttribution 4.0 International License,whichpermits use, sharing, adaptation, distribution and reproductionin any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate ifchangesweremade. The images or other third partymaterial in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to thematerial. Ifmaterial is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtainpermission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.Correspondence: Madoka Ono (madoka.ono.d7@tohoku.ac.jp)1Graduate School of Engineering, Tohoku University, Sendai, Miyagi 980-8579,Japan2Research Institute for Electronic Science, Hokkaido University, Sapporo 001-0020 Hokkaido, JapanFull list of author information is available at the end of the article1234567890():,;1234567890():,;1234567890():,;1234567890():,;http://orcid.org/0000-0002-1216-2731http://orcid.org/0000-0002-1216-2731http://orcid.org/0000-0002-1216-2731http://orcid.org/0000-0002-1216-2731http://orcid.org/0000-0002-1216-2731http://orcid.org/0000-0002-3134-512Xhttp://orcid.org/0000-0002-3134-512Xhttp://orcid.org/0000-0002-3134-512Xhttp://orcid.org/0000-0002-3134-512Xhttp://orcid.org/0000-0002-3134-512Xhttp://orcid.org/0000-0001-9596-2680http://orcid.org/0000-0001-9596-2680http://orcid.org/0000-0001-9596-2680http://orcid.org/0000-0001-9596-2680http://orcid.org/0000-0001-9596-2680http://creativecommons.org/licenses/by/4.0/mailto:madoka.ono.d7@tohoku.ac.jpgas media. In this study, we explored the higher-pressureregion by using a hydrostatic iso-pressure apparatuscapable of reaching 0.98 GPa at 1800 °C using Ar-gasmedia, which is one of the only two in the world15. Wefabricated large-sized glass samples at pressures above0.2 GPa and measured the Rayleigh scattering. Togetherwith the investigation of the properties and the structuralchange using various techniques such as Raman scatter-ing, high-energy X-ray scattering, and the refractive indexwith dispersion measurement, we experimentally observethe minimum Rayleigh scattering at slightly lower than1 GPa and the structural change similar to which waspredicted as topological pruning. This paper elucidatesthe crucial distinction between ascertaining whether thediverse physical attributes of silica glass, encompassing itsoptical characteristics, emanate from an averaged struc-ture centralized on 4 Å or arise from a medium-rangestructural order spanning approximately 8 Å.Materials and MethodsThe original glass used for pressure application is thesynthetic silica glass, AQ (product of AGC Inc, Japan)with OH concentration of approximately 50 wt. ppm,fluorine (F) and chlorine (Cl) of less than 1 wt. ppb. Thefictive temperature of this original silica glass was 1330 K.For hot compression process, the glass rods of 50 mm indiameter and 70 mm in length was put into a carboncrucible in the hydrostatic isothermal pressure machine(HIP) of JUTEM CO. Ltd, Japan15. The pressure mediumis Ar-gas. Due to the limitation of the machine, thepressure of 0.2, 0.5, 0.75, and at highest, 0.98 GPa wereapplied at 2073 K. After holding both the pressure and thetemperature for 1 hr, the electricity was shut down andthe temperature rapidly went down for 1000 K within10minutes, then cooled down to room temperature. Thepressure was released without any further compressionduring cooling. The example of the pressure and tem-perature profiles are shown in Fig. S1. For higher pressure(from 1 to 4 GPa), a solid cubic anvil high-pressuresynthesis method16 was used to compress the silica glassat 1373 K. The sample size for the solid pressure appa-ratus was 4.3 mm in diameter and 5mm in length at thelargest. The surface of the compressed sample pieces wastaken away to avoid a crystallized surface at the interfaceof the glass and the pressure medium. We preparedanother silica glass samples by changing their fictivetemperature, Tf, by heat treatment at 1340–1673 K beingkept for 2–960 h. The details of the preparation methodfor these glasses are explained in13. Tf of the sample wasdetermined from an infrared absorption peak at around2260 cm−1 17.The density was measured with the Archimedeanmethod using sample pieces larger than 10mm in length.The refractive index and its dispersion were measured atroom temperature using a refractometer (KPR-2000,Shimadzu, Kyoto, Japan) for blocks with 10mm or 5mmcube with six optically polished surfaces. Numerical dataof the plots in Fig. 1 are listed in Table. S1. Abbe numberis defined as (nD –1)/(nF – nC) with the refractive index nat each wavelength of D (587.56 nm), F (486.13 nm), andC (656.27 nm), respectively. Raman spectra were mea-sured using Raman spectroscopy (RENISHOW INVIA)under a microscope with 532 nm CW laser light. Polishedplates of 1 mm thick were used. Since Raman spectro-scopy is applicable to small-sized samples, we measuredglass that we quenched under 1 to 4 GPa at 1373 K usingFig. 1 Pressure dependence of density, refractive index and its dispersion. The pressure dependence of the hot compressed silica glass on (a)the density (black dots, plotted by the left axis) and the refractive index of silica glass (empty dots, plotted against the right axis) at d line (587.5 nm).b The dispersion of the refractive index of each sample. c shows the Abbe number which is defined in the context to quantify the amount of therefractive index dispersion. The higher the Abbe number, the smaller the wavelength dispersion of the sample.Ono et al. NPG Asia Materials            (2025) 17:5 Page 2 of 8     5 cubic multianvil apparatus. The glass made under 0.2 to0.98 GPa were by HIP at 2073 K. To examine the repro-ducibility and comparability, those glass made by HIPwere cross-measured by the other Raman measurementsystem (HORIBA T64000MW/vis-NIR) with non-pressurized glass for intensity normalization. The dataof 1373 K were found to be comparable with 2073 K. Thedetails of the measurement, analysis conditions and thedata connectivity are explained in Fig. S3.Rayleigh scattering intensity measurements were con-ducted with 5 × 5 × 5mm cubic sample pieces withoptically polished surfaces. They were immersed in arefractive index matching liquid (nD= 1.4580 at 298 K) ina cylindrical cell, to minimize stray light from the surfaceof the sample and the cell. The scattered light at thenormal angle to the incident laser light with a wavelengthof 488 nm was inserted into a 10-cm monochromator anddetected with a silicon PIN detector (ET-2030,Electronics-Optics Technologies Inc.) whose signal wasamplified by the current preamplifier (LI-76, NF Elec-tronic Instruments) and put into lock-in amplifier system(EG&G Park model 5209). The light intensity of twoidentical pieces, which were cut out from each HIPsample were measured several times to confirm the datareproducibility. The time-average signal over 100ms wasmeasured as the Rayleigh scattering intensity. To obtainthe absolute value of the Rayleigh scattering loss for HIPsilica glasses, we compared the intensity with that of thesilica glass samples with different Tf with no HIP treat-ment, Tf= 1342 and 1634 K. Further details of the Ray-leigh scattering measurement can be found also in12.The X-ray diffraction experiments were performed intransmission geometry using beamline BL04B2 at Spring-8 with an incident energy of 61.4 keV. The incident X-rayintensity was monitored using an Ar-filled ionizationchamber, and the scattered X-rays were detected by a Gedetector. A vacuum chamber was used to suppress airscattering, and the data was corrected using a standardprogram18. Each data set was normalized to give theFaber–Ziman total structure factor, S(q)19.Results and DiscussionsMonotonic change of properties by compressive pressureThe density and the refractive index of the silica glassshowed a monotonic increase when the applied pressureis increased. The closed and open circles in Fig. 1a cor-respond to the density and the refractive index, respec-tively. The samples were hot compressed at 2073 K.Pressure monotonically densifies the glass and induces ahigh refractive index. The highest densification in thisregion is 3%, while the refractive index increased by 1% at0.98 GPa. The change of the density is consistent to thepreviously reported value of around 5% at 1 GPa(1373 K)20. Figure 1b shows the refractive index at variouswavelengths for each sample. They are dispersion curvesof the refractive index of each sample with differentpressure. The detailed data of the refractive index is in Fig.S2. The cm-sized samples enabled us to obtain precisedata, which systematically changes by pressure. Figure 1cshows the Abbe number ν of each refractive index dis-persion curve. The value of ν linearly grows for higherpressure except for irregularity for ambient pressure of0.1MPa. The 0.1MPa sample is the original silica glasswhich had been annealed for more than one month toobtain low Tf, while other pressure-quenched samplesshould have higher values due to the rapid cooling of theHIP machine. Since low Tf leads to higher ν, irregularityhas occurred. The systematic increase of ν by increasingpressure, i.e. increasing refractive index, indicates thathomogeneity grew with the increase of the pressure asfound in ref. 21. They made quenched glass using solidpressure medium under higher pressure of 7.7 GPa atvaried temperature up to 1473 K. Our data covers asmaller region of the pressure, but within consistency totheir data. In ref. 21 they concluded that the increase of ν isdue to the homogeneity growth by pressure. The increaseof the refractive index and ν from ambient pressure up to0.98 GPa at 2073 K show the densification and thehomogenization of the electric dipoles which govern thebandgap of silica glass (corresponding to the chargetransfer transition from 2p-band of Oxygen to Si d-band)occur monotonically by pressure.Non-monotonic change of properties by compressivepressureOn the contrary to the monotonic pressure dependenceof density and the Si-O charge transfer oscillator, thestructure of longer-distance, such as three-membered SiO2rings, showed non-monotonous trends against pressure.Figure 2a shows the Raman scattering data of the hotcompressed glass. The data of the HIP-compressed sam-ples are plotted with circles, while the ones compressed at1373 K using multianvil cell pressure machine with solidmedium of tungsten carbide (plotted with triangles) whichcan cover up to 4 GPa. Although the process temperatureis not as high as 2073 K, they showed good connectivity at1 GPa [S3]. The connectivity goes well with ref. 20 whichindicated that temperatures higher than 1173 K lead tosimilar structural changes by pressure. The effective samplesize of the multianvil samples with the polished vitrifiedsurface is small (smaller than 2mm) due to the surfacecrystallization from the attached pressure media. However,the Raman spectroscopy in-situ heating is possible untheder microscope on the polished vitrified region. The insertfigure in Fig. 2a shows the Raman spectra with main peakof 450 cm−1, small D1 (500 cm−1) which corresponds tothe number density of 4-membered rings, while D2(610 cm−1) peak is attributable to the number density ofOno et al. NPG Asia Materials            (2025) 17:5 Page 3 of 8     5 3-membered SiO2 ring structures22. The increase of thesmall ring structures in silica glass is known as the sign ofinstability of the glass network structure22,23. Thus, thenumber density of 3-membered ring structures is analyzedfrom the area of the D2 peak [S3]. They are plotted againstpressure in Fig. 2a. The area of the D2 peak minimized ataround 1GPa, which indicates that the 3-membered ringstructure decreases by pressure up to 1 GPa from ambientpressure. Then, it increases for higher pressure from 1GPaup to 4 GPa. The result indicates the structure becomesmost stable at around 1 GPa. The color shows the tem-perature at which each Raman measurement is done. Opencircles in Fig. 2a are the area of the D2 peak measured atroom temperature. The temperature dependence of thenumber density of 3-membered ring structure, which areshown by different colored dots is also the smallest at1 GPa. The least variance at 1 GPa against temperatureimplies the structure becomes stable against heat at thispressure. D1 peak intensity shows similar minimum at1 GPa, but since D1 peak is hardly deconvoluted from themain peak, and 4-membered ring structure is not alwaysthought of as the representative of the unstable structure23,we relied onto the indication of the pressure- andtemperature-dependence of the D2 peak.Fig. 2 Optical properties of hot compressed silica glass with non-monotonic pressure dependence. a Number density of the 3-memberedring in silica glass, estimated from the area of D2 peak normalized by ω3 peak height. The insert figure shows the Raman scattering spectra of theoriginal silica glass made under ambient pressure (black) with the one quenched at 4 GPa (red). The data for P= 1, 2, 3, and 4 GPa are of the samplemade by using solid pressure medium at 1373 K (triangles). While for P= 0.2 to 0.98 GPa are made by HIP at 2073 K (open and closed circles). Data forP= 0.1 MPa corresponds to the original glass made under ambient pressure (squares). The color of the point shows the temperature at which wemeasured the Raman spectra. Temperatures at which we measured in-situ are the room temperature (black), 100 °C (light blue), 200 °C (green), 300 °C(orange), 400 °C (red), 500 °C (dark red). b Rayleigh scattering coefficient of the silica glasses hot compressed with Ar gas media. Filled circles are newdata while the open ones are taken from ref. 12. Note that the pressure range of lower figure is up to 1 GPa, while upper axis is up to 4 GPa.Ono et al. NPG Asia Materials            (2025) 17:5 Page 4 of 8     5 Figure 2b shows the pressure dependence of the Ray-leigh scattering coefficient of the HIP silica glass withpressure from ambient to 0.98 GPa. The filled dots inFig. 2b correspond to the newly obtained data in thispaper. The details of the measurement scheme can befound in ref. 12. The open circles are taken from ref. 12,correspond to the data of samples made under pressurefor a duration time of 2 hours. The previous values areconsistent with the newly obtained data. It is found thatthe Rayleigh scattering coefficient decreases for pressurehigher than 0.2 GPa, but a broad minimum exists ataround 0.5–0.8 GPa. Within the data of the pressureduration time of 1 hr, the minimum Rayleigh scatteringcoefficient is 0.4 dB/km/μm−4, which corresponds to0.1 dB/km at 1.55 μm12. The minimum value is similar tothose obtained for experimental conditions of 0.2 GPa for3.5 h of pressure duration time. It is noteworthy thatRayleigh scattering rapidly decreases by pressure of about0.2 GPa, but above that pressure, the derivative is notsignificantly large. From positron annihilation lifetimespectroscopy (PALS), the void size of these hotcompressed glasses showed small values for 0.2 GPa and0.98 GPa compared to that of ambient glass (initial glassbefore hot compression) [S4]. However, saturation of thevoid sizes at pressure over 0.2 GPa was also observed. Thepressure dependence was not as clear as Raman andRayleigh. The results of PALS are plotted in Fig. S4.Overall, the trend observed by Raman and Rayleighindicate that silica glass which is hot compressed ataround 1 GPa or slightly less has stable and homogeneousintermediate range structure.Structure and ordering change by pressureThe presence of two distinct types of pressure depen-dence suggests that two different structural changes occurin the silica glass network as the pressure is altered. Toinvestigate the structural factors responsible for thepressure dependence of these properties, we performedhigh-energy X-ray scattering (XRS) measurements atSPring-8, using the BL04B2 beamline. The black curve inFig. 3a shows the observed X-ray total structure factorS(q) calculated from the XRS pattern of the initial silicaglass with ambient pressure (0.1 MPa). The red curve inFig. 3a is that of 0.98 GPa. As is seen, XRS patterns for qvalues larger than 2 Å−1 are almost identical. On thecontrary, the S(q) in the range of q= 0 to 2 Å⁻¹ showssystematic changes with pressure, as seen in Fig. 3b, whichis an expanded view of Fig. 3a. The peak called FSDP(First Sharp Diffraction Peak of q around 1.5 Å−1)24,25shifts to higher q by increasing pressure. The peak posi-tion and width of the FSDP as a function of pressure wereanalyzed by fitting Lorentzian curves and are shown inFig. 3c, d, respectively. FSDP is attributed to the inter-mediate range ordering of the neighboring Si-O-Sichains26,27. Thus, the higher q shift of FSDP peak corre-sponds to the decrease of the intrachain distance bypressure, which is reasonable. The width of FSDPbecomes smaller with increased pressure. Such sharpen-ing of FSDP is previously reported as the indication ofhomogenization20,28,29. They reported linear change in qposition and FWHM (full-width half maximum) of FSDPoccurs for pressure from ambient up to 8 GPa at 1373 Kand 1473 K. But in our case, we observe a small kink ontheir trend at 0.75 GPa, even though they are roughlylinear to the pressure. This observation was made possibleby the precisely controlled pressure conditions and thehigh accuracy of the XRS measurement technique.Furthermore, we clearly observed the disappearance ofthe peak at the shoulder of FSDP (we call it SP, which islocated at q= 0.8 Å−1), by increasing pressure. Theobservation was successful for the first time owing to thehigh accuracy of the experimental techniques. Since theXRS pattern for 0.98 GPa sample seems to have only asingle FSDP with no sign of an additional shoulder peakat 0.8 Å−1, we subtracted the XRS patterns of 0.1 to0.75 GPa by that of 0.98 GPa. XRS spectra with noobvious sign of SP is typical of silica glass with higherpressure24–29. The insert figure of Fig. 3e shows thesubtracted patterns. The disappearance of SP by pres-sure is prominent at the lower pressure region, andgradually saturates around 0.75 to 0.98 GPa. It is indi-cative of the structural ordering with the scale of around8 Å, disappears by pressure within this range. The 8 Åscale can be interpreted as the distance between atomicstructures that form voids in silica glass, which havebeen observed to be approximately 5 to 6 Å in diameter[13, S4], with electron wavefunctions surrounding theatoms. XRS corresponds to the position of the nucleus,while positronium observes the ionic radius as the wallof the voids. Notably, the disappearance of SP and thelocal minimum of FWHM of FSDP occurs at a pressureslightly lower than 1 GPa (around 0.8 GPa), whichcoincides with the minimum of Rayleigh scattering andthe number density of 3-membered rings. In Fig. 4,schematic pictures of the structure and voids in silicaglass corresponding to the pressure region are drawn.From the experimental results which have a minimum ataround 0.8 GPa, it is likely that the structure grows to behomogeneous by the extinction of large voids and small(3- and 4-membered) unstable rings by increasingpressure, as shown by the two left figures in Fig. 4. Thetrend saturates at pressure slightly lower than 1 GPa.When pressure is higher than 1 GPa, homogeneity ofstructures that sharpen FSDP (scale of 4 Å) continues togrow as shown by the homogeneous size of the voids inFig. 4. But higher pressure generates 3- and 4 memberedrings and their stability against heat gradually decreases.Due to the generation, densified parts should appear at aOno et al. NPG Asia Materials            (2025) 17:5 Page 5 of 8     5 pressure higher than 1 GPa. Figure 4 indicates thatdensity and refractive index determined by the structurepart (colored part in Fig. 4) are supposed to increasecontinuously with higher pressure, while large voids andsmall rings should show different pressure dependence.Such behavior of the minimization of the number ofFig. 4 Schematic pictures of the silica glass structure hot compressed under each pressure range at 2073 K. Colored parts show where theSiO2 network occupies the space while white-painted circles show voids.Fig. 3 The X-ray total structure factors, S(q) of of hot compressed silica glasses. a S(q) of initial glass (black) compared to the one hotcompressed under 980 MPa (red). b S(q) of glasses hot compressed under each denoted pressure (the numbers are in MPa unit). q range is 0 to 2 Å−1.The peak around 1.5 Å−1 is FSDP (First Sharp Diffraction Peak), while the one at the arrow is SP (Shoulder Peak). c The peak position of FSDP for eachS(q) profile. d shows the HWHM (half width of half maximum) of the FSDP. The example of fitting FSDP with a Lorentzian curve is shown in the insertof Fig. 3d. e The difference of the SP peak height from that of 980 MPa. The insert figure shows ΔS(q) = (S(q) of each spectrum) – (S(q) of 980 MPa).Ono et al. NPG Asia Materials            (2025) 17:5 Page 6 of 8     5 3-membered rings together with large voids at a peculiarpressure qualitatively matches to what is predicted astopological pruning14. In the simulated model, smallrings appear close by large voids, and they disappear inpairs when pressure is applied. The structure is homo-genized due to the reconstruction of the structures. MDpredicted the silica glass network is topologically prunedto the most at the optimal pressure of 4 GPa. In thiswork, we observed the topological pruning phenomenaexperimentally. However, the pressure at which the voidand 3-membered rings disappeared was 0.8 GPa and ismuch lower than 4 GPa. This experimental finding givesa delight future to realizing low-loss silica core glassfiber because 1 GPa is somewhat achievable pressure forindustry; for example, rolling contact pressure formanufacturing steels can be 2 GPa, even though thetemperature is not as high as 2000 K30,31.In comparison to the behavior of SP, the sharpening ofFSDP, increase of density, refractive index and abbenumber show no minima. They continue to change belowand above 1 GPa. It is also interesting to recall the decayconstant of density and refractive index under 0.2 GPa isvery fast, less than 1 milli second at melting temperature,while those of void radius and Rayleigh scattering inten-sity were found to be longer than 1 hour12. Since SPcorresponds to longer distance ordering (8 Å) comparedto FSDP of intermediate distance ordering (4 Å), the trendof longer relaxation time for longer distance ordering isunderstandable. Overall, the experimental findings in thiswork strongly indicate the presence of two distinct typesof structure order with different relaxation constants.So far, the structures and properties of silica glassquenched under pressure of more than 1 GPa using asolid medium were investigated. With the bulk glasssamples with large size obtainable by HIP, it was possibleto observe the minimization of Rayleigh scattering and thegrowth and maximization of homogeneity at the pressureat slightly less than 1 GPa. The longer distance (8 Å)ordering seems to disappear under pressure above 1 GPa.However, the shorter distance ordering shows a con-tinuous increase for higher pressure. This has been indi-cated for different temperatures as 1273 and 1373 K20,25,the intermediate to short-range order structure whichrelates to optical oscillator strength is supposed to growits homogeneity above 1 GPa and its distribution is sup-pressed at least up to 8 GPa. By the topological pruning, itis predicted that generation of the 5- to 6-coordination ofSilicon sites instead of 4 starts to happen above 4 GPa andtriggers inhomogeneity in the glass structure. Consideringthat 5 to 6-coordinated silicon is generated under 20 GPaat room temperature32, but none exists up to 7.7 GPa at1573 K29,33, whether the coordination number of Si wouldincrease at a pressure above 1 GPa at melting temperatureshould be investigated.SummaryThrough measurements of density, refractive index andits dispersion, Rayleigh scattering, Raman scatteringspectra, and XRS of hot compressed silica glass, we suc-cessfully identified two distinct pressure-dependent opti-cal properties, each arising from different structuralordering scales. The optical Rayleigh scattering is mini-mized at pressure around 0.8 GPa, which is coincident tothe minimization of the 3-membered ring structuredetected by Raman spectroscopy. They are well-explainedas topological pruning phenomena, through XRS mea-surement which revealed the extinction of 8 Å ordering atthe similar pressure. PALS does not contradictthese results. The temperature dependence of 3- and4-membered rings-structure also becomes minimum,indicating the structure becomes most stable at thispressure. Such homogeneous structure with stability waspredicted to occur at 4 GPa molecular dynamics simula-tion as “topological pruning”, and we observed directlyand experimentally the phenomena. But the experimentalresults have a plausible implication that the topologicalpruning, in reality, happens at slightly lower pressure than1 GPa. The value is realistic for industrial application. Itwill be interesting to explore how shorter ordering scalecontinuously homogenize silica glass structure above1 GPa. But the local homogeneity which determines Si-Ooscillators, which corresponds to the width of the FSDP inXRS spectra, is independent of the long-ranged opticalscattering which determines the quality and loss of silicaglass fiber. Overall, this work reveals two distinctpressure-dependent phenomena in silica glass for hotcompression: the local homogeneity of the long-rangeorder structure, on the scale of about 8 Å correspondingto voids and related to Rayleigh scattering, occurs below1 GPa; local and microscopic homogeneity increases from1 GPa to around 4 GPa. The pressure dependence of eachstructural ordering scale is independent and cannot bedetected solely through FSDP observations.AcknowledgementsM. Ono thank Dr. N. Nakagawa of JUTEM CO Ltd. for his support on high-pressure HIP machines. This work is supported by Japan Society for thePromotion of Science (JSPS) for its support through KAKENHI Grant Nos.20H05880, 21H01835, 21K19016, and 24K01371. The high-energy X-ray totalscattering experiments at SPring-8 were approved by the Japan SynchrotronRadiation Research Institute under proposal nos. 2020A1698, 2021A1189, and2022A1261.Author details1Graduate School of Engineering, Tohoku University, Sendai, Miyagi 980-8579,Japan. 2Research Institute for Electronic Science, Hokkaido University, Sapporo001-0020 Hokkaido, Japan. 3National Institute of Advanced Industrial Scienceand Technology (AIST), Aichi 463-8560, Japan. 4Japan Synchrotron RadiationResearch Institute (JASRI), Kouto, Hyogo 679-5198, Japan. 5Faculty of Materialsfor Energy, Shimane University, Matsue, Shimane 690-8504, Japan. 6NationalInstitute for Materials Science (NIMS), Tsukuba, Ibaraki 305-0047, Japan.7Department of Applied Chemistry & Biotechnology, Chiba University, Inage,Chiba 263-8522, JapanOno et al. NPG Asia Materials            (2025) 17:5 Page 7 of 8     5 Author contributionsM.O. designed the study, took primary responsibility for writing and revisingthe manuscript. Y. T. contributed to the preparation of samples, measurement,and analyses. M. F. supervised multianvil cell pressure machine tests, while H. Y.and K. O. supported experiments at Spring-8. S. K. performed the X-rayscattering spectroscopy and its analysis. M. F. conducted and analyzed thepositron annihilation lifetime spectroscopy. J. N. revised the manuscript. Allauthors discussed the results and commented on the manuscript.Data availabilityData is available upon request to the corresponding author, M. O.Conflict of interestWe declare there is no conflict of financial interests in relation to the workdescribed here.Ethics approval and consent to participateThis paper is not applicable as it does not report on or involve any animals,humans, human data, human tissue or plants.Publisher’s noteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.Supplementary information The online version contains supplementarymaterial available at https://doi.org/10.1038/s41427-025-00589-5.Received: 22 October 2024 Revised: 20 December 2024 Accepted: 21January 2025References1. Crisp, J. & Elliott, B. Introduction to Fiber Optics. Third Edition Elsevier Ltd. (2005).2. Kanamori, H. Fifty Year History of Optical Fibers. SEI TECHNICAL REVIEW 91,15–22 (2020).3. Kanamori, H. Transmission of loss of Optical fibers; Achievements in halfcentury. IEICE Trans. Commun. E104–B, 922–933 (2021).4. Ikuta, Y. et al. New silica glass (AQF) for 157-nm lithography. Optical Micro-lithography XIII, (5 July 2000); https://doi.org/10.1117/12.388990.5. Toombs, J. T. 3D PRINTING Volumetric additive manufacturing of silicaglass with microscale computed axial lithography. Taylor Sci. 376,308–312 (2022).6. Dieks, D. Communication by EPR devices. Phys. Lett. A 92, 271–272 (1982).7. Azuma, K. et al. Quantum repeaters: From quantum networks to the quantuminternet. Rev. Mod. Phys. 95, 045006 (2023).8. Agrawal, G. P. Optical Communication: Its History and Recent Progress. In: Al-Amri, M., El-Gomati, M., Zubairy, M. (eds) Optics in Our Time. Springer, (2016).9. Saito, K. et al. Limit of the Rayleigh Scattering loss in silica fiber. Appl. Phys. Lett.83, 5175 (2003).10. Saito, K., Kakiuchida, H. & Ikushima, A. J. Light-scattering study of the glasstransition in silica, with practical implications. J. Appl. Phys. 84, 3107–3112(1998).11. Saito, K., Kakiuchida, H. & Ikushima, A. J. Fictive temperature dependence ofthe density fluctuations. J. Appl. Phys. 94, 4824–4827 (2003).12. Ono, M., Aoyama, S., Fujinami, M. & Ito, S. Significant suppression of Rayleighscattering loss in silica glass formed by the compression of its melted phase.Opt. express 26, 7942–7948 (2018).13. Ono, M., Hara, K., Fujinami, M. & Ito, S. Void structure in silica glass with differentfictive temperatures observed with positron annihilation lifetime spectro-scopy. Appl. Phys. Lett. 101, 164103 (2012).14. Yang, Y., Homma, O., Urata, S., Ono, M. & Mauro, J. C. Topological pruningenables ultra-low Rayleigh scattering in pressure-quenched silica glass. npjComput. Mater. 6, 139 (2020).15. JUTEM CO Ltd., Yamaguchi, Japan; https://www.jutem.co.jp/.16. Fujioka, M. et al. High-Pressure Diffusion Control: Na Extraction from NaAlB14.Chem. Mater. 35, 3008–3014 (2023).17. Agarwal, A., Davis, K. M. & Tomozawa, M. A simple IR spectroscopic method fordetermining fictive temperature of silica glasses. J. Non-Cryst. Solids 185,191–198 (1995).18. Ohara, K., Onodera, Y., Murakami, M. & Kohara, S. Structure of disorderedmaterials under ambient to extreme conditions revealed by synchrotron x-raydiffraction techniques at SPring-8—recent instrumentation and synergic col-laboration with modelling and topological analyses. J. Phys.: Condens. Matter33, 383001 (2021).19. Faber, T. E. & Ziman, J. M. A theory of the electrical properties of liquid metalsIII. the resistivity of binary alloys. Philos. Mag.: A J. Theor. Exp. Appl. Phys. 11,153–173 (1964).20. Guerette, M. et al. Structure and Properties of Silica Glass Densified in ColdCompression and Hot Compression. Sci. Rep. 5, 15343 (2018).21. Masuno, A. et al. Higher refractive index and lower wavelength dispersion ofSiO2 glass by structural ordering evolution via densification at a higher tem-perature. RSC Adv. 6, 19144–19149 (2016).22. Geissberger, A. E. & Galeener, F. L. Raman studies of vitreous SiO2 versus fictivetemperature. Phys. Rev. B 28, 3266–3271 (1983).23. Uchino, T., Kitagawa, Y. & Yoko, T. Structure, energies, and vibrational prop-erties of silica rings in SiO2 glass. Phys. Rev. B 61, 234–240 (2000).24. Susman, S. et al. Intermediate-range order in permanently densified vitreousSiO2: A neutron-diffraction and molecular-dynamics study. Phys. Rev. B 43,1194 (1991).25. Inamura, Y., Katayama, Y., Utsumi, W. & Funakoshi, K. Transformations in theIntermediate-Range Structure of SiO2 Glass under High Pressure and Tem-perature. Phys. Rev. Lett. 93, 015501 (2004).26. Mei, Q., Benmore, C. J. & Weber, J. K. R. Structure of Liquid SiO2: A Measure-ment by High-Energy X-Ray Diffraction. Phys. Rev. Lett. 98, 057802 (2007).27. Hirata, A. et al. Direct observation of the atomic density fluctuation originatingfrom the first sharp diffraction peak in SiO2 glass. NPG Asia Mater. 16, 25 (2024).28. Onodera, Y. et al. Understanding diffraction patterns of glassy, liquid andamorphous materials via persistent homology analyses. J. Ceram. Soc. J. 127,853–863 (2019).29. Onodera, Y. et al. Structure and properties of densified silica glass: char-acterizing the order within disorder. NPG Asia Mater. 12, 85 (2020).30. Uppaluri, R. & Helm, D. Thermomechanical characterization of 22MnB5 steelswith special emphasis on stress relaxation and creep behavior. Mater. Sci. Eng.:A 658, 301–308 (2016).31. Okayasu, M. & Yang, L. Influence of microstructure on the mechanical prop-erties and hydrogen embrittlement characteristics of 1800 MPa grade hot-stamped 22MnB5 steel. J. Mat. Sci. 54, 5061–5073 (2019).32. Sato, T. & Funamori, N. High-pressure structural transformation of SiO2 glass upto 100 GPa. Phys. Rev. B 82, 184102 (2010).33. Sato, S. et al. Synthesis of hyperordered permanently densified silica glasses byhot compression above the glass transition temperature. J. Ceram. Soc. Jpn.132, 427–433 (2024).Ono et al. NPG Asia Materials            (2025) 17:5 Page 8 of 8     5 https://doi.org/10.1038/s41427-025-00589-5https://doi.org/10.1117/12.388990https://www.jutem.co.jp/ Direct observation of the topological pruning in silica glass network; the key for realizing extreme transparency Introduction Materials and Methods Results and Discussions Monotonic change of properties by compressive pressure Non-monotonic change of properties by compressive pressure Structure and ordering change by pressure Summary Acknowledgements Acknowledgements