optical properties of geo_2 glass and optical fibers

Optical properties of GeO2 glass and optical fibers

Shigeki Sakaguchi and Shin-ichi Todoroki

The optical properties of GeO2 glass were investigated to clarify its potentiality as an optical fibermaterial. Glass samples were prepared by the flame hydrolysis and the melting techniques, and theirRayleigh scattering and infrared absorption properties were examined. Optical fibers composed of apure GeO2 core and an F-doped GeO2 cladding were drawn to clarify the scattering loss characteristics.The Rayleigh-scattering intensity obtained from spectral loss measurements on the fibers agreed withthat measured in bulk samples, and the intensity relative to that of SiO2 glass was found to be approx-imately 3. These results suggest that a minimum loss of 0.15 dBykm is expected at a wavelength of 2mm. © 1997 Optical Society of America

Key words: GeO2 glass, optical fiber, Rayleigh scattering, infrared absorption, optical loss.

1. Introduction

It is known that GeO2 glass has advantageous opticalproperties such as superior transparency in the mid-infrared region1,2 and a large nonlinearity3 comparedwith SiO2 glass. GeO2 glass has attracted particu-lar attention as a potential material for optical fiberswith less transmission loss than SiO2 fibers.1,2 Be-cause of this, fabrication techniques for GeO2 fibershave been developed and Sb-doped GeO2 fibers havebeen realized.4 However, the light-transmissioncharacteristics measured in the fibers do not seem toreflect the properties expected on the basis of bulk-glass measurements.4 The light-scattering~Rayleigh-scattering! loss seems to be especially highcompared with the estimation. Therefore it isworthwhile clarifying the optical properties of GeO2glass, including Rayleigh scattering and infrared ab-sorption, in both bulk samples and optical fibers toconfirm its potentiality.

Rayleigh scattering, which is dominated by densityfluctuation in single-component glasses such asGeO2, is related to glass properties5 such as refractiveindex and glass transition temperature.6 It is pos-sible to evaluate Rayleigh scattering in relation torefractive-index dispersion.7 In addition, the scat-tering appears as wavelength-dependent4 spectral

The authors are with the Opto-electronics Laboratories, NipponTelegraph and Telephone Corporation, Tokai, Ibaraki 319-11, Ja-pan.

Received 4 March 1996; revised manuscript received 20 Novem-ber 1996.

0003-6935y97y276809-06$10.00y0© 1997 Optical Society of America

loss in optical fibers.8 One can confirm the inherentscattering intensity by comparing the intensity inbulk samples with that in fibers. To this end it isnecessary to fabricate pure GeO2 core fibers.

Although pure GeO2 core fibers have been devel-oped with SiO2-doped GeO2 cladding for the purposeof Raman amplification, they were insufficient for usein clarifying the scattering properties because theirtransmission loss was too high.9 It seems that theformation of SiO2-doped GeO2 cladding suffers fromexcess scattering owing to a certain structural changein the glass such as a Ge coordination shift from 4 to6 ~Ref. 10!. By contrast, F doping appears to beuseful because the reduction in the refractive indexhas been demonstrated although it was not high.11

This report deals with optical properties in GeO2glass and optical fibers focusing on Rayleigh scatter-ing. Pure GeO2 core fibers with F-doped claddingwere prepared by the flame hydrolysis technique.We examined the scattering properties in the fibersby comparing them with those in bulk-glass samples.

2. Experimental Procedure

A. GeO2 Glasses

We prepared glass samples by the flame hydrolysistechnique: GeCl4 was hydrolyzed in an H2–O2 flameto produce glass fine particles ~soot!, and soot bodieswere formed by depositing the soot on a rotating seedin an axial direction.4 They were consolidated intotransparent glass rods in a flow of He–O2 ~flow ratio,1:0.1! at ;850 °C.12 We obtained glass rods thatwere typically 8 mm in diameter and 80 mm inlength.

20 September 1997 y Vol. 36, No. 27 y APPLIED OPTICS 6809

For comparison we also prepared glass samples forthe scattering measurement by a conventional melt-ing method. We melted high-purity GeO2 powder~purity . 99.9999%! using Pt crucibles at a temper-ature of 1550 °C for 1 hour and subsequently cooledto room temperature at a rate of 20 °Cymin ~furnacecooled below 800 °C!. Glass rods obtained were op-tically polished to approximately 6 mm 3 6 mm 3 20mm in size.

We measured the Rayleigh scattering in these sam-ples by launching vertically polarized Ar-laser lightat the 488-mm wavelength.13 Scattered light inten-sities from the samples were measured every 5° inthe 30°–150° scattering angle region by a photomul-tiplier with an analyzer. No filter was used for lightdetection. The samples were immersed in di-n-butyl-phthalate ~nD, 1.49! to minimize the effect ofrefraction and reflection at the sample surface. Themeasured intensities were compared with that ofSiO2 glass, which is used for optical fiber application,as a standard.14

Infrared absorption measurement for GeO2 glassprepared by flame hydrolysis was performed with acommercial Fourier transform infrared spectrometer.Specimens cut from a single glass rod were polishedto various thicknesses of 0.5, 5, and 50 mm.

B. GeO2 Fibers

Optical fibers were drawn from preforms composed ofa pure GeO2 core and an F-doped GeO2 cladding.The cladding layer was formed by overcoating soot ona transparent core glass rod by flame hydrolysis and,subsequently, consolidating it in an He–SF6 flow~flow ratio, 1:0.04! at ;850 °C. The drawn fiberswere coated in-line with thermally curable siliconeresin.

We measured the transmission loss of fibers with alength of approximately 500 m using a conventionalcut-back method. A light from a halogen lamp waslaunched into the fiber through a monochrometer andwas detected by a Si–Ge composite photodetector.

3. Results

A. Rayleigh Scattering

Figure 1 shows examples of the angular distributionsof the scattered light intensity measured in the GeO2glasses. The distribution for SiO2 glass is alsoshown for comparison. The intensities are almostconstant independent of scattering angle, indicatingthat the glasses are optically homogeneous.15 Theintensity of GeO2 glass prepared by melting isslightly lower than that of glass prepared by flamehydrolysis.

We evaluated the scattering intensity for theseglasses with the relative intensity, Rsc 5 R90yR90~SiO2!, where R90 is the scattered light intensity at90°. We determined the R90 value by taking an av-erage of intensities in the range of 45°–135° in thepresent experiment because the vertically polarizedlight is launched so that the scattered light intensityshould be constant as long as the glass is homoge-

6810 APPLIED OPTICS y Vol. 36, No. 27 y 20 September 1997

neous. The Rsc value obtained from six glass sam-ples prepared by flame hydrolysis is 3.13 1 ~10.32y20.40!. With the glass prepared by the meltingmethod, the Rsc value obtained from two samples is2.66 6 0.08.

B. Infrared Absorption

Figure 2 shows the infrared absorption spectrum forGeO2 glass. The spectrum for SiO2 glass, which isthe same material as that used as a standard for thescattering measurement, is also shown for compari-son. From this spectrum the infrared absorption forGeO2 glass is represented by the dotted line and isgiven as

aIR 5 0.4 3 1011 exp~258yl!, (1)

where aIR is the infrared absorption loss in decibelsper kilometer and l is the wavelength in microme-

Fig. 1. Angular distributions of scattered light intensities ob-tained from GeO2 glass.

Fig. 2. Infrared absorption properties measured in GeO2 glassprepared by flame hydrolysis. ~a! GeO2 glass ~this research!, ~b!GeO2 glass ~published data!, and ~c! SiO2 glass ~this research!.Dotted lines indicate the estimated infrared absorption tail.

ters. Although the absorption is very close to thepublished data of 1.0 3 1011 exp~260yl!, ~Ref. 16!, itis slightly small.

C. Spectral Loss of Optical Fiber

Four fibers with pure GeO2 core–F-doped GeO2 clad-ding were successfully drawn, and their transmissionloss spectra were measured. An example of a crosssection of a fiber coated with silicone resin is shown inFig. 3. Although we tried to measure the refractive-index difference with a commercial preform analyzer,we could not obtain a reliable value. It is probablethat the refractive-index difference of approximately0.2% is formed.11

Figure 4 shows an example of the spectral loss of afiber. The loss is measured in the wavelength rangeof 0.55–1.4 mm. Because of a large OH absorptionnear 1.4 mm, the loss spectrum is not obtained atwavelengths above 1.4 mm. The minimum loss of10.7 dBykm is obtained at 1.05 mm. The minimum

Fig. 3. Cross section of an optical fiber with a pure GeO2 core andF-doped GeO2 cladding. The core and cladding of the fiber are 80and 125 mm in diameter, respectively, and the fiber is coated withsilicone resin.

Fig. 4. Spectral loss obtained from an optical fiber with a pureGeO2 core and F-doped GeO2 cladding.

losses for the four fibers are scattered in a range10–20 dBykm, and they are observed at approxi-mately 1 mm.

Figure 5 shows the plot of transmission loss versusl24 for the fiber shown in Fig. 4. As shown in Fig. 5,a linear relation is obtained as a 5 Ayl4 1 B, whereA and B are experimental constants. The slope A forthis fiber can be determined as 2.33 6 0.04 ~dBykm!mm4. The l-independent loss B is 8.4 6 0.2 dBykm.

Regarding this fiber, a scattering measurementwas performed for the residual neckdown part of thepreform. The obtained Rsc value was 3.16. Thisvalue is also close to that obtained from bulk samples.

The slope representing l24-dependent loss ob-tained from the four fibers falls in a narrow bandfrom 2.23 to 2.92 ~dBykm! mm4 with an average valueof 2.54 ~dBykm! mm4. The l-independent loss varieslargely from 8.4 to 19.9 dBykm for the four fibers.

4. Discussion

We evaluated the Rayleigh-scattering intensity forGeO2 glass prepared by the flame hydrolysis tech-nique using an intensity relative to that of SiO2 glass,Rsc, and the Rsc value of 3.13 is obtained. In thetransmission loss of fibers, the Rsc is reflected in theratio of the slope for loss-l24 plot. For example, aslope value of 2.33 ~dBykm! mm4 is obtained for anoptical fiber with a pure GeO2 core as shown in Fig. 5.By comparing this value with the value of 0.76 ~dBykm! mm4 obtained for a pure SiO2 core fiber,17 weobtained an Rsc value of 3.07. This value is close tothat obtained from bulk samples. The averagevalue, 2.54 ~dBykm! mm4, for the slope obtained fromthe four fibers corresponds to 3.34. This value isvery close to the value of 3.13 obtained from bulksamples.

As described above, the agreement of Rsc betweenbulk glass and fibers indicates that the inherent scat-tering intensity relative to that of SiO2 glass fallsnear 3. With the relative intensity of 3.13 obtainedfrom bulk glass, the scattering coefficient of approx-imately 2.3 ~dBykm! mm4 is expected in GeO2 fibers.

Rayleigh scattering in a single-component glasssuch as GeO2 glass is dominated by density fluctua-

Fig. 5. Plot of transmission loss versus wavelength24 for the fibershown in Fig. 4.


tion. The scattering loss resulting from density fluc-tuation is related to glass properties and isrepresented by6

ad 58p3

3l4 n8p2kTgbT, (2)

where ad is the scattering loss owing to density fluc-tuation, n is the refractive index, p is the photoelasticconstant, k is Boltzmann’s constant, Tg is the glass-transition temperature, and bT is the isothermalcompressibility at Tg.

The scattering loss has been estimated on the basisof the above relation. Estimated losses are 2.63~Devyatykh et al.18!, 1.13 ~Olshansky and Scherer1!,and 0.79 ~dBykm! mm4 ~Lines16!. When these esti-mations are represented by the relative scattering Rd5 adyad ~SiO2!, values of 3.5, 1.8, and 1.7 are ob-tained for the respective estimations, assuming thatad ~SiO2! is 0.76, 0.63 ~Ref. 1!, and 0.47 ~dBykm! mm4

~Ref. 16!, respectively.Because the Rsc value of approximately 3 is ob-

tained in the present result, there are some differ-ences between the previous estimations based ondensity fluctuation and the present result. The Rdvalue of 3.5 is very close to the present result, but thevalues of approximately 1.7 are almost half thepresent result. This discrepancy seems to indicatethe effect of inelastic-scattering contributions, mainlyRaman scattering. In the present scattering mea-surement for bulk samples a photomultiplier is usedto detect scattered light without any filters so that thescattered light includes not only a Rayleigh compo-nent but also a Raman component.

The Raman-scattering loss in GeO2 glass is esti-mated by Lines to be 0.14 ~dBykm! mm4 ~Ref. 19!.This value is very large compared with the value of0.044 for SiO2 glass.19 When these values are takeninto account, the evaluation for Rd improves slightlyto 1.8. This relative intensity is still too small com-pared with the present result. However, the scat-tering loss due to density fluctuation is also related toexperimental parameters for refractive-index disper-sion given by20

n2 2 1 5EdE0

E02 2 E2 2

El2

E2 , (3)

where Ed is the electronic oscillator strength, E0 isthe average electronic gap, El is the lattice oscillatorstrength, and E is the photon energy. Because n8p2

is proportional to Ed2yE0

4, ~Ref. 20! the scatteringloss is represented by21

ad ,1l4

Ed2

E04 TgbT. (4)

Thus one can evaluate Rd with approximation ~4!,using Ed, E0, Tg, and bT.

The values of Ed ~eV! and E0 ~eV! are obtained bythe measurement of refractive-index dispersion, andthey are 15.82 and 10.52 for GeO2 and 14.73 and


13.34 for SiO2, respectively.21 The values of Tg ~°C!are 544 for GeO2 and 1180 for SiO2, respectively.21

As for bT, although the value for the melt at justabove the glass transition should be used, it seems tobe valid to use the value of the glass at just below theglass transition, according to the fact that the ratio ofthese values is comparable for many materials.22

The value of bT is obtained by the measurement oftemperaure dependence of Young’s modulus, asshown in Fig. 6.23 The Young’s modulus K at Tg isgiven by K 5 K0 1 aTg, where K0 is Young’s modulusat room temperature and a, for simplicity, is the tem-perature coefficient. From this result, we find thatbT is derived by bT 5 3 ~1 2 2n!yK, where n is Pois-son’s ratio. The values of bT 3 10211 Pa21 are 3.86for GeO2 and 2.39 for SiO2, with 0.21 and 0.164 for n,respectively.23

With the above numerical values, Rd of 2.2 is ob-tained. This evaluation gives an improved valuecompared with previous estimations although it isstill small compared with the present experiment.As discussed earlier in this section, the evaluation ofrelative scattering based on density fluctuation offersvalues somewhat smaller than the experiment.This result seems to indicate that the effect of Ramanscattering is unexpectedly large.

Raman scattering also exhibits l24-dependent lossand is strong in GeO2 glass.3 Thus the effect of thiscomponent appears in the slope of loss–l24 plots forspectral loss in optical fibers. The slope includesboth Rayleigh- and Raman-scattering losses.Therefore the relative scattering obtained from bulksamples and from fibers should agree with eachother. Present result shows good agreement.

Finally, it seems valid that the inherent Rsc valuefalls near 3 because the Rsc value measured in bulksamples agrees with that obtained in the fiber lossspectrum. The overall scattering intensity, includ-ing the Raman effect, has practical meaning in termsof fiber application.

As described above, the measured scattering inten-sity is large compared with previous evaluations,while the infrared absorption is close to the reported

Fig. 6. Temperature dependence of Young’s modulus for GeO2

and SiO2 glasses ~Ref. 23!.

value. By using these values, one can give the spec-tral loss in GeO2 fiber:

a 5 2.3yl4 1 0.4 3 1011 exp~258yl! ~dBykm!. (5)

Figure 7 shows the loss curve described by Eq. ~5!.The minimum loss of 0.15 dBykm is expected at 2 mm.The minimum loss estimated in this study is almosttwice the previously reported value owing to the largescattering loss. Based on this estimation, low-lossGeO2 fibers cannot be expected.

In Fig. 7 the measured loss shown in Fig. 4 is alsoincluded. The large loss in this fiber seems to be dueto imperfections between the core and the cladding.The cladding layer is formed by overcoating the sooton core glass. In this process, the core glass surfaceis exposed to H2–O2 flame. Because GeO2 glass isstrongly sensitive to H2O, the H2O attack inducessurface damage such as crystallization in addition toH2O penetration. Such crystallization induces im-perfection scattering loss in the fiber. As shown inFig. 5, the scattering loss is given by l24-dependentand l-independent losses. The surface damagecaused by overcoating seems to induce l-independentloss. The degree of such damage varies in each fiber,resulting in the variation of minimum loss.

In a loss spectrum shown in Fig. 4, absorptionpeaks are seen in a region from 0.9 to 1.3 mm inaddition to GeOH absorption at 1.4 mm. The absorp-tion peaks are due to silicone resin used as protectivecoating.24,25 Because of low refractive-index differ-ence between core and cladding, waveguide structureis not firm enough to avoid the effect of silicone. Ina region below 0.9 mm there is no additional loss; i.e.,scattering loss dominates. Thus the slope value isanalyzed in this region. The loss at 1.4 mm is so highthat the spectrum is not obtained in the region abovethis wavelength.

The minimum loss of the fibers drawn in thepresent experiment is not reduced to the expectedvalue. However, because of the development of thefabrication technique for pure GeO2 core fibers, it willhelp to use them in fiber amplifiers for widebandoperation, including the 1.3–1.5 mm bands, based onRaman scattering.26,27

Fig. 7. Spectral loss expected in GeO2 optical fiber.

As for the scattering characteristics of GeO2 glassshown in Fig. 1, it can be seen that there is a slightdifference in scattering intensity depending on thepreparation method. The intensity of GeO2 glassprepared by the melting method is ;10% smallerthan that of one prepared by the flame hydrolysismethod. Although the cause of this difference is notclarified in the present research, a certain differencein glass structure seems to occur.

In the flame hydrolysis method, the soot ~fine glassparticles! is consolidated into transparent glass.12

Because of this process, the glass seems to have acertain granular structure. In contrast, the glassprepared by the melting method seems not to containsuch a structure because of melting at high temper-ature. In addition, because this glass undergoes aslow-cooling process, it seems that the fictive temper-ature at which the density fluctuation is frozen issufficiently close to Tg to minimize the density fluc-tuation. The value of Tg, which is sensitive to ther-mal history and impurities, for the glass prepared bymelting is 560 6 6 °C based on differential thermalanalysis. This value is slightly high compared with544 °C for the glass prepared by flame hydrolysis.Because the effect of Tg is small, the effect of thegranular structure is highly plausible.

Angular distributions of scattered light intensityexhibit almost constant independence of the scatter-ing angle, as shown in Fig. 1. However, intensitiesare slightly high in very high- and low-angle regions.They seem to be due to inhomogeneity or refraction atthe sample surface. The sample surface is damagedeasily because of strong sensitivity to H2O. If thesample contains inhomogeneity inside its body, theangular distribution should exhibit a characteristicpattern, forward or backward scattering dependingon the nature of inhomogeneity.15 The distributionsshown in Fig. 1 do not exhibit such a characteristicpattern. However, an index-matching liquid is usedin the measurement to minimize the effect of surfacerefraction and reflection.28 The refractive index ofthe liquid is suitable for the measurement of SiO2,but it seems to be low for the measurement of GeO2.However, these effects, both inhomogeneity and re-fraction, are small in the vicinity of right angle.Therefore it seems to be valid that R90 is representedby an average of intensities in the region close to rightangle.

5. Conclusion

Optical properties including Rayleigh scattering andinfrared absorption are determined. The scatteringintensities relative to that of SiO2 for bulk samplesand fibers are in good agreement and exhibit a valueof approximately 3. This implies that a scatteringcoefficient of 2.3 ~dBykm! mm4 can be expected. In-frared absorption is determined to be 0.4 3 1011

exp~258yl! dBykm. On the basis of these results, aminimum loss of 0.15 dBykm at 2 mm is expected.

The authors thank Shuichi Shibata at Tokyo Insti-tute of Technology for useful discussions, Hiroaki


Hiratsuka for continuous encouragement, KazukoHashimoto for glass preparation, and KiyoshiOikawa for fiber loss measurements.

References1. R. Olshansky and G. W. Scherer, “High GeO2 optical

waveguides”, in Tech. Dig. 5th European Conference on Opti-cal Communications, ~The Netherland, 1979!, p. 12.5-1.

2. T. Miyashita and T. Manabe, “Infrared optical fibers,” IEEE J.Quantum Electron. QE-18, 1432–1450 ~1982!.

3. F. L. Galeener, J. C. Mikkelsen Jr., R. H. Geils, and W. J.Mosby, “The relative Raman cross sections of vitreous SiO2,GeO2, B2O3, and P2O5,” Appl. Phys. Lett. 32, 34–36 ~1978!.

4. H. Takahashi and I. Sugimoto, “A germanium-oxide glass op-tical fiber prepared by a VAD method,” IEEE J. LightwaveTechnol. LT-2, 613–616 ~1984!.

5. J. Schroeder, R. Mohr, P. B. Macedo, and C. J. Montrose,“Rayleigh and Brillouin scattering in K2O-SiO2 glasses,”J. Am. Ceram. Soc. 56, 510–514 ~1973!.

6. D. A. Pinnow, T. C. Rich, F. W. Ostermayer Jr., and M. DiDo-menico Jr., “Fundamental optical attenuation limits in theliquid and glassy state with application to fiber opticalwaveguide materials,” Appl. Phys. Lett. 22, 527–529 ~1973!.

7. L. G. Van Uitert and S. H. Wemple, “ZnCl2 glass: a potentialultralow-loss optical fiber material,” Appl. Phys. Lett. 33,57–59 ~1978!.

8. K. Inada, “A new graphical method relating to optical fiberattenuation,” Opt. Commun. 19, 437–439 ~1976!.

9. K. Nakamura, S. Shibuya, I. Sugimoto, H. Takahashi, and Y.Mitsuhashi, “Fiber Raman laser emission in a pure GeO2 glassfiber,” in Technical Digest 12th European Conference on Op-tical Communications, Vol. 1 ~Telefonica, Barcelona, Spain,1986!, p. 11.

10. K. Takahashi, N. Mochida, H. Matsui, S. Takeuchi, and Y.Goshi, “Properties and structure of glasses in the systemsSiO2-PO5y2 and GeO2-PO5y2,” ~in Japanese!, J. Ceram. Soc.Jpn. 84, 482–490 ~1976!.

11. T. Hosaka, S. Sudo, and K. Okamoto, “Dispersion of pure GeO2

glass core and F-doped GeO2 glass cladding single-mode opti-cal fibre,” Electron. Lett. 23, 24–26 ~1987!.

12. S. Sakaguchi, “Consolidation of GeO2 soot body prepared byflame hydrolysis reaction,” J. Non-Cryst. Solids 171, 228–235~1994!.

13. K. Shiraki and M. Ohasi, “Optical properties of sodium almino-silicate glass,” J. Non-Cryst. Solids 149, 243–248 ~1992!.

14. T. Izawa, N. Shibata, and A. Takeda, “Optical attenuation inpure and doped fused silica in the IR wavelength region,” Appl.Phys. Lett. 31, 33–35 ~1977!.


15. H. C. van de Hulst, Light Scattering by Small Particles ~Dover,New York, 1981!, Chap. 6, p. 63.

16. M. E. Lines, “Scattering losses in optic fiber materials. I: anew parameterization,” J. Appl. Phys. 55, 4052–4057 ~1984!.

17. H. Kanamori, H. Yokota, G. Tanaka, M. Watanabe, Y. Ish-iguro, I. Yoshida, T. Kakii, S. Itoh, Y. Asano, and S. Tanaka,“Transmission characteristics and reliability of pure-silica-core single-mode fibers,” IEEE J. Lightwave Technol. LT-4,1144–1149 ~1986!.

18. G. G. Devyatykh, E. M. Dianov, N. S. Karpychev, S. M. Maza-vin, U. M. Mashinskii, V. B. Neustruev, A. V. Nikolaichik,A. M. Prokhorov, A. I. Ritus, N. I. Sokolov, and A. S. Yushin,“Material dispersion and Rayleigh scattering in glassy germa-nium oxide, a substance with promising application in low-lossoptical fiber waveguides,” Sov. J. Quantum. Electron. 10, 900–902 ~1980!.

19. M. E. Lines, “Absolute Raman intensities in glasses. I: the-ory,” J. Non-Cryst. Solids 89, 143–162 ~1987!.

20. S. H. Wemple, “Material dispersion in optical fibers,” Appl.Opt. 18, 31–35 ~1979!.

21. S. Sakaguchi, S. Todoroki, and N. Rigout, “Refractive indexdispersion in ternary germanate glasses,” Jpn. J. Appl. Phys.34, 5615–5620 ~1995!.

22. D. M. Krol, K. B. Lyons, S. A. Brawer, and C. R. Kurkjian,“High-temperature light scattering and the glass transition invitrious silica,” Phys. Rev. B 33, 4196–4202 ~1986!.

23. S. Sakaguchi and S. Todoroki, “Rayleigh scattering in silicaglass,” J. Am. Ceram. Soc. 79, 2821–2824 ~1996!.

24. H. Takahashi and I. Sugimoto, “Silicone-resin-cladgermanium-oxide glass optical fiber,” Jpn. J. Appl. Phys. 22,L313–L314 ~1983!.

25. S. Sakguchi, H. Itoh, F. Hanawa, and T. Kimura, “Drawing-induced 1.53-mm wavelength optical loss in single-mode fibersdrawn at high speed,” Appl. Phys. Lett. 47, 344–346 ~1985!.

26. H. Takahashi, J. Chang, K. Nakamura, I. Sugimoto, T. Tak-abayashi, A. Oyobe, and Y. Fujii, “Efficient single-pass Ramangeneration in a GeO2 optical fiber and its application to mea-surement of chromatic dispersion,” Opt. Lett. 11, 383–385~1986!.

27. D. R. Dykaar, S. G. Grubb, J. Simpson, T. Strasser, A. Veng-sarkar, J. M. Borick, W. Y. Cheng, and S. B. Darak, “2.5 GbysRaman amplifier at 1.3 mm in germanosilicate fibers,” in Op-tical Fiber Communication Conference, Vol. 8 of 1995, OSATechnical Digest Series ~Optical Society of America, Washing-ton, D.C., 1995!, p. PD1-1.

28. H. N. Daglish, “Light scattering in selected optical glasses,”Glass Technol. 11, 30–35 ~1970!.

optical properties of geo_2 glass and optical fibers

Documents