Available online at www.sciencedirect.com
Scripta Materialia 62 (2010) 806–809 www.elsevier.com/locate/scriptamat
Metal oxide doping eﬀects on Raman spectra and third-order nonlinear susceptibilities of thallium–tellurite glasses Tomokatsu Hayakawa,a,b,c,* Motohiro Koduka,a Masayuki Nogami,a Jean Rene´ Ducle`re,b Andrei P. Mirgorodskyb and Philippe Thomasb a b
Department of Frontier Materials, Nagoya Institute of Technology, Gokiso, Showa, Nagoya, Aichi 466-8555, Japan Science des Procedes Ceramiques et de Traitements de Surface (SPCTS), UMR 6638 CNRS, Faculte des Sciences, Universite de Limoges, 123 avenue Albert Thomas, 87060 Limoges Cedex, France c Toyota Physical and Chemical Research Institute, 41-1 Yokomichi, Nagakute, Nagakute, Aichi 480-1192, Japan Received 28 August 2009; revised 25 January 2010; accepted 28 January 2010 Available online 4 February 2010
Third-order nonlinear optical susceptibilities (v(3)) of thallium–tellurite (Tl2O–TeO2) glasses doped separately with various metal oxides (MOY/X; M = 22Ti(IV), 30Zn(II), 31Ga(III), 82Pb(II) and 83Bi(III)) were investigated using a femtosecond Z-scan technique. Surprisingly, the highest change in refractive index was obtained for M = Ti other than Pb and Bi, which was attributed to the formation of the TeO2/TiO2 glass structure without destruction of the initial three-dimensional network, accounting for a strong dielectric response intrinsic to pure TeO2 glass. Ó 2010 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved. Keywords: Amorphous oxide; Optical activity and birefringence; Short-range ordering; Quenching
Currently, tellurium dioxide-based glasses with high refractive indexes and third-order hypersusceptibilities are attracting increasing attention as candidates for nonlinear optical materials applicable to optical wavelength converters, optical switching, Raman ampliﬁcation, etc. [1–4]. They have a number of very promising characteristics, including third-order optical nonlinearity (several orders of magnitude larger than that of conventional silicate and borate glasses [5,6]), as well as low melting point, low glass transition temperature, a wide optical window from visible to infrared regions and low phonon energy for optical activators. Starting with several binary systems (e.g. alkali tellurite , alkalineearth tellurite , PbO–TeO2 , ZnO–TeO2 , Tl2O–TeO2 [11,12]), researches have recently extended to ternary systems that include tellurium dioxide as a main constituent (e.g. TiO2–Nb2O5–TeO2 , ZnO– Nb2O5–TeO2 [14–16], Bi2O3(PbO)–WO3–TeO2 [17,18]). We investigated third-order nonlinear optical susceptibilities (v(3)) using Sheik-Bahae’s Z-scan  for * Corresponding author. Address: Department of Frontier Materials, Nagoya Institute of Technology, Gokiso, Showa, Nagoya, Aichi 4668555, Japan. Tel./fax: +81 52 735 5110; e-mail: [email protected]
various metal oxide (MOY/X)-doped Tl2O–TeO2 glasses (M = 22Ti(IV), 30Zn(II), 31Ga(III), 82Pb(II) and 83Bi(III)), as well as their Raman scattering spectra. Comprehensive data on refractive index n0 and third-order optical susceptibility revealed that a small addition of MXOY can strongly inﬂuence n0 and the real part of v(3) (or Rev(3)). It should be underlined that both n0 and Rev(3) stand out as being markedly larger for M = Ti than for other modiﬁers. According to the results published in Refs. [20,21], it was concluded that the addition of TiO2 to TeO2 can provide reinforcement of the glassy framework, owing to the creation of a Te–O–Ti linkage. The results of the present study allow us to assume that, in parallel with Ti–O–Te bridges, TeO3+1 fragments contribute to the formation of a 3-dimensional network, thus giving rise to a marked increase in n0 and Rev(3). In this paper, we consider the shape of the Raman scattering spectra, when manifesting intermediate TeO3+1 units, as being one qualitative criterium for estimating the variation of Rev(3) values of TeO2-based glasses. The glass samples were prepared using TeO2, Tl2CO3 and MXOY powders. TeO2 ﬁne powder was obtained by a thermal decomposition of orthotelluric acid (H6TeO6, Aldrich, 99.9%) at 560 °C for 24 h. The mixture was melted at 700 or 800 °C for 20 min in an ambient atmo-
1359-6462/$ - see front matter Ó 2010 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.scriptamat.2010.01.055
T. Hayakawa et al. / Scripta Materialia 62 (2010) 806–809 100 80
B 40 E2 E1
Relative Raman Intensity (SiO =1 at peak)
sphere in a platinum crucible and quenched into a brass ring placed on a brass plate warmed at 150 °C to obtain cylindrical samples 10 mm wide and 1–3 mm thick. The quenched glasses were systematically annealed at a temperature 50 °C below the glass transition temperature (Tg) for 24 h in order to release the thermal stresses resulting from the quenching. Each of the glass compositions was determined, along with their thermal and mechanical stabilities. Table 1 lists the glass composition studied, notated as MOY/X–TlO0.5–TeO2, Tg and linear refractive index n0. The wavelength for the refractive index measurement was 632.8 nm from an He–Ne laser in an automatic ellipsometer (FiveLab, MARY-102). In this study, to evaluate v(3) using the Z-scan technique [15,19], a femtosecond pulse laser from a regenerative Ti:Sapphire laser system (Spectra Physics, Harricane) was used at a wavelength of 800 nm with a 1 kHz repetition rate and approximately 170 fs pulse duration. The planar sample with optical ﬂats was moved at 0.5 mm intervals along a 30 mm optical path before and behind the focal position concentrated on by the lens (f = 200 mm). Simultaneously, the optical transmittance was measured with an optical power meter system (Newport, 2930-C/818-SL silicon photodiodes), where the diﬀraction length zR = 11.2 mm and the beam waist radius x0 = 53 lm. The samples were thin enough to satisfy the condition L << n0zR. Close and open Z-scan measurement using the femtosecond laser was applied to estimate third-order nonlinear optical susceptibilities (the real and imaginary parts of v(3), respectively) for the glass samples studied. To test the accuracy of our Z-scan measurement, a ZnSe plate was also examined as a reference for both the real and imaginary parts of v(3) . Raman data were collected using a micro-Raman spectrophotometer (JASCO, NRS-2000). An Ar+ laser with a wavelength of 514.5 nm and 20 mW power was employed as an excitation source. The Raman spectra were ﬁrst calibrated by temperature (Bose–Einstein), refractive index and instrument corrections . The spectra were then compared to each other by both the frequency position and the relative Raman intensity (R.R.I.) criteria. Figure 1(a–e) shows the Raman spectra of MOY/XTlO0.5–TeO2 glasses studied, in which TlO0.5 was partially substituted with MOY/X (5 or 10 mol.%). Each spectrum was decomposed with Gaussian functions (A–E) and baseline-corrected. As was repeatedly stated in a series of papers [20,21], the large band around 450 cm 1 (E = E1 + E2(+ E3)) is intrinsic to TeO2-based glasses having well polymerized network structures (i.e.
20 0 100 80
60 40 E2 E 1
20 0 100 80
60 E3 E2
Raman Shift / cm
Figure 1. Raman spectra of the MOY/X–TlO0.5–TeO2 glasses studied (M = (a) Ti(IV), (b) Zn(II), (c) Ga(III), (d) Pb(II) and (e) Bi(III)), whose intensities were calibrated and compared with the peak intensity of pure silica (SiO2 = 1 at 435 cm 1) . Structures shown are trigonal pyramid (tp) TeO3 (750–760 cm 1, peak A), intermediate TeO3+1 or isolated TeO32 (690–725 cm 1, peak B), and TeO4 disphenoid-like structure (660–650 cm 1, peak C; 600 cm 1, peak D). Peaks E1, E2 and E3 are assigned to the framework-type constitution (involving Te–O–Te, Te–O–M and M–O–M bridges), and a peak around 300 cm 1 to a bending vibration of isolated [TeO32 ] ortho-tellurite groups .
reﬂecting the occurrence of the covalently bonded Te–O–Te, Te–O–M and M–O–M bridges). Therefore, a ratio of band intensities, n = IE/(IA + IB + IC + ID) can be considered as a factor indicating the degree of polymerization of a given glass. For the systems under consideration, this factor was found to be equal to 0.164, 0.125, 0.113, 0.115 and 0.099 for M = Ti, Zn, Ga, Pb and Bi, respectively.
Table 1. Glass composition, Tg, n0, c, Rev(3) and R.R.I. at peak for the MOY/X–TlO0.5–TeO2 glasses studied. 18
c ( 10
5TiO2–30TlO0.5–65TeO2 10ZnO–30TlO0.5–60TeO2 10GaO1.5–36TlO0.5–54TeO2 5PbO–28.5TlO0.5–66.5TeO2 5BiO1.5–38TlO0.5–57TeO2
209 206 175 197 161
2.119 1.945 1.947 1.986 1.970
4.6 ± 0.6 2.6 ± 0.4 2.7 ± 0.4 4.6 ± 0.6 4.5 ± 0.6
Ti(IV) Zn(II) 31 Ga(III) 82 Pb(II) 83 Bi(III) 30
m 2 W 1)
Rev(3) ( 10
5.2 ± 0.7 2.5 ± 0.3 2.6 ± 0.4 4.5 ± 0.6 4.4 ± 0.6
R.R.I at peak (cm 1) (SiO2 = [email protected]
435 cm 1) 97 65 61 76 98
@697.5 @730.0 @732.5 @710.0 @722.0
T. Hayakawa et al. / Scripta Materialia 62 (2010) 806–809
In Figure 2, the Raman peak position is compared with that of a binary Tl2O–TeO2 system with an equivalent TeO2 content. The introduction of MOY/X oxides, with M being diﬀerent from Ti, results in the shift of the Raman band to higher frequencies. In contrast, TiO2 as a third component is distinguishable from these other oxides, as the Raman band moves to lower frequencies. In analyzing this fact, it should be recalled that a binary system of Tl2O–TeO2 with high Tl2O content (>40 mol.% of TlO0.5) has an island-like structure consisting of [TeO3]2 entities with Tl+ ions [12,24]. For this system, isolated [TeO3]2 anions produce the most intense Raman scattering peak at 720–725 cm 1 [12,24], This peak (labeled B) is clearly seen in Figure 1(e), and corresponds to the highest concentration of the complementary oxygen atoms delivered into TeO2 by MOY/X modiﬁers. Actually, as explained in Ref. , if the cation strength of Te is higher than that of atoms M, the oxygen atoms (initially belonging to the modiﬁer) would vigorously attack the coordination spheres around atoms Te, thus augmenting the number of terminal Te–O bonds. As a result, the TeO4 disphenoid-like fragments of the glassy TeO2 framework transform into isolated [TeO3]2 ortho-tellurite groups, thus destroying the initial structure. In turn, the dielectric electronic response becomes signiﬁcantly shorter, and the Rev(3) value decreases . However, as just mentioned above, the structural destruction of the glass framework would not occur if the strength of cation M was close to that of the Te atoms. In such a case, the structure would keep its frame-like constitution, and would consist of Te– O–Te, Te–O–M and M–O–M bridges, which resemble each other quite closely in their chemical aspects. Evidently, in such a structure, the presence of [TeO3]2 anions would not occur, and the Raman spectrum would keep the shape inherent to the TeO2 glassy structure [20,26], i.e. would be dominated by the band centered near 650 cm 1 (labeled C in Fig. 1). Consequently, in the binary (1 x)TeO2–xTl2O glassy system (where 0 < x < 0.5), the bands C and B would coexist, manifesting simultaneous occurrence of structural fragments with the framework-like constitu-
tion (i.e. mainly made from the Te–O–Te bridges), or with the ortho-tellurite constitution (involving the isolated [TeO3]2 anions) . The appearance of a second modiﬁer MOY/X in this system would induce some structural variations which would depend on the strength of the M cation. Actually, the second modiﬁer can drive the tellurite glassy constitution to the framework-type (dominated by band C) or to the island-type (dominated by band B). Correspondingly, it can be expected that the Rev(3) value would increase or decrease, respectively. In our case, only M = Ti4+ would correspond to the former case, whereas the weaker M = Bi3+, Pb2+, Ga3+ and Zn2+ cations would correspond to the latter case. Consequently, when adding TiO2 to the Tl2O–TeO2 glass, the resulting system would tend towards the framework-type constitution (involving Te–O–Te, Te– O–Ti and Ti–O–Ti bridges), thus progressively transforming the [TeO3]2 anions into TeO4 disphenoids via the TeO3+1 intermediate grouping. This tendency is suggested via the displacement of the band B to the lowerfrequency side (Fig. 1a). According to Ref. , this should result in an increase in the Rev(3) value. This is precisely what we observe in Figure 3. Table 1 also lists the R.R.I. of each glass, in comparison with SiO2 . The addition of either Bi2O3 or TiO2 modiﬁer clearly led to the highest Raman intensities. However, broader Raman spectra, which may be more suitable for Raman ampliﬁcation , were obtained in the case of M = Ti, Ga and Zn (see also Fig. 1), although the latter two systems had relatively weak Raman intensities (the Raman band width DRS in the range from 600 to 800 cm 1 is shown in each spectrum of Fig. 1). Figure 3(a) shows a close/open Z-scan data of TiO2– TlO0.5–TeO2 glass. Before and behind the focal point, the transmittance data exhibited a valley and a peak. The transmittance change DT was used to calculate the real part of the third-order nonlinear susceptibility Rev(3). Table 1 also summarizes Rev(3) and the corresponding nonlinear refractive index coeﬃcient c (n = n0 + cI; I = incident power density (W m–2)) . 7 TeO3 6 (terminal)
TeO4 disphenoids 640 100
TeO2 content (mol%)
Figure 2. Raman band peak frequencies of MOY/X–TlO0.5–TeO2 glasses (M = Ti(IV), Zn(II), Ga(III), Pb(II) and Bi(III)), in comparison with those of the TlO0.5–TeO2 binary system (circle, Ref. ) with an equivalent TeO2 content. In the ternary systems, TlO0.5 was partially substituted with MOY/X (5 or 10 mol.%).
Raman Band Frequency (cm-1)
5PbO Normalized transmittance
TlO0.5(+MOY/X) content (mol%)
1 0.9 0.8 0.7 0.6 -12
Sample Position, z / mm
740 680 660 720 700 -1 Raman Band Frequency / cm
Figure 3. (a) Closed/open Z-scan data of the TiO2–TlO0.5–TeO2 glass studied. The dotted line shows a least-squares ﬁtting result with the theoretical equation (A2) from Ref. . (b) Rev(3) as a function of the Raman band frequency for various MOY/X-doped TlO0.5–TeO2 glasses.
T. Hayakawa et al. / Scripta Materialia 62 (2010) 806–809
The largest Rev(3) was obtained for the 5TiO2–30TlO0.5– 65TeO2 glass. This enhancement of the third-order nonlinear susceptibility by the introduction of TiO2 is supposed to be caused by the reinforcement of the TeO2 network structure, mainly composed of TeO3+1, implicating a Te–O–Ti linkage, the latter having a more covalent nature than the Te–O –Tl+ linkage. The relationship of Rev(3) with the Raman peak frequency is shown in Figure 3(b). Importantly, the lower the frequency, the more enhanced the Rev(3) value. In fact, the TiO2–TlO0.5–TeO2 system has the highest n value (n = 0.164), giving evidence for a more polymerized three-dimensional amorphous network. Consistent with recent theoretical predictions [25,29], an increasing polymerization of TeO2-based glasses is found to augment their nonlinear optical susceptibility. Finally, in reference to the literature [30–33], bands B and C are possibly composed not only of [TeO3]2 and TeO4 units but also of TiOn units. However, because of a small amount of TiO2 (5 mol.%) in comparison with the TeO2 component (65 mol.%), the contribution of TiOn units is not visible in the Raman spectrum. This is the case for the TiO2–TeO2 binary system . According to Soulis et al. , elemental Ti is considered to substitute for elemental Te (where a disphenoid-like fragment will be converted to TiO6 octahedron) and form an O–TeIV–O–TiVI–O–TeIV–O linkage due to the covalent bondings of Ti–O–Te (the Roman superscript denotes the coordination number). This study has focused on MOY/X–TlO0.5–TeO2 ternary systems. Although the eﬀect of elemental Tl on TiOn unit is still unclear, we emphasize here again that the introduction of a TiO2 component to binary TlO0.5–TeO2 system mainly aﬀects the Te–O–Te inorganic polymer chains and can reinforce the glass network via Ti–O–Te bonding, which is evidenced by the signiﬁcant shift in the Raman peak position (band B) and the increased Rev(3) values. In summary, we have established a correlation between measured third-order nonlinear optical susceptibility Rev(3) and the structural peculiarities of MOY/X-doped Tl2O–TeO2 glasses (M = Ti(IV), Zn(II), Ga(III) , Pb(II) and Bi(III)), deduced from their Raman scattering spectra. The materials under study were found to consist of TeO3 + 1 units, thus representing intermediate systems between the two types of TeO2–TlO0.5–MOY/X glasses: those of island-type structures (M = Pb(II), Zn(II), Ga(III) and Bi(III)) and those of framework-type structures (M = Ti(IV)). It is concluded that the incorporation of TiO2 into Tl2O–TeO2 glasses is highly favorable for glass polymerization, resulting in an increase in their third-order hypersusceptibilities. This research was ﬁnancially supported by a Grantin-Aid for Scientiﬁc Research (No. 19018011) from the Ministry of Education, Science, and Culture of Japan.  A. Mori, J. Ceram. Soc. Jpn. 116 (2008) 1040.  J. Li, Z. Sun, X. Zhu, H. Zeng, Z. Xu, Z. Wang, J. Lin, W. Huang, R.S. Armstrong, P.A. Lay, Opt. Mater. 25 (2004) 401.  C. Rivero, R. Stegeman, K. Richardson, G. Stegemen, G. Turri, M. Bass, P. Thomas, M. Udovic, T. Cardinal, E. Fargin, J. Appl. Phys. 101 (2007) 023526.
 M.A. Cubinskii, L.D. Merkle, Opt. Lett. 29 (2004) 992.  V. Dimitrov, T. Komatsu, J. Non-Cryst. Solids 249 (1999) 160.  S.H. Kim, T. Yoko, S. Sakka, J. Am. Ceram. Soc. 76 (1993) 2486.  S. Ghosh, A. Ghosh, J. Chem. Phys. 126 (2007) 184509.  H. Desirena, A. Schulzgen, S. Sabet, G. Ramos-Oritz, E. de la Rosa, M. Peyghambarian, Opt. Mater. 31 (2009) 784.  C. Duverger, M. Bouazaoui, S. Turrell, J. Non-Cryst. Solids 220 (1997) 169.  S. Dai, C. Yu, G. Zhou, J. Zhang, G. Wang, L. Hu, J. Lumin. 117 (2006) 39.  B. Jeansannetas, S. Blanchandin, P. Thomas, P. Marchet, J.C. Champarnaud-Mesjard, T. Merle-Me´jean, B. Frit, V. Nazabal, E. Fargin, G.L. Flem, M.O. Martin, B. Bousquet, L. Canioni, S.L. Boiteux, P. Segonds, L. Sarger, J. Solid State Chem. 146 (1999) 329.  O. Noguera, T. Merle-Me´jean, A.P. Mirgorodsky, P. Thomas, J.C. Champarnaud-Mesjard, J. Phys. Chem. Solids 65 (2004) 981.  M.A. Villegas, J.M. Ferna´ndez Navarro, J. Eur. Ceram. Soc. 27 (2007) 2715.  J. Lin, W. Huang, Z. Sun, C.S. Ray, D.E. Day, J. NonCryst. Solid 336 (2004) 189.  T. Hayakawa, M. Hayakawa, M. Nogami, J. Alloys Comp. 451 (2008) 77.  T. Hayakawa, M. Hayakawa, M. Nogami, J. Ceram. Soc. Japan 116 (2008) 1092.  Y. Chen, Q. Nie, T. Xu, S. Dai, X. Wang, X. Shen, J. Non-Cryst. Solids 354 (2008) 3468.  R. Balda, J. Ferna´ndez, J.M. Ferna´ndez-Navarro, Opt. Exp. 17 (2009) 8781.  B.-M. Sheik-Bahae, A.A. Said, T. Wei, D.J. Hagan, E.W. van Styland, IEEE J. Quant. Electron QE-26 (1990) 760.  M. Soulis, A.P. Mirgorodsky, T. Merle-Me´jean, O. Masson, P. Thomas, M. Udovic, J. Non-Cryst. Solids 354 (2008) 143.  M. Udovic, P. Thomas, A.P. Mirgorodsky, O. Durand, M. Soulis, O. Masson, T. Merle-Me´jean, J.C. Champarnaud-Mesjard, J. Solid State Chem. 179 (2006) 3252.  H.P. Li, C.H. Kam, Y.L. Lam, W. Ji, Opt. Mater. 15 (2001) 237.  V.G. Plotnichenko, V.O. Sokolov, V.V. Koltashev, E.M. Dianov, I.A. Grishin, M.F. Churbanov, Opt. Lett. 30 (2005) 1156.  T. Sekiya, N. Mochida, A. Ohtsuka, J. Non-Cryst. Solids 168 (1994) 106.  A.P. Mirgorodsky, M. Soulis, P. Thomas, T. MerleMe´jean, M. Smirnov, Phys. Rev. B 73 (2006) 134206.  O. Noguera, T. Merle-Me´jean, A.P. Mirgorodsky, M. Smirnov, P. Thomas, J.-C. Champarnaud-Mesjard, J. Non-Cryst. Solids 330 (2003) 50.  R. Stegeman, C. Rivero, K. Richardson, G. Stegeman, P. Delfyett Jr., T. Guo, A. Pope, A. Schulte, T. Cardinal, P. Thomas, J.C. Champarnaud-Mesjard, Opt. Exp. 13 (2005) 1144.  R.W. Boyd, Nonlinear Optics, second ed., Academic Press, San Diego, 2003.  O. Noguera, M. Smirnov, A.P. Mirgorodsky, T. Merle-Me´jean, P. Thomas, J.-C. Champarnaud-Mesjard, J. Non-Cryst. Solids 345–346 (2004) 734.  A. Bertoluzza, A. Marinangeli, M.A. Morelli, R. Simoni, J. Non-Cryst. Solids 4 (1988) 149.  C.E. Bamberger, G.M. Begun, O.B. Cavin, J. Solid State Chem. 73 (1981) 317.  T. Cardinal, E. Fargin, G. Le Flem, M. Couzi, J. Solid State Chem. 120 (1995) 151.  H.V. Alberto, N.A. de Campos, B.O. Mysen, Phys. Chem. Glasses 36 (1995) 114.