Inversion in the temperature coefficient of the optical path length close to the glass transition temperature in tellurite glasses S. M. Lima, L. H. C. Andrade, E. A. Falcão, A. Steimacher, N. G. C. Astrath et al. Citation: Appl. Phys. Lett. 94, 251903 (2009); doi: 10.1063/1.3155210 View online: http://dx.doi.org/10.1063/1.3155210 View Table of Contents: http://apl.aip.org/resource/1/APPLAB/v94/i25 Published by the AIP Publishing LLC. Additional information on Appl. Phys. Lett. Journal Homepage: http://apl.aip.org/ Journal Information: http://apl.aip.org/about/about_the_journal Top downloads: http://apl.aip.org/features/most_downloaded Information for Authors: http://apl.aip.org/authors Downloaded 11 Jul 2013 to 186.217.234.138. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://apl.aip.org/about/rights_and_permissions http://apl.aip.org/?ver=pdfcov http://oasc12039.247realmedia.com/RealMedia/ads/click_lx.ads/www.aip.org/pt/adcenter/pdfcover_test/L-37/2074845429/x01/AIP-PT/APL_PDFCoverPg_061913/FreeContentHand_1640x440.jpg/6c527a6a7131454a5049734141754f37?x http://apl.aip.org/search?sortby=newestdate&q=&searchzone=2&searchtype=searchin&faceted=faceted&key=AIP_ALL&possible1=S. M. Lima&possible1zone=author&alias=&displayid=AIP&ver=pdfcov http://apl.aip.org/search?sortby=newestdate&q=&searchzone=2&searchtype=searchin&faceted=faceted&key=AIP_ALL&possible1=L. H. C. Andrade&possible1zone=author&alias=&displayid=AIP&ver=pdfcov http://apl.aip.org/search?sortby=newestdate&q=&searchzone=2&searchtype=searchin&faceted=faceted&key=AIP_ALL&possible1=E. A. Falc�o&possible1zone=author&alias=&displayid=AIP&ver=pdfcov http://apl.aip.org/search?sortby=newestdate&q=&searchzone=2&searchtype=searchin&faceted=faceted&key=AIP_ALL&possible1=A. Steimacher&possible1zone=author&alias=&displayid=AIP&ver=pdfcov http://apl.aip.org/search?sortby=newestdate&q=&searchzone=2&searchtype=searchin&faceted=faceted&key=AIP_ALL&possible1=N. G. C. Astrath&possible1zone=author&alias=&displayid=AIP&ver=pdfcov http://apl.aip.org/?ver=pdfcov http://link.aip.org/link/doi/10.1063/1.3155210?ver=pdfcov http://apl.aip.org/resource/1/APPLAB/v94/i25?ver=pdfcov http://www.aip.org/?ver=pdfcov http://apl.aip.org/?ver=pdfcov http://apl.aip.org/about/about_the_journal?ver=pdfcov http://apl.aip.org/features/most_downloaded?ver=pdfcov http://apl.aip.org/authors?ver=pdfcov Inversion in the temperature coefficient of the optical path length close to the glass transition temperature in tellurite glasses S. M. Lima,1,a� L. H. C. Andrade,1 E. A. Falcão,2 A. Steimacher,2 N. G. C. Astrath,2 A. N. Medina,2 M. L. Baesso,2 R. C. Oliveira,3 J. C. S. Moraes,3 K. Yukimitu,3 and E. B. Araújo3 1Grupo de Espectroscopia Óptica e Fototérmica, Universidade Estadual de Mato Grosso do Sul, C.P. 351, CEP 79804-970 Dourados, Mato Grosso do Sul, Brazil 2Grupo de Estudos dos Efeitos Fototérmicos, Universidade Estadual de Maringá, Av. Colombo 5790, CEP 87020-900 Maringá, Paraná, Brazil 3Grupo de Vidros e Cerâmicas, Universidade Estadual Paulista, C.P. 31, CEP 15385-000 Ilha Solteira, São Paulo, Brazil �Received 16 December 2008; accepted 24 May 2009; published online 22 June 2009� In this study, thermal lens spectrometry was applied to determine the thermo-optical properties of fragile tellurite glasses as a function of temperature, close to the glass transition region. The results showed an inversion from positive to negative values in the temperature coefficient of the optical path length occurring after the glass transition temperature, which is the region where structural changes from the TeO4 trigonal bipyramidal unit to a TeO3 trigonal pyramid containing nonbridging oxygen take place. In addition, the thermal diffusivity values as a function of temperature exhibited behaviors that were related to thermodynamic and kinetic structural changes in the glass. © 2009 American Institute of Physics. �DOI: 10.1063/1.3155210� Tellurite �TeO2-based� glasses are of scientific and tech- nological interest because of their low melting temperatures, good optical transmission in the visible and infrared regions �up to about 7 �m�, high refractive index, high dielectric constant, and large third-order nonlinear susceptibility. These properties suggest that those materials are suitable for appli- cations involving third-harmonic generation or optical Kerr effects.1–5 Tellurite glasses are also strong candidates for superhigh-speed optical switches or shutters, as well as promising materials for fiber-optic applications.6 An interest- ing aspect of these glasses is their thermodynamic and fragile behaviors close to the glass transition temperature,1–3 which are not yet well understood, despite their direct relationship to the structural changes.7,8 The magnitude of the thermal diffusivity �D� and the temperature coefficient of the optical path length �ds /dT� define whether an optical material can be used in optical systems, for instance, in laser windows, second- and third- harmonic generation, and high-power laser-active medium.9 Another important aspect is that in many applications, the nonradiative relaxation processes induce significant tempera- ture variation in the optical devices, which indicates that it is important to know the behavior of the thermo-optical param- eters over a wide temperature range, up to the glass transition temperature �Tg�. For tellurite glasses, which exhibit struc- tural changes even below Tg, the determination of these properties as a function of temperature may contribute to better understanding of the figure of merit of this material in terms of its application in the optoelectronic area.1–3 The two-beam mode-mismatched thermal-lens �TL� method has been used to determine the thermo-optical prop- erties of several optical materials as a function of temperature.10,11 The remote character may make the TL technique a valuable tool for the complete characterization of transparent materials as a function of temperature. Therefore, in this study, the TL method was applied to determine the thermo-optical properties of three different tel- lurite glasses as a function of temperature. The nominal com- positions of the glasses studied were, in mol %: 80TeO2–20Li2O �TeLi�, 80TeO2–15Li2O–5TiO2 �TeLiTi- 5�, and 80TeO2–10Li2O–10TiO2 �TeLiTi-10�. The focus of the study was to investigate the tellurite glass fragility close to Tg by analyzing the thermo-optical parameters along the glass transition region. The influence of TiO2 on the thermo- optical properties close to Tg is also discussed. Measure- ments with thermal relaxation calorimetry �TRC�, optical in- terferometry �OI�, and infrared absorption spectroscopy were also performed. In the TL experiment, the change in intensity of the probe beam is proportional to the TL-induced phase shift, which is given by11,4 � = − Pabs K�p � ds dT , �1� where �p is the probe beam wavelength, Pabs= PALeff is the absorbed power of the excitation beam, P is the excitation power, A is the optical absorption coefficient at the excitation wavelength, Leff= �1−exp�−AL�� /A is the effective sample thickness, L is the sample thickness, K=�CpD is the thermal conductivity, � is the density, Cp is the specific heat, and � is the fraction of the absorbed energy converted into heat. For samples with no fluorescent characteristics, such as tellurite glasses, �=1. In our measurements, an Ar+ laser at 514 nm was used to excite the samples and consequently to create the TL effect, and a HeNe laser at �p=632.8 nm was used to probe this effect. Figure 1�a� shows a typical curve for TeLi glass at 270 °C, which is very similar to that obtained at room tem- perature. The observed increase in the intensity of the probe a�Author to whom correspondence should be addressed. Electronic ad- dresses: smlima@pesquisador.cnpq.br and smlima@uems.br. APPLIED PHYSICS LETTERS 94, 251903 �2009� 0003-6951/2009/94�25�/251903/3/$25.00 © 2009 American Institute of Physics94, 251903-1 Downloaded 11 Jul 2013 to 186.217.234.138. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://apl.aip.org/about/rights_and_permissions http://dx.doi.org/10.1063/1.3155210 http://dx.doi.org/10.1063/1.3155210 http://dx.doi.org/10.1063/1.3155210 beam means that ds /dT�0. By fitting the experimental curve with the TL time-resolved analytical equation given in Ref. 10, both � and the characteristic time response, tc =woe 2 /4D, were obtained. Here, woe is the excitation beam spot size �radius� at the sample position, at L /2. Conse- quently, the D values were calculated. We used woe =48.5 �m. The three tellurite glasses exhibited similar val- ues at room temperature, D= �2.9�0.1��10−3 cm2 /s, indi- cating that no significant changes are observed when the glass structure is modified by replacing TiO2 with Li2O. Comparing with other glasses, tellurite has a D value around 10% higher than that of chalcogenide �2.6�10−3 cm2 /s� �Ref. 12� and approximately half the value of aluminosili- cates ��5.7�10−3 cm2 /s�.13 To determine the thermal con- ductivity, we measured the specific heat using the TRC method.14 The value �0.47�0.02�J /g K at room temperature was not dependent on the glass composition. Assuming that the tellurite glasses studied have the same density ��=4.825 g /cm3�, K= �6.6�0.5��10−3 W /K cm was de- termined. This value is much lower than that of aluminosili- cates ��15�10−3 W /K cm�.13 As defined in Eq. �1�, ds /dT can be determined by normalizing the obtained � parameter by the absorbed excitation power �Pabs� and using the K and �p values. Thus, the value of ds /dT for the three glasses was 12.3�10−6 K−1, which is similar to that of aluminosilicates.13 For the TL measurements as a function of temperature, the glasses were placed in a furnace so that the sample tem- perature could be increased to cross Tg. The TL curves were obtained by scanning the temperature with a ramp rate of 0.5 °C /min. The time interval between each consecutive la- ser shot was about 30 s, which was the appropriate condition to obtain a complete TL relaxation between the events. In addition, it should be stressed that the laser-induced tempera- ture rise in the sample necessary to obtain a detectable TL signal is very low, on the order of 10−2 K. Therefore, the furnace temperature monitored very close to the sample po- sition can be assumed to be the respective sample tempera- ture for each TL datum shown. The TL transients were fitted in the same way as described above, so that D�T� and ds /dT�T� could be determined. The curves �a�, �b�, �c�, and �d� in Fig. 1 show the TL experimental transients obtained for TeLi glass at 270, 280, 290, and 311 °C, respectively, with the same excitation power ��48.6 mW�. The solid lines represent the theoretical fittings. An inversion in the TL curve behavior �from ds /dT�0, curve �a�, to ds /dT 0, curve �d��, occurred around Tg. This inversion was also ob- served in the TeLiTi-5 and TeLiTi-10 glasses, but at different temperatures, so Tg was different for each sample. The ex- planation for this effect is that tellurite glasses combine ds /dT�0 and dn /dT 0 because of their high values of both the thermal expansion coefficient and the refractive in- dex around Tg. Figure 2 shows the D�T� values for the three glasses studied. The lack of data between 280 and 300 °C for TeLi, between 305 and 325 °C for TeLiTi-5, and between 335 and 360 °C for TeLiTi-10 is due to the fact that over these tem- perature intervals, the TL effect goes to zero as a conse- quence of the inversion of ds /dT values, as shown in Fig. 1, curves �b� and �c�. Note that in Fig. 2 there are three distinct regions for the thermal diffusivity behavior: �i� a monotonic trend from room temperature up to the region close to Tg; �ii� a significant increase after Tg; and �iii� a subsequent increase after passing through a minimum in the region of the inver- sion of ds /dT. The thermal diffusivity behavior from room temperature up to Tg is similar to the differential scanning calorimetry �DSC� curves, independently of the tellurite glass studied, so they do not exhibit either endothermic or exothermic trends in this temperature range. A similar obser- vation was reported for fluoride glasses.10,11 The Tg values can be determined by D�T� curves, as indicated in Fig. 2, and were 264, 288, and 318 °C for the TeLi, TeLiTi-5, and TeLiTi-10 glasses, respectively. These values are in agree- ment with those determined by the DSC method, which were 264, 285, and 312 °C, respectively. The observed increase in D�T� values for temperatures above Tg may be related to the structural change from the TeO4 trigonal bipyramid �TBP� unit to the TeO3 trigonal pyramid �TP� containing nonbridging oxygen. The occur- rence of this structural change in tellurite glass was previ- ously confirmed by high-temperature Raman measurements, which showed a decrease in the TBP Raman peak intensity of around 770 cm−1 and a corresponding increase in the TP Raman peak intensity near 670 cm−1.2 The structural alter- ation above Tg increases the glass viscosity,2 producing an increase in both D and heat capacity Cp=Cpe−Cpg values, in which Cpe and Cpg are the heat capacities of supercooled liquids and glasses, respectively.1–3 The Cp was measured with the TRC, and the result obtained was Cpe /Cpg=1.55, which is similar to the value of �1.6 for fragile glasses found in the literature.1–3 This parameter was used by some 0 10 20 30 40 50 0.94 0.96 0.98 1.00 1.02 1.04 (d) (c) (b) (a) I(t )/I o Time (ms) FIG. 1. TL signal for TeLi glass at 270 °C �a�, 280 °C �b�, 290 °C �c�, and 311 °C �d�. The error for each experimental point was lower than 0.5%. The solid lines represent the curve fittings with the TL analytical equation, as described in Ref. 12. We used Pe=48.6 mW. 25 50 250 300 350 400 0 5 10 2 4 6 0 4 8 Tg ~ 289ºC Tg ~ 265ºC Tg ~ 320ºC TeLiTi-10 Temperature (ºC) TeLi D (1 0- 3 cm 2 /s ) TeLiTi-5 FIG. 2. Thermal diffusivity values, D�T�, for TeLi, TeLiTi-5, and TeLiTi-10 glasses. 251903-2 Lima et al. Appl. Phys. Lett. 94, 251903 �2009� Downloaded 11 Jul 2013 to 186.217.234.138. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://apl.aip.org/about/rights_and_permissions authors to define a fragile character of the tellurite glass.1–3 On the contrary, the so-called strong liquids, as pointed out by Angell,7,8 tend to have smaller changes in heat capacity at Tg. For instance, a typical fluoride glass has Cpe /Cpg on the order of 1.1.1–3 For higher temperatures, D�T� values in- creased from a minimum to a maximum, from where we were not able to go further because of the deterioration of the optical quality of the sample. This behavior may also be related to the structural changes with a strong increase in the viscosity of the glass. In order to discuss ds /dT�T�, we wrote it in terms of ds /dQ�T� as: ds /dQ�T�= ��Cp�−1ds /dT�T�. This is under- stood as the sample characteristic response denoting how the optical path changes with the laser-induced heat deposited per unit of volume.10 Figure 3 shows ds /dQ�T� for the TeLi, TeLiTi-5, and TeLiTi-10 glasses. The inset shows ds /dQ�T� from room temperature up to the maximum, corresponding to Tg values. These are compared with those obtained by the D�T� and DSC methods. The increase in ds /dQ�T� is fol- lowed by a decrease, crossing the zero line and assuming negative values. In order to understand this behavior it is important to remember that ds /dT is related to the thermo- optical coefficient �dn /dT� by15 ds dT = �n − 1��1 + ��� + dn dT , �2� where n is the refractive index, � is the Poisson’s ratio, and � is the linear thermal expansion coefficient. Because the first term on the right side of Eq. �3� is always positive, the term responsible for the signal of ds /dT is dn /dT. Remembering that dn /dT ��−��, in which � is the temperature coefficient of the electronic polarizability and �=3� is the volumetric thermal expansion coefficient, and also that � of tellurite glasses did not change when the temperature varied from room temperature to Tg �not shown�, we conclude that the observed ds /dQ �or ds /dT� behavior up to Tg is related to an increase in the electronic polarizability. This is similar to the observations reported by Prod’homme16 for oxide glasses. Tellurite glasses have structural units of TeO4 and TeO3 that are connected weakly with each other, and thus the interme- diate structure varies easily with increasing temperature, which can change the electronic polarizability of the glasses.2 Another interesting observation is that � exhibits a stronger temperature dependence for the glasses containing TiO2, indicating that this metal produces significant changes in the glass structure. Above Tg, the structural change �from TeO4 TBP units to TeO3 TP units containing nonbridging oxygen� is more pro- nounced, resulting in a considerable increase in the volumet- ric thermal expansion coefficient �as observed by the tem- perature behavior of the specific heat�, causing a negative increase in dn /dT. This explains the inversion from positive to negative in ds /dQ values, which was also observed in the OI measurements via visual inspection, in which the interfer- ence fringes inverted their dislocation direction for tempera- tures above Tg. In conclusion, the TL method was successfully applied to measure the thermo-optical properties of fragile tellurite glasses as a function of temperature. The measurements pro- vided Tg values in good agreement with those obtained by the DSC method. The observed inversion in the ds /dQ �or ds /dT� parameter above Tg may be associated with both the strong variation in the volumetric thermal expansion coeffi- cient and the structural change from TeO4 TBP units to a TeO3 TP containing nonbridging oxygen. A significant change in the thermal diffusivity occurred above Tg because of the increase in the glass viscosity. Our results also showed that when the sample is heated above room temperature, the TeLiTi-5 and TeLiTi-10 glasses, with TiO2 in their composi- tion, undergo a higher beam deformation than that of the TeLi glass. We are grateful to the Brazilian National Research Council �CNPq�, FAPESP, PROPP-UEMS, and the Araucária Foundation for financial support of this study. 1S. K. Lee, M. Tatsumisago, and T. Minami, Phys. Chem. Glasses 35, 226 �1994�. 2T. Komatsu, T. Noguchi, and R. Sato, J. Am. Ceram. Soc. 80, 1327 �1997�. 3K. Putz and P. F. Green, J. Non-Cryst. Solids 337, 254 �2004�. 4R. El-Mallawany, J. Appl. Phys. 72, 1774 �1992�. 5C. Y. Zahra and A. M. Zahra, J. Non-Cryst. Solids 190, 251 �1995�. 6M. Yamada, A. Mori, K. Kobayashi, H. Ono, T. Kanamori, K. Oikawa, Y. Nishida, and Y. Ohishi, IEEE Photonics Technol. Lett. 10, 1244 �1998�. 7C. A. Angell, J. Non-Cryst. Solids 73, 1 �1985�. 8C. A. Angell, J. Non-Cryst. Solids 131, 13 �1991�. 9M. L. Baesso, J. Shen, and R. D. Snook, J. Appl. Phys. 75, 3732 �1994�. 10S. M. Lima, T. Catunda, R. Lebullenger, A. C. Hernandes, M. L. Baesso, A. C. Bento, and L. C. M. Miranda, Phys. Rev. B 60, 15173 �1999�. 11J. A. Sampaio, S. M. Lima, T. Catunda, A. N. Medina, A. C. Bento, and M. L. Baesso, J. Non-Cryst. Solids 304, 315 �2002�. 12S. M. Lima, J. A. Sampaio, T. Catunda, A. S. S. de Camargo, L. A. O. Nunes, M. L. Baesso, and D. W. Hewak, J. Non-Cryst. Solids 284, 274 �2001�. 13M. L. Baesso, A. C. Bento, A. A. Andrade, T. Catunda, E. Pecoraro, L. A. O. Nunes, J. A. Sampaio, and S. Gama, Phys. Rev. B 57, 10545 �1998�. 14A. N. Medina, A. M. F. Caldeira, A. C. Bento, M. L. Baesso, J. A. Sam- paio, T. Catunda, and F. G. Gandra, J. Non-Cryst. Solids 304, 299 �2002�. 15S. M. Lima, J. A. Sampaio, T. Catunda, A. C. Bento, L. C. M. Miranda, and M. L. Baesso, J. Non-Cryst. Solids 273, 215 �2000�. 16L. Prod’homme, Phys. Chem. Glasses 1, 119 �1960�. 0 100 200 300 400 -90 -60 -30 0 30 0 100 200 300 5 10 15 20 TeLi TeLiTi-5 TeLiTi-10 ds /d Q (1 0- 6 cm 3 /J ) Temperature ( ºC ) Tg = 289 ºC Tg = 318 ºC Tg = 260 ºC ds /d Q (1 0- 6 cm 3 /J ) Temperature ( ºC ) FIG. 3. Temperature dependence of ds /dQ. The error for each experimental point is around 7%. The inset shows the region between room temperature and Tg. 251903-3 Lima et al. Appl. Phys. Lett. 94, 251903 �2009� Downloaded 11 Jul 2013 to 186.217.234.138. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://apl.aip.org/about/rights_and_permissions http://dx.doi.org/10.1016/j.jnoncrysol.2004.04.009 http://dx.doi.org/10.1063/1.351649 http://dx.doi.org/10.1016/0022-3093(95)00277-4 http://dx.doi.org/10.1109/68.705604 http://dx.doi.org/10.1016/0022-3093(85)90334-5 http://dx.doi.org/10.1016/0022-3093(91)90266-9 http://dx.doi.org/10.1063/1.356045 http://dx.doi.org/10.1103/PhysRevB.60.15173 http://dx.doi.org/10.1016/S0022-3093(02)01040-2 http://dx.doi.org/10.1016/S0022-3093(01)00414-8 http://dx.doi.org/10.1103/PhysRevB.57.10545 http://dx.doi.org/10.1016/S0022-3093(02)01038-4 http://dx.doi.org/10.1016/S0022-3093(00)00169-1