Yb3+ codoped flurotellurite glasses

Yb3+ codoped flurotellurite glasses

Accepted Manuscript Title: Removal of hydroxyl groups from Er3+ /Yb3+ codoped flurotellurite glasses Authors: A. Maaoui, M. Haouari, N. Bel Haj Mohame...

1MB Sizes 6 Downloads 55 Views

Accepted Manuscript Title: Removal of hydroxyl groups from Er3+ /Yb3+ codoped flurotellurite glasses Authors: A. Maaoui, M. Haouari, N. Bel Haj Mohamed, H. Ben Ouda, A. Bulou PII: DOI: Reference:

S0025-5408(17)31401-0 http://dx.doi.org/doi:10.1016/j.materresbull.2017.05.020 MRB 9335

To appear in:

MRB

Received date: Revised date: Accepted date:

12-4-2017 5-5-2017 7-5-2017

Please cite this article as: A.Maaoui, M.Haouari, N.Bel Haj Mohamed, H.Ben Ouda, A.Bulou, Removal of hydroxyl groups from Er3+/Yb3+ codoped flurotellurite glasses, Materials Research Bulletinhttp://dx.doi.org/10.1016/j.materresbull.2017.05.020 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Removal of hydroxyl groups from Er3+/Yb3+ codoped flurotellurite glasses A. Maaoui a* , M. Haouari a,*, N. Bel Haj Mohamed a H. Ben Oudaa, A. Bulou b a) Université de Monastir, Faculté des Sciences de Monastir, Laboratoire des Interfaces et des Matériaux Avancés (LIMA), 5000, Avenue de l’Environnement, Monastir, Tunisia b) Université du Maine, Institut des Molécules et Matériaux du Mans, UMR CNRS 6283, Avenue Olivier Messiaen, 72085, Le Mans Cedex 9, France

*Correspondent authors: Amir maaoui: [email protected] Mohamed Haouari : [email protected]

Graphical abstract:

1

Highlights 

  

In this work, we propose a simple and safe method for an efficient removal of hydroxyl groups from Er3+/Yb3+ doped fluorotellurite glasses having the basic molar composition: TeO2-ZnONb2O5-BaF2:Er2O3/Yb2O3 We have shown that using a convenient amount of BaF 2 as drying agent helped to remove partially hydroxyl groups. For an efficient removal of hydroxyl groups, we have subjected the glasses to an additional heattreatment in two steps. The obtained results revealed an efficient removal of water from the glasses which manifested itself in a drastic reduction of the absorption coefficient and the concentration of hydroxyl groups and a significant increase of Er3+ emission under down or up conversion excitation. This study showed that this procedure might be a veritable alternative for drying tellurite glasses.

Abstract: In this work, we present a safe and relatively simple procedure to eliminate hydroxyl groups from Er3+ and Yb3+ codoped new flurotellurite glasses. It was shown that adding a convenient amount of BaF2 as drying agent reduced the absorption coefficient and the concentration of hydroxyl groups inside the glass samples. However, to remove efficiently water from the glasses, it was necessary to subject the samples to supplementary heat treatment in two steps. A drastic decrease of the density of hydroxyl groups accompanied with a significant enhancement of the erbium emission was observed after the heat-treatment procedure. The obtained results showed that this method might be a veritable alternative to obtain nearly water free tellurite glasses suitable for many optical devices with small sizes or thickness. However, for a good use of this procedure, we tried to elucidate what has happened at the microscopic scale for the glass taking as a guide the available explanations dealing with the removal of water from silica-based glasses. Keywords: Er3+/Yb3+ codoped flurotellurite glasses; hydroxyl groups; water removal; drying agent (BaF2); heat-treatment. Introduction TeO2 based glasses have been the subject of intensive investigations during the three last decades because of their useful optical, thermal and electric properties [1-3]. As compared to others oxides glasses, these materials have a low phonon energy, high thermal stability, high linear and non-linear refractive index, good transparency from UV to MIR which make of them suitable candidates for photorefractive materials, nonlinear optical devices, up-conversion lasers and optical fiber amplifiers [4–6]. Although tellurite glasses possess much interesting qualities, however their use as luminescent materials or for the processing of optical fiber for telecommunications is limited by the level of 2

impurities that they contain. Among these impurities, hydroxyl groups constitute one of the major factors which significantly reduces the quantum yield of luminescence of rare earth ions in the glass matrix and decreases the transparency in the near infrared (NIR) of optical fibers. Indeed, the fundamental vibrations of OH - groups give rise to intense absorption bands in the near IR region between 2500 and 3600 cm-1 (2.7 - 4 m). The overtones of these OH - vibrations are overlapping with the emissions of many rare earth ions, which results in a serious quenching of these emissions. Therefore, several researchers were interested by removing or, at least reducing the amount of these impurities in glasses in order to improve their performance for the applications mentioned above [7]. However, this problem is still inefficiently resolved, and the full elimination of hydroxyl groups remains one of the most urgent goals to be achieved. The suitable key to reach this objective lies in understanding the mechanisms by which water enter in the glass matrix and by which it can be eliminated. Indeed, hydroxyl groups are incorporated inside the glass matrix during the different stages of the preparation. In particular, these radicals are commonly found as impurities in the as-received raw materials with significant amount. Additionally, they may diffuse from the ambient uncontrolled atmosphere into the glass host during the melting stage. Therefore, the natural first step to remove water is to use raw materials with high purity and subject them to profound dehydration before any other operation. To eliminate successfully the residual OH groups from these glasses, many authors have proposed the melting in a controlled atmosphere to reduce moisture content and to create an oxidizing atmosphere avoiding tellurium reduction to its metallic state. For that reason, some authors suggested the dehydration in an ultra-dry atmosphere [8-9]. These authors used a mixed N2 and O2 gas with the same moisture level as the purge in the furnace on the fiber-drawing tower during the fiber drawing. Another group found that working in a dry glove box filled with a mixture of O2 and N2 reduces the amount of hydroxyl groups by 13 order of magnitude [10]. Other more complex methods were also proposed to remove water from the glass. Among the practical procedure used in this field, reactive atmosphere processing (RAP) was suggested as an effective approach for dehydrating fluoride glasses [11-12]. Therefore, some authors groups used Cl2 and O2 whereas others groups used more dangerous gases such as NF3, HF, SF6, CCl4 [13-15]. Although these methods seem to reduce the concentration of hydroxyl impurities in the final glass, however, either they are complicate or they use hazardous chemicals. Additionally they cannot totally extract water from the glass host. Therefore, others supplementary steps are still necessary. In this work, we present a useful and relatively safe method to remove water from new flurotellurite glasses without the need of hazardous gases and chemicals. The first step that we 3

have realized was the use of barium fluoride (BaF2) as drying agent because it is less toxic than others chemicals such as CdF2, CaF2, SrF2, ZnF2 and ZnCl2, and it is less expensive than LaF3. Secondly, we have subjected the glasses to further annealing. The thermal stability of these glasses was investigated by differential scanning calorimetry (DSC). XRD was used to detect the growth of any crystalline phase inside the glass after each annealing step. Raman, FTIR and downshifted or up-converted photoluminescence measurements were carried out to examine the effect of BaF2 and annealing on the internal glass structure and to evaluate the effectiveness of our procedure for removing hydroxyl groups from the final glass. 2. Experimental 2.1. Sample preparation Er3+ and Yb3+ codoped flurotellurite glasses (hereafter referred as TNZF1, TNZF2 and TNZF3), were synthesized by the conventional melt quenching method. The starting materials were from Sigma Aldrich Chemicals with higher purity. The glass starting chemicals were weighted using electronic balance. Subsequently, they were mixed and thoroughly ground in fine powder using a ceramic mortar. The obtained batches were then putted in an alumina crucible and heated in an electric furnace at 900 °C for 30 min at ambient atmosphere. During the melting process, the crucible was slightly shaken to achieve homogeneity of the final glass. After that, the melt was quenched rapidly into cylindrical stainless steel mold preheated at 300 °C in order to relieve the mechanical stress in glass samples. Before any further operation, the samples were annealed in a second furnace for 3 h at 300 °C in order to relieve internal mechanical strain. Then, the furnace was switched-off and allowed to cool to room temperature. The obtained glasses were transparent and bubble-free. For optical measurements, the samples were polished using various grades of abrasive paper to obtain smooth and transparent parallel slides with cylindrical shape of 15 mm diameter and 3 mm thickness. For further removal of hydroxyl groups, the samples were subjected to two steps annealing according to the diagram of Fig. 1. 2.2. Characterization techniques Glass transition (Tg) and crystallization onset (Tx) temperatures were measured at a heating rate of 10 °C/min via DSC measurements using an apparatus from SETARAM (LABSYS EVO DTA/DSC). XRD patterns were obtained with CuKα radiation on a PANalytical X’pert Pro diffractometer equipped with a X'celerator detector. The apparatus used for Raman measurements was a JOBIN-YVON T64000 multichannel spectrometer associated with an Olympus BX41 microscope and coupled with a CCD matrix cooled to 140K to limit the background noise. The excitation wavelength was the 457.1 nm radiation of an argon/krypton laser with a power of 100 mW. The resolution was better than 0.7 4

cm-1. The calibration of the spectrum was carried out using a silicon crystal having a Raman line at 520 cm-1. For FTIR measurements, the glass is reduced in fine powder. Therefore, 1.0 mg of each glass sample was mixed with 100 mg of KBr to prepare pellets with 1 mm thickness. FTIR measurements were achieved in the 400-4000 cm-1 range by using a Nicolet 510 FT-IR spectrometer having a resolution of 0.4 cm-1. Photoluminescence spectra were recorded using the 475.1 nm line argon/krypton laser source. The up converted emission was obtained using a diode laser emitting at 980 nm as excitation source. 3. Results and discussion 3.1. Physical and thermal Properties As it is clear from Table 1, the density of the glass increases with increasing the amount of BaF2. In general, the replacement of oxygen by fluorine leads to a decrease of the density and an increase in the molar volume and the structure of the glass becomes loose [16-17]. However, in the case of the present study, the increase of the molar volume may be screened by the molar weight of BaF2 (175.3) which is greater than that of TeO2 (159.6). Fig. 2 shows the diffractograms of the TNZF2 glass sample subjected to different heat treatment procedures. The appearance of a broad, halo band, and the absence of any sharp peak in these XRD patterns confirm the vitreous nature of the material. Thermal stability is required for fiber drawing where the nucleation and the subsequent crystal growth must be avoided. Therefore, it is of great interest to investigate the thermal stability of these glasses. Glass forming ability and thermal stability of glasses can be evaluated from the characteristic temperatures Tg, Tx and Tf, which correspond to glass transition, crystallization onset and glass melting respectively. Several criteria were proposed in literature to evaluate the glass stability [18-24]. The most useful parameter that combines both nucleation and growth aspects of phase transformation is that suggested by Hruby and expressed as [18]: 𝑯𝒓 =

𝑻𝒙 −𝑻𝒈 𝑻𝒇 −𝑻𝒙

(1)

According to this parameter, the larger difference (Tx –Tg) and the smaller temperature interval (Tf –Tx) hamper the processes of crystallization and consequently facilitate glass formation. However, because of the difficulty to evaluate the melting temperature for all the samples, we have used the criterion ∆T = Tx -Tg introduced by Dietzel [25], which is often used to evaluate the glass forming ability and the possibility of its thermal processing without inducing crystallization. A glass with ΔT > 100°C is considered stable for fiber drawing [26]. Therefore, we have carried out DSC measurements on the as quenched glass and the obtained traces are presented in Fig 3. The curves of sample TZNF1 and TZNF2 present only one 5

crystallization peak whereas two exothermic peaks are seen in the DSC trace of sample TZNF3. The appearance of these peaks reveals the presence of two different crystalline phases due to the great proportion of fluorine. However, only the lowest temperature peak is considered when discussing glass stability. The values of Tg and Tx estimated from the DSC curves are given in Table 2. Since the values of ΔT are greater than 100°C, the obtained glasses may be convenient for fiber drawing. 3.2. Raman spectral analysis Raman spectroscopy is a useful spectroscopic tool for exploring the structure and the presence of functional groups in the glass host. In addition, it is usually used to determine the maximum phonon energy of the materials. Therefore, we have recorded the Raman spectra of the present glass system in the spectral range 100–1100 cm−1 (Fig. 4). All the samples were operated at the same condition for comparing Raman intensity and frequency shift. As we can see from Fig. 4, the Raman spectra of the entire flurotellurite glasses exhibit four Raman bands named A, B, C and D. The three peaks (A–C) are the characteristic Raman modes observed in most of the tellurite glasses [27-29]. Indeed, the peak A around 446 cm-1, is assigned to the bending vibrations of Te–O–Te or O–Te–O linkages. The peak B at about 678 cm-1 is attributed to the stretching vibration of TeO4 trigonal bipyramid (tbp) units, whereas the peaks C near 767 cm-1 could be attributed to the stretching vibrations of TeO3 trigonal pyramidal (tp) units. The band D at nearly 860 cm-1 is assigned to the vibrational mode of Nb–O in the NbO6 octahedra or Nb– O–Te linkage [30]. The presence of these bands indicates that the glassy network is composed of different structural units such as NbO6, TeO4 and TeO3. As we can see in Fig. 4, the intensity ratio (ITeO3/ITeO4) increases with adding BaF2 suggesting the transformation of trigonal bipyramidal TeO4 through TeO3+1 to trigonal pyramidal TeO3. This is due to partial replacement of some O2- ions by F- because of their comparable ionic radius. Therefore, the network structures will be composed of the mixed Te(O, F) 3 and Te(O, F)4 units [31]. More precisely, the fluoride ions replace bridging-oxygen (BO) ions somewhere in the TeO4 (tbp), which becomes Te(O, F)4 with bridging fluoride (BF) ions. This will stabilize the chain structure where two TeO4 (tbps) are connected by the BF. Finally, it is worth to note that the cut-off frequency of vibrational modes related to the maximum phonon energy of the glass network remains almost constant and equal to 767 cm-1 (band C), whereas the shoulder in the high-energy side of this band decreases from 890 to 856 cm-1 when adding BaF2. This displacement is ascribed to the partial substitution of oxygen by fluorine, which results in a relative reduction of the intensities of the peaks composing this band. Indeed, Ba2+ and F− may decrease the field strength and loosen the structure of the basic 6

TeO2 unit, which results in a decrease of the IR cut-off frequency and a redshift of the MIR absorption edge [32]. This leads to a reduction of the multiphonon relaxation of rare earth ions in these glasses with respect to other oxide glasses, which makes them useful for obtaining high efficiency lasers and fiber amplifiers since this reduces the non-radiative transitions and enhance the luminescence efficiency of the dopant ions. 3.3. Removal of hydroxyl groups 3.3.1. Melting conditions In glasses melted at ambient atmosphere, the concentration of hydroxyl groups is not due only to the incorporation of water from ambient atmosphere into the melt but it is determined by the water content of the raw materials [33]. Therefore, an increase of the melting time and temperature may improve the elimination of hydroxyl groups. However, for higher melting temperature and long melting time, glass constituents can evaporate together with water leading to an important deviation from the targeted final composition of the glass. Therefore, one must choose the convenient melting conditions that enhance glass drying without significant composition modification. In the case of the present study, we have chosen melting the starting chemicals at 900 °C for 30 min in order to allow the homogenization of the melt and the evacuation of the great fraction of volatile gases such as water vapor or hydrogen fluoride without losing the basic components. However, during the glass melting, only the surface of the glass melt can be dehydrated although drying may be improved if we work in dried-air atmospheres. Moreover, pre-drying before melting may be helpful since it eliminates a great quantity of water from the raw materials. 3.3.2. Effect of BaF2 To analyze the effect of adding BaF2, FTIR measurements were conducted on the different samples. As it is clear from Fig. 5, the absorption band characteristic for hydroxyl groups at nearly 2.9 μm (3450 cm-1) is reduced by the addition of BaF2 before increasing again. Since the quenching rate due to water content is found to be proportional to the absorption coefficient of OH radicals [11], it is convenient to get an estimation of this coefficient and the concentration of these groups in the final product. This was done by means of spectroscopic data and the BeerLambert law. Under these conditions, the absorption coefficient of hydroxyl groups at the maximum of the absorption band is given by: 𝜶𝑶𝑯 = −

𝐥𝐧(

𝑰 ) 𝑰𝟎

𝒍

(2)

where l is the thickness (in cm), I0 and I are the incident and the transmitted light intensities respectively. The concentration of OH groups is thus estimated using the equation [34]: 7

𝑵𝑶𝑯 =

𝑵𝑨 𝜺

𝜶𝑶𝑯

(3)

where NA is the Avogadro number and ε is the molar absorptivity of the free OH groups in the glass and here a value of 4.91x104 cm2/mol was used [35]. Table 3 lists the calculated absorption coefficient, the OH concentration for the different glass samples. These results clearly show the benefic effect of adding BaF2 for hydroxyl suppression. Indeed, OH- ions can be easily replaced by fluorine during the melting according to the equation: 𝐎𝐇− + 𝐅 − → 𝐇𝐅 + 𝐎𝟐−

(4)

When adding BaF2, the removal of hydroxyl groups happens according to the chemical reaction: 𝐁𝐚𝐅𝟐 + 𝐇𝟐 𝐎 ⟶ 𝐁𝐚𝐎 + 𝟐𝐇𝐅

(5)

The generated hydrogen fluoride gas (HF) is removed naturally during the melting process. As seen in Fig. 6, the removal of water manifests itself in the increase of the down-shifted or up-converted emission of Er3+ ions. Unfortunately, however, when the amount of the fluorine agent exceeds certain threshold (here), the absorption coefficient and the OH concentration increases again. This may be due to the great amount of fluorine in the final composition, which makes the glass more hygroscopic. In addition, introducing large amount of fluorides into tellurite glasses results in a significant decrease of both the refractive index and the nonlinear refractive index and a weakening of the glass network and the reduction of its chemical durability. Moreover, when the concentration of fluorides becomes so high, the glass will have a strong tendency to be crystallized easily during the annealing process, which limits its use for fiber drawing. As mentioned above, this is clearly seen in the XRD diffractogram of the sample TZNF3, which contains the great amount of BaF2 (Fig 2). Therefore, it is convenient to choose the appropriate amount of fluoride to get the desirable properties. 3.3.3. Effect of annealing To further remove water, we have proceeded with annealing the glass samples for long times in two steps (Fig. 1). To evaluate the OH-concentration in the glasses before and after each heat-treatment stage, the IR transmission spectra of the as-melted and dried glasses were recorded. As is indicated in Fig. 7 and Table 3, there is a significant decrease of the absorption band relative to hydroxyl groups during the first stage and the dehydration of the glasses was improved in the second stage. The effect of the removal of hydroxyl groups was also confirmed by the significant improvement of the emission of the samples, which became brighter after each heat-treatment step (Fig. 8). By comparing, our results to those obtained by others groups (Table 4), we may suggest that this relatively simple and safe drying procedure is very encouraging to prepare glasses and to process optical fibers nearly water free. However, to 8

make good use of this method, it is necessary to understand the mechanisms that are behind this dehydration. Although the lack of a convenient model that may explain appropriately the obtained results, perhaps due to infrequent similar studies and to the sophistical techniques usually used to probe molecules diffusion in solids, we may propose the following ideas to try to understand what has happened to the glass samples at the microscopic scale during heat treatment. Firstly, as we have used BaF2 as a drying agent in the starting composition, probably a minor fraction was remaining after the melting step since reaction required more time to be completed. During the heat treatment, the reaction given in equation (5) might take place leading to the regeneration of hydrogen fluoride, which is assumed to have higher mobility than bound fluorine [40], and can diffuse through network before being evacuated outside the glass. The generation of HF inside the glass might be due to the network bound fluorine and water molecule. Similar process was also proposed to account for the generation of hydrogen fluoride in silica-based glasses [41-42]. This process is schematized by the following equation: ≡ 𝐒𝐢 − 𝐂𝐥 + 𝐇𝟐 𝐎 → ≡ 𝐓𝐞 − 𝐎𝐇 + 𝐇𝐂𝐥

(6)

Secondly, as suggested for silica-based glasses [43], water molecules could be regenerated when glass is annealed at relatively high temperature via the following process: 𝐓𝐞 − 𝐎𝐇 + 𝐇𝐎 − 𝐓𝐞 ⇋ 𝐓𝐞 − 𝐎 − 𝐓𝐞 + 𝐇𝟐 𝐎

(7)

The generated water molecules might diffuse through the sample before reaching its surface [44]. This leaded to the dehydration of the glass and a drastic decrease of the absorption band relative to hydroxyl groups in the IR spectra. Another process that could occur at relatively elevated temperature is the reaction between hydride (Te − H) and hydroxyl (Te − OH) impurities which might generate hydrogen molecules. In a similar way, it was suggested to explain the removal of hydroxyl groups from silicate glasses, that chemical reaction between Si − H and Si − OH produces hydrogen molecules, which may diffuse through interstices of the glass network [45]. Molecules diffusion is a common phenomenon usually observed inside silica glasses and is thoroughly investigated [46-52]. However, to the best of our knowledge, there are no similar studies interested with the diffusion of such molecules inside a TeO2 based glasses. Therefore, we will take as a guide silica based glasses although the explanation we present may stimulate discussions about this subject. Diffusion of molecules from one region to another in presence of chemical reaction is driven by the following equation [51]: 𝝏𝑪 𝝏𝒕

𝝏𝟐 𝑪

𝝏𝑺

= 𝑫 𝝏𝒙𝟐 − 𝝏𝒕 9

(8)

In this equation, C is the concentration of diffusing molecules, S is the concentration of reacting immobile species and D is the diffusion coefficient of molecules. On the other hand, the diffusion coefficient can be expressed by the following Arrhenius equation [53]: −𝑸

𝑫(𝑻) = 𝑫𝟎 𝐞𝐱𝐩 (𝒌 𝑻) 𝑩

(9)

In the above equation, Q is the activation energy of diffusion, kB, the Boltzmann constant, T the absolute temperature and D0 a factor also known as frequency factor. This equation shows that diffusion coefficient increases with temperature. However, diffusion of water molecules across the glass may be limited by the reverse reaction with tellurium/oxygen network according to equation (7). Therefore, the reaction of H2O molecules with bridging oxygen is an active step during movement of water. In fact, it was suggested that water diffuse inside glass network as H2O molecules or OH groups [54]. However, because the rate of diffusion is influenced by the removal or supply of reacting species, water releasing from the glass surface will displace the equilibrium in equation 7 toward the elimination of hydroxyl groups and the construction of Te − O − Te bridges. Indeed, it was observed in silica glasses that reaction of water with the glass network became faster than diffusion only at temperatures exceeding 600°C [55]. Moreover, the interaction of hydrogen fluoride with the silica glass network is known to be weak [56], which facilitate its diffusion and extraction. In the diffusion process, molecules can jump in solids from one site to another thanks to the thermal vibrations of the atoms in the lattice. This means that annealing at a convenient temperature may activate the diffusion and the removal of water molecules from the glass matrix. Besides, diffusion coefficient is also related to viscosity constant  and the diffusive jump distance  according to the Eyring equation [57]: 𝑫(𝑻) =

𝒌𝑩 𝑻 𝜼𝝀

(10)

From this equation, we can see that the diffusion of molecules becomes more important when decreasing the viscosity and the distance . This may happens when annealing the glass at a convenient temperature, which lowers the threshold for diffusion. Additionally, diffusion is enhanced when the diffusing molecules find open pathways to reach the surface. This is found in the large free volume usually present in glasses because they have more open network structures as compared to crystalline materials [58]. In the case of tellurite glasses, the addition of Nb, Ba and Zn ions to TeO2 generates large amount of non-bridging oxygen atoms and disturb the glass network, which becomes more disordered inserting very close lacunas due to the increase of inter-atomic spacing [59-60]. Therefore, water diffusivity is controlled by the concentrations of diffusing species and of diffusion pathways, which are determined by the 10

concentrations of hydroxyl groups and network modifier cations. Indeed, the change of the oxidation state of Te from TeO4 to TeO3 as revealed from Raman study, produces voids incrementing thus the interstitial space within the glass network. Therefore, open pathways across the glass structure will exist facilitating thus the diffusion of molecules. Additionally, heating at relatively high temperature will expand chemical bonds, which results in more free space required for the enhancement of molecules diffusion. This is why the obtained results at the second step of the drying procedure seems to be more encouraging and the reduction of either the absorption coefficient and the concentration of hydroxyl groups is clearer. Since diffusion is a time dependent operation, some eventual mechanisms slowing the overall process such as reactions between diffusing molecules and the lattice, have to be taken into account. For this reason, a long annealing time is requested to achieve a reasonable drying of the glass. At the same time, because the samples were relatively small, the diffusing molecules would reach their surface more easily. Indeed, it was suggested that the removal rate of hydrogen and water from the vitreous silica is a function of the diffusion coefficient, the physical solubility and the geometry of the sample [61]. Therefore, this method may be appropriate to remove water from small samples for low OH photonic devices or optical fibers. In general, a convenient shaping of the glass may improve the removal of hydroxyl groups after annealing. Based on the obtained results, we may recommend this procedure to remove water from optical fiber drawn from flurotellurite glasses. In particular, sample TZNF1 may be a good candidate for applications in fiber drawing or luminescent device since it has good thermal stability against crystallization, weak absorption coefficient, and lower concentration of hydroxyl groups after the proposed drying procedure. 4. Conclusion We have studied the effect of BaF2 and annealing on the water content in flurotellurite glasses. The obtained results show clearly that using BaF2 with a convenient amount together with annealing the glass for a relatively long time at a moderate temperature can constitute a veritable alternative to obtain nearly water free glass with enhanced optical and thermal properties suitable for high-quality optical fiber processing. However, to understand the mechanism of the removal of hydroxyl groups, a supplementary theoretical and experimental work should be done since no more similar studies are available in literature although different methods were proposed to get water free tellurite glasses. In particular, quadrupole mass spectrometry or oxygen tracer experiments and theoretical simulation using molecular dynamics procedure may be very helpful to better clarify this subject. 11

References: [1] R. El‐Mallawany, The optical properties of tellurite glasses, J.Appl. Phys., 72, (1992) 17741777 [2] D. Manzani, J. L.Ferrari, F. C. Polachini,Y. Messaddeq, S. J. L. Ribeiro, 1.5 µm and visible up-conversion emissions in Er3+/Yb3+ co-doped tellurite glasses and optical fibers for photonic applications, J. Mater. Chem., 22 (2012) 16540. [3] D. Souri, Seebeck coefficient of tellurite–vanadate glasses containing molybdenum, J. Phys. D: Appl. Phys., 41 (2008) 105102. [4] G. Murugan, T. Suzuki, Y. Ohishi, Raman characteristics and nonlinear optical properties of tellurite and phosphotellurite glasses containing heavy metal oxides with ultrabroad Raman bands, J. App. Phys., 100 (2006) 0231071-76 [5] H. Lin, S. Jiang, J. Wu, F. Song, N.r Peyghambarian, E. Pun, Er 3+ doped Na2O–Nb2O5– TeO2 glasses for optical waveguide laser and amplifier, J. Phys. D: Appl. Phys. 36 (2003) 812– 817 [6] H. Lin, G.Meredith, Shibin Jiang, Xiang Peng, Tao Luo, N Peyghambarian, E. Pun, Optical transitions and visible upconversion in Er3+ doped niobic tellurite glass, J. Appl. Phys., 93 (2003) 186-191. [7] P.F. Wang, W.N. Li, B. Peng, M. Lu, Effect of dehydration techniques on the fluorescence spectral features and OH absorption of heavy metals containing fluoride tellurite glasses, J. Non-Cryst. Solids., 358 (2012) 788-793. [8] I. Savelii, F. Desevedavy , J.C. Jules, G. Gadret, J. Fatome, B. Kibler, H. Kawashima, Y. Ohishi, F. Smektala, Management of OH absorption in tellurite optical fibers and related supercontinuum generation, Opt. Mater., 35, (2013) 1595–1599. [9] X. Feng, J. Shi, M. Segura, N. M. White, P. Kannan, W. H. Loh, L. Calvez, X. Zhang, and L. Brilland, Halo-tellurite glass fiber with low OH content for 2-5µm mid-infrared nonlinear applications , Opt. Express., 21 (2013) 18949-18954. [10] H. Tilanka M. A. W. Beckmann, C. Schiele, D. Manzani, L. Wondraczek, S. Afshar V., T. M. Monro, H. Ebendorff-Heidepriem, Lead-germanate glasses and fibers: a practical alternative to tellurite for nonlinear fiber applications, Opt. Mater. Express., 3 (2013) 1488-1503. [11] M. Boivin, M. El-Amraoui, S. Poliquin, R. Vallee, Y. Messaddeq, Advances in methods of purification and dispersion measurement applicable to tellurite-based glasses, Opt. Mater. Express, 6 (2016) 1079-1086. [12] A. E. Comyns (Ed.) Fluoride glasses, (John Wiley & Sons) (1989).

12

[13] J. H. Campbell et al., Continuous melting of phosphate laser glasses J. Non-Cryst. Solids, 263 (2000) 342-57. [14] J. Massera, W. Haldeman, Z. Jackson, Z. Rivero-Baleine, Y. Petit, K. Richardson, Processing of Tellurite-Based Glass with Low OH Content, J. Am. Ceram. Soc., 94 (2011) 130–136. [15] P. Josh, B. Richards, A. Jha, Reduction of OH ions in tellurite glasses using chlorine and oxygen gases, J. Mater. Res., 28 (2013) 3228-3233. [16] G. Wang, S. Dai, J. Zhang, S. Xu, L. Hu, Z. Jiang, Effect of F- ions on physical and spectroscopic properties of Yb3+-doped TeO2-based glasses, J. Lumin. 113 (2005) 27–32. [17] S. Ahmmad, E. kondaul, S. Rahman, Effect of F ions on physical and optical properties of fluorine substituted zinc arsenic tellurite glasses, IOP Conf. Series: Mater. Sci. Eng. 73,, (2015) 012080. [18] Hrubý A. Evaluation of glass-forming tendency by means of DTA. Czech J. Phys B., 22, (1972) 187-93. [19] D.R. Uhlmann, Kinetics of glass formation and devitrification behavior. J. Phys. Colloques, 43 (1982) C9-175 - C9-190. [20] D.R. Uhlmann, H. Yinnon, Glass: Science and Technology, Academic Press Inc., San Diego, CA, (1983). [21] M. Saad, M. Poulain, Glass Forming Ability Criterion, Mater. Sci. Forum, 19-20 (1987) 11-18. [22] MC. Weinberg, An assessment of glass stability criteria. Phys. Chem. Glass, 35 (1994) 119-23. [23] Z. Lu, C Liu. Glass formation criteria for various glass-forming systems. Phys. Rev. Lett., 91 (2003) 5504-05. [24] L.F. M. Nascimento, L. A. Souza, E. B. Ferreira, E. D. Zanotto, Can glass stability parameters infer glass forming ability, J. Non-Cryst. Solids, 351 (2005) 3296–08. [25] A. Dietzel. Glass structure and glass properties. Glasstech, 22 (1968) 41. [26] R. El-Mallawany, Tellurite Glasses Handbook Physical Properites and Data (CRC Press LLC, (2002) [27] V.O. Sokolov, V.G. Plotnichenko, V.V. Koltashev, E.M. Dianov, J. Non-Cryst. Solids, 352 (2006) 5618-5632. [28] S. Murugan, Y. Ohishi, TeO2–BaO–SrO–Nb2O5 glasses: a new glass system for waveguide devices applications, J. Non-Cryst. Solids, 34 (2004) 86-92.

13

[29] T. Sekiya, N. Mochida, A. Ohtsuka, M. Tonokawa, Raman spectra of MO 1/z-TeO2 (M = Li, Na, K, Rb, Cs and T1) glasses , J. Non-Cryst. Solids, 144 (1992) 128. [30] T. Sekiya, N. Mochida, A. Soejima, Raman spectra of binary tellurite glasses containing tri- or tetra-valent cations, J. Non-Cryst. Solids, 191 (1995) 115. [31] S. Ahmmad, M.A. Samee, A. Edukondalu, S. Rahman, Physical and optical properties of zinc arsenic tellurite glasses, Results in Physics, 2 (2012) 175-181. [32] R. Wang, X. Meng, F. Yin, Y. Feng, G. Qin, W. Qin, Heavily erbium-doped low-hydroxyl flurotellurite glasses for 2.7 μm laser applications, Opt. Mater. Express, 3 (2013) 1127-1136 [33] H. Ebendorff-Heidepriem, K. Kuan, M. R. Oermann, K. Knight, and T. M. Monro, Extruded tellurite glass and fibers with low OH content for mid-infrared applications, Opt. Mater. Express, 2 (2012) 432-442. [34] X. Feng, S. Tanabe, T. Hanada, hydroxyl groups in erbium doped germanotellurite glasses, J. Non Cryst. Solids, 281 (2001) 48-54. [35] S. Dai, J. Zhang, C. Yu, G. Zhou, G. Wang, L. Hu, Effect of hydroxyl groups on nonradiative decay of Er3+: 4I13/2 → 4I15/2 transition in zinc tellurite glasses, Mater. Lett., 59, (2005) 2333- 2336. [36] L. Gomes, J. Lousteau, D. Milanese, G. C. Scarpignato, S. D. Jackson, Energy transfer and energy level decay processes in Tm3+-doped tellurite glass, J. App. Phys, 111 (2012) 063105. [37] Y. Zhou, N. Gai, F. Chen, g. yang, effect of hydroxyl groups in erbium-doped telluriteand bismuth-based glasses, Opt fiber technol, 16 (2010) 318–322 [38] V.A.G. Rivera, M. El-Amraoui, Y. Ledemi, Y. Messaddeq, E. Marega Jr. Expanding broadband emission in the near-IR via energy transfer between Er3+–Tm3+ co-doped telluriteglasses, J. Lumin, 145 (2014) 787–792. [39] F. Chen, T. Wei , X. Jing, Y. Tian, J. Zhang, S. Xu, Investigation of mid-infrared emission characteristics and energy transfer dynamics in Er 3+ doped oxyfluoride tellurite glass, Sci. Rep., 5, (2015) 10676. [40] L. Némec, J. GOTZ, Infrared Absorption of OH- in E Glass, J. Am. Ceram. Soc. 53, (1970) 526. [41] K. Kajihara, M. Hirano, L. Skuja, H. Hosono, Reactivity of SiCl and SiF groups in SiO 2 glass with mobile interstitial O2 and H2O molecules, J. Non Cryst. Solids, 353 (2007) 514–517. [42] M. Fokine, Formation of thermally stable chemical composition gratings in optical fibers, J. Opt. Soc. Am. B, 19 (2002) 1759-1765. [43] R. H. Doremus, Diffusion of water in silica glass, J. Mater. Res., 10, (1995) 2379-2389.

14

[44] X. Feng, J. Shi, M. Segura, N. M. White, P. Kannan, W. H. Loh, L. Calvez, X. Zhang, L. Brilland , Halo-tellurite glass fiber with low OH content for 2-5µm mid-infrared nonlinear applications , Opt. Express., 21(2013) 18949-18954. [45] Z. Yongheng, G. Zhenan, The study of removing hydroxyl from silica glass, J. Non Cryst. Solids, 352 (2006) 4030-4033. [46] H. Behrens, M. Nowak, The mechanisms of water diffusion in polymerized silicate melts, Contrib Mineral Petrol, 126 (1997) 377-385. [47] R. H. Doremus, diffusion-controlled reaction of water with glass, J. Non Cryst. Solids, 55, (1983) 143-147. [48] H.A. Schaeffer, Transport phenomena and diffusion anomalies in glass, Ceramic Materials, 64 (2012) 156-161. [49] H. Li, M. Tomozawa, Effects of water in simulated borosilicate-based nuclear waste glasses on their properties, J. Non-Cryst. Solids 195 (1996) 188-198. [50] T. H. Elmer, Engineered Materials Handbook, 4, Ceramic and Glasses, Copyright 1932, ASM International, Materials Park, OH 44073-0002, p 427-32 [51] R. H. Doremus, Diffusion of water in silica glass, J. Mater. Res., 10 (1995) 2379-2389. [52] G. Iovino, S. Agnello, F. M. Gelardi, Dependence of O2 diffusion dynamics on pressure and temperature in silica nanoparticles, J. Nanopart. Res., 15 (2013) 1876 [53] S. Kostinski, R. Pandey, S. Gowtham, U. Pernisz, A. Kostinski, Diffusion of Water Molecules in Amorphous Silica, IEEE Electron Device Letters, 33 (2012) 863-865. [54] Y. Zhang, E. M. Stolper, and G. J. Wasserburg, Diffusion of water in rhyolitic glasses Geochimica et Cosmhimica Am, 55 (1991) 441-456. [55] R. H. Doremus. Diffusion of reactive molecules in solids and melts. John Wiley and Sons, New York, 2002. [56] I. G. Sanchez, Fabrication and Applications of low OH Photonic Crystal Fibres, thesis, University of Bath, 2012. [57] H. Eyring, J. Chem. Phys., 3 (1935) 107. [58] H. Kohara, H. Ohno, M. Takata, T. Usuki, H. Morita, K. Suzuya, J. Akola, L. Pusztai, Lead silicate glasses: Binary network-former glasses with large amounts of free volume, Phys. Rev. B, 82 (2010) 134209-134215. [59] R. El-Mallawany I. Abbas Ahmed, Thermal properties of multicomponent tellurite glass, J. Mater Sci, 43, (2008), 5131-5138. [60] A. Kaura, A. Khanna, V. G. Sathe, F. Gonzalez, B. Orti, Optical, thermal, and structural properties of Nb2O5–TeO2 and WO3–TeO2 glasses, Phase Transitions, 86, (2013) 598-619. 15

[61] vd Steen, G. H. A. M.. Introduction and removal of hydroxyl groups in vitreous silica Eindhoven: Technische Hogeschool Eindhoven, 1976, DOI: 10.6100/IR76787. [62] V.V. Dorofeev, A.N. Moiseev, M.F. Churbanov, G.E. Snopatin, A.V. Chilyasov, I.A. Kraev, A.S. Lobanov, T.V. Kotereva, L.A. Ketkovaa, A.A. Pushkin, V.V. Gerasimenko, V.G. Plotnichenko, A.F. Kosolapov, E.M. Dianov, High-purity TeO2–WO3–(La2O3, Bi2O3) glasses for fiber-optics, Optical Materials 33 (2011) 1911–1915. Figure Caption

16

Figures

Fig. 1: The heat-treatment procedure for the different glass samples. 2700

a)

TNZF1 Before heat treatment TNZF1 Heat treatment 1 TNZF1 Heat treatment 2

1800

Intensity (a.u)

900 10 2700

20

30

40

b)

50

60

TNZF2 Before heat treatment TNZF2 Heat treatment 1 TNZF2 Heat treatment 2

1800

900 10 2700

20

30

40

c)

50

60

TNZF3 Before heat treatment TNZF3 Heat treatment 1 TNZF3 Heat treatment 2

1800 900 10

20

30

40

degree

50

60

Figure 2: a) XRD diagram of the sample TNZF1. b) XRD diagram of the sample TNZF2. c) XRD diagram of the sample TNZF3

17

TNZF1 TNZF2 TNZF3

Tg

0

Tc1 T

Heat Flow exo

c2

-4

Tf

-8

-12

-16

-20 300

400

500

600

700

800

900

Temperature (°C) Fig. 3: DSC curves for the different glass samples.

25000

B

TNZF1 TNZF2 TNZF3

C

Intensity

20000

A

15000

D

10000

5000

200

400

600

800

1000

Raman shift cm-1

Fig. 4: Raman spectra of TNZF glasses.

18

TNZF1 TNZF2 TNZF3

-1

Absorption coefficient (cm )

3

2

1

0 3000

3200

3400

3600

3800

4000

Wavenumber (cm-1)

Fig. 5: Hydroxyl absorption band of the different glass samples before heat treatment. TNZF1 TNZF2 TNZF3

12000

Intensity (u.a)

8000

4000

520

540

560

580

600

620

640

500

660

680

TNZF1 TNZF2 TNZF3

400 300 200 100 0 520

540

560

580

600

620

640

660

680

Wavelength (nm) Fig. 6: a) the luminescence spectra of the different glasses under 475.1 nm excitation before heat treatment. b) the upconversion luminescence spectra of the different glasses under 980 nm excitation before heat treatment.

19

TNZF1 TNZF2 TNZF3

-1

Absorption coefficient (cm )

0,14

0,07

0,00 3100

3200

3300

3400

3500

3600

3700

3800

3900

TNZF1 TNZF2 TNZF3

0,06 0,04 0,02 0,00 3200

3300

3400

3500

3600

3700

3800

-1

Wavenumber (cm )

Fig. 7: a) Hydroxyl absorption band of the different glass samples after the first heat-treatment step. b) Hydroxyl absorption band of the different glass samples after the second heat-treatment step. 18000

Before heat treatment Heat treatment T1 Heat treatment T2

Intensity (u.a)

12000

6000

0 520

560

600

1200

640

680

Before heat treatment Heat treatment T1 Heat treatment T2

800

400

0 520

560

600

640

680

Wavelength (nm) Fig. 8: a) the luminescence spectra under 475.1 nm excitation of the TNZF2 glass

sample after each drying step. b) the upconversion luminescence spectra of the TNZF2 glass under 980 nm excitation after each drying step.

20

Table 1: Densities ρ, erbium (NEr) of ytterbium (NYb) concentrations in the different glass samples.

20 cm- NYb(1020 cm- ρ (g cmEr(10 Molar composition (%) N 3 3 3

Label

)

)

)

97.2 4.797

:1.4Er2O3 2.23 TNZ 60.1TeO2-17.2ZnO-8.6Nb2O5-10.8BaF2 /1.9Yb 2O3 F1 55.4TeO2-15.8ZnO-7.9Nb2O5- 2.22 TNZ 17.5BaF2:1.4Er2O3/1.9Yb2O3 F2 51.4TeO2-14.7ZnO-7.3Nb2O5-- 2.19 TNZ 23.2BaF2:1.4Er2O3/1.9Yb2O3 F3

2.77 4.871 2.75

5.565

Table 2: Glass transition (Tg), onset (Tx) and maximum (Tc) temperature of the crystallization, and deduced from DSC curves of the different glass samples.

Samples

Tg(°C)

TC1(°C)

TC1(°C)

Tx (°C)

∆T (°C)

TNZF1

438

604

****

575

137

TNZF2

439

610

****

565

126

TNZF3

440

628

659

609

169

21

Table 3: Absorption coefficients and densities of hydroxyl groups in different tellurite glasses subjected to different drying procedures. matrix

OH (cm- NOH 1 cm-3) )

Drying method

2.706

3.319

1.430

1.754

TNZF3

1.732

2.124

TNZF1

0.176

0.216

0.137

0.168

TNZF3

0.161

0.197

TNZF1

0.039

0.048

0.063

0.077

0.040

0.049

TNZF1 As-quenched

TNZF2

Heat-treatment 1

TNZF2

Heat-treatment 2

TNZF2 TNZF3

(1019

Ref

Present work

Present work

Present work

Nb2O5–ZnO–TeO2–AlF3– PbF2–ZrF4

Melting with stirring and dry 1.09 O2 + CCl4 bubbling

1.17

7

Germanotellurite glass

Mixture of dry oxygen and 1.99 CCl4

2.44

34

high purity oxygen

1.2

14.7

35

O2 atmosphère

0.008

0.1

8

0.1574

0.2

36

75TeO2 – 20ZnO La2O3 :Er2O3



80TeO2–10ZnO–10Na2O

76 TeO2 – 12.5 ZnF2 – 5 Bi2O3 – 6.5 GeO2 : Tm2O3 a dry glovebox 70TeO2–10Bi2O3–20ZnF2– 0.04BaO

melted in O2-rich atmosphere with preheat treatment with --NH4F–HF

0.32

14

80TeO2–10Na2O–10ZnO

dry Cl2/O2 gas mixture

11

15

70TeO2–15B2O3–10GeO2– 5Na2O– 1 wt%Er2O3

O2 bubbling

3.794

0.65

37

0.674

0.8

38

0.9

1.1

39

TeO2 -39 RF2 (R=Ba, Mg, Using great Fluorides Zn)-11 NaF 74 TeO2-10ZnO-10NaO2-5 GeO2-1Er2O3

amount

Dry O2 flow

22

of

75TeO2-25WO3- 69TeO2Dry O2 flow 23WO3- 8La2O3

1.2

23

62