Spectroscopic properties of Sm doped sodium ...

88 downloads 0 Views 2MB Size Report
0.6. 141.21. TNS4. 79.0. 20.0. 1.0. 141.97. TNS5. 78.8. 20.0. 1.2. 142.34. TNS6. 78.5. 20.0. 1.5. 142.91. S.Q. Mawlud et al. / Optical Materials 69 (2017) 318e327.
Optical Materials 69 (2017) 318e327

Contents lists available at ScienceDirect

Optical Materials journal homepage: www.elsevier.com/locate/optmat

Spectroscopic properties of Sm3þ doped sodium-tellurite glasses: Judd-Ofelt analysis Saman Q. Mawlud a, b, Mudhafar M. Ameen a, Md. Rahim Sahar b, *, Zahra Ashur Said Mahraz b, Kasim F. Ahmed a, b a b

Department of Physics, College of Education, University of Salahaddin, Erbil, Kurdistan Region, Iraq Advanced Optical Material Research Group, Department of Physics, Faculty of Science, University of Technology Malaysia, Skudai, Johor, Malaysia

a r t i c l e i n f o

a b s t r a c t

Article history: Received 7 January 2017 Received in revised form 7 March 2017 Accepted 10 April 2017

Modifying the optical response of rare earth doped inorganic glasses for diverse optical applications is the current challenge in materials science and technology. We report the enhancement of the visible emissions of the Sm3þ ions doped sodium-tellurite (TNS) glasses. The impacts of varying Sm3þ ions concentration on the spectroscopic properties of such glass samples are evaluated. Synthesized glass samples are characterized via emission and absorption measurements. The UVeViseNIR absorption spectra revealed nine absorption peaks which are assigned to the transitions from the ground level (6H5/ 6 4 6 6 6 6 6 6 6 3þ ions. Emission 2) to P3/2, I11/2, F11/2, F9/2, F7/2, F5/2, F3/2, H15/2 and F1/2 excited energy levels of Sm spectra of the prepared glass under 404 nm excitation wavelength consisted of four bands centered at 561 nm, 598 nm, 643 nm and 704 nm which are originated from 4G5/2/6HJ (J ¼ 5/2, 7/2, 9/2 and 11/2) transitions. The experimental oscillator strengths, fexp are calculated from the area under absorption bands. Using Judd-Ofelt theory and fit process of least square, the phenomenological intensity parameters Ul (l ¼ 2, 4, 6) are obtained. In order to evaluate potential applications of Sm3þ ions in telluride glasses, the spectroscopic parameters: radiative transition probability AR, branching ratio BR, radiative life time tr and stimulated emission cross section sl for each band are calculated. These glass compositions could be a potential candidate for lasers. © 2017 Elsevier B.V. All rights reserved.

Keywords: Tellurite glasses Samarium ions Optical absorption Luminescence JuddeOfelt analysis

1. Introduction Rare earth (RE) ion activated materials are considered as one of the most interesting research areas due to their various applications, e.g. lasers, sensors, telecommunication, display devices, etc [1]. Recent years have witnessed a tremendous increase in research activities related to glasses doped with rare earth ions in various forms such as network formers, modifiers or luminescent ions [2e4]. In these systems, there are a number of interesting relationships between the active ions and the host glasses. Among these, one important point is that glasses with low phonon energies are of interest as hosts for infrared and infrared to visible upconversion lasers, because the glass host with low phonon energy can reduce the non-radiative loss due to the mutiphonon relaxation and thus achieves strong upconversion luminescence.

* Corresponding author. E-mail addresses: [email protected] (S.Q. Mawlud), mrahim057@ gmail.com (Md.R. Sahar). http://dx.doi.org/10.1016/j.optmat.2017.04.022 0925-3467/© 2017 Elsevier B.V. All rights reserved.

Consequently, it is important to select a host material for which the maximum phonon energy is as low as possible. In tellurite based glasses, this phonon energy is reasonably large (~700 cm1). Moreover, tellurite glasses are considered as excellent materials for hosting lasing ions due to their better thermal and chemical stability, low melting temperature, high thermal expansion, good infrared transmission, high refractive index and capable of incorporating large concentrations of rare earth ions into the matrix [5e7]. However, tellurium dioxide (TeO2) itself is only a conditional glass-former, which requires a special fast-quenching procedures to vitrify. Due to the difficulty of vitrifying TeO2 alone by traditional method, the high transparent tellurite glasses are obtained by introducing other oxide such as transition metal oxides, alkaline oxides and alkaline-earth oxides or any other glass former [1]. In fact, sodium oxide (Na2O) is nominated as the best modifier amongst the other alkali oxides. Additionally, Na2O presents the most glass-forming ability on the basis of stability against crystallization. Among active RE ions, the trivalent samarium (Sm3þ) ions in glassy matrix exhibits efficient fluorescence in a wide spectral

S.Q. Mawlud et al. / Optical Materials 69 (2017) 318e327

range from ultraviolet to infrared region which can be used as undersea communications, optical storage materials, in highdensity memories, colour displays, medical diagnostics and solidstate laser [8e10]. Glasses doped with Sm3þ ions exhibit relatively high quantum efficiency because of the large energy gap between the 4G5/2 level and the next lower lying energy level 6F11/2, which is approximately 7200 cm1. Additionally, the Sm3þ ions exhibit broad emission bands due to 4G5/2/6HJ (J ¼ 5/2, 7/2, 9/2, 11/2) transitions in any host matrix [11,12]. It is also well known that the intensities of emission bands of Sm3þ ion in glasses depend on its concentration and glass composition. In order to obtain optimum emission characteristics for device applications, the characteristic features of host as well as concentration dependent studies of Sm3þ ions are essential. Over the years, a vast amount of data on the spectroscopic properties of trivalent lanthanide ions in glasses was collected. However, the synthesis and optical analysis of Sm3þ ions doped tellurite glasses have seldom been studied. The complicated electronic structure of the 4f5 configuration is responsible for this neglect. Usually, the JuddeOfelt (J-O) theory has been used for the analysis of optical spectra of RE ions in various hosts. J-O theory is a useful tool for characterizing the radiative transition probability for different RE doped materials and to approximate the intensities of transitions in RE [13,14]. Three intensity parameters, Ul (l ¼ 2, 4, 6) can be defined by this theory which are very sensitive to RE environment. These parameters are further used to estimate the other radiative parameters such as branching ratio, radiative transition probability and stimulated emission cross-section. Understanding the optical properties of RE doped tellurite glasses are prerequisite for optical applications. Therefore, in present work, a comprehensive analysis of spectroscopic properties of Sm3þ:TNS glasses by optical absorption and luminescence spectra were reported. The aim of present study is to synthesize the Sm3þ:TNS glasses with varying concentration of Sm2O3, to examine the energy levels to calculate the oscillator strength and J-O intensity parameters. Various radiative properties such as radiative transition probability, radiative lifetime, stimulated emission cross-section and optical gain are determined taking the J-O intensity parameters into account. All measured and calculated results were compared with similar Sm3þ ions doped glass systems.

20 min in order to reduce the batch blanket coverage on the top of the glass and enlarges the free non coverage glass melt surface [15]. After that, the batch transferred to another controlled electric furnace for melting it at 900  C for about 40 min. The melts were poured on to a stainless steel mold and annealed at 250  C for 3 h. After three hours, the furnace is switched off and the samples are allowed to cool down gradually to room temperature (25  C). Finally, the samples are polished until the appropriate thickness (2.5 ± 0.01 mm) for structural and optical measurements. The amorphous nature of glasses are examined via a Bruker D8 Advance X-ray diffractometer (XRD) which uses CuKa radiations (l ¼ 1.54 Ao) at 40 kV and 100 mA. The optical absorption of the prepared samples measured at a room temperature by using UVeViseNIR (Shimadzu 1301PC) spectrophotometer and the used wavelength range was 200e1000 nm. Perkin-Elmer LS-55 luminescence spectrometer is used to record the emission and excitation spectra, and pulsed xenon lamp operates as an excitation source, the luminescence spectra in the range of wavelength 200e900 nm under the excitation wavelength 404 nm is used in present research. 3. Results and discussion 3.1. Physical parameter Density of the glass samples was determined by using Archimedes' method. The weight of the samples was measured by using an accurate 4 digit sensitive analytical electronic balance. Precisa XT220A, toluene was used as an immersed liquid with the density (rt ¼ 0.8669 g cm3) at room temperature. The density r was calculated using the expression [16]:





The glass samples have nominal composition (80-x)TeO220Na2O-xSm2O3 where x ¼ 0.0, 0.3, 0.6, 1.0, 1.2 and 1.5 mol% are prepared using melt-quenching technique. The glass samples code and their compositions ratio are listed in Table 1. The raw materials with required proportion are weighted using a very sensitive weighing machine (Electronic Balance Precisa 205A SCS). The total weight of each batch of glass is 15 g and calculated in mol percent (mol%). A milling machine was used for mixing the chemical compositions before putting inside a Platinum crucible of about 30 ml capacities. Then the batch was preheated at 250  C for about

Sample Code

Glass composition (mol%) TeO2

Na2O

Sm2O3

TNS1 TNS2 TNS3 TNS4 TNS5 TNS6

80.0 79.7 79.4 79.0 78.8 78.5

20.0 20.0 20.0 20.0 20.0 20.0

0.0 0.3 0.6 1.0 1.2 1.5

Molar Mass [g mol1]

140.07 140.64 141.21 141.97 142.34 142.91



 rt þ ra

 (1)

M

(2)

r

where M is the molar weight and r is the density of the glass. The ionic packing density (Vt) is calculated using Makishima and Mackenzie approach [17,18].

 Vt ¼

 X 1 * ðVi *xi Þ VM

(3)

where xi is the mole fraction (mol%) and Vi is packing density parameter (m3/mol). For an oxide glass of the form MxOY, the value of Vi yields [18],

 Vi ¼

Table 1 Glasses samples code and their concentration ratio.

Wa Wa  Wt

where Wa and Wt are weight of the glass sample in air and in toluene respectively, ra is the density of air and rt is the density of the toluene. The molar volume (VM) of the glasses was calculated from density values according to [17]:

VM ¼

2. Experimental procedure

319

 i 4pNA h 3 XrM þ Yro3 3

(4)

where NA is Avogadro's number (mol1), rM and ro are the Shannon's ionic radius of metal and oxygen, respectively. The refractive index (n) of glass in terms of optical band gap (Eopt) is obtained by using [19,20].

n2  1 ¼1 n2 þ 2

rffiffiffiffiffiffiffiffiffi Eopt 20

The molar refractivity (RM) can be obtained from Ref. [21],

(5)

320

RM ¼

S.Q. Mawlud et al. / Optical Materials 69 (2017) 318e327

n2  1 ðVM Þ n2 þ 2

(6)

Polarizability (ae) is related to the refractive index through the Lorentz-Lorentz equation and can be expressed as [21],

n2  1 4 ðVM Þ ¼ pNA ae 3 n2 þ 2

(7)

The effects of Sm3þ ion on the physical properties of the prepared glass samples were summarized in Table 2. Fig. 1 shows the variation of Sm2O3 concentration plotted against density and molar volume of the prepared glass. The density of Sm3þ:TNS glass system increases from 4.903 g cm3 for x ¼ 0 mol% to 5.019 g cm3 for x ¼ 1.2 mol% of Sm3þ ion. Higher molecular mass of Sm3O2 (348.74 g mol1) and those for TeO2 (159.60 g mol1) might be responsible for increase its density [22]. The decrease in density at x ¼ 1.5 mol% might be caused by Sm3þ ions which take a part in the structure of glass and make the density to be decreased [23]. Moreover, the increase in molar volume results in an increase in packing density of oxygen [24]. This behavior of molar volume is in consistent with the increasing behavior of rigidity and compactness of glass samples [25]. The variation of refractive index can be explained in term of disorder effect in the glass system. The plot of refractive index and density versus average molar mass are is shown in Fig. 2. At the beginning of insertion Sm2O3 from 0.3 to 0.6 mol%, which cause the increase of molar mass and then the refractive index gradually decreases. This decrement is due to the formation of TeO3þ1 polyhedral from TeO3 tbp in the glass network. In this part, the Sm atom is less impact of comparison toward glass lattices and consequently resulted in a lower lattice strain. Further addition of Sm2O3 up to 1.5 mol% shows that refractive index increased which reflects to the formation of NBO. This increment is due to the alteration of structural network with the transformation of TeO4 tbp to TeO3þ1 tp as has been observed in the result of optical energy band gap. The structure changes towards more disordered state with the substitution of Sm atoms [26e28]. Indirect band gap is determined due to bridging the gap between valence and conduction bands with contribution of lattice vibrations (phonons). Indirect band gap is first discussed by Davis and Mott [29] as,

Fig. 1. The Sm2O3concentration (mol%) dependent density and molar volume.

Fig. 2. The average molar mass dependent of density and refractive index.

usually follows the Urbach rule [30],

aðyÞ ¼ a0 expðhy=ЕU Þ

 r hy  Eg aðyÞ ¼ B hy

(8)

where a is the frequency (y) dependent absorption, B is a constant, Eg is the optical band gap and the index r signifies the nature of the electronic transitions which takes a value 2 for indirect allowed transitions. A graph between (ahy)1/r versus photon energy hy called Tauc plot is generated by substituting the value of r ¼ 0 in Equation (8) to obtain the optical band gap for indirect (Eind) allowed transitions. The linear portions of the curves are extrapolated to (ahy)1/2 ¼ 0 to determine the values of Eind. The fundamental absorption edge in most amorphous materials

(9)

where a0 is a constant and EU is the Urbach energy which is calculated by the inverse slope of the lna against hy in the lower photon energy level. This exponential behavior is due to the band tails associated with the valence and conduction bands which extends into the band gap. Fig. 3 shows the curves in determination of the indirect band gap and Fig. 4 illustrates the behavior of Urbach energy of the prepared glass system. Indirect band gap energies were listed in Table 2 along with the Urbach energy. The values of Eind were found to lie between 2.544 eV and 2.793 eV for the studied glass samples. The indirect optical band gap energies are found to increase with

Table 2 Physical and optical properties of Sm3þ:TNS glasses. Sample Name

r [gm cm3]

VM [cm3 mol1]

Vt

Eu [ev]

Eind [ev]

n

RM [cm3mol1]

ae  10^18 [cm3]

TNS1 TNS2 TNS3 TNS4 TNS5 TNS6

4.903 4.914 4.920 4.965 5.019 4.942

28.571 28.623 28.700 28.593 28.359 28.919

0.4012 0.4017 0.4019 0.4051 0.4092 0.4025

0.312 0.256 0.226 0.217 0.209 0.309

2.544 2.695 2.769 2.793 2.712 2.613

2.385 2.354 2.349 2.355 2.362 2.367

17.424 17.233 17.245 17.224 17.131 17.509

4.839 4.786 4.790 4.784 4.758 4.863

S.Q. Mawlud et al. / Optical Materials 69 (2017) 318e327

Fig. 3. Plot of ðaZuÞ1=2 versus Zu of the prepared Sm3þ:TNS glass system.

321

Fig. 5. The Sm2O3concentration (mol%) dependent band gap and Urbach energy.

Fig. 4. Plot of ln a versus Zu of the prepared Sm3þ:TNS glass system.

Fig. 6. Typical XRD pattern of Sm3þ:TNS glasses system.

the increase in Sm2O3 concentration up to 1.0 mol% and decrease thereafter. Shifting of the absorption band to a lower energy can be related to the formation of non-bridging oxygen (NBO) which binds exited electrons of NBO less tightly than bridging oxygen [27], and this cause the glass network to become less rigid. Consequently, transfer of electrons from the valence band to the conduction band becomes easier. The existence of a band tailing in the forbidden energy band gap in glass and amorphous materials represent the disorder in the material, the decrease in EU from 0.312 to 0.209 eV with the increase of Sm3þ ion exemplifies the decreasing degree of disorder present in the glasses [19]. Fig. 5 presents the variation of indirect band gap and Urbach energy of the prepared samples with different concentration of Sm3þ ions. From this figure, it was concluded that these behaviors signify the more extension of the localized states within the gap [31].

ray diffraction patterns confirms that there are no well-defined planes in the structure on/or around the constituent atoms were regularly arranged [27].

3.2. X-ray diffraction (XRD) X-ray diffraction technique is used to determine the vitreous states of the sample. Fig. 6 shows the diffraction patterns of all the prepared glass samples. The broad humps over the region 20 -35 for 2q values confirms the amorphous nature of the glass sample [19], which indicates the absence of long range atomic arrangement and the periodicity of the three dimensional network in the quenched sample. The absence of sharp diffracted beams in the X-

3.3. Absorption spectra The absorption spectra of the prepared glass system recorded at room temperature in the wavelength range 400e1800 nm are shown in Fig. 7. Where nine significant absorption peaks were observed due to the 4f5-4f5 electronic transitions from the 6H5/2 ground manifold to different excited manifolds of Sm3þ ions. The absorption bands are due to transition from 6H5/2 to the excited state 6P3/2, 4I11/2, 6F11/2, 6F9/2, 6F7/2, 6F5/2, 6F3/2, 6H15/2 and 6F1/2 centered at about 400, 474, 949, 1086, 1239, 1387, 1491 and 1597 nm. The majority of the transitions in the spectra originate from induced electric-dipole interactions with the selection rule DJ  6. However, certain transitions contain magnetic-dipole contribution that follow the selection rule as DJ ¼ 0, ±1. The 4f electrons are very effectively shielded by the filled 5s and 5p orbitals, and gives rise to sharp and intense bands. The partial energy level diagram of Sm3þ ion is schematically shown in Fig. 8 in order to discuss the down-conversion process and the involved mechanisms. Ground state absorption and excited state absorption are the important involved mechanisms. The nature of the rare eartheoxygen (REeO) bonding of the

322

S.Q. Mawlud et al. / Optical Materials 69 (2017) 318e327

prepared glasses can be expressed by the nephelauxetic ratios (b) and bonding parameters (d), these parameters have been calculated for the Sm3þ: TNS glasses using the following relation [32]:



vc va

(10)

where vc is the wavenumber (in cm1) of a particular transition of the RE ion under investigation and va is the wavenumber (in cm1) for the corresponding transition of an aquo-ion [33]. From the 

average values of b(referred as b) the bonding parameter d is calculated using the following formula [32]: 



1b

(11)

b

The bonding will be covalent or ionic depending upon the positive or negative sign of the d. The b and d values for all the prepared samples are shown in Table 3. Nephelauxetic effect occurs due to metal-ligand bond formation, that is, overlap between the metal and ligand orbitals forming larger molecular orbital leading to delocalization of the electron cloud over a larger area. From Table 3 it can be observed that for all samples the bonding parameter (d) is negative, so the studied glasses show bonds with an ionic character. Our results also show that, the d values decrease with decreasing the Sm3þ content of the samples. This is due to the fact that decreasing the number of free ions from the glass network leads to the decrease of its metallic character. These results are in agreement with the previously published data [34,35], where Sm3þ ions doped glasses were reported to have bonds with an ionic character.

Fig. 7. A typical UVeViseNIR absorption spectra of all prepared glasses.

3.4. Luminescence spectra The luminescence spectra in the range of 200e900 nm at the excitation of 404 nm were shown in Fig. 9. The emission spectra exhibit four emission bands centered at 561 nm, 598 nm, 643 nm and 704 nm. The emission peaks assigned to the transition of 4G5/ 6 4 6 4 6 4 6 2 / H5/2, G5/2 / H7/2, G5/2 / H9/2 and G5/2 / H11/2 respectively [36]. From these emissions band a possible of moderate green, moderate yellow, intense orange and feeble red color could be expected. Fig. 10 shows the simplified emission energy diagram of Sm3þ:TNS glass, it can be seen that, as the Sm3þ ions are pumped with 404 nm excitation wavelength, they excites to 6P3/2 level, then some of Sm3þ ions were found relaxed non-radioactively

Fig. 8. Energy level diagram of absorption transitions of Sm3þ:TNS glass matrix.

Table 3 The transitions, energy levels, nephelauxetic and bonding parameters of the Sm3þ:TNS glasses. Transition

Energy Level (cm1)

Aquo-ion [31]

TNS2

TNS3

TNS4

TNS5

TNS6

24961 21142 10550 9201 8070 7208 6698 6440 6266

25021 21275 10539 9195 8070 7206 6698 6452 6272

25000 21200 10548 9197 8070 7205 6698 6447 6266

25001 21275 10542 9196 8068 7210 6700 6450 6278

25022 21277 10553 9193 8068 7213 6705 6448 6276

24999 21096 10517 9136 7977 7131 6641 6508 6397

b

9.0110

9.0205

9.0152

9.0212

9.0235

e

`

1.0012

1.0023

1.0017

1.0024

1.0026

e

d

0.1216

0.2278

0.1681

0.2345

0.2602

e

6

H5/2 H5/2 6 H5/2 6 H5/2 6 H5/2 6 H5/2 6 H5/2 6 H5/2 6 H5/2 6

b

/ / / / / / / / /

6

P3/2 I11/2 6 F11/2 6 F9/2 6 F7/2 6 F5/2 6 F3/2 6 H15/2 6 F1/2 4

S.Q. Mawlud et al. / Optical Materials 69 (2017) 318e327

Fig. 9. Down conversion photoluminescence spectra of Sm3þ:TNS glass system.

323

Fig. 11. Emission peak wavelength and energy of the transition for Sm3þ:TNS glass sample excited with 404 nm.

configuration of the Sm3þ ion have been analyzed using the J-O theory [19,28]. According to J-O theory the calculated oscillator strengths (fcal), induced electric-dipole transitions from initial state jJ to the final state j0 J 0 depends on three parameters Ul (l ¼ 2, 4, 6) as:

 fcal ¼

8p2 mcv 3hð2J þ 1Þ

2 # " 2

2 X n þ2  Ul jJ

U l

j0 J 0 9n

(12)

l¼2;4;6

where n is the wavenumber (cm1) of the transition from ground statejJ to excited state j0 J 0 , n is the refractive index, c is the velocity of light in vacuum, m is the mass of an electron, ðn2 þ 2Þ2 =9n is the Lorentz local field correction accounts for the dipole-dipole transition, J is the total angular momentum of the ground state, Ul is the

2



J-O intensity parameter and U l are the squared doubly reduced matrix elements of the unit tensor operator, which are evaluated from the intermediate coupling approximation for a transition from (jJ) to (j0 J 0 ). The experimental oscillator strength (fexp) is directly proportional to the area under the absorption curve and is defined as [31,32]. Fig. 10. Schematic energy level diagram of Sm3þ ion shows ground state absorption and excited state absorption mechanisms.

to a lower levels 4G7/2, 4F3/2 and then they decay to 6H9/2. Meanwhile, some of the Sm3þ ions were found relaxed non-radioactively to 4G5/2 level that lead to populate corresponding level and emit moderate green, moderate yellow, intense orange and feeble red color. All emission peak wavelength and energy of the transition under 404 nm excitation were shown in Fig. 11. Table 4 shows the florescence peak center and full width at half maxima for all the luminescence transition which will be used for calculating the radiative parameters of Sm3þ:TNS glasses under investigation. From the luminescence spectra, the normalized fluorescence intensity of Sm3þ ions at transitions band of 4G5/2 / 6H5/2, 4G5/ 6 4 6 4 6 2 / H7/2, G5/2 / H9/2 and G5/2 / H11/2 was increased with addition of Sm3þions as shown in Fig. 12. 3.5. Oscillator strengths, Judd-Ofelt analysis and radiative properties The radiative transitions probability associated within the 4f5

fexp ¼

2:303mc2 N pe2

Z

εðvÞdv ¼ 4:318  109

Z εðvÞdv

(13)

where N is the Avagadro's number, εðvÞ is the molar absorptive of the band at a wave number v (cm1), e is the electron charge. In fact, the estimation of J-O parameters is a problem of solving simultaneous linear equations. The accuracy of the fit is given by root mean square error dRMS defined by nonlinear Gauss-Seidel routine based on the least square fitting [37,38].

dRMS

rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2 . XN f ðN  3Þ ¼  f exp cal i¼1

(14)

where N is the number of absorption band involved in the J-O calculation. The RMSerror will be listed with the obtained results, which is given in the following [39],

RMSerror ¼ where

dRMS fRMS

 100%

(15)

324

S.Q. Mawlud et al. / Optical Materials 69 (2017) 318e327

Table 4 Florescence peak center and full width at half maxima (FWHM, nm) of Sm3þ:TNS glasses under investigation. Transition G5/2/

TNS2 Peak Center cm1

FWHM nm

Peak Center cm1

FWHM nm

Peak Center cm1

FWHM nm

Peak Center cm1

FWHM nm

Peak Center cm1

FWHM nm

6

17705.7 16671.4 15509.9 14215.0

10.0 12.4 13.2 14.1

17754.1 16663.1 15511.1 14214.8

8.3 13.7 12.9 18.9

17756.0 16660.3 15501.7 14198.5

9.0 12.3 13.3 16.3

17725.8 16663.3 15507.0 14193.3

7.7 12.1 13.1 14.1

17757.9 16655.0 15503.9 14183.6

9.9 12.5 13.4 16.0

TNS3

TNS4

TNS5

TNS6

4

H5/2 H7/2 H9/2 6 H11/2 6 6

hypersensitive transition (HSTs) being very sensitive to small changes in environment around Ln ions obeys the selection rules jDJj  2, jDL  2j and DS ¼ 0. The absorption bands of the 6H5/ 6 6 6 6 6 6 6 2 / F3/2, H5/2 / F5/2, H5/2 / F7/2 and H5/2 / F9/2 are hypersensitive in nature as they are more intense as compared to other transition. This indicates low site symmetry around Sm3þ ion in present glass systems. It can be observed, both the experimental and calculated oscillator strength showed the highest value at the transition 6H5/2 / 6F3/2, 6H5/2 / 6F5/2, 6H5/2 / 6F7/2 and 6H5/ 6 2 / F9/2. This is due to the hypersensitive transition which always has the highest intensity in spectra. This highest intensity of HSTs values suggests the low site symmetry around Sm3þ ions because low site symmetry will assist to increase the intensity of absorption line strength. The intensity of the HSTs transition of Sm3þ ions decrease with the decrease of covalency between the Sm3þ ions and the surrounding ligand environment [33,35]. The dRMS deviation gives the quality of the fit between experimental and calculated oscillator strengths. The dRMS values are found to be (0.28, 0.51, 1.17, 5.82 and 4.09  106) corresponding to TNS2, TNS3, TNS4, TNS5 and TNS6 glasses, respectively. Our result implies the good fit of the measured spectral intensities. The oscillator strength values of the prepared TNS glasses are similar to the other reported Sm3þ doped glasses [25,33e35]. The J-O intensity parameters are host dependent and important in exploring the glass structure and transition rate of the rare earth ion energy levels. The trends of the J-O parameters in the present work were found in increasing order of U2>U4>U6 for the prepared Sm3þ:TNS glasses. The J-O intensity parameter reveals a variation with increasing concentration of Sm3þ ions. This relation has been plotted out and the result is shown in Fig. 13. The U2 , J-O intensity parameter of the Sm3þ:TNS glasses were found to be associated with the covalency, structural change and symmetry of the ligand field around Sm3þ site [36]. TNS3 has the largest value of U2 compared to other glass samples, which indicates that sites

Fig. 12. Normalized luminescence intensity as a function of emission peak wavelength of Sm3þ:TNS glass excited at 404 nm.

fRMS

vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi u N uX 2 . ¼t f exp N

(16)

i¼1

The experimental and calculated oscillator strengths of the various absorption bands along with dRMS deviation are presented in Table 5. The strong dependence of the oscillator strengths on Sm2O3 contents is shows in Table 5 implies that the non-symmetric component of the electric field acting on the Sm3þ:TNS glasses is relatively high. Few of the transitions corresponding to the Sm3þ ions are known as hypersensitive transition (HSTs) being very sensitive to the surrounding environment of ions. The

Table 5 The values of the experimental and calculated oscillator strength (  106) of Sm3þ in Sm3þ:TNS glass matrix with different concentrations. dRMS and fRMS are given in units of  106. Transition

6

H5/2 H5/2 H5/2 6 H5/2 6 H5/2 6 H5/2 6 H5/2 6 H5/2 6 H5/2 6 6

/ / / / / / / / /

6

P3/2 I11/2 F11/2 6 F9/2 6 F7/2 6 F5/2 6 F3/2 6 H15/2 6 F1/2 4 6

TNS2

TNS3

TNS4

TNS5

TNS6

fexp

fcal

fexp

fcal

fexp

fcal

fexp

fcal

fexp

fcal

0.16 0.64 0.61 6.41 14.09 8.56 2.86 0.91 0.093

1.36 0.57 1.39 8.57 12.53 6.67 3.08 0.71 0.14

1.77 3.18 2.79 23.42 45.12 30.59 8.65 2.56 0.26

4.72 1.88 4.56 28.15 41.75 23.13 10.44 2.29 0.16

1.62 5.83 7.52 64.93 123.05 97.15 25.35 5.49 0.71

1.46 4.97 12.06 75.21 116.11 72.47 35.73 6.05 4.72

0.89 10.96 11.45 98.38 189.77 126.26 35.37 10.36 1.42

1.95 7.94 19.24 118.41 174.45 94.96 43.06 9.67 0.93

5.09 16.36 21.63 161.14 309.12 203.28 57.96 13.71 2.44

3.13 27.91 31.67 194.62 285.62 156.59 83.42 15.91 1.89

0.28

0.51

1.17

5.82

4.09

fRMS

11.44

39.45

110.55

161.62

263.58

RMSerror %

2.45

1.30

1.06

3.60

1.55

dRMS

S.Q. Mawlud et al. / Optical Materials 69 (2017) 318e327

325

presented in Table 6. The values of U4/U6 are found to vary in the range of 1.28  1020 cm2 to 1.61  1020 cm2. It is observed that, the magnitude of U4/U6 for TNS4 glass sample indicates optimum features among the prepared glasses. The best set of Ul parameters was determined by a standard least-square fitting of the theoretical oscillator strength values to the measured ones. The radiative properties can be calculated through the Ul values. The radiative transition probability (A) is expressed as [25].

  A jJ; j0 J 0 ¼ Aed þ Amd ¼

64p4 3

3hl ð2J þ 1Þ



!  2 n n2 þ 2 3 Sed þ c Smd 9

(17)

where Aed is electric-dipole and Amd is magnetic-dipole contribu-

Fig. 13. Variation of J-O parameters Ul for Sm3þions in the prepared systems as a function of Sm2O3.

tions, respectively. ðn2 þ 2Þ2 =9n is the local field correction for the electric dipole transitions, and c3 is the local field correction for magnetic dipole transitions. Sed and Smd correspond to electric and magnetic dipole line strengths, which are expressed as

X

Sed ¼ e2



2

(18)

l¼2;4;6



occupied by Sm ions in TNS3 are the most asymmetric in nature and the chemical bond between Sm3þ ions and ligand ion is the most covalent. On the other hand, the decrease of U2 beyond 0.6 mol% Sm2O3 indicates the increase in the symmetrical degree around Sm3þ ion site. The J-O intensity parameters U4 and U6 refer to the viscosity of the glass matrix and dielectric of the media, which are affected by the vibronic transitions of the RE ions bound to the ligand atoms [37,38]. The higher values of U4 and U6 suggests that the structural network will be more open and becoming less rigid. This would also mean that the NBO is formed, which reflects the deformation of TeO4 tbp structural unit to TeO3þ1 polyhedral or TeO3 tp structural unit. The spectroscopic quality factor c ¼ U4 =U6 is used to characterize the quality of the prepared glasses [6,35]. The J-O parameters values, trends and the spectroscopic quality factor of the prepared Sm3þ:TNS glasses are quite comparable to the other reported Sm3þdoped glass systems and are



Ul jJ

U l

j0 J 0

and

Smd ¼

 2 e2 h2 Ul jJkL þ 2Skj0 J 0 2 2 2 16p m c

(19)

The total radiative transition probability AT is given as the sum of AðjJ; j0 J 0 Þ terms calculated over all terminal levels and is given as

AT ðjJÞ ¼

X   A jJ; j0 J 0

(20)

The radiative life time (tR ) of the jJ level is given by

tR ðjJÞ ¼ ½AT ðjJÞ1

(21)

The branching ratio (bR ) can be obtained from the equation given below

Table 6 Judd-Ofelt intensity parameters and spectroscopic quality factor (c ¼ U4 =U6 ) of varying Sm3þ concentration in the TNS glass matrix. Intensity parameters Ul (  1020 cm2)

Glass Code

TNS2 TNS3 TNS4 TNS5 TNS6 Zinc tellurite Phosphotellurite Fluoro-tellurite Tungsten tellurite

U2

U4

U6

2.486 2.735 0.802 1.58 0.318 0.85 4.54 4.94 6.68

0.668 0.233 0.719 0.947 0.152 0.01 1.12 3.15 3.78

0.521 0.171 0.449 0.714 0.117 0.02 1.34 1.75 1.86

Quality Factor

Trends of Ul

Ref.

1.28 1.36 1.61 1.32 1.31 0.50 0.84 1.8 2.03

U2 > U4 > U6 U2 > U4 > U6 U2 > U4 > U6 U2 > U4 > U6 U2 > U4 > U6 U2 > U6 > U4 U2 > U6 > U4 U2 > U4 > U6 U2 > U4 > U6

Present Present Present Present Present [38] [39] [40] [41]

Work Work Work Work Work

Table 7 P Total radiative transition probability (AT s1), branching ratios (bR %), relaxation rates ( AT S1) and relative lifetimes (tR ms) for Sm3þ:TNS glasses. Transition 4G5/2/

TNS2 AT

bR

AT

bR

AT

bR

AT

bR

AT

bR

6

H5/2 6 H7/2 6 H9/2 6 H11/2 P AT (S1)

30.19 78.59 47.94 43.29

0.151 0.393 0.239 0.217

26.67 66.80 41.16 36.72

0.156 0.391 0.241 0.215

26.32 61.34 38.98 33.36

0.165 0.384 0.244 0.209

24.55 61.85 38.10 33.97

0.155 0.391 0.241 0.215

27.30 69.42 42.47 38.07

0.154 0.392 0.239 0.214

200.03

171.37

159.99

158.49

177.28

tR (ms)

4.99

5.83

6.25

6.31

5.64

TNS3

TNS4

TNS5

TNS6

326

S.Q. Mawlud et al. / Optical Materials 69 (2017) 318e327

Table 8 Stimulated emission cross-section (sEp  1022cm2), gain bandwidth sEp  FWHM(  1028 cm2) and optical gain (sEp  tR  1025 cm2s1) of Sm3þ:TNS glasses under investigation. Transition 4

4

4

4

G5/2 / 6H5/2

sEp sEp sEp sEp sEp sEp sEp sEp sEp sEp sEp sEp

G5/2 / 6H7/2

G5/2 / 6H9/2

G5/2 / 6H11/2





bR jJ; j0 J 0 ¼

  A jJ; j0 J 0 AT ðjJÞ

TNS2

TNS3

TNS4

TNS5

TNS6

4.87

5.01

4.29

4.98

4.27

 FWHM

4.88

4.15

3.85

3.82

4.22

 tR

0.24

0.29

0.19

0.22

0.17 4.36

3.79

3.89

4.04

4.06

 FWHM

6.21

5.35

4.97

4.89

5.46

 tR

0.25

0.23

0.17

0.18

0.17 4.95

6.29

5.52

4.99

4.97

 FWHM

8.28

7.12

6.63

6.52

7.27

 tR

0.31

0.32

0.22

0.21

0.22 6.49

8.34

5.35

5.81

6.57

 FWHM

11.7

10.1

9.42

9.29

10.4

 tR

0.42

0.31

0.25

0.29

0.26

(22)

The relative values of the branching ratios can be obtained from the areas under the emission curves. The peak stimulated emission cross-section (sEp ) can be calculated using the expression:

sEp ¼

l4p A 8pcn2 Dleff

(23)

where lp is the emission transition peak wavelength and Dleff is the effective line width of the transition and is given by

Dleff ¼

1 Imax

Z IðlÞdl

(24)

where I is the fluorescence intensity and Imax is the intensity at band maximum. From the obtained J-O intensity parameter values, the other radiative parameters such as radiative transition probability (A), stimulated emission cross-section (sEp ) and branching ratios (bR ) for the excited levels of Sm3þ ions corresponding to the 4G5/ 6 4 6 4 6 4 6 2 / H5/2, G5/2 / H7/2, G5/2 / H9/2 and G5/2 / H11/2 transition shave also have been calculated referring earlier reports [19,25,39,40] and the results are presented in Table 7 and Table 8. The values of Amd depend on the matrix elements kL þ2Sǁ which are host independent. The levels having the relatively large values of Aed, Amd and bR and energy gap of more than few phonons may exhibit laser action [41]. From the values of radiative transition probabilities in Table 7, it is noticed that 4G5/2 / 6H7/2 transition has the highest radiative transition probabilities compared to other transitions; hence this transition is very useful for laser emission [41]. The predicated branching ratios are found to be high for those transitions having maximum radiative transition probabilities values. As seen from Table 7, the 6H7/2 transition exhibits higher bR values for the prepared glass systems. The luminescence branching ratio is a critical parameter to the laser designer because it characterizes the possibility of attaining stimulated emission from any specific transition. As expected, stimulated emission cross-section is very sensitive to the increase of Sm3þ concentration which confirms the presence of non-radiative excitation transfer from interaction between rare earth ions when concentration increases. These results show sEp is more sensitive to rare earth concentration rather than glassy composition. Stimulated emission cross-section sEp and FWHM are important parameters in an optical amplifier's

Fig. 14. Variation of stimulated emission cross-section sEp as a function of Sm2O3 concentration.

achieving broad band and high gain amplification. Bandwidth properties of the optical amplifier can be evaluated from the product sEp  FWHM. The results in Table 8 show that high amplification can be seen in glasses weakly doped with Sm3þ ions. It is clear from Table 8 that adding Sm3þ to tellurite glass enhances the value of sEp  FWHM. Thus the glass composition plays a key role in optical amplifiers. The variation of stimulated emission crosssection (sEp ) as a function of Sm2O3 concentration is shown in Fig. 14. The values of stimulated emission cross section for the 4G5/ 6 22 to 4.36  1022 cm2 2 / H7/2 transition increase from 3.79  10 and on the other hand those of the stimulated emission cross section for the 4G5/2 / 6H9/2 transition decrease from 6.29  1022 to 4.95  1022 cm2 with an increase in Sm2O3 concentration from 0.3 to 1.5 mol%.

4. Conclusion A Sm3þ:TNS glasses were obtained by conventional melt quenching technique. X-ray diffraction confirmed the amorphous nature. The J-O intensity parameters were calculated, as well as the radiative transition rates, branching ratios, radiative lifetimes and stimulated emission cross-sections. Multichannel transition emissions were observed in the Vis regions. The J-O parameters

S.Q. Mawlud et al. / Optical Materials 69 (2017) 318e327

evaluated from the measured absorption spectra of these glasses were used to calculate radiative properties of Sm3þ ions. The J-O intensity parameters values follows the trend U2>U4>U6 for all the TNS glass samples. The negative values of bonding parameter (d) and comparatively high values of U2 in the present glasses shows the ionic nature of SmeO bonds and more asymmetric nature of Sm3þ sites. The spectroscopic quality factor (U4 =U6 ) is in the range of 1.28e1.61 which shows that prepared TNS glasses have relatively more spectroscopic nature. The emission spectra recorded by monitoring the excitation at 404 nm using CW laser in the titled glasses gives four emission bands in visible region. Among them 4 G5/2 / 6H7/2 is more intense and 4G5/2 / 6H7/2 is moderate for all the glasses and are responsible for the distinct orangeered emission. This result is in consistent with the con-focal photoluminescence images recorded for all TNS glasses. Based on the emission spectra, high stimulated emission cross-sections and branching ratios observed for 4G5/2 / 6H7/2 transition for all these glasses suggest the feasibility of using these materials as visible lasers in orangeered (598 nm) visible region. From the measured stimulated emission cross-sections and branching ratios it was concluded that 1 mol% of Sm3þ ion concentration is optimum in TNS glasses for the development of orange-red visible lasers (598 nm) in principle. Acknowledgment The authors gratefully acknowledge the financial support from Ministry of Higher Education, RMC, UTM and University of Salahaddin/Ministry of Higher Education/KRG through the research grant (vote 4B182 and 03E23) are highly appreciated. References [1] Z.A.S. Mahraz, M.R. Sahar, S.K. Ghoshal, M.R. Dousti, R.J. Amjad, Silver nanoparticles enhanced luminescence of Er3þ ions in boro-tellurite glasses, Mater. Lett. 112 (2013) 136. [2] Y. Zhou, X.H. He, B. Yan, Self-assembled RE2(MO4)3: Ln3þ(RE¼ Y, La, Gd, Lu; M¼ W, Mo; Ln¼ Yb/Er, Yb/Tm) hierarchical microcrystals: hydrothermal synthesis and up-conversion luminescence, Opt. Mater. 36 (2014) 602. [3] S. Thomas, R. George, S.N. Rasool, M. Rathaiah, V. Venkatramu, C. Joseph, N.V. Unnikrishnan, Optical properties of Sm3þ ions in zinc potassium fluorophosphate glasses, Opt. Mater. 36 (2013) 242. [4] V. Thomas, R.G.S. Sofin, M. Allen, H. Thomas, P.R. Biju, G. Jose, N.V. Unnikrishnan, Optical analysis of samarium doped sodium bismuth silicate glass, Spectrochim. acta A 171 (2017) 144. [5] P. Babu, H.J. Seo, C.R. Kesavulu, K.H. Jang, C.K. Jayasankar, Thermal and optical properties of Er3þ-doped oxyfluorotellurite glasses, J. Lumine. 129 (2009) 444. [6] H. Nii, K. Ozaki, M. Herren, M. Morita, Up-conversion fluorescence of Er3þ and Yb3þ-doped TeO2-based oxide glass and single crystals, J. Lumine. 76 (1998) 116. [7] R.A. El-Mallawany, Tellurite Glasses Handbook: physical Properties and Data, CRC Press, 2011. [8] I. Pal, A. Agarwal, S. Sanghi, M.P. Aggarwal, Structural, absorption and fluorescence spectral analysis of Pr3þ ions doped zinc bismuth borate glasses, J. Alloy. Compd. 509 (2011) 7625. [9] M. Olivier, P. Pirasteh, J.L. Doualan, P. Camy, H. Lhermite, J.L. Adam, V. Nazabal, Pr3þ-doped ZBLA fluoride glasses for visible laser emission, Opt. Mater. 33 (2011) 980. [10] C.R. Kesavulu, C.K. Jayasankar, Spectroscopic properties of Sm3þ ions in lead fluorophosphate glasses, J. Lumine. 132 (2012) 2802. [11] M.C. Farries, P.R. Morkel, J.E. Townsend, Sm3þ-doped glass laser operating at 651 nm, Electron. Lett. 24 (1988) 709. [12] G. Turky, M. Dawy, Spectral and electrical properties of ternary (TeO2-V2O5-

327

Sm2O3) glasses, Mater. Chem. Phys. 77 (2003) 48. [13] B.R. Judd, Optical absorption intensities of rare-earth ions, Phys. Rev. 127 (1962) 750. [14] G.S. Ofelt, Intensities of crystal spectra of rare-earth ions, J. Chem. Phys. 37 (1962) 511. [15] M.R. Sahar, K. Sulhadi, M.S. Rohani, Spectroscopic studies of TeO2-ZnO-Er2O3 glass system, J. Mater. Scien. 42 (2007) 824. [16] S.A.M. Azmi, M.R. Sahar, S.K. Ghoshal, R. Arifin, Modification of structural and physical properties of samarium doped zinc phosphate glasses due to the inclusion of nickel oxide nanoparticles, J. Non-Cryst. Solid 411 (2015) 53. [17] N.M. Yusoff, M.R. Sahar, Effect of silver nanoparticles incorporated with samarium-doped magnesium tellurite glasses, Phys. B Condense Matt 456 (2015) 191. [18] S.k. Ahmmad, M.A. Samee, A. Edukondalu, S. Rahman, Physical and optical properties of zinc arsenic tellurite glasses, Results Phys. 2 (2012) 175. [19] G. Upender, S. Ramesh, M. Prasad, V.G. Sathe, V.C. Mouli, Optical band gap, glass transition temperature and structural studies of (100-2x) TeO2-xAg2OxWO3 glass system, J. Alloy. Compd. 504 (2010) 468. [20] P.G. Pavani, K. Sadhana, V.C. Mouli, Optical, physical and structural studies of boro-zinc tellurite glasses, Phys. B Condens Matt 406 (2011) 1242. [21] B. Eraiah, Optical properties of lead-tellurite glasses doped with samarium trioxide, Bull. Mater. Sci. 33 (2010) 391. [22] N. Jaba, A. Mermet, E. Duval, B. Champagnon, Raman spectroscopy studies of Er3þ-doped zinc tellurite glasses, J. Non-Crysta Solids 351 (2005) 833. [23] V.G. Plotnichenko, V.O. Sokolov, V.V. Koltashev, E.M. Dianov, I.A. Grishin, M.F. Churbanov, Raman band intensities of tellurite glasses, Opt. Lette 30 (2005) 1156. [24] N.N. Aye, Study of xP2O5-(1-x-y)V2O5-yCuO semi-conducting glass system, Univ. Res. J. 4 (2011) 39. [25] S.H. Kim, T. Yoko, Nonlinear optical properties of TeO2-based glasses: MOxTeO2 (M¼ Sc, Ti, V, Nb, Mo, Ta, and W) binary glasses, J. Am. Ceram. Soc. 78 (1995) 1061. [26] V.A.G. Rivera, D. Manzani, Y. Messaddeq, L.A.O. Nunes, E. Marega Jr., Study of Er3þ fluorescence on tellurite glasses containing Ag nanoparticles, J. Phys. Conf. Ser. 274 (2011) 012123. [27] H.A.A. Sidek, S. Rosmawati, Z.A. Talib, M.K. Halimah, W.M. Daud, Synthesis and optical properties of ZnO-TeO2 glass system, Am. J. Appl. Sci. 6 (2009) 1489. [28] P. Chimalawong, J. Kaewkhao, C. Kedkaew, P. Limsuwan, Optical and electronic polarizability investigation of Nd3þ-doped soda-lime silicate glasses, J. Phys. Chemis. Solid 71 (2010) 965. [29] E.A. Davis, N. Mott, Conduction in non-crystalline systems V. Conductivity, optical absorption and photoconductivity in amorphous semiconductors, Philos. Mag. 22 (1970) 0903. [30] F. Urbach, The long-wavelength edge of photographic sensitivity and of the electronic absorption of solids, Phys. Rev. 92 (1953) 1324. [31] M.K. Halimah, W.M. Daud, H.A.A. Sidek, A.W. Zaidan, A.S. Zainal, Optical properties of ternary tellurite glasses, Mater. Sci. Pol. 28 (2010) 173. [32] S.P. Sinha, Complexes of the Rare Earths, Elsevier, 2013. [33] W.T. Carnall, P.R. Fields, K. Rajnak, Electronic energy levels in the trivalent lanthanide aquo ions. I. Pr3þ, Nd3þ, Pm3þ, Sm3þ, Dy3þ, Ho3þ, Er3þ, and Tm3þ, J. Chem. Phys. 49 (1968) 4424. [34] C.K. Jayasankar, P. Babu, Optical properties of Sm3þ ions in lithium borate and lithium fluoroborate glasses, J. Alloy. Compd. 307 (2000) 82. [35] R. Praveena, V. Venkatramu, P. Babu, C.K. Jayasankar, Fluorescence spectroscopy of Sm3þ ions in P2O5ePbOeNb2O5 glasses, Phys. B Condens. Matt 403 (2008) 3527. [36] G. Lakshminarayana, J. Qiu, Photoluminescence of Pr3þ, Sm3þ and Dy3þ: SiO2Al2O3-LiF-GdF3 glass ceramics and Sm3þ, Dy3þ: GeO2-B2O3-ZnO-LaF3 glasses, Phys. B Condens. Matt 404 (2009) 1169. [37] G.H. Jia, C.Y. Tu, J.F. Li, Z.J. Zhu, Z.Y. You, Y. Wang, B.C. Wu, Spectroscopy of GdAl3(BO3)4:Tm3þ crystal, J. Appl. Phys. 96 (2004). [38] W. Guo, Y. Chen, Y. Lin, Z. Luo, X. Gong, Y. Huang, Spectroscopic properties and laser performance of Tm3þ-doped NaLa (MoO4)2 crystal, J. Appl. Phys. 103 (2008) 3106. [39] I. Jlassi, H. Elhouichet, M. Ferid, C. Barthou, JuddeOfelt analysis and improvement of thermal and optical properties of tellurite glasses by adding P2O5, J. Lumine. 130 (2010) 2394. [40] A. Ghosh, R. Debnath, JuddeOfelt analysis of Er3þ activated lead free fluoroetellurite glass, Opt. Mater. 31 (2009) 604. [41] V. Venkatramu, P. Babu, C.K. Jayasankar, Th Trȍster, W. Sievers, G. Wortmann, Optical spectroscopy of Sm3þ ions in phosphate and fluorophosphate glasses, Opt. Mater. 29 (2007) 1429.