Photoluminescence and spectral performance of manganese ions in zinc phosphate and barium phosphate host glasses

Journal of Non-Crystalline Solids 458 (2017) 1–14

Contents lists available at ScienceDirect

Journal of Non-Crystalline Solids journal homepage: www.elsevier.com/locate/jnoncrysol

Photoluminescence and spectral performance of manganese ions in zinc phosphate and barium phosphate host glasses M.A. Marzouk a,⁎, Y.M. Hamdy b, H.A. Elbatal a a b

Glass Research Department, National Research Centre, 33 EL Bohouth St. (former EL Tahrir St.), Dokki, Giza, P.O. 12622, Egypt Spectroscopy Department, National Research Centre, 33 El Bohouth St. (former EL Tahrir), Dokki, Giza, P.O.12622, Egypt

a r t i c l e

i n f o

Article history: Received 4 October 2016 Received in revised form 29 November 2016 Accepted 6 December 2016 Available online xxxx Keywords: Zinc phosphate Barium phosphate Glass MnO2 Photoluminescence

a b s t r a c t Zinc and barium metaphosphate glasses containing dopants of MnO2 (2, 4, 8%) were prepared. Collective studies were carried out through optical, photoluminescence and Fourier transform infrared (FTIR) spectral measurements for the prepared glasses before and after gamma irradiation. Electron spin resonance (ESR), thermal expansion and crystallization studies were also investigated. Optical spectra for undoped and MnO2–glasses reveal strong ultraviolet (UV) absorption due to trace iron impurity in the materials used for preparation of glasses and Mn2+ ions are not identified due to spin–forbidden nature of their spectra. Upon gamma irradiation, the MnO2–doped glasses reveal extension of the UV absorption exhibiting two peaks together with the generation of an induced broad visible band centered at 570–580 nm. Extra two glasses doped with 8% MnO2 were melted under reducing condition and they showed characteristic band at 410 nm related to Mn2+ ions absorption. Photoluminescence (PL) spectra reveal two excitation peaks in MnO2–doped zinc phosphate glasses at 348 and a sharp peak at 410 nm while the MnO2–doped barium phosphate glasses exhibit a sharp peak at 410 nm and a curvature at 348 nm. The emission spectra show a broad band centered at 620 nm in both glasses but with a curvature at 605–610 nm in zinc phosphate glasses which are correlated with the transition 4T1g → 6A1g.The efficiency of the photoluminescence increases with the MnO2 content and the heat treatment. The values of the optical band gap (Eopt) and Urbach energy (ΔE) were calculated. FTIR spectra reveal vibrational bands due to phosphate network mainly of metaphosphate groups but in zinc phosphate glasses, the mid spectra show compact network because of the ability of Zn2+ ions to participate partly as ZnO4 or P–O–Zn linkages. Thermal expansion data show different responses. Zinc phosphate glasses show anomalous behavior related to the change in the coordination of zinc cations while barium phosphate glasses exhibit normal expansion because Ba2+ ions exist only as modifiers. The crystallization behavior reflects also anomalous behavior. The zinc phosphate glass crystallizes in two crystalline species, mainly zinc metaphosphate phase and another zinc phosphate (3ZnO·P2O5) phase while the barium phosphate glass crystallizes only in barium metaphosphate phase. © 2016 Published by Elsevier B.V.

1. Introduction It has been established that the introduction of any divalent, trivalent or tetravalent oxides to phosphate glasses increases their chemical durability while generally keeping their relatively lower preparation temperatures and their excellent and unique optical, thermal and electrical properties [1–3]. The addition of divalent oxides to P2O5 causes the breaking of the tetrahedral building PO4 blocks and the conversion of the phosphate network to a system of entangled linear chains of phosphorus–oxygen tetrahedral cross linked by divalent ions [4]. Some of the alkaline earth oxides (CaO, SrO, BaO) when added to P2O5 result in the creation of non-bridging oxygens at the expense of bridging oxygens and the mentioned alkaline earth cations are solely acting as ⁎ Corresponding author. E-mail address: [email protected] (M.A. Marzouk).

http://dx.doi.org/10.1016/j.jnoncrysol.2016.12.013 0022-3093/© 2016 Published by Elsevier B.V.

modifiers [3]. Other divalent oxides (e.g. ZnO, MgO) combine with P2O5 – and the two binary systems of ZnO–P2O5 and MgO–P2O5 have been classified as anomalous because of discontinuities in composition–property trends near the metaphosphate (50 mol% P2O5) composition [5,6]. Such anomalous behavior can be related to several reasons including the change in the coordination number of Zn2+ or Mg2+ or the ability of their oxides to form partly structural building units (ZnO4, MgO4) or to other factors such as relatively high field strength of those divalent cations or to the change in the polarizability of the neighboring oxygen ions [5,6]. Manganese as one of 3d transition metals can exist in glasses in two valences, the divalent and trivalent states and the ratio of each depends on the glass type and composition and condition of melting [7]. The trivalent manganese ions exhibit a broad nearly symmetrical visible band at 450–540 nm which is related to a spin–allowed 5Eg → 5T2g transition of octahedral symmetry [8,9]. The divalent manganese ions are known

2

M.A. Marzouk et al. / Journal of Non-Crystalline Solids 458 (2017) 1–14

to exhibit four weak spin–forbidden peaks which vary in their positions with their symmetry of either octahedral or tetrahedral coordination. Both Mn3+ and Mn2+ ions are well–known paramagnetic ions, Mn3+ ion has a large anisotropy due to its spin–orbit interaction of the 3d orbital whereas such anisotropy energy of Mn2+ ion is small because its orbital angular momentum is zero [7,10]. When glasses are subjected to ionizing gamma irradiation (such as γ–rays, UV radiation), some of the glasses suffer darkening or the formation of induced color centers which can be identified by optical and/or E.S.R measurements [11,12]. It is accepted that the glasses containing some transition metal ions (e.g. Cu2 +, V5 +) exhibit some shielding effects and their optical spectral curves remain unchanged and closely parallel upon successive gamma irradiation [13,14]. Also, some glasses containing heavy metal oxides (e.g. PbO, Bi2O3, BaO) show the same shielding behavior upon gamma irradiation because of the heavy masses of their cations which retard the free motion of excited electrons or positive holes during the irradiation process [15,16]. The main objective of this present work is to characterize and compare collective optical, photoluminescence, FTIR spectral analysis together with ESR and thermal properties of manganese–doped within both zinc meta–phosphate and barium meta–phosphate glasses. The present study also includes the investigation of the effects of gamma irradiation on the combined optical, photoluminescence and FTIR spectral properties of the two glass systems containing varying added dopants (2, 4, 8 wt%) of manganese. Two further selected glasses (8% MnO2) were prepared at reducing condition and their main spectral properties were measured and characterized. A final study in this work is focused on the crystallization behavior of some selected zinc phosphate and barium phosphate glasses upon controlled thermal two step heat treatment regime.

2.2. Preparation of the corresponding glass–ceramic derivatives The heat treatment temperatures are based on differential thermal analysis (DTA) measurements (Fig. 1) using SDTQ600 under N2 gas. Glass samples were subjected to controlled thermal heat-treatment through a two-step regime (Table 2). The glasses were first heat treated slowly at a rate of 5 °C/min to reach 450 °C and kept at this temperature for 12 h which was sufficient to provide sufficient nucleation sites. The muffle was then raised to reach 600 °C and kept at this second hold temperature for 6 h. The muffle furnace with the heat-treated samples inside was switched off and then left to cool to room temperature at a rate of 30 °C/h. 2.3. X-ray diffraction analysis Identification of the crystalline phases formed during controlled heat-treatment was performed by X-ray diffraction analysis using a Bruker AXS diffractometer (CD8-ADVANCE) with CuKa radiation, operating at 40 kV and 10 mA. The diffraction data were recorded for 2θ values between 4° and 70° and the scanning rate was 10°/min. 2.4. Thermal expansion measurements The thermal expansion characteristics of the glasses were measured through specified samples using a recording dilatometer (type NETZCH – 1–402 PC Geräebau GmbH, Selb Germany) with a heating rate of 10 °C/min up to the dilatometric softening temperature. The thermal data for the transformation and softening temperatures were collected to be used for controlled thermal heat treatment of the glasses to their corresponding glass–ceramic derivatives.

2. Materials and methods

2.5. Optical absorption measurements

2.1. Preparation of the glasses

Optical (UV–visible) absorption measurements were carried out on polished samples (2 mm ±0.1 mm) using a recording spectrophotometer (type JASCO V-570, Japan) covering the range from 200 to 1100 nm.

The glasses were prepared from pure laboratory chemicals. The materials include ammonium dihydrogen ortho phosphate (NH4H2PO4), 99% Winlab Ltd., Harborough, Leicestershire, UK) was the source of P2O5, zinc oxide was introduced as ZnO (BDH, Laboratory reagent, England) and anhydrous heavy barium carbonate (BaCO3) (MERCK G. Darmstadt, Germany, Fe 0.001%) for BaO and MnO 2 (Fluka, Buchs, LUKA – Switzerland, Fe 0.3%) was added as such. The accurately weighed batches in Table 1 were melted under atmospheric condition in alumina crucibles at 1100 °C for 90 min in SiC heated furnace (Vecstar, Chesterfield, UK). The melts were rotated at intervals to reach complete mixing and homogeneity. The homogenized melts were poured into preheated stainless steel molds of the required dimensions. The prepared glassy samples were immediately transferred to an annealing muffle regulated at 300 °C to obtain glasses free from stresses or strains. The muffle was switched after 1 h with the samples inside and left to cool to room temperature at a rate of 30 °C/h. Table 1 depicts the chemical compositions of the prepared glasses.

2.6. Photoluminescence measurements Photoluminescence measurements were recorded at room temperature under the excitation wavelength of 410 nm in the spectral range 500–750 nm using a fluorescence spectrophotometer (type JASCO, FP -6500, Japan) equipped with a xenon flash lamp as the excitation light source. The scan speed is 0.15 step − 1 with a step length of 0.25 nm and slit width 0.2 nm. 2.7. Calculations of optical band gap (Eopt), Urbach energy (ΔE) and refractive index Mott and Davis [17] model has been applied to determine the optical band gap energies (Eopt) of the prepared glasses using Eq. (1) , by

Table 1 Chemical composition in mol% and optical parameters of the prepared glasses. Sample no.

(1) (2) (3) (4) (5) (6) (7) (8)

P2O5

50 50 50 50 50 50 50 50

ZnO

50 50 50 50 – – – –

BaO

– – – – 50 50 50 50

ΔE (eV)

MnO2

Eopt (eV)

Added in wt%

0 MR (±0.05)

10 MR (±0.031)

0 MR (±0.035)

10 MR (±0.052)

n 0 MR (±0.011)

10 MR (±0.007)

– 2 4 8 – 2 4 8

4.149 4.100 4.048 4.018 3.920 3.849 3.797 3.614

3.199 3.160 3.137 3.088 3.385 3.250 3.190 3.103

0.405 0.570 0.625 0.675 0.346 0.485 0.496 0.519

0.743 0.621 0.567 0.547 0.960 0.683 0.547 0.436

2.141 2.150 2.160 2.166 2.185 2.199 2.210 2.248

2.345 2.355 2.361 2.373 2.300 2.332 2.347 2.369

M.A. Marzouk et al. / Journal of Non-Crystalline Solids 458 (2017) 1–14

3

plots, the value of the Urbach energy ΔE can be calculated from the reciprocal of the slope of the linear region of such plots. The refractive index, n, can be calculated from the value of Eopt using the formula proposed by Dimitrov and Sakka [19] as follows: rffiffiffiffiffiffiffiffi n2 −1 Eopt ¼ 1− 20 n2 þ 2

ð3Þ

2.8. Fourier transform infrared absorption measurements (FTIR) FT infrared absorption spectra of the prepared glasses and glass-ceramics were measured at room temperature in the wave number range 400–4000 cm−1 by a Fourier transform computerized infrared spectrometer type (FTIR 4600 JASCO Corp Japan)using the KBr disc technique. The glasses were examined in the form of pulverized powder which was mixed with KBr with the ratio 1:100 mg glass powder to KBr, respectively. The weighed mixtures were then subjected to a pressure 5 tons/cm2 to produce clear homogeneous discs. FTIR measurements were also carried out for the glass-ceramics in the powder form. The same measurements were repeated for the glasses after gamma irradiation (10 Mrad). 2.9. E.S.R spectral measurements Electron spin resonance spectra for some selected MnO2–doped glasses as powder were recorded at room temperature on an ESR spectrometer (Bruker, E500, Germany) operating at 9.808 GHz and using 100 kHz field modulation. The magnetic field was scanned from 480 to 6840 G. 2.10. Irradiation facility An Indian 60Co gamma cell (2000 Ci) was used as a gamma ray source with a dose rate of 1.5 Gy/s (150 rads/s) at 30 °C. The glass samples were placed in the gamma cell in the manner that each sample was subjected to the same gamma irradiation dose (10 Mrad = 10 × 104 Gy). Fig. 1. DTA of undoped and MnO2 doped (a) zinc phosphate (b) barium phosphate glasses.

plotting (αhν) n as a function of photon energy hν, one can find the optical energy band gaps (Eopt)
n αhv ¼ B hv−Eopt

ð1Þ

where Eopt is the optical energy gap, B is a constant called the band tailing parameter and n is an index which can be assumed to have values of 1/2, 3/2, 2 and 3, depending on the nature of the electronic transition responsible for the absorption. The Urbach energy can be calculated from the following Eq. (2) [18].   hν α ðνÞ ¼ B exp ΔE

ð2Þ

where B is a constant, ΔE is the Urbach energy and v is the frequency of radiation. Plots were also drawn between ln(α) and hν and from these

3. Results 3.1. Optical absorption spectra of the studied glasses 3.1.1. Optical absorption spectra of undoped and MnO2–doped zinc phosphate glasses before and after gamma irradiation Fig. 2(a) illustrates the UV–visible optical absorption spectra of undoped and MnO2–doped zinc phosphate glasses before and after gamma irradiation. The spectrum of the undoped zinc phosphate glass reveals only strong UV absorption with a peak at 210 nm and without any further absorption to the end of the measurement at 1100 nm. On introducing MnO2 in increasing contents (2, 4, 8%) the UV peak increases in intensity and shifts slightly in position reaching about 220 nm and without any further observed absorption to the rest of measurement at 1100 nm. Fig. 2(a) shows the optical spectra of the studied undoped and MnO2–doped zinc phosphate glasses after gamma irradiation. The

Table 2 Heat treatment Scheme with two-step regime. Glass

Base binary glasses Sample 2% MnO2 Sample 4% MnO2 Sample 8% MnO2

First nucleation temperature (°C)

Holding time (hours)

Second crystal growth temperature (°C)

Holding time (hours)

ZnP

BaP

ZnP

BaP

ZnP

BaP

ZnP

BaP

450 450 450 450

450 450 450 450

12 12 12 12

12 12 12 12

600 600 600 600

600 600 600 600

6 6 6 6

6 6 6 6

4

M.A. Marzouk et al. / Journal of Non-Crystalline Solids 458 (2017) 1–14

3.1.3. Optical absorption spectra MnO2–doped barium and zinc phosphate glasses prepared under reducing conditions Fig. 3(a & b) illustrates the optical spectra of the two selected zinc and barium phosphate glasses containing dopant MnO2 (8%) melted under reducing conditions. The optical spectra of the two glasses reveal strong and broad UV absorption with peak at 240–242 nm and followed by a small peak at 380 nm and a small distinct peak at 407–410 nm and with no further absorption to the rest of measurements. After irradiation the optical spectrum of the reduced zinc phosphate glass (8% MnO2) show the extension of the UV absorption with the resolution of two peaks at 242 and 335 nm and with more distinction of the band at 410 nm together with the generation of an induced visible broad band centered at 565 nm. The spectrum of the irradiated reduced barium phosphate glass reveals the extension of UV absorption with two peaks at 240 and 348 nm together with appearance of the band at 407 nm and the induced visible band centered at 452 nm.

3.2. Photoluminescence spectra of the studied glasses Fig. 4(a) illustrates the PL spectra of undoped and MnO2 doped zinc phosphate glasses before gamma irradiation. The PL spectra before irradiation reveal two sharp excitation peaks at 348 and 410 nm while the emission spectra show a very broad emission band with a small peak at 605 and a broad sharp peak at 620 nm. The PL data show an obvious increase of the intensity of the peaks with the increase of MnO2 content in the glass together with a slight shift to higher wavelength.

Fig. 2. UV–visible absorption spectra before and after 10 Mrad γ-irradiation of undoped and MnO2–doped (a) zinc phosphate (b) barium phosphate glasses.

optical spectrum of the irradiated undoped glass reveals strong and broader sharp UV absorption with a peak at 240 nm and with no further absorption. The MnO 2 –doped samples after irradiation show two strong UV absorption bands at 240 and 300 nm together with the generation of an induced broad visible band centered at 580 nm.

3.1.2. Optical absorption spectra of undoped and MnO2–doped barium phosphate glasses before and after irradiation Fig. 2(b) reveals the optical (UV–visible) absorption spectra of undoped and MnO2–doped barium phosphate glasses before and after gamma irradiation. The optical spectrum of the undoped barium phosphate glass shows distinct UV absorption band with a peak at about 215 nm and with no further absorption to the rest of measurement. The MnO2–doped barium phosphate glasses reveal shifting of the UV peak accompanying the broadness of the absorption to reach 240 nm at the highest MnO2 content and no visible absorption bands are identified to the rest of measurement. Fig. 2(b) also shows the optical spectra of barium phosphate glasses after gamma irradiation. The optical spectra of the irradiated glasses reveal the extension of the UV absorption with the appearance of two peaks at 240 and 295 nm with the introduction of MnO2 beside the generation of an induced medium visible band centered at 570 nm.

Fig. 3. UV–visible absorption spectra before and after 10 Mrad γ-irradiation of the reduced 8% MnO2–doped (a) zinc phosphate (b) barium phosphate glasses.

M.A. Marzouk et al. / Journal of Non-Crystalline Solids 458 (2017) 1–14

Fig. 4. PL spectra of MnO2–doped zinc phosphate glasses (a) before and (b) after 10 Mrad γ-irradiation.

Fig. 4(b) illustrates the PL spectra of the undoped and MnO2–doped zinc phosphate glasses after gamma irradiation (dose 10 Mrad). It is obvious that only two pronounced changes are identified after irradiation. The first change is the decrease of the intensity of the peaks in both the excitation and emission. The second change is the almost disappearance of the excitation peak or curvature at 348 nm. Fig. 5(a) illustrates the PL spectra of undoped and MnO2–doped barium phosphate glasses before gamma irradiation. The PL spectra reveal a sharp peak at 410 nm and a curvature peak at 348 nm and the emission spectra shows a very broad strong band with its peak at 620 nm. It is observed that the intensity of all the peaks progressively increases with the MnO2 content. Fig. 5(b) shows the PL spectra of the undoped and MnO2–doped barium phosphate glasses after gamma irradiation (dose 10 Mrad). It is obvious that the same PL spectra before irradiation remain with the same features with only one variation. The increase of intensity of the peaks is observed to be sharp at first on increasing the concentration from 2% to 4% and then attains almost constancy or approaches saturation at 8% MnO2. Photoluminescence spectra of reduced manganese–doped (8 mol%) zinc phosphate and barium phosphate glasses shown in Fig. 6(a & b) prepared with reducing agent (sugar) as carried out by Konidakis et al. [20] reveal the following results: (a) The zinc phosphate glass (reduced 8% Mn) shows an excitation spectrum consisting of three peaks at 350, 410 and 516 nm (the first one is two split, the second is very sharp, and the

5

Fig. 5. PL spectra of MnO2–doped barium phosphate glasses (a) before and (b) after 10 Mrad γ-irradiation.

third is broad but with medium intensity). The emission spectrum reveals a very broad highly distinct peak at 610 nm. (b) The barium phosphate glass (reduced 8% Mn) reveals an excitation peaks at 354, 410 and 538 nm. The emission spectrum shows a very broad band with a peak at 632 nm. After irradiation, the photoluminescence spectra of the two glasses are highly affected and reveal marked decrease of the excitation and emission peaks.

3.3. Photoluminescence spectra of the glass-ceramics Fig. 7(a & b) reveals the PL spectra of the heat treated glass-ceramic derivatives. The PL spectra of the zinc phosphate glass ceramics show distinct excitation peaks at 350, 410, 445 and 495 nm and the emission spectra show a distinct broad peak at 560 nm. It is evident that the intensity of all the peaks decreases sharply from 2% to 4% MnO2 and then with moderated decrease on increasing the MnO2 to 8%. The PL spectra of the barium phosphate glass–ceramics show two excitation peaks at 354 and 410 nm and the emission spectra reveal one very broad band centered at 625 nm. On increasing the MnO2 content from 2 to 4% MnO2, the intensity of the excitation and emission peaks slightly decreases but on further increase to 8% MnO2, the intensity sharply decreases.

6

M.A. Marzouk et al. / Journal of Non-Crystalline Solids 458 (2017) 1–14

Fig. 6. PL spectra before and after 10 Mrad γ-irradiation of reduced 8% MnO2–doped (a) zinc phosphate (b) barium phosphate glasses.

Fig. 7. PL spectra of the heat treated MnO2–doped (a) zinc phosphate (b) barium phosphate glasses.

3.4. Optical parameters, band gap (Eopt), Urbach energy (ΔE) and refractive index (n) 3.5. FT infrared absorption spectra The values of the optical band gap (Eopt) are obtained by extrapolating from the linear region of the plots of the quantity (αhν)1/2 against the photon energy (hν) as shown in Fig. 8(a & b) and listed in Table 1. Before irradiation, it is clear from Table 1 that the values of optical band gap for undoped and MnO2 doped zinc phosphate or barium phosphate glasses decrease with increasing MnO2 content. The value of Eopt is a maximum for zero MnO2 content. The observed optical band gap value is 4.149 eV for undoped zinc phosphate glasses while it is 3.920 eV for undoped barium phosphate glasses. The introduction of MnO2 causes a shift in absorption edge towards lower energies. The observed values of Eopt lie in the range of 4.1–4.018 eV for MnO2 doped zinc phosphate glasses and 3.849–3.614 eV for MnO2 doped barium phosphate glasses. After 10 Mrad gamma irradiation, an observed slight decrease in the values of optical band gap from 3.199 to 3.088 eV for undoped and MnO2 zinc phosphate glasses and from 3.385 to 3.103 eV for undoped and MnO2 doped barium phosphate glasses. As shown in Table 1, the Urbach energy (ΔE) values before irradiation are observed to increase with increasing the amount of MnO2 for both barium and zinc phosphate glasses. An opposite behavior is observed after 10 Mrad gamma irradiation. Δ E values increase with the irradiation dose and with MnO2 content. It is noticed that the results of the calculated values of refractive index slightly increase with increasing the MnO2 percent from 2 up to 8%.

Fig. 9 shows the FTIR spectra of the undoped zinc phosphate glass and MnO2–doped samples before and after gamma irradiation. The IR spectrum of the undoped zinc phosphate glass reveals the following IR spectral details: (a) The appearance of a distinct and broad far–IR band extending from 400 to about 670 cm−1 with a peak at 480 cm−1. (b) A medium band is identified at about 746 cm−1. (c) A very broad and connected or composite band is highly identified to be extending from about 800 to 1500 cm− 1 with four peaks at about 900, 1100, 1250 and 1375 cm−1. (d) A medium band is observed with a peak at about 1640 cm−1. The IR spectra of MnO2–doped zinc phosphate glasses Fig. 9(a) show somewhat similar IR absorption bands but with limited modifications specifically in the very broad wide connected band. The vibrational peak at about 1375 cm−1 is almost disappeared and the highest peak at about 1120 cm−1 remains to be the prominent peak in all the measured IR spectra. It is found necessary and profitable to make a deconvolution to the IR spectrum of the base undoped glass to identify all the individual possible vibrational modes which are connected or superimposed. The deconvoluted vibrational peaks identified in the IR spectrum of the

M.A. Marzouk et al. / Journal of Non-Crystalline Solids 458 (2017) 1–14

7

(i) With 2% MnO2, the far–IR band decreases its sharpness and with the appearance of another absorption band at about 420 cm−1. Also, a small broad band is identified at about 1440 cm−1. (ii) With 4 and 8% MnO2, both the band at 736 cm−1 and the band at 950 cm−1 increase in intensity and the second band approaches to the intensity of the band at 1130 cm−1. A further additional strong sharp band is identified at about 1350 cm−1. After gamma irradiation (Fig. 10b), the FTIR spectra show distinct variations as can be summarized as follows: (a) The IR spectrum of the undoped glass shows the connection of the mid region from about 800 to about 1800 cm−1 but the intensity of the broad far–IR band becomes higher than the intensity of the connected very broad band. (b) The IR spectrum of the 2% MnO2–doped sample shows the growth of the two bands at about 1280 and 1400 cm−1. (c) The IR spectra of the 4 and 8% MnO2 samples reveal the decrease of the intensity of the peak at about 910 cm−1 and the increase of the band at about 1630 cm−1. 3.6. Electron spin resonance data Fig. 11 shows the E.S.R. spectra of representative 2% MnO2 doped samples of both zinc and barium phosphate glasses. The ESR data indicate that the two MnO2–doped glasses of zinc phosphate and barium phosphate glasses reveal specific characteristic signals nearly on the same place but with relatively different intensities. Also, it is obvious that upon gamma irradiation, the sample of MnO2 doped zinc phosphate is slightly affected while with the MnO2–doped barium phosphate, the signal shows obvious high response or change with irradiation. 3.7. Thermal expansion data Fig. 8. Optical band gap of undoped and MnO2–doped zinc phosphate barium phosphate glasses (a) before and (b) after 10 Mrad γ-irradiation.

undoped zinc phosphate glass are as follows: 414, 521, 746, 884, 959, 1121, 1259, 1376 and 1641 cm−1. Fig. 9(b) reveals the FTIR spectra after gamma irradiation with a dose of 10 Mrad. The changes in the IR spectra upon gamma irradiation are summarized as follows: (i) The IR spectrum of the undoped glass shows obvious increase of the intensities of the two peaks at about 950 and 1275 cm−1. (ii) The IR spectra of the MnO2 doped glasses reveal the disappearance of the peak at about 1250 cm−1. Fig. 10 illustrates the FTIR spectra of undoped and MnO2–doped barium phosphate glasses before (a) and after gamma irradiation (b). Before irradiation (Fig. 10a) the IR spectrum of the undoped binary barium phosphate glass shows the following spectral features: (a) The appearance of a broad and strong far–IR band extending from 400 to about 680 cm−1 and centered at about 500 cm−1. (b) A second broad band is identified at 736 cm−1. (c) A very broad and highly distinguished band is observed extending from about 800 to 1500 cm−1 with four peaks at 909, 1016, 1121 and 1229 cm−1. (d) A medium sharp band is observed with a peak at 1634 cm−1. The glasses doped with MnO2 reveal distinct variations with the increase of the MnO2 content as can be summarized as follows:

Fig. 12(a) shows the thermal expansion of the undoped and selected MnO 2 –doped (4, 8%) zinc phosphate glasses. The thermal expansion data reveal parallel curves behavior consisting of the first part from room temperature up to about 130 °C in which the thermal expansion coefficient of the glass decreases with temperature to that specified temperature. After that, the thermal expansion coefficient progressively increases with temperature but with an obvious high rate up to 300 °C and then becomes slower and above 400 °C sharply increases to reach the dilatometric softening point after which the thermal curve drops sharply. It is obvious that the dilatometric softening temperature decreases with the increase of MnO 2 from 420, 410, 405 °C. The appearance of a reverse variation in the thermal expansion coefficient with temperature in this specific glass system confirms the assumption that the studied ZnO–P 2 O 5 of the metaphosphate composition (50 mol of each) glass shows obvious anomalous behavior. Fig. 12(b) reveals the thermal expansion curves of the undoped and selected MnO2–doped (4%, 8%) – barium phosphate glasses. The thermal expansion data show different behavior than that for zinc phosphate samples. The thermal expansion at first increases with lower rate from room temperature up to about 150 °C, then increases with faster rate until reaching about 430 °C, after which the thermal expansion coefficient sharply increases up to the dilatometric softening temperature and then drops sharply. The dilatometric softening temperature decreases progressively with MnO2 content (453 °C → 445 °C). 3.8. X-ray diffraction data (crystallization of the studied glasses) Fig. (13) illustrates the x-ray diffraction patterns of the heat treated samples of zinc phosphate and barium phosphate. The x-ray data

8

M.A. Marzouk et al. / Journal of Non-Crystalline Solids 458 (2017) 1–14

Fig. 9. FTIR absorption spectra of undoped and MnO2–zinc phosphate glasses (a) before and (b) after 10 Mrad γ-irradiation.

indicate that heat-treatment produces two crystalline phases of zinc metaphosphate (ZnO·P2O5) or Zn(PO3)2 with 61.4% card No. 010587(D) and the second phase of zinc phosphate of composition (3ZnO·P2O5 or Zn3(PO4)2) card No. 76-0604 (C). The barium phosphate glass is identified to crystallize only in one phase of barium metaphosphate BaO·P2O5 (or Ba(PO3)2) card No. 72-0374 (C). The detailed crystallization data for the two glass systems after the introduction of increasing MnO2 contents as shown in X-ray diffraction patterns (Fig. 13) data are as follows: (i) When 2% MnO 2 is added to zinc phosphate glass system, the crystalline phases are zinc meta-phosphate (ZnO-P 2 O 5 ) with 70.4% beside the appearance of a new crystalline phase of manganese phosphate hydrate MnPO4 (H2 O) with 29.6%. (ii) When 8% MnO2 is added to the ZnO·P2O5 glass system, three crystalline phases are identified by x-ray diffraction analysis: a major zinc meta-phosphate phase with 61.5% card No. 30-1488 (N) and a second zinc meta-phosphate hexagonal phase card No. 01-0875 (D) with 19.0% and the third is manganese phosphate hydrate phase MnPO4(H2O) with 19.4%. (iii) When 2% MnO2 is added to the BaO–P2O5 glass, the identified crystalline phase after heat treatment is the main barium metaphosphate phase card No. 2-0374 with 80.2% beside an additional new barium manganese phosphate phase (BaMn(P2O7)) with 19.8%. (iv) When the added MnO2 reaches 8% to the BaO·P2O5 glass, the identified crystalline phases after heat treatment are two, the main phase is barium meta-phosphate (Ba(P2O5)) with 73.0%

and a secondary phase of barium manganese phosphate (BaMn(P2O7)) with 27.0%.

4. Discussion 4.1. Interpretation of the optical spectra of the undoped and MnO2–doped glasses 4.1.1. Interpretation of the optical spectra of undoped glasses The experimental optical spectra of both the undoped zinc phosphate and undoped barium phosphate glasses reveal two strong UV absorption bands at 210 or 215 nm in the two respective studied glasses. Some scientists [21,22] had identified strong UV absorption in various undoped commercial glasses and this UV absorption was related to unavoidable contaminated trace impurities (mainly iron ions) within the raw materials used in the preparation of such commercial glasses or during the contact of molten commercial glasses with building furnace refractories during industrial melting of glass. Duffy [23] and Ehrt et al. [24,25] have recognized strong charge transfer UV absorption in glasses and have assumed that the presence of trace metal ion impurities (such as Fe3 +, Cr6 +) could impair the transmission properties of the glasses even if present in the ppm level. They have attributed such UV absorption to an electron transfer mechanism involving the transition of an electron from the orbital of a coordinated oxygen atom to an orbital of the metal ion. In subsequent studies, the UV absorption bands identified in various undoped phosphate glasses are related to trace ferric ions present as impurities in glasses [26,27].

M.A. Marzouk et al. / Journal of Non-Crystalline Solids 458 (2017) 1–14

9

Fig. 10. FTIR absorption spectra of undoped and MnO2–doped barium phosphate glasses (a) before and (b) after 10 Mrad γ-irradiation.

Based on previous considerations, the identified UV bands in both studied undoped zinc phosphate and barium phosphate glasses can be related to absorption from trace impurities of ferric (Fe3+) ions.

4.1.2. Interpretation of the optical spectra of MnO2–doped glasses before and after gamma irradiation All the MnO 2 –oped zinc phosphate and barium phosphate glasses reveal before gamma irradiation, the same UV absorption band as the undoped glasses without any further visible absorption to the rest of measurements at 1100 nm. This result indicates primarily the absence of any absorption due to trivalent manganese ions which are known to appear in the visible region of the spectrum (480–540 nm) as indicated by various studies [28,29]. The second observation is the absence of any distinct absorption bands due to divalent manganese ions which are expected to be present in the studied two varieties of MnO 2 –doped host phosphate glasses. Divalent manganese ions are known to exhibit weak peaks within the visible region due to spin–forbidden nature of the expected peaks from the (d5 ) configuration [7–9]. Bingham and Parke [10] assumed that the clear identification of absorption for divalent manganese ions was when the concentration of manganese ions reached 10%. The identified UV absorption bands from the MnO2–doped glasses can be related to the same attribution as that for the undoped glasses to originate from trace iron impurity. Upon gamma irradiation, the UV absorption spectra of all the studied glasses are observed to be mostly extended with the appearance of two peaks at about 220 and about 310 nm and with the resolution of a broad visible band centered at about 560–580 nm. The effects of gamma irradiation on the optical spectra of MnO2– doped glasses can thus be interpreted as follows:

(i) The extension of the UV absorption with the generation of two strong peaks can be related to photochemical reactions of some ferrous iron (Fe2 +)+ ions, present due to reducing action of phosphate glasses, with generated positive holes during the irradiation process and the formation of additional Fe3 + ions or (Fe2+)+ ions as suggested by Moncke and Ehrt [30]. (ii) The generation of an additional broad visible band can be related to photochemical reactions between some divalent manganese (Mn2+ ion) and formed positive holes by the irradiation process and the formation of Mn3 + ions or (Mn2 +)+ as suggested by Moncke and Ehrt [30], Elbatal et al. [29] and more recently by Ehrt [31].

4.1.3. Interpretation of the optical spectra of reduced MnO2–doped glasses The optical spectra of the two selected reduced MnO2–doped (8%) zinc phosphate and barium phosphate glasses reveal additionally a distinct small band at about 407–410 nm. This specific near visible band is related to d–d transition of Mn2+ ions (d5) as previously mentioned by several authors [7,9,10,20]. This indicates and confirms the assumption of Bingham and Parke [10] that the clear identification of absorption due to divalent manganese (Mn2+) ions necessitates the concentration of manganese ions should be high and with reducing condition for melting. 4.2. Interpretation of the photoluminescence spectra It is very important to have clear insight of previous relevant publications on photoluminescence studies for manganese doped glasses and specifically in phosphate glasses before interpretation of the present luminescence spectra. The following points are a summary of some selected related studies:

10

M.A. Marzouk et al. / Journal of Non-Crystalline Solids 458 (2017) 1–14

Fig. 11. ESR of 2% MnO2 doped (a) zinc phosphate and (b) barium phosphate before and after 10 Mrad γ-irradiation. Fig. 12. Thermal expansion of undoped and MnO2–doped (a) zinc phosphate (b) barium phosphate glasses.

(a) Reisfeld et al. [32] reported the presence of large quantum efficiencies of Mn2 + luminescence at room temperature in phosphate glasses. (b) On the other hand, Faber et al. [33] have identified Mn3 + luminescence in silicate and borate glasses. (c) Machado et al. [34] studied optical properties of doped manganese (1–20%) in barium phosphate glasses and obtained both Mn2 + and Mn3 + species as indicated by the absorption spectra but only one luminescence center was observed and ascribed to Mn2 + ions. (d) Krishna Mohan et al. [9] studied spectroscopic and dielectric properties of MnO doped (0–2.5 mol%) in PbO–Nb2O5–P2O5 glass system. They identified two absorption bands at 525 and 420 nm, corresponding to the transitions 6A1g(S) → 4T1g (G) and 6A1 (S) → 4 T2(G) of Mn2+ ions. The first band is related to octahedral transition whereas the second one is due to tetrahedral transition of divalent manganese ions. When the MnO increased beyond 1 mol%, these two bands appear to be masked with a new absorption at 490 nm which is identified to 5 Eg → 5T2g transition of Mn3+ ions. They identified two luminescence bands at 620 and 540 nm assigned to 4G → 6S transitions. The first band at 620 nm is related to spin forbidden transition of octahedral Mn2+ ions while the second band at 540 nm is related to spin allowed transition of tetrahedral Mn2 + ions. At higher MnO content, the second band decreases due to either reduction of tetrahedral Mn2+ content or due to the conversion of a part of manganese ions from Mn2+ state to Mn3+ state. (e) Kesavulu et al. [35] carried out EPR and photoluminescence of Mn2 + ions (0.5 mol%) in lithium–potassium borophosphate

glasses and identified a green emission band at 582 nm which was assigned to a transition from 4T1g → 6A1g of Mn2+ ions. (f) Spectral studies by Kiran et al. [36] on Mn2 + ions (1–5 mol%) doped in sodium–lead borophosphate glasses indicated the appearance of four weak bands at 410, 445, 490 and 535 nm related to octahedral symmetry of Mn2+ ions. They identified a single broad luminescence emission band at 560 nm. (g) Moncke et al. [37] studied spectral, EPR and fluorescence spectroscopy of Mn2 + ions in metaphosphate fluoride–phosphate and borosilicate glasses. The glasses were melted under reducing conditions. Optical absorption data reveal two bands at 410 and 530 nm due to d–d electronic transitions of Mn2+(d5) and Mn3+ (d4). Fluorescence emission between 590 and 680 nm in fluoride phosphate and low alkaline borosilicate glasses designates octahedral coordination and the green emission at 525 nm for the alkaline borosilicate glasses indicates tetrahedral coordination of Mn2+ ions. (h) The spectral and dielectric properties of manganese ions (1–5 mol%) in new host glassed from Na2SO4–B2O3–P2O5 were studied by Kumar et al. [38]. The optical spectra revealed four weak bands at 410, 433, 490 and 335 nm and photoluminescence spectra showed a single emission band at 590 nm due to octahedral Mn 2 + ions. Based on the previous considerations by different authors the following points are reached:

M.A. Marzouk et al. / Journal of Non-Crystalline Solids 458 (2017) 1–14

11

structural network arrangement is quite different for both the ZnO-P2O5 and BaO-P2O5 glassy system and in their crystalline glass–ceramic derivatives as evident and revealed by the different situation or coordination of manganese ions in them. It is obvious that the luminescence spectra are observed to be variable in the crystalline glass–ceramic derivatives while the base parent glasses show almost no variation. This indicates that the glass–ceramics contain distinct microcrystalline structural units reflecting that the ZnO–P2O5 glass ceramic structural units are different because it comprises additional structural ZnO4 units while BaO–P2O5 glass ceramic consists of Ba2 + in modifying position linking to phosphate groups. These variations in the structural units are obviously affecting the housing of Mn2+ ions and hence the difference in the luminescence spectra. 4.3. Optical parameters, band gap (Eopt), Urbach energy (ΔE) and refractive index (n) From the experimental obtained results of optical band gap, Urbach energy (ΔE) and refractive index a number of facts can be concluded:

Fig. 13. XRD of the heat-treated glasses of undoped and MnO2 doped (a) zinc phosphate (b) barium phosphate samples.

(a) The luminescence spectra show some limited variations with the type or composition of the glasses. The identified luminescence emission spectra of manganese ions in lead glasses at 620 and 450 nm are related to octahedral and tetrahedral Mn2+ ions, respectively [9]. In borosilicate glasses, the identified emission band at 525 nm is related to tetrahedral Mn2+ ions [37]. (b) The identified emission sharp band in all the studied MnO2– doped zinc phosphate glasses or barium phosphate glasses can be related to octahedral Mn2+ ions as reached by various authors [9,37]. (c) The reduced MnO2 (8%) doped glasses reveal excitation peaks at about 350–354, 410, 516–538 nm while the emission band is at 610 nm in zinc phosphate glass and at 632 nm in barium phosphate glass. This result indicates that the emission peak is related to octahedral Mn2+ ions. (d) The glass ceramic derivatives of the two glass systems show some variations in the excitation and emission spectra. The MnO2–doped ZnO-P2O5 glass ceramic reveals four excitation peaks at 350, 410, 445 and 495 nm and with the emission band at 560 nm indicating the presence of Mn2+ ions in tetrahedral coordination. On the other hand, the MnO2–doped barium phosphate glass ceramic shows two excitation peaks at 350 and 410 nm while the emission peak is at 625 nm indicating octahedral Mn2 + ions. These different results for the glass ceramics from the two systems seem to reflect and confirm that the

(1) The introduction of MnO2 causes the progressive increase in the concentration of non-bridging oxygens leading to a decrease in the bridging oxygen (P–O–P). (2) The addition of multivalent metal ions results in the increase of the bonding defect which leads to an increase of the localization degree of electrons thereby the increase of donor center in the glass matrix which enhances the shift of absorption edge towards longer wavelengths [39]. (3) The data can be related to suggested changes of network structure, which are brought by the introduction of transition metal ions Zn2 + or heavy metal ions Ba2+ ions. The values of Eopt in zinc phosphate glasses are higher than that for barium phosphate glasses. The structural differences which are the result of the different site occupation, where Zn2+ ions can partly share as structural ZnO4 units while Ba2+ ions are solely modifiers hence cause a different structural arrangement of the phosphate network. Such changes depend on the type, nature and concentration of introduced divalent cations [40]. (4) Many previous studies confirmed that [41,42], the addition of heavy metal oxides to the glass structure causes a deeper band tail extended in the gap and thereby leading to a decrease in the value of Eopt. The presence of heavy metal ions leads to a decrease in the ionic character of the covalent bond which causes a small energy gap of the corresponding material of the prepared glasses. Based on the previous considerations, it can be concluded that the variations of Eopt and ΔE could be attributed to the network structural differences which are brought by the existence of different metal ions. The increase of the refractive index can be related to the changes in glass composition and the change in the polarizability of the oxygens. 4.4. Interpretation of the FTIR spectra For better understanding and interpretation of the FTIR spectra of the two glass systems of undoped and MnO2–doped zinc phosphate and barium phosphate glasses, the following parameters are to be considered: (1) It is recognized that the main structural forming oxide in the two studied glasses is P2O5 with 50 mol%. In this case, the structural PO4 tetrahedra are considered to be the main structural building units which retain 4-fold coordination throughout the full composition range [43,44]. The addition of alkali oxide or divalent network, leads to linear phosphate chains [44,45]. This linear

12

M.A. Marzouk et al. / Journal of Non-Crystalline Solids 458 (2017) 1–14

chain structure involves the cleavage of P–O–P linkages and the creation of non-bridging oxygens (NBOs) in the glass. The divalent ions serve as ionic cross-link between the NBOs of two different chains. This is true for the alkaline earth cations (Ca2+, Sr2+, Ba2+). But some divalent oxides (e.g. ZnO, MgO) are assumed to be able to form structural building groups (ZnO4, MgO4). It is therefore assumed that the studied two varieties of phosphate glasses with BaO or ZnO possess structural network of different arrangements which are expected to be more clear and identified through FTIR infrared absorption spectral measurements. (2) The close inspection of the experimental FTIR spectra of the two undoped zinc phosphate and barium phosphate glass indicates that some differences are identified including the intensities of the various vibrational bands and with the zinc phosphate glass the mid vibrating bands from about 800 to about 1800 cm− 1 are observed to be connected or compacted than the IR spectrum of the barium phosphate glass. (3) It should be remembered that the main vibrational bands in the two different glasses are related to phosphate network as the constitutional building units as the two basic host glasses have 50 mol% P2O5 and thus it can be expected that the main vibrational units belong to the metaphosphate chain and keeping in mind that ZnO may share in P–O–Zn linkages or ZnO4 groups while BaO is solely acting as modifying oxide. The interpretation of the deconvoluted vibrational bands of both undoped zinc phosphate and barium phosphate glasses can be summarized as follows [27,43–45]:

in the bond angles and/or bond lengths of the phosphate groups which are developed by the effect of gamma irradiation in rearrangement process as previously reached by several authors [49–51] (d) The FTIR spectra of the undoped barium phosphate glass appear less compact than that for zinc phosphate glass due to the presence of Ba2+ ions in modifying or linking positions. (e) The addition of MnO2 to barium phosphate glass produces a new stronger IR band at about 1300 cm−1 together with the increase of the intensity of the band at about 900 cm−1. These changes can be related to depolymerization effects including first the appearance of the band at 1300 cm−1 which can be related to different P_O groups which are participating in the coordination environment of the Mn2+ ions as suggested by Metwalli et al. [52]. The increase of the intensity of the band at about 900 cm−1 can be related to asymmetric vibrations of P\\O\\P groups linked with linear metaphosphate chains [45]. This change is considered as a depolymerization or rearrangement process. (f) Gamma irradiation is observed to cause compaction of the IR spectrum of the undoped barium phosphate glass including the changes in the intensity of the vibrating bands. This result can be related to some sort of rearrangement process. The MnO2–doped glass reveal limited changes upon gamma irradiation including the reappearance of the vibrations of the IR band at 1640 cm−1 specifically at high concentration of MnO2 which is correlated with OH or POH vibrations. 4.6. Interpretation of the E.S.R spectra

−1

can be related to modifier cation vibra(a) The peak at 414 cm tions (Ba2+, Zn2+). (b) The peak at 521 cm−1 can be related to harmonics or bending vibrations of O_P_O linkages. (c) The middle peak at 746 cm−1 is related to symmetric stretching vibrations of frequency of P\\O\\P bonding or symmetric stretching mode of bridging oxygens in Q2 units. (d) The peak at 884 cm−1 is related to metaphosphate group vibrations. (e) The peaks at 989 and 1121 cm−1 are related to the asymmetric stretching vibrations of frequency of P\\O\\P groups attached to metaphosphate groups. (f) The peak at 1259 cm−1 is attributed to the PO2 asymmetric stretching vibrations of metaphosphate units. (g) The peak at 1376 cm−1 is related to P_O vibrations and (PO2) asymmetrical stretching mode of nonbridging oxygens in Q2 units. (h) The peak at 1641 cm−1 is related to bending vibrations of H2O molecules or P-OH bridges. 4.5. Interpretation of the effect of the MnO2 content and gamma irradiation on the FTIR spectra of the zinc phosphate and barium phosphate glasses (a) The appearance of compact FTIR spectrum for the undoped zinc phosphate glass which is quite different than the IR of undoped barium phosphate glass can be related to the ability of Zn2 + ions to form ZnO4 or P–O–Zn bridges via phosphate chains linkages in the glass network as revealed by several authors [46–48]. (b) The disappearance of the IR band at 1376 cm−1 with the introduction of MnO2 can be related to depolymerization process by the decrease of isolated P_O groups into the metaphosphate chain as visualized by the increase of the IR bands at about 989 and 1250 cm−1. (c) Gamma irradiation causes the increase of the two IR bands at 1115 and 915 cm−1 which are related to asymmetric stretching vibrations of P\\O\\P bonding or symmetric mode of bridging oxygen in Q2 units. This behavior can be related to some changes

The E.S.R spectra of the two selected studied glasses containing 2% MnO2 of the two variant systems reveal specific characteristic equal signals for both of them with the same positions but with slightly different intensities. Similar results have been reached by various glass scientists [35,37]. They have agreed that the observed signals belong to resonance of Mn2+ ions. The different response of the two variant samples with gamma irradiation can be related to the stability of the zinc phosphate glass structure containing additional network units such as ZnO4 and/or P-O-Zn while the barium phosphate glass consists of Ba2+ ions in modifying position as modifiers and therefore this glass is more affected by gamma radiation than that of zinc phosphate glass. 4.7. Interpretation of the thermal expansion data The experimental thermal expansion results reveal two different behaviors for the two studied phosphate systems. The barium phosphate glasses exhibit normal continuous increase of the thermal expansion with temperature in two rates before reaching the dilatometric softening temperature. On the other hand, the zinc phosphate glasses reveal anomalous behavior by decreasing of the thermal expansion at first within low temperatures up to 130 °C and followed by a reverse increase in thermal expansion until reaching the dilatometric softening temperatures of the different glasses. These different responses of the thermal expansion with type or composition of the glass can be understood and explained on the following basis: (a) Thermal expansion of glasses is considered to be a complex property connected with the network structural constituents, the bonding strength of the various cations with oxygens and with the magnitude and distribution of forces acting in the system and reflects any change of the distribution of forces which increases thermal vibrations [53,54]. A solid is considered as an array of oscillating masses. (b) Rawson [55] described normal expansion in solids (including glasses) as being due to the increasing amplitude of the atomic

M.A. Marzouk et al. / Journal of Non-Crystalline Solids 458 (2017) 1–14

vibrations of the constituents. From lattice energy considerations, the ionic vibrations are anharmonic and as a result with increasing the amplitude of vibration, the inter-ionic distance increases [56]. The different ions vibrate independently of each other. (c) It is accepted that glass normally expands on heating. However, there are some glasses which show zero or negative expansion on heating. Some of the glasses and glass–ceramics from the system of alkaline earth oxide–borates reveal zero or negative thermal expansion coefficients and several authors have recommended them for various commercial applications [57]. (d) The experimental thermal changes identified in the thermal expansion behavior in the two glass systems can be interpreted as follows: (i) The thermal expansion data of all the barium phosphate glasses (undoped or MnO2–doped samples) reveal like ordinary phosphate glasses small thermal expansion at low temperature and then the rate of thermal expansion sharply increases until reaching the transformation range and softening temperature. This behavior is correlated with the structural arrangement of the barium phosphate glass in which barium ions (Ba2+) are behaving only as modifiers. (ii) The thermal expansion of all the zinc phosphate glasses show an anomalous behavior where the thermal expansion decreases at low temperature up to about 130 °C and then afterwards the thermal expansion increases until reaching the transformation range and dilatmetric softening temperature. This behavior can also be discussed in relation to the structural role of Zn2+ ions in the glasses. At first ZnO can enter the glass structure in network forming units (as ZnO4) whereby the glass network becomes compact and thermal expansion at first is observed to decrease. ZnO can also exist in modifying groups ZnO6 which is similar to the same effect of BaO. Thus the anomalous change in thermal expansion of zinc phosphate glass can thus be related to the change in the coordination of Zn2+ ions. (iii) Tischendorf et al. [58] have suggested that zinc atoms in metaphosphate glass are exhibiting tetrahedral coordination and change to higher coordination after the metaphosphate composition and this assumption could explain the dilatometric softening temperatures of the ZnO–P2O5 glass system. (iv) Similar assumptions have been advanced for the anomalous behavior of glasses from ZnO–P2O5 glass system [59].

13

This can be realized and interpreted by the same assumptions as the thermal expansion data. (c) For the barium phosphate glass which is known to have Ba2 + ions solely as modifiers and hence the crystallization of which reveal only one phase of barium metaphosphate crystalline species as the chemical composition of the studied glass without any variations. Experimental X-ray diffraction data indicate that the addition of MnO2 causes distinct changes in the separated crystalline phases and also they are quite different with the type of host phosphate glass and this behavior can be realized and interpreted as follows: (a) The addition of 2% MnO2 to the ZnO–P2O5 glass causes the persistence of the crystallization of hexagonal zinc meta-phosphate phase with the increase of its percent to 70.4% instead of 61.4%. But on the other hand, the manganese ions are involved in the separation of manganese phosphate hydrate crystal phase and the disappearance of the other zinc phosphate phase of the formula 3ZnO–P2O5. It is obvious that the manganese ions at first is not acting as nucleating agent only but are involved in the development and formation of manganese crystalline phase and also changing the crystallographic type of zinc meta-phosphate phase. On increasing the MnO2 to 8%, the abundant identified crystalline phases are two, the zinc meta-phosphate species with the hexagonal type prevailing, beside the persistence of manganese phosphate hydrate phase with 19.4%. This result for ZnO·P2O5 glass is supporting the assumption that Zn2+ ions are situated in two coordinations, tetrahedral and octahedral coordinations and their ability to share in the crystallization behavior is different as revealed in increasing the MnO2 content. (b) On the other hand, the addition of 2% MnO2 to the BaO–P2O5 glasses causes the separation of an additional crystalline barium manganese phosphate phase beside the main phase of barium phosphate and the same main phase is continued with the increase of MnO2 to 8% but with the increase of the second phase of barium manganese phosphate from 19.8% to 27%. This result indicates that the manganese ions are introduced in an additional phase but did not change the crystallographic details of the main barium phosphate phase which supports the assumption that Ba2 + ions are solely situated in only network forming positions. 5. Conclusion

4.8. Interpretation of the crystallization behavior The x-ray diffraction patterns data (Fig. 12) indicate the crystallization of the two studied zinc phosphate glass and barium phosphate glass in varying behaviors. These crystallization results can be realized and interpreted as follows: (a) The voluminous crystallization of the two studied glasses can be related to the presence of P2O5 in appreciable percent of 50 mol% with the ability of P2O5 to induce nucleation and crystallization as suggested by McMillan [60]. (b) The different crystallization behavior of the two glass systems can be related with the possibility of suggested formation of additional constitutional network forming units. For the ZnO–P2O5 glass system it is assumed that Zn2+ ions can participate in the structure in two ways by forming structural building ZnO4 units and as network modifying ZnO6 units. This assumption is introduced to explain the anomalous behavior of thermal expansion of this glass system. Also, the crystallization data indicate the formation of two phases, main zinc metaphosphate phase (ZnO·P2O5) and secondary phase of zinc phosphate 3ZnO·P2O5.

MnO2–doped (2,4, 8%) zinc and barium phosphate glasses were prepared by melting under ordinary atmospheric condition and their optical, structural, photoluminescence, thermal and crystallization properties were investigated. Optical absorption spectra before irradiation showed strong UV absorption and no indication of divalent or trivalent manganese ions. Specific samples melted with reducing agent indicate the appearance of specific absorption of Mn2 + ions at 410 nm. Gamma irradiation produces extension of UV absorption and the generation of an induced visible band due to Mn3 +. These defects are assumed to be generated by photochemical reactions. Photoluminescence spectra show characteristic excitation and emission peaks due to Mn2 + ions. The PL spectra of the crystallized glass ceramic samples show sharper peaks reflecting the distinct crystallinity of the samples. Also, they show variations than the parent glasses which are correlated with the different structural units between the two systems. The variation of Eopt and Δ E values could be attributed to the network structural differences brought by the existence of transition metal ion (manganese) in different oxidation or coordination states. The study of FTIR spectra show vibrational bands due to mainly to metaphosphate groups but exhibit

14

M.A. Marzouk et al. / Journal of Non-Crystalline Solids 458 (2017) 1–14

compactness or condensation in the mid region of zinc phosphate glasses due to sharing of Zn2 + ions in structural building ZnO 4 units. The thermal expansion data reveal anomalous behavior in the ZnO–P2O5 data while the BaO–P2O5 glasses behave in normal expansion. The crystallization behavior data reveal similar trend with ZnO–P 2O 5 glasses leading to two crystalline species of zinc metaphosphate and other zinc phosphate phases while the barium phosphate glasses crystallize in barium metaphosphate phase only. This trend confirms the assumption of the different structural building arrangements of the two glass systems (ZnO–P2O5 and BaO–P2O5). References [1] J.A. Wilder, Glasses and glass ceramics for sealing to aluminum alloys, J. Non-Cryst. Solids 38–39 (1990) 879–884. [2] S.W. Martin, Review of the structures of phosphate glasses, European journal of solid state, Inorg. Chem. 28 (1991) 163–205. [3] R.K. Brow, Review: the Structure of simple phosphate glasses, J. Non-Cryst. Solids 263–264 (2000) 1–28. [4] J.E. Shelby, Introduction to Glass Science and Technology, second ed. The Royal Society of Chemistry, Cambridge, UK, 2005. [5] T. Kubo, J. Cha, H. Takebe, M. Kuwabara, Thermal properties and structure of zinc phosphate glasses, Phys. Chem. Glasses Eur. J. Glass Sci. Technol. B 50 (1) (2009) 15–18. [6] C.E. Smith, R.K. Brow, The properties and structure of zinc magnesium phosphate glasses, J. Non-Cryst. Solids 390 (2014) 51–58. [7] C.R. Bamford, Colour Generation and Control in Glass, Glass Science & Technology, 2, Elsevier Publishing Company, Amsterdam, 1977. [8] N.A. Ghoneim, F.A. Moustaffa, A.H. Zahran, F.M. Ezzeldin, Gamma-ray interaction with lead borate and lead silicate glasses containing manganese, J. Am. Ceram. Soc. 66 (1983) 447–451. [9] N. Krishna Mohan, M. Rami Reddy, C.K. Jayasankar, N. Veeraiah, Spectroscopic and dielectric studies on MnO doped PbO–Nb2O5–P2O5 glass system, J. Alloys Compd. 458 (2008) 66–76. [10] K. Bingham, S. Parke, Absorption and fluorescence spectra of divalence manganese in glasses, Phys. Chem. Glasses 6 (1965) 224. [11] A. Bishay, Radiation induced color centers in multicomponent glasses, J. Non-Cryst. Solids 3 (1970) 54–114. [12] E.J. Friebele, in: D.R. Uhlmann, N.J. Kreidl (Eds.), Optical Properties of Glasses, American Ceramic Society, Westerville, OH, U. S. A. 1991, pp. 205–262. [13] F.H. ElBatal, Gamma ray interaction with copper-doped sodium phosphate glasses, J. Mater. Sci. 43 (2008) 1070–1079. [14] H.A. ElBatal, N.A. Ghoneim, F.M. Ezz ElDin, Effect of thermal expansion of some ternary borate glasses, Nucl. Inst. Methods Phys. Res. B 88 (1994) 411–417. [15] G. Sharma, K. Singh, Manupriya, S. Mohan, H. Singh, S. Bindra, Effects of gamma irradiation on optical and structural properties of PbO–Bi2O3–B2O3 glasses, Radiat. Phys. Chem. 75 (2006) 959–966. [16] M.A. Marzouk, F.H. ElBatal, W.H. Eisa, N.A. Ghoneim, Comparative spectral and shielding studies of binary borate glasses with the heavy metal oxides SrO, CdO, BaO, PbO or Bi2O3 before and after gamma irradiation, J. Non-Cryst. Solids 387 (2014) 155–160. [17] N.F. Mott, E.A. Davis, Electronic Processes in Non-Crystalline Materials, second ed. Clarendon Press, Oxford, 1979. [18] F. Urbach, The long-wavelength edge of photographic sensitivity and of the electronic absorption of solids, Phys. Rev. 92 (1953) 1324. [19] V. Dimitrov, S. Sakka, Electronic oxide polarizability and optical basicity of simple oxides. I, J. Appl. Phys. 79 (1996) 1736–1740. [20] I. Konidakis, C.E. Varsamis, E.I. Kamitsos, D. Möncke, D. Ehrt, Structure and properties of mixed strontium–manganese metaphosphate glasses, J. Phys. Chem. C 114 (2014) 9125–9138. [21] G.H. Sigel, R.J. Ginther, The effect of iron on the ultraviolet absorption of high purity soda-silica glass, Glass Technol. 9 (1968) 66–69. [22] G.H. Sigel Jr., M.J. Marrone, Photoluminescence in as-drawn and irradiated silica optical fibers: an assessment of the role of non-bridging oxygen defect centers, J. NonCryst. Solids 45 (1981) 235–247. [23] J.A. Duffy, Charge transfer spectra of metal ions in glass, Phys. Chem. Glasses 38 (1997) 289–292. [24] D. Ehrt, H. Ebendorff-Heidepriem, H. Albrecht, T. Schröder, W. Triebel, Charge transfer transitions and radiation effects in glasses in the deep-UV range, Glastech. Ber. Glass Sci. Technol. 68 (C1) (1995) 521–526. [25] D. Ehrt, P. Ebeling, U. Natura, UV Transmission and radiation-induced defects in phosphate and fluoride–phosphate glasses, J. Non-Cryst. Solids 263–264 (2000) 240–250. [26] A.M. Abdelghany, F.H. ElBatal, M.A. Azooz, M.A. Ouis, H.A. ElBatal, Optical and infrared absorption spectra of 3d transition metal ions-doped sodium borophosphate glasses and effect of gamma irradiation, Spectrochim. Acta A Mol. Biomol. Spectrosc. 98 (2012) 148–155. [27] M.A. Ouis, H.A. ElBatal, A.M. Abdelghany, A.H. Hammad, Structural and optical properties of CuO in zinc phosphate glasses and effects of gamma irradiation, J. Mol. Struct. 1103 (2016) 224–231.

[28] F.H. ElBatal, M.A. Marzouk, A.M. Abdel Ghany, Gamma rays interaction with bismuth borate glasses doped by transition metal ions, J. Mater. Sci. 46 (2011) 5140–5152. [29] F.H. ElBatal, M.A. Ouis, R.M.M. Morsi, S.Y. Marzouk, Interaction of gamma rays with some sodium phosphate glasses containing cobalt, J. Non-Cryst. Solids 356 (2010) 46. [30] D. Möncke, D. Ehrt, Irradiation induced defects in glasses resulting in the photoionization of polyvalent dopants, Opt. Mater. 25 (2004) 425–437. [31] D. Ehrt, Review: phosphate and fluoride phosphate optical glass properties, structure and applications, Phys. Chem. Glasses Eur. J. Glass Sci. Technol. B 56 (6) (2015) 217–234. [32] R. Reisfeld, A. Kisilev, C.K. Jørgensen, Luminescence of manganese(II) in 24 phosphate glasses, Chem. Phys. Lett. 111 (1984) 19–24. [33] A.J. Faber, A. van Die, G. Blasse, W.F. van der Weg, Luminescence of manganese of different valencies in oxide glasses, Phys. Chem. Glasses 28 (1987) 150. [34] I.E.C. Machado, L. Prado, L. Gomes, J.M. Prison, J.R. Martinelli, Optical properties of manganese in barium phosphate glasses, J. Non-Cryst. Solids 348 (2004) 113–117. [35] C.R. Kesavulu, R.S. Muralidhara, J.L. Rao, R.V. Anavekar, R.P.S. Chakradhar, EPR and photoluminescence studies on lithium-potassium borophosphate glasses doped with Mn2+ ions, J. Alloys Compd. 486 (2009) 46–50. [36] N. Kiran, C.R. Kesavulu, A. Suresh Kumar, J.L. Rao, Spectral studies on Mn2+ ions doped in sodium–lead borophosphate glasses, Phys. B Condens. Matter 406 (2011) 3816–3820. [37] D. Möncke, E.I. Kamitsos, A. Herrmann, D. Ehrt, M. Friedrich, Bonding and ion–ion interactions of Mn2+ ions in fluoride-phosphate and boro-silicate glasses probed by EPR and fluorescence spectroscopy, J. Non-Cryst. Solids 357 (2011) 2542–2551. [38] A.V. Ravi Kumar, C. Srinivasa Rao, N. Narasimha Rao, V. Ravi Kumar, I.V. Kityk, N. Veeraiah, Influence of valence and coordination of manganese ions on spectral and dielectric features of Na2SO4–B2O3–P2O5 glasses, J. Non-Cryst. Solids 358 (2012) 1278–1286. [39] P. Chimalawong, J. Kaewkhao, P. Limsuwan, Optical investigation and electronic polarizability of Nd3+ doped soda-lime silicate glasses, Energy Res. J. 1 (2) (2010) 176–181. [40] H.A.A. Sidek, M. Hamezan, A.W. Zaidan, Z.A. Talib, K. Kaida, Optical characterization of lead-bismuth phosphate glasses, Am. J. Appl. Sci. 2 (8) (2005) 1266–1269. [41] P. Sharma, M. Vashistha, I.P. Jain, Optical properties of Ge20Se80-XBiX thin films, J. Optoelectron. Adv. Mater. 7 (5) (2005) 2647–2654. [42] M.A. Marzouk, F.H. ElBatal, H.A. ElBatal, Effect of TiO2 on the optical, structural and crystallization behavior of barium borate glasses, Opt. Mater. 57 (2016) 14–22. [43] A. Abdel-Kader, A.A. Higazy, M.M. Elkholy, Compositional dependence of infrared absorption spectra studies for TeO2-P2O5 and TeO2-P2O5-Bi2O3 glasses, J. Mater. Sci. Mater. Electron. 2 (1991) 157–163. [44] A.M. Efimov, IR fundamental spectra and structure of pyrophosphate glasses along the 2ZnO·P2O5–2Me2O·P2O5 join (Me being Na and Li), J. Non-Cryst. Solids 209 (1997) 209–227. [45] Y.M. Moustafa, K. El-Egili, Infrared spectra of sodium phosphate glasses, J. Non-Cryst. Solids 240 (1998) 144–153. [46] S. Toyoda, S. Fujino, K. Morinaga, Density, viscosity and surface tension of 50RO– 50P2O5 (R: Mg, Ca, Sr, Ba, and Zn) glass melts, J. Non-Cryst. Solids 321 (3) (2003) 169–174. [47] E. Mansour, G. El-Damrawi, Electrical properties and FTIR spectra of ZnO–PbO–P2O5 glasses, Phys. B Condens. Matter 405 (2010) 2137–2143. [48] M.H. Wan, P.S. Wong, R. Hussin, H.O. Lintang, S. Endud, Structural and luminescence properties of Mn2+ ions doped calcium zinc borophosphate glasses, J. Alloys Compd. 595 (2014) 39–45. [49] W. Primak, Mechanism for the radiation compaction of vitreous silica, J. Appl. Phys. 43 (1972) 2745–2754. [50] L.W. Hobbs, A.N. Sreeram, C.E. Jesurum, B.A. Berger, Structural freedom, topological disorder, and the irradiation-induced amorphization of ceramic structures, Nucl. Inst. Methods Phys. Res. B 116 (1996) 18–25. [51] F. Piao, W.G. Oldham, E.E. Haller, The mechanism of radiation-induced compaction in vitreous silica, J. Non-Cryst. Solids 276 (2000) 61–71. [52] E. Metwalli, M. Karabulut, D.L. Sidebottom, M.M. Morsi, R.K. Brow, Properties and structure of copper ultraphosphate glasses, J. Non-Cryst. Solids 344 (2004) 128–134. [53] W.A. Weyl, E.C. Marboe, The Constitution of Glasses, A Dynamic Interpretation, 1, Academic Press, New York, 1962 353. [54] D.G. Holloway, The Physical Properties of Glass, Wykeham, London, 1973 36. [55] H. Rawson, Properties and Applications of Glass, Glass Science and Technology, 3, Elsevier, Amsterdam, 1980 69. [56] F.M. Ezz ElDin, H.A. ElBatal, Radiation effects on thermal expansion of some ternary alkali-silicate glasses, J. Non-Cryst. Solids 152 (1993) 195–200. [57] C. Rüssel, D. Tauch, R. Garkova, S. Woltz, G. Völksch, Phase separation and crystallisation in borate glasses, Phys. Chem. Glasses Eur. J. Glass Sci. Technol. B 47 (4) (2006) 397–404. [58] B. Tischendorf, J.U. Otaigbe, J.W. Wiench, M. Pruski, B.C. Sales, A study of short and intermediate range order in zinc phosphate glasses, J. Non-Cryst. Solids 282 (2001) 147–158. [59] R.M. Hovhannisyan, Binary alkaline-earth borates: phase diagrams correction and low thermal expansion of crystallized stoichiometric glass compositions, Phys. Chem. Glasses Eur. J. Glass Sci. Technol. B 47 (4) (2006) 460–465. [60] P.W. McMillan, Glass – Ceramic, second ed. Academic Press, London, 1979.