Crystallization of Anatase TiO₂ in Niobium Potassium Phosphate Glasses

Abstract

In this work, the glass forming ability was studied in potassium phosphate glasses with increasing amounts of TiO₂ in order to obtain a glass-ceramic with photocatalytic properties. The first studied series has been the binary system (100-x) KPO₃-xTiO₂. Homogeneous and transparent glasses could be obtained with x varying from 10 to 30 mole%. Since the photocatalytic anatase phase could not be precipitated in this system, the ternary system KPO₃-TiO₂-Nb₂O₅ was investigated in order to incorporate higher TiO₂ contents without spontaneous crystallization under cooling. Thermal properties of all glass samples were investigated by DSC and allowed identifying an increase of glass transition temperatures with increasing TiO₂. For all compositions, exothermic events related with crystallization were also observed and suitable heat-treatments resulted in specific crystalline phases identified by X-ray diffraction. Selective precipitation of the anatase titanium oxide was successfully obtained from the glass compostion 35KPO₃-25Nb₂O₅-40TiO₂ (KN25T40) after heat treatment at 720°C for 2 h, suggesting the possibility of obtaining a glass-ceramic for photocatalytic applications. Structural investigations by Raman were also performed on glasses and glass-ceramics and allowed to point out the glass intermediary behavior of TiO₂ in the phosphate vitreous network where TiO₄ and TiO₆ octahedra are inserted inside the phosphate network with TiO₆ clusters identified at higher TiO₂ contents. Raman analysis also identified anatase TiO₂ in the KN25T40 glass-ceramic.

Keywords:
glass; phosphate; titanium oxide; anatase

1. Introduction

Titanium dioxide is considered one of the most important semiconductor oxide with interesting chemical, electrical and optical properties. Since its discovery as photocatalytic agent observed by Fujishima and Honda in 1972, research on TiO₂ applications has attracting considerable interest in several fields of science due to its excellent properties, including chemical stability, photostability and appropriate electronic band structure^1,2.

Titanium dioxide thin films have been formed on glass, steel and other surfaces by a wide range of techniques, especially by sol-gel and chemical vapour deposition^3,4. However, the properties of these materials produced by deposition or coating techniques can change over time by surface damage and thus a recoating process can be necessary. On the other hand, TiO₂ crystallites precipitated in the glass matrix can exhibit stable physical and chemical properties even with surface polishing⁵.

There are three allotropic forms of TiO₂: anatase (tetragonal), brookite (orthorhombic) and rutile (tetragonal) among which rutile is the most thermodynamically stable phase for bulk TiO₂ under most conditions. However, the rutile activity as a photocatalytic compound is generally poor. Anatase TiO₂ is considered to be the active photocatalytic component based on charge carrier dynamics, chemical properties and the activity of photocatalytic degradation of organic compounds^1,3.

When a glass-ceramic is obtained by heat-treatment of the mother glass, both the disordered glass regions and ordered crystalline regions are presents. The final material can exhibit not only the benefits of the glass material but also the unique physical properties originated from the crystalline phase. However, only a few works reported phosphate glass-ceramics containing anatase TiO₂ in crystal volume fraction by heat treatment and this lack is in part attributed to the difficulty of promoting controlled and selective crystallization of anatase in phosphate-based glasses with high TiO₂ contents since titania is known to act as a nucleating agent of other crystalline phases⁶.

Selective crystallization of the glass CaO-Bi₂O₃-B₂O₃-Al₂O₃-TiO₂ containing 20% TiO₂ under heat-treatment using an infrared halogen lamp is reported but allowed only precipitation of the rutile phase⁷. Recently, selective crystallization of TiO₂ by controlled heat-treatment and SiO₂ addition in the glass 14TiO₂-23ZnO-45B₂O₃-18Al₂O₃ was also reported⁸ but XRD results indicate that TiO₂ was not the unique crystalline phase after heat treatment and Al₄B₂O₉ was precipitated as a subcrystalline phase.

In this work, TiO₂-containing glass compositions were investigated for preferential anatase crystallization. First, the binary system (100-x)KPO₃-xTiO₂ was investigated and glass samples were obtained from x=10 to x=30. These glasses were characterized by DSC and Raman spectroscopy and specific heat-treatments were performed but anatase could not be precipitated under heat-treatments. Adding Nb₂O₅ to this binary system allowed incorporating higher TiO₂ contents (40%) and anatase TiO₂ could be successfully precipitated in the mother glass.

2. Experimental Part

Glass samples were synthesized from the precursors: titanium dioxide 99,9% (TiO₂) from Vetec, potassium phosphate monobasic 99,9% (KH₂PO₄) from Synth and niobium oxide 99,8% (Nb₂O₅) from Sigma-Aldrich by conventional melt-quenching: The starting powders were weighted using an analytical balance and grinded in an agate mortar. The batches were melted for 2 hours in a platinum crucible between 1100 and 1450°C depending on the TiO₂ content. Finally, the melts were quenched in a steel mold preheated 20°C below the glass transition temperature and kept at this temperature for 8 hours before slow cooling inside the furnace.First, samples in the binary system (100-x) KPO₃-xTiO₂ with x varying from 10 to 30 (mole%) were prepared using the methodology described above. In a second step, a ternary system was investigated with the introduction of niobium oxide as glass intermediary in order to increase the TiO₂ content in the final glass using the same methodology. Molar compositions, melting temperature and visual aspect of the glass samples are presented in Table 1. DSC curves were performed on bulk glass samples of 30 mg in Pt/Rh covered crucibles between 200ºC and 1100ºC at 10ºC/min under N₂ atmosphere. These thermal analyzes were obtained using a DSC/TG calorimeter STA 449 F3 Jupiter from Netzsch. X-ray diffraction measurements were performed on powder samples using a Rigaku ultima IV diffractometer working at 40KV and 30mA between 10º and 70º in continuous mode of 0,02º/s. The crystalline phases were identified according to X-ray powder diffraction patterns (PDF file)⁹. Finally, Raman spectra were collected on bulk glass samples between 100cm^-1 and 1200cm^-1 using a LabRam Micro-Raman from Horiba Jobin-Yvon operating at 632,8 nm with a He-Ne laser.

Table 1
Molar compositions, characteristic temperatures and visual aspect of the glass samples.

Sample	Molar Compositions (% mol)			Melting Temperature	Picture
	KPO3	TiO2	Nb2O5	(°C)
KT10	90	10	___	1100
KT20	80	20	___	1200
KT30	70	30	___	1300
KN25T40	35	40	25	1450
KN20T40	40	40	20	1500

3. Results

Transparent and homogeneous samples were obtained in the binary system (100-x)KPO₃ - xTiO₂ with x varying from 10 to 30 mole% by melt-quenching with an increasing brownish color for higher TiO₂ contents. As expected, an increase of the melt viscosity with increasing TiO₂ content was observed during synthesis and higher melting temperatures were used for a suitable casting of the melt. However, partial or total crystallization of the samples was observed for TiO₂ contents above 30 mole% under these experimental conditions. After X-ray diffraction analysis of the crystalline phases obtained by heat-treatment in these glasses, Nb₂O₅ was added and samples containing 40 mole% of TiO₂ could be prepared. In fact, several compositions were tested by varying the KPO₃:Nb₂O₅ ratio (50:10; 40:20; 35:25; 30:30) with constant TiO₂ content of 40%. Compositions 50KPO₃-10Nb₂O₅-40TiO₂ and 30KPO₃-30Nb₂O₅-40TiO₂ crystallized under quenching while compositions 40KPO₃-20Nb₂O₅-40TiO₂ and 35KPO₃-25Nb₂O₅-40TiO₂ sucessfully vitrified. Table 1 resumes all investigated compositions, required melting temperatures and the visual aspect of the final samples. DSC curves for all bulk glass samples are presented in Figure 1 and Table 2 resumes the characteristic temperatures Tg, Tx and Tf where Tg is the glass transition temperature, Tx the onset of crystallization and Tf the melting temperature as well as the thermal stability parameter Tx-Tg. These thermal data point out that the addition of TiO₂ results in a significant increase in Tg values from 330°C to 510°C in the binary system KPO₃-TiO₂ with an increase of the thermal stability parameter from 10% to 20% and further decrease for higher contents. Nb₂O₅ incorporation also results in a clear increase of Tg identified around 650°C for 20 and 25 mole% of niobium oxide. Crystallization events were observed for all samples as well as melting for glasses KT10 and KT30 at the same temperature. Heat-treatments were applied near the crystallization temperatures in order to identify the crystallization events observed by DSC and check the ability of these glass compositions to precipitate titanium dioxide crystalline phases and especially anatase, which is known for its efficient photocatalytic properties. The crystalline phase identifications performed by X-ray diffraction are summarized in Figure 2 and Figure 3 as well as Table 2. Sample KT10 only exhibited precipitation of long-chains potassium metaphosphate KPO₃ whereas sample KT20 suffered crystallization of potassium pyrophosphate K₄P₂O₇ and cyclic potassium metaphosphate K₃P₃O₉. For sample KT30, phases KPO₃ and KTiOPO₄ were detected. Finally, for samples modified with niobium oxide, anatase TiO₂ could be identified as the first precipitated phase in sample KN25T40 heat-treated below the first crystallization peak whereas heat-treatments at higher temperatures induce crystallization of the mother glass with formation of K₂TiNb₂P₂O₁₃. For the other niobium titanium phosphate glass KN20T40, crystalline phases K₂TiNb₂P₂O₁₃, K₃Nb₆P₄O₂₆ and rutile TiO₂ were identified.

Table 2
Characteristic temperatures, thermal stability and crystalline phases identified in heat-treated glass samples.

Samples	Characteristic temperatures (°C)					Thermal stability parameter	Crystalline phase after heat-treatments
	Tg	Tx1	Tx2	Tx3	Tf	Tx1-Tg
KT10	330	475	620	-	780	145	KPO3
KT20	440	705	740	-	-	265	K4P2O7
							K3P3O9
KT30	510	650	660	-	770	140	KPO3
							KTiOPO4
KN20T40	670	730	-	-	-	60	Rutile TiO2
							K3Nb6P4O26
KN25T40							K2TiNb2P2O13
	650	720	790	870		70	Anatase TiO2
							K2TiNb2P2O13

• - Event not observed.

Figure 1
DSC curves of glass samples in the binary system KPO₃-TiO₂ (10-30% TiO₂) and ternary KPO₃-Nb₂O₅- TiO₂ (40% TiO₂).

Figure 2
X-ray diffraction patterns of heat-treated glasses in the binary system KPO₃-TiO₂.

Figure 3
X-ray diffraction patterns of heat-treated glasses in the ternary system KPO₃-Nb₂O₅-TiO₂.

Raman spectra of the glass samples and crystalline references KPO₃, TiO₂ and Nb₂O₅ in the frequency region between 100 and 1500 cm⁻¹ were recorded to probe the structural evolution with TiO₂ addition and to compare the Raman features of precursor glass and glass-ceramic KN25T40 (Figure 4). In the binary system KPO₃-TiO₂, the first clear change in Raman spectra is the progressive vanishing of vibrational bands centered at 1150cm^-1 and 680cm^-1 and commonly attributed to symmetric stretching modes of P-O bonds in P-O-P linkages and Q² tetrahedra in metaphosphate compounds^10-11. On the other hand, TiO₂ addition results in new Raman signals centered around 1250cm^-1 and 520cm^-1 which shifts to lower wavenumbers. These signals are reported to be due to P=O terminal bonds in PO₄ tetrahedra wherein the other three oxygens are linked to another cation (P-O-X where X=P or Ti) and vibrational modes of distorted TiO₆ respectively¹⁰. Other Raman features appear with TiO₂ addition at 290cm^-1, 735cm^-1, 863cm^-1, 935cm^-1 and 977cm^-1 and were attributed to bending modes of TiO₆ octahedra¹², stretching modes of TiO₆^10,13, stretching modes of TiO₄^14-15, Ti-O stretchings of the axial Ti-O bond in TiO₅ pyramidal units^10,16-17 and PO₄ orthophosphate units respectively¹⁸. Precursor glass composition KN25T40 is dominated by broad Raman bands centered around 200 cm^-1, 650 cm^-1 and 825 cm^-1 as for crystalline Nb₂O₅ whereas the glass-ceramic exhibits sharper Raman signals at 240 cm^-1 and 640 cm^-1 as for crystalline anatase TiO₂.

Figure 4
Raman spectra of glasses, glass-ceramic KN25T40 and crystalline references TiO₂, Nb₂O₅ and KPO₃.

4. Discussion

Since the main objective of this work was the obtaining of glass-ceramics containing anatase crystallites for photocatalytic applications, the glass forming ability of the binary system KPO₃-TiO₂ was investigated with increasing amounts of TiO₂. Homogeneous glass samples were obtained by melt-quenching from TiO₂ contents ranging from 10 to 30 mole% whereas higher contents lead to crystallized samples under these experimental conditions. As presented in Table 1, these glasses also exhibit an increasing brownish color attributed to partial titanium reduction from Ti⁴⁺ to Ti³⁺. It is suggested that the increasing melting temperatures can induce oxygen loss of the melt and titanium reduction as described for non-stochiometric crystalline titanium oxides. The strong absorption in the visible is due to both d-d internal electronic transition of Ti³⁺ ( d¹ electronic configuration) and electronic polaron transitions between reduced Ti³⁺ and oxidized Ti⁴⁺ species^19-20. These increasing melting temperatures and increasing viscosity of the resulting melt observed during synthesis bring a first indication of the effective insertion of titanium ions inside the phosphate chains and higher network connectivity and field strength. A detailed structural description can be extracted from a general analysis of thermal results, Raman data and crystalline phase identification obtained by X-ray diffraction. Since the results obtained from each technique point out the same structural evolution, these data are discussed simultaneously. The first important point is the clear increase of glass transition temperatures from 330ºC to 510ºC for TiO₂ contents ranging from 10 to 30 mole%. This behavior is commonly related with an increase in the glass network connectivity as well as in the network bond strength as reported in many structural studies concerning transition metal oxide-containing phosphate glasses. In fact, these metallic oxides usually exhibit strong metal-oxygen bonds and high coordination number. In addition, their intermediary behavior in phosphate network usually allow their partial or complete insertion inside the phosphate covalent chains, resulting in cross-linking bonds between these chains. In our case, the Tg increase with composition can be attributed to a progressive insertion of titanium oxide units such as TiO₄ and/or TiO₆ between PO₄ tetrahedra. These species were identified by the Raman bands observed at 863cm^-1 and 520cm^-1 respectively. The former Raman band is often attributed to Nb-O-P linkages²¹ but is better attributed to TiO₄ units in these glasses since niobium free samples prepared in the binary system KPO₃-TiO₂ also exhibit this signal. The potassium metaphosphate network depolymerisation is also supported by the progressive vanishing of Raman bands at 680cm^-1 and 1150cm^-1 related with P-O-P and P-O terminal bonds in Q² tetrahedra. Another important result supporting this structural evolution is the preferential precipitation of potassium metaphosphate KPO₃ in sample KT10 at 475ºC which melts at 780ºC whereas sample KT20 exhibits crystallization of potassium pyrophosphate K₄P₂O₇ and cyclic metaphosphate K₃P₃O₉. The other crystallization event observed for KT10 can be related with precipitation of titanium oxide with high melting temperature not observed below 1000ºC. These titanium oxide polyhedra can increase the network connectivity by cross-linked Ti-O-P bonds. In fact, the appearance of the Raman band around 1250cm^-1 and attributed to P=O bonds means that all other P-O linkages in these PO₄ tetrahedra are bridging bonds like P-O-Ti bonds resulting in this non-resonant P=O species. Based on the well-known glass intermediary behavior of Nb₂O₅ in phosphate glasses, it is also suggested that niobium oxide incorporation promotes bridging Nb-O-P bonds and higher glass network connectivity, in agreement with higher Tg temperatures for these glasses. In addition, The Raman band centered at 863cm^-1 and attributed to TiO₄ units enhances from 10% to 20% of TiO₂ but further decreases for 30% but a new band at 935cm^-1 clearly appears for KT30. Since this Raman signal is known to be due to the axial Ti-O bond in pyramidal TiO₅ units, it can be suggested that titanium ions are preferentially four-fold and six-fold coordinated for lower TiO₂ contents whereas TiO₆ and TiO₅ units dominate for higher titanium oxide concentrations. The precipitation of titanium-containing crystalline phases could not be detected by X-ray diffraction of the heat-treated samples KT10 and KT20 but the thermal behavior of sample KT30 also gives important structural clues explaining the limited glass forming ability of this binary system. The Raman spectrum of this sample is dominated by signals centered at 290, 520, 935 and 1250cm^-1 related with TiO₆, TiO₅ and PO₄ units whereas the DSC curve exhibits two crystallization peaks and a melting event. Heat-treatment of this sample resulted in precipitation of crystalline phases KPO₃ and KTiOPO₄. Since the Tg temperature of this glass is higher than compositions KT10 and KT20, it appears that titanium oxide polyhedra effectively modified the phosphate network through P-O-Ti bonds but the thermal treatment induces a spontaneous phase separation of a potassium metaphosphate structure and potassium titanium orthophosphate constituted of TiO₆ octahedra and orthophosphate PO₄^3- ions. The KPO₃ formation has been detected by X-ray diffraction as well as melting event observed at 770ºC. This phase separation tendency for these higher TiO₂ contents is in agreement with the spontaneous crystallization of melts with higher TiO₂ contents by melt-quenching.

Anyway, these binary compositions were not able to precipitate TiO₂ anatase by heat-treatment and niobium oxide Nb₂O₅ was added in order to investigate the glass forming ability of compositions with 40 mole% of TiO₂. Samples with 10% and 30% of Nb₂O₅ crystallized under quenching while compositions with 20% and 25% successfully vitrified (KN20T40 and KN25T40). For sample KN20T40, only one crystallization peak was observed by DSC with Tx at 730ºC and heat-treatments were performed at 750ºC and 1050ºC in order to identify the first crystalline phase as well as other possible precipitations not detected by thermal analysis. The first heat-treatment resulted in a diffraction pattern that could not be identified using the ICSD database whereas heat-treatment at higher temperatures allowed characterizing the potassium niobium phosphate K₃Nb₆P₄O₂₆, potassium niobium titanium phosphate K₂TiNb₂P₂O₁₃ and rutile TiO₂. For this composition, these results clearly point out that these crystalline phase precipitation are not related with distinct crystallization peaks and can hardly be selectively precipitated by suitable heat-treatments. The chemical composition of these phases suggests crystallization of the mother glass with rutile TiO₂. In any case, anatase TiO₂ could not be detected for this composition. On the other hand, sample KN25T40 exhibits three distinct exothermic events starting at 720ºC, 790ºC and 870ºC attributed to precipitation of several crystalline phases in the glass. For this reason, heat-treatments were performed at 720ºC, 780ºC and 880ºC and the resulting samples were characterized by X-ray diffraction. As shown in Figure 3 and Table 2, these results allowed attributing the first peak to precipitation of anatase while the second exothermic event is related with crystallization of K₂TiNb₂P₂O₁₃. A stoichiometric analysis of the starting glass composition (35KPO₃-25Nb₂O₅-40TiO₂ corresponding to K₇Ti₈Nb₁₀P₇O₆₂ average formula) and final crystalline phases TiO₂ and K₂TiNb₂P₂O₁₃ suggest that approximately half of the starting titanium ions is precipitated as anatase resulting in a final composition of the residual glass next to K₇Ti₄Nb₁₀P₇O₅₄ which is close to the composition of the crystalline phase K₂TiNb₂P₂O₁₃. For this composition, it has been shown that anatase can be preferentially precipitated as the first crystalline phase in these glasses resulting in a glass-ceramic containing only anatase crystallites. The Raman data presented in Figure 4 also support this assumption. The Raman spectrum of the precursor glass KN25T40 exhibits intense signals around 230cm^-1, 750cm^-1 and 850cm^-1 also observed for crystalline Nb₂O₅ and attributed to bending and stretching modes of NbO₆ octahedra. The higher polarizability of Nb-O bonds when compared to P-O and Ti-O bonds explains why the former bonds dominate the spectrum when compared to the latter one. On the other hand, The Raman spectrum of the glass-ceramic KN25T40 obtained from heat-treatment of the starting glass at 720ºC shows a very distinct signal with dominant narrow bands centered at 150cm^-1 and 650cm^-1 characteristic of bending and stretching modes of the crystalline phase anatase.

5. Conclusion

Homogeneous and transparent titanophosphate glasses were obtained in the binary KPO₃-TiO₂ and ternary KPO₃-TiO₂-Nb₂O₅ systems by the melt-quenching method. 30 mole% of TiO₂ was successfully incorporated without devitrification in the binary system and the structural changes were monitored by DSC, Raman and X-ray diffraction of the heat-treated glasses. Titanium oxide polyhedra are inserted inside the phosphate chains and increase the network connectivity. However, for higher TiO₂ contents, a phase separation has been identified without precipitation of the desired anatase TiO₂ phase. Addition of Nb₂O₅ as an intermediary compound led to important structural changes, increasing the stability against devitrification and homogeneous glasses containing 40% TiO₂ could be obtained. The glass composition 35KPO₃-25Nb₂O₅-40TiO₂ exhibits a preferential precipitation of anatase in the mother glass resulting in glass-ceramics with a potential photocatalytic activity.

6. Acknowledgments

The authors would like to thank funding agencies CAPES, CNPq, FAPEMIG and FINEP for financial support.

7. References

• ¹
  Ye M, Vennerberg D, Lin C, Lin Z. Nanostructured TiO2 architectures for environmental and energy applications. Journal of Nanoscience Letters 2012;2:1-35.
• ²
  Fujishima A, Honda K. Electrochemical photolysis of water at a semiconductor electrode. Nature 1972;238(5358):37-38.
• ³
  Gupta SM, Tripathi M. A review of TiO2 nanoparticles Chinese Science Bulletin 2011;56:1639-1657.
• ⁴
  Page K, Palgrave RG, Parkin IP, Wilson M, Savin SLP, Chadwick AV. Titania and silver-titania composite films on glass-potent antimicrobial coatings. Journal of Materials Chemistry 2007;17(1):95-104.
• ⁵
  Yazawa T, Machida F, Oki K, Mineshige A, Kobune M. Novel porous TiO2 glass-ceramics with highly photocatalytic ability. Ceramics International 2009;35(4):1693-1697.
• ⁶
  Masai H, Fujiwara T, Mori H, Komatsu T. Fabrication of TiO2 nanocrystallized glass. Applied Physics Letters 2007;90(8):081907.
• ⁷
  Masai H, Kanamori E, Takahashi Y, Fujiwara T. Surface crystallization of CaO-Bi2O3-B2O3-Al2O3-TiO2 glass using IR furnace. Journal of Non-Crystalline Solids 2010;356(52-54):2977-2979.
• ⁸
  Yoshida K, Masai H, Takahashi Y, Ihara R, Fujiwara T. Selective crystallization of anatase and rutile by control of heat-treatment conditions and SiO2 addition in TiO2-crystallized glass. Journal of the Ceramic Society of Japan 2013;121(1420):999-1003.
• ⁹
  International Centre for Diffraction Data-Joint Committee of Powder Diffraction Standards. Powder diffraction file (PDF) Newton Square: International Centre for Diffraction Data-Joint Committee of Powder Diffraction Standards; 2000.
• ¹⁰
  Cardinal T, Fargin E, Flem E, Couzi M. Raman Scattering and XAFS Study of Optically Nonlinear Glasses of the TiO2-NaPO3-Na2B4O7 System. Journal of Solid State Chemistry 1995;120(1):151-156.
• ¹¹
  Kaur M, Singh A, Thakur V, Singh L. Effect of TiO2 substitution on optical and structural aspects of phosphate glasses. Journal of Molecular Structure 2015;1089:95-101.
• ¹²
  Farrow LA, Vogel EM. Raman spectra of phosphate and silicate glasses doped with the cations Ti, Nb and Bi. Journal of Non-Crystalline Solids 1992;143:59-64.
• ¹³
  Dias AG, Skakle JMS, Gibson IR, Lopes MA, Santos JD. In situ thermal and structural characterization of bioactive calcium phosphate glass ceramics containing TiO2 and MgO oxides: High temperature - XRD studies. Journal of Non-Crystalline Solids 2005;351(10-11):810-817.
• ¹⁴
  Sakka S, Miyaji F, Fukumi K. Structure of binary K2O-TiO2 and Cs2O-TiO2 glasses. Journal of Non-Crystalline Solids 1989;112(1-3):64-68.
• ¹⁵
  Zheng K, Liao J, Wang X, Zhang Z. Raman spectroscopic study of the structural properties of CaO-MgO-SiO2-TiO2 slags. Journal of Non-Crystalline Solids 2013;376:209-215.
• ¹⁶
  Magyari K, Stefan R, Vulpoi A, Baia L. Bioactivity evolution of calcium-free borophosphate glass with addition of titanium dioxide. Journal of Non-Crystalline Solids 2015;410:112-117.
• ¹⁷
  Tiwari B, Pandey M, Sudarsan V, Deb SK, Kothiyal GP. Study of structural modification of sodium aluminophosphate glasses with TiO2 addition through Raman and NMR spectroscopy. Physica B: Condensed Matter 2009;404(1):47-51.
• ¹⁸
  Morikawa H, Lee S, Kasuga T, Brauer DS. Effects of magnesium for calcium substitution in P2O5-CaO-TiO2 glasses. Journal of Non-Crystalline Solids 2013;380:53-59.
• ¹⁹
  El Batal FH, Ibrahim S, Marzouk MA. UV-visible, infrared absorption spectra of undoped and TiO2-doped lead phosphate glasses and the effect of gamma irradiation. Radiation Effects and Defects in Solids 2012;167(4):256-267.
• ²⁰
  Ramachandra Rao MV, Gandhi Y, Srinivasa Rao L, Sahayabaskaran G, Veeraiah N. Electrical and spectroscopic properties of LiF-Bi2O3-P2O5:TiO2 glass system. Materials Chemistry and Physics 2011;126(1-2):58-68.
• ²¹
  Sene FF, Martinelli JR, Gomes L. Synthesis and characterization of niobium phosphate glasses containing barium and potassium. Journal of Non-Crystalline Solids 2004;348:30-37.

Publication Dates

• Publication in this collection
  16 Feb 2017
• Date of issue
  Mar-Apr 2017

History

• Received
  31 Mar 2016
• Reviewed
  11 Oct 2016
• Accepted
  25 Jan 2017