3. 1. Raman spectroscopy
The structure of prepared glassy phosphates of the system 20Na2O−(50−x)Na2WO4−xTiO2−30P2O5 (x= 0, 1, 5, 8, 10, 15, 20, and 25 mol%) were studied by using Raman scattering, spectroscopy, which carried out in the wavenumber range between 1200 cm-1 and 150 cm-1 and the results are gathered in Figure 1. The bands of the Raman spectra are attributed according to literature [45−47]. Therefore, the principal band positions and their attributions are gathered in Table 1. Where the main bands are checked at different positions: 1095 cm-1, 935 cm-1, 885 cm-1, 750 cm-1, 650−570 cm-1, 390 cm-1 and 255 cm-1. From Figure 1, it was observed that the main bands were located from 935 cm−1 to 750 cm−1. These are associated with asymmetric and symmetric vibration modes of non-bridging oxygen on different P−tetrahedra. Besides, the systematic decrease in peak frequency for P−O bonds on 3−dimensional phosphates (Q2), P2O7 (Q1) dimer and isolated PO4 (Q0). The band around 885 cm−1 is associated with the symmetric vibration mode of the W–O bond of the WO6 and/or W−O−P bond. The band of low intensity, located at 750 cm-1, is associated with the symmetric vibration modes of the P–O–P bond of the Q1 units also appeared. The bands located in the low region between 650 cm-1 and 255 cm−1, are generally assigned as stretching modes of the [PO4] units.
Table 1
Raman band assignments in the 1200−150 cm-1 range frequency for the studied phosphate glasses.
Band position (cm-1)
|
Band assignment
|
1095
|
νs(PO2)−, Q2
|
935
|
νas(P−O−P)/ νas(P−O−M) (M= W. Ti)
|
885
|
Ο−Μ−Ο/ΜΟ4/ΜΟ6 (M= W. Ti)
|
750
|
νs(P−O−P), Q1
|
650−570
|
νs(P−O−P), Q2
|
390
|
νs(M−O−M)(M= W, Ti)
|
255
|
δ δ(PO4)
|
Even after the substitution of Na2WO4 by TiO2 in the system, it appeared that the spectra have changed significantly with an increasing of TiO2 content. From composition (x= 1 mol%), it was observed that two supplementary bands at 935 cm−1 and 885 cm−1are, respectively, dependent on the TiO4 and TiO6. That their intensities decreased at TiO2 mol% increased (region marked with red color, Figure 1) [48]. The relative intensity of the peak assigned to Q1 decreased when TiO2 content increased. Besides, it looked that the intensity of the bands between 450 cm−1 and 180 cm−1 started decreasing when the quantity of TiO2 increased (region marked with green color as mentioned in Figure 1).
At increasing TiO2 content increased up to 5 mol% and more, the band in the position 1095 cm−1 was observed with an intensity decreasing with increasing the quantity of TiO2. The vibrations in a range from 1050 cm−1 to 1095 cm−1 are associated with the symmetric vibration mode of nonbridging oxygens in pyrophosphate structural units (P2O7)4− [49]. In the region of lower wave numbers, a band at 650 cm−1 is observed; the intensity decreases with increasing content of TiO2. According to literature [50], it could be designed to control the vibration mode of Ti−O bonds in TiO6 octahedral units. It can be predicted by the existence of several structural units in different glassy frameworks. Besides, some suggestions showed that the titanium ions could be occupied by distorted octahedrals, which contributed to the modification glassy matrix, and then the NBOs increased. Because of the existence of TiO2 mol% in high quantities (x> 5), the structural units TiO4 break the P=O bonds, which leads to the forming of new non−bridging oxygen ions. The structural units [TiO6/2]2 will be formed [51]. In additions, the TiO6 structural units increase at high TiO2 quantities.
From the analysis Figure 1, the Raman spectra attributed, the vibration intensity identifying the metaphosphate Q2 units is significantly reduced up to x = 8 mol% and the pyrophosphate Q1 units predominate for x ≥ 8. But the exact determination, the quantity of Q1 and Q2 units is difficult. Also, the higher polarizability of the Ti−O bonds compared with the P−O bonds has been influenced by the intensity attributed to Raman bands. In spite of increasing TiO2 to 20 mol% and 25 mol%, it can find glassy phosphates that have the metaphosphate Q2 units [50]. It assumes that increasing the TiO2 mol% in the glassy framework led to partly changing the metaphosphate structure to a pyrophosphate, then an orthophosphate. At a quantity of TiO2 (x<8 mol%) the metaphosphate units are predominant in the glassy structural. After that, the pyrophosphate units will be raised, also the orthophosphate units. Therefore, the TiO6 units interlink with different phosphate units (metaphosphate, pyrophosphate and orthophosphate) leading to formation of new bonds P−O−Ti, which participate to increase the durability for prepared glasses (see section 3.2).
3. 2. Chemical durability
3. 2. a. Weight loss
Glassy phosphates of simple compositions are very vulnerable. Their corrosion resistance changes only very slightly with condensation. Certainly an improvement appears in the polyphosphate → metaphosphate → ultraphosphate series associated with the finding of the phosphorus atoms at the Q3 site, slowing the water diffusion step-by-steric effect. However, the durability of these compositions remains insufficient to hope for potential applications. An effective improvement involves an increase in the cross-linking of the phosphate entities, which form the glassy network to make it as tight as possible to aqueous attacks. To develop phosphate glasses resistant to corrosion by water, it is necessary to choose not only the type of phosphate anions which will be used for the structural building, but also the addition cations or the substitution anions. A compromise between these glassy components makes it possible to optimize the cross-linking of the entity phosphates while taking care to obtain vitrifiable compositions [52–53].
Data collected from durability tests are given in curve forms, as seen in Figure 2. Where, the weight loss at the initial surface (g.cm-2) is dependent on the time (h). For the first time, it was observed that the glasses had the same dissolution behavior (see Figure 2). To simplify knowing the curve variation, it followed, one method is based on dividing the curves. For this reason, it could separate into two regions: the first one is at t<100h, where the dissolution is dominated by a linear behavior. But, the second one, between 100h and 250h, which is scarred by a slow-down in the dissolution rate. Regarding the substitution of Na2WO4 by TiO2, it was checked the dissolution comportment of the synthesized glasses banks on the TiO2 mol%. At the initial situation of dissolution of the glassy composition (x = 0 mol%), the dissolution was faster than other glassy compositions (x> 0 mol%). Because of this, it could conclude that the substitution of Na2WO4 by TiO2 influences the solubility of the studied glasses. Furthermore, insertion of TiO2 mol% in glassy phosphate improved the durability of glassy framework. In previous studies, it has been tested some oxides; Fe2O3 and TiO2 for improving the durability of the glassy phosphates [54–55]. From these previous works, it has been concluded that the insertion of Fe2O3 or TiO2 into glassy networks enhanced the durability by the creation of a new P−O−Fe/Ti bonds, which lead to increasing of network rigidity. For similar roads, the insertion of different quantities of the quantity of the TiO2 into our studied glasses has improved of durability of the studied glassy network, through creation of new P−O−Ti bonds. This hypothesis was corroborated by Raman spectroscopy, by identification of the new bonds (P−O−Ti). Also, it was confirmed by thermal examination (DSC). Where, at the integration of TiO2 in glassy networks, Tg increased.
Insertion of TiO2 into glasses improves the strengthening of the chains formed in the glassy network of the Na2O−Na2WO4−TiO2−P2O5 system. This explanation is supported through the correlation of chemical durability with thermal and Raman analysis.
3. 2. B. Ph Variation
Corrosion by water of alkaline phosphate glasses of compositions (1−x)P2O5−xNa2O) can be summed up by the involvement of two mechanisms (hydration and hydrolysis) The importance of each one varies as a function of composition (x) and the degree of depolymerization [52]. Hydration is connected to the diffusion of water, forming a hydration layer. Hydration is predominant when the glassy structure consists mainly of Q2 units forming chains, i.e. around x = 0.5 (metaphosphate and polyphosphate). It generally begins with the chain ends formed by hydroxylated Q1 entities, giving free access to penetration of water between the chains. Well-known that the hydrated chains detach intact from the glassy surface to migrate into the solution. But the hydrolysis mechanism is characterized by:
i) low-energy reaction when it is an interdiffusional phase with an ion-exchange reaction H+↔Na+, called "dealkalinization" or "ion-exchange", between the ions of the solution and the Na+ ions located among the phosphate entities;
ii) much more energetic reaction regarding causing the rupture of a P−O−P bond with binding of H+ and OH− ions from the dissociation of water. It is in particular responsible for dissolving the ultraphosphates. But the chains were relieved into the solution, starting with their ends [52].
Figure 3 exposes the evolution of the pH curves as a function of the submersion time in demineralized water at a temperature of 30 ° C. In the same way, for analyzing the curves of chemical durability, it’s going to analyze the pH curves. From the pH curves, it could be clearly differentiated into two regions: i) at t<48 h it’s characterized by a hard decrease in pH curves; ii) at 48 h>, the pH curves show a horizontal linear (behavior is nearly constant) with a distinction among the curves of the studied glasses. From this attitude of the curves, it can be suggested that the development of pH values is established on the chemical composition of the glassy studied. In the literature, the evolution of pH is affected by the forming of phosphoric acid, which is survived by the H2PO4− units, in water [54]. Generally, through first contact, the glass with the solution (water), the alkali ions and the phosphate ions created on the glassy surface have been leached into the solution. Therefore, the ion interchange between ions formed on the glassy surface and those of a solution; it could determine the acid−base character of the solution. In effect, the sudden decrease in pH curves associated with the influx of the phosphate units (H2PO4−) from glassy surface to the solution (first region). But when the H2PO4− units slowdown because of saturation of the solution, the pH curves behavior becomes constant (second region). At TiO2 mol% increased (x>5) into the glassy network. The pH curves decreased (only for the first region). But for the second region, saturation rapidly happened, as the quantity of TiO2 increased. Because of this, it was verified that the insertion of the TiO2 mol% into the glassy framework enhanced the durability for the studied glasses.
In conclusion, our study of chemical durability interested in the produced glasses could distinguish between two regions: the first one is described by a rapid alteration. Where thermodynamic disequilibrium happens between the glassy surface and the changing solution is important, which leads to an abrupt decrease in the pH value. The second one, is dominated by corrosion, which is checked with the saturation of the altering solution. Where, the saturation phenomenon has happened because of the founding of the protective layer against diffusion on the glassy surface.
2. 3. Crystallization Kinetics By Dsc
Thermal analysis of glasses by DSC makes it prospective to identify their behavior as a function of temperature. When we analyze glasses using several heating speeds (β), we can get information on their crystallization. From these non-isothermal measurements, we can follow the development of the crystallization temperature as a function of the heating rate (Tc= (β)).This variation allowed us to decide the parameters characterizing the crystallization of materials, including the activation energy of the crystallization (Ec).
The activation energy of the crystallization (Ec) is determined from the decide form of Kissinger equation (1), which was reported by Matusita and Saka [56]. This crystallization activation energy (Ec) will be investigated to establish the crystallization kinetics mechanism of the studied glasses.
Ln(T2p/β)= Εc/(RTP) + constant (1)
Where; β, R and Tp are, respectively, the heating rate β (5°C/min, 8°C/min, 10°C/min, 12°C/min), the universal gas constant, and the peak temperature.
From the value of the activation energy (Ec), the Avrami exponent (n) was calculated using Augis–Bennett equation (2) [57]:
n= (2.5/ΔT)×(RT2P/Ec) (2)
With; n and ΔT are, respectively, the crystallization index or Avrami exponent and the full width of the exothermic DSC peak at half-maximum intensity. The Avrami exponent (n) indicated the growing and nucleation mechanisms. In the literature, (n) is nominated to crystallization pattern, n= 1 designed to surface crystallization, n= 2 means mono-dimensional crystallization, n= 3 means bi-dimensional crystallization, n= 4 designs to crystallization for glassy materials were in three-dimensional [58–59].
Figure 4 exhibits the DSC thermograms corresponding to the glassy formulation (x= 5) for all suggested heating rates. It was checked that each DSC thermogram presented a crystallization peak. But in comparison between the crystallization peaks, it noted, slowly shifted to the superior temperature value as heating rates increased. The obtained values, corresponding crystallization temperature (Tp) of the synthesized glass (x = 5 mol %) are gathered in Table 2. By examination of the crystallization temperature (Tp) and heating rates, it has been achieved from the Figure 5, which exposes the plot of ln(Tp2/β) versus 1/Tp. Also, it could be determined Ec from the slope of the variation, the parameter ln(Tp2/β) as a function of 1/Tp. The obtained Ec value is around ≈ 144.77 kJ/mol. The Avrami parameter (n) is found to be about 2, which suggests that the crystallization of the prepared glasses is controlled by a mono-dimensional mechanism [60–62].
Table 2
Crystallization temperature of the glass (x = 5 mol %) at different heat crystallization rates.
Glass
|
β(°C/min)
|
|
Tp
|
|
ΔT
|
|
Avrami parameter (n)
|
20Na2O−45Na2WO4−5TiO2−30P2O5
|
5
8
|
|
558
566
|
|
18
20
|
|
2.43
2.25
|
|
10
|
|
570
|
|
21
|
|
2.17
|
|
12
|
|
573
|
|
23
|
|
2.01
|