Much research has focused on biomedical glasses and glasses as promising materials for diverse applications  -  . Phosphate glasses have interesting characteristics and properties such as a low melting point, high thermal expansion coefficient, and bioactivity, including the concept of the degradation of biomaterials, which make them useful as biomaterials. Several studies have shown that single phosphate glasses do not have good chemical stability compared to phosphate glasses with several components and have also demonstrated that the macroscopic properties of phosphate bioglasses can be improved, by making small changes in the molar concentration of the modifying oxides and the intermediate network. The latter make it possible to reinforce the structure of the vitreous network while at the same time releasing an amount of calcium and phosphate, during the attack by an SBF solution, necessary to accelerate the regeneration of damaged tissue and improve chemical resistance  . Through the use of phosphorus pentoxide (P2O5) as the initial network, and sodium oxide (Na2O) with calcium oxide (CaO) as network modifiers, followed by adding other oxides (Al2O3, Fe2O3, ZnO and TiO2), the control of degradation may be achieved. Phosphate bioglasses can react with bone tissue through the formation of a hydroxyapatite layer, which is equivalent to the mineral phase of bone; that will then be involved in the process of bone regeneration via a set of physical and chemical reactions  . These surface reactions are also responsible for the degradation of the bioglass after implantation    . The aim of the present study was to study the structural changes, chemical resistance modification and properties of bioglasses in the Al2O3-CaO-NaO-P2O5 system, focusing particularly on the effect of the addition of CaO and Al2O3 to phosphate oxide glasses for use in the medical field. So the study of series of glasses of composition xAl2O3-(40-x)CaO-10NaO-50P2O5 (with 0 ≤ x ≤ 10 moles%) indicates that the substitution of CaO by Al2O3 in the glass network entrained a decrease of the chemical durability, and a important change from metaphosphate (Q2) and pyrophosphate structural units (Q1) toward short isolated orthophosphate units (Q0) when the Al2O3 content reached 10 mol%, confirmed by IR spectrum and X-Ray diffraction.
2. Experimental Procedures
Phosphate glasses were produced by the direct melting of a mixture of (NH4)2HPO4, CaCO3, Na2O and Al2O3 in suitable proportions. The reagents were ground together and then introduced into a porcelain crucible. Then, they were heated initially to a temperature of 300˚C for 2 hours and then 500˚C for 1 h to complete decomposition. The reaction mixture was then heated at 900˚C for 40 min and finally at 1080˚C for 30 minutes to obtain a homogeneous liquid. This was then poured onto an aluminum plate which had a temperature of 200˚C to avoid thermal shock. This procedure provided pellets 5 - 10 mm in diameter and 1 to 3 mm thick. The prepared samples were attacked with distilled water at 90˚C for 20 days to determine the dissolution rate estimated from the mass loss. IR spectroscopy analysis was done in a frequency range between 400 cm−1 and 1300 cm−1 with a resolution of 2 cm−1 using a Fourier transform Vertex 70 spectrometer and recorded on a DTGS detector (deuterium triglycine sulfate). The samples were ground and mixed with KBr, which is transparent to infrared. The ratio of the material to KBr in the pellets was 10% to 90% by weight. The analysis by X-ray diffraction was used to identify the structure of the glasses annealed at 550˚C and 660˚C for 48 hours. The samples were analyzed by an X’Pert Pro MPD Panalytical diffractometer. The microstructure of the glass samples was characterized using a scanning electron microscope (SEM).
3.1. Ternary Diagram
As can be seen in the ternary diagram in Figure 1, the location of the studied glasses indicates that these glasses are theoretically formed of metaphosphate and pyrophosphate groups. The composition of each sample is given in Table 1.
3.2. Chemical Durability
The dissolution rate (DR) calculated for the series of bioglasses xAl2O3-(40 - x)CaO-10Na2O-50P2O5 (with 0 ≤ x ≤ 10, mol%) is defined as the loss of
Figure 1. Localization of the studied samples in the ternary diagram of (Na2O∙P2O5)-(Al2O3∙P2O5)-(CaO∙P2O5). Table 1 gives the corresponding compositions in the quaternary system Al2O3-CaO-Na2O-P2O5.
glass mass after immersion in 100 mL of distilled water at 90˚C for 20 days, and expressed as g∙cm−2∙min−1. The average dissolution rates, which are shown in Figure 2 and Table 2, were measured with respect to the glass surface and the time of exposure. The results show that the chemical durability was improved after increasing the molar percentage of CaO to the detriment of Al2O3   .
3.3. Infrared Spectra
The IR spectra of the glasses xAl2O3-(40 - x)CaO-10Na2O-50P2O5 (with 0 ≤ x ≤ 10) are shown in Figure 3, and the vibration bands of the assignments are given
Figure 2. The chemical durability of phosphate glasses versus the Al2O3 content.
Table 1. Glass composition expressed in terms of quaternary systems.
Table 2. The composition of the glasses in mol% and some characteristics of the quaternary glasses xAl2O3-(40 - x)CaO-10Na2O-50P2O5.
Figure 3. IR spectra of the series of xAl2O3-(40 - x)CaO-10Na2O-50P2O5 glasses, x = 0, 5, 7.5 and 10.
in Table 3. All vibration bands are located between 400 and 1300 cm−1. The vibration bands located around 700 - 780 cm−1 are attributed to the vibration mode usym(P-O-P) of pyrophosphate groups (Q1), while the bands between 900 - 924 cm−1 are attributed to the vibration mode uasym(P-O-P) of pyrophosphate groups (Q1). When the Al2O3 content reached 10 mol%, we observed the appearance of a band at 980 cm−1, assigned to the vibration mode uasym(P-O-P) of isolated orthophosphate groups Q0, to the detriment of bands at 750 - 780 cm−1 and 900 - 924 cm−1 attributed to pyrophosphate groups, which became single shoulders. The band in the frequency range between 1225 - 1280 cm−1 is assigned to the vibration mode uasym(PO2) of metaphosphate groups without a bridging oxygen (Q2)    , which became weaker when the value of x was greater than 7.5 mol%. The vibration bands located around 1070 and 1120 cm−1, characteristic of the vibration modes uym(PO2) and uasym(PO3) of tetrahedral units, Q2 + Q1, which are a mixture of pyrophosphate and metaphosphate groups, also underwent a decrease in intensity when the Al2O3 content exceeded 5 mol%.
3.4. X-Ray Diffraction
X-ray diffraction confirmed the vitreous character of all the samples. The XRD spectra of samples S050 to S1050, annealed at 550˚C and 660˚C for 48 h are shown in Figure 4. Several phases were detected, including calcium metaphosphate or
Figure 4. XRD patterns for the glass samples S050, S7.550 and S1050 after heat treatment for 48 h in an air atmosphere at 550˚C and 600˚C.
cyclic metaphosphate (Ca(PO3)2, NaPO3 and Ca2P4O11 in sample S050, while the same figure shows calcium pyrophosphate (Ca2P2O7), sodium-aluminum pyrophosphate (NaAlP2O7) and Ca(PO3)2 in sample S7.550      . However, for the compound S1050, there is a radical structural change which indicates the predominance of orthophosphates phase type AlPO4 and eventually Ca3(PO4)2.
3.5. Scanning Electron Microscopy
Figure 5 shows SEM micrographs of sample S050 with the compositions 40CaO∙10Na2O∙50P2O5 (Figure 5(a)), as well as samples S7.550 and S1050 with the compositions 7.5Al2O3∙32.5CaO∙10Na2O∙50P2O5 (Figure 5(b)) and 10Al2O3∙ 30CaO∙10Na2O∙50P2O5 (Figure 5(c)), respectively. The SEM micro graph in
(a) (b) (c) (d)
Figure 5. SEM micrograph showing the visual structure of the samples S050 (a); S7.550 (b) and S1050(c) before attack and of the samples S050 (d); S7.550 (e) and S1050 (f) after attack in distillated water at 90˚C for 20 days.
Figure 5 illustrates the morphology of the glasses, before and after immersion in distillated water for 20 consecutive days, considered in this work. It shows the existence of two phases, one crystalline and the other glass     . It also indicates the formation of agglomerates of the crystalline phase from one micrometer to a few tens of micrometers in size. The SEM micrograph also indicates that the number of crystallites increased from S050 to S1050 when the Al2O3
Table 3. The assignments of different vibration bands of the IR spectra of the quaternary xAl2O3-(40-x)CaO-10NaO-50P2O5.
content increased at the expense of the CaO content.
In this study, we prepared a glass series with different percentages of phosphate oxides; Al2O3, Na2O, CaO and P2O5 were used as the basic constituents. In this series, we substituted the CaO oxide with Al2O3 oxide while keeping the percentage of P2O5 and Na2O oxides constant. The improved chemical durability was attributed to the replacement of the easily hydrated P-O-P bonds by covalent and resistant Ca-O-P bonds. However, the substitution of CaO with Al2O3 in the glass network led to a considerable decrease in chemical durability. The IR spectra indicated a radical structural change from metaphosphate or cyclic metaphosphate and pyrophosphate towards short majority isolated orthophosphate groups when the Al2O3 content was above 7.5 mol%. Also, the structure deduced from the vibrational spectroscopy is compatible with the localizations of the analysed compounds S050 to S7.550 inside the ternary diagram given in Figure 1 with the exception of compound S1050 which predominantly contains isolated orthophosphate groups This is seems to due to the physical and chemical properties of the intermediate oxide that participates in the glass formation (melting temperature, rate of Ca + Al/P.)    . The X-ray diffraction spectra confirmed the presence of Ca2(P2O7), CaP4O11, Ca(PO3)2 and Ca3(PO4)2 crystalline phases in the S050 sample and NaAl(P2O7), Ca2(P2O7) and Ca(PO3)2 crystalline phases in the S7.550 sample, while in the S1050 sample, the structure obtained is formed of almost crystalline isolated orthophosphate phase AlPO4 with some trace of Ca3(PO4)2. The SEM micrographs (Figure 5) indicated that the number of crystallites increased from S050 to S1050 when the Al2O3 content increased at the expense of the CaO content. The increase in crystallites in phosphate network glasses, in general, improves the chemical durability   . In our case, we observed the opposite. It was also noted in these micrographs that crystallite metaphosphate cyclic chains predominated in S050, to different sizes of crystallites observed in S7.550 and finally to a larger number of crystallites of different sizes in S1050, likely dominated by small crystallite sizes assigned to isolated orthophosphate groups. The increase in short chain isolated orthophosphate groups in the glass network at the expense of cyclical chain metaphosphate groups and pyrophosphate chains when the Al2O3 content exceeded 7.5 mol% can be explained by the fact that we were close to the border area between crystal and glass. The number of crystallites of different sizes generally increased and exceeded the equilibrium that must be established between the glass and the crystallites, which led to a significant decrease in chemical durability  . Indeed the compound S1050, has an opaque white appearance which explains its frontal position between the glass and the crystal. However, the favorable formation of the isolated orthophosphate groups when the CaO oxide is substituted by Al2O3 oxide, and especially when the Al2O3 level reaches 10 mol%, can be of great utility in the formation of the apatite layers in an organic-substance-free a cellular simulated body fluid (SBF), able of regenerating bone.
The structure and chemical durability of a glass series composed of xAl2O3-(40 − x)CaO-10Na2O-50P2O5 (with 0≤ x ≤10) (mol%) were investigated using various techniques such as IR, X-ray diffraction and SEM. The improved chemical durability in these glasses was attributed to the replacement of the easily hydrated P-O-P bonds by covalent and resistant Ca-O-P bands. However, the substitution of CaO by Al2O3 in the glass network led to a considerable decrease in the chemical durability. The IR spectra indicated a radical structure change from metaphosphate or cyclic metaphosphate and pyrophosphate towards short isolated orthophosphate groups when the Al2O3 content was above 7.5 mol%. Consequently, we can predict the depolymerization of the large phosphate network into isolated short chains of the orthophosphate (Q˚) type. The SEM micrographs of S050, S7.550 and S1050 illustrate that the number of crystallites increased from S050 to S1050 when the Al2O3 content increased at the expense of the CaO content. The increase in the number of crystallites that led to the formation of isolated short chains of the orthophosphate (Q˚) type almost certainly caused disproportionality between the glass and the crystallites, which caused a significant decrease in chemical durability. Hence, a better understanding of glass corrosion is very relevant to the industry in the development of technical bioglasses to achieve both good performance and, at the same time, provide calcium ion and Orthophosphate ions, in an aqueous solution or SBF solution, necessary for the formation of effective hydroxyapatite layers in bone regeneration in biomedical applications.
The authors wish to thank National Center for Scientific and Technical Research [Division of Technical Support Unit for Scientific Research (TSUSR) Rabat, Morocco] for their assistance to the realization of this work. We also thank Pr. R. Elouatib (Laboratory physic and chemistry of inorganic materials) for the support that has brought us.