Generation and bleaching of E′-centers induced in a-SiO₂ by γ-irradiation

Abstract

Radiation indicator as well as radiation shielding of extremely high dose have been proposed. Three different types of a-SiO₂ namely, G₁-xerogel, G₂-fused and G₃-natural silica were monitored by EPR before and after γ-irradiation. The E′-center has been used for EPR radiation characterization of a-SiO₂ with assuming that the signal intensity changes with γ-irradiation differently as those of other EPR signals do. Formation and decay of the E′-center are closely related with its precursor, diamagnetic oxygen vacancies. Gamma ray of large dose (500 kGy) creates oxygen vacancies giving up-and-down irradiation effects, which, therefore, might be useful for high dose radiation indicator (G₁) and radiation shielding (G₂ and G₃).

Introduction

The process of applications of high irradiation doses have been developed in the area of industry, medicine and agriculture [1], optical fibers [2–5] and it can be used for photonics [6]. The effect of irradiation in inducing local structural changes (point defects) in wide band gap insulators like amorphous SiO₂ is currently an attractive research field owing to the large utilization of silica-based optical components for laser and electronic applications [7, 8]. It is well accepted that the generation of defects by irradiation takes place from the transformation of native precursor centers [9–11]. These conversion processes can be activated by diverse channels; directly by the bond cleavage at the precursor site, or indirectly by trapping atomic or molecular diffusing species released due to ionization process.

It is accepted that, electron paramagnetic resonance (EPR) spectroscopy is one of the most useful tools to investigate the irradiation-induced transformations involving paramagnetic defects, i.e. centers having an unpaired electrons. Indeed, as the signals observed in the EPR spectra are considered as fingerprint of paramagnetic centers, they could provide information on the nature and the kinetics of structural changes occurring at the sites of these defects in amorphous or crystalline silica matrix [2]. Together with a specific interest in the structural modifications of the quartz matrix, one kind of radiation defects in SiO₂, namely E′-defect centers consisting of a hole trapping by a neutral oxygen mono-vacancy, attracting wide interest in EPR as an useful two-level spin system [12].

Recently, the three-fold coordinated Si with one unpaired electron (≡Si^·) constitutes the widely accepted basic configuration of the paramagnetic E′-centers in crystalline and amorphous SiO₂ (a-SiO₂) [13]. In spite of this assignment, the identification of E′-center generation mechanism is still an open question, as well as the involvement of one or more precursors. The oxygen-deficient O⋯Si–Si⋯O structure (where ⋯ represent three covalent bonds of Si with neighboring O atoms and—represents the covalent bond between Si) has been suggested as a possible precursors of E′-center in a-SiO₂ [14, 15].

One more still unanswered question regards the mechanisms of generation of the E′ centers by irradiation. Diverse processes have been postulated like the activation of one or concurrently more precursors that give rise to an exponential growth of defects [16]; the activation of the unperturbed matrix that is correlated to growth of defects without saturation [17, 18]; and finally, a generation of the E′-centers with successive conversion to other defects concurrently with a back conversion to recover the starting structure. These processes have been associated with sublinear dependence of the concentration on the irradiation dose [19].

In the present work we report experimental investigations on the generation and bleaching of E′-centers and weak satellite peaks induced by either γ-irradiation or thermal-treatment in some selected varieties of a-SiO₂. The generation of E′-centers was monitored by EPR measurements in each variety in a dose range from 1 to 500 kGy. Also, the present work includes FTIR to study the effect of both water molecules and hydroxyl groups on the behavior of EPR signal intensity. As a final point, an effort has been given to found possibility to use one of these studied materials as candidate for radiation indicator or radiation shielding purposes.

Materials, instruments, and methods

Various high purity silica samples were employed in our experiments. They are grouped in three types: xerogel (G₁), fused (G₂) and natural (G₃) amorphous-silica as reported in Table 1.

Table 1 Sample list, code name, g-factor and A-constant (hyperfine interaction)

• a. Preparation of xerogel-silica: The sol preparation should lead to lowest residual SiOH quantity which should still be sufficient to form the structural network. The studied G₁ sample was prepared by the alkoxide method. Pure silica gels sintering in the range between 60 and 700 °C, were prepared using procedure via the hydrolysis and condensation of (CH₃CH₂O)₄Si tetraethoxysilane TEOS, ethanol and distilled water, in the presence of hydrochloric acid, with molar ratios; 1:6:10:3 for TEOS: CH₃CH₂OH:H₂O:HCl. These solutions were filtered, followed by stirring for 1 h at room temperature. The resultant homogeneous solutions for preparing bulk materials were filled in moulds and aged in a drying oven type GFL 71.5, at 60 °C for 21 days until no shrinkage was observed. Densification and crystallization of gel were obtained, by sintering in air for 2 h at temperature ranging from 200 °C up to 700 °C, in a muffle furnace type (Lento) with a heating rate 1.5 °C/min. The final products were monolithic, clear, transparent, and cracks-free and then they were converted to fine powder using a pestle and mortar.

• b. Synthetic fused silica (G₂) was obtained from (Reidel-de Haën, Germany), which contains <1 ppm total cationic impurities other than hydrogen. This variety constitutes of lumbered bars of pure Z-growth material from which 100 g were taken and crushed for use in the EPR investigations.

• c. The variety considered here as G₃, is commercial natural sand deposit. The sample of commercial origin is the best raw material for sand deposits is collected from Abu Zeinema, Sinaii, Egypt. The sample was washed by water, sieved to obtain the 0.1–0.25 mm grain fractions and treated with deionizer water to remove organic materials. After cleaning, the separated grains were soaked in 6 mol/L HCl acid solution for about 1 h to remove any contaminated traces from feldspar as well as the outer layers of the grains. The samples were thoroughly washed with water and dried at 100 °C. Finally, any magnetic minerals in this sample were removed by a magnetic separator.

Irradiation procedure

The irradiation of powdered samples was performed with γ-rays in the ⁶⁰Co gamma cell 220 Excel at the central spatial position. This place is the most uniform isodose in the chamber, and it’s accurately calibration was made using the standard NPL alanine reference dosimeter. The temperature during γ-ray irradiation was adjusted at 35 °C. The absorbed dos rate at the time of irradiation was about 3.06 kGy/h. Samples were irradiated at room temperature and the radiation doses delivered to samples ranged from 1 to 500 kGy to generate paramagnetic centers.

Thermal treatment

Isothermal treatment for native samples

Heat-treatment was carried out at 300 °C for holding times (1, 2 and 3 h) in an electrically—heated furnace. After heat-treatment, the samples were left to cool from the specific temperature to room temperature (~25 °C). The EPR spectra of the samples after heat-treatment show an inducing for different EPR signal lines which vary with varying the exposure time.

Post-irradiation stability

To study the effect of long storage time at room temperature (~25 ± 2 °C), the EPR intensities of previously irradiated samples at g ~2 were measured every few days (1–3 days) until it reached 120 days. However, the effect of thermal bleaching was carried out by studying the changes in EPR intensity for the same irradiated samples and at the same g value. The data were recorded after being bleached at 200 °C for different time intervals (10 min) during 130 min.

EPR measurements

Prior to all the measurements, the samples were ground and sieved (~125 µm). EPR measurements were performed at room temperature with a Bruker EMX spectrometer (X-band) working at 9.714 GHz and with modulation frequency 100 kHz. The cavity used was the standard Bruker ER 4102 rectangular cavity. E′-center was identified under non-saturation condition at microwave power: P = 1.2 mW, and with modulation amplitude: m. a. = 4 Gauss, time constant 81.92 ms, receiver gain 8.93 × 10³. Samples were inserted in EPR tubes and measured at the above instrument parameters. All EPR measurements were carried out at lab temperature (25 ± 2 °C). The spin concentration was estimated by double integration of the EPR signal and by comparison with a reference sample of known concentration DPPH (α, α-diphenyl β-picrylhydrazyl) [20].

FTIR measurements

FTIR absorbance spectra were recorded on a Perkin Elmer Spectrum GX FTIR within the spectral range 400–4,000 cm⁻¹ using potassium bromide (300 mg) containing pulverized sample (1.5 mg). These pellets were pressed in a vacuum die at ~680 MPa to produce transparent discs. The prepared discs were immediately measured.

Results

In the present investigation no distinct spectral features are observed in the EPR spectra for G₁ sample before irradiation. This deserves several comments in comparison to the unirradiated G₂ and G₃ samples which possess variant EPR intensity signals as shown in Fig. 1.

Fig. 1

EPR spectra of non-irradiated samples, G₁, G₂ andG₃

Effect of γ-irradiation doses

Gamma-ray irradiation was made for three aliquots of each material in order to examine the change of EPR signal shape and its intensities with irradiation dose. Figure 2a shows the EPR intensity of G₁ as a function of absorbed dose. It is observed that, when the irradiation dose exceeds 10 kGy, an obvious EPR signal (E′-center) was induced. The observed linear increase of EPR intensity with dose indicates propagation in creation of new E′ defects with successive γ-ray doses. Also the results indicate that the EPR signal induced by irradiation for sample G₁ are created by some hole-type centers may be related with NBOHCs. In Fig. 2b, c typical EPR spectra can be observed for the samples G₂ and G₃, respectively, after been irradiated at different doses. It is evident that radiation induced E′-centers as well as a weak satellite (precursor) EPR peaks (P₁ and P₂) are identified as n Fig. 3a, b. It is important to notice that, the EPR intensity for E′ defect center challenges the EPR intensity signal of the precursors. This indicates the negative effect of precursors on the creation of new E′ defects with γ-ray dose.

Fig. 2

EPR spectra at different doses: a G₁ b G₂ and c G₃

Fig. 3

E′-centers and its precursor (P ₁ and P ₂) of irradiated samples at 20 kGy, a G₂, b G₃

Moreover, by comparing the EPR spectra before and after irradiation for G₁, G₂ and G₃ samples (Figs. 1 and 2) it is observed that there are changes of the induced spectral features for G₁, while the signals detected in the fused material (G₂) and natural material (G₃) have nearly the same spectral features. The results show that, the differences of the principle g-value are: ∆g _G1 = 2.01648 ± 0.00099, ∆g _G2 = 2.00574 ± 0.00085 and ∆g _G3 = 2.00577 ± 0.00055. These values are in agreement with those reported for the E′-center in the irradiated commercial a-SiO₂ and it can be hence concluded that the defects induced by irradiation in our investigated materials are of the E′ type [8, 21, 22].

Response curve

Figure 4 illustrates the growth of the E′ concentration with different irradiation doses (1–500 kGy) for the investigated samples. According to the data tabulated in Table 2 and represented in Fig. 4, it can be concluded that, the concentration of E′-centers induced in G₁ are slightly hard to irradiation that means propagation of E′ centers with irradiation is unlimited. However, the amplitude of the EPR spectra from the E′-center for samples G₂ and G₃ come up to a saturated level with a dose of 20 kGy.

Fig. 4

Dose response curves of G₁, G₂ and G₃ samples at different doses from 1 to 500 kGy

Table 2 E′ concentration at different doses for different samples

FTIR bands

Before describing the total FTIR spectra of the investigated samples, it is found fruitful to mention about the formation sites of water or OH groups in glass.

The principal mechanism of interaction of water with silicon-oxygen network is known to involve breaking the Si–O–Si bridges by water molecules dissolved in a melt. An increase in the total H₂O content is assumed to result in a consecutive increase in the fraction of molecular water, whereas the fraction of water in the form of the hydroxyl group is found to reach to saturation when reaching the total H₂O content of 4–6 % [23]. Among properties influenced by the water-related properties, it is the absorption in the near- and mid- IR region of the spectrum that is of principal importance in our work.

The absorbance spectrum of the G₁ a-silica in the region 400–4,000 cm⁻¹ is investigated (Fig. 5). Several bands are observed throughout the 1,500–3,700 cm⁻¹ region which is assumed to be due to the components of the stretching modes of the bound hydroxyl groups in three different structural sites [24]. The peak located at ~853 cm⁻¹ is lower in intensity for sample G₁ while it is very sharp for samples G₂ and G₃ in the FTIR spectrum. This indicates that the Si(OH)₄ is not converted into SiO₂ in the initial state of preparing the (G₁) material. There are Si–OH stretching peaks in the region of approximately 870–950 cm⁻¹. These peaks can be considered as an indication of water in the glass structure and is typically associated with the [SiO₄] in the glass surface [25, 26]. In addition, the presence of adsorbed water is indicated by the broad absorption band at ~1,300–1,440 cm⁻¹ only for glass G₁, which can be assigned to the bending mode of water molecules. The obvious decrease of the broad and stronger absorption bands in the region between about 1,400 and 3,800 cm⁻¹ after the irradiation (Fig. 5b), corresponding to the fundamental stretching vibrations of different hydroxyl groups, may be explained by the condensation of first Si–OH groups leading to Si–O–Si links and second C–OH groups leading to C–O–C bands. In addition, radiation dose (500 kGy) may be is sufficient to evacuate OH groups and the inorganic part becomes sufficiently rigid to form a real trap for the remaining OH groups, preventing their evacuation.

Fig. 5

FTIR spectra, a FTIR absorbance for the native investigated samples, b FTIR absorbance for the G₁ sample before and after irradiation (500 kGy)

The FTIR absorption bands with inherent frequencies 1577, 1790, 2285, 2423 and 1450, 1554, 1661, 1752, 1820, 1935, 2111, 2179, 2285, 2370–2830 cm⁻¹ in the spectra of samples G₂ and G₃, respectively, are assumed to be due to the combination modes and overtones of the silica matrix [27]. In the spectrum of low-water-content silica glasses (G₂) at wavenumbers from 3,000 down to 1,300 cm⁻¹, all bands are commonly assigned to the multiphonon modes of the water-free SiO₂ matrix [28]. For type G₃ sample, some water-related bands are identified. Of these, three bands with frequencies from 3,445 to 3,680 cm⁻¹ are assigned to the stretching modes of hydroxyl groups in two different structural sites. Bands similar in locations of 3,467 cm⁻¹ were reported earlier for high-water content silicate glasses [29]. Commonly, a band around 3,450–3,500 cm⁻¹ of high-water-content silica was assigned to a certain stretching mode of H₂O molecule [23, 30]. Usually, Si–Si bond rocking is observed in silicate glasses within 410–490 cm⁻¹ range [31]. Therefore, the band at 510 cm⁻¹ for sample G₃ can be ascribed to the Si–O–Si bending mode.

Isothermal treatment for native samples

Experimental data (Fig. 1) indicate that the paramagnetic centers are present in our samples before isothermal treatment (preexisting defects) except G₁ sample. The EPR signals of E′-center for the non-irradiated G₁ as a function of the duration times (1–3 h) at 300 °C are given in Fig. 6a. It can be seen that the isothermal-treatment at 300 °C for 1 h shows detectable EPR signal of E′-centers which are gradually decreased with increasing the duration time of isothermal treatment. Also, we can observed that (Fig. 6b, c), after successive isothermal-pulses for glasses G₂ and G₃ of total duration time of 2 h at T = 300 °C the intensity of E′-center are not affected however affect other induced EPR signals of no interest for the present study. However, behind an overall duration time of 3 h, there is a remarkable increase in EPR intensity accompanied with slight shift to high Gaussian value.

Fig. 6

EPR spectra of glasses after been exposed to different times 1, 2 and 3 h at 300 °C before irradiation: a G₁, b G₂ and c G₃

Decay at room temperature

The decay behavior of the EPR spectra for the investigated samples was recorded immediately after the given irradiation dose (20 kGy) and over 120 days (Fig. 7). The recorded results indicate that sample G₁ shows high decay reaching about 40.77 % from its initial value within first 40 days, and then the degree of decay increases (~61.49 %). Similarly, sample G₂ shows low stability within 40 days (the decay of G₂ was about 26.15 % from initial value after that the result show stability (plateau shape), while G₃ shows high stability within the same region.

Fig. 7

Peak-to-peak heights (H _PP) of G₁, G₂ and G₃ as a function of post-irradiation storage time during 120 days after irradiation (20 kGy)

Thermal bleaching

The effects of isothermal treatment on the stability of the irradiated samples (G₁, G₂ and G₃) up to 130 min at 200 °C are observed in Fig. 8. The interplay between generation and annealing gives rise to a zigzag-shaped behavior for sample G₁. However, for samples G₂ and G₃ there are a slight increase at early duration times followed by permanent stability (plateau shape).

Fig. 8

Change of the concentration of E′ centers versus heating time (130 min) for the irradiated samples (20 kGy)

Discussion

Extended studies of irradiated bulk a-SiO₂ have identified two oxygen-associated trapped-hole centers (OHCs), in addition to the generic E′-center. These are the nonbridging oxygen hole center (NBOHC) [32] and the peroxy radical [33]. The latter has been assumed to comprise an O₂ ⁻ molecular ion [33] bonded to a single silicon in the glass network [34]. An apparent analogous of the NBOHC has been described in α-quartz [35]. On the other hand, no obvious counter parts of the E′₂, E′₄ centers have been reported in amorphous silicon dioxide [36].

EPR studies of xerogel a-SiO₂

There is a great deal of interest in producing silica-based glasses because of their wide applications in different important fields [1–6]. As well, materials prepared by sol–gel method could be obtained in homogeneous form with high purity, but, it is very difficult to compare the structure of different sol–gel prepared glasses because of the various experimental conditions and different precursors used.

It is well known that, the E′-defect center is an important class of radiation-induced paramagnetic defects in silica which have g values slightly less than 2.0023 and have long spin–lattice relaxation times and was first reported by Weeks [12] and later by Slisbee [14]. Thus, the anomalous peak at g = 2.01648 (Fig. 2a) which is assumed to be related to the xerogel sample (G₁) may be attributed to overriding the non-bridging oxygen hole centers (NBOHCs) defect, where the OH concentration is considered even in the wet case and, consequently, the peak should be related to the wet variant of the centers produced by radiolysis of OH group and stabilized by the presence of hydrogen. With increasing irradiation dose, the concentration of induced defects is exceeded and a rapid concentration is observed (Table 2) suggesting that the existence of a defect generation process is effective at higher doses. This indicates that the production of the E′ centers is a two-step process, i.e. at low irradiation dose changes NBOHCs into intermediate configurations which are then converted into E′-centers by increasing irradiation doses. Although, the observed weak field splitting at the lower dose (less than 10 kGy) has been explained [37, 38] as the result of the admixture of excited states into the ground state. In addition, the weak irradiation effect on the EPR intensity observed at early irradiation doses can be related to be due to the pertinent radiation levels which are too small to generate new vacancies. However, increasing irradiation doses, new vacancies are consequently generated.

EPR studies of fused and natural a-SiO₂ samples (G₂ and G₃)

In the present work, we report on the generation of E′-centers in oxygen deficient materials. The materials considered here are a-SiO₂. One of these is synthesized by vapor axial deposition technique (G₂), while a second material is commercial a-SiO₂ sample (G₃). After irradiation of the varieties considered, we have estimated a comparable variance in both EPR signal intensities and spin concentrations. Thus, it is most important to examine the behavior of the different paramagnetic centers in fused and commercial a-SiO₂ structures under different irradiation doses.

After inspection of the spectra illustrated in Figs. 1 and 2b, c, an evolution of the EPR lines shape is observed after gamma-irradiation for samples G₂ and G₃. Figure 3a, b are characterized by a central and two outer satellite (precursor) lines. Both outer lines can be ascribed as a spectrum of hyperfine structure (hfs) splitting of the hydrogen nucleus of water in the form of hydroxyl groups [18]. Hfs splitting of such a kind has also been observed by several authors [19, 37, 38].

The main peak is E′, this peak was reported in several investigations concerning magnetic properties of silicon dioxide under peak to peak excitation [39], and assigned to radiative de-excitation of self trapped excitons. However, a detailed investigation of the E′ concentration (Table 2) reached by different irradiation doses (1–500 kGy) shows that, the line shape variation occurs when the E′ concentration is larger than 10¹⁶ spins/cm³ and is slightly enhanced by increasing irradiation. The present results are in general agreements with recent studies about interaction of radiation on pure silica [40, 41].

In early studies [22] of point defects in quartz, interstitial protons (H⁺) and alkalis (Li⁺ and Na⁺) which are known to be present in all quartz species as charge compensators are possible candidates for the ions being moved by γ-irradiation. These interstitial alkali ions (Li⁺ and Na⁺) are initially located adjacent to the ubiquitous Al³⁺ substitutional ions, where they act as charge compensators [22]. Thus, once the sample is irradiated, the interstitial alkali ions are released; they diffuse along the c-axis channels and become firmly trapped at unidentified sites within the lattice. Increasing irradiation frees the interstitial alkalis from these traps and allows them to return to Al³⁺ sites, converts the induced E′-centers back into their precursor form and return the sample to its as-grown state (Fig. 2b, c). These results indicate that movement of the interstitial alkali ion away from the Al³⁺ sites by radiation (up to 20 kGy) accompanies the production of the E′-centers. Increasing irradiation dose (up to 50 kGy) return the interstitial alkali ions to the Al³⁺ sites, and restores the precursor state of the E′-centers.

The slow growth with irradiation indicates that the process which transforms the precursor defect to E′-center may involve the displacement of ions within the lattice. However, the narrow EPR line-widths (G₂), long spin–lattice relaxation times, and small negative g shifts indicate that the E′-centers involve electrons associated with oxygen vacancies, are in direct correspondence to the P₁ and P₂ centers in quartz. Moreover, the observed line broadening in the sample G₃ can be attributed to spin–spin dipolar interaction which competes with the inhomogeneous broadening at high irradiation doses.

As regard to Table 2, the concentration of E′ centers induced in the G₂ sample is observed to be lower than that induced in G₃, reflecting the lower oxygen deficiency of the G₂ with respect to the other varieties considered. This result suggests that the site precursors of the E′ defect is oxygen- deficient. In addition, the strike similarity between the g value of G₂ and G₃ (E′-center) suggests that similar Si-sp³ hybrid orbitals are involved. On the other hand, the orthorhombic components of the E′-center could indicate that a weak interaction of the unpaired electron with the atoms disposed close to the defect occurs.

It is worth to explain the obvious variance in intensity of E′ peak between samples G₂ than G₃ (Fig. 2b, c) is not surprising. Because, both samples are incorporated with traces of metallic impurities such as Fe₂O₃, but, as the sample G₂ (fused a-SiO₂) does not contain any multivalent elements to accept electrons released from Fe²⁺ during oxidation. It may be concluded that oxidation of Fe²⁺ to Fe³⁺ is limited to the surface where oxygen is available. But, through irradiation process the diffusion of oxygen in the sample surface increases leading to remarkable increase of the peak intensity [42].

The EPR data reported in Fig. 2b, c evidence the existence of obvious shift toward high g resonance value at 20 kGy while at 50 kGy the shift disappears for both investigated samples (G₂ and G₃). It is possible to conclude that, at 20 kGy there is shift in the equilibrium concentration of both Fe²⁺ and Fe³⁺ towards the Fe³⁺ state (Eq. 1). However, at 50 kGy the Fe²⁺ formation overcomes the concentration of Fe³⁺

$$ {\text{Fe}}^{ 2+ } + hv \leftrightarrow {\text{ Fe}}^{ 3+ } + {\text{ e}}^{ - } . $$

(1)

A further contribution to this anomalous shift behavior is given by ionization interactions (Fe³⁺ ↔ Fe²⁺) which add a term to the local magnetic field [43]. Moreover, the different shift observed for either E′-center or precursors being due to the dependence of the second order corrections on the square of the hyperfine splitting [44]. Finally, from the consideration of experimental results, it must be emphasized, however, that for g ≈ 2.0 Fe²⁺ does not itself produce the signal, but acts as a center for the paramagnetic defect which is produced on irradiations.

The concentrations of the main E′-defect center and its satellites for fused (G₂) and natural (G₃) samples feature a good radiation resistance because they are unaffected by irradiation dose when exceeding 20 kGy. This finding evidences and confirms the role of the induced precursors, which possess particular resistant to the formation of E′-center during irradiation [45]. Also, it is observed that typically the concentration of the E′-centers induced by γ-irradiation remains nearly constant at high dose, which indicates that, at such high doses there are some sort of equilibrium between generation and annihilation of E′ defect centers.

Response with dose

To have a personal radiation indicator, its response should ideally be proportional to dose. Figure 4 illustrates this response of the G₁ a-SiO₂ samples caused by different irradiation doses (1–500 kGy). As it is shown in Table 2 and Fig. 4 there are an active and pronounced linear increase in spin density values at early irradiation doses (1–200 kGy), but at high irradiation doses (when exceeding 200 kGy) there is a reduction and delay in spin density values. It is evident that the radiation-induced sensitivity (RIS) which corresponds to the increase in spin density values is due to more creation of induced E′-centers with increasing irradiation doses. At high irradiation doses (more than 200 kGy), when the rate and amount of spin density values are reduced, it is unlikely that the formation of NBOHCs site can be assisted by γ-irradiation. Also, we can not ignore the poorness of the precursor lines which enhances the activity of the spin density for G₁ glass.

However, the growth kinetics for samples G₂ and G₃ reported in Fig. 4 evidence the existence of a dose range in which it is possible to induce E′ defect with increasing irradiation. This range is 1–20 kGy, and for these irradiation doses the generation mechanism overcomes the destruction ones, so that a model to describe the irradiation response has to consider both processes. Exceeding this dose (20 kGy), the curves are reaching the equilibrium state with increasing dose.

In spite of the variations in the spin density values, the response curves of Fig. 4 show specific spin density values increasing linearly with the absorbed doses, resulting in response curve of xerogel a.SiO₂ glass (G₁) of adequate behavior for dosimetric purposes. However, G₂ and G₃ samples can be consider as a good candidate for radiation shielding.

Effect of isothermal-treatment

The room temperature structure of a-SiO₂ is known to change subtly with either irradiation or thermal treatment [46]. We have used FTIR spectroscopy (Fig. 5) as a sensitive probe to identify the effect on structure of both water content and thermal history. Also, we have carried out an investigation of the effect of these two parameters on point defects structure induced through isothermal-treatment for those defects which can be seen by EPR spectroscopy.

The effect of isothermal-treatment on the EPR intensity of the prepared pure silica gel derived sample (G₁) is shown in Fig. 6a. These observations might be interpreted by assuming that, at lower isothermal-treatment (1 h) it rests some residual water and hydroxyl groups which prevent the density to increase, allowing to inducing NBOHCs defects. Additional thermal-treatments at 300 °C were able to gradually loose some of residual water and hydroxyl groups from the gel which turn out to be more densified. This gives rise to the structure collapses, and the pores are expected to be closed and fused together, and a smooth material was obtained. This morphology was assumed to condensation, strengthening and densification of the gel at higher temperature [47]. With increasing the duration time (3 h), the water content as well as solvent molecules showed drastic decrease giving rise to a relatively fine and dense material, allowing the concentration of bridging oxygens to overcomes the concentration of NBOHCs.

Because there was insufficient time during deposition to carry out the very slow relaxation processes for natural sample (G₃), accordingly, there are significant structural differences (hydroxyl groups and water molecules traces) between samples G₂ and G₃ as seen in Fig. 5. Consequently, it would be interesting to compare the EPR spectra of a low water content sample (G₂) with those of the dry oxide of Si (G₃).

The anomalous trend of the EPR signal intensity when heat-treating both G₂ and G₃ samples for 1 and 2 h at 300 °C involves a stationary effect. This phenomenon is explained assuming a movement of an oxygen atom process out from its regular site of a-SiO₂ which increases the thermal stability of E′ defect center causing the observed stationary phenomenon. This lower sensitivity may indicate that the overlapping signal is related to the presence of oxygen vacancies which annealed when samples are heat-treated at this duration time (2 h). Though, the observed increase in EPR intensity after successive treatment of total duration 3 h at T = 300 °C, may be assumed to the transformation of some diamagnetic center to paramagnetic centers, as well activating the spins of numerous preexisting E′ defects at such isothermal-treatment (3 h). Although, the smaller shifts of the EPR lines are consistent with the expected rigidity of the regular structure and/or with their assignment to regular structure in the otherwise more disordered network. Also, the broad lines shift in various directions that are consistent with reduction in the Si–O–Si angles of the glass matrix.

Environmental and thermal bleaching of E′ defect center

Environmental bleaching

It is of great importance from the dosimetric point of view to investigate the time dependence of radiation induced free radicals. As shown in Fig. 7, the investigated samples after been irradiated up to 20 kGy were used to check stabilities of radiation induced point defects during 120 days at room temperature (RT) ~25 °C. We have reproduced EPR data immediately after been irradiated, where a large concentration of E′ defects exist. Throughout 120 days, there is a fast decay in EPR signal intensity during the first 40 days followed by slight decay for sample G₁. In order to explain this phenomenon, it may be assumed that the damage process in this sample consists of primary formation of highly mobile electrons and holes caused by irradiation. These mobile electrons then either produce an electron scavenger or become localized or solvated electrons are responsible for the observed early fast decay after irradiation, which is reasonably well correlated with holes trapped on nonbridging oxygen atoms. Trapping of holes means that the Si ions are no longer bound by Columbic forces to oxygen atoms and are diffused away and trap electrons.

Likewise, Fig. 7 illustrates an obvious decay for the first 30 days followed by some sort of stability in EPR signal intensity for samples G₂ and G₃. This observed behavior can be interpreted by recalling that, oxygen vacancies are present in the a-SiO₂ network prior to irradiation and ~90 % of them have local geometries that favor an E′-center. In addition, the energy levels of these defects are shallow so that they are first to capture holes that are generated by irradiation. Hence, immediately after irradiation, the predominant defect is E′, and, the holes in the shallow levels of E′-centers are re-emitted and reabsorbed gradually. Some of these migrating holes get captured by the deeper states of vacancies that become precursor centers. In addition, the neutralization occurs by a Si electron jumping to an ‘electron level’ of the defect which can be assumed to be an oxygen vacancy. The neutralization process creates a dipole which is then destroyed when the electron jumps back to the matrix under reverse bias. The experimental results verified the existence of just such a level and the formation of meta-stable dipole [40]. Furthermore, it is well known that [ 47 ], the number of preexisting E′ defects in high water content materials are very poor when compared with the number created by γ-ray doses, and relaxes more rapidly to its new equilibrium defect concentration than does a sample with low [OH]. This phenomenon can explain the variant behaviors in both response and fading curves for the three investigated samples. Generally, the EPR signals for entire defect centers are assumed to decay with time. This indicates that, the decay has converted part of the centers to an unidentified non-paramagnetic form, but not back to the initial defect.

Thermal bleaching

The EPR signal intensity for sample G₁ (Fig. 8) is observed to increase in the early time of isothermal-treatment and could be explained as being the result of the accumulation of paramagnetic centers, The annealing also brought about a noticeable change in the shape of the E′-center consistent with an evolution of the surviving centers, and/or at this isothermal annealing, all the intermediate configurations are converted to E′ defect center through this interval period (20 min). However, after 40 min, these centers become unstable, later then; their influence is negligible due to spin conversion of paramagnetic centers created by γ-irradiation on the surface of amorphous SiO₂ grains. Adding together in high-OH silica such as sample G₁, the rapid annealing of E′ has been attributed to a reaction of these defects with radiolytic atomic hydrogen [48], which also anneals out in this temperature range.

However, the results given in Fig. 8 show that the E′-centers for samples G₂ and G₃ are more stable than sample G₁, which make them good candidate as glass shielding material for radiation processing. This indicates that, the annealing of E′ centers involve two step processes, i.e. the early isothermal-treatment changes precursor defects into intermediate configuration which are then converted to E′-centers, and, at this duration of isothermal treatment the interstitial alkali ions are released causing the slight increase observed in the EPR signal intensity (during the first 20 min). The striking stability feature of G₂ and G₃ is pertinent to the presence of water-content in these samples, but due to the presence of appreciable H₂O-content in sample (G₂), the formation of E′ defect centers is reduced [49]. We can deduce this relationship from the observation of the hydroxyl group and molecular water illustrated in Fig. 5. Beside that, these faint changes in thermal bleaching process indicate that, the process of transformation of precursor defects may involve displacement of ions within the network.

It can thus be assumed that, the thermal bleaching mechanisms which govern the various anneals curves of Fig. 8 indicate the contrasting behaviors of radiolytic atomic hydrogen in high-OH silica [48] and of peroxy radical in low-OH materials [32]. In high-OH silica, it is proposed that the displaced oxygen reacts with radiolytic hydrogen to form water. It is the diffusion of these interstitial water molecules which governs the thermal bleaching of E′ [36].

Conclusions

• The concentration of E′-centers is observed to increase with increasing irradiation dose (1–500 kGy) for sol–gel sample (no saturation effect) i.e. slightly radiation hard.

• It is interested to note that the precursors of the investigated materials might be particular resistant to the formation of E′-center by ionization as in G₂ and G₃.

• Isothermal treatments of the samples are found to be able to induce differences in the EPR features.

• FTIR study elucidates the bonding system of the constituent atoms and groups such as Si, O and OH that throw light to the expected structure intense.

• These finding evidences that the investigated samples G₁ may be used as radiation indicator, as well as G₂ and G₃ are good candidate for radiation shielding purposes.