Introduction

As a result of traditional disposal practices of agricultural, industrial and sewage wastes, there has been an increase in the toxicity of water bodies which affects both human and wild health. Most industrial wastes containing toxic compounds are discharged into natural water bodies without adequate treatment [1,2,3,4,5,6,7] Phenol-based compounds are the chemical residues of byproducts/wastes of several industries such as petroleum refineries, paper mills, pharmaceuticals, pesticides, coke plants, polymeric resin, coal conversion and paint [8, 9]. Therefore, numerous studies have been investigated for the treatment of phenolic wastes before their release in the nature [10, 11]. The literature reports on three techniques for the elimination of phenols from waters: biological, physical and chemical. The most widely applicable techniques are coagulation, precipitation, ion exchange, filtration, adsorption, solvent extraction and advanced oxidation processes (AOPs) [12,13,14,15,16]. Several studies have been extensively highlighted the use of the photocatalysis process, one of the AOPs, in hydrogen production and water purification [17,18,19]. It is especially adequate for the depollution of low liquid volumes containing low pollutants concentrations. This technique is based on the efficiency of a semiconductor, which could be photo-illuminated with energy greater than its band-gap producing electron–hole (\({\text{e}}^{ - } - {\text{h}}^{ + }\)) pairs. These latest can emigrate to the semiconductor surface generating radicals that can initiate redox reactions [20]. The effectiveness of these reactions is related to the advantage of the high reactive radicals \(^{ \cdot } {\text{HO}}\) in driving oxidation processes, which are convenient for performing complete degradation and mineralization of even most less and persistent reactive compounds.

The synthesis of porous photocatalysts assembled, or immobilized by nanostructured semiconductors, might be a tricky manner to design porous materials with higher photo-activity properties. This could be the case when fixing a semiconductor onto solids such as zeolites, diatomite, activated carbon, kenyaite or clays [17, 21,22,23,24]. Clay minerals are promising photocatalyst supports because of their good adsorption capacity, their high mechanical and chemical stability, which reduces the aggregation of photocatalyst nanoparticles, leading to increased photocatalytic efficiency. Numerous layered clay-based catalysts have been extensively investigated such as modified organoclays, acid-modified clays, semiconductor-clay heterojunctions, pillared clays and delaminated clay heterostructures, characterized by higher quality porous morphologies [25,26,27].

The noble metals (e.g., Au, Pt, Pd, Ag, Cu) have received much interest for various applications in biotechnology, catalysis and environment [28,29,30,31]. It was previously reported the uses of such metals as photodeposition tools, more precisely the photodegradation of dyes, acids, phenol [32,33,34,35] and NO [36], and the H2 production [37,38,39]. Among noble metals selected as electron traps, silver turned out to be highly beneficial for industrial applications, due to its low cost and easy preparation [40]. Silver-deposits on TiO2 have been shown to mineralize mono-, di- and poly-carboxylic acids [41] and remove the 2-propanol [42], and urea and chloroform [43]. Simultaneously, the nano size of silver particles, loading amount, preparation method, nature, pH and pollutant concentration influence the photocatalytic performance of silver-semiconductor nanomaterials [44]. Heretofore, silver-TiO2 nanocomposites have been prepared by various chemical and physical techniques, like magnetron sputtering, electron beam evaporation, molecular precursor, sol–gel and photo-deposition [45,46,47].

Pyrophyllite, an aluminum silicate hydroxide [Al2(OH)2Si4O10], is a layered clay with crystal structure (Fig. 1) consists of an octahedral Al–O sheet sandwiched between two tetrahedral Si–O sheets, with each octahedral aluminum bonded with tetrahedral silicon by apical oxygen and with adjacent aluminum by two hydroxyl groups [48]. Pyrophyllite has been widely used in many industries, including refractory bricks, plastics, papers, and ceramics [49]. In addition, several studies have been conducted on use this clay as an adsorbent for various pollutants in aqueous medium [50,51,52,53].

Fig. 1
figure 1

Structure of pyrophyllite

Full size image

As to continue the research in the field of the pyrophyllite-nanocomposites, we report in the present paper on the sol–gel synthesis of the TiO2 such nanocomposites and their photocatalytic performance. The prepared photocatalysts (SG and 1.5%Ag/SG) were tested in the photodegradation of phenol under UV irradiation. Moreover, the optimization of parameters (catalyst amount, pH, temperature, and H2O2 concentration) by response surface methodology in the phenol degradation with 1.5%Ag/SG was also studied.

Experimental

Materials

The local clay used in this paper has previously been investigated [54,55,56,57,58,59,60,61]. It is gray type clay and mainly composed of pyrophyllite-2M [Al2(Si4O10)(OH)2], with weight composition: SiO2 58%wt, Al2O3 26%wt, Fe2O3 2.7%wt, K2O 1.6%wt, Na2O 1.5%wt, MgO 0.8%wt, TiO2 0.6%wt and CaO 0.12%wt. It was used in the present work after acid decarbonasition as follows [62]: (1) solution made up of 100 mL hydrochloric acid (0.5 M) and 2 g of raw clay was aged for 1 h at room temperature. (2) After several washes, the resultant material was dried at 110 °C for 24 h and sieved with a 63 μm mesh (we refer to it below as PT). All chemicals that have been used in this present work are Sigma-Aldrich products: Titania powder (AEROXIDE® TiO2 P25), titanium (IV) t-butoxide 97%wt, phenol 99.5%wt, ethanol 99.5%wt, sodium hydroxide 98%wt, hydrochloric acid 37%wt, muconic acid 97%wt, formic acid 98–100%wt, maleic acid 99%wt, hydrogen peroxide 30%wt, and acetonitrile 99.8%wt. Milli-Q water was used in all the processes.

Preparation of the photocatalysts

SG nanocomposite material was synthesized via sol–gel technique as follows: (1) solution made up of 40 mL ethanol and 17 mL titanium (IV) t-butoxide Ti(OCH2CH2CH2CH3)4 was dropwise added to another solution of 20 mL ethanol in which 2 g decarbonated clay was dispersed. (2) 30 mL ethanol diluted with 20 mL Milli-Q water was dropwise added until the formation of a dense gel. (3) The resultant gel was aged for 24 h. (4) After washing cycles, the gel was dried at 100 °C for 24 h. (5) The dried gel was sieved (63 μm) and then calcined at 750 °C for 4 h.

Silver was deposited into SG by photo-irradiation using a mercury UV lamp (400 W), isopropanol as a sacrificial reagent and silver nitrate as a source of Ag+ ions. Photodeposition of 1.5%wt of Ag during 6 h was carried out in an anoxic medium achieved with a constant N2 flow. This photocatalyst is referred to hereinafter as 1.5%Ag/SG.

Characterization and analytical techniques

X-ray diffraction (XRD) patterns were recorded on a diffractometer (Siemens D-500, λCu Kα = 1.5418 Å) in the region of 5–80°. The morphology of the materials was performed by scanning electron microscopy (SEM) using a JEOL JSM-5400 microscope coupled with energy-dispersive x-ray spectroscopy (EDX). BET specific surface area was measured by N2 adsorption at − 196 °C using a Micromeritics ASAP 2010 system. X-ray photoelectron spectroscopy (XPS) characterization was conducted on a spectrometer (Leybold-Heraeus LHS-10). UV–Vis diffuse-reflectance spectrophotometry (UV–Vis DRS) analysis was performed in the range of 250–800 nm through a spectrophotometer (Varian Cary 5 UV-NIR) using polytetrafluoroethylene as reference. Concentrations of phenol and intermediate products were measured by high-performance liquid chromatography (HPLC) by means of a C18 column (0.25 m × 0.046 m ID, 5*10–6 m particles), detector (UV, λ = 270 nm) and water-acetonitrile (70:30) as mobile phase. The determination of total organic carbon (TOC) was performed using Shimadzu TOC-VSCN analyzer. The toxicity of treated solutions was investigated by the Microtox toxicity test (standard ISO 11348-2:2007) as the inhibition of bioluminescence of Vibrio fischeri marinebacteria. The inhibition below 10% indicates a solution is nontoxic.

Photocatalysis experiments of phenol

The experiments were performed in a 500 mL Pyrex cylindrical reactor. 200 mL phenol solution at 50 mg L−1 and study amounts of photocatalyst were stirred and continuously bubbled by air (air flow = 400 mL min−1). The reactor was irradiated from the outside using a UV lamp (Philips Solarium HB175) equipped with four fluorescent tubes (15 W, 365 nm).

Analysis of response surface modeling

Central composite design (CCD) was postulated to specify the operating conditions for optimizing the phenol degradation yield. The interaction between dependent and independent variables was studied following the second-degree polynomial:

$$\begin{aligned} Y & = \, \alpha_{0} + \, \alpha_{1} X_{1} + \, \alpha_{2} X_{2} + \, \alpha_{3} X_{3} + \, \alpha_{4} X_{4} + \, \alpha_{12} X_{1} X_{2} + \, \alpha_{13} X_{1} X_{3} + \, \alpha_{14} X_{1} X_{4} + \, \alpha_{23} X_{2} X_{3} + \, \alpha_{24} X_{2} X_{4} \\ & \quad + \alpha_{34} X_{3} X_{4} + \alpha_{11} X_{1}^{2} + \, \alpha_{22} X_{2}^{2} + \, \alpha_{33} X_{3}^{2} + \alpha_{44} X_{4}^{2} \\ \end{aligned}$$
(1)

where Xi denotes the independent variables (catalyst amount, pH, temperature and H2O2 concentration), Y is the predicted response (phenol degradation yield), αi is the linear effect, αij is the interaction effect, αii is the quadratic effect and α0 is a constant offset term.

The Design Expert Software Version 11.0 (State-Ease, Minneapolis, USA) was used to analyze the experimental results [63].

Results and discussions

Characterizations

Figure 2 presents the diffractograms of the clay PT and photocatalysts. The PT sample exhibited (Fig. 2a) characteristic peaks of pyrophyllite phase Al2(Si4O10)(OH)2 at 2θ = 9.62, 12.37, 20.85, 21.29, 26.63, 29.11, 34.96, 47.99, 56.43, 59.95, 62.37, 68.48 (Ref. code: 00-046-1308) and characteristic peaks of quartz impurities at 2θ = 26.63, 36.53, 39.46, 42.44, 45.79, 50.13 (Ref. code: 01-089-8934), in good agreement with the literature [64,65,66]. The diffractogram of photocatalysts (Fig. 2b) shows anatase TiO2 peaks at 2θ = 68.8°, 63.2°, 62.7°, 55.1°, 53.9°, 48.1°, 37.9° and 25.3° (JCPDS No. 21-1272) [67,68,69,70,71,72,73,74,75,76,77,78], rutile TiO2 peaks at 2θ = 56.6°, 54.3°, 36.1° and 27.4° (JCPDS No. 75-1753) [72] and characteristic clay’s peaks; but otherwise, no crystalline phases that might result from silver was observed in the 1.5%Ag/SG catalyst, which might be assigned to the low-doped silver [40, 74].

Fig. 2
figure 2

XRD diffractograms of PT (a) and investigated photocatalysts (b)

Full size image

Table 1 summarizes the surface area and the ratio distribution of the TiO2 Anatase–rutile forms in terms of particle size and composition for the photocatalysts. The results show that PT and SG have a surface area of 9.14 m2 g−1 and 16.6 m2 g−1, respectively. This increase in the surface area indicates the successful pillaring of the TiO2 nanoparticles into the PT layers. In fact, the synthesis of SG was carried out in an aqueous medium containing titanium (IV) t-butoxide and PT, so the Ti ions were transformed into Ti(OH)4 intermediate. After the calcination step at 750 °C, the Ti(OH)4 was transformed into TiO2. Based on this transformation, we assumed that the ion exchange process was first initiated in the PT interlayers and, therefore, that TiO2 pillaring is possible within the layered PT. Indeed, the exchangeable cations in the PT interlayers were substituted by Ti4+ ions which were hydrolyzed to Ti(OH)4 and then converted to TiO2 after calcinations [79]. In addition, photodeposition negatively affected the surface area and crystal size of SG, probably due to the additional incorporation of Ag particles into the SG pores. The synthesis of SG and 1.5%Ag/SG implies the generation of mesopores as a significant issue, which can be confirmed by the important participation of mesoporosity to the external surface area attributing non-microporous part (Table 1).

Table 1 Anatase–rutile ratio, crystalline size, surface area, and band gap of photocatalysts
Full size table

The mesopores are essentially distributed within the diameter range of 3.5–4.5 nm (Fig. 3a). The porous structure of the SG and 1.5%Ag/SG, which is due to the clay layers [75], was also proven by the N2 adsorption–desorption isotherms represented in Fig. 3b. Both photocatalysts showed a type II isotherm according to the IUPAC classification with a H3 hysteresis loop [76, 77]. This type of isotherm has been discussed in several investigations, described by porous materials due to a house-of-cards structure produced by the plate-like aggregates [69, 78].

Fig. 3
figure 3

Pore size distribution (a) and N2 adsorption–desorption isotherms of prepared photocatalysts (b)

Full size image

To confirm that the TiO2 crystals and Ag particles were well dispersed in the photocatalysts, SEM analysis (Fig. 4) was performed. The SEM image of PT (Fig. 4a) shows that it is layered clay, while the SEM image of SG (Fig. 4b) insures that TiO2 crystals were incorporated into the clay layers. Figure 4c illustrates the Ag and Ti SEM mapping images and semiquantitative EDX analysis of 1.5%Ag/SG. It can be seen that Ti and Ag were homogeneously dispersed along the surface. Moreover, The EDX shows good photodeposition of Ag metal on SG as confirmed by its weight percentage very near 1.5%.

Fig. 4
figure 4

SEM photographs: a PT and b SG; c EDX mapping image of 1.5%Ag/SG

Full size image

The XPS analysis was performed for 1.5%Ag/SG in order to specify the oxidation state of Ti, Ag and O (Fig. 5). The Ag 3d XPS spectrum (Fig. 5b) of 1.5%Ag/SG shows two peaks at 373.8 eV and 367.8 eV attributed to the Ag 3d3/2 and 3d5/2 core levels, respectively [44, 70, 80]. The gap energy separation between Ag 3d3/2 and 3d5/2 is equal to 6.0 eV demonstrating the presence of silver in the chemical state of Ag° [81]. The absence of Ag+ ions indicates their reduction to Ag (Eq. 2) [70, 80, 81]. The peaks of Ag 3d shift negatively compared to bulk Ag° ones (374 or 374.3 eV for Ag 3d3/2 and 368 or 368.3 eV for Ag 3d5/2), indicating that the electrons may migrate from TiO2 to Ag, which confirms that there is a strong TiO2–Ag interaction in the interface of the photocatalyst [82, 83]. As shown in Fig. 5c, the Ti 2p XPS spectra of 1.5%Ag/SG with two peaks at binding energies of 458.0 eV and 463.8 eV, correspond to the Ti 2p3/2 and Ti 2p1/2 orbits, respectively [69]. The gap of 5.8 eV between these orbits confirms the existence of the Ti4+ oxidation state reported in TiO2 [44, 74, 84]. From Fig. 5d, the O 1 s XPS spectra of 1.5%Ag/SG show a peak located at 529.7 eV which is ascribed to lattice oxygen (Ti–O bond) comes from TiO2 [84].

$$({\text{SG}},{\text{ Ag}}^{ + } ) \, + {\text{ e}}^{ - } \to { 1}.{5}\% {\text{Ag}}/{\text{SG}}$$
(2)
Fig. 5
figure 5

XPS analysis of 1.5%Ag/SG: a Full spectrum, b Ag 3d XPS spectrum, c Ti 2p XPS spectrum, d O 1 s XPS spectrum

Full size image

UV–Vis DRS analysis of the photocatalysts (Fig. 6, Table 1) was performed and their band gap values were determined using Tauc’s plot following this formula (αℎv)2 = A(ℎv − Eg) [15]. The difference observed between band gap energies of SG (3.17 eV) and 1.5%Ag/SG (3.12 eV) is not relevant. This shifting in band gap is maybe due to the metallic gray color of 1.5%Ag/SG sample [85] or the creation of lattice plasmons [86,87,88]. In addition, a particular feature to note in the UV–Vis DRS spectra in Fig. 6 is the rise of baseline of 1.5%Ag/SG’s spectrum with respect to that of SG, probably due the presence of silver plasmons.

Fig. 6
figure 6

UV–Vis DRS spectra of investigated photocatalysts

Full size image

Photo-activity of sol–gel nanocomposites

The photocatalyst amount in these tests was fixed at 1 g L−1. Before testing SG, 1.5%Ag/SG and the reference AEROXIDE® TiO2 P25, preliminary experiences of the adsorption for 1 h were carried out (Fig. 7). Apparently, the maximum equilibrium time was around 30 min; 1.5%Ag/SG and AEROXIDE® TiO2 P25 exhibited low adsorption efficiency, while SG exhibited the highest one (5.3%), which may be explained by the clay content (37.3%).

Fig. 7
figure 7

Profile of removal efficiency of phenol by adsorption onto photocatalysts at [catalyst] = 1 g mol−1, room temperature, and not adjusted pH

Full size image

Figure 8 shows the profile of phenol degradation and mineralization and the plot of Ln(C/C0) versus time with the investigated photocatalysts. It can be seen that SG showed 91% degradation and 81% mineralization. The photocatalytic test with this photocatalyst showed an activity ~ 1.5-times better than that with AEROXIDE® TiO2 P25. In fact, the plot of Ln(C/C0) versus time (Fig. 8b) gives respectively the kapp values of 0.019 min−1 and 0.014 min−1. The main explanation for the high SG activity is probably related to the structural modifications due to TiO2 that act to promote the mobility of charge separation in the interstitial interior in order to promote the photo-oxidation process by avoiding recombination between the loaded TiO2 and PT [89]. Doping of SG with 1.5% of silver enhanced the degradation efficiency from 91 to 95% with kapp value of 0.022 min−1 and the mineralization from 81 to 83%. The increased activity of 1.5%Ag/SG was caused by the photo-deposition of Ag, which increases the separation of generated electron–hole pairs, thus avoiding their recombination [74] (Scheme 1). In order to improve the photo-activity of 1.5%Ag/SG, we optimized the operating parameters (catalyst amount, pH, temperature, H2O2 concentration) in the next section.

Fig. 8
figure 8

Profile of degradation and mineralization of phenol versus time over photocatalysts at [catalyst] = 1 g mol.−1, room temperature, and not adjusted pH (a); Plot of Ln(C/C0) versus time (b)

Full size image
Scheme 1.
scheme 1

Diagram of the electron–hole pairs separation and corresponding redox reactions in the 1.5%Ag/SG nanoparticles

Full size image

Optimization of phenol photodegradation with 1.5%Ag/SG

The CCD model was adopted for investigating the optimization of the individual effects of the parameters screened concerning the photodegradation of phenol by UV–1.5%Ag/SG system. Table 2 summarized the factors, ranges and levels of selected variables.

Table 2 Coded and real values of the parameters for CCD design
Full size table

Based on the experimental design matrix proposed by CCD design (Table 3), the regression between the phenol degradation efficiency, coded variables, polynomials and corresponding interactions yields a second degree polynomial model (Eq. 3) which shows that the terms X1, X3, X4, X1X2 and X1X3 affect it positively, while X2, X3 X4, X2X4, X2X3, X1X4,, X42, X32, X22 and X12 affect it negatively. Catalyst amount (X1) has a stronger effect between studied variables.

$$\begin{aligned} {{Y}} & = { 78}.{73 } + { 14}.{{56 X}}_{{1}} - { 1}.{{52 X}}_{{2}} + { 7}.0{{1 X}}_{{3}} + { 7}.{7}0{{ X}}_{{4}} + { 2}.{{57 X}}_{{1}} {{X}}_{{2}} + \, 0.{{17 X}}_{{1}} {{X}}_{{3}} - { 2}.{{69 X}}_{{1}} {{X}}_{{4}} \\ & \quad - {1}.{{69 X}}_{{2}} {{X}}_{{3}} - \, 0.{{26 X}}_{{2}} {{X}}_{{4}} - \, 0.{{96 X}}_{{3}} {{X}}_{{4}} - { 1}.{{75 X}}_{1}^{2} - { 2}.{{92 X}}_{2}^{2} - { 4}.{{57 X}}_{3}^{2} - { 4}.{{28 X}}_{4}^{2} \\ \end{aligned}$$
(3)
Table 3 Design matrix with the experimental and predicted responses of phenol degradation (%)
Full size table

To confirm the quality of the fit for the determined coefficients, analysis of variance (ANOVA) was considered and the results are summarized in Table 4. According to such analysis, the low p value of (< 0.0001) and the F-value of 73.31 are higher than critical F values for a 95% confidence level (F9.10 = 3.02 or F9.7 = 3.8), suggesting that the chosen model fits well [90,91,92,93,94].

Table 4 Analysis of variance (ANOVA) for CCD design
Full size table

Additionally, it can be seen that all terms, except X1X3 and X2X4, such as linear (X4, X3, X2 and X1), interactions (X3X4, X2X3, X1X4 and X1X2) and quadratic (X12, X22, X32 and X42) are statistically significant (P value less than 0.05) [95]. The quality of the fit of the regression model can be confirmed by the coefficient of determination (R2), which suggests that the variations in the degradation efficiency are considered by the model. Adjusted R2 (R2adj) is also an indication of the quality of fit. It changes the R2 considering the number of predictors in the model [26]. The fit statistics results are described in Table 5. The obtained R2 and R2adj were high and very close, justifying that the model was well adjusted to the experimental data [27, 53].

Table 5 Fit statistics results
Full size table

The 3D and 2D response surface plots were used to show information on the effect of interactions on the % phenol degradation. The response surface plot based on interaction X1X2 (Fig. 9) has been selected in fact that its effect is positive and significant according to the model equation and ANOVA test, respectively. Thus, the optimum conditions for phenol degradation under the postulated model are described in Table 6.

Fig. 9
figure 9

The 3D (a) and 2D (b) response surface plots for the interactive effect of pH and catalyst amount

Full size image
Table 6 Optimized parameters obtained from response surface analysis
Full size table

Under these optimum conditions, an experimental experience was carried out (Fig. 10). It can be seen that the degradation efficiency was 100% with mineralization of 91% for 120 min of reaction (Fig. 10a). Additionally, the treated solution was not toxic based on the bioluminescence inhibition test of Vibrio Fischeri bacteria, it showed 5.8% of inhibition which is less than the normalized one of 10%. The analysis of the curve ln(C0/C) = f(t) under optimized conditions (Fig. 10b) showed Kapp of 0.08 min−1, which is ~ 4-times faster than that found before optimization (0.022 min−1).

Fig. 10
figure 10

At optimized parameters ([catalyst] = 2 g mol−1, pH = 5, T = 40 °C, and [H2O2] = 8 mmol L−1): profile of degradation, mineralization and toxicity (a) and plot of Ln(C/C0) versus time with 1.5%Ag/SG (b)

Full size image

Figure 11 shows the evaluation of degradation and mineralization at optimal conditions without photocatalyst during 120 min of irradiation. It can be seen that the photolysis of H2O2 molecules was not negligible, but has not mainly affected the above optimized degradation, with respect to the obtained percentages of degradation (17%) and mineralization (8%). These results further suggest that 1.5%Ag/SG is very active in the UV degradation of phenol.

Fig. 11
figure 11

Profile of degradation and mineralization with UV/H2O2 system, at pH = 5, T = 40 °C, and H2O2] = 8 mmol L−1

Full size image

Figure 12 presents the profile of intermediates concentration (hydroquinone, p-benzoquinone, catechol, maleic acid, formic acid, muconic acid) at optimum conditions with 1.5%Ag/SG during 120 min of irradiation. The concentrations of the intermediates increased with irradiation time until reaching an optimum, after which they started to decrease. This evolution can be attributed to the continuous radical degradation of phenol until mineralization to H2O and CO2. The aromatic intermediates resulting from the degradation of phenol (catechol, hydroquinone, p-benzoquinone) can be converted into carboxylic acids like oxalic, formic, maleic and acetic [96, 97]. The final concentration of intermediates was very low, thus confirming the non-toxicity and mineralization of the treated solutions. Scheme 2 illustrated the proposed pathway of studied phenol degradation.

Fig. 12
figure 12

Profile of intermediates concentration at optimized parameters

Full size image
Scheme 2.
scheme 2

Proposed pathways for photodegradation of phenol over 1.5%Ag/SG

Full size image

The investigated phenol degradation at optimized conditions might be described in the presence of silver-TiO2 particles and H2O2 agent for which the following mechanism may thus be suggested:

$${\text{Ag}}/{\text{SG}} + h\nu \to {\text{Ag}}/{\text{SG}}\left[ {{\text{h}}_{{{\text{VB}}}}^{ + } ,{\text{ e}}_{{{\text{CB}}}}^{ - } } \right]$$
(4)
$${\text{H}}_{{2}} {\text{O }} + {\text{ h}}^{ + } \to {\text{ OH}}^{ \cdot } + {\text{H}}^{ + }$$
(5)
$${\text{O}}_{{2}} + {\text{ e}}^{ - } \to {\text{ O}}_{2}^{ - \cdot }$$
(6)
$${\text{O}}_{2}^{ - \cdot } + {\text{ H}}^{ + } \to {\text{OOH}}^{ \cdot }$$
(7)
$${\text{Ag}}/{\text{SG}} + {\text{ H}}_{{2}} {\text{O}}_{{2}} { \leftrightarrows }{\text{Ag}}/{\text{SG}}({\text{HO}}_{2}^{ - } ) \, + {\text{ H}}^{ + }$$
(8)
$${\text{Ag}}/{\text{SG}}({\text{HO}}_{2}^{ - } ){ \leftrightarrows }{\text{Ag}}/{\text{SG}}({\text{OH}}^{ - } ) \, + { 1}/{\text{2 O}}_{{2}}$$
(9)
$${\text{Ag}}/{\text{SG}} + {\text{ H}}_{{2}} {\text{O}}_{{2}} { \leftrightarrows }{\text{Ag}}^{ + } /{\text{SG}} + {\text{ OH}}^{ \cdot } + {\text{ OH}}^{ - }$$
(10)
$${\text{Ag}}^{ + } /{\text{SG}}({\text{OH}}^{ - } ){ \leftrightarrows }{\text{Ag}}/{\text{SG}} + {\text{ OH}}^{ \cdot }$$
(11)
$${\text{Ag}}^{ + } /{\text{SG}}({\text{HO}}_{2}^{ - } ){ \leftrightarrows }{\text{Ag}}/{\text{SG}} + {\text{ OOH}}^{ \cdot }$$
(12)
$${\text{OH}}^{ \cdot } + {\text{ H}}_{{2}} {\text{O}}_{{2}} { \leftrightarrows }{\text{H}}_{{2}} {\text{O }} + {\text{ OOH}}^{ \cdot }$$
(13)
$${\text{OH}}^{ \cdot } + {\text{ OOH}}^{ \cdot } { \leftrightarrows } {\text{H}}_{{2}} {\text{O }} + {\text{ O}}_{{2}}$$
(14)
$${\text{H}}^{ + } + {\text{OH}}^{ - } { \leftrightarrows }{\text{H}}_{{2}} {\text{O}}$$
(15)
$${\text{C}}_{{6}} {\text{H}}_{{5}} {\text{OH}}\left( {{\text{OOH}}^{ \cdot } /{\text{OH}}^{ \cdot } } \right) \, \to {\text{ oxidation products }} \to {\text{ H}}_{{2}} {\text{O }} + {\text{ CO}}_{{2}}$$
(16)

Conclusion

The heterogeneous photocatalysis process was investigated in the degradation of phenol under UV irradiation using sol–gel nanocomposites as photocatalysts. Natural pyrophyllite and titanium (IV) t-butoxide were used as support and precursor, respectively. The TiO2-clay nanocomposite (SG), the silver-photodeposited nanocomposite 1.5%Ag/SG and the reference AEROXIDE® TiO2 P25 were characterized by XRD, XPS, UV–Vis DRS and BET surface area. The degradation and TOC mineralization with SG at catalyst amount of 1 g L−1 and phenol concentration of 50 mg L−1 was higher than that found with AEROXIDE® TiO2 P25. The photocatalytic performance of the studied reaction was enhanced using 1.5%Ag/SG; in fact, the degradation and TOC mineralization efficiencies were increased from 91 and 81% to 94.2 and 82.3%, respectively. 4-level composite central design (CCD) in response surface methodology (RSM), was used to optimize the degradation of phenol with 1.5%Ag/SG. Various process parameters were optimized by using CCD, viz., catalyst amount (X1: 0.5–2.5 g L−1), solution pH (X2: 2–10), temperature (X3: 20–60 °C), and Hydrogen peroxide concentration (X4: 4–12 mmol L−1). According to CCD design, the second-degree polynomial model correlates the parameters to % degradation of phenol. The significance of this model and regression coefficients was examined by the analysis of variance (ANOVA). This analysis indicated that the selected model was well fitted and correlated to experimental data (p value < 0.001, F-value = 73.31, R2 = 0.9959). The optimum parameters were photocatalyst amount of 2 g L−1, solution pH of 5, and the temperature of 40 °C and hydrogen peroxide concentration of 8 mmol L−1. The highest % degradation efficiency of 100%, with a constant rate ~ 4-times greater than that found before optimization, was observed based on the interaction X1X2. Furthermore, non-toxicity and good mineralization (91%) were also observed under optimized conditions.