Decolorization of Orange-G Aqueous Solutions over C 60 / MCM-41 Photocatalysts

: The majority of the photocatalysts studied for azo-dye degradation are based on semiconductor materials. Studies reported on non-semiconducting materials are very scarce. In the present work, we studied the fullerene (C 60 ) ability to accelerate photodegradation of the dye’s azo bond in the presence of ascorbic acid. A series of C 60 supported on ordered mesoporous silica (MCM-41) catalysts, containing 1, 3, 6, 9, and 12 wt % of fullerene C 60 , was studied using Orange G (OG) as representative azo-dye. This study showed that partial decolorization is achieved in the dark by simple adsorption of the dye on the bare surface of the carrier. The extent of decolorization increases with the irradiation of the suspension due to photocatalytic degradation of the azo-bond. This is maximized over the sample containing 3 wt % of C 60 and it has been attributed to the best combination of the extent of the dye adsorption with the high intrinsic photocatalytic activity of small C 60 clusters predominated in this sample. This catalyst proved to be quite stable upon ﬁve subsequent photocatalytic cycles, losing less than 5% of its initial activity. No degradation of OG takes place in the absence of ascorbic acid.


Introduction
Discharge of industrial effluents containing hazardous contaminants, even at low concentrations, causes a negative impact on the environment [1]. Out of the various pollutants, dyes rank among the most notorious organic contaminants that are discharged into the environment as part of textile, leather, paints, paper, and dye manufacturing [2]. These are synthetic, water soluble, and dispersible organic compounds, which may adversely affect aquatic environment both aesthetically and chemically. Several studies have been addressed to the decolorization of aqueous media via adsorption of dyes on high surface area solids [1,[3][4][5][6]. Aqueous media decolorization is also obtained by photocatalytic oxidation of such pollutants over UV irradiated titanium dioxide [7][8][9][10].
Photocatalytic oxidation belongs to advanced oxidation processes (AOPs), which combine powerful oxidizing agents (catalytic initiators) with UV or near-UV light for the removal of resistant organic pollutants and xenobiotics [11][12][13][14][15]. Among them, photocatalytic oxidation has been proven to be a promising technology for water treatment in order to destruct a variety of organic and inorganic contaminants [15][16][17]. The majority of the photocatalysts studied are based on semiconductor materials [18,19]. Studies reported on non-semiconducting materials are very scarce [20][21][22].
UV irradiation produces holes and electrons in the valence and conduction bands of semiconductor materials, respectively. The holes oxidize OHto OH free radicals. These are oxidative agents with high oxidative potential and may mineralize the dyes. Hydrogen peroxide also produces hydroxyl radicals and therefore sometimes is added to the reactor for achieving faster photocatalytic oxidation [23]. Photocatalytic degradation of dyes can be also achieved by singlet oxygen which is 1 eV stronger oxidative agent in comparison to the oxygen fundamental state and therefore strongly electrophilic [24].
Fullerenes have attracted special interest as photocatalysts due to their unique structure and properties [25,26]. In particular, the photoreactivity of C 60 has received considerable attention due to its electron acceptability, intense absorption in near-UV light, and broad absorbance in the visible light region. A number of studies of our group have been reported dealing with C 60 as a photocatalyst [27][28][29][30][31][32]. However, there are very few studies on the photocatalytic performance of C 60 in aqueous solutions because it exhibits very low solubility in this solvent. Immobilization of C 60 on high surface area supports is expected not only to inhibit aggregation of C 60 molecules but also allows recycling of the supported photocatalyst and prevents potential environmental release of fullerene as there is an emerging consensus concerning the in vivo toxicity of these carbon nanoparticles [33].
Non-semiconducting materials like γ-Al 2 O 3 , SiO 2 , and MCM-41 exhibit high surface areas suitable to immobilize highly dispersed C 60 moieties [27,[29][30][31][32]34]. Recently, the photocatalytic degradation of methyl orange has been achieved even under visible irradiation over C 60 /SiO 2 photo-catalyst in the presence of ascorbic acid [20]. The main advantage of using fullerene for photocatalytic water decontamination lies in its efficiency to activate reactive oxygen species even under visible light. The latter species are able to degrade micropollutants found in water [34] A critical question at this point is whether the decolorizing activity of supported C 60 is limited to methyl orange, eventually related to its molecular structure, or can it be applied generally for decolorizing aqueous systems. This is the reason for which the present study deals with decolorization of aqueous solutions containing Orange G. Orange G is an azo-dye with a wide range of applications (coloring wools and silks; pigmentation of paper, leather, and timber; printing ink; pH indicators; etc.) [10]. It is usually present, as a sodium salt, in two tautomeric forms in aqueous solutions. Organic solvents favor the azo form. In contrast, the hydrozon form predominates in aqueous solutions.
In the present article, we attempt to develop non-semiconducting materials for the decolorization of aqueous media with two main characteristics: very high specific surface area enhancing the adsorption of dyes and quite high photocatalytic activity. Specifically, we studied the decolorization of orange G aqueous solutions over C 60 /MCM-41 photocatalysts containing 1, 3, 6, 9, and 12 wt % of C 60 in the presence of ascorbic acid. The investigation of the relative contribution of adsorption and photocatalysis to the whole decolorization, the determination of the best catalyst composition and a contribution to the decolorization reaction network are among the goals of the present work. To the best of our knowledge, it is the first time that one is attempting to combine the C 60 photocatalytic performance with the high adsorption capacity of high surface area ordered mesoporous silica (MCM-41) for decolorization of an aqueous solution.

Photocatalytic Tests
The photocatalytic tests were performed at 400 mL aqueous suspensions containing 25 µM of orange G (OG, Merck, Darmstadt, Germany), 0.5 mM of ascorbic acid (AA, Sigma, China), and 300 mg of the photocatalyst. A glass semi-batch photoreactor was used (Heraeus, Hanau, Germany) equipped with magnetic stirrer (Figure S1), sampling system and a proper socket for glass electrode to measure the suspension's pH. The lamp, TQ 150 (power: 150W), situated in a glass cooling jacket filled with re-circulated deionized water, is located inside the reactor. Thus, the emitted UV irradiation in wavelengths lower than 300 nm was cut off ( Figure S2). Before each experiment the photoreactor was covered by aluminum paper to avoid the effect of external light. The reactor was fed with 15 mL/min air regulated by a mass flow controller (Brooks, Holland). The temperature of the suspension during the whole experimental procedure was maintained at 25±0.5 • C and the pH≈4. Liquid samples were taken at various time intervals. These were filtered (Millipore cellulose filters 0.22 nm, Ireland) for removing catalytic particles and then their spectra were recorded in the range 200-600 nm using a UV-Vis spectrophotometer (Varian, model Cary 3, Australia).
Each kinetic experiment was initially carried out for more than 180 min in dark in order to determine the extent of decolorization which is due to the adsorption of the Orange G on the photocatalyst's surface. During this time, liquid samples were taken at various time intervals and analyzed as mentioned above. Then the lamp was switched on and kinetic procedure was performed under irradiation in order to monitor decolorization via photocatalytic oxidation.

Physicochemical Characteristics of the Photocatalysts
Detailed physicochemical characterization of the studied photocatalysts presented in our previous works [31,32], revealed that the SSA BET of the C 60 /MCM-41 samples was slightly decreased (Table 1) and the mesoporous structure of MCM-41 remained intact upon deposition of C 60 on its surface. The nature of the C 60 species has not been changed upon the deposition of C 60 on the MCM-41 surface or upon the development of the support-supported phase interactions during heating. The C 60 was highly dispersed within the mesopores or the external surface of MCM-41 as clusters or crystallites/aggregates of very small size. Calculation of the size of C 60 crystallites using the Scherer's equation gave the following values: 1C 60 /MCM-41 & 3C 60 /MCM-41: < 4nm, 6C 60 /MCM-41: 18nm, 9C 60 /MCM-41: 45 nm, 12C 60 /MCM-41: 62 nm. The mesopore size of all the photocatalysts remained similar to that of the MCM-41 support. This finding in combination with the XRD results inferred that a preferential growth of C 60 crystallites in one direction, namely a cylindrical growth inside the support pores takes place upon deposition of C 60 on the MCM-41 surface.

Decolorization by Adsorption
The electronic spectra of OG, OG-AA, and AA aqueous solutions are seen in Figure 1. The peaks at 475 and 410 nm, responsible for the color, are attributed respectively to the azo-bonds of hydrozon and azo form and the peaks at 250 and 330 nm are assigned to benzene and naphthalene rings of OG, Appl. Sci. 2019, 9,1958 4 of 12 respectively [35,36]. It should be noted that ascorbic acid absorption peak (265 nm) interferes with the benzene ring peak of azo dye located at 250 nm. On the other hand, we were obliged to doubly dilute the initial AA aqueous solution ( Figure 1) in order to record correctly its spectrum. This was not necessary in the case of a solution containing both substances indicating that an interaction takes place between them in the solution. rings of OG, respectively [35,36]. It should be noted that ascorbic acid absorption peak (265 nm) interferes with the benzene ring peak of azo dye located at 250 nm. On the other hand, we were obliged to doubly dilute the initial AA aqueous solution ( Figure 1) in order to record correctly its spectrum. This was not necessary in the case of a solution containing both substances indicating that an interaction takes place between them in the solution.     rings of OG, respectively [35,36]. It should be noted that ascorbic acid absorption peak (265 nm) interferes with the benzene ring peak of azo dye located at 250 nm. On the other hand, we were obliged to doubly dilute the initial AA aqueous solution ( Figure 1) in order to record correctly its spectrum. This was not necessary in the case of a solution containing both substances indicating that an interaction takes place between them in the solution.    The kinetics of decolorization in the absence of light over two samples with different C 60 loadings (3 and 9 wt % C 60 ) are illustrated in Figure 3 (left part). The absorbance at 475 nm has been taken Appl. Sci. 2019, 9,1958 5 of 12 as a measure of decolorization. We observe that the extent of decolorization due to the adsorption depends on the photocatalyst composition. To study the adsorption kinetics, the pseudo-first-order and pseudo-second-order kinetic models were both examined according to Equations (1) and (2), respectively [37].
where, q e and q t (mg g −1 ) symbolize the amount of OG absorbed per gram of photocatalyst at adsorption equilibrium and each time t (min), respectively. k A1 (min −1 ) and k A2 (g mg −1 min −1 ) represent pseudo-first-order and pseudo-second-order rate constant, respectively. The aforementioned kinetic analysis of OG adsorption experiments showed that a pseudo second order kinetic law (Equation (2)) describes much better the experimental data (R 2 > 0.99). This finding has been also reported for the adsorption kinetics of OG on the surface of other adsorbents [37]. The calculated k A2 values are given in Table 1. Inspection of this observe that the absorbance in the above range decreases significantly up to 60 min and afterwards the decrease is slowing down and absorbance remains almost constant for the period of 120-180 min. The kinetics of decolorization in the absence of light over two samples with different C60 loadings (3 and 9 wt % C60) are illustrated in Figure 3 (left part). The absorbance at 475 nm has been taken as a measure of decolorization. We observe that the extent of decolorization due to the adsorption depends on the photocatalyst composition. To study the adsorption kinetics, the pseudo-first-order and pseudo-second-order kinetic models were both examined according to Equations (1) and (2), respectively [37].
where, qe and qt (mg g -1 ) symbolize the amount of OG absorbed per gram of photocatalyst at adsorption equilibrium and each time t (min), respectively. kA1 (min -1 ) and kA2 (g mg -1 min -1 ) represent pseudo-first-order and pseudo-second-order rate constant, respectively. The aforementioned kinetic analysis of OG adsorption experiments showed that a pseudo second order kinetic law (Equation 2) describes much better the experimental data (R 2 >0.99). This finding has been also reported for the adsorption kinetics of OG on the surface of other adsorbents [37]. The calculated kA2 values are given in Table 1. Inspection of this table shows that the adsorption rate  Inspection of Table 1 shows that the extent of adsorption follows similar trend to those of specific surface area of the catalysts and the adsorption kinetic constant, as well. However, the decrease in the specific adsorption is abrupt from the second to the third sample (36.5 to 11.2 µmol g −1 ) whereas the corresponding decrease in the specific surface area is not so large (831 to 813 m 2 g −1 ). This suggests an additional reason for the decrease of the adsorption. One may propose that the decrease in the specific adsorption is also due to coverage of adsorption sites, situated on the surface of MCM-41, by the supported C 60 clusters as the concentration of the latter is duplicated from the second to the third sample. The relatively low dispersion of the C 60 supported clusters in the samples 9C 60 /MCM-41 and 12C 60 /MCM-41 [31,32] leaves uncovered sufficient portion of MCM-41 surface, justifying the relatively small additional decrease in the specific adsorption observed from the third to the fifth sample. Therefore, we consider that the decolorization through adsorption is taking place mainly on the bare surface of the carrier.
In a previous work [10], studying the mechanism of OG adsorption on the surface of TiO 2 , we had concluded that the adsorption of OG occurs mainly through the one of the two -SO 3 − groups of its structure, presumably due to steric constraints, on the positively charged surface sites. This has been also confirmed by Tang et al. [37], who studied the selective adsorption of anionic dyes on silver bromide-based nanomaterials and concluded that electrostatic attraction of -SO 3 − groups by positively charged surface sites is the dominating contribution to the adsorption of these dyes. It should be noticed that the above mentioned nanomaterials exhibited higher SA (~58 µmol/g) than that achieved over the best of our photocatalyst (3C 60 /MCM-41). This can be explained taking into account that although C 60 /MCM-41 photocatalysts are characterized by too high SSA BET values, the silica nature of their support implies low surface density of positively charged sites at pH >~3 [38], where the adsorption and photocatalytic experiments took place.   One may observe that in both cases irradiation induces further reduction of the absorbance due to the photocatalytic decomposition of Orange G via the azo-bonds splitting. However, this is not the case in the absence of ascorbic acid (Figure 3). This reveals the key role of the ascorbic acid in the photocatalytic degradation mechanism of Orange G.

Decolorization through Photocatalytic Splitting of Azo-Bonds
In order to further study decolorization through the photocatalytic splitting of azo-bonds, we determined the rate constants (kR) of photocatalytic reaction for the samples studied. Assuming that  One may observe that in both cases irradiation induces further reduction of the absorbance due to the photocatalytic decomposition of Orange G via the azo-bonds splitting. However, this is not the case in the absence of ascorbic acid (Figure 3). This reveals the key role of the ascorbic acid in the photocatalytic degradation mechanism of Orange G.
In order to further study decolorization through the photocatalytic splitting of azo-bonds, we determined the rate constants (k R ) of photocatalytic reaction for the samples studied. Assuming that the concentration of O 2 in the reaction mixture remains constant and equal to saturation due to the continuous feed with air, a pseudo-first-order photoreaction with respect to the Orange G (Equation 3) could be easily derived [10,39,40].
where A 0 and A stand, respectively, for the absorption at zero time and irradiation time t. The linear plots, -ln(A/A 0 ) vs. t, obtained for all the samples studied corroborated the aforementioned assumption. Correlation coefficients (R 2 ) greater than 0.93 have been obtained in all cases. An example, over the sample 9C 60 /MCM-41, is illustrated in Figure 5.   ) illustrates the photocatalytic decolorization kinetics obtained over two samples with different C60 loadings (3 and 9 wt % C60) when the photoreactor lamp was switched on. One may observe that in both cases irradiation induces further reduction of the absorbance due to the photocatalytic decomposition of Orange G via the azo-bonds splitting. However, this is not the case in the absence of ascorbic acid (Figure 3). This reveals the key role of the ascorbic acid in the photocatalytic degradation mechanism of Orange G.
In order to further study decolorization through the photocatalytic splitting of azo-bonds, we determined the rate constants (kR) of photocatalytic reaction for the samples studied. Assuming that the concentration of O2 in the reaction mixture remains constant and equal to saturation due to the continuous feed with air, a pseudo-first-order photoreaction with respect to the Orange G (Equation 3) could be easily derived [10,39,40].
where A0 and A stand, respectively, for the absorption at zero time and irradiation time t. The linear plots, -ln(A/A0) vs. t, obtained for all the samples studied corroborated the aforementioned assumption. Correlation coefficients (R 2 ) greater than 0.93 have been obtained in all cases. An example, over the sample 9C60/MCM-41, is illustrated in Figure 5.  The calculated photocatalytic reaction rate constant values are compiled in Table 1. These values are in good agreement with those reported in the recent literature for similar reaction conditions [39][40][41][42][43]. The k R value was initially increased with the C 60 loading of the photocatalysts approaching its maximum over the 3C 60 /MCM41 sample. Then it was decreased considerably up to the sample 6C 60 /MCM41 and afterwards was slightly decreased with the fullerene content. The relatively high value for the photocatalytic reaction rate constant obtained over the sample 3C 60 /MCM41 can be effortlessly attributed to the very small C 60 clusters, not detectable by XRD, supported on MCM-41 [31]. These clusters, in effect, exhibited also very high intrinsic photocatalytic activity in other photocatalytic reactions as in the olefins photoxidation [31].
By increasing the C 60 content from 3 to 6 wt % the C 60 clusters size is increasing [32], bringing about a significant decrease in the intrinsic activity. The further decrease in the value of the rate constant determined over the photocatalysts with C 60 loading 9 and 12 wt % (Table 1) could be attributed to the additional size increase of C 60 clusters and therefore to the decrease of their intrinsic activity. However, the decrease of reaction constant value observed as the C 60 loading of the catalysts increases from 6 to 12 wt % is relatively low. This could be attributed to the partial compensation of the intrinsic activity decrease by the increase of photo sensitizer content per gram of catalyst.
The high values obtained for the specific adsorption of Orange G and the specific rate of photocatalytic splitting of azo-bonds over the sample 3C 60 /MCM-41 justify the relatively high value obtained for the percentage decolorization of the Orange G aqueous solutions ( Table 1). The trend for this parameter follows the trend obtained for the rate constant of the catalytic splitting of azo-bonds (Table 1).

Interesting Characteristics of Decolorization Procedure
In order to investigate some details of the decolorization procedure, we have monitored the change with time in the absorption bands at 330 and 250 nm assigned to the naphthalene and benzene (interfered by AA) rings, respectively ( Figure 6). Inspection of Figures 3 and 6 shows that similar kinetics are followed concerning the decrease of the intensity of the absorption band at 475 nm with those observed for the absorption bands at 330 and 250 nm over the sample 3C 60 /MCM41 taken as an example. This is valid in the dark and under irradiation as well. Concentrating our attention to the experiments taking place in the absence of light, we observe that the normalized absorption is decreasing from 1 to about 0.43-0.45 at t=200 min for the three bands at 475, 330 and 250 nm assigned, respectively, to azo-bonds, naphthalene, and benzene (and ascorbic acid) rings. This suggests that the adsorption of Orange G on the solid surface is not accompanied by its decomposition and selective adsorption of some of the resulting molecules. The relatively faster initial decrease of the absorbance at 250 nm ascribed to benzene ring of the azo-dye and to AA could be attributed to the lower molecular weight of the latter (about 2.5 times lighter than OG) which facilitates its mass transfer from the solution to the catalyst surface.
Concerning the experiments taken place under irradiation, similar kinetics is observed for the decolorization (absorbance decrease at 475 nm, Figure 3) and absorbance decrease at 330 and 250 nm attributed, respectively, to naphthalene rings and to the benzene ring plus AA ( Figure 6). In both cases, indeed, the normalized absorption is decreasing to 0.2-0.25 after 140 minutes of irradiation. This presumably suggests full destruction of Orange G, not only the simple cleavage of azo-bonds, as well as of the ascorbic acid.  According to our adsorption and photocatalytic kinetic results, it seems that adsorption is the rate determining step (kR>>kA2), see Table 1. This finding probably explains why complete decolorization has not been achieved in the reaction time for which photocatalytic degradation of OG has been studied ( Figure 3).
As a critical point in relevant studies, the reusability of the photocatalysts was developed, five subsequent cycles of OG photo-destruction were performed over the same sample of 3C60/MCM-41 catalyst, which showed the highest photocatalytic activity. The catalyst was thoroughly washed with triple-distilled water and dried at 110 °C overnight between the subsequent photocatalytic experiments. The results presented in Figure 7 reveal that this catalyst is quite stable losing less than 5% of its initial activity after five cycles. According to our adsorption and photocatalytic kinetic results, it seems that adsorption is the rate determining step (k R >>k A2 ), see Table 1. This finding probably explains why complete decolorization has not been achieved in the reaction time for which photocatalytic degradation of OG has been studied ( Figure 3).
As a critical point in relevant studies, the reusability of the photocatalysts was developed, five subsequent cycles of OG photo-destruction were performed over the same sample of 3C 60 /MCM-41 catalyst, which showed the highest photocatalytic activity. The catalyst was thoroughly washed with triple-distilled water and dried at 110 • C overnight between the subsequent photocatalytic experiments. The results presented in Figure 7 reveal that this catalyst is quite stable losing less than 5% of its initial activity after five cycles. subsequent cycles of OG photo-destruction were performed over the same sample of 3C60/MCM-41 catalyst, which showed the highest photocatalytic activity. The catalyst was thoroughly washed with triple-distilled water and dried at 110 °C overnight between the subsequent photocatalytic experiments. The results presented in Figure 7 reveal that this catalyst is quite stable losing less than 5% of its initial activity after five cycles. Taking into account our experimental findings, as well as the relevant literature [20,44,45], we can propose the following tentative reaction network for the photocatalytic degradation of OG over the C60/MCM-41 catalysts. Irradiation causes the excitation of C60 from the fundamental singlet state, 1 C60, to the excited singlet state 1 C60*. It is very probable that this state is transformed to the excited triplet state, 3 C60*, by intersystem crossing [46]. Electron transfer occurs from the 3 C60* to OG, being in the hydrozone form, adsorbed on catalyst surface and from AA to the 3 C60*.This electron circulation Taking into account our experimental findings, as well as the relevant literature [20,44,45], we can propose the following tentative reaction network for the photocatalytic degradation of OG over the C 60 /MCM-41 catalysts. Irradiation causes the excitation of C 60 from the fundamental singlet state, 1 C 60 , to the excited singlet state 1 C 60 *. It is very probable that this state is transformed to the excited triplet state, 3 C 60 *, by intersystem crossing [46]. Electron transfer occurs from the 3 C 60 * to OG, being in the hydrozone form, adsorbed on catalyst surface and from AA to the 3 C 60 *.This electron circulation activates the N-N bond and causes the decomposition of hydrozone structure towards 7-hydroxy-8-(hydroxyamino)naphthalene-1,3-disulfonic species and aniline. These intermediate products are further mineralized producing SO 4 2− , NO 3 − , NO 2 − , NH 4 + , and CO 2 [10]. Scheme 1 depicts the above described reaction network. activates the N-N bond and causes the decomposition of hydrozone structure towards 7-hydroxy-8-(hydroxyamino)naphthalene-1,3-disulfonic species and aniline. These intermediate products are further mineralized producing SO4 2− , NO3 − , NO2 − , NH4 + , and CO2 [10]. Scheme 1 depicts the above described reaction network.

Conclusions
A series of C60/MCM-41 photocatalysts was studied for the first time in azo-dye removal from water. All these photocatalysts were proven to be active in the decolorization of orange G aqueous solutions.
The decolorization is achieved partly in the dark via adsorption of orange G on the bare MCM-41 surface following pseudo-second order kinetics. The adsorption in this stage is not accompanied by decomposition. The specific adsorption of the dye decreases as the fullerene content increases because photocatalyst specific surface area decreases and the surface of MCM-41 is

Conclusions
A series of C 60 /MCM-41 photocatalysts was studied for the first time in azo-dye removal from water. All these photocatalysts were proven to be active in the decolorization of orange G aqueous solutions.
The decolorization is achieved partly in the dark via adsorption of orange G on the bare MCM-41 surface following pseudo-second order kinetics. The adsorption in this stage is not accompanied by decomposition. The specific adsorption of the dye decreases as the fullerene content increases because photocatalyst specific surface area decreases and the surface of MCM-41 is covered by an increasing amount of the sensitizer.
The extent of decolorization is increased by irradiation in the presence of ascorbic acid and it is due to the photocatalytic degradation of the azo dye which is following first order kinetics.
The photocatalytic activity increases initially with C 60 loading. It takes its maximum value over the sample containing 3 wt % C 60 and then it significantly decreases as the C 60 loading increases to 6 wt %. Further increase of C 60 loading (up to 12 wt %) is accompanied by a slight decrease of photocatalytic activity being the result of a compromise between the increase of active phase loading and the lowering of its dispersion and thus intrinsic activity. The high photocatalytic activity exhibited by the sample 3C 60 /MCM-41 is attributed to the very small C 60 clusters supported on MCM-41, which exhibit very high intrinsic activity.
The 3C 60 /MCM-41 catalyst proved to be quite stable upon five subsequent photocatalytic cycles, losing less than 5% of its initial activity.
The crucial role of ascorbic acid for photocatalytic degradation of azo-dyes over supported C 60 catalysts has been demonstrated once again and a reaction network has been proposed.
Supplementary Materials: The following are available online at http://www.mdpi.com/2076-3417/9/9/1958/s1, Figure S1: Photocatalytic reactor (Heraeus) with the TQ 150 lamp situated in a glass cooling jacket filled with re-circulated deionized water and immersed in the reaction mixture, Figure S2: Relative spectral radiant flux of low pressure Hg lamp (TQ 150) being in a glass cooling jacket filled with re-circulated de-ionized water.