**Radical Intermediates in Photoinduced Reactions on TiO2 (An EPR Spin Trapping Study)**

#### **Dana Dvoranová, Zuzana Barbieriková and Vlasta Brezová**

**Abstract:** The radical intermediates formed upon UVA irradiation of titanium dioxide suspensions in aqueous and non-aqueous environments were investigated applying the EPR spin trapping technique. The results showed that the generation of reactive species and their consecutive reactions are influenced by the solvent properties (e.g., polarity, solubility of molecular oxygen, rate constant for the reaction of hydroxyl radicals with the solvent). The formation of hydroxyl radicals, evidenced as the corresponding spin-adducts, dominated in the irradiated TiO2 aqueous suspensions. The addition of 17O-enriched water caused changes in the EPR spectra reflecting the interaction of an unpaired electron with the 17O nucleus. The photoexcitation of TiO2 in non-aqueous solvents (dimethylsulfoxide, acetonitrile, methanol and ethanol) in the presence of 5,5-dimethyl-1 pyrroline *N*-oxide spin trap displayed a stabilization of the superoxide radical anions generated via electron transfer reaction to molecular oxygen, and various oxygen- and carbon-centered radicals from the solvents were generated. The character and origin of the carbon-centered spin-adducts was confirmed using nitroso spin trapping agents.

Reprinted from *Molecules.* Cite as: Dvoranová, D.; Barbieriková, Z.; Brezová, V. Radical Intermediates in Photoinduced Reactions on TiO2 (An EPR Spin Trapping Study). *Molecules* **2014**, *19*, 17279-17304.

#### **1. Introduction**

Among the materials previously studied as potential photocatalysts, titanium dioxide meets the criteria for industrial-scale utilization. Stability, low cost, relatively low toxicity and appropriate photocatalytic activity predispose TiO2 to a wide range of applications in various areas (gas sensors, photocatalysts, solar cells, thin film capacitors, self-cleaning surfaces, *etc.*) [1–6]. Especially attractive are nowadays the prospects of titania photocatalysts applications in the remediation of polluted water, soil and air, or in unconventional organic syntheses [7–13]. Consequently, all titanium dioxide polymorphs (anatase, brookite, rutile) have been intensively studied regarding their ability to produce, upon UVA photoexcitation, electron (e<sup>í</sup> ) and hole (h+ ) pairs further involved in the consecutive chemical reactions [1,2,5,14–16]. In general, the photoactivity of TiO2 is determined by the processes of electron/hole pair generation, recombination, interfacial transfer and by the surface reactions of these charge carriers with the species adsorbed on the surface of the photocatalyst [1,2,5,10,15–17]. The photoinduced processes on TiO2 nanoparticles upon ultra-band gap irradiation are also well influenced by the bulk structure, surface properties and the electronic structure of the photocatalyst [5]. The reactions of photogenerated holes with the adsorbed hydroxide anions and water molecules lead to the formation of highly reactive hydroxyl radicals, which, together with the hole itself, can initiate the oxidative degradation of organic pollutants down to water and carbon dioxide [1,2,5,10,16]. The efficient production of hydroxyl radicals and their non-selective reactions with organic and inorganic pollutants represent a crucial point considering the application of photocatalytic processes in water and air purification [1]. Recent investigations have revealed different mechanisms on anatase and rutile surfaces [18], as well as the role of surface-bridging oxygens of TiO2 on the • OH formation associated with the oxidation of surface hydroxide anions and water molecules by the photogenerated holes [19,20]. The presence of molecular oxygen also plays a substantial role in the photoinduced processes on irradiated TiO2 surfaces, as it enables an effective charge carriers separation. The electrons trapped transiently on the surface or on the next-to-surface defects can react with the adsorbed oxygen molecules [21–23]. The consecutive reactions of the so generated O2 •í are influenced by the solvent properties [24]. Although the superoxide radical anion is quite stable in the aprotic solvents [25], in aqueous solutions the reaction with protons is favorable, and hydrogen peroxide is formed and involved in further photocatalytic processes, Equations (1)–(6) [26]:

O2 •í + H+ ĺ • O2H (1)

$$\text{H}\_2\text{"O}\_2\text{H} \rightarrow \text{H}\_2\text{O}\_2 + \text{O}\_2\tag{2}$$

$$\rm H\_2O\_2 + O\_2{}^{\bullet -} \rightarrow \rm 3OH + O\_2 + OH^- \tag{3}$$

$$\text{H}\_2\text{O}\_2 \xrightarrow{hv} 2\text{ 'OH} \tag{4}$$

$$\mathrm{H\_2O\_2 + e^- \to \text{'OH + OH^-}}\tag{5}$$

$$\rm H\_2O\_2 + h^+ + OH^- \rightarrow H\_2O + \rm O\_2H \tag{6}$$

Consequently, superoxide radical anion and hydrogen peroxide as the most important products of the molecular oxygen reduction play an important role in the complex mechanism of Reactive Oxygen Species (ROS, e.g., • OH, • O2H or singlet oxygen) generation on the irradiated TiO2 surfaces [1,18,26–28].

EPR spectroscopy occupies an exclusive position in the investigation of titania photocatalysts, providing a characterization of paramagnetic centers produced via the trapped photogenerated electrons and holes [29–36], and of titanium dioxide materials with transition-metal ions doping [17,34]. A majority of the research exploiting EPR spectroscopy deals with the investigation of reactive radical intermediates produced in the irradiated TiO2 particulate systems where the application of an indirect spin trapping technique is inevitable [37–45]. This method is based on the chemical reaction of a diamagnetic spin trap (ST) with a short-lived radical, producing a more stable nitroxide radical, *i.e.*, spin-adduct, using nitrones, *N*-oxides and nitroso compounds as the spin trapping agents (Figure 1). Spin traps possessing *N*-oxide and nitrone groups are mainly applied in the identification of hydroxyl radicals generation, as well as other oxygen-, nitrogen- and sulfur-centered reactive radicals, however the information on the structure of carbon-centered radicals trapped with these agents is limited, and the application of nitroso spin traps is necessary to bring the knowledge on other nuclei in the vicinity of the trapped carbon [46,47]. Successful assignment of measured EPR spectra of spin-adducts requires a thorough interpretation of the acquired data and careful choice of the spin trapping agent for the specific experimental conditions [48].

**Figure 1.** Overview of the spin trapping agents applied in *in situ* EPR investigations.

In the literature hydroxyl radicals are frequently declared as the most important reactive species generated upon TiO2 irradiation, in spite of the addition of organic co-solvent to the reaction systems, which effects the character and amount of radicals formed. The main aim of our study was to point on the formation of radical intermediates in aqueous TiO2 suspensions, as well as in suspensions prepared in an organic solvent (DMSO, acetonitrile, methanol, ethanol) and thus provide a straightforward comparison of the radical species detected under UVA exposure of TiO2 suspended in water and in aprotic and protic polar solvents exploiting the *in situ* EPR spin trapping technique. The oxidation of sterically hindered amines to the corresponding nitroxide radicals via ROS photogenerated in the aqueous and acetonitrile TiO2 suspensions was also monitored by *in situ* EPR spectroscopy.

#### **2. Results and Discussion**

#### *2.1. Spin Trapping in the Aqueous TiO2 Suspensions*

Despite the fact that the detection of hydroxyl radicals upon UVA irradiation of the aerated aqueous TiO2 suspensions in the presence of 5,5-dimethyl-1-pyrroline *N*-oxide (DMPO) spin trap represents a frequently applied EPR technique [38–45,49], in order to bring the complete information on the radical species generated in the irradiated TiO2 in different media we also report the results of EPR spin trapping experiments in aqueous TiO2 suspensions using different spin trapping agents. As expected, immediately after the irradiation started, a typical four-line EPR signal attributed to • DMPO-OH spin-adduct with spin Hamiltonian parameters (*a*N = 1.497 mT, *a*H = 1.477 mT; *g* = 2.0057 [50]) was generated in the system TiO2/DMPO/H2O/air, as is shown in Figure 2a. The concentration profile of • DMPO-OH during the *in situ* EPR spin trapping experiments is strongly influenced by TiO2 loading, UVA radiation dose, as well as by the initial oxygen and spin trap concentrations [38,44].

The primary source of the • OH radicals in the irradiated aqueous TiO2 suspensions is the oxidation of OH<sup>í</sup> and H2O by the photogenerated holes, however further reactions of the reactive oxygen species generated in the system leading to • OH cannot be excluded (Equations (1)–(6)). The EPR spin trapping technique is assumed to detect the photogenerated hydroxyl radicals on the photocatalysts' surfaces, based on the previous comparison with the quantification of • OH via fluorescence detection using the hydroxylation of terephthalic acid, by which the bulk • OH are detected [43].

**Figure 2.** The sets of individual EPR spectra (magnetic field sweep width, *SW* = 8 mT) monitored upon continuous UVA irradiation (Ȝmax = 365 nm; irradiance 15 mW·cm<sup>í</sup><sup>2</sup> ) of aerated TiO2 P25 suspensions in the presence of spin trapping agent DMPO: (**a**) water; (**b**) mixed solvent water/dimethylsulfoxide (5:1 v:v); (**c**) acetonitrile. TiO2 concentration 0.167 mg·mL<sup>í</sup><sup>1</sup> , *c*0,DMPO = 0.035 M.

Figure 3 illustrates experimental and simulated EPR spectra of • DMPO-OH measured in the TiO2 suspensions prepared either using ordinary water, or water enriched with the magnetically active 17O (13%–17% atom.) nucleus. The EPR spectrum depicted in Figure 3b is fully compatible with the presence of both spin-adducts, *i.e.*, • DMPO-OH and • DMPO-17OH [51], unambiguously identifying the adsorbed and close-to-surface water molecules as the source of hydroxyl radicals. Recently, we conducted EPR spin trapping experiments using aerated aqueous suspensions of Ti17O2 (containing up to 90% atom. 17O) with DMPO, and the EPR spectra of • DMPO-OH corresponded to the interaction of one nitrogen nucleus (*a*N = 1.492 mT) and one hydrogen nucleus (*a*H = 1.476 mT) with an unpaired electron [52]. No evidence of a hyperfine coupling from 17O was found, consequently the lattice oxygens from TiO2 were excluded as the source of hydroxyl radicals trapped by DMPO under the given experimental conditions [52].

**Figure 3.** Experimental (**black**) and simulated (**red**) EPR spectra (*SW* = 8 mT) obtained upon irradiation (Ȝmax = 365 nm; irradiance 15 mW·cm<sup>í</sup><sup>2</sup> ; exposure 400 s) of the aerated aqueous TiO2 P25 suspensions in the presence of spin trapping agent DMPO (TiO2 concentration 0.167 mg·mL<sup>í</sup><sup>1</sup> , *c*0,DMPO = 0.035 M): (**a**) ordinary water; (**b**) water enriched with H2 17O (13%–17% atom.). Simulation parameters (hfcc in mT): (a) • DMPO–OH (*a*N = 1.497, *a*H = 1.477; *g* = 2.0057); (b) linear combination of spin-adducts, *i.e.*, • DMPO–OH (relative concentration in %, 82) and • DMPO-17OH (*a*N = 1.494, *a*H = 1.480, *a*17O = 0.467; *g* = 2.0057; 18).

Previously, the possibility of a direct oxidation of the spin trapping agent DMPO via photogenerated holes to a radical cation DMPO•+, which subsequently reacts with water molecules forming a so-called imposter spin-adduct • DMPO-OH, was supposed [41,43,44]. In addition, the degradation of the low-stability • DMPO-O2H spin-adduct, theoretically also generated in the studied system, results in the • DMPO-OH formation [53]. However, the generation of the surface hydroxyl radicals can be evidenced by the addition of dimethylsulfoxide (DMSO) to the aqueous TiO2 suspensions, since the rapid reaction of hydroxyl radicals with DMSO (Table 1) produces methyl radicals [54], detectable in the reaction with spin trap (ST) as the corresponding carbon-centered spin-adduct, Equations (7) and (8):

$$\text{(CH}\_3\text{)}\_2\text{SO} + \text{'OH} \rightarrow \text{CH}\_3\text{(OH)}\text{SO} + \text{'CH}\_3\text{}\tag{7}$$

$$\text{"CH}\_3 + \text{ST} \to \text{"ST-CH}\_3 \tag{8}$$


**Table 1.** Bimolecular rate constants for the reaction of hydroxyl radical with the selected solvents [54], decay constants and lifetime of singlet oxygen (1 ¨g) in selected solvents [55].
