: post-radiation saturation with argon. Symbols ȕ and Ȗ denote the position of the interacting hydrogen nuclei.

These oxygen-centered radical species can be easily identified using the DMPO spin trap. Due to the higher concentration of dissolved oxygen in these solvents, the EPR spectra of spin-adducts measured in aerated methanol and ethanol TiO2 suspensions are characterized by a significant line broadening, which hinders a detailed simulation analysis (Figure 8a,b blue lines). However, the post-radiation saturation of the TiO2 suspensions with argon, and subsequent measurement of EPR spectra provide the EPR signals of sufficient quality (Figure 8a,b black lines).

**Figure 8.** Experimental (**blue**) 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 TiO2 P25 (0.167 mg·mL<sup>í</sup><sup>1</sup> ) suspensions in organic solvents containing DMPO spin trap (*c*0,DMPO = 0.035 M), along with EPR spectra measured after post-radiation saturation with argon (**black**) and their simulations (**red**): (**a**) methanol; (**b**) ethanol. The spin Hamiltonian parameters of corresponding spin-adducts are listed in Table 4. (a) • DMPO–O2 í (rel. conc. in %, 33), • DMPO–OCH3 (35), • DMPO–OCH2OH (21), • DMPO–CH2OH (7), • DMPOdegr (4); (b) • DMPO–O2 í (17), • DMPO–OCH2CH3 (59), • DMPO–OR (10), • DMPO–CR1 (9), • DMPO–CR2 (5).

Simulations of the EPR spectra measured using this experimental procedure in the system TiO2/DMPO/methanol revealed the presence of individual DMPO spin-adducts corresponding to • DMPO-O2 í , • DMPO-OCH3, • DMPO-CH2OH, • DMPO-OCH2OH and a triplet signal assigned to the DMPO degradation product (Figure 8a red line). The spin Hamiltonian parameters of the identified spin-adducts are gathered in Table 4. We assume that the radical species • CH2OH and • OCH2OH are produced via complex reactions of • OCH3 radicals with methanol molecules and

molecular oxygen [62], an alternative mechanism of • CH2OH generation represents the hydrogen abstraction from methanol via hydroxyl radicals (Table 1, Equations (16) and (17)).

$$\text{CHrO'} + \text{CHrOH} \rightarrow \text{CHrOH} + \text{'CHrOH} \tag{16}$$

$$\text{C}^{\cdot}\text{CH}\text{OH} + \text{'OH} \rightarrow \text{'C}\text{H}\text{ZnO} + \text{H}\text{O} \tag{17}$$

Simulation of the EPR spectra obtained in the irradiated suspensions TiO2/DMPO/ethanol after the saturation with argon evidenced the presence of spin-adducts characteristic for • DMPO-O2 í , • DMPO-OCH2CH3, • DMPO-OR, and two carbon-centered spin-adducts with slightly differing hyperfine coupling constants (Figure 8b, Table 4). Nitrosodurene spin trapping agent was used in the analogous experiments to identify the carbon-centered radicals in the irradiated methanol or ethanol TiO2 suspensions (spectra not shown). In methanol only the generation of • ND-CH2OH was evidenced. The EPR spectra monitored in TiO2/ethanol suspensions are compatible with • ND-CH(OH)CH3 and • ND-CH3 spin-adducts in good accordance with the reactions of ethoxy, 1-hydroxyethyl and 2-hydroxyethyl radicals in ethanol (Equations (18)–(20)) published previously [62], and also with two DMPO carbon-centered spin-adducts detected:

$$\text{CHrCHzO}^{\bullet} + \text{CHrCHzOH} \rightarrow \text{CHrCHzOH} + \text{CHrCHOH} \tag{18}$$

$$\text{CH}\_3\text{CHOH} + \text{CH}\_3\text{CH}\_2\text{OH} \rightarrow \text{CH}\_2\text{CH}\_2\text{OH} + \text{CH}\_3\text{CH}\_2\text{OH} \tag{19}$$

$$\text{CH}\_2\text{'CH}\_2\text{OH} \rightarrow \text{'CH}\_3 + \text{CH}\_2\text{O} \tag{20}$$

#### *2.3. Oxidation of Sterically Hindered Amine in TiO2 Suspensions*

The irradiation of titanium dioxide nanoparticles in the presence of molecular oxygen results in the generation of singlet oxygen, but the specific mechanism of <sup>1</sup> O2 formation is not straightforward and alternative reaction pathways have been suggested [88–90]. The direct detection of <sup>1</sup> O2 is based on the phosphorescence measurement at 1270 nm corresponding to the radiative transition O2( 1 ¨g)ĺO2( 3 Σg) [89,91]. The principle of the indirect techniques of <sup>1</sup> O2 monitoring is the specific reaction with an organic compound generating a product detectable by a suitable method [92], supported by the application of <sup>1</sup> O2 scavengers and traps, or using the effect of deuterated solvents. The photoinduced formation of singlet oxygen in the homogeneous systems is frequently monitored also by EPR spectroscopy, detecting the generation of nitroxide radicals derived from 4(R)-2,2,6,6-tetramethylpiperide *N*-oxyl (R = hydroxy, oxo) produced by the oxidation of corresponding sterically hindered amines (SHA) [84,93,94]. Although this method is widely used for the singlet oxygen detection, many questions arise concerning its selectivity [95]. Particular problems may appear when a numerous ROS or other reactive species are formed in the studied system and their interaction with SHA cannot be excluded, e.g., in the irradiated TiO2 suspensions. The detailed analysis of paramagnetic species generated in homogeneous ACN solutions and TiO2 suspensions in the presence SHA and ROS was performed previously in our laboratory [74].

The concentration of molecular oxygen in aqueous TiO2 suspensions play an important role during the oxidation of 4-oxo-2,2,6,6-tetramethylpiperidine (TMPO) to the radical product 2,2,6,6-tetramethylpiperidine *N*-oxyl (Tempone; *a*N = 1.617 mT, *a*13C(413C) = 0.610 mT; *g* = 2.0054). In the photoexcited system TiO2/TMPO/water/air the concentrations of Tempone was very low, and the prolonged irradiation led to a total disappearance of the EPR signal (data not shown). However, the saturation of the aqueous TiO2 suspension by oxygen led to the continuous growth of the EPR signal of Tempone (*a*N = 1.479 mT; *g* = 2.0057) as shown in Figure 9a. The addition of sodium azide, a widely used water-soluble singlet oxygen quencher, to the TiO2/TMPO/water/O2 systems completely suppressed the Tempone generation. However this result should be very carefully analyzed, since besides the singlet oxygen, azide anions also react very fast with the hydroxyl radicals (*k* = 1.4 × 1010 Mí<sup>1</sup> s<sup>í</sup><sup>1</sup> [55]) producing the azide radical • N3 [96] detected here as the corresponding spin-adduct • DMPO-N3 in the photoexcited system TiO2/DMPO/water/NaN3/air (Table 2). Despite the limited water solubility of ȕ-carotene, an analogous inhibition of TMPO photooxidation was observed also when ȕ-carotene as an effective singlet oxygen quencher [97] was added to the TiO2 suspensions (Figure 9b). However, due to the lack of specificity the alternative reaction pathways of ȕ-carotene with the radical species generated in the irradiated titania suspensions must be considered [98]. The role of hydroxyl radicals in the SHA oxidation was further demonstrated in the mixed solvent containing DMSO, where the total inhibition of Tempone formation was found, due to the effective scavenging of hydroxyl radicals by DMSO (Figure 9c). The increased concentration of dissolved molecular oxygen in acetonitrile (Table 3), as well as longer lifetime of <sup>1</sup> O2 in this solvent (Table 1) resulted in the effective oxidation of TMPO to Tempone. The higher <sup>3</sup> O2 concentration in the systems TiO2/TMPO/ACN/air is reflected also in the spectral linewidth growth with not-resolved 13C-satellites (Figure 9d).

**Figure 9.** The sets of individual EPR spectra (*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 sterically hindered amine TMPO: (**a**) oxygenated water; (**b**) oxygenated water saturated with ȕ-carotene; (**c**) oxygenated mixed solvent water/DMSO (5:1 v:v); (**d**) aerated ACN. TiO2 concentration 0.167 mg·mL<sup>í</sup><sup>1</sup> , *c*0,TMPO = 0.008 M.

#### **3. Experimental Section**

The commercial titanium dioxide Aeroxide® P25 (Evonic Degussa, Essen, Germany) was used and stock suspensions containing 1 mg TiO2 mL<sup>í</sup><sup>1</sup> were prepared in redistilled water, dimethylsulfoxide (Merck, Darmstadt, Germany, SeccoSolv®, max. 0.025% H2O), acetonitrile (Merck, SeccoSolv®, max. 0.005% H2O), methanol (spectroscopic grade, Lachema, Brno, Czech Republic), and ethanol (for UV spectroscopy, MikroChem, Pezinok, Slovak Republic). The isotopically enriched water-17O (20%–24.9% atom. 17O) and deuterated DMSO-*d*6, both from Sigma-Aldrich (Buchs, Switzerland), were used as co-solvents. The stock TiO2 suspensions were homogenized for 1 min using ultrasound (Ultrasonic Compact Cleaner TESON 1; Tesla, PiešĢany, Slovak Republic). The spin trapping agent 5,5-dimethyl-1-pyrroline *N*-oxide (DMPO, Sigma-Aldrich) was distilled prior to use. 5-(Diisopropoxyphosphoryl)-5-methyl-1-pyrroline *N*-oxide (DIPPMPO, Enzo Life Sciences, Farmingdale, NY, USA), 5-(ethoxycarbonyl)-5-methyl-1-pyrroline *N*-oxide (EMPO; Enzo Life Sciences), Į-(4-pyridyl-1-oxide)-*N*-*tert*-butylnitrone (POBN; Janssen Chimica, Geel, Belgium), 2,3,5,6,-tetramethylnitrosobenzene (nitrosodurene, ND, Sigma-Aldrich) and 3,5 dibromo-4-nitrosobenzene sulfonate (DBNBS, Sigma-Aldrich) were used without extra purification. All spin traps were stored at í18 °C. The stock solutions of the spin trapping agents were prepared in studied solvents, apart from the ND, characteristic with a limited solubility in polar solvents, which was applied in a saturated suspension directly before the specific experiments. The concentrations of spin traps applied were chosen in order to minimize the undesired photochemical reactions of the spin traps and to gain the effective trapping of photogenerated radical species. The sterically hindered amine 4-oxo-2,2,6,6-tetramethylpiperidine (TMPO, Merck-Schuchardt, Hohenbrunn, Germany) was used as supplied. Sodium azide (analytical grade, Sigma-Aldrich) and ȕ-carotene (UV grade, Sigma-Aldrich) were applied as the singlet oxygen quenchers. Concentrations of the photogenerated paramagnetic species were determined using solutions of 4-oxo-2,2,6,6 tetramethylpiperidine *N*-oxyl (Tempone, Sigma-Aldrich) as the calibration standards.

The TiO2 P25 suspensions containing the spin trapping agent or the TMPO was mixed and carefully saturated with air or oxygen using a slight gas stream immediately before the EPR measurement. So prepared samples were transferred to a small quartz Àat cell (WG 808-Q, optical cell length 0.04 cm; Wilmad-LabGlass, Vineland, NJ, USA) optimized for the TE102 cavity (Bruker, Rheinstetten, Germany) of the spectrometer X-band EPR spectrometer (EMXplus, Bruker). During the EPR photochemical experiments the samples were irradiated at 295 K directly in the EPR resonator, and the EPR spectra were recorded *in situ* during a continuous photoexcitation or after a defined exposure. As an irradiation source a UV LED monochromatic radiator (Ȝmax = 365 nm; Bluepoint LED, Hönle UV Technology, Gräfelfing/München, Germany) was used. The irradiance value (Ȝmax = 365 nm; 15 mW·cm<sup>í</sup><sup>2</sup> ) within the EPR cavity was determined using a UVX radiometer (UVP, Upland, CA, USA). In some cases, argon saturation needed to be applied after the irradiation of the aerated suspensions prior to the subsequent EPR experiment to get better resolved spectra by suppressing the line-broadening effect of molecular oxygen.

Typical EPR spectrometer settings in a standard photochemical experiment were: microwave frequency, ~9.424 GHz; microwave power, 10.53 mW; center field, 335.6 mT; sweep width, 8–16 mT; gain, 1 × 10<sup>5</sup> to 1 × 10<sup>6</sup> ; modulation amplitude, 0.05–0.1 mT; scan, 20 s; time constant, 10.24 ms. The *g*-values (±0.0001) were determined using a built-in magnetometer. The EPR spectra so obtained were analyzed and simulated using the Bruker software WinEPR and SimFonia and the Winsim2002 [99].

#### **4. Conclusions**

The EPR spin trapping experiments using a variety of spin trapping agents (DMPO, EMPO, DIPPMPO, POBN, DBNBS and ND) were performed to identify reactive intermediates formed upon irradiation of TiO2 suspended in water and organic solvents. The role of water in the photoinduced generation of the hydroxyl radical spin-adduct • DMPO-OH in aerated aqueous TiO2 systems was evidenced using 17O-enriched water. Application of a water-soluble nitroso spin trapping agent DBNBS confirmed the production of methyl radicals when DMSO was added to the aqueous TiO2 suspensions and the addition of DMSO-*d*6 revealed also the origin of these radicals. The photoexcitation of TiO2 in non-aqueous solvents (DMSO, ACN, methanol and ethanol) in the presence of spin trapping agents showed the stabilization of superoxide radical anions generated via electron transfer reaction to molecular oxygen, as well as the production of various oxygen- and carbon-centered radicals from the solvents. The oxidation of sterically hindered amine TMPO to radical Tempone via ROS was monitored in aqueous and acetonitrile TiO2 suspensions.

The results obtained demonstrate that indirect EPR spectroscopy techniques represent valuable tools for the characterization of radical intermediates generated in irradiated TiO2 suspensions. However, a careful selection of the experimental conditions and a precise analysis of the experimental EPR spectra considering alternative reaction pathways is an important aspect of any successful application of these indirect techniques in the characterization of TiO2 photoactivity.

#### **Supplementary Materials**

Supplementary materials can be accessed at: http://www.mdpi.com/1420-3049/19/11/17279/s1.

#### **Acknowledgments**

This work was financially supported by Scientific Grant Agency of the Slovak Republic (Project VEGA 1/0289/12) and Slovak University of Technology in Bratislava Young Researcher Grant (Z. Barbieriková).

#### **Author Contributions**

Dana Dvoranová and Vlasta Brezová designed experimental research; Dana Dvoranová, Zuzana Barbieriková and Vlasta Brezová performed analysis of experimental data and wrote the paper. All authors read and approved the final manuscript.

#### **Conflicts of Interest**

The authors declare no conflict of interest.

#### **References**


*Sample Availability*: Not available.
