Search

The Surfaces Special Issue entitled “Surface Aspects of Semiconductor Photochemistry” is mainly devoted to the 7th International Conference on Semiconductor Photochemistry (SP7), which was held on 11–14 September 2019 in Milano, Italy, in the beautiful Renaissance “Ca’ Granda” main building of the University Milan [...]

BiVO4 is an efficient and stable visible active photoanode material. However, due to its unfavorable conduction band position which falls below the H2 reduction potential, it fails to carry out the complete water splitting reaction. On the other hand, larger band gap TiO2 is able to photocatalytically split water, thanks to its negative conduction band energy. Aiming at verifying the possibility of sensitizing TiO2 with BiVO4 and employing the so obtained composite material in photocatalytic water splitting under visible light, we prepared and photoelectrochemically characterized TiO2/BiVO4 heterojunction electrodes. The photocatalytic reduction of methyl viologen, an electron acceptor probe with a reduction potential close to that of protons, was used to evaluate the reducing ability of the photoactive materials under visible light. An apparently counterintuitive electron transfer from photoexcited BiVO4 to the TiO2 conduction band occurs in the TiO2/BiVO4 heterojunction, resulting in TiO2 sensitization and production of highly reductive electrons, which appears to be favored by the band alignment occurring at the heterojunction.

The effect of Pt deposition on TiO2 and of Pt particle size on the photocatalytic steam reforming of methanol was studied by in situ attenuated total reflectance infrared spectroscopy (ATR-IR). Two 0.5 wt.% Pt/TiO2 samples were investigated, one possessing Pt nanoparticles of ca. 4 nm mean size, the other Pt clusters of ca. 1.3 nm mean size showing significantly different photoactivity in terms of both hydrogen production rate and selectivity to CO, CO2 and all other by-products. The presence of Pt nanoparticles strongly affected both the adsorption/desorption and the reactivity properties of the TiO2 surface. Moreover, the variation of the IR spectrum background upon UV–vis irradiation proved that the photopromoted electrons can be trapped by the Pt particles with the consequent increase of electron-hole separation. Reducing the Pt size from nanoparticles to clusters increased the rate of methanol and water absorption and hindered the detrimental formation of irreversibly adsorbed CO on Pt. All of these aspects contribute to increase the photocatalytic performance of Pt cluster-decorated TiO2 with respect to Pt nanoparticles containing TiO2. Finally, prolonged exposure of all samples to methanol/water vapour in the dark led to the formation of unreactive formate which persisted also under UV–vis irradiation. By contrast, this spectator species did not form when the sample was exposed to methanol/water vapour under UV–vis irradiation.

Aiming at clarifying the interplay on TiO2 photoactivity between particle morphology and surface fluorination, the photocatalytic performance of anatase nanocrystals, characterized by a pseudo-spherical shape or a nanosheet structure, is investigated in both a reduction and an oxidation reaction, either in the absence or in the presence of added fluoride anions. Cr(VI) photocatalytic reduction is strongly favored by a large exposure of anatase {001} facets; however, surface fluorination leads in this case to a morphology-independent photoactivity decrease, due to the decreased adsorption of the reaction substrate. More interestingly, a beneficial synergistic effect between the platelet-like anatase morphology and TiO2 surface fluorination is clearly outlined in Rhodamine B photocatalytic degradation, possibly resulting from the intrinsic ability of fluorinated {001} anatase facets of boosting •OH radical mediated oxidation paths, due to their larger amount of surface –OH groups, as revealed using Fourier-transform infrared spectroscopy.

A novel Ni-based 2,2′-bisdipyrrinato complex bearing peripheral benzoic acid groups (Ni-bisdpmCOOH) has been synthesized and characterized. Electrochemical investigations revealed a high energy LUMO level that, combined with its absorption properties extending up to the NIR region, is compatible with its use as a sensitizer in thermodynamically uphill photocatalytic reactions, such as proton and carbon dioxide photoreduction. A sensitized photo-anode was prepared by adsorbing Ni-bisdpmCOOH onto a TiO2 film, which showed an incident photon to current efficiency, measured at 0.62 V vs. RHE, extending up to the visible light region. This demonstrates that electronically excited Ni-bisdpmCOOH is effectively able to inject electrons into the conduction band of TiO2, thus opening the way for effective TiO2 sensitization or MOF implementation of the molecule itself for energetic applications.

Nanocrystalline materials are widely employed as photocatalysts for water splitting applications. In particular, cadmium selenide quantum dots (QDs) attracted growing interest for their superior photon to H2 conversion efficiency and their stability under irradiation. A direct insight is here presented on the effects that CdSe QDs size has on their photoactivity in hydrogen evolution from Na2SO3—containing aqueous solutions under visible light irradiation. Four highly monodisperse CdSe quantum dots with different nanocrystal diameters were synthesized and, after exchange of the capping ligand, employed as photocatalyst in H2 production in the absence of any co-catalyst. Their size-dependent photocatalytic activity is discussed in relation to their absorption properties, their conduction band energy and the recombination probability of the photogenerated charges inside the nanocrystals. High quantum efficiency in H2 generation results from a balance among these three QDs features, their activity scale being not exclusively dominated by the photocatalyst conduction band position.

Bismuth vanadate (BiVO4) with a band gap of ∼2.4 eV has emerged as one of the visible photocatalysts that can absorb light below 520 nm. The electron/hole pairs that are generated following BiVO4 band gap excitation are effective for water splitting, especially when BiVO4 is combined with other metal oxides such as WO3. We report a solution processed method for designing transparent WO3/BiVO4 heterojunction electrodes and observe a synergistic effect on the photoelectrochemical activity of WO3/BiVO4, with the combined system performing dramatically better than either individual component. Using ultrafast transient absorption spectroscopy, we elucidated the electronic interaction between WO3 and excited BiVO4. Moreover, the photocatalytic reduction of thionine by WO3/BiVO4 as well as by each individual oxide component is used to track electron injection processes and determine the energetics of the studied systems. In the composite WO3/BiVO4 film a shifted quasi-Fermi level results, due to electronic equil...

The time-resolved photoluminescence (PL) in the nanosecond time scale of TiO2 materials undoped or codoped with nitrogen and fluorine (N,F-doping) and modified by noble metals (NM, i.e. Au or Pt) nanoparticles (NPs) photodeposition has been systematically investigated in relation to their photocatalytic activity in hydrogen production. The main aim of the study is to elucidate the origin of the NM-dependent synergistic effects in photoactivity produced by N,F-doping of TiO2 and NM NPs deposition on the oxide surface. While TiO2 doping with fluorine and nitrogen introduces new stabilized luminescent defective trap states below the conduction band revealed by long-living PL components, the presence of NM NPs on the TiO2 surface produces a PL intensity suppression, which is more relevant for Au rather than for Pt-NPs containing materials. Time-resolved PL analysis indicates that the electron transfer occurring at the TiO2/metal interface is affected by both the defective structure of anatase N,F-doped TiO2 and the type of NM (Au or Pt). In particular, Au rather than Pt NPs appear to strongly interact with the charge carriers trapped at surface defect sites, gold NPs being expected to preferentially grow on such sites during photodeposition. Furthermore, plasmonic gold NPs excitation upon PL light absorption is evidenced by the PL spectral shape variation observed only in the case of Au/TiO2. Thus, the larger synergistic effect on photocatalytic hydrogen production observed upon Au NPs photodeposition on N,F-doped TiO2 results from the opening of a new efficient electron transfer path from luminescent defective trap states on doped TiO2 to Au NPs, where proton reduction occurs.

The effect of TiO 2 surface fluorination in the photocatalytic degradation of the azo dye Acid Red 1 (AR1) was investigated employing a series of TiO 2 photocatalysts including P25, a rutile and high surface area anatase samples. The effects of (i) pH lowering to 3.7 and (ii) TiO 2 surface fluorination at this pH on the reaction rate under 254 nm or polychromatic, mainly visible light, irradiation and on AR1 adsorption on the photocatalyst surface were taken into account separately. In the case of P25 and rutile TiO 2 photocatalysts the reaction rate did not vary upon pH lowering in the absence of fluoride, but almost doubled upon fluorination at pH 3.7, an effect to be attributed exclusively to the surface fluorine-induced modification of the reaction paths. By contrast, in the case of high surface area anatase samples a pH lowering from natural conditions down to 3.7 led to a remarkable increase of the reaction rate, which however was clearly inhibited upon surface fluorination at pH 3.7, with a trend perfectly matching that of AR1 adsorption on anatase. Thus, reaction paths favored by direct interaction of the substrate with the photocatalyst surface appear to prevail for high surface area anatase photocatalysts with respect to those which take advantage of a less reactive (and less adsorptive) fluorinated photocatalyst surface.

Evolution of pure hydrogen and oxygen by photocatalytic water splitting was attained from the opposite sides of a composite Pt/Ti/TiO2 photoelectrode. The TiO2 films were prepared by radio frequency (RF)-Magnetron Sputtering at different deposition time ranging from 1 up to 8 h and then characterized by X-ray diffraction (XRD), scanning electron microscopy (SEM) and ultraviolet-visible-near infrared (UV-vis-NIR) diffuse reflectance spectroscopy. The photocatalytic activity was evaluated by incident photon to current efficiency (IPCE) measurements and by photocatalytic water splitting measurements in a two-compartment cell. The highest H2 production rate was attained with the photoelectrode prepared by 6 h-long TiO2 deposition thanks to its high content in the rutile polymorph, which is active under visible light. By contrast, the photoactivity dropped for longer deposition time, because of the increased probability of electron-hole recombination due to the longer electron transfer path.

The effect of the crystalline phase of TiO 2 (anatase, rutile and brookite) on its photocatalytic activity in hydrogen production from methanol-water vapours has been investigated by testing a series of both home-made and commercial TiO 2 photocatalysts, either bare or surface-modified by deposition of a fixed amount, i.e. 1 wt%, of platinum as co-catalyst. For all of the TiO 2 samples the rate of hydrogen production increased by one order of magnitude upon Pt deposition, because of the ability of Pt to enhance the separation of photoproduced electron-hole pairs. Under irradiation in the 350-450 nm wavelength range, brookite and anatase showed similar photoactivities, both superior to that of rutile. By contrast, rutile, possessing a narrower band gap, was active also under visible light (λ > 400 nm), whereas no hydrogen evolution was observed with anatase and brookite under such conditions. Surface area proved to be a key parameter, strongly influencing photoactivity. However, as the particle size became ultra-small, the semiconductor absorption edge was blue-shifted because of size quantisation effects, with a consequent decrease in hydrogen production rate due to the smaller portion of incident photons absorbed by the photocatalyst. © The Royal Society of Chemistry and Owner Societies 2011.

Anatase–rutile and anatase–brookite–rutile composite nanocrystals were synthesized by a controlled sol–gel reaction followed by hydrothermal growth and by mechanically mixing pure crystalline phase precursors, respectively. Their physicochemical and photocatalytic properties were investigated in comparison with those of several home-made and commercial pure anatase, rutile and mixed phase TiO 2 samples. In particular, the phase composition and the crystallite size were determined by X-ray diffraction measurements. The photocatalytic degradation of the azo dye Acid Red 1 (AR1) and of formic acid (FA) in aqueous suspension were employed as test reactions. Hydrogen peroxide evolution, the main reductive process occurring in parallel, was also monitored during the runs. The investigated photocatalysts exhibit different photoactivity scales towards the two organic substrates, depending on the prevailing photodegradation path. High surface area anatase samples were particularly active in FA degradation, principally occurring through direct interaction with photoproduced valence band holes, whereas rutile samples, even possessing high surface areas, but exhibiting poor wettability, were scarcely active. A good linear correlation was found between H 2 O 2 evolution and the rate of AR1 photodegradation on the investigated series of photocatalysts, whereas no H 2 O 2 was detected during FA mineralization, apart in the case of a low surface area anatase sample, whose peculiar photocatalytic properties were ascribed to the presence of Ti 3+ ions in its structure.

The rate of 1,4-dichlorobenzene (1,4-DCB) degradation and mineralization in the aqueous phase was investigated either under direct photolysis or photocatalysis in the presence of commercial or sol-gel synthesized TiO2, or under sonolysis at 20 kHz with different power inputs. Two lamps, both emitting in the 340-400 nm wavelength range with different energy, were employed as irradiation sources. Photocatalysis ensured faster removal of 1,4-DCB with respect to sonolysis and direct photolysis. The highest degradation and mineralization rate was attained with the combined use of photocatalysis and sonolysis, i.e. under sonophotocatalytic conditions. The efficiency of the employed advanced oxidation techniques in 1,4-DCB degradation is discussed also in relation to their energy consumption, which might be decisive for their practical application.

The photocatalytic degradation of formic acid (FA) and benzoic acid (BA), chosen as model organic molecules with acidic properties, was investigated in TiO2 and ZnO water suspensions under different experimental conditions. Hydrogen peroxide evolution, formed through a reductive pathway started by conduction band electrons, was also simultaneously monitored during the degradation runs. The effect of different initial amounts of substrates and the dependence of the reaction rate on the initial pH of the TiO2 suspensions was interpreted under the light of a pseudo-steady state Langmuir–Hinshelwood rate form and of the electrostatic interactions occurring at the water–semiconductor interface. ZnO appeared a more effective photocatalyst than TiO2 for BA, but not for FA degradation. A much higher amount of hydrogen peroxide was detected in ZnO irradiated suspensions, both in the presence and in the absence of the substrates, mainly because of its lower photocatalytic decomposition rate on such oxide. The rate of hydrogen peroxide evolution during the photocatalytic oxidation of BA on TiO2 could be related to the rate of the oxidation process, while H2O2 could not be detected during the photocatalytic degradation of FA on this oxide, mainly because of the reduced shielding ability of this substrate.

The mechanism of atrazine degradation in aqueous phase was investigated under sonolysis at 20 kHz, ozonation, photolysis at 254 nm and photocatalysis in the presence of TiO2, employed either separately or in combination. Ozonation and photocatalysis induced atrazine de-alkylation, followed by slower de-chlorination, while direct photolysis at 254 nm produced very efficient de-chlorination. Simultaneous sonolysis had beneficial effects on ozonation and photocatalysis, especially by increasing the rate of photocatalytic de-alkylation, and no effect on the unimolecular photolytic de-chlorination of atrazine. Although complete atrazine mineralization could not be attained, because of the stability of the s-triazine ring toward oxidation, atrazine degradation and overall detoxification, as related to the disappearance of chlorinated by-products, proceeded at the highest rate when photolysis at 254 nm was combined with ozonation.

Photocatalysis on TiO2, with irradiation wavelength λirr=254 nm or 315 nm

The degradation of 2-chlorophenol and of the two azo dyes acid orange 8 and acid red 1 in aqueous solution was investigated kinetically under sonolysis at 20 kHz and under photocatalysis in the presence of titanium dioxide particles, as well as under simultaneous sonolysis and photocatalysis, i.e. sonophotocatalysis. The influence on the degradation and mineralisation rates of the initial substrate concentration and of the photocatalyst amount was systematically investigated to ascertain the origin of the synergistic effect observed between the two degradation techniques. The evolution of hydrogen peroxide during kinetic runs was also monitored. Small amounts of Fe(III) were found to affect both the adsorption equilibria on the semiconductor and the degradation paths. Ultrasound may modify the rate of photocatalytic degradation by promoting the deaggregation of the photocatalyst, by inducing the desorption of organic substrates and degradation intermediates from the photocatalyst surface and, mainly, by favouring the scission of the photocatalytically and sonolytically produced H(2)O(2), with a consequent increase of oxidising species in the aqueous phase.

Aiming at understanding how plasmonic reactions depend on important parameters such as metal loading and strong metal-support interaction (SMSI), we report the plasmonic photodegradation of formic acid (FA) under green LED irradiation employing three TiO2 supports (stoichiometric TiO2, N-doped TiO2, black TiO2) modified with Au nanoparticles (NPs) 3-6 nm in size. The rate of FA photo-oxidation follows different trends depending on Au loading for stoichiometric and doped Au/TiO2 materials. In the first case, the only contribution of hot electron transfer produces a volcano-shaped curve of photoreaction rates with increasing the Au loading. When TiO2 contains intra-bandgap states the photoactivity increases linearly with the Au NPs amount, thanks to the concomitant enhancement produced by hot electron transfer and plasmonic resonant energy transfer (PRET). The role of PRET is supported by Finite-Difference Time-Domain simulations, which show that the increase of both A u NPs inter-distance and of SMSI enhances the probability of charge carrier generation at the Au/TiO2 interface.

Photocatalysis and sonolysis exhibit synergistic effects in the degradation of organic molecules in aqueous suspension, when low ultrasound frequency (i.e. 20 kHz) is employed. The degradation of the azo dye Acid Orange 8 in aqueous suspensions was systematically evaluated under sonolysis, photocatalysis and sonophotocatalysis as a function of dye concentration, amount and type of photocatalyst (TiO2 or ZnO) and catalase addition, in terms of pseudo-first-order degradation rate constants. The evolution of hydrogen peroxide was also monitored. Synergistic effects of the combination of ultrasound and photocatalysis should mainly involve the aqueous phase and be due to an increased concentration of reactive radicals, consequent to the action of ultrasound on the peroxide species produced by both photocatalysis and sonolysis.

The degradation of 2-chlorophenol in water was kinetically investigated using the following different techniques, employed either separately or simultaneously, always with the same experimental set-up: light irradiation (315-400 nm), sonication, photocatalysis with different types of TiO2, photocatalysis with sonication. The influence of the reaction volume and of different gas mixtures, containing Ar and O2, as well as O2/O3, was also investigated. Finally, an energetic comparison among these different techniques was performed, focused on an industrial application of some of them.

The photoreversible EZ isomerization of a series of protonated N -benzylideneanilines (X-C 6 H 5 CH=NC 6 H 5 -Y; X ≡ H, N(CH 3 ) 2 ; Y ≡ H, NO 2 ) was investigated in concentrated sulphuric acid solution. The UV absorption spectra of the E isomers are practically independent of substituents on benzene rings and suggest a planar conformation for protonated N -benzylideneanilines. The UV spectra of Z isomers were obtained from photostationary state conditions at two wavelengths. Also the quantum yields of E → Z and Z → E photoisomerization are not markedly affected by the ring substituents and are compatible with a photoreversible process occurring through a common excited state.

Aiming at discerning the role of fluorine from that of nitrogen as a dopant in N,F-codoped TiO2, a series of HF-doped TiO2 photocatalysts were investigated in the decomposition of formic and acetic acid in aqueous suspensions, also as a function of the irradiation wavelength (action spectra analysis), in comparison with recent results obtained with an analogous series of NH4F-doped TiO2 photocatalysts. Visible light absorption around 420 nm, which was found to be inactive in acetic acid decomposition, is definitely shown to be associated with nitrogen doping, whereas the enhanced absorption at ca. 365 nm, increasing with increasing calcination temperature, can be unambiguously attributed to structural modifications induced by fluorine doping. Action spectra analysis confirms that this absorption is active in acetic acid decomposition, in both HF- and NH4F-doped TiO2 photocatalysts. From time-resolved photoluminescence (PL) spectroscopy analysis, a clear correlation is outlined between the photoactivity of...

Metal-loaded TiO 2 is, by far, one of the most important class of photocatalysts in hydrogen production through photoreforming of organics and water photosplitting. In this study anatase loaded with Au and Pt nanoparticles (Au/TiO 2 and Pt/TiO 2 ) by an impregnation-reduction method was investigated as for morphological, electronic (XPS) and photocatalytic properties in hydrogen production by methanol photoreforming. The electron and hole trapping centers, Ti 3+ and O − , respectively, formed under UV–vis irradiation of the photocatalysts, were studied by in situ electron spin resonance (ESR) spectroscopy. The nature of the loaded metal affected both the H 2 evolution rate and the distribution of the methanol oxidation products. The better performance of Pt/TiO 2 is attributable to the greater ability of Pt with respect to Au to act as electron sink, slowering the recombination of photoproduced electron–hole couples. Direct evidence of this effect was obtained by ESR analysis, showing that the amount of Ti 3+ active sites follows the order TiO 2 > Au/TiO 2 ≫ Pt/TiO 2 , thus confirming easier electron transfer from Ti 3+ to Pt, where the H + reduction to H 2 occurs.

Propagation radicals generated during the photopolymerization and photo-cross-linking of a trimethacrylate and a vinylmethacrylate monomer have been characterized by EPR and ENDOR spectroscopy at different temperatures. Also, radical decay has been investigated at 363–423 K. The EPR spectral shape reflects a fast exchange process between two radical conformations and is sensitive to the mobility of the polymeric structure surrounding radicals. ENDOR spectra show 1H-matrix lines and a band around 45 MHz, which has been attributed to the interaction of the unpaired electron with the hydrogen nuclei of a freely rotating methyl group. Radical decay and dynamic phenomena involving radicals have been interpreted in the light of a dispersive kinetic model, which allows the evaluation of useful parameters for polymer characterization.

The photo-cross-linking of 18.4-20.0 μmol of diallyl oxydiethylene dicarbonate ( I ) grafted onto cellulose (per square contimetre of apparent cellulose surface) was examined kinetically at 30±2 °C, on exposure to an incident polychromatic radiation flux of (2.05±0.08) × 10 −7 einstein s −1 cm −2 , in the presence of catalytic amounts (20-200 ppm) of three vanadium- ( V ) organometallic compounds. As with the diacrylate systems studied previously, the kinetics of photo-cross-linking of grafted I followed a multi-relaxation spectrum in terms of two parameters (width of spectrum β and mean lifetime τ). The β values (0.42±0.04) were almost the same in all experiments and were similar to that (0.44±0.02) observed for the thermal decay of double bonds of pre-irradiated I in bulk. The β parameter therefore seems to reflect the cross-linked character of the linear polymer pattern, rather than the cross-linking process, either thermal or photocatalytic. In contrast, the τ values reflect, in a very sensitive manner, the effect of the concentration and type of photocatalytic system employed. An acceleration factor of about 190 was reached with 200 ppm vanadium (synergic mixture of vanadium (V) isopropoxide and isobutoxide). Some relevant features of the photo-cross-linking mechanism are discussed in the light of the proposed kinetic model.

Direct experimental evidence of the higher concentration of hydroxyl radicals generated on fluorinated titanium dioxide (F–TiO2) under irradiation was obtained by spin-trapping EPR measurements. The faster photoinduced bleaching of the azo dye Acid Red 1 (AR1) observed in the presence of F–TiO2 was explained by the high affinity of the azo double bond towards ˙OH radicals. Moreover, the pronounced decrease of the AR1 bleaching rate by addition of 2-propanol, as hydroxyl radicals scavenger, on F–TiO2 and not on naked TiO2 demonstrated that on fluorinated titania AR1 is mainly degraded via ˙OH radical attack.

Photochemical grafting of diallyl oxydiethylene dicarbonate onto cellulose (in the form of filter paper) was investigated kinetically at 20 ± 2 °C with 1,2-diphenyl-2,2-dimethoxyethanone as photoinitiator. Irradiations were carried out polychromatically with an incident radiation flux of (1.25 ± 0.07) × 10−7 einstein s−1 cm−2. The ratio R between the molar concentrations of photoinitiator and monomer was varied in the range 0.032–0.150 and the n/S value (n moles of monomer adsorbed per unit apparent surface S) was varied in the range 8.40–43.8 μmol cm−2. Two consecutive rate processes were shown by the kinetic curves, with quantum yields differing by about two orders of magnitude. Quantum yields Φ1 for the first, almost constant, rate period were measured as a function of R and n/S. Kinetic features of the grafting process of diallylic monomer are discussed in the light of a general mechanism proposed for grafting of diacrylates and the photochemical reactivity of the allylic vs. the acrylic function.

The geometrical photoisomerization reaction of the title molecule ( trans -I) has been studied in methanol/water 90/10 wt.% mixtures containing sulphuric or hydrochloric acid. Quantum yields of the process, measured under acidic conditions in the range 0.095 N + ] trans -I as a function of pH, experimental quantum yields have been interpreted in the light of prototropic equilibria in the excited state. Mechanistic models are proposed which are able to fit results obtained in the investigated experimental conditions.

The degradation of methyl tert-butyl ether (MTBE) in water was kinetically investigated in a O(2)/Ar 80:20 atmosphere employing either sonolysis at 20 kHz, or photocatalysis on TiO(2) (with 315 nmlambda(irr)400 nm), or simultaneous sonolysis and photocatalysis (i.e. sonophotocatalysis), as degradation techniques. In all investigated conditions, MTBE concentration decreased according to a first order rate law; under ultrasound the degradation rate was stirring-dependent. The time profile of the reaction intermediates gave information on the reaction paths prevailing under the different experimental conditions. The energy consumption of the employed degradation techniques was also evaluated, which might be decisive for their practical application.

The visible light-induced decomposition of two series of 4-diethylaminoazobenzenes has been studied in solvents of different polarity, in the presence and absence of O2, in order to clarify the reaction mechanism. Two different paths have been recognised, whose importance depends on the nature of the solvent; both involve the formation of organic free radicals. Decomposition is much faster in acetone than in the other solvents. A strong inhibiting effect of molecular oxygen has been observed in all cases.

The stability to visible light of five 4-diethylaminoazobenzenes was investigated in different solvents under various experimental conditions, as well as in solid crystalline form and on dyed poly(ethyleneterephthalate). Though generally exhibiting relatively high photostability, all investigated dyes underwent photodegradation in acetone solution. Two different reaction paths were recognised in this solvent, depending on the nature and position of the substituting groups on the second aromatic ring, both involving the formation and subsequent fragmentation of organic free radicals, as suggested by the strong inhibiting effect of molecular oxygen on the photodegradation rate.

l.nng-livcd Impped rndlcals prodnced during Ihe pholopolymerizalion ' or bulane-I,4-diol diaCT}'1I1te (DODA) were studied b}' EPR aod ENOOR speclro 'coPY IInd tbcir Ihermul dCCll)' "'us followed by EPR measuremeob al different Icrnpenilures (40-120 °(:) l'or killelic ~tud). The EPR signal ~howcd !he superimposition 01' Iwo dill'erenl poHems, Il Ihree-line nnd a single-line speclrum. BOlh EPR pllllcrns decayed following flrsl-ordcr kinelics in Ihe investigated lemperalure mnge. Aclh'alion parameters of the d~cay were ublained. rIle EPR pnllern werc nllribulcd III Ihe same radical species silllaled in l1uid and in eros -lillked regions Clf Ihe phOIOpol}meriLCd BODA. Radicals 01' Ihe lalter kind undergo eleclron spin exchange slrong enough lo wasll out the h)'perfine splilling. The single-lIne widlh is mainly determjDed by electron spin dipole-dipole inleractioos. Thl' ENDOR re~ponse is unI)' of Ihe malrix killd, t~'pic,,1 o\' radicals in a solid pllase. l'hl' preseol model WAS 9150 compured with receDI titeralure rep(lns.

The formation of trapped radicals under UV irradiation of ethylene glycol dimethacrylate in the presence of different amounts of photoinitiator, as well as radical decay at different temperatures (90–150 °C) after the end of irradiation, have been studied by electron paramagnetic resonance (EPR) spectroscopy. A fitting procedure of the EPR signals revealed that the nine-line EPR spectrum can be attributed to the propagating radical, undergoing a fast exchange process between two conformations. This radical also gives an ENDOR spectrum even at room temperature, thus suggesting that the polymer structure is essentially solid. At least one different radical species is present in highly photo-cross-linked samples. At temperatures above 100 °C, the hyperfine structure of the nine-line EPR spectrum is partly washed out, owing to a spin–spin exchange phenomenon between radicals. This becomes more evident with increasing degree of polymerization and cross-linking during radical decay kinetic runs and also affects the rate of exchange between the two radical conformations.

Radicals trapped during the photopolymerization and photo-cross-linking of 2-ethyl-2-hydroxymethylpropane-1,3-diol triacrylic ester have been studied by EPR spectroscopy at different temperatures (25–180 °C) and their thermal decay has been followed at 120 °C and 150 °C. Performing a fitting procedure on the EPR signals showed that they are composed of two kinds of patterns, which superimpose with the same g value. One of them is a single-line spectrum, while the other is hyperfine-structured. An exchange between two radical conformations is revealed by the latter, with the rate depending both on temperature and on the molar fraction of radicals giving the single-line spectrum. This suggests that both species are involved in the exchange mechanism. Termination occurs preferentially between radicals giving the same EPR pattern. Second-order termination rate constants depend on the photoinitiator concentration in the irradiated photoreactive mixture and give information about the microstructure of the polymer.

The formation of trapped radicals under UV irradiation during the photopolymerization of butane-1,4-diyl diacrylate in the presence of different amounts of photoinitiator, as well as the radical decay after the end of irradiation, have been studied by EPR spectroscopy. Direct evidence that higher photoinitiator radical concentrations greatly favour radical–radical encounters is obtained. The formation of the same radical species in two different polymeric phases, one essentially rigid and the other more fluid, is at the origin of the two observed superimposed EPR patterns. The ratio R between them decreases with increasing photoinitiator concentration, owing to the formation of shorter and consequently more mobile polymer chains. In contrast, the more rigid phase seems to dominate in photopolymerized diallyl 3-oxapentane-1,5-dicarbonate. ENDOR measurements further support the correlation proposed by us between EPR spectral shape and the rigidity of the polymeric phase.

Direct evidence of the presence of extremely long-lived radicals produced in the photopolymerization of 1,6-hexanediol diacrylate and tetraethyleneglycol diacrylate has been obtained by EPR spectroscopy. As radical decay in thermal after-treatment is accompanied by additional polymerization and cross-linking, kinetic studies were carried out at different temperatures (40-120 °C) both by EPR spectroscopy and by analysis of the parallel disappearance of double bonds. Apart from EPR data at 120 °C, all experimental results show that both radical decay and double-bond conversion are first-order reactions, and kinetic constants evaluated for the two processes are practically coincident at the same temperature. Activation parameters for thermal after-treatment of the two diacrylates are almost the same, owing to the similarity in their chain length and mobility. The behaviour of the system at 120 °C, accounted for by a multiple relaxation model, is interpreted as being due to prevailing cross-linking reactions in the vitrified network of trapped radicals.

The solvent dependence of the kinetics of the geometrical isomerization of trans -3-(2-hydroxybenzylidene)-4,5-dihydrofuran-2(3 H )-one and the subsequent intramolecular acylation of the cis substrate was studied. Even if the quantum yields Φ 1 and Φ 2 of the two processes increase on increasing the polarity of the methanol solutions, addition of water is found to quench dramatically Φ 1 and Φ 2 . This behaviour is interpreted in terms of the proton exchange equilibria of the excited singlet state. A detailed analysis of the dependence of the fluorescence on the pH was carried out and a mechanistic model which is able to describe satisfactorily the experimental results is proposed.

The influence of radiation intensity (from 1,08 · 10−7 to 2,48 · 10−4 einstein.s−1 · 1−1) on the photopolymerization of diallyl oxydiethylene dicarbonate was investigated systematically not only in the initial stages of reaction, but also at high monomer conversion (x), in order to detect possible variations in the polymerization mechanism. In some of the runs irradiation was stopped at x = 0,55–0,65 and thermal after-treatments were carried out both in the presence and absence of air by following kinetically x-values during the self-decelerating cross-linking. To rationalize experimental data a model is proposed which assumes two basically different mechanisms occurring simultaneously (bimolecular termination and radical trapping), but with different “weights” as a function of time. Long living radicals produced by irradiation continue to react in the dark. The rate of this reaction, which is enhanced in the presence of oxygen, is fitted by a relaxation model that considers double bonds, particularly of pendant groups, as traps, with increasing lifetimes, able to transfer to radical sites. The role of oxygen is also discussed.

Photoinitiated grafting and graft polymerization of liquid mixtures, adsorbed onto polypropylene, polycaprolactam, and poly(ethylene terephthalate) films, containing one of four acryloxy-substituted aromatic diazenes and one of four diacrylate comonomers, in the presence of 1,2-diphenyl-2,2-dimethoxyethanone as photoinitiator, were investigated kinetically at 30 ± 2°C. Irradiation was carried out polychromatically, with impinging photoenergy from 2.1 × 10−8 to 20.5 × 10−8 einstein s−1 cm−2; in some of the runs the ultraviolet radiation was filtered. The ratio R between the molar concentration of photoinitiator and the sum of concentrations of dye and diacrylate varied between 0.005 and 0.095; the ratio M between the molar concentration of dye and diacrylate varied between 0.005 and 0.046. The moles of dye and diacrylate n initially deposited per unit apparent polymeric surface S varied between 2 and 74 μmol cm−2. No selective effect was shown by the presence of dyes in the comonomer mixture. The surface density of grafted molecules at the end of the grafting process was not affected by the photoinitiator concentration (for 0.030 < R < 0.095), by diacrylate or dye concentrations, or by the kind of polymer substrate. This parameter, on the contrary, clearly depended on n/S, and linearly up to n/S ≅ 30 μmol cm−2. Quantum efficiencies for the two consecutive kinetic processes of grafting and graft polymerization (Φ1 and Φ2, respectively) were evaluated. The dependency of Φ1 on R, as well as of Φ2 on n/S, are critically discussed on the basis of the proposed mechanism involving grafting of an oligomeric chain (Φ1 values up to the order of 102) followed by a step-by-step graft polymerization (Φ2 limiting values of the order of unity). The sensitizing effect of acryloxy-substituted aromatic diazenes on both Φ1 and Φ2 is evidenced.

Asymmetric membranes were prepared by photochemical grafting of 28.1 μmol cm−2 of 1,6-hexanediol diacrylate onto cellulose in the form of filter paper, in the presence of 1,2-diphenyl-2,2-dimethoxyethanone as a photoinitiator, the molar ratio of photoinitiator to acrylated monomer being fixed at 0.086. A fixed polychromatic irradiation flux of 20.2×10−8 einsteins s−1 cm−2 was used. After 10 min irradiation, a nearly quantitative graft yield was reached. By multiple internal reflection IR analysis, a structure with one pendant acrylate group per monomer unit was assigned to the grafted polymer; at longer irradiation times, a further decrease in unsaturation resulting from photo-cross-linking between pendant acrylate groups was observed. Permeabilities P and diffusivities D at 25 °C for hydrogen and carbon monoxide were measured for membrane specimens obtained at different irradiation times. An increase in P reflects a decrease in unsaturation, while the D are independent of irradiation time. The selectivity of them embranes is not influenced by photo-cross-linking.

Photoinitiated grafting and graft polymerization of 1,6-hexanediol diacrylate (I) and 4-(N-ethyl,N-2-acryloxyethyl)amino-4′-nitro-azobenzene (II) as co-monomers with a fixed molar ratio I:II equal to 10, in the presence of 1,2-diphenyl-2,2-dimethoxyethanone (III) as photoinitiator with a ratio R of its molar concentration of 0.005-0.38 with respect to the sum of molar concentrations of I and II, were studied using a polypropylene ribbon as a model system of a macroporous substrate. By polychromatic irradiation at 45°C with energy inputs of 90 mW cm−2, ultrathin layers of co-polymer, able to function as asymmetric membranes, could thus be obtained. Mean polychromatic quantum yields of grafting and graft polymerization, as a function of R, were also measured. Membrane specimens with constant values of grafted (200±20) and graft polymerized (500±80) molecules of monomer units in co-polymer per unit apparent surface (A2) have been subjected to measurements of permeabilities (P) and diffusivities (D) at 25°C for oxygen and nitrogen, as a function of R. Results show that only an optimum R minimizes P and D, and a narrow range of R-values allows a normal membrane behaviour.

Photochemically induced grafting (1) and graft polymerization (2) from N,N-dimethylformamide solutions of four acryloxy-substituted aromatic diazenes, chosen as model molecules of acryloxy azo dyes, were investigated kinetically at 25°C onto polypropylene, polycaprolactam, and poly(ethylene terephtalate) films, by polychromatic irradiations. Quantum efficiencies at zero concentration of dyes and lifetimes of excited states for processes (1) and (2) were evaluated from experimental data by Stern–Volmer plots. The surface density of grafted molecules at the end of process (1) was not affected by the concentration of dyes in solution, but varied sensibly with the chemical nature of polymers and dyes. The relevant parameters are discussed on the basis of the proposed mechanism.

Photochemical grafting onto cellulose and successive photo-cross-linking of 2.00 – 37.00 mg cm−2 of an epoxydiacrylate prepolymer based on glycidyl ether of bisphenol A, containing 25 wt./wt.% trimethylolpropane triacrylate, were investiated kinetically at 30 °C in the presence of 1,2-diphenyl-2,2-dimethoxyethanone as a photoinitiator, with the weight concentration ratio of photoinitiator to prepolymer varied between 0.070 and 1.115. Irradiations were carried out polychromatically, in air or under a stream of nitrogen, with incident radiation of flux I ranging from 2.1 × 10−8 to 20.5 × 10−8 einsteins s−1 cm−2. Graft yields as a function of irradiation time were determined gravimetrically: two consecutive kinetic processes were observed which had rates linearly dependent on I. The quantum yields Φ1 of the first constant-rate period showed a linear dependence on photoinitiator concentration, while the quantum yields Φ2 for the second constant-rate process showed an inverse dependence on n/S (moles of prepolymer initially deposited per unit apparent cellulose surface). Kinetic analysis of the degree of unsaturation by multiple internal reflection IR spectroscopy was carried out in parallel and the results indicated that a structure with one pendant acrylate group per monomer unit may be assigned to the grafted polymer. At longer irradiation times, further decrease in unsaturation resulting from photo-cross-linking was observed which had a rate linearly dependent on I0.5, suggesting a bimolecular termination for this process. In the light of this information, the mechanism of photografting of poly-functional acrylate monomers onto polymeric surfaces is discussed.

The possibility of immobilizing reagents and catalysts by photochemical grafting of their composites with an epoxy-diacrylate resin has been investigated. Two model systems have been examined: (i) the catalytic decomposition of aqueous sodium hypochlorite by cobalt peroxide, with oxygen production; (ii) the reaction of alkali metals with alcohols. For case (i) the catalyst was embedded, in the form of a mixture with 90% of zeolite 13X, into the photografted polymer matrix; while for case (ii) the alkali metal reagent, supported onto 90% of zeolite 13X, was incorporated into the microporous membrane. Kinetics of reaction (i) were studied in standard conditions (1.0 M NaOCl; 1.0 M NaCl; 0.25 M NaOH) between 288.2 to 318.2 K, and the relation between pseudp-first order kinetic constants and amount of catalyst incorporated into the membrane determined. Isotope effects for reaction (ii), relative to H2, D2, HD production, were measured with mixtures of deuterated and non-deuterated CH3OH(D), C2H5OH(D), n-C3H7OH(D), and i-C3H7OH(D) at 25 °C, as a function of volume percent of non-deuterated alkanols in the liquid phase. For reaction (i) reactivity was found to be substantially the same in the membrane reactor as in ordinary chemical conditions: loss of catalytic activity did not exceed 15%. For reaction (ii) isotope effects, the origin of which is discussed, fully coincided with those measured in the absence of the polymer matrix.

Kinetic studies of photochemical I y induced grafting [1] and graft polymerization [2] onto polypropylene fibres from N, N–dimethylformamide solutions of four acryloxy–substituted aromatic diazenes, chosen as model molecules of acrylated azo dyes, were carried ou t at 25°C by monochromatic irradiation at various wavelengths. The effect of dye concentration was systematically investigated, and quantum efficiencies at zero concentration of dyes, as well as lifetimes for processes (1) and (2), were evaluated using Stern– Volmer plots. The marked influence of wavelength and the constancy of the relatively long lifetimes at the various wavelengths support the hypothesis that quantum yields reflect efficiencies of conversion from the states populated by light absorption to reactive upper triplets.

Photochemical grafting of 1,6-hexanediol diacrylate, diethylene glycol diacrylate and tetraethylene glycol diacrylate onto cellulose in the form of filter paper was investigated kinetically at 30 ± 2 °C with 1,2-diphenyl-2,2-dimethoxyethanone as a photoinitiator, under experimental conditions suitable for the preparation of membrane systems. Irradiations were carried out polychromatically, with incident radiation of flux (2.5 × 10−8)−(22.6 × 10−8) einsteins s−1 cm−2. The ratio R between the molar concentrations of photoinitiator and monomer were varied between 0.005 and 0.095, while the n/S value (n, moles of diacrylate monomer deposited per unit apparent surface S) was varied between 7.61 and 75.0 μmol cm−2. Two consecutive constant-rate processes were indicated by the kinetic curves. The quantum yields Φ1 and Φ2 for the first and second constant-rate periods were calculated. They are almost insensitive to the difference in diacrylate monomer structure; the Φ1 show a linear dependence on R, while the Φ2 show an inverse dependence on n/S. The relevance of photochemical parameters in the proposed reaction mechanism is discussed.

The photocatalytic activity of Au modified titanium dioxide was evaluated in the photodegradation of the azo dye Acid Red 1 (AR1) under 254 nm irradiation. Noble metal nanoparticles were deposited on TiO2 either through depositionprecipitation (DP), or by immobilisation of preformed metallic sols (polyvinylalcohol (PVA)/NaBH4 or tetrakis(hydroxymethyl)phosphonium chloride (THPC)/NaOH systems). Gold nanoparticles on the photocatalyst surface had dimensions of around 3–4 nm in diameter, as determined by HRTEM analysis, and exhibited visible light plasmon absorption. THPC Au/TiO2 appears to be the most photoactive amongst the photocatalysts with a 1 wt.% Au loading, while among THPC samples with different Au loadings (0.5–20.0 wt.%) the maximum photoactivity was attained with 5 wt.% Au/TiO2. The higher AR1 photodegradation rate observed on Au/TiO2 at basic pH can be related to the higher concentration of hydroxyl anions at the interface: these are able to effectively scavenge photoproduced valence band holes, possibly in competition with Au0 oxidation to Au+.

Photo-induced grafting and graft polymerization onto polymers promise to become an elegant and powerful method for modifying surface properties. With the wide variety of vinyl and other monomers available today, photochemical grafting may become an established technical method to impart desired surface characteristics: from anti-static behavior to thermal stability, from adhesion to bacteriocidal properties, from water repellency to hygroscopicity, from soil resistance to flame retardant properties, from optical imaging to dyeability and so on. Photochemistry may consequently play a very important role in this field, by promoting knowledge, by studying applications, as well as optimizing conditions of industrial processes, potentially capable of being developed.