Introduction

The occurrence of contaminants of environmental concern (CEC) in surface waters has been extensively reported (Bound and Voulvoulis 2004; Emblidge and De Lorenzo 2006; Vecchiato et al. 2015). These compounds have the potential to impact human and ecosystem health (U.S. Environmental Protection Agency 2016), and therefore, their fate and transport in environmental systems is of interest to scientists and engineers. Photochemical processes, such as direct and sensitized (or indirect) photolysis, are important mechanisms for the degradation of CEC in both natural and engineered aquatic systems (Schwarzenbach et al. 2003; Vione et al. 2010).

Through direct photolysis, pollutants react by absorbing light and ultimately undergoing degradation. The efficacy of this process depends on the overlap between the light spectrum, the absorption spectra of the contaminants, and their degradation quantum yields (Braun et al. 1991; Lester et al. 2013; Schwarzenbach et al. 2003). Indirect photolysis involves the degradation of the CEC via reactions with reactive intermediates (RIs). The RIs of environmental importance include singlet oxygen (1O2), hydroxyl radicals (•OH), and triplet-excited states (3OM*) (Rosario-Ortiz and Canonica 2016). These RIs are mainly formed by photosensitization of dissolved organic matter (DOM) (Vione et al. 2010). Light absorption by DOM leads to the formation of triplet-excited states (3OM*) by intersystem crossing following the formation of singlet-excited states (1OM*) (Mostafa et al. 2014). Energy transfer between 3OM* and triplet-state molecular oxygen (3O2) forms 1O2 (Passananti et al. 2014). The formation of •OH is less clear. It is known that •OH can form via photolysis of nitrate and nitrite; however, photolysis of DOM also forms •OH and other species which can also behave like •OH (McKay and Rosario-Ortiz 2015; Rosario-Ortiz and Canonica 2016; Vione et al. 2014). There have been numerous reports on the quantification of quantum yields for the production of these RIs from DOM samples and also wastewater-derived organic matter (Haag et al. 1984; Lester et al. 2013; McKay and Rosario-Ortiz 2015; Cawley et al. 2015; Dalrymple et al. 2010, Zhang et al. 2014, Zhou et al. 2017).

Despite the production of RI from naturally occurring DOM, the kinetics of contaminant degradation under typical (low) DOC concentrations is slow (t1/2, minutes∼hours) (Avetta et al. 2013; Gomis et al. 2014). Alternatively, urban biowaste-derived sensitizing materials contain soluble bio-organic substances (UW-BOSs), which exhibit similar physicochemical properties to DOM and have been reported as promising chemical auxiliaries for a number of technological applications including environmental remediation processes (Avetta et al. 2015).

UW-BOSs are described as mixtures of macromolecules with average molecular weight ranging from 67 to 463 kg mol−1 (Avetta et al. 2012). Chemical composition data also show that UW-BOS are macromolecules combining a complex lignin-derived structure with several functionalities and carbon types of different polarities, formed by long aliphatic carbon chains substituted by aromatic rings and functional groups such as COOH, CON, C=O, PhOH, O-alkyl, O-aryl, OCO, OMe, and NRR′, (R and R’ = alkyl or H) (Prevot et al. 2011). Structural similarities between UW-BOS and natural organic matter (NOM) are expected. In fact, according to Avetta et al. (2012), UW-BOSs are the likely remains of the main constituents of the sources of bio-organic refuse that were not completely mineralized during aging under aerobic fermentation conditions. Detailed information on the composition and functional groups of different UW-BOSs can be found elsewhere (Arques and Prevot 2015). Regarding safety, previous acute toxicity assays indicated the absence of harmful effects to the test-organism Vibrio fischeri after adding UW-BOS to water (Avetta et al. 2012; Avetta et al. 2013). Nevertheless, toxicity assays with organisms from different trophic levels are still needed.

The capability of UW-BOS to promote the photo-degradation of various pollutants has been studied, and several encouraging results have been obtained (Avetta et al. 2013; Avetta et al. 2015; Avetta et al. 2012; Prevot et al. 2011, 2010). Most of these studies focus on the implementation of photo-Fenton processes at circumneutral conditions, possible through the stabilization of iron species in the presence of UW-BOS, while few studies demonstrated the involvement of reactive oxygenated species in the elimination of different contaminants in the presence of UW-BOS. Prevot et al. (2010) showed a relevantly lower dye abatement rate in the presence of pure N2 compared to air, which suggests a reaction mechanism involving •OH and 1O2. Prevot et al. (2011) monitored these species during the degradation of 4-chlorophenol in the presence of UW-BOS by EPR spectroscopy, concluding that UW-BOS are potential photosensitizers for applications in environmental remediation. However, quantitative measurements of RI generated by UW-BOS are presently unavailable.

We addressed this knowledge gap by measuring the quantum yields of 1O2 and •OH formation, as well as quantum yield coefficients of TMP degradation (indicative of 3OM* formation) for two UW-BOS samples, both under ultraviolet irradiation (UV) at 254 nm and solar-simulated light. This quantitative information is critical for assessing the potential of these materials in photochemical wastewater treatment. Quantum yield values were compared with those obtained for a standard organic matter Suwannee River Natural Organic Matter (SRNOM). Additionally, we investigated the effect of UW-BOS on the degradation of amicarbazone (AMZ), a herbicide extensively used in large-scale cultivations and a potential emerging concern water pollutant.

Experimental

Chemicals

All solutions were prepared using water purified by reverse osmosis (Milli-Q). Technical grade amicarbazone (AMZ, 241.3 g mol−1, minimum purity 95.4%) was supplied by Arysta LifeScience Corp. and used as a standard in liquid chromatography (HPLC) analysis and in photochemical experiments. SRNOM was obtained from the International Humic Substances Society (catalog number 1R101N).

Two UW-BOS samples, identified by the acronyms CVT230 and CVDF110, were obtained from the process lines of ACEA Pinerolese waste treatment plant in Pinerolo (Italy) and have been isolated from different source materials: CVT230 from home gardening and park trimming residue (GR) piles aerated for 230 days and CVDF110 from 35/55/10 w/w/w urban waste organic humic fraction/GR/urban sewage sludge mix aerated for 110 days.

SRNOM and UW-BOS were dissolved in 10 mmol L−1 phosphate buffer (pH 7.2) at a concentration of 10 mg L−1. Bovine catalase (Sigma Aldrich), furfuryl alcohol-FFA (TCI America), para-chlorobenzoic acid-pCBA (ICN Biomedicals Inc.), 2,4,6-trimethylphenol-TMP (Sigma Aldrich), benzene (Alpha Aesar), sodium nitrate and sodium phosphate (Fisher Scientific), and phosphoric acid (85%, J.T. Baker) were all of reagent grade purity. For HPLC analysis, acetonitrile (HPLC grade) and methanol (HPLC grade) were purchased from Honeywell.

Irradiation apparatus

Solar simulation experiments were conducted using a Model 94041 Oriel solar simulator equipped with a 1000-W lamp and air mass 1.5 global filter (Newport Corp.). The total irradiance was 68 W m−2 in the wavelength range 290–400 nm. Solar simulator irradiance was measured with a USB2000 spectroradiometer (Ocean Optics Inc.). Samples were irradiated in 2-mL pyrex vials with no headspace and exposed to radiation while in a water bath maintained at 20 °C.

Irradiation by a low-pressure (LP) UV lamp was performed using a semi-collimated beam apparatus with four 15-W LP lamps (ozone-free, General Electric #G15 T8), emitting at 254 nm. Petri dishes (70 × 50 mm) were used as exposure vessels in which the samples were continuously stirred. A radiometer (International Light Inc., Model 1700/SED 240/W) was used to measure UV irradiance at the surface of the liquid samples; the incident irradiance was 2.00 mW cm−2.

Measurements and calculation of apparent quantum yields

Furfuryl alcohol (FFA) was used as a chemical probe to measure the formation of 1O2 from NOM (SRNOM or UW-BOS)-containing solutions, following previously established procedures (Mostafa and Rosario-Ortiz 2013) for both LP-UV and simulated sunlight systems, since FFA is commonly used to measure 1O2 formation at wavelengths >300 nm, and its direct photolysis at 254 nm can be considered insignificant for UV experiments (Lester et al. 2013). FFA was added to the samples at an initial concentration of 22.5 μmol L−1 and its concentration was monitored after light exposure using liquid chromatography. Methanol (0.1 mol L−1) was added in order to quench photochemically generated •OH radicals, in order not to overestimate 1O2 concentration; the reaction between FFA and •OH radicals follow a second-order rate law with k •OH,FFA = 1.5 × 1010 L mol−1 s−1 (Mostafa and Rosario-Ortiz 2013).

Samples were exposed over 4 h under simulated sunlight and over 1 h under the LP-UV lamp. Plotting ln[FFA]/[FFA]0 versus irradiation time generates a straight line as expected from a pseudo-first-order reaction. The slope obtained is divided by the second-order reaction rate constant between 1O2 and FFA (1.2 × 108 L mol−1 s−1) according to Haag et al. (1984) to obtain the steady-state singlet oxygen concentration, [1O2]ss. This value was used to determine the 1O2 formation quantum yield [Φ(+1O2)], according to Eq. 1:

$$ {\left[{}{}^1{\mathrm{O}}_2\right]}_{\mathrm{ss}}=\frac{{\mathrm{R}}_{\mathrm{a}}\ \Phi \left(+{}{}^1{O}_2\right)}{\Sigma {\mathrm{k}}_{1\mathrm{O}2,\mathrm{X}}\left[\mathrm{X}\right]+{\mathrm{k}}_{\mathrm{d},\mathrm{H}2\mathrm{O}}} $$
(1)

where R a is the rate of light absorption by DOM over the wavelength range, which is a function of the solution absorbance and irradiance; k 1O2,X is the second-order reaction rate constant between 1O2 and any species X present in the water matrix (including but not limited to NOM); and k d,H2Ocorresponds to the rate constant of 1O2 decay to the ground state due to physical quenching by water molecules (k d,H2O = 2.5 × 105 s−1) (Mostafa and Rosario-Ortiz 2013). The overall decay of singlet oxygen is dominated by k d,H2O in such a way that the physical or chemical quenching due to OM or other species (k 1O2,X[X]) can generally be ignored at environmentally relevant concentrations (Mostafa and Rosario-Ortiz 2013). R a is calculated by Eq. 2:

$$ {R}_a=\int \frac{E_P^0\ \left(1-{10}^{-{Abs}_{\lambda} z}\right)}{z}\ d\lambda $$
(2)

where E P 0 is the irradiance (measured with an Ocean Optics USB2000 spectroradiometer), Abs λ is the solution absorbance at wavelength λ, and z is the optical path length.

For the LP-UV-irradiated systems, the formation quantum yield of hydroxyl radicals [Φ(+•OH)] from DOM (SRNOM or UW-BOS)-containing solutions was measured using para-chlorobenzoic acid (pCBA) as a chemical probe. The choice of this method was based on the higher photostability of pCBA at 254 nm in comparison with other probe molecules, like phenol (Lester et al. 2013). In brief, pCBA was added to the samples at an initial concentration of 2.23 × 10−2 mmol L−1 and the solution was exposed to UV irradiation for up to 1 h. The slope of a plot of ln[pCBA]/[pCBA]0 as a function of irradiation time yielded k pCBA, assuming pseudo-first-order pCBA concentration decay. The value of Φ(+•OH) at 254 nm was then estimated as R pCBA/R a , where R pCBA corresponds to the initial pCBA degradation rate (R pCBA = k pCBA[pCBA]0), and R a is the rate of UV radiation absorbed by DOM at 254 nm.

For the simulated sunlight-irradiated systems, the formation rate of •OH radicals (R f •OH) from DOM (SRNOM or UW-BOS)-containing solutions was measured using the Dorfman reaction (Dorfman et al. 1962), i.e., the hydroxylation of benzene to phenol. Briefly, a 10-mg L−1 DOM (SRNOM or UW-BOS) solution was spiked with 3 mmol L−1 benzene and irradiated for 4 h, with samples collected every hour. R f •OH was calculated based on the rate of phenol production divided by the known reaction yield (0.63), as detailed elsewhere (McKay and Rosario-Ortiz 2015). The corresponding formation quantum yields were obtained by dividing R f •OH by the rate of light absorption by DOM over the wavelength range (R a ). To quantify •OH formation due solely to UW-BOS or NOM (and not photo-Fenton processes), catalase (20 U mL−1) was added to each sample 30 min before irradiation. Catalase decomposes photochemically generated H2O2 and (the minor amount) produced as a byproduct of benzene hydroxylation (Larson and Zepp 1988).

For both LP-UV and simulated sunlight systems, TMP was used as a probe to assess the formation of 3OM*; its degradation is an indicative of 3OM* formation (Mostafa et al. 2014; Cawley et al. 2015; Canonica and Freiburghaus 2001; Cawley et al. 2009; Richard et al. 2004; Golanoski et al. 2012). TMP was spiked at an initial concentration of 3.67 × 10−3 mmol L−1 into DOM (SRNOM or UW-BOS)-containing solutions, which were irradiated under LP-UV (1 h) or simulated sunlight (4 h). Direct photolysis of TMP at 254 nm and above was found to be slow (Lester et al. 2013). The quantum yield coefficient f TMP (L Einstein−1), was calculated according to previous studies (Sharpless et al. 2014; McCabe and Arnold 2016; McKay et al. 2017), as the TMP degradation rate constant (k TMP) divided by the rate of light absorption by DOM over the wavelength range (R a ) [f TMP = k TMP/R a , L Einstein−1]. The slope of a plot of ln[TMP]/[TMP]0 as a function of irradiation time yields k TMP, assuming pseudo-first-order TMP concentration decay. No •OH or 1O2 quenching agents were added in the experiments carried with TMP, based on the conclusions of Lester et al. (2013). These authors carried out TMP degradation experiments in the presence of Suwanee River humic acid (SRHA) under different conditions (in the absence of scavengers; with 50 mmol L−1 tert-butanol, with 1 mmol L−1 FFA, and after sparging with either O2 or N2) and concluded that the reactions of TMP with •OH and 1O2 were insignificant.

AMZ degradation experiments

Additional experiments were performed to evaluate the effect of UW-BOS on AMZ degradation under LP-UV and simulated solar light. These experiments were performed in solutions containing 10 mg L−1 of UW-BOS or SRNOM in phosphate buffer. AMZ was spiked to achieve a starting concentration of 2.07 × 10−2 mmol L−1.

Analytical methods

UV-Vis absorption spectra were measured with a Varian Cary 100 Bio spectrophotometer (Agilent) using a 1-cm path length quartz cuvette, in the range 190–820 nm and 1 nm intervals.

Total organic carbon (TOC) was measured using a TOC-VSCH (Shimadzu Corp., Japan) analyzer. The method detection level (MDL) was 0.2 mgC L−1.

An HPLC-UV (Agilent 1200) system equipped with an Agilent Eclipse XDB-C18 (4.6 mm inner diameter × 250 mm, 5 μm) column was used for monitoring AMZ, phenol, pCBA, FFA, and TMP. AMZ concentrations over time during the degradation experiments were determined at 224 nm by gradient elution with a phosphate buffer (pH 2.8)/acetonitrile solution, at a flow rate of 1.0 mL min−1. The following gradient was used: 0–3 min (15% acetonitrile), 3–20 min (15–40% acetonitrile), 20–22 min (40% acetonitrile), 22–23 min (40–15% acetonitrile), and 23–25 min (15% acetonitrile). The mobile phase used to determine pCBA was 45% phosphate buffer (pH 2.8)/55% methanol at a flow rate of 1 mL min−1, with isocratic elution and detection at 234 nm. A flow rate of 1 mL min−1 and isocratic elution with 70% phosphate buffer (pH 2.8)/30% methanol was used to monitor FFA at 219 nm. Finally, TMP was monitored at 220 nm, using a phosphate buffer (pH 2.8)/acetonitrile solution as the mobile phase at 1 mL min−1, according to the following gradient elution: 0–6 min (60–80% methanol), 6–10 min (80% methanol), and 10–11 min (80–60% methanol).

Results and discussion

BOS characterization and optical parameters

The absorption spectra of UW-BOS are shown in Fig. 1. For comparison, the spectrum for SRNOM is also presented. The UV-Vis spectra of UW-BOS are featureless, with an exponential decrease with increasing wavelength and extend into the visible. These observations are similar to what is seen for SRNOM (Korshin et al. 1997). On a carbon molar basis, UW-BOSs exhibit slightly higher absorptivities than SRNOM. This is further discussed in the following sections.

Fig. 1
figure 1

Absorption spectra of CVT230, CVDF110, and SRNOM measured at a DOC concentration of ∼10 mg L−1 in 10 mmol L−1 phosphate buffer, pH 7.2

Full size image

The specific UV absorbance at 254 nm (SUVA254) for the samples is shown in Table 1. SUVA254 is defined as the absorbance of the sample divided by the DOC. SUVA254 is an indicator of the amount of aromatic carbon present in the dissolved NOM, which has been shown to be an important indicator of DOC reactivity (Weishaar et al. 2003). The results show that UW-BOS samples exhibit higher SUVA254 values and therefore have higher aromaticity than SRNOM. Overall, the SUVA254 values for the UW-BOS were generally higher in comparison to values found in the literature for NOM, on average 1.5 times higher (Bodhipaksha et al. 2015; Cawley et al. 2015).

Table 1 DOC concentrations and optical parameters measured for UW-BOS and SRNOM samples
Full size table

The E2:E3 ratio denotes the ratio of the absorbances at 250 and 365 nm. This parameter can be used as a surrogate for average molecular weight of DOM, decreasing with increasing molecular weight (Helms et al. 2008; Peuravouri and Pihlaja 1997). The calculated E2:E3 ratios for UW-BOS (Table 1) are in agreement with that obtained by Gomis et al. (2015). Based on the SUVA254 and E2:E3 values in Table 1 and in the literature, UW-BOS samples appear to be more aromatic and of a higher molar weight than NOM, for which the average E2:E3 ratios are about 4.8 (considering twelve different NOM types) (McKay et al. 2016; Mostafa et al. 2014).

In our case, singlet oxygen formation quantum yields [Φ(+1O2)] seem to increase with increasing E2:E3 ratios (Online Resource, Fig. S1), in agreement with previous studies (Mostafa and Rosario-Ortiz 2013). Moreover, this trend was also observed for Φ(+•OH), the hydroxyl radical formation quantum yield, a result also observed by McCabe and Arnold (2016). (Online Resource, Fig. S2). Analogously, the negative trend of f TMP with E2:E3 suggests that 3OM* formation increases with increasing E2:E3 ratios (Online Resource, Fig. S3). These results suggest that low-molecular-weight DOM is more photoactive than high-molecular-weight DOM.

The spectral slope (S) was determined from the slope of a plot of ln(α λ) vs. (λ-λ r) according to Eq. 3, where α λ refers to the linear Napierian absorption coefficients from 300 to 600 nm and λ r is a reference wavelength (λ r = 300 nm) (Twardowski et al. 2004). Values of α λ were calculated according to Eq. 4, where A λ is the absorbance at wavelength λ and ℓ is the optical path length:

$$ \upalpha \left(\uplambda \right)=\upalpha \left(\uplambda \mathrm{r}\right)\ {\mathrm{e}}^{-\mathrm{S}\left(\uplambda -\uplambda \mathrm{r}\right)} $$
(3)
$$ \upalpha \left(\uplambda \right)=\frac{2.303\ \mathrm{A}\left(\uplambda \right)}{\ell } $$
(4)

Both the E2:E3 ratio and the spectral slope are inversely proportional to molecular size (McKay et al. 2016). As obtained for the E2:E3 ratio, the quantum yields of formation of singlet oxygen and hydroxyl radicals seem to increase with increasing spectral slope; the same can be concluded for the formation of 3OM* (Online Resource, Figs. S4, S5, and S6).

RI quantum yields

Determination of RI formation quantum yields from UW-BOS is of importance to identify the potential use of these materials in water remediation, considering the significant role of reactive species towards the degradation of refractory compounds. Figure 2 summarizes the values of Φ(+1O2), Φ(+•OH), and f TMP for CVT230 and CVDF110 under simulated solar light and UV radiation. It is worth noting that both UW-BOS samples exhibit similar quantum yields under simulated sunlight, but not under UV irradiation. Similarly to SRNOM, UW-BOS exhibited quantum yields for 1O2 about an order of magnitude greater than those for •OH. The quantum yield coefficients for TMP degradation were about a factor of two greater for UW-BOS than SRNOM, both under UV and simulated solar irradiation. Moreover, singlet oxygen is generated with quantum yields three orders of magnitude higher than hydroxyl radical formation, for all types of organic matter. In fact, Vione et al. (2010) found that 1O2 steady-state concentrations were, on average, two orders of magnitude higher than •OH upon irradiation of water from eight different lakes. The authors attribute this fact to the fast formation of 1O2 relative to •OH, because the transformation rate constants of 1O2 (105 s−1—collision with solvent) and •OH (104 to 105 s−1—reaction with dissolved compounds) are comparable. It is worth observing that the direct comparison of Φ(+1O2) or Φ(+•OH) with f TMP is not possible. In fact, 3OM* is scavenged mostly by dissolved oxygen, and only a small fraction of this species would react with TMP, since a very low concentration of this substrate was used.

Fig. 2
figure 2

Quantum yields for (a) 1O2 formation, (b) •OH formation, and (c) quantum yield coefficient of TMP (indicative of 3OM* formation) from SRNOM and UW-BOS under simulated sunlight and 254-nm UV radiation

Full size image

UW-BOSs present lower Φ(+1O2) and Φ(+•OH) and higher f TMP in comparison with values reported for NOM (Cawley et al. 2015; Dalrymple et al. 2010). In fact, under simulated sunlight, the two UW-BOS samples showed Φ(+1O2) and Φ(+•OH) on average 1.1 and 2.9 times lower than the corresponding values obtained for SRNOM. On the other hand, UW-BOS present f TMP values 2.2 times higher than those obtained for SRNOM. UW-BOSs have a lower efficiency of RI generation compared to SRNOM as determined by measured quantum yields. Although UW-BOS and SRNOM are different in terms of OM source, UW-BOSs are likely higher in aromaticity and molecular weight than SRNOM (see E2:E3 and S values in Table 1). Past research has demonstrated that natural water samples and DOM isolates of various origins exhibit decreasing photoreactivity with increasing molecular weight, which is a potential explanation for the trends observed in this study.

Under low pressure UV irradiation, the values obtained for RI formation quantum yields follow the same trend observed under solar light: the generation of 1O2 and •OH radicals by SRNOM was more efficient in comparison with UW-BOS samples, while UW-BOS degrade TMP more efficiently than SRNOM. Moreover, quantum yields determined with the LP-UV system were greater than those determined with the solar-driven systems. This is consistent with previous work that compared RI quantum yields between simulated sunlight and an LP-UV system (Lester et al. 2013). In our study, for the UW-BOS system under 254-nm irradiation, the quantum yields of 1O2 and •OH formation and the quantum yield coefficient of TMP degradation were on average 2.4, 420, and 2.7 times higher than those observed under simulated solar light, respectively. Although speculative, reasons for increased Φ(+1O2) and f TMP under UV irradiation include excitation into higher energy excited states, a larger fraction of which is able to oxidize TMP and form 1O2. Considering •OH, quantum yields under UV irradiation were almost three orders of magnitude greater than those measured with solar irradiation. This could be due to increased direct photolysis of photochemically produced H2O2, as well as an increased rate of photo-ionization of hydroxybenzoic acids, which has been hypothesized to produce •OH (Sun et al. 2015).

Influence on pollutants degradation

Several studies have shown that UW-BOS can enhance pollutant removal. For example, more than 65% removal of 4-chlorophenol was observed in the presence of 100 mg L−1 of UW-BOS under simulated solar light (Prevot et al. 2010). Also, 65% removal was achieved for ethyl orange in the presence of UW-BOS under sunlight (Prevot et al. 2011). It is worth mentioning that these pollutants virtually do not undergo direct photolysis (Prevot et al. 2010, 2011). Gomis et al. (2013) reported that the pseudo-first-order reaction rate constants of crystal violet degradation increased from 5.8 × 10−4 to 3.5 × 10−3 min−1 in the presence of UW-BOS under sunlight. Likewise, an increase in the removals of naphthalene sulfonates from 15 to 30% was obtained when UW-BOS were added to the medium (Avetta et al. 2012). In contrast, Gomis et al. (2014) found that the degradation of a mixture of six emerging pollutants that were prone to direct photolysis was hampered in the presence of UW-BOS. The authors attributed this fact to light screening by these materials.

Peixoto and Teixeira (2014) showed that despite the low molar absorptivity of amicarbazone at 254 nm (ελ = 52 L mol−1 cm−1), irradiation from LP-Hg lamps can degrade AMZ. For example, the half-life of a neutral pH AMZ solution at [AMZ]0 = 100 mg L−1 was ∼685 min when irradiated with a 36-W lamp (Peixoto and Teixeira 2014). In our study, amicarbazone-containing solutions were irradiated under simulated sunlight in the presence of SRNOM or UW-BOS in order to compare the contributions of these sources of RI species to herbicide degradation. The results are shown in Table 2. Significant direct photolysis of AMZ under sunlight is unlikely, as reported elsewhere (Silva et al. 2015). Therefore, the solar-driven AMZ degradation is expected to be due to the indirect photolysis promoted by DOM.

Table 2 Results of photolysis experiments: k, pseudo first-order degradation rate constant; t 1/2, half-life; and R 2, determination coefficient
Full size table

AMZ removal followed pseudo-first-order behavior in solutions containing SRNOM or UW-BOS using both simulated sunlight and LP-UV as irradiation sources. Under solar light, a slight increase in AMZ degradation rate was observed in the presence of UW-BOS. On the other hand, for the UV-irradiated system, the AMZ half-lives are on average 1.4 times lower in the presence of UW-BOS. Despite these differences, we would argue that organic matter isolated from urban-bio waste and naturally occurring DOM (i.e. SRNOM) are comparable in terms of their capacities of sensitizing organic pollutants. Considering their application in treatment of polluted waters, conditions required for high removal depend on the type of radiation source used. If mixtures of AMZ and UW-BOS were exposed to solar radiation, increased concentrations of UW-BOS would increase removal rates, because AMZ does not absorb light (and thus undergo direct photolysis) in the solar spectrum. Conversely, if LP-UV radiation is used, it is desirable to have low concentrations of UW-BOS due to light screening of AMZ.

The rate of photo-degradation of a pollutant (P) is the sum of direct and indirect photolysis, the last one mainly by interactions with 1O2, •OH, and 3OM* (Eq. 5):

$$ -{r}_P=\left({k}_{directphot}+{k}_{\bullet \mathrm{OH},\mathrm{P}}\left[{}{}^{\bullet}\mathrm{O}\mathrm{H}\right]+{k}_{O_2,\mathrm{P}}\left[{}{}^1\mathrm{O}_2\right]+{k}_{{}{}^3 O{M}^{\ast },\mathrm{P}}\left[{}{}^3\mathrm{O}{\mathrm{M}}^{\ast}\right]+{k}_{\mathrm{others}}\right)\left[\mathrm{P}\right] $$
(5)

As a consequence, higher values of −r P are expected with the addition of UW-BOS for large values of the second-order rate constant\( {k}_{{}{}^3\mathrm{O}{\mathrm{M}}^{\ast },\mathrm{P}} \) since UW-BOS present high values of f TMP (indicative of 3OM* formation).

Zeng and Arnold (2013) studied the contributions of direct photolysis, 1O2, •OH radicals, 3OM*, and others species to pesticide photo-degradation. For pesticides that were inefficiently degraded by direct photolysis, the authors showed that the indirect photolysis induced by 3OM* was the dominating process. In effect, the percent contributions of 3OM* were 66% (atrazine), 55% (cyanazine), 44% (acetochlor), 47% (alachlor), 50% (metolachlor), 62% (mesotrione), 79% (diuron), 67% (isoproturon), and 55% (metoxuron) in surface water samples collected from US Geological Survey (USGS) Cottonwood Lake study area, Jamestown, North Dakota.

Conclusions

In this study, optical properties and reactivity of UW-BOS under solar light and UV irradiation (254 nm) was detailed, which will contribute to understand the potential use of these bio-organic materials in engineered photochemical treatment processes. UW-BOS extracted from gardening-park trimming residues, urban waste, and sewage sludge are capable of generating 3OM* more efficiently than natural organic matter (NOM), as a result, the presence of UW-BOS could favor the degradation of compounds of environmental concern, like pesticides, which exhibit large second-order rate constants with 3OM*.

Through the assessment of UW-BOS photosensitizing properties and RI formation quantum yields, UW-BOSs were shown to be not necessarily better than naturally occurring DOM, i.e., SRNOM. However, because they do sensitize the production of RI, they could potentially be used in engineered treatment scenarios, from both the economic and ecological point of view. The RI formation quantum yields and quantum yield coefficient values determined in this study from UW-BOS, reported here for the first time in literature, will contribute to estimate the potential of these species in environmental aquatic photochemistry.