This is the submitted version of the following article: Serrà A., Zhang Y., Sepúlveda B., Gómez E., Nogués J., Michler J., Philippe L.. Highly reduced ecotoxicity of ZnO-based micro/nanostructures on aquatic biota: Influence of architecture, chemical composition, fixation, and photocatalytic efficiency. Water Research, (2020). 169. 115210: - . 10.1016/j.watres.2019.115210, which has been published in final form at https://dx.doi.org/10.1016/j.watres.2019.115210 © https://dx.doi.org/10.1016/j.watres.2019.115210. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ 1 Highly reduced ecotoxicity of ZnO-based micro/nanostructures on aquatic 2 biota: Influence of architecture, chemical composition, fixation, and photo- 3 catalytic efficiency 4 Albert Serrà ‖,*, Yue Zhang &, Borja Sepúlveda &, Elvira Gómez $, @, Josep Nogués &, #, Jo- 5 hann Michler ‖, Laetitia Philippe ‖ 6 7 ‖ Empa, Swiss Federal Laboratories for Materials Science and Technology, Laboratory for Mechanics of 8 Materials and Nanostructures, Feuerwerkerstrasse 39, CH-3602 Thun, Switzerland. 9 & Catalan Institute of Nanoscience and Nanotechnology (ICN2), CSIC and BIST, Campus UAB, Bel- 10 laterra, E-08193 Barcelona, Spain. 11 $ Grup d’Electrodeposició de Capes Primes i Nanoestructures (GE-CPN), Departament de Ciència de 12 Materials i Química Física, Universitat de Barcelona, Martí i Franquès, 1, E-08028, Barcelona, Catalo- 13 nia, Spain. 14 @ Institute of Nanoscience and Nanotechnology (IN2UB), Universitat de Barcelona, Barcelona, Catalo- 15 nia, Spain. 16 # ICREA, Pg. Lluís Companys 23, E-08010, Barcelona, Spain. 17 Corresponding author: albert.serraramos@empa.ch (A.S.) 18 19 ABSTRACT 20 Developing efficient sunlight photocatalysts with enhanced photocorrosion resistance and minimal eco- 21 toxicological effects on aquatic biota is critical to combat water contamination. Here, the role of chem- 22 ical composition, architecture, and fixation on the ecotoxicological effects on microalgae of different 23 ZnO and ZnO@ZnS based water decontamination photocatalysts was analyzed in depth. In particular, 24 the ecotoxicological effects of films, nanoparticles and biomimetic micro/nano-ferns were carefully 1 25 assessed by correlating the algae’s viability to the Zn(II) release, the photocatalyst–microalgae interac- 26 tion, and the production of reactive oxygen species (ROS). The results showed a drastic improvement 27 in algal viability for supported ZnO@ZnS core@shell micro/nanoferns, as their ecotoxicity after 96 h 28 light exposure was significantly lower (3.7–10.0% viability loss) compared to the ZnO films (18.4– 29 35.5% loss), ZnO micro/nanoferns (28.5-53.5% loss), ZnO nanoparticles (48.3–91.7% loss) or 30 ZnO@ZnS nanoparticles (8.6-19.2% loss) for catalysts concentrations ranging from 25 mg L-1 to 400 31 mg L-1. In particular, the ZnO@ZnS micro/nanoferns with a concentration of 400 mg L-1 exhibited ex- 32 cellent photocatalytic efficiency to mineralize a multi-pollutant solution (81.4±0.3% mineralization ef- 33 ficiency after 210 min under UV-filtered visible light irradiation) and minimal photocorrosion (< 5% of 34 photocatalyst dissolution after 96 h of UV-filtered visible light irradiation). Remarkably, the ZnO@ZnS 35 micro/nanoferns showed lower loss of algal viability (9.8±1.1%) after 96 h of light exposure, with min- 36 imal reduction in microalgal biomass (9.1±1.0%), as well as in the quantity of chlorophyll-a (9.5±1.0%), 37 carotenoids (8.6±0.9%) and phycocyanin (5.6±0.6%). Altogether, the optimized ZnO@ZnS core@shell 38 micro/nanoferns represent excellent ecofriendly photocatalysts for water remediation in complex media, 39 as they combine enhanced sunlight remediation efficiency, minimal adverse effects on biological micro- 40 organisms, high reusability and easy recyclability. 41 GRAPHICAL ABSTRACT 42 43 KEYWORDS: ecotoxicity, ZnO-based photocatalysts, sunlight photocatalysis, microalgae, persis- 44 tent organic pollutants 2 45 1. INTRODUCTION 46 The decontamination of wastewater is one of the main challenges of this century since high levels of 47 pollution resulting from anthropological and industrial activities have important negative effects on eco- 48 systems and on human life and health (Fu et al., 2011; McMullan et al., 2011; Gleik, 1993; Jury, 2007). 49 Currently, the main processes for the decontamination of wastewater are largely ineffective for the elim- 50 ination of many pollutants, especially persistent organic pollutants (POPs) (Chong, et al., 2010; Malato, 51 et al., 2009; Moreira, et al., 2017, Serrà, et al., 2019). POPs include pesticides, drugs, hormones, personal 52 hygiene products, etc. Given the severe negative effects of POPs on biota, their efficient elimination is 53 critical. In phytoplankton communities in particular, many POPs can directly or indirectly facilitate the 54 growth of microorganisms—for example, by stimulating cyanobacterial blooms—and consequently 55 modify population growth (Harris, et al., 2016; Everaert, et al., 2015; Bettinetti, et al., 2012; Maule, et 56 al., 1984). 57 In this context, photocatalysis offers an excellent decontamination strategy by exploiting the solar en- 58 ergy for the total remediation of POPs to provide a clean, green, and sustainable approach (Lee, et al., 59 2016; Batista, et al., 2017). This is motivating the rapid proliferation of new efficient photocatalysts, 60 especially semiconducting materials such as TiO2 or ZnO, that enable rapid water decontamination. 61 Many of these photocatalysts are also effective as antifungal and antibacterial multitasking platforms 62 (Liu, et al., 2018; Hatamie, et al., 2015). Current photocatalyst design is focused on developing new 63 materials, architectures, and configurations to improve their mineralization efficiency, especially under 64 sunlight irradiation. Researchers have also invested significant effort into identifying and understanding 65 the remediation mechanisms, including the production of reactive species and different intermediates 66 (Serrà, et al., 2019; Liu, et al., 2018; Hatamie, et al., 2015, Yadav, et al., 2016; Arshad, et al. 2017). 67 However, researchers have rarely analyzed the potentially disastrous side effects of photocatalysts, es- 68 pecially for aquatic microorganisms. Therefore, it is crucial to develop materials that affect minimally 69 the biota throughout their whole life cycle by considering all the potential impacts on the ecosystems in 70 which they will be used. 3 71 In the particular case of ZnO photocatalysts, their extensive use in a wide variety of industrial applica- 72 tions and consumer products increases the risk of release into aquatic environments, where they can 73 cause ecotoxicological effects (Mortimer, et al, 2010; Djeramane, et al. 2018; Subashchandrabose, et al. 74 2013). The toxicity can be especially relevant for some microorganisms, such as microalgae, which play 75 a fundamental role in the ecosystem as primary food sources for aquatic biota (Singh, et al., 2017; Mani, 76 et al., 2014; Khan, et al., 2011). Therefore, although ZnO photocatalysts could seem a clean and eco- 77 nomical method for water decontamination, they can also have adverse effects on ecosystems and human 78 life. The main route of ZnO contamination is the release and accumulation of the micro- and nano-ZnO 79 structures in aquatic environments, and the subsequent liberation of Zn(II) caused by the extremely high 80 ZnO photocorrosion (Mortimer, et al, 2010; Djeramane, et al. 2018; Subashchandrabose, et al. 2013). 81 However, in the particular case of microalgae, the ZnO damage could be more complex and involve the 82 following factors: (i) physical or mechanical damage because of the direct contact of micro- and nano- 83 materials with the fragile microalgae; (ii) the high level of dissolved Zn(II) in the aquatic media due to 84 the high capability of algae to adsorb metal ions, and (iii) the generation of reactive oxygen species 85 (ROS) during exposure of ZnO to sunlight. Moreover, diverse factors, such as shape, porosity, size, 86 surface coating, exposure mode and time, have determinant effects on the ecotoxicology of these pho- 87 tocatalysts in aquatic microorganisms (Bondarenko, et al., 2013; Hou, et al., 2018; Aruoja, et al., 2009; 88 Ma, et al., 2013; Miao, et al., 2010; Wong, et al., 2010). Consequently, there is an urgent need to find 89 and thoroughly analyze different photocatalytic ZnO architectures which could combine excellent pol- 90 lutant degradation efficiency and minimal ecotoxicity on algae. 91 In this work, we present a comprehensive ecotoxicological and photocatalytic analysis of different ZnO 92 and ZnO@ZnS core@shell micro/nanostructures as a way to highlight the importance of integrating 93 ecotoxicity and photocatalysis efficiency in the water decontamination process, by considering the 94 whole life cycle of the photocatalyst. To analyze the effect of the photocatalyst architecture on micro- 95 algae ecotoxicity, three different systems were evaluated: (i) ZnO films, which are characterized by a 96 continuous supported/fixed structure with low active surface area, minimal direct interaction/contact 97 with aquatic microorganisms, low photocatalytic activity, and low ROS production efficiency under 4 98 sunlight irradiation; (ii) supported/fixed ZnO fern-like microstructured arrays, showing high active sur- 99 face areas due to their fractal architecture, which potentially increases the direct interaction with aquatic 100 microorganisms, as well as the pollutant adsorption and light trapping capability, thus increasing the 101 photocatalytic and ROS production efficiencies under sunlight irradiation; and (iii) ZnO nanoparticles, 102 being one of the most relevant ZnO residues in many industrial applications and consumer products, 103 which exhibit high surface area, strong direct interaction with aquatic microorganisms, and moderate 104 photocatalytic and ROS production efficiencies under sunlight irradiation. Additionally, the effect of 105 the chemical composition was also analyzed by comparing ZnO and ZnO@ZnS core@shell photocata- 106 lysts. This analysis is motivated by the recent demonstration of the improved photocatalytic performance 107 provided by the ZnS shell due to the increased ROS production under sunlight irradiation while signif- 108 icantly reducing the photocorrosion activity (Serrà, et al., 2019; Ranjith, et al., 2018). Importantly, here 109 we demonstrate that optimized ZnO@ZnS core@shell fern-like biomimetic microleaf arrays supported 110 on solid substrates can efficiently combine low ecotoxicological effects with highly enhanced photo- 111 catalytic efficiency compared to other ZnO and ZnO@ZnS core@shell micro- and nanostructures. To 112 demonstrate these features, we have investigated the photocorrosion resistance, and the dose- and time- 113 dependent toxicity of these nanostructures with different architectures in Spirulina (Arthrospira) platen- 114 sis (microalgae), to specifically determine the relationship between the ecotoxicological effects and the 115 photocatalyst size, shape, composition, configuration (dispersed vs. supported), and efficiency. 116 2. EXPERIMENTAL SECTION 117 2.1. Synthesis of ZnO-based photocatalysts 118 ZnO films were hydrothermally grown on a glass substrate using solution of 0.01 M hexamethylenetet- 119 ramine (#398160; Sigma-Aldrich, > 99.0%) and 0.1 M Zn(NO3)2 (#398160; Sigma-Aldrich, > 99.0%), 120 with the pH was adjusted to 10, for 30 min at 65°C under magnetic stirring conditions (600 rpm) 121 (Vessalli, et al., 2017; Mizuta, et al., 2006). 122 ZnO fern-like microleaf arrays were potentiostatically electrodeposited at -1.0 V (vs. Ag/AgCl/KCl 123 (3 M)) – 28 C cm-2 of circulated charge density – using a conductive fluorine-doped tin oxide film on a 5 124 glass substrate with a classical three-electrode electrochemical cell, an Autolab with PGSTAT30 poten- 125 tiostat/galvanostat (Metrohm Autolab; Netherlands), and the NOVA software (Version 2.1.4; Metrohm 126 Autolab; Netherlands). Working, counter, and reference electrodes were the fluorine-doped tin oxide 127 film on a glass substrate, a Pt mesh, and Ag/AgCl/KCl (3 M), respectively. The electrochemical medium 128 was a 0.5 mM ZnCl2 (#14422; Fluka, > 98.0%) + 0.1 M KCl (#P9333; Sigma-Aldrich, > 99.0%) oxygen- 129 saturated (bubbled 45 min before and during the electrodeposition) aqueous solution (pH = 7 in standard 130 conditions) maintained at 80°C. The electrodeposition process was performed under strong stirring con- 131 ditions (400 rpm of magnetic stirring and 12 L min-1 of oxygen bubbling). 132 To form ZnO@ZnS core@shell microstructures, the ZnO micro/nanoferns were immersed in an aque- 133 ous solution of 30 mM thioacetamide (TAA) – CH3CSNH2 – (#163678; Sigma-Aldrich, 98%) at 50°C 134 in a water bath for 4, 8, and 12 h. These structures are denoted as ZnO@ZnS(4h), ZnO@ZnS(8h), and 135 ZnO@ZnS(12h) micro/nanoferns, respectively. The ZnO and ZnO@ZnS core@shell micro/nanofern 136 arrays were then exhaustively washed with water and ethanol, dried at room temperature, and annealed 137 for 2 h at 400°C in an argon atmosphere. The annealing treatment was performed using a rapid thermal 138 annealing equipment (Advance Riko Mila 5050; Japan). The heating ramp rate was set at 10°C min-1. 139 The ZnO@ZnS nanoparticles were fabricated using commercial 20 nm ZnO nanoparticles (99.5%; Io- 140 LiTec Nanomaterials; Germany) as seeds, following the same sulfidation process with thioacetamide 141 for 4 h. 142 2.2. Characterization of ZnO-based photocatalysts 143 The surface morphologies and the elemental composition were examined by Hitachi S-4800 and H- 144 4100FE field-emission scanning electron microscopy (FE-SEM; Hitachi; Japan) equipped with energy- 145 dispersive X-ray spectroscopy detector. The ZnO and ZnO/ZnS nanoparticles were analyzed using a 146 high-resolution HR-TEM Jeol JEM 2100 (Jeol; Japan) equipped with a LaB6 source operated at 200 kV 147 (images were recorded with Digital Micrograph v.1.82.80 software). Nanoparticles were dispersed in 148 ethanol, and then a droplet of the suspension was poured in Holey Carbon covered copper TEM grids 149 (300 Mesh Cu, Agar Scientific; United Kingdom) prior to HR-TEM observation. The particle size 6 150 distribution was evaluated by analyzing more than 120 nanoparticles. The specific surface area of each 151 biomimetic photocatalyst was determined by the Brunauer–Emmett–Teller (BET) method from N2 ad- 152 sorption-desorption isotherms at 77 K using a Micrometrics Tristar-II (Micrometrics; Canada). The 153 structural characterization was conducted using X-ray diffraction (XRD, Bruker D8 Discover diffrac- 154 tometer; Bruker; United States) in the Bragg–Brentano configuration with CuKα radiation. The optical 155 and electrical properties of the photocatalysts were analyzed by recording the UV-vis diffused reflec- 156 tance spectra (DRS) and photoluminescence. DRS were measured using a UV-vis PerkinElmer Lambda 157 900 UV spectrophotometer (PerkinElmer; United States). The photoluminescence was acquired with a 158 custom-made set-up based on a narrow band filtered (FB360-10; Thorlabs; United States) light emitting 159 diode with central emission at 365 nm (M365FP1; Thorlabs; United States) as excitation source. The 160 back scattered luminescence was long pass filtered (FEL0400; Thorlabs; United States) and detected by 161 an Andor 193i spectrometer with an Andor Idus camera (Andor technology, United Kingdom). 162 2.3. Photocatalytic efficiency of the ZnO-based photocatalysts 163 The photocatalytic activity was examined by monitoring the decomposition of a complex solution of 164 three different POPs (5 ppm of methylene blue (MB) (#M9140; Sigma-Aldrich, > 82 %) + 5 ppm of 165 4-nitrophenol (4-NP) (#73560; Honeywell Fluka, > 99 %) + 5 ppm of Rhodamine B (Rh-B) (#83689; 166 Sigma-Aldrich, > 98 %) pollutants) under a 75 W Xe lamp setup with UV-filtered simulated sunlight 167 (light intensity of 678 ± 11 lx) or natural UV-filtered sunlight (average light intensity 1471 ± 275 lx). 168 Longpass filters (cut-on wavelength region: 400 nm to 2200 nm) were introduced to limit wavelength 169 radiation and to avoid direct photolysis of the pollutants. Pollutant solutions were prepared in algal cul- 170 ture medium. The photocatalysts (400 mg L1) were immersed first in each pollutant solution in dark 171 conditions for 60 min to reach adsorption-desorption equilibrium before starting the light actuation. The 172 photoremediation process was followed at 30°C and under argon bubbling using the following the com- 173 parison of the total organic content (TOC) prior to the start of irradiation and after irradiating the sample 174 for 210 min, by using the high-temperature combustion method on a catalyst (Pt-Al2O3) in a tubular 175 flow microreactor operated at 680°C, with a stream of hydrocarbon free air to oxidize the organic car- 176 bon, using TOC analyzer (model TOC-VCSH; Shimadzu Corporation; Japan) with a high-sensibility 7 177 column. In addition, the ZnO-based photocatalysts were recycled/reused for eight consecutive cycles to 178 mineralize the three multi-pollutant solution under UV-filtered natural sunlight to test their reusability 179 and recyclability properties. Each experiment was repeated four times to ensure accuracy and reproduc- 180 ibility. 181 2.4. Reactive oxygen species identification 182 Chemical selective radical quenchers were used to determine the formation of hydroxyl radicals (●OH), 183 oxygen superoxide ions (O2-), and singlet oxygen (1O2) by the ZnO-based photocatalysts under UV- 184 filtered simulated sunlight (> 400 nm, light intensity of 678 ± 11 lx) at 30°C. The hydroxyl radical (●OH) 185 concentration was measured by following the time-dependent reduction of the fluorescence peak at 515 186 nm (λex = 303 nm) in quartz cuvettes containing 8 µM of fluorescein sodium salt (#30181; 187 Supelco/Sigma-Aldrich) using an AMINCO-Bowman Series 2 spectrofluorometer (Thermo Electron; 188 United States). The reaction between 100 µM of XTT [2,3-Bis(2-methoxy-4-nitro-5-sulfophenyl)-2H- 189 tetrazolium-5-carboxanilide] sodium salt (#X4251; Supelco/Sigma-Aldrich, > 90%) and the formed ox- 190 ygen superoxide ions (O2-) allowed to identify the formation of oxygen superoxide ions by measuring 191 the characteristic absorption peak of XTT-formazan (resulting from the reduction of XTT by superoxide 192 ions) at 475 nm by using a UV-1800 Shimadzu UV-vis spectrophotometer (Shimadzu Corporation; Ja- 193 pan). The singlet oxygen (1O2) formation was determined by a highly selective singlet oxygen sensor 194 green (SOSG) reagent (Invitrogen; United States). The formation of singlet oxygen was monitored by 195 measuring the time-dependent photoluminescence intensity at 535 nm (λex = 488 nm) using an 196 AMINCO-Bowman Series 2 spectrofluorometer (Thermo Electron; United States), which corresponds 197 to the formation of the endoperoxide SOSG-1O2 produced by the reaction of the anthracene part of SOSG 198 reagent (5mM in methanol) with the singlet oxygen. 199 2.5. Photostability and photocorrosion resistance of the ZnO-based photocatalysts 200 The photostability and photocorrosion activity of the different ZnO-based photocatalysts were evaluated 201 by (i) determining the evolution of the concentration of Zn(II) ions in the algal culture medium, (ii) 202 measuring the BET surface area, and (iii) examining the surface by SEM after the irradiation of each 8 203 photocatalyst with UV-filtered simulated sunlight during 96 h (light intensity of 658 ± 25 lx). The algal 204 culture medium consists on: 4.5 g L-1 of NaHCO3 (#13433; Fluka, > 99 %), 0.5 g L-1 of K2HPO4 (#60356; 205 Fluka, > 99 %), 1.5 g L-1 of NaNO3 (#S5506; Sigma-Aldrich, > 99 %), 1.0 g L-1 of K2SO4 (#P0772; 206 Sigma-Aldrich, > 99 %), 1.0 g L-1 of NaCl (#S7563; Sigma-Aldrich, > 99.5 %), 1.2 g L-1 of MgSO4 207 (#746452; Sigma-Aldrich, > 99.5 %), and 0.04 g L-1 CaCl2 (#793639; Sigma-Aldrich, > 96 %), 0.001 g 208 L-1 of FeSO4·7H2O (#F8263; Sigma-Aldrich, > 99 %) (Wang, et al, 2007; Lone, et al., 2013). The 209 concentration of Zn(II) ions in the algal culture medium was determined by measuring the absorbance 210 at 620 nm, associated to the Zn(II)-Zincon complex, in a quartz cuvette with an optical length of 1 cm 211 using a UV-1800 Shimadzu UV-vis spectrophotometer (Shimadzu Corporation; Japan). Prior to meas- 212 uring the absorbance, 100 µL of algal culture medium were incubated at 20°C for 10 min with 900 µL 213 of 40 µM Zincon monosodium salt (#96440; Sigma-Aldrich) in borate buffer (50 mM, pH = 9). Metal 214 stock solution of Zn(II) was prepared by dissolving the appropriate amount of ZnCl2 (#14422; Fluka, > 215 98.0%) in algal culture medium. To evaluate the possible interference of iron ions from the algal medium 216 in the determination of the Zn(II) ions, metal stock solutions of Zn(II) were also prepared in MilliQ 217 water. Note that Fe(III) ions can interfere when they are present in concentrations higher than 5 mg L-1, 218 while Fe(II) does not interfere. Our results indicate that the algae medium did not interfere in the deter- 219 mination of Zn(II) using the Zincon monosodium salt due to the high selectivity for Zn(II) ions and the 220 low concentration of ferric ions (<1 mg L-1 in our case). 221 2.6. Ecotoxicity of the ZnO-based photocatalysts 222 The microalgae Spirulina (Arthrospira) platensis –"paracas" type –was cultivated in an algal culture 223 medium (pH 9.8) under a temperature of 30°C, solar illumination, and air bubbling. The bioassays were 224 conducted using ZnO-based photocatalysts at the following concentrations: 0, 25, 50, 100, 200, and 400 225 mg L-1. Irradiation experiments were performed to analyze the effect of (i) the architecture, (ii) the 226 chemical composition, and (iii) the fixation of photocatalysts on microalgae: 227 (i) Architecture effect: Different concentrations of ZnO films, ZnO nanoparticles, and ZnO mi- 228 cro/nanoferns were immersed in a microalgae culture and irradiated with continuous simulated 9 229 sunlight (light intensity of 740 ± 15 lx) for 8 h per day for four days. The microalgae viability, 230 biomass reduction, and photosynthetic pigment reduction were measured after 6, 12, 24, 72, and 231 96 h. 232 (ii) Composition effect: Different concentrations of ZnO and ZnO@ZnS nanoparticles and mi- 233 cro/nanoferns were immersed in a microalgae culture and irradiated with continuous simulated 234 sunlight (light intensity of 740 ± 15 lx) for 8 h per day for four days. The microalgae viability, 235 biomass reduction, and photosynthetic pigment reduction were measured after 6, 12, 24, 72, and 236 96 h. 237 (iii) Fixation effect: Different concentrations of fixed and non-fixed ZnO@ZnS micro/nanoferns (sul- 238 fidation time of 4 h) were immersed in the microalgae culture and then were irradiated 8 h per 239 day with continuous simulated sunlight (light intensity of 740 ± 15 lx) for 8 h per day for four 240 days. The microalgae viability, biomass reduction, and photosynthetic pigment reduction were 241 measured after 6, 12, 24, 72, and 96 h. 242 The microalgae survival was estimated as the percentage of viability loss in microalgae with respect to 243 the control (0 mg L-1 of photocatalyst). For this purpose, 1 mL of each suspension was loaded in a 244 Sedgewick-Rafter cell (SPI supplies, Graticules S50; Structure Probe Inc, United States) to count the 245 number of microalgae non-distorted in shape or size. The quantification of the biomass reduction was 246 determined by optical density measurements at 560 nm using a UV-1800 Shimadzu UV-vis spectropho- 247 tometer (Shimadzu Corporation; Japan) during the incubation period. Spirulina (arthrospira) platensis 248 include chlorophyll-a, carotenoid, and phycocyanin photosynthetic pigments. The chlorophyll-a, and 249 carotenoid pigments were measured using the Lichtenthaler and Wellburn method (Deniz, et al., 2011; 250 Khan, et al., 1987; Dere, et al., 1998). During the incubation period, 3 mL of suspensions were taken 251 and centrifuged at 5000 rpm for 10 min. Then the sample was washed with phosphate buffer saline (0.1 252 M, pH = 7) solution three times. After discarding the supernatant, ethanol (96% v/v) was added to the 253 microalgae residues and thoroughly mixed, and the pigments were extracted in ethanol at 65°C for 90 254 min. Then, the absorbance of the supernatant in the ethanol solution at 470, 649, and 665 nm was meas- 255 ured using a UV-1800 Shimadzu UV-vis spectrophotometer (Shimadzu Corporation; Japan). The 10 256 concentration of chlorophyll-a, and carotenoids was calculated according to Lichtenthaler and Wellburn 257 equations (Deniz, et al., 2011; Khan, et al., 1987; Dere, et al., 1998). For the phycocyanin concentration 258 determination, 0.5 mL of microalgae suspension were taken and centrifuged at 10,000 rpm for 10 min 259 and washed with phosphate buffer saline solution (0.1 M, pH = 7) three times. The pellets were then 260 resuspended in phosphate buffer saline solution (0.1 M, pH = 7) and ultrasonicated for 30 min to break 261 the microalgae filaments. Then, the resultant suspension was centrifuged at 4°C 10,000 rpm for 5 min, 262 and the absorbance of the supernatant solution at 615 and 652 nm was measured using a UV-1800 Shi- 263 madzu UV-vis spectrophotometer (Shimadzu Corporation; Japan). The concentration of phycocyanin 264 was calculated according to the Bennett and Bogorad equation (Moraes, et al., 2011; Bennett, et al., 265 1973). 266 3. RESULTS AND DISCUSSION 267 3.1. Synthesis and characterization of the different ZnO-based structures 268 With the aim to identify the photocatalysts that can combine highly efficient POPs mineralization and 269 minimal toxicity on aquatic biota, we have analyzed in this work the roles of shape, chemical composi- 270 tion, and fixation of different ZnO-based micro/nanoarchitectures (Figure 1) on microalgae The effect 271 of the shape has been established by comparing: (i) ZnO films chemically deposited on glass; (ii) ZnO 272 dendritical micro/nanoferns electrodeposited on fluorine-doped tin oxide films on a glass substrate; and 273 (iii) commercial ZnO nanoparticles. The effect of the chemical composition has been assessed by com- 274 paring both ZnO and ZnO@ZnS core@shell micro/nanoferns and nanoparticles, which were prepared 275 by chemical sulfidation and thermal annealing (Figure 2a). Finally, the effect of fixation has been ana- 276 lyzed by comparing fixed and non-fixed ZnO@ZnS micro/nanoferns (i.e., detached from the fluorine- 277 doped tin oxide film on a glass substrate). 11 278 279 Figure 1: (a) Schematic representation of the different ZnO architectures. FE-SEM micrographs of (b) 280 the ZnO films (scale bar: 2 µm) and (c) the ZnO micro/nanoferns (scale bar: 5 µm). (d) HR-TEM mi- 281 crographs of the ZnO nanoparticles (scale bar: 100 nm). 282 As depicted in the FE-SEM micrographs, the chemically deposited ZnO films on glass formed a rough 283 layer with an average thickness of 40 ± 3 µm (Figure 1b). In contrast, the electrodeposited ZnO in 284 strong stirring conditions (high flux of oxygen and magnetic stirring) resulted in well-defined fractal 285 and dendritical ZnO micro/nanoferns (Figure 1c) with an average thickness of 53 ± 2 µm. On the other 286 hand, commercial ZnO nanoparticles were mainly spherical but showed a heterogeneous size distribu- 287 tion (diameters of 70 ± 26 nm) (Figure 1d). 288 The sulfidation process was easily observed macroscopically by the color change of the photocatalysts 289 from dark gray to golden (Figure S1a), which got darker for longer sulfidation times. Macroscopic color 290 changes were also observed in ZnO commercial nanoparticles when the ZnO@ZnS core@shell was 291 formed, changing from white to yellow (Figure S1b). At the microscopic level, important changes were 292 also seen in the architecture, morphology, and roughness of the photocatalyst. A substantial increase of 293 diameter in the central trunk and ramifications of micro/nanoferns – approximately 1.2 and 3.8 times 12 294 higher after 4 and 12 h of sulfidation, respectively – and a noticeable increase in the roughness (Figure 295 2b-d, Figure S2) was detected compared to the pristine ZnO micro/nanoferns, which can be explained 296 by the Kirkendall effect (Serrà, et al., 2019; Ranjith, et al., 2018). The fractal and dendritical architecture 297 was significantly reduced and damaged when the sulfidation was extended over 4 h (Figure 2c, d). In 298 addition, the formation of ZnS shells with varying thicknesses depending on sulfidation time has also 299 been confirmed by means of elemental mapping of the central trunk of ZnO@ZnS micro/nanoferns 300 (Figure S2c-e), resulting in an increase in shell thickness when the sulfidation time increases. Note that, 301 for a given sulfidation time, the ZnS shell thicknesses differ depending on the area of the micro/nanofern 302 that is analyzed (i.e., central trunk, primary ramification, or secondary ramification) due to variations in 303 surface reactivity and stability. Despite the increase in surface roughness, the overall micro/nanoferns 304 architecture became more compact during the sulfidation process, which resulted in a significant reduc- 305 tion in the BET surface area values, falling from 68.2 m2g-1 (ZnO micro/nanoferns) to 30.1 m2g-1 306 (ZnO@ZnS(12h) micro/nanoferns) (Table 1, Figure S3). However, interestingly, the slight increase in 307 roughness without affecting the architecture at short sulfidation times slightly increased the accessible 308 surface area up to 70.4 m2g-1 in the ZnO@ZnS(4h) micro/nanoferns, which may improve the photocata- 309 lytic activity by increasing the accessible reactive sites. In the case of the nanoparticles, a weak modifi- 310 cation was observed in the particle size distribution –76 ± 21 nm for ZnO@ZnS nanoparticles– by the 311 formation of a ZnS layer of approximately 8 nm in thickness (Figure 2e). In addition, Table 1 shows 312 that the ZnS to ZnO ratio increased considerably for longer sulfidation times, demonstrating that sulfi- 313 dation is a volumetric process as a consequence of the thermally activated sulfur diffusion. 13 314 315 Figure 2: (a) Schematic representation of the sulfidation process. FE-SEM micrographs of (b) the 316 ZnO@ZnS(4h), (c) ZnO@ZnS(8h), and (d) ZnO@ZnS(12h) micro/nanoferns. Scale bar: 5 µm. (e) HR- 317 TEM micrographs of the ZnO@ZnS(4h) nanoparticles. Scale bar: 20 nm. 318 The X-ray diffraction pattern (XRD) of the ZnO architectures (Figure S4) matched perfectly with the 319 standard hexagonal wurtzite ZnO structure (JCPDS card No. 36-1451). However, the XRD measure- 320 ments showed that ZnO films had a (101) preferred orientation, while in the ZnO nanoparticles and 321 micro/nanoferns had a (002) preferred orientation. After the sulfidation process, the formation of the 322 ZnS layer was confirmed by the detection of two extra peaks, which perfectly corresponded to the cubic 14 323 ZnS blende structure (JCPDS card No. 65-1691), at 2θ = 28.55o (111) and 33.87o (200). A significant 324 reduction in the intensity of ZnO was also observed, with a higher reduction as the sulfidation time 325 increased, due to the substitution of ZnO by ZnS. The obtained data confirmed the formation of a ZnS 326 shell over the ZnO surface due to the volumetric substitution of ZnO by ZnS. Therefore, controlling the 327 sulfidation time is crucial for synthesizing well-defined ZnO@ZnS core@shell architectures. 328 The UV-vis DRS spectra (Figure S5) of the ZnO photocatalysts (film, nanoparticles, and micro/nano- 329 ferns) showed a strong absorption in the UV region and a significantly lower intensity in the visible 330 region due to the wide band-gap of ZnO. After the sulfidation process, the absorption band of ZnO was 331 extended to the visible domain due to the formation of a two-phase heterojunction. Interestingly, the 4h 332 sulfidation process yielded the highest absorbance in the visible range. The large difference in the lattice 333 parameters between ZnO and ZnS is responsible of generating a large mechanical stress at the interface. 334 Such stress can induce drastic changes in the band structure with a substantial reduction of the bandgap, 335 as it has been predicted by numerical calculations (Torabi, et al., 2015). The incorporation of the S atoms 336 can also generate additional lattice imperfections giving rise to impurity levels within the band-gap. The 337 optical band-gap of the ZnO-based photocatalysts (Table 1) was calculated using Tauc relation (Figure 338 S6). Note that the obtained values agree well with the data reported in the literature (Serrà, et al., 2019; 339 Ranjith, et al., 2018). It is worth noting that the largest bandgap reduction is observed for the 340 ZnO@ZnS(4h) micro/nanoferns. The increase of the sulfidation time clearly favored the dominant op- 341 tical behavior of the ZnS, thereby widening again the bandgap. The photoluminescence measurements 342 (Figure S7) showed an important intensity increase in the visible and near-infrared regions for the 343 ZnO@ZnS(4h) micro/nanoferns compared to the pristine structures. The luminescence intensity decays 344 in the 8h sulfidated sample, probably due to the decrease of the lattice imperfections and the lower 345 surface area of the micro/nano-ferns. In contrast, the deterioration of the fern-like structures for longer 346 (12h) sulfidation time generated by the large mechanical stress in the structures causing cracks and 347 fractures, resulted in an increase of the luminescence intensity compared to the 8h sulfidation time. On 348 the other hand, the sulfidation of the nanoparticles induced a luminescence reduction after 4h sulfidation 349 time. This is possibly due to the isotropic sulfidation dynamics in the dispersed nanoparticles under 15 350 agitation, which enables the formation of a ZnS shell with lower mechanical stress and the reduction in 351 the number of defects. 16 352 Table 1. Ratios of ZnS to ZnO, BET surface area, band-gap energy, and natural sunlight photocatalytic efficiency of the ZnO-based photocatalysts (photocatalyst 353 dosage = 400 mg L-1, Temperature = 30.0 ± 0.1 oC). ZnS to ZnO ratio ZnS thickness / BET surface Band-gap Mineralization efficiency Mineralization efficiency Photocatalyst / at. % nm area / m2 g-1 energy / eV (fresh photocatalyst) / % (after 8th recycling cycles) ZnO film - - 8.1 3.24 ± 0.12 4.3 ± 0.5 2.2 ± 0.3 ZnO nanoparticles - - 51.2 3.28 ± 0.09 17.9 ± 0.1 11.2 ± 0.4 ZnO@ZnS(4h) nanoparti- 34.6 ± 1.2 8 54.7 2.93 ± 0.05 61.7 ± 0.9 49.8 ± 0.4 cles ZnO micro/nanoferns - - 68.2 3.30 ± 0.07 20.1 ± 0.4 12.9 ± 0.5 ZnO@ZnS(4h) 25.8 ± 1.2 50-90 70.4 2.77 ± 0.07 81.4 ± 0.3 79.9 ± 0.7 micro/nanoferns ZnO@ZnS(8h) 52.4 ± 1.2 90-170 54.2 2.96 ± 0.09 71.0 ± 0.6 68.2 ± 0.6 micro/nanoferns ZnO@ZnS(12h) 75.7 ± 1.2 350-400 30.1 3.17 ± 0.06 39.9 ± 0.9 37.2 ± 0.4 micro/nanoferns 354 17 355 3.2. Photocatalytic efficiency of ZnO-based structures 356 The photocatalytic performance of each photocatalyst was evaluated by means of UV-filtered natural 357 and simulated sunlight irradiation (see supporting information). As expected, the ZnO@ZnS core@shell 358 photocatalysts with well-defined biomimetic micro/nanoferns exhibited excellent photocatalytic effi- 359 ciency (Tables 1 and S1) in mineralizing a multi-pollutant solution. Moreover, the ZnO@ZnS 360 core@shell(4h) micro/nanoferns showed excellent reusability and recyclability properties (Table 1), as 361 the mineralization efficiency was virtually constant after their reuse for eight consecutive times. Note 362 that the ZnO@ZnS core@shell(4h) nanoparticles exhibited worse recycling properties, possibly due to 363 nanoparticle loss during the recycling process. 364 3.3. Reactive oxygen species identification 365 It is well known that ROS generated during the photocatalyst irradiation are not only efficient to miner- 366 alize organic compounds, but also to damage microorganisms. However, microorganisms must have 367 direct contact/interaction with photocatalysts to suffer effective photo-damage due to the very short life- 368 time of the ROS. Although, hydroxyl radicals are the main actors in the photo-mineralization of organic 369 pollutants and the photo-damage of microorganisms, oxygen superoxide ions and singlet oxygen can 370 also be relevant in for both processes (Anastasescu, et al., 2018). To assess the role of the different ROS, 371 we first investigated the kinetics and concentration of hydroxyl radicals using fluorescein salt as selec- 372 tive radical quencher (Figure 3a, 3b, and S8). Figure 3a shows the time-dependent evolution of fluo- 373 rescein concentration under simulated sunlight by following the photoluminescence peak at 515 nm (λex 374 = 303 nm). The kinetics of hydroxyl generation (Figure 3b) was evaluated by assuming zero order 375 kinetics and equimolar reaction stoichiometry between hydroxyl radicals and fluorescein molecules. 376 The kinetics of hydroxyl formation was also investigated in the absence of photocatalysts as the photol- 377 ysis of water can also generate hydroxyl radicals. The hydroxyl formation by the ZnO@ZnS(4h) mi- 378 cro/nanoferns was approximately 45, 30, 1.8, 44, 1.8 and 2.6 times higher than that obtained for ZnO 379 films, ZnO nanoparticles, ZnO@ZnS(4h) nanoparticles, ZnO, ZnO@ZnS(8h), and ZnO@ZnS(12h) mi- 380 cro/nanoferns, respectively. The photolytic formation of hydroxyl radicals from water was negligible, 18 381 as it was 380 times lower than that obtained for the ZnO@ZnS(4h) micro/nanoferns. Next, the formation 382 of oxygen superoxide ions was determined by monitoring spectroscopically the reaction of XTT and 383 oxygen superoxide ions (Figure 3c). As can be seen in Figure 3c, ZnO@ZnS core@shell photocatalysts 384 presented a relatively low activity to generate oxygen superoxide ions compared to the ZnO photocata- 385 lysts. Note that superoxide ions have a low oxidation power, which provides a relatively low activity to 386 mineralize the POPs or to damage the microorganism cell wall. In acidic media, the superoxide radical 387 can react with protons to generate hydroxyl radicals. However, this process is negligible in alkaline 388 solutions such as microalgae media. In addition, the formation of singlet oxygen molecules was moni- 389 tored by determining the consumption of SOSG reagent (Figure 3d) and the formation of endoperoxide 390 compound (Figure 3e). According to these experiments, ZnO@ZnS core@shell photocatalysts also pre- 391 sented a relatively low activity to produce singlet oxygen compared to ZnO. Consequently, the high 392 mineralization efficiency of the ZnO@ZnS core@shell photocatalysts can be attributed to the high pro- 393 duction of hydroxyl radicals, which can also damage microorganisms in direct contact to the photocata- 394 lysts. 395 3.4. Photostability and anti-photocorrosion of the different ZnO-based structures 396 A well-known problem of the ZnO-based photocatalysts is their high photocorrosion, which considera- 397 bly affects their efficiency and potential use (Han, et al., 2014; Weng, et al., 2014). The photocorrosion 398 triggers the release of Zn(II) ions (i.e., new pollutants) into water, which exert important negative effects 399 on ecosystems, especially on biota. Therefore, for practical applications photocatalysts require high pho- 400 tocorrosion resistance, chemical stability, and high reusability. 401 402 403 19 404 405 Figure 3: (a) Time-dependent photoluminescence spectra of 8 µM fluorescein (FL) and (b) kinetic con- 406 stant of hydroxyl radical formation using ZnO-based photocatalysts. Time-dependent UV-vis spectra of 20 407 (c) the formation of XTT-formazan, which indicates the formation of oxygen superoxide ions, and (d) 408 SOSG consumption using ZnO-based photocatalysts. (e) Time-dependent photoluminescence spectra of 409 endoperoxide formation, which indicate the formation of singlet oxygen, using ZnO-based photocata- 410 lysts. Photocatalyst dosage = 400 mg L-1; temperature = 30.0 ± 0.1 oC; irradiation: UV-filtered simulated 411 sunlight (> 400 nm, light intensity of 678 ± 11 lx). 412 To evaluate the photostability of the ZnO-based photocatalysts, the time-dependent concentration of 413 Zn(II) ions in the algal culture medium containing 400 mg L-1 of each photocatalyst was determined 414 after an irradiation time of 96 h with UV-filtered simulated sunlight. As Figure 4 shows, the photocor- 415 rosion mainly depends on the photocatalyst composition, being especially high in the case of unmodified 416 ZnO films, nanoparticles and micro/nanoferns. Note that the Zn(II) concentration in aqueous media in- 417 dicates the dissolution of 91%, 78%, and 76% of the ZnO film, the nanoparticles, and the micro/nano- 418 ferns, respectively. In addition, the photocatalyst shape and architecture also affects the photocorrosion 419 effect, since nanoparticles and micro/nanoferns, which have larger surface area and a greater ability to 420 trap light, present a lower photocorrosion than ZnO films. These results confirm that the photocorrosion 421 activity is strongly related to crystal morphology (i.e., different facets) and the physical properties of the 422 ZnO surface, having more photo-stability ZnO micro/nanomaterials with a (002) preferred orientation 423 (Debroye, et al., 2017; Ishioka, et al., 2017). In contrast, ZnO@ZnS(4h) core@shell architectures exhib- 424 ited excellent anti-photocorrosive properties, with a ZnO dissolution of less than 5%, due to the effective 425 transfer of photogenerated holes from the ZnO core to the ZnS shell, both protecting the surface oxygen 426 of ZnO from the solution and preventing the attack of the surface oxygen atom by the holes transported 427 to the catalyst/solution interface (Torabi, et al., 2015, Yu, et al., 2015). However, after longer sulfidation 428 times, the ZnS behavior predominates and, therefore, the photocorrosion activity is significantly in- 429 creased, as expected from the well-known photo-instability of pure ZnS. Therefore, the photocorrosion 430 resistance strongly depends on the shell thickness that serves as a protective coating. This behavior has 431 also been observed in the case of the ZnS@ZnO core@shell photocatalysts (Serrà, et al., 2019; Ranjith, 432 et al., 2018). The photocatalyst dissolution is clearly confirmed by observing the changes in the surface 433 morphology of each photocatalyst after 96 h of continuous irradiation in a fresh algal culture medium. 21 434 As can be seen in Figure S10, the surface morphology of the ZnO micro/nanoferns exhibited a deterio- 435 ration of the catalyst surface accompanied by the detachment of photocatalyst fragments after 96 h of 436 continuous irradiation. On the other hand, only roughness increase was observed in the core@shell het- 437 erostructures, which was more relevant for the structures with longer sulfidation times. Therefore, the 438 formation of a thin ZnS layer considerably increased the photocorrosion resistance, although the Zn(II) 439 concentration in the solution raised when the sulfidation time was increased. 440 441 Figure 4: Percentage of dissolved Zn (relative to the initial amount of Zn in each ZnO-based photocata- 442 lyst) after the continuous irradiation of 400 mg L-1 of the ZnO-based photocatalysts in a fresh algae 443 culture medium under UV-filtered simulated sunlight (temperature = 30.0 ± 0.1 oC). Error bars indicate 444 standard deviations of the four replicated experiments. 445 3.5. Ecotoxicological effects of the different ZnO-based photocatalysts on microalgae 446 Although the interaction of micro- and nano-structures with algae can favor microalgae growth, in gen- 447 eral, photocatalysts can present negative effects on microalgae viability. Several studies have analyzed 448 the effect of ZnO-based nanomaterials on plants, algae, and other living organisms, attributing ZnO 449 toxicity mainly to the release of Zn(II) ions as a consequence of the poor photostability (Bondarenko, et 450 al., 2013; Hou, et al., 2018; Aruoja, et al., 2009; Ma, et al., 2013; Miao, et al., 2010; Wong, et al., 2010). 451 It is generally accepted that the ZnO toxicity is mainly produced by the solubilization of Zn(II) ions, 22 452 which form metal bindings to SH-groups of proteins, especially in the plasma membranes, and inhibit 453 the cell or microorganisms growth. However, most studies ignore other important effects such as the 454 aggregation or sedimentation of the nanostructures, or the organism-nanostructure interaction, thus at- 455 tributing that ZnO-based nanomaterials have greater toxicity than bulk materials solely due to the solu- 456 bilized Zn(II) ions (Bondarenko, et al., 2013; Hou, et al., 2018; Aruoja, et al., 2009; Ma, et al., 2013; 457 Miao, et al., 2010; Wong, et al., 2010). However, we demonstrate here that the toxicity of ZnO-based 458 photocatalysts is motivated by at least three different factors: (i) the release of inorganic toxic substances 459 (e.g., Zn(II)) as a consequence of their poor chemical and photochemical stability, (ii) the mechanical 460 effects as a consequence of the microalgae-photocatalyst interaction, in which the shape and catalyst 461 fixation have an important role, and (iii) the photogeneration of ROS (Serrà, et al., 2019; Ranjith, et al., 462 2018). 463 The reduction in microalgae viability, biomass, and photosynthetic pigments of Spirulina platensis (Fig- 464 ure 5) demonstrated that the photocatalyst architecture and shape have a determining effect on micro- 465 algae viability. The percentage of maximum microalgae death at 96 h was 48.3±4.6 %, 72.5±3.9 %, 466 84.8±3.7 %, 88.8±1.2 %, and 91.7±1.7 %, at 25, 50, 100, 200, and 400 mg L -1 of ZnO nanoparticles, 467 respectively. These values were approximately 1.7 and 2.6 times higher than those obtained with the 468 same amount of ZnO micro/nanoferns and films, respectively. In addition, the reduction in biomass 469 exhibited exactly the same trend as the microalgae viability. The different amounts of released Zn(II) 470 due to the different surface areas and surface stability, significantly influenced the photocatalyst’s tox- 471 icity on microalgae. The toxicity of unmodified ZnO photocatalysts was clearly indicated by the red and 472 brown microalgae clumps after 96 h of incubation (Figure S11). However, toxicity did not depend ex- 473 clusively on the release of Zn(II) ions; because the release rate of Zn(II) for films was approximately 474 1.2 times greater than that for nanoparticles and micro/nanoferns. The microalgae-photocatalyst inter- 475 action and the production of ROS also determined its toxicity. Microalgae-photocatalyst interaction can 476 be achieved by the internalization of photocatalysts within the microalgae, which depends upon the type 477 of microalgae and the size of photocatalyst, or by the simple physical contact between the entities. In 478 the study reported here, no internalization of nanoparticles or micro/nanoferns fragments occurred after 23 479 96 h of incubation, since the electronic microscopy analysis of the residues resulting from the dissolution 480 of dried microalgae did not present any photocatalyst fragments. Therefore, the greater toxicity of na- 481 noparticles and micro/nanoferns can be explained by the sum of the release of Zn(II) ions, the greater 482 interaction of the micro/nanoferns, especially the nanoparticles with the microalgae, and the photogen- 483 eration of ROS. Note that the effect of ROS also requires the direct physical interaction of microalgae 484 and photocatalysts. 485 The significant effect of the Zn(II) release was clearly demonstrated when ZnO@ZnS core@shell mi- 486 cro/nanoferns were used since the loss in viability (400 mg L-1 of photocatalyst at 96 h of exposure) was 487 approximately 5-6 times lower for ZnO@ZnS(4h), ZnO@ZnS(8h), and ZnO@ZnS(12h) micro/nano- 488 ferns, when compared with unmodified ZnO micro/nanoferns (Figure 6). The same behavior was also 489 observed when ZnO@ZnS(4h) nanoparticles were used (Figure S12). 490 Regarding the architecture/shape effect, the comparison of ZnO@ZnS(4h) nanoparticles and mi- 491 cro/nanoferns show that the negative effects on microalgae were slightly higher for nanoparticles, de- 492 spite having a virtually identical release of Zn(II) ions. Thus, the greater contact and interaction between 493 the microalgae and the photocatalysts in the case of nanoparticles have greater negative effects on mi- 494 croalgae cultivation due to the photo-damage produced by the photogenerated ROS. 495 Concerning the fixation state, non-fixed ZnO@ZnS(4h) micro/nanoferns (i.e., detached from the fluo- 496 rine-doped tin oxide film on a glass substrate) exhibited at least 3 times higher negative effect on all of 497 the analyzed microalgae viability parameters compared to the fixed-micro/nanoferns, as can be seen in 498 Figure 6. The interaction between the photocatalyst and the microalgae, and surely the mechanical ef- 499 fects (i.e., mechanical destruction) during the air bubbling of the cultivation process, also affected the 500 development of Spirulina microalgae. Finally, it is worth comparing the results of the ZnO@ZnS mi- 501 cro/nanoferns for different sulfidation times. Although the release of Zn(II) was substantially more pro- 502 nounced for long sulfidation times, these structures showed a slightly lower viability reduction compared 503 to the ZnO@ZnS(4h) micro/nanoferns. The slight increase of the ecotoxicity in the later structure can 504 then be attributed to its enhanced photocatalytic efficacy to produce hydroxyl radicals, which are the 24 505 main mediators of the efficient pollutant (photo)mineralization (Serrà, et al., 2019). When the 506 ZnO@ZnS(4h) micro/nanoferns were fixed, the effect was weak due to the very short lifetime of the 507 hydroxyl radicals. However, the detrimental effect could be more pronounced when the micro/nanoferns 508 were not fixed, thereby enabling a closer interaction with the algae. 509 The quantification of the photosynthetic pigments can also provide important information to determine 510 the effects that inhibit microalgae growth. To show effect of the different ZnO and ZnO@ZnS structures, 511 the variation in the content of photosynthetic chlorophyll-a, carotenoids, and phycocyanin pigments was 512 also evaluated, as Zn(II) can replace Mg(II) in chlorophyll molecules, thereby inhibiting the photosyn- 513 thesis process. However, it should be taken into account that the effects of Zn(II) concentration might 514 be complex. Low Zn(II) concentrations can stimulate the production of photosynthetic molecules, but 515 the production inhibition is significant when a certain amount is exceeded, thus decreasing the concen- 516 tration of chlorophyll-a and carotenoids. In addition, the Zn(II) concentration can present also significant 517 effects on the photosynthesis of phycocyanin, possibly as a consequence of the Zn(II) ions blocking the 518 enzyme activity in the photosynthetic synthesis route (Bondarenko, et al., 2013; Hou, et al., 2018; Aru- 519 oja, et al., 2009; Ma, et al., 2013; Miao, et al., 2010; Wong, et al., 2010). In the case of the ZnO and 520 ZnO@ZnS structures, the reduction of the photosynthetic pigments (Figure 5c-e, S13, and S14) fol- 521 lowed the same trend as microalgae viability and the reduction of biomass: the longer the exposure time 522 and the concentration, the greater the effect due to the higher amount of released Zn(II). Therefore, all 523 the analyzed parameters to determine the ecotoxicological effects of ZnO-based photocatalysts followed 524 the same trend. The results show that Zn(II) release is the main toxicity effect, but shape, architecture, 525 fixation, and ROS photogeneration must be considered as well, as discussed above. 526 527 25 528 26 529 Figure 5: Percentage of loss in (a) the microalgae viability, (b) the microalgae biomass, (c) the chloro- 530 phyll-a of the microalgae, (d) the carotenoids of the microalgae, and (e) the phycocyanin pigment of the 531 microalgae for the treatment with 25, 50, 100, 200, and 400 mg L-1 of ZnO films, nanoparticles, and 532 bioinspired micro/nanoferns after an exposure time of 6, 24, 48, 72, and 96 h. The microalgae cultures 533 were irradiated 8 h per day with continuous simulated sunlight (light intensity of 740 ± 15 lx), starting 534 the first irradiation cycle when photocatalysts were introduced in microalgae culture. Percentages are 535 relative to the control (microalgae culture without catalyst) cultivated in the same experimental condi- 536 tions. Error bars indicate standard deviations of the four replicated experiments. 27 537 538 Figure 6: Percentage of loss in (a, c) the microalgae viability and (b, d) the microalgae biomass reduc- 539 tion of microalgae for the treatment with 25, 50, 100, 200, and 400 mg L-1 of ZnO-based bioinspired 540 micro/nanoferns after exposure times of 6, 24, 48, 72, and 96 h. The microalgae cultures were irradiated 541 8 h per day with continuous simulated sunlight (light intensity of 740 ± 15 lx), starting the first irradia- 542 tion cycle when photocatalysts were introduced in microalgae culture. Percentages are relative to the 543 control (microalgae culture without catalyst) cultivated in the same experimental conditions. Error bars 544 indicate standard deviations of the four replicated experiments. 28 545 4. Conclusion 546 Realizing efficient and totally clean photocatalysts for water decontamination requires the development 547 of new materials with minimal impact on ecosystems, which involves integrating improved photocata- 548 lytic performance with reduced ecotoxicological effects throughout the catalyst life cycle. With this in 549 mind, the effects of architecture, chemical composition, and fixation on the ecotoxicity of different ZnO- 550 based sunlight photocatalysis on microalgae were analyzed and discussed, and the following conclusions 551 can be highlighted: 552 (i) Effects of architecture. ZnO micro/nanofern architectures had ecotoxicological effects on micro- 553 algae comparable to those of film architectures but significantly lower than nanoparticles. Photo- 554 catalyst architecture is also relevant to avoid or promote (depending on the application) the inter- 555 action between the microorganisms and the photocatalysts. Such interactions determine the photo- 556 inactivation and photo-damage produced by the generated ROS under sunlight irradiation. 557 (ii) Effects of composition. We show that controlling the ZnS shell thickness is critical, since long 558 sulfidation times affect the surface morphology, photocorrosion resistance, and photocatalytic 559 performance of the micro/nanofern architecture. Most importantly, the reduced photocorrosion 560 enabled a large decrease in the release of Zn(II) ions into the environment, thus improving the 561 reusability and recyclability of the core@shell photocatalysts for water remediation and causing 562 a remarkably lower ecotoxicity on microalgae – at least five times lower – relative to non-modi- 563 fied ZnO. However, ZnO@ZnS core@shell photocatalysts produced a significantly greater num- 564 ber of hydroxyl radicals, which can play important roles in the photo-inactivation and photo- 565 damage of microorganisms when photocatalysts and microorganisms are in direct contact. 566 (iii) Effects of fixation. Fixation of the photocatalysts also facilitated their recyclability and reduced 567 the interaction with microalgae. Most importantly, the reduced photochemical and mechanical 568 interactions in micro/nanofern architectures improved the viability reduction at least three-fold. 569 In conclusion, the optimized ZnO@ZnS(4h) fern architectures are excellent ecofriendly candidates for 570 photocatalytic water remediation in complex saline and biological media given their excellent 29 571 mineralization efficiency, outstanding photostability, photocorrosion resistance, and minimal ecotoxi- 572 cological effects on aquatic biota. 573 Acknowledgments 574 This work was partially supported by the Metrohm foundation. Partial funding from the 2017-SGR-292 575 project from the Generalitat de Catalunya, and the PCIN2016-093 and TEC2017-85059-C3-2-R pro- 576 jects (co-financed by the Fondo Europeo de Desarrollo Regional, FEDER) from the Spanish Ministerio 577 de Economía y Competitividad (MINECO) is also acknowledged. ICN2 is funded by the CERCA Pro- 578 gramme/Generalitat de Catalunya. ICN2 also acknowledges the support from the Severo Ochoa Pro- 579 gram (MINECO, Grant No. SEV-2017-0706). Albert Serrà would like to acknowledge funding from the 580 EMPAPOSTDOCS-II program. The EMPAPOSTDOCS-II programme has received funding from the 581 European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Cu- 582 rie grant agreement number 754364. Yue Zhang acknowledges the China Scholarship Council (CSC) 583 for financial support (201608310112). 584 Appendix A. Supplementary data 585 Supplementary material related to this article can be found, in the online version, a 586 References 587 Anastasescu, C., Negrila, C., Angelescu, D.G., Atkinson, I., Anastasescu, M., Spataru, N., 588 Zaharescu, M., Balint, I., 2018. Particularities of photocatalysis and formation of reactive 589 oxygen species on insulators and semiconductors: Cases of SiO2, TiO2 and their compo- 590 site SiO2-TiO2. Catal. Sci. Technol. 8, 5657–5668. https://doi.org/10.1039/c8cy00991k 591 Arshad, A., Iqbal, J., Siddiq, M., Mansoor, Q., Ismail, M., Mehmood, F., Ajmal, M., Abid, Z., 2017. 592 Graphene nanoplatelets induced tailoring in photocatalytic activity and antibacterial characteris- 593 tics of MgO/graphene nanoplatelets nanocomposites. J. Appl. Phys. 121. 594 https://doi.org/10.1063/1.4972970 595 Aruoja, V., Dubourguier, H.C., Kasemets, K., Kahru, A., 2009. Toxicity of nanoparticles of CuO, ZnO 596 and TiO2to microalgae Pseudokirchneriella subcapitata. Sci. Total Environ. 407, 1461–1468. 597 https://doi.org/10.1016/j.scitotenv.2008.10.053 30 598 Batista, L.M.B., dos Santos, A.J., da Silva, D.R., Alves, A.P. de M., Garcia-Segura, S., Martínez- 599 Huitle, C.A., 2017. Solar photocatalytic application of NbO2OH as alternative photocatalyst for 600 water treatment. Sci. Total Environ. 596–597, 79–86. https://doi.org/10.1016/j.sci- 601 totenv.2017.04.019 602 Bennett, A., Bogorad, L., 1973. Complementary Chromatic Adaptation in a Filamentous Blue-Green 603 Alga. J. Cell Biol. 58, 419 LP-435. https://doi.org/10.1083/jcb.58.2.419 604 Bettinetti, R., Quadroni, S., Manca, M., Piscia, R., Volta, P., Guzzella, L., Roscioli, C., Galassi, S., 605 2012. Seasonal fluctuations of DDTs and PCBs in zooplankton and fish of Lake Maggiore 606 (Northern Italy). Chemosphere 88, 344–351. https://doi.org/10.1016/j.chemosphere.2012.03.009 607 Bondarenko, O., Juganson, K., Ivask, A., Kasemets, K., Mortimer, M., Kahru, A., 2013. Toxicity of 608 Ag, CuO and ZnO nanoparticles to selected environmentally relevant test organisms and mam- 609 malian cells in vitro: A critical review. Arch. Toxicol. 87, 1181–1200. 610 https://doi.org/10.1007/s00204-013-1079-4 611 Chong, M.N., Jin, B., Chow, C.W.K., Saint, C., 2010. Recent developments in photocatalytic water 612 treatment technology: A review. Water Res. 44, 2997–3027. https://doi.org/10.1016/j.wa- 613 tres.2010.02.039 614 Debroye, E., Van Loon, J., Yuan, H., Janssen, K.P.F., Lou, Z., Kim, S., Majima, T., Roeffaers, M.B.J., 615 2017. Facet-Dependent Photoreduction on Single ZnO Crystals. J. Phys. Chem. Lett. 8, 340–346. 616 https://doi.org/10.1021/acs.jpclett.6b02577 617 Deniz, F., Saygideger, S.D., Karaman, S., 2011. Response to copper and sodium chloride excess in 618 Spirulina sp. (Cyanobacteria). Bull. Environ. Contam. Toxicol. 87, 11–15. 619 https://doi.org/10.1007/s00128-011-0300-5 620 Dere, S., Gunes, T., Sivaci, R., 1998. Spectrophotometric Determination of Chlorophyll - A, B and 621 Total Carotenoid Contents of Some Algae Species Using Different Solvents. Turkey J. Bot. 22, 622 13–17. https://doi.org/10.1080/19393550802541200 623 Djearamane, S., Lim, Y.M., Wong, L.S., Lee, P.F., 2018. Cytotoxic effects of zinc oxide nanoparticles 624 on cyanobacterium Spirulina (Arthrospira) platensis. PeerJ 6, e4682. 625 https://doi.org/10.7717/peerj.4682 626 Everaert, G., De Laender, F., Goethals, P.L.M., Janssen, C.R., 2015. Relative contribution of persis- 627 tent organic pollutants to marine phytoplankton biomass dynamics in the North Sea and the Kat- 628 tegat. Chemosphere 134, 76–83. https://doi.org/10.1016/j.chemosphere.2015.03.084 629 Fu, F., Wang, Q., 2011. Removal of heavy metal ions from wastewaters: A review. J. Environ. Ma- 630 nage. 92, 407–418. https://doi.org/10.1016/j.jenvman.2010.11.011 631 Han, C., Yang, M.Q., Weng, B., Xu, Y.J., 2014. Improving the photocatalytic activity and anti-photo- 632 corrosion of semiconductor ZnO by coupling with versatile carbon. Phys. Chem. Chem. Phys. 633 16, 16891–16903. https://doi.org/10.1039/c4cp02189d 634 Harris, T.D., Smith, V.H., 2016. Do persistent organic pollutants stimulate cyanobacterial blooms? Inl. 635 Waters 6, 124–130. https://doi.org/10.5268/IW-6.2.887 636 Hatamie, A., Khan, A., Golabi, M., Turner, A.P.F., Beni, V., Mak, W.C., Sadollahkhani, A., Alnoor, 637 H., Zargar, B., Bano, S., Nur, O., Willander, M., 2015. Zinc Oxide Nanostructure-Modified 31 638 Textile and Its Application to Biosensing, Photocatalysis, and as Antibacterial Material. 639 Langmuir 31, 10913–10921. https://doi.org/10.1021/acs.langmuir.5b02341 640 Hou, J., Wu, Y., Li, X., Wei, B., Li, S., Wang, X., 2018. Toxic effects of different types of zinc oxide 641 nanoparticles on algae, plants, invertebrates, vertebrates and microorganisms. Chemosphere 193, 642 852–860. https://doi.org/10.1016/j.chemosphere.2017.11.077 643 Ishioka, J., Kogure, K., Ofuji, K., Kawaguchi, K., Jeem, M., Kato, T., Shibayama, T., Watanabe, S., 644 2017. In situ direct observation of photocorrosion in ZnO crystals in ionic liquid using a laser- 645 equipped high-voltage electron microscope. AIP Adv. 7. https://doi.org/10.1063/1.4979726 646 Jury, W.A., Vaux, H.J., 2007. The Emerging Global Water Crisis: Managing Scarcity and Conflict Be- 647 tween Water Users. Adv. Agron. 95, 1–76. https://doi.org/10.1016/S0065-2113(07)95001-4 648 Jury, W.A., Vaux, H.J., 2007. The Emerging Global Water Crisis: Managing Scarcity and Conflict Be- 649 tween Water Users. Adv. Agron. 95, 1–76. https://doi.org/10.1016/S0065-2113(07)95001-4 650 Khan, M.S., Zaidi, R., Goel, R., Musarrat, J., 2011 Biomanagement of metal-contaminated soild, 651 Spinger. https://doi.org/10.1007/978-94-007-1914-9 652 Khan, M.U., Mitchell, K., 1987. Chlorophylls Carotenoids. Methods Enzymol. 148, 350–382. 653 https://doi.org/10.1016/0076-6879(87)48036-1 654 Lee, K.M., Lai, C.W., Ngai, K.S., Juan, J.C., 2016. Recent developments of zinc oxide based photo- 655 catalyst in water treatment technology: A review. Water Res. 88, 428–448. 656 https://doi.org/10.1016/j.watres.2015.09.045 657 Liu, J., Wang, Y., Ma, J., Peng, Y., Wang, A., 2018. A review on bidirectional analogies between the 658 photocatalysis and antibacterial properties of ZnO. J. Alloys Compd. 783, 898–918. 659 https://doi.org/10.1016/j.jallcom.2018.12.330 660 Lone, J.A., Kumar, A., Kundu, S., Lone, F.A., Suseela, M.R., 2013. Characterization of tolerance limit 661 in spirulina platensis in relation to nanoparticles. Water. Air. Soil Pollut. 224. 662 https://doi.org/10.1007/s11270-013-1670-6 663 Ma, H., Williams, P.L., Diamond, S.A., 2013. Ecotoxicity of manufactured ZnO nanoparticles - A re- 664 view. Environ. Pollut. 172, 76–85. https://doi.org/10.1016/j.envpol.2012.08.011 665 Malato, S., Fernández-Ibáñez, P., Maldonado, M.I., Blanco, J., Gernjak, W., 2009. Decontamination 666 and disinfection of water by solar photocatalysis: Recent overview and trends. Catal. Today 147, 667 1–59. https://doi.org/10.1016/j.cattod.2009.06.018 668 Mani, D., Kumar, C., 2014. Biotechnological advances in bioremediation of heavy metals contami- 669 nated ecosystems: An overview with special reference to phytoremediation. Int. J. Environ. Sci. 670 Technol. 11, 843–872. https://doi.org/10.1007/s13762-013-0299-8 671 Maule, A., Wright, S.J.L., 1984. Herbicide effects on the population growth of some green algae and 672 cyanobacteria. J. Appl. Bacteriol. 57, 369–379. https://doi.org/10.1111/j.1365- 673 2672.1984.tb01403.x 674 Miao, A.J., Zhang, X.Y., Luo, Z., Chen, C.S., Chin, W.C., Santschi, P.H., Quigg, A., 2010. Zinc ox- 675 ide-engineered nanoparticles: Dissolution and toxicity to marine phytoplankton. Environ. Toxi- 676 col. Chem. 29, 2814–2822. https://doi.org/10.1002/etc.340 32 677 Mizuta, T., Ishibashi, T., Minemoto, T., Takakura, H., Hamakawa, Y., 2006. Chemical deposition of 678 zinc oxide thin films on silicon substrate. Thin Solid Films 515, 2458–2463. 679 https://doi.org/10.1016/j.tsf.2006.06.035 680 Moraes, C.C., Sala, L., Cerveira, G.P., Kalil, S.J., 2011. C-Phycocyanin extraction from Spirulina 681 platensis Wet. Biomass. 28, 45–49. http://dx.doi.org/10.1590/S0104-66322011000100006 682 Moreira, F.C., Boaventura, R.A.R., Brillas, E., Vilar, V.J.P., 2017. Electrochemical advanced oxida- 683 tion processes: A review on their application to synthetic and real wastewaters. Appl. Catal. B 684 Environ. 202, 217–261. https://doi.org/10.1016/j.apcatb.2016.08.037 685 Mortimer, M., Kasemets, K., Kahru, A., 2010. Toxicity of ZnO and CuO nanoparticles to ciliated pro- 686 tozoa Tetrahymena thermophila. Toxicology 269, 182–189. 687 https://doi.org/10.1016/j.tox.2009.07.007 688 Ranjith, K.S., Castillo, R.B., Sillanpaa, M., Rajendra Kumar, R.T., 2018. Effective shell wall thickness 689 of vertically aligned ZnO-ZnS core-shell nanorod arrays on visible photocatalytic and photo 690 sensing properties. Appl. Catal. B Environ. 237, 128–139. https://doi.org/10.1016/j.ap- 691 catb.2018.03.099 692 Serrà, A., Zhang, Y., Sepúlveda, B., Gómez, E., Nogués, J., Michler, J., Philippe, L., 2019. Highly ac- 693 tive ZnO-based biomimetic fern-like microleaves for photocatalytic water decontamination using 694 sunlight. Appl. Catal. B Environ. 248, 129–146. https://doi.org/10.1016/j.apcatb.2019.02.017 695 Singh, J.S., Seneviratne, G., Health, M.C., 2017 Agro-Environmental Sustainability. Springer. 696 https://doi.org/10.1007/978-3-319-49727-3 697 Subashchandrabose, S.R., Ramakrishnan, B., Megharaj, M., Venkateswarlu, K., Naidu, R., 2013. Mix- 698 otrophic cyanobacteria and microalgae as distinctive biological agents for organic pollutant deg- 699 radation. Environ. Int. 51, 59–72. https://doi.org/10.1016/j.envint.2012.10.007 700 T, R., McMullan, R, M., P, N., 2011. Remediation of dyes in textile effluent: a critical review on cur- 701 rent treatment technologies with a proposed alternative. Bioresour Technol 77, 247–255. 702 https://doi.org/10.1016/S0960-8524(00)00080-8 703 Torabi, A., Staroverov, V.N., 2015. Band Gap Reduction in ZnO and ZnS by Creating Layered 704 ZnO/ZnS Heterostructures. J. Phys. Chem. Lett. 6, 2075–2080. 705 https://doi.org/10.1021/acs.jpclett.5b00687 706 Vessalli, B.A., Zito, C.A., Perfecto, T.M., Volanti, D.P., Mazon, T., 2017. ZnO nanorods/graphene ox- 707 ide sheets prepared by chemical bath deposition for volatile organic compounds detection. J. Al- 708 loys Compd. 696, 996–1003. https://doi.org/10.1016/j.jallcom.2016.12.075 709 Wang, C.Y., Fu, C.C., Liu, Y.C., 2007. Effects of using light-emitting diodes on the cultivation of 710 Spirulina platensis. Biochem. Eng. J. 37, 21–25. https://doi.org/10.1016/j.bej.2007.03.004 711 Weng, B., Yang, M.Q., Zhang, N., Xu, Y.J., 2014. Toward the enhanced photoactivity and photosta- 712 bility of ZnO nanospheres via intimate surface coating with reduced graphene oxide. J. Mater. 713 Chem. A 2, 9380–9389. https://doi.org/10.1039/c4ta01077a 714 Wong, S.W.Y., Leung, P.T.Y., Djurišić, A.B., Leung, K.M.Y., 2010. Toxicities of nano zinc oxide to 715 five marine organisms: Influences of aggregate size and ion solubility. Anal. Bioanal. Chem. 716 396, 609–618. https://doi.org/10.1007/s00216-009-3249-z 33 717 Yadav, H.M., Kim, J.S., Pawar, S.H., 2016. Developments in photocatalytic antibacterial activity of 718 nano TiO2: A review. Korean J. Chem. Eng. 33, 1989–1998. https://doi.org/10.1007/s11814- 719 016-0118-2 720 Yu, L., Chen, W., Li, D., Wang, J., Shao, Y., He, M., Wang, P., Zheng, X., 2015. Applied Catalysis 721 B : Environmental Inhibition of photocorrosion and photoactivity enhancement for ZnO via spe- 722 cific hollow ZnO core / ZnS shell structure. Applied Catal. B, Environ. 164, 453–461. 723 https://doi.org/10.1016/j.apcatb.2014.09.055 724 725 726 727 34 Highly reduced ecotoxicity of ZnO-based micro/nanostructures on aquatic biota: Influence of architecture, chemical composition, fixation, and photocatalytic efficiency Albert Serrà ‖,*, Yue Zhang &, Borja Sepúlveda &, Elvira Gómez $, @, Josep Nogués &, #, Johann Michler ‖, Laetitia Philippe ‖ ‖ Empa, Swiss Federal Laboratories for Materials Science and Technology, Laboratory for Mechanics of Materials and Nanostructures, Feuerwerkerstrasse 39, CH-3602 Thun, Switzerland. & Catalan Institute of Nanoscience and Nanotechnology (ICN2), CSIC and BIST, Campus UAB, Bellaterra, E-08193 Barcelona, Spain. $ Grup d’Electrodeposició de Capes Primes i Nanoestructures (GE-CPN), Departament de Ciència de Materials i Química Física, Universitat de Barcelona, Martí i Franquès, 1, E-08028, Barcelona, Catalonia, Spain. @ Institute of Nanoscience and Nanotechnology (IN2UB), Universitat de Barcelona, Barcelona, Catalonia, Spain. # ICREA, Pg. Lluís Companys 23, E-08010, Barcelona, Spain. Corresponding author: albert.serraramos@empa.ch (A.S.) 1 Photograph of ZnO-based photocatalysts Figure S1: Photograph of (a) the as-prepared ZnO-based ferns and (b) the ZnO-based nanoparticles. 2 Scanning electron microscopy images of ZnO and ZnO@ZnS(12h) nano/microferns Figure S2: Representative FE-SEM micrographs of (a) the ZnO and (b) the ZnO@ZnS(12h) surfaces of the micro/nanoferns. Scale bar: 1 µm. EDS mapping of (c) the ZnO@ZnS(4h), (d)the ZnO@ZnS(8h), and (e) the ZnO@ZnS(12h) micro/nanoferns. Scale bar: 400 nm. 3 Brunauer-Emmett-Teller surface areas Figure S3: N2 adsorption-desorption isotherms of the different ZnO-based photocatalysts. 4 XRD of ZnO-based photocatalysts Figure S4: (a) XRD patterns of (i) the ZnO films and (ii) the ZnO nanoparticles and (iii) the ZnO@ZnS(4h) nanoparticles. (b) XRD patterns of ZnO-based bioinspired fern-shape microleaves. The diffraction peaks that correspond to the fluorine-doped tin oxide substrate structures are indicated with an *. The main peaks corresponding to ZnO and ZnS are highlighted by . 5 UV-vis DRS absorption of ZnO-based photocatalysts Figure S5: UV-vis DRS absorption spectra of (a) ZnO films, (b) ZnO-based micro/nanoferns, and (c) ZnO-based nanoparticles. 6 Tauc plots of the ZnO-based photocatalysts Figure S6: Tauc plots from the UV-vis analysis of (a) ZnO films, (b) ZnO-based nanoparticles, (c) the ZnO nano/microferns, (d) the ZnO@ZnS(4h) nano/microferns, (e) the ZnO@ZnS(8h) nano/microferns, and (f) the ZnO@ZnS(12h) nano/microferns. 7 Photoluminescence spectra of ZnO-based photocatalysts Figure S7: Room temperature photoluminescence spectra of (a) the ZnO films, (b) the ZnO-based micro/nanoferns, (c) the ZnO-based nanoparticles. 8 Hydroxyl radical trapping experiments Figure S8: (a) Calibration of Fluorescein sodium in aqueous solution using the PL intensity at 515 nm (λex = 303). (b) Time-dependent evolution of the Fluorescein sodium concentration under UV-filtered simulated sunlight in the absence and using 400 mg L-1 of ZnO-based photocatalysts (temperature = 30.0 ±0.1 oC). 9 Mineralization efficiencies of ZnO-based photocatalysts The performance of each photocatalyst to degrade the multi-pollutant solution was evaluated by quantifying the reduction in TOC (Figure S9) after the irradiation of each sample under natural sunlight during 210 min. As Tables 1 and S1 show, the ZnO@ZnS(4h) core@shell micro/nanoferns were the most efficient photocatalyst. Such differences in the photocatalytic performance can be explained by the combination of three synergetic elements: (i) the electronic properties, which determine the ability of the photocatalysts to produce reactive oxygen spices (ROS) when irradiated with visible light, and (ii) the shape and architecture, which strongly determine not only the affinity and surface reactivity of the photocatalysts, but also their light trapping ability. In that sense, large surface areas, abundant active sites, and an enhanced ability to trap light and pollutants of the ZnO micro/nanoferns due to their dendritical architecture were responsible for the improved activity of micro/nanoferns compared with that of ZnO films and ZnO nanoparticles. By extension, the increased photocatalytic degradation and mineralization efficiencies of the ZnO@ZnS core@shell nanoparticles and micro/nanoferns compared to those of pristine ZnO architectures resulted from their narrower bandgap and, in turn, from their exceptional ability to harvest larger fractions of visible light. Among the various studied photocatalysts, ZnO@ZnS(4h) core@shell micro/nanoferns are the best candidates for photodecontamination in terms of mineralization efficiency, since they combine a narrower bandgap and a dendritical microstructure. A longer sulfidation did not improve the electronic properties, and the microfern architecture was affected by a reduction in the active and accessible surface area, which in conjunction reduced the catalytic performance. 10 Table S1: Photocatalytic degradation efficiency of MB (5 ppm) + 4-NP (5 ppm) + Rh-B (5 ppm) polluted solutions under artificial sunlight (λ> 400 nm), and natural sunlight (λ> 400 nm), by using the ZnO film, the ZnO and ZnO@ZnS core@shell nanoparticles and the ZnO and ZnO@ZnS core@shell bioinspired fern-shape microleaves (0.4 mg mL-1). Artificial UV-filtered Natural UV-filtered TOC of pollutant sunlight irradiation sunlight irradiation Photocatalyst solution without (210 min λ> 400 nm) (210 min λ> 400 nm) irradiation / ppm Mineralization Mineralization efficiency / % efficiency / % ZnO film 15.14 ± 0.02 5.1 ± 0.7 4.3 ± 0.5 ZnO nanoparticles 15.26 ± 0.13 19.2 ± 0.4 17.9 ± 0.1 ZnO@ZnS(4h) 15.19 ± 0.09 68.4 ± 0.2 61.7 ± 0.9 nanoparticles ZnO micro/nanoferns 15.18 ± 0.07 24.1 ± 0.3 20.1 ± 0.4 ZnO@ZnS(4h) 15.34 ± 0.11 86.9 ± 0.8 81.4 ± 0.3 micro/nanoferns ZnO@ZnS(8h) 15.05 ± 0.09 76.5 ± 1.2 71.0 ± 0.6 micro/nanoferns ZnO@ZnS(12h) 15.44 ± 0.12 45.9 ± 0.9 39.9 ± 0.9 micro/nanoferns Figure S9: Mineralization efficiency after 210 min of artificial and natural (λ > 400 nm) sunlight irradiation for the ZnO film (dark golden), the ZnO nanoparticles (dark cyan), the ZnO@ZnS(4h) nanoparticles (dark violet), the ZnO micro/nanoferns (gray), ZnO@ZnS(4h) micro/nanoferns (blue), the ZnO@ZnS(8h) micro/nanoferns (green), and the ZnO@ZnS(12h) micro/nanoferns (red). 11 FE-SEM micrographs of reused micro/nanofern photocatalysts Figure S10: FE-SEM micrographs of fern-like bioinspired microleaves of (a, b) ZnO, (c, d) ZnO@ZnS(4h), (e, f) ZnO@ZnS(8h), and (g, h) ZnO@ZnS(12h) after their continuous irradiation in fresh water during (a, c, e , and g) 0 and (b, d , f, and h) 96 h under artificial sunlight (λ>400 nm). Scale bar: 500 nm (a, c), 350 (e, g) or 90 nm (b, d, f, and h). 12 Spirulina (Arthrospira) platensis tank Figure S11: (a) Spirulina (Arthrospira) platensis tank. (b) Dried clumps which appeared after the incubation of 400 mg L-1 of ZnO nanoparticles with microalgae during 96 h. (c) Dried microalgae after the incubation of 400 mg L-1 of ZnO@ZnS(4h) micro/nanoferns with microalgae during 96 h. 13 Ecotoxicological effect of ZnO and ZnO@ZnS(4h) nanoparticles on microalgae Figure S12: Percentage of loss in (a) microalgae viability, (b) microalgae biomass, (c) chlorophyll-a of microalgae, (d) carotenoids of microalgae, and (e) phycocyanin pigment of microalgae for the treatment with 25, 50, 100, 200, and 400 mg L-1 of ZnO and ZnO@ZnS(4h) nanoparticles, after exposure times of 6, 24, 48, 72, and 96 h. The microalgae cultures were irradiated 8 h per day with continuous UV-filtered simulated sunlight (light intensity of 740 ± 15 lx), starting the first irradiation cycle when photocatalysts were introduced in microalgae culture. Percentages are relative to the control (microalgae culture without catalyst) cultivated in the same experimental conditions. 14 Percentage of photosynthetic pigments of Spirulina microalgae treated with ZnO- based biomimetic microferns Figure S13: Percentage of loss in (a) chlorophyll-a of the microalgae, (b) carotenoids of the microalgae, and (c) phycocyanin pigment of the microalgae for the treatment with 25, 50, 100, 200, and 400 mg L-1 of ZnO-based bioinspired micro/nanoferns after exposure times of 6, 24, 48, 72, and 96 h. The microalgae cultures were irradiated 8 h per day with continuous simulated sunlight (light intensity of 740 ± 15 lx), starting the first irradiation cycle when photocatalysts were introduced in microalgae culture. Percentages are relative to the control (microalgae culture without catalyst) cultivated in the same experimental conditions. 15 Figure S14: Percentage of loss in (a) chlorophyll-a of the microalgae, (b) carotenoids of the microalgae, and (c) phycocyanin pigment of the microalgae for the treatment with 25, 50, 100, 200, and 400 mg L-1 of fixed and non-fixed ZnO@ZnS(4h) micro/nanoferns after an exposure time of 6, 24, 48, 72, and 96 h. The microalgae cultures were irradiated 8 h per day with continuous simulated sunlight (light intensity of 740 ± 15 lx), starting the first irradiation cycle when photocatalysts were introduced in microalgae culture. Percentages are relative to the control (microalgae culture without catalyst) cultivated in the same experimental conditions. 16