A sustainable natural clam shell derived photocatalyst for the effective adsorption and photodegradation of organic dyes

In response to an increasing desire for modern industries to be both green and sustainable, there has been increasing research focus on the reutilization of natural waste materials to effectively remove and degrade toxic wastewater effluents. One interesting food industry waste product is clam shell. Here a new photocatalytic nanomaterial derived from marine clam shells was successfully prepared and characterized. Thereafter the material was applied for the removal of two target dyes from aqueous solution, where the effect of both catalyst dose and initial dye concentration on adsorption and photocatalysis was investigated. The maximum adsorption capacities of methylene blue (100 mg/L) and Congo red (500 mg/L) were 123.45 mg/g and 679.91 mg/g, respectively, where adsorption followed pseudo second order kinetics predominantly via a chemical adsorption process. The photodegradation removal efficiencies of the two dye solutions under visible light irradiation were 99.6% and 83.3% for MB and CR, respectively. The excellent degradation performance in a mixed dye solution, with strong degradation capability and low cost, demonstrated that the clam shell catalyst material was a good candidate for practical field remediation of dye contaminated wastewater.

A sustainable natural clam shell derived photocatalyst for the effective adsorption and photodegradation of organic dyes Ting Qu 1,5 , Xinxin Yao 2,4,5 , Gary Owens 3 , Liangjun Gao 2 & Hailong Zhang 1,2* In response to an increasing desire for modern industries to be both green and sustainable, there has been increasing research focus on the reutilization of natural waste materials to effectively remove and degrade toxic wastewater effluents. One interesting food industry waste product is clam shell. Here a new photocatalytic nanomaterial derived from marine clam shells was successfully prepared and characterized. Thereafter the material was applied for the removal of two target dyes from aqueous solution, where the effect of both catalyst dose and initial dye concentration on adsorption and photocatalysis was investigated. The maximum adsorption capacities of methylene blue (100 mg/L) and Congo red (500 mg/L) were 123.45 mg/g and 679.91 mg/g, respectively, where adsorption followed pseudo second order kinetics predominantly via a chemical adsorption process. The photodegradation removal efficiencies of the two dye solutions under visible light irradiation were 99.6% and 83.3% for MB and CR, respectively. The excellent degradation performance in a mixed dye solution, with strong degradation capability and low cost, demonstrated that the clam shell catalyst material was a good candidate for practical field remediation of dye contaminated wastewater.
While water is essential for all life, the rapid growth in modern human civilizations globally, concurrent with vigorous and rapid industrialization has resulted in significant pollution of water resources through the discharge of waste materials, especially by the dye industry. Water resource contamination by dye wastewaters is becoming increasingly serious 1 . Methylene blue (MB), Congo red (CR) and many other organic dyes are widely and routinely used in large quantities by various dye industries. These dye molecules have a complex molecular structure containing many diverse functional groups such as amino, hydroxyl and the benzene ring 2 , which while giving the dye strong coloring ability also results in strong toxicity and poor biodegradability 3 . These latter features are of concern because the efflux of large concentrations of dyes into the natural environment may thus lead to environmental harm. Hence, development of new materials to treat dye wastewater is urgently required to mitigate environmental pollution. While many non-destructive methods, such as adsorption, membrane filtration, and precipitation, have previously been successfully employed to remove dye from wastewater, the disadvantage of non-destructive methods are that do they nor destroy the dye but merely concentrate the dye in either sludge or silt, which still needs further treatment to remove the hazard [4][5][6] . In comparison, destructive methods, can directly decompose large dyes into non-toxic smaller molecules, which can be directly discharged into the environment. Of these destructive methods, photocatalytic reduction has emerged as the most popular. For example, bismuth oxychloride nanomaterials were synthesised via a hydrothermal method, and used for the degradation of Congo red with a 72% photocatalytic removal efficiency 7 . A Fe 2 O 3 -CeO 2 nanocomposite prepared by a Preparation of clam shell catalyst. Clam shells were initially washed with tap water to remove any obvious contamination from the surface of the clam shell. Thereafter, clam shells were immersed in a solution of 1 M sodium hydroxide and heated to 85 °C in a water bath remove any residual meat from the surface of the clam shells, then, the samples were rinsed with fresh water (pH = 7), and evaporated until dryness at 80 °C on a stove. The clam shells were soaked in 1 M hydrochloric acid solution in a clam pot overnight, washed with deionized water several times and dried. Finally, the clam shell was calcined in a muffle furnace for 2 h at 800, 900, and 1000 °C respectively. After calcination, the white powder obtained was ground, sieved (100 mesh) and stored in a desiccator for later use.
Characterization. Specific surface area and pore size distribution of samples, was calculated using the BET method from N 2 adsorption/desorption isotherms obtained using a static volumetric adsorption analyzer (Micromeritics ASAP 2010, Shanghai, China). Morphology and microstructure of the as-prepared samples was determined via a scanning electron microscope (SEM; S4800, Hitachi, Japan) and high-resolution transmission electron microscope (HR-TEM; Joel-2100, JEOL, Japan). Crystal structures were analyzed by X-ray diffraction (XRD, RIGAKU miniflex/600, Japan) in the range between 20° and 80°. Thermogravimetric analysis was conducted on a TGA Q5000 (American TA instrument Q Series) to better understand phase transformations during www.nature.com/scientificreports/ calcination. Optical properties of samples, were examined using a Cary 500 UV-vis NIR spectrophotometer in the range 200-800 nm. The chemical states of elements at the near surface of samples were also studied using X-ray photoelectron spectroscopy (XPS, Phi 5000C ESCA system, USA).

Photocatalytic experiment.
As-prepared photocatalyst (20 mg) was added into an aqueous solution of either MB or CR (40 mL) and the solution was magnetically stirred in the dark. for 2 h for the reaction to achieve adsorption equilibrium. During this time aliquots of the supernatant solution were removed at regular intervals (i.e., 0, 5, 10, 20, 30, 60, 90 and 120 min) an analyzed for the residual dye concentration. To assess adsorption performance, residual dye concentrations were determined photometrically. After 2 h dark reaction, a light source (xenon lamp, HSX-F300) was used to initiate photocatalytic experiments. The vertical distance between the light source and the reactor was 10 cm, and supernatant samples were again taken at regularly intervals to assess photocatalytic performance. In the photocatalytic system, a cooling reflux device was also connected to the reaction vessel to eliminate the influence of heat generated by light on the experiment.
Determination of dye concentration. After the reaction single dye solution samples were initially centrifuged at 4000 r min -1 for 10 min, and the absorbance of the supernatant subsequently measured using an UVvis spectrophotometer. The wavelengths of absorbance maxima for the two dyes were λ MB : 664, and λ CR : 497 nm. The removal efficiency (R%) of adsorbed dye was calculated by Eq. (1), where the amount of dye adsorbed on the catalyst q e (mg g -1 ) was calculated by Eq. (2).
where C 0 and C e (mg L -1 ) were respectively, the initial and equilibrium dye concentrations in solution, m (g) was the mass of photocataclyst and V (L) was the volume of dye solution. In order to study reaction kinetics, the rate constant k was obtained by curve fitting data to Eq. (3) where C' 0 is the initial dye concentration after dark reaction and C t is the dye concentration at light irradiation time t.
Adsorption kinetics under dark reaction. Adsorption kinetics of the single dye system was further studied by adding catalyst to the dye solution under dark conditions with constant magnetic stirring and measuring the residual dye concentration in the solution at fixed time periods. The kinetic data so obtained was fit to both the pseudo first order (PFO Eq. 4) and second order (PSO Eq. 5) kinetics to evaluate the kinetics of adsorption.
where q t (mg g -1 ) is the dye adsorption capacity at time (t); and k 1 (min -1 ) and k 2 (g mg -1 min -1 ) are the rate constants for PFO and PSO models, respectively.

Results and discussion
Characterization of catalyst. Morphology. The changes in clam shell powder morphology following calcination at three different temperatures were monitored using SEM and TEM (Fig. 1). Before calcination, clam shell powder had a spongy porous structure (Fig. a,b), due to presence of organic matter and a calcium carbonate skeleton after soaking in acid. However, after calcination at 800 °C, the clam shell surface became fragmented and bulky, due to the consumption of organic matter and growth of crystal particles (Fig. 1c,d). As calcination temperature further increased (900 and 1000 °C), clam shell powder morphology and surface structure became more compact. The increase in particle size was attributed to particle sintering, resulting in Ostwald ripening, which causes growth of larger particles at the expense of smaller ones. This was consistent with previous results which had shown an increase in average surface grain size with calcination temperature 19 .
More detailed morphological characteristics of clam shell calcined at 1000 °C were obtained by HR-TEM ( Fig. 1i-k), and showed that after high temperature calcination the clam shell catalyst (CSC) had a layered structure where the distance between two adjacent lattices at the interface was 0.286 nm, which was consistent with XRD results (X-ray diffraction analysis) corresponding to the CaO (111) crystal face. In addition, the observed heterojunctions between dislocations in the crystal structure may be due to the rearrangement of additional atomic layers produced by the presence of trace metals in the CSC during calcination 20 . This may enable greater www.nature.com/scientificreports/ electron migration to the conductive band of CaO, since the energy band gap is reduced and charge carrier separation is increased, therefore improving photocatalytic performance.
BET analysis. All samples exhibited Type IV isotherms (Fig. S1 a and b) with hysteresis loops. The presence of small hysteresis loops at P/P 0 > 0.4 indicated the existence of both microporous and mesoporous structures 21 .
After calcination, specific surface area, pore volume and average pore diameter of the catalyst all decreased (Table S1). This was because as calcination temperature increased, the inner part of the particles shrunk more, leading to an overall decrease in the pore size of the material. However, specific surface area of the catalyst did increase slightly between 800 and 1000 °C, which was attributed to the formation of micro pores following decomposition of calcium carbonate. Adherence to the Type IV isotherm model clearly indicated the formation of multilayers on the catalyst surface during high temperature reactions, which was also clearly observed in TEM imagery ( Fig. 1  Thermogravimetric analysis. Thermogravimetric analysis provides further information on the transformation of clam shell powder during calcination. The initial mass lost from clam shell was only 0.47% due to the loss of water between 30-200 °C (Fig. 2b). Thereafter, the curve decreased slightly between 200-500 °C, due to the decomposition and removal of organic components of the clam shell, whereas weight loss between 500-610 °C was attributed to a phase transformation from aragonite to calcite. Thereafter, the mass of clam shell decreased significantly between 610 and 800 °C, due to further conversion of calcite to CaO and CO 2 , during calcination 10 . At 800-900 °C, further mass loss was caused by removal of any residual intermediate products from incomplete conversion of CaCO 3 to CaO, and thereafter the slight decrease observed at 900-1000 °C, may be due to the presence of a small amount of unconverted CaCO 3 and further mass loss of trace metals. All these results were in good agreement with both XRD (X-ray diffraction analysis) and XPS (XPS analysis) analysis. www.nature.com/scientificreports/ Ultraviolet visible spectrum analysis. Diffuse reflectance spectroscopy in the UV-vis range was employed to investigate the optical properties of uncalcined clam shell powder and clam shell calcined at three different temperatures (Fig. 2c), as well as the light absorption characteristics and band gap energy of the CSCs (Fig. 2d). The wavelength absorption maxima of uncalcinated, 800, 900 and 1000 °C samples were 277, 271, 276 and 273 nm, respectively (Fig. 2c), corresponding to the optical band gap obtained from Tauc diagram (Fig. 2d). The optical band gap of the catalysts decreased with an increase in calcination temperature, and thus the lowest band gap (2.64 eV) was obtained at 1000 °C. As the calcination temperature increased, the weight of clam shell decreased, and the specific gravity of trace metal elements increased. Therefore, the decrease in optical band gap with increased calcination temperature was attributed to the appearance of localised energy states in the CaO band gap and the existence of a higher content of lattice defects, i.e., increases in trace metals and oxygen vacancies 23 . High resolution TEM (Morphology) had also confirmed the existence of a large number of lattice defects.
XPS analysis. X-ray photoelectric spectroscopy (XPS) was used to analyze the valence state and content of surface elements in clam shell powder before and after high temperature (1000 °C) calcination, by comparing XPS spectra of clam shell treated at 800 and 1000 °C (Fig. 3). Survey scans (Fig. 3a), indicated that the main components of clam shells at 800 and 1000 °C were C, O, Ca and a small amount of Na. It is likely that traces of other metals such as magnesium (Mg) are also present but these would be at levels well below the detection limit of XPS (i.e., < 0.01%). For clam shell powder calcined at 800 °C, the XPS spectra (Fig. 3b) exhibited many superimposed C1s peaks (C1-C3), where C1 with a binding energy of 285 eV, was attributed to amorphous carbon/adventitious carbon, a small standard C2 peak at 286.1 eV, corresponded to the C-O, and the C3 peak at 289.4 eV corresponded to a binding energy commonly associated with metal carbonate MCO 3 , and was thus attributed to CaCO 3 . In comparison, for clam shell calcined at 1000 °C (Fig. 3e), the binding energy of the adventitious carbon peak C1 occurred at 285 eV, the C2 peak for organic carbon (C-O bond) occurred at 286 eV, the C3 peak attributable to COOR at 288 eV, and the C4 peak corresponding to MCO 3 at 289.4 eV. However, compared with the sample calcined at 800 °C, the sample calcined at 1000 °C, exhibited two extra weak peaks (C5 and C6) at 292 and 295 eV, respectively, which were attributed to compounds formed by carbon and/or halogen. The observed peak (O1) at 530.8 eV for the 800 °C calcined sample in the O1s XPS spectra was attributed to CO 3 2of CaCO 3 (Fig. 3c); whereas the binding energy of the peak at 531.6 eV (peak O2) for the clam shell calcined at 1000 °C (Fig. 3f). corresponded to CaO.
Calcium typically exhibits two characteristic binding energies, i.e., Ca2p 3/2 (346 eV) and Ca2p 1/2 (349.6 eV) 24 . Here high-energy resolution spectra of Ca2p peaks showed a complex pattern that could be deconvoluted into www.nature.com/scientificreports/ three (Ca1, Ca2 and Ca3) constituent peaks (Fig. 3d,g). For sample calcined at 800 °C, peak Ca1 (346 eV), corresponded to CaCO 3 and peak Ca2 (347.1 eV) was associated with CaO. In comparison, for the 1000 °C calcined sample the area of the CaO peak (Ca2) increased, indicating that the content of CaO was relatively larger when the sample was pyrolyzed at the higher temperature.
Adsorption-photodegradation performance. Effect of different catalysts. While the total adsorptionphotodegradation removal efficiency of MB and CR for calcined clam shells at 1000 °C reached 99.7 and 90.8%, respectively, under dark conditions, the removal efficiencies of MB and CR reached only 25.9 and 70.5%, respectively (Fig. 4). In comparison the total absorption-photodegradation removal efficiency of untreated clam shells (MB: 15.9%, CR: 10.4%), and those calcined at 800 °C (MB: 53.9%, CR: 25.9%), or 900 °C, (MB: 85.7%, CR: 81.1%) were lower. This indicated that the adsorption and/or photocatalytic properties of uncalcined clam shell or clam shell calcined at temperatures < 1000 °C were poorer than clam shell calcined at 1000 °C. It can also be clearly seen (Fig. 4) that the adsorption performance (dark environment) of clam shell calcined at different temperatures was consistently better than that of non-calcined clam shell. This suggested that decomposition of residual organic matter in the original clam shell via heat treatment during calcination, results in significantly more pore structure. Where the increase in adsorption performance with calcination temperature, could be attributed to increased pore forming effects from CO 2 generated by thermal decomposition of CaCO 3 , the main component of clam shell. The photocatalytic activity (under light) of the clam shell powder was also enhanced www.nature.com/scientificreports/ by an increase in calcination temperature. This may be due to an increase in the relative content of calcium oxide due to heating and the interaction of trace transition metals in the clam shell. HR-TEM images confirmed the presences of a significant number of dislocations and defects in the CaO crystal structure. Overall, results from the adsorption-photodegradation study indicated that the best photocatalytic performance was obtained when clam shell powder was calcined at 1000 °C. This was attributed to synergetic effects of enhanced absorption and increased heterojunctions among some dislocations in the crystal structures, which also improved photodegradation efficiency. Photodegradation kinetics of both MB and CR by CSCs varied with the calcination temperature of the catalyst (Fig. 4b,d). The kinetic data for the degradation of both MB and CR best fit the pseudo first order kinetic model (R 2 approaching 1), where the rate constant (k) increased with calcination temperature (Table 1), and the higher the k value, the better the photocatalytic performance and the greater the photodegradation amount. Thus, in all subsequent experiments, only clam shell calcined at 1000 °C was investigated.
The photodegradation of both MB and CR by 1000 °C calcined CSC occurs in the visible light region of the absorption spectrum (Fig. 4e,f). As reaction time increased, peaks of MB and CR both decreased sharply, and some peaks disappeared or were slightly shifted in wavelength, which indicated that both MB and CR molecules were being destroyed, and that new substances (degradation products) were also formed 10 .
Adsorption performances. The adsorption performance of 1000 °C calcined clam shell powder varied with the initial concentrations of MB and CR within 2 h. (Fig. S2) Over 2 h the concentration of both MB or CR decreased gradually after exposure to the CSC until the adsorption equilibrium was reached. When the initial dye concentration was low, equilibrium was reached relatively quickly, within 30 min. For example, for MB a sharp decrease in MB concentration was observed in the first 10 min followed by a slower adsorption until the MB concentration stabilized. This is because when the initial dye concentration is low, the adsorption reaction can quickly reach equilibrium because there are a large number of sites available on the surface of CSC for adsorption relative to the number of MB molecules. As the initial concentration of dye increases, a more competitive adsorption for MB molecules for surface sites is induced. Thus, as more active sites become occupied, it becomes  www.nature.com/scientificreports/ increasingly more difficult to adsorb dye due to a weaker attraction between the surface of CSC and the MB molecules, which leads to slowing of dye adsorption rate in the latter stages of the reaction. In order to further investigate the adsorption mechanism of the MB dye on to the CSC, both pseudo first order (PFO) (Eq. 4) and pseudo second order (PSO) (Eq. 5) kinetic models were used for data analysis. The kinetic data for MB adsorption on to CSC fit both the PFO (Fig S2(c)) and PSO (Fig. S2(d)) models well with good linear behavior. However, the goodness of fit (R 2 ) for the PSO model was slightly higher than that for the PFO model, and the adsorption capacity at equilibrium obtained from PSO fitting was also closer to the experimental value (Table S2). The corresponding experimental adsorption capacities for initial MB concentrations of 50, 75 and 100 ppm were 60.52, 96.72 and 123.45 mg g -1 , respectively. Therefore, the PSO model was more suitable for describing dye adsorption to CSC. Based on the underlying assumptions of the PSO equation model this indicates that MB participated in chemical adsorption, involving the sharing or exchange of electrons between the hydrophilic edge sites of the CSC and the dye cations 25 . As the initial MB concentration increased, the goodness of fit of adsorption capacity (q e ) also increased, and became very close to the experimentally derived q e value, but the adsorption rate (k 2 value) decreased. This decrease in rate constant is due to the intense adsorption competition between relatively limited surface-active sites and a large number of MB molecules in a higher concentration dye solution, resulting in a relatively low-rate constant 26 .
Similarly, the initial adsorption capacities for CR at 400, 455 and 500 ppm were 592.76, 653.26 and 679.91 mg g -1 , respectively where fitting of the kinetic data to the same two models (Fig. S2 (g) and (h)) showed that the PSO model better described the adsorption of CR on the CSC, with a calculated adsorption amount at equilibrium closer to the experimental value (Table S2). Like MB, since the PSO model assumes chemical adsorption as the dominant adsorption process 27,28 , which involves the sharing or exchange of electrons between the hydrophilic edge sites of the CSC and the dye ions, this result also suggests that negatively charged CR ions were removed from solution via a chemical adsorption process. This adsorption may be caused by a combination of π-π stacking, hydrogen bonding, and van der Waals forces. In addition, the k 1 values for CR were much lower than those calculated for MB, which indicated that the CSC had a higher affinity for the negatively charged dye 29 .
Effect of catalyst dose. The catalyst dose is an important practical consideration and also plays an important role in determining the efficiency of the photocatalytic reaction. An increase in catalyst dose will increase the overall concentration of active substance and hence increase the number of active sites in the reaction system. Thus, the probability of contact of dye molecules with the active sites increases, as does the reaction probability, and rate of reaction, leading to improved dye removal efficiency. For photodegradation removal processes there will generally be an increase in removal efficiency with an increase in dose until, at a certain concentration, the increase in catalyst amount reduces the light transmittance of the suspension 30 and hence the removal efficiency.
For MB and CR (Fig. 5) as the amount of CSC increased from 20, 40, 60 to 80 mg, the removal efficiencies of MB were 99.7, 99.8, 99.9 and 99.9%, respectively, and 83.3, 88.6, 96.7% and 97.0% for CR, respectively. Thus, for both dyes' removal efficiency increases with catalyst dose, being almost a negligible for MB and only modest for CR. This indicated that even the lowest dose used here had an adsorption capacity well above the amount of dye in solution. Even though the increase in MB removal efficiency with dose was not very significant, the time www.nature.com/scientificreports/ required to achieve this removal became less as dose increased (Fig. 5b). Since even a 20 mg catalyst dose could achieve a strong catalytic effect, in 40 mL dye solutions a catalyst dose of 0.5 g L -1 was selected for all subsequent experiments. According to the experimental data, enough binding sites can be reached at 20 mg. Although the removal efficiency further increases with an increase in the amount of catalyst, the higher concentration is not essential for performance and are thus a waste of catalyst, since nearly all of the additional binding sites are not used. The best fit kinetic parameters (Table 2) for MB and CR degradation under different catalyst doses showed that the degradation of both MB and CR followed pseudo first order kinetics, with good correlation (R 2 close to 1).
The k value for MB increased slightly in the range of 20-40 mg, and decreased between 40 and 80 mg. This was potentially because the increase of catalyst above 40 mg reduced the light transmittance of the suspension, making it more difficult for photons to reach the CSC surface, and the amount of •OH free radical formed between dye molecules and catalyst surface became relatively small. However, for CR, the k value increased significantly between 20 and 60 mg, and slightly increased from 60 to 80 mg.
Effect of initial dye concentration. The removal efficiencies of MB dye at initial dye concentrations of 50, 75 and 100 ppm were 99.8, 99.7 and 99.6%, respectively (Fig. 6). Although the removal efficiency did not significantly change with initial dye concentration, the time required for the photodegradation become longer (Fig. 6). Likewise, after 120 min of illumination, the overall removal efficiency of CR dye decreased with an increase in the initial dye concentration, being 98.2, 90.8 and 83.3%, for CR concentration of 400, 450 and 500 ppm respectively. This is because the number of •OH radicals formed on the catalyst surface and the number of interactions with dye molecules determine the overall efficiency of photocatalytic degradation. With an increase in the number of dye molecules in solution, the number of dye molecules contacting the catalyst surface also increases, causing active sites on the catalyst surface to be covered by them, which will affect the arrival of photons on the surface of the catalyst, and will reduce the catalytic efficiency, resulting in less •OH radical generation on the catalyst surface. Therefore, increased dye concentration can lead to a reduction in both degradation efficiency and reaction rate 31 .
The variation in best fit degradation kinetic parameters for both MB and CR with different concentrations are shown in Table 3. The k values of both MB and CR increased as the initial dye concentration decreased. The  Influence of free radical scavengers on catalyst performance. Since the photodegradation efficiency of the catalyst is likely to depend on the concentration of free radicals at the surface, free radical scavengers can affect photodegradation performance. Here, in order to examine the factors affecting the performance of the catalyst, isopropanol (IP) as a •OH radical scavenger, p-benzoquinone (BQ) as an •O 2 radical scavenger, and ammonium oxalate (AO) as a hole (h + ) scavenger were added to the reaction system 36,37 . Under illumination from a xenon lamp, the removal efficiencies of MB with BQ, AO and IP were 88.6%, 93.9% and 95.0%, respectively (Fig. 7) compared to a MB removal efficiency of 99.9 with no added scavengers. Similarly, the removal efficiencies for CR after adding BQ, AO and IP were 64.9, 83.5 and 88.4% respectively, compared to 90.8% without any added scavenger.
The presence of BQ, significantly inhibited the removal efficiencies of both MB and CR This was because BQ leads to a decrease in the amount of superoxide radicals present, which leads to a significant decrease in the photodegradation rate (Table 4). In comparison, AO and IP induced only slight decreases in the removal efficiencies of MB and CR. These experimental results imply that while the formation of •O 2 − and h + in the photocatalytic reaction are both essential for a high removal efficiency, •O 2 − is the dominant reactive species in this reaction. These observations allow a removal mechanism to be proposed. When the CSC is irradiated by a xenon lamp, h + electron holes are generated in the collision gap. Oxygen molecules in the mixed solution form •O 2 − superoxide radicals by electron removal, and h + react with water to form •OH radical 38 .
The addition of hydrogen peroxide (H 2 O 2 ) is often used as an amendment to enhance photocatalytic degradation of organic materials by enhancing free radical generation. Here, under H 2 O 2 assisted photocatalysis, the removal efficiencies of MB and CR were 99.9% and 97.1%, respectively, indicating that compared to the blank samples (99.6% and 90.8%), the addition of H 2 O 2 significantly improved the photodegradation removal efficiency. The reaction rate constants of the two dyes were also close to each other, indicating that these two dyes were removed via similar mechanisms.  39 , and the further reaction of HO 2 and H 2 O 2 forms OH radical, which is the decisive step of the degradation rate of organic pollutants in aqueous solution 40 . The reaction mechanism is summarized as follows: In the process of photocatalytic degradation of dyes, the electrons generated by the catalyst can react with O 2 to form •O 2 − , and then the products can react with water to form hydrogenated oxygen radicals (HO 2 •) and hydroxyl radicals (OH -). The generated HO 2 • rearranges to form oxygen (O 2 ) and peroxide hydrogen (H 2 O 2 ) 41 , and H 2 O 2 reacts with superoxide radical to form hydroxyl radical (OH -). The generated h + interacts with water to form a highly reactive hydroxyl radical, the hole itself can also attack dye molecules to produce by-products as well. After the formation of free radicals, the reactive substance (•O 2 − ) attacks dye molecules until the end of decolourisation and ring opening reaction, effectively degrading the dye into smaller intermediates and the final products (CO 2 and H 2 O) [42][43][44] . The possible photocatalytic reactions are as follows: Proposed photocatalytic mechanism. Based on the above experimental studies, a possible mechanism for the photocatalytic degradation of dyes by the CSC produced here was proposed (Fig S3). Under light irradiation, the electrons gain sufficient energy to become excited and transition from the valence band to the conduction band, and form excitons (electron-hole pairs). These excitons are known to act in four potential ways: (1) intralattice recombination, (2) recombination at surface active sites, (3) hole (h + ) oxidizes external dyes, and (4) electron (e -) reduces external dyes. High photocatalytic activity can be achieved through the minimum photocatalytic system of electron hole recombination, and the combination of effective absorption and enhanced electron transition from heterojunctions among some dislocations in the crystal structures 45,46 . It was previously reported that photocatalytic degradation efficiency can also be further improved via effective size reduction of CaO nanoparticles and the innate nano properties of the initial raw materials 47 . Hence, calcium oxide nanoparticles synthesized by calcination of marine biomass, such as shells or other biomass materials containing metal oxides and trace transition metals, which can act as excellent natural photocatalysts for the degradation of dyes from the textile industries. In contrast, while the photodegradation efficiency of 50 mg CaO derived from eggshell biomass catalyst can reach 96.2% at an initial MB dye concentration of 20 mg/L through parameter optimization 11 , here, the photodegradation efficiency using a 20 mg CSC derived from marine biomass reached 99.6% for an initial MB dye solution concentration of 100 mg/L.   www.nature.com/scientificreports/ Feasibility of practical application. In the textile industry, dye wastewater typically contains many kinds of different dyes which need to be simultaneously removed. Hence here, the feasibility of applying a CSC for practical wastewater treatment was evaluated in a mixed dye solution. The results indicated that 20 mg of CSC could easily degrade both a mixed dark colored binary dye solution (100 ppm MB + 100 ppm CR) and a ternary dye system (75 ppm MB + 75 ppm CR + 75 ppm rhodamine B (Rh B)) into a transparent solution (Fig S4). Analysis of the absorption spectra of the mixed dye solution system under visible-light irradiation, showed that the initial peaks of the mixed dye solution decreased sharply as reaction time increases, with some peaks shifting slightly or completely disappearing, which strongly suggested that mixtures of dye molecules were destroyed or degraded, and new substances were formed.

Conclusion
A novel photocatalyst nanomaterial derived from natural clam shell was successfully prepared via a facile calcination process. The as prepared material was then evaluated for its potential to treat dye contaminated wastewater. Under xenon lamp illumination, the overall absorbance-photocatalytic removal efficiencies for 100 mg L -1 MB and 500 mg L -1 CR reached 99.6 and 83.3%, respectively. Increasing the calcination temperature enhanced the catalytic performance of the catalyst. This was attributed to synergetic effects of increased absorption together with increased heterojunctions among some dislocations in the crystal structures. Increased numbers of heterojunction can greatly reduce the energy required for electron transition, making it much easier to form holes and excited electrons. Consequently, the inexpensive clam shell photocatalyst, synthesized in this work from a marine biowaste, may have great potential for future practical large-scale industrial wastewater treatment applications.